diff options
237 files changed, 7392 insertions, 1760 deletions
@@ -1,10 +1,10 @@ Meta: 1 Name: zfs Branch: 1.0 -Version: 2.1.10 +Version: 2.1.15 Release: 1 Release-Tags: relext License: CDDL Author: OpenZFS -Linux-Maximum: 6.2 +Linux-Maximum: 6.7 Linux-Minimum: 3.10 diff --git a/cmd/zdb/zdb.c b/cmd/zdb/zdb.c index f424dd771f70..ef3920d23f5b 100644 --- a/cmd/zdb/zdb.c +++ b/cmd/zdb/zdb.c @@ -3102,13 +3102,22 @@ dump_znode_sa_xattr(sa_handle_t *hdl) (void) printf("\tSA xattrs: %d bytes, %d entries\n\n", sa_xattr_size, sa_xattr_entries); while ((elem = nvlist_next_nvpair(sa_xattr, elem)) != NULL) { + boolean_t can_print = !dump_opt['P']; uchar_t *value; uint_t cnt, idx; (void) printf("\t\t%s = ", nvpair_name(elem)); nvpair_value_byte_array(elem, &value, &cnt); + + for (idx = 0; idx < cnt; ++idx) { + if (!isprint(value[idx])) { + can_print = B_FALSE; + break; + } + } + for (idx = 0; idx < cnt; ++idx) { - if (isprint(value[idx])) + if (can_print) (void) putchar(value[idx]); else (void) printf("\\%3.3o", value[idx]); @@ -4806,7 +4815,7 @@ dump_label(const char *dev) if (nvlist_size(config, &size, NV_ENCODE_XDR) != 0) size = buflen; - /* If the device is a cache device clear the header. */ + /* If the device is a cache device read the header. */ if (!read_l2arc_header) { if (nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_STATE, &l2cache) == 0 && diff --git a/cmd/zed/agents/zfs_mod.c b/cmd/zed/agents/zfs_mod.c index f67fd96af045..ecf20418e494 100644 --- a/cmd/zed/agents/zfs_mod.c +++ b/cmd/zed/agents/zfs_mod.c @@ -142,6 +142,17 @@ zfs_unavail_pool(zpool_handle_t *zhp, void *data) } /* + * Write an array of strings to the zed log + */ +static void lines_to_zed_log_msg(char **lines, int lines_cnt) +{ + int i; + for (i = 0; i < lines_cnt; i++) { + zed_log_msg(LOG_INFO, "%s", lines[i]); + } +} + +/* * Two stage replace on Linux * since we get disk notifications * we can wait for partitioned disk slice to show up! @@ -195,6 +206,8 @@ zfs_process_add(zpool_handle_t *zhp, nvlist_t *vdev, boolean_t labeled) boolean_t is_mpath_wholedisk = B_FALSE; uint_t c; vdev_stat_t *vs; + char **lines = NULL; + int lines_cnt = 0; if (nvlist_lookup_string(vdev, ZPOOL_CONFIG_PATH, &path) != 0) return; @@ -209,8 +222,12 @@ zfs_process_add(zpool_handle_t *zhp, nvlist_t *vdev, boolean_t labeled) } (void) nvlist_lookup_string(vdev, ZPOOL_CONFIG_PHYS_PATH, &physpath); + + update_vdev_config_dev_sysfs_path(vdev, path, + ZPOOL_CONFIG_VDEV_ENC_SYSFS_PATH); (void) nvlist_lookup_string(vdev, ZPOOL_CONFIG_VDEV_ENC_SYSFS_PATH, &enc_sysfs_path); + (void) nvlist_lookup_uint64(vdev, ZPOOL_CONFIG_WHOLE_DISK, &wholedisk); (void) nvlist_lookup_uint64(vdev, ZPOOL_CONFIG_OFFLINE, &offline); (void) nvlist_lookup_uint64(vdev, ZPOOL_CONFIG_FAULTED, &faulted); @@ -377,6 +394,22 @@ zfs_process_add(zpool_handle_t *zhp, nvlist_t *vdev, boolean_t labeled) if (is_mpath_wholedisk) { /* Don't label device mapper or multipath disks. */ + zed_log_msg(LOG_INFO, + " it's a multipath wholedisk, don't label"); + if (zpool_prepare_disk(zhp, vdev, "autoreplace", &lines, + &lines_cnt) != 0) { + zed_log_msg(LOG_INFO, + " zpool_prepare_disk: could not " + "prepare '%s' (%s)", fullpath, + libzfs_error_description(g_zfshdl)); + if (lines_cnt > 0) { + zed_log_msg(LOG_INFO, + " zfs_prepare_disk output:"); + lines_to_zed_log_msg(lines, lines_cnt); + } + libzfs_free_str_array(lines, lines_cnt); + return; + } } else if (!labeled) { /* * we're auto-replacing a raw disk, so label it first @@ -399,10 +432,18 @@ zfs_process_add(zpool_handle_t *zhp, nvlist_t *vdev, boolean_t labeled) * If this is a request to label a whole disk, then attempt to * write out the label. */ - if (zpool_label_disk(g_zfshdl, zhp, leafname) != 0) { - zed_log_msg(LOG_INFO, " zpool_label_disk: could not " + if (zpool_prepare_and_label_disk(g_zfshdl, zhp, leafname, + vdev, "autoreplace", &lines, &lines_cnt) != 0) { + zed_log_msg(LOG_INFO, + " zpool_prepare_and_label_disk: could not " "label '%s' (%s)", leafname, libzfs_error_description(g_zfshdl)); + if (lines_cnt > 0) { + zed_log_msg(LOG_INFO, + " zfs_prepare_disk output:"); + lines_to_zed_log_msg(lines, lines_cnt); + } + libzfs_free_str_array(lines, lines_cnt); (void) zpool_vdev_online(zhp, fullpath, ZFS_ONLINE_FORCEFAULT, &newstate); @@ -457,6 +498,8 @@ zfs_process_add(zpool_handle_t *zhp, nvlist_t *vdev, boolean_t labeled) DEV_BYID_PATH, new_devid); } + libzfs_free_str_array(lines, lines_cnt); + /* * Construct the root vdev to pass to zpool_vdev_attach(). While adding * the entire vdev structure is harmless, we construct a reduced set of @@ -597,8 +640,6 @@ zfs_iter_vdev(zpool_handle_t *zhp, nvlist_t *nvl, void *data) */ if (nvlist_lookup_string(nvl, dp->dd_prop, &path) != 0 || strcmp(dp->dd_compare, path) != 0) { - zed_log_msg(LOG_INFO, " %s: no match (%s != vdev %s)", - __func__, dp->dd_compare, path); return; } if (dp->dd_new_vdev_guid != 0 && dp->dd_new_vdev_guid != guid) { diff --git a/cmd/zed/agents/zfs_retire.c b/cmd/zed/agents/zfs_retire.c index b4794e31193f..b2b28ef67f13 100644 --- a/cmd/zed/agents/zfs_retire.c +++ b/cmd/zed/agents/zfs_retire.c @@ -415,6 +415,11 @@ zfs_retire_recv(fmd_hdl_t *hdl, fmd_event_t *ep, nvlist_t *nvl, FM_EREPORT_PAYLOAD_ZFS_VDEV_GUID, &vdev_guid) != 0) return; + if (vdev_guid == 0) { + fmd_hdl_debug(hdl, "Got a zero GUID"); + return; + } + if (spare) { int nspares = find_and_remove_spares(zhdl, vdev_guid); fmd_hdl_debug(hdl, "%d spares removed", nspares); @@ -444,14 +449,16 @@ zfs_retire_recv(fmd_hdl_t *hdl, fmd_event_t *ep, nvlist_t *nvl, return; /* Remove the vdev since device is unplugged */ + int remove_status = 0; if (l2arc || (strcmp(class, "resource.fs.zfs.removed") == 0)) { - int status = zpool_vdev_remove_wanted(zhp, devname); + remove_status = zpool_vdev_remove_wanted(zhp, devname); fmd_hdl_debug(hdl, "zpool_vdev_remove_wanted '%s'" - ", ret:%d", devname, status); + ", err:%d", devname, libzfs_errno(zhdl)); } /* Replace the vdev with a spare if its not a l2arc */ - if (!l2arc && (!fmd_prop_get_int32(hdl, "spare_on_remove") || + if (!l2arc && !remove_status && + (!fmd_prop_get_int32(hdl, "spare_on_remove") || replace_with_spare(hdl, zhp, vdev) == B_FALSE)) { /* Could not handle with spare */ fmd_hdl_debug(hdl, "no spare for '%s'", devname); diff --git a/cmd/zed/zed.d/Makefile.am b/cmd/zed/zed.d/Makefile.am index 2c8173b3e769..1905a92078dd 100644 --- a/cmd/zed/zed.d/Makefile.am +++ b/cmd/zed/zed.d/Makefile.am @@ -21,6 +21,7 @@ dist_zedexec_SCRIPTS = \ scrub_finish-notify.sh \ statechange-led.sh \ statechange-notify.sh \ + statechange-slot_off.sh \ vdev_clear-led.sh \ vdev_attach-led.sh \ pool_import-led.sh \ @@ -39,6 +40,7 @@ zedconfdefaults = \ scrub_finish-notify.sh \ statechange-led.sh \ statechange-notify.sh \ + statechange-slot_off.sh \ vdev_clear-led.sh \ vdev_attach-led.sh \ pool_import-led.sh \ diff --git a/cmd/zed/zed.d/statechange-slot_off.sh b/cmd/zed/zed.d/statechange-slot_off.sh new file mode 100755 index 000000000000..06acce93b8aa --- /dev/null +++ b/cmd/zed/zed.d/statechange-slot_off.sh @@ -0,0 +1,64 @@ +#!/bin/sh +# shellcheck disable=SC3014,SC2154,SC2086,SC2034 +# +# Turn off disk's enclosure slot if it becomes FAULTED. +# +# Bad SCSI disks can often "disappear and reappear" causing all sorts of chaos +# as they flip between FAULTED and ONLINE. If +# ZED_POWER_OFF_ENCLOSURE_SLOT_ON_FAULT is set in zed.rc, and the disk gets +# FAULTED, then power down the slot via sysfs: +# +# /sys/class/enclosure/<enclosure>/<slot>/power_status +# +# We assume the user will be responsible for turning the slot back on again. +# +# Note that this script requires that your enclosure be supported by the +# Linux SCSI Enclosure services (SES) driver. The script will do nothing +# if you have no enclosure, or if your enclosure isn't supported. +# +# Exit codes: +# 0: slot successfully powered off +# 1: enclosure not available +# 2: ZED_POWER_OFF_ENCLOSURE_SLOT_ON_FAULT disabled +# 3: vdev was not FAULTED +# 4: The enclosure sysfs path passed from ZFS does not exist +# 5: Enclosure slot didn't actually turn off after we told it to + +[ -f "${ZED_ZEDLET_DIR}/zed.rc" ] && . "${ZED_ZEDLET_DIR}/zed.rc" +. "${ZED_ZEDLET_DIR}/zed-functions.sh" + +if [ ! -d /sys/class/enclosure ] ; then + # No JBOD enclosure or NVMe slots + exit 1 +fi + +if [ "${ZED_POWER_OFF_ENCLOSURE_SLOT_ON_FAULT}" != "1" ] ; then + exit 2 +fi + +if [ "$ZEVENT_VDEV_STATE_STR" != "FAULTED" ] ; then + exit 3 +fi + +if [ ! -f "$ZEVENT_VDEV_ENC_SYSFS_PATH/power_status" ] ; then + exit 4 +fi + +# Turn off the slot and wait for sysfs to report that the slot is off. +# It can take ~400ms on some enclosures and multiple retries may be needed. +for i in $(seq 1 20) ; do + echo "off" | tee "$ZEVENT_VDEV_ENC_SYSFS_PATH/power_status" + + for j in $(seq 1 5) ; do + if [ "$(cat $ZEVENT_VDEV_ENC_SYSFS_PATH/power_status)" == "off" ] ; then + break 2 + fi + sleep 0.1 + done +done + +if [ "$(cat $ZEVENT_VDEV_ENC_SYSFS_PATH/power_status)" != "off" ] ; then + exit 5 +fi + +zed_log_msg "powered down slot $ZEVENT_VDEV_ENC_SYSFS_PATH for $ZEVENT_VDEV_PATH" diff --git a/cmd/zed/zed.d/zed.rc b/cmd/zed/zed.d/zed.rc index 227b26c26b50..91ceea79e5c5 100644 --- a/cmd/zed/zed.d/zed.rc +++ b/cmd/zed/zed.d/zed.rc @@ -143,3 +143,8 @@ ZED_SYSLOG_SUBCLASS_EXCLUDE="history_event" # Disabled by default, 1 to enable and 0 to disable. #ZED_SYSLOG_DISPLAY_GUIDS=1 +## +# Power off the drive's slot in the enclosure if it becomes FAULTED. This can +# help silence misbehaving drives. This assumes your drive enclosure fully +# supports slot power control via sysfs. +#ZED_POWER_OFF_ENCLOSURE_SLOT_ON_FAULT=1 diff --git a/cmd/zed/zed_event.c b/cmd/zed/zed_event.c index 9eaad0e92fbb..e384b16dc8f0 100644 --- a/cmd/zed/zed_event.c +++ b/cmd/zed/zed_event.c @@ -35,6 +35,7 @@ #include "zed_strings.h" #include "agents/zfs_agents.h" +#include <libzutil.h> #define MAXBUF 4096 @@ -907,6 +908,25 @@ _zed_event_add_time_strings(uint64_t eid, zed_strings_t *zsp, int64_t etime[]) } } + +static void +_zed_event_update_enc_sysfs_path(nvlist_t *nvl) +{ + char *vdev_path; + + if (nvlist_lookup_string(nvl, FM_EREPORT_PAYLOAD_ZFS_VDEV_PATH, + &vdev_path) != 0) { + return; /* some other kind of event, ignore it */ + } + + if (vdev_path == NULL) { + return; + } + + update_vdev_config_dev_sysfs_path(nvl, vdev_path, + FM_EREPORT_PAYLOAD_ZFS_VDEV_ENC_SYSFS_PATH); +} + /* * Service the next zevent, blocking until one is available. */ @@ -954,6 +974,17 @@ zed_event_service(struct zed_conf *zcp) zed_log_msg(LOG_WARNING, "Failed to lookup zevent class (eid=%llu)", eid); } else { + /* + * Special case: If we can dynamically detect an enclosure sysfs + * path, then use that value rather than the one stored in the + * vd->vdev_enc_sysfs_path. There have been rare cases where + * vd->vdev_enc_sysfs_path becomes outdated. However, there + * will be other times when we can not dynamically detect the + * sysfs path (like if a disk disappears) and have to rely on + * the old value for things like turning on the fault LED. + */ + _zed_event_update_enc_sysfs_path(nvl); + /* let internal modules see this event first */ zfs_agent_post_event(class, NULL, nvl); diff --git a/cmd/zfs/zfs_main.c b/cmd/zfs/zfs_main.c index 2d81ef31c4ac..a7b2a6430419 100644 --- a/cmd/zfs/zfs_main.c +++ b/cmd/zfs/zfs_main.c @@ -535,7 +535,7 @@ usage(boolean_t requested) show_properties = B_TRUE; if (show_properties) { - (void) fprintf(fp, + (void) fprintf(fp, "%s", gettext("\nThe following properties are supported:\n")); (void) fprintf(fp, "\n\t%-14s %s %s %s\n\n", diff --git a/cmd/zpool/os/freebsd/zpool_vdev_os.c b/cmd/zpool/os/freebsd/zpool_vdev_os.c index aa66d29fa604..08f382fb527b 100644 --- a/cmd/zpool/os/freebsd/zpool_vdev_os.c +++ b/cmd/zpool/os/freebsd/zpool_vdev_os.c @@ -116,3 +116,24 @@ after_zpool_upgrade(zpool_handle_t *zhp) "details.\n"), zpool_get_name(zhp)); } } + +int +zpool_power_current_state(zpool_handle_t *zhp, char *vdev) +{ + + (void) zhp; + (void) vdev; + /* Enclosure slot power not supported on FreeBSD yet */ + return (-1); +} + +int +zpool_power(zpool_handle_t *zhp, char *vdev, boolean_t turn_on) +{ + + (void) zhp; + (void) vdev; + (void) turn_on; + /* Enclosure slot power not supported on FreeBSD yet */ + return (ENOTSUP); +} diff --git a/cmd/zpool/os/linux/zpool_vdev_os.c b/cmd/zpool/os/linux/zpool_vdev_os.c index da87aa79f365..cfaeef56a20e 100644 --- a/cmd/zpool/os/linux/zpool_vdev_os.c +++ b/cmd/zpool/os/linux/zpool_vdev_os.c @@ -410,3 +410,258 @@ void after_zpool_upgrade(zpool_handle_t *zhp) { } + +/* + * Read from a sysfs file and return an allocated string. Removes + * the newline from the end of the string if there is one. + * + * Returns a string on success (which must be freed), or NULL on error. + */ +static char *zpool_sysfs_gets(char *path) +{ + int fd; + struct stat statbuf; + char *buf = NULL; + ssize_t count = 0; + fd = open(path, O_RDONLY); + if (fd < 0) + return (NULL); + + if (fstat(fd, &statbuf) != 0) { + close(fd); + return (NULL); + } + + buf = calloc(sizeof (*buf), statbuf.st_size + 1); + if (buf == NULL) { + close(fd); + return (NULL); + } + + /* + * Note, we can read less bytes than st_size, and that's ok. Sysfs + * files will report their size is 4k even if they only return a small + * string. + */ + count = read(fd, buf, statbuf.st_size); + if (count < 0) { + /* Error doing read() or we overran the buffer */ + close(fd); + free(buf); + return (NULL); + } + + /* Remove trailing newline */ + if (buf[count - 1] == '\n') + buf[count - 1] = 0; + + close(fd); + + return (buf); +} + +/* + * Write a string to a sysfs file. + * + * Returns 0 on success, non-zero otherwise. + */ +static int zpool_sysfs_puts(char *path, char *str) +{ + FILE *file; + + file = fopen(path, "w"); + if (!file) { + return (-1); + } + + if (fputs(str, file) < 0) { + fclose(file); + return (-2); + } + fclose(file); + return (0); +} + +/* Given a vdev nvlist_t, rescan its enclosure sysfs path */ +static void +rescan_vdev_config_dev_sysfs_path(nvlist_t *vdev_nv) +{ + update_vdev_config_dev_sysfs_path(vdev_nv, + fnvlist_lookup_string(vdev_nv, ZPOOL_CONFIG_PATH), + ZPOOL_CONFIG_VDEV_ENC_SYSFS_PATH); +} + +/* + * Given a power string: "on", "off", "1", or "0", return 0 if it's an + * off value, 1 if it's an on value, and -1 if the value is unrecognized. + */ +static int zpool_power_parse_value(char *str) +{ + if ((strcmp(str, "off") == 0) || (strcmp(str, "0") == 0)) + return (0); + + if ((strcmp(str, "on") == 0) || (strcmp(str, "1") == 0)) + return (1); + + return (-1); +} + +/* + * Given a vdev string return an allocated string containing the sysfs path to + * its power control file. Also do a check if the power control file really + * exists and has correct permissions. + * + * Example returned strings: + * + * /sys/class/enclosure/0:0:122:0/10/power_status + * /sys/bus/pci/slots/10/power + * + * Returns allocated string on success (which must be freed), NULL on failure. + */ +static char * +zpool_power_sysfs_path(zpool_handle_t *zhp, char *vdev) +{ + char *enc_sysfs_dir = NULL; + char *path = NULL; + nvlist_t *vdev_nv = zpool_find_vdev(zhp, vdev, NULL, NULL, NULL); + + if (vdev_nv == NULL) { + return (NULL); + } + + /* Make sure we're getting the updated enclosure sysfs path */ + rescan_vdev_config_dev_sysfs_path(vdev_nv); + + if (nvlist_lookup_string(vdev_nv, ZPOOL_CONFIG_VDEV_ENC_SYSFS_PATH, + &enc_sysfs_dir) != 0) { + return (NULL); + } + + if (asprintf(&path, "%s/power_status", enc_sysfs_dir) == -1) + return (NULL); + + if (access(path, W_OK) != 0) { + free(path); + path = NULL; + /* No HDD 'power_control' file, maybe it's NVMe? */ + if (asprintf(&path, "%s/power", enc_sysfs_dir) == -1) { + return (NULL); + } + + if (access(path, R_OK | W_OK) != 0) { + /* Not NVMe either */ + free(path); + return (NULL); + } + } + + return (path); +} + +/* + * Given a path to a sysfs power control file, return B_TRUE if you should use + * "on/off" words to control it, or B_FALSE otherwise ("0/1" to control). + */ +static boolean_t +zpool_power_use_word(char *sysfs_path) +{ + if (strcmp(&sysfs_path[strlen(sysfs_path) - strlen("power_status")], + "power_status") == 0) { + return (B_TRUE); + } + return (B_FALSE); +} + +/* + * Check the sysfs power control value for a vdev. + * + * Returns: + * 0 - Power is off + * 1 - Power is on + * -1 - Error or unsupported + */ +int +zpool_power_current_state(zpool_handle_t *zhp, char *vdev) +{ + char *val; + int rc; + + char *path = zpool_power_sysfs_path(zhp, vdev); + if (path == NULL) + return (-1); + + val = zpool_sysfs_gets(path); + if (val == NULL) { + free(path); + return (-1); + } + + rc = zpool_power_parse_value(val); + free(val); + free(path); + return (rc); +} + +/* + * Turn on or off the slot to a device + * + * Device path is the full path to the device (like /dev/sda or /dev/sda1). + * + * Return code: + * 0: Success + * ENOTSUP: Power control not supported for OS + * EBADSLT: Couldn't read current power state + * ENOENT: No sysfs path to power control + * EIO: Couldn't write sysfs power value + * EBADE: Sysfs power value didn't change + */ +int +zpool_power(zpool_handle_t *zhp, char *vdev, boolean_t turn_on) +{ + char *sysfs_path; + const char *val; + int rc; + int timeout_ms; + + rc = zpool_power_current_state(zhp, vdev); + if (rc == -1) { + return (EBADSLT); + } + + /* Already correct value? */ + if (rc == (int)turn_on) + return (0); + + sysfs_path = zpool_power_sysfs_path(zhp, vdev); + if (sysfs_path == NULL) + return (ENOENT); + + if (zpool_power_use_word(sysfs_path)) { + val = turn_on ? "on" : "off"; + } else { + val = turn_on ? "1" : "0"; + } + + rc = zpool_sysfs_puts(sysfs_path, (char *)val); + + free(sysfs_path); + if (rc != 0) { + return (EIO); + } + + /* + * Wait up to 30 seconds for sysfs power value to change after + * writing it. + */ + timeout_ms = zpool_getenv_int("ZPOOL_POWER_ON_SLOT_TIMEOUT_MS", 30000); + for (int i = 0; i < MAX(1, timeout_ms / 200); i++) { + rc = zpool_power_current_state(zhp, vdev); + if (rc == (int)turn_on) + return (0); /* success */ + + fsleep(0.200); /* 200ms */ + } + + /* sysfs value never changed */ + return (EBADE); +} diff --git a/cmd/zpool/zpool.d/ses b/cmd/zpool/zpool.d/ses index b51fe31894ab..8be3eb57acf2 100755 --- a/cmd/zpool/zpool.d/ses +++ b/cmd/zpool/zpool.d/ses @@ -32,10 +32,18 @@ for i in $scripts ; do val="" case $i in enc) - val=$(ls "$VDEV_ENC_SYSFS_PATH/../../" 2>/dev/null) + if echo "$VDEV_ENC_SYSFS_PATH" | grep -q '/sys/bus/pci/slots' ; then + val="$VDEV_ENC_SYSFS_PATH" + else + val="$(ls """$VDEV_ENC_SYSFS_PATH/../../""" 2>/dev/null)" + fi ;; slot) - val=$(cat "$VDEV_ENC_SYSFS_PATH/slot" 2>/dev/null) + if echo "$VDEV_ENC_SYSFS_PATH" | grep -q '/sys/bus/pci/slots' ; then + val="$(basename """$VDEV_ENC_SYSFS_PATH""")" + else + val="$(cat """$VDEV_ENC_SYSFS_PATH/slot""" 2>/dev/null)" + fi ;; encdev) val=$(ls "$VDEV_ENC_SYSFS_PATH/../device/scsi_generic" 2>/dev/null) diff --git a/cmd/zpool/zpool_iter.c b/cmd/zpool/zpool_iter.c index abfa2b7f6b90..82250f692700 100644 --- a/cmd/zpool/zpool_iter.c +++ b/cmd/zpool/zpool_iter.c @@ -439,39 +439,23 @@ static void vdev_run_cmd(vdev_cmd_data_t *data, char *cmd) { int rc; - char *argv[2] = {cmd, 0}; - char *env[5] = {"PATH=/bin:/sbin:/usr/bin:/usr/sbin", NULL, NULL, NULL, - NULL}; + char *argv[2] = {cmd}; + char **env; char **lines = NULL; int lines_cnt = 0; int i; - /* Setup our custom environment variables */ - rc = asprintf(&env[1], "VDEV_PATH=%s", - data->path ? data->path : ""); - if (rc == -1) { - env[1] = NULL; + env = zpool_vdev_script_alloc_env(data->pool, data->path, data->upath, + data->vdev_enc_sysfs_path, NULL, NULL); + if (env == NULL) goto out; - } - - rc = asprintf(&env[2], "VDEV_UPATH=%s", - data->upath ? data->upath : ""); - if (rc == -1) { - env[2] = NULL; - goto out; - } - - rc = asprintf(&env[3], "VDEV_ENC_SYSFS_PATH=%s", - data->vdev_enc_sysfs_path ? - data->vdev_enc_sysfs_path : ""); - if (rc == -1) { - env[3] = NULL; - goto out; - } /* Run the command */ rc = libzfs_run_process_get_stdout_nopath(cmd, argv, env, &lines, &lines_cnt); + + zpool_vdev_script_free_env(env); + if (rc != 0) goto out; @@ -483,10 +467,6 @@ vdev_run_cmd(vdev_cmd_data_t *data, char *cmd) out: if (lines != NULL) libzfs_free_str_array(lines, lines_cnt); - - /* Start with i = 1 since env[0] was statically allocated */ - for (i = 1; i < ARRAY_SIZE(env); i++) - free(env[i]); } /* @@ -571,6 +551,10 @@ for_each_vdev_run_cb(void *zhp_data, nvlist_t *nv, void *cb_vcdl) if (nvlist_lookup_string(nv, ZPOOL_CONFIG_PATH, &path) != 0) return (1); + /* Make sure we're getting the updated enclosure sysfs path */ + update_vdev_config_dev_sysfs_path(nv, path, + ZPOOL_CONFIG_VDEV_ENC_SYSFS_PATH); + nvlist_lookup_string(nv, ZPOOL_CONFIG_VDEV_ENC_SYSFS_PATH, &vdev_enc_sysfs_path); diff --git a/cmd/zpool/zpool_main.c b/cmd/zpool/zpool_main.c index 2311d4f046f6..c50a8d309741 100644 --- a/cmd/zpool/zpool_main.c +++ b/cmd/zpool/zpool_main.c @@ -345,7 +345,7 @@ get_usage(zpool_help_t idx) return (gettext("\tattach [-fsw] [-o property=value] " "<pool> <device> <new-device>\n")); case HELP_CLEAR: - return (gettext("\tclear [-nF] <pool> [device]\n")); + return (gettext("\tclear [[--power]|[-nF]] <pool> [device]\n")); case HELP_CREATE: return (gettext("\tcreate [-fnd] [-o property=value] ... \n" "\t [-O file-system-property=value] ... \n" @@ -381,9 +381,11 @@ get_usage(zpool_help_t idx) "[-T d|u] [pool] ... \n" "\t [interval [count]]\n")); case HELP_OFFLINE: - return (gettext("\toffline [-f] [-t] <pool> <device> ...\n")); + return (gettext("\toffline [--power]|[[-f][-t]] <pool> " + "<device> ...\n")); case HELP_ONLINE: - return (gettext("\tonline [-e] <pool> <device> ...\n")); + return (gettext("\tonline [--power][-e] <pool> <device> " + "...\n")); case HELP_REPLACE: return (gettext("\treplace [-fsw] [-o property=value] " "<pool> <device> [new-device]\n")); @@ -392,7 +394,7 @@ get_usage(zpool_help_t idx) case HELP_REOPEN: return (gettext("\treopen [-n] <pool>\n")); case HELP_INITIALIZE: - return (gettext("\tinitialize [-c | -s] [-w] <pool> " + return (gettext("\tinitialize [-c | -s | -u] [-w] <pool> " "[<device> ...]\n")); case HELP_SCRUB: return (gettext("\tscrub [-s | -p] [-w] <pool> ...\n")); @@ -402,7 +404,7 @@ get_usage(zpool_help_t idx) return (gettext("\ttrim [-dw] [-r <rate>] [-c | -s] <pool> " "[<device> ...]\n")); case HELP_STATUS: - return (gettext("\tstatus [-c [script1,script2,...]] " + return (gettext("\tstatus [--power] [-c [script1,script2,...]] " "[-igLpPstvxD] [-T d|u] [pool] ... \n" "\t [interval [count]]\n")); case HELP_UPGRADE: @@ -486,6 +488,77 @@ print_prop_cb(int prop, void *cb) } /* + * Given a leaf vdev name like 'L5' return its VDEV_CONFIG_PATH like + * '/dev/disk/by-vdev/L5'. + */ +static const char * +vdev_name_to_path(zpool_handle_t *zhp, char *vdev) +{ + nvlist_t *vdev_nv = zpool_find_vdev(zhp, vdev, NULL, NULL, NULL); + if (vdev_nv == NULL) { + return (NULL); + } + return (fnvlist_lookup_string(vdev_nv, ZPOOL_CONFIG_PATH)); +} + +static int +zpool_power_on(zpool_handle_t *zhp, char *vdev) +{ + return (zpool_power(zhp, vdev, B_TRUE)); +} + +static int +zpool_power_on_and_disk_wait(zpool_handle_t *zhp, char *vdev) +{ + int rc; + + rc = zpool_power_on(zhp, vdev); + if (rc != 0) + return (rc); + + zpool_disk_wait(vdev_name_to_path(zhp, vdev)); + + return (0); +} + +static int +zpool_power_on_pool_and_wait_for_devices(zpool_handle_t *zhp) +{ + nvlist_t *nv; + const char *path = NULL; + int rc; + + /* Power up all the devices first */ + FOR_EACH_REAL_LEAF_VDEV(zhp, nv) { + path = fnvlist_lookup_string(nv, ZPOOL_CONFIG_PATH); + if (path != NULL) { + rc = zpool_power_on(zhp, (char *)path); + if (rc != 0) { + return (rc); + } + } + } + + /* + * Wait for their devices to show up. Since we powered them on + * at roughly the same time, they should all come online around + * the same time. + */ + FOR_EACH_REAL_LEAF_VDEV(zhp, nv) { + path = fnvlist_lookup_string(nv, ZPOOL_CONFIG_PATH); + zpool_disk_wait(path); + } + + return (0); +} + +static int +zpool_power_off(zpool_handle_t *zhp, char *vdev) +{ + return (zpool_power(zhp, vdev, B_FALSE)); +} + +/* * Display usage message. If we're inside a command, display only the usage for * that command. Otherwise, iterate over the entire command table and display * a complete usage message. @@ -519,7 +592,7 @@ usage(boolean_t requested) (strcmp(current_command->name, "get") == 0) || (strcmp(current_command->name, "list") == 0))) { - (void) fprintf(fp, + (void) fprintf(fp, "%s", gettext("\nthe following properties are supported:\n")); (void) fprintf(fp, "\n\t%-19s %s %s\n\n", @@ -548,12 +621,13 @@ usage(boolean_t requested) } /* - * zpool initialize [-c | -s] [-w] <pool> [<vdev> ...] + * zpool initialize [-c | -s | -u] [-w] <pool> [<vdev> ...] * Initialize all unused blocks in the specified vdevs, or all vdevs in the pool * if none specified. * * -c Cancel. Ends active initializing. * -s Suspend. Initializing can then be restarted with no flags. + * -u Uninitialize. Clears initialization state. * -w Wait. Blocks until initializing has completed. */ int @@ -569,12 +643,14 @@ zpool_do_initialize(int argc, char **argv) struct option long_options[] = { {"cancel", no_argument, NULL, 'c'}, {"suspend", no_argument, NULL, 's'}, + {"uninit", no_argument, NULL, 'u'}, {"wait", no_argument, NULL, 'w'}, {0, 0, 0, 0} }; pool_initialize_func_t cmd_type = POOL_INITIALIZE_START; - while ((c = getopt_long(argc, argv, "csw", long_options, NULL)) != -1) { + while ((c = getopt_long(argc, argv, "csuw", long_options, + NULL)) != -1) { switch (c) { case 'c': if (cmd_type != POOL_INITIALIZE_START && @@ -594,6 +670,15 @@ zpool_do_initialize(int argc, char **argv) } cmd_type = POOL_INITIALIZE_SUSPEND; break; + case 'u': + if (cmd_type != POOL_INITIALIZE_START && + cmd_type != POOL_INITIALIZE_UNINIT) { + (void) fprintf(stderr, gettext("-u cannot be " + "combined with other options\n")); + usage(B_FALSE); + } + cmd_type = POOL_INITIALIZE_UNINIT; + break; case 'w': wait = B_TRUE; break; @@ -620,8 +705,8 @@ zpool_do_initialize(int argc, char **argv) } if (wait && (cmd_type != POOL_INITIALIZE_START)) { - (void) fprintf(stderr, gettext("-w cannot be used with -c or " - "-s\n")); + (void) fprintf(stderr, gettext("-w cannot be used with -c, -s" + "or -u\n")); usage(B_FALSE); } @@ -2042,11 +2127,13 @@ typedef struct status_cbdata { boolean_t cb_explain; boolean_t cb_first; boolean_t cb_dedup_stats; + boolean_t cb_print_unhealthy; boolean_t cb_print_status; boolean_t cb_print_slow_ios; boolean_t cb_print_vdev_init; boolean_t cb_print_vdev_trim; vdev_cmd_data_list_t *vcdl; + boolean_t cb_print_power; } status_cbdata_t; /* Return 1 if string is NULL, empty, or whitespace; return 0 otherwise. */ @@ -2240,6 +2327,35 @@ health_str_to_color(const char *health) } /* + * Called for each leaf vdev. Returns 0 if the vdev is healthy. + * A vdev is unhealthy if any of the following are true: + * 1) there are read, write, or checksum errors, + * 2) its state is not ONLINE, or + * 3) slow IO reporting was requested (-s) and there are slow IOs. + */ +static int +vdev_health_check_cb(void *hdl_data, nvlist_t *nv, void *data) +{ + status_cbdata_t *cb = data; + vdev_stat_t *vs; + uint_t vsc; + (void) hdl_data; + + if (nvlist_lookup_uint64_array(nv, ZPOOL_CONFIG_VDEV_STATS, + (uint64_t **)&vs, &vsc) != 0) + return (1); + + if (vs->vs_checksum_errors || vs->vs_read_errors || + vs->vs_write_errors || vs->vs_state != VDEV_STATE_HEALTHY) + return (1); + + if (cb->cb_print_slow_ios && vs->vs_slow_ios) + return (1); + + return (0); +} + +/* * Print out configuration state as requested by status_callback. */ static void @@ -2257,7 +2373,8 @@ print_status_config(zpool_handle_t *zhp, status_cbdata_t *cb, const char *name, const char *state; char *type; char *path = NULL; - char *rcolor = NULL, *wcolor = NULL, *ccolor = NULL; + char *rcolor = NULL, *wcolor = NULL, *ccolor = NULL, + *scolor = NULL; if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_CHILDREN, &child, &children) != 0) @@ -2284,6 +2401,15 @@ print_status_config(zpool_handle_t *zhp, status_cbdata_t *cb, const char *name, state = gettext("AVAIL"); } + /* + * If '-e' is specified then top-level vdevs and their children + * can be pruned if all of their leaves are healthy. + */ + if (cb->cb_print_unhealthy && depth > 0 && + for_each_vdev_in_nvlist(nv, vdev_health_check_cb, cb) == 0) { + return; + } + printf_color(health_str_to_color(state), "\t%*s%-*s %-8s", depth, "", cb->cb_namewidth - depth, name, state); @@ -2298,6 +2424,9 @@ print_status_config(zpool_handle_t *zhp, status_cbdata_t *cb, const char *name, if (vs->vs_checksum_errors) ccolor = ANSI_RED; + if (vs->vs_slow_ios) + scolor = ANSI_BLUE; + if (cb->cb_literal) { printf(" "); printf_color(rcolor, "%5llu", @@ -2330,9 +2459,30 @@ print_status_config(zpool_handle_t *zhp, status_cbdata_t *cb, const char *name, } if (cb->cb_literal) - printf(" %5llu", (u_longlong_t)vs->vs_slow_ios); + printf_color(scolor, " %5llu", + (u_longlong_t)vs->vs_slow_ios); else - printf(" %5s", rbuf); + printf_color(scolor, " %5s", rbuf); + } + if (cb->cb_print_power) { + if (children == 0) { + /* Only leaf vdevs have physical slots */ + switch (zpool_power_current_state(zhp, (char *) + fnvlist_lookup_string(nv, + ZPOOL_CONFIG_PATH))) { + case 0: + printf_color(ANSI_RED, " %5s", + gettext("off")); + break; + case 1: + printf(" %5s", gettext("on")); + break; + default: + printf(" %5s", "-"); + } + } else { + printf(" %5s", "-"); + } } } @@ -5374,19 +5524,6 @@ get_interval_count_filter_guids(int *argc, char **argv, float *interval, } /* - * Floating point sleep(). Allows you to pass in a floating point value for - * seconds. - */ -static void -fsleep(float sec) -{ - struct timespec req; - req.tv_sec = floor(sec); - req.tv_nsec = (sec - (float)req.tv_sec) * NANOSEC; - nanosleep(&req, NULL); -} - -/* * Terminal height, in rows. Returns -1 if stdout is not connected to a TTY or * if we were unable to determine its size. */ @@ -6874,10 +7011,12 @@ zpool_do_split(int argc, char **argv) return (ret); } - +#define POWER_OPT 1024 /* - * zpool online <pool> <device> ... + * zpool online [--power] <pool> <device> ... + * + * --power: Power on the enclosure slot to the drive (if possible) */ int zpool_do_online(int argc, char **argv) @@ -6888,13 +7027,21 @@ zpool_do_online(int argc, char **argv) int ret = 0; vdev_state_t newstate; int flags = 0; + boolean_t is_power_on = B_FALSE; + struct option long_options[] = { + {"power", no_argument, NULL, POWER_OPT}, + {0, 0, 0, 0} + }; /* check options */ - while ((c = getopt(argc, argv, "e")) != -1) { + while ((c = getopt_long(argc, argv, "e", long_options, NULL)) != -1) { switch (c) { case 'e': flags |= ZFS_ONLINE_EXPAND; break; + case POWER_OPT: + is_power_on = B_TRUE; + break; case '?': (void) fprintf(stderr, gettext("invalid option '%c'\n"), optopt); @@ -6902,6 +7049,9 @@ zpool_do_online(int argc, char **argv) } } + if (libzfs_envvar_is_set("ZPOOL_AUTO_POWER_ON_SLOT")) + is_power_on = B_TRUE; + argc -= optind; argv += optind; @@ -6921,6 +7071,29 @@ zpool_do_online(int argc, char **argv) return (1); for (i = 1; i < argc; i++) { + vdev_state_t oldstate; + boolean_t avail_spare, l2cache; + int rc; + + if (is_power_on) { + rc = zpool_power_on_and_disk_wait(zhp, argv[i]); + if (rc == ENOTSUP) { + (void) fprintf(stderr, + gettext("Power control not supported\n")); + } + if (rc != 0) + return (rc); + } + + nvlist_t *tgt = zpool_find_vdev(zhp, argv[i], &avail_spare, + &l2cache, NULL); + if (tgt == NULL) { + ret = 1; + continue; + } + uint_t vsc; + oldstate = ((vdev_stat_t *)fnvlist_lookup_uint64_array(tgt, + ZPOOL_CONFIG_VDEV_STATS, &vsc))->vs_state; if (zpool_vdev_online(zhp, argv[i], flags, &newstate) == 0) { if (newstate != VDEV_STATE_HEALTHY) { (void) printf(gettext("warning: device '%s' " @@ -6934,6 +7107,17 @@ zpool_do_online(int argc, char **argv) (void) printf(gettext("use 'zpool " "replace' to replace devices " "that are no longer present\n")); + if ((flags & ZFS_ONLINE_EXPAND)) { + (void) printf(gettext("%s: failed " + "to expand usable space on " + "unhealthy device '%s'\n"), + (oldstate >= VDEV_STATE_DEGRADED ? + "error" : "warning"), argv[i]); + if (oldstate >= VDEV_STATE_DEGRADED) { + ret = 1; + break; + } + } } } else { ret = 1; @@ -6946,12 +7130,15 @@ zpool_do_online(int argc, char **argv) } /* - * zpool offline [-ft] <pool> <device> ... + * zpool offline [-ft]|[--power] <pool> <device> ... + * * * -f Force the device into a faulted state. * * -t Only take the device off-line temporarily. The offline/faulted * state will not be persistent across reboots. + * + * --power Power off the enclosure slot to the drive (if possible) */ /* ARGSUSED */ int @@ -6963,9 +7150,15 @@ zpool_do_offline(int argc, char **argv) int ret = 0; boolean_t istmp = B_FALSE; boolean_t fault = B_FALSE; + boolean_t is_power_off = B_FALSE; + + struct option long_options[] = { + {"power", no_argument, NULL, POWER_OPT}, + {0, 0, 0, 0} + }; /* check options */ - while ((c = getopt(argc, argv, "ft")) != -1) { + while ((c = getopt_long(argc, argv, "ft", long_options, NULL)) != -1) { switch (c) { case 'f': fault = B_TRUE; @@ -6973,6 +7166,9 @@ zpool_do_offline(int argc, char **argv) case 't': istmp = B_TRUE; break; + case POWER_OPT: + is_power_off = B_TRUE; + break; case '?': (void) fprintf(stderr, gettext("invalid option '%c'\n"), optopt); @@ -6980,6 +7176,20 @@ zpool_do_offline(int argc, char **argv) } } + if (is_power_off && fault) { + (void) fprintf(stderr, + gettext("-0 and -f cannot be used together\n")); + usage(B_FALSE); + return (1); + } + + if (is_power_off && istmp) { + (void) fprintf(stderr, + gettext("-0 and -t cannot be used together\n")); + usage(B_FALSE); + return (1); + } + argc -= optind; argv += optind; @@ -6999,8 +7209,22 @@ zpool_do_offline(int argc, char **argv) return (1); for (i = 1; i < argc; i++) { - if (fault) { - uint64_t guid = zpool_vdev_path_to_guid(zhp, argv[i]); + uint64_t guid = zpool_vdev_path_to_guid(zhp, argv[i]); + if (is_power_off) { + /* + * Note: we have to power off first, then set REMOVED, + * or else zpool_vdev_set_removed_state() returns + * EAGAIN. + */ + ret = zpool_power_off(zhp, argv[i]); + if (ret != 0) { + (void) fprintf(stderr, "%s %s %d\n", + gettext("unable to power off slot for"), + argv[i], ret); + } + zpool_vdev_set_removed_state(zhp, guid, VDEV_AUX_NONE); + + } else if (fault) { vdev_aux_t aux; if (istmp == B_FALSE) { /* Force the fault to persist across imports */ @@ -7023,7 +7247,7 @@ zpool_do_offline(int argc, char **argv) } /* - * zpool clear <pool> [device] + * zpool clear [-nF]|[--power] <pool> [device] * * Clear all errors associated with a pool or a particular device. */ @@ -7035,13 +7259,20 @@ zpool_do_clear(int argc, char **argv) boolean_t dryrun = B_FALSE; boolean_t do_rewind = B_FALSE; boolean_t xtreme_rewind = B_FALSE; + boolean_t is_power_on = B_FALSE; uint32_t rewind_policy = ZPOOL_NO_REWIND; nvlist_t *policy = NULL; zpool_handle_t *zhp; char *pool, *device; + struct option long_options[] = { + {"power", no_argument, NULL, POWER_OPT}, + {0, 0, 0, 0} + }; + /* check options */ - while ((c = getopt(argc, argv, "FnX")) != -1) { + while ((c = getopt_long(argc, argv, "FnX", long_options, + NULL)) != -1) { switch (c) { case 'F': do_rewind = B_TRUE; @@ -7052,6 +7283,9 @@ zpool_do_clear(int argc, char **argv) case 'X': xtreme_rewind = B_TRUE; break; + case POWER_OPT: + is_power_on = B_TRUE; + break; case '?': (void) fprintf(stderr, gettext("invalid option '%c'\n"), optopt); @@ -7059,6 +7293,9 @@ zpool_do_clear(int argc, char **argv) } } + if (libzfs_envvar_is_set("ZPOOL_AUTO_POWER_ON_SLOT")) + is_power_on = B_TRUE; + argc -= optind; argv += optind; @@ -7099,6 +7336,14 @@ zpool_do_clear(int argc, char **argv) return (1); } + if (is_power_on) { + if (device == NULL) { + zpool_power_on_pool_and_wait_for_devices(zhp); + } else { + zpool_power_on_and_disk_wait(zhp, device); + } + } + if (zpool_clear(zhp, device, policy) != 0) ret = 1; @@ -7549,19 +7794,20 @@ print_scan_scrub_resilver_status(pool_scan_stat_t *ps) zfs_nicebytes(ps->pss_processed, processed_buf, sizeof (processed_buf)); - assert(ps->pss_func == POOL_SCAN_SCRUB || - ps->pss_func == POOL_SCAN_RESILVER); + int is_resilver = ps->pss_func == POOL_SCAN_RESILVER; + int is_scrub = ps->pss_func == POOL_SCAN_SCRUB; + assert(is_resilver || is_scrub); /* Scan is finished or canceled. */ if (ps->pss_state == DSS_FINISHED) { secs_to_dhms(end - start, time_buf); - if (ps->pss_func == POOL_SCAN_SCRUB) { + if (is_scrub) { (void) printf(gettext("scrub repaired %s " "in %s with %llu errors on %s"), processed_buf, time_buf, (u_longlong_t)ps->pss_errors, ctime(&end)); - } else if (ps->pss_func == POOL_SCAN_RESILVER) { + } else if (is_resilver) { (void) printf(gettext("resilvered %s " "in %s with %llu errors on %s"), processed_buf, time_buf, (u_longlong_t)ps->pss_errors, @@ -7569,10 +7815,10 @@ print_scan_scrub_resilver_status(pool_scan_stat_t *ps) } return; } else if (ps->pss_state == DSS_CANCELED) { - if (ps->pss_func == POOL_SCAN_SCRUB) { + if (is_scrub) { (void) printf(gettext("scrub canceled on %s"), ctime(&end)); - } else if (ps->pss_func == POOL_SCAN_RESILVER) { + } else if (is_resilver) { (void) printf(gettext("resilver canceled on %s"), ctime(&end)); } @@ -7582,7 +7828,7 @@ print_scan_scrub_resilver_status(pool_scan_stat_t *ps) assert(ps->pss_state == DSS_SCANNING); /* Scan is in progress. Resilvers can't be paused. */ - if (ps->pss_func == POOL_SCAN_SCRUB) { + if (is_scrub) { if (pause == 0) { (void) printf(gettext("scrub in progress since %s"), ctime(&start)); @@ -7592,7 +7838,7 @@ print_scan_scrub_resilver_status(pool_scan_stat_t *ps) (void) printf(gettext("\tscrub started on %s"), ctime(&start)); } - } else if (ps->pss_func == POOL_SCAN_RESILVER) { + } else if (is_resilver) { (void) printf(gettext("resilver in progress since %s"), ctime(&start)); } @@ -7634,17 +7880,27 @@ print_scan_scrub_resilver_status(pool_scan_stat_t *ps) scanned_buf, issued_buf, total_buf); } - if (ps->pss_func == POOL_SCAN_RESILVER) { + if (is_resilver) { (void) printf(gettext("\t%s resilvered, %.2f%% done"), processed_buf, 100 * fraction_done); - } else if (ps->pss_func == POOL_SCAN_SCRUB) { + } else if (is_scrub) { (void) printf(gettext("\t%s repaired, %.2f%% done"), processed_buf, 100 * fraction_done); } if (pause == 0) { + /* + * Only provide an estimate iff: + * 1) the time remaining is valid, and + * 2) the issue rate exceeds 10 MB/s, and + * 3) it's either: + * a) a resilver which has started repairs, or + * b) a scrub which has entered the issue phase. + */ if (total_secs_left != UINT64_MAX && - issue_rate >= 10 * 1024 * 1024) { + issue_rate >= 10 * 1024 * 1024 && + ((is_resilver && ps->pss_processed > 0) || + (is_scrub && issued > 0))) { (void) printf(gettext(", %s to go\n"), time_buf); } else { (void) printf(gettext(", no estimated " @@ -8596,6 +8852,10 @@ status_callback(zpool_handle_t *zhp, void *data) printf_color(ANSI_BOLD, " %5s", gettext("SLOW")); } + if (cbp->cb_print_power) { + printf_color(ANSI_BOLD, " %5s", gettext("POWER")); + } + if (cbp->vcdl != NULL) print_cmd_columns(cbp->vcdl, 0); @@ -8643,11 +8903,13 @@ status_callback(zpool_handle_t *zhp, void *data) if (nerr == 0) (void) printf(gettext("errors: No known data " "errors\n")); - else if (!cbp->cb_verbose) + else if (!cbp->cb_verbose) { + color_start(ANSI_RED); (void) printf(gettext("errors: %llu data " "errors, use '-v' for a list\n"), (u_longlong_t)nerr); - else + color_end(); + } else print_error_log(zhp); } @@ -8662,10 +8924,11 @@ status_callback(zpool_handle_t *zhp, void *data) } /* - * zpool status [-c [script1,script2,...]] [-igLpPstvx] [-T d|u] [pool] ... - * [interval [count]] + * zpool status [-c [script1,script2,...]] [-igLpPstvx] [--power] [-T d|u] ... + * [pool] [interval [count]] * * -c CMD For each vdev, run command CMD + * -e Display only unhealthy vdevs * -i Display vdev initialization status. * -g Display guid for individual vdev name. * -L Follow links when resolving vdev path name. @@ -8677,6 +8940,7 @@ status_callback(zpool_handle_t *zhp, void *data) * -D Display dedup status (undocumented) * -t Display vdev TRIM status. * -T Display a timestamp in date(1) or Unix format + * --power Display vdev enclosure slot power status * * Describes the health status of all pools or some subset. */ @@ -8690,8 +8954,14 @@ zpool_do_status(int argc, char **argv) status_cbdata_t cb = { 0 }; char *cmd = NULL; + struct option long_options[] = { + {"power", no_argument, NULL, POWER_OPT}, + {0, 0, 0, 0} + }; + /* check options */ - while ((c = getopt(argc, argv, "c:igLpPsvxDtT:")) != -1) { + while ((c = getopt_long(argc, argv, "c:eigLpPsvxDtT:", long_options, + NULL)) != -1) { switch (c) { case 'c': if (cmd != NULL) { @@ -8717,6 +8987,9 @@ zpool_do_status(int argc, char **argv) } cmd = optarg; break; + case 'e': + cb.cb_print_unhealthy = B_TRUE; + break; case 'i': cb.cb_print_vdev_init = B_TRUE; break; @@ -8750,6 +9023,9 @@ zpool_do_status(int argc, char **argv) case 'T': get_timestamp_arg(*optarg); break; + case POWER_OPT: + cb.cb_print_power = B_TRUE; + break; case '?': if (optopt == 'c') { print_zpool_script_list("status"); diff --git a/cmd/zpool/zpool_util.h b/cmd/zpool/zpool_util.h index da75866f5145..8fb389d6113f 100644 --- a/cmd/zpool/zpool_util.h +++ b/cmd/zpool/zpool_util.h @@ -124,6 +124,10 @@ vdev_cmd_data_list_t *all_pools_for_each_vdev_run(int argc, char **argv, void free_vdev_cmd_data_list(vdev_cmd_data_list_t *vcdl); +void free_vdev_cmd_data(vdev_cmd_data_t *data); + +int vdev_run_cmd_simple(char *path, char *cmd); + int check_device(const char *path, boolean_t force, boolean_t isspare, boolean_t iswholedisk); boolean_t check_sector_size_database(char *path, int *sector_size); @@ -131,6 +135,9 @@ void vdev_error(const char *fmt, ...); int check_file(const char *file, boolean_t force, boolean_t isspare); void after_zpool_upgrade(zpool_handle_t *zhp); +int zpool_power(zpool_handle_t *zhp, char *vdev, boolean_t turn_on); +int zpool_power_current_state(zpool_handle_t *zhp, char *vdev); + #ifdef __cplusplus } #endif diff --git a/cmd/zpool/zpool_vdev.c b/cmd/zpool/zpool_vdev.c index 3d83da641ecb..c8daa730e721 100644 --- a/cmd/zpool/zpool_vdev.c +++ b/cmd/zpool/zpool_vdev.c @@ -373,6 +373,10 @@ make_leaf_vdev(nvlist_t *props, const char *arg, boolean_t is_primary) verify(nvlist_add_string(vdev, ZPOOL_CONFIG_PATH, path) == 0); verify(nvlist_add_string(vdev, ZPOOL_CONFIG_TYPE, type) == 0); + /* Lookup and add the enclosure sysfs path (if exists) */ + update_vdev_config_dev_sysfs_path(vdev, path, + ZPOOL_CONFIG_VDEV_ENC_SYSFS_PATH); + if (strcmp(type, VDEV_TYPE_DISK) == 0) verify(nvlist_add_uint64(vdev, ZPOOL_CONFIG_WHOLE_DISK, (uint64_t)wholedisk) == 0); @@ -921,6 +925,15 @@ zero_label(char *path) return (0); } +static void +lines_to_stderr(char *lines[], int lines_cnt) +{ + int i; + for (i = 0; i < lines_cnt; i++) { + fprintf(stderr, "%s\n", lines[i]); + } +} + /* * Go through and find any whole disks in the vdev specification, labelling them * as appropriate. When constructing the vdev spec, we were unable to open this @@ -932,7 +945,7 @@ zero_label(char *path) * need to get the devid after we label the disk. */ static int -make_disks(zpool_handle_t *zhp, nvlist_t *nv) +make_disks(zpool_handle_t *zhp, nvlist_t *nv, boolean_t replacing) { nvlist_t **child; uint_t c, children; @@ -1017,6 +1030,8 @@ make_disks(zpool_handle_t *zhp, nvlist_t *nv) */ if (!is_exclusive && !is_spare(NULL, udevpath)) { char *devnode = strrchr(devpath, '/') + 1; + char **lines = NULL; + int lines_cnt = 0; ret = strncmp(udevpath, UDISK_ROOT, strlen(UDISK_ROOT)); if (ret == 0) { @@ -1028,9 +1043,27 @@ make_disks(zpool_handle_t *zhp, nvlist_t *nv) /* * When labeling a pool the raw device node name * is provided as it appears under /dev/. + * + * Note that 'zhp' will be NULL when we're creating a + * pool. */ - if (zpool_label_disk(g_zfs, zhp, devnode) == -1) + if (zpool_prepare_and_label_disk(g_zfs, zhp, devnode, + nv, zhp == NULL ? "create" : + replacing ? "replace" : "add", &lines, + &lines_cnt) != 0) { + (void) fprintf(stderr, + gettext( + "Error preparing/labeling disk.\n")); + if (lines_cnt > 0) { + (void) fprintf(stderr, + gettext("zfs_prepare_disk output:\n")); + lines_to_stderr(lines, lines_cnt); + } + + libzfs_free_str_array(lines, lines_cnt); return (-1); + } + libzfs_free_str_array(lines, lines_cnt); /* * Wait for udev to signal the device is available @@ -1067,19 +1100,19 @@ make_disks(zpool_handle_t *zhp, nvlist_t *nv) } for (c = 0; c < children; c++) - if ((ret = make_disks(zhp, child[c])) != 0) + if ((ret = make_disks(zhp, child[c], replacing)) != 0) return (ret); if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_SPARES, &child, &children) == 0) for (c = 0; c < children; c++) - if ((ret = make_disks(zhp, child[c])) != 0) + if ((ret = make_disks(zhp, child[c], replacing)) != 0) return (ret); if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_L2CACHE, &child, &children) == 0) for (c = 0; c < children; c++) - if ((ret = make_disks(zhp, child[c])) != 0) + if ((ret = make_disks(zhp, child[c], replacing)) != 0) return (ret); return (0); @@ -1740,7 +1773,7 @@ split_mirror_vdev(zpool_handle_t *zhp, char *newname, nvlist_t *props, return (NULL); } - if (!flags.dryrun && make_disks(zhp, newroot) != 0) { + if (!flags.dryrun && make_disks(zhp, newroot, B_FALSE) != 0) { nvlist_free(newroot); return (NULL); } @@ -1861,7 +1894,7 @@ make_root_vdev(zpool_handle_t *zhp, nvlist_t *props, int force, int check_rep, /* * Run through the vdev specification and label any whole disks found. */ - if (!dryrun && make_disks(zhp, newroot) != 0) { + if (!dryrun && make_disks(zhp, newroot, replacing) != 0) { nvlist_free(newroot); return (NULL); } diff --git a/config/Rules.am b/config/Rules.am index 3b24e3630102..8c484d18c2a0 100644 --- a/config/Rules.am +++ b/config/Rules.am @@ -41,6 +41,7 @@ AM_CPPFLAGS += -D_REENTRANT AM_CPPFLAGS += -D_FILE_OFFSET_BITS=64 AM_CPPFLAGS += -D_LARGEFILE64_SOURCE AM_CPPFLAGS += -DLIBEXECDIR=\"$(libexecdir)\" +AM_CPPFLAGS += -DZFSEXECDIR=\"$(zfsexecdir)\" AM_CPPFLAGS += -DRUNSTATEDIR=\"$(runstatedir)\" AM_CPPFLAGS += -DSBINDIR=\"$(sbindir)\" AM_CPPFLAGS += -DSYSCONFDIR=\"$(sysconfdir)\" diff --git a/config/always-compiler-options.m4 b/config/always-compiler-options.m4 index 5046ce0ddb83..fb78c93d4895 100644 --- a/config/always-compiler-options.m4 +++ b/config/always-compiler-options.m4 @@ -181,6 +181,62 @@ AC_DEFUN([ZFS_AC_CONFIG_ALWAYS_CC_INFINITE_RECURSION], [ ]) dnl # +dnl # Check if kernel cc supports -Winfinite-recursion option. +dnl # +AC_DEFUN([ZFS_AC_CONFIG_ALWAYS_KERNEL_CC_INFINITE_RECURSION], [ + AC_MSG_CHECKING([whether $KERNEL_CC supports -Winfinite-recursion]) + + saved_cc="$CC" + saved_flags="$CFLAGS" + CC="gcc" + CFLAGS="$CFLAGS -Werror -Winfinite-recursion" + + AS_IF([ test -n "$KERNEL_CC" ], [ + CC="$KERNEL_CC" + ]) + AS_IF([ test -n "$KERNEL_LLVM" ], [ + CC="clang" + ]) + + AC_COMPILE_IFELSE([AC_LANG_PROGRAM([], [])], [ + KERNEL_INFINITE_RECURSION=-Winfinite-recursion + AC_DEFINE([HAVE_KERNEL_INFINITE_RECURSION], 1, + [Define if compiler supports -Winfinite-recursion]) + AC_MSG_RESULT([yes]) + ], [ + KERNEL_INFINITE_RECURSION= + AC_MSG_RESULT([no]) + ]) + + CC="$saved_cc" + CFLAGS="$saved_flags" + AC_SUBST([KERNEL_INFINITE_RECURSION]) +]) + +dnl # +dnl # Check if cc supports -Wformat-overflow option. +dnl # +AC_DEFUN([ZFS_AC_CONFIG_ALWAYS_CC_FORMAT_OVERFLOW], [ + AC_MSG_CHECKING([whether $CC supports -Wformat-overflow]) + + saved_flags="$CFLAGS" + CFLAGS="$CFLAGS -Werror -Wformat-overflow" + + AC_COMPILE_IFELSE([AC_LANG_PROGRAM([], [])], [ + FORMAT_OVERFLOW=-Wformat-overflow + AC_DEFINE([HAVE_FORMAT_OVERFLOW], 1, + [Define if compiler supports -Wformat-overflow]) + AC_MSG_RESULT([yes]) + ], [ + FORMAT_OVERFLOW= + AC_MSG_RESULT([no]) + ]) + + CFLAGS="$saved_flags" + AC_SUBST([FORMAT_OVERFLOW]) +]) + +dnl # dnl # Check if gcc supports -fno-omit-frame-pointer option. dnl # AC_DEFUN([ZFS_AC_CONFIG_ALWAYS_CC_NO_OMIT_FRAME_POINTER], [ @@ -221,3 +277,34 @@ AC_DEFUN([ZFS_AC_CONFIG_ALWAYS_CC_NO_IPA_SRA], [ CFLAGS="$saved_flags" AC_SUBST([NO_IPA_SRA]) ]) + +dnl # +dnl # Check if kernel cc supports -fno-ipa-sra option. +dnl # +AC_DEFUN([ZFS_AC_CONFIG_ALWAYS_KERNEL_CC_NO_IPA_SRA], [ + AC_MSG_CHECKING([whether $KERNEL_CC supports -fno-ipa-sra]) + + saved_cc="$CC" + saved_flags="$CFLAGS" + CC="gcc" + CFLAGS="$CFLAGS -Werror -fno-ipa-sra" + + AS_IF([ test -n "$KERNEL_CC" ], [ + CC="$KERNEL_CC" + ]) + AS_IF([ test -n "$KERNEL_LLVM" ], [ + CC="clang" + ]) + + AC_COMPILE_IFELSE([AC_LANG_PROGRAM([], [])], [ + KERNEL_NO_IPA_SRA=-fno-ipa-sra + AC_MSG_RESULT([yes]) + ], [ + KERNEL_NO_IPA_SRA= + AC_MSG_RESULT([no]) + ]) + + CC="$saved_cc" + CFLAGS="$saved_flags" + AC_SUBST([KERNEL_NO_IPA_SRA]) +]) diff --git a/config/kernel-acl.m4 b/config/kernel-acl.m4 index 6e92da97d0fe..3ae5dc6b6dbc 100644 --- a/config/kernel-acl.m4 +++ b/config/kernel-acl.m4 @@ -172,7 +172,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_INODE_OPERATIONS_GET_ACL], [ ZFS_LINUX_TEST_SRC([inode_operations_get_acl], [ #include <linux/fs.h> - struct posix_acl *get_acl_fn(struct inode *inode, int type) + static struct posix_acl *get_acl_fn(struct inode *inode, int type) { return NULL; } static const struct inode_operations @@ -184,7 +184,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_INODE_OPERATIONS_GET_ACL], [ ZFS_LINUX_TEST_SRC([inode_operations_get_acl_rcu], [ #include <linux/fs.h> - struct posix_acl *get_acl_fn(struct inode *inode, int type, + static struct posix_acl *get_acl_fn(struct inode *inode, int type, bool rcu) { return NULL; } static const struct inode_operations @@ -196,7 +196,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_INODE_OPERATIONS_GET_ACL], [ ZFS_LINUX_TEST_SRC([inode_operations_get_inode_acl], [ #include <linux/fs.h> - struct posix_acl *get_inode_acl_fn(struct inode *inode, int type, + static struct posix_acl *get_inode_acl_fn(struct inode *inode, int type, bool rcu) { return NULL; } static const struct inode_operations @@ -236,11 +236,26 @@ dnl # dnl # 6.2 API change, dnl # set_acl() second paramter changed to a struct dentry * dnl # +dnl # 6.3 API change, +dnl # set_acl() first parameter changed to struct mnt_idmap * +dnl # AC_DEFUN([ZFS_AC_KERNEL_SRC_INODE_OPERATIONS_SET_ACL], [ + ZFS_LINUX_TEST_SRC([inode_operations_set_acl_mnt_idmap_dentry], [ + #include <linux/fs.h> + + static int set_acl_fn(struct mnt_idmap *idmap, + struct dentry *dent, struct posix_acl *acl, + int type) { return 0; } + + static const struct inode_operations + iops __attribute__ ((unused)) = { + .set_acl = set_acl_fn, + }; + ],[]) ZFS_LINUX_TEST_SRC([inode_operations_set_acl_userns_dentry], [ #include <linux/fs.h> - int set_acl_fn(struct user_namespace *userns, + static int set_acl_fn(struct user_namespace *userns, struct dentry *dent, struct posix_acl *acl, int type) { return 0; } @@ -252,7 +267,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_INODE_OPERATIONS_SET_ACL], [ ZFS_LINUX_TEST_SRC([inode_operations_set_acl_userns], [ #include <linux/fs.h> - int set_acl_fn(struct user_namespace *userns, + static int set_acl_fn(struct user_namespace *userns, struct inode *inode, struct posix_acl *acl, int type) { return 0; } @@ -264,7 +279,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_INODE_OPERATIONS_SET_ACL], [ ZFS_LINUX_TEST_SRC([inode_operations_set_acl], [ #include <linux/fs.h> - int set_acl_fn(struct inode *inode, struct posix_acl *acl, + static int set_acl_fn(struct inode *inode, struct posix_acl *acl, int type) { return 0; } static const struct inode_operations @@ -281,17 +296,24 @@ AC_DEFUN([ZFS_AC_KERNEL_INODE_OPERATIONS_SET_ACL], [ AC_DEFINE(HAVE_SET_ACL, 1, [iops->set_acl() exists]) AC_DEFINE(HAVE_SET_ACL_USERNS, 1, [iops->set_acl() takes 4 args]) ],[ - ZFS_LINUX_TEST_RESULT([inode_operations_set_acl_userns_dentry], [ + ZFS_LINUX_TEST_RESULT([inode_operations_set_acl_mnt_idmap_dentry], [ AC_MSG_RESULT(yes) AC_DEFINE(HAVE_SET_ACL, 1, [iops->set_acl() exists]) - AC_DEFINE(HAVE_SET_ACL_USERNS_DENTRY_ARG2, 1, - [iops->set_acl() takes 4 args, arg2 is struct dentry *]) + AC_DEFINE(HAVE_SET_ACL_IDMAP_DENTRY, 1, + [iops->set_acl() takes 4 args, arg1 is struct mnt_idmap *]) ],[ - ZFS_LINUX_TEST_RESULT([inode_operations_set_acl], [ + ZFS_LINUX_TEST_RESULT([inode_operations_set_acl_userns_dentry], [ AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_SET_ACL, 1, [iops->set_acl() exists, takes 3 args]) + AC_DEFINE(HAVE_SET_ACL, 1, [iops->set_acl() exists]) + AC_DEFINE(HAVE_SET_ACL_USERNS_DENTRY_ARG2, 1, + [iops->set_acl() takes 4 args, arg2 is struct dentry *]) ],[ - ZFS_LINUX_REQUIRE_API([i_op->set_acl()], [3.14]) + ZFS_LINUX_TEST_RESULT([inode_operations_set_acl], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_SET_ACL, 1, [iops->set_acl() exists, takes 3 args]) + ],[ + ZFS_LINUX_REQUIRE_API([i_op->set_acl()], [3.14]) + ]) ]) ]) ]) diff --git a/config/kernel-automount.m4 b/config/kernel-automount.m4 index f7bb63c68154..52f1931b748e 100644 --- a/config/kernel-automount.m4 +++ b/config/kernel-automount.m4 @@ -8,7 +8,7 @@ dnl # AC_DEFUN([ZFS_AC_KERNEL_SRC_AUTOMOUNT], [ ZFS_LINUX_TEST_SRC([dentry_operations_d_automount], [ #include <linux/dcache.h> - struct vfsmount *d_automount(struct path *p) { return NULL; } + static struct vfsmount *d_automount(struct path *p) { return NULL; } struct dentry_operations dops __attribute__ ((unused)) = { .d_automount = d_automount, }; diff --git a/config/kernel-bio.m4 b/config/kernel-bio.m4 index 18620ca5b7e4..b22c1a3de7e1 100644 --- a/config/kernel-bio.m4 +++ b/config/kernel-bio.m4 @@ -247,7 +247,7 @@ dnl # AC_DEFUN([ZFS_AC_KERNEL_SRC_BIO_END_IO_T_ARGS], [ ZFS_LINUX_TEST_SRC([bio_end_io_t_args], [ #include <linux/bio.h> - void wanted_end_io(struct bio *bio) { return; } + static void wanted_end_io(struct bio *bio) { return; } bio_end_io_t *end_io __attribute__ ((unused)) = wanted_end_io; ], []) ]) diff --git a/config/kernel-blkdev.m4 b/config/kernel-blkdev.m4 index 28e5364581ea..8e9e638b125a 100644 --- a/config/kernel-blkdev.m4 +++ b/config/kernel-blkdev.m4 @@ -16,12 +16,90 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV_GET_BY_PATH], [ ]) ]) +dnl # +dnl # 6.5.x API change, +dnl # blkdev_get_by_path() takes 4 args +dnl # +AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV_GET_BY_PATH_4ARG], [ + ZFS_LINUX_TEST_SRC([blkdev_get_by_path_4arg], [ + #include <linux/fs.h> + #include <linux/blkdev.h> + ], [ + struct block_device *bdev __attribute__ ((unused)) = NULL; + const char *path = "path"; + fmode_t mode = 0; + void *holder = NULL; + struct blk_holder_ops h; + + bdev = blkdev_get_by_path(path, mode, holder, &h); + ]) +]) + +dnl # +dnl # 6.8.x API change +dnl # bdev_open_by_path() replaces blkdev_get_by_path() +dnl # +AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV_BDEV_OPEN_BY_PATH], [ + ZFS_LINUX_TEST_SRC([bdev_open_by_path], [ + #include <linux/fs.h> + #include <linux/blkdev.h> + ], [ + struct bdev_handle *bdh __attribute__ ((unused)) = NULL; + const char *path = "path"; + fmode_t mode = 0; + void *holder = NULL; + struct blk_holder_ops h; + + bdh = bdev_open_by_path(path, mode, holder, &h); + ]) +]) + AC_DEFUN([ZFS_AC_KERNEL_BLKDEV_GET_BY_PATH], [ - AC_MSG_CHECKING([whether blkdev_get_by_path() exists]) + AC_MSG_CHECKING([whether blkdev_get_by_path() exists and takes 3 args]) ZFS_LINUX_TEST_RESULT([blkdev_get_by_path], [ AC_MSG_RESULT(yes) ], [ - ZFS_LINUX_TEST_ERROR([blkdev_get_by_path()]) + AC_MSG_RESULT(no) + AC_MSG_CHECKING([whether blkdev_get_by_path() exists and takes 4 args]) + ZFS_LINUX_TEST_RESULT([blkdev_get_by_path_4arg], [ + AC_DEFINE(HAVE_BLKDEV_GET_BY_PATH_4ARG, 1, + [blkdev_get_by_path() exists and takes 4 args]) + AC_MSG_RESULT(yes) + ], [ + AC_MSG_RESULT(no) + AC_MSG_CHECKING([whether bdev_open_by_path() exists]) + ZFS_LINUX_TEST_RESULT([bdev_open_by_path], [ + AC_DEFINE(HAVE_BDEV_OPEN_BY_PATH, 1, + [bdev_open_by_path() exists]) + AC_MSG_RESULT(yes) + ], [ + ZFS_LINUX_TEST_ERROR([blkdev_get_by_path()]) + ]) + ]) + ]) +]) + +dnl # +dnl # 6.5.x API change +dnl # blk_mode_t was added as a type to supercede some places where fmode_t +dnl # is used +dnl # +AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV_BLK_MODE_T], [ + ZFS_LINUX_TEST_SRC([blk_mode_t], [ + #include <linux/fs.h> + #include <linux/blkdev.h> + ], [ + blk_mode_t m __attribute((unused)) = (blk_mode_t)0; + ]) +]) + +AC_DEFUN([ZFS_AC_KERNEL_BLKDEV_BLK_MODE_T], [ + AC_MSG_CHECKING([whether blk_mode_t is defined]) + ZFS_LINUX_TEST_RESULT([blk_mode_t], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_BLK_MODE_T, 1, [blk_mode_t is defined]) + ], [ + AC_MSG_RESULT(no) ]) ]) @@ -41,12 +119,58 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV_PUT], [ ]) ]) +dnl # +dnl # 6.5.x API change. +dnl # blkdev_put() takes (void* holder) as arg 2 +dnl # +AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV_PUT_HOLDER], [ + ZFS_LINUX_TEST_SRC([blkdev_put_holder], [ + #include <linux/fs.h> + #include <linux/blkdev.h> + ], [ + struct block_device *bdev = NULL; + void *holder = NULL; + + blkdev_put(bdev, holder); + ]) +]) + +dnl # +dnl # 6.8.x API change +dnl # bdev_release() replaces blkdev_put() +dnl # +AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV_BDEV_RELEASE], [ + ZFS_LINUX_TEST_SRC([bdev_release], [ + #include <linux/fs.h> + #include <linux/blkdev.h> + ], [ + struct bdev_handle *bdh = NULL; + bdev_release(bdh); + ]) +]) + AC_DEFUN([ZFS_AC_KERNEL_BLKDEV_PUT], [ AC_MSG_CHECKING([whether blkdev_put() exists]) ZFS_LINUX_TEST_RESULT([blkdev_put], [ AC_MSG_RESULT(yes) ], [ - ZFS_LINUX_TEST_ERROR([blkdev_put()]) + AC_MSG_RESULT(no) + AC_MSG_CHECKING([whether blkdev_put() accepts void* as arg 2]) + ZFS_LINUX_TEST_RESULT([blkdev_put_holder], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_BLKDEV_PUT_HOLDER, 1, + [blkdev_put() accepts void* as arg 2]) + ], [ + AC_MSG_RESULT(no) + AC_MSG_CHECKING([whether bdev_release() exists]) + ZFS_LINUX_TEST_RESULT([bdev_release], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_BDEV_RELEASE, 1, + [bdev_release() exists]) + ], [ + ZFS_LINUX_TEST_ERROR([blkdev_put()]) + ]) + ]) ]) ]) @@ -104,6 +228,33 @@ AC_DEFUN([ZFS_AC_KERNEL_BLKDEV_CHECK_DISK_CHANGE], [ ]) dnl # +dnl # 6.5.x API change +dnl # disk_check_media_change() was added +dnl # +AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV_DISK_CHECK_MEDIA_CHANGE], [ + ZFS_LINUX_TEST_SRC([disk_check_media_change], [ + #include <linux/fs.h> + #include <linux/blkdev.h> + ], [ + struct block_device *bdev = NULL; + bool error; + + error = disk_check_media_change(bdev->bd_disk); + ]) +]) + +AC_DEFUN([ZFS_AC_KERNEL_BLKDEV_DISK_CHECK_MEDIA_CHANGE], [ + AC_MSG_CHECKING([whether disk_check_media_change() exists]) + ZFS_LINUX_TEST_RESULT([disk_check_media_change], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_DISK_CHECK_MEDIA_CHANGE, 1, + [disk_check_media_change() exists]) + ], [ + AC_MSG_RESULT(no) + ]) +]) + +dnl # dnl # bdev_kobj() is introduced from 5.12 dnl # AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV_BDEV_KOBJ], [ @@ -443,9 +594,36 @@ AC_DEFUN([ZFS_AC_KERNEL_BLKDEV_GET_ERESTARTSYS], [ ]) ]) +dnl # +dnl # 6.5.x API change +dnl # BLK_STS_NEXUS replaced with BLK_STS_RESV_CONFLICT +dnl # +AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV_BLK_STS_RESV_CONFLICT], [ + ZFS_LINUX_TEST_SRC([blk_sts_resv_conflict], [ + #include <linux/blkdev.h> + ],[ + blk_status_t s __attribute__ ((unused)) = BLK_STS_RESV_CONFLICT; + ]) +]) + +AC_DEFUN([ZFS_AC_KERNEL_BLKDEV_BLK_STS_RESV_CONFLICT], [ + AC_MSG_CHECKING([whether BLK_STS_RESV_CONFLICT is defined]) + ZFS_LINUX_TEST_RESULT([blk_sts_resv_conflict], [ + AC_DEFINE(HAVE_BLK_STS_RESV_CONFLICT, 1, [BLK_STS_RESV_CONFLICT is defined]) + AC_MSG_RESULT(yes) + ], [ + AC_MSG_RESULT(no) + ]) + ]) +]) + AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV], [ ZFS_AC_KERNEL_SRC_BLKDEV_GET_BY_PATH + ZFS_AC_KERNEL_SRC_BLKDEV_GET_BY_PATH_4ARG + ZFS_AC_KERNEL_SRC_BLKDEV_BDEV_OPEN_BY_PATH ZFS_AC_KERNEL_SRC_BLKDEV_PUT + ZFS_AC_KERNEL_SRC_BLKDEV_PUT_HOLDER + ZFS_AC_KERNEL_SRC_BLKDEV_BDEV_RELEASE ZFS_AC_KERNEL_SRC_BLKDEV_REREAD_PART ZFS_AC_KERNEL_SRC_BLKDEV_INVALIDATE_BDEV ZFS_AC_KERNEL_SRC_BLKDEV_LOOKUP_BDEV @@ -458,6 +636,9 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV], [ ZFS_AC_KERNEL_SRC_BLKDEV_ISSUE_SECURE_ERASE ZFS_AC_KERNEL_SRC_BLKDEV_BDEV_KOBJ ZFS_AC_KERNEL_SRC_BLKDEV_PART_TO_DEV + ZFS_AC_KERNEL_SRC_BLKDEV_DISK_CHECK_MEDIA_CHANGE + ZFS_AC_KERNEL_SRC_BLKDEV_BLK_STS_RESV_CONFLICT + ZFS_AC_KERNEL_SRC_BLKDEV_BLK_MODE_T ]) AC_DEFUN([ZFS_AC_KERNEL_BLKDEV], [ @@ -476,4 +657,7 @@ AC_DEFUN([ZFS_AC_KERNEL_BLKDEV], [ ZFS_AC_KERNEL_BLKDEV_ISSUE_SECURE_ERASE ZFS_AC_KERNEL_BLKDEV_BDEV_KOBJ ZFS_AC_KERNEL_BLKDEV_PART_TO_DEV + ZFS_AC_KERNEL_BLKDEV_DISK_CHECK_MEDIA_CHANGE + ZFS_AC_KERNEL_BLKDEV_BLK_STS_RESV_CONFLICT + ZFS_AC_KERNEL_BLKDEV_BLK_MODE_T ]) diff --git a/config/kernel-block-device-operations.m4 b/config/kernel-block-device-operations.m4 index 84e39dc8a2f6..4ff20b9c413d 100644 --- a/config/kernel-block-device-operations.m4 +++ b/config/kernel-block-device-operations.m4 @@ -5,7 +5,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_BLOCK_DEVICE_OPERATIONS_CHECK_EVENTS], [ ZFS_LINUX_TEST_SRC([block_device_operations_check_events], [ #include <linux/blkdev.h> - unsigned int blk_check_events(struct gendisk *disk, + static unsigned int blk_check_events(struct gendisk *disk, unsigned int clearing) { (void) disk, (void) clearing; return (0); @@ -34,7 +34,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_BLOCK_DEVICE_OPERATIONS_RELEASE_VOID], [ ZFS_LINUX_TEST_SRC([block_device_operations_release_void], [ #include <linux/blkdev.h> - void blk_release(struct gendisk *g, fmode_t mode) { + static void blk_release(struct gendisk *g, fmode_t mode) { (void) g, (void) mode; return; } @@ -49,12 +49,42 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_BLOCK_DEVICE_OPERATIONS_RELEASE_VOID], [ ], [], []) ]) +dnl # +dnl # 5.9.x API change +dnl # +AC_DEFUN([ZFS_AC_KERNEL_SRC_BLOCK_DEVICE_OPERATIONS_RELEASE_1ARG], [ + ZFS_LINUX_TEST_SRC([block_device_operations_release_void_1arg], [ + #include <linux/blkdev.h> + + static void blk_release(struct gendisk *g) { + (void) g; + return; + } + + static const struct block_device_operations + bops __attribute__ ((unused)) = { + .open = NULL, + .release = blk_release, + .ioctl = NULL, + .compat_ioctl = NULL, + }; + ], [], []) +]) + AC_DEFUN([ZFS_AC_KERNEL_BLOCK_DEVICE_OPERATIONS_RELEASE_VOID], [ - AC_MSG_CHECKING([whether bops->release() is void]) + AC_MSG_CHECKING([whether bops->release() is void and takes 2 args]) ZFS_LINUX_TEST_RESULT([block_device_operations_release_void], [ AC_MSG_RESULT(yes) ],[ - ZFS_LINUX_TEST_ERROR([bops->release()]) + AC_MSG_RESULT(no) + AC_MSG_CHECKING([whether bops->release() is void and takes 1 arg]) + ZFS_LINUX_TEST_RESULT([block_device_operations_release_void_1arg], [ + AC_MSG_RESULT(yes) + AC_DEFINE([HAVE_BLOCK_DEVICE_OPERATIONS_RELEASE_1ARG], [1], + [Define if release() in block_device_operations takes 1 arg]) + ],[ + ZFS_LINUX_TEST_ERROR([bops->release()]) + ]) ]) ]) @@ -66,7 +96,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_BLOCK_DEVICE_OPERATIONS_REVALIDATE_DISK], [ ZFS_LINUX_TEST_SRC([block_device_operations_revalidate_disk], [ #include <linux/blkdev.h> - int blk_revalidate_disk(struct gendisk *disk) { + static int blk_revalidate_disk(struct gendisk *disk) { (void) disk; return(0); } @@ -92,6 +122,7 @@ AC_DEFUN([ZFS_AC_KERNEL_BLOCK_DEVICE_OPERATIONS_REVALIDATE_DISK], [ AC_DEFUN([ZFS_AC_KERNEL_SRC_BLOCK_DEVICE_OPERATIONS], [ ZFS_AC_KERNEL_SRC_BLOCK_DEVICE_OPERATIONS_CHECK_EVENTS ZFS_AC_KERNEL_SRC_BLOCK_DEVICE_OPERATIONS_RELEASE_VOID + ZFS_AC_KERNEL_SRC_BLOCK_DEVICE_OPERATIONS_RELEASE_1ARG ZFS_AC_KERNEL_SRC_BLOCK_DEVICE_OPERATIONS_REVALIDATE_DISK ]) diff --git a/config/kernel-commit-metadata.m4 b/config/kernel-commit-metadata.m4 index 7df9b980290e..49bffbf609d2 100644 --- a/config/kernel-commit-metadata.m4 +++ b/config/kernel-commit-metadata.m4 @@ -7,7 +7,7 @@ dnl # AC_DEFUN([ZFS_AC_KERNEL_SRC_COMMIT_METADATA], [ ZFS_LINUX_TEST_SRC([export_operations_commit_metadata], [ #include <linux/exportfs.h> - int commit_metadata(struct inode *inode) { return 0; } + static int commit_metadata(struct inode *inode) { return 0; } static struct export_operations eops __attribute__ ((unused))={ .commit_metadata = commit_metadata, }; diff --git a/config/kernel-cpu_has_feature.m4 b/config/kernel-cpu_has_feature.m4 new file mode 100644 index 000000000000..608faf0f89fe --- /dev/null +++ b/config/kernel-cpu_has_feature.m4 @@ -0,0 +1,29 @@ +dnl # +dnl # cpu_has_feature() may referencing GPL-only cpu_feature_keys on powerpc +dnl # + +dnl # +dnl # Checking if cpu_has_feature is exported GPL-only +dnl # +AC_DEFUN([ZFS_AC_KERNEL_SRC_CPU_HAS_FEATURE], [ + ZFS_LINUX_TEST_SRC([cpu_has_feature], [ + #include <linux/version.h> + #if LINUX_VERSION_CODE >= KERNEL_VERSION(4, 7, 0) + #include <asm/cpu_has_feature.h> + #else + #include <asm/cputable.h> + #endif + ], [ + return cpu_has_feature(CPU_FTR_ALTIVEC) ? 0 : 1; + ], [], [ZFS_META_LICENSE]) +]) +AC_DEFUN([ZFS_AC_KERNEL_CPU_HAS_FEATURE], [ + AC_MSG_CHECKING([whether cpu_has_feature() is GPL-only]) + ZFS_LINUX_TEST_RESULT([cpu_has_feature_license], [ + AC_MSG_RESULT(no) + ], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_CPU_HAS_FEATURE_GPL_ONLY, 1, + [cpu_has_feature() is GPL-only]) + ]) +]) diff --git a/config/kernel-current-time.m4 b/config/kernel-current-time.m4 index 3ceb5f63efa9..ab7d9c5cedba 100644 --- a/config/kernel-current-time.m4 +++ b/config/kernel-current-time.m4 @@ -2,12 +2,15 @@ dnl # dnl # 4.9, current_time() added dnl # 4.18, return type changed from timespec to timespec64 dnl # +dnl # Note that we don't care about the return type in this check. If we have +dnl # to implement a fallback, we'll know we're <4.9, which was timespec. +dnl # AC_DEFUN([ZFS_AC_KERNEL_SRC_CURRENT_TIME], [ ZFS_LINUX_TEST_SRC([current_time], [ #include <linux/fs.h> ], [ struct inode ip __attribute__ ((unused)); - ip.i_atime = current_time(&ip); + (void) current_time(&ip); ]) ]) diff --git a/config/kernel-dentry-operations.m4 b/config/kernel-dentry-operations.m4 index dd470d7607b4..500f61e26aee 100644 --- a/config/kernel-dentry-operations.m4 +++ b/config/kernel-dentry-operations.m4 @@ -98,7 +98,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_D_REVALIDATE_NAMEIDATA], [ #include <linux/dcache.h> #include <linux/sched.h> - int revalidate (struct dentry *dentry, + static int revalidate (struct dentry *dentry, struct nameidata *nidata) { return 0; } static const struct dentry_operations diff --git a/config/kernel-dirty-inode.m4 b/config/kernel-dirty-inode.m4 index dc7667fa4881..2ef8658748ca 100644 --- a/config/kernel-dirty-inode.m4 +++ b/config/kernel-dirty-inode.m4 @@ -8,7 +8,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_DIRTY_INODE], [ ZFS_LINUX_TEST_SRC([dirty_inode_with_flags], [ #include <linux/fs.h> - void dirty_inode(struct inode *a, int b) { return; } + static void dirty_inode(struct inode *a, int b) { return; } static const struct super_operations sops __attribute__ ((unused)) = { diff --git a/config/kernel-encode-fh-inode.m4 b/config/kernel-encode-fh-inode.m4 index 9d4ba5f0f61f..b3ec040b5e95 100644 --- a/config/kernel-encode-fh-inode.m4 +++ b/config/kernel-encode-fh-inode.m4 @@ -7,7 +7,7 @@ dnl # AC_DEFUN([ZFS_AC_KERNEL_SRC_ENCODE_FH_WITH_INODE], [ ZFS_LINUX_TEST_SRC([export_operations_encode_fh], [ #include <linux/exportfs.h> - int encode_fh(struct inode *inode, __u32 *fh, int *max_len, + static int encode_fh(struct inode *inode, __u32 *fh, int *max_len, struct inode *parent) { return 0; } static struct export_operations eops __attribute__ ((unused))={ .encode_fh = encode_fh, diff --git a/config/kernel-evict-inode.m4 b/config/kernel-evict-inode.m4 index 66f10492de54..87082c9a2839 100644 --- a/config/kernel-evict-inode.m4 +++ b/config/kernel-evict-inode.m4 @@ -6,7 +6,7 @@ dnl # AC_DEFUN([ZFS_AC_KERNEL_SRC_EVICT_INODE], [ ZFS_LINUX_TEST_SRC([evict_inode], [ #include <linux/fs.h> - void evict_inode (struct inode * t) { return; } + static void evict_inode (struct inode * t) { return; } static struct super_operations sops __attribute__ ((unused)) = { .evict_inode = evict_inode, }; diff --git a/config/kernel-fallocate.m4 b/config/kernel-fallocate.m4 index 815602d3e2c6..95186dada453 100644 --- a/config/kernel-fallocate.m4 +++ b/config/kernel-fallocate.m4 @@ -11,7 +11,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_FALLOCATE], [ ZFS_LINUX_TEST_SRC([file_fallocate], [ #include <linux/fs.h> - long test_fallocate(struct file *file, int mode, + static long test_fallocate(struct file *file, int mode, loff_t offset, loff_t len) { return 0; } static const struct file_operations diff --git a/config/kernel-filemap-splice-read.m4 b/config/kernel-filemap-splice-read.m4 new file mode 100644 index 000000000000..4c83b31d738a --- /dev/null +++ b/config/kernel-filemap-splice-read.m4 @@ -0,0 +1,25 @@ +AC_DEFUN([ZFS_AC_KERNEL_SRC_COPY_SPLICE_READ], [ + dnl # + dnl # Kernel 6.5 - generic_file_splice_read was removed in favor + dnl # of copy_splice_read for the .splice_read member of the + dnl # file_operations struct. + dnl # + ZFS_LINUX_TEST_SRC([has_copy_splice_read], [ + #include <linux/fs.h> + + struct file_operations fops __attribute__((unused)) = { + .splice_read = copy_splice_read, + }; + ],[]) +]) + +AC_DEFUN([ZFS_AC_KERNEL_COPY_SPLICE_READ], [ + AC_MSG_CHECKING([whether copy_splice_read() exists]) + ZFS_LINUX_TEST_RESULT([has_copy_splice_read], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_COPY_SPLICE_READ, 1, + [copy_splice_read exists]) + ],[ + AC_MSG_RESULT(no) + ]) +]) diff --git a/config/kernel-filemap.m4 b/config/kernel-filemap.m4 new file mode 100644 index 000000000000..745928168f92 --- /dev/null +++ b/config/kernel-filemap.m4 @@ -0,0 +1,26 @@ +dnl # +dnl # filemap_range_has_page was not available till 4.13 +dnl # +AC_DEFUN([ZFS_AC_KERNEL_SRC_FILEMAP], [ + ZFS_LINUX_TEST_SRC([filemap_range_has_page], [ + #include <linux/fs.h> + ],[ + struct address_space *mapping = NULL; + loff_t lstart = 0; + loff_t lend = 0; + bool ret __attribute__ ((unused)); + + ret = filemap_range_has_page(mapping, lstart, lend); + ]) +]) + +AC_DEFUN([ZFS_AC_KERNEL_FILEMAP], [ + AC_MSG_CHECKING([whether filemap_range_has_page() is available]) + ZFS_LINUX_TEST_RESULT([filemap_range_has_page], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_FILEMAP_RANGE_HAS_PAGE, 1, + [filemap_range_has_page() is available]) + ],[ + AC_MSG_RESULT(no) + ]) +]) diff --git a/config/kernel-flush_dcache_page.m4 b/config/kernel-flush_dcache_page.m4 new file mode 100644 index 000000000000..2340c386ef57 --- /dev/null +++ b/config/kernel-flush_dcache_page.m4 @@ -0,0 +1,26 @@ +dnl # +dnl # Starting from Linux 5.13, flush_dcache_page() becomes an inline +dnl # function and may indirectly referencing GPL-only cpu_feature_keys on +dnl # powerpc +dnl # + +dnl # +dnl # Checking if flush_dcache_page is exported GPL-only +dnl # +AC_DEFUN([ZFS_AC_KERNEL_SRC_FLUSH_DCACHE_PAGE], [ + ZFS_LINUX_TEST_SRC([flush_dcache_page], [ + #include <asm/cacheflush.h> + ], [ + flush_dcache_page(0); + ], [], [ZFS_META_LICENSE]) +]) +AC_DEFUN([ZFS_AC_KERNEL_FLUSH_DCACHE_PAGE], [ + AC_MSG_CHECKING([whether flush_dcache_page() is GPL-only]) + ZFS_LINUX_TEST_RESULT([flush_dcache_page_license], [ + AC_MSG_RESULT(no) + ], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_FLUSH_DCACHE_PAGE_GPL_ONLY, 1, + [flush_dcache_page() is GPL-only]) + ]) +]) diff --git a/config/kernel-fpu.m4 b/config/kernel-fpu.m4 index e79632bf9b0c..9961bf206667 100644 --- a/config/kernel-fpu.m4 +++ b/config/kernel-fpu.m4 @@ -91,6 +91,13 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_FPU], [ __kernel_fpu_end(); ], [], [ZFS_META_LICENSE]) + ZFS_LINUX_TEST_SRC([kernel_neon], [ + #include <asm/neon.h> + ], [ + kernel_neon_begin(); + kernel_neon_end(); + ], [], [ZFS_META_LICENSE]) + ZFS_LINUX_TEST_SRC([fpu_internal], [ #if defined(__x86_64) || defined(__x86_64__) || \ defined(__i386) || defined(__i386__) @@ -186,18 +193,28 @@ AC_DEFUN([ZFS_AC_KERNEL_FPU], [ AC_DEFINE(KERNEL_EXPORTS_X86_FPU, 1, [kernel exports FPU functions]) ],[ - ZFS_LINUX_TEST_RESULT([fpu_internal], [ - AC_MSG_RESULT(internal) - AC_DEFINE(HAVE_KERNEL_FPU_INTERNAL, 1, - [kernel fpu internal]) + dnl # + dnl # ARM neon symbols (only on arm and arm64) + dnl # could be GPL-only on arm64 after Linux 6.2 + dnl # + ZFS_LINUX_TEST_RESULT([kernel_neon_license],[ + AC_MSG_RESULT(kernel_neon_*) + AC_DEFINE(HAVE_KERNEL_NEON, 1, + [kernel has kernel_neon_* functions]) ],[ - ZFS_LINUX_TEST_RESULT([fpu_xsave_internal], [ - AC_MSG_RESULT(internal with internal XSAVE) - AC_DEFINE(HAVE_KERNEL_FPU_XSAVE_INTERNAL, 1, - [kernel fpu and XSAVE internal]) - ],[ - AC_MSG_RESULT(unavailable) - ]) + ZFS_LINUX_TEST_RESULT([fpu_internal], [ + AC_MSG_RESULT(internal) + AC_DEFINE(HAVE_KERNEL_FPU_INTERNAL, 1, + [kernel fpu internal]) + ],[ + ZFS_LINUX_TEST_RESULT([fpu_xsave_internal], [ + AC_MSG_RESULT(internal with internal XSAVE) + AC_DEFINE(HAVE_KERNEL_FPU_XSAVE_INTERNAL, 1, + [kernel fpu and XSAVE internal]) + ],[ + AC_MSG_RESULT(unavailable) + ]) + ]) ]) ]) ]) diff --git a/config/kernel-fsync-bdev.m4 b/config/kernel-fsync-bdev.m4 new file mode 100644 index 000000000000..c47e236f705f --- /dev/null +++ b/config/kernel-fsync-bdev.m4 @@ -0,0 +1,36 @@ +dnl # +dnl # 6.6 API change, +dnl # fsync_bdev was removed in favor of sync_blockdev +dnl # +AC_DEFUN([ZFS_AC_KERNEL_SRC_SYNC_BDEV], [ + ZFS_LINUX_TEST_SRC([fsync_bdev], [ + #include <linux/blkdev.h> + ],[ + fsync_bdev(NULL); + ]) + + ZFS_LINUX_TEST_SRC([sync_blockdev], [ + #include <linux/blkdev.h> + ],[ + sync_blockdev(NULL); + ]) +]) + +AC_DEFUN([ZFS_AC_KERNEL_SYNC_BDEV], [ + AC_MSG_CHECKING([whether fsync_bdev() exists]) + ZFS_LINUX_TEST_RESULT([fsync_bdev], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_FSYNC_BDEV, 1, + [fsync_bdev() is declared in include/blkdev.h]) + ],[ + AC_MSG_CHECKING([whether sync_blockdev() exists]) + ZFS_LINUX_TEST_RESULT([sync_blockdev], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_SYNC_BLOCKDEV, 1, + [sync_blockdev() is declared in include/blkdev.h]) + ],[ + ZFS_LINUX_TEST_ERROR( + [neither fsync_bdev() nor sync_blockdev() exist]) + ]) + ]) +]) diff --git a/config/kernel-fsync.m4 b/config/kernel-fsync.m4 index d198191d3ab9..c155f8af81a8 100644 --- a/config/kernel-fsync.m4 +++ b/config/kernel-fsync.m4 @@ -5,7 +5,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_FSYNC], [ ZFS_LINUX_TEST_SRC([fsync_without_dentry], [ #include <linux/fs.h> - int test_fsync(struct file *f, int x) { return 0; } + static int test_fsync(struct file *f, int x) { return 0; } static const struct file_operations fops __attribute__ ((unused)) = { @@ -16,7 +16,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_FSYNC], [ ZFS_LINUX_TEST_SRC([fsync_range], [ #include <linux/fs.h> - int test_fsync(struct file *f, loff_t a, loff_t b, int c) + static int test_fsync(struct file *f, loff_t a, loff_t b, int c) { return 0; } static const struct file_operations diff --git a/config/kernel-generic_fillattr.m4 b/config/kernel-generic_fillattr.m4 index 0acd5d53103f..f5323f0dcb9f 100644 --- a/config/kernel-generic_fillattr.m4 +++ b/config/kernel-generic_fillattr.m4 @@ -4,7 +4,14 @@ dnl # dnl # generic_fillattr in linux/fs.h now requires a struct user_namespace* dnl # as the first arg, to support idmapped mounts. dnl # -AC_DEFUN([ZFS_AC_KERNEL_SRC_GENERIC_FILLATTR_USERNS], [ +dnl # 6.3 API +dnl # generic_fillattr() now takes struct mnt_idmap* as the first argument +dnl # +dnl # 6.6 API +dnl # generic_fillattr() now takes u32 as second argument, representing a +dnl # request_mask for statx +dnl # +AC_DEFUN([ZFS_AC_KERNEL_SRC_GENERIC_FILLATTR], [ ZFS_LINUX_TEST_SRC([generic_fillattr_userns], [ #include <linux/fs.h> ],[ @@ -13,16 +20,49 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_GENERIC_FILLATTR_USERNS], [ struct kstat *k = NULL; generic_fillattr(userns, in, k); ]) + + ZFS_LINUX_TEST_SRC([generic_fillattr_mnt_idmap], [ + #include <linux/fs.h> + ],[ + struct mnt_idmap *idmap = NULL; + struct inode *in = NULL; + struct kstat *k = NULL; + generic_fillattr(idmap, in, k); + ]) + + ZFS_LINUX_TEST_SRC([generic_fillattr_mnt_idmap_reqmask], [ + #include <linux/fs.h> + ],[ + struct mnt_idmap *idmap = NULL; + struct inode *in = NULL; + struct kstat *k = NULL; + generic_fillattr(idmap, 0, in, k); + ]) ]) -AC_DEFUN([ZFS_AC_KERNEL_GENERIC_FILLATTR_USERNS], [ - AC_MSG_CHECKING([whether generic_fillattr requires struct user_namespace*]) - ZFS_LINUX_TEST_RESULT([generic_fillattr_userns], [ +AC_DEFUN([ZFS_AC_KERNEL_GENERIC_FILLATTR], [ + AC_MSG_CHECKING( + [whether generic_fillattr requires struct mnt_idmap* and request_mask]) + ZFS_LINUX_TEST_RESULT([generic_fillattr_mnt_idmap_reqmask], [ AC_MSG_RESULT([yes]) - AC_DEFINE(HAVE_GENERIC_FILLATTR_USERNS, 1, - [generic_fillattr requires struct user_namespace*]) + AC_DEFINE(HAVE_GENERIC_FILLATTR_IDMAP_REQMASK, 1, + [generic_fillattr requires struct mnt_idmap* and u32 request_mask]) ],[ - AC_MSG_RESULT([no]) + AC_MSG_CHECKING([whether generic_fillattr requires struct mnt_idmap*]) + ZFS_LINUX_TEST_RESULT([generic_fillattr_mnt_idmap], [ + AC_MSG_RESULT([yes]) + AC_DEFINE(HAVE_GENERIC_FILLATTR_IDMAP, 1, + [generic_fillattr requires struct mnt_idmap*]) + ],[ + AC_MSG_CHECKING([whether generic_fillattr requires struct user_namespace*]) + ZFS_LINUX_TEST_RESULT([generic_fillattr_userns], [ + AC_MSG_RESULT([yes]) + AC_DEFINE(HAVE_GENERIC_FILLATTR_USERNS, 1, + [generic_fillattr requires struct user_namespace*]) + ],[ + AC_MSG_RESULT([no]) + ]) + ]) ]) ]) diff --git a/config/kernel-get-link.m4 b/config/kernel-get-link.m4 index e4f478e37c18..1f8f5b0c8b72 100644 --- a/config/kernel-get-link.m4 +++ b/config/kernel-get-link.m4 @@ -5,7 +5,7 @@ dnl # AC_DEFUN([ZFS_AC_KERNEL_SRC_GET_LINK], [ ZFS_LINUX_TEST_SRC([inode_operations_get_link], [ #include <linux/fs.h> - const char *get_link(struct dentry *de, struct inode *ip, + static const char *get_link(struct dentry *de, struct inode *ip, struct delayed_call *done) { return "symlink"; } static struct inode_operations iops __attribute__ ((unused)) = { @@ -15,7 +15,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_GET_LINK], [ ZFS_LINUX_TEST_SRC([inode_operations_get_link_cookie], [ #include <linux/fs.h> - const char *get_link(struct dentry *de, struct + static const char *get_link(struct dentry *de, struct inode *ip, void **cookie) { return "symlink"; } static struct inode_operations iops __attribute__ ((unused)) = { @@ -25,7 +25,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_GET_LINK], [ ZFS_LINUX_TEST_SRC([inode_operations_follow_link], [ #include <linux/fs.h> - const char *follow_link(struct dentry *de, + static const char *follow_link(struct dentry *de, void **cookie) { return "symlink"; } static struct inode_operations iops __attribute__ ((unused)) = { @@ -35,7 +35,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_GET_LINK], [ ZFS_LINUX_TEST_SRC([inode_operations_follow_link_nameidata], [ #include <linux/fs.h> - void *follow_link(struct dentry *de, struct + static void *follow_link(struct dentry *de, struct nameidata *nd) { return (void *)NULL; } static struct inode_operations iops __attribute__ ((unused)) = { diff --git a/config/kernel-inode-create.m4 b/config/kernel-inode-create.m4 index a6ea11fb61b2..95f8aa2d5220 100644 --- a/config/kernel-inode-create.m4 +++ b/config/kernel-inode-create.m4 @@ -1,5 +1,23 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_CREATE], [ dnl # + dnl # 6.3 API change + dnl # The first arg is changed to struct mnt_idmap * + dnl # + ZFS_LINUX_TEST_SRC([create_mnt_idmap], [ + #include <linux/fs.h> + #include <linux/sched.h> + + static int inode_create(struct mnt_idmap *idmap, + struct inode *inode ,struct dentry *dentry, + umode_t umode, bool flag) { return 0; } + + static const struct inode_operations + iops __attribute__ ((unused)) = { + .create = inode_create, + }; + ],[]) + + dnl # dnl # 5.12 API change that added the struct user_namespace* arg dnl # to the front of this function type's arg list. dnl # @@ -7,7 +25,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_CREATE], [ #include <linux/fs.h> #include <linux/sched.h> - int inode_create(struct user_namespace *userns, + static int inode_create(struct user_namespace *userns, struct inode *inode ,struct dentry *dentry, umode_t umode, bool flag) { return 0; } @@ -24,7 +42,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_CREATE], [ #include <linux/fs.h> #include <linux/sched.h> - int inode_create(struct inode *inode ,struct dentry *dentry, + static int inode_create(struct inode *inode ,struct dentry *dentry, umode_t umode, bool flag) { return 0; } static const struct inode_operations @@ -35,19 +53,28 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_CREATE], [ ]) AC_DEFUN([ZFS_AC_KERNEL_CREATE], [ - AC_MSG_CHECKING([whether iops->create() takes struct user_namespace*]) - ZFS_LINUX_TEST_RESULT([create_userns], [ + AC_MSG_CHECKING([whether iops->create() takes struct mnt_idmap*]) + ZFS_LINUX_TEST_RESULT([create_mnt_idmap], [ AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_IOPS_CREATE_USERNS, 1, - [iops->create() takes struct user_namespace*]) + AC_DEFINE(HAVE_IOPS_CREATE_IDMAP, 1, + [iops->create() takes struct mnt_idmap*]) ],[ AC_MSG_RESULT(no) - AC_MSG_CHECKING([whether iops->create() passes flags]) - ZFS_LINUX_TEST_RESULT([create_flags], [ + AC_MSG_CHECKING([whether iops->create() takes struct user_namespace*]) + ZFS_LINUX_TEST_RESULT([create_userns], [ AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_IOPS_CREATE_USERNS, 1, + [iops->create() takes struct user_namespace*]) ],[ - ZFS_LINUX_TEST_ERROR([iops->create()]) + AC_MSG_RESULT(no) + + AC_MSG_CHECKING([whether iops->create() passes flags]) + ZFS_LINUX_TEST_RESULT([create_flags], [ + AC_MSG_RESULT(yes) + ],[ + ZFS_LINUX_TEST_ERROR([iops->create()]) + ]) ]) ]) ]) diff --git a/config/kernel-inode-getattr.m4 b/config/kernel-inode-getattr.m4 index f62e82f5230a..5f7ce1ad9a5d 100644 --- a/config/kernel-inode-getattr.m4 +++ b/config/kernel-inode-getattr.m4 @@ -1,5 +1,25 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_INODE_GETATTR], [ dnl # + dnl # Linux 6.3 API + dnl # The first arg of getattr I/O operations handler type + dnl # is changed to struct mnt_idmap* + dnl # + ZFS_LINUX_TEST_SRC([inode_operations_getattr_mnt_idmap], [ + #include <linux/fs.h> + + static int test_getattr( + struct mnt_idmap *idmap, + const struct path *p, struct kstat *k, + u32 request_mask, unsigned int query_flags) + { return 0; } + + static const struct inode_operations + iops __attribute__ ((unused)) = { + .getattr = test_getattr, + }; + ],[]) + + dnl # dnl # Linux 5.12 API dnl # The getattr I/O operations handler type was extended to require dnl # a struct user_namespace* as its first arg, to support idmapped @@ -8,7 +28,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_INODE_GETATTR], [ ZFS_LINUX_TEST_SRC([inode_operations_getattr_userns], [ #include <linux/fs.h> - int test_getattr( + static int test_getattr( struct user_namespace *userns, const struct path *p, struct kstat *k, u32 request_mask, unsigned int query_flags) @@ -27,7 +47,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_INODE_GETATTR], [ ZFS_LINUX_TEST_SRC([inode_operations_getattr_path], [ #include <linux/fs.h> - int test_getattr( + static int test_getattr( const struct path *p, struct kstat *k, u32 request_mask, unsigned int query_flags) { return 0; } @@ -41,7 +61,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_INODE_GETATTR], [ ZFS_LINUX_TEST_SRC([inode_operations_getattr_vfsmount], [ #include <linux/fs.h> - int test_getattr( + static int test_getattr( struct vfsmount *mnt, struct dentry *d, struct kstat *k) { return 0; } @@ -55,37 +75,48 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_INODE_GETATTR], [ AC_DEFUN([ZFS_AC_KERNEL_INODE_GETATTR], [ dnl # - dnl # Kernel 5.12 test + dnl # Kernel 6.3 test dnl # - AC_MSG_CHECKING([whether iops->getattr() takes user_namespace]) - ZFS_LINUX_TEST_RESULT([inode_operations_getattr_userns], [ + AC_MSG_CHECKING([whether iops->getattr() takes mnt_idmap]) + ZFS_LINUX_TEST_RESULT([inode_operations_getattr_mnt_idmap], [ AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_USERNS_IOPS_GETATTR, 1, - [iops->getattr() takes struct user_namespace*]) + AC_DEFINE(HAVE_IDMAP_IOPS_GETATTR, 1, + [iops->getattr() takes struct mnt_idmap*]) ],[ AC_MSG_RESULT(no) - dnl # - dnl # Kernel 4.11 test + dnl # Kernel 5.12 test dnl # - AC_MSG_CHECKING([whether iops->getattr() takes a path]) - ZFS_LINUX_TEST_RESULT([inode_operations_getattr_path], [ + AC_MSG_CHECKING([whether iops->getattr() takes user_namespace]) + ZFS_LINUX_TEST_RESULT([inode_operations_getattr_userns], [ AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_PATH_IOPS_GETATTR, 1, - [iops->getattr() takes a path]) + AC_DEFINE(HAVE_USERNS_IOPS_GETATTR, 1, + [iops->getattr() takes struct user_namespace*]) ],[ AC_MSG_RESULT(no) dnl # - dnl # Kernel < 4.11 test + dnl # Kernel 4.11 test dnl # - AC_MSG_CHECKING([whether iops->getattr() takes a vfsmount]) - ZFS_LINUX_TEST_RESULT([inode_operations_getattr_vfsmount], [ + AC_MSG_CHECKING([whether iops->getattr() takes a path]) + ZFS_LINUX_TEST_RESULT([inode_operations_getattr_path], [ AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_VFSMOUNT_IOPS_GETATTR, 1, - [iops->getattr() takes a vfsmount]) + AC_DEFINE(HAVE_PATH_IOPS_GETATTR, 1, + [iops->getattr() takes a path]) ],[ AC_MSG_RESULT(no) + + dnl # + dnl # Kernel < 4.11 test + dnl # + AC_MSG_CHECKING([whether iops->getattr() takes a vfsmount]) + ZFS_LINUX_TEST_RESULT([inode_operations_getattr_vfsmount], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_VFSMOUNT_IOPS_GETATTR, 1, + [iops->getattr() takes a vfsmount]) + ],[ + AC_MSG_RESULT(no) + ]) ]) ]) ]) diff --git a/config/kernel-inode-lookup.m4 b/config/kernel-inode-lookup.m4 index 1a56e69b04aa..c7373056422c 100644 --- a/config/kernel-inode-lookup.m4 +++ b/config/kernel-inode-lookup.m4 @@ -6,7 +6,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_LOOKUP_FLAGS], [ #include <linux/fs.h> #include <linux/sched.h> - struct dentry *inode_lookup(struct inode *inode, + static struct dentry *inode_lookup(struct inode *inode, struct dentry *dentry, unsigned int flags) { return NULL; } static const struct inode_operations iops diff --git a/config/kernel-inode-setattr.m4 b/config/kernel-inode-setattr.m4 new file mode 100644 index 000000000000..69289e897be6 --- /dev/null +++ b/config/kernel-inode-setattr.m4 @@ -0,0 +1,87 @@ +AC_DEFUN([ZFS_AC_KERNEL_SRC_INODE_SETATTR], [ + dnl # + dnl # Linux 6.3 API + dnl # The first arg of setattr I/O operations handler type + dnl # is changed to struct mnt_idmap* + dnl # + ZFS_LINUX_TEST_SRC([inode_operations_setattr_mnt_idmap], [ + #include <linux/fs.h> + + static int test_setattr( + struct mnt_idmap *idmap, + struct dentry *de, struct iattr *ia) + { return 0; } + + static const struct inode_operations + iops __attribute__ ((unused)) = { + .setattr = test_setattr, + }; + ],[]) + + dnl # + dnl # Linux 5.12 API + dnl # The setattr I/O operations handler type was extended to require + dnl # a struct user_namespace* as its first arg, to support idmapped + dnl # mounts. + dnl # + ZFS_LINUX_TEST_SRC([inode_operations_setattr_userns], [ + #include <linux/fs.h> + + static int test_setattr( + struct user_namespace *userns, + struct dentry *de, struct iattr *ia) + { return 0; } + + static const struct inode_operations + iops __attribute__ ((unused)) = { + .setattr = test_setattr, + }; + ],[]) + + ZFS_LINUX_TEST_SRC([inode_operations_setattr], [ + #include <linux/fs.h> + + static int test_setattr( + struct dentry *de, struct iattr *ia) + { return 0; } + + static const struct inode_operations + iops __attribute__ ((unused)) = { + .setattr = test_setattr, + }; + ],[]) +]) + +AC_DEFUN([ZFS_AC_KERNEL_INODE_SETATTR], [ + dnl # + dnl # Kernel 6.3 test + dnl # + AC_MSG_CHECKING([whether iops->setattr() takes mnt_idmap]) + ZFS_LINUX_TEST_RESULT([inode_operations_setattr_mnt_idmap], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_IDMAP_IOPS_SETATTR, 1, + [iops->setattr() takes struct mnt_idmap*]) + ],[ + AC_MSG_RESULT(no) + dnl # + dnl # Kernel 5.12 test + dnl # + AC_MSG_CHECKING([whether iops->setattr() takes user_namespace]) + ZFS_LINUX_TEST_RESULT([inode_operations_setattr_userns], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_USERNS_IOPS_SETATTR, 1, + [iops->setattr() takes struct user_namespace*]) + ],[ + AC_MSG_RESULT(no) + + AC_MSG_CHECKING([whether iops->setattr() exists]) + ZFS_LINUX_TEST_RESULT([inode_operations_setattr], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_IOPS_SETATTR, 1, + [iops->setattr() exists]) + ],[ + AC_MSG_RESULT(no) + ]) + ]) + ]) +]) diff --git a/config/kernel-inode-times.m4 b/config/kernel-inode-times.m4 index 9c016c790081..4d861596ed0b 100644 --- a/config/kernel-inode-times.m4 +++ b/config/kernel-inode-times.m4 @@ -27,6 +27,73 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_INODE_TIMES], [ memset(&ip, 0, sizeof(ip)); ts = ip.i_mtime; ]) + + dnl # + dnl # 6.6 API change + dnl # i_ctime no longer directly accessible, must use + dnl # inode_get_ctime(ip), inode_set_ctime*(ip) to + dnl # read/write. + dnl # + ZFS_LINUX_TEST_SRC([inode_get_ctime], [ + #include <linux/fs.h> + ],[ + struct inode ip; + + memset(&ip, 0, sizeof(ip)); + inode_get_ctime(&ip); + ]) + + ZFS_LINUX_TEST_SRC([inode_set_ctime_to_ts], [ + #include <linux/fs.h> + ],[ + struct inode ip; + struct timespec64 ts = {0}; + + memset(&ip, 0, sizeof(ip)); + inode_set_ctime_to_ts(&ip, ts); + ]) + + dnl # + dnl # 6.7 API change + dnl # i_atime/i_mtime no longer directly accessible, must use + dnl # inode_get_mtime(ip), inode_set_mtime*(ip) to + dnl # read/write. + dnl # + ZFS_LINUX_TEST_SRC([inode_get_atime], [ + #include <linux/fs.h> + ],[ + struct inode ip; + + memset(&ip, 0, sizeof(ip)); + inode_get_atime(&ip); + ]) + ZFS_LINUX_TEST_SRC([inode_get_mtime], [ + #include <linux/fs.h> + ],[ + struct inode ip; + + memset(&ip, 0, sizeof(ip)); + inode_get_mtime(&ip); + ]) + + ZFS_LINUX_TEST_SRC([inode_set_atime_to_ts], [ + #include <linux/fs.h> + ],[ + struct inode ip; + struct timespec64 ts = {0}; + + memset(&ip, 0, sizeof(ip)); + inode_set_atime_to_ts(&ip, ts); + ]) + ZFS_LINUX_TEST_SRC([inode_set_mtime_to_ts], [ + #include <linux/fs.h> + ],[ + struct inode ip; + struct timespec64 ts = {0}; + + memset(&ip, 0, sizeof(ip)); + inode_set_mtime_to_ts(&ip, ts); + ]) ]) AC_DEFUN([ZFS_AC_KERNEL_INODE_TIMES], [ @@ -47,4 +114,58 @@ AC_DEFUN([ZFS_AC_KERNEL_INODE_TIMES], [ AC_DEFINE(HAVE_INODE_TIMESPEC64_TIMES, 1, [inode->i_*time's are timespec64]) ]) + + AC_MSG_CHECKING([whether inode_get_ctime() exists]) + ZFS_LINUX_TEST_RESULT([inode_get_ctime], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_INODE_GET_CTIME, 1, + [inode_get_ctime() exists in linux/fs.h]) + ],[ + AC_MSG_RESULT(no) + ]) + + AC_MSG_CHECKING([whether inode_set_ctime_to_ts() exists]) + ZFS_LINUX_TEST_RESULT([inode_set_ctime_to_ts], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_INODE_SET_CTIME_TO_TS, 1, + [inode_set_ctime_to_ts() exists in linux/fs.h]) + ],[ + AC_MSG_RESULT(no) + ]) + + AC_MSG_CHECKING([whether inode_get_atime() exists]) + ZFS_LINUX_TEST_RESULT([inode_get_atime], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_INODE_GET_ATIME, 1, + [inode_get_atime() exists in linux/fs.h]) + ],[ + AC_MSG_RESULT(no) + ]) + + AC_MSG_CHECKING([whether inode_set_atime_to_ts() exists]) + ZFS_LINUX_TEST_RESULT([inode_set_atime_to_ts], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_INODE_SET_ATIME_TO_TS, 1, + [inode_set_atime_to_ts() exists in linux/fs.h]) + ],[ + AC_MSG_RESULT(no) + ]) + + AC_MSG_CHECKING([whether inode_get_mtime() exists]) + ZFS_LINUX_TEST_RESULT([inode_get_mtime], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_INODE_GET_MTIME, 1, + [inode_get_mtime() exists in linux/fs.h]) + ],[ + AC_MSG_RESULT(no) + ]) + + AC_MSG_CHECKING([whether inode_set_mtime_to_ts() exists]) + ZFS_LINUX_TEST_RESULT([inode_set_mtime_to_ts], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_INODE_SET_MTIME_TO_TS, 1, + [inode_set_mtime_to_ts() exists in linux/fs.h]) + ],[ + AC_MSG_RESULT(no) + ]) ]) diff --git a/config/kernel-is_owner_or_cap.m4 b/config/kernel-is_owner_or_cap.m4 index a90cf3da641d..4e9c002b77f2 100644 --- a/config/kernel-is_owner_or_cap.m4 +++ b/config/kernel-is_owner_or_cap.m4 @@ -16,12 +16,20 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_INODE_OWNER_OR_CAPABLE], [ (void) inode_owner_or_capable(ip); ]) - ZFS_LINUX_TEST_SRC([inode_owner_or_capable_idmapped], [ + ZFS_LINUX_TEST_SRC([inode_owner_or_capable_userns], [ #include <linux/fs.h> ],[ struct inode *ip = NULL; (void) inode_owner_or_capable(&init_user_ns, ip); ]) + + ZFS_LINUX_TEST_SRC([inode_owner_or_capable_mnt_idmap], [ + #include <linux/fs.h> + #include <linux/mnt_idmapping.h> + ],[ + struct inode *ip = NULL; + (void) inode_owner_or_capable(&nop_mnt_idmap, ip); + ]) ]) AC_DEFUN([ZFS_AC_KERNEL_INODE_OWNER_OR_CAPABLE], [ @@ -35,12 +43,21 @@ AC_DEFUN([ZFS_AC_KERNEL_INODE_OWNER_OR_CAPABLE], [ AC_MSG_CHECKING( [whether inode_owner_or_capable() takes user_ns]) - ZFS_LINUX_TEST_RESULT([inode_owner_or_capable_idmapped], [ + ZFS_LINUX_TEST_RESULT([inode_owner_or_capable_userns], [ AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_INODE_OWNER_OR_CAPABLE_IDMAPPED, 1, + AC_DEFINE(HAVE_INODE_OWNER_OR_CAPABLE_USERNS, 1, [inode_owner_or_capable() takes user_ns]) ],[ - ZFS_LINUX_TEST_ERROR([capability]) + AC_MSG_RESULT(no) + AC_MSG_CHECKING( + [whether inode_owner_or_capable() takes mnt_idmap]) + ZFS_LINUX_TEST_RESULT([inode_owner_or_capable_mnt_idmap], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_INODE_OWNER_OR_CAPABLE_IDMAP, 1, + [inode_owner_or_capable() takes mnt_idmap]) + ], [ + ZFS_LINUX_TEST_ERROR([capability]) + ]) ]) ]) ]) diff --git a/config/kernel-make-request-fn.m4 b/config/kernel-make-request-fn.m4 index f17416acca67..4d20dd45c4a1 100644 --- a/config/kernel-make-request-fn.m4 +++ b/config/kernel-make-request-fn.m4 @@ -4,7 +4,7 @@ dnl # AC_DEFUN([ZFS_AC_KERNEL_SRC_MAKE_REQUEST_FN], [ ZFS_LINUX_TEST_SRC([make_request_fn_void], [ #include <linux/blkdev.h> - void make_request(struct request_queue *q, + static void make_request(struct request_queue *q, struct bio *bio) { return; } ],[ blk_queue_make_request(NULL, &make_request); @@ -12,7 +12,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_MAKE_REQUEST_FN], [ ZFS_LINUX_TEST_SRC([make_request_fn_blk_qc_t], [ #include <linux/blkdev.h> - blk_qc_t make_request(struct request_queue *q, + static blk_qc_t make_request(struct request_queue *q, struct bio *bio) { return (BLK_QC_T_NONE); } ],[ blk_queue_make_request(NULL, &make_request); @@ -20,7 +20,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_MAKE_REQUEST_FN], [ ZFS_LINUX_TEST_SRC([blk_alloc_queue_request_fn], [ #include <linux/blkdev.h> - blk_qc_t make_request(struct request_queue *q, + static blk_qc_t make_request(struct request_queue *q, struct bio *bio) { return (BLK_QC_T_NONE); } ],[ struct request_queue *q __attribute__ ((unused)); @@ -29,7 +29,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_MAKE_REQUEST_FN], [ ZFS_LINUX_TEST_SRC([blk_alloc_queue_request_fn_rh], [ #include <linux/blkdev.h> - blk_qc_t make_request(struct request_queue *q, + static blk_qc_t make_request(struct request_queue *q, struct bio *bio) { return (BLK_QC_T_NONE); } ],[ struct request_queue *q __attribute__ ((unused)); diff --git a/config/kernel-mkdir.m4 b/config/kernel-mkdir.m4 index 6667ed04fa4c..367f100094d3 100644 --- a/config/kernel-mkdir.m4 +++ b/config/kernel-mkdir.m4 @@ -3,6 +3,22 @@ dnl # Supported mkdir() interfaces checked newest to oldest. dnl # AC_DEFUN([ZFS_AC_KERNEL_SRC_MKDIR], [ dnl # + dnl # 6.3 API change + dnl # mkdir() takes struct mnt_idmap * as the first arg + dnl # + ZFS_LINUX_TEST_SRC([mkdir_mnt_idmap], [ + #include <linux/fs.h> + + static int mkdir(struct mnt_idmap *idmap, + struct inode *inode, struct dentry *dentry, + umode_t umode) { return 0; } + static const struct inode_operations + iops __attribute__ ((unused)) = { + .mkdir = mkdir, + }; + ],[]) + + dnl # dnl # 5.12 API change dnl # The struct user_namespace arg was added as the first argument to dnl # mkdir() @@ -10,7 +26,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_MKDIR], [ ZFS_LINUX_TEST_SRC([mkdir_user_namespace], [ #include <linux/fs.h> - int mkdir(struct user_namespace *userns, + static int mkdir(struct user_namespace *userns, struct inode *inode, struct dentry *dentry, umode_t umode) { return 0; } @@ -31,7 +47,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_MKDIR], [ ZFS_LINUX_TEST_SRC([inode_operations_mkdir], [ #include <linux/fs.h> - int mkdir(struct inode *inode, struct dentry *dentry, + static int mkdir(struct inode *inode, struct dentry *dentry, umode_t umode) { return 0; } static const struct inode_operations @@ -43,25 +59,36 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_MKDIR], [ AC_DEFUN([ZFS_AC_KERNEL_MKDIR], [ dnl # - dnl # 5.12 API change - dnl # The struct user_namespace arg was added as the first argument to - dnl # mkdir() of the iops structure. + dnl # 6.3 API change + dnl # mkdir() takes struct mnt_idmap * as the first arg dnl # - AC_MSG_CHECKING([whether iops->mkdir() takes struct user_namespace*]) - ZFS_LINUX_TEST_RESULT([mkdir_user_namespace], [ + AC_MSG_CHECKING([whether iops->mkdir() takes struct mnt_idmap*]) + ZFS_LINUX_TEST_RESULT([mkdir_mnt_idmap], [ AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_IOPS_MKDIR_USERNS, 1, - [iops->mkdir() takes struct user_namespace*]) + AC_DEFINE(HAVE_IOPS_MKDIR_IDMAP, 1, + [iops->mkdir() takes struct mnt_idmap*]) ],[ - AC_MSG_RESULT(no) - - AC_MSG_CHECKING([whether iops->mkdir() takes umode_t]) - ZFS_LINUX_TEST_RESULT([inode_operations_mkdir], [ + dnl # + dnl # 5.12 API change + dnl # The struct user_namespace arg was added as the first argument to + dnl # mkdir() of the iops structure. + dnl # + AC_MSG_CHECKING([whether iops->mkdir() takes struct user_namespace*]) + ZFS_LINUX_TEST_RESULT([mkdir_user_namespace], [ AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_MKDIR_UMODE_T, 1, - [iops->mkdir() takes umode_t]) + AC_DEFINE(HAVE_IOPS_MKDIR_USERNS, 1, + [iops->mkdir() takes struct user_namespace*]) ],[ - ZFS_LINUX_TEST_ERROR([mkdir()]) + AC_MSG_RESULT(no) + + AC_MSG_CHECKING([whether iops->mkdir() takes umode_t]) + ZFS_LINUX_TEST_RESULT([inode_operations_mkdir], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_MKDIR_UMODE_T, 1, + [iops->mkdir() takes umode_t]) + ],[ + ZFS_LINUX_TEST_ERROR([mkdir()]) + ]) ]) ]) ]) diff --git a/config/kernel-mknod.m4 b/config/kernel-mknod.m4 index ffe45106003a..6ad3453aaf0a 100644 --- a/config/kernel-mknod.m4 +++ b/config/kernel-mknod.m4 @@ -1,5 +1,23 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_MKNOD], [ dnl # + dnl # 6.3 API change + dnl # The first arg is now struct mnt_idmap* + dnl # + ZFS_LINUX_TEST_SRC([mknod_mnt_idmap], [ + #include <linux/fs.h> + #include <linux/sched.h> + + static int tmp_mknod(struct mnt_idmap *idmap, + struct inode *inode ,struct dentry *dentry, + umode_t u, dev_t d) { return 0; } + + static const struct inode_operations + iops __attribute__ ((unused)) = { + .mknod = tmp_mknod, + }; + ],[]) + + dnl # dnl # 5.12 API change that added the struct user_namespace* arg dnl # to the front of this function type's arg list. dnl # @@ -7,7 +25,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_MKNOD], [ #include <linux/fs.h> #include <linux/sched.h> - int tmp_mknod(struct user_namespace *userns, + static int tmp_mknod(struct user_namespace *userns, struct inode *inode ,struct dentry *dentry, umode_t u, dev_t d) { return 0; } @@ -19,12 +37,20 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_MKNOD], [ ]) AC_DEFUN([ZFS_AC_KERNEL_MKNOD], [ - AC_MSG_CHECKING([whether iops->mknod() takes struct user_namespace*]) - ZFS_LINUX_TEST_RESULT([mknod_userns], [ + AC_MSG_CHECKING([whether iops->mknod() takes struct mnt_idmap*]) + ZFS_LINUX_TEST_RESULT([mknod_mnt_idmap], [ AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_IOPS_MKNOD_USERNS, 1, - [iops->mknod() takes struct user_namespace*]) + AC_DEFINE(HAVE_IOPS_MKNOD_IDMAP, 1, + [iops->mknod() takes struct mnt_idmap*]) ],[ AC_MSG_RESULT(no) + AC_MSG_CHECKING([whether iops->mknod() takes struct user_namespace*]) + ZFS_LINUX_TEST_RESULT([mknod_userns], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_IOPS_MKNOD_USERNS, 1, + [iops->mknod() takes struct user_namespace*]) + ],[ + AC_MSG_RESULT(no) + ]) ]) ]) diff --git a/config/kernel-proc-operations.m4 b/config/kernel-proc-operations.m4 index df216222ecc2..3ae8ce2b6d0d 100644 --- a/config/kernel-proc-operations.m4 +++ b/config/kernel-proc-operations.m4 @@ -7,14 +7,14 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_PROC_OPERATIONS], [ ZFS_LINUX_TEST_SRC([proc_ops_struct], [ #include <linux/proc_fs.h> - int test_open(struct inode *ip, struct file *fp) { return 0; } - ssize_t test_read(struct file *fp, char __user *ptr, + static int test_open(struct inode *ip, struct file *fp) { return 0; } + static ssize_t test_read(struct file *fp, char __user *ptr, size_t size, loff_t *offp) { return 0; } - ssize_t test_write(struct file *fp, const char __user *ptr, + static ssize_t test_write(struct file *fp, const char __user *ptr, size_t size, loff_t *offp) { return 0; } - loff_t test_lseek(struct file *fp, loff_t off, int flag) + static loff_t test_lseek(struct file *fp, loff_t off, int flag) { return 0; } - int test_release(struct inode *ip, struct file *fp) + static int test_release(struct inode *ip, struct file *fp) { return 0; } const struct proc_ops test_ops __attribute__ ((unused)) = { diff --git a/config/kernel-put-link.m4 b/config/kernel-put-link.m4 index 4234861f3347..8ab318cbff8c 100644 --- a/config/kernel-put-link.m4 +++ b/config/kernel-put-link.m4 @@ -4,7 +4,7 @@ dnl # AC_DEFUN([ZFS_AC_KERNEL_SRC_PUT_LINK], [ ZFS_LINUX_TEST_SRC([put_link_cookie], [ #include <linux/fs.h> - void put_link(struct inode *ip, void *cookie) + static void put_link(struct inode *ip, void *cookie) { return; } static struct inode_operations iops __attribute__ ((unused)) = { @@ -14,7 +14,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_PUT_LINK], [ ZFS_LINUX_TEST_SRC([put_link_nameidata], [ #include <linux/fs.h> - void put_link(struct dentry *de, struct + static void put_link(struct dentry *de, struct nameidata *nd, void *ptr) { return; } static struct inode_operations iops __attribute__ ((unused)) = { diff --git a/config/kernel-reclaim_state.m4 b/config/kernel-reclaim_state.m4 new file mode 100644 index 000000000000..9936b3c1001f --- /dev/null +++ b/config/kernel-reclaim_state.m4 @@ -0,0 +1,26 @@ +AC_DEFUN([ZFS_AC_KERNEL_SRC_RECLAIMED], [ + dnl # + dnl # 6.4 API change + dnl # The reclaimed_slab of struct reclaim_state + dnl # is renamed to reclaimed + dnl # + ZFS_LINUX_TEST_SRC([reclaim_state_reclaimed], [ + #include <linux/swap.h> + static const struct reclaim_state + rs __attribute__ ((unused)) = { + .reclaimed = 100, + }; + ],[]) +]) + +AC_DEFUN([ZFS_AC_KERNEL_RECLAIMED], [ + AC_MSG_CHECKING([whether struct reclaim_state has reclaimed field]) + ZFS_LINUX_TEST_RESULT([reclaim_state_reclaimed], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_RECLAIM_STATE_RECLAIMED, 1, + [struct reclaim_state has reclaimed]) + ],[ + AC_MSG_RESULT(no) + ]) +]) + diff --git a/config/kernel-register_sysctl_table.m4 b/config/kernel-register_sysctl_table.m4 new file mode 100644 index 000000000000..a5e934f56d29 --- /dev/null +++ b/config/kernel-register_sysctl_table.m4 @@ -0,0 +1,27 @@ +dnl # +dnl # Linux 6.5 removes register_sysctl_table +dnl # +AC_DEFUN([ZFS_AC_KERNEL_SRC_REGISTER_SYSCTL_TABLE], [ + ZFS_LINUX_TEST_SRC([has_register_sysctl_table], [ + #include <linux/sysctl.h> + + static struct ctl_table dummy_table[] = { + {} + }; + + ],[ + struct ctl_table_header *h + __attribute((unused)) = register_sysctl_table(dummy_table); + ]) +]) + +AC_DEFUN([ZFS_AC_KERNEL_REGISTER_SYSCTL_TABLE], [ + AC_MSG_CHECKING([whether register_sysctl_table exists]) + ZFS_LINUX_TEST_RESULT([has_register_sysctl_table], [ + AC_MSG_RESULT([yes]) + AC_DEFINE(HAVE_REGISTER_SYSCTL_TABLE, 1, + [register_sysctl_table exists]) + ],[ + AC_MSG_RESULT([no]) + ]) +]) diff --git a/config/kernel-rename.m4 b/config/kernel-rename.m4 index 302db43f5748..34f5b664e2b7 100644 --- a/config/kernel-rename.m4 +++ b/config/kernel-rename.m4 @@ -6,7 +6,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_RENAME], [ dnl # ZFS_LINUX_TEST_SRC([inode_operations_rename_flags], [ #include <linux/fs.h> - int rename_fn(struct inode *sip, struct dentry *sdp, + static int rename_fn(struct inode *sip, struct dentry *sdp, struct inode *tip, struct dentry *tdp, unsigned int flags) { return 0; } @@ -24,7 +24,22 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_RENAME], [ dnl # ZFS_LINUX_TEST_SRC([inode_operations_rename_userns], [ #include <linux/fs.h> - int rename_fn(struct user_namespace *user_ns, struct inode *sip, + static int rename_fn(struct user_namespace *user_ns, struct inode *sip, + struct dentry *sdp, struct inode *tip, struct dentry *tdp, + unsigned int flags) { return 0; } + + static const struct inode_operations + iops __attribute__ ((unused)) = { + .rename = rename_fn, + }; + ],[]) + + dnl # + dnl # 6.3 API change - the first arg is now struct mnt_idmap* + dnl # + ZFS_LINUX_TEST_SRC([inode_operations_rename_mnt_idmap], [ + #include <linux/fs.h> + static int rename_fn(struct mnt_idmap *idmap, struct inode *sip, struct dentry *sdp, struct inode *tip, struct dentry *tdp, unsigned int flags) { return 0; } @@ -36,21 +51,30 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_RENAME], [ ]) AC_DEFUN([ZFS_AC_KERNEL_RENAME], [ - AC_MSG_CHECKING([whether iops->rename() takes struct user_namespace*]) - ZFS_LINUX_TEST_RESULT([inode_operations_rename_userns], [ + AC_MSG_CHECKING([whether iops->rename() takes struct mnt_idmap*]) + ZFS_LINUX_TEST_RESULT([inode_operations_rename_mnt_idmap], [ AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_IOPS_RENAME_USERNS, 1, - [iops->rename() takes struct user_namespace*]) + AC_DEFINE(HAVE_IOPS_RENAME_IDMAP, 1, + [iops->rename() takes struct mnt_idmap*]) ],[ AC_MSG_RESULT(no) - AC_MSG_CHECKING([whether iop->rename() wants flags]) - ZFS_LINUX_TEST_RESULT([inode_operations_rename_flags], [ + AC_MSG_CHECKING([whether iops->rename() takes struct user_namespace*]) + ZFS_LINUX_TEST_RESULT([inode_operations_rename_userns], [ AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_RENAME_WANTS_FLAGS, 1, - [iops->rename() wants flags]) + AC_DEFINE(HAVE_IOPS_RENAME_USERNS, 1, + [iops->rename() takes struct user_namespace*]) ],[ AC_MSG_RESULT(no) + + AC_MSG_CHECKING([whether iops->rename() wants flags]) + ZFS_LINUX_TEST_RESULT([inode_operations_rename_flags], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_RENAME_WANTS_FLAGS, 1, + [iops->rename() wants flags]) + ],[ + AC_MSG_RESULT(no) + ]) ]) ]) ]) diff --git a/config/kernel-setattr-prepare.m4 b/config/kernel-setattr-prepare.m4 index 24245aa53448..e02d6263e9c9 100644 --- a/config/kernel-setattr-prepare.m4 +++ b/config/kernel-setattr-prepare.m4 @@ -27,26 +27,48 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_SETATTR_PREPARE], [ int error __attribute__ ((unused)) = setattr_prepare(userns, dentry, attr); ]) + + dnl # + dnl # 6.3 API change + dnl # The first arg of setattr_prepare() is changed to struct mnt_idmap* + dnl # + ZFS_LINUX_TEST_SRC([setattr_prepare_mnt_idmap], [ + #include <linux/fs.h> + ], [ + struct dentry *dentry = NULL; + struct iattr *attr = NULL; + struct mnt_idmap *idmap = NULL; + int error __attribute__ ((unused)) = + setattr_prepare(idmap, dentry, attr); + ]) ]) AC_DEFUN([ZFS_AC_KERNEL_SETATTR_PREPARE], [ - AC_MSG_CHECKING([whether setattr_prepare() is available and accepts struct user_namespace*]) - ZFS_LINUX_TEST_RESULT_SYMBOL([setattr_prepare_userns], + AC_MSG_CHECKING([whether setattr_prepare() is available and accepts struct mnt_idmap*]) + ZFS_LINUX_TEST_RESULT_SYMBOL([setattr_prepare_mnt_idmap], [setattr_prepare], [fs/attr.c], [ AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_SETATTR_PREPARE_USERNS, 1, - [setattr_prepare() accepts user_namespace]) + AC_DEFINE(HAVE_SETATTR_PREPARE_IDMAP, 1, + [setattr_prepare() accepts mnt_idmap]) ], [ - AC_MSG_RESULT(no) - - AC_MSG_CHECKING([whether setattr_prepare() is available, doesn't accept user_namespace]) - ZFS_LINUX_TEST_RESULT_SYMBOL([setattr_prepare], - [setattr_prepare], [fs/attr.c], [ + AC_MSG_CHECKING([whether setattr_prepare() is available and accepts struct user_namespace*]) + ZFS_LINUX_TEST_RESULT_SYMBOL([setattr_prepare_userns], + [setattr_prepare], [fs/attr.c], [ AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_SETATTR_PREPARE_NO_USERNS, 1, - [setattr_prepare() is available, doesn't accept user_namespace]) + AC_DEFINE(HAVE_SETATTR_PREPARE_USERNS, 1, + [setattr_prepare() accepts user_namespace]) ], [ AC_MSG_RESULT(no) + + AC_MSG_CHECKING([whether setattr_prepare() is available, doesn't accept user_namespace]) + ZFS_LINUX_TEST_RESULT_SYMBOL([setattr_prepare], + [setattr_prepare], [fs/attr.c], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_SETATTR_PREPARE_NO_USERNS, 1, + [setattr_prepare() is available, doesn't accept user_namespace]) + ], [ + AC_MSG_RESULT(no) + ]) ]) ]) ]) diff --git a/config/kernel-show-options.m4 b/config/kernel-show-options.m4 index 93bd5fbfbb24..fd62f30086dc 100644 --- a/config/kernel-show-options.m4 +++ b/config/kernel-show-options.m4 @@ -5,7 +5,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_SHOW_OPTIONS], [ ZFS_LINUX_TEST_SRC([super_operations_show_options], [ #include <linux/fs.h> - int show_options(struct seq_file * x, struct dentry * y) { + static int show_options(struct seq_file * x, struct dentry * y) { return 0; }; diff --git a/config/kernel-shrink.m4 b/config/kernel-shrink.m4 index 0c702153e8c4..6580b08d5ff2 100644 --- a/config/kernel-shrink.m4 +++ b/config/kernel-shrink.m4 @@ -8,9 +8,6 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_SUPER_BLOCK_S_SHRINK], [ ZFS_LINUX_TEST_SRC([super_block_s_shrink], [ #include <linux/fs.h> - int shrink(struct shrinker *s, struct shrink_control *sc) - { return 0; } - static const struct super_block sb __attribute__ ((unused)) = { .s_shrink.seeks = DEFAULT_SEEKS, @@ -19,12 +16,44 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_SUPER_BLOCK_S_SHRINK], [ ],[]) ]) +dnl # +dnl # 6.7 API change +dnl # s_shrink is now a pointer. +dnl # +AC_DEFUN([ZFS_AC_KERNEL_SRC_SUPER_BLOCK_S_SHRINK_PTR], [ + ZFS_LINUX_TEST_SRC([super_block_s_shrink_ptr], [ + #include <linux/fs.h> + static unsigned long shrinker_cb(struct shrinker *shrink, + struct shrink_control *sc) { return 0; } + static struct shrinker shrinker = { + .count_objects = shrinker_cb, + .scan_objects = shrinker_cb, + .seeks = DEFAULT_SEEKS, + }; + static const struct super_block + sb __attribute__ ((unused)) = { + .s_shrink = &shrinker, + }; + ],[]) +]) + AC_DEFUN([ZFS_AC_KERNEL_SUPER_BLOCK_S_SHRINK], [ AC_MSG_CHECKING([whether super_block has s_shrink]) ZFS_LINUX_TEST_RESULT([super_block_s_shrink], [ AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_SUPER_BLOCK_S_SHRINK, 1, + [have super_block s_shrink]) ],[ - ZFS_LINUX_TEST_ERROR([sb->s_shrink()]) + AC_MSG_RESULT(no) + AC_MSG_CHECKING([whether super_block has s_shrink pointer]) + ZFS_LINUX_TEST_RESULT([super_block_s_shrink_ptr], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_SUPER_BLOCK_S_SHRINK_PTR, 1, + [have super_block s_shrink pointer]) + ],[ + AC_MSG_RESULT(no) + ZFS_LINUX_TEST_ERROR([sb->s_shrink()]) + ]) ]) ]) @@ -57,7 +86,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SHRINK_CONTROL_HAS_NID], [ AC_DEFUN([ZFS_AC_KERNEL_SRC_REGISTER_SHRINKER_VARARG], [ ZFS_LINUX_TEST_SRC([register_shrinker_vararg], [ #include <linux/mm.h> - unsigned long shrinker_cb(struct shrinker *shrink, + static unsigned long shrinker_cb(struct shrinker *shrink, struct shrink_control *sc) { return 0; } ],[ struct shrinker cache_shrinker = { @@ -72,7 +101,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_REGISTER_SHRINKER_VARARG], [ AC_DEFUN([ZFS_AC_KERNEL_SRC_SHRINKER_CALLBACK], [ ZFS_LINUX_TEST_SRC([shrinker_cb_shrink_control], [ #include <linux/mm.h> - int shrinker_cb(struct shrinker *shrink, + static int shrinker_cb(struct shrinker *shrink, struct shrink_control *sc) { return 0; } ],[ struct shrinker cache_shrinker = { @@ -84,7 +113,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_SHRINKER_CALLBACK], [ ZFS_LINUX_TEST_SRC([shrinker_cb_shrink_control_split], [ #include <linux/mm.h> - unsigned long shrinker_cb(struct shrinker *shrink, + static unsigned long shrinker_cb(struct shrinker *shrink, struct shrink_control *sc) { return 0; } ],[ struct shrinker cache_shrinker = { @@ -96,6 +125,25 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_SHRINKER_CALLBACK], [ ]) ]) +dnl # +dnl # 6.7 API change +dnl # register_shrinker has been replaced by shrinker_register. +dnl # +AC_DEFUN([ZFS_AC_KERNEL_SRC_SHRINKER_REGISTER], [ + ZFS_LINUX_TEST_SRC([shrinker_register], [ + #include <linux/shrinker.h> + static unsigned long shrinker_cb(struct shrinker *shrink, + struct shrink_control *sc) { return 0; } + ],[ + struct shrinker cache_shrinker = { + .count_objects = shrinker_cb, + .scan_objects = shrinker_cb, + .seeks = DEFAULT_SEEKS, + }; + shrinker_register(&cache_shrinker); + ]) +]) + AC_DEFUN([ZFS_AC_KERNEL_SHRINKER_CALLBACK],[ dnl # dnl # 6.0 API change @@ -133,14 +181,36 @@ AC_DEFUN([ZFS_AC_KERNEL_SHRINKER_CALLBACK],[ dnl # cs->shrink() is logically split in to dnl # cs->count_objects() and cs->scan_objects() dnl # - AC_MSG_CHECKING([if cs->count_objects callback exists]) + AC_MSG_CHECKING( + [whether cs->count_objects callback exists]) ZFS_LINUX_TEST_RESULT( - [shrinker_cb_shrink_control_split],[ - AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_SPLIT_SHRINKER_CALLBACK, 1, - [cs->count_objects exists]) + [shrinker_cb_shrink_control_split],[ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_SPLIT_SHRINKER_CALLBACK, 1, + [cs->count_objects exists]) ],[ + AC_MSG_RESULT(no) + + AC_MSG_CHECKING( + [whether shrinker_register exists]) + ZFS_LINUX_TEST_RESULT([shrinker_register], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_SHRINKER_REGISTER, 1, + [shrinker_register exists]) + + dnl # We assume that the split shrinker + dnl # callback exists if + dnl # shrinker_register() exists, + dnl # because the latter is a much more + dnl # recent addition, and the macro + dnl # test for shrinker_register() only + dnl # works if the callback is split + AC_DEFINE(HAVE_SPLIT_SHRINKER_CALLBACK, + 1, [cs->count_objects exists]) + ],[ + AC_MSG_RESULT(no) ZFS_LINUX_TEST_ERROR([shrinker]) + ]) ]) ]) ]) @@ -174,10 +244,12 @@ AC_DEFUN([ZFS_AC_KERNEL_SHRINK_CONTROL_STRUCT], [ AC_DEFUN([ZFS_AC_KERNEL_SRC_SHRINKER], [ ZFS_AC_KERNEL_SRC_SUPER_BLOCK_S_SHRINK + ZFS_AC_KERNEL_SRC_SUPER_BLOCK_S_SHRINK_PTR ZFS_AC_KERNEL_SRC_SHRINK_CONTROL_HAS_NID ZFS_AC_KERNEL_SRC_SHRINKER_CALLBACK ZFS_AC_KERNEL_SRC_SHRINK_CONTROL_STRUCT ZFS_AC_KERNEL_SRC_REGISTER_SHRINKER_VARARG + ZFS_AC_KERNEL_SRC_SHRINKER_REGISTER ]) AC_DEFUN([ZFS_AC_KERNEL_SHRINKER], [ diff --git a/config/kernel-strlcpy.m4 b/config/kernel-strlcpy.m4 new file mode 100644 index 000000000000..c31cf52d78b0 --- /dev/null +++ b/config/kernel-strlcpy.m4 @@ -0,0 +1,47 @@ +dnl # +dnl # 6.8.x replaced strlcpy with strscpy. Check for both so we can provide +dnl # appropriate fallbacks. +dnl # +AC_DEFUN([ZFS_AC_KERNEL_SRC_STRLCPY], [ + ZFS_LINUX_TEST_SRC([kernel_has_strlcpy], [ + #include <linux/string.h> + ], [ + const char *src = "goodbye"; + char dst[32]; + size_t len; + len = strlcpy(dst, src, sizeof (dst)); + ]) +]) + +AC_DEFUN([ZFS_AC_KERNEL_SRC_STRSCPY], [ + ZFS_LINUX_TEST_SRC([kernel_has_strscpy], [ + #include <linux/string.h> + ], [ + const char *src = "goodbye"; + char dst[32]; + ssize_t len; + len = strscpy(dst, src, sizeof (dst)); + ]) +]) + +AC_DEFUN([ZFS_AC_KERNEL_STRLCPY], [ + AC_MSG_CHECKING([whether strlcpy() exists]) + ZFS_LINUX_TEST_RESULT([kernel_has_strlcpy], [ + AC_MSG_RESULT([yes]) + AC_DEFINE(HAVE_KERNEL_STRLCPY, 1, + [strlcpy() exists]) + ], [ + AC_MSG_RESULT([no]) + ]) +]) + +AC_DEFUN([ZFS_AC_KERNEL_STRSCPY], [ + AC_MSG_CHECKING([whether strscpy() exists]) + ZFS_LINUX_TEST_RESULT([kernel_has_strscpy], [ + AC_MSG_RESULT([yes]) + AC_DEFINE(HAVE_KERNEL_STRSCPY, 1, + [strscpy() exists]) + ], [ + AC_MSG_RESULT([no]) + ]) +]) diff --git a/config/kernel-symlink.m4 b/config/kernel-symlink.m4 index d90366d04b72..804fceab28f0 100644 --- a/config/kernel-symlink.m4 +++ b/config/kernel-symlink.m4 @@ -1,5 +1,21 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_SYMLINK], [ dnl # + dnl # 6.3 API change that changed the first arg + dnl # to struct mnt_idmap* + dnl # + ZFS_LINUX_TEST_SRC([symlink_mnt_idmap], [ + #include <linux/fs.h> + #include <linux/sched.h> + static int tmp_symlink(struct mnt_idmap *idmap, + struct inode *inode ,struct dentry *dentry, + const char *path) { return 0; } + + static const struct inode_operations + iops __attribute__ ((unused)) = { + .symlink = tmp_symlink, + }; + ],[]) + dnl # dnl # 5.12 API change that added the struct user_namespace* arg dnl # to the front of this function type's arg list. dnl # @@ -7,7 +23,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_SYMLINK], [ #include <linux/fs.h> #include <linux/sched.h> - int tmp_symlink(struct user_namespace *userns, + static int tmp_symlink(struct user_namespace *userns, struct inode *inode ,struct dentry *dentry, const char *path) { return 0; } @@ -19,12 +35,19 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_SYMLINK], [ ]) AC_DEFUN([ZFS_AC_KERNEL_SYMLINK], [ - AC_MSG_CHECKING([whether iops->symlink() takes struct user_namespace*]) - ZFS_LINUX_TEST_RESULT([symlink_userns], [ + AC_MSG_CHECKING([whether iops->symlink() takes struct mnt_idmap*]) + ZFS_LINUX_TEST_RESULT([symlink_mnt_idmap], [ AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_IOPS_SYMLINK_USERNS, 1, - [iops->symlink() takes struct user_namespace*]) + AC_DEFINE(HAVE_IOPS_SYMLINK_IDMAP, 1, + [iops->symlink() takes struct mnt_idmap*]) ],[ - AC_MSG_RESULT(no) + AC_MSG_CHECKING([whether iops->symlink() takes struct user_namespace*]) + ZFS_LINUX_TEST_RESULT([symlink_userns], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_IOPS_SYMLINK_USERNS, 1, + [iops->symlink() takes struct user_namespace*]) + ],[ + AC_MSG_RESULT(no) + ]) ]) ]) diff --git a/config/kernel-timer.m4 b/config/kernel-timer.m4 index 403cff3f4189..c710e804be0b 100644 --- a/config/kernel-timer.m4 +++ b/config/kernel-timer.m4 @@ -18,7 +18,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_TIMER_SETUP], [ int data; }; - void task_expire(struct timer_list *tl) + static void task_expire(struct timer_list *tl) { struct my_task_timer *task_timer = from_timer(task_timer, tl, timer); @@ -31,7 +31,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_TIMER_SETUP], [ ZFS_LINUX_TEST_SRC([timer_list_function], [ #include <linux/timer.h> - void task_expire(struct timer_list *tl) {} + static void task_expire(struct timer_list *tl) {} ],[ struct timer_list tl; tl.function = task_expire; diff --git a/config/kernel-tmpfile.m4 b/config/kernel-tmpfile.m4 index 0e1deb3612f3..7439514186e4 100644 --- a/config/kernel-tmpfile.m4 +++ b/config/kernel-tmpfile.m4 @@ -4,12 +4,25 @@ dnl # Add support for i_op->tmpfile dnl # AC_DEFUN([ZFS_AC_KERNEL_SRC_TMPFILE], [ dnl # + dnl # 6.3 API change + dnl # The first arg is now struct mnt_idmap * + dnl # + ZFS_LINUX_TEST_SRC([inode_operations_tmpfile_mnt_idmap], [ + #include <linux/fs.h> + static int tmpfile(struct mnt_idmap *idmap, + struct inode *inode, struct file *file, + umode_t mode) { return 0; } + static struct inode_operations + iops __attribute__ ((unused)) = { + .tmpfile = tmpfile, + }; + ],[]) dnl # 6.1 API change dnl # use struct file instead of struct dentry dnl # ZFS_LINUX_TEST_SRC([inode_operations_tmpfile], [ #include <linux/fs.h> - int tmpfile(struct user_namespace *userns, + static int tmpfile(struct user_namespace *userns, struct inode *inode, struct file *file, umode_t mode) { return 0; } static struct inode_operations @@ -23,7 +36,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_TMPFILE], [ dnl # ZFS_LINUX_TEST_SRC([inode_operations_tmpfile_dentry_userns], [ #include <linux/fs.h> - int tmpfile(struct user_namespace *userns, + static int tmpfile(struct user_namespace *userns, struct inode *inode, struct dentry *dentry, umode_t mode) { return 0; } static struct inode_operations @@ -33,7 +46,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_TMPFILE], [ ],[]) ZFS_LINUX_TEST_SRC([inode_operations_tmpfile_dentry], [ #include <linux/fs.h> - int tmpfile(struct inode *inode, struct dentry *dentry, + static int tmpfile(struct inode *inode, struct dentry *dentry, umode_t mode) { return 0; } static struct inode_operations iops __attribute__ ((unused)) = { @@ -44,23 +57,29 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_TMPFILE], [ AC_DEFUN([ZFS_AC_KERNEL_TMPFILE], [ AC_MSG_CHECKING([whether i_op->tmpfile() exists]) - ZFS_LINUX_TEST_RESULT([inode_operations_tmpfile], [ + ZFS_LINUX_TEST_RESULT([inode_operations_tmpfile_mnt_idmap], [ AC_MSG_RESULT(yes) AC_DEFINE(HAVE_TMPFILE, 1, [i_op->tmpfile() exists]) - AC_DEFINE(HAVE_TMPFILE_USERNS, 1, [i_op->tmpfile() has userns]) - ],[ - ZFS_LINUX_TEST_RESULT([inode_operations_tmpfile_dentry_userns], [ + AC_DEFINE(HAVE_TMPFILE_IDMAP, 1, [i_op->tmpfile() has mnt_idmap]) + ], [ + ZFS_LINUX_TEST_RESULT([inode_operations_tmpfile], [ AC_MSG_RESULT(yes) AC_DEFINE(HAVE_TMPFILE, 1, [i_op->tmpfile() exists]) AC_DEFINE(HAVE_TMPFILE_USERNS, 1, [i_op->tmpfile() has userns]) - AC_DEFINE(HAVE_TMPFILE_DENTRY, 1, [i_op->tmpfile() uses old dentry signature]) ],[ - ZFS_LINUX_TEST_RESULT([inode_operations_tmpfile_dentry], [ + ZFS_LINUX_TEST_RESULT([inode_operations_tmpfile_dentry_userns], [ AC_MSG_RESULT(yes) AC_DEFINE(HAVE_TMPFILE, 1, [i_op->tmpfile() exists]) + AC_DEFINE(HAVE_TMPFILE_USERNS, 1, [i_op->tmpfile() has userns]) AC_DEFINE(HAVE_TMPFILE_DENTRY, 1, [i_op->tmpfile() uses old dentry signature]) ],[ - ZFS_LINUX_REQUIRE_API([i_op->tmpfile()], [3.11]) + ZFS_LINUX_TEST_RESULT([inode_operations_tmpfile_dentry], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_TMPFILE, 1, [i_op->tmpfile() exists]) + AC_DEFINE(HAVE_TMPFILE_DENTRY, 1, [i_op->tmpfile() uses old dentry signature]) + ],[ + ZFS_LINUX_REQUIRE_API([i_op->tmpfile()], [3.11]) + ]) ]) ]) ]) diff --git a/config/kernel-vfs-direct_IO.m4 b/config/kernel-vfs-direct_IO.m4 index 82583d52fcbc..7b7b91f979f9 100644 --- a/config/kernel-vfs-direct_IO.m4 +++ b/config/kernel-vfs-direct_IO.m4 @@ -5,7 +5,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_VFS_DIRECT_IO], [ ZFS_LINUX_TEST_SRC([direct_io_iter], [ #include <linux/fs.h> - ssize_t test_direct_IO(struct kiocb *kiocb, + static ssize_t test_direct_IO(struct kiocb *kiocb, struct iov_iter *iter) { return 0; } static const struct address_space_operations @@ -17,7 +17,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_VFS_DIRECT_IO], [ ZFS_LINUX_TEST_SRC([direct_io_iter_offset], [ #include <linux/fs.h> - ssize_t test_direct_IO(struct kiocb *kiocb, + static ssize_t test_direct_IO(struct kiocb *kiocb, struct iov_iter *iter, loff_t offset) { return 0; } static const struct address_space_operations @@ -29,7 +29,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_VFS_DIRECT_IO], [ ZFS_LINUX_TEST_SRC([direct_io_iter_rw_offset], [ #include <linux/fs.h> - ssize_t test_direct_IO(int rw, struct kiocb *kiocb, + static ssize_t test_direct_IO(int rw, struct kiocb *kiocb, struct iov_iter *iter, loff_t offset) { return 0; } static const struct address_space_operations @@ -41,7 +41,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_VFS_DIRECT_IO], [ ZFS_LINUX_TEST_SRC([direct_io_iovec], [ #include <linux/fs.h> - ssize_t test_direct_IO(int rw, struct kiocb *kiocb, + static ssize_t test_direct_IO(int rw, struct kiocb *kiocb, const struct iovec *iov, loff_t offset, unsigned long nr_segs) { return 0; } diff --git a/config/kernel-vfs-iov_iter.m4 b/config/kernel-vfs-iov_iter.m4 index e0617faab02c..ff560ff3eef0 100644 --- a/config/kernel-vfs-iov_iter.m4 +++ b/config/kernel-vfs-iov_iter.m4 @@ -6,8 +6,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_VFS_IOV_ITER], [ #include <linux/fs.h> #include <linux/uio.h> ],[ - int type __attribute__ ((unused)) = - ITER_IOVEC | ITER_KVEC | ITER_BVEC | ITER_PIPE; + int type __attribute__ ((unused)) = ITER_KVEC; ]) ZFS_LINUX_TEST_SRC([iov_iter_advance], [ @@ -93,6 +92,14 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_VFS_IOV_ITER], [ struct iov_iter iter = { 0 }; __attribute__((unused)) enum iter_type i = iov_iter_type(&iter); ]) + + ZFS_LINUX_TEST_SRC([iter_iov], [ + #include <linux/fs.h> + #include <linux/uio.h> + ],[ + struct iov_iter iter = { 0 }; + __attribute__((unused)) const struct iovec *iov = iter_iov(&iter); + ]) ]) AC_DEFUN([ZFS_AC_KERNEL_VFS_IOV_ITER], [ @@ -201,4 +208,19 @@ AC_DEFUN([ZFS_AC_KERNEL_VFS_IOV_ITER], [ AC_DEFINE(HAVE_VFS_IOV_ITER, 1, [All required iov_iter interfaces are available]) ]) + + dnl # + dnl # Kernel 6.5 introduces the iter_iov() function that returns the + dnl # __iov member of an iov_iter*. The iov member was renamed to this + dnl # __iov member, and is intended to be accessed via the helper + dnl # function now. + dnl # + AC_MSG_CHECKING([whether iter_iov() is available]) + ZFS_LINUX_TEST_RESULT([iter_iov], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_ITER_IOV, 1, + [iter_iov() is available]) + ],[ + AC_MSG_RESULT(no) + ]) ]) diff --git a/config/kernel-vfs-iterate.m4 b/config/kernel-vfs-iterate.m4 index 172118eac87b..2e396daa1c0f 100644 --- a/config/kernel-vfs-iterate.m4 +++ b/config/kernel-vfs-iterate.m4 @@ -1,7 +1,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_VFS_ITERATE], [ ZFS_LINUX_TEST_SRC([file_operations_iterate_shared], [ #include <linux/fs.h> - int iterate(struct file *filp, struct dir_context * context) + static int iterate(struct file *filp, struct dir_context * context) { return 0; } static const struct file_operations fops @@ -12,7 +12,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_VFS_ITERATE], [ ZFS_LINUX_TEST_SRC([file_operations_iterate], [ #include <linux/fs.h> - int iterate(struct file *filp, + static int iterate(struct file *filp, struct dir_context *context) { return 0; } static const struct file_operations fops @@ -27,7 +27,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_VFS_ITERATE], [ ZFS_LINUX_TEST_SRC([file_operations_readdir], [ #include <linux/fs.h> - int readdir(struct file *filp, void *entry, + static int readdir(struct file *filp, void *entry, filldir_t func) { return 0; } static const struct file_operations fops diff --git a/config/kernel-vfs-rw-iterate.m4 b/config/kernel-vfs-rw-iterate.m4 index 000353ec15b0..cb20ed03099a 100644 --- a/config/kernel-vfs-rw-iterate.m4 +++ b/config/kernel-vfs-rw-iterate.m4 @@ -5,9 +5,9 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_VFS_RW_ITERATE], [ ZFS_LINUX_TEST_SRC([file_operations_rw], [ #include <linux/fs.h> - ssize_t test_read(struct kiocb *kiocb, struct iov_iter *to) + static ssize_t test_read(struct kiocb *kiocb, struct iov_iter *to) { return 0; } - ssize_t test_write(struct kiocb *kiocb, struct iov_iter *from) + static ssize_t test_write(struct kiocb *kiocb, struct iov_iter *from) { return 0; } static const struct file_operations diff --git a/config/kernel-writepage_t.m4 b/config/kernel-writepage_t.m4 new file mode 100644 index 000000000000..a82cf370c9d4 --- /dev/null +++ b/config/kernel-writepage_t.m4 @@ -0,0 +1,26 @@ +AC_DEFUN([ZFS_AC_KERNEL_SRC_WRITEPAGE_T], [ + dnl # + dnl # 6.3 API change + dnl # The writepage_t function type now has its first argument as + dnl # struct folio* instead of struct page* + dnl # + ZFS_LINUX_TEST_SRC([writepage_t_folio], [ + #include <linux/writeback.h> + static int putpage(struct folio *folio, + struct writeback_control *wbc, void *data) + { return 0; } + writepage_t func = putpage; + ],[]) +]) + +AC_DEFUN([ZFS_AC_KERNEL_WRITEPAGE_T], [ + AC_MSG_CHECKING([whether int (*writepage_t)() takes struct folio*]) + ZFS_LINUX_TEST_RESULT([writepage_t_folio], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_WRITEPAGE_T_FOLIO, 1, + [int (*writepage_t)() takes struct folio*]) + ],[ + AC_MSG_RESULT(no) + ]) +]) + diff --git a/config/kernel-xattr-handler.m4 b/config/kernel-xattr-handler.m4 index b6cbfa155007..32f58c70a500 100644 --- a/config/kernel-xattr-handler.m4 +++ b/config/kernel-xattr-handler.m4 @@ -68,7 +68,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_XATTR_HANDLER_GET], [ ZFS_LINUX_TEST_SRC([xattr_handler_get_dentry_inode], [ #include <linux/xattr.h> - int get(const struct xattr_handler *handler, + static int get(const struct xattr_handler *handler, struct dentry *dentry, struct inode *inode, const char *name, void *buffer, size_t size) { return 0; } static const struct xattr_handler @@ -80,7 +80,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_XATTR_HANDLER_GET], [ ZFS_LINUX_TEST_SRC([xattr_handler_get_xattr_handler], [ #include <linux/xattr.h> - int get(const struct xattr_handler *handler, + static int get(const struct xattr_handler *handler, struct dentry *dentry, const char *name, void *buffer, size_t size) { return 0; } static const struct xattr_handler @@ -92,7 +92,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_XATTR_HANDLER_GET], [ ZFS_LINUX_TEST_SRC([xattr_handler_get_dentry], [ #include <linux/xattr.h> - int get(struct dentry *dentry, const char *name, + static int get(struct dentry *dentry, const char *name, void *buffer, size_t size, int handler_flags) { return 0; } static const struct xattr_handler @@ -104,7 +104,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_XATTR_HANDLER_GET], [ ZFS_LINUX_TEST_SRC([xattr_handler_get_dentry_inode_flags], [ #include <linux/xattr.h> - int get(const struct xattr_handler *handler, + static int get(const struct xattr_handler *handler, struct dentry *dentry, struct inode *inode, const char *name, void *buffer, size_t size, int flags) { return 0; } @@ -179,10 +179,25 @@ dnl # dnl # Supported xattr handler set() interfaces checked newest to oldest. dnl # AC_DEFUN([ZFS_AC_KERNEL_SRC_XATTR_HANDLER_SET], [ + ZFS_LINUX_TEST_SRC([xattr_handler_set_mnt_idmap], [ + #include <linux/xattr.h> + + static int set(const struct xattr_handler *handler, + struct mnt_idmap *idmap, + struct dentry *dentry, struct inode *inode, + const char *name, const void *buffer, + size_t size, int flags) + { return 0; } + static const struct xattr_handler + xops __attribute__ ((unused)) = { + .set = set, + }; + ],[]) + ZFS_LINUX_TEST_SRC([xattr_handler_set_userns], [ #include <linux/xattr.h> - int set(const struct xattr_handler *handler, + static int set(const struct xattr_handler *handler, struct user_namespace *mnt_userns, struct dentry *dentry, struct inode *inode, const char *name, const void *buffer, @@ -197,7 +212,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_XATTR_HANDLER_SET], [ ZFS_LINUX_TEST_SRC([xattr_handler_set_dentry_inode], [ #include <linux/xattr.h> - int set(const struct xattr_handler *handler, + static int set(const struct xattr_handler *handler, struct dentry *dentry, struct inode *inode, const char *name, const void *buffer, size_t size, int flags) @@ -211,7 +226,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_XATTR_HANDLER_SET], [ ZFS_LINUX_TEST_SRC([xattr_handler_set_xattr_handler], [ #include <linux/xattr.h> - int set(const struct xattr_handler *handler, + static int set(const struct xattr_handler *handler, struct dentry *dentry, const char *name, const void *buffer, size_t size, int flags) { return 0; } @@ -224,7 +239,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_XATTR_HANDLER_SET], [ ZFS_LINUX_TEST_SRC([xattr_handler_set_dentry], [ #include <linux/xattr.h> - int set(struct dentry *dentry, const char *name, + static int set(struct dentry *dentry, const char *name, const void *buffer, size_t size, int flags, int handler_flags) { return 0; } static const struct xattr_handler @@ -240,53 +255,63 @@ AC_DEFUN([ZFS_AC_KERNEL_XATTR_HANDLER_SET], [ dnl # The xattr_handler->set() callback was changed to 8 arguments, and dnl # struct user_namespace* was inserted as arg #2 dnl # - AC_MSG_CHECKING([whether xattr_handler->set() wants dentry, inode, and user_namespace]) - ZFS_LINUX_TEST_RESULT([xattr_handler_set_userns], [ + dnl # 6.3 API change, + dnl # The xattr_handler->set() callback 2nd arg is now struct mnt_idmap * + dnl # + AC_MSG_CHECKING([whether xattr_handler->set() wants dentry, inode, and mnt_idmap]) + ZFS_LINUX_TEST_RESULT([xattr_handler_set_mnt_idmap], [ AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_XATTR_SET_USERNS, 1, - [xattr_handler->set() takes user_namespace]) - ],[ - dnl # - dnl # 4.7 API change, - dnl # The xattr_handler->set() callback was changed to take both - dnl # dentry and inode. - dnl # - AC_MSG_RESULT(no) - AC_MSG_CHECKING([whether xattr_handler->set() wants dentry and inode]) - ZFS_LINUX_TEST_RESULT([xattr_handler_set_dentry_inode], [ + AC_DEFINE(HAVE_XATTR_SET_IDMAP, 1, + [xattr_handler->set() takes mnt_idmap]) + ], [ + AC_MSG_CHECKING([whether xattr_handler->set() wants dentry, inode, and user_namespace]) + ZFS_LINUX_TEST_RESULT([xattr_handler_set_userns], [ AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_XATTR_SET_DENTRY_INODE, 1, - [xattr_handler->set() wants both dentry and inode]) + AC_DEFINE(HAVE_XATTR_SET_USERNS, 1, + [xattr_handler->set() takes user_namespace]) ],[ dnl # - dnl # 4.4 API change, - dnl # The xattr_handler->set() callback was changed to take a - dnl # xattr_handler, and handler_flags argument was removed and - dnl # should be accessed by handler->flags. + dnl # 4.7 API change, + dnl # The xattr_handler->set() callback was changed to take both + dnl # dentry and inode. dnl # AC_MSG_RESULT(no) - AC_MSG_CHECKING( - [whether xattr_handler->set() wants xattr_handler]) - ZFS_LINUX_TEST_RESULT([xattr_handler_set_xattr_handler], [ + AC_MSG_CHECKING([whether xattr_handler->set() wants dentry and inode]) + ZFS_LINUX_TEST_RESULT([xattr_handler_set_dentry_inode], [ AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_XATTR_SET_HANDLER, 1, - [xattr_handler->set() wants xattr_handler]) + AC_DEFINE(HAVE_XATTR_SET_DENTRY_INODE, 1, + [xattr_handler->set() wants both dentry and inode]) ],[ dnl # - dnl # 2.6.33 API change, - dnl # The xattr_handler->set() callback was changed - dnl # to take a dentry instead of an inode, and a - dnl # handler_flags argument was added. + dnl # 4.4 API change, + dnl # The xattr_handler->set() callback was changed to take a + dnl # xattr_handler, and handler_flags argument was removed and + dnl # should be accessed by handler->flags. dnl # AC_MSG_RESULT(no) AC_MSG_CHECKING( - [whether xattr_handler->set() wants dentry]) - ZFS_LINUX_TEST_RESULT([xattr_handler_set_dentry], [ + [whether xattr_handler->set() wants xattr_handler]) + ZFS_LINUX_TEST_RESULT([xattr_handler_set_xattr_handler], [ AC_MSG_RESULT(yes) - AC_DEFINE(HAVE_XATTR_SET_DENTRY, 1, - [xattr_handler->set() wants dentry]) + AC_DEFINE(HAVE_XATTR_SET_HANDLER, 1, + [xattr_handler->set() wants xattr_handler]) ],[ - ZFS_LINUX_TEST_ERROR([xattr set()]) + dnl # + dnl # 2.6.33 API change, + dnl # The xattr_handler->set() callback was changed + dnl # to take a dentry instead of an inode, and a + dnl # handler_flags argument was added. + dnl # + AC_MSG_RESULT(no) + AC_MSG_CHECKING( + [whether xattr_handler->set() wants dentry]) + ZFS_LINUX_TEST_RESULT([xattr_handler_set_dentry], [ + AC_MSG_RESULT(yes) + AC_DEFINE(HAVE_XATTR_SET_DENTRY, 1, + [xattr_handler->set() wants dentry]) + ],[ + ZFS_LINUX_TEST_ERROR([xattr set()]) + ]) ]) ]) ]) @@ -300,7 +325,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_XATTR_HANDLER_LIST], [ ZFS_LINUX_TEST_SRC([xattr_handler_list_simple], [ #include <linux/xattr.h> - bool list(struct dentry *dentry) { return 0; } + static bool list(struct dentry *dentry) { return 0; } static const struct xattr_handler xops __attribute__ ((unused)) = { .list = list, @@ -310,7 +335,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_XATTR_HANDLER_LIST], [ ZFS_LINUX_TEST_SRC([xattr_handler_list_xattr_handler], [ #include <linux/xattr.h> - size_t list(const struct xattr_handler *handler, + static size_t list(const struct xattr_handler *handler, struct dentry *dentry, char *list, size_t list_size, const char *name, size_t name_len) { return 0; } static const struct xattr_handler @@ -322,7 +347,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_XATTR_HANDLER_LIST], [ ZFS_LINUX_TEST_SRC([xattr_handler_list_dentry], [ #include <linux/xattr.h> - size_t list(struct dentry *dentry, + static size_t list(struct dentry *dentry, char *list, size_t list_size, const char *name, size_t name_len, int handler_flags) { return 0; } diff --git a/config/kernel.m4 b/config/kernel.m4 index 7806da7a8fc9..14c6a233a922 100644 --- a/config/kernel.m4 +++ b/config/kernel.m4 @@ -69,6 +69,7 @@ AC_DEFUN([ZFS_AC_KERNEL_TEST_SRC], [ ZFS_AC_KERNEL_SRC_INODE_OWNER_OR_CAPABLE ZFS_AC_KERNEL_SRC_XATTR ZFS_AC_KERNEL_SRC_ACL + ZFS_AC_KERNEL_SRC_INODE_SETATTR ZFS_AC_KERNEL_SRC_INODE_GETATTR ZFS_AC_KERNEL_SRC_INODE_SET_FLAGS ZFS_AC_KERNEL_SRC_INODE_SET_IVERSION @@ -130,7 +131,7 @@ AC_DEFUN([ZFS_AC_KERNEL_TEST_SRC], [ ZFS_AC_KERNEL_SRC_KSTRTOUL ZFS_AC_KERNEL_SRC_PERCPU ZFS_AC_KERNEL_SRC_CPU_HOTPLUG - ZFS_AC_KERNEL_SRC_GENERIC_FILLATTR_USERNS + ZFS_AC_KERNEL_SRC_GENERIC_FILLATTR ZFS_AC_KERNEL_SRC_MKNOD ZFS_AC_KERNEL_SRC_SYMLINK ZFS_AC_KERNEL_SRC_BIO_MAX_SEGS @@ -139,11 +140,25 @@ AC_DEFUN([ZFS_AC_KERNEL_TEST_SRC], [ ZFS_AC_KERNEL_SRC_SYSFS ZFS_AC_KERNEL_SRC_SET_SPECIAL_STATE ZFS_AC_KERNEL_SRC_STANDALONE_LINUX_STDARG + ZFS_AC_KERNEL_SRC_STRLCPY + ZFS_AC_KERNEL_SRC_STRSCPY ZFS_AC_KERNEL_SRC_PAGEMAP_FOLIO_WAIT_BIT ZFS_AC_KERNEL_SRC_ADD_DISK ZFS_AC_KERNEL_SRC_KTHREAD ZFS_AC_KERNEL_SRC_ZERO_PAGE ZFS_AC_KERNEL_SRC___COPY_FROM_USER_INATOMIC + ZFS_AC_KERNEL_SRC_FILEMAP + ZFS_AC_KERNEL_SRC_WRITEPAGE_T + ZFS_AC_KERNEL_SRC_RECLAIMED + ZFS_AC_KERNEL_SRC_REGISTER_SYSCTL_TABLE + ZFS_AC_KERNEL_SRC_COPY_SPLICE_READ + ZFS_AC_KERNEL_SRC_SYNC_BDEV + case "$host_cpu" in + powerpc*) + ZFS_AC_KERNEL_SRC_CPU_HAS_FEATURE + ZFS_AC_KERNEL_SRC_FLUSH_DCACHE_PAGE + ;; + esac AC_MSG_CHECKING([for available kernel interfaces]) ZFS_LINUX_TEST_COMPILE_ALL([kabi]) @@ -186,6 +201,7 @@ AC_DEFUN([ZFS_AC_KERNEL_TEST_RESULT], [ ZFS_AC_KERNEL_INODE_OWNER_OR_CAPABLE ZFS_AC_KERNEL_XATTR ZFS_AC_KERNEL_ACL + ZFS_AC_KERNEL_INODE_SETATTR ZFS_AC_KERNEL_INODE_GETATTR ZFS_AC_KERNEL_INODE_SET_FLAGS ZFS_AC_KERNEL_INODE_SET_IVERSION @@ -247,7 +263,7 @@ AC_DEFUN([ZFS_AC_KERNEL_TEST_RESULT], [ ZFS_AC_KERNEL_KSTRTOUL ZFS_AC_KERNEL_PERCPU ZFS_AC_KERNEL_CPU_HOTPLUG - ZFS_AC_KERNEL_GENERIC_FILLATTR_USERNS + ZFS_AC_KERNEL_GENERIC_FILLATTR ZFS_AC_KERNEL_MKNOD ZFS_AC_KERNEL_SYMLINK ZFS_AC_KERNEL_BIO_MAX_SEGS @@ -256,11 +272,25 @@ AC_DEFUN([ZFS_AC_KERNEL_TEST_RESULT], [ ZFS_AC_KERNEL_SYSFS ZFS_AC_KERNEL_SET_SPECIAL_STATE ZFS_AC_KERNEL_STANDALONE_LINUX_STDARG + ZFS_AC_KERNEL_STRLCPY + ZFS_AC_KERNEL_STRSCPY ZFS_AC_KERNEL_PAGEMAP_FOLIO_WAIT_BIT ZFS_AC_KERNEL_ADD_DISK ZFS_AC_KERNEL_KTHREAD ZFS_AC_KERNEL_ZERO_PAGE ZFS_AC_KERNEL___COPY_FROM_USER_INATOMIC + ZFS_AC_KERNEL_FILEMAP + ZFS_AC_KERNEL_WRITEPAGE_T + ZFS_AC_KERNEL_RECLAIMED + ZFS_AC_KERNEL_REGISTER_SYSCTL_TABLE + ZFS_AC_KERNEL_COPY_SPLICE_READ + ZFS_AC_KERNEL_SYNC_BDEV + case "$host_cpu" in + powerpc*) + ZFS_AC_KERNEL_CPU_HAS_FEATURE + ZFS_AC_KERNEL_FLUSH_DCACHE_PAGE + ;; + esac ]) dnl # diff --git a/config/zfs-build.m4 b/config/zfs-build.m4 index 2ab6765c3a30..1ab48c6ed1c9 100644 --- a/config/zfs-build.m4 +++ b/config/zfs-build.m4 @@ -81,7 +81,7 @@ AC_DEFUN([ZFS_AC_DEBUG], [ AC_DEFUN([ZFS_AC_DEBUGINFO_ENABLE], [ DEBUG_CFLAGS="$DEBUG_CFLAGS -g -fno-inline $NO_IPA_SRA" - KERNEL_DEBUG_CFLAGS="$KERNEL_DEBUG_CFLAGS -fno-inline $NO_IPA_SRA" + KERNEL_DEBUG_CFLAGS="$KERNEL_DEBUG_CFLAGS -fno-inline $KERNEL_NO_IPA_SRA" KERNEL_MAKE="$KERNEL_MAKE CONFIG_DEBUG_INFO=y" DEBUGINFO_ZFS="_with_debuginfo" @@ -211,12 +211,15 @@ AC_DEFUN([ZFS_AC_CONFIG_ALWAYS], [ ZFS_AC_CONFIG_ALWAYS_CC_NO_CLOBBERED ZFS_AC_CONFIG_ALWAYS_CC_INFINITE_RECURSION + ZFS_AC_CONFIG_ALWAYS_KERNEL_CC_INFINITE_RECURSION ZFS_AC_CONFIG_ALWAYS_CC_IMPLICIT_FALLTHROUGH ZFS_AC_CONFIG_ALWAYS_CC_FRAME_LARGER_THAN ZFS_AC_CONFIG_ALWAYS_CC_NO_FORMAT_TRUNCATION ZFS_AC_CONFIG_ALWAYS_CC_NO_FORMAT_ZERO_LENGTH + ZFS_AC_CONFIG_ALWAYS_CC_FORMAT_OVERFLOW ZFS_AC_CONFIG_ALWAYS_CC_NO_OMIT_FRAME_POINTER ZFS_AC_CONFIG_ALWAYS_CC_NO_IPA_SRA + ZFS_AC_CONFIG_ALWAYS_KERNEL_CC_NO_IPA_SRA ZFS_AC_CONFIG_ALWAYS_CC_ASAN ZFS_AC_CONFIG_ALWAYS_TOOLCHAIN_SIMD ZFS_AC_CONFIG_ALWAYS_SYSTEM diff --git a/configure.ac b/configure.ac index 2671434afc72..cb339ccd48a2 100644 --- a/configure.ac +++ b/configure.ac @@ -222,6 +222,7 @@ AC_CONFIG_FILES([ tests/zfs-tests/cmd/mmap_exec/Makefile tests/zfs-tests/cmd/mmap_libaio/Makefile tests/zfs-tests/cmd/mmap_seek/Makefile + tests/zfs-tests/cmd/mmap_sync/Makefile tests/zfs-tests/cmd/mmapwrite/Makefile tests/zfs-tests/cmd/nvlist_to_lua/Makefile tests/zfs-tests/cmd/randfree_file/Makefile diff --git a/contrib/bash_completion.d/zfs.in b/contrib/bash_completion.d/zfs.in index 41ce2f871e34..f97fa51330c4 100644 --- a/contrib/bash_completion.d/zfs.in +++ b/contrib/bash_completion.d/zfs.in @@ -69,7 +69,7 @@ __zfs_match_snapshot() else if [ "$cur" != "" ] && __zfs_list_datasets "$cur" &> /dev/null then - $__ZFS_CMD list -H -o name -s name -t filesystem -r "$cur" | tail -n +2 + $__ZFS_CMD list -H -o name -s name -t filesystem,volume -r "$cur" | tail -n +2 # We output the base dataset name even though we might be # completing a command that can only take a snapshot, because it # prevents bash from considering the completion finished when it diff --git a/contrib/initramfs/scripts/zfs b/contrib/initramfs/scripts/zfs index 4ce739fda704..3c51b53eec46 100644 --- a/contrib/initramfs/scripts/zfs +++ b/contrib/initramfs/scripts/zfs @@ -326,7 +326,7 @@ mount_fs() # Need the _original_ datasets mountpoint! mountpoint=$(get_fs_value "$fs" mountpoint) - ZFS_CMD="mount.zfs -o zfsutil" + ZFS_CMD="mount -o zfsutil -t zfs" if [ "$mountpoint" = "legacy" ] || [ "$mountpoint" = "none" ]; then # Can't use the mountpoint property. Might be one of our # clones. Check the 'org.zol:mountpoint' property set in @@ -343,7 +343,7 @@ mount_fs() fi # Don't use mount.zfs -o zfsutils for legacy mountpoint if [ "$mountpoint" = "legacy" ]; then - ZFS_CMD="mount.zfs" + ZFS_CMD="mount -t zfs" fi # Last hail-mary: Hope 'rootmnt' is set! mountpoint="" @@ -914,7 +914,7 @@ mountroot() echo " not specified on the kernel command line." echo "" echo "Manually mount the root filesystem on $rootmnt and then exit." - echo "Hint: Try: mount.zfs -o zfsutil ${ZFS_RPOOL-rpool}/ROOT/system $rootmnt" + echo "Hint: Try: mount -o zfsutil -t zfs ${ZFS_RPOOL-rpool}/ROOT/system $rootmnt" shell fi diff --git a/contrib/pam_zfs_key/pam_zfs_key.c b/contrib/pam_zfs_key/pam_zfs_key.c index 0db119382e7b..3137037707b5 100644 --- a/contrib/pam_zfs_key/pam_zfs_key.c +++ b/contrib/pam_zfs_key/pam_zfs_key.c @@ -548,16 +548,11 @@ zfs_key_config_modify_session_counter(pam_handle_t *pamh, errno); return (-1); } - size_t runtime_path_len = strlen(runtime_path); - size_t counter_path_len = runtime_path_len + 1 + 10; - char *counter_path = malloc(counter_path_len + 1); - if (!counter_path) { + + char *counter_path; + if (asprintf(&counter_path, "%s/%u", runtime_path, config->uid) == -1) return (-1); - } - counter_path[0] = 0; - strcat(counter_path, runtime_path); - snprintf(counter_path + runtime_path_len, counter_path_len, "/%d", - config->uid); + const int fd = open(counter_path, O_RDWR | O_CLOEXEC | O_CREAT | O_NOFOLLOW, S_IRUSR | S_IWUSR); diff --git a/copy-builtin b/copy-builtin index cd6f259092ed..18cc741b58e7 100755 --- a/copy-builtin +++ b/copy-builtin @@ -43,32 +43,8 @@ config ZFS If unsure, say N. EOF -add_after() -{ - FILE="$1" - MARKER="$2" - NEW="$3" - - while IFS='' read -r LINE - do - printf "%s\n" "$LINE" - - if [ -n "$MARKER" ] && [ "$LINE" = "$MARKER" ] - then - printf "%s\n" "$NEW" - MARKER='' - if IFS='' read -r LINE - then - [ "$LINE" != "$NEW" ] && printf "%s\n" "$LINE" - fi - fi - done < "$FILE" > "$FILE.new" - - mv "$FILE.new" "$FILE" -} - -add_after "$KERNEL_DIR/fs/Kconfig" 'if BLOCK' 'source "fs/zfs/Kconfig"' -add_after "$KERNEL_DIR/fs/Makefile" 'endif' 'obj-$(CONFIG_ZFS) += zfs/' +sed -i '/source "fs\/ext2\/Kconfig\"/i\source "fs/zfs/Kconfig"' "$KERNEL_DIR/fs/Kconfig" +echo 'obj-$(CONFIG_ZFS) += zfs/' >> "$KERNEL_DIR/fs/Makefile" echo "$0: done. now you can build the kernel with ZFS support." >&2 echo "$0: make sure you enable ZFS support (CONFIG_ZFS) before building." >&2 diff --git a/include/libzfs.h b/include/libzfs.h index 214a188f9474..6c335bbc4af9 100644 --- a/include/libzfs.h +++ b/include/libzfs.h @@ -311,6 +311,8 @@ _LIBZFS_H int zpool_vdev_remove_wanted(zpool_handle_t *, const char *); extern int zpool_vdev_fault(zpool_handle_t *, uint64_t, vdev_aux_t); extern int zpool_vdev_degrade(zpool_handle_t *, uint64_t, vdev_aux_t); +extern int zpool_vdev_set_removed_state(zpool_handle_t *, uint64_t, + vdev_aux_t); extern int zpool_vdev_clear(zpool_handle_t *, uint64_t); extern nvlist_t *zpool_find_vdev(zpool_handle_t *, const char *, boolean_t *, @@ -318,6 +320,15 @@ extern nvlist_t *zpool_find_vdev(zpool_handle_t *, const char *, boolean_t *, extern nvlist_t *zpool_find_vdev_by_physpath(zpool_handle_t *, const char *, boolean_t *, boolean_t *, boolean_t *); extern int zpool_label_disk(libzfs_handle_t *, zpool_handle_t *, const char *); +extern int zpool_prepare_disk(zpool_handle_t *zhp, nvlist_t *vdev_nv, + const char *prepare_str, char **lines[], int *lines_cnt); +extern int zpool_prepare_and_label_disk(libzfs_handle_t *hdl, + zpool_handle_t *, const char *, nvlist_t *vdev_nv, const char *prepare_str, + char **lines[], int *lines_cnt); +extern char ** zpool_vdev_script_alloc_env(const char *pool_name, + const char *vdev_path, const char *vdev_upath, + const char *vdev_enc_sysfs_path, const char *opt_key, const char *opt_val); +extern void zpool_vdev_script_free_env(char **env); extern uint64_t zpool_vdev_path_to_guid(zpool_handle_t *zhp, const char *path); const char *zpool_get_state_str(zpool_handle_t *); diff --git a/include/libzutil.h b/include/libzutil.h index 15024a4e8888..af0f74318729 100644 --- a/include/libzutil.h +++ b/include/libzutil.h @@ -78,6 +78,7 @@ extern int zpool_find_config(void *, const char *, nvlist_t **, importargs_t *, extern const char * const * zpool_default_search_paths(size_t *count); extern int zpool_read_label(int, nvlist_t **, int *); extern int zpool_label_disk_wait(const char *, int); +extern int zpool_disk_wait(const char *); struct udev_device; @@ -143,6 +144,8 @@ extern void zfs_niceraw(uint64_t, char *, size_t); extern void zpool_dump_ddt(const ddt_stat_t *, const ddt_histogram_t *); extern int zpool_history_unpack(char *, uint64_t, uint64_t *, nvlist_t ***, uint_t *); +extern void fsleep(float sec); +extern int zpool_getenv_int(const char *env, int default_val); struct zfs_cmd; int zfs_ioctl_fd(int fd, unsigned long request, struct zfs_cmd *zc); @@ -184,9 +187,65 @@ _LIBZUTIL_H void zfs_setproctitle(const char *fmt, ...); typedef int (*pool_vdev_iter_f)(void *, nvlist_t *, void *); int for_each_vdev_cb(void *zhp, nvlist_t *nv, pool_vdev_iter_f func, void *data); +int for_each_vdev_macro_helper_func(void *zhp_data, nvlist_t *nv, void *data); +int for_each_real_leaf_vdev_macro_helper_func(void *zhp_data, nvlist_t *nv, + void *data); +/* + * Often you'll want to iterate over all the vdevs in the pool, but don't want + * to use for_each_vdev() since it requires a callback function. + * + * Instead you can use FOR_EACH_VDEV(): + * + * zpool_handle_t *zhp // Assume this is initialized + * nvlist_t *nv + * ... + * FOR_EACH_VDEV(zhp, nv) { + * const char *path = NULL; + * nvlist_lookup_string(nv, ZPOOL_CONFIG_PATH, &path); + * printf("Looking at vdev %s\n", path); + * } + * + * Note: FOR_EACH_VDEV runs in O(n^2) time where n = number of vdevs. However, + * there's an upper limit of 256 vdevs per dRAID top-level vdevs (TLDs), 255 for + * raidz2 TLDs, a real world limit of ~500 vdevs for mirrors, so this shouldn't + * really be an issue. + * + * Here are some micro-benchmarks of a complete FOR_EACH_VDEV loop on a RAID0 + * pool: + * + * 100 vdevs = 0.7ms + * 500 vdevs = 17ms + * 750 vdevs = 40ms + * 1000 vdevs = 82ms + * + * The '__nv += 0' at the end of the for() loop gets around a "comma or + * semicolon followed by non-blank" checkstyle error. Note on most compliers + * the '__nv += 0' can just be replaced with 'NULL', but gcc on Centos 7 + * will give a 'warning: statement with no effect' error if you do that. + */ +#define __FOR_EACH_VDEV(__zhp, __nv, __func) { \ + __nv = zpool_get_config(__zhp, NULL); \ + VERIFY0(nvlist_lookup_nvlist(__nv, ZPOOL_CONFIG_VDEV_TREE, &__nv)); \ + } \ + for (nvlist_t *__root_nv = __nv, *__state = (nvlist_t *)0; \ + for_each_vdev_cb(&__state, __root_nv, __func, &__nv) == 1; \ + __nv += 0) + +#define FOR_EACH_VDEV(__zhp, __nv) \ + __FOR_EACH_VDEV(__zhp, __nv, for_each_vdev_macro_helper_func) + +/* + * "real leaf" vdevs are leaf vdevs that are real devices (disks or files). + * This excludes leaf vdevs like like draid spares. + */ +#define FOR_EACH_REAL_LEAF_VDEV(__zhp, __nv) \ + __FOR_EACH_VDEV(__zhp, __nv, for_each_real_leaf_vdev_macro_helper_func) + int for_each_vdev_in_nvlist(nvlist_t *nvroot, pool_vdev_iter_f func, void *data); void update_vdevs_config_dev_sysfs_path(nvlist_t *config); +_LIBZUTIL_H void update_vdev_config_dev_sysfs_path(nvlist_t *nv, + const char *path, const char *key); #ifdef __cplusplus } #endif diff --git a/include/os/freebsd/spl/sys/mod_os.h b/include/os/freebsd/spl/sys/mod_os.h index 46ea2d15ac6e..b62331b3bad0 100644 --- a/include/os/freebsd/spl/sys/mod_os.h +++ b/include/os/freebsd/spl/sys/mod_os.h @@ -92,6 +92,12 @@ #define param_set_max_auto_ashift_args(var) \ CTLTYPE_U64, &var, 0, param_set_max_auto_ashift, "QU" +#define spa_taskq_read_param_set_args(var) \ + CTLTYPE_STRING, NULL, 0, spa_taskq_read_param, "A" + +#define spa_taskq_write_param_set_args(var) \ + CTLTYPE_STRING, NULL, 0, spa_taskq_write_param, "A" + #define fletcher_4_param_set_args(var) \ CTLTYPE_STRING, NULL, 0, fletcher_4_param, "A" diff --git a/include/os/freebsd/spl/sys/vnode.h b/include/os/freebsd/spl/sys/vnode.h index b7ac12f3d676..3eb77af491a5 100644 --- a/include/os/freebsd/spl/sys/vnode.h +++ b/include/os/freebsd/spl/sys/vnode.h @@ -36,7 +36,11 @@ struct xucred; typedef struct flock flock64_t; typedef struct vnode vnode_t; typedef struct vattr vattr_t; +#if __FreeBSD_version < 1400093 typedef enum vtype vtype_t; +#else +#define vtype_t __enum_uint8(vtype) +#endif #include <sys/types.h> #include <sys/queue.h> diff --git a/include/os/freebsd/zfs/sys/zfs_znode_impl.h b/include/os/freebsd/zfs/sys/zfs_znode_impl.h index 3d93525b45ab..120884116c16 100644 --- a/include/os/freebsd/zfs/sys/zfs_znode_impl.h +++ b/include/os/freebsd/zfs/sys/zfs_znode_impl.h @@ -118,7 +118,8 @@ extern minor_t zfsdev_minor_alloc(void); #define Z_ISLNK(type) ((type) == VLNK) #define Z_ISDIR(type) ((type) == VDIR) -#define zn_has_cached_data(zp) vn_has_cached_data(ZTOV(zp)) +#define zn_has_cached_data(zp, start, end) \ + vn_has_cached_data(ZTOV(zp)) #define zn_flush_cached_data(zp, sync) vn_flush_cached_data(ZTOV(zp), sync) #define zn_rlimit_fsize(zp, uio) \ vn_rlimit_fsize(ZTOV(zp), GET_UIO_STRUCT(uio), zfs_uio_td(uio)) diff --git a/include/os/linux/kernel/linux/blkdev_compat.h b/include/os/linux/kernel/linux/blkdev_compat.h index 02a269a89fff..912919f4c8db 100644 --- a/include/os/linux/kernel/linux/blkdev_compat.h +++ b/include/os/linux/kernel/linux/blkdev_compat.h @@ -170,7 +170,11 @@ bi_status_to_errno(blk_status_t status) return (ENOLINK); case BLK_STS_TARGET: return (EREMOTEIO); +#ifdef HAVE_BLK_STS_RESV_CONFLICT + case BLK_STS_RESV_CONFLICT: +#else case BLK_STS_NEXUS: +#endif return (EBADE); case BLK_STS_MEDIUM: return (ENODATA); @@ -204,7 +208,11 @@ errno_to_bi_status(int error) case EREMOTEIO: return (BLK_STS_TARGET); case EBADE: +#ifdef HAVE_BLK_STS_RESV_CONFLICT + return (BLK_STS_RESV_CONFLICT); +#else return (BLK_STS_NEXUS); +#endif case ENODATA: return (BLK_STS_MEDIUM); case EILSEQ: @@ -326,6 +334,9 @@ zfs_check_media_change(struct block_device *bdev) return (0); } #define vdev_bdev_reread_part(bdev) zfs_check_media_change(bdev) +#elif defined(HAVE_DISK_CHECK_MEDIA_CHANGE) +#define vdev_bdev_reread_part(bdev) disk_check_media_change(bdev->bd_disk) +#define zfs_check_media_change(bdev) disk_check_media_change(bdev->bd_disk) #else /* * This is encountered if check_disk_change() and bdev_check_media_change() @@ -376,6 +387,12 @@ vdev_lookup_bdev(const char *path, dev_t *dev) #endif } +#if defined(HAVE_BLK_MODE_T) +#define blk_mode_is_open_write(flag) ((flag) & BLK_OPEN_WRITE) +#else +#define blk_mode_is_open_write(flag) ((flag) & FMODE_WRITE) +#endif + /* * Kernels without bio_set_op_attrs use bi_rw for the bio flags. */ diff --git a/include/os/linux/kernel/linux/dcache_compat.h b/include/os/linux/kernel/linux/dcache_compat.h index c90135fd3818..f87f1653ab4f 100644 --- a/include/os/linux/kernel/linux/dcache_compat.h +++ b/include/os/linux/kernel/linux/dcache_compat.h @@ -40,6 +40,21 @@ #endif /* + * Starting from Linux 5.13, flush_dcache_page() becomes an inline function + * and under some configurations, may indirectly referencing GPL-only + * cpu_feature_keys on powerpc. Override this function when it is detected + * being GPL-only. + */ +#if defined __powerpc__ && defined HAVE_FLUSH_DCACHE_PAGE_GPL_ONLY +#include <linux/simd_powerpc.h> +#define flush_dcache_page(page) do { \ + if (!cpu_has_feature(CPU_FTR_COHERENT_ICACHE) && \ + test_bit(PG_dcache_clean, &(page)->flags)) \ + clear_bit(PG_dcache_clean, &(page)->flags); \ + } while (0) +#endif + +/* * 2.6.30 API change, * The const keyword was added to the 'struct dentry_operations' in * the dentry structure. To handle this we define an appropriate diff --git a/include/os/linux/kernel/linux/simd_aarch64.h b/include/os/linux/kernel/linux/simd_aarch64.h index 50937e97ced1..cd1a78240016 100644 --- a/include/os/linux/kernel/linux/simd_aarch64.h +++ b/include/os/linux/kernel/linux/simd_aarch64.h @@ -43,9 +43,15 @@ #include <sys/types.h> #include <asm/neon.h> +#if (defined(HAVE_KERNEL_NEON) && defined(CONFIG_KERNEL_MODE_NEON)) #define kfpu_allowed() 1 #define kfpu_begin() kernel_neon_begin() #define kfpu_end() kernel_neon_end() +#else +#define kfpu_allowed() 0 +#define kfpu_begin() do {} while (0) +#define kfpu_end() do {} while (0) +#endif #define kfpu_init() 0 #define kfpu_fini() ((void) 0) diff --git a/include/os/linux/kernel/linux/simd_powerpc.h b/include/os/linux/kernel/linux/simd_powerpc.h index 108cef22f56f..422b85af3ac2 100644 --- a/include/os/linux/kernel/linux/simd_powerpc.h +++ b/include/os/linux/kernel/linux/simd_powerpc.h @@ -77,6 +77,17 @@ #define kfpu_fini() ((void) 0) /* + * Linux 4.7 makes cpu_has_feature to use jump labels on powerpc if + * CONFIG_JUMP_LABEL_FEATURE_CHECKS is enabled, in this case however it + * references GPL-only symbol cpu_feature_keys. Therefore we overrides this + * interface when it is detected being GPL-only. + */ +#if defined(CONFIG_JUMP_LABEL_FEATURE_CHECKS) && \ + defined(HAVE_CPU_HAS_FEATURE_GPL_ONLY) +#define cpu_has_feature(feature) early_cpu_has_feature(feature) +#endif + +/* * Check if AltiVec instruction set is available */ static inline boolean_t diff --git a/include/os/linux/kernel/linux/vfs_compat.h b/include/os/linux/kernel/linux/vfs_compat.h index 91e908598fbb..045135fe9226 100644 --- a/include/os/linux/kernel/linux/vfs_compat.h +++ b/include/os/linux/kernel/linux/vfs_compat.h @@ -344,7 +344,8 @@ static inline void zfs_gid_write(struct inode *ip, gid_t gid) * 4.9 API change */ #if !(defined(HAVE_SETATTR_PREPARE_NO_USERNS) || \ - defined(HAVE_SETATTR_PREPARE_USERNS)) + defined(HAVE_SETATTR_PREPARE_USERNS) || \ + defined(HAVE_SETATTR_PREPARE_IDMAP)) static inline int setattr_prepare(struct dentry *dentry, struct iattr *ia) { @@ -399,6 +400,15 @@ func(struct user_namespace *user_ns, const struct path *path, \ return (func##_impl(user_ns, path, stat, request_mask, \ query_flags)); \ } +#elif defined(HAVE_IDMAP_IOPS_GETATTR) +#define ZPL_GETATTR_WRAPPER(func) \ +static int \ +func(struct mnt_idmap *user_ns, const struct path *path, \ + struct kstat *stat, u32 request_mask, unsigned int query_flags) \ +{ \ + return (func##_impl(user_ns, path, stat, request_mask, \ + query_flags)); \ +} #else #error #endif @@ -450,8 +460,21 @@ zpl_is_32bit_api(void) * 5.12 API change * To support id-mapped mounts, generic_fillattr() was modified to * accept a new struct user_namespace* as its first arg. + * + * 6.3 API change + * generic_fillattr() first arg is changed to struct mnt_idmap * + * + * 6.6 API change + * generic_fillattr() gets new second arg request_mask, a u32 type + * */ -#ifdef HAVE_GENERIC_FILLATTR_USERNS +#ifdef HAVE_GENERIC_FILLATTR_IDMAP +#define zpl_generic_fillattr(idmap, ip, sp) \ + generic_fillattr(idmap, ip, sp) +#elif defined(HAVE_GENERIC_FILLATTR_IDMAP_REQMASK) +#define zpl_generic_fillattr(idmap, rqm, ip, sp) \ + generic_fillattr(idmap, rqm, ip, sp) +#elif defined(HAVE_GENERIC_FILLATTR_USERNS) #define zpl_generic_fillattr(user_ns, ip, sp) \ generic_fillattr(user_ns, ip, sp) #else diff --git a/include/os/linux/kernel/linux/xattr_compat.h b/include/os/linux/kernel/linux/xattr_compat.h index 30403fe87397..3ffd0016911b 100644 --- a/include/os/linux/kernel/linux/xattr_compat.h +++ b/include/os/linux/kernel/linux/xattr_compat.h @@ -134,19 +134,34 @@ fn(const struct xattr_handler *handler, struct dentry *dentry, \ #endif /* + * 6.3 API change, + * The xattr_handler->set() callback was changed to take the + * struct mnt_idmap* as the first arg, to support idmapped + * mounts. + */ +#if defined(HAVE_XATTR_SET_IDMAP) +#define ZPL_XATTR_SET_WRAPPER(fn) \ +static int \ +fn(const struct xattr_handler *handler, struct mnt_idmap *user_ns, \ + struct dentry *dentry, struct inode *inode, const char *name, \ + const void *buffer, size_t size, int flags) \ +{ \ + return (__ ## fn(user_ns, inode, name, buffer, size, flags)); \ +} +/* * 5.12 API change, * The xattr_handler->set() callback was changed to take the * struct user_namespace* as the first arg, to support idmapped * mounts. */ -#if defined(HAVE_XATTR_SET_USERNS) +#elif defined(HAVE_XATTR_SET_USERNS) #define ZPL_XATTR_SET_WRAPPER(fn) \ static int \ fn(const struct xattr_handler *handler, struct user_namespace *user_ns, \ struct dentry *dentry, struct inode *inode, const char *name, \ const void *buffer, size_t size, int flags) \ { \ - return (__ ## fn(inode, name, buffer, size, flags)); \ + return (__ ## fn(user_ns, inode, name, buffer, size, flags)); \ } /* * 4.7 API change, @@ -160,7 +175,7 @@ fn(const struct xattr_handler *handler, struct dentry *dentry, \ struct inode *inode, const char *name, const void *buffer, \ size_t size, int flags) \ { \ - return (__ ## fn(inode, name, buffer, size, flags)); \ + return (__ ## fn(kcred->user_ns, inode, name, buffer, size, flags));\ } /* * 4.4 API change, @@ -174,7 +189,8 @@ static int \ fn(const struct xattr_handler *handler, struct dentry *dentry, \ const char *name, const void *buffer, size_t size, int flags) \ { \ - return (__ ## fn(dentry->d_inode, name, buffer, size, flags)); \ + return (__ ## fn(kcred->user_ns, dentry->d_inode, name, \ + buffer, size, flags)); \ } /* * 2.6.33 API change, @@ -187,7 +203,8 @@ static int \ fn(struct dentry *dentry, const char *name, const void *buffer, \ size_t size, int flags, int unused_handler_flags) \ { \ - return (__ ## fn(dentry->d_inode, name, buffer, size, flags)); \ + return (__ ## fn(kcred->user_ns, dentry->d_inode, name, buffer, \ + size, flags)); \ } #else #error "Unsupported kernel" diff --git a/include/os/linux/spl/sys/Makefile.am b/include/os/linux/spl/sys/Makefile.am index 450baffc395e..ccb22bddb0d1 100644 --- a/include/os/linux/spl/sys/Makefile.am +++ b/include/os/linux/spl/sys/Makefile.am @@ -34,6 +34,7 @@ KERNEL_H = \ signal.h \ simd.h \ stat.h \ + string.h \ strings.h \ sunddi.h \ sysmacros.h \ diff --git a/include/os/linux/spl/sys/cred.h b/include/os/linux/spl/sys/cred.h index b7d3f38d70bb..501bd4566cb7 100644 --- a/include/os/linux/spl/sys/cred.h +++ b/include/os/linux/spl/sys/cred.h @@ -45,6 +45,8 @@ typedef struct cred cred_t; #define SGID_TO_KGID(x) (KGIDT_INIT(x)) #define KGIDP_TO_SGIDP(x) (&(x)->val) +extern zidmap_t *zfs_get_init_idmap(void); + extern void crhold(cred_t *cr); extern void crfree(cred_t *cr); extern uid_t crgetuid(const cred_t *cr); diff --git a/include/os/linux/spl/sys/kmem_cache.h b/include/os/linux/spl/sys/kmem_cache.h index 48006ec5d27e..c4aeebb81558 100644 --- a/include/os/linux/spl/sys/kmem_cache.h +++ b/include/os/linux/spl/sys/kmem_cache.h @@ -111,7 +111,7 @@ typedef struct spl_kmem_magazine { uint32_t skm_refill; /* Batch refill size */ struct spl_kmem_cache *skm_cache; /* Owned by cache */ unsigned int skm_cpu; /* Owned by cpu */ - void *skm_objs[0]; /* Object pointers */ + void *skm_objs[]; /* Object pointers */ } spl_kmem_magazine_t; typedef struct spl_kmem_obj { diff --git a/include/os/linux/spl/sys/shrinker.h b/include/os/linux/spl/sys/shrinker.h index d472754be4f4..bca4c850694a 100644 --- a/include/os/linux/spl/sys/shrinker.h +++ b/include/os/linux/spl/sys/shrinker.h @@ -29,12 +29,13 @@ /* * Due to frequent changes in the shrinker API the following - * compatibility wrappers should be used. They are as follows: + * compatibility wrapper should be used. * - * SPL_SHRINKER_DECLARE(varname, countfunc, scanfunc, seek_cost); + * shrinker = spl_register_shrinker(name, countfunc, scanfunc, seek_cost); + * spl_unregister_shrinker(shrinker); * - * SPL_SHRINKER_DECLARE is used to declare a shrinker with the name varname, - * which is passed to spl_register_shrinker()/spl_unregister_shrinker(). + * spl_register_shrinker is used to create and register a shrinker with the + * given name. * The countfunc returns the number of free-able objects. * The scanfunc returns the number of objects that were freed. * The callbacks can return SHRINK_STOP if further calls can't make any more @@ -57,57 +58,28 @@ * ...scan objects in the cache and reclaim them... * } * - * SPL_SHRINKER_DECLARE(my_shrinker, my_count, my_scan, DEFAULT_SEEKS); + * static struct shrinker *my_shrinker; * * void my_init_func(void) { - * spl_register_shrinker(&my_shrinker); + * my_shrinker = spl_register_shrinker("my-shrinker", + * my_count, my_scan, DEFAULT_SEEKS); + * } + * + * void my_fini_func(void) { + * spl_unregister_shrinker(my_shrinker); * } */ -#ifdef HAVE_REGISTER_SHRINKER_VARARG -#define spl_register_shrinker(x) register_shrinker(x, "zfs-arc-shrinker") -#else -#define spl_register_shrinker(x) register_shrinker(x) -#endif -#define spl_unregister_shrinker(x) unregister_shrinker(x) +typedef unsigned long (*spl_shrinker_cb) + (struct shrinker *, struct shrink_control *); -/* - * Linux 3.0 to 3.11 Shrinker API Compatibility. - */ -#if defined(HAVE_SINGLE_SHRINKER_CALLBACK) -#define SPL_SHRINKER_DECLARE(varname, countfunc, scanfunc, seek_cost) \ -static int \ -__ ## varname ## _wrapper(struct shrinker *shrink, struct shrink_control *sc)\ -{ \ - if (sc->nr_to_scan != 0) { \ - (void) scanfunc(shrink, sc); \ - } \ - return (countfunc(shrink, sc)); \ -} \ - \ -static struct shrinker varname = { \ - .shrink = __ ## varname ## _wrapper, \ - .seeks = seek_cost, \ -} +struct shrinker *spl_register_shrinker(const char *name, + spl_shrinker_cb countfunc, spl_shrinker_cb scanfunc, int seek_cost); +void spl_unregister_shrinker(struct shrinker *); +#ifndef SHRINK_STOP +/* 3.0-3.11 compatibility */ #define SHRINK_STOP (-1) - -/* - * Linux 3.12 and later Shrinker API Compatibility. - */ -#elif defined(HAVE_SPLIT_SHRINKER_CALLBACK) -#define SPL_SHRINKER_DECLARE(varname, countfunc, scanfunc, seek_cost) \ -static struct shrinker varname = { \ - .count_objects = countfunc, \ - .scan_objects = scanfunc, \ - .seeks = seek_cost, \ -} - -#else -/* - * Linux 2.x to 2.6.22, or a newer shrinker API has been introduced. - */ -#error "Unknown shrinker callback" #endif #endif /* SPL_SHRINKER_H */ diff --git a/include/os/linux/spl/sys/string.h b/include/os/linux/spl/sys/string.h new file mode 100644 index 000000000000..f44bf23eb326 --- /dev/null +++ b/include/os/linux/spl/sys/string.h @@ -0,0 +1,50 @@ +/* + * Copyright (C) 2007-2010 Lawrence Livermore National Security, LLC. + * Produced at Lawrence Livermore National Laboratory (cf, DISCLAIMER). + * Written by Brian Behlendorf <behlendorf1@llnl.gov>. + * UCRL-CODE-235197 + * + * This file is part of the SPL, Solaris Porting Layer. + * + * The SPL is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + * + * The SPL is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + * + * You should have received a copy of the GNU General Public License along + * with the SPL. If not, see <http://www.gnu.org/licenses/>. + */ + +#ifndef _SPL_STRING_H +#define _SPL_STRING_H + +#include <linux/string.h> + +/* Fallbacks for kernel missing strlcpy */ +#ifndef HAVE_KERNEL_STRLCPY + +#if defined(HAVE_KERNEL_STRSCPY) +/* + * strscpy is strlcpy, but returns an error on truncation. strlcpy is defined + * to return strlen(src), so detect error and override it. + */ +static inline size_t +strlcpy(char *dest, const char *src, size_t size) +{ + ssize_t ret = strscpy(dest, src, size); + if (likely(ret > 0)) + return ((size_t)ret); + return (strlen(src)); +} +#else +#error "no strlcpy fallback available" +#endif + +#endif /* HAVE_KERNEL_STRLCPY */ + +#endif /* _SPL_STRING_H */ diff --git a/include/os/linux/spl/sys/types.h b/include/os/linux/spl/sys/types.h index b44c94518750..9f85685fac77 100644 --- a/include/os/linux/spl/sys/types.h +++ b/include/os/linux/spl/sys/types.h @@ -38,7 +38,7 @@ typedef unsigned long ulong_t; typedef unsigned long long u_longlong_t; typedef long long longlong_t; -typedef unsigned long intptr_t; +typedef long intptr_t; typedef unsigned long long rlim64_t; typedef struct task_struct kthread_t; @@ -54,4 +54,18 @@ typedef ulong_t pgcnt_t; typedef int major_t; typedef int minor_t; +struct user_namespace; +#ifdef HAVE_IOPS_CREATE_IDMAP +#include <linux/refcount.h> +struct mnt_idmap { + struct user_namespace *owner; + refcount_t count; +}; +typedef struct mnt_idmap zidmap_t; +#else +typedef struct user_namespace zidmap_t; +#endif + +extern zidmap_t *zfs_init_idmap; + #endif /* _SPL_TYPES_H */ diff --git a/include/os/linux/spl/sys/uio.h b/include/os/linux/spl/sys/uio.h index 439eec986236..68fab0314616 100644 --- a/include/os/linux/spl/sys/uio.h +++ b/include/os/linux/spl/sys/uio.h @@ -146,4 +146,16 @@ zfs_uio_iov_iter_init(zfs_uio_t *uio, struct iov_iter *iter, offset_t offset, } #endif +#if defined(HAVE_ITER_IOV) +#define zfs_uio_iter_iov(iter) iter_iov((iter)) +#else +#define zfs_uio_iter_iov(iter) (iter)->iov +#endif + +#if defined(HAVE_IOV_ITER_TYPE) +#define zfs_uio_iov_iter_type(iter) iov_iter_type((iter)) +#else +#define zfs_uio_iov_iter_type(iter) (iter)->type +#endif + #endif /* SPL_UIO_H */ diff --git a/include/os/linux/zfs/sys/trace_acl.h b/include/os/linux/zfs/sys/trace_acl.h index 21bcefa4ea26..656552749b9f 100644 --- a/include/os/linux/zfs/sys/trace_acl.h +++ b/include/os/linux/zfs/sys/trace_acl.h @@ -58,9 +58,10 @@ DECLARE_EVENT_CLASS(zfs_ace_class, __field(uint64_t, z_size) __field(uint64_t, z_pflags) __field(uint32_t, z_sync_cnt) + __field(uint32_t, z_sync_writes_cnt) + __field(uint32_t, z_async_writes_cnt) __field(mode_t, z_mode) __field(boolean_t, z_is_sa) - __field(boolean_t, z_is_mapped) __field(boolean_t, z_is_ctldir) __field(uint32_t, i_uid) @@ -90,9 +91,10 @@ DECLARE_EVENT_CLASS(zfs_ace_class, __entry->z_size = zn->z_size; __entry->z_pflags = zn->z_pflags; __entry->z_sync_cnt = zn->z_sync_cnt; + __entry->z_sync_writes_cnt = zn->z_sync_writes_cnt; + __entry->z_async_writes_cnt = zn->z_async_writes_cnt; __entry->z_mode = zn->z_mode; __entry->z_is_sa = zn->z_is_sa; - __entry->z_is_mapped = zn->z_is_mapped; __entry->z_is_ctldir = zn->z_is_ctldir; __entry->i_uid = KUID_TO_SUID(ZTOI(zn)->i_uid); @@ -114,18 +116,18 @@ DECLARE_EVENT_CLASS(zfs_ace_class, TP_printk("zn { id %llu unlinked %u atime_dirty %u " "zn_prefetch %u blksz %u seq %u " "mapcnt %llu size %llu pflags %llu " - "sync_cnt %u mode 0x%x is_sa %d " - "is_mapped %d is_ctldir %d inode { " - "uid %u gid %u ino %lu nlink %u size %lli " + "sync_cnt %u sync_writes_cnt %u async_writes_cnt %u " + "mode 0x%x is_sa %d is_ctldir %d " + "inode { uid %u gid %u ino %lu nlink %u size %lli " "blkbits %u bytes %u mode 0x%x generation %x } } " "ace { type %u flags %u access_mask %u } mask_matched %u", __entry->z_id, __entry->z_unlinked, __entry->z_atime_dirty, __entry->z_zn_prefetch, __entry->z_blksz, __entry->z_seq, __entry->z_mapcnt, __entry->z_size, - __entry->z_pflags, __entry->z_sync_cnt, __entry->z_mode, - __entry->z_is_sa, __entry->z_is_mapped, - __entry->z_is_ctldir, __entry->i_uid, - __entry->i_gid, __entry->i_ino, __entry->i_nlink, + __entry->z_pflags, __entry->z_sync_cnt, + __entry->z_sync_writes_cnt, __entry->z_async_writes_cnt, + __entry->z_mode, __entry->z_is_sa, __entry->z_is_ctldir, + __entry->i_uid, __entry->i_gid, __entry->i_ino, __entry->i_nlink, __entry->i_size, __entry->i_blkbits, __entry->i_bytes, __entry->i_mode, __entry->i_generation, __entry->z_type, __entry->z_flags, __entry->z_access_mask, diff --git a/include/os/linux/zfs/sys/zfs_vnops_os.h b/include/os/linux/zfs/sys/zfs_vnops_os.h index 47f91e4a6cf4..a29f35d258e4 100644 --- a/include/os/linux/zfs/sys/zfs_vnops_os.h +++ b/include/os/linux/zfs/sys/zfs_vnops_os.h @@ -54,8 +54,12 @@ extern int zfs_mkdir(znode_t *dzp, char *dirname, vattr_t *vap, extern int zfs_rmdir(znode_t *dzp, char *name, znode_t *cwd, cred_t *cr, int flags); extern int zfs_readdir(struct inode *ip, zpl_dir_context_t *ctx, cred_t *cr); -extern int zfs_getattr_fast(struct user_namespace *, struct inode *ip, - struct kstat *sp); +#ifdef HAVE_GENERIC_FILLATTR_IDMAP_REQMASK +extern int zfs_getattr_fast(zidmap_t *, u32 request_mask, struct inode *ip, + struct kstat *sp); +#else +extern int zfs_getattr_fast(zidmap_t *, struct inode *ip, struct kstat *sp); +#endif extern int zfs_setattr(znode_t *zp, vattr_t *vap, int flag, cred_t *cr); extern int zfs_rename(znode_t *sdzp, char *snm, znode_t *tdzp, char *tnm, cred_t *cr, int flags); @@ -68,9 +72,9 @@ extern void zfs_inactive(struct inode *ip); extern int zfs_space(znode_t *zp, int cmd, flock64_t *bfp, int flag, offset_t offset, cred_t *cr); extern int zfs_fid(struct inode *ip, fid_t *fidp); -extern int zfs_getpage(struct inode *ip, struct page *pl[], int nr_pages); +extern int zfs_getpage(struct inode *ip, struct page *pp); extern int zfs_putpage(struct inode *ip, struct page *pp, - struct writeback_control *wbc); + struct writeback_control *wbc, boolean_t for_sync); extern int zfs_dirty_inode(struct inode *ip, int flags); extern int zfs_map(struct inode *ip, offset_t off, caddr_t *addrp, size_t len, unsigned long vm_flags); diff --git a/include/os/linux/zfs/sys/zfs_znode_impl.h b/include/os/linux/zfs/sys/zfs_znode_impl.h index de46fc8f2bd8..9b9ac7a4f189 100644 --- a/include/os/linux/zfs/sys/zfs_znode_impl.h +++ b/include/os/linux/zfs/sys/zfs_znode_impl.h @@ -47,9 +47,16 @@ extern "C" { #endif +#if defined(HAVE_FILEMAP_RANGE_HAS_PAGE) #define ZNODE_OS_FIELDS \ inode_timespec_t z_btime; /* creation/birth time (cached) */ \ struct inode z_inode; +#else +#define ZNODE_OS_FIELDS \ + inode_timespec_t z_btime; /* creation/birth time (cached) */ \ + struct inode z_inode; \ + boolean_t z_is_mapped; /* we are mmap'ed */ +#endif /* * Convert between znode pointers and inode pointers @@ -70,7 +77,14 @@ extern "C" { #define Z_ISDEV(type) (S_ISCHR(type) || S_ISBLK(type) || S_ISFIFO(type)) #define Z_ISDIR(type) S_ISDIR(type) -#define zn_has_cached_data(zp) ((zp)->z_is_mapped) +#if defined(HAVE_FILEMAP_RANGE_HAS_PAGE) +#define zn_has_cached_data(zp, start, end) \ + filemap_range_has_page(ZTOI(zp)->i_mapping, start, end) +#else +#define zn_has_cached_data(zp, start, end) \ + ((zp)->z_is_mapped) +#endif + #define zn_flush_cached_data(zp, sync) write_inode_now(ZTOI(zp), sync) #define zn_rlimit_fsize(zp, uio) (0) diff --git a/include/os/linux/zfs/sys/zpl.h b/include/os/linux/zfs/sys/zpl.h index ac9815d4ee00..69b6c1185bb1 100644 --- a/include/os/linux/zfs/sys/zpl.h +++ b/include/os/linux/zfs/sys/zpl.h @@ -64,7 +64,10 @@ extern int zpl_xattr_security_init(struct inode *ip, struct inode *dip, const struct qstr *qstr); #if defined(CONFIG_FS_POSIX_ACL) #if defined(HAVE_SET_ACL) -#if defined(HAVE_SET_ACL_USERNS) +#if defined(HAVE_SET_ACL_IDMAP_DENTRY) +extern int zpl_set_acl(struct mnt_idmap *idmap, struct dentry *dentry, + struct posix_acl *acl, int type); +#elif defined(HAVE_SET_ACL_USERNS) extern int zpl_set_acl(struct user_namespace *userns, struct inode *ip, struct posix_acl *acl, int type); #elif defined(HAVE_SET_ACL_USERNS_DENTRY_ARG2) @@ -186,13 +189,15 @@ zpl_dir_emit_dots(struct file *file, zpl_dir_context_t *ctx) #if defined(HAVE_INODE_OWNER_OR_CAPABLE) #define zpl_inode_owner_or_capable(ns, ip) inode_owner_or_capable(ip) -#elif defined(HAVE_INODE_OWNER_OR_CAPABLE_IDMAPPED) +#elif defined(HAVE_INODE_OWNER_OR_CAPABLE_USERNS) #define zpl_inode_owner_or_capable(ns, ip) inode_owner_or_capable(ns, ip) +#elif defined(HAVE_INODE_OWNER_OR_CAPABLE_IDMAP) +#define zpl_inode_owner_or_capable(idmap, ip) inode_owner_or_capable(idmap, ip) #else #error "Unsupported kernel" #endif -#ifdef HAVE_SETATTR_PREPARE_USERNS +#if defined(HAVE_SETATTR_PREPARE_USERNS) || defined(HAVE_SETATTR_PREPARE_IDMAP) #define zpl_setattr_prepare(ns, dentry, ia) setattr_prepare(ns, dentry, ia) #else /* @@ -202,4 +207,35 @@ zpl_dir_emit_dots(struct file *file, zpl_dir_context_t *ctx) #define zpl_setattr_prepare(ns, dentry, ia) setattr_prepare(dentry, ia) #endif +#ifdef HAVE_INODE_GET_CTIME +#define zpl_inode_get_ctime(ip) inode_get_ctime(ip) +#else +#define zpl_inode_get_ctime(ip) (ip->i_ctime) +#endif +#ifdef HAVE_INODE_SET_CTIME_TO_TS +#define zpl_inode_set_ctime_to_ts(ip, ts) inode_set_ctime_to_ts(ip, ts) +#else +#define zpl_inode_set_ctime_to_ts(ip, ts) (ip->i_ctime = ts) +#endif +#ifdef HAVE_INODE_GET_ATIME +#define zpl_inode_get_atime(ip) inode_get_atime(ip) +#else +#define zpl_inode_get_atime(ip) (ip->i_atime) +#endif +#ifdef HAVE_INODE_SET_ATIME_TO_TS +#define zpl_inode_set_atime_to_ts(ip, ts) inode_set_atime_to_ts(ip, ts) +#else +#define zpl_inode_set_atime_to_ts(ip, ts) (ip->i_atime = ts) +#endif +#ifdef HAVE_INODE_GET_MTIME +#define zpl_inode_get_mtime(ip) inode_get_mtime(ip) +#else +#define zpl_inode_get_mtime(ip) (ip->i_mtime) +#endif +#ifdef HAVE_INODE_SET_MTIME_TO_TS +#define zpl_inode_set_mtime_to_ts(ip, ts) inode_set_mtime_to_ts(ip, ts) +#else +#define zpl_inode_set_mtime_to_ts(ip, ts) (ip->i_mtime = ts) +#endif + #endif /* _SYS_ZPL_H */ diff --git a/include/sys/dmu.h b/include/sys/dmu.h index 7bdd42e8bed4..12bd887200e9 100644 --- a/include/sys/dmu.h +++ b/include/sys/dmu.h @@ -778,6 +778,9 @@ dmu_tx_t *dmu_tx_create(objset_t *os); void dmu_tx_hold_write(dmu_tx_t *tx, uint64_t object, uint64_t off, int len); void dmu_tx_hold_write_by_dnode(dmu_tx_t *tx, dnode_t *dn, uint64_t off, int len); +void dmu_tx_hold_append(dmu_tx_t *tx, uint64_t object, uint64_t off, int len); +void dmu_tx_hold_append_by_dnode(dmu_tx_t *tx, dnode_t *dn, uint64_t off, + int len); void dmu_tx_hold_free(dmu_tx_t *tx, uint64_t object, uint64_t off, uint64_t len); void dmu_tx_hold_free_by_dnode(dmu_tx_t *tx, dnode_t *dn, uint64_t off, diff --git a/include/sys/dmu_objset.h b/include/sys/dmu_objset.h index 7ade2dc91247..fffcbcfcafd0 100644 --- a/include/sys/dmu_objset.h +++ b/include/sys/dmu_objset.h @@ -72,6 +72,10 @@ struct dmu_tx; */ #define OBJSET_CRYPT_PORTABLE_FLAGS_MASK (0) +#if defined(__clang__) +#pragma clang diagnostic push +#pragma clang diagnostic ignored "-Wgnu-variable-sized-type-not-at-end" +#endif typedef struct objset_phys { dnode_phys_t os_meta_dnode; zil_header_t os_zil_header; @@ -88,6 +92,9 @@ typedef struct objset_phys { char os_pad1[OBJSET_PHYS_SIZE_V3 - OBJSET_PHYS_SIZE_V2 - sizeof (dnode_phys_t)]; } objset_phys_t; +#if defined(__clang__) +#pragma clang diagnostic pop +#endif typedef int (*dmu_objset_upgrade_cb_t)(objset_t *); diff --git a/include/sys/dmu_tx.h b/include/sys/dmu_tx.h index ad3f1b0e47ca..e8886fd4e6f9 100644 --- a/include/sys/dmu_tx.h +++ b/include/sys/dmu_tx.h @@ -90,6 +90,7 @@ enum dmu_tx_hold_type { THT_ZAP, THT_SPACE, THT_SPILL, + THT_APPEND, THT_NUMTYPES }; diff --git a/include/sys/dnode.h b/include/sys/dnode.h index 20b7c2aaf2be..39bbdae44339 100644 --- a/include/sys/dnode.h +++ b/include/sys/dnode.h @@ -120,7 +120,11 @@ extern "C" { #define DN_MAX_LEVELS (DIV_ROUND_UP(DN_MAX_OFFSET_SHIFT - SPA_MINBLOCKSHIFT, \ DN_MIN_INDBLKSHIFT - SPA_BLKPTRSHIFT) + 1) -#define DN_BONUS(dnp) ((void*)((dnp)->dn_bonus + \ +/* + * Use the flexible array instead of the fixed length one dn_bonus + * to address memcpy/memmove fortify error + */ +#define DN_BONUS(dnp) ((void*)((dnp)->dn_bonus_flexible + \ (((dnp)->dn_nblkptr - 1) * sizeof (blkptr_t)))) #define DN_MAX_BONUS_LEN(dnp) \ ((dnp->dn_flags & DNODE_FLAG_SPILL_BLKPTR) ? \ @@ -266,6 +270,10 @@ typedef struct dnode_phys { sizeof (blkptr_t)]; blkptr_t dn_spill; }; + struct { + blkptr_t __dn_ignore4; + uint8_t dn_bonus_flexible[]; + }; }; } dnode_phys_t; diff --git a/include/sys/fs/zfs.h b/include/sys/fs/zfs.h index 111e70ece151..84f5aee59d81 100644 --- a/include/sys/fs/zfs.h +++ b/include/sys/fs/zfs.h @@ -1173,6 +1173,7 @@ typedef enum pool_initialize_func { POOL_INITIALIZE_START, POOL_INITIALIZE_CANCEL, POOL_INITIALIZE_SUSPEND, + POOL_INITIALIZE_UNINIT, POOL_INITIALIZE_FUNCS } pool_initialize_func_t; diff --git a/include/sys/spa.h b/include/sys/spa.h index fedadab459b7..42f7fec0fc1f 100644 --- a/include/sys/spa.h +++ b/include/sys/spa.h @@ -785,6 +785,7 @@ extern int bpobj_enqueue_free_cb(void *arg, const blkptr_t *bp, dmu_tx_t *tx); #define SPA_ASYNC_L2CACHE_REBUILD 0x800 #define SPA_ASYNC_L2CACHE_TRIM 0x1000 #define SPA_ASYNC_REBUILD_DONE 0x2000 +#define SPA_ASYNC_DETACH_SPARE 0x4000 /* device manipulation */ extern int spa_vdev_add(spa_t *spa, nvlist_t *nvroot); @@ -971,6 +972,8 @@ extern int spa_import_progress_set_state(uint64_t pool_guid, /* Pool configuration locks */ extern int spa_config_tryenter(spa_t *spa, int locks, void *tag, krw_t rw); extern void spa_config_enter(spa_t *spa, int locks, const void *tag, krw_t rw); +extern void spa_config_enter_mmp(spa_t *spa, int locks, const void *tag, + krw_t rw); extern void spa_config_exit(spa_t *spa, int locks, const void *tag); extern int spa_config_held(spa_t *spa, int locks, krw_t rw); diff --git a/include/sys/vdev_initialize.h b/include/sys/vdev_initialize.h index 81d39ebebcb2..942fc71c52ae 100644 --- a/include/sys/vdev_initialize.h +++ b/include/sys/vdev_initialize.h @@ -33,6 +33,7 @@ extern "C" { #endif extern void vdev_initialize(vdev_t *vd); +extern void vdev_uninitialize(vdev_t *vd); extern void vdev_initialize_stop(vdev_t *vd, vdev_initializing_state_t tgt_state, list_t *vd_list); extern void vdev_initialize_stop_all(vdev_t *vd, diff --git a/include/sys/vdev_raidz_impl.h b/include/sys/vdev_raidz_impl.h index 908723da0c2a..b5c9a4f09ace 100644 --- a/include/sys/vdev_raidz_impl.h +++ b/include/sys/vdev_raidz_impl.h @@ -130,7 +130,7 @@ typedef struct raidz_row { uint64_t rr_offset; /* Logical offset for *_io_verify() */ uint64_t rr_size; /* Physical size for *_io_verify() */ #endif - raidz_col_t rr_col[0]; /* Flexible array of I/O columns */ + raidz_col_t rr_col[]; /* Flexible array of I/O columns */ } raidz_row_t; typedef struct raidz_map { @@ -139,7 +139,7 @@ typedef struct raidz_map { int rm_nskip; /* RAIDZ sectors skipped for padding */ int rm_skipstart; /* Column index of padding start */ const raidz_impl_ops_t *rm_ops; /* RAIDZ math operations */ - raidz_row_t *rm_row[0]; /* flexible array of rows */ + raidz_row_t *rm_row[]; /* flexible array of rows */ } raidz_map_t; diff --git a/include/sys/zfs_context.h b/include/sys/zfs_context.h index 235a73d5d782..74bef8281e57 100644 --- a/include/sys/zfs_context.h +++ b/include/sys/zfs_context.h @@ -57,6 +57,7 @@ extern "C" { #include <sys/disp.h> #include <sys/debug.h> #include <sys/random.h> +#include <sys/string.h> #include <sys/strings.h> #include <sys/byteorder.h> #include <sys/list.h> diff --git a/include/sys/zfs_znode.h b/include/sys/zfs_znode.h index 0df8a0e4b19a..48dab671dd1c 100644 --- a/include/sys/zfs_znode.h +++ b/include/sys/zfs_znode.h @@ -188,7 +188,6 @@ typedef struct znode { boolean_t z_atime_dirty; /* atime needs to be synced */ boolean_t z_zn_prefetch; /* Prefetch znodes? */ boolean_t z_is_sa; /* are we native sa? */ - boolean_t z_is_mapped; /* are we mmap'ed */ boolean_t z_is_ctldir; /* are we .zfs entry */ boolean_t z_suspended; /* extra ref from a suspend? */ uint_t z_blksz; /* block size in bytes */ @@ -198,6 +197,8 @@ typedef struct znode { uint64_t z_size; /* file size (cached) */ uint64_t z_pflags; /* pflags (cached) */ uint32_t z_sync_cnt; /* synchronous open count */ + uint32_t z_sync_writes_cnt; /* synchronous write count */ + uint32_t z_async_writes_cnt; /* asynchronous write count */ mode_t z_mode; /* mode (cached) */ kmutex_t z_acl_lock; /* acl data lock */ zfs_acl_t *z_acl_cached; /* cached acl */ diff --git a/lib/libnvpair/libnvpair.abi b/lib/libnvpair/libnvpair.abi index 4f961c83667d..fa21e485f6ed 100644 --- a/lib/libnvpair/libnvpair.abi +++ b/lib/libnvpair/libnvpair.abi @@ -589,14 +589,6 @@ </function-decl> </abi-instr> <abi-instr address-size='64' path='../../module/nvpair/nvpair.c' language='LANG_C99'> - <typedef-decl name='__u_short' type-id='8efea9e5' id='46c660f8'/> - <typedef-decl name='__u_int' type-id='f0981eeb' id='8ae6822f'/> - <typedef-decl name='__quad_t' type-id='bd54fe1a' id='2632227a'/> - <typedef-decl name='__u_quad_t' type-id='7359adad' id='5f3d50a6'/> - <typedef-decl name='u_short' type-id='46c660f8' id='32580e96'/> - <typedef-decl name='u_int' type-id='8ae6822f' id='48f7c3f5'/> - <typedef-decl name='quad_t' type-id='2632227a' id='f5ef0660'/> - <typedef-decl name='u_quad_t' type-id='5f3d50a6' id='bd226ac0'/> <typedef-decl name='bool_t' type-id='3ff5601b' id='310a70df'/> <enum-decl name='xdr_op' id='6badf1b8'> <underlying-type type-id='9cac1fee'/> @@ -655,6 +647,14 @@ </class-decl> <typedef-decl name='XDR' type-id='755707df' id='bc407f0e'/> <typedef-decl name='xdrproc_t' type-id='94d188f0' id='c28db3e9'/> + <typedef-decl name='__u_short' type-id='8efea9e5' id='46c660f8'/> + <typedef-decl name='__u_int' type-id='f0981eeb' id='8ae6822f'/> + <typedef-decl name='__quad_t' type-id='bd54fe1a' id='2632227a'/> + <typedef-decl name='__u_quad_t' type-id='7359adad' id='5f3d50a6'/> + <typedef-decl name='u_short' type-id='46c660f8' id='32580e96'/> + <typedef-decl name='u_int' type-id='8ae6822f' id='48f7c3f5'/> + <typedef-decl name='quad_t' type-id='2632227a' id='f5ef0660'/> + <typedef-decl name='u_quad_t' type-id='5f3d50a6' id='bd226ac0'/> <pointer-type-def type-id='bc407f0e' size-in-bits='64' id='17fd1621'/> <pointer-type-def type-id='755707df' size-in-bits='64' id='812c6697'/> <qualified-type-def type-id='26a90f95' const='yes' id='57de658a'/> @@ -1771,6 +1771,63 @@ <var-decl name='nvprt_custops' type-id='7be54adb' visibility='default'/> </data-member> </class-decl> + <typedef-decl name='__re_long_size_t' type-id='7359adad' id='ba516949'/> + <typedef-decl name='reg_syntax_t' type-id='7359adad' id='1b72c3b3'/> + <class-decl name='re_pattern_buffer' size-in-bits='512' is-struct='yes' visibility='default' id='19fc9a8c'> + <data-member access='public' layout-offset-in-bits='0'> + <var-decl name='buffer' type-id='33976309' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='64'> + <var-decl name='allocated' type-id='ba516949' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='128'> + <var-decl name='used' type-id='ba516949' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='192'> + <var-decl name='syntax' type-id='1b72c3b3' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='256'> + <var-decl name='fastmap' type-id='26a90f95' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='320'> + <var-decl name='translate' type-id='cf536864' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='384'> + <var-decl name='re_nsub' type-id='b59d7dce' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='448'> + <var-decl name='can_be_null' type-id='f0981eeb' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='449'> + <var-decl name='regs_allocated' type-id='f0981eeb' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='451'> + <var-decl name='fastmap_accurate' type-id='f0981eeb' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='452'> + <var-decl name='no_sub' type-id='f0981eeb' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='453'> + <var-decl name='not_bol' type-id='f0981eeb' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='454'> + <var-decl name='not_eol' type-id='f0981eeb' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='455'> + <var-decl name='newline_anchor' type-id='f0981eeb' visibility='default'/> + </data-member> + </class-decl> + <typedef-decl name='regex_t' type-id='19fc9a8c' id='aca3bac8'/> + <typedef-decl name='regoff_t' type-id='95e97e5e' id='54a2a2a8'/> + <class-decl name='regmatch_t' size-in-bits='64' is-struct='yes' naming-typedef-id='1b941664' visibility='default' id='4f932615'> + <data-member access='public' layout-offset-in-bits='0'> + <var-decl name='rm_so' type-id='54a2a2a8' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='32'> + <var-decl name='rm_eo' type-id='54a2a2a8' visibility='default'/> + </data-member> + </class-decl> + <typedef-decl name='regmatch_t' type-id='4f932615' id='1b941664'/> <typedef-decl name='int8_t' type-id='2171a512' id='ee31ee44'/> <typedef-decl name='int16_t' type-id='03896e23' id='23bd8cb5'/> <typedef-decl name='int32_t' type-id='33f57a65' id='3ff5601b'/> @@ -1880,63 +1937,6 @@ <var-decl name='_unused2' type-id='664ac0b7' visibility='default'/> </data-member> </class-decl> - <typedef-decl name='__re_long_size_t' type-id='7359adad' id='ba516949'/> - <typedef-decl name='reg_syntax_t' type-id='7359adad' id='1b72c3b3'/> - <class-decl name='re_pattern_buffer' size-in-bits='512' is-struct='yes' visibility='default' id='19fc9a8c'> - <data-member access='public' layout-offset-in-bits='0'> - <var-decl name='buffer' type-id='33976309' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='64'> - <var-decl name='allocated' type-id='ba516949' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='128'> - <var-decl name='used' type-id='ba516949' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='192'> - <var-decl name='syntax' type-id='1b72c3b3' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='256'> - <var-decl name='fastmap' type-id='26a90f95' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='320'> - <var-decl name='translate' type-id='cf536864' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='384'> - <var-decl name='re_nsub' type-id='b59d7dce' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='448'> - <var-decl name='can_be_null' type-id='f0981eeb' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='449'> - <var-decl name='regs_allocated' type-id='f0981eeb' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='451'> - <var-decl name='fastmap_accurate' type-id='f0981eeb' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='452'> - <var-decl name='no_sub' type-id='f0981eeb' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='453'> - <var-decl name='not_bol' type-id='f0981eeb' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='454'> - <var-decl name='not_eol' type-id='f0981eeb' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='455'> - <var-decl name='newline_anchor' type-id='f0981eeb' visibility='default'/> - </data-member> - </class-decl> - <typedef-decl name='regex_t' type-id='19fc9a8c' id='aca3bac8'/> - <typedef-decl name='regoff_t' type-id='95e97e5e' id='54a2a2a8'/> - <class-decl name='regmatch_t' size-in-bits='64' is-struct='yes' naming-typedef-id='1b941664' visibility='default' id='4f932615'> - <data-member access='public' layout-offset-in-bits='0'> - <var-decl name='rm_so' type-id='54a2a2a8' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='32'> - <var-decl name='rm_eo' type-id='54a2a2a8' visibility='default'/> - </data-member> - </class-decl> - <typedef-decl name='regmatch_t' type-id='4f932615' id='1b941664'/> <typedef-decl name='size_t' type-id='7359adad' id='b59d7dce'/> <pointer-type-def type-id='aa12d1ba' size-in-bits='64' id='822cd80b'/> <qualified-type-def type-id='822cd80b' restrict='yes' id='e75a27e9'/> @@ -2452,17 +2452,6 @@ <parameter type-id='95e97e5e'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='fprintf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='e75a27e9'/> - <parameter type-id='9d26089a'/> - <parameter is-variadic='yes'/> - <return type-id='95e97e5e'/> - </function-decl> - <function-decl name='printf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='80f4b756'/> - <parameter is-variadic='yes'/> - <return type-id='95e97e5e'/> - </function-decl> <function-decl name='malloc' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='b59d7dce'/> <return type-id='eaa32e2f'/> @@ -2491,6 +2480,19 @@ <parameter type-id='80f4b756'/> <return type-id='b59d7dce'/> </function-decl> + <function-decl name='__fprintf_chk' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='e75a27e9'/> + <parameter type-id='95e97e5e'/> + <parameter type-id='9d26089a'/> + <parameter is-variadic='yes'/> + <return type-id='95e97e5e'/> + </function-decl> + <function-decl name='__printf_chk' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='95e97e5e'/> + <parameter type-id='80f4b756'/> + <parameter is-variadic='yes'/> + <return type-id='95e97e5e'/> + </function-decl> <function-type size-in-bits='64' id='9f88f76e'> <parameter type-id='196db161'/> <parameter type-id='eaa32e2f'/> @@ -3124,14 +3126,15 @@ </abi-instr> <abi-instr address-size='64' path='assert.c' language='LANG_C99'> <var-decl name='aok' type-id='95e97e5e' mangled-name='aok' visibility='default' elf-symbol-id='aok'/> - <function-decl name='vfprintf' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='abort' visibility='default' binding='global' size-in-bits='64'> + <return type-id='48b5725f'/> + </function-decl> + <function-decl name='__vfprintf_chk' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='e75a27e9'/> + <parameter type-id='95e97e5e'/> <parameter type-id='9d26089a'/> <parameter type-id='b7f2d5e6'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='abort' visibility='default' binding='global' size-in-bits='64'> - <return type-id='48b5725f'/> - </function-decl> </abi-instr> </abi-corpus> diff --git a/lib/libnvpair/libnvpair.c b/lib/libnvpair/libnvpair.c index 2e9ea1c174e9..3162fc4d6d3d 100644 --- a/lib/libnvpair/libnvpair.c +++ b/lib/libnvpair/libnvpair.c @@ -200,6 +200,17 @@ nvprint_##type_and_variant(nvlist_prtctl_t pctl, void *private, \ return (1); \ } +/* + * Workaround for GCC 12+ with UBSan enabled deficencies. + * + * GCC 12+ invoked with -fsanitize=undefined incorrectly reports the code + * below as violating -Wformat-overflow. + */ +#if defined(__GNUC__) && !defined(__clang__) && \ + defined(ZFS_UBSAN_ENABLED) && defined(HAVE_FORMAT_OVERFLOW) +#pragma GCC diagnostic push +#pragma GCC diagnostic ignored "-Wformat-overflow" +#endif NVLIST_PRTFUNC(boolean, int, int, "%d") NVLIST_PRTFUNC(boolean_value, boolean_t, int, "%d") NVLIST_PRTFUNC(byte, uchar_t, uchar_t, "0x%2.2x") @@ -214,6 +225,10 @@ NVLIST_PRTFUNC(uint64, uint64_t, u_longlong_t, "0x%llx") NVLIST_PRTFUNC(double, double, double, "0x%f") NVLIST_PRTFUNC(string, char *, char *, "%s") NVLIST_PRTFUNC(hrtime, hrtime_t, hrtime_t, "0x%llx") +#if defined(__GNUC__) && !defined(__clang__) && \ + defined(ZFS_UBSAN_ENABLED) && defined(HAVE_FORMAT_OVERFLOW) +#pragma GCC diagnostic pop +#endif /* * Generate functions to print array-valued nvlist members. diff --git a/lib/libshare/os/linux/nfs.c b/lib/libshare/os/linux/nfs.c index a7bcbd13856c..3f93d9b763bf 100644 --- a/lib/libshare/os/linux/nfs.c +++ b/lib/libshare/os/linux/nfs.c @@ -180,8 +180,9 @@ foreach_nfs_host_cb(const char *opt, const char *value, void *pcookie) { int error; const char *access; - char *host_dup, *host, *next; + char *host_dup, *host, *next, *v6Literal; nfs_host_cookie_t *udata = (nfs_host_cookie_t *)pcookie; + int cidr_len; #ifdef DEBUG fprintf(stderr, "foreach_nfs_host_cb: key=%s, value=%s\n", opt, value); @@ -204,10 +205,46 @@ foreach_nfs_host_cb(const char *opt, const char *value, void *pcookie) host = host_dup; do { - next = strchr(host, ':'); - if (next != NULL) { - *next = '\0'; - next++; + if (*host == '[') { + host++; + v6Literal = strchr(host, ']'); + if (v6Literal == NULL) { + free(host_dup); + return (SA_SYNTAX_ERR); + } + if (v6Literal[1] == '\0') { + *v6Literal = '\0'; + next = NULL; + } else if (v6Literal[1] == '/') { + next = strchr(v6Literal + 2, ':'); + if (next == NULL) { + cidr_len = + strlen(v6Literal + 1); + memmove(v6Literal, + v6Literal + 1, + cidr_len); + v6Literal[cidr_len] = '\0'; + } else { + cidr_len = next - v6Literal - 1; + memmove(v6Literal, + v6Literal + 1, + cidr_len); + v6Literal[cidr_len] = '\0'; + next++; + } + } else if (v6Literal[1] == ':') { + *v6Literal = '\0'; + next = v6Literal + 2; + } else { + free(host_dup); + return (SA_SYNTAX_ERR); + } + } else { + next = strchr(host, ':'); + if (next != NULL) { + *next = '\0'; + next++; + } } error = udata->callback(udata->filename, diff --git a/lib/libuutil/libuutil.abi b/lib/libuutil/libuutil.abi index 21418ec1d4a0..7b1a2cd3af60 100644 --- a/lib/libuutil/libuutil.abi +++ b/lib/libuutil/libuutil.abi @@ -774,16 +774,23 @@ <parameter type-id='b59d7dce'/> <return type-id='eaa32e2f'/> </function-decl> - <function-decl name='mbstowcs' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='access' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='80f4b756'/> + <parameter type-id='95e97e5e'/> + <return type-id='95e97e5e'/> + </function-decl> + <function-decl name='__mbstowcs_chk' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='f1358bc3'/> <parameter type-id='9d26089a'/> <parameter type-id='b59d7dce'/> + <parameter type-id='b59d7dce'/> <return type-id='b59d7dce'/> </function-decl> - <function-decl name='wcstombs' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='__wcstombs_chk' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='266fe297'/> <parameter type-id='598aab80'/> <parameter type-id='b59d7dce'/> + <parameter type-id='b59d7dce'/> <return type-id='b59d7dce'/> </function-decl> <function-decl name='mkdir' visibility='default' binding='global' size-in-bits='64'> @@ -791,17 +798,13 @@ <parameter type-id='e1c52942'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='access' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='80f4b756'/> - <parameter type-id='95e97e5e'/> - <return type-id='95e97e5e'/> - </function-decl> </abi-instr> <abi-instr address-size='64' path='os/linux/getexecname.c' language='LANG_C99'> - <function-decl name='readlink' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='__readlink_chk' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='9d26089a'/> <parameter type-id='266fe297'/> <parameter type-id='b59d7dce'/> + <parameter type-id='b59d7dce'/> <return type-id='79a0948f'/> </function-decl> </abi-instr> @@ -827,10 +830,11 @@ <parameter type-id='95e97e5e'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='read' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='__read_chk' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='95e97e5e'/> <parameter type-id='eaa32e2f'/> <parameter type-id='b59d7dce'/> + <parameter type-id='b59d7dce'/> <return type-id='79a0948f'/> </function-decl> <function-decl name='get_system_hostid' mangled-name='get_system_hostid' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='get_system_hostid'> @@ -879,6 +883,26 @@ <var-decl name='mnt_minor' type-id='3502e3ff' visibility='default'/> </data-member> </class-decl> + <class-decl name='mntent' size-in-bits='320' is-struct='yes' visibility='default' id='56fe4a37'> + <data-member access='public' layout-offset-in-bits='0'> + <var-decl name='mnt_fsname' type-id='26a90f95' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='64'> + <var-decl name='mnt_dir' type-id='26a90f95' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='128'> + <var-decl name='mnt_type' type-id='26a90f95' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='192'> + <var-decl name='mnt_opts' type-id='26a90f95' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='256'> + <var-decl name='mnt_freq' type-id='95e97e5e' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='288'> + <var-decl name='mnt_passno' type-id='95e97e5e' visibility='default'/> + </data-member> + </class-decl> <class-decl name='stat64' size-in-bits='1152' is-struct='yes' visibility='default' id='0bbec9cd'> <data-member access='public' layout-offset-in-bits='0'> <var-decl name='st_dev' type-id='35ed8932' visibility='default'/> @@ -933,26 +957,6 @@ <typedef-decl name='__nlink_t' type-id='7359adad' id='80f0b9df'/> <typedef-decl name='__blksize_t' type-id='bd54fe1a' id='d3f10a7f'/> <typedef-decl name='__blkcnt64_t' type-id='bd54fe1a' id='4e711bf1'/> - <class-decl name='mntent' size-in-bits='320' is-struct='yes' visibility='default' id='56fe4a37'> - <data-member access='public' layout-offset-in-bits='0'> - <var-decl name='mnt_fsname' type-id='26a90f95' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='64'> - <var-decl name='mnt_dir' type-id='26a90f95' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='128'> - <var-decl name='mnt_type' type-id='26a90f95' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='192'> - <var-decl name='mnt_opts' type-id='26a90f95' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='256'> - <var-decl name='mnt_freq' type-id='95e97e5e' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='288'> - <var-decl name='mnt_passno' type-id='95e97e5e' visibility='default'/> - </data-member> - </class-decl> <pointer-type-def type-id='0c544dc0' size-in-bits='64' id='394fc496'/> <pointer-type-def type-id='56fe4a37' size-in-bits='64' id='b6b61d2f'/> <qualified-type-def type-id='b6b61d2f' restrict='yes' id='3cad23cd'/> @@ -1019,6 +1023,7 @@ </function-decl> </abi-instr> <abi-instr address-size='64' path='timestamp.c' language='LANG_C99'> + <typedef-decl name='nl_item' type-id='95e97e5e' id='03b79a94'/> <class-decl name='tm' size-in-bits='448' is-struct='yes' visibility='default' id='dddf6ca2'> <data-member access='public' layout-offset-in-bits='0'> <var-decl name='tm_sec' type-id='95e97e5e' visibility='default'/> @@ -1055,7 +1060,6 @@ </data-member> </class-decl> <typedef-decl name='time_t' type-id='65eda9c0' id='c9d12d66'/> - <typedef-decl name='nl_item' type-id='95e97e5e' id='03b79a94'/> <qualified-type-def type-id='c9d12d66' const='yes' id='588b3216'/> <pointer-type-def type-id='588b3216' size-in-bits='64' id='9f201474'/> <qualified-type-def type-id='dddf6ca2' const='yes' id='e824a34f'/> @@ -1071,11 +1075,6 @@ <parameter type-id='03b79a94'/> <return type-id='26a90f95'/> </function-decl> - <function-decl name='printf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='80f4b756'/> - <parameter is-variadic='yes'/> - <return type-id='95e97e5e'/> - </function-decl> <function-decl name='time' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='b2eb2c3f'/> <return type-id='c9d12d66'/> @@ -1091,31 +1090,21 @@ <parameter type-id='9f201474'/> <return type-id='d915a820'/> </function-decl> + <function-decl name='__printf_chk' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='95e97e5e'/> + <parameter type-id='80f4b756'/> + <parameter is-variadic='yes'/> + <return type-id='95e97e5e'/> + </function-decl> </abi-instr> <abi-instr address-size='64' path='uu_alloc.c' language='LANG_C99'> <type-decl name='char' size-in-bits='8' id='a84c031d'/> - <class-decl name='__va_list_tag' size-in-bits='192' is-struct='yes' visibility='default' id='d5027220'> - <data-member access='public' layout-offset-in-bits='0'> - <var-decl name='gp_offset' type-id='f0981eeb' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='32'> - <var-decl name='fp_offset' type-id='f0981eeb' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='64'> - <var-decl name='overflow_arg_area' type-id='eaa32e2f' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='128'> - <var-decl name='reg_save_area' type-id='eaa32e2f' visibility='default'/> - </data-member> - </class-decl> - <type-decl name='int' size-in-bits='32' id='95e97e5e'/> <type-decl name='unsigned int' size-in-bits='32' id='f0981eeb'/> <type-decl name='unsigned long int' size-in-bits='64' id='7359adad'/> <type-decl name='variadic parameter type' id='2c1145c5'/> <type-decl name='void' id='48b5725f'/> <typedef-decl name='uint_t' type-id='f0981eeb' id='3502e3ff'/> <typedef-decl name='size_t' type-id='7359adad' id='b59d7dce'/> - <pointer-type-def type-id='d5027220' size-in-bits='64' id='b7f2d5e6'/> <pointer-type-def type-id='a84c031d' size-in-bits='64' id='26a90f95'/> <qualified-type-def type-id='a84c031d' const='yes' id='9b45d938'/> <pointer-type-def type-id='9b45d938' size-in-bits='64' id='80f4b756'/> @@ -1151,13 +1140,6 @@ <parameter is-variadic='yes'/> <return type-id='26a90f95'/> </function-decl> - <function-decl name='vsnprintf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='26a90f95'/> - <parameter type-id='b59d7dce'/> - <parameter type-id='80f4b756'/> - <parameter type-id='b7f2d5e6'/> - <return type-id='95e97e5e'/> - </function-decl> <function-decl name='malloc' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='b59d7dce'/> <return type-id='eaa32e2f'/> @@ -1166,18 +1148,6 @@ <parameter type-id='eaa32e2f'/> <return type-id='48b5725f'/> </function-decl> - <function-decl name='memcpy' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='eaa32e2f'/> - <parameter type-id='eaa32e2f'/> - <parameter type-id='b59d7dce'/> - <return type-id='eaa32e2f'/> - </function-decl> - <function-decl name='memset' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='eaa32e2f'/> - <parameter type-id='95e97e5e'/> - <parameter type-id='b59d7dce'/> - <return type-id='eaa32e2f'/> - </function-decl> <function-decl name='strlen' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='80f4b756'/> <return type-id='b59d7dce'/> @@ -1201,6 +1171,7 @@ <array-type-def dimensions='1' type-id='a84c031d' size-in-bits='512' id='59daf3ef'> <subrange length='64' type-id='7359adad' id='b10be967'/> </array-type-def> + <type-decl name='int' size-in-bits='32' id='95e97e5e'/> <type-decl name='long int' size-in-bits='64' id='bd54fe1a'/> <type-decl name='short int' size-in-bits='16' id='a2185560'/> <type-decl name='signed char' size-in-bits='8' id='28577a57'/> @@ -1326,6 +1297,7 @@ </data-member> </class-decl> <typedef-decl name='ulong_t' type-id='7359adad' id='ee1f298e'/> + <typedef-decl name='uintptr_t' type-id='7359adad' id='e475ab95'/> <union-decl name='pthread_mutexattr_t' size-in-bits='32' naming-typedef-id='8afd6070' visibility='default' id='7300eb00'> <data-member access='public'> <var-decl name='__size' type-id='8e0573fd' visibility='default'/> @@ -1388,7 +1360,6 @@ <typedef-decl name='__int8_t' type-id='28577a57' id='2171a512'/> <typedef-decl name='__uint8_t' type-id='002ac4a6' id='c51d6389'/> <typedef-decl name='__uint32_t' type-id='f0981eeb' id='62f1140c'/> - <typedef-decl name='uintptr_t' type-id='7359adad' id='e475ab95'/> <pointer-type-def type-id='0e01899c' size-in-bits='64' id='4d98cd5a'/> <pointer-type-def type-id='fba6cb51' size-in-bits='64' id='32adbf30'/> <pointer-type-def type-id='428b67b3' size-in-bits='64' id='bf311473'/> @@ -1633,6 +1604,20 @@ <class-decl name='_IO_codecvt' is-struct='yes' visibility='default' is-declaration-only='yes' id='a4036571'/> <class-decl name='_IO_marker' is-struct='yes' visibility='default' is-declaration-only='yes' id='010ae0b9'/> <class-decl name='_IO_wide_data' is-struct='yes' visibility='default' is-declaration-only='yes' id='79bd3751'/> + <class-decl name='__va_list_tag' size-in-bits='192' is-struct='yes' visibility='default' id='d5027220'> + <data-member access='public' layout-offset-in-bits='0'> + <var-decl name='gp_offset' type-id='f0981eeb' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='32'> + <var-decl name='fp_offset' type-id='f0981eeb' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='64'> + <var-decl name='overflow_arg_area' type-id='eaa32e2f' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='128'> + <var-decl name='reg_save_area' type-id='eaa32e2f' visibility='default'/> + </data-member> + </class-decl> <type-decl name='unnamed-enum-underlying-type-32' is-anonymous='yes' size-in-bits='32' alignment-in-bits='32' id='9cac1fee'/> <type-decl name='unsigned short int' size-in-bits='16' id='8efea9e5'/> <typedef-decl name='uu_dprintf_t' type-id='0538fe4f' id='2367d595'/> @@ -1757,6 +1742,7 @@ <pointer-type-def type-id='bb4788fa' size-in-bits='64' id='cecf4ea7'/> <pointer-type-def type-id='010ae0b9' size-in-bits='64' id='e4c6fa61'/> <pointer-type-def type-id='79bd3751' size-in-bits='64' id='c65a1f29'/> + <pointer-type-def type-id='d5027220' size-in-bits='64' id='b7f2d5e6'/> <qualified-type-def type-id='80f4b756' restrict='yes' id='9d26089a'/> <pointer-type-def type-id='2367d595' size-in-bits='64' id='ed73b5ca'/> <class-decl name='_IO_codecvt' is-struct='yes' visibility='default' is-declaration-only='yes' id='a4036571'/> @@ -1789,22 +1775,24 @@ <parameter type-id='95e97e5e'/> <return type-id='26a90f95'/> </function-decl> - <function-decl name='fprintf' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='strdup' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='80f4b756'/> + <return type-id='26a90f95'/> + </function-decl> + <function-decl name='__fprintf_chk' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='e75a27e9'/> + <parameter type-id='95e97e5e'/> <parameter type-id='9d26089a'/> <parameter is-variadic='yes'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='vfprintf' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='__vfprintf_chk' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='e75a27e9'/> + <parameter type-id='95e97e5e'/> <parameter type-id='9d26089a'/> <parameter type-id='b7f2d5e6'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='strdup' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='80f4b756'/> - <return type-id='26a90f95'/> - </function-decl> </abi-instr> <abi-instr address-size='64' path='uu_ident.c' language='LANG_C99'> <function-decl name='strchr' visibility='default' binding='global' size-in-bits='64'> @@ -2144,13 +2132,6 @@ <parameter type-id='3502e3ff' name='uflags'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='snprintf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='26a90f95'/> - <parameter type-id='b59d7dce'/> - <parameter type-id='80f4b756'/> - <parameter is-variadic='yes'/> - <return type-id='95e97e5e'/> - </function-decl> <function-decl name='clock_gettime' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='a1c3b834'/> <parameter type-id='3d83ba87'/> @@ -2160,6 +2141,12 @@ <parameter type-id='80f4b756'/> <return type-id='95e97e5e'/> </function-decl> + <function-decl name='__open_too_many_args' visibility='default' binding='global' size-in-bits='64'> + <return type-id='48b5725f'/> + </function-decl> + <function-decl name='__open_missing_mode' visibility='default' binding='global' size-in-bits='64'> + <return type-id='48b5725f'/> + </function-decl> </abi-instr> <abi-instr address-size='64' path='uu_pname.c' language='LANG_C99'> <function-decl name='getexecname' mangled-name='getexecname' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='getexecname'> diff --git a/lib/libzfs/libzfs.abi b/lib/libzfs/libzfs.abi index 13ce19df9b34..7d5f07d8fa55 100644 --- a/lib/libzfs/libzfs.abi +++ b/lib/libzfs/libzfs.abi @@ -1,14 +1,10 @@ <abi-corpus version='2.0' architecture='elf-amd-x86_64' soname='libzfs.so.4'> <elf-needed> <dependency name='libzfs_core.so.3'/> - <dependency name='libuuid.so.1'/> - <dependency name='libblkid.so.1'/> - <dependency name='libudev.so.1'/> <dependency name='libnvpair.so.3'/> - <dependency name='libtirpc.so.3'/> <dependency name='libuutil.so.3'/> <dependency name='libm.so.6'/> - <dependency name='libcrypto.so.1.1'/> + <dependency name='libcrypto.so.3'/> <dependency name='libz.so.1'/> <dependency name='libc.so.6'/> </elf-needed> @@ -333,6 +329,8 @@ <elf-symbol name='zpool_open_canfail' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zpool_open_silent' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zpool_pool_state_to_name' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> + <elf-symbol name='zpool_prepare_and_label_disk' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> + <elf-symbol name='zpool_prepare_disk' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zpool_print_unsup_feat' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zpool_prop_align_right' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zpool_prop_column_name' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> @@ -380,6 +378,9 @@ <elf-symbol name='zpool_vdev_remove' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zpool_vdev_remove_cancel' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zpool_vdev_remove_wanted' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> + <elf-symbol name='zpool_vdev_script_alloc_env' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> + <elf-symbol name='zpool_vdev_script_free_env' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> + <elf-symbol name='zpool_vdev_set_removed_state' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zpool_vdev_split' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zpool_wait' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zpool_wait_status' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> @@ -538,11 +539,6 @@ <parameter type-id='80f4b756'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='fputs' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='9d26089a'/> - <parameter type-id='e75a27e9'/> - <return type-id='95e97e5e'/> - </function-decl> <function-decl name='flock' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='95e97e5e'/> <parameter type-id='95e97e5e'/> @@ -601,8 +597,9 @@ <parameter type-id='80f4b756'/> <return type-id='f09217ba'/> </function-decl> - <function-decl name='fgets' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='__fgets_chk' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='266fe297'/> + <parameter type-id='b59d7dce'/> <parameter type-id='95e97e5e'/> <parameter type-id='e75a27e9'/> <return type-id='26a90f95'/> @@ -1762,6 +1759,54 @@ <var-decl name='cl_haszonedchild' type-id='c19b74c3' visibility='default'/> </data-member> </class-decl> + <typedef-decl name='__re_long_size_t' type-id='7359adad' id='ba516949'/> + <typedef-decl name='reg_syntax_t' type-id='7359adad' id='1b72c3b3'/> + <class-decl name='re_pattern_buffer' size-in-bits='512' is-struct='yes' visibility='default' id='19fc9a8c'> + <data-member access='public' layout-offset-in-bits='0'> + <var-decl name='buffer' type-id='33976309' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='64'> + <var-decl name='allocated' type-id='ba516949' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='128'> + <var-decl name='used' type-id='ba516949' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='192'> + <var-decl name='syntax' type-id='1b72c3b3' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='256'> + <var-decl name='fastmap' type-id='26a90f95' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='320'> + <var-decl name='translate' type-id='cf536864' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='384'> + <var-decl name='re_nsub' type-id='b59d7dce' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='448'> + <var-decl name='can_be_null' type-id='f0981eeb' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='449'> + <var-decl name='regs_allocated' type-id='f0981eeb' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='451'> + <var-decl name='fastmap_accurate' type-id='f0981eeb' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='452'> + <var-decl name='no_sub' type-id='f0981eeb' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='453'> + <var-decl name='not_bol' type-id='f0981eeb' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='454'> + <var-decl name='not_eol' type-id='f0981eeb' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='455'> + <var-decl name='newline_anchor' type-id='f0981eeb' visibility='default'/> + </data-member> + </class-decl> + <typedef-decl name='regex_t' type-id='19fc9a8c' id='aca3bac8'/> + <typedef-decl name='uintptr_t' type-id='7359adad' id='e475ab95'/> <union-decl name='pthread_mutex_t' size-in-bits='320' naming-typedef-id='7a6844eb' visibility='default' id='70681f9b'> <data-member access='public'> <var-decl name='__data' type-id='4c734837' visibility='default'/> @@ -1910,54 +1955,6 @@ <var-decl name='_unused2' type-id='664ac0b7' visibility='default'/> </data-member> </class-decl> - <typedef-decl name='__re_long_size_t' type-id='7359adad' id='ba516949'/> - <typedef-decl name='reg_syntax_t' type-id='7359adad' id='1b72c3b3'/> - <class-decl name='re_pattern_buffer' size-in-bits='512' is-struct='yes' visibility='default' id='19fc9a8c'> - <data-member access='public' layout-offset-in-bits='0'> - <var-decl name='buffer' type-id='33976309' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='64'> - <var-decl name='allocated' type-id='ba516949' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='128'> - <var-decl name='used' type-id='ba516949' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='192'> - <var-decl name='syntax' type-id='1b72c3b3' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='256'> - <var-decl name='fastmap' type-id='26a90f95' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='320'> - <var-decl name='translate' type-id='cf536864' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='384'> - <var-decl name='re_nsub' type-id='b59d7dce' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='448'> - <var-decl name='can_be_null' type-id='f0981eeb' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='449'> - <var-decl name='regs_allocated' type-id='f0981eeb' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='451'> - <var-decl name='fastmap_accurate' type-id='f0981eeb' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='452'> - <var-decl name='no_sub' type-id='f0981eeb' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='453'> - <var-decl name='not_bol' type-id='f0981eeb' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='454'> - <var-decl name='not_eol' type-id='f0981eeb' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='455'> - <var-decl name='newline_anchor' type-id='f0981eeb' visibility='default'/> - </data-member> - </class-decl> - <typedef-decl name='regex_t' type-id='19fc9a8c' id='aca3bac8'/> - <typedef-decl name='uintptr_t' type-id='7359adad' id='e475ab95'/> <typedef-decl name='size_t' type-id='7359adad' id='b59d7dce'/> <pointer-type-def type-id='aa12d1ba' size-in-bits='64' id='822cd80b'/> <pointer-type-def type-id='ec1ed955' size-in-bits='64' id='dca988a5'/> @@ -2736,11 +2733,6 @@ <parameter type-id='80f4b756'/> <return type-id='26a90f95'/> </function-decl> - <function-decl name='strcpy' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='26a90f95'/> - <parameter type-id='80f4b756'/> - <return type-id='26a90f95'/> - </function-decl> <function-decl name='strchr' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='80f4b756'/> <parameter type-id='95e97e5e'/> @@ -2802,6 +2794,18 @@ </enum-decl> <typedef-decl name='zpool_prop_t' type-id='af1ba157' id='5d0c23fb'/> <typedef-decl name='uint_t' type-id='f0981eeb' id='3502e3ff'/> + <typedef-decl name='regoff_t' type-id='95e97e5e' id='54a2a2a8'/> + <class-decl name='regmatch_t' size-in-bits='64' is-struct='yes' naming-typedef-id='1b941664' visibility='default' id='4f932615'> + <data-member access='public' layout-offset-in-bits='0'> + <var-decl name='rm_so' type-id='54a2a2a8' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='32'> + <var-decl name='rm_eo' type-id='54a2a2a8' visibility='default'/> + </data-member> + </class-decl> + <typedef-decl name='regmatch_t' type-id='4f932615' id='1b941664'/> + <typedef-decl name='__sighandler_t' type-id='03347643' id='8cdd9566'/> + <typedef-decl name='ssize_t' type-id='41060289' id='79a0948f'/> <class-decl name='sigaction' size-in-bits='1216' is-struct='yes' visibility='default' id='fe391c48'> <data-member access='public' layout-offset-in-bits='0'> <var-decl name='__sigaction_handler' type-id='ac5ab598' visibility='default'/> @@ -3010,18 +3014,6 @@ </class-decl> <typedef-decl name='siginfo_t' type-id='d8149419' id='cb681f62'/> <typedef-decl name='sigset_t' type-id='b9c97942' id='daf33c64'/> - <typedef-decl name='regoff_t' type-id='95e97e5e' id='54a2a2a8'/> - <class-decl name='regmatch_t' size-in-bits='64' is-struct='yes' naming-typedef-id='1b941664' visibility='default' id='4f932615'> - <data-member access='public' layout-offset-in-bits='0'> - <var-decl name='rm_so' type-id='54a2a2a8' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='32'> - <var-decl name='rm_eo' type-id='54a2a2a8' visibility='default'/> - </data-member> - </class-decl> - <typedef-decl name='regmatch_t' type-id='4f932615' id='1b941664'/> - <typedef-decl name='__sighandler_t' type-id='03347643' id='8cdd9566'/> - <typedef-decl name='ssize_t' type-id='41060289' id='79a0948f'/> <qualified-type-def type-id='822cd80b' restrict='yes' id='e75a27e9'/> <qualified-type-def type-id='9b23c9ad' restrict='yes' id='8c85230f'/> <qualified-type-def type-id='80f4b756' restrict='yes' id='9d26089a'/> @@ -3253,24 +3245,6 @@ <parameter type-id='80f4b756'/> <return type-id='822cd80b'/> </function-decl> - <function-decl name='printf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='80f4b756'/> - <parameter is-variadic='yes'/> - <return type-id='95e97e5e'/> - </function-decl> - <function-decl name='snprintf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='26a90f95'/> - <parameter type-id='b59d7dce'/> - <parameter type-id='80f4b756'/> - <parameter is-variadic='yes'/> - <return type-id='95e97e5e'/> - </function-decl> - <function-decl name='asprintf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='8c85230f'/> - <parameter type-id='9d26089a'/> - <parameter is-variadic='yes'/> - <return type-id='95e97e5e'/> - </function-decl> <function-decl name='fputc' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='95e97e5e'/> <parameter type-id='822cd80b'/> @@ -3283,13 +3257,6 @@ <parameter type-id='e75a27e9'/> <return type-id='41060289'/> </function-decl> - <function-decl name='fread' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='1b7446cd'/> - <parameter type-id='b59d7dce'/> - <parameter type-id='b59d7dce'/> - <parameter type-id='e75a27e9'/> - <return type-id='b59d7dce'/> - </function-decl> <function-decl name='rewind' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='822cd80b'/> <return type-id='48b5725f'/> @@ -3311,12 +3278,6 @@ <parameter type-id='b59d7dce'/> <return type-id='eaa32e2f'/> </function-decl> - <function-decl name='memcpy' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='eaa32e2f'/> - <parameter type-id='eaa32e2f'/> - <parameter type-id='b59d7dce'/> - <return type-id='eaa32e2f'/> - </function-decl> <function-decl name='strdup' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='80f4b756'/> <return type-id='26a90f95'/> @@ -3340,12 +3301,6 @@ <parameter type-id='95e97e5e'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='read' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='95e97e5e'/> - <parameter type-id='eaa32e2f'/> - <parameter type-id='b59d7dce'/> - <return type-id='79a0948f'/> - </function-decl> <function-decl name='getpid' visibility='default' binding='global' size-in-bits='64'> <return type-id='3629bad8'/> </function-decl> @@ -3357,6 +3312,40 @@ <parameter type-id='80f4b756'/> <return type-id='95e97e5e'/> </function-decl> + <function-decl name='__open_too_many_args' visibility='default' binding='global' size-in-bits='64'> + <return type-id='48b5725f'/> + </function-decl> + <function-decl name='__open_missing_mode' visibility='default' binding='global' size-in-bits='64'> + <return type-id='48b5725f'/> + </function-decl> + <function-decl name='__printf_chk' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='95e97e5e'/> + <parameter type-id='80f4b756'/> + <parameter is-variadic='yes'/> + <return type-id='95e97e5e'/> + </function-decl> + <function-decl name='__asprintf_chk' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='8c85230f'/> + <parameter type-id='95e97e5e'/> + <parameter type-id='9d26089a'/> + <parameter is-variadic='yes'/> + <return type-id='95e97e5e'/> + </function-decl> + <function-decl name='__fread_chk' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='1b7446cd'/> + <parameter type-id='b59d7dce'/> + <parameter type-id='b59d7dce'/> + <parameter type-id='b59d7dce'/> + <parameter type-id='e75a27e9'/> + <return type-id='b59d7dce'/> + </function-decl> + <function-decl name='__read_chk' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='95e97e5e'/> + <parameter type-id='eaa32e2f'/> + <parameter type-id='b59d7dce'/> + <parameter type-id='b59d7dce'/> + <return type-id='79a0948f'/> + </function-decl> <function-type size-in-bits='64' id='ee076206'> <return type-id='48b5725f'/> </function-type> @@ -3401,13 +3390,13 @@ <typedef-decl name='zprop_list_t' type-id='bd9b4291' id='bdb8ac4f'/> <class-decl name='renameflags' size-in-bits='32' is-struct='yes' visibility='default' id='7aee5792'> <data-member access='public' layout-offset-in-bits='0'> - <var-decl name='recursive' type-id='95e97e5e' visibility='default'/> + <var-decl name='recursive' type-id='f0981eeb' visibility='default'/> </data-member> <data-member access='public' layout-offset-in-bits='1'> - <var-decl name='nounmount' type-id='95e97e5e' visibility='default'/> + <var-decl name='nounmount' type-id='f0981eeb' visibility='default'/> </data-member> <data-member access='public' layout-offset-in-bits='2'> - <var-decl name='forceunmount' type-id='95e97e5e' visibility='default'/> + <var-decl name='forceunmount' type-id='f0981eeb' visibility='default'/> </data-member> </class-decl> <typedef-decl name='renameflags_t' type-id='7aee5792' id='067170c2'/> @@ -3479,55 +3468,6 @@ <var-decl name='mnt_mntopts' type-id='26a90f95' visibility='default'/> </data-member> </class-decl> - <union-decl name='pthread_mutexattr_t' size-in-bits='32' naming-typedef-id='8afd6070' visibility='default' id='7300eb00'> - <data-member access='public'> - <var-decl name='__size' type-id='8e0573fd' visibility='default'/> - </data-member> - <data-member access='public'> - <var-decl name='__align' type-id='95e97e5e' visibility='default'/> - </data-member> - </union-decl> - <typedef-decl name='pthread_mutexattr_t' type-id='7300eb00' id='8afd6070'/> - <typedef-decl name='int64_t' type-id='0c9942d2' id='9da381c4'/> - <typedef-decl name='__int64_t' type-id='bd54fe1a' id='0c9942d2'/> - <typedef-decl name='__gid_t' type-id='f0981eeb' id='d94ec6d9'/> - <typedef-decl name='__time_t' type-id='bd54fe1a' id='65eda9c0'/> - <class-decl name='tm' size-in-bits='448' is-struct='yes' visibility='default' id='dddf6ca2'> - <data-member access='public' layout-offset-in-bits='0'> - <var-decl name='tm_sec' type-id='95e97e5e' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='32'> - <var-decl name='tm_min' type-id='95e97e5e' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='64'> - <var-decl name='tm_hour' type-id='95e97e5e' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='96'> - <var-decl name='tm_mday' type-id='95e97e5e' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='128'> - <var-decl name='tm_mon' type-id='95e97e5e' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='160'> - <var-decl name='tm_year' type-id='95e97e5e' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='192'> - <var-decl name='tm_wday' type-id='95e97e5e' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='224'> - <var-decl name='tm_yday' type-id='95e97e5e' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='256'> - <var-decl name='tm_isdst' type-id='95e97e5e' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='320'> - <var-decl name='tm_gmtoff' type-id='bd54fe1a' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='384'> - <var-decl name='tm_zone' type-id='80f4b756' visibility='default'/> - </data-member> - </class-decl> - <typedef-decl name='time_t' type-id='65eda9c0' id='c9d12d66'/> <class-decl name='group' size-in-bits='256' is-struct='yes' visibility='default' id='01a1b934'> <data-member access='public' layout-offset-in-bits='0'> <var-decl name='gr_name' type-id='26a90f95' visibility='default'/> @@ -3585,6 +3525,55 @@ <var-decl name='pw_shell' type-id='26a90f95' visibility='default'/> </data-member> </class-decl> + <union-decl name='pthread_mutexattr_t' size-in-bits='32' naming-typedef-id='8afd6070' visibility='default' id='7300eb00'> + <data-member access='public'> + <var-decl name='__size' type-id='8e0573fd' visibility='default'/> + </data-member> + <data-member access='public'> + <var-decl name='__align' type-id='95e97e5e' visibility='default'/> + </data-member> + </union-decl> + <typedef-decl name='pthread_mutexattr_t' type-id='7300eb00' id='8afd6070'/> + <typedef-decl name='int64_t' type-id='0c9942d2' id='9da381c4'/> + <typedef-decl name='__int64_t' type-id='bd54fe1a' id='0c9942d2'/> + <typedef-decl name='__gid_t' type-id='f0981eeb' id='d94ec6d9'/> + <typedef-decl name='__time_t' type-id='bd54fe1a' id='65eda9c0'/> + <class-decl name='tm' size-in-bits='448' is-struct='yes' visibility='default' id='dddf6ca2'> + <data-member access='public' layout-offset-in-bits='0'> + <var-decl name='tm_sec' type-id='95e97e5e' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='32'> + <var-decl name='tm_min' type-id='95e97e5e' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='64'> + <var-decl name='tm_hour' type-id='95e97e5e' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='96'> + <var-decl name='tm_mday' type-id='95e97e5e' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='128'> + <var-decl name='tm_mon' type-id='95e97e5e' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='160'> + <var-decl name='tm_year' type-id='95e97e5e' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='192'> + <var-decl name='tm_wday' type-id='95e97e5e' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='224'> + <var-decl name='tm_yday' type-id='95e97e5e' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='256'> + <var-decl name='tm_isdst' type-id='95e97e5e' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='320'> + <var-decl name='tm_gmtoff' type-id='bd54fe1a' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='384'> + <var-decl name='tm_zone' type-id='80f4b756' visibility='default'/> + </data-member> + </class-decl> + <typedef-decl name='time_t' type-id='65eda9c0' id='c9d12d66'/> <typedef-decl name='uid_t' type-id='cc5fcceb' id='354978ed'/> <pointer-type-def type-id='fba6cb51' size-in-bits='64' id='32adbf30'/> <pointer-type-def type-id='f20fbd51' size-in-bits='64' id='a3681dea'/> @@ -4459,12 +4448,6 @@ <parameter type-id='80f4b756'/> <return type-id='a195f4a3'/> </function-decl> - <function-decl name='fprintf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='e75a27e9'/> - <parameter type-id='9d26089a'/> - <parameter is-variadic='yes'/> - <return type-id='95e97e5e'/> - </function-decl> <function-decl name='strtol' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='9d26089a'/> <parameter type-id='8c85230f'/> @@ -4480,12 +4463,6 @@ <function-decl name='abort' visibility='default' binding='global' size-in-bits='64'> <return type-id='48b5725f'/> </function-decl> - <function-decl name='strncpy' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='26a90f95'/> - <parameter type-id='80f4b756'/> - <parameter type-id='b59d7dce'/> - <return type-id='26a90f95'/> - </function-decl> <function-decl name='strrchr' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='80f4b756'/> <parameter type-id='95e97e5e'/> @@ -4506,12 +4483,6 @@ <parameter type-id='9d26089a'/> <return type-id='26a90f95'/> </function-decl> - <function-decl name='ioctl' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='95e97e5e'/> - <parameter type-id='7359adad'/> - <parameter is-variadic='yes'/> - <return type-id='95e97e5e'/> - </function-decl> <function-decl name='strftime' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='266fe297'/> <parameter type-id='b59d7dce'/> @@ -4524,6 +4495,19 @@ <parameter type-id='f099ad08'/> <return type-id='d915a820'/> </function-decl> + <function-decl name='__fprintf_chk' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='e75a27e9'/> + <parameter type-id='95e97e5e'/> + <parameter type-id='9d26089a'/> + <parameter is-variadic='yes'/> + <return type-id='95e97e5e'/> + </function-decl> + <function-decl name='ioctl' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='95e97e5e'/> + <parameter type-id='7359adad'/> + <parameter is-variadic='yes'/> + <return type-id='95e97e5e'/> + </function-decl> <function-type size-in-bits='64' id='7e291ce6'> <parameter type-id='eaa32e2f'/> <parameter type-id='80f4b756'/> @@ -4536,7 +4520,7 @@ <array-type-def dimensions='1' type-id='a84c031d' size-in-bits='448' id='6093ff7c'> <subrange length='56' type-id='7359adad' id='f8137894'/> </array-type-def> - <class-decl name='differ_info' size-in-bits='9024' is-struct='yes' visibility='default' id='d41965ee'> + <class-decl name='differ_info' size-in-bits='9088' is-struct='yes' visibility='default' id='d41965ee'> <data-member access='public' layout-offset-in-bits='0'> <var-decl name='zhp' type-id='9200a744' visibility='default'/> </data-member> @@ -4577,18 +4561,21 @@ <var-decl name='timestamped' type-id='c19b74c3' visibility='default'/> </data-member> <data-member access='public' layout-offset-in-bits='8832'> - <var-decl name='shares' type-id='9c313c2d' visibility='default'/> + <var-decl name='no_mangle' type-id='c19b74c3' visibility='default'/> </data-member> <data-member access='public' layout-offset-in-bits='8896'> + <var-decl name='shares' type-id='9c313c2d' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='8960'> <var-decl name='zerr' type-id='95e97e5e' visibility='default'/> </data-member> - <data-member access='public' layout-offset-in-bits='8928'> + <data-member access='public' layout-offset-in-bits='8992'> <var-decl name='cleanupfd' type-id='95e97e5e' visibility='default'/> </data-member> - <data-member access='public' layout-offset-in-bits='8960'> + <data-member access='public' layout-offset-in-bits='9024'> <var-decl name='outputfd' type-id='95e97e5e' visibility='default'/> </data-member> - <data-member access='public' layout-offset-in-bits='8992'> + <data-member access='public' layout-offset-in-bits='9056'> <var-decl name='datafd' type-id='95e97e5e' visibility='default'/> </data-member> </class-decl> @@ -4626,6 +4613,13 @@ <parameter type-id='ee78f675'/> <return type-id='95e97e5e'/> </function-decl> + <function-decl name='color_start' mangled-name='color_start' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='color_start'> + <parameter type-id='80f4b756'/> + <return type-id='48b5725f'/> + </function-decl> + <function-decl name='color_end' mangled-name='color_end' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='color_end'> + <return type-id='48b5725f'/> + </function-decl> <function-decl name='zfs_show_diffs' mangled-name='zfs_show_diffs' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zfs_show_diffs'> <parameter type-id='9200a744' name='zhp'/> <parameter type-id='95e97e5e' name='outfd'/> @@ -4650,6 +4644,11 @@ <parameter type-id='4051f5e7'/> <return type-id='95e97e5e'/> </function-decl> + <function-decl name='fputs' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='9d26089a'/> + <parameter type-id='e75a27e9'/> + <return type-id='95e97e5e'/> + </function-decl> <function-decl name='pipe2' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='7292109c'/> <parameter type-id='95e97e5e'/> @@ -4779,31 +4778,26 @@ <parameter type-id='37e3bd22' name='inuse'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='memset' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='eaa32e2f'/> - <parameter type-id='95e97e5e'/> - <parameter type-id='b59d7dce'/> - <return type-id='eaa32e2f'/> - </function-decl> - <function-decl name='fstat64' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='95e97e5e'/> - <parameter type-id='62f7a03d'/> - <return type-id='95e97e5e'/> - </function-decl> - <function-decl name='pread64' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='pwrite64' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='95e97e5e'/> <parameter type-id='eaa32e2f'/> <parameter type-id='b59d7dce'/> <parameter type-id='724e4de6'/> <return type-id='79a0948f'/> </function-decl> - <function-decl name='pwrite64' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='__pread64_chk' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='95e97e5e'/> <parameter type-id='eaa32e2f'/> <parameter type-id='b59d7dce'/> <parameter type-id='724e4de6'/> + <parameter type-id='b59d7dce'/> <return type-id='79a0948f'/> </function-decl> + <function-decl name='fstat64' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='95e97e5e'/> + <parameter type-id='62f7a03d'/> + <return type-id='95e97e5e'/> + </function-decl> <function-type size-in-bits='64' id='baa42fef'> <parameter type-id='eaa32e2f'/> <parameter type-id='80f4b756'/> @@ -4902,6 +4896,9 @@ </class-decl> <typedef-decl name='proto_table_t' type-id='f4c8e1ed' id='f1bd64e2'/> <typedef-decl name='tpool_t' type-id='88d1b7f9' id='b1bbf10d'/> + <typedef-decl name='DIR' type-id='20cd73f2' id='54a5d683'/> + <typedef-decl name='mode_t' type-id='e1c52942' id='d50d396c'/> + <typedef-decl name='__compar_fn_t' type-id='585e1de9' id='aba7edd8'/> <class-decl name='dirent64' size-in-bits='2240' is-struct='yes' visibility='default' id='5725d813'> <data-member access='public' layout-offset-in-bits='0'> <var-decl name='d_ino' type-id='71288a47' visibility='default'/> @@ -5015,9 +5012,6 @@ <typedef-decl name='__fsblkcnt64_t' type-id='7359adad' id='95fe1a02'/> <typedef-decl name='__fsfilcnt64_t' type-id='7359adad' id='0c3a4dde'/> <typedef-decl name='__fsword_t' type-id='bd54fe1a' id='6028cbfe'/> - <typedef-decl name='DIR' type-id='20cd73f2' id='54a5d683'/> - <typedef-decl name='mode_t' type-id='e1c52942' id='d50d396c'/> - <typedef-decl name='__compar_fn_t' type-id='585e1de9' id='aba7edd8'/> <pointer-type-def type-id='54a5d683' size-in-bits='64' id='f09217ba'/> <pointer-type-def type-id='5725d813' size-in-bits='64' id='07b96073'/> <pointer-type-def type-id='9b293607' size-in-bits='64' id='77bf1784'/> @@ -5071,9 +5065,6 @@ <parameter type-id='9cf59a50'/> <return type-id='48b5725f'/> </function-decl> - <function-decl name='use_color' mangled-name='use_color' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='use_color'> - <return type-id='95e97e5e'/> - </function-decl> <function-decl name='mkdirp' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='80f4b756'/> <parameter type-id='d50d396c'/> @@ -5240,13 +5231,19 @@ <parameter type-id='aba7edd8'/> <return type-id='48b5725f'/> </function-decl> - <function-decl name='statfs64' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='rmdir' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='80f4b756'/> - <parameter type-id='7fd094c8'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='rmdir' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='__openat_too_many_args' visibility='default' binding='global' size-in-bits='64'> + <return type-id='48b5725f'/> + </function-decl> + <function-decl name='__openat_missing_mode' visibility='default' binding='global' size-in-bits='64'> + <return type-id='48b5725f'/> + </function-decl> + <function-decl name='statfs64' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='80f4b756'/> + <parameter type-id='7fd094c8'/> <return type-id='95e97e5e'/> </function-decl> <function-type size-in-bits='64' id='c5c76c9c'> @@ -5258,10 +5255,10 @@ <type-decl name='long long unsigned int' size-in-bits='64' id='3a47d82b'/> <class-decl name='splitflags' size-in-bits='64' is-struct='yes' visibility='default' id='dc01bf52'> <data-member access='public' layout-offset-in-bits='0'> - <var-decl name='dryrun' type-id='95e97e5e' visibility='default'/> + <var-decl name='dryrun' type-id='f0981eeb' visibility='default'/> </data-member> <data-member access='public' layout-offset-in-bits='1'> - <var-decl name='import' type-id='95e97e5e' visibility='default'/> + <var-decl name='import' type-id='f0981eeb' visibility='default'/> </data-member> <data-member access='public' layout-offset-in-bits='32'> <var-decl name='name_flags' type-id='95e97e5e' visibility='default'/> @@ -5410,7 +5407,8 @@ <enumerator name='POOL_INITIALIZE_START' value='0'/> <enumerator name='POOL_INITIALIZE_CANCEL' value='1'/> <enumerator name='POOL_INITIALIZE_SUSPEND' value='2'/> - <enumerator name='POOL_INITIALIZE_FUNCS' value='3'/> + <enumerator name='POOL_INITIALIZE_UNINIT' value='3'/> + <enumerator name='POOL_INITIALIZE_FUNCS' value='4'/> </enum-decl> <typedef-decl name='pool_initialize_func_t' type-id='5c246ad4' id='7063e1ab'/> <enum-decl name='pool_trim_func' id='54ed608a'> @@ -5880,6 +5878,12 @@ <parameter type-id='9d774e0b' name='aux'/> <return type-id='95e97e5e'/> </function-decl> + <function-decl name='zpool_vdev_set_removed_state' mangled-name='zpool_vdev_set_removed_state' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zpool_vdev_set_removed_state'> + <parameter type-id='4c81de99' name='zhp'/> + <parameter type-id='9c313c2d' name='guid'/> + <parameter type-id='9d774e0b' name='aux'/> + <return type-id='95e97e5e'/> + </function-decl> <function-decl name='zpool_vdev_attach' mangled-name='zpool_vdev_attach' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zpool_vdev_attach'> <parameter type-id='4c81de99' name='zhp'/> <parameter type-id='80f4b756' name='old_disk'/> @@ -6052,11 +6056,6 @@ <parameter type-id='95e97e5e'/> <return type-id='3a47d82b'/> </function-decl> - <function-decl name='realpath' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='9d26089a'/> - <parameter type-id='266fe297'/> - <return type-id='26a90f95'/> - </function-decl> <function-decl name='memcmp' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='eaa32e2f'/> <parameter type-id='eaa32e2f'/> @@ -6075,6 +6074,12 @@ <parameter type-id='b59d7dce'/> <return type-id='95e97e5e'/> </function-decl> + <function-decl name='__realpath_chk' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='9d26089a'/> + <parameter type-id='266fe297'/> + <parameter type-id='b59d7dce'/> + <return type-id='26a90f95'/> + </function-decl> <function-decl name='munmap' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='eaa32e2f'/> <parameter type-id='b59d7dce'/> @@ -6864,21 +6869,10 @@ <parameter type-id='a3681dea' name='stream_avl'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='sprintf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='26a90f95'/> - <parameter type-id='80f4b756'/> - <parameter is-variadic='yes'/> - <return type-id='95e97e5e'/> - </function-decl> <function-decl name='perror' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='80f4b756'/> <return type-id='48b5725f'/> </function-decl> - <function-decl name='strcat' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='26a90f95'/> - <parameter type-id='80f4b756'/> - <return type-id='26a90f95'/> - </function-decl> <function-decl name='strndup' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='80f4b756'/> <parameter type-id='b59d7dce'/> @@ -7081,6 +7075,12 @@ <qualified-type-def type-id='d33f11cb' restrict='yes' id='5c53ba29'/> <pointer-type-def type-id='ffa52b96' size-in-bits='64' id='76c8174b'/> <pointer-type-def type-id='f3d87113' size-in-bits='64' id='0d2a0670'/> + <function-decl name='zpool_label_disk' mangled-name='zpool_label_disk' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zpool_label_disk'> + <parameter type-id='b0382bb3'/> + <parameter type-id='4c81de99'/> + <parameter type-id='80f4b756'/> + <return type-id='95e97e5e'/> + </function-decl> <function-decl name='zfs_version_kernel' mangled-name='zfs_version_kernel' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zfs_version_kernel'> <parameter type-id='26a90f95'/> <parameter type-id='95e97e5e'/> @@ -7095,6 +7095,10 @@ <function-decl name='libzfs_load_module' mangled-name='libzfs_load_module' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='libzfs_load_module'> <return type-id='95e97e5e'/> </function-decl> + <function-decl name='zfs_get_underlying_path' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='80f4b756'/> + <return type-id='26a90f95'/> + </function-decl> <function-decl name='zpool_prop_unsupported' mangled-name='zpool_prop_unsupported' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zpool_prop_unsupported'> <parameter type-id='80f4b756'/> <return type-id='c19b74c3'/> @@ -7261,23 +7265,45 @@ <function-decl name='zfs_version_print' mangled-name='zfs_version_print' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zfs_version_print'> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='color_start' mangled-name='color_start' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='color_start'> - <parameter type-id='26a90f95' name='color'/> - <return type-id='48b5725f'/> - </function-decl> - <function-decl name='color_end' mangled-name='color_end' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='color_end'> - <return type-id='48b5725f'/> + <function-decl name='use_color' mangled-name='use_color' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='use_color'> + <return type-id='95e97e5e'/> </function-decl> <function-decl name='printf_color' mangled-name='printf_color' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='printf_color'> - <parameter type-id='26a90f95' name='color'/> + <parameter type-id='80f4b756' name='color'/> <parameter type-id='26a90f95' name='format'/> <parameter is-variadic='yes'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='pow' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='a0eb0f08'/> - <parameter type-id='a0eb0f08'/> - <return type-id='a0eb0f08'/> + <function-decl name='zpool_vdev_script_alloc_env' mangled-name='zpool_vdev_script_alloc_env' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zpool_vdev_script_alloc_env'> + <parameter type-id='80f4b756' name='pool_name'/> + <parameter type-id='80f4b756' name='vdev_path'/> + <parameter type-id='80f4b756' name='vdev_upath'/> + <parameter type-id='80f4b756' name='vdev_enc_sysfs_path'/> + <parameter type-id='80f4b756' name='opt_key'/> + <parameter type-id='80f4b756' name='opt_val'/> + <return type-id='9b23c9ad'/> + </function-decl> + <function-decl name='zpool_vdev_script_free_env' mangled-name='zpool_vdev_script_free_env' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zpool_vdev_script_free_env'> + <parameter type-id='9b23c9ad' name='env'/> + <return type-id='48b5725f'/> + </function-decl> + <function-decl name='zpool_prepare_disk' mangled-name='zpool_prepare_disk' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zpool_prepare_disk'> + <parameter type-id='4c81de99' name='zhp'/> + <parameter type-id='5ce45b60' name='vdev_nv'/> + <parameter type-id='80f4b756' name='prepare_str'/> + <parameter type-id='c0563f85' name='lines'/> + <parameter type-id='7292109c' name='lines_cnt'/> + <return type-id='95e97e5e'/> + </function-decl> + <function-decl name='zpool_prepare_and_label_disk' mangled-name='zpool_prepare_and_label_disk' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zpool_prepare_and_label_disk'> + <parameter type-id='b0382bb3' name='hdl'/> + <parameter type-id='4c81de99' name='zhp'/> + <parameter type-id='80f4b756' name='name'/> + <parameter type-id='5ce45b60' name='vdev_nv'/> + <parameter type-id='80f4b756' name='prepare_str'/> + <parameter type-id='c0563f85' name='lines'/> + <parameter type-id='7292109c' name='lines_cnt'/> + <return type-id='95e97e5e'/> </function-decl> <function-decl name='__ctype_toupper_loc' visibility='default' binding='global' size-in-bits='64'> <return type-id='24f95ba5'/> @@ -7296,25 +7322,6 @@ <parameter type-id='d33f11cb'/> <return type-id='48b5725f'/> </function-decl> - <function-decl name='vfprintf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='e75a27e9'/> - <parameter type-id='9d26089a'/> - <parameter type-id='b7f2d5e6'/> - <return type-id='95e97e5e'/> - </function-decl> - <function-decl name='vsnprintf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='26a90f95'/> - <parameter type-id='b59d7dce'/> - <parameter type-id='80f4b756'/> - <parameter type-id='b7f2d5e6'/> - <return type-id='95e97e5e'/> - </function-decl> - <function-decl name='vasprintf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='8c85230f'/> - <parameter type-id='9d26089a'/> - <parameter type-id='b7f2d5e6'/> - <return type-id='95e97e5e'/> - </function-decl> <function-decl name='strtod' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='9d26089a'/> <parameter type-id='8c85230f'/> @@ -7334,11 +7341,10 @@ <parameter type-id='b59d7dce'/> <return type-id='b59d7dce'/> </function-decl> - <function-decl name='waitpid' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='3629bad8'/> - <parameter type-id='7292109c'/> + <function-decl name='access' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='80f4b756'/> <parameter type-id='95e97e5e'/> - <return type-id='3629bad8'/> + <return type-id='95e97e5e'/> </function-decl> <function-decl name='dup2' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='95e97e5e'/> @@ -7374,6 +7380,31 @@ <function-decl name='vfork' visibility='default' binding='global' size-in-bits='64'> <return type-id='3629bad8'/> </function-decl> + <function-decl name='pow' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='a0eb0f08'/> + <parameter type-id='a0eb0f08'/> + <return type-id='a0eb0f08'/> + </function-decl> + <function-decl name='__vfprintf_chk' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='e75a27e9'/> + <parameter type-id='95e97e5e'/> + <parameter type-id='9d26089a'/> + <parameter type-id='b7f2d5e6'/> + <return type-id='95e97e5e'/> + </function-decl> + <function-decl name='__vasprintf_chk' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='8c85230f'/> + <parameter type-id='95e97e5e'/> + <parameter type-id='9d26089a'/> + <parameter type-id='b7f2d5e6'/> + <return type-id='95e97e5e'/> + </function-decl> + <function-decl name='waitpid' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='3629bad8'/> + <parameter type-id='7292109c'/> + <parameter type-id='95e97e5e'/> + <return type-id='3629bad8'/> + </function-decl> <function-type size-in-bits='64' id='c70fa2e8'> <parameter type-id='95e97e5e'/> <parameter type-id='eaa32e2f'/> @@ -7382,6 +7413,9 @@ </abi-instr> <abi-instr address-size='64' path='os/linux/libzfs_mount_os.c' language='LANG_C99'> <pointer-type-def type-id='7359adad' size-in-bits='64' id='1d2c2b85'/> + <function-decl name='geteuid' visibility='default' binding='global' size-in-bits='64'> + <return type-id='cc5fcceb'/> + </function-decl> <function-decl name='mount' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='80f4b756'/> <parameter type-id='80f4b756'/> @@ -7395,9 +7429,6 @@ <parameter type-id='95e97e5e'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='geteuid' visibility='default' binding='global' size-in-bits='64'> - <return type-id='cc5fcceb'/> - </function-decl> <function-decl name='zfs_parse_mount_options' mangled-name='zfs_parse_mount_options' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zfs_parse_mount_options'> <parameter type-id='26a90f95' name='mntopts'/> <parameter type-id='1d2c2b85' name='mntflags'/> @@ -7568,12 +7599,6 @@ <parameter type-id='95e97e5e'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='zpool_label_disk' mangled-name='zpool_label_disk' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zpool_label_disk'> - <parameter type-id='b0382bb3' name='hdl'/> - <parameter type-id='4c81de99' name='zhp'/> - <parameter type-id='80f4b756' name='name'/> - <return type-id='95e97e5e'/> - </function-decl> </abi-instr> <abi-instr address-size='64' path='os/linux/libzfs_util_os.c' language='LANG_C99'> <typedef-decl name='__useconds_t' type-id='f0981eeb' id='4e80d4b1'/> @@ -7588,11 +7613,6 @@ <parameter type-id='3d83ba87'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='access' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='80f4b756'/> - <parameter type-id='95e97e5e'/> - <return type-id='95e97e5e'/> - </function-decl> <function-decl name='usleep' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='4e80d4b1'/> <return type-id='95e97e5e'/> diff --git a/lib/libzfs/libzfs_dataset.c b/lib/libzfs/libzfs_dataset.c index f2219d1c3d26..09352af0fcee 100644 --- a/lib/libzfs/libzfs_dataset.c +++ b/lib/libzfs/libzfs_dataset.c @@ -1017,6 +1017,7 @@ zfs_valid_proplist(libzfs_handle_t *hdl, zfs_type_t type, nvlist_t *nvl, nvlist_t *ret; int chosen_normal = -1; int chosen_utf = -1; + int set_maxbs = 0; if (nvlist_alloc(&ret, NV_UNIQUE_NAME, 0) != 0) { (void) no_memory(hdl); @@ -1234,12 +1235,17 @@ zfs_valid_proplist(libzfs_handle_t *hdl, zfs_type_t type, nvlist_t *nvl, (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } + /* save the ZFS_PROP_RECORDSIZE during create op */ + if (zpool_hdl == NULL && prop == ZFS_PROP_RECORDSIZE) { + set_maxbs = intval; + } break; } case ZFS_PROP_SPECIAL_SMALL_BLOCKS: { - int maxbs = SPA_OLD_MAXBLOCKSIZE; + int maxbs = + set_maxbs == 0 ? SPA_OLD_MAXBLOCKSIZE : set_maxbs; char buf[64]; if (zpool_hdl != NULL) { @@ -1756,7 +1762,8 @@ zfs_prop_set_list(zfs_handle_t *zhp, nvlist_t *props) nvlist_t *nvl; int nvl_len = 0; int added_resv = 0; - zfs_prop_t prop = 0; + zfs_prop_t prop; + boolean_t nsprop = B_FALSE; nvpair_t *elem; (void) snprintf(errbuf, sizeof (errbuf), @@ -1803,6 +1810,7 @@ zfs_prop_set_list(zfs_handle_t *zhp, nvlist_t *props) elem = nvlist_next_nvpair(nvl, elem)) { prop = zfs_name_to_prop(nvpair_name(elem)); + nsprop |= zfs_is_namespace_prop(prop); assert(cl_idx < nvl_len); /* @@ -1903,8 +1911,7 @@ zfs_prop_set_list(zfs_handle_t *zhp, nvlist_t *props) * if one of the options handled by the generic * Linux namespace layer has been modified. */ - if (zfs_is_namespace_prop(prop) && - zfs_is_mounted(zhp, NULL)) + if (nsprop && zfs_is_mounted(zhp, NULL)) ret = zfs_mount(zhp, MNTOPT_REMOUNT, 0); } } diff --git a/lib/libzfs/libzfs_pool.c b/lib/libzfs/libzfs_pool.c index 29f077841da0..b3e8948cefcb 100644 --- a/lib/libzfs/libzfs_pool.c +++ b/lib/libzfs/libzfs_pool.c @@ -2224,8 +2224,8 @@ xlate_init_err(int err) } /* - * Begin, suspend, or cancel the initialization (initializing of all free - * blocks) for the given vdevs in the given pool. + * Begin, suspend, cancel, or uninit (clear) the initialization (initializing + * of all free blocks) for the given vdevs in the given pool. */ static int zpool_initialize_impl(zpool_handle_t *zhp, pool_initialize_func_t cmd_type, @@ -2251,11 +2251,16 @@ zpool_initialize_impl(zpool_handle_t *zhp, pool_initialize_func_t cmd_type, vdev_guids, &errlist); if (err != 0) { - if (errlist != NULL) { - vd_errlist = fnvlist_lookup_nvlist(errlist, - ZPOOL_INITIALIZE_VDEVS); + if (errlist != NULL && nvlist_lookup_nvlist(errlist, + ZPOOL_INITIALIZE_VDEVS, &vd_errlist) == 0) { goto list_errors; } + + if (err == EINVAL && cmd_type == POOL_INITIALIZE_UNINIT) { + zfs_error_aux(zhp->zpool_hdl, dgettext(TEXT_DOMAIN, + "uninitialize is not supported by kernel")); + } + (void) zpool_standard_error(zhp->zpool_hdl, err, dgettext(TEXT_DOMAIN, "operation failed")); goto out; @@ -2803,6 +2808,9 @@ zpool_vdev_is_interior(const char *name) return (B_FALSE); } +/* + * Lookup the nvlist for a given vdev. + */ nvlist_t * zpool_find_vdev(zpool_handle_t *zhp, const char *path, boolean_t *avail_spare, boolean_t *l2cache, boolean_t *log) @@ -2810,6 +2818,7 @@ zpool_find_vdev(zpool_handle_t *zhp, const char *path, boolean_t *avail_spare, char *end; nvlist_t *nvroot, *search, *ret; uint64_t guid; + boolean_t __avail_spare, __l2cache, __log; verify(nvlist_alloc(&search, NV_UNIQUE_NAME, KM_SLEEP) == 0); @@ -2825,6 +2834,18 @@ zpool_find_vdev(zpool_handle_t *zhp, const char *path, boolean_t *avail_spare, verify(nvlist_lookup_nvlist(zhp->zpool_config, ZPOOL_CONFIG_VDEV_TREE, &nvroot) == 0); + /* + * User can pass NULL for avail_spare, l2cache, and log, but + * we still need to provide variables to vdev_to_nvlist_iter(), so + * just point them to junk variables here. + */ + if (!avail_spare) + avail_spare = &__avail_spare; + if (!l2cache) + l2cache = &__l2cache; + if (!log) + log = &__log; + *avail_spare = B_FALSE; *l2cache = B_FALSE; if (log != NULL) @@ -3228,21 +3249,23 @@ zpool_vdev_fault(zpool_handle_t *zhp, uint64_t guid, vdev_aux_t aux) } /* - * Mark the given vdev degraded. + * Generic set vdev state function */ -int -zpool_vdev_degrade(zpool_handle_t *zhp, uint64_t guid, vdev_aux_t aux) +static int +zpool_vdev_set_state(zpool_handle_t *zhp, uint64_t guid, vdev_aux_t aux, + vdev_state_t state) { zfs_cmd_t zc = {"\0"}; char msg[1024]; libzfs_handle_t *hdl = zhp->zpool_hdl; (void) snprintf(msg, sizeof (msg), - dgettext(TEXT_DOMAIN, "cannot degrade %llu"), (u_longlong_t)guid); + dgettext(TEXT_DOMAIN, "cannot set %s %llu"), + zpool_state_to_name(state, aux), (u_longlong_t)guid); (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); zc.zc_guid = guid; - zc.zc_cookie = VDEV_STATE_DEGRADED; + zc.zc_cookie = state; zc.zc_obj = aux; if (zfs_ioctl(hdl, ZFS_IOC_VDEV_SET_STATE, &zc) == 0) @@ -3252,6 +3275,27 @@ zpool_vdev_degrade(zpool_handle_t *zhp, uint64_t guid, vdev_aux_t aux) } /* + * Mark the given vdev degraded. + */ +int +zpool_vdev_degrade(zpool_handle_t *zhp, uint64_t guid, vdev_aux_t aux) +{ + return (zpool_vdev_set_state(zhp, guid, aux, VDEV_STATE_DEGRADED)); +} + +/* + * Mark the given vdev as in a removed state (as if the device does not exist). + * + * This is different than zpool_vdev_remove() which does a removal of a device + * from the pool (but the device does exist). + */ +int +zpool_vdev_set_removed_state(zpool_handle_t *zhp, uint64_t guid, vdev_aux_t aux) +{ + return (zpool_vdev_set_state(zhp, guid, aux, VDEV_STATE_REMOVED)); +} + +/* * Returns TRUE if the given nvlist is a vdev that was originally swapped in as * a hot spare. */ diff --git a/lib/libzfs/libzfs_sendrecv.c b/lib/libzfs/libzfs_sendrecv.c index 8f496b20b89f..c5e62d35c302 100644 --- a/lib/libzfs/libzfs_sendrecv.c +++ b/lib/libzfs/libzfs_sendrecv.c @@ -981,39 +981,63 @@ send_print_verbose(FILE *fout, const char *tosnap, const char *fromsnap, { if (parsable) { if (fromsnap != NULL) { - (void) fprintf(fout, "incremental\t%s\t%s", - fromsnap, tosnap); + (void) fprintf(fout, dgettext(TEXT_DOMAIN, + "incremental\t%s\t%s"), fromsnap, tosnap); } else { - (void) fprintf(fout, "full\t%s", - tosnap); +/* + * Workaround for GCC 12+ with UBSan enabled deficencies. + * + * GCC 12+ invoked with -fsanitize=undefined incorrectly reports the code + * below as violating -Wformat-overflow. + */ +#if defined(__GNUC__) && !defined(__clang__) && \ + defined(ZFS_UBSAN_ENABLED) && defined(HAVE_FORMAT_OVERFLOW) +#pragma GCC diagnostic push +#pragma GCC diagnostic ignored "-Wformat-overflow" +#endif + (void) fprintf(fout, dgettext(TEXT_DOMAIN, + "full\t%s"), tosnap); +#if defined(__GNUC__) && !defined(__clang__) && \ + defined(ZFS_UBSAN_ENABLED) && defined(HAVE_FORMAT_OVERFLOW) +#pragma GCC diagnostic pop +#endif } + (void) fprintf(fout, "\t%llu", (longlong_t)size); } else { if (fromsnap != NULL) { if (strchr(fromsnap, '@') == NULL && strchr(fromsnap, '#') == NULL) { (void) fprintf(fout, dgettext(TEXT_DOMAIN, - "send from @%s to %s"), - fromsnap, tosnap); + "send from @%s to %s"), fromsnap, tosnap); } else { (void) fprintf(fout, dgettext(TEXT_DOMAIN, - "send from %s to %s"), - fromsnap, tosnap); + "send from %s to %s"), fromsnap, tosnap); } } else { (void) fprintf(fout, dgettext(TEXT_DOMAIN, - "full send of %s"), - tosnap); + "full send of %s"), tosnap); + } + if (size != 0) { + char buf[16]; + zfs_nicebytes(size, buf, sizeof (buf)); +/* + * Workaround for GCC 12+ with UBSan enabled deficencies. + * + * GCC 12+ invoked with -fsanitize=undefined incorrectly reports the code + * below as violating -Wformat-overflow. + */ +#if defined(__GNUC__) && !defined(__clang__) && \ + defined(ZFS_UBSAN_ENABLED) && defined(HAVE_FORMAT_OVERFLOW) +#pragma GCC diagnostic push +#pragma GCC diagnostic ignored "-Wformat-overflow" +#endif + (void) fprintf(fout, dgettext(TEXT_DOMAIN, + " estimated size is %s"), buf); +#if defined(__GNUC__) && !defined(__clang__) && \ + defined(ZFS_UBSAN_ENABLED) && defined(HAVE_FORMAT_OVERFLOW) +#pragma GCC diagnostic pop +#endif } - } - - if (parsable) { - (void) fprintf(fout, "\t%llu", - (longlong_t)size); - } else if (size != 0) { - char buf[16]; - zfs_nicebytes(size, buf, sizeof (buf)); - (void) fprintf(fout, dgettext(TEXT_DOMAIN, - " estimated size is %s"), buf); } (void) fprintf(fout, "\n"); } diff --git a/lib/libzfs/libzfs_util.c b/lib/libzfs/libzfs_util.c index 7c4d310782eb..256ae9d1b778 100644 --- a/lib/libzfs/libzfs_util.c +++ b/lib/libzfs/libzfs_util.c @@ -2121,3 +2121,196 @@ printf_color(const char *color, char *format, ...) return (rc); } + +/* PATH + 5 env vars + a NULL entry = 7 */ +#define ZPOOL_VDEV_SCRIPT_ENV_COUNT 7 + +/* + * There's a few places where ZFS will call external scripts (like the script + * in zpool.d/ and `zfs_prepare_disk`). These scripts are called with a + * reduced $PATH, and some vdev specific environment vars set. This function + * will allocate an populate the environment variable array that is passed to + * these scripts. The user must free the arrays with zpool_vdev_free_env() when + * they are done. + * + * The following env vars will be set (but value could be blank): + * + * POOL_NAME + * VDEV_PATH + * VDEV_UPATH + * VDEV_ENC_SYSFS_PATH + * + * In addition, you can set an optional environment variable named 'opt_key' + * to 'opt_val' if you want. + * + * Returns allocated env[] array on success, NULL otherwise. + */ +char ** +zpool_vdev_script_alloc_env(const char *pool_name, + const char *vdev_path, const char *vdev_upath, + const char *vdev_enc_sysfs_path, const char *opt_key, const char *opt_val) +{ + char **env = NULL; + int rc; + + env = calloc(ZPOOL_VDEV_SCRIPT_ENV_COUNT, sizeof (*env)); + if (!env) + return (NULL); + + env[0] = strdup("PATH=/bin:/sbin:/usr/bin:/usr/sbin"); + if (!env[0]) + goto error; + + /* Setup our custom environment variables */ + rc = asprintf(&env[1], "POOL_NAME=%s", pool_name ? pool_name : ""); + if (rc == -1) { + env[1] = NULL; + goto error; + } + + rc = asprintf(&env[2], "VDEV_PATH=%s", vdev_path ? vdev_path : ""); + if (rc == -1) { + env[2] = NULL; + goto error; + } + + rc = asprintf(&env[3], "VDEV_UPATH=%s", vdev_upath ? vdev_upath : ""); + if (rc == -1) { + env[3] = NULL; + goto error; + } + + rc = asprintf(&env[4], "VDEV_ENC_SYSFS_PATH=%s", + vdev_enc_sysfs_path ? vdev_enc_sysfs_path : ""); + if (rc == -1) { + env[4] = NULL; + goto error; + } + + if (opt_key != NULL) { + rc = asprintf(&env[5], "%s=%s", opt_key, + opt_val ? opt_val : ""); + if (rc == -1) { + env[5] = NULL; + goto error; + } + } + + return (env); + +error: + for (int i = 0; i < ZPOOL_VDEV_SCRIPT_ENV_COUNT; i++) + free(env[i]); + + free(env); + + return (NULL); +} + +/* + * Free the env[] array that was allocated by zpool_vdev_script_alloc_env(). + */ +void +zpool_vdev_script_free_env(char **env) +{ + for (int i = 0; i < ZPOOL_VDEV_SCRIPT_ENV_COUNT; i++) + free(env[i]); + + free(env); +} + +/* + * Prepare a disk by (optionally) running a program before labeling the disk. + * This can be useful for installing disk firmware or doing some pre-flight + * checks on the disk before it becomes part of the pool. The program run is + * located at ZFSEXECDIR/zfs_prepare_disk + * (E.x: /usr/local/libexec/zfs/zfs_prepare_disk). + * + * Return 0 on success, non-zero on failure. + */ +int +zpool_prepare_disk(zpool_handle_t *zhp, nvlist_t *vdev_nv, + const char *prepare_str, char **lines[], int *lines_cnt) +{ + const char *script_path = ZFSEXECDIR "/zfs_prepare_disk"; + const char *pool_name; + int rc = 0; + + /* Path to script and a NULL entry */ + char *argv[2] = {(char *)script_path}; + char **env = NULL; + char *path = NULL, *enc_sysfs_path = NULL; + char *upath; + *lines_cnt = 0; + + if (access(script_path, X_OK) != 0) { + /* No script, nothing to do */ + return (0); + } + + (void) nvlist_lookup_string(vdev_nv, ZPOOL_CONFIG_PATH, &path); + (void) nvlist_lookup_string(vdev_nv, ZPOOL_CONFIG_VDEV_ENC_SYSFS_PATH, + &enc_sysfs_path); + + upath = zfs_get_underlying_path(path); + pool_name = zhp ? zpool_get_name(zhp) : NULL; + + env = zpool_vdev_script_alloc_env(pool_name, path, upath, + enc_sysfs_path, "VDEV_PREPARE", prepare_str); + + free(upath); + + if (env == NULL) { + return (ENOMEM); + } + + rc = libzfs_run_process_get_stdout(script_path, argv, env, lines, + lines_cnt); + + zpool_vdev_script_free_env(env); + + return (rc); +} + +/* + * Optionally run a script and then label a disk. The script can be used to + * prepare a disk for inclusion into the pool. For example, it might update + * the disk's firmware or check its health. + * + * The 'name' provided is the short name, stripped of any leading + * /dev path, and is passed to zpool_label_disk. vdev_nv is the nvlist for + * the vdev. prepare_str is a string that gets passed as the VDEV_PREPARE + * env variable to the script. + * + * The following env vars are passed to the script: + * + * POOL_NAME: The pool name (blank during zpool create) + * VDEV_PREPARE: Reason why the disk is being prepared for inclusion: + * "create", "add", "replace", or "autoreplace" + * VDEV_PATH: Path to the disk + * VDEV_UPATH: One of the 'underlying paths' to the disk. This is + * useful for DM devices. + * VDEV_ENC_SYSFS_PATH: Path to the disk's enclosure sysfs path, if available. + * + * Note, some of these values can be blank. + * + * Return 0 on success, non-zero otherwise. + */ +int +zpool_prepare_and_label_disk(libzfs_handle_t *hdl, zpool_handle_t *zhp, + const char *name, nvlist_t *vdev_nv, const char *prepare_str, + char **lines[], int *lines_cnt) +{ + int rc; + char vdev_path[MAXPATHLEN]; + (void) snprintf(vdev_path, sizeof (vdev_path), "%s/%s", DISK_ROOT, + name); + + /* zhp will be NULL when creating a pool */ + rc = zpool_prepare_disk(zhp, vdev_nv, prepare_str, lines, lines_cnt); + if (rc != 0) + return (rc); + + rc = zpool_label_disk(hdl, zhp, name); + return (rc); +} diff --git a/lib/libzfs_core/libzfs_core.abi b/lib/libzfs_core/libzfs_core.abi index 1b03a5c42ef4..089cf48ae9ef 100644 --- a/lib/libzfs_core/libzfs_core.abi +++ b/lib/libzfs_core/libzfs_core.abi @@ -6,7 +6,6 @@ <dependency name='libblkid.so.1'/> <dependency name='libudev.so.1'/> <dependency name='libnvpair.so.3'/> - <dependency name='libtirpc.so.3'/> <dependency name='libc.so.6'/> <dependency name='ld-linux-x86-64.so.2'/> </elf-needed> @@ -159,8 +158,11 @@ <elf-symbol name='efi_type' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='efi_use_whole_disk' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='efi_write' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> + <elf-symbol name='for_each_real_leaf_vdev_macro_helper_func' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='for_each_vdev_cb' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='for_each_vdev_in_nvlist' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> + <elf-symbol name='for_each_vdev_macro_helper_func' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> + <elf-symbol name='fsleep' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='get_system_hostid' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='getexecname' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='getextmntent' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> @@ -255,6 +257,7 @@ <elf-symbol name='tpool_suspended' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='tpool_wait' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='update_vdev_config_dev_strs' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> + <elf-symbol name='update_vdev_config_dev_sysfs_path' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='update_vdevs_config_dev_sysfs_path' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zfs_append_partition' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zfs_dev_flush' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> @@ -278,9 +281,11 @@ <elf-symbol name='zfs_strip_partition' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zfs_strip_path' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zpool_default_search_paths' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> + <elf-symbol name='zpool_disk_wait' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zpool_dump_ddt' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zpool_find_config' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zpool_find_import_blkid' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> + <elf-symbol name='zpool_getenv_int' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zpool_history_unpack' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zpool_label_disk_wait' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> <elf-symbol name='zpool_open_func' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/> @@ -407,12 +412,6 @@ <parameter type-id='c43b27a6' name='vtoc'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='sprintf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='26a90f95'/> - <parameter type-id='80f4b756'/> - <parameter is-variadic='yes'/> - <return type-id='95e97e5e'/> - </function-decl> <function-decl name='write' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='95e97e5e'/> <parameter type-id='eaa32e2f'/> @@ -440,16 +439,32 @@ <type-decl name='unsigned long int' size-in-bits='64' id='7359adad'/> </abi-instr> <abi-instr address-size='64' path='assert.c' language='LANG_C99'> + <class-decl name='__va_list_tag' size-in-bits='192' is-struct='yes' visibility='default' id='d5027220'> + <data-member access='public' layout-offset-in-bits='0'> + <var-decl name='gp_offset' type-id='f0981eeb' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='32'> + <var-decl name='fp_offset' type-id='f0981eeb' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='64'> + <var-decl name='overflow_arg_area' type-id='eaa32e2f' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='128'> + <var-decl name='reg_save_area' type-id='eaa32e2f' visibility='default'/> + </data-member> + </class-decl> + <pointer-type-def type-id='d5027220' size-in-bits='64' id='b7f2d5e6'/> <var-decl name='aok' type-id='95e97e5e' mangled-name='aok' visibility='default' elf-symbol-id='aok'/> - <function-decl name='vfprintf' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='abort' visibility='default' binding='global' size-in-bits='64'> + <return type-id='48b5725f'/> + </function-decl> + <function-decl name='__vfprintf_chk' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='e75a27e9'/> + <parameter type-id='95e97e5e'/> <parameter type-id='9d26089a'/> <parameter type-id='b7f2d5e6'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='abort' visibility='default' binding='global' size-in-bits='64'> - <return type-id='48b5725f'/> - </function-decl> </abi-instr> <abi-instr address-size='64' path='atomic.c' language='LANG_C99'> <typedef-decl name='int8_t' type-id='2171a512' id='ee31ee44'/> @@ -895,16 +910,18 @@ <parameter type-id='d50d396c' name='mode'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='mbstowcs' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='__mbstowcs_chk' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='f1358bc3'/> <parameter type-id='9d26089a'/> <parameter type-id='b59d7dce'/> + <parameter type-id='b59d7dce'/> <return type-id='b59d7dce'/> </function-decl> - <function-decl name='wcstombs' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='__wcstombs_chk' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='266fe297'/> <parameter type-id='598aab80'/> <parameter type-id='b59d7dce'/> + <parameter type-id='b59d7dce'/> <return type-id='b59d7dce'/> </function-decl> <function-decl name='mkdir' visibility='default' binding='global' size-in-bits='64'> @@ -1024,15 +1041,8 @@ <abi-instr address-size='64' path='page.c' language='LANG_C99'> <var-decl name='pagesize' type-id='b59d7dce' mangled-name='pagesize' visibility='default' elf-symbol-id='pagesize'/> </abi-instr> - <abi-instr address-size='64' path='strlcat.c' language='LANG_C99'> - <function-decl name='memcpy' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='eaa32e2f'/> - <parameter type-id='eaa32e2f'/> - <parameter type-id='b59d7dce'/> - <return type-id='eaa32e2f'/> - </function-decl> - </abi-instr> <abi-instr address-size='64' path='timestamp.c' language='LANG_C99'> + <typedef-decl name='nl_item' type-id='95e97e5e' id='03b79a94'/> <class-decl name='tm' size-in-bits='448' is-struct='yes' visibility='default' id='dddf6ca2'> <data-member access='public' layout-offset-in-bits='0'> <var-decl name='tm_sec' type-id='95e97e5e' visibility='default'/> @@ -1069,7 +1079,6 @@ </data-member> </class-decl> <typedef-decl name='time_t' type-id='65eda9c0' id='c9d12d66'/> - <typedef-decl name='nl_item' type-id='95e97e5e' id='03b79a94'/> <qualified-type-def type-id='c9d12d66' const='yes' id='588b3216'/> <pointer-type-def type-id='588b3216' size-in-bits='64' id='9f201474'/> <qualified-type-def type-id='dddf6ca2' const='yes' id='e824a34f'/> @@ -1144,6 +1153,40 @@ <var-decl name='tpa_tid' type-id='4051f5e7' visibility='default'/> </data-member> </class-decl> + <class-decl name='__cancel_jmp_buf_tag' size-in-bits='576' is-struct='yes' visibility='default' id='8901473c'> + <data-member access='public' layout-offset-in-bits='0'> + <var-decl name='__cancel_jmp_buf' type-id='379a1ab7' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='512'> + <var-decl name='__mask_was_saved' type-id='95e97e5e' visibility='default'/> + </data-member> + </class-decl> + <class-decl name='__pthread_unwind_buf_t' size-in-bits='832' is-struct='yes' naming-typedef-id='4423cf7f' visibility='default' id='a0abc656'> + <data-member access='public' layout-offset-in-bits='0'> + <var-decl name='__cancel_jmp_buf' type-id='f5da478b' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='576'> + <var-decl name='__pad' type-id='209ef23f' visibility='default'/> + </data-member> + </class-decl> + <typedef-decl name='__pthread_unwind_buf_t' type-id='a0abc656' id='4423cf7f'/> + <union-decl name='__atomic_wide_counter' size-in-bits='64' naming-typedef-id='f3b40860' visibility='default' id='613ce450'> + <data-member access='public'> + <var-decl name='__value64' type-id='3a47d82b' visibility='default'/> + </data-member> + <data-member access='public'> + <var-decl name='__value32' type-id='e7f43f72' visibility='default'/> + </data-member> + </union-decl> + <class-decl name='__anonymous_struct__' size-in-bits='64' is-struct='yes' is-anonymous='yes' visibility='default' id='e7f43f72'> + <data-member access='public' layout-offset-in-bits='0'> + <var-decl name='__low' type-id='f0981eeb' visibility='default'/> + </data-member> + <data-member access='public' layout-offset-in-bits='32'> + <var-decl name='__high' type-id='f0981eeb' visibility='default'/> + </data-member> + </class-decl> + <typedef-decl name='__atomic_wide_counter' type-id='613ce450' id='f3b40860'/> <typedef-decl name='__cpu_mask' type-id='7359adad' id='49ef3ffd'/> <class-decl name='cpu_set_t' size-in-bits='1024' is-struct='yes' naming-typedef-id='8037c762' visibility='default' id='1f20d231'> <data-member access='public' layout-offset-in-bits='0'> @@ -1176,10 +1219,10 @@ <typedef-decl name='__jmp_buf' type-id='5d4efd44' id='379a1ab7'/> <class-decl name='__pthread_cond_s' size-in-bits='384' is-struct='yes' visibility='default' id='c987b47c'> <data-member access='public' layout-offset-in-bits='0'> - <var-decl name='' type-id='ac5ab595' visibility='default'/> + <var-decl name='__wseq' type-id='f3b40860' visibility='default'/> </data-member> <data-member access='public' layout-offset-in-bits='64'> - <var-decl name='' type-id='ac5ab596' visibility='default'/> + <var-decl name='__g1_start' type-id='f3b40860' visibility='default'/> </data-member> <data-member access='public' layout-offset-in-bits='128'> <var-decl name='__g_refs' type-id='0d532ec1' visibility='default'/> @@ -1197,30 +1240,6 @@ <var-decl name='__g_signals' type-id='0d532ec1' visibility='default'/> </data-member> </class-decl> - <union-decl name='__anonymous_union__1' size-in-bits='64' is-anonymous='yes' visibility='default' id='ac5ab595'> - <data-member access='public'> - <var-decl name='__wseq' type-id='3a47d82b' visibility='default'/> - </data-member> - <data-member access='public'> - <var-decl name='__wseq32' type-id='e7f43f72' visibility='default'/> - </data-member> - </union-decl> - <class-decl name='__anonymous_struct__' size-in-bits='64' is-struct='yes' is-anonymous='yes' visibility='default' id='e7f43f72'> - <data-member access='public' layout-offset-in-bits='0'> - <var-decl name='__low' type-id='f0981eeb' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='32'> - <var-decl name='__high' type-id='f0981eeb' visibility='default'/> - </data-member> - </class-decl> - <union-decl name='__anonymous_union__2' size-in-bits='64' is-anonymous='yes' visibility='default' id='ac5ab596'> - <data-member access='public'> - <var-decl name='__g1_start' type-id='3a47d82b' visibility='default'/> - </data-member> - <data-member access='public'> - <var-decl name='__g1_start32' type-id='e7f43f72' visibility='default'/> - </data-member> - </union-decl> <class-decl name='__sigset_t' size-in-bits='1024' is-struct='yes' naming-typedef-id='b9c97942' visibility='default' id='2616147f'> <data-member access='public' layout-offset-in-bits='0'> <var-decl name='__val' type-id='d2baa450' visibility='default'/> @@ -1232,23 +1251,6 @@ <var-decl name='sched_priority' type-id='95e97e5e' visibility='default'/> </data-member> </class-decl> - <class-decl name='__cancel_jmp_buf_tag' size-in-bits='576' is-struct='yes' visibility='default' id='8901473c'> - <data-member access='public' layout-offset-in-bits='0'> - <var-decl name='__cancel_jmp_buf' type-id='379a1ab7' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='512'> - <var-decl name='__mask_was_saved' type-id='95e97e5e' visibility='default'/> - </data-member> - </class-decl> - <class-decl name='__pthread_unwind_buf_t' size-in-bits='832' is-struct='yes' naming-typedef-id='4423cf7f' visibility='default' id='a0abc656'> - <data-member access='public' layout-offset-in-bits='0'> - <var-decl name='__cancel_jmp_buf' type-id='f5da478b' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='576'> - <var-decl name='__pad' type-id='209ef23f' visibility='default'/> - </data-member> - </class-decl> - <typedef-decl name='__pthread_unwind_buf_t' type-id='a0abc656' id='4423cf7f'/> <pointer-type-def type-id='8901473c' size-in-bits='64' id='eb91b7ea'/> <pointer-type-def type-id='4423cf7f' size-in-bits='64' id='ba7c727c'/> <pointer-type-def type-id='b9c97942' size-in-bits='64' id='bbf06c47'/> @@ -1267,8 +1269,6 @@ <qualified-type-def type-id='0897719a' const='yes' id='c4a7b189'/> <pointer-type-def type-id='c4a7b189' size-in-bits='64' id='36fca399'/> <qualified-type-def type-id='36fca399' restrict='yes' id='37e4897b'/> - <qualified-type-def type-id='a9c79a1f' const='yes' id='cd087e36'/> - <pointer-type-def type-id='cd087e36' size-in-bits='64' id='e05e8614'/> <qualified-type-def type-id='e05e8614' restrict='yes' id='0be2e71c'/> <pointer-type-def type-id='8037c762' size-in-bits='64' id='d74a6869'/> <qualified-type-def type-id='7292109c' restrict='yes' id='6942f6a4'/> @@ -1306,16 +1306,6 @@ <parameter type-id='9cf59a50' name='tpool'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='__sysconf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='95e97e5e'/> - <return type-id='bd54fe1a'/> - </function-decl> - <function-decl name='pthread_sigmask' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='95e97e5e'/> - <parameter type-id='5a8729d0'/> - <parameter type-id='65e6ec45'/> - <return type-id='95e97e5e'/> - </function-decl> <function-decl name='pthread_create' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='cc338b26'/> <parameter type-id='e1815e87'/> @@ -1468,6 +1458,16 @@ <parameter type-id='0be2e71c'/> <return type-id='95e97e5e'/> </function-decl> + <function-decl name='__sysconf' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='95e97e5e'/> + <return type-id='bd54fe1a'/> + </function-decl> + <function-decl name='pthread_sigmask' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='95e97e5e'/> + <parameter type-id='5a8729d0'/> + <parameter type-id='65e6ec45'/> + <return type-id='95e97e5e'/> + </function-decl> <function-type size-in-bits='64' id='cd5d79f4'> <parameter type-id='eaa32e2f'/> <return type-id='eaa32e2f'/> @@ -1726,7 +1726,8 @@ <enumerator name='POOL_INITIALIZE_START' value='0'/> <enumerator name='POOL_INITIALIZE_CANCEL' value='1'/> <enumerator name='POOL_INITIALIZE_SUSPEND' value='2'/> - <enumerator name='POOL_INITIALIZE_FUNCS' value='3'/> + <enumerator name='POOL_INITIALIZE_UNINIT' value='3'/> + <enumerator name='POOL_INITIALIZE_FUNCS' value='4'/> </enum-decl> <typedef-decl name='pool_initialize_func_t' type-id='5c246ad4' id='7063e1ab'/> <enum-decl name='pool_trim_func' id='54ed608a'> @@ -1865,7 +1866,7 @@ <var-decl name='drr_payloadlen' type-id='8f92235e' visibility='default'/> </data-member> <data-member access='public' layout-offset-in-bits='64'> - <var-decl name='drr_u' type-id='ac5ab597' visibility='default'/> + <var-decl name='drr_u' type-id='ac5ab595' visibility='default'/> </data-member> </class-decl> <enum-decl name='__anonymous_enum__' is-anonymous='yes' id='08f5ca17'> @@ -1883,7 +1884,7 @@ <enumerator name='DRR_REDACT' value='10'/> <enumerator name='DRR_NUMTYPES' value='11'/> </enum-decl> - <union-decl name='__anonymous_union__' size-in-bits='2432' is-anonymous='yes' visibility='default' id='ac5ab597'> + <union-decl name='__anonymous_union__' size-in-bits='2432' is-anonymous='yes' visibility='default' id='ac5ab595'> <data-member access='public'> <var-decl name='drr_begin' type-id='09fcdc01' visibility='default'/> </data-member> @@ -2987,10 +2988,17 @@ <parameter type-id='95e97e5e'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='read' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='__open_too_many_args' visibility='default' binding='global' size-in-bits='64'> + <return type-id='48b5725f'/> + </function-decl> + <function-decl name='__open_missing_mode' visibility='default' binding='global' size-in-bits='64'> + <return type-id='48b5725f'/> + </function-decl> + <function-decl name='__read_chk' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='95e97e5e'/> <parameter type-id='eaa32e2f'/> <parameter type-id='b59d7dce'/> + <parameter type-id='b59d7dce'/> <return type-id='79a0948f'/> </function-decl> </abi-instr> @@ -3158,21 +3166,23 @@ <parameter type-id='822cd80b'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='fgets' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='266fe297'/> - <parameter type-id='95e97e5e'/> - <parameter type-id='e75a27e9'/> - <return type-id='26a90f95'/> - </function-decl> <function-decl name='strstr' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='80f4b756'/> <parameter type-id='80f4b756'/> <return type-id='26a90f95'/> </function-decl> - <function-decl name='readlink' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='__fgets_chk' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='266fe297'/> + <parameter type-id='b59d7dce'/> + <parameter type-id='95e97e5e'/> + <parameter type-id='e75a27e9'/> + <return type-id='26a90f95'/> + </function-decl> + <function-decl name='__readlink_chk' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='9d26089a'/> <parameter type-id='266fe297'/> <parameter type-id='b59d7dce'/> + <parameter type-id='b59d7dce'/> <return type-id='79a0948f'/> </function-decl> <function-decl name='zfs_strip_partition' mangled-name='zfs_strip_partition' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zfs_strip_partition'> @@ -3210,19 +3220,18 @@ <class-decl name='blkid_struct_dev' is-struct='yes' visibility='default' is-declaration-only='yes' id='86223623'/> <class-decl name='blkid_struct_dev_iterate' is-struct='yes' visibility='default' is-declaration-only='yes' id='d88420d6'/> <class-decl name='udev_list_entry' is-struct='yes' visibility='default' is-declaration-only='yes' id='e7dbdca3'/> - <typedef-decl name='__useconds_t' type-id='f0981eeb' id='4e80d4b1'/> - <typedef-decl name='__clockid_t' type-id='95e97e5e' id='08f9a87a'/> - <typedef-decl name='clockid_t' type-id='08f9a87a' id='a1c3b834'/> <typedef-decl name='blkid_dev' type-id='8433f053' id='f47b023a'/> <typedef-decl name='blkid_cache' type-id='940e3afc' id='0882dfdf'/> <typedef-decl name='blkid_dev_iterate' type-id='b8fa2efc' id='f4760fa7'/> + <typedef-decl name='__useconds_t' type-id='f0981eeb' id='4e80d4b1'/> + <typedef-decl name='__clockid_t' type-id='95e97e5e' id='08f9a87a'/> + <typedef-decl name='clockid_t' type-id='08f9a87a' id='a1c3b834'/> <pointer-type-def type-id='0882dfdf' size-in-bits='64' id='2e3e7caa'/> <pointer-type-def type-id='f47b023a' size-in-bits='64' id='d87f9b75'/> <pointer-type-def type-id='09286066' size-in-bits='64' id='940e3afc'/> <pointer-type-def type-id='86223623' size-in-bits='64' id='8433f053'/> <pointer-type-def type-id='d88420d6' size-in-bits='64' id='b8fa2efc'/> <qualified-type-def type-id='62f7a03d' restrict='yes' id='f1cadedf'/> - <pointer-type-def type-id='a9c79a1f' size-in-bits='64' id='3d83ba87'/> <pointer-type-def type-id='e7dbdca3' size-in-bits='64' id='deabd0d3'/> <class-decl name='blkid_struct_cache' is-struct='yes' visibility='default' is-declaration-only='yes' id='09286066'/> <class-decl name='blkid_struct_dev' is-struct='yes' visibility='default' is-declaration-only='yes' id='86223623'/> @@ -3306,11 +3315,6 @@ <parameter type-id='b59d7dce'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='stat64' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='9d26089a'/> - <parameter type-id='f1cadedf'/> - <return type-id='95e97e5e'/> - </function-decl> <function-decl name='clock_gettime' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='a1c3b834'/> <parameter type-id='3d83ba87'/> @@ -3320,6 +3324,11 @@ <parameter type-id='4e80d4b1'/> <return type-id='95e97e5e'/> </function-decl> + <function-decl name='stat64' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='9d26089a'/> + <parameter type-id='f1cadedf'/> + <return type-id='95e97e5e'/> + </function-decl> <function-decl name='zfs_dev_flush' mangled-name='zfs_dev_flush' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zfs_dev_flush'> <parameter type-id='95e97e5e' name='fd'/> <return type-id='95e97e5e'/> @@ -3341,6 +3350,16 @@ <parameter type-id='95e97e5e' name='timeout_ms'/> <return type-id='95e97e5e'/> </function-decl> + <function-decl name='zpool_disk_wait' mangled-name='zpool_disk_wait' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zpool_disk_wait'> + <parameter type-id='80f4b756' name='path'/> + <return type-id='95e97e5e'/> + </function-decl> + <function-decl name='update_vdev_config_dev_sysfs_path' mangled-name='update_vdev_config_dev_sysfs_path' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='update_vdev_config_dev_sysfs_path'> + <parameter type-id='5ce45b60' name='nv'/> + <parameter type-id='80f4b756' name='path'/> + <parameter type-id='80f4b756' name='key'/> + <return type-id='48b5725f'/> + </function-decl> </abi-instr> <abi-instr address-size='64' path='os/linux/zutil_setproctitle.c' language='LANG_C99'> <function-decl name='warnx' visibility='default' binding='global' size-in-bits='64'> @@ -3398,23 +3417,10 @@ <parameter type-id='95e97e5e' name='wholedisk'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='snprintf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='26a90f95'/> - <parameter type-id='b59d7dce'/> - <parameter type-id='80f4b756'/> - <parameter is-variadic='yes'/> - <return type-id='95e97e5e'/> - </function-decl> <function-decl name='getenv' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='80f4b756'/> <return type-id='26a90f95'/> </function-decl> - <function-decl name='memset' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='eaa32e2f'/> - <parameter type-id='95e97e5e'/> - <parameter type-id='b59d7dce'/> - <return type-id='eaa32e2f'/> - </function-decl> <function-decl name='strcmp' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='80f4b756'/> <parameter type-id='80f4b756'/> @@ -3468,20 +3474,6 @@ <class-decl name='_IO_marker' is-struct='yes' visibility='default' is-declaration-only='yes' id='010ae0b9'/> <class-decl name='_IO_wide_data' is-struct='yes' visibility='default' is-declaration-only='yes' id='79bd3751'/> <class-decl name='__dirstream' is-struct='yes' visibility='default' is-declaration-only='yes' id='20cd73f2'/> - <class-decl name='__va_list_tag' size-in-bits='192' is-struct='yes' visibility='default' id='d5027220'> - <data-member access='public' layout-offset-in-bits='0'> - <var-decl name='gp_offset' type-id='f0981eeb' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='32'> - <var-decl name='fp_offset' type-id='f0981eeb' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='64'> - <var-decl name='overflow_arg_area' type-id='eaa32e2f' visibility='default'/> - </data-member> - <data-member access='public' layout-offset-in-bits='128'> - <var-decl name='reg_save_area' type-id='eaa32e2f' visibility='default'/> - </data-member> - </class-decl> <class-decl name='tpool' size-in-bits='2496' is-struct='yes' visibility='default' id='88d1b7f9'> <data-member access='public' layout-offset-in-bits='0'> <var-decl name='tp_forw' type-id='9cf59a50' visibility='default'/> @@ -3671,6 +3663,8 @@ <var-decl name='__glibc_reserved' type-id='16dc656a' visibility='default'/> </data-member> </class-decl> + <typedef-decl name='DIR' type-id='20cd73f2' id='54a5d683'/> + <typedef-decl name='uintptr_t' type-id='7359adad' id='e475ab95'/> <class-decl name='dirent64' size-in-bits='2240' is-struct='yes' visibility='default' id='5725d813'> <data-member access='public' layout-offset-in-bits='0'> <var-decl name='d_ino' type-id='71288a47' visibility='default'/> @@ -3836,10 +3830,10 @@ <var-decl name='sigev_notify' type-id='95e97e5e' visibility='default'/> </data-member> <data-member access='public' layout-offset-in-bits='128'> - <var-decl name='_sigev_un' type-id='ac5ab598' visibility='default'/> + <var-decl name='_sigev_un' type-id='ac5ab596' visibility='default'/> </data-member> </class-decl> - <union-decl name='__anonymous_union__' size-in-bits='384' is-anonymous='yes' visibility='default' id='ac5ab598'> + <union-decl name='__anonymous_union__' size-in-bits='384' is-anonymous='yes' visibility='default' id='ac5ab596'> <data-member access='public'> <var-decl name='_pad' type-id='73b82f0f' visibility='default'/> </data-member> @@ -3956,8 +3950,6 @@ <var-decl name='tv_nsec' type-id='03085adc' visibility='default'/> </data-member> </class-decl> - <typedef-decl name='DIR' type-id='20cd73f2' id='54a5d683'/> - <typedef-decl name='uintptr_t' type-id='7359adad' id='e475ab95'/> <pointer-type-def type-id='54a5d683' size-in-bits='64' id='f09217ba'/> <pointer-type-def type-id='aa12d1ba' size-in-bits='64' id='822cd80b'/> <qualified-type-def type-id='822cd80b' restrict='yes' id='e75a27e9'/> @@ -3966,7 +3958,6 @@ <pointer-type-def type-id='bb4788fa' size-in-bits='64' id='cecf4ea7'/> <pointer-type-def type-id='010ae0b9' size-in-bits='64' id='e4c6fa61'/> <pointer-type-def type-id='79bd3751' size-in-bits='64' id='c65a1f29'/> - <pointer-type-def type-id='d5027220' size-in-bits='64' id='b7f2d5e6'/> <pointer-type-def type-id='e4957c49' size-in-bits='64' id='924bbc81'/> <qualified-type-def type-id='924bbc81' const='yes' id='5499dcde'/> <pointer-type-def type-id='5499dcde' size-in-bits='64' id='2236d41c'/> @@ -4210,6 +4201,18 @@ <parameter type-id='f095e320' name='pco'/> <return type-id='95e97e5e'/> </function-decl> + <function-decl name='for_each_vdev_macro_helper_func' mangled-name='for_each_vdev_macro_helper_func' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='for_each_vdev_macro_helper_func'> + <parameter type-id='eaa32e2f' name='state'/> + <parameter type-id='5ce45b60' name='nv'/> + <parameter type-id='eaa32e2f' name='last_nv'/> + <return type-id='95e97e5e'/> + </function-decl> + <function-decl name='for_each_real_leaf_vdev_macro_helper_func' mangled-name='for_each_real_leaf_vdev_macro_helper_func' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='for_each_real_leaf_vdev_macro_helper_func'> + <parameter type-id='eaa32e2f' name='state'/> + <parameter type-id='5ce45b60' name='nv'/> + <parameter type-id='eaa32e2f' name='last_nv'/> + <return type-id='95e97e5e'/> + </function-decl> <function-decl name='for_each_vdev_cb' mangled-name='for_each_vdev_cb' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='for_each_vdev_cb'> <parameter type-id='eaa32e2f' name='zhp'/> <parameter type-id='5ce45b60' name='nv'/> @@ -4268,25 +4271,6 @@ <parameter type-id='18c91f9e'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='fprintf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='e75a27e9'/> - <parameter type-id='9d26089a'/> - <parameter is-variadic='yes'/> - <return type-id='95e97e5e'/> - </function-decl> - <function-decl name='vsnprintf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='26a90f95'/> - <parameter type-id='b59d7dce'/> - <parameter type-id='80f4b756'/> - <parameter type-id='b7f2d5e6'/> - <return type-id='95e97e5e'/> - </function-decl> - <function-decl name='asprintf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='8c85230f'/> - <parameter type-id='9d26089a'/> - <parameter is-variadic='yes'/> - <return type-id='95e97e5e'/> - </function-decl> <function-decl name='strtoull' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='9d26089a'/> <parameter type-id='8c85230f'/> @@ -4308,11 +4292,6 @@ <parameter type-id='95e97e5e'/> <return type-id='48b5725f'/> </function-decl> - <function-decl name='realpath' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='9d26089a'/> - <parameter type-id='266fe297'/> - <return type-id='26a90f95'/> - </function-decl> <function-decl name='strncmp' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='80f4b756'/> <parameter type-id='80f4b756'/> @@ -4328,30 +4307,51 @@ <parameter type-id='95e97e5e'/> <return type-id='26a90f95'/> </function-decl> - <function-decl name='ioctl' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='sysconf' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='95e97e5e'/> - <parameter type-id='7359adad'/> + <return type-id='bd54fe1a'/> + </function-decl> + <function-decl name='geteuid' visibility='default' binding='global' size-in-bits='64'> + <return type-id='cc5fcceb'/> + </function-decl> + <function-decl name='__fprintf_chk' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='e75a27e9'/> + <parameter type-id='95e97e5e'/> + <parameter type-id='9d26089a'/> <parameter is-variadic='yes'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='fstat64' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='__asprintf_chk' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='8c85230f'/> <parameter type-id='95e97e5e'/> - <parameter type-id='62f7a03d'/> + <parameter type-id='9d26089a'/> + <parameter is-variadic='yes'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='pread64' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='__realpath_chk' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='9d26089a'/> + <parameter type-id='266fe297'/> + <parameter type-id='b59d7dce'/> + <return type-id='26a90f95'/> + </function-decl> + <function-decl name='__pread64_chk' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='95e97e5e'/> <parameter type-id='eaa32e2f'/> <parameter type-id='b59d7dce'/> <parameter type-id='724e4de6'/> + <parameter type-id='b59d7dce'/> <return type-id='79a0948f'/> </function-decl> - <function-decl name='sysconf' visibility='default' binding='global' size-in-bits='64'> + <function-decl name='ioctl' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='95e97e5e'/> - <return type-id='bd54fe1a'/> + <parameter type-id='7359adad'/> + <parameter is-variadic='yes'/> + <return type-id='95e97e5e'/> </function-decl> - <function-decl name='geteuid' visibility='default' binding='global' size-in-bits='64'> - <return type-id='cc5fcceb'/> + <function-decl name='fstat64' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='95e97e5e'/> + <parameter type-id='62f7a03d'/> + <return type-id='95e97e5e'/> </function-decl> <function-type size-in-bits='64' id='baa42fef'> <parameter type-id='eaa32e2f'/> @@ -4430,6 +4430,9 @@ <parameter type-id='b59d7dce' name='buflen'/> <return type-id='48b5725f'/> </function-decl> + <function-decl name='__ctype_b_loc' visibility='default' binding='global' size-in-bits='64'> + <return type-id='c59e1ef0'/> + </function-decl> <function-decl name='powl' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='e095c704'/> <parameter type-id='e095c704'/> @@ -4439,14 +4442,12 @@ <parameter type-id='a0eb0f08'/> <return type-id='a0eb0f08'/> </function-decl> - <function-decl name='__ctype_b_loc' visibility='default' binding='global' size-in-bits='64'> - <return type-id='c59e1ef0'/> - </function-decl> </abi-instr> <abi-instr address-size='64' path='zutil_pool.c' language='LANG_C99'> <array-type-def dimensions='1' type-id='853fd5dc' size-in-bits='32768' id='b505fc2f'> <subrange length='64' type-id='7359adad' id='b10be967'/> </array-type-def> + <type-decl name='float' size-in-bits='32' id='a6c45d85'/> <class-decl name='ddt_stat' size-in-bits='512' is-struct='yes' visibility='default' id='65242dfe'> <data-member access='public' layout-offset-in-bits='0'> <var-decl name='dds_blocks' type-id='9c313c2d' visibility='default'/> @@ -4484,6 +4485,9 @@ <pointer-type-def type-id='ec92d602' size-in-bits='64' id='932720f8'/> <qualified-type-def type-id='853fd5dc' const='yes' id='764c298c'/> <pointer-type-def type-id='764c298c' size-in-bits='64' id='dfe59052'/> + <qualified-type-def type-id='a9c79a1f' const='yes' id='cd087e36'/> + <pointer-type-def type-id='cd087e36' size-in-bits='64' id='e05e8614'/> + <pointer-type-def type-id='a9c79a1f' size-in-bits='64' id='3d83ba87'/> <function-decl name='zpool_dump_ddt' mangled-name='zpool_dump_ddt' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zpool_dump_ddt'> <parameter type-id='dfe59052' name='dds_total'/> <parameter type-id='932720f8' name='ddh'/> @@ -4497,9 +4501,13 @@ <parameter type-id='4dd26a40' name='numrecords'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='printf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='80f4b756'/> - <parameter is-variadic='yes'/> + <function-decl name='fsleep' mangled-name='fsleep' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='fsleep'> + <parameter type-id='a6c45d85' name='sec'/> + <return type-id='48b5725f'/> + </function-decl> + <function-decl name='zpool_getenv_int' mangled-name='zpool_getenv_int' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zpool_getenv_int'> + <parameter type-id='80f4b756' name='env'/> + <parameter type-id='95e97e5e' name='default_val'/> <return type-id='95e97e5e'/> </function-decl> <function-decl name='realloc' visibility='default' binding='global' size-in-bits='64'> @@ -4507,5 +4515,16 @@ <parameter type-id='b59d7dce'/> <return type-id='eaa32e2f'/> </function-decl> + <function-decl name='nanosleep' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='e05e8614'/> + <parameter type-id='3d83ba87'/> + <return type-id='95e97e5e'/> + </function-decl> + <function-decl name='__printf_chk' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='95e97e5e'/> + <parameter type-id='80f4b756'/> + <parameter is-variadic='yes'/> + <return type-id='95e97e5e'/> + </function-decl> </abi-instr> </abi-corpus> diff --git a/lib/libzfsbootenv/libzfsbootenv.abi b/lib/libzfsbootenv/libzfsbootenv.abi index 0ddd41d0630e..8a7fa782d8a0 100644 --- a/lib/libzfsbootenv/libzfsbootenv.abi +++ b/lib/libzfsbootenv/libzfsbootenv.abi @@ -1,16 +1,7 @@ <abi-corpus version='2.0' architecture='elf-amd-x86_64' soname='libzfsbootenv.so.1'> <elf-needed> <dependency name='libzfs.so.4'/> - <dependency name='libzfs_core.so.3'/> - <dependency name='libuuid.so.1'/> - <dependency name='libblkid.so.1'/> - <dependency name='libudev.so.1'/> - <dependency name='libuutil.so.3'/> - <dependency name='libm.so.6'/> - <dependency name='libcrypto.so.1.1'/> - <dependency name='libz.so.1'/> <dependency name='libnvpair.so.3'/> - <dependency name='libtirpc.so.3'/> <dependency name='libc.so.6'/> </elf-needed> <elf-function-symbols> @@ -289,18 +280,6 @@ <parameter type-id='9b23c9ad' name='device'/> <return type-id='95e97e5e'/> </function-decl> - <function-decl name='fprintf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='e75a27e9'/> - <parameter type-id='9d26089a'/> - <parameter is-variadic='yes'/> - <return type-id='95e97e5e'/> - </function-decl> - <function-decl name='asprintf' visibility='default' binding='global' size-in-bits='64'> - <parameter type-id='8c85230f'/> - <parameter type-id='9d26089a'/> - <parameter is-variadic='yes'/> - <return type-id='95e97e5e'/> - </function-decl> <function-decl name='free' visibility='default' binding='global' size-in-bits='64'> <parameter type-id='eaa32e2f'/> <return type-id='48b5725f'/> @@ -319,6 +298,20 @@ <parameter type-id='80f4b756'/> <return type-id='b59d7dce'/> </function-decl> + <function-decl name='__fprintf_chk' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='e75a27e9'/> + <parameter type-id='95e97e5e'/> + <parameter type-id='9d26089a'/> + <parameter is-variadic='yes'/> + <return type-id='95e97e5e'/> + </function-decl> + <function-decl name='__asprintf_chk' visibility='default' binding='global' size-in-bits='64'> + <parameter type-id='8c85230f'/> + <parameter type-id='95e97e5e'/> + <parameter type-id='9d26089a'/> + <parameter is-variadic='yes'/> + <return type-id='95e97e5e'/> + </function-decl> </abi-instr> <abi-instr address-size='64' path='lzbe_pair.c' language='LANG_C99'> <type-decl name='short int' size-in-bits='16' id='a2185560'/> diff --git a/lib/libzutil/os/freebsd/zutil_import_os.c b/lib/libzutil/os/freebsd/zutil_import_os.c index 7c48e06f9315..a57a2db57eba 100644 --- a/lib/libzutil/os/freebsd/zutil_import_os.c +++ b/lib/libzutil/os/freebsd/zutil_import_os.c @@ -249,6 +249,23 @@ zfs_dev_flush(int fd __unused) } void +update_vdev_config_dev_sysfs_path(nvlist_t *nv, const char *path, + const char *key) +{ + (void) nv; + (void) path; + (void) key; +} + +void update_vdevs_config_dev_sysfs_path(nvlist_t *config) { } + +int +zpool_disk_wait(const char *path) +{ + + (void) path; + return (ENOTSUP); +} diff --git a/lib/libzutil/os/linux/zutil_import_os.c b/lib/libzutil/os/linux/zutil_import_os.c index 6c406d373a0c..ebf20956a213 100644 --- a/lib/libzutil/os/linux/zutil_import_os.c +++ b/lib/libzutil/os/linux/zutil_import_os.c @@ -188,25 +188,17 @@ zpool_open_func(void *arg) if (rn->rn_labelpaths) { char *path = NULL; char *devid = NULL; - char *env = NULL; rdsk_node_t *slice; avl_index_t where; - int timeout; int error; if (label_paths(rn->rn_hdl, rn->rn_config, &path, &devid)) return; - env = getenv("ZPOOL_IMPORT_UDEV_TIMEOUT_MS"); - if ((env == NULL) || sscanf(env, "%d", &timeout) != 1 || - timeout < 0) { - timeout = DISK_LABEL_WAIT; - } - /* * Allow devlinks to stabilize so all paths are available. */ - zpool_label_disk_wait(rn->rn_name, timeout); + zpool_disk_wait(rn->rn_name); if (path != NULL) { slice = zutil_alloc(hdl, sizeof (rdsk_node_t)); @@ -701,6 +693,20 @@ zpool_label_disk_wait(const char *path, int timeout_ms) } /* + * Simplified version of zpool_label_disk_wait() where we wait for a device + * to appear using the default timeouts. + */ +int +zpool_disk_wait(const char *path) +{ + int timeout; + timeout = zpool_getenv_int("ZPOOL_IMPORT_UDEV_TIMEOUT_MS", + DISK_LABEL_WAIT); + + return (zpool_label_disk_wait(path, timeout)); +} + +/* * Encode the persistent devices strings * used for the vdev disk label */ @@ -781,20 +787,37 @@ no_dev: * Rescan the enclosure sysfs path for turning on enclosure LEDs and store it * in the nvlist * (if applicable). Like: * vdev_enc_sysfs_path: '/sys/class/enclosure/11:0:1:0/SLOT 4' + * + * If an old path was in the nvlist, and the rescan can not find a new path, + * then keep the old path, since the disk may have been removed. + * + * path: The vdev path (value from ZPOOL_CONFIG_PATH) + * key: The nvlist_t name (like ZPOOL_CONFIG_VDEV_ENC_SYSFS_PATH) */ -static void -update_vdev_config_dev_sysfs_path(nvlist_t *nv, char *path) +void +update_vdev_config_dev_sysfs_path(nvlist_t *nv, const char *path, + const char *key) { char *upath, *spath; + char *oldpath = NULL; + + (void) nvlist_lookup_string(nv, key, &oldpath); /* Add enclosure sysfs path (if disk is in an enclosure). */ upath = zfs_get_underlying_path(path); spath = zfs_get_enclosure_sysfs_path(upath); if (spath) { - nvlist_add_string(nv, ZPOOL_CONFIG_VDEV_ENC_SYSFS_PATH, spath); + (void) nvlist_add_string(nv, key, spath); } else { - nvlist_remove_all(nv, ZPOOL_CONFIG_VDEV_ENC_SYSFS_PATH); + /* + * We couldn't dynamically scan the disk's enclosure sysfs path. + * This could be because the disk went away. If there's an old + * enclosure sysfs path in the nvlist, then keep using it. + */ + if (!oldpath) { + (void) nvlist_remove_all(nv, key); + } } free(upath); @@ -812,7 +835,8 @@ sysfs_path_pool_vdev_iter_f(void *hdl_data, nvlist_t *nv, void *data) return (1); /* Rescan our enclosure sysfs path for this vdev */ - update_vdev_config_dev_sysfs_path(nv, path); + update_vdev_config_dev_sysfs_path(nv, path, + ZPOOL_CONFIG_VDEV_ENC_SYSFS_PATH); return (0); } @@ -901,7 +925,8 @@ update_vdev_config_dev_strs(nvlist_t *nv) (void) nvlist_add_string(nv, ZPOOL_CONFIG_PHYS_PATH, vds.vds_devphys); } - update_vdev_config_dev_sysfs_path(nv, path); + update_vdev_config_dev_sysfs_path(nv, path, + ZPOOL_CONFIG_VDEV_ENC_SYSFS_PATH); } else { /* Clear out any stale entries. */ (void) nvlist_remove_all(nv, ZPOOL_CONFIG_DEVID); diff --git a/lib/libzutil/zutil_import.c b/lib/libzutil/zutil_import.c index 98f138957ba6..b9a0b67f2b8c 100644 --- a/lib/libzutil/zutil_import.c +++ b/lib/libzutil/zutil_import.c @@ -1836,6 +1836,104 @@ zpool_find_config(void *hdl, const char *target, nvlist_t **configp, return (0); } +/* Return if a vdev is a leaf vdev. Note: draid spares are leaf vdevs. */ +static boolean_t +vdev_is_leaf(nvlist_t *nv) +{ + uint_t children = 0; + nvlist_t **child; + + (void) nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_CHILDREN, + &child, &children); + + return (children == 0); +} + +/* Return if a vdev is a leaf vdev and a real device (disk or file) */ +static boolean_t +vdev_is_real_leaf(nvlist_t *nv) +{ + char *type = NULL; + if (!vdev_is_leaf(nv)) + return (B_FALSE); + + (void) nvlist_lookup_string(nv, ZPOOL_CONFIG_TYPE, &type); + if ((strcmp(type, VDEV_TYPE_DISK) == 0) || + (strcmp(type, VDEV_TYPE_FILE) == 0)) { + return (B_TRUE); + } + + return (B_FALSE); +} + +/* + * This function is called by our FOR_EACH_VDEV() macros. + * + * state: State machine status (stored inside of a (nvlist_t *)) + * nv: The current vdev nvlist_t we are iterating over. + * last_nv: The previous vdev nvlist_t we returned to the user in + * the last iteration of FOR_EACH_VDEV(). We use it + * to find the next vdev nvlist_t we should return. + * real_leaves_only: Only return leaf vdevs. + * + * Returns 1 if we found the next vdev nvlist_t for this iteration. 0 if + * we're still searching for it. + */ +static int +__for_each_vdev_macro_helper_func(void *state, nvlist_t *nv, void *last_nv, + boolean_t real_leaves_only) +{ + enum {FIRST_NV = 0, NEXT_IS_MATCH = 1, STOP_LOOKING = 2}; + + /* The very first entry in the NV list is a special case */ + if (*((nvlist_t **)state) == (nvlist_t *)FIRST_NV) { + if (real_leaves_only && !vdev_is_real_leaf(nv)) + return (0); + + *((nvlist_t **)last_nv) = nv; + *((nvlist_t **)state) = (nvlist_t *)STOP_LOOKING; + return (1); + } + + /* + * We came across our last_nv, meaning the next one is the one we + * want + */ + if (nv == *((nvlist_t **)last_nv)) { + /* Next iteration of this function will return the nvlist_t */ + *((nvlist_t **)state) = (nvlist_t *)NEXT_IS_MATCH; + return (0); + } + + /* + * We marked NEXT_IS_MATCH on the previous iteration, so this is the one + * we want. + */ + if (*(nvlist_t **)state == (nvlist_t *)NEXT_IS_MATCH) { + if (real_leaves_only && !vdev_is_real_leaf(nv)) + return (0); + + *((nvlist_t **)last_nv) = nv; + *((nvlist_t **)state) = (nvlist_t *)STOP_LOOKING; + return (1); + } + + return (0); +} + +int +for_each_vdev_macro_helper_func(void *state, nvlist_t *nv, void *last_nv) +{ + return (__for_each_vdev_macro_helper_func(state, nv, last_nv, B_FALSE)); +} + +int +for_each_real_leaf_vdev_macro_helper_func(void *state, nvlist_t *nv, + void *last_nv) +{ + return (__for_each_vdev_macro_helper_func(state, nv, last_nv, B_TRUE)); +} + /* * Internal function for iterating over the vdevs. * diff --git a/lib/libzutil/zutil_pool.c b/lib/libzutil/zutil_pool.c index 734650f3cffc..eeb7c589ee05 100644 --- a/lib/libzutil/zutil_pool.c +++ b/lib/libzutil/zutil_pool.c @@ -28,6 +28,7 @@ #include <string.h> #include <sys/nvpair.h> #include <sys/fs/zfs.h> +#include <math.h> #include <libzutil.h> @@ -143,3 +144,33 @@ zpool_history_unpack(char *buf, uint64_t bytes_read, uint64_t *leftover, *leftover = bytes_read; return (0); } + +/* + * Floating point sleep(). Allows you to pass in a floating point value for + * seconds. + */ +void +fsleep(float sec) +{ + struct timespec req; + req.tv_sec = floor(sec); + req.tv_nsec = (sec - (float)req.tv_sec) * NANOSEC; + nanosleep(&req, NULL); +} + +/* + * Get environment variable 'env' and return it as an integer. + * If 'env' is not set, then return 'default_val' instead. + */ +int +zpool_getenv_int(const char *env, int default_val) +{ + char *str; + int val; + str = getenv(env); + if ((str == NULL) || sscanf(str, "%d", &val) != 1 || + val < 0) { + val = default_val; + } + return (val); +} diff --git a/man/Makefile.am b/man/Makefile.am index 64650c2b988a..2608461625d4 100644 --- a/man/Makefile.am +++ b/man/Makefile.am @@ -104,7 +104,8 @@ dist_man_MANS = \ nodist_man_MANS = \ man8/zed.8 \ - man8/zfs-mount-generator.8 + man8/zfs-mount-generator.8 \ + man8/zfs_prepare_disk.8 SUBSTFILES += $(nodist_man_MANS) diff --git a/man/man4/zfs.4 b/man/man4/zfs.4 index 71a95c3bd812..d8a62384fd61 100644 --- a/man/man4/zfs.4 +++ b/man/man4/zfs.4 @@ -1712,7 +1712,7 @@ completes in order to verify the checksums of all blocks which have been resilvered. This is enabled by default and strongly recommended. . -.It Sy zfs_rebuild_vdev_limit Ns = Ns Sy 33554432 Ns B Po 32MB Pc Pq ulong +.It Sy zfs_rebuild_vdev_limit Ns = Ns Sy 67108864 Ns B Po 64 MiB Pc Pq ulong Maximum amount of I/O that can be concurrently issued for a sequential resilver per leaf device, given in bytes. . @@ -1831,6 +1831,13 @@ When we cross this limit from above it is because we are issuing verification I/ In this case (unless the metadata scan is done) we stop issuing verification I/O and start scanning metadata again until we get to the hard limit. . +.It Sy zfs_scan_report_txgs Ns = Ns Sy 0 Ns | Ns 1 Pq uint +When reporting resilver throughput and estimated completion time use the +performance observed over roughly the last +.Sy zfs_scan_report_txgs +TXGs. +When set to zero performance is calculated over the time between checkpoints. +. .It Sy zfs_scan_strict_mem_lim Ns = Ns Sy 0 Ns | Ns 1 Pq int Enforce tight memory limits on pool scans when a sequential scan is in progress. When disabled, the memory limit may be exceeded by fast disks. @@ -1839,7 +1846,7 @@ When disabled, the memory limit may be exceeded by fast disks. Freezes a scrub/resilver in progress without actually pausing it. Intended for testing/debugging. . -.It Sy zfs_scan_vdev_limit Ns = Ns Sy 4194304 Ns B Po 4MB Pc Pq int +.It Sy zfs_scan_vdev_limit Ns = Ns Sy 16777216 Ns B Po 16 MiB Pc Pq int Maximum amount of data that can be concurrently issued at once for scrubs and resilvers per leaf device, given in bytes. . @@ -2209,6 +2216,16 @@ If .Sy 0 , generate a system-dependent value close to 6 threads per taskq. . +.It Sy zio_taskq_read Ns = Ns Sy fixed,1,8 null scale null Pq charp +Set the queue and thread configuration for the IO read queues. +This is an advanced debugging parameter. +Don't change this unless you understand what it does. +. +.It Sy zio_taskq_write Ns = Ns Sy batch fixed,1,5 scale fixed,1,5 Pq charp +Set the queue and thread configuration for the IO write queues. +This is an advanced debugging parameter. +Don't change this unless you understand what it does. +. .It Sy zvol_inhibit_dev Ns = Ns Sy 0 Ns | Ns 1 Pq uint Do not create zvol device nodes. This may slightly improve startup time on diff --git a/man/man8/.gitignore b/man/man8/.gitignore index f2fc702147e9..a468f9cbf9d3 100644 --- a/man/man8/.gitignore +++ b/man/man8/.gitignore @@ -1,2 +1,3 @@ /zed.8 /zfs-mount-generator.8 +/zfs_prepare_disk.8 diff --git a/man/man8/zfs.8 b/man/man8/zfs.8 index 23220b7f3ee6..df6029096016 100644 --- a/man/man8/zfs.8 +++ b/man/man8/zfs.8 @@ -545,7 +545,7 @@ access for a set of IP addresses and to enable root access for system on the .Ar tank/home file system: -.Dl # Nm zfs Cm set Sy sharenfs Ns = Ns ' Ns Ar rw Ns =@123.123.0.0/16,root= Ns Ar neo Ns ' tank/home +.Dl # Nm zfs Cm set Sy sharenfs Ns = Ns ' Ns Ar rw Ns =@123.123.0.0/16:[::1],root= Ns Ar neo Ns ' tank/home .Pp If you are using DNS for host name resolution, specify the fully-qualified hostname. diff --git a/man/man8/zfs_prepare_disk.8.in b/man/man8/zfs_prepare_disk.8.in new file mode 100644 index 000000000000..2a741531e415 --- /dev/null +++ b/man/man8/zfs_prepare_disk.8.in @@ -0,0 +1,70 @@ +.\" +.\" Developed at Lawrence Livermore National Laboratory (LLNL-CODE-403049). +.\" Copyright (C) 2023 Lawrence Livermore National Security, LLC. +.\" Refer to the OpenZFS git commit log for authoritative copyright attribution. +.\" +.\" The contents of this file are subject to the terms of the +.\" Common Development and Distribution License Version 1.0 (CDDL-1.0). +.\" You can obtain a copy of the license from the top-level file +.\" "OPENSOLARIS.LICENSE" or at <http://opensource.org/licenses/CDDL-1.0>. +.\" You may not use this file except in compliance with the license. +.\" +.\" Developed at Lawrence Livermore National Laboratory (LLNL-CODE-403049) +.\" +.Dd August 30, 2023 +.Dt ZFS_PREPARE_DISK 8 +.Os +. +.Sh NAME +.Nm zfs_prepare_disk +.Nd special script that gets run before bringing a disk into a pool +.Sh DESCRIPTION +.Nm +is an optional script that gets called by libzfs before bringing a disk into a +pool. +It can be modified by the user to run whatever commands are necessary to prepare +a disk for inclusion into the pool. +For example, users can add lines to +.Nm zfs_prepare_disk +to do things like update the drive's firmware or check the drive's health. +.Nm zfs_prepare_disk +is optional and can be removed if not needed. +libzfs will look for the script at @zfsexecdir@/zfs_prepare_disk. +. +.Ss Properties +.Nm zfs_prepare_disk +will be passed the following environment variables: +.sp +.Bl -tag -compact -width "VDEV_ENC_SYSFS_PATH" +. +.It Nm POOL_NAME +.No Name of the pool +.It Nm VDEV_PATH +.No Path to the disk (like /dev/sda) +.It Nm VDEV_PREPARE +.No Reason why the disk is being prepared for inclusion +('create', 'add', 'replace', or 'autoreplace'). +This can be useful if you only want the script to be run under certain actions. +.It Nm VDEV_UPATH +.No Path to one of the underlying devices for the +disk. +For multipath this would return one of the /dev/sd* paths to the disk. +If the device is not a device mapper device, then +.Nm VDEV_UPATH +just returns the same value as +.Nm VDEV_PATH +.It Nm VDEV_ENC_SYSFS_PATH +.No Path to the disk's enclosure sysfs path, if available +.El +.Pp +Note that some of these variables may have a blank value. +.Nm POOL_NAME +is blank at pool creation time, for example. +.Sh ENVIRONMENT +.Nm zfs_prepare_disk +runs with a limited $PATH. +.Sh EXIT STATUS +.Nm zfs_prepare_disk +should return 0 on success, non-zero otherwise. +If non-zero is returned, the disk will not be included in the pool. +. diff --git a/man/man8/zpool-clear.8 b/man/man8/zpool-clear.8 index 0b256b28bd21..19861a319000 100644 --- a/man/man8/zpool-clear.8 +++ b/man/man8/zpool-clear.8 @@ -36,6 +36,7 @@ .Sh SYNOPSIS .Nm zpool .Cm clear +.Op Fl -power .Ar pool .Oo Ar device Oc Ns … . @@ -52,6 +53,16 @@ Pools with enabled which have been suspended cannot be resumed. While the pool was suspended, it may have been imported on another host, and resuming I/O could result in pool damage. +.Bl -tag -width Ds +.It Fl -power +Power on the devices's slot in the storage enclosure and wait for the device +to show up before attempting to clear errors. +This is done on all the devices specified. +Alternatively, you can set the +.Sy ZPOOL_AUTO_POWER_ON_SLOT +environment variable to always enable this behavior. +Note: This flag currently works on Linux only. +.El . .Sh SEE ALSO .Xr zdb 8 , diff --git a/man/man8/zpool-initialize.8 b/man/man8/zpool-initialize.8 index 0a108180dbbe..ada00bb1be90 100644 --- a/man/man8/zpool-initialize.8 +++ b/man/man8/zpool-initialize.8 @@ -36,7 +36,7 @@ .Sh SYNOPSIS .Nm zpool .Cm initialize -.Op Fl c Ns | Ns Fl s +.Op Fl c Ns | Ns Fl s | Ns Fl u .Op Fl w .Ar pool .Oo Ar device Oc Ns … @@ -60,6 +60,14 @@ initialized, the command will fail and no suspension will occur on any device. Initializing can then be resumed by running .Nm zpool Cm initialize with no flags on the relevant target devices. +.It Fl u , -uninit +Clears the initialization state on the specified devices, or all eligible +devices if none are specified. +If the devices are being actively initialized the command will fail. +After being cleared +.Nm zpool Cm initialize +with no flags can be used to re-initialize all unallocoated regions on +the relevant target devices. .It Fl w , -wait Wait until the devices have finished initializing before returning. .El diff --git a/man/man8/zpool-offline.8 b/man/man8/zpool-offline.8 index 9b2cf59cf414..011cefed2f13 100644 --- a/man/man8/zpool-offline.8 +++ b/man/man8/zpool-offline.8 @@ -36,12 +36,13 @@ .Sh SYNOPSIS .Nm zpool .Cm offline -.Op Fl ft +.Op Fl Sy -power Ns | Ns Op Fl Sy ft .Ar pool .Ar device Ns … .Nm zpool .Cm online -.Op Fl e +.Op Fl Sy -power +.Op Fl Sy e .Ar pool .Ar device Ns … . @@ -50,7 +51,7 @@ .It Xo .Nm zpool .Cm offline -.Op Fl ft +.Op Fl Sy -power Ns | Ns Op Fl Sy ft .Ar pool .Ar device Ns … .Xc @@ -60,6 +61,9 @@ While the is offline, no attempt is made to read or write to the device. This command is not applicable to spares. .Bl -tag -width Ds +.It Fl -power +Power off the device's slot in the storage enclosure. +This flag currently works on Linux only .It Fl f Force fault. Instead of offlining the disk, put it into a faulted state. @@ -73,6 +77,7 @@ Upon reboot, the specified physical device reverts to its previous state. .It Xo .Nm zpool .Cm online +.Op Fl -power .Op Fl e .Ar pool .Ar device Ns … @@ -80,6 +85,13 @@ Upon reboot, the specified physical device reverts to its previous state. Brings the specified physical device online. This command is not applicable to spares. .Bl -tag -width Ds +.It Fl -power +Power on the device's slot in the storage enclosure and wait for the device +to show up before attempting to online it. +Alternatively, you can set the +.Sy ZPOOL_AUTO_POWER_ON_SLOT +environment variable to always enable this behavior. +This flag currently works on Linux only .It Fl e Expand the device to use all available space. If the device is part of a mirror or raidz then all devices must be expanded diff --git a/man/man8/zpool-status.8 b/man/man8/zpool-status.8 index 7c825f69d8e2..11d0696ad5eb 100644 --- a/man/man8/zpool-status.8 +++ b/man/man8/zpool-status.8 @@ -36,7 +36,7 @@ .Sh SYNOPSIS .Nm zpool .Cm status -.Op Fl DigLpPstvx +.Op Fl DeigLpPstvx .Op Fl T Sy u Ns | Ns Sy d .Op Fl c Op Ar SCRIPT1 Ns Oo , Ns Ar SCRIPT2 Oc Ns … .Oo Ar pool Oc Ns … @@ -57,6 +57,8 @@ and the estimated time to completion. Both of these are only approximate, because the amount of data in the pool and the other workloads on the system can change. .Bl -tag -width Ds +.It Fl -power +Display vdev enclosure slot power status (on or off). .It Fl c Op Ar SCRIPT1 Ns Oo , Ns Ar SCRIPT2 Oc Ns … Run a script (or scripts) on each vdev and include the output as a new column in the @@ -67,6 +69,8 @@ See the option of .Nm zpool Cm iostat for complete details. +.It Fl e +Only show unhealthy vdevs (not-ONLINE or with errors). .It Fl i Display vdev initialization status. .It Fl g diff --git a/man/man8/zpool.8 b/man/man8/zpool.8 index e5d7c8515177..591c7772c749 100644 --- a/man/man8/zpool.8 +++ b/man/man8/zpool.8 @@ -424,7 +424,7 @@ rpool 14.6G 54.9G 4 55 250K 2.69M .El . .Sh ENVIRONMENT VARIABLES -.Bl -tag -compact -width "ZPOOL_IMPORT_UDEV_TIMEOUT_MS" +.Bl -tag -compact -width "ZPOOL_STATUS_NON_NATIVE_ASHIFT_IGNORE" .It Sy ZFS_ABORT Cause .Nm @@ -436,6 +436,23 @@ Use ANSI color in and .Nm zpool iostat output. +.It Sy ZPOOL_AUTO_POWER_ON_SLOT +Automatically attempt to turn on the drives enclosure slot power to a drive when +running the +.Nm zpool Cm online +or +.Nm zpool Cm clear +commands. +This has the same effect as passing the +.Fl -power +option to those commands. +.It Sy ZPOOL_POWER_ON_SLOT_TIMEOUT_MS +The maximum time in milliseconds to wait for a slot power sysfs value +to return the correct value after writing it. +For example, after writing "on" to the sysfs enclosure slot power_control file, +it can take some time for the enclosure to power down the slot and return +"on" if you read back the 'power_control' value. +Defaults to 30 seconds (30000ms) if not set. .It Sy ZPOOL_IMPORT_PATH The search path for devices or files to use with the pool. This is a colon-separated list of directories in which diff --git a/module/Kbuild.in b/module/Kbuild.in index 1507965c5750..7675d614fa34 100644 --- a/module/Kbuild.in +++ b/module/Kbuild.in @@ -44,4 +44,5 @@ endif subdir-asflags-y := $(ZFS_MODULE_CFLAGS) $(ZFS_MODULE_CPPFLAGS) subdir-ccflags-y := $(ZFS_MODULE_CFLAGS) $(ZFS_MODULE_CPPFLAGS) + endif diff --git a/module/icp/Makefile.in b/module/icp/Makefile.in index ce84999ad1cf..0c5cb7c1fe12 100644 --- a/module/icp/Makefile.in +++ b/module/icp/Makefile.in @@ -69,6 +69,8 @@ OBJECT_FILES_NON_STANDARD_aesni-gcm-x86_64.o := y OBJECT_FILES_NON_STANDARD_sha256_impl.o := y OBJECT_FILES_NON_STANDARD_sha512_impl.o := y +UBSAN_SANITIZE_modhash.o := n + ICP_DIRS = \ api \ core \ diff --git a/module/icp/algs/edonr/edonr.c b/module/icp/algs/edonr/edonr.c index 7a3ba30c0582..baf8bb885a6c 100644 --- a/module/icp/algs/edonr/edonr.c +++ b/module/icp/algs/edonr/edonr.c @@ -343,9 +343,11 @@ Q256(size_t bitlen, const uint32_t *data, uint32_t *restrict p) * which only goes over it by a hair (1248 bytes on ARM32). */ #include <sys/isa_defs.h> /* for _ILP32 */ -#ifdef _ILP32 /* We're 32-bit, assume small stack frames */ +#if defined(_ILP32) /* We're 32-bit, assume small stack frames */ +#if defined(__GNUC__) && !defined(__clang__) #pragma GCC diagnostic ignored "-Wframe-larger-than=" #endif +#endif #if defined(__IBMC__) && defined(_AIX) && defined(__64BIT__) static inline size_t diff --git a/module/icp/algs/skein/skein_block.c b/module/icp/algs/skein/skein_block.c index 7ba165a48511..3ad52da5f6a3 100644 --- a/module/icp/algs/skein/skein_block.c +++ b/module/icp/algs/skein/skein_block.c @@ -30,7 +30,9 @@ * the #pragma here to ignore the warning. */ #if defined(_ILP32) || defined(__powerpc) /* Assume small stack */ +#if defined(__GNUC__) && !defined(__clang__) #pragma GCC diagnostic ignored "-Wframe-larger-than=" +#endif /* * We're running on 32-bit, don't unroll loops to save stack frame space * diff --git a/module/lua/ldebug.c b/module/lua/ldebug.c index da005c44376e..c724e4632ee5 100644 --- a/module/lua/ldebug.c +++ b/module/lua/ldebug.c @@ -112,10 +112,11 @@ static const char *upvalname (Proto *p, int uv) { static const char *findvararg (CallInfo *ci, int n, StkId *pos) { int nparams = clLvalue(ci->func)->p->numparams; - if (n >= ci->u.l.base - ci->func - nparams) + int nvararg = cast_int(ci->u.l.base - ci->func) - nparams; + if (n <= -nvararg) return NULL; /* no such vararg */ else { - *pos = ci->func + nparams + n; + *pos = ci->func + nparams - n; return "(*vararg)"; /* generic name for any vararg */ } } @@ -127,7 +128,7 @@ static const char *findlocal (lua_State *L, CallInfo *ci, int n, StkId base; if (isLua(ci)) { if (n < 0) /* access to vararg values? */ - return findvararg(ci, -n, pos); + return findvararg(ci, n, pos); else { base = ci->u.l.base; name = luaF_getlocalname(ci_func(ci)->p, n, currentpc(ci)); diff --git a/module/lua/ldo.c b/module/lua/ldo.c index a9835c4f571d..1bb9cda2abc6 100644 --- a/module/lua/ldo.c +++ b/module/lua/ldo.c @@ -171,7 +171,8 @@ static void seterrorobj (lua_State *L, int errcode, StkId oldtop) { /* * Silence infinite recursion warning which was added to -Wall in gcc 12.1 */ -#if defined(HAVE_INFINITE_RECURSION) +#if defined(__GNUC__) && !defined(__clang__) && \ + defined(HAVE_KERNEL_INFINITE_RECURSION) #pragma GCC diagnostic push #pragma GCC diagnostic ignored "-Winfinite-recursion" #endif @@ -197,7 +198,8 @@ l_noret luaD_throw (lua_State *L, int errcode) { } } -#if defined(HAVE_INFINITE_RECURSION) +#if defined(__GNUC__) && !defined(__clang__) && \ + defined(HAVE_INFINITE_RECURSION) #pragma GCC diagnostic pop #endif diff --git a/module/lua/lfunc.h b/module/lua/lfunc.h index 59a4fa75c46e..638971bdd055 100644 --- a/module/lua/lfunc.h +++ b/module/lua/lfunc.h @@ -13,10 +13,10 @@ #define sizeCclosure(n) (cast(int, sizeof(CClosure)) + \ - cast(int, sizeof(TValue)*((n)-1))) + cast(int, sizeof(TValue)*((n)))) #define sizeLclosure(n) (cast(int, sizeof(LClosure)) + \ - cast(int, sizeof(TValue *)*((n)-1))) + cast(int, sizeof(TValue *)*((n)))) LUAI_FUNC Proto *luaF_newproto (lua_State *L); diff --git a/module/lua/lobject.h b/module/lua/lobject.h index a16b8d62eb4b..ede697cfc51e 100644 --- a/module/lua/lobject.h +++ b/module/lua/lobject.h @@ -514,14 +514,14 @@ typedef struct UpVal { typedef struct CClosure { ClosureHeader; lua_CFunction f; - TValue upvalue[1]; /* list of upvalues */ + TValue upvalue[]; /* list of upvalues */ } CClosure; typedef struct LClosure { ClosureHeader; struct Proto *p; - UpVal *upvals[1]; /* list of upvalues */ + UpVal *upvals[]; /* list of upvalues */ } LClosure; diff --git a/module/os/freebsd/zfs/zfs_ctldir.c b/module/os/freebsd/zfs/zfs_ctldir.c index 5bd2e1510ddb..cfc4bab2fa96 100644 --- a/module/os/freebsd/zfs/zfs_ctldir.c +++ b/module/os/freebsd/zfs/zfs_ctldir.c @@ -204,6 +204,10 @@ sfs_vgetx(struct mount *mp, int flags, uint64_t parent_id, uint64_t id, return (error); } +#if __FreeBSD_version >= 1400077 + vn_set_state(vp, VSTATE_CONSTRUCTED); +#endif + *vpp = vp; return (0); } @@ -675,6 +679,17 @@ zfsctl_root_readdir(struct vop_readdir_args *ap) ASSERT3S(vp->v_type, ==, VDIR); + /* + * FIXME: this routine only ever emits 3 entries and does not tolerate + * being called with a buffer too small to handle all of them. + * + * The check below facilitates the idiom of repeating calls until the + * count to return is 0. + */ + if (zfs_uio_offset(&uio) == 3 * sizeof (entry)) { + return (0); + } + error = sfs_readdir_common(zfsvfs->z_root, ZFSCTL_INO_ROOT, ap, &uio, &dots_offset); if (error != 0) { @@ -801,6 +816,9 @@ static struct vop_vector zfsctl_ops_root = { #if __FreeBSD_version >= 1300121 .vop_fplookup_vexec = VOP_EAGAIN, #endif +#if __FreeBSD_version >= 1300139 + .vop_fplookup_symlink = VOP_EAGAIN, +#endif .vop_open = zfsctl_common_open, .vop_close = zfsctl_common_close, .vop_ioctl = VOP_EINVAL, @@ -1127,6 +1145,9 @@ static struct vop_vector zfsctl_ops_snapdir = { #if __FreeBSD_version >= 1300121 .vop_fplookup_vexec = VOP_EAGAIN, #endif +#if __FreeBSD_version >= 1300139 + .vop_fplookup_symlink = VOP_EAGAIN, +#endif .vop_open = zfsctl_common_open, .vop_close = zfsctl_common_close, .vop_getattr = zfsctl_snapdir_getattr, @@ -1150,7 +1171,7 @@ zfsctl_snapshot_inactive(struct vop_inactive_args *ap) { vnode_t *vp = ap->a_vp; - VERIFY3S(vrecycle(vp), ==, 1); + vrecycle(vp); return (0); } @@ -1234,6 +1255,11 @@ static struct vop_vector zfsctl_ops_snapshot = { #if __FreeBSD_version >= 1300121 .vop_fplookup_vexec = VOP_EAGAIN, #endif +#if __FreeBSD_version >= 1300139 + .vop_fplookup_symlink = VOP_EAGAIN, +#endif + .vop_open = zfsctl_common_open, + .vop_close = zfsctl_common_close, .vop_inactive = zfsctl_snapshot_inactive, #if __FreeBSD_version >= 1300045 .vop_need_inactive = vop_stdneed_inactive, diff --git a/module/os/freebsd/zfs/zfs_debug.c b/module/os/freebsd/zfs/zfs_debug.c index dad342b06fc1..b75cf092182f 100644 --- a/module/os/freebsd/zfs/zfs_debug.c +++ b/module/os/freebsd/zfs/zfs_debug.c @@ -30,7 +30,7 @@ typedef struct zfs_dbgmsg { list_node_t zdm_node; time_t zdm_timestamp; int zdm_size; - char zdm_msg[1]; /* variable length allocation */ + char zdm_msg[]; } zfs_dbgmsg_t; list_t zfs_dbgmsgs; @@ -159,7 +159,7 @@ __zfs_dbgmsg(char *buf) DTRACE_PROBE1(zfs__dbgmsg, char *, buf); - size = sizeof (zfs_dbgmsg_t) + strlen(buf); + size = sizeof (zfs_dbgmsg_t) + strlen(buf) + 1; zdm = kmem_zalloc(size, KM_SLEEP); zdm->zdm_size = size; zdm->zdm_timestamp = gethrestime_sec(); diff --git a/module/os/freebsd/zfs/zfs_ioctl_os.c b/module/os/freebsd/zfs/zfs_ioctl_os.c index 7f7e2b72c51a..effc11518c57 100644 --- a/module/os/freebsd/zfs/zfs_ioctl_os.c +++ b/module/os/freebsd/zfs/zfs_ioctl_os.c @@ -59,7 +59,7 @@ zfs_vfs_ref(zfsvfs_t **zfvp) return (error); } -int +boolean_t zfs_vfs_held(zfsvfs_t *zfsvfs) { return (zfsvfs->z_vfs != NULL); diff --git a/module/os/freebsd/zfs/zfs_znode.c b/module/os/freebsd/zfs/zfs_znode.c index 1debc3ec3dfa..92e3bdd2e6e9 100644 --- a/module/os/freebsd/zfs/zfs_znode.c +++ b/module/os/freebsd/zfs/zfs_znode.c @@ -153,6 +153,9 @@ zfs_znode_cache_constructor(void *buf, void *arg, int kmflags) zp->z_xattr_cached = NULL; zp->z_xattr_parent = 0; zp->z_vnode = NULL; + zp->z_sync_writes_cnt = 0; + zp->z_async_writes_cnt = 0; + return (0); } @@ -172,6 +175,9 @@ zfs_znode_cache_destructor(void *buf, void *arg) ASSERT3P(zp->z_acl_cached, ==, NULL); ASSERT3P(zp->z_xattr_cached, ==, NULL); + + ASSERT0(atomic_load_32(&zp->z_sync_writes_cnt)); + ASSERT0(atomic_load_32(&zp->z_async_writes_cnt)); } @@ -457,6 +463,8 @@ zfs_znode_alloc(zfsvfs_t *zfsvfs, dmu_buf_t *db, int blksz, zp->z_blksz = blksz; zp->z_seq = 0x7A4653; zp->z_sync_cnt = 0; + zp->z_sync_writes_cnt = 0; + zp->z_async_writes_cnt = 0; #if __FreeBSD_version >= 1300139 atomic_store_ptr(&zp->z_cached_symlink, NULL); #endif diff --git a/module/os/linux/spl/Makefile.in b/module/os/linux/spl/Makefile.in index b2325f91b4a7..ad2dc6e3eccc 100644 --- a/module/os/linux/spl/Makefile.in +++ b/module/os/linux/spl/Makefile.in @@ -8,6 +8,7 @@ $(MODULE)-objs += ../os/linux/spl/spl-kmem-cache.o $(MODULE)-objs += ../os/linux/spl/spl-kstat.o $(MODULE)-objs += ../os/linux/spl/spl-proc.o $(MODULE)-objs += ../os/linux/spl/spl-procfs-list.o +$(MODULE)-objs += ../os/linux/spl/spl-shrinker.o $(MODULE)-objs += ../os/linux/spl/spl-taskq.o $(MODULE)-objs += ../os/linux/spl/spl-thread.o $(MODULE)-objs += ../os/linux/spl/spl-trace.o diff --git a/module/os/linux/spl/spl-cred.c b/module/os/linux/spl/spl-cred.c index f81b9540a639..d407fc66b2de 100644 --- a/module/os/linux/spl/spl-cred.c +++ b/module/os/linux/spl/spl-cred.c @@ -145,6 +145,18 @@ crgetgid(const cred_t *cr) return (KGID_TO_SGID(cr->fsgid)); } +/* Return the initial user ns or nop_mnt_idmap */ +zidmap_t * +zfs_get_init_idmap(void) +{ +#ifdef HAVE_IOPS_CREATE_IDMAP + return ((zidmap_t *)&nop_mnt_idmap); +#else + return ((zidmap_t *)&init_user_ns); +#endif +} + +EXPORT_SYMBOL(zfs_get_init_idmap); EXPORT_SYMBOL(crhold); EXPORT_SYMBOL(crfree); EXPORT_SYMBOL(crgetuid); diff --git a/module/os/linux/spl/spl-generic.c b/module/os/linux/spl/spl-generic.c index 508fb9d4c7f7..2cb5251d731a 100644 --- a/module/os/linux/spl/spl-generic.c +++ b/module/os/linux/spl/spl-generic.c @@ -225,8 +225,10 @@ __div_u64(uint64_t u, uint32_t v) * replacements for libgcc-provided functions and will never be called * directly. */ +#if defined(__GNUC__) && !defined(__clang__) #pragma GCC diagnostic push #pragma GCC diagnostic ignored "-Wmissing-prototypes" +#endif /* * Implementation of 64-bit unsigned division for 32-bit machines. @@ -425,7 +427,9 @@ __aeabi_ldivmod(int64_t u, int64_t v) EXPORT_SYMBOL(__aeabi_ldivmod); #endif /* __arm || __arm__ */ +#if defined(__GNUC__) && !defined(__clang__) #pragma GCC diagnostic pop +#endif #endif /* BITS_PER_LONG */ diff --git a/module/os/linux/spl/spl-kmem-cache.c b/module/os/linux/spl/spl-kmem-cache.c index 5a318e0a5391..8624d0d9ccc7 100644 --- a/module/os/linux/spl/spl-kmem-cache.c +++ b/module/os/linux/spl/spl-kmem-cache.c @@ -28,6 +28,7 @@ #include <sys/timer.h> #include <sys/vmem.h> #include <sys/wait.h> +#include <sys/string.h> #include <linux/slab.h> #include <linux/swap.h> #include <linux/prefetch.h> @@ -183,8 +184,11 @@ kv_free(spl_kmem_cache_t *skc, void *ptr, int size) * of that infrastructure we are responsible for incrementing it. */ if (current->reclaim_state) +#ifdef HAVE_RECLAIM_STATE_RECLAIMED + current->reclaim_state->reclaimed += size >> PAGE_SHIFT; +#else current->reclaim_state->reclaimed_slab += size >> PAGE_SHIFT; - +#endif vfree(ptr); } @@ -1017,10 +1021,20 @@ spl_cache_grow(spl_kmem_cache_t *skc, int flags, void **obj) ASSERT0(flags & ~KM_PUBLIC_MASK); ASSERT(skc->skc_magic == SKC_MAGIC); ASSERT((skc->skc_flags & KMC_SLAB) == 0); - might_sleep(); + *obj = NULL; /* + * Since we can't sleep attempt an emergency allocation to satisfy + * the request. The only alterative is to fail the allocation but + * it's preferable try. The use of KM_NOSLEEP is expected to be rare. + */ + if (flags & KM_NOSLEEP) + return (spl_emergency_alloc(skc, flags, obj)); + + might_sleep(); + + /* * Before allocating a new slab wait for any reaping to complete and * then return so the local magazine can be rechecked for new objects. */ diff --git a/module/os/linux/spl/spl-kstat.c b/module/os/linux/spl/spl-kstat.c index b5666e78842b..02874050c77d 100644 --- a/module/os/linux/spl/spl-kstat.c +++ b/module/os/linux/spl/spl-kstat.c @@ -32,6 +32,7 @@ #include <sys/vmem.h> #include <sys/cmn_err.h> #include <sys/sysmacros.h> +#include <sys/string.h> static kmutex_t kstat_module_lock; static struct list_head kstat_module_list; diff --git a/module/os/linux/spl/spl-proc.c b/module/os/linux/spl/spl-proc.c index c4af27a7fcd7..81dd5d25a1d0 100644 --- a/module/os/linux/spl/spl-proc.c +++ b/module/os/linux/spl/spl-proc.c @@ -46,6 +46,10 @@ static unsigned long table_min = 0; static unsigned long table_max = ~0; static struct ctl_table_header *spl_header = NULL; +#ifndef HAVE_REGISTER_SYSCTL_TABLE +static struct ctl_table_header *spl_kmem = NULL; +static struct ctl_table_header *spl_kstat = NULL; +#endif static struct proc_dir_entry *proc_spl = NULL; static struct proc_dir_entry *proc_spl_kmem = NULL; static struct proc_dir_entry *proc_spl_kmem_slab = NULL; @@ -624,6 +628,7 @@ static struct ctl_table spl_table[] = { .mode = 0644, .proc_handler = &proc_dohostid, }, +#ifdef HAVE_REGISTER_SYSCTL_TABLE { .procname = "kmem", .mode = 0555, @@ -634,9 +639,11 @@ static struct ctl_table spl_table[] = { .mode = 0555, .child = spl_kstat_table, }, +#endif {}, }; +#ifdef HAVE_REGISTER_SYSCTL_TABLE static struct ctl_table spl_dir[] = { { .procname = "spl", @@ -648,21 +655,64 @@ static struct ctl_table spl_dir[] = { static struct ctl_table spl_root[] = { { - .procname = "kernel", - .mode = 0555, - .child = spl_dir, + .procname = "kernel", + .mode = 0555, + .child = spl_dir, }, {} }; +#endif + +static void spl_proc_cleanup(void) +{ + remove_proc_entry("kstat", proc_spl); + remove_proc_entry("slab", proc_spl_kmem); + remove_proc_entry("kmem", proc_spl); + remove_proc_entry("taskq-all", proc_spl); + remove_proc_entry("taskq", proc_spl); + remove_proc_entry("spl", NULL); + +#ifndef HAVE_REGISTER_SYSCTL_TABLE + if (spl_kstat) { + unregister_sysctl_table(spl_kstat); + spl_kstat = NULL; + } + if (spl_kmem) { + unregister_sysctl_table(spl_kmem); + spl_kmem = NULL; + } +#endif + if (spl_header) { + unregister_sysctl_table(spl_header); + spl_header = NULL; + } +} int spl_proc_init(void) { int rc = 0; +#ifdef HAVE_REGISTER_SYSCTL_TABLE spl_header = register_sysctl_table(spl_root); if (spl_header == NULL) return (-EUNATCH); +#else + spl_header = register_sysctl("kernel/spl", spl_table); + if (spl_header == NULL) + return (-EUNATCH); + + spl_kmem = register_sysctl("kernel/spl/kmem", spl_kmem_table); + if (spl_kmem == NULL) { + rc = -EUNATCH; + goto out; + } + spl_kstat = register_sysctl("kernel/spl/kstat", spl_kstat_table); + if (spl_kstat == NULL) { + rc = -EUNATCH; + goto out; + } +#endif proc_spl = proc_mkdir("spl", NULL); if (proc_spl == NULL) { @@ -703,15 +753,8 @@ spl_proc_init(void) goto out; } out: - if (rc) { - remove_proc_entry("kstat", proc_spl); - remove_proc_entry("slab", proc_spl_kmem); - remove_proc_entry("kmem", proc_spl); - remove_proc_entry("taskq-all", proc_spl); - remove_proc_entry("taskq", proc_spl); - remove_proc_entry("spl", NULL); - unregister_sysctl_table(spl_header); - } + if (rc) + spl_proc_cleanup(); return (rc); } @@ -719,13 +762,5 @@ out: void spl_proc_fini(void) { - remove_proc_entry("kstat", proc_spl); - remove_proc_entry("slab", proc_spl_kmem); - remove_proc_entry("kmem", proc_spl); - remove_proc_entry("taskq-all", proc_spl); - remove_proc_entry("taskq", proc_spl); - remove_proc_entry("spl", NULL); - - ASSERT(spl_header != NULL); - unregister_sysctl_table(spl_header); + spl_proc_cleanup(); } diff --git a/module/os/linux/spl/spl-shrinker.c b/module/os/linux/spl/spl-shrinker.c new file mode 100644 index 000000000000..d5c8da471cbb --- /dev/null +++ b/module/os/linux/spl/spl-shrinker.c @@ -0,0 +1,115 @@ +/* + * Copyright (C) 2007-2010 Lawrence Livermore National Security, LLC. + * Copyright (C) 2007 The Regents of the University of California. + * Produced at Lawrence Livermore National Laboratory (cf, DISCLAIMER). + * Written by Brian Behlendorf <behlendorf1@llnl.gov>. + * UCRL-CODE-235197 + * + * This file is part of the SPL, Solaris Porting Layer. + * + * The SPL is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + * + * The SPL is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + * + * You should have received a copy of the GNU General Public License along + * with the SPL. If not, see <http://www.gnu.org/licenses/>. + * + * Solaris Porting Layer (SPL) Shrinker Implementation. + */ + +#include <sys/kmem.h> +#include <sys/shrinker.h> + +#ifdef HAVE_SINGLE_SHRINKER_CALLBACK +/* 3.0-3.11: single shrink() callback, which we wrap to carry both functions */ +struct spl_shrinker_wrap { + struct shrinker shrinker; + spl_shrinker_cb countfunc; + spl_shrinker_cb scanfunc; +}; + +static int +spl_shrinker_single_cb(struct shrinker *shrinker, struct shrink_control *sc) +{ + struct spl_shrinker_wrap *sw = (struct spl_shrinker_wrap *)shrinker; + + if (sc->nr_to_scan != 0) + (void) sw->scanfunc(&sw->shrinker, sc); + return (sw->countfunc(&sw->shrinker, sc)); +} +#endif + +struct shrinker * +spl_register_shrinker(const char *name, spl_shrinker_cb countfunc, + spl_shrinker_cb scanfunc, int seek_cost) +{ + struct shrinker *shrinker; + + /* allocate shrinker */ +#if defined(HAVE_SHRINKER_REGISTER) + /* 6.7: kernel will allocate the shrinker for us */ + shrinker = shrinker_alloc(0, name); +#elif defined(HAVE_SPLIT_SHRINKER_CALLBACK) + /* 3.12-6.6: we allocate the shrinker */ + shrinker = kmem_zalloc(sizeof (struct shrinker), KM_SLEEP); +#elif defined(HAVE_SINGLE_SHRINKER_CALLBACK) + /* 3.0-3.11: allocate a wrapper */ + struct spl_shrinker_wrap *sw = + kmem_zalloc(sizeof (struct spl_shrinker_wrap), KM_SLEEP); + shrinker = &sw->shrinker; +#else + /* 2.x-2.6.22, or a newer shrinker API has been introduced. */ +#error "Unknown shrinker API" +#endif + + if (shrinker == NULL) + return (NULL); + + /* set callbacks */ +#ifdef HAVE_SINGLE_SHRINKER_CALLBACK + sw->countfunc = countfunc; + sw->scanfunc = scanfunc; + shrinker->shrink = spl_shrinker_single_cb; +#else + shrinker->count_objects = countfunc; + shrinker->scan_objects = scanfunc; +#endif + + /* set params */ + shrinker->seeks = seek_cost; + + /* register with kernel */ +#if defined(HAVE_SHRINKER_REGISTER) + shrinker_register(shrinker); +#elif defined(HAVE_REGISTER_SHRINKER_VARARG) + register_shrinker(shrinker, name); +#else + register_shrinker(shrinker); +#endif + + return (shrinker); +} +EXPORT_SYMBOL(spl_register_shrinker); + +void +spl_unregister_shrinker(struct shrinker *shrinker) +{ +#if defined(HAVE_SHRINKER_REGISTER) + shrinker_free(shrinker); +#elif defined(HAVE_SPLIT_SHRINKER_CALLBACK) + unregister_shrinker(shrinker); + kmem_free(shrinker, sizeof (struct shrinker)); +#elif defined(HAVE_SINGLE_SHRINKER_CALLBACK) + unregister_shrinker(shrinker); + kmem_free(shrinker, sizeof (struct spl_shrinker_wrap)); +#else +#error "Unknown shrinker API" +#endif +} +EXPORT_SYMBOL(spl_unregister_shrinker); diff --git a/module/os/linux/spl/spl-thread.c b/module/os/linux/spl/spl-thread.c index 16d2ca1b133b..faf06775e50f 100644 --- a/module/os/linux/spl/spl-thread.c +++ b/module/os/linux/spl/spl-thread.c @@ -26,6 +26,7 @@ #include <sys/thread.h> #include <sys/kmem.h> #include <sys/tsd.h> +#include <sys/string.h> /* * Thread interfaces diff --git a/module/os/linux/zfs/abd_os.c b/module/os/linux/zfs/abd_os.c index 6067950d5cd7..da3e9fbb91ca 100644 --- a/module/os/linux/zfs/abd_os.c +++ b/module/os/linux/zfs/abd_os.c @@ -60,8 +60,16 @@ #ifdef _KERNEL #include <linux/kmap_compat.h> #include <linux/scatterlist.h> +#endif + +#ifdef _KERNEL +#if defined(MAX_ORDER) +#define ABD_MAX_ORDER (MAX_ORDER) +#elif defined(MAX_PAGE_ORDER) +#define ABD_MAX_ORDER (MAX_PAGE_ORDER) +#endif #else -#define MAX_ORDER 1 +#define ABD_MAX_ORDER (1) #endif typedef struct abd_stats { @@ -71,7 +79,7 @@ typedef struct abd_stats { kstat_named_t abdstat_scatter_cnt; kstat_named_t abdstat_scatter_data_size; kstat_named_t abdstat_scatter_chunk_waste; - kstat_named_t abdstat_scatter_orders[MAX_ORDER]; + kstat_named_t abdstat_scatter_orders[ABD_MAX_ORDER]; kstat_named_t abdstat_scatter_page_multi_chunk; kstat_named_t abdstat_scatter_page_multi_zone; kstat_named_t abdstat_scatter_page_alloc_retry; @@ -139,7 +147,7 @@ struct { wmsum_t abdstat_scatter_cnt; wmsum_t abdstat_scatter_data_size; wmsum_t abdstat_scatter_chunk_waste; - wmsum_t abdstat_scatter_orders[MAX_ORDER]; + wmsum_t abdstat_scatter_orders[ABD_MAX_ORDER]; wmsum_t abdstat_scatter_page_multi_chunk; wmsum_t abdstat_scatter_page_multi_zone; wmsum_t abdstat_scatter_page_alloc_retry; @@ -149,7 +157,7 @@ struct { #define abd_for_each_sg(abd, sg, n, i) \ for_each_sg(ABD_SCATTER(abd).abd_sgl, sg, n, i) -unsigned zfs_abd_scatter_max_order = MAX_ORDER - 1; +unsigned zfs_abd_scatter_max_order = ABD_MAX_ORDER - 1; /* * zfs_abd_scatter_min_size is the minimum allocation size to use scatter @@ -271,7 +279,7 @@ abd_alloc_chunks(abd_t *abd, size_t size) struct page *page, *tmp_page = NULL; gfp_t gfp = __GFP_NOWARN | GFP_NOIO; gfp_t gfp_comp = (gfp | __GFP_NORETRY | __GFP_COMP) & ~__GFP_RECLAIM; - int max_order = MIN(zfs_abd_scatter_max_order, MAX_ORDER - 1); + int max_order = MIN(zfs_abd_scatter_max_order, ABD_MAX_ORDER - 1); int nr_pages = abd_chunkcnt_for_bytes(size); int chunks = 0, zones = 0; size_t remaining_size; @@ -728,7 +736,7 @@ abd_kstats_update(kstat_t *ksp, int rw) wmsum_value(&abd_sums.abdstat_scatter_data_size); as->abdstat_scatter_chunk_waste.value.ui64 = wmsum_value(&abd_sums.abdstat_scatter_chunk_waste); - for (int i = 0; i < MAX_ORDER; i++) { + for (int i = 0; i < ABD_MAX_ORDER; i++) { as->abdstat_scatter_orders[i].value.ui64 = wmsum_value(&abd_sums.abdstat_scatter_orders[i]); } @@ -757,7 +765,7 @@ abd_init(void) wmsum_init(&abd_sums.abdstat_scatter_cnt, 0); wmsum_init(&abd_sums.abdstat_scatter_data_size, 0); wmsum_init(&abd_sums.abdstat_scatter_chunk_waste, 0); - for (i = 0; i < MAX_ORDER; i++) + for (i = 0; i < ABD_MAX_ORDER; i++) wmsum_init(&abd_sums.abdstat_scatter_orders[i], 0); wmsum_init(&abd_sums.abdstat_scatter_page_multi_chunk, 0); wmsum_init(&abd_sums.abdstat_scatter_page_multi_zone, 0); @@ -767,7 +775,7 @@ abd_init(void) abd_ksp = kstat_create("zfs", 0, "abdstats", "misc", KSTAT_TYPE_NAMED, sizeof (abd_stats) / sizeof (kstat_named_t), KSTAT_FLAG_VIRTUAL); if (abd_ksp != NULL) { - for (i = 0; i < MAX_ORDER; i++) { + for (i = 0; i < ABD_MAX_ORDER; i++) { snprintf(abd_stats.abdstat_scatter_orders[i].name, KSTAT_STRLEN, "scatter_order_%d", i); abd_stats.abdstat_scatter_orders[i].data_type = @@ -797,7 +805,7 @@ abd_fini(void) wmsum_fini(&abd_sums.abdstat_scatter_cnt); wmsum_fini(&abd_sums.abdstat_scatter_data_size); wmsum_fini(&abd_sums.abdstat_scatter_chunk_waste); - for (int i = 0; i < MAX_ORDER; i++) + for (int i = 0; i < ABD_MAX_ORDER; i++) wmsum_fini(&abd_sums.abdstat_scatter_orders[i]); wmsum_fini(&abd_sums.abdstat_scatter_page_multi_chunk); wmsum_fini(&abd_sums.abdstat_scatter_page_multi_zone); diff --git a/module/os/linux/zfs/arc_os.c b/module/os/linux/zfs/arc_os.c index f96cd1271ee5..19540221d688 100644 --- a/module/os/linux/zfs/arc_os.c +++ b/module/os/linux/zfs/arc_os.c @@ -219,7 +219,11 @@ arc_shrinker_scan(struct shrinker *shrink, struct shrink_control *sc) arc_reduce_target_size(ptob(sc->nr_to_scan)); arc_wait_for_eviction(ptob(sc->nr_to_scan), B_FALSE); if (current->reclaim_state != NULL) +#ifdef HAVE_RECLAIM_STATE_RECLAIMED + current->reclaim_state->reclaimed += sc->nr_to_scan; +#else current->reclaim_state->reclaimed_slab += sc->nr_to_scan; +#endif /* * We are experiencing memory pressure which the arc_evict_zthr was @@ -243,8 +247,7 @@ arc_shrinker_scan(struct shrinker *shrink, struct shrink_control *sc) return (sc->nr_to_scan); } -SPL_SHRINKER_DECLARE(arc_shrinker, - arc_shrinker_count, arc_shrinker_scan, DEFAULT_SEEKS); +static struct shrinker *arc_shrinker = NULL; int arc_memory_throttle(spa_t *spa, uint64_t reserve, uint64_t txg) @@ -347,14 +350,18 @@ arc_lowmem_init(void) * reclaim from the arc. This is done to prevent kswapd from * swapping out pages when it is preferable to shrink the arc. */ - spl_register_shrinker(&arc_shrinker); + arc_shrinker = spl_register_shrinker("zfs-arc-shrinker", + arc_shrinker_count, arc_shrinker_scan, DEFAULT_SEEKS); + VERIFY(arc_shrinker); + arc_set_sys_free(allmem); } void arc_lowmem_fini(void) { - spl_unregister_shrinker(&arc_shrinker); + spl_unregister_shrinker(arc_shrinker); + arc_shrinker = NULL; } int diff --git a/module/os/linux/zfs/policy.c b/module/os/linux/zfs/policy.c index 5a52092bb90a..8d508bcb4999 100644 --- a/module/os/linux/zfs/policy.c +++ b/module/os/linux/zfs/policy.c @@ -124,7 +124,7 @@ secpolicy_vnode_any_access(const cred_t *cr, struct inode *ip, uid_t owner) if (crgetuid(cr) == owner) return (0); - if (zpl_inode_owner_or_capable(kcred->user_ns, ip)) + if (zpl_inode_owner_or_capable(zfs_init_idmap, ip)) return (0); #if defined(CONFIG_USER_NS) diff --git a/module/os/linux/zfs/vdev_disk.c b/module/os/linux/zfs/vdev_disk.c index 60b111c59f23..3ec4f0f09696 100644 --- a/module/os/linux/zfs/vdev_disk.c +++ b/module/os/linux/zfs/vdev_disk.c @@ -41,8 +41,28 @@ #include <linux/blk-cgroup.h> #endif +/* + * Linux 6.8.x uses a bdev_handle as an instance/refcount for an underlying + * block_device. Since it carries the block_device inside, its convenient to + * just use the handle as a proxy. For pre-6.8, we just emulate this with + * a cast, since we don't need any of the other fields inside the handle. + */ +#ifdef HAVE_BDEV_OPEN_BY_PATH +typedef struct bdev_handle zfs_bdev_handle_t; +#define BDH_BDEV(bdh) ((bdh)->bdev) +#define BDH_IS_ERR(bdh) (IS_ERR(bdh)) +#define BDH_PTR_ERR(bdh) (PTR_ERR(bdh)) +#define BDH_ERR_PTR(err) (ERR_PTR(err)) +#else +typedef void zfs_bdev_handle_t; +#define BDH_BDEV(bdh) ((struct block_device *)bdh) +#define BDH_IS_ERR(bdh) (IS_ERR(BDH_BDEV(bdh))) +#define BDH_PTR_ERR(bdh) (PTR_ERR(BDH_BDEV(bdh))) +#define BDH_ERR_PTR(err) (ERR_PTR(err)) +#endif + typedef struct vdev_disk { - struct block_device *vd_bdev; + zfs_bdev_handle_t *vd_bdh; krwlock_t vd_lock; } vdev_disk_t; @@ -74,9 +94,25 @@ typedef struct dio_request { struct bio *dr_bio[0]; /* Attached bio's */ } dio_request_t; +#ifdef HAVE_BLK_MODE_T +static blk_mode_t +#else static fmode_t -vdev_bdev_mode(spa_mode_t spa_mode) +#endif +vdev_bdev_mode(spa_mode_t spa_mode, boolean_t exclusive) { +#ifdef HAVE_BLK_MODE_T + blk_mode_t mode = 0; + + if (spa_mode & SPA_MODE_READ) + mode |= BLK_OPEN_READ; + + if (spa_mode & SPA_MODE_WRITE) + mode |= BLK_OPEN_WRITE; + + if (exclusive) + mode |= BLK_OPEN_EXCL; +#else fmode_t mode = 0; if (spa_mode & SPA_MODE_READ) @@ -85,6 +121,10 @@ vdev_bdev_mode(spa_mode_t spa_mode) if (spa_mode & SPA_MODE_WRITE) mode |= FMODE_WRITE; + if (exclusive) + mode |= FMODE_EXCL; +#endif + return (mode); } @@ -183,20 +223,52 @@ static void vdev_disk_kobj_evt_post(vdev_t *v) { vdev_disk_t *vd = v->vdev_tsd; - if (vd && vd->vd_bdev) { - spl_signal_kobj_evt(vd->vd_bdev); + if (vd && vd->vd_bdh) { + spl_signal_kobj_evt(BDH_BDEV(vd->vd_bdh)); } else { vdev_dbgmsg(v, "vdev_disk_t is NULL for VDEV:%s\n", v->vdev_path); } } +static zfs_bdev_handle_t * +vdev_blkdev_get_by_path(const char *path, spa_mode_t mode, void *holder) +{ +#if defined(HAVE_BDEV_OPEN_BY_PATH) + return (bdev_open_by_path(path, + vdev_bdev_mode(mode, B_TRUE), holder, NULL)); +#elif defined(HAVE_BLKDEV_GET_BY_PATH_4ARG) + return (blkdev_get_by_path(path, + vdev_bdev_mode(mode, B_TRUE), holder, NULL)); +#else + return (blkdev_get_by_path(path, + vdev_bdev_mode(mode, B_TRUE), holder)); +#endif +} + +static void +vdev_blkdev_put(zfs_bdev_handle_t *bdh, spa_mode_t mode, void *holder) +{ +#if defined(HAVE_BDEV_RELEASE) + return (bdev_release(bdh)); +#elif defined(HAVE_BLKDEV_PUT_HOLDER) + return (blkdev_put(BDH_BDEV(bdh), holder)); +#else + return (blkdev_put(BDH_BDEV(bdh), + vdev_bdev_mode(mode, B_TRUE))); +#endif +} + static int vdev_disk_open(vdev_t *v, uint64_t *psize, uint64_t *max_psize, uint64_t *logical_ashift, uint64_t *physical_ashift) { - struct block_device *bdev; - fmode_t mode = vdev_bdev_mode(spa_mode(v->vdev_spa)); + zfs_bdev_handle_t *bdh; +#ifdef HAVE_BLK_MODE_T + blk_mode_t mode = vdev_bdev_mode(spa_mode(v->vdev_spa), B_FALSE); +#else + fmode_t mode = vdev_bdev_mode(spa_mode(v->vdev_spa), B_FALSE); +#endif hrtime_t timeout = MSEC2NSEC(zfs_vdev_open_timeout_ms); vdev_disk_t *vd; @@ -221,10 +293,11 @@ vdev_disk_open(vdev_t *v, uint64_t *psize, uint64_t *max_psize, boolean_t reread_part = B_FALSE; rw_enter(&vd->vd_lock, RW_WRITER); - bdev = vd->vd_bdev; - vd->vd_bdev = NULL; + bdh = vd->vd_bdh; + vd->vd_bdh = NULL; - if (bdev) { + if (bdh) { + struct block_device *bdev = BDH_BDEV(bdh); if (v->vdev_expanding && bdev != bdev_whole(bdev)) { vdev_bdevname(bdev_whole(bdev), disk_name + 5); /* @@ -246,15 +319,16 @@ vdev_disk_open(vdev_t *v, uint64_t *psize, uint64_t *max_psize, reread_part = B_TRUE; } - blkdev_put(bdev, mode | FMODE_EXCL); + vdev_blkdev_put(bdh, mode, zfs_vdev_holder); } if (reread_part) { - bdev = blkdev_get_by_path(disk_name, mode | FMODE_EXCL, + bdh = vdev_blkdev_get_by_path(disk_name, mode, zfs_vdev_holder); - if (!IS_ERR(bdev)) { - int error = vdev_bdev_reread_part(bdev); - blkdev_put(bdev, mode | FMODE_EXCL); + if (!BDH_IS_ERR(bdh)) { + int error = + vdev_bdev_reread_part(BDH_BDEV(bdh)); + vdev_blkdev_put(bdh, mode, zfs_vdev_holder); if (error == 0) { timeout = MSEC2NSEC( zfs_vdev_open_timeout_ms * 2); @@ -297,11 +371,11 @@ vdev_disk_open(vdev_t *v, uint64_t *psize, uint64_t *max_psize, * subsequent attempts are expected to eventually succeed. */ hrtime_t start = gethrtime(); - bdev = ERR_PTR(-ENXIO); - while (IS_ERR(bdev) && ((gethrtime() - start) < timeout)) { - bdev = blkdev_get_by_path(v->vdev_path, mode | FMODE_EXCL, + bdh = BDH_ERR_PTR(-ENXIO); + while (BDH_IS_ERR(bdh) && ((gethrtime() - start) < timeout)) { + bdh = vdev_blkdev_get_by_path(v->vdev_path, mode, zfs_vdev_holder); - if (unlikely(PTR_ERR(bdev) == -ENOENT)) { + if (unlikely(BDH_PTR_ERR(bdh) == -ENOENT)) { /* * There is no point of waiting since device is removed * explicitly @@ -310,52 +384,54 @@ vdev_disk_open(vdev_t *v, uint64_t *psize, uint64_t *max_psize, break; schedule_timeout(MSEC_TO_TICK(10)); - } else if (unlikely(PTR_ERR(bdev) == -ERESTARTSYS)) { + } else if (unlikely(BDH_PTR_ERR(bdh) == -ERESTARTSYS)) { timeout = MSEC2NSEC(zfs_vdev_open_timeout_ms * 10); continue; - } else if (IS_ERR(bdev)) { + } else if (BDH_IS_ERR(bdh)) { break; } } - if (IS_ERR(bdev)) { - int error = -PTR_ERR(bdev); + if (BDH_IS_ERR(bdh)) { + int error = -BDH_PTR_ERR(bdh); vdev_dbgmsg(v, "open error=%d timeout=%llu/%llu", error, (u_longlong_t)(gethrtime() - start), (u_longlong_t)timeout); - vd->vd_bdev = NULL; + vd->vd_bdh = NULL; v->vdev_tsd = vd; rw_exit(&vd->vd_lock); return (SET_ERROR(error)); } else { - vd->vd_bdev = bdev; + vd->vd_bdh = bdh; v->vdev_tsd = vd; rw_exit(&vd->vd_lock); } + struct block_device *bdev = BDH_BDEV(vd->vd_bdh); + /* Determine the physical block size */ - int physical_block_size = bdev_physical_block_size(vd->vd_bdev); + int physical_block_size = bdev_physical_block_size(bdev); /* Determine the logical block size */ - int logical_block_size = bdev_logical_block_size(vd->vd_bdev); + int logical_block_size = bdev_logical_block_size(bdev); /* Clear the nowritecache bit, causes vdev_reopen() to try again. */ v->vdev_nowritecache = B_FALSE; /* Set when device reports it supports TRIM. */ - v->vdev_has_trim = bdev_discard_supported(vd->vd_bdev); + v->vdev_has_trim = bdev_discard_supported(bdev); /* Set when device reports it supports secure TRIM. */ - v->vdev_has_securetrim = bdev_secure_discard_supported(vd->vd_bdev); + v->vdev_has_securetrim = bdev_secure_discard_supported(bdev); /* Inform the ZIO pipeline that we are non-rotational */ - v->vdev_nonrot = blk_queue_nonrot(bdev_get_queue(vd->vd_bdev)); + v->vdev_nonrot = blk_queue_nonrot(bdev_get_queue(bdev)); /* Physical volume size in bytes for the partition */ - *psize = bdev_capacity(vd->vd_bdev); + *psize = bdev_capacity(bdev); /* Physical volume size in bytes including possible expansion space */ - *max_psize = bdev_max_capacity(vd->vd_bdev, v->vdev_wholedisk); + *max_psize = bdev_max_capacity(bdev, v->vdev_wholedisk); /* Based on the minimum sector size set the block size */ *physical_ashift = highbit64(MAX(physical_block_size, @@ -375,9 +451,9 @@ vdev_disk_close(vdev_t *v) if (v->vdev_reopening || vd == NULL) return; - if (vd->vd_bdev != NULL) { - blkdev_put(vd->vd_bdev, - vdev_bdev_mode(spa_mode(v->vdev_spa)) | FMODE_EXCL); + if (vd->vd_bdh != NULL) { + vdev_blkdev_put(vd->vd_bdh, spa_mode(v->vdev_spa), + zfs_vdev_holder); } rw_destroy(&vd->vd_lock); @@ -786,10 +862,10 @@ vdev_disk_io_trim(zio_t *zio) #if defined(HAVE_BLKDEV_ISSUE_SECURE_ERASE) if (zio->io_trim_flags & ZIO_TRIM_SECURE) { - return (-blkdev_issue_secure_erase(vd->vd_bdev, + return (-blkdev_issue_secure_erase(BDH_BDEV(vd->vd_bdh), zio->io_offset >> 9, zio->io_size >> 9, GFP_NOFS)); } else { - return (-blkdev_issue_discard(vd->vd_bdev, + return (-blkdev_issue_discard(BDH_BDEV(vd->vd_bdh), zio->io_offset >> 9, zio->io_size >> 9, GFP_NOFS)); } #elif defined(HAVE_BLKDEV_ISSUE_DISCARD) @@ -798,7 +874,7 @@ vdev_disk_io_trim(zio_t *zio) if (zio->io_trim_flags & ZIO_TRIM_SECURE) trim_flags |= BLKDEV_DISCARD_SECURE; #endif - return (-blkdev_issue_discard(vd->vd_bdev, + return (-blkdev_issue_discard(BDH_BDEV(vd->vd_bdh), zio->io_offset >> 9, zio->io_size >> 9, GFP_NOFS, trim_flags)); #else #error "Unsupported kernel" @@ -828,7 +904,7 @@ vdev_disk_io_start(zio_t *zio) * If the vdev is closed, it's likely due to a failed reopen and is * in the UNAVAIL state. Nothing to be done here but return failure. */ - if (vd->vd_bdev == NULL) { + if (vd->vd_bdh == NULL) { rw_exit(&vd->vd_lock); zio->io_error = ENXIO; zio_interrupt(zio); @@ -856,7 +932,7 @@ vdev_disk_io_start(zio_t *zio) break; } - error = vdev_disk_io_flush(vd->vd_bdev, zio); + error = vdev_disk_io_flush(BDH_BDEV(vd->vd_bdh), zio); if (error == 0) { rw_exit(&vd->vd_lock); return; @@ -895,7 +971,7 @@ vdev_disk_io_start(zio_t *zio) } zio->io_target_timestamp = zio_handle_io_delay(zio); - error = __vdev_disk_physio(vd->vd_bdev, zio, + error = __vdev_disk_physio(BDH_BDEV(vd->vd_bdh), zio, zio->io_size, zio->io_offset, rw, 0); rw_exit(&vd->vd_lock); @@ -918,8 +994,8 @@ vdev_disk_io_done(zio_t *zio) vdev_t *v = zio->io_vd; vdev_disk_t *vd = v->vdev_tsd; - if (!zfs_check_disk_status(vd->vd_bdev)) { - invalidate_bdev(vd->vd_bdev); + if (!zfs_check_disk_status(BDH_BDEV(vd->vd_bdh))) { + invalidate_bdev(BDH_BDEV(vd->vd_bdh)); v->vdev_remove_wanted = B_TRUE; spa_async_request(zio->io_spa, SPA_ASYNC_REMOVE); } diff --git a/module/os/linux/zfs/zfs_ctldir.c b/module/os/linux/zfs/zfs_ctldir.c index c45644a69153..b51df6a2ebe5 100644 --- a/module/os/linux/zfs/zfs_ctldir.c +++ b/module/os/linux/zfs/zfs_ctldir.c @@ -118,6 +118,7 @@ typedef struct { spa_t *se_spa; /* pool spa */ uint64_t se_objsetid; /* snapshot objset id */ struct dentry *se_root_dentry; /* snapshot root dentry */ + krwlock_t se_taskqid_lock; /* scheduled unmount taskqid lock */ taskqid_t se_taskqid; /* scheduled unmount taskqid */ avl_node_t se_node_name; /* zfs_snapshots_by_name link */ avl_node_t se_node_objsetid; /* zfs_snapshots_by_objsetid link */ @@ -144,6 +145,7 @@ zfsctl_snapshot_alloc(const char *full_name, const char *full_path, spa_t *spa, se->se_objsetid = objsetid; se->se_root_dentry = root_dentry; se->se_taskqid = TASKQID_INVALID; + rw_init(&se->se_taskqid_lock, NULL, RW_DEFAULT, NULL); zfs_refcount_create(&se->se_refcount); @@ -160,6 +162,7 @@ zfsctl_snapshot_free(zfs_snapentry_t *se) zfs_refcount_destroy(&se->se_refcount); kmem_strfree(se->se_name); kmem_strfree(se->se_path); + rw_destroy(se->se_taskqid_lock); kmem_free(se, sizeof (zfs_snapentry_t)); } @@ -335,7 +338,9 @@ snapentry_expire(void *data) return; } + rw_enter(&se->se_taskqid_lock, RW_WRITER); se->se_taskqid = TASKQID_INVALID; + rw_exit(&se->se_taskqid_lock); (void) zfsctl_snapshot_unmount(se->se_name, MNT_EXPIRE); zfsctl_snapshot_rele(se); @@ -359,8 +364,18 @@ snapentry_expire(void *data) static void zfsctl_snapshot_unmount_cancel(zfs_snapentry_t *se) { - if (taskq_cancel_id(system_delay_taskq, se->se_taskqid) == 0) { - se->se_taskqid = TASKQID_INVALID; + int err = 0; + rw_enter(&se->se_taskqid_lock, RW_WRITER); + err = taskq_cancel_id(system_delay_taskq, se->se_taskqid); + /* + * if we get ENOENT, the taskq couldn't be found to be + * canceled, so we can just mark it as invalid because + * it's already gone. If we got EBUSY, then we already + * blocked until it was gone _anyway_, so we don't care. + */ + se->se_taskqid = TASKQID_INVALID; + rw_exit(&se->se_taskqid_lock); + if (err == 0) { zfsctl_snapshot_rele(se); } } @@ -371,14 +386,29 @@ zfsctl_snapshot_unmount_cancel(zfs_snapentry_t *se) static void zfsctl_snapshot_unmount_delay_impl(zfs_snapentry_t *se, int delay) { - ASSERT3S(se->se_taskqid, ==, TASKQID_INVALID); if (delay <= 0) return; zfsctl_snapshot_hold(se); + rw_enter(&se->se_taskqid_lock, RW_WRITER); + /* + * If this condition happens, we managed to: + * - dispatch once + * - want to dispatch _again_ before it returned + * + * So let's just return - if that task fails at unmounting, + * we'll eventually dispatch again, and if it succeeds, + * no problem. + */ + if (se->se_taskqid != TASKQID_INVALID) { + rw_exit(&se->se_taskqid_lock); + zfsctl_snapshot_rele(se); + return; + } se->se_taskqid = taskq_dispatch_delay(system_delay_taskq, snapentry_expire, se, TQ_SLEEP, ddi_get_lbolt() + delay * HZ); + rw_exit(&se->se_taskqid_lock); } /* @@ -468,7 +498,9 @@ zfsctl_inode_alloc(zfsvfs_t *zfsvfs, uint64_t id, zp->z_atime_dirty = B_FALSE; zp->z_zn_prefetch = B_FALSE; zp->z_is_sa = B_FALSE; +#if !defined(HAVE_FILEMAP_RANGE_HAS_PAGE) zp->z_is_mapped = B_FALSE; +#endif zp->z_is_ctldir = B_TRUE; zp->z_sa_hdl = NULL; zp->z_blksz = 0; @@ -478,15 +510,17 @@ zfsctl_inode_alloc(zfsvfs_t *zfsvfs, uint64_t id, zp->z_pflags = 0; zp->z_mode = 0; zp->z_sync_cnt = 0; + zp->z_sync_writes_cnt = 0; + zp->z_async_writes_cnt = 0; ip->i_generation = 0; ip->i_ino = id; ip->i_mode = (S_IFDIR | S_IRWXUGO); ip->i_uid = SUID_TO_KUID(0); ip->i_gid = SGID_TO_KGID(0); ip->i_blkbits = SPA_MINBLOCKSHIFT; - ip->i_atime = now; - ip->i_mtime = now; - ip->i_ctime = now; + zpl_inode_set_atime_to_ts(ip, now); + zpl_inode_set_mtime_to_ts(ip, now); + zpl_inode_set_ctime_to_ts(ip, now); ip->i_fop = fops; ip->i_op = ops; #if defined(IOP_XATTR) diff --git a/module/os/linux/zfs/zfs_debug.c b/module/os/linux/zfs/zfs_debug.c index 98c9923d5927..595806373162 100644 --- a/module/os/linux/zfs/zfs_debug.c +++ b/module/os/linux/zfs/zfs_debug.c @@ -30,7 +30,7 @@ typedef struct zfs_dbgmsg { procfs_list_node_t zdm_node; uint64_t zdm_timestamp; int zdm_size; - char zdm_msg[1]; /* variable length allocation */ + char zdm_msg[]; /* variable length allocation */ } zfs_dbgmsg_t; procfs_list_t zfs_dbgmsgs; @@ -134,7 +134,7 @@ __set_error(const char *file, const char *func, int line, int err) void __zfs_dbgmsg(char *buf) { - int size = sizeof (zfs_dbgmsg_t) + strlen(buf); + int size = sizeof (zfs_dbgmsg_t) + strlen(buf) + 1; zfs_dbgmsg_t *zdm = kmem_zalloc(size, KM_SLEEP); zdm->zdm_size = size; zdm->zdm_timestamp = gethrestime_sec(); diff --git a/module/os/linux/zfs/zfs_ioctl_os.c b/module/os/linux/zfs/zfs_ioctl_os.c index 79b9d777dc88..767d3a377dcc 100644 --- a/module/os/linux/zfs/zfs_ioctl_os.c +++ b/module/os/linux/zfs/zfs_ioctl_os.c @@ -288,6 +288,8 @@ zfsdev_detach(void) #define ZFS_DEBUG_STR "" #endif +zidmap_t *zfs_init_idmap; + static int __init openzfs_init(void) { @@ -311,6 +313,8 @@ openzfs_init(void) printk(KERN_NOTICE "ZFS: Posix ACLs disabled by kernel\n"); #endif /* CONFIG_FS_POSIX_ACL */ + zfs_init_idmap = (zidmap_t *)zfs_get_init_idmap(); + return (0); } diff --git a/module/os/linux/zfs/zfs_vfsops.c b/module/os/linux/zfs/zfs_vfsops.c index da897f120c0b..75901b9e47dc 100644 --- a/module/os/linux/zfs/zfs_vfsops.c +++ b/module/os/linux/zfs/zfs_vfsops.c @@ -1192,7 +1192,7 @@ zfs_prune_aliases(zfsvfs_t *zfsvfs, unsigned long nr_to_scan) int objects = 0; int i = 0, j = 0; - zp_array = kmem_zalloc(max_array * sizeof (znode_t *), KM_SLEEP); + zp_array = vmem_zalloc(max_array * sizeof (znode_t *), KM_SLEEP); mutex_enter(&zfsvfs->z_znodes_lock); while ((zp = list_head(&zfsvfs->z_all_znodes)) != NULL) { @@ -1228,7 +1228,7 @@ zfs_prune_aliases(zfsvfs_t *zfsvfs, unsigned long nr_to_scan) zrele(zp); } - kmem_free(zp_array, max_array * sizeof (znode_t *)); + vmem_free(zp_array, max_array * sizeof (znode_t *)); return (objects); } @@ -1238,12 +1238,18 @@ zfs_prune_aliases(zfsvfs_t *zfsvfs, unsigned long nr_to_scan) * and inode caches. This can occur when the ARC needs to free meta data * blocks but can't because they are all pinned by entries in these caches. */ +#if defined(HAVE_SUPER_BLOCK_S_SHRINK) +#define S_SHRINK(sb) (&(sb)->s_shrink) +#elif defined(HAVE_SUPER_BLOCK_S_SHRINK_PTR) +#define S_SHRINK(sb) ((sb)->s_shrink) +#endif + int zfs_prune(struct super_block *sb, unsigned long nr_to_scan, int *objects) { zfsvfs_t *zfsvfs = sb->s_fs_info; int error = 0; - struct shrinker *shrinker = &sb->s_shrink; + struct shrinker *shrinker = S_SHRINK(sb); struct shrink_control sc = { .nr_to_scan = nr_to_scan, .gfp_mask = GFP_KERNEL, @@ -1254,7 +1260,7 @@ zfs_prune(struct super_block *sb, unsigned long nr_to_scan, int *objects) #if defined(HAVE_SPLIT_SHRINKER_CALLBACK) && \ defined(SHRINK_CONTROL_HAS_NID) && \ defined(SHRINKER_NUMA_AWARE) - if (sb->s_shrink.flags & SHRINKER_NUMA_AWARE) { + if (shrinker->flags & SHRINKER_NUMA_AWARE) { *objects = 0; for_each_online_node(sc.nid) { *objects += (*shrinker->scan_objects)(shrinker, &sc); diff --git a/module/os/linux/zfs/zfs_vnops_os.c b/module/os/linux/zfs/zfs_vnops_os.c index ae0401e60e54..b3c35a034e7e 100644 --- a/module/os/linux/zfs/zfs_vnops_os.c +++ b/module/os/linux/zfs/zfs_vnops_os.c @@ -198,7 +198,7 @@ zfs_open(struct inode *ip, int mode, int flag, cred_t *cr) ZFS_VERIFY_ZP(zp); /* Honor ZFS_APPENDONLY file attribute */ - if ((mode & FMODE_WRITE) && (zp->z_pflags & ZFS_APPENDONLY) && + if (blk_mode_is_open_write(mode) && (zp->z_pflags & ZFS_APPENDONLY) && ((flag & O_APPEND) == 0)) { ZFS_EXIT(zfsvfs); return (SET_ERROR(EPERM)); @@ -244,43 +244,46 @@ zfs_close(struct inode *ip, int flag, cred_t *cr) } #if defined(_KERNEL) + +static int zfs_fillpage(struct inode *ip, struct page *pp); + /* * When a file is memory mapped, we must keep the IO data synchronized - * between the DMU cache and the memory mapped pages. What this means: - * - * On Write: If we find a memory mapped page, we write to *both* - * the page and the dmu buffer. + * between the DMU cache and the memory mapped pages. Update all mapped + * pages with the contents of the coresponding dmu buffer. */ void update_pages(znode_t *zp, int64_t start, int len, objset_t *os) { - struct inode *ip = ZTOI(zp); - struct address_space *mp = ip->i_mapping; - struct page *pp; - uint64_t nbytes; - int64_t off; - void *pb; + struct address_space *mp = ZTOI(zp)->i_mapping; + int64_t off = start & (PAGE_SIZE - 1); - off = start & (PAGE_SIZE-1); for (start &= PAGE_MASK; len > 0; start += PAGE_SIZE) { - nbytes = MIN(PAGE_SIZE - off, len); + uint64_t nbytes = MIN(PAGE_SIZE - off, len); - pp = find_lock_page(mp, start >> PAGE_SHIFT); + struct page *pp = find_lock_page(mp, start >> PAGE_SHIFT); if (pp) { if (mapping_writably_mapped(mp)) flush_dcache_page(pp); - pb = kmap(pp); - (void) dmu_read(os, zp->z_id, start + off, nbytes, - pb + off, DMU_READ_PREFETCH); + void *pb = kmap(pp); + int error = dmu_read(os, zp->z_id, start + off, + nbytes, pb + off, DMU_READ_PREFETCH); kunmap(pp); - if (mapping_writably_mapped(mp)) - flush_dcache_page(pp); + if (error) { + SetPageError(pp); + ClearPageUptodate(pp); + } else { + ClearPageError(pp); + SetPageUptodate(pp); + + if (mapping_writably_mapped(mp)) + flush_dcache_page(pp); + + mark_page_accessed(pp); + } - mark_page_accessed(pp); - SetPageUptodate(pp); - ClearPageError(pp); unlock_page(pp); put_page(pp); } @@ -291,38 +294,44 @@ update_pages(znode_t *zp, int64_t start, int len, objset_t *os) } /* - * When a file is memory mapped, we must keep the IO data synchronized - * between the DMU cache and the memory mapped pages. What this means: - * - * On Read: We "read" preferentially from memory mapped pages, - * else we default from the dmu buffer. - * - * NOTE: We will always "break up" the IO into PAGESIZE uiomoves when - * the file is memory mapped. + * When a file is memory mapped, we must keep the I/O data synchronized + * between the DMU cache and the memory mapped pages. Preferentially read + * from memory mapped pages, otherwise fallback to reading through the dmu. */ int mappedread(znode_t *zp, int nbytes, zfs_uio_t *uio) { struct inode *ip = ZTOI(zp); struct address_space *mp = ip->i_mapping; - struct page *pp; - int64_t start, off; - uint64_t bytes; + int64_t start = uio->uio_loffset; + int64_t off = start & (PAGE_SIZE - 1); int len = nbytes; int error = 0; - void *pb; - start = uio->uio_loffset; - off = start & (PAGE_SIZE-1); for (start &= PAGE_MASK; len > 0; start += PAGE_SIZE) { - bytes = MIN(PAGE_SIZE - off, len); + uint64_t bytes = MIN(PAGE_SIZE - off, len); - pp = find_lock_page(mp, start >> PAGE_SHIFT); + struct page *pp = find_lock_page(mp, start >> PAGE_SHIFT); if (pp) { - ASSERT(PageUptodate(pp)); + /* + * If filemap_fault() retries there exists a window + * where the page will be unlocked and not up to date. + * In this case we must try and fill the page. + */ + if (unlikely(!PageUptodate(pp))) { + error = zfs_fillpage(ip, pp); + if (error) { + unlock_page(pp); + put_page(pp); + return (error); + } + } + + ASSERT(PageUptodate(pp) || PageDirty(pp)); + unlock_page(pp); - pb = kmap(pp); + void *pb = kmap(pp); error = zfs_uiomove(pb + off, bytes, UIO_READ, uio); kunmap(pp); @@ -338,9 +347,11 @@ mappedread(znode_t *zp, int nbytes, zfs_uio_t *uio) len -= bytes; off = 0; + if (error) break; } + return (error); } #endif /* _KERNEL */ @@ -1010,7 +1021,7 @@ top: mutex_enter(&zp->z_lock); may_delete_now = atomic_read(&ZTOI(zp)->i_count) == 1 && - !(zp->z_is_mapped); + !zn_has_cached_data(zp, 0, LLONG_MAX); mutex_exit(&zp->z_lock); /* @@ -1098,7 +1109,8 @@ top: &xattr_obj_unlinked, sizeof (xattr_obj_unlinked)); delete_now = may_delete_now && !toobig && atomic_read(&ZTOI(zp)->i_count) == 1 && - !(zp->z_is_mapped) && xattr_obj == xattr_obj_unlinked && + !zn_has_cached_data(zp, 0, LLONG_MAX) && + xattr_obj == xattr_obj_unlinked && zfs_external_acl(zp) == acl_obj; } @@ -1663,8 +1675,12 @@ out: */ /* ARGSUSED */ int -zfs_getattr_fast(struct user_namespace *user_ns, struct inode *ip, +#ifdef HAVE_GENERIC_FILLATTR_IDMAP_REQMASK +zfs_getattr_fast(zidmap_t *user_ns, u32 request_mask, struct inode *ip, struct kstat *sp) +#else +zfs_getattr_fast(zidmap_t *user_ns, struct inode *ip, struct kstat *sp) +#endif { znode_t *zp = ITOZ(ip); zfsvfs_t *zfsvfs = ITOZSB(ip); @@ -1676,7 +1692,11 @@ zfs_getattr_fast(struct user_namespace *user_ns, struct inode *ip, mutex_enter(&zp->z_lock); +#ifdef HAVE_GENERIC_FILLATTR_IDMAP_REQMASK + zpl_generic_fillattr(user_ns, request_mask, ip, sp); +#else zpl_generic_fillattr(user_ns, ip, sp); +#endif /* * +1 link count for root inode with visible '.zfs' directory. */ @@ -2430,15 +2450,16 @@ top: if ((mask & ATTR_ATIME) || zp->z_atime_dirty) { zp->z_atime_dirty = B_FALSE; - ZFS_TIME_ENCODE(&ip->i_atime, atime); + inode_timespec_t tmp_atime = zpl_inode_get_atime(ip); + ZFS_TIME_ENCODE(&tmp_atime, atime); SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_ATIME(zfsvfs), NULL, &atime, sizeof (atime)); } if (mask & (ATTR_MTIME | ATTR_SIZE)) { ZFS_TIME_ENCODE(&vap->va_mtime, mtime); - ZTOI(zp)->i_mtime = zpl_inode_timestamp_truncate( - vap->va_mtime, ZTOI(zp)); + zpl_inode_set_mtime_to_ts(ZTOI(zp), + zpl_inode_timestamp_truncate(vap->va_mtime, ZTOI(zp))); SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_MTIME(zfsvfs), NULL, mtime, sizeof (mtime)); @@ -2446,8 +2467,8 @@ top: if (mask & (ATTR_CTIME | ATTR_SIZE)) { ZFS_TIME_ENCODE(&vap->va_ctime, ctime); - ZTOI(zp)->i_ctime = zpl_inode_timestamp_truncate(vap->va_ctime, - ZTOI(zp)); + zpl_inode_set_ctime_to_ts(ZTOI(zp), + zpl_inode_timestamp_truncate(vap->va_ctime, ZTOI(zp))); SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_CTIME(zfsvfs), NULL, ctime, sizeof (ctime)); } @@ -3434,7 +3455,7 @@ top: } static void -zfs_putpage_commit_cb(void *arg) +zfs_putpage_sync_commit_cb(void *arg) { struct page *pp = arg; @@ -3442,13 +3463,26 @@ zfs_putpage_commit_cb(void *arg) end_page_writeback(pp); } +static void +zfs_putpage_async_commit_cb(void *arg) +{ + struct page *pp = arg; + znode_t *zp = ITOZ(pp->mapping->host); + + ClearPageError(pp); + end_page_writeback(pp); + atomic_dec_32(&zp->z_async_writes_cnt); +} + /* * Push a page out to disk, once the page is on stable storage the * registered commit callback will be run as notification of completion. * - * IN: ip - page mapped for inode. - * pp - page to push (page is locked) - * wbc - writeback control data + * IN: ip - page mapped for inode. + * pp - page to push (page is locked) + * wbc - writeback control data + * for_sync - does the caller intend to wait synchronously for the + * page writeback to complete? * * RETURN: 0 if success * error code if failure @@ -3458,7 +3492,8 @@ zfs_putpage_commit_cb(void *arg) */ /* ARGSUSED */ int -zfs_putpage(struct inode *ip, struct page *pp, struct writeback_control *wbc) +zfs_putpage(struct inode *ip, struct page *pp, struct writeback_control *wbc, + boolean_t for_sync) { znode_t *zp = ITOZ(ip); zfsvfs_t *zfsvfs = ITOZSB(ip); @@ -3469,6 +3504,7 @@ zfs_putpage(struct inode *ip, struct page *pp, struct writeback_control *wbc) caddr_t va; int err = 0; uint64_t mtime[2], ctime[2]; + inode_timespec_t tmp_ts; sa_bulk_attr_t bulk[3]; int cnt = 0; struct address_space *mapping; @@ -3556,6 +3592,16 @@ zfs_putpage(struct inode *ip, struct page *pp, struct writeback_control *wbc) zfs_rangelock_exit(lr); if (wbc->sync_mode != WB_SYNC_NONE) { + /* + * Speed up any non-sync page writebacks since + * they may take several seconds to complete. + * Refer to the comment in zpl_fsync() (when + * HAVE_FSYNC_RANGE is defined) for details. + */ + if (atomic_load_32(&zp->z_async_writes_cnt) > 0) { + zil_commit(zfsvfs->z_log, zp->z_id); + } + if (PageWriteback(pp)) #ifdef HAVE_PAGEMAP_FOLIO_WAIT_BIT folio_wait_bit(page_folio(pp), PG_writeback); @@ -3581,6 +3627,8 @@ zfs_putpage(struct inode *ip, struct page *pp, struct writeback_control *wbc) * was in fact not skipped and should not be counted as if it were. */ wbc->pages_skipped--; + if (!for_sync) + atomic_inc_32(&zp->z_async_writes_cnt); set_page_writeback(pp); unlock_page(pp); @@ -3602,6 +3650,8 @@ zfs_putpage(struct inode *ip, struct page *pp, struct writeback_control *wbc) #endif ClearPageError(pp); end_page_writeback(pp); + if (!for_sync) + atomic_dec_32(&zp->z_async_writes_cnt); zfs_rangelock_exit(lr); ZFS_EXIT(zfsvfs); return (err); @@ -3618,15 +3668,19 @@ zfs_putpage(struct inode *ip, struct page *pp, struct writeback_control *wbc) &zp->z_pflags, 8); /* Preserve the mtime and ctime provided by the inode */ - ZFS_TIME_ENCODE(&ip->i_mtime, mtime); - ZFS_TIME_ENCODE(&ip->i_ctime, ctime); + tmp_ts = zpl_inode_get_mtime(ip); + ZFS_TIME_ENCODE(&tmp_ts, mtime); + tmp_ts = zpl_inode_get_ctime(ip); + ZFS_TIME_ENCODE(&tmp_ts, ctime); zp->z_atime_dirty = B_FALSE; zp->z_seq++; err = sa_bulk_update(zp->z_sa_hdl, bulk, cnt, tx); zfs_log_write(zfsvfs->z_log, tx, TX_WRITE, zp, pgoff, pglen, 0, - zfs_putpage_commit_cb, pp); + for_sync ? zfs_putpage_sync_commit_cb : + zfs_putpage_async_commit_cb, pp); + dmu_tx_commit(tx); zfs_rangelock_exit(lr); @@ -3638,6 +3692,16 @@ zfs_putpage(struct inode *ip, struct page *pp, struct writeback_control *wbc) * performance reasons. */ zil_commit(zfsvfs->z_log, zp->z_id); + } else if (!for_sync && atomic_load_32(&zp->z_sync_writes_cnt) > 0) { + /* + * If the caller does not intend to wait synchronously + * for this page writeback to complete and there are active + * synchronous calls on this file, do a commit so that + * the latter don't accidentally end up waiting for + * our writeback to complete. Refer to the comment in + * zpl_fsync() (when HAVE_FSYNC_RANGE is defined) for details. + */ + zil_commit(zfsvfs->z_log, zp->z_id); } dataset_kstats_update_write_kstats(&zfsvfs->z_kstat, pglen); @@ -3657,6 +3721,7 @@ zfs_dirty_inode(struct inode *ip, int flags) zfsvfs_t *zfsvfs = ITOZSB(ip); dmu_tx_t *tx; uint64_t mode, atime[2], mtime[2], ctime[2]; + inode_timespec_t tmp_ts; sa_bulk_attr_t bulk[4]; int error = 0; int cnt = 0; @@ -3701,9 +3766,12 @@ zfs_dirty_inode(struct inode *ip, int flags) SA_ADD_BULK_ATTR(bulk, cnt, SA_ZPL_CTIME(zfsvfs), NULL, &ctime, 16); /* Preserve the mode, mtime and ctime provided by the inode */ - ZFS_TIME_ENCODE(&ip->i_atime, atime); - ZFS_TIME_ENCODE(&ip->i_mtime, mtime); - ZFS_TIME_ENCODE(&ip->i_ctime, ctime); + tmp_ts = zpl_inode_get_atime(ip); + ZFS_TIME_ENCODE(&tmp_ts, atime); + tmp_ts = zpl_inode_get_mtime(ip); + ZFS_TIME_ENCODE(&tmp_ts, mtime); + tmp_ts = zpl_inode_get_ctime(ip); + ZFS_TIME_ENCODE(&tmp_ts, ctime); mode = ip->i_mode; zp->z_mode = mode; @@ -3747,7 +3815,9 @@ zfs_inactive(struct inode *ip) if (error) { dmu_tx_abort(tx); } else { - ZFS_TIME_ENCODE(&ip->i_atime, atime); + inode_timespec_t tmp_atime; + tmp_atime = zpl_inode_get_atime(ip); + ZFS_TIME_ENCODE(&tmp_atime, atime); mutex_enter(&zp->z_lock); (void) sa_update(zp->z_sa_hdl, SA_ZPL_ATIME(zfsvfs), (void *)&atime, sizeof (atime), tx); @@ -3766,55 +3836,45 @@ zfs_inactive(struct inode *ip) * Fill pages with data from the disk. */ static int -zfs_fillpage(struct inode *ip, struct page *pl[], int nr_pages) +zfs_fillpage(struct inode *ip, struct page *pp) { - znode_t *zp = ITOZ(ip); zfsvfs_t *zfsvfs = ITOZSB(ip); - objset_t *os; - struct page *cur_pp; - u_offset_t io_off, total; - size_t io_len; - loff_t i_size; - unsigned page_idx; - int err; + loff_t i_size = i_size_read(ip); + u_offset_t io_off = page_offset(pp); + size_t io_len = PAGE_SIZE; - os = zfsvfs->z_os; - io_len = nr_pages << PAGE_SHIFT; - i_size = i_size_read(ip); - io_off = page_offset(pl[0]); + ASSERT3U(io_off, <, i_size); if (io_off + io_len > i_size) io_len = i_size - io_off; - /* - * Iterate over list of pages and read each page individually. - */ - page_idx = 0; - for (total = io_off + io_len; io_off < total; io_off += PAGESIZE) { - caddr_t va; + void *va = kmap(pp); + int error = dmu_read(zfsvfs->z_os, ITOZ(ip)->z_id, io_off, + io_len, va, DMU_READ_PREFETCH); + if (io_len != PAGE_SIZE) + memset((char *)va + io_len, 0, PAGE_SIZE - io_len); + kunmap(pp); - cur_pp = pl[page_idx++]; - va = kmap(cur_pp); - err = dmu_read(os, zp->z_id, io_off, PAGESIZE, va, - DMU_READ_PREFETCH); - kunmap(cur_pp); - if (err) { - /* convert checksum errors into IO errors */ - if (err == ECKSUM) - err = SET_ERROR(EIO); - return (err); - } + if (error) { + /* convert checksum errors into IO errors */ + if (error == ECKSUM) + error = SET_ERROR(EIO); + + SetPageError(pp); + ClearPageUptodate(pp); + } else { + ClearPageError(pp); + SetPageUptodate(pp); } - return (0); + return (error); } /* - * Uses zfs_fillpage to read data from the file and fill the pages. + * Uses zfs_fillpage to read data from the file and fill the page. * * IN: ip - inode of file to get data from. - * pl - list of pages to read - * nr_pages - number of pages to read + * pp - page to read * * RETURN: 0 on success, error code on failure. * @@ -3823,24 +3883,22 @@ zfs_fillpage(struct inode *ip, struct page *pl[], int nr_pages) */ /* ARGSUSED */ int -zfs_getpage(struct inode *ip, struct page *pl[], int nr_pages) +zfs_getpage(struct inode *ip, struct page *pp) { - znode_t *zp = ITOZ(ip); zfsvfs_t *zfsvfs = ITOZSB(ip); - int err; - - if (pl == NULL) - return (0); + znode_t *zp = ITOZ(ip); + int error; ZFS_ENTER(zfsvfs); ZFS_VERIFY_ZP(zp); - err = zfs_fillpage(ip, pl, nr_pages); - - dataset_kstats_update_read_kstats(&zfsvfs->z_kstat, nr_pages*PAGESIZE); + error = zfs_fillpage(ip, pp); + if (error == 0) + dataset_kstats_update_read_kstats(&zfsvfs->z_kstat, PAGE_SIZE); ZFS_EXIT(zfsvfs); - return (err); + + return (error); } /* diff --git a/module/os/linux/zfs/zfs_znode.c b/module/os/linux/zfs/zfs_znode.c index f3475b4d9794..d1e64d0a5cbb 100644 --- a/module/os/linux/zfs/zfs_znode.c +++ b/module/os/linux/zfs/zfs_znode.c @@ -134,6 +134,9 @@ zfs_znode_cache_constructor(void *buf, void *arg, int kmflags) zp->z_acl_cached = NULL; zp->z_xattr_cached = NULL; zp->z_xattr_parent = 0; + zp->z_sync_writes_cnt = 0; + zp->z_async_writes_cnt = 0; + return (0); } @@ -151,9 +154,12 @@ zfs_znode_cache_destructor(void *buf, void *arg) rw_destroy(&zp->z_xattr_lock); zfs_rangelock_fini(&zp->z_rangelock); - ASSERT(zp->z_dirlocks == NULL); - ASSERT(zp->z_acl_cached == NULL); - ASSERT(zp->z_xattr_cached == NULL); + ASSERT3P(zp->z_dirlocks, ==, NULL); + ASSERT3P(zp->z_acl_cached, ==, NULL); + ASSERT3P(zp->z_xattr_cached, ==, NULL); + + ASSERT0(atomic_load_32(&zp->z_sync_writes_cnt)); + ASSERT0(atomic_load_32(&zp->z_async_writes_cnt)); } static int @@ -524,6 +530,7 @@ zfs_znode_alloc(zfsvfs_t *zfsvfs, dmu_buf_t *db, int blksz, uint64_t links; uint64_t z_uid, z_gid; uint64_t atime[2], mtime[2], ctime[2], btime[2]; + inode_timespec_t tmp_ts; uint64_t projid = ZFS_DEFAULT_PROJID; sa_bulk_attr_t bulk[12]; int count = 0; @@ -540,7 +547,9 @@ zfs_znode_alloc(zfsvfs_t *zfsvfs, dmu_buf_t *db, int blksz, ASSERT3P(zp->z_xattr_cached, ==, NULL); zp->z_unlinked = B_FALSE; zp->z_atime_dirty = B_FALSE; +#if !defined(HAVE_FILEMAP_RANGE_HAS_PAGE) zp->z_is_mapped = B_FALSE; +#endif zp->z_is_ctldir = B_FALSE; zp->z_suspended = B_FALSE; zp->z_sa_hdl = NULL; @@ -549,6 +558,8 @@ zfs_znode_alloc(zfsvfs_t *zfsvfs, dmu_buf_t *db, int blksz, zp->z_blksz = blksz; zp->z_seq = 0x7A4653; zp->z_sync_cnt = 0; + zp->z_sync_writes_cnt = 0; + zp->z_async_writes_cnt = 0; zfs_znode_sa_init(zfsvfs, zp, db, obj_type, hdl); @@ -591,9 +602,12 @@ zfs_znode_alloc(zfsvfs_t *zfsvfs, dmu_buf_t *db, int blksz, if (zp->z_pflags & ZFS_XATTR) zp->z_xattr_parent = parent; - ZFS_TIME_DECODE(&ip->i_atime, atime); - ZFS_TIME_DECODE(&ip->i_mtime, mtime); - ZFS_TIME_DECODE(&ip->i_ctime, ctime); + ZFS_TIME_DECODE(&tmp_ts, atime); + zpl_inode_set_atime_to_ts(ip, tmp_ts); + ZFS_TIME_DECODE(&tmp_ts, mtime); + zpl_inode_set_mtime_to_ts(ip, tmp_ts); + ZFS_TIME_DECODE(&tmp_ts, ctime); + zpl_inode_set_ctime_to_ts(ip, tmp_ts); ZFS_TIME_DECODE(&zp->z_btime, btime); ip->i_ino = zp->z_id; @@ -1174,6 +1188,7 @@ zfs_rezget(znode_t *zp) uint64_t gen; uint64_t z_uid, z_gid; uint64_t atime[2], mtime[2], ctime[2], btime[2]; + inode_timespec_t tmp_ts; uint64_t projid = ZFS_DEFAULT_PROJID; znode_hold_t *zh; @@ -1266,9 +1281,12 @@ zfs_rezget(znode_t *zp) zfs_uid_write(ZTOI(zp), z_uid); zfs_gid_write(ZTOI(zp), z_gid); - ZFS_TIME_DECODE(&ZTOI(zp)->i_atime, atime); - ZFS_TIME_DECODE(&ZTOI(zp)->i_mtime, mtime); - ZFS_TIME_DECODE(&ZTOI(zp)->i_ctime, ctime); + ZFS_TIME_DECODE(&tmp_ts, atime); + zpl_inode_set_atime_to_ts(ZTOI(zp), tmp_ts); + ZFS_TIME_DECODE(&tmp_ts, mtime); + zpl_inode_set_mtime_to_ts(ZTOI(zp), tmp_ts); + ZFS_TIME_DECODE(&tmp_ts, ctime); + zpl_inode_set_ctime_to_ts(ZTOI(zp), tmp_ts); ZFS_TIME_DECODE(&zp->z_btime, btime); if ((uint32_t)gen != ZTOI(zp)->i_generation) { @@ -1376,21 +1394,24 @@ zfs_zinactive(znode_t *zp) boolean_t zfs_relatime_need_update(const struct inode *ip) { - inode_timespec_t now; + inode_timespec_t now, tmp_atime, tmp_ts; gethrestime(&now); + tmp_atime = zpl_inode_get_atime(ip); /* * In relatime mode, only update the atime if the previous atime * is earlier than either the ctime or mtime or if at least a day * has passed since the last update of atime. */ - if (zfs_compare_timespec(&ip->i_mtime, &ip->i_atime) >= 0) + tmp_ts = zpl_inode_get_mtime(ip); + if (zfs_compare_timespec(&tmp_ts, &tmp_atime) >= 0) return (B_TRUE); - if (zfs_compare_timespec(&ip->i_ctime, &ip->i_atime) >= 0) + tmp_ts = zpl_inode_get_ctime(ip); + if (zfs_compare_timespec(&tmp_ts, &tmp_atime) >= 0) return (B_TRUE); - if ((hrtime_t)now.tv_sec - (hrtime_t)ip->i_atime.tv_sec >= 24*60*60) + if ((hrtime_t)now.tv_sec - (hrtime_t)tmp_atime.tv_sec >= 24*60*60) return (B_TRUE); return (B_FALSE); @@ -1413,7 +1434,7 @@ void zfs_tstamp_update_setup(znode_t *zp, uint_t flag, uint64_t mtime[2], uint64_t ctime[2]) { - inode_timespec_t now; + inode_timespec_t now, tmp_ts; gethrestime(&now); @@ -1421,7 +1442,8 @@ zfs_tstamp_update_setup(znode_t *zp, uint_t flag, uint64_t mtime[2], if (flag & ATTR_MTIME) { ZFS_TIME_ENCODE(&now, mtime); - ZFS_TIME_DECODE(&(ZTOI(zp)->i_mtime), mtime); + ZFS_TIME_DECODE(&tmp_ts, mtime); + zpl_inode_set_mtime_to_ts(ZTOI(zp), tmp_ts); if (ZTOZSB(zp)->z_use_fuids) { zp->z_pflags |= (ZFS_ARCHIVE | ZFS_AV_MODIFIED); @@ -1430,7 +1452,8 @@ zfs_tstamp_update_setup(znode_t *zp, uint_t flag, uint64_t mtime[2], if (flag & ATTR_CTIME) { ZFS_TIME_ENCODE(&now, ctime); - ZFS_TIME_DECODE(&(ZTOI(zp)->i_ctime), ctime); + ZFS_TIME_DECODE(&tmp_ts, ctime); + zpl_inode_set_ctime_to_ts(ZTOI(zp), tmp_ts); if (ZTOZSB(zp)->z_use_fuids) zp->z_pflags |= ZFS_ARCHIVE; } @@ -1628,7 +1651,7 @@ zfs_free_range(znode_t *zp, uint64_t off, uint64_t len) * Zero partial page cache entries. This must be done under a * range lock in order to keep the ARC and page cache in sync. */ - if (zp->z_is_mapped) { + if (zn_has_cached_data(zp, off, off + len - 1)) { loff_t first_page, last_page, page_len; loff_t first_page_offset, last_page_offset; diff --git a/module/os/linux/zfs/zio_crypt.c b/module/os/linux/zfs/zio_crypt.c index 50e93909659f..3860885292c5 100644 --- a/module/os/linux/zfs/zio_crypt.c +++ b/module/os/linux/zfs/zio_crypt.c @@ -229,8 +229,27 @@ zio_crypt_key_init(uint64_t crypt, zio_crypt_key_t *key) ASSERT(key != NULL); ASSERT3U(crypt, <, ZIO_CRYPT_FUNCTIONS); +/* + * Workaround for GCC 12+ with UBSan enabled deficencies. + * + * GCC 12+ invoked with -fsanitize=undefined incorrectly reports the code + * below as violating -Warray-bounds + */ +#if defined(__GNUC__) && !defined(__clang__) && \ + ((!defined(_KERNEL) && defined(ZFS_UBSAN_ENABLED)) || \ + defined(CONFIG_UBSAN)) +#pragma GCC diagnostic push +#pragma GCC diagnostic ignored "-Warray-bounds" +#endif keydata_len = zio_crypt_table[crypt].ci_keylen; bzero(key, sizeof (zio_crypt_key_t)); +#if defined(__GNUC__) && !defined(__clang__) && \ + ((!defined(_KERNEL) && defined(ZFS_UBSAN_ENABLED)) || \ + defined(CONFIG_UBSAN)) +#pragma GCC diagnostic pop +#endif + memset(key, 0, sizeof (zio_crypt_key_t)); + rw_init(&key->zk_salt_lock, NULL, RW_DEFAULT, NULL); /* fill keydata buffers and salt with random data */ ret = random_get_bytes((uint8_t *)&key->zk_guid, sizeof (uint64_t)); @@ -284,7 +303,6 @@ zio_crypt_key_init(uint64_t crypt, zio_crypt_key_t *key) key->zk_crypt = crypt; key->zk_version = ZIO_CRYPT_KEY_CURRENT_VERSION; key->zk_salt_count = 0; - rw_init(&key->zk_salt_lock, NULL, RW_DEFAULT, NULL); return (0); diff --git a/module/os/linux/zfs/zpl_ctldir.c b/module/os/linux/zfs/zpl_ctldir.c index 9b526afd0002..cb8824cc8946 100644 --- a/module/os/linux/zfs/zpl_ctldir.c +++ b/module/os/linux/zfs/zpl_ctldir.c @@ -40,7 +40,7 @@ static int zpl_common_open(struct inode *ip, struct file *filp) { - if (filp->f_mode & FMODE_WRITE) + if (blk_mode_is_open_write(filp->f_mode)) return (-EACCES); return (generic_file_open(ip, filp)); @@ -101,7 +101,11 @@ zpl_root_readdir(struct file *filp, void *dirent, filldir_t filldir) */ /* ARGSUSED */ static int -#ifdef HAVE_USERNS_IOPS_GETATTR +#ifdef HAVE_IDMAP_IOPS_GETATTR +zpl_root_getattr_impl(struct mnt_idmap *user_ns, + const struct path *path, struct kstat *stat, u32 request_mask, + unsigned int query_flags) +#elif defined(HAVE_USERNS_IOPS_GETATTR) zpl_root_getattr_impl(struct user_namespace *user_ns, const struct path *path, struct kstat *stat, u32 request_mask, unsigned int query_flags) @@ -112,8 +116,16 @@ zpl_root_getattr_impl(const struct path *path, struct kstat *stat, { struct inode *ip = path->dentry->d_inode; -#if defined(HAVE_GENERIC_FILLATTR_USERNS) && defined(HAVE_USERNS_IOPS_GETATTR) +#if (defined(HAVE_USERNS_IOPS_GETATTR) || defined(HAVE_IDMAP_IOPS_GETATTR)) +#ifdef HAVE_GENERIC_FILLATTR_USERNS generic_fillattr(user_ns, ip, stat); +#elif defined(HAVE_GENERIC_FILLATTR_IDMAP) + generic_fillattr(user_ns, ip, stat); +#elif defined(HAVE_GENERIC_FILLATTR_IDMAP_REQMASK) + generic_fillattr(user_ns, request_mask, ip, stat); +#else + (void) user_ns; +#endif #else generic_fillattr(ip, stat); #endif @@ -304,6 +316,10 @@ static int zpl_snapdir_rename2(struct user_namespace *user_ns, struct inode *sdip, struct dentry *sdentry, struct inode *tdip, struct dentry *tdentry, unsigned int flags) +#elif defined(HAVE_IOPS_RENAME_IDMAP) +zpl_snapdir_rename2(struct mnt_idmap *user_ns, struct inode *sdip, + struct dentry *sdentry, struct inode *tdip, struct dentry *tdentry, + unsigned int flags) #else zpl_snapdir_rename2(struct inode *sdip, struct dentry *sdentry, struct inode *tdip, struct dentry *tdentry, unsigned int flags) @@ -325,7 +341,9 @@ zpl_snapdir_rename2(struct inode *sdip, struct dentry *sdentry, return (error); } -#if !defined(HAVE_RENAME_WANTS_FLAGS) && !defined(HAVE_IOPS_RENAME_USERNS) +#if (!defined(HAVE_RENAME_WANTS_FLAGS) && \ + !defined(HAVE_IOPS_RENAME_USERNS) && \ + !defined(HAVE_IOPS_RENAME_IDMAP)) static int zpl_snapdir_rename(struct inode *sdip, struct dentry *sdentry, struct inode *tdip, struct dentry *tdentry) @@ -352,6 +370,9 @@ static int #ifdef HAVE_IOPS_MKDIR_USERNS zpl_snapdir_mkdir(struct user_namespace *user_ns, struct inode *dip, struct dentry *dentry, umode_t mode) +#elif defined(HAVE_IOPS_MKDIR_IDMAP) +zpl_snapdir_mkdir(struct mnt_idmap *user_ns, struct inode *dip, + struct dentry *dentry, umode_t mode) #else zpl_snapdir_mkdir(struct inode *dip, struct dentry *dentry, umode_t mode) #endif @@ -384,7 +405,11 @@ zpl_snapdir_mkdir(struct inode *dip, struct dentry *dentry, umode_t mode) */ /* ARGSUSED */ static int -#ifdef HAVE_USERNS_IOPS_GETATTR +#ifdef HAVE_IDMAP_IOPS_GETATTR +zpl_snapdir_getattr_impl(struct mnt_idmap *user_ns, + const struct path *path, struct kstat *stat, u32 request_mask, + unsigned int query_flags) +#elif defined(HAVE_USERNS_IOPS_GETATTR) zpl_snapdir_getattr_impl(struct user_namespace *user_ns, const struct path *path, struct kstat *stat, u32 request_mask, unsigned int query_flags) @@ -397,8 +422,16 @@ zpl_snapdir_getattr_impl(const struct path *path, struct kstat *stat, zfsvfs_t *zfsvfs = ITOZSB(ip); ZPL_ENTER(zfsvfs); -#if defined(HAVE_GENERIC_FILLATTR_USERNS) && defined(HAVE_USERNS_IOPS_GETATTR) +#if (defined(HAVE_USERNS_IOPS_GETATTR) || defined(HAVE_IDMAP_IOPS_GETATTR)) +#ifdef HAVE_GENERIC_FILLATTR_USERNS + generic_fillattr(user_ns, ip, stat); +#elif defined(HAVE_GENERIC_FILLATTR_IDMAP) generic_fillattr(user_ns, ip, stat); +#elif defined(HAVE_GENERIC_FILLATTR_IDMAP_REQMASK) + generic_fillattr(user_ns, request_mask, ip, stat); +#else + (void) user_ns; +#endif #else generic_fillattr(ip, stat); #endif @@ -439,7 +472,9 @@ const struct file_operations zpl_fops_snapdir = { const struct inode_operations zpl_ops_snapdir = { .lookup = zpl_snapdir_lookup, .getattr = zpl_snapdir_getattr, -#if defined(HAVE_RENAME_WANTS_FLAGS) || defined(HAVE_IOPS_RENAME_USERNS) +#if (defined(HAVE_RENAME_WANTS_FLAGS) || \ + defined(HAVE_IOPS_RENAME_USERNS) || \ + defined(HAVE_IOPS_RENAME_IDMAP)) .rename = zpl_snapdir_rename2, #else .rename = zpl_snapdir_rename, @@ -530,6 +565,10 @@ static int zpl_shares_getattr_impl(struct user_namespace *user_ns, const struct path *path, struct kstat *stat, u32 request_mask, unsigned int query_flags) +#elif defined(HAVE_IDMAP_IOPS_GETATTR) +zpl_shares_getattr_impl(struct mnt_idmap *user_ns, + const struct path *path, struct kstat *stat, u32 request_mask, + unsigned int query_flags) #else zpl_shares_getattr_impl(const struct path *path, struct kstat *stat, u32 request_mask, unsigned int query_flags) @@ -543,8 +582,16 @@ zpl_shares_getattr_impl(const struct path *path, struct kstat *stat, ZPL_ENTER(zfsvfs); if (zfsvfs->z_shares_dir == 0) { -#if defined(HAVE_GENERIC_FILLATTR_USERNS) && defined(HAVE_USERNS_IOPS_GETATTR) +#if (defined(HAVE_USERNS_IOPS_GETATTR) || defined(HAVE_IDMAP_IOPS_GETATTR)) +#ifdef HAVE_GENERIC_FILLATTR_USERNS + generic_fillattr(user_ns, path->dentry->d_inode, stat); +#elif defined(HAVE_GENERIC_FILLATTR_IDMAP) generic_fillattr(user_ns, path->dentry->d_inode, stat); +#elif defined(HAVE_GENERIC_FILLATTR_IDMAP_REQMASK) + generic_fillattr(user_ns, request_mask, ip, stat); +#else + (void) user_ns; +#endif #else generic_fillattr(path->dentry->d_inode, stat); #endif @@ -556,7 +603,10 @@ zpl_shares_getattr_impl(const struct path *path, struct kstat *stat, error = -zfs_zget(zfsvfs, zfsvfs->z_shares_dir, &dzp); if (error == 0) { -#if defined(HAVE_GENERIC_FILLATTR_USERNS) && defined(HAVE_USERNS_IOPS_GETATTR) +#ifdef HAVE_GENERIC_FILLATTR_IDMAP_REQMASK + error = -zfs_getattr_fast(user_ns, request_mask, ZTOI(dzp), + stat); +#elif (defined(HAVE_USERNS_IOPS_GETATTR) || defined(HAVE_IDMAP_IOPS_GETATTR)) error = -zfs_getattr_fast(user_ns, ZTOI(dzp), stat); #else error = -zfs_getattr_fast(kcred->user_ns, ZTOI(dzp), stat); diff --git a/module/os/linux/zfs/zpl_file.c b/module/os/linux/zfs/zpl_file.c index 38d2bd147007..c880e015ae1c 100644 --- a/module/os/linux/zfs/zpl_file.c +++ b/module/os/linux/zfs/zpl_file.c @@ -165,17 +165,56 @@ static int zpl_fsync(struct file *filp, loff_t start, loff_t end, int datasync) { struct inode *inode = filp->f_mapping->host; + znode_t *zp = ITOZ(inode); + zfsvfs_t *zfsvfs = ITOZSB(inode); cred_t *cr = CRED(); int error; fstrans_cookie_t cookie; + /* + * The variables z_sync_writes_cnt and z_async_writes_cnt work in + * tandem so that sync writes can detect if there are any non-sync + * writes going on and vice-versa. The "vice-versa" part to this logic + * is located in zfs_putpage() where non-sync writes check if there are + * any ongoing sync writes. If any sync and non-sync writes overlap, + * we do a commit to complete the non-sync writes since the latter can + * potentially take several seconds to complete and thus block sync + * writes in the upcoming call to filemap_write_and_wait_range(). + */ + atomic_inc_32(&zp->z_sync_writes_cnt); + /* + * If the following check does not detect an overlapping non-sync write + * (say because it's just about to start), then it is guaranteed that + * the non-sync write will detect this sync write. This is because we + * always increment z_sync_writes_cnt / z_async_writes_cnt before doing + * the check on z_async_writes_cnt / z_sync_writes_cnt here and in + * zfs_putpage() respectively. + */ + if (atomic_load_32(&zp->z_async_writes_cnt) > 0) { + ZPL_ENTER(zfsvfs); + zil_commit(zfsvfs->z_log, zp->z_id); + ZPL_EXIT(zfsvfs); + } + error = filemap_write_and_wait_range(inode->i_mapping, start, end); + + /* + * The sync write is not complete yet but we decrement + * z_sync_writes_cnt since zfs_fsync() increments and decrements + * it internally. If a non-sync write starts just after the decrement + * operation but before we call zfs_fsync(), it may not detect this + * overlapping sync write but it does not matter since we have already + * gone past filemap_write_and_wait_range() and we won't block due to + * the non-sync write. + */ + atomic_dec_32(&zp->z_sync_writes_cnt); + if (error) return (error); crhold(cr); cookie = spl_fstrans_mark(); - error = -zfs_fsync(ITOZ(inode), datasync, cr); + error = -zfs_fsync(zp, datasync, cr); spl_fstrans_unmark(cookie); crfree(cr); ASSERT3S(error, <=, 0); @@ -255,15 +294,10 @@ zpl_uio_init(zfs_uio_t *uio, struct kiocb *kiocb, struct iov_iter *to, #if defined(HAVE_VFS_IOV_ITER) zfs_uio_iov_iter_init(uio, to, pos, count, skip); #else -#ifdef HAVE_IOV_ITER_TYPE - zfs_uio_iovec_init(uio, to->iov, to->nr_segs, pos, - iov_iter_type(to) & ITER_KVEC ? UIO_SYSSPACE : UIO_USERSPACE, + zfs_uio_iovec_init(uio, zfs_uio_iter_iov(to), to->nr_segs, pos, + zfs_uio_iov_iter_type(to) & ITER_KVEC ? + UIO_SYSSPACE : UIO_USERSPACE, count, skip); -#else - zfs_uio_iovec_init(uio, to->iov, to->nr_segs, pos, - to->type & ITER_KVEC ? UIO_SYSSPACE : UIO_USERSPACE, - count, skip); -#endif #endif } @@ -579,7 +613,6 @@ static int zpl_mmap(struct file *filp, struct vm_area_struct *vma) { struct inode *ip = filp->f_mapping->host; - znode_t *zp = ITOZ(ip); int error; fstrans_cookie_t cookie; @@ -594,9 +627,12 @@ zpl_mmap(struct file *filp, struct vm_area_struct *vma) if (error) return (error); +#if !defined(HAVE_FILEMAP_RANGE_HAS_PAGE) + znode_t *zp = ITOZ(ip); mutex_enter(&zp->z_lock); zp->z_is_mapped = B_TRUE; mutex_exit(&zp->z_lock); +#endif return (error); } @@ -609,29 +645,16 @@ zpl_mmap(struct file *filp, struct vm_area_struct *vma) static inline int zpl_readpage_common(struct page *pp) { - struct inode *ip; - struct page *pl[1]; - int error = 0; fstrans_cookie_t cookie; ASSERT(PageLocked(pp)); - ip = pp->mapping->host; - pl[0] = pp; cookie = spl_fstrans_mark(); - error = -zfs_getpage(ip, pl, 1); + int error = -zfs_getpage(pp->mapping->host, pp); spl_fstrans_unmark(cookie); - if (error) { - SetPageError(pp); - ClearPageUptodate(pp); - } else { - ClearPageError(pp); - SetPageUptodate(pp); - flush_dcache_page(pp); - } - unlock_page(pp); + return (error); } @@ -688,19 +711,42 @@ zpl_readahead(struct readahead_control *ractl) static int zpl_putpage(struct page *pp, struct writeback_control *wbc, void *data) { - struct address_space *mapping = data; + boolean_t *for_sync = data; fstrans_cookie_t cookie; ASSERT(PageLocked(pp)); ASSERT(!PageWriteback(pp)); cookie = spl_fstrans_mark(); - (void) zfs_putpage(mapping->host, pp, wbc); + (void) zfs_putpage(pp->mapping->host, pp, wbc, *for_sync); spl_fstrans_unmark(cookie); return (0); } +#ifdef HAVE_WRITEPAGE_T_FOLIO +static int +zpl_putfolio(struct folio *pp, struct writeback_control *wbc, void *data) +{ + (void) zpl_putpage(&pp->page, wbc, data); + return (0); +} +#endif + +static inline int +zpl_write_cache_pages(struct address_space *mapping, + struct writeback_control *wbc, void *data) +{ + int result; + +#ifdef HAVE_WRITEPAGE_T_FOLIO + result = write_cache_pages(mapping, wbc, zpl_putfolio, data); +#else + result = write_cache_pages(mapping, wbc, zpl_putpage, data); +#endif + return (result); +} + static int zpl_writepages(struct address_space *mapping, struct writeback_control *wbc) { @@ -722,8 +768,9 @@ zpl_writepages(struct address_space *mapping, struct writeback_control *wbc) * we run it once in non-SYNC mode so that the ZIL gets all the data, * and then we commit it all in one go. */ + boolean_t for_sync = (sync_mode == WB_SYNC_ALL); wbc->sync_mode = WB_SYNC_NONE; - result = write_cache_pages(mapping, wbc, zpl_putpage, mapping); + result = zpl_write_cache_pages(mapping, wbc, &for_sync); if (sync_mode != wbc->sync_mode) { ZPL_ENTER(zfsvfs); ZPL_VERIFY_ZP(zp); @@ -739,7 +786,7 @@ zpl_writepages(struct address_space *mapping, struct writeback_control *wbc) * details). That being said, this is a no-op in most cases. */ wbc->sync_mode = sync_mode; - result = write_cache_pages(mapping, wbc, zpl_putpage, mapping); + result = zpl_write_cache_pages(mapping, wbc, &for_sync); } return (result); } @@ -756,7 +803,9 @@ zpl_writepage(struct page *pp, struct writeback_control *wbc) if (ITOZSB(pp->mapping->host)->z_os->os_sync == ZFS_SYNC_ALWAYS) wbc->sync_mode = WB_SYNC_ALL; - return (zpl_putpage(pp, wbc, pp->mapping)); + boolean_t for_sync = (wbc->sync_mode == WB_SYNC_ALL); + + return (zpl_putpage(pp, wbc, &for_sync)); } /* @@ -924,7 +973,7 @@ __zpl_ioctl_setflags(struct inode *ip, uint32_t ioctl_flags, xvattr_t *xva) !capable(CAP_LINUX_IMMUTABLE)) return (-EPERM); - if (!zpl_inode_owner_or_capable(kcred->user_ns, ip)) + if (!zpl_inode_owner_or_capable(zfs_init_idmap, ip)) return (-EACCES); xva_init(xva); @@ -1093,7 +1142,11 @@ const struct file_operations zpl_file_operations = { .read_iter = zpl_iter_read, .write_iter = zpl_iter_write, #ifdef HAVE_VFS_IOV_ITER +#ifdef HAVE_COPY_SPLICE_READ + .splice_read = copy_splice_read, +#else .splice_read = generic_file_splice_read, +#endif .splice_write = iter_file_splice_write, #endif #else diff --git a/module/os/linux/zfs/zpl_inode.c b/module/os/linux/zfs/zpl_inode.c index dd634f70eb75..4bd185baeb19 100644 --- a/module/os/linux/zfs/zpl_inode.c +++ b/module/os/linux/zfs/zpl_inode.c @@ -131,6 +131,9 @@ static int #ifdef HAVE_IOPS_CREATE_USERNS zpl_create(struct user_namespace *user_ns, struct inode *dir, struct dentry *dentry, umode_t mode, bool flag) +#elif defined(HAVE_IOPS_CREATE_IDMAP) +zpl_create(struct mnt_idmap *user_ns, struct inode *dir, + struct dentry *dentry, umode_t mode, bool flag) #else zpl_create(struct inode *dir, struct dentry *dentry, umode_t mode, bool flag) #endif @@ -174,6 +177,9 @@ static int #ifdef HAVE_IOPS_MKNOD_USERNS zpl_mknod(struct user_namespace *user_ns, struct inode *dir, struct dentry *dentry, umode_t mode, +#elif defined(HAVE_IOPS_MKNOD_IDMAP) +zpl_mknod(struct mnt_idmap *user_ns, struct inode *dir, + struct dentry *dentry, umode_t mode, #else zpl_mknod(struct inode *dir, struct dentry *dentry, umode_t mode, #endif @@ -224,7 +230,10 @@ zpl_mknod(struct inode *dir, struct dentry *dentry, umode_t mode, #ifdef HAVE_TMPFILE static int -#ifndef HAVE_TMPFILE_DENTRY +#ifdef HAVE_TMPFILE_IDMAP +zpl_tmpfile(struct mnt_idmap *userns, struct inode *dir, + struct file *file, umode_t mode) +#elif !defined(HAVE_TMPFILE_DENTRY) zpl_tmpfile(struct user_namespace *userns, struct inode *dir, struct file *file, umode_t mode) #else @@ -317,6 +326,9 @@ static int #ifdef HAVE_IOPS_MKDIR_USERNS zpl_mkdir(struct user_namespace *user_ns, struct inode *dir, struct dentry *dentry, umode_t mode) +#elif defined(HAVE_IOPS_MKDIR_IDMAP) +zpl_mkdir(struct mnt_idmap *user_ns, struct inode *dir, + struct dentry *dentry, umode_t mode) #else zpl_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) #endif @@ -386,6 +398,10 @@ static int zpl_getattr_impl(struct user_namespace *user_ns, const struct path *path, struct kstat *stat, u32 request_mask, unsigned int query_flags) +#elif defined(HAVE_IDMAP_IOPS_GETATTR) +zpl_getattr_impl(struct mnt_idmap *user_ns, + const struct path *path, struct kstat *stat, u32 request_mask, + unsigned int query_flags) #else zpl_getattr_impl(const struct path *path, struct kstat *stat, u32 request_mask, unsigned int query_flags) @@ -402,7 +418,9 @@ zpl_getattr_impl(const struct path *path, struct kstat *stat, u32 request_mask, * XXX query_flags currently ignored. */ -#ifdef HAVE_USERNS_IOPS_GETATTR +#ifdef HAVE_GENERIC_FILLATTR_IDMAP_REQMASK + error = -zfs_getattr_fast(user_ns, request_mask, ip, stat); +#elif (defined(HAVE_USERNS_IOPS_GETATTR) || defined(HAVE_IDMAP_IOPS_GETATTR)) error = -zfs_getattr_fast(user_ns, ip, stat); #else error = -zfs_getattr_fast(kcred->user_ns, ip, stat); @@ -441,9 +459,12 @@ zpl_getattr_impl(const struct path *path, struct kstat *stat, u32 request_mask, ZPL_GETATTR_WRAPPER(zpl_getattr); static int -#ifdef HAVE_SETATTR_PREPARE_USERNS +#ifdef HAVE_USERNS_IOPS_SETATTR zpl_setattr(struct user_namespace *user_ns, struct dentry *dentry, struct iattr *ia) +#elif defined(HAVE_IDMAP_IOPS_SETATTR) +zpl_setattr(struct mnt_idmap *user_ns, struct dentry *dentry, + struct iattr *ia) #else zpl_setattr(struct dentry *dentry, struct iattr *ia) #endif @@ -454,7 +475,13 @@ zpl_setattr(struct dentry *dentry, struct iattr *ia) int error; fstrans_cookie_t cookie; - error = zpl_setattr_prepare(kcred->user_ns, dentry, ia); +#ifdef HAVE_SETATTR_PREPARE_USERNS + error = zpl_setattr_prepare(user_ns, dentry, ia); +#elif defined(HAVE_SETATTR_PREPARE_IDMAP) + error = zpl_setattr_prepare(user_ns, dentry, ia); +#else + error = zpl_setattr_prepare(zfs_init_idmap, dentry, ia); +#endif if (error) return (error); @@ -470,7 +497,8 @@ zpl_setattr(struct dentry *dentry, struct iattr *ia) vap->va_ctime = ia->ia_ctime; if (vap->va_mask & ATTR_ATIME) - ip->i_atime = zpl_inode_timestamp_truncate(ia->ia_atime, ip); + zpl_inode_set_atime_to_ts(ip, + zpl_inode_timestamp_truncate(ia->ia_atime, ip)); cookie = spl_fstrans_mark(); error = -zfs_setattr(ITOZ(ip), vap, 0, cr); @@ -489,10 +517,14 @@ static int #ifdef HAVE_IOPS_RENAME_USERNS zpl_rename2(struct user_namespace *user_ns, struct inode *sdip, struct dentry *sdentry, struct inode *tdip, struct dentry *tdentry, - unsigned int flags) + unsigned int rflags) +#elif defined(HAVE_IOPS_RENAME_IDMAP) +zpl_rename2(struct mnt_idmap *user_ns, struct inode *sdip, + struct dentry *sdentry, struct inode *tdip, struct dentry *tdentry, + unsigned int rflags) #else zpl_rename2(struct inode *sdip, struct dentry *sdentry, - struct inode *tdip, struct dentry *tdentry, unsigned int flags) + struct inode *tdip, struct dentry *tdentry, unsigned int rflags) #endif { cred_t *cr = CRED(); @@ -500,7 +532,7 @@ zpl_rename2(struct inode *sdip, struct dentry *sdentry, fstrans_cookie_t cookie; /* We don't have renameat2(2) support */ - if (flags) + if (rflags) return (-EINVAL); crhold(cr); @@ -514,7 +546,9 @@ zpl_rename2(struct inode *sdip, struct dentry *sdentry, return (error); } -#if !defined(HAVE_RENAME_WANTS_FLAGS) && !defined(HAVE_IOPS_RENAME_USERNS) +#if !defined(HAVE_IOPS_RENAME_USERNS) && \ + !defined(HAVE_RENAME_WANTS_FLAGS) && \ + !defined(HAVE_IOPS_RENAME_IDMAP) static int zpl_rename(struct inode *sdip, struct dentry *sdentry, struct inode *tdip, struct dentry *tdentry) @@ -527,6 +561,9 @@ static int #ifdef HAVE_IOPS_SYMLINK_USERNS zpl_symlink(struct user_namespace *user_ns, struct inode *dir, struct dentry *dentry, const char *name) +#elif defined(HAVE_IOPS_SYMLINK_IDMAP) +zpl_symlink(struct mnt_idmap *user_ns, struct inode *dir, + struct dentry *dentry, const char *name) #else zpl_symlink(struct inode *dir, struct dentry *dentry, const char *name) #endif @@ -693,7 +730,7 @@ zpl_link(struct dentry *old_dentry, struct inode *dir, struct dentry *dentry) return (-EMLINK); crhold(cr); - ip->i_ctime = current_time(ip); + zpl_inode_set_ctime_to_ts(ip, current_time(ip)); /* Must have an existing ref, so igrab() cannot return NULL */ VERIFY3P(igrab(ip), !=, NULL); @@ -745,6 +782,8 @@ const struct inode_operations zpl_dir_inode_operations = { .mknod = zpl_mknod, #if defined(HAVE_RENAME_WANTS_FLAGS) || defined(HAVE_IOPS_RENAME_USERNS) .rename = zpl_rename2, +#elif defined(HAVE_IOPS_RENAME_IDMAP) + .rename = zpl_rename2, #else .rename = zpl_rename, #endif diff --git a/module/os/linux/zfs/zpl_xattr.c b/module/os/linux/zfs/zpl_xattr.c index 364cd34c1639..1fc80da6d014 100644 --- a/module/os/linux/zfs/zpl_xattr.c +++ b/module/os/linux/zfs/zpl_xattr.c @@ -510,7 +510,7 @@ zpl_xattr_set_dir(struct inode *ip, const char *name, const void *value, error = -zfs_write_simple(xzp, value, size, pos, NULL); out: if (error == 0) { - ip->i_ctime = current_time(ip); + zpl_inode_set_ctime_to_ts(ip, current_time(ip)); zfs_mark_inode_dirty(ip); } @@ -725,9 +725,11 @@ __zpl_xattr_user_get(struct inode *ip, const char *name, ZPL_XATTR_GET_WRAPPER(zpl_xattr_user_get); static int -__zpl_xattr_user_set(struct inode *ip, const char *name, +__zpl_xattr_user_set(zidmap_t *user_ns, + struct inode *ip, const char *name, const void *value, size_t size, int flags) { + (void) user_ns; char *xattr_name; int error; /* xattr_resolve_name will do this for us if this is defined */ @@ -794,9 +796,11 @@ __zpl_xattr_trusted_get(struct inode *ip, const char *name, ZPL_XATTR_GET_WRAPPER(zpl_xattr_trusted_get); static int -__zpl_xattr_trusted_set(struct inode *ip, const char *name, +__zpl_xattr_trusted_set(zidmap_t *user_ns, + struct inode *ip, const char *name, const void *value, size_t size, int flags) { + (void) user_ns; char *xattr_name; int error; @@ -863,9 +867,11 @@ __zpl_xattr_security_get(struct inode *ip, const char *name, ZPL_XATTR_GET_WRAPPER(zpl_xattr_security_get); static int -__zpl_xattr_security_set(struct inode *ip, const char *name, +__zpl_xattr_security_set(zidmap_t *user_ns, + struct inode *ip, const char *name, const void *value, size_t size, int flags) { + (void) user_ns; char *xattr_name; int error; /* xattr_resolve_name will do this for us if this is defined */ @@ -889,7 +895,7 @@ zpl_xattr_security_init_impl(struct inode *ip, const struct xattr *xattrs, int error = 0; for (xattr = xattrs; xattr->name != NULL; xattr++) { - error = __zpl_xattr_security_set(ip, + error = __zpl_xattr_security_set(NULL, ip, xattr->name, xattr->value, xattr->value_len, 0); if (error < 0) @@ -954,7 +960,8 @@ zpl_set_acl_impl(struct inode *ip, struct posix_acl *acl, int type) */ if (ip->i_mode != mode) { ip->i_mode = mode; - ip->i_ctime = current_time(ip); + zpl_inode_set_ctime_to_ts(ip, + current_time(ip)); zfs_mark_inode_dirty(ip); } @@ -1004,6 +1011,9 @@ int #ifdef HAVE_SET_ACL_USERNS zpl_set_acl(struct user_namespace *userns, struct inode *ip, struct posix_acl *acl, int type) +#elif defined(HAVE_SET_ACL_IDMAP_DENTRY) +zpl_set_acl(struct mnt_idmap *userns, struct dentry *dentry, + struct posix_acl *acl, int type) #elif defined(HAVE_SET_ACL_USERNS_DENTRY_ARG2) zpl_set_acl(struct user_namespace *userns, struct dentry *dentry, struct posix_acl *acl, int type) @@ -1013,6 +1023,8 @@ zpl_set_acl(struct inode *ip, struct posix_acl *acl, int type) { #ifdef HAVE_SET_ACL_USERNS_DENTRY_ARG2 return (zpl_set_acl_impl(d_inode(dentry), acl, type)); +#elif defined(HAVE_SET_ACL_IDMAP_DENTRY) + return (zpl_set_acl_impl(d_inode(dentry), acl, type)); #else return (zpl_set_acl_impl(ip, acl, type)); #endif /* HAVE_SET_ACL_USERNS_DENTRY_ARG2 */ @@ -1108,7 +1120,7 @@ zpl_init_acl(struct inode *ip, struct inode *dir) return (PTR_ERR(acl)); if (!acl) { ip->i_mode &= ~current_umask(); - ip->i_ctime = current_time(ip); + zpl_inode_set_ctime_to_ts(ip, current_time(ip)); zfs_mark_inode_dirty(ip); return (0); } @@ -1256,7 +1268,8 @@ __zpl_xattr_acl_get_default(struct inode *ip, const char *name, ZPL_XATTR_GET_WRAPPER(zpl_xattr_acl_get_default); static int -__zpl_xattr_acl_set_access(struct inode *ip, const char *name, +__zpl_xattr_acl_set_access(zidmap_t *mnt_ns, + struct inode *ip, const char *name, const void *value, size_t size, int flags) { struct posix_acl *acl; @@ -1270,8 +1283,14 @@ __zpl_xattr_acl_set_access(struct inode *ip, const char *name, if (ITOZSB(ip)->z_acl_type != ZFS_ACLTYPE_POSIX) return (-EOPNOTSUPP); - if (!zpl_inode_owner_or_capable(kcred->user_ns, ip)) +#if defined(HAVE_XATTR_SET_USERNS) || defined(HAVE_XATTR_SET_IDMAP) + if (!zpl_inode_owner_or_capable(mnt_ns, ip)) + return (-EPERM); +#else + (void) mnt_ns; + if (!zpl_inode_owner_or_capable(zfs_init_idmap, ip)) return (-EPERM); +#endif if (value) { acl = zpl_acl_from_xattr(value, size); @@ -1295,7 +1314,8 @@ __zpl_xattr_acl_set_access(struct inode *ip, const char *name, ZPL_XATTR_SET_WRAPPER(zpl_xattr_acl_set_access); static int -__zpl_xattr_acl_set_default(struct inode *ip, const char *name, +__zpl_xattr_acl_set_default(zidmap_t *mnt_ns, + struct inode *ip, const char *name, const void *value, size_t size, int flags) { struct posix_acl *acl; @@ -1309,8 +1329,14 @@ __zpl_xattr_acl_set_default(struct inode *ip, const char *name, if (ITOZSB(ip)->z_acl_type != ZFS_ACLTYPE_POSIX) return (-EOPNOTSUPP); - if (!zpl_inode_owner_or_capable(kcred->user_ns, ip)) +#if defined(HAVE_XATTR_SET_USERNS) || defined(HAVE_XATTR_SET_IDMAP) + if (!zpl_inode_owner_or_capable(mnt_ns, ip)) + return (-EPERM); +#else + (void) mnt_ns; + if (!zpl_inode_owner_or_capable(zfs_init_idmap, ip)) return (-EPERM); +#endif if (value) { acl = zpl_acl_from_xattr(value, size); diff --git a/module/os/linux/zfs/zvol_os.c b/module/os/linux/zfs/zvol_os.c index cef047bec6f6..87288190e7c0 100644 --- a/module/os/linux/zfs/zvol_os.c +++ b/module/os/linux/zfs/zvol_os.c @@ -492,7 +492,11 @@ out: } static int +#ifdef HAVE_BLK_MODE_T +zvol_open(struct gendisk *disk, blk_mode_t flag) +#else zvol_open(struct block_device *bdev, fmode_t flag) +#endif { zvol_state_t *zv; int error = 0; @@ -507,10 +511,14 @@ retry: /* * Obtain a copy of private_data under the zvol_state_lock to make * sure that either the result of zvol free code path setting - * bdev->bd_disk->private_data to NULL is observed, or zvol_free() + * disk->private_data to NULL is observed, or zvol_os_free() * is not called on this zv because of the positive zv_open_count. */ +#ifdef HAVE_BLK_MODE_T + zv = disk->private_data; +#else zv = bdev->bd_disk->private_data; +#endif if (zv == NULL) { rw_exit(&zvol_state_lock); return (SET_ERROR(-ENXIO)); @@ -590,14 +598,15 @@ retry: } } - error = -zvol_first_open(zv, !(flag & FMODE_WRITE)); + error = -zvol_first_open(zv, !(blk_mode_is_open_write(flag))); if (drop_namespace) mutex_exit(&spa_namespace_lock); } if (error == 0) { - if ((flag & FMODE_WRITE) && (zv->zv_flags & ZVOL_RDONLY)) { + if ((blk_mode_is_open_write(flag)) && + (zv->zv_flags & ZVOL_RDONLY)) { if (zv->zv_open_count == 0) zvol_last_close(zv); @@ -612,14 +621,25 @@ retry: rw_exit(&zv->zv_suspend_lock); if (error == 0) +#ifdef HAVE_BLK_MODE_T + disk_check_media_change(disk); +#else zfs_check_media_change(bdev); +#endif return (error); } static void -zvol_release(struct gendisk *disk, fmode_t mode) +#ifdef HAVE_BLOCK_DEVICE_OPERATIONS_RELEASE_1ARG +zvol_release(struct gendisk *disk) +#else +zvol_release(struct gendisk *disk, fmode_t unused) +#endif { +#if !defined(HAVE_BLOCK_DEVICE_OPERATIONS_RELEASE_1ARG) + (void) unused; +#endif zvol_state_t *zv; boolean_t drop_suspend = B_TRUE; @@ -674,7 +694,13 @@ zvol_ioctl(struct block_device *bdev, fmode_t mode, switch (cmd) { case BLKFLSBUF: +#ifdef HAVE_FSYNC_BDEV fsync_bdev(bdev); +#elif defined(HAVE_SYNC_BLOCKDEV) + sync_blockdev(bdev); +#else +#error "Neither fsync_bdev() nor sync_blockdev() found" +#endif invalidate_bdev(bdev); rw_enter(&zv->zv_suspend_lock, RW_READER); diff --git a/module/zcommon/Makefile.in b/module/zcommon/Makefile.in index ebc538440445..614968a42be9 100644 --- a/module/zcommon/Makefile.in +++ b/module/zcommon/Makefile.in @@ -26,3 +26,7 @@ $(MODULE)-$(CONFIG_X86) += zfs_fletcher_intel.o $(MODULE)-$(CONFIG_X86) += zfs_fletcher_sse.o $(MODULE)-$(CONFIG_X86) += zfs_fletcher_avx512.o $(MODULE)-$(CONFIG_ARM64) += zfs_fletcher_aarch64_neon.o + +ifeq ($(CONFIG_ARM64),y) +CFLAGS_REMOVE_zfs_fletcher_aarch64_neon.o += -mgeneral-regs-only +endif diff --git a/module/zfs/Makefile.in b/module/zfs/Makefile.in index 653ea0da9bcc..d9b86890b5f5 100644 --- a/module/zfs/Makefile.in +++ b/module/zfs/Makefile.in @@ -154,4 +154,13 @@ ifeq ($(CONFIG_ALTIVEC),y) $(obj)/vdev_raidz_math_powerpc_altivec.o: c_flags += -maltivec endif +ifeq ($(CONFIG_ARM64),y) +CFLAGS_REMOVE_vdev_raidz_math_aarch64_neon.o += -mgeneral-regs-only +CFLAGS_REMOVE_vdev_raidz_math_aarch64_neonx2.o += -mgeneral-regs-only +endif + +UBSAN_SANITIZE_zap_leaf.o := n +UBSAN_SANITIZE_zap_micro.o := n +UBSAN_SANITIZE_sa.o := n + include $(mfdir)/../os/linux/zfs/Makefile diff --git a/module/zfs/abd.c b/module/zfs/abd.c index 8ee8e7e57420..754974a559b6 100644 --- a/module/zfs/abd.c +++ b/module/zfs/abd.c @@ -109,7 +109,6 @@ void abd_verify(abd_t *abd) { #ifdef ZFS_DEBUG - ASSERT3U(abd->abd_size, >, 0); ASSERT3U(abd->abd_size, <=, SPA_MAXBLOCKSIZE); ASSERT3U(abd->abd_flags, ==, abd->abd_flags & (ABD_FLAG_LINEAR | ABD_FLAG_OWNER | ABD_FLAG_META | ABD_FLAG_MULTI_ZONE | @@ -118,6 +117,7 @@ abd_verify(abd_t *abd) IMPLY(abd->abd_parent != NULL, !(abd->abd_flags & ABD_FLAG_OWNER)); IMPLY(abd->abd_flags & ABD_FLAG_META, abd->abd_flags & ABD_FLAG_OWNER); if (abd_is_linear(abd)) { + ASSERT3U(abd->abd_size, >, 0); ASSERT3P(ABD_LINEAR_BUF(abd), !=, NULL); } else if (abd_is_gang(abd)) { uint_t child_sizes = 0; @@ -130,6 +130,7 @@ abd_verify(abd_t *abd) } ASSERT3U(abd->abd_size, ==, child_sizes); } else { + ASSERT3U(abd->abd_size, >, 0); abd_verify_scatter(abd); } #endif @@ -369,7 +370,20 @@ abd_gang_add_gang(abd_t *pabd, abd_t *cabd, boolean_t free_on_free) * will retain all the free_on_free settings after being * added to the parents list. */ +#ifdef ZFS_DEBUG + /* + * If cabd had abd_parent, we have to drop it here. We can't + * transfer it to pabd, nor we can clear abd_size leaving it. + */ + if (cabd->abd_parent != NULL) { + (void) zfs_refcount_remove_many( + &cabd->abd_parent->abd_children, + cabd->abd_size, cabd); + cabd->abd_parent = NULL; + } +#endif pabd->abd_size += cabd->abd_size; + cabd->abd_size = 0; list_move_tail(&ABD_GANG(pabd).abd_gang_chain, &ABD_GANG(cabd).abd_gang_chain); ASSERT(list_is_empty(&ABD_GANG(cabd).abd_gang_chain)); @@ -407,7 +421,6 @@ abd_gang_add(abd_t *pabd, abd_t *cabd, boolean_t free_on_free) */ if (abd_is_gang(cabd)) { ASSERT(!list_link_active(&cabd->abd_gang_link)); - ASSERT(!list_is_empty(&ABD_GANG(cabd).abd_gang_chain)); return (abd_gang_add_gang(pabd, cabd, free_on_free)); } ASSERT(!abd_is_gang(cabd)); diff --git a/module/zfs/arc.c b/module/zfs/arc.c index 6900b6b134d9..1180853da038 100644 --- a/module/zfs/arc.c +++ b/module/zfs/arc.c @@ -946,7 +946,7 @@ static void l2arc_hdr_restore(const l2arc_log_ent_phys_t *le, l2arc_dev_t *dev); /* L2ARC persistence write I/O routines. */ -static void l2arc_log_blk_commit(l2arc_dev_t *dev, zio_t *pio, +static uint64_t l2arc_log_blk_commit(l2arc_dev_t *dev, zio_t *pio, l2arc_write_callback_t *cb); /* L2ARC persistence auxiliary routines. */ @@ -8415,7 +8415,7 @@ l2arc_write_eligible(uint64_t spa_guid, arc_buf_hdr_t *hdr) static uint64_t l2arc_write_size(l2arc_dev_t *dev) { - uint64_t size, dev_size, tsize; + uint64_t size; /* * Make sure our globals have meaningful values in case the user @@ -8432,18 +8432,23 @@ l2arc_write_size(l2arc_dev_t *dev) if (arc_warm == B_FALSE) size += l2arc_write_boost; + /* We need to add in the worst case scenario of log block overhead. */ + size += l2arc_log_blk_overhead(size, dev); + if (dev->l2ad_vdev->vdev_has_trim && l2arc_trim_ahead > 0) { + /* + * Trim ahead of the write size 64MB or (l2arc_trim_ahead/100) + * times the writesize, whichever is greater. + */ + size += MAX(64 * 1024 * 1024, + (size * l2arc_trim_ahead) / 100); + } + /* * Make sure the write size does not exceed the size of the cache * device. This is important in l2arc_evict(), otherwise infinite * iteration can occur. */ - dev_size = dev->l2ad_end - dev->l2ad_start; - tsize = size + l2arc_log_blk_overhead(size, dev); - if (dev->l2ad_vdev->vdev_has_trim && l2arc_trim_ahead > 0) - tsize += MAX(64 * 1024 * 1024, - (tsize * l2arc_trim_ahead) / 100); - - if (tsize >= dev_size) { + if (size > dev->l2ad_end - dev->l2ad_start) { cmn_err(CE_NOTE, "l2arc_write_max or l2arc_write_boost " "plus the overhead of log blocks (persistent L2ARC, " "%llu bytes) exceeds the size of the cache device " @@ -8452,8 +8457,19 @@ l2arc_write_size(l2arc_dev_t *dev) dev->l2ad_vdev->vdev_guid, L2ARC_WRITE_SIZE); size = l2arc_write_max = l2arc_write_boost = L2ARC_WRITE_SIZE; + if (l2arc_trim_ahead > 1) { + cmn_err(CE_NOTE, "l2arc_trim_ahead set to 1"); + l2arc_trim_ahead = 1; + } + if (arc_warm == B_FALSE) size += l2arc_write_boost; + + size += l2arc_log_blk_overhead(size, dev); + if (dev->l2ad_vdev->vdev_has_trim && l2arc_trim_ahead > 0) { + size += MAX(64 * 1024 * 1024, + (size * l2arc_trim_ahead) / 100); + } } return (size); @@ -9074,22 +9090,9 @@ l2arc_evict(l2arc_dev_t *dev, uint64_t distance, boolean_t all) buflist = &dev->l2ad_buflist; - /* - * We need to add in the worst case scenario of log block overhead. - */ - distance += l2arc_log_blk_overhead(distance, dev); - if (vd->vdev_has_trim && l2arc_trim_ahead > 0) { - /* - * Trim ahead of the write size 64MB or (l2arc_trim_ahead/100) - * times the write size, whichever is greater. - */ - distance += MAX(64 * 1024 * 1024, - (distance * l2arc_trim_ahead) / 100); - } - top: rerun = B_FALSE; - if (dev->l2ad_hand >= (dev->l2ad_end - distance)) { + if (dev->l2ad_hand + distance > dev->l2ad_end) { /* * When there is no space to accommodate upcoming writes, * evict to the end. Then bump the write and evict hands @@ -9283,7 +9286,7 @@ out: */ ASSERT3U(dev->l2ad_hand + distance, <, dev->l2ad_end); if (!dev->l2ad_first) - ASSERT3U(dev->l2ad_hand, <, dev->l2ad_evict); + ASSERT3U(dev->l2ad_hand, <=, dev->l2ad_evict); } } @@ -9549,7 +9552,13 @@ l2arc_write_buffers(spa_t *spa, l2arc_dev_t *dev, uint64_t target_sz) uint64_t asize = vdev_psize_to_asize(dev->l2ad_vdev, psize); - if ((write_asize + asize) > target_sz) { + /* + * If the allocated size of this buffer plus the max + * size for the pending log block exceeds the evicted + * target size, terminate writing buffers for this run. + */ + if (write_asize + asize + + sizeof (l2arc_log_blk_phys_t) > target_sz) { full = B_TRUE; mutex_exit(hash_lock); break; @@ -9669,8 +9678,14 @@ l2arc_write_buffers(spa_t *spa, l2arc_dev_t *dev, uint64_t target_sz) * arcstat_l2_{size,asize} kstats are updated * internally. */ - if (l2arc_log_blk_insert(dev, hdr)) - l2arc_log_blk_commit(dev, pio, cb); + if (l2arc_log_blk_insert(dev, hdr)) { + /* + * l2ad_hand will be adjusted in + * l2arc_log_blk_commit(). + */ + write_asize += + l2arc_log_blk_commit(dev, pio, cb); + } zio_nowait(wzio); } @@ -10820,7 +10835,7 @@ l2arc_dev_hdr_update(l2arc_dev_t *dev) * This function allocates some memory to temporarily hold the serialized * buffer to be written. This is then released in l2arc_write_done. */ -static void +static uint64_t l2arc_log_blk_commit(l2arc_dev_t *dev, zio_t *pio, l2arc_write_callback_t *cb) { l2arc_log_blk_phys_t *lb = &dev->l2ad_log_blk; @@ -10933,6 +10948,8 @@ l2arc_log_blk_commit(l2arc_dev_t *dev, zio_t *pio, l2arc_write_callback_t *cb) dev->l2ad_log_ent_idx = 0; dev->l2ad_log_blk_payload_asize = 0; dev->l2ad_log_blk_payload_start = 0; + + return (asize); } /* diff --git a/module/zfs/dmu_recv.c b/module/zfs/dmu_recv.c index 98ca2b3bcec1..6eb1009a788b 100644 --- a/module/zfs/dmu_recv.c +++ b/module/zfs/dmu_recv.c @@ -71,6 +71,12 @@ int zfs_recv_write_batch_size = 1024 * 1024; static char *dmu_recv_tag = "dmu_recv_tag"; const char *recv_clone_name = "%recv"; +typedef enum { + ORNS_NO, + ORNS_YES, + ORNS_MAYBE +} or_need_sync_t; + static int receive_read_payload_and_next_header(dmu_recv_cookie_t *ra, int len, void *buf); @@ -121,6 +127,9 @@ struct receive_writer_arg { uint8_t or_iv[ZIO_DATA_IV_LEN]; uint8_t or_mac[ZIO_DATA_MAC_LEN]; boolean_t or_byteorder; + + /* Keep track of DRR_FREEOBJECTS right after DRR_OBJECT_RANGE */ + or_need_sync_t or_need_sync; }; typedef struct dmu_recv_begin_arg { @@ -1524,17 +1533,19 @@ receive_handle_existing_object(const struct receive_writer_arg *rwa, } /* - * The dmu does not currently support decreasing nlevels - * or changing the number of dnode slots on an object. For - * non-raw sends, this does not matter and the new object - * can just use the previous one's nlevels. For raw sends, - * however, the structure of the received dnode (including - * nlevels and dnode slots) must match that of the send - * side. Therefore, instead of using dmu_object_reclaim(), - * we must free the object completely and call - * dmu_object_claim_dnsize() instead. + * The dmu does not currently support decreasing nlevels or changing + * indirect block size if there is already one, same as changing the + * number of of dnode slots on an object. For non-raw sends this + * does not matter and the new object can just use the previous one's + * parameters. For raw sends, however, the structure of the received + * dnode (including indirects and dnode slots) must match that of the + * send side. Therefore, instead of using dmu_object_reclaim(), we + * must free the object completely and call dmu_object_claim_dnsize() + * instead. */ - if ((rwa->raw && drro->drr_nlevels < doi->doi_indirection) || + if ((rwa->raw && ((doi->doi_indirection > 1 && + indblksz != doi->doi_metadata_block_size) || + drro->drr_nlevels < doi->doi_indirection)) || dn_slots != doi->doi_dnodesize >> DNODE_SHIFT) { err = dmu_free_long_object(rwa->os, drro->drr_object); if (err != 0) @@ -1658,10 +1669,22 @@ receive_object(struct receive_writer_arg *rwa, struct drr_object *drro, /* object was freed and we are about to allocate a new one */ object_to_hold = DMU_NEW_OBJECT; } else { + /* + * If the only record in this range so far was DRR_FREEOBJECTS + * with at least one actually freed object, it's possible that + * the block will now be converted to a hole. We need to wait + * for the txg to sync to prevent races. + */ + if (rwa->or_need_sync == ORNS_YES) + txg_wait_synced(dmu_objset_pool(rwa->os), 0); + /* object is free and we are about to allocate a new one */ object_to_hold = DMU_NEW_OBJECT; } + /* Only relevant for the first object in the range */ + rwa->or_need_sync = ORNS_NO; + /* * If this is a multi-slot dnode there is a chance that this * object will expand into a slot that is already used by @@ -1856,6 +1879,9 @@ receive_freeobjects(struct receive_writer_arg *rwa, if (err != 0) return (err); + + if (rwa->or_need_sync == ORNS_MAYBE) + rwa->or_need_sync = ORNS_YES; } if (next_err != ESRCH) return (next_err); @@ -2298,6 +2324,8 @@ receive_object_range(struct receive_writer_arg *rwa, bcopy(drror->drr_mac, rwa->or_mac, ZIO_DATA_MAC_LEN); rwa->or_byteorder = byteorder; + rwa->or_need_sync = ORNS_MAYBE; + return (0); } diff --git a/module/zfs/dmu_send.c b/module/zfs/dmu_send.c index cd9ecc07fd5c..0dd1ec210a1d 100644 --- a/module/zfs/dmu_send.c +++ b/module/zfs/dmu_send.c @@ -2797,6 +2797,7 @@ dmu_send(const char *tosnap, const char *fromsnap, boolean_t embedok, } if (err == 0) { + owned = B_TRUE; err = zap_lookup(dspp.dp->dp_meta_objset, dspp.to_ds->ds_object, DS_FIELD_RESUME_TOGUID, 8, 1, @@ -2810,21 +2811,24 @@ dmu_send(const char *tosnap, const char *fromsnap, boolean_t embedok, sizeof (dspp.saved_toname), dspp.saved_toname); } - if (err != 0) + /* Only disown if there was an error in the lookups */ + if (owned && (err != 0)) dsl_dataset_disown(dspp.to_ds, dsflags, FTAG); kmem_strfree(name); } else { err = dsl_dataset_own(dspp.dp, tosnap, dsflags, FTAG, &dspp.to_ds); + if (err == 0) + owned = B_TRUE; } - owned = B_TRUE; } else { err = dsl_dataset_hold_flags(dspp.dp, tosnap, dsflags, FTAG, &dspp.to_ds); } if (err != 0) { + /* Note: dsl dataset is not owned at this point */ dsl_pool_rele(dspp.dp, FTAG); return (err); } diff --git a/module/zfs/dmu_tx.c b/module/zfs/dmu_tx.c index 1eed0526b51d..063934f39493 100644 --- a/module/zfs/dmu_tx.c +++ b/module/zfs/dmu_tx.c @@ -291,6 +291,53 @@ dmu_tx_count_write(dmu_tx_hold_t *txh, uint64_t off, uint64_t len) } static void +dmu_tx_count_append(dmu_tx_hold_t *txh, uint64_t off, uint64_t len) +{ + dnode_t *dn = txh->txh_dnode; + int err = 0; + + if (len == 0) + return; + + (void) zfs_refcount_add_many(&txh->txh_space_towrite, len, FTAG); + + if (dn == NULL) + return; + + /* + * For i/o error checking, read the blocks that will be needed + * to perform the append; first level-0 block (if not aligned, i.e. + * if they are partial-block writes), no additional blocks are read. + */ + if (dn->dn_maxblkid == 0) { + if (off < dn->dn_datablksz && + (off > 0 || len < dn->dn_datablksz)) { + err = dmu_tx_check_ioerr(NULL, dn, 0, 0); + if (err != 0) { + txh->txh_tx->tx_err = err; + } + } + } else { + zio_t *zio = zio_root(dn->dn_objset->os_spa, + NULL, NULL, ZIO_FLAG_CANFAIL); + + /* first level-0 block */ + uint64_t start = off >> dn->dn_datablkshift; + if (P2PHASE(off, dn->dn_datablksz) || len < dn->dn_datablksz) { + err = dmu_tx_check_ioerr(zio, dn, 0, start); + if (err != 0) { + txh->txh_tx->tx_err = err; + } + } + + err = zio_wait(zio); + if (err != 0) { + txh->txh_tx->tx_err = err; + } + } +} + +static void dmu_tx_count_dnode(dmu_tx_hold_t *txh) { (void) zfs_refcount_add_many(&txh->txh_space_towrite, @@ -331,6 +378,42 @@ dmu_tx_hold_write_by_dnode(dmu_tx_t *tx, dnode_t *dn, uint64_t off, int len) } /* + * Should be used when appending to an object and the exact offset is unknown. + * The write must occur at or beyond the specified offset. Only the L0 block + * at provided offset will be prefetched. + */ +void +dmu_tx_hold_append(dmu_tx_t *tx, uint64_t object, uint64_t off, int len) +{ + dmu_tx_hold_t *txh; + + ASSERT0(tx->tx_txg); + ASSERT3U(len, <=, DMU_MAX_ACCESS); + + txh = dmu_tx_hold_object_impl(tx, tx->tx_objset, + object, THT_APPEND, off, DMU_OBJECT_END); + if (txh != NULL) { + dmu_tx_count_append(txh, off, len); + dmu_tx_count_dnode(txh); + } +} + +void +dmu_tx_hold_append_by_dnode(dmu_tx_t *tx, dnode_t *dn, uint64_t off, int len) +{ + dmu_tx_hold_t *txh; + + ASSERT0(tx->tx_txg); + ASSERT3U(len, <=, DMU_MAX_ACCESS); + + txh = dmu_tx_hold_dnode_impl(tx, dn, THT_APPEND, off, DMU_OBJECT_END); + if (txh != NULL) { + dmu_tx_count_append(txh, off, len); + dmu_tx_count_dnode(txh); + } +} + +/* * This function marks the transaction as being a "net free". The end * result is that refquotas will be disabled for this transaction, and * this transaction will be able to use half of the pool space overhead @@ -638,6 +721,26 @@ dmu_tx_dirty_buf(dmu_tx_t *tx, dmu_buf_impl_t *db) if (blkid == 0) match_offset = TRUE; break; + case THT_APPEND: + if (blkid >= beginblk && (blkid <= endblk || + txh->txh_arg2 == DMU_OBJECT_END)) + match_offset = TRUE; + + /* + * THT_WRITE used for bonus and spill blocks. + */ + ASSERT(blkid != DMU_BONUS_BLKID && + blkid != DMU_SPILL_BLKID); + + /* + * They might have to increase nlevels, + * thus dirtying the new TLIBs. Or the + * might have to change the block size, + * thus dirying the new lvl=0 blk=0. + */ + if (blkid == 0) + match_offset = TRUE; + break; case THT_FREE: /* * We will dirty all the level 1 blocks in @@ -1421,6 +1524,8 @@ dmu_tx_fini(void) EXPORT_SYMBOL(dmu_tx_create); EXPORT_SYMBOL(dmu_tx_hold_write); EXPORT_SYMBOL(dmu_tx_hold_write_by_dnode); +EXPORT_SYMBOL(dmu_tx_hold_append); +EXPORT_SYMBOL(dmu_tx_hold_append_by_dnode); EXPORT_SYMBOL(dmu_tx_hold_free); EXPORT_SYMBOL(dmu_tx_hold_free_by_dnode); EXPORT_SYMBOL(dmu_tx_hold_zap); diff --git a/module/zfs/dnode.c b/module/zfs/dnode.c index ed75c3bdf698..efebc443a210 100644 --- a/module/zfs/dnode.c +++ b/module/zfs/dnode.c @@ -1773,7 +1773,14 @@ dnode_try_claim(objset_t *os, uint64_t object, int slots) } /* - * Checks if the dnode contains any uncommitted dirty records. + * Checks if the dnode itself is dirty, or is carrying any uncommitted records. + * It is important to check both conditions, as some operations (eg appending + * to a file) can dirty both as a single logical unit, but they are not synced + * out atomically, so checking one and not the other can result in an object + * appearing to be clean mid-way through a commit. + * + * Do not change this lightly! If you get it wrong, dmu_offset_next() can + * detect a hole where there is really data, leading to silent corruption. */ boolean_t dnode_is_dirty(dnode_t *dn) @@ -1781,7 +1788,8 @@ dnode_is_dirty(dnode_t *dn) mutex_enter(&dn->dn_mtx); for (int i = 0; i < TXG_SIZE; i++) { - if (multilist_link_active(&dn->dn_dirty_link[i])) { + if (multilist_link_active(&dn->dn_dirty_link[i]) || + !list_is_empty(&dn->dn_dirty_records[i])) { mutex_exit(&dn->dn_mtx); return (B_TRUE); } @@ -1891,7 +1899,7 @@ dnode_set_blksz(dnode_t *dn, uint64_t size, int ibs, dmu_tx_t *tx) if (ibs == dn->dn_indblkshift) ibs = 0; - if (size >> SPA_MINBLOCKSHIFT == dn->dn_datablkszsec && ibs == 0) + if (size == dn->dn_datablksz && ibs == 0) return (0); rw_enter(&dn->dn_struct_rwlock, RW_WRITER); @@ -1914,24 +1922,25 @@ dnode_set_blksz(dnode_t *dn, uint64_t size, int ibs, dmu_tx_t *tx) if (ibs && dn->dn_nlevels != 1) goto fail; - /* resize the old block */ - err = dbuf_hold_impl(dn, 0, 0, TRUE, FALSE, FTAG, &db); - if (err == 0) { - dbuf_new_size(db, size, tx); - } else if (err != ENOENT) { - goto fail; - } - - dnode_setdblksz(dn, size); dnode_setdirty(dn, tx); - dn->dn_next_blksz[tx->tx_txg&TXG_MASK] = size; + if (size != dn->dn_datablksz) { + /* resize the old block */ + err = dbuf_hold_impl(dn, 0, 0, TRUE, FALSE, FTAG, &db); + if (err == 0) { + dbuf_new_size(db, size, tx); + } else if (err != ENOENT) { + goto fail; + } + + dnode_setdblksz(dn, size); + dn->dn_next_blksz[tx->tx_txg & TXG_MASK] = size; + if (db) + dbuf_rele(db, FTAG); + } if (ibs) { dn->dn_indblkshift = ibs; - dn->dn_next_indblkshift[tx->tx_txg&TXG_MASK] = ibs; + dn->dn_next_indblkshift[tx->tx_txg & TXG_MASK] = ibs; } - /* release after we have fixed the blocksize in the dnode */ - if (db) - dbuf_rele(db, FTAG); rw_exit(&dn->dn_struct_rwlock); return (0); diff --git a/module/zfs/dsl_deadlist.c b/module/zfs/dsl_deadlist.c index d5fe2ee56804..9827eb14728d 100644 --- a/module/zfs/dsl_deadlist.c +++ b/module/zfs/dsl_deadlist.c @@ -859,7 +859,7 @@ void dsl_deadlist_merge(dsl_deadlist_t *dl, uint64_t obj, dmu_tx_t *tx) { zap_cursor_t zc, pzc; - zap_attribute_t za, pza; + zap_attribute_t *za, *pza; dmu_buf_t *bonus; dsl_deadlist_phys_t *dlp; dmu_object_info_t doi; @@ -874,28 +874,31 @@ dsl_deadlist_merge(dsl_deadlist_t *dl, uint64_t obj, dmu_tx_t *tx) return; } + za = kmem_alloc(sizeof (*za), KM_SLEEP); + pza = kmem_alloc(sizeof (*pza), KM_SLEEP); + mutex_enter(&dl->dl_lock); /* * Prefetch up to 128 deadlists first and then more as we progress. * The limit is a balance between ARC use and diminishing returns. */ for (zap_cursor_init(&pzc, dl->dl_os, obj), i = 0; - (perror = zap_cursor_retrieve(&pzc, &pza)) == 0 && i < 128; + (perror = zap_cursor_retrieve(&pzc, pza)) == 0 && i < 128; zap_cursor_advance(&pzc), i++) { - dsl_deadlist_prefetch_bpobj(dl, pza.za_first_integer, - zfs_strtonum(pza.za_name, NULL)); + dsl_deadlist_prefetch_bpobj(dl, pza->za_first_integer, + zfs_strtonum(pza->za_name, NULL)); } for (zap_cursor_init(&zc, dl->dl_os, obj); - (error = zap_cursor_retrieve(&zc, &za)) == 0; + (error = zap_cursor_retrieve(&zc, za)) == 0; zap_cursor_advance(&zc)) { - uint64_t mintxg = zfs_strtonum(za.za_name, NULL); - dsl_deadlist_insert_bpobj(dl, za.za_first_integer, mintxg, tx); + uint64_t mintxg = zfs_strtonum(za->za_name, NULL); + dsl_deadlist_insert_bpobj(dl, za->za_first_integer, mintxg, tx); VERIFY0(zap_remove_int(dl->dl_os, obj, mintxg, tx)); if (perror == 0) { - dsl_deadlist_prefetch_bpobj(dl, pza.za_first_integer, - zfs_strtonum(pza.za_name, NULL)); + dsl_deadlist_prefetch_bpobj(dl, pza->za_first_integer, + zfs_strtonum(pza->za_name, NULL)); zap_cursor_advance(&pzc); - perror = zap_cursor_retrieve(&pzc, &pza); + perror = zap_cursor_retrieve(&pzc, pza); } } VERIFY3U(error, ==, ENOENT); @@ -908,6 +911,9 @@ dsl_deadlist_merge(dsl_deadlist_t *dl, uint64_t obj, dmu_tx_t *tx) bzero(dlp, sizeof (*dlp)); dmu_buf_rele(bonus, FTAG); mutex_exit(&dl->dl_lock); + + kmem_free(za, sizeof (*za)); + kmem_free(pza, sizeof (*pza)); } /* diff --git a/module/zfs/dsl_scan.c b/module/zfs/dsl_scan.c index f3c639b0d04e..f0a851ff53a9 100644 --- a/module/zfs/dsl_scan.c +++ b/module/zfs/dsl_scan.c @@ -37,6 +37,7 @@ #include <sys/dmu_tx.h> #include <sys/dmu_objset.h> #include <sys/arc.h> +#include <sys/arc_impl.h> #include <sys/zap.h> #include <sys/zio.h> #include <sys/zfs_context.h> @@ -126,12 +127,21 @@ static boolean_t scan_ds_queue_contains(dsl_scan_t *scn, uint64_t dsobj, static void scan_ds_queue_insert(dsl_scan_t *scn, uint64_t dsobj, uint64_t txg); static void scan_ds_queue_remove(dsl_scan_t *scn, uint64_t dsobj); static void scan_ds_queue_sync(dsl_scan_t *scn, dmu_tx_t *tx); -static uint64_t dsl_scan_count_data_disks(vdev_t *vd); +static uint64_t dsl_scan_count_data_disks(spa_t *spa); extern int zfs_vdev_async_write_active_min_dirty_percent; static int zfs_scan_blkstats = 0; /* + * 'zpool status' uses bytes processed per pass to report throughput and + * estimate time remaining. We define a pass to start when the scanning + * phase completes for a sequential resilver. Optionally, this value + * may be used to reset the pass statistics every N txgs to provide an + * estimated completion time based on currently observed performance. + */ +static uint_t zfs_scan_report_txgs = 0; + +/* * By default zfs will check to ensure it is not over the hard memory * limit before each txg. If finer-grained control of this is needed * this value can be set to 1 to enable checking before scanning each @@ -147,7 +157,7 @@ int zfs_scan_strict_mem_lim = B_FALSE; * overload the drives with I/O, since that is protected by * zfs_vdev_scrub_max_active. */ -unsigned long zfs_scan_vdev_limit = 4 << 20; +unsigned long zfs_scan_vdev_limit = 16 << 20; int zfs_scan_issue_strategy = 0; int zfs_scan_legacy = B_FALSE; /* don't queue & sort zios, go direct */ @@ -450,11 +460,12 @@ dsl_scan_init(dsl_pool_t *dp, uint64_t txg) /* * Calculate the max number of in-flight bytes for pool-wide - * scanning operations (minimum 1MB). Limits for the issuing - * phase are done per top-level vdev and are handled separately. + * scanning operations (minimum 1MB, maximum 1/4 of arc_c_max). + * Limits for the issuing phase are done per top-level vdev and + * are handled separately. */ - scn->scn_maxinflight_bytes = MAX(zfs_scan_vdev_limit * - dsl_scan_count_data_disks(spa->spa_root_vdev), 1ULL << 20); + scn->scn_maxinflight_bytes = MIN(arc_c_max / 4, MAX(1ULL << 20, + zfs_scan_vdev_limit * dsl_scan_count_data_disks(spa))); avl_create(&scn->scn_queue, scan_ds_queue_compare, sizeof (scan_ds_t), offsetof(scan_ds_t, sds_node)); @@ -584,6 +595,8 @@ dsl_scan_init(dsl_pool_t *dp, uint64_t txg) } spa_scan_stat_init(spa); + vdev_scan_stat_init(spa->spa_root_vdev); + return (0); } @@ -742,6 +755,7 @@ dsl_scan_setup_sync(void *arg, dmu_tx_t *tx) scn->scn_last_checkpoint = 0; scn->scn_checkpointing = B_FALSE; spa_scan_stat_init(spa); + vdev_scan_stat_init(spa->spa_root_vdev); if (DSL_SCAN_IS_SCRUB_RESILVER(scn)) { scn->scn_phys.scn_ddt_class_max = zfs_scrub_ddt_class_max; @@ -2797,8 +2811,9 @@ dsl_scan_visit(dsl_scan_t *scn, dmu_tx_t *tx) } static uint64_t -dsl_scan_count_data_disks(vdev_t *rvd) +dsl_scan_count_data_disks(spa_t *spa) { + vdev_t *rvd = spa->spa_root_vdev; uint64_t i, leaves = 0; for (i = 0; i < rvd->vdev_children; i++) { @@ -3638,6 +3653,16 @@ dsl_scan_sync(dsl_pool_t *dp, dmu_tx_t *tx) } /* + * Disabled by default, set zfs_scan_report_txgs to report + * average performance over the last zfs_scan_report_txgs TXGs. + */ + if (!dsl_scan_is_paused_scrub(scn) && zfs_scan_report_txgs != 0 && + tx->tx_txg % zfs_scan_report_txgs == 0) { + scn->scn_issued_before_pass += spa->spa_scan_pass_issued; + spa_scan_stat_init(spa); + } + + /* * It is possible to switch from unsorted to sorted at any time, * but afterwards the scan will remain sorted unless reloaded from * a checkpoint after a reboot. @@ -3693,12 +3718,13 @@ dsl_scan_sync(dsl_pool_t *dp, dmu_tx_t *tx) taskqid_t prefetch_tqid; /* - * Recalculate the max number of in-flight bytes for pool-wide - * scanning operations (minimum 1MB). Limits for the issuing - * phase are done per top-level vdev and are handled separately. + * Calculate the max number of in-flight bytes for pool-wide + * scanning operations (minimum 1MB, maximum 1/4 of arc_c_max). + * Limits for the issuing phase are done per top-level vdev and + * are handled separately. */ - scn->scn_maxinflight_bytes = MAX(zfs_scan_vdev_limit * - dsl_scan_count_data_disks(spa->spa_root_vdev), 1ULL << 20); + scn->scn_maxinflight_bytes = MIN(arc_c_max / 4, MAX(1ULL << 20, + zfs_scan_vdev_limit * dsl_scan_count_data_disks(spa))); if (scnp->scn_ddt_bookmark.ddb_class <= scnp->scn_ddt_class_max) { @@ -3759,6 +3785,9 @@ dsl_scan_sync(dsl_pool_t *dp, dmu_tx_t *tx) if (scn->scn_is_sorted) { scn->scn_checkpointing = B_TRUE; scn->scn_clearing = B_TRUE; + scn->scn_issued_before_pass += + spa->spa_scan_pass_issued; + spa_scan_stat_init(spa); } zfs_dbgmsg("scan complete txg %llu", (longlong_t)tx->tx_txg); @@ -4485,6 +4514,9 @@ ZFS_MODULE_PARAM(zfs, zfs_, scan_strict_mem_lim, INT, ZMOD_RW, ZFS_MODULE_PARAM(zfs, zfs_, scan_fill_weight, INT, ZMOD_RW, "Tunable to adjust bias towards more filled segments during scans"); +ZFS_MODULE_PARAM(zfs, zfs_, scan_report_txgs, UINT, ZMOD_RW, + "Tunable to report resilver performance over the last N txgs"); + ZFS_MODULE_PARAM(zfs, zfs_, resilver_disable_defer, INT, ZMOD_RW, "Process all resilvers immediately"); /* END CSTYLED */ diff --git a/module/zfs/mmp.c b/module/zfs/mmp.c index f67a4eb22a2d..139bb0acd277 100644 --- a/module/zfs/mmp.c +++ b/module/zfs/mmp.c @@ -444,7 +444,7 @@ mmp_write_uberblock(spa_t *spa) uint64_t offset; hrtime_t lock_acquire_time = gethrtime(); - spa_config_enter(spa, SCL_STATE, mmp_tag, RW_READER); + spa_config_enter_mmp(spa, SCL_STATE, mmp_tag, RW_READER); lock_acquire_time = gethrtime() - lock_acquire_time; if (lock_acquire_time > (MSEC2NSEC(MMP_MIN_INTERVAL) / 10)) zfs_dbgmsg("MMP SCL_STATE acquisition pool '%s' took %llu ns " diff --git a/module/zfs/spa.c b/module/zfs/spa.c index 1ed79eed3e8b..81a6547896ac 100644 --- a/module/zfs/spa.c +++ b/module/zfs/spa.c @@ -33,6 +33,7 @@ * Copyright 2017 Joyent, Inc. * Copyright (c) 2017, Intel Corporation. * Copyright (c) 2021, Colm Buckley <colm@tuatha.org> + * Copyright (c) 2023 Hewlett Packard Enterprise Development LP. */ /* @@ -150,7 +151,7 @@ static const char *const zio_taskq_types[ZIO_TASKQ_TYPES] = { * and interrupt) and then to reserve threads for ZIO_PRIORITY_NOW I/Os that * need to be handled with minimum delay. */ -const zio_taskq_info_t zio_taskqs[ZIO_TYPES][ZIO_TASKQ_TYPES] = { +static zio_taskq_info_t zio_taskqs[ZIO_TYPES][ZIO_TASKQ_TYPES] = { /* ISSUE ISSUE_HIGH INTR INTR_HIGH */ { ZTI_ONE, ZTI_NULL, ZTI_ONE, ZTI_NULL }, /* NULL */ { ZTI_N(8), ZTI_NULL, ZTI_SCALE, ZTI_NULL }, /* READ */ @@ -1110,6 +1111,275 @@ spa_taskqs_fini(spa_t *spa, zio_type_t t, zio_taskq_type_t q) tqs->stqs_taskq = NULL; } +#ifdef _KERNEL +/* + * The READ and WRITE rows of zio_taskqs are configurable at module load time + * by setting zio_taskq_read or zio_taskq_write. + * + * Example (the defaults for READ and WRITE) + * zio_taskq_read='fixed,1,8 null scale null' + * zio_taskq_write='batch fixed,1,5 scale fixed,1,5' + * + * Each sets the entire row at a time. + * + * 'fixed' is parameterised: fixed,Q,T where Q is number of taskqs, T is number + * of threads per taskq. + * + * 'null' can only be set on the high-priority queues (queue selection for + * high-priority queues will fall back to the regular queue if the high-pri + * is NULL. + */ +static const char *const modes[ZTI_NMODES] = { + "fixed", "batch", "scale", "null" +}; + +/* Parse the incoming config string. Modifies cfg */ +static int +spa_taskq_param_set(zio_type_t t, char *cfg) +{ + int err = 0; + + zio_taskq_info_t row[ZIO_TASKQ_TYPES] = {{0}}; + + char *next = cfg, *tok, *c; + + /* + * Parse out each element from the string and fill `row`. The entire + * row has to be set at once, so any errors are flagged by just + * breaking out of this loop early. + */ + uint_t q; + for (q = 0; q < ZIO_TASKQ_TYPES; q++) { + /* `next` is the start of the config */ + if (next == NULL) + break; + + /* Eat up leading space */ + while (isspace(*next)) + next++; + if (*next == '\0') + break; + + /* Mode ends at space or end of string */ + tok = next; + next = strchr(tok, ' '); + if (next != NULL) *next++ = '\0'; + + /* Parameters start after a comma */ + c = strchr(tok, ','); + if (c != NULL) *c++ = '\0'; + + /* Match mode string */ + uint_t mode; + for (mode = 0; mode < ZTI_NMODES; mode++) + if (strcmp(tok, modes[mode]) == 0) + break; + if (mode == ZTI_NMODES) + break; + + /* Invalid canary */ + row[q].zti_mode = ZTI_NMODES; + + /* Per-mode setup */ + switch (mode) { + + /* + * FIXED is parameterised: number of queues, and number of + * threads per queue. + */ + case ZTI_MODE_FIXED: { + /* No parameters? */ + if (c == NULL || *c == '\0') + break; + + /* Find next parameter */ + tok = c; + c = strchr(tok, ','); + if (c == NULL) + break; + + /* Take digits and convert */ + unsigned long long nq; + if (!(isdigit(*tok))) + break; + err = ddi_strtoull(tok, &tok, 10, &nq); + /* Must succeed and also end at the next param sep */ + if (err != 0 || tok != c) + break; + + /* Move past the comma */ + tok++; + /* Need another number */ + if (!(isdigit(*tok))) + break; + /* Remember start to make sure we moved */ + c = tok; + + /* Take digits */ + unsigned long long ntpq; + err = ddi_strtoull(tok, &tok, 10, &ntpq); + /* Must succeed, and moved forward */ + if (err != 0 || tok == c || *tok != '\0') + break; + + /* + * sanity; zero queues/threads make no sense, and + * 16K is almost certainly more than anyone will ever + * need and avoids silly numbers like UINT32_MAX + */ + if (nq == 0 || nq >= 16384 || + ntpq == 0 || ntpq >= 16384) + break; + + const zio_taskq_info_t zti = ZTI_P(ntpq, nq); + row[q] = zti; + break; + } + + case ZTI_MODE_BATCH: { + const zio_taskq_info_t zti = ZTI_BATCH; + row[q] = zti; + break; + } + + case ZTI_MODE_SCALE: { + const zio_taskq_info_t zti = ZTI_SCALE; + row[q] = zti; + break; + } + + case ZTI_MODE_NULL: { + /* + * Can only null the high-priority queues; the general- + * purpose ones have to exist. + */ + if (q != ZIO_TASKQ_ISSUE_HIGH && + q != ZIO_TASKQ_INTERRUPT_HIGH) + break; + + const zio_taskq_info_t zti = ZTI_NULL; + row[q] = zti; + break; + } + + default: + break; + } + + /* Ensure we set a mode */ + if (row[q].zti_mode == ZTI_NMODES) + break; + } + + /* Didn't get a full row, fail */ + if (q < ZIO_TASKQ_TYPES) + return (SET_ERROR(EINVAL)); + + /* Eat trailing space */ + if (next != NULL) + while (isspace(*next)) + next++; + + /* If there's anything left over then fail */ + if (next != NULL && *next != '\0') + return (SET_ERROR(EINVAL)); + + /* Success! Copy it into the real config */ + for (q = 0; q < ZIO_TASKQ_TYPES; q++) + zio_taskqs[t][q] = row[q]; + + return (0); +} + +static int +spa_taskq_param_get(zio_type_t t, char *buf, boolean_t add_newline) +{ + int pos = 0; + + /* Build paramater string from live config */ + const char *sep = ""; + for (uint_t q = 0; q < ZIO_TASKQ_TYPES; q++) { + const zio_taskq_info_t *zti = &zio_taskqs[t][q]; + if (zti->zti_mode == ZTI_MODE_FIXED) + pos += sprintf(&buf[pos], "%s%s,%u,%u", sep, + modes[zti->zti_mode], zti->zti_count, + zti->zti_value); + else + pos += sprintf(&buf[pos], "%s%s", sep, + modes[zti->zti_mode]); + sep = " "; + } + + if (add_newline) + buf[pos++] = '\n'; + buf[pos] = '\0'; + + return (pos); +} + +#ifdef __linux__ +static int +spa_taskq_read_param_set(const char *val, zfs_kernel_param_t *kp) +{ + char *cfg = kmem_strdup(val); + int err = spa_taskq_param_set(ZIO_TYPE_READ, cfg); + kmem_free(cfg, strlen(val)+1); + return (-err); +} +static int +spa_taskq_read_param_get(char *buf, zfs_kernel_param_t *kp) +{ + return (spa_taskq_param_get(ZIO_TYPE_READ, buf, TRUE)); +} + +static int +spa_taskq_write_param_set(const char *val, zfs_kernel_param_t *kp) +{ + char *cfg = kmem_strdup(val); + int err = spa_taskq_param_set(ZIO_TYPE_WRITE, cfg); + kmem_free(cfg, strlen(val)+1); + return (-err); +} +static int +spa_taskq_write_param_get(char *buf, zfs_kernel_param_t *kp) +{ + return (spa_taskq_param_get(ZIO_TYPE_WRITE, buf, TRUE)); +} +#else +/* + * On FreeBSD load-time parameters can be set up before malloc() is available, + * so we have to do all the parsing work on the stack. + */ +#define SPA_TASKQ_PARAM_MAX (128) + +static int +spa_taskq_read_param(ZFS_MODULE_PARAM_ARGS) +{ + char buf[SPA_TASKQ_PARAM_MAX]; + int err; + + (void) spa_taskq_param_get(ZIO_TYPE_READ, buf, FALSE); + err = sysctl_handle_string(oidp, buf, sizeof (buf), req); + if (err || req->newptr == NULL) + return (err); + return (spa_taskq_param_set(ZIO_TYPE_READ, buf)); +} + +static int +spa_taskq_write_param(ZFS_MODULE_PARAM_ARGS) +{ + char buf[SPA_TASKQ_PARAM_MAX]; + int err; + + (void) spa_taskq_param_get(ZIO_TYPE_WRITE, buf, FALSE); + err = sysctl_handle_string(oidp, buf, sizeof (buf), req); + if (err || req->newptr == NULL) + return (err); + return (spa_taskq_param_set(ZIO_TYPE_WRITE, buf)); +} +#endif +#endif /* _KERNEL */ + /* * Dispatch a task to the appropriate taskq for the ZFS I/O type and priority. * Note that a type may have multiple discrete taskqs to avoid lock contention @@ -6261,6 +6531,16 @@ spa_tryimport(nvlist_t *tryconfig) spa->spa_config_source = SPA_CONFIG_SRC_SCAN; } + /* + * spa_import() relies on a pool config fetched by spa_try_import() + * for spare/cache devices. Import flags are not passed to + * spa_tryimport(), which makes it return early due to a missing log + * device and missing retrieving the cache device and spare eventually. + * Passing ZFS_IMPORT_MISSING_LOG to spa_tryimport() makes it fetch + * the correct configuration regardless of the missing log device. + */ + spa->spa_import_flags |= ZFS_IMPORT_MISSING_LOG; + error = spa_load(spa, SPA_LOAD_TRYIMPORT, SPA_IMPORT_EXISTING); /* @@ -6747,9 +7027,11 @@ spa_vdev_attach(spa_t *spa, uint64_t guid, nvlist_t *nvroot, int replacing, if (!spa_feature_is_enabled(spa, SPA_FEATURE_DEVICE_REBUILD)) return (spa_vdev_exit(spa, NULL, txg, ENOTSUP)); - if (dsl_scan_resilvering(spa_get_dsl(spa))) + if (dsl_scan_resilvering(spa_get_dsl(spa)) || + dsl_scan_resilver_scheduled(spa_get_dsl(spa))) { return (spa_vdev_exit(spa, NULL, txg, ZFS_ERR_RESILVER_IN_PROGRESS)); + } } else { if (vdev_rebuild_active(rvd)) return (spa_vdev_exit(spa, NULL, txg, @@ -6987,7 +7269,7 @@ spa_vdev_attach(spa_t *spa, uint64_t guid, nvlist_t *nvroot, int replacing, * Detach a device from a mirror or replacing vdev. * * If 'replace_done' is specified, only detach if the parent - * is a replacing vdev. + * is a replacing or a spare vdev. */ int spa_vdev_detach(spa_t *spa, uint64_t guid, uint64_t pguid, int replace_done) @@ -7294,6 +7576,10 @@ spa_vdev_initialize_impl(spa_t *spa, uint64_t guid, uint64_t cmd_type, vd->vdev_initialize_state != VDEV_INITIALIZE_ACTIVE) { mutex_exit(&vd->vdev_initialize_lock); return (SET_ERROR(ESRCH)); + } else if (cmd_type == POOL_INITIALIZE_UNINIT && + vd->vdev_initialize_thread != NULL) { + mutex_exit(&vd->vdev_initialize_lock); + return (SET_ERROR(EBUSY)); } switch (cmd_type) { @@ -7306,6 +7592,9 @@ spa_vdev_initialize_impl(spa_t *spa, uint64_t guid, uint64_t cmd_type, case POOL_INITIALIZE_SUSPEND: vdev_initialize_stop(vd, VDEV_INITIALIZE_SUSPENDED, vd_list); break; + case POOL_INITIALIZE_UNINIT: + vdev_uninitialize(vd); + break; default: panic("invalid cmd_type %llu", (unsigned long long)cmd_type); } @@ -8210,7 +8499,8 @@ spa_async_thread(void *arg) * If any devices are done replacing, detach them. */ if (tasks & SPA_ASYNC_RESILVER_DONE || - tasks & SPA_ASYNC_REBUILD_DONE) { + tasks & SPA_ASYNC_REBUILD_DONE || + tasks & SPA_ASYNC_DETACH_SPARE) { spa_vdev_resilver_done(spa); } @@ -9986,4 +10276,13 @@ ZFS_MODULE_PARAM(zfs_livelist_condense, zfs_livelist_condense_, zthr_cancel, INT ZFS_MODULE_PARAM(zfs_livelist_condense, zfs_livelist_condense_, new_alloc, INT, ZMOD_RW, "Whether extra ALLOC blkptrs were added to a livelist entry while it " "was being condensed"); + +#ifdef _KERNEL +ZFS_MODULE_VIRTUAL_PARAM_CALL(zfs_zio, zio_, taskq_read, + spa_taskq_read_param_set, spa_taskq_read_param_get, ZMOD_RD, + "Configure IO queues for read IO"); +ZFS_MODULE_VIRTUAL_PARAM_CALL(zfs_zio, zio_, taskq_write, + spa_taskq_write_param_set, spa_taskq_write_param_get, ZMOD_RD, + "Configure IO queues for write IO"); +#endif /* END CSTYLED */ diff --git a/module/zfs/spa_misc.c b/module/zfs/spa_misc.c index a57f0727db31..113943026d59 100644 --- a/module/zfs/spa_misc.c +++ b/module/zfs/spa_misc.c @@ -494,8 +494,9 @@ spa_config_tryenter(spa_t *spa, int locks, void *tag, krw_t rw) return (1); } -void -spa_config_enter(spa_t *spa, int locks, const void *tag, krw_t rw) +static void +spa_config_enter_impl(spa_t *spa, int locks, const void *tag, krw_t rw, + int mmp_flag) { (void) tag; int wlocks_held = 0; @@ -510,7 +511,8 @@ spa_config_enter(spa_t *spa, int locks, const void *tag, krw_t rw) continue; mutex_enter(&scl->scl_lock); if (rw == RW_READER) { - while (scl->scl_writer || scl->scl_write_wanted) { + while (scl->scl_writer || + (!mmp_flag && scl->scl_write_wanted)) { cv_wait(&scl->scl_cv, &scl->scl_lock); } } else { @@ -529,6 +531,27 @@ spa_config_enter(spa_t *spa, int locks, const void *tag, krw_t rw) } void +spa_config_enter(spa_t *spa, int locks, const void *tag, krw_t rw) +{ + spa_config_enter_impl(spa, locks, tag, rw, 0); +} + +/* + * The spa_config_enter_mmp() allows the mmp thread to cut in front of + * outstanding write lock requests. This is needed since the mmp updates are + * time sensitive and failure to service them promptly will result in a + * suspended pool. This pool suspension has been seen in practice when there is + * a single disk in a pool that is responding slowly and presumably about to + * fail. + */ + +void +spa_config_enter_mmp(spa_t *spa, int locks, const void *tag, krw_t rw) +{ + spa_config_enter_impl(spa, locks, tag, rw, 1); +} + +void spa_config_exit(spa_t *spa, int locks, const void *tag) { (void) tag; @@ -2564,7 +2587,6 @@ spa_scan_stat_init(spa_t *spa) spa->spa_scan_pass_scrub_spent_paused = 0; spa->spa_scan_pass_exam = 0; spa->spa_scan_pass_issued = 0; - vdev_scan_stat_init(spa->spa_root_vdev); } /* diff --git a/module/zfs/vdev.c b/module/zfs/vdev.c index 4b9d7e7c0506..57259b8ce88e 100644 --- a/module/zfs/vdev.c +++ b/module/zfs/vdev.c @@ -28,7 +28,7 @@ * Copyright 2017 Joyent, Inc. * Copyright (c) 2017, Intel Corporation. * Copyright (c) 2019, Datto Inc. All rights reserved. - * Copyright [2021] Hewlett Packard Enterprise Development LP + * Copyright (c) 2021, 2023 Hewlett Packard Enterprise Development LP. */ #include <sys/zfs_context.h> @@ -2646,6 +2646,17 @@ vdev_reopen(vdev_t *vd) } /* + * Recheck if resilver is still needed and cancel any + * scheduled resilver if resilver is unneeded. + */ + if (!vdev_resilver_needed(spa->spa_root_vdev, NULL, NULL) && + spa->spa_async_tasks & SPA_ASYNC_RESILVER) { + mutex_enter(&spa->spa_async_lock); + spa->spa_async_tasks &= ~SPA_ASYNC_RESILVER; + mutex_exit(&spa->spa_async_lock); + } + + /* * Reassess parent vdev's health. */ vdev_propagate_state(vd); @@ -3983,11 +3994,18 @@ vdev_remove_wanted(spa_t *spa, uint64_t guid) return (spa_vdev_state_exit(spa, NULL, SET_ERROR(ENODEV))); /* - * If the vdev is already removed, then don't do anything. + * If the vdev is already removed, or expanding which can trigger + * repartition add/remove events, then don't do anything. */ - if (vd->vdev_removed) + if (vd->vdev_removed || vd->vdev_expanding) return (spa_vdev_state_exit(spa, NULL, 0)); + /* + * Confirm the vdev has been removed, otherwise don't do anything. + */ + if (vd->vdev_ops->vdev_op_leaf && !zio_wait(vdev_probe(vd, NULL))) + return (spa_vdev_state_exit(spa, NULL, SET_ERROR(EEXIST))); + vd->vdev_remove_wanted = B_TRUE; spa_async_request(spa, SPA_ASYNC_REMOVE); @@ -4085,9 +4103,19 @@ vdev_online(spa_t *spa, uint64_t guid, uint64_t flags, vdev_state_t *newstate) if (wasoffline || (oldstate < VDEV_STATE_DEGRADED && - vd->vdev_state >= VDEV_STATE_DEGRADED)) + vd->vdev_state >= VDEV_STATE_DEGRADED)) { spa_event_notify(spa, vd, NULL, ESC_ZFS_VDEV_ONLINE); + /* + * Asynchronously detach spare vdev if resilver or + * rebuild is not required + */ + if (vd->vdev_unspare && + !dsl_scan_resilvering(spa->spa_dsl_pool) && + !dsl_scan_resilver_scheduled(spa->spa_dsl_pool) && + !vdev_rebuild_active(tvd)) + spa_async_request(spa, SPA_ASYNC_DETACH_SPARE); + } return (spa_vdev_state_exit(spa, vd, 0)); } diff --git a/module/zfs/vdev_indirect.c b/module/zfs/vdev_indirect.c index 8762855d46aa..9e4c115f212c 100644 --- a/module/zfs/vdev_indirect.c +++ b/module/zfs/vdev_indirect.c @@ -270,7 +270,7 @@ typedef struct indirect_split { */ indirect_child_t *is_good_child; - indirect_child_t is_child[1]; /* variable-length */ + indirect_child_t is_child[]; } indirect_split_t; /* diff --git a/module/zfs/vdev_initialize.c b/module/zfs/vdev_initialize.c index 6ffd0d618fdd..5d90fd67cc2f 100644 --- a/module/zfs/vdev_initialize.c +++ b/module/zfs/vdev_initialize.c @@ -101,6 +101,39 @@ vdev_initialize_zap_update_sync(void *arg, dmu_tx_t *tx) } static void +vdev_initialize_zap_remove_sync(void *arg, dmu_tx_t *tx) +{ + uint64_t guid = *(uint64_t *)arg; + + kmem_free(arg, sizeof (uint64_t)); + + vdev_t *vd = spa_lookup_by_guid(tx->tx_pool->dp_spa, guid, B_FALSE); + if (vd == NULL || vd->vdev_top->vdev_removing || !vdev_is_concrete(vd)) + return; + + ASSERT3S(vd->vdev_initialize_state, ==, VDEV_INITIALIZE_NONE); + ASSERT3U(vd->vdev_leaf_zap, !=, 0); + + vd->vdev_initialize_last_offset = 0; + vd->vdev_initialize_action_time = 0; + + objset_t *mos = vd->vdev_spa->spa_meta_objset; + int error; + + error = zap_remove(mos, vd->vdev_leaf_zap, + VDEV_LEAF_ZAP_INITIALIZE_LAST_OFFSET, tx); + VERIFY(error == 0 || error == ENOENT); + + error = zap_remove(mos, vd->vdev_leaf_zap, + VDEV_LEAF_ZAP_INITIALIZE_STATE, tx); + VERIFY(error == 0 || error == ENOENT); + + error = zap_remove(mos, vd->vdev_leaf_zap, + VDEV_LEAF_ZAP_INITIALIZE_ACTION_TIME, tx); + VERIFY(error == 0 || error == ENOENT); +} + +static void vdev_initialize_change_state(vdev_t *vd, vdev_initializing_state_t new_state) { ASSERT(MUTEX_HELD(&vd->vdev_initialize_lock)); @@ -127,8 +160,14 @@ vdev_initialize_change_state(vdev_t *vd, vdev_initializing_state_t new_state) dmu_tx_t *tx = dmu_tx_create_dd(spa_get_dsl(spa)->dp_mos_dir); VERIFY0(dmu_tx_assign(tx, TXG_WAIT)); - dsl_sync_task_nowait(spa_get_dsl(spa), vdev_initialize_zap_update_sync, - guid, tx); + + if (new_state != VDEV_INITIALIZE_NONE) { + dsl_sync_task_nowait(spa_get_dsl(spa), + vdev_initialize_zap_update_sync, guid, tx); + } else { + dsl_sync_task_nowait(spa_get_dsl(spa), + vdev_initialize_zap_remove_sync, guid, tx); + } switch (new_state) { case VDEV_INITIALIZE_ACTIVE: @@ -149,6 +188,10 @@ vdev_initialize_change_state(vdev_t *vd, vdev_initializing_state_t new_state) spa_history_log_internal(spa, "initialize", tx, "vdev=%s complete", vd->vdev_path); break; + case VDEV_INITIALIZE_NONE: + spa_history_log_internal(spa, "uninitialize", tx, + "vdev=%s", vd->vdev_path); + break; default: panic("invalid state %llu", (unsigned long long)new_state); } @@ -605,6 +648,24 @@ vdev_initialize(vdev_t *vd) } /* + * Uninitializes a device. Caller must hold vdev_initialize_lock. + * Device must be a leaf and not already be initializing. + */ +void +vdev_uninitialize(vdev_t *vd) +{ + ASSERT(MUTEX_HELD(&vd->vdev_initialize_lock)); + ASSERT(vd->vdev_ops->vdev_op_leaf); + ASSERT(vdev_is_concrete(vd)); + ASSERT3P(vd->vdev_initialize_thread, ==, NULL); + ASSERT(!vd->vdev_detached); + ASSERT(!vd->vdev_initialize_exit_wanted); + ASSERT(!vd->vdev_top->vdev_removing); + + vdev_initialize_change_state(vd, VDEV_INITIALIZE_NONE); +} + +/* * Wait for the initialize thread to be terminated (cancelled or stopped). */ static void @@ -760,6 +821,7 @@ vdev_initialize_restart(vdev_t *vd) } EXPORT_SYMBOL(vdev_initialize); +EXPORT_SYMBOL(vdev_uninitialize); EXPORT_SYMBOL(vdev_initialize_stop); EXPORT_SYMBOL(vdev_initialize_stop_all); EXPORT_SYMBOL(vdev_initialize_stop_wait); diff --git a/module/zfs/vdev_label.c b/module/zfs/vdev_label.c index ec6bbc6fc610..277c14ec1ad7 100644 --- a/module/zfs/vdev_label.c +++ b/module/zfs/vdev_label.c @@ -468,6 +468,9 @@ vdev_config_generate(spa_t *spa, vdev_t *vd, boolean_t getstats, if (vd->vdev_isspare) fnvlist_add_uint64(nv, ZPOOL_CONFIG_IS_SPARE, 1); + if (flags & VDEV_CONFIG_L2CACHE) + fnvlist_add_uint64(nv, ZPOOL_CONFIG_ASHIFT, vd->vdev_ashift); + if (!(flags & (VDEV_CONFIG_SPARE | VDEV_CONFIG_L2CACHE)) && vd == vd->vdev_top) { fnvlist_add_uint64(nv, ZPOOL_CONFIG_METASLAB_ARRAY, @@ -1100,6 +1103,16 @@ vdev_label_init(vdev_t *vd, uint64_t crtxg, vdev_labeltype_t reason) POOL_STATE_L2CACHE) == 0); VERIFY(nvlist_add_uint64(label, ZPOOL_CONFIG_GUID, vd->vdev_guid) == 0); + + /* + * This is merely to facilitate reporting the ashift of the + * cache device through zdb. The actual retrieval of the + * ashift (in vdev_alloc()) uses the nvlist + * spa->spa_l2cache->sav_config (populated in + * spa_ld_open_aux_vdevs()). + */ + VERIFY(nvlist_add_uint64(label, ZPOOL_CONFIG_ASHIFT, + vd->vdev_ashift) == 0); } else { uint64_t txg = 0ULL; diff --git a/module/zfs/vdev_rebuild.c b/module/zfs/vdev_rebuild.c index 9dfbe0cf6f30..b180fa14682e 100644 --- a/module/zfs/vdev_rebuild.c +++ b/module/zfs/vdev_rebuild.c @@ -34,6 +34,7 @@ #include <sys/zio.h> #include <sys/dmu_tx.h> #include <sys/arc.h> +#include <sys/arc_impl.h> #include <sys/zap.h> /* @@ -116,13 +117,12 @@ unsigned long zfs_rebuild_max_segment = 1024 * 1024; * segment size is also large (zfs_rebuild_max_segment=1M). This helps keep * the queue depth short. * - * 32MB was selected as the default value to achieve good performance with - * a large 90-drive dRAID HDD configuration (draid2:8d:90c:2s). A sequential - * rebuild was unable to saturate all of the drives using smaller values. - * With a value of 32MB the sequential resilver write rate was measured at - * 800MB/s sustained while rebuilding to a distributed spare. + * 64MB was observed to deliver the best performance and set as the default. + * Testing was performed with a 106-drive dRAID HDD pool (draid2:11d:106c) + * and a rebuild rate of 1.2GB/s was measured to the distribute spare. + * Smaller values were unable to fully saturate the available pool I/O. */ -unsigned long zfs_rebuild_vdev_limit = 32 << 20; +unsigned long zfs_rebuild_vdev_limit = 64 << 20; /* * Automatically start a pool scrub when the last active sequential resilver @@ -754,6 +754,7 @@ vdev_rebuild_thread(void *arg) { vdev_t *vd = arg; spa_t *spa = vd->vdev_spa; + vdev_t *rvd = spa->spa_root_vdev; int error = 0; /* @@ -786,9 +787,6 @@ vdev_rebuild_thread(void *arg) vr->vr_pass_bytes_scanned = 0; vr->vr_pass_bytes_issued = 0; - vr->vr_bytes_inflight_max = MAX(1ULL << 20, - zfs_rebuild_vdev_limit * vd->vdev_children); - uint64_t update_est_time = gethrtime(); vdev_rebuild_update_bytes_est(vd, 0); @@ -805,6 +803,17 @@ vdev_rebuild_thread(void *arg) vr->vr_scan_msp = msp; /* + * Calculate the max number of in-flight bytes for top-level + * vdev scanning operations (minimum 1MB, maximum 1/4 of + * arc_c_max shared by all top-level vdevs). Limits for the + * issuing phase are done per top-level vdev and are handled + * separately. + */ + uint64_t limit = (arc_c_max / 4) / MAX(rvd->vdev_children, 1); + vr->vr_bytes_inflight_max = MIN(limit, MAX(1ULL << 20, + zfs_rebuild_vdev_limit * vd->vdev_children)); + + /* * Removal of vdevs from the vdev tree may eliminate the need * for the rebuild, in which case it should be canceled. The * vdev_rebuild_cancel_wanted flag is set until the sync task diff --git a/module/zfs/vdev_trim.c b/module/zfs/vdev_trim.c index 92daed48f3d5..c0ce2ac28dc5 100644 --- a/module/zfs/vdev_trim.c +++ b/module/zfs/vdev_trim.c @@ -23,6 +23,7 @@ * Copyright (c) 2016 by Delphix. All rights reserved. * Copyright (c) 2019 by Lawrence Livermore National Security, LLC. * Copyright (c) 2021 Hewlett Packard Enterprise Development LP + * Copyright 2023 RackTop Systems, Inc. */ #include <sys/spa.h> @@ -572,6 +573,7 @@ vdev_trim_ranges(trim_args_t *ta) uint64_t extent_bytes_max = ta->trim_extent_bytes_max; uint64_t extent_bytes_min = ta->trim_extent_bytes_min; spa_t *spa = vd->vdev_spa; + int error = 0; ta->trim_start_time = gethrtime(); ta->trim_bytes_done = 0; @@ -591,19 +593,32 @@ vdev_trim_ranges(trim_args_t *ta) uint64_t writes_required = ((size - 1) / extent_bytes_max) + 1; for (uint64_t w = 0; w < writes_required; w++) { - int error; - error = vdev_trim_range(ta, VDEV_LABEL_START_SIZE + rs_get_start(rs, ta->trim_tree) + (w *extent_bytes_max), MIN(size - (w * extent_bytes_max), extent_bytes_max)); if (error != 0) { - return (error); + goto done; } } } - return (0); +done: + /* + * Make sure all TRIMs for this metaslab have completed before + * returning. TRIM zios have lower priority over regular or syncing + * zios, so all TRIM zios for this metaslab must complete before the + * metaslab is re-enabled. Otherwise it's possible write zios to + * this metaslab could cut ahead of still queued TRIM zios for this + * metaslab causing corruption if the ranges overlap. + */ + mutex_enter(&vd->vdev_trim_io_lock); + while (vd->vdev_trim_inflight[0] > 0) { + cv_wait(&vd->vdev_trim_io_cv, &vd->vdev_trim_io_lock); + } + mutex_exit(&vd->vdev_trim_io_lock); + + return (error); } static void @@ -922,11 +937,6 @@ vdev_trim_thread(void *arg) } spa_config_exit(spa, SCL_CONFIG, FTAG); - mutex_enter(&vd->vdev_trim_io_lock); - while (vd->vdev_trim_inflight[0] > 0) { - cv_wait(&vd->vdev_trim_io_cv, &vd->vdev_trim_io_lock); - } - mutex_exit(&vd->vdev_trim_io_lock); range_tree_destroy(ta.trim_tree); diff --git a/module/zfs/zfs_ioctl.c b/module/zfs/zfs_ioctl.c index a4b391cbea12..f441328f3018 100644 --- a/module/zfs/zfs_ioctl.c +++ b/module/zfs/zfs_ioctl.c @@ -3985,7 +3985,8 @@ zfs_ioc_pool_initialize(const char *poolname, nvlist_t *innvl, nvlist_t *outnvl) if (!(cmd_type == POOL_INITIALIZE_CANCEL || cmd_type == POOL_INITIALIZE_START || - cmd_type == POOL_INITIALIZE_SUSPEND)) { + cmd_type == POOL_INITIALIZE_SUSPEND || + cmd_type == POOL_INITIALIZE_UNINIT)) { return (SET_ERROR(EINVAL)); } diff --git a/module/zfs/zfs_vnops.c b/module/zfs/zfs_vnops.c index b9498d17ee2f..0987fd0f7bb7 100644 --- a/module/zfs/zfs_vnops.c +++ b/module/zfs/zfs_vnops.c @@ -68,7 +68,9 @@ zfs_fsync(znode_t *zp, int syncflag, cred_t *cr) if (zfsvfs->z_os->os_sync != ZFS_SYNC_DISABLED) { ZFS_ENTER(zfsvfs); ZFS_VERIFY_ZP(zp); + atomic_inc_32(&zp->z_sync_writes_cnt); zil_commit(zfsvfs->z_log, zp->z_id); + atomic_dec_32(&zp->z_sync_writes_cnt); ZFS_EXIT(zfsvfs); } tsd_set(zfs_fsyncer_key, NULL); @@ -102,7 +104,7 @@ zfs_holey_common(znode_t *zp, ulong_t cmd, loff_t *off) hole = B_FALSE; /* Flush any mmap()'d data to disk */ - if (zn_has_cached_data(zp)) + if (zn_has_cached_data(zp, 0, file_sz - 1)) zn_flush_cached_data(zp, B_FALSE); lr = zfs_rangelock_enter(&zp->z_rangelock, 0, UINT64_MAX, RL_READER); @@ -275,7 +277,8 @@ zfs_read(struct znode *zp, zfs_uio_t *uio, int ioflag, cred_t *cr) error = mappedread_sf(zp, nbytes, uio); else #endif - if (zn_has_cached_data(zp) && !(ioflag & O_DIRECT)) { + if (zn_has_cached_data(zp, zfs_uio_offset(uio), + zfs_uio_offset(uio) + nbytes - 1) && !(ioflag & O_DIRECT)) { error = mappedread(zp, nbytes, uio); } else { error = dmu_read_uio_dbuf(sa_get_db(zp->z_sa_hdl), @@ -686,7 +689,8 @@ zfs_write(znode_t *zp, zfs_uio_t *uio, int ioflag, cred_t *cr) zfs_uioskip(uio, nbytes); tx_bytes = nbytes; } - if (tx_bytes && zn_has_cached_data(zp) && + if (tx_bytes && + zn_has_cached_data(zp, woff, woff + tx_bytes - 1) && !(ioflag & O_DIRECT)) { update_pages(zp, woff, tx_bytes, zfsvfs->z_os); } diff --git a/module/zfs/zil.c b/module/zfs/zil.c index aaf509a2fc73..a4f7c008935d 100644 --- a/module/zfs/zil.c +++ b/module/zfs/zil.c @@ -226,11 +226,10 @@ zil_init_log_chain(zilog_t *zilog, blkptr_t *bp) */ static int zil_read_log_block(zilog_t *zilog, boolean_t decrypt, const blkptr_t *bp, - blkptr_t *nbp, void *dst, char **end) + blkptr_t *nbp, char **begin, char **end, arc_buf_t **abuf) { enum zio_flag zio_flags = ZIO_FLAG_CANFAIL; arc_flags_t aflags = ARC_FLAG_WAIT; - arc_buf_t *abuf = NULL; zbookmark_phys_t zb; int error; @@ -247,7 +246,7 @@ zil_read_log_block(zilog_t *zilog, boolean_t decrypt, const blkptr_t *bp, ZB_ZIL_OBJECT, ZB_ZIL_LEVEL, bp->blk_cksum.zc_word[ZIL_ZC_SEQ]); error = arc_read(NULL, zilog->zl_spa, bp, arc_getbuf_func, - &abuf, ZIO_PRIORITY_SYNC_READ, zio_flags, &aflags, &zb); + abuf, ZIO_PRIORITY_SYNC_READ, zio_flags, &aflags, &zb); if (error == 0) { zio_cksum_t cksum = bp->blk_cksum; @@ -262,23 +261,23 @@ zil_read_log_block(zilog_t *zilog, boolean_t decrypt, const blkptr_t *bp, */ cksum.zc_word[ZIL_ZC_SEQ]++; + uint64_t size = BP_GET_LSIZE(bp); if (BP_GET_CHECKSUM(bp) == ZIO_CHECKSUM_ZILOG2) { - zil_chain_t *zilc = abuf->b_data; + zil_chain_t *zilc = (*abuf)->b_data; char *lr = (char *)(zilc + 1); - uint64_t len = zilc->zc_nused - sizeof (zil_chain_t); if (bcmp(&cksum, &zilc->zc_next_blk.blk_cksum, - sizeof (cksum)) || BP_IS_HOLE(&zilc->zc_next_blk)) { + sizeof (cksum)) || BP_IS_HOLE(&zilc->zc_next_blk) || + zilc->zc_nused < sizeof (*zilc) || + zilc->zc_nused > size) { error = SET_ERROR(ECKSUM); } else { - ASSERT3U(len, <=, SPA_OLD_MAXBLOCKSIZE); - bcopy(lr, dst, len); - *end = (char *)dst + len; + *begin = lr; + *end = lr + zilc->zc_nused - sizeof (*zilc); *nbp = zilc->zc_next_blk; } } else { - char *lr = abuf->b_data; - uint64_t size = BP_GET_LSIZE(bp); + char *lr = (*abuf)->b_data; zil_chain_t *zilc = (zil_chain_t *)(lr + size) - 1; if (bcmp(&cksum, &zilc->zc_next_blk.blk_cksum, @@ -286,15 +285,11 @@ zil_read_log_block(zilog_t *zilog, boolean_t decrypt, const blkptr_t *bp, (zilc->zc_nused > (size - sizeof (*zilc)))) { error = SET_ERROR(ECKSUM); } else { - ASSERT3U(zilc->zc_nused, <=, - SPA_OLD_MAXBLOCKSIZE); - bcopy(lr, dst, zilc->zc_nused); - *end = (char *)dst + zilc->zc_nused; + *begin = lr; + *end = lr + zilc->zc_nused; *nbp = zilc->zc_next_blk; } } - - arc_buf_destroy(abuf, &abuf); } return (error); @@ -362,7 +357,6 @@ zil_parse(zilog_t *zilog, zil_parse_blk_func_t *parse_blk_func, uint64_t blk_count = 0; uint64_t lr_count = 0; blkptr_t blk, next_blk; - char *lrbuf, *lrp; int error = 0; bzero(&next_blk, sizeof (blkptr_t)); @@ -382,13 +376,13 @@ zil_parse(zilog_t *zilog, zil_parse_blk_func_t *parse_blk_func, * If the log has been claimed, stop if we encounter a sequence * number greater than the highest claimed sequence number. */ - lrbuf = zio_buf_alloc(SPA_OLD_MAXBLOCKSIZE); zil_bp_tree_init(zilog); for (blk = zh->zh_log; !BP_IS_HOLE(&blk); blk = next_blk) { uint64_t blk_seq = blk.blk_cksum.zc_word[ZIL_ZC_SEQ]; int reclen; - char *end = NULL; + char *lrp, *end; + arc_buf_t *abuf = NULL; if (blk_seq > claim_blk_seq) break; @@ -404,8 +398,10 @@ zil_parse(zilog_t *zilog, zil_parse_blk_func_t *parse_blk_func, break; error = zil_read_log_block(zilog, decrypt, &blk, &next_blk, - lrbuf, &end); + &lrp, &end, &abuf); if (error != 0) { + if (abuf) + arc_buf_destroy(abuf, &abuf); if (claimed) { char name[ZFS_MAX_DATASET_NAME_LEN]; @@ -418,20 +414,25 @@ zil_parse(zilog_t *zilog, zil_parse_blk_func_t *parse_blk_func, break; } - for (lrp = lrbuf; lrp < end; lrp += reclen) { + for (; lrp < end; lrp += reclen) { lr_t *lr = (lr_t *)lrp; reclen = lr->lrc_reclen; ASSERT3U(reclen, >=, sizeof (lr_t)); - if (lr->lrc_seq > claim_lr_seq) + if (lr->lrc_seq > claim_lr_seq) { + arc_buf_destroy(abuf, &abuf); goto done; + } error = parse_lr_func(zilog, lr, arg, txg); - if (error != 0) + if (error != 0) { + arc_buf_destroy(abuf, &abuf); goto done; + } ASSERT3U(max_lr_seq, <, lr->lrc_seq); max_lr_seq = lr->lrc_seq; lr_count++; } + arc_buf_destroy(abuf, &abuf); } done: zilog->zl_parse_error = error; @@ -441,7 +442,6 @@ done: zilog->zl_parse_lr_count = lr_count; zil_bp_tree_fini(zilog); - zio_buf_free(lrbuf, SPA_OLD_MAXBLOCKSIZE); return (error); } @@ -1593,6 +1593,7 @@ zil_lwb_write_issue(zilog_t *zilog, lwb_t *lwb) wsz = P2ROUNDUP_TYPED(lwb->lwb_nused, ZIL_MIN_BLKSZ, uint64_t); ASSERT3U(wsz, <=, lwb->lwb_sz); zio_shrink(lwb->lwb_write_zio, wsz); + wsz = lwb->lwb_write_zio->io_size; } else { wsz = lwb->lwb_sz; @@ -2848,7 +2849,14 @@ static void zil_commit_itx_assign(zilog_t *zilog, zil_commit_waiter_t *zcw) { dmu_tx_t *tx = dmu_tx_create(zilog->zl_os); - VERIFY0(dmu_tx_assign(tx, TXG_WAIT)); + + /* + * Since we are not going to create any new dirty data, and we + * can even help with clearing the existing dirty data, we + * should not be subject to the dirty data based delays. We + * use TXG_NOTHROTTLE to bypass the delay mechanism. + */ + VERIFY0(dmu_tx_assign(tx, TXG_WAIT | TXG_NOTHROTTLE)); itx_t *itx = zil_itx_create(TX_COMMIT, sizeof (lr_t)); itx->itx_sync = B_TRUE; diff --git a/module/zfs/zio.c b/module/zfs/zio.c index 700f8791045f..c367ef7211aa 100644 --- a/module/zfs/zio.c +++ b/module/zfs/zio.c @@ -2287,7 +2287,7 @@ zio_nowait(zio_t *zio) ASSERT3P(zio->io_executor, ==, NULL); if (zio->io_child_type == ZIO_CHILD_LOGICAL && - zio_unique_parent(zio) == NULL) { + list_is_empty(&zio->io_parent_list)) { zio_t *pio; /* diff --git a/module/zstd/Makefile.in b/module/zstd/Makefile.in index 598409ca1627..4d9398efd575 100644 --- a/module/zstd/Makefile.in +++ b/module/zstd/Makefile.in @@ -42,4 +42,4 @@ gensymbols: @OBJDUMP@ -t lib/zstd.o | awk '$$2 == "g" && !/ zfs_/ {print "#define\t" $$6 " zfs_" $$6}' | sort >> include/zstd_compat_wrapper.h checksymbols: - @OBJDUMP@ -t lib/zstd.o | awk '/file format/ {print} $$2 == "g" && !/ zfs_/ {++ret; print} END {exit ret}' + @OBJDUMP@ -t lib/zstd.o | awk '/file format/ {print} $$2 == "g" && (!/ zfs_/ && !/ __pfx_zfs_/) {++ret; print} END {exit ret}' diff --git a/rpm/generic/zfs-dkms.spec.in b/rpm/generic/zfs-dkms.spec.in index 22beb6b68ae3..cd85dd28cf56 100644 --- a/rpm/generic/zfs-dkms.spec.in +++ b/rpm/generic/zfs-dkms.spec.in @@ -24,6 +24,7 @@ BuildRoot: %{_tmppath}/%{name}-%{version}-%{release}-root-%(%{__id_u} -n) BuildArch: noarch Requires: dkms >= 2.2.0.3 +Requires(pre): dkms >= 2.2.0.3 Requires(post): dkms >= 2.2.0.3 Requires(preun): dkms >= 2.2.0.3 Requires: gcc, make, perl, diffutils @@ -68,46 +69,93 @@ fi %defattr(-,root,root) /usr/src/%{module}-%{version} +%pre +echo "Running pre installation script: $0. Parameters: $*" +# We don't want any other versions lingering around in dkms. +# Tests with 'dnf' showed that in case of reinstall, or upgrade +# the preun scriptlet removed the version we are trying to install. +# Because of this, find all zfs dkms sources in /var/lib/dkms and +# remove them, if we find a matching version in dkms. + +dkms_root=/var/lib/dkms +if [ -d ${dkms_root}/%{module} ]; then + cd ${dkms_root}/%{module} + for x in [[:digit:]]*; do + [ -d "$x" ] || continue + otherver="$x" + opath="${dkms_root}/%{module}/${otherver}" + if [ "$otherver" != %{version} ]; then + # This is a workaround for a broken 'dkms status', we caused in a previous version. + # One day it might be not needed anymore, but it does not hurt to keep it. + if dkms status -m %{module} -v "$otherver" 2>&1 | grep "${opath}/source/dkms.conf does not exist" + then + echo "ERROR: dkms status is broken!" >&2 + if [ -L "${opath}/source" -a ! -d "${opath}/source" ] + then + echo "Trying to fix it by removing the symlink: ${opath}/source" >&2 + echo "You should manually remove ${opath}" >&2 + rm -f "${opath}/source" || echo "Removal failed!" >&2 + fi + fi + if [ `dkms status -m %{module} -v "$otherver" | grep -c %{module}` -gt 0 ]; then + echo "Removing old %{module} dkms modules version $otherver from all kernels." + dkms remove -m %{module} -v "$otherver" --all ||: + fi + fi + done + cd ${dkms_root} +fi + +# Uninstall this version of zfs dkms modules before installation of the package. +if [ `dkms status -m %{module} -v %{version} | grep -c %{module}` -gt 0 ]; then + echo "Removing %{module} dkms modules version %{version} from all kernels." + dkms remove -m %{module} -v %{version} --all ||: +fi + %post -for POSTINST in /usr/lib/dkms/common.postinst; do - if [ -f $POSTINST ]; then - $POSTINST %{module} %{version} - exit $? - fi - echo "WARNING: $POSTINST does not exist." -done -echo -e "ERROR: DKMS version is too old and %{module} was not" -echo -e "built with legacy DKMS support." -echo -e "You must either rebuild %{module} with legacy postinst" -echo -e "support or upgrade DKMS to a more current version." -exit 1 +echo "Running post installation script: $0. Parameters: $*" +# Add the module to dkms, as reccommended in the dkms man page. +# This is generally rpm specfic. +# But this also may help, if we have a broken 'dkms status'. +# Because, if the sources are available and only the symlink pointing +# to them is missing, this will resolve the situation +echo "Adding %{module} dkms modules version %{version} to dkms." +dkms add -m %{module} -v %{version} %{!?not_rpm:--rpm_safe_upgrade} ||: + +# After installing the package, dkms install this zfs version for the current kernel. +# Force the overwriting of old modules to avoid diff warnings in dkms status. +# Or in case of a downgrade to overwrite newer versions. +# Or if some other backed up versions have been restored before. +echo "Installing %{module} dkms modules version %{version} for the current kernel." +dkms install --force -m %{module} -v %{version} ||: %preun -# Are we doing an upgrade? +dkms_root="/var/lib/dkms/%{module}/%{version}" +echo "Running pre uninstall script: $0. Parameters: $*" +# In case of upgrade we do nothing. See above comment in pre hook. if [ "$1" = "1" -o "$1" = "upgrade" ] ; then - # Yes we are. Are we upgrading to a new ZFS version? - NEWEST_VER=$(dkms status zfs | tr -d , | sort -r -V | awk '/installed/{print $2; exit}') - if [ "$NEWEST_VER" != "%{version}" ] ; then - # Yes, it's a new ZFS version. We'll uninstall the old module - # later on in this script. - true - else - # No, it's probably an upgrade of the same ZFS version - # to a new distro (zfs-dkms-0.7.12.fc28->zfs-dkms-0.7.12.fc29). - # Don't remove our modules, since the rebuild for the new - # distro will automatically delete the old modules. - exit 0 - fi + echo "This is an upgrade. Skipping pre uninstall action." + exit 0 fi -# If we're here then we're doing an uninstall (not upgrade). -CONFIG_H="/var/lib/dkms/%{module}/%{version}/*/*/%{module}_config.h" -SPEC_META_ALIAS="@PACKAGE@-@VERSION@-@RELEASE@" -DKMS_META_ALIAS=`cat $CONFIG_H 2>/dev/null | - awk -F'"' '/META_ALIAS\s+"/ { print $2; exit 0 }'` -if [ "$SPEC_META_ALIAS" = "$DKMS_META_ALIAS" ]; then - echo -e - echo -e "Uninstall of %{module} module ($SPEC_META_ALIAS) beginning:" - dkms remove -m %{module} -v %{version} --all %{!?not_rpm:--rpm_safe_upgrade} +# Check if we uninstall the package. In that case remove the dkms modules. +# '0' is the value for the first parameter for rpm packages. +# 'remove' or 'purge' are the possible names for deb packages. +if [ "$1" = "0" -o "$1" = "remove" -o "$1" = "purge" ] ; then + if [ `dkms status -m %{module} -v %{version} | grep -c %{module}` -gt 0 ]; then + echo "Removing %{module} dkms modules version %{version} from all kernels." + dkms remove -m %{module} -v %{version} --all %{!?not_rpm:--rpm_safe_upgrade} && exit 0 + fi + # If removing the modules failed, it might be because of the broken 'dkms status'. + if dkms status -m %{module} -v %{version} 2>&1 | grep "${dkms_root}/source/dkms.conf does not exist" + then + echo "ERROR: dkms status is broken!" >&2 + echo "You should manually remove ${dkms_root}" >&2 + echo "WARNING: installed modules in /lib/modules/`uname -r`/extra could not be removed automatically!" >&2 + fi +else + echo "Script parameter $1 did not match any removal condition." fi + exit 0 + diff --git a/scripts/Makefile.am b/scripts/Makefile.am index 047ae7eaca6d..f3c133a29ae8 100644 --- a/scripts/Makefile.am +++ b/scripts/Makefile.am @@ -9,6 +9,9 @@ dist_pkgdata_SCRIPTS = \ zloop.sh \ zfs-helpers.sh +dist_zfsexec_SCRIPTS = \ + zfs_prepare_disk + EXTRA_SCRIPTS = \ commitcheck.sh \ common.sh.in \ diff --git a/scripts/zfs_prepare_disk b/scripts/zfs_prepare_disk new file mode 100755 index 000000000000..02aa9f8a7728 --- /dev/null +++ b/scripts/zfs_prepare_disk @@ -0,0 +1,17 @@ +#!/bin/sh +# +# This is an optional helper script that is automatically called by libzfs +# before a disk is about to be added into the pool. It can be modified by +# the user to run whatever commands are necessary to prepare a disk for +# inclusion into the pool. For example, users can add lines to this +# script to do things like update the drive's firmware or check the drive's +# health. The script is optional and can be removed if it is not needed. +# +# See the zfs_prepare_disk(8) man page for details. +# +# Example: +# +# echo "Prepare disk $VDEV_PATH ($VDEV_UPATH) for $VDEV_PREPARE in $POOL_NAME" +# + +exit 0 diff --git a/tests/runfiles/common.run b/tests/runfiles/common.run index 554cf96f883a..5097e08b1b55 100644 --- a/tests/runfiles/common.run +++ b/tests/runfiles/common.run @@ -37,7 +37,7 @@ tests = ['alloc_class_001_pos', 'alloc_class_002_neg', 'alloc_class_003_pos', 'alloc_class_004_pos', 'alloc_class_005_pos', 'alloc_class_006_pos', 'alloc_class_007_pos', 'alloc_class_008_pos', 'alloc_class_009_pos', 'alloc_class_010_pos', 'alloc_class_011_neg', 'alloc_class_012_pos', - 'alloc_class_013_pos'] + 'alloc_class_013_pos', 'alloc_class_014_neg', 'alloc_class_015_pos'] tags = ['functional', 'alloc_class'] [tests/functional/arc] @@ -407,7 +407,7 @@ tests = ['zpool_import_001_pos', 'zpool_import_002_pos', 'import_cachefile_mirror_detached', 'import_cachefile_paths_changed', 'import_cachefile_shared_device', - 'import_devices_missing', + 'import_devices_missing', 'import_log_missing', 'import_paths_changed', 'import_rewind_config_changed', 'import_rewind_device_replaced'] @@ -431,6 +431,7 @@ tests = ['zpool_initialize_attach_detach_add_remove', 'zpool_initialize_start_and_cancel_neg', 'zpool_initialize_start_and_cancel_pos', 'zpool_initialize_suspend_resume', + 'zpool_initialize_uninit', 'zpool_initialize_unsupported_vdevs', 'zpool_initialize_verify_checksums', 'zpool_initialize_verify_initialized'] @@ -456,7 +457,8 @@ tests = ['zpool_replace_001_neg', 'replace-o_ashift', 'replace_prop_ashift'] tags = ['functional', 'cli_root', 'zpool_replace'] [tests/functional/cli_root/zpool_resilver] -tests = ['zpool_resilver_bad_args', 'zpool_resilver_restart'] +tests = ['zpool_resilver_bad_args', 'zpool_resilver_restart', + 'zpool_resilver_concurrent'] tags = ['functional', 'cli_root', 'zpool_resilver'] [tests/functional/cli_root/zpool_scrub] @@ -480,7 +482,7 @@ tags = ['functional', 'cli_root', 'zpool_split'] [tests/functional/cli_root/zpool_status] tests = ['zpool_status_001_pos', 'zpool_status_002_pos', - 'zpool_status_features_001_pos'] + 'zpool_status_008_pos', 'zpool_status_features_001_pos'] tags = ['functional', 'cli_root', 'zpool_status'] [tests/functional/cli_root/zpool_sync] @@ -573,7 +575,7 @@ tests = ['compress_001_pos', 'compress_002_pos', 'compress_003_pos', tags = ['functional', 'compression'] [tests/functional/cp_files] -tests = ['cp_files_001_pos'] +tests = ['cp_files_001_pos', 'cp_stress'] tags = ['functional', 'cp_files'] [tests/functional/crtime] @@ -669,7 +671,8 @@ tests = ['migration_001_pos', 'migration_002_pos', 'migration_003_pos', tags = ['functional', 'migration'] [tests/functional/mmap] -tests = ['mmap_write_001_pos', 'mmap_read_001_pos', 'mmap_seek_001_pos'] +tests = ['mmap_mixed', 'mmap_read_001_pos', 'mmap_seek_001_pos', + 'mmap_write_001_pos', 'mmap_sync_001_pos'] tags = ['functional', 'mmap'] [tests/functional/mount] @@ -823,9 +826,9 @@ tests = ['recv_dedup', 'recv_dedup_encrypted_zvol', 'rsend_001_pos', 'send-c_mixed_compression', 'send-c_stream_size_estimate', 'send-c_embedded_blocks', 'send-c_resume', 'send-cpL_varied_recsize', 'send-c_recv_dedup', 'send-L_toggle', - 'send_encrypted_incremental.ksh', 'send_encrypted_hierarchy', - 'send_encrypted_props', 'send_encrypted_truncated_files', - 'send_freeobjects', 'send_realloc_files', + 'send_encrypted_incremental.ksh', 'send_encrypted_freeobjects', + 'send_encrypted_hierarchy', 'send_encrypted_props', + 'send_encrypted_truncated_files', 'send_freeobjects', 'send_realloc_files', 'send_realloc_encrypted_files', 'send_spill_block', 'send_holds', 'send_hole_birth', 'send_mixed_raw', 'send-wR_encrypted_zvol', 'send_partial_dataset', 'send_invalid', 'send_doall', diff --git a/tests/runfiles/linux.run b/tests/runfiles/linux.run index 94c1cbbc3f9f..c60b00ca119d 100644 --- a/tests/runfiles/linux.run +++ b/tests/runfiles/linux.run @@ -53,7 +53,7 @@ tags = ['functional', 'cli_root', 'zfs_mount'] [tests/functional/cli_root/zfs_share:Linux] tests = ['zfs_share_005_pos', 'zfs_share_007_neg', 'zfs_share_009_neg', - 'zfs_share_012_pos'] + 'zfs_share_012_pos', 'zfs_share_013_pos'] tags = ['functional', 'cli_root', 'zfs_share'] [tests/functional/cli_root/zfs_sysfs:Linux] diff --git a/tests/runfiles/sanity.run b/tests/runfiles/sanity.run index fb39fa54b9d5..0a3d42cb269d 100644 --- a/tests/runfiles/sanity.run +++ b/tests/runfiles/sanity.run @@ -547,6 +547,7 @@ tests = ['recv_dedup', 'recv_dedup_encrypted_zvol', 'rsend_001_pos', 'rsend_014_pos', 'rsend_016_neg', 'send-c_verify_contents', 'send-c_volume', 'send-c_zstreamdump', 'send-c_recv_dedup', 'send-L_toggle', 'send_encrypted_hierarchy', 'send_encrypted_props', + 'send_encrypted_freeobjects', 'send_encrypted_truncated_files', 'send_freeobjects', 'send_holds', 'send_mixed_raw', 'send-wR_encrypted_zvol', 'send_partial_dataset', 'send_invalid'] diff --git a/tests/test-runner/bin/test-runner.py.in b/tests/test-runner/bin/test-runner.py.in index a652d3d4a0ff..5c868d945a8c 100755 --- a/tests/test-runner/bin/test-runner.py.in +++ b/tests/test-runner/bin/test-runner.py.in @@ -33,7 +33,7 @@ from subprocess import PIPE from subprocess import Popen from subprocess import check_output from threading import Timer -from time import time, CLOCK_MONOTONIC_RAW +from time import time, CLOCK_MONOTONIC from os.path import exists BASEDIR = '/var/tmp/test_results' @@ -62,7 +62,7 @@ clock_gettime.argtypes = [ctypes.c_int, ctypes.POINTER(timespec)] def monotonic_time(): t = timespec() - if clock_gettime(CLOCK_MONOTONIC_RAW, ctypes.pointer(t)) != 0: + if clock_gettime(CLOCK_MONOTONIC, ctypes.pointer(t)) != 0: errno_ = ctypes.get_errno() raise OSError(errno_, os.strerror(errno_)) return t.tv_sec + t.tv_nsec * 1e-9 diff --git a/tests/test-runner/bin/zts-report.py.in b/tests/test-runner/bin/zts-report.py.in index 432899c21f4d..44a798850a30 100755 --- a/tests/test-runner/bin/zts-report.py.in +++ b/tests/test-runner/bin/zts-report.py.in @@ -164,8 +164,8 @@ summary = { # reasons listed above can be used. # known = { - 'casenorm/mixed_none_lookup_ci': ['FAIL', '7633'], - 'casenorm/mixed_formd_lookup_ci': ['FAIL', '7633'], + 'casenorm/mixed_none_lookup_ci': ['FAIL', 7633], + 'casenorm/mixed_formd_lookup_ci': ['FAIL', 7633], 'cli_root/zfs_unshare/zfs_unshare_002_pos': ['SKIP', na_reason], 'cli_root/zfs_unshare/zfs_unshare_006_pos': ['SKIP', na_reason], 'cli_root/zpool_import/import_rewind_device_replaced': @@ -175,7 +175,7 @@ known = { 'privilege/setup': ['SKIP', na_reason], 'refreserv/refreserv_004_pos': ['FAIL', known_reason], 'rootpool/setup': ['SKIP', na_reason], - 'rsend/rsend_008_pos': ['SKIP', '6066'], + 'rsend/rsend_008_pos': ['SKIP', 6066], 'vdev_zaps/vdev_zaps_007_pos': ['FAIL', known_reason], } @@ -183,17 +183,20 @@ if sys.platform.startswith('freebsd'): known.update({ 'cli_root/zfs_receive/receive-o-x_props_override': ['FAIL', known_reason], + 'cli_root/zpool_resilver/zpool_resilver_concurrent': + ['SKIP', na_reason], 'cli_root/zpool_wait/zpool_wait_trim_basic': ['SKIP', trim_reason], 'cli_root/zpool_wait/zpool_wait_trim_cancel': ['SKIP', trim_reason], 'cli_root/zpool_wait/zpool_wait_trim_flag': ['SKIP', trim_reason], 'link_count/link_count_001': ['SKIP', na_reason], + 'mmap/mmap_sync_001_pos': ['SKIP', na_reason], }) elif sys.platform.startswith('linux'): known.update({ - 'casenorm/mixed_formd_lookup': ['FAIL', '7633'], - 'casenorm/mixed_formd_delete': ['FAIL', '7633'], - 'casenorm/sensitive_formd_lookup': ['FAIL', '7633'], - 'casenorm/sensitive_formd_delete': ['FAIL', '7633'], + 'casenorm/mixed_formd_lookup': ['FAIL', 7633], + 'casenorm/mixed_formd_delete': ['FAIL', 7633], + 'casenorm/sensitive_formd_lookup': ['FAIL', 7633], + 'casenorm/sensitive_formd_delete': ['FAIL', 7633], 'removal/removal_with_zdb': ['SKIP', known_reason], }) @@ -210,56 +213,56 @@ elif sys.platform.startswith('linux'): # reasons listed above can be used. # maybe = { + 'threadsappend/threadsappend_001_pos': ['FAIL', 6136], 'chattr/setup': ['SKIP', exec_reason], 'crtime/crtime_001_pos': ['SKIP', statx_reason], 'cli_root/zdb/zdb_006_pos': ['FAIL', known_reason], 'cli_root/zfs_destroy/zfs_destroy_dev_removal_condense': ['FAIL', known_reason], 'cli_root/zfs_get/zfs_get_004_pos': ['FAIL', known_reason], - 'cli_root/zfs_get/zfs_get_009_pos': ['SKIP', '5479'], + 'cli_root/zfs_get/zfs_get_009_pos': ['SKIP', 5479], 'cli_root/zfs_rollback/zfs_rollback_001_pos': ['FAIL', known_reason], 'cli_root/zfs_rollback/zfs_rollback_002_pos': ['FAIL', known_reason], 'cli_root/zfs_share/setup': ['SKIP', share_reason], 'cli_root/zfs_snapshot/zfs_snapshot_002_neg': ['FAIL', known_reason], 'cli_root/zfs_unshare/setup': ['SKIP', share_reason], 'cli_root/zpool_add/zpool_add_004_pos': ['FAIL', known_reason], - 'cli_root/zpool_destroy/zpool_destroy_001_pos': ['SKIP', '6145'], - 'cli_root/zpool_import/zpool_import_missing_003_pos': ['SKIP', '6839'], + 'cli_root/zpool_destroy/zpool_destroy_001_pos': ['SKIP', 6145], + 'cli_root/zpool_import/zpool_import_missing_003_pos': ['SKIP', 6839], 'cli_root/zpool_initialize/zpool_initialize_import_export': - ['FAIL', '11948'], + ['FAIL', 11948], 'cli_root/zpool_labelclear/zpool_labelclear_removed': ['FAIL', known_reason], 'cli_root/zpool_trim/setup': ['SKIP', trim_reason], - 'cli_root/zpool_upgrade/zpool_upgrade_004_pos': ['FAIL', '6141'], + 'cli_root/zpool_upgrade/zpool_upgrade_004_pos': ['FAIL', 6141], 'delegate/setup': ['SKIP', exec_reason], 'fallocate/fallocate_punch-hole': ['SKIP', fspacectl_reason], - 'history/history_004_pos': ['FAIL', '7026'], - 'history/history_005_neg': ['FAIL', '6680'], - 'history/history_006_neg': ['FAIL', '5657'], + 'history/history_004_pos': ['FAIL', 7026], + 'history/history_005_neg': ['FAIL', 6680], + 'history/history_006_neg': ['FAIL', 5657], 'history/history_008_pos': ['FAIL', known_reason], 'history/history_010_pos': ['SKIP', exec_reason], 'io/mmap': ['SKIP', fio_reason], 'largest_pool/largest_pool_001_pos': ['FAIL', known_reason], 'mmp/mmp_on_uberblocks': ['FAIL', known_reason], 'pyzfs/pyzfs_unittest': ['SKIP', python_deps_reason], - 'pool_checkpoint/checkpoint_discard_busy': ['FAIL', '11946'], + 'pool_checkpoint/checkpoint_discard_busy': ['FAIL', 11946], + 'pam/setup': ['SKIP', "pamtester might be not available"], 'projectquota/setup': ['SKIP', exec_reason], 'removal/removal_condense_export': ['FAIL', known_reason], - 'reservation/reservation_008_pos': ['FAIL', '7741'], - 'reservation/reservation_018_pos': ['FAIL', '5642'], + 'reservation/reservation_008_pos': ['FAIL', 7741], + 'reservation/reservation_018_pos': ['FAIL', 5642], 'snapshot/clone_001_pos': ['FAIL', known_reason], - 'snapshot/snapshot_009_pos': ['FAIL', '7961'], - 'snapshot/snapshot_010_pos': ['FAIL', '7961'], - 'snapused/snapused_004_pos': ['FAIL', '5513'], + 'snapshot/snapshot_009_pos': ['FAIL', 7961], + 'snapshot/snapshot_010_pos': ['FAIL', 7961], + 'snapused/snapused_004_pos': ['FAIL', 5513], 'tmpfile/setup': ['SKIP', tmpfile_reason], - 'threadsappend/threadsappend_001_pos': ['FAIL', '6136'], 'trim/setup': ['SKIP', trim_reason], 'upgrade/upgrade_projectquota_001_pos': ['SKIP', project_id_reason], 'user_namespace/setup': ['SKIP', user_ns_reason], 'userquota/setup': ['SKIP', exec_reason], - 'vdev_zaps/vdev_zaps_004_pos': ['FAIL', '6935'], - 'zvol/zvol_ENOSPC/zvol_ENOSPC_001_pos': ['FAIL', '5848'], - 'pam/setup': ['SKIP', "pamtester might be not available"], + 'vdev_zaps/vdev_zaps_004_pos': ['FAIL', known_reason], + 'zvol/zvol_ENOSPC/zvol_ENOSPC_001_pos': ['FAIL', 5848], } if sys.platform.startswith('freebsd'): @@ -271,23 +274,29 @@ if sys.platform.startswith('freebsd'): ['FAIL', known_reason], 'cli_root/zpool_import/zpool_import_012_pos': ['FAIL', known_reason], 'delegate/zfs_allow_003_pos': ['FAIL', known_reason], - 'inheritance/inherit_001_pos': ['FAIL', '11829'], + 'inheritance/inherit_001_pos': ['FAIL', 11829], 'resilver/resilver_restart_001': ['FAIL', known_reason], - 'pool_checkpoint/checkpoint_big_rewind': ['FAIL', '12622'], - 'pool_checkpoint/checkpoint_indirect': ['FAIL', '12623'], + 'pool_checkpoint/checkpoint_big_rewind': ['FAIL', 12622], + 'pool_checkpoint/checkpoint_indirect': ['FAIL', 12623], + 'snapshot/snapshot_002_pos': ['FAIL', 14831], }) elif sys.platform.startswith('linux'): maybe.update({ 'cli_root/zfs_rename/zfs_rename_002_pos': ['FAIL', known_reason], 'cli_root/zpool_reopen/zpool_reopen_003_pos': ['FAIL', known_reason], - 'fault/auto_spare_shared': ['FAIL', '11889'], + 'fault/auto_online_002_pos': ['FAIL', 11889], + 'fault/auto_replace_001_pos': ['FAIL', 14851], + 'fault/auto_spare_002_pos': ['FAIL', 11889], + 'fault/auto_spare_multiple': ['FAIL', 11889], + 'fault/auto_spare_shared': ['FAIL', 11889], + 'fault/decompress_fault': ['FAIL', 11889], 'io/io_uring': ['SKIP', 'io_uring support required'], 'limits/filesystem_limit': ['SKIP', known_reason], 'limits/snapshot_limit': ['SKIP', known_reason], 'mmp/mmp_active_import': ['FAIL', known_reason], 'mmp/mmp_exported_import': ['FAIL', known_reason], 'mmp/mmp_inactive_import': ['FAIL', known_reason], - 'zvol/zvol_misc/zvol_misc_snapdev': ['FAIL', '12621'], + 'zvol/zvol_misc/zvol_misc_snapdev': ['FAIL', 12621], 'zvol/zvol_misc/zvol_misc_volmode': ['FAIL', known_reason], }) @@ -318,7 +327,7 @@ if os.environ.get('CI') == 'true': }) maybe.update({ - 'events/events_002_pos': ['FAIL', '11546'], + 'events/events_002_pos': ['FAIL', 11546], }) elif sys.platform.startswith('linux'): maybe.update({ @@ -442,13 +451,13 @@ if __name__ == "__main__": if test in known: if known[test][1] == na_reason: continue - elif known[test][1].isdigit(): - expect = issue_url + known[test][1] + elif isinstance(known[test][1], int): + expect = f"{issue_url}{known[test][1]}" else: expect = known[test][1] elif test in maybe: - if maybe[test][1].isdigit(): - expect = issue_url + maybe[test][1] + if isinstance(maybe[test][1], int): + expect = f"{issue_url}{maybe[test][1]}" else: expect = maybe[test][1] elif setup in known and known[setup][0] == "SKIP" and setup != test: diff --git a/tests/zfs-tests/cmd/Makefile.am b/tests/zfs-tests/cmd/Makefile.am index d1c29fcd1c62..7ec4cb6190a8 100644 --- a/tests/zfs-tests/cmd/Makefile.am +++ b/tests/zfs-tests/cmd/Makefile.am @@ -20,6 +20,7 @@ SUBDIRS = \ mmap_exec \ mmap_libaio \ mmap_seek \ + mmap_sync \ mmapwrite \ nvlist_to_lua \ randwritecomp \ diff --git a/tests/zfs-tests/cmd/mmap_sync/.gitignore b/tests/zfs-tests/cmd/mmap_sync/.gitignore new file mode 100644 index 000000000000..c721f472ba44 --- /dev/null +++ b/tests/zfs-tests/cmd/mmap_sync/.gitignore @@ -0,0 +1 @@ +/mmap_sync diff --git a/tests/zfs-tests/cmd/mmap_sync/Makefile.am b/tests/zfs-tests/cmd/mmap_sync/Makefile.am new file mode 100644 index 000000000000..313e8db5c0f4 --- /dev/null +++ b/tests/zfs-tests/cmd/mmap_sync/Makefile.am @@ -0,0 +1,6 @@ +include $(top_srcdir)/config/Rules.am + +pkgexecdir = $(datadir)/@PACKAGE@/zfs-tests/bin + +pkgexec_PROGRAMS = mmap_sync +mmap_sync_SOURCES = mmap_sync.c diff --git a/tests/zfs-tests/cmd/mmap_sync/mmap_sync.c b/tests/zfs-tests/cmd/mmap_sync/mmap_sync.c new file mode 100644 index 000000000000..226e71be2f57 --- /dev/null +++ b/tests/zfs-tests/cmd/mmap_sync/mmap_sync.c @@ -0,0 +1,152 @@ +/* + * CDDL HEADER START + * + * The contents of this file are subject to the terms of the + * Common Development and Distribution License (the "License"). + * You may not use this file except in compliance with the License. + * + * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE + * or http://opensource.org/licenses/CDDL-1.0. + * See the License for the specific language governing permissions + * and limitations under the License. + * + * When distributing Covered Code, include this CDDL HEADER in each + * file and include the License file at usr/src/OPENSOLARIS.LICENSE. + * If applicable, add the following below this CDDL HEADER, with the + * fields enclosed by brackets "[]" replaced with your own identifying + * information: Portions Copyright [yyyy] [name of copyright owner] + * + * CDDL HEADER END + */ + +#include <stdlib.h> +#include <stdio.h> +#include <string.h> +#include <sys/mman.h> +#include <sys/stat.h> +#include <sys/time.h> +#include <fcntl.h> +#include <unistd.h> +#include <time.h> + +static void +cleanup(char *file) +{ + (void) remove(file); +} + +int +main(int argc, char *argv[]) +{ + char *testdir = getenv("TESTDIR"); + if (!testdir) { + fprintf(stderr, "environment variable TESTDIR not set\n"); + return (1); + } + + struct stat st; + umask(0); + if (stat(testdir, &st) != 0 && + mkdir(testdir, 0777) != 0) { + perror("mkdir"); + return (1); + } + + if (argc > 3) { + fprintf(stderr, "usage: %s " + "[run time in mins] " + "[max msync time in ms]\n", argv[0]); + return (1); + } + + int run_time_mins = 1; + if (argc >= 2) { + run_time_mins = atoi(argv[1]); + } + + int max_msync_time_ms = 1000; + if (argc >= 3) { + max_msync_time_ms = atoi(argv[2]); + } + + char filepath[512]; + filepath[0] = '\0'; + char *file = &filepath[0]; + + (void) snprintf(file, 512, "%s/msync_file", testdir); + + const int LEN = 8; + cleanup(file); + + int fd = open(file, O_CREAT | O_RDWR, S_IRUSR | S_IWUSR | + S_IRGRP | S_IROTH); + + if (fd == -1) { + (void) fprintf(stderr, "%s: %s: ", argv[0], file); + perror("open"); + return (1); + } + + if (ftruncate(fd, LEN) != 0) { + perror("ftruncate"); + cleanup(file); + return (1); + } + + void *ptr = mmap(NULL, LEN, PROT_READ | PROT_WRITE, MAP_SHARED, fd, 0); + + if (ptr == MAP_FAILED) { + perror("mmap"); + cleanup(file); + return (1); + } + + struct timeval tstart; + gettimeofday(&tstart, NULL); + + long long x = 0LL; + + for (;;) { + *((long long *)ptr) = x; + x++; + + struct timeval t1, t2; + gettimeofday(&t1, NULL); + if (msync(ptr, LEN, MS_SYNC|MS_INVALIDATE) != 0) { + perror("msync"); + cleanup(file); + return (1); + } + + gettimeofday(&t2, NULL); + + double elapsed = (t2.tv_sec - t1.tv_sec) * 1000.0; + elapsed += ((t2.tv_usec - t1.tv_usec) / 1000.0); + if (elapsed > max_msync_time_ms) { + fprintf(stderr, "slow msync: %f ms\n", elapsed); + if (munmap(ptr, LEN) != 0) + perror("munmap"); + cleanup(file); + return (1); + } + + double elapsed_start = (t2.tv_sec - tstart.tv_sec) * 1000.0; + elapsed_start += ((t2.tv_usec - tstart.tv_usec) / 1000.0); + if (elapsed_start > run_time_mins * 60 * 1000) { + break; + } + } + + if (munmap(ptr, LEN) != 0) { + perror("munmap"); + cleanup(file); + return (1); + } + + if (close(fd) != 0) { + perror("close"); + } + + cleanup(file); + return (0); +} diff --git a/tests/zfs-tests/include/commands.cfg b/tests/zfs-tests/include/commands.cfg index 78802c9fb942..8ac38dfd8934 100644 --- a/tests/zfs-tests/include/commands.cfg +++ b/tests/zfs-tests/include/commands.cfg @@ -207,6 +207,7 @@ export ZFSTEST_FILES='badsend mmap_exec mmap_libaio mmap_seek + mmap_sync mmapwrite nvlist_to_lua randfree_file diff --git a/tests/zfs-tests/include/libtest.shlib b/tests/zfs-tests/include/libtest.shlib index 89c6382dedc3..9e7f46d5f7d1 100644 --- a/tests/zfs-tests/include/libtest.shlib +++ b/tests/zfs-tests/include/libtest.shlib @@ -37,6 +37,12 @@ . ${STF_SUITE}/include/tunables.cfg +# On AlmaLinux 9 we will see $PWD = '.' instead of the full path. This causes +# some tests to fail. Fix it up here. +if [ "$PWD" = "." ] ; then + PWD="$(readlink -f $PWD)" +fi + # # Apply constrained path when available. This is required since the # PATH may have been modified by sudo's secure_path behavior. @@ -3356,6 +3362,18 @@ function is_te_enabled fi } +# Return the number of CPUs (cross-platform) +function get_num_cpus +{ + if is_linux ; then + grep -c '^processor' /proc/cpuinfo + elif is_freebsd; then + sysctl -n kern.smp.cpus + else + psrinfo | wc -l + fi +} + # Utility function to determine if a system has multiple cpus. function is_mp { @@ -4268,7 +4286,7 @@ function arcstat_quiescence # stat echo while $do_once || [ $stat1 -ne $stat2 ] || [ $stat2 -eq 0 ]; do typeset stat1=$(get_arcstat $stat) - sleep 2 + sleep 0.5 typeset stat2=$(get_arcstat $stat) do_once=false done diff --git a/tests/zfs-tests/tests/functional/alloc_class/Makefile.am b/tests/zfs-tests/tests/functional/alloc_class/Makefile.am index 7cffb2eac450..82fd9f340269 100644 --- a/tests/zfs-tests/tests/functional/alloc_class/Makefile.am +++ b/tests/zfs-tests/tests/functional/alloc_class/Makefile.am @@ -14,7 +14,9 @@ dist_pkgdata_SCRIPTS = \ alloc_class_010_pos.ksh \ alloc_class_011_neg.ksh \ alloc_class_012_pos.ksh \ - alloc_class_013_pos.ksh + alloc_class_013_pos.ksh \ + alloc_class_014_neg.ksh \ + alloc_class_015_pos.ksh dist_pkgdata_DATA = \ alloc_class.cfg \ diff --git a/tests/zfs-tests/tests/functional/alloc_class/alloc_class_013_pos.ksh b/tests/zfs-tests/tests/functional/alloc_class/alloc_class_013_pos.ksh index 2ce22a624291..790a47f26576 100755 --- a/tests/zfs-tests/tests/functional/alloc_class/alloc_class_013_pos.ksh +++ b/tests/zfs-tests/tests/functional/alloc_class/alloc_class_013_pos.ksh @@ -42,7 +42,8 @@ log_must display_status "$TESTPOOL" log_must zfs create -o dedup=on -V 2G $TESTPOOL/$TESTVOL -log_must eval "new_fs $ZVOL_DEVDIR/$TESTPOOL/$TESTVOL >/dev/null 2>&1" +block_device_wait "$ZVOL_DEVDIR/$TESTPOOL/$TESTVOL" +log_must eval "new_fs $ZVOL_DEVDIR/$TESTPOOL/$TESTVOL >/dev/null" sync_pool log_must zpool list -v $TESTPOOL diff --git a/tests/zfs-tests/tests/functional/alloc_class/alloc_class_014_neg.ksh b/tests/zfs-tests/tests/functional/alloc_class/alloc_class_014_neg.ksh new file mode 100755 index 000000000000..1b52014fd2d9 --- /dev/null +++ b/tests/zfs-tests/tests/functional/alloc_class/alloc_class_014_neg.ksh @@ -0,0 +1,38 @@ +#!/bin/ksh -p + +# +# This file and its contents are supplied under the terms of the +# Common Development and Distribution License ("CDDL"), version 1.0. +# You may only use this file in accordance with the terms of version +# 1.0 of the CDDL. +# +# A full copy of the text of the CDDL should have accompanied this +# source. A copy of the CDDL is also available via the Internet at +# http://www.illumos.org/license/CDDL. +# + +. $STF_SUITE/tests/functional/alloc_class/alloc_class.kshlib + +# +# DESCRIPTION: +# Setting the special_small_blocks property greater than recordsize fails. +# + +verify_runnable "global" + +claim="Setting the special_small_blocks property greater than recordsize fails" + +log_assert $claim +log_onexit cleanup +log_must disk_setup + +for size in 512 4096 32768 131072 524288 1048576 +do + let bigger=$size*2 + log_mustnot zpool create -O recordsize=$size \ + -O special_small_blocks=$bigger \ + $TESTPOOL raidz $ZPOOL_DISKS special mirror \ + $CLASS_DISK0 $CLASS_DISK1 +done + +log_pass $claim diff --git a/tests/zfs-tests/tests/functional/alloc_class/alloc_class_015_pos.ksh b/tests/zfs-tests/tests/functional/alloc_class/alloc_class_015_pos.ksh new file mode 100755 index 000000000000..49c468af6702 --- /dev/null +++ b/tests/zfs-tests/tests/functional/alloc_class/alloc_class_015_pos.ksh @@ -0,0 +1,45 @@ +#!/bin/ksh -p + +# +# This file and its contents are supplied under the terms of the +# Common Development and Distribution License ("CDDL"), version 1.0. +# You may only use this file in accordance with the terms of version +# 1.0 of the CDDL. +# +# A full copy of the text of the CDDL should have accompanied this +# source. A copy of the CDDL is also available via the Internet at +# http://www.illumos.org/license/CDDL. +# + +. $STF_SUITE/tests/functional/alloc_class/alloc_class.kshlib + +# +# DESCRIPTION: +# Can set special_small_blocks property less than or equal to recordsize. +# + +verify_runnable "global" + +claim="Can set special_small_blocks property less than or equal to recordsize" + +log_assert $claim +log_onexit cleanup +log_must disk_setup + +for size in 8192 32768 131072 524288 1048576 +do + let smaller=$size/2 + log_must zpool create -O recordsize=$size \ + -O special_small_blocks=$smaller \ + $TESTPOOL raidz $ZPOOL_DISKS special mirror \ + $CLASS_DISK0 $CLASS_DISK1 + log_must zpool destroy -f "$TESTPOOL" + + log_must zpool create -O recordsize=$size \ + -O special_small_blocks=$size \ + $TESTPOOL raidz $ZPOOL_DISKS special mirror \ + $CLASS_DISK0 $CLASS_DISK1 + log_must zpool destroy -f "$TESTPOOL" +done + +log_pass $claim diff --git a/tests/zfs-tests/tests/functional/cli_root/zfs_load-key/zfs_load-key_common.kshlib b/tests/zfs-tests/tests/functional/cli_root/zfs_load-key/zfs_load-key_common.kshlib index f7461437c615..3a8cba7b0f16 100644 --- a/tests/zfs-tests/tests/functional/cli_root/zfs_load-key/zfs_load-key_common.kshlib +++ b/tests/zfs-tests/tests/functional/cli_root/zfs_load-key/zfs_load-key_common.kshlib @@ -124,7 +124,10 @@ if not httpd: with open('$HTTPS_PORT_FILE', 'w') as portf: print(port, file=portf) -httpd.socket = ssl.wrap_socket(httpd.socket, server_side=True, keyfile='/$TESTPOOL/snakeoil.key', certfile='$SSL_CA_CERT_FILE', ssl_version=ssl.PROTOCOL_TLS) +sslctx = ssl.SSLContext(ssl.PROTOCOL_TLS_SERVER) +sslctx.check_hostname = False +sslctx.load_cert_chain(certfile='$SSL_CA_CERT_FILE', keyfile='/$TESTPOOL/snakeoil.key') +httpd.socket = httpd.socket = sslctx.wrap_socket(httpd.socket, server_side=True) os.chdir('$STF_SUITE/tests/functional/cli_root/zfs_load-key') diff --git a/tests/zfs-tests/tests/functional/cli_root/zfs_share/Makefile.am b/tests/zfs-tests/tests/functional/cli_root/zfs_share/Makefile.am index bf33ed038d78..35332f822e6c 100644 --- a/tests/zfs-tests/tests/functional/cli_root/zfs_share/Makefile.am +++ b/tests/zfs-tests/tests/functional/cli_root/zfs_share/Makefile.am @@ -14,6 +14,7 @@ dist_pkgdata_SCRIPTS = \ zfs_share_010_neg.ksh \ zfs_share_011_pos.ksh \ zfs_share_012_pos.ksh \ + zfs_share_013_pos.ksh \ zfs_share_concurrent_shares.ksh dist_pkgdata_DATA = \ diff --git a/tests/zfs-tests/tests/functional/cli_root/zfs_share/zfs_share_007_neg.ksh b/tests/zfs-tests/tests/functional/cli_root/zfs_share/zfs_share_007_neg.ksh index 29ca9a143a27..c64157cee601 100755 --- a/tests/zfs-tests/tests/functional/cli_root/zfs_share/zfs_share_007_neg.ksh +++ b/tests/zfs-tests/tests/functional/cli_root/zfs_share/zfs_share_007_neg.ksh @@ -51,7 +51,7 @@ function cleanup { set -A badopts \ "r0" "r0=machine1" "r0=machine1:machine2" \ - "-g" "-b" "-c" "-d" "--invalid" \ + "-g" "-b" "-c" "-d" "--invalid" "rw=[::1]a:[::2]" "rw=[::1" \ "$TESTPOOL" "$TESTPOOL/$TESTFS" "$TESTPOOL\$TESTCTR\$TESTFS1" log_assert "Verify that invalid share parameters and options are caught." diff --git a/tests/zfs-tests/tests/functional/cli_root/zfs_share/zfs_share_013_pos.ksh b/tests/zfs-tests/tests/functional/cli_root/zfs_share/zfs_share_013_pos.ksh new file mode 100755 index 000000000000..150eddac0ebb --- /dev/null +++ b/tests/zfs-tests/tests/functional/cli_root/zfs_share/zfs_share_013_pos.ksh @@ -0,0 +1,80 @@ +#!/bin/ksh -p +# +# CDDL HEADER START +# +# The contents of this file are subject to the terms of the +# Common Development and Distribution License (the "License"). +# You may not use this file except in compliance with the License. +# +# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE +# or http://www.opensolaris.org/os/licensing. +# See the License for the specific language governing permissions +# and limitations under the License. +# +# When distributing Covered Code, include this CDDL HEADER in each +# file and include the License file at usr/src/OPENSOLARIS.LICENSE. +# If applicable, add the following below this CDDL HEADER, with the +# fields enclosed by brackets "[]" replaced with your own identifying +# information: Portions Copyright [yyyy] [name of copyright owner] +# +# CDDL HEADER END +# + +# +# Copyright (c) 2020, Felix Dörre +# + +. $STF_SUITE/include/libtest.shlib + +# +# DESCRIPTION: +# Verify that NFS share options including ipv6 literals are parsed and propagated correctly. +# + +verify_runnable "global" + +function cleanup +{ + log_must zfs set sharenfs=off $TESTPOOL/$TESTFS + is_shared $TESTPOOL/$TESTFS && \ + log_must unshare_fs $TESTPOOL/$TESTFS +} + +log_onexit cleanup + +cleanup + +log_must zfs set sharenfs="rw=[::1]" $TESTPOOL/$TESTFS +output=$(showshares_nfs 2>&1) +log_must grep "::1(" <<< "$output" > /dev/null + +log_must zfs set sharenfs="rw=[2::3]" $TESTPOOL/$TESTFS +output=$(showshares_nfs 2>&1) +log_must grep "2::3(" <<< "$output" > /dev/null + +log_must zfs set sharenfs="rw=[::1]:[2::3]" $TESTPOOL/$TESTFS +output=$(showshares_nfs 2>&1) +log_must grep "::1(" <<< "$output" > /dev/null +log_must grep "2::3(" <<< "$output" > /dev/null + +log_must zfs set sharenfs="rw=[::1]/64" $TESTPOOL/$TESTFS +output=$(showshares_nfs 2>&1) +log_must grep "::1/64(" <<< "$output" > /dev/null + +log_must zfs set sharenfs="rw=[2::3]/128" $TESTPOOL/$TESTFS +output=$(showshares_nfs 2>&1) +log_must grep "2::3/128(" <<< "$output" > /dev/null + +log_must zfs set sharenfs="rw=[::1]/32:[2::3]/128" $TESTPOOL/$TESTFS +output=$(showshares_nfs 2>&1) +log_must grep "::1/32(" <<< "$output" > /dev/null +log_must grep "2::3/128(" <<< "$output" > /dev/null + +log_must zfs set sharenfs="rw=[::1]:[2::3]/64:[2a01:1234:1234:1234:aa34:234:1234:1234]:1.2.3.4/24" $TESTPOOL/$TESTFS +output=$(showshares_nfs 2>&1) +log_must grep "::1(" <<< "$output" > /dev/null +log_must grep "2::3/64(" <<< "$output" > /dev/null +log_must grep "2a01:1234:1234:1234:aa34:234:1234:1234(" <<< "$output" > /dev/null +log_must grep "1\\.2\\.3\\.4/24(" <<< "$output" > /dev/null + +log_pass "NFS share ip address propagated correctly." diff --git a/tests/zfs-tests/tests/functional/cli_root/zfs_share/zfs_share_concurrent_shares.ksh b/tests/zfs-tests/tests/functional/cli_root/zfs_share/zfs_share_concurrent_shares.ksh index dbaaf39b65d4..7354a2b5a355 100755 --- a/tests/zfs-tests/tests/functional/cli_root/zfs_share/zfs_share_concurrent_shares.ksh +++ b/tests/zfs-tests/tests/functional/cli_root/zfs_share/zfs_share_concurrent_shares.ksh @@ -47,31 +47,20 @@ verify_runnable "global" function cleanup { wait - for fs in $(seq 0 50) + for fs in {0..50} do - log_must zfs set sharenfs=off $TESTPOOL/$TESTFS1/$fs - log_must zfs set sharenfs=off $TESTPOOL/$TESTFS2/$fs - log_must zfs set sharenfs=off $TESTPOOL/$TESTFS3/$fs - unshare_fs $TESTPOOL/$TESTFS1/$fs - unshare_fs $TESTPOOL/$TESTFS2/$fs - unshare_fs $TESTPOOL/$TESTFS3/$fs - - if mounted $TESTPOOL/$TESTFS1/$fs; then - log_must zfs unmount $TESTPOOL/$TESTFS1/$fs - fi - if mounted $TESTPOOL/$TESTFS2/$fs; then - log_must zfs unmount $TESTPOOL/$TESTFS2/$fs - fi - if mounted $TESTPOOL/$TESTFS3/$fs; then - log_must zfs unmount $TESTPOOL/$TESTFS3/$fs - fi - - datasetexists $TESTPOOL/$TESTFS1/$fs && \ - destroy_dataset $TESTPOOL/$TESTFS1/$fs -f - datasetexists $TESTPOOL/$TESTFS2/$fs && \ - destroy_dataset $TESTPOOL/$TESTFS2/$fs -f - datasetexists $TESTPOOL/$TESTFS3/$fs && \ - destroy_dataset $TESTPOOL/$TESTFS3/$fs -f + for pfs in $TESTFS1 $TESTFS2 $TESTFS3 + do + log_must zfs set sharenfs=off $TESTPOOL/$pfs/$fs + unshare_fs $TESTPOOL/$pfs/$fs + + if mounted $TESTPOOL/$pfs/$fs; then + log_must zfs unmount $TESTPOOL/$pfs/$fs + fi + + datasetexists $TESTPOOL/$pfs/$fs && \ + destroy_dataset $TESTPOOL/$pfs/$fs -f + done done log_must zfs share -a @@ -79,7 +68,7 @@ function cleanup function create_filesystems { - for fs in $(seq 0 50) + for fs in {0..50} do log_must zfs create -p $TESTPOOL/$TESTFS1/$fs log_must zfs create -p $TESTPOOL/$TESTFS2/$fs @@ -87,6 +76,12 @@ function create_filesystems done } +function sub_fail +{ + log_note $$: "$@" + exit 1 +} + # # Main test routine. # @@ -99,32 +94,74 @@ function test_share # filesystem typeset mntp=$(get_prop mountpoint $filesystem) not_shared $mntp || \ - log_fail "File system $filesystem is already shared." + sub_fail "File system $filesystem is already shared." zfs set sharenfs=on $filesystem || \ - log_fail "zfs set sharenfs=on $filesystem failed." - is_shared $mntp || \ - log_fail "File system $filesystem is not shared (set sharenfs)." + sub_fail "zfs set sharenfs=on $filesystem failed." + + # + # Verify 'zfs share' results in a shared mount. We check this + # multiple times because of Fedora 37+ it's been observed in + # the CI that the share may not be immediately reported. + # + for retry in $(seq 1 10); do + is_shared $mntp && break + + log_note "Wait $retry / 10 for is_shared $mntp (set sharenfs)" + + if [[ $retry -eq 10 ]]; then + sub_fail "File system $filesystem is not shared (set sharenfs)." + fi + + sleep 1 + done # - # Verify 'zfs share' works as well. + # Verify 'zfs unshare' works as well. # zfs unshare $filesystem || \ - log_fail "zfs unshare $filesystem failed." + sub_fail "zfs unshare $filesystem failed." is_shared $mntp && \ - log_fail "File system $filesystem is still shared." + sub_fail "File system $filesystem is still shared." + zfs share $filesystem || \ - log_fail "zfs share $filesystem failed." - is_shared $mntp || \ - log_fail "file system $filesystem is not shared (zfs share)." + sub_fail "zfs share $filesystem failed." + + # + # Verify 'zfs share' results in a shared mount. We check this + # multiple times because of Fedora 37+ it's been observed in + # the CI that the share may not be immediately reported. + # + for retry in $(seq 1 10); do + is_shared $mntp && break + + log_note "Wait $retry / 10 for is_shared $mntp (zfs share)" + + if [[ $retry -eq 10 ]]; then + sub_fail "File system $filesystem is not shared (zfs share)." + fi + + sleep 1 + done #log_note "Sharing a shared file system fails." zfs share $filesystem && \ - log_fail "zfs share $filesystem did not fail" + sub_fail "zfs share $filesystem did not fail" + return 0 } +function unshare_fs_nolog +{ + typeset fs=$1 + + if is_shared $fs || is_shared_smb $fs; then + zfs unshare $fs || + sub_fail "zfs unshare $fs: $?" + fi +} + # # Set the main process id so that we know to capture # failures from child processes and allow the parent process @@ -137,20 +174,16 @@ log_onexit cleanup create_filesystems child_pids=() -for fs in $(seq 0 50) +for fs in {0..50} do - test_share $TESTPOOL/$TESTFS1/$fs & - child_pids+=($!) - log_note "$TESTPOOL/$TESTFS1/$fs ==> $!" - test_share $TESTPOOL/$TESTFS2/$fs & - child_pids+=($!) - log_note "$TESTPOOL/$TESTFS2/$fs ==> $!" - test_share $TESTPOOL/$TESTFS3/$fs & - child_pids+=($!) - log_note "$TESTPOOL/$TESTFS3/$fs ==> $!" + for pfs in $TESTFS1 $TESTFS2 $TESTFS3 + do + test_share $TESTPOOL/$pfs/$fs & + child_pids+=($!) + log_note "$TESTPOOL/$pfs/$fs ==> $!" + done done -wait_for_children "${child_pids[@]}" || - log_fail "multithreaded share test failed" +log_must wait_for_children "${child_pids[@]}" log_note "Verify 'zfs share -a' succeeds." @@ -158,17 +191,16 @@ log_note "Verify 'zfs share -a' succeeds." # Unshare each of the file systems. # child_pids=() -for fs in $(seq 0 50) +for fs in {0..50} do - unshare_fs $TESTPOOL/$TESTFS1/$fs & - child_pids+=($!) - unshare_fs $TESTPOOL/$TESTFS2/$fs & - child_pids+=($!) - unshare_fs $TESTPOOL/$TESTFS3/$fs & - child_pids+=($!) + for pfs in $TESTFS1 $TESTFS2 $TESTFS3 + do + unshare_fs_nolog $TESTPOOL/$pfs/$fs & + child_pids+=($!) + log_note "$TESTPOOL/$pfs/$fs (unshare) ==> $!" + done done -wait_for_children "${child_pids[@]}" || - log_fail "multithreaded unshare failed" +log_must wait_for_children "${child_pids[@]}" # # Try a zfs share -a and verify all file systems are shared. @@ -181,21 +213,13 @@ log_must zfs share -a # unset __ZFS_POOL_EXCLUDE -for fs in $(seq 0 50) +for fs in {0..50} do - is_shared $TESTPOOL/$TESTFS1/$fs || \ - log_fail "File system $TESTPOOL/$TESTFS1/$fs is not shared" - is_shared $TESTPOOL/$TESTFS2/$fs || \ - log_fail "File system $TESTPOOL/$TESTFS2/$fs is not shared" - is_shared $TESTPOOL/$TESTFS3/$fs || \ - log_fail "File system $TESTPOOL/$TESTFS3/$fs is not shared" - - is_exported $TESTPOOL/$TESTFS1/$fs || \ - log_fail "File system $TESTPOOL/$TESTFS1/$fs is not exported" - is_exported $TESTPOOL/$TESTFS2/$fs || \ - log_fail "File system $TESTPOOL/$TESTFS2/$fs is not exported" - is_exported $TESTPOOL/$TESTFS3/$fs || \ - log_fail "File system $TESTPOOL/$TESTFS3/$fs is not exported" + for pfs in $TESTFS1 $TESTFS2 $TESTFS3 + do + log_must is_shared $TESTPOOL/$pfs/$fs + log_must is_exported $TESTPOOL/$pfs/$fs + done done -log_pass "'zfs share [ -a ] <filesystem>' succeeds as root." +log_pass "'zfs share [-a] <filesystem>' succeeds as root." diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_import/Makefile.am b/tests/zfs-tests/tests/functional/cli_root/zpool_import/Makefile.am index a8c9a31dcfdc..4230ec5575f6 100644 --- a/tests/zfs-tests/tests/functional/cli_root/zpool_import/Makefile.am +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_import/Makefile.am @@ -12,6 +12,7 @@ dist_pkgdata_SCRIPTS = \ import_cachefile_paths_changed.ksh \ import_cachefile_shared_device.ksh \ import_devices_missing.ksh \ + import_log_missing.ksh \ import_paths_changed.ksh \ import_rewind_config_changed.ksh \ import_rewind_device_replaced.ksh \ diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_log_missing.ksh b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_log_missing.ksh new file mode 100755 index 000000000000..f12cac78540f --- /dev/null +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_log_missing.ksh @@ -0,0 +1,75 @@ +#!/bin/ksh -p + +# +# This file and its contents are supplied under the terms of the +# Common Development and Distribution License ("CDDL"), version 1.0. +# You may only use this file in accordance with the terms of version +# 1.0 of the CDDL. +# +# A full copy of the text of the CDDL should have accompanied this +# source. A copy of the CDDL is also available via the Internet at +# http://www.illumos.org/license/CDDL. +# + +. $STF_SUITE/tests/functional/cli_root/zpool_import/zpool_import.kshlib + +# +# DESCRIPTION: +# Import with missing log device should not remove spare/cache. +# +# STRATEGY: +# 1. Create a pool. +# 2. Add spare, cache and log devices to the pool. +# 3. Export the pool. +# 4. Remove the log device. +# 5. Import the pool with -m flag. +# 6. Verify that spare and cache are still present in the pool. +# + +verify_runnable "global" + +log_onexit cleanup + +function test_missing_log +{ + typeset poolcreate="$1" + typeset cachevdev="$2" + typeset sparevdev="$3" + typeset logvdev="$4" + typeset missingvdev="$4" + + log_note "$0: pool '$poolcreate', adding $cachevdev, $sparevdev," \ + "$logvdev then moving away $missingvdev." + + log_must zpool create $TESTPOOL1 $poolcreate + + log_must zpool add $TESTPOOL1 cache $cachevdev spare $sparevdev \ + log $logvdev + + log_must_busy zpool export $TESTPOOL1 + + log_must mv $missingvdev $BACKUP_DEVICE_DIR + + log_must zpool import -m -d $DEVICE_DIR $TESTPOOL1 + + CACHE_PRESENT=$(zpool status -v $TESTPOOL1 | grep $cachevdev) + + SPARE_PRESENT=$(zpool status -v $TESTPOOL1 | grep $sparevdev) + + if [ -z "$CACHE_PRESENT"] || [ -z "SPARE_PRESENT"] + then + log_fail "cache/spare vdev missing after importing with missing" \ + "log device" + fi + + # Cleanup + log_must zpool destroy $TESTPOOL1 + + log_note "" +} + +log_must mkdir -p $BACKUP_DEVICE_DIR + +test_missing_log "$VDEV0" "$VDEV1" "$VDEV2" "$VDEV3" + +log_pass "zpool import succeeded with missing log device" diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_initialize/Makefile.am b/tests/zfs-tests/tests/functional/cli_root/zpool_initialize/Makefile.am index 3968902ec36d..483c1c2f5914 100644 --- a/tests/zfs-tests/tests/functional/cli_root/zpool_initialize/Makefile.am +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_initialize/Makefile.am @@ -10,6 +10,7 @@ dist_pkgdata_SCRIPTS = \ zpool_initialize_start_and_cancel_neg.ksh \ zpool_initialize_start_and_cancel_pos.ksh \ zpool_initialize_suspend_resume.ksh \ + zpool_initialize_uninit.ksh \ zpool_initialize_unsupported_vdevs.ksh \ zpool_initialize_verify_checksums.ksh \ zpool_initialize_verify_initialized.ksh diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_initialize/zpool_initialize_uninit.ksh b/tests/zfs-tests/tests/functional/cli_root/zpool_initialize/zpool_initialize_uninit.ksh new file mode 100755 index 000000000000..17f776cfbc20 --- /dev/null +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_initialize/zpool_initialize_uninit.ksh @@ -0,0 +1,141 @@ +#!/bin/ksh -p +# +# CDDL HEADER START +# +# The contents of this file are subject to the terms of the +# Common Development and Distribution License (the "License"). +# You may not use this file except in compliance with the License. +# +# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE +# or https://opensource.org/licenses/CDDL-1.0. +# See the License for the specific language governing permissions +# and limitations under the License. +# +# When distributing Covered Code, include this CDDL HEADER in each +# file and include the License file at usr/src/OPENSOLARIS.LICENSE. +# If applicable, add the following below this CDDL HEADER, with the +# fields enclosed by brackets "[]" replaced with your own identifying +# information: Portions Copyright [yyyy] [name of copyright owner] +# +# CDDL HEADER END +# + +# +# Copyright (c) 2016 by Delphix. All rights reserved. +# Copyright (C) 2023 Lawrence Livermore National Security, LLC. +# +. $STF_SUITE/include/libtest.shlib +. $STF_SUITE/tests/functional/cli_root/zpool_initialize/zpool_initialize.kshlib + +# +# DESCRIPTION: +# Starting, stopping, uninitializing, and restart an initialize works. +# +# STRATEGY: +# 1. Create a one-disk pool. +# 2. Verify uninitialize succeeds for uninitialized pool. +# 3. Verify pool wide cancel|suspend + uninit +# a. Start initializing and verify that initializing is active. +# b. Verify uninitialize fails when actively initializing. +# c. Cancel or suspend initializing and verify that initializing is not active. +# d. Verify uninitialize succeeds after being cancelled. +# 4. Verify per-disk cancel|suspend + uninit +# + +DISK1="$(echo $DISKS | cut -d' ' -f1)" +DISK2="$(echo $DISKS | cut -d' ' -f2)" +DISK3="$(echo $DISKS | cut -d' ' -f3)" + +function status_check # pool disk1-state disk2-state disk3-state +{ + typeset pool="$1" + typeset disk1_state="$2" + typeset disk2_state="$3" + typeset disk3_state="$4" + + state=$(zpool status -i "$pool" | grep "$DISK1" | grep "$disk1_state") + if [[ -z "$state" ]]; then + log_fail "DISK1 state; expected='$disk1_state' got '$state'" + fi + + state=$(zpool status -i "$pool" | grep "$DISK2" | grep "$disk2_state") + if [[ -z "$state" ]]; then + log_fail "DISK2 state; expected='$disk2_state' got '$state'" + fi + + state=$(zpool status -i "$pool" | grep "$DISK3" | grep "$disk3_state") + if [[ -z "$state" ]]; then + log_fail "DISK3 state; expected='$disk3_state' got '$state'" + fi +} + +function status_check_all # pool disk-state +{ + typeset pool="$1" + typeset disk_state="$2" + + status_check "$pool" "$disk_state" "$disk_state" "$disk_state" +} + +# 1. Create a one-disk pool. +log_must zpool create -f $TESTPOOL $DISK1 $DISK2 $DISK3 +status_check_all $TESTPOOL "uninitialized" + +# 2. Verify uninitialize succeeds for uninitialized pool. +log_must zpool initialize -u $TESTPOOL +status_check_all $TESTPOOL "uninitialized" + +# 3. Verify pool wide cancel + uninit +log_must zpool initialize $TESTPOOL +status_check_all $TESTPOOL "[[:digit:]]* initialized" + +log_mustnot zpool initialize -u $TESTPOOL +status_check_all $TESTPOOL "[[:digit:]]* initialized" + +log_must zpool initialize -c $TESTPOOL +status_check_all $TESTPOOL "uninitialized" + +log_must zpool initialize -u $TESTPOOL +status_check_all $TESTPOOL "uninitialized" + +# 3. Verify pool wide suspend + uninit +log_must zpool initialize $TESTPOOL +status_check_all $TESTPOOL "[[:digit:]]* initialized" + +log_mustnot zpool initialize -u $TESTPOOL +status_check_all $TESTPOOL "[[:digit:]]* initialized" + +log_must zpool initialize -s $TESTPOOL +status_check_all $TESTPOOL "suspended" + +log_must zpool initialize -u $TESTPOOL +status_check_all $TESTPOOL "uninitialized" + +# 4. Verify per-disk cancel|suspend + uninit +log_must zpool initialize $TESTPOOL +status_check_all $TESTPOOL "[[:digit:]]* initialized" + +log_must zpool initialize -c $TESTPOOL $DISK1 +log_must zpool initialize -s $TESTPOOL $DISK2 +log_mustnot zpool initialize -u $TESTPOOL $DISK3 +status_check $TESTPOOL "uninitialized" "suspended" "[[:digit:]]* initialized" + +log_must zpool initialize -u $TESTPOOL $DISK1 +status_check $TESTPOOL "uninitialized" "suspended" "[[:digit:]]* initialized" + +log_must zpool initialize -u $TESTPOOL $DISK2 +status_check $TESTPOOL "uninitialized" "uninitialized" "[[:digit:]]* initialized" + +log_must zpool initialize $TESTPOOL $DISK1 +status_check $TESTPOOL "[[:digit:]]* initialized" "uninitialized" "[[:digit:]]* initialized" + +log_must zpool initialize $TESTPOOL $DISK2 +status_check_all $TESTPOOL "[[:digit:]]* initialized" + +log_must zpool initialize -s $TESTPOOL +status_check_all $TESTPOOL "suspended" + +log_must zpool initialize -u $TESTPOOL $DISK1 $DISK2 $DISK3 +status_check_all $TESTPOOL "uninitialized" + +log_pass "Initialize start + cancel/suspend + uninit + start works" diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_resilver/Makefile.am b/tests/zfs-tests/tests/functional/cli_root/zpool_resilver/Makefile.am index 2cec5335fbae..7ca9e81c1867 100644 --- a/tests/zfs-tests/tests/functional/cli_root/zpool_resilver/Makefile.am +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_resilver/Makefile.am @@ -3,7 +3,8 @@ dist_pkgdata_SCRIPTS = \ setup.ksh \ cleanup.ksh \ zpool_resilver_bad_args.ksh \ - zpool_resilver_restart.ksh + zpool_resilver_restart.ksh \ + zpool_resilver_concurrent.ksh dist_pkgdata_DATA = \ zpool_resilver.cfg diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_resilver/zpool_resilver_concurrent.ksh b/tests/zfs-tests/tests/functional/cli_root/zpool_resilver/zpool_resilver_concurrent.ksh new file mode 100755 index 000000000000..4c3b09796869 --- /dev/null +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_resilver/zpool_resilver_concurrent.ksh @@ -0,0 +1,101 @@ +#!/bin/ksh -p +# +# CDDL HEADER START +# +# The contents of this file are subject to the terms of the +# Common Development and Distribution License (the "License"). +# You may not use this file except in compliance with the License. +# +# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE +# or http://www.opensolaris.org/os/licensing. +# See the License for the specific language governing permissions +# and limitations under the License. +# +# When distributing Covered Code, include this CDDL HEADER in each +# file and include the License file at usr/src/OPENSOLARIS.LICENSE. +# If applicable, add the following below this CDDL HEADER, with the +# fields enclosed by brackets "[]" replaced with your own identifying +# information: Portions Copyright [yyyy] [name of copyright owner] +# +# CDDL HEADER END +# + +# +# Copyright (c) 2023 Hewlett Packard Enterprise Development LP. +# + +. $STF_SUITE/include/libtest.shlib +. $STF_SUITE/tests/functional/redundancy/redundancy.kshlib + +# +# DESCRIPTION: +# Verify 'zpool clear' doesn't cause concurrent resilvers +# +# STRATEGY: +# 1. Create N(10) virtual disk files. +# 2. Create draid pool based on the virtual disk files. +# 3. Fill the filesystem with directories and files. +# 4. Force-fault 2 vdevs and verify distributed spare is kicked in. +# 5. Free the distributed spare by replacing the faulty drive. +# 6. Run zpool clear and verify that it does not initiate 2 resilvers +# concurrently while distributed spare gets kicked in. +# + +verify_runnable "global" + +typeset -ir devs=10 +typeset -ir nparity=1 +typeset -ir ndata=8 +typeset -ir dspare=1 + +function cleanup +{ + poolexists "$TESTPOOL" && destroy_pool "$TESTPOOL" + + for i in {0..$devs}; do + log_must rm -f "$BASEDIR/vdev$i" + done + + for dir in $BASEDIR; do + if [[ -d $dir ]]; then + log_must rm -rf $dir + fi + done + + zed_stop + zed_cleanup +} + +log_assert "Verify zpool clear on draid pool doesn't cause concurrent resilvers" +log_onexit cleanup + +setup_test_env $TESTPOOL draid${nparity}:${ndata}d:${dspare}s $devs + +# ZED needed for sequential resilver +zed_setup +log_must zed_start + +log_must zpool offline -f $TESTPOOL $BASEDIR/vdev5 +log_must wait_vdev_state $TESTPOOL draid1-0-0 "ONLINE" 60 +log_must zpool wait -t resilver $TESTPOOL +log_must zpool offline -f $TESTPOOL $BASEDIR/vdev6 + +log_must zpool labelclear -f $BASEDIR/vdev5 +log_must zpool labelclear -f $BASEDIR/vdev6 + +log_must zpool replace -w $TESTPOOL $BASEDIR/vdev5 +sync_pool $TESTPOOL + +log_must zpool events -c +log_must zpool clear $TESTPOOL +log_must wait_vdev_state $TESTPOOL draid1-0-0 "ONLINE" 60 +log_must zpool wait -t resilver $TESTPOOL +log_must zpool wait -t scrub $TESTPOOL + +nof_resilver=$(zpool events | grep -c resilver_start) +if [ $nof_resilver = 1 ] ; then + log_must verify_pool $TESTPOOL + log_pass "zpool clear on draid pool doesn't cause concurrent resilvers" +else + log_fail "FAIL: sequential and healing resilver initiated concurrently" +fi diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_status/Makefile.am b/tests/zfs-tests/tests/functional/cli_root/zpool_status/Makefile.am index 5553061c67b3..8dff69381a4e 100644 --- a/tests/zfs-tests/tests/functional/cli_root/zpool_status/Makefile.am +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_status/Makefile.am @@ -4,4 +4,5 @@ dist_pkgdata_SCRIPTS = \ cleanup.ksh \ zpool_status_001_pos.ksh \ zpool_status_002_pos.ksh \ + zpool_status_008_pos.ksh \ zpool_status_features_001_pos.ksh diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_status/zpool_status_002_pos.ksh b/tests/zfs-tests/tests/functional/cli_root/zpool_status/zpool_status_002_pos.ksh index e2751b112597..976906bce0c9 100755 --- a/tests/zfs-tests/tests/functional/cli_root/zpool_status/zpool_status_002_pos.ksh +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_status/zpool_status_002_pos.ksh @@ -51,7 +51,7 @@ else fi set -A args "" "-x" "-v" "-x $testpool" "-v $testpool" "-xv $testpool" \ - "-vx $testpool" + "-vx $testpool" "-e $testpool" "-es $testpool" log_assert "Executing 'zpool status' with correct options succeeds" @@ -64,4 +64,6 @@ while [[ $i -lt ${#args[*]} ]]; do (( i = i + 1 )) done +cleanup + log_pass "'zpool status' with correct options succeeded" diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_status/zpool_status_008_pos.ksh b/tests/zfs-tests/tests/functional/cli_root/zpool_status/zpool_status_008_pos.ksh new file mode 100755 index 000000000000..6be2ad5a7410 --- /dev/null +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_status/zpool_status_008_pos.ksh @@ -0,0 +1,104 @@ +#!/bin/ksh -p + +# +# CDDL HEADER START +# +# This file and its contents are supplied under the terms of the +# Common Development and Distribution License ("CDDL"), version 1.0. +# You may only use this file in accordance with the terms of version +# 1.0 of the CDDL. +# +# A full copy of the text of the CDDL should have accompanied this +# source. A copy of the CDDL is also available via the Internet at +# http://www.illumos.org/license/CDDL. +# +# CDDL HEADER END +# + +# +# Copyright (c) 2024 by Lawrence Livermore National Security, LLC. +# + +. $STF_SUITE/include/libtest.shlib + +# +# DESCRIPTION: +# Verify 'zpool status -e' only shows unhealthy devices. +# +# STRATEGY: +# 1. Create zpool +# 2. Force DEGRADE, FAULT, or inject slow IOs for vdevs +# 3. Verify vdevs are reported correctly with -e and -s +# 4. Verify parents are reported as DEGRADED +# 5. Verify healthy children are not reported +# + +function cleanup +{ + log_must set_tunable64 ZIO_SLOW_IO_MS $OLD_SLOW_IO + zinject -c all + poolexists $TESTPOOL2 && destroy_pool $TESTPOOL2 + log_must rm -f $all_vdevs +} + +log_assert "Verify 'zpool status -e'" + +log_onexit cleanup + +all_vdevs=$(echo $TESTDIR/vdev{1..6}) +log_must mkdir -p $TESTDIR +log_must truncate -s $MINVDEVSIZE $all_vdevs + +OLD_SLOW_IO=$(get_tunable ZIO_SLOW_IO_MS) + +for raid_type in "draid2:3d:6c:1s" "raidz2"; do + + log_must zpool create -f $TESTPOOL2 $raid_type $all_vdevs + + # Check DEGRADED vdevs are shown. + log_must check_vdev_state $TESTPOOL2 $TESTDIR/vdev4 "ONLINE" + log_must zinject -d $TESTDIR/vdev4 -A degrade $TESTPOOL2 + log_must eval "zpool status -e $TESTPOOL2 | grep $TESTDIR/vdev4 | grep DEGRADED" + + # Check FAULTED vdevs are shown. + log_must check_vdev_state $TESTPOOL2 $TESTDIR/vdev5 "ONLINE" + log_must zinject -d $TESTDIR/vdev5 -A fault $TESTPOOL2 + log_must eval "zpool status -e $TESTPOOL2 | grep $TESTDIR/vdev5 | grep FAULTED" + + # Check no ONLINE vdevs are shown + log_mustnot eval "zpool status -e $TESTPOOL2 | grep ONLINE" + + # Check no ONLINE slow vdevs are show. Then mark IOs greater than + # 10ms slow, delay IOs 20ms to vdev6, check slow IOs. + log_must check_vdev_state $TESTPOOL2 $TESTDIR/vdev6 "ONLINE" + log_mustnot eval "zpool status -es $TESTPOOL2 | grep ONLINE" + + log_must set_tunable64 ZIO_SLOW_IO_MS 10 + log_must zinject -d $TESTDIR/vdev6 -D20:100 $TESTPOOL2 + log_must mkfile 1048576 /$TESTPOOL2/testfile + sync_pool $TESTPOOL2 + log_must set_tunable64 ZIO_SLOW_IO_MS $OLD_SLOW_IO + + # Check vdev6 slow IOs are only shown when requested with -s. + log_mustnot eval "zpool status -e $TESTPOOL2 | grep $TESTDIR/vdev6 | grep ONLINE" + log_must eval "zpool status -es $TESTPOOL2 | grep $TESTDIR/vdev6 | grep ONLINE" + + # Pool level and top-vdev level status must be DEGRADED. + log_must eval "zpool status -e $TESTPOOL2 | grep $TESTPOOL2 | grep DEGRADED" + log_must eval "zpool status -e $TESTPOOL2 | grep $raid_type | grep DEGRADED" + + # Check that healthy vdevs[1-3] aren't shown with -e. + log_must check_vdev_state $TESTPOOL2 $TESTDIR/vdev1 "ONLINE" + log_must check_vdev_state $TESTPOOL2 $TESTDIR/vdev2 "ONLINE" + log_must check_vdev_state $TESTPOOL2 $TESTDIR/vdev3 "ONLINE" + log_mustnot eval "zpool status -es $TESTPOOL2 | grep $TESTDIR/vdev1 | grep ONLINE" + log_mustnot eval "zpool status -es $TESTPOOL2 | grep $TESTDIR/vdev2 | grep ONLINE" + log_mustnot eval "zpool status -es $TESTPOOL2 | grep $TESTDIR/vdev3 | grep ONLINE" + + log_must zinject -c all + log_must zpool status -es $TESTPOOL2 + + zpool destroy $TESTPOOL2 +done + +log_pass "Verify zpool status -e shows only unhealthy vdevs" diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_trim/zpool_trim_start_and_cancel_pos.ksh b/tests/zfs-tests/tests/functional/cli_root/zpool_trim/zpool_trim_start_and_cancel_pos.ksh index fbb0c291046c..19781137d3ca 100755 --- a/tests/zfs-tests/tests/functional/cli_root/zpool_trim/zpool_trim_start_and_cancel_pos.ksh +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_trim/zpool_trim_start_and_cancel_pos.ksh @@ -35,7 +35,7 @@ DISK1=${DISKS%% *} log_must zpool create -f $TESTPOOL $DISK1 -log_must zpool trim $TESTPOOL +log_must zpool trim -r 1 $TESTPOOL [[ -z "$(trim_progress $TESTPOOL $DISK1)" ]] && \ log_fail "TRIM did not start" diff --git a/tests/zfs-tests/tests/functional/cp_files/.gitignore b/tests/zfs-tests/tests/functional/cp_files/.gitignore index eac05e155378..1f29418c0175 100644 --- a/tests/zfs-tests/tests/functional/cp_files/.gitignore +++ b/tests/zfs-tests/tests/functional/cp_files/.gitignore @@ -1 +1,2 @@ -/cp_files +cp_files +seekflood diff --git a/tests/zfs-tests/tests/functional/cp_files/Makefile.am b/tests/zfs-tests/tests/functional/cp_files/Makefile.am index 06c31f5f3f92..f01f1c3402a2 100644 --- a/tests/zfs-tests/tests/functional/cp_files/Makefile.am +++ b/tests/zfs-tests/tests/functional/cp_files/Makefile.am @@ -4,10 +4,12 @@ pkgdatadir = $(datadir)/@PACKAGE@/zfs-tests/tests/functional/cp_files dist_pkgdata_SCRIPTS = \ cp_files_001_pos.ksh \ + cp_stress.ksh \ cleanup.ksh \ setup.ksh pkgexecdir = $(datadir)/@PACKAGE@/zfs-tests/tests/functional/cp_files -pkgexec_PROGRAMS = cp_files +pkgexec_PROGRAMS = cp_files seekflood cp_files_SOURCES= cp_files.c +seekflood_SOURCES = seekflood.c diff --git a/tests/zfs-tests/tests/functional/cp_files/cp_stress.ksh b/tests/zfs-tests/tests/functional/cp_files/cp_stress.ksh new file mode 100755 index 000000000000..43bb8ab572d2 --- /dev/null +++ b/tests/zfs-tests/tests/functional/cp_files/cp_stress.ksh @@ -0,0 +1,73 @@ +#! /bin/ksh -p +# +# CDDL HEADER START +# +# The contents of this file are subject to the terms of the +# Common Development and Distribution License (the "License"). +# You may not use this file except in compliance with the License. +# +# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE +# or https://opensource.org/licenses/CDDL-1.0. +# See the License for the specific language governing permissions +# and limitations under the License. +# +# When distributing Covered Code, include this CDDL HEADER in each +# file and include the License file at usr/src/OPENSOLARIS.LICENSE. +# If applicable, add the following below this CDDL HEADER, with the +# fields enclosed by brackets "[]" replaced with your own identifying +# information: Portions Copyright [yyyy] [name of copyright owner] +# +# CDDL HEADER END + +# +# Copyright (c) 2023 by Lawrence Livermore National Security, LLC. +# + +. $STF_SUITE/include/libtest.shlib + +# +# DESCRIPTION: +# +# https://github.com/openzfs/zfs/issues/15526 identified a dirty dnode +# SEEK_HOLE/SEEK_DATA bug. https://github.com/openzfs/zfs/pull/15571 +# fixed the bug, and was backported to 2.1.14 and 2.2.2. +# +# This test is to ensure that the bug, as understood, will not recur. +# +# STRATEGY: +# +# 1. Run the 'seekflood' binary, for creation of files with timing +# characteristics that can trigger #15526. +# 2. A single run is not always a trigger, so run repeatedly. + +verify_runnable "global" + +function cleanup +{ + rm -rf /$TESTDIR/cp_stress +} + +log_assert "Run the 'seekflood' binary repeatedly to try to trigger #15526" + +log_onexit cleanup + +log_must mkdir /$TESTPOOL/cp_stress + +MYPWD="$PWD" +cd /$TESTPOOL/cp_stress +CPUS=$(get_num_cpus) + +if is_freebsd ; then + # 'seekflood' takes longer on FreeBSD and can timeout the test + RUNS=3 +else + RUNS=10 +fi + +for i in $(seq 1 $RUNS) ; do + # Each run takes around 12 seconds. + log_must $STF_SUITE/tests/functional/cp_files/seekflood 2000 $CPUS +done +cd "$MYPWD" + +log_pass "No corruption detected" diff --git a/tests/zfs-tests/tests/functional/cp_files/seekflood.c b/tests/zfs-tests/tests/functional/cp_files/seekflood.c new file mode 100644 index 000000000000..02c2c8e6eca5 --- /dev/null +++ b/tests/zfs-tests/tests/functional/cp_files/seekflood.c @@ -0,0 +1,180 @@ +/* + * SPDX-License-Identifier: MIT + * + * Copyright (c) 2023, Rob Norris <robn@despairlabs.com> + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + * + */ + +#ifndef _GNU_SOURCE +#define _GNU_SOURCE +#endif + +#include <fcntl.h> +#include <unistd.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <errno.h> +#include <sys/stat.h> +#include <sys/wait.h> + +#define DATASIZE (4096) +char data[DATASIZE]; + +static int +_open_file(int n, int wr) +{ + char buf[256]; + int fd; + + snprintf(buf, sizeof (buf), "testdata_%d_%d", getpid(), n); + + if ((fd = open(buf, wr ? (O_WRONLY | O_CREAT) : O_RDONLY, + wr ? (S_IRUSR | S_IWUSR) : 0)) < 0) { + fprintf(stderr, "Error: open '%s' (%s): %s\n", + buf, wr ? "write" : "read", strerror(errno)); + exit(1); + } + + return (fd); +} + +static void +_write_file(int n, int fd) +{ + /* write a big ball of stuff */ + ssize_t nwr = write(fd, data, DATASIZE); + if (nwr < 0) { + fprintf(stderr, "Error: write '%d_%d': %s\n", + getpid(), n, strerror(errno)); + exit(1); + } else if (nwr < DATASIZE) { + fprintf(stderr, "Error: write '%d_%d': short write\n", getpid(), + n); + exit(1); + } +} + +static int +_seek_file(int n, int fd) +{ + struct stat st; + if (fstat(fd, &st) < 0) { + fprintf(stderr, "Error: fstat '%d_%d': %s\n", getpid(), n, + strerror(errno)); + exit(1); + } + + /* + * A zero-sized file correctly has no data, so seeking the file is + * pointless. + */ + if (st.st_size == 0) + return (0); + + /* size is real, and we only write, so SEEK_DATA must find something */ + if (lseek(fd, 0, SEEK_DATA) < 0) { + if (errno == ENXIO) + return (1); + fprintf(stderr, "Error: lseek '%d_%d': %s\n", + getpid(), n, strerror(errno)); + exit(2); + } + + return (0); +} + +int +main(int argc, char **argv) +{ + int nfiles = 0; + int nthreads = 0; + + if (argc < 3 || (nfiles = atoi(argv[1])) == 0 || + (nthreads = atoi(argv[2])) == 0) { + printf("usage: seekflood <nfiles> <threads>\n"); + exit(1); + } + + memset(data, 0x5a, DATASIZE); + + /* fork off some flood threads */ + for (int i = 0; i < nthreads; i++) { + if (!fork()) { + /* thread main */ + + /* create zero file */ + int fd = _open_file(0, 1); + _write_file(0, fd); + close(fd); + + int count = 0; + + int h = 0, i, j, rfd, wfd; + for (i = 0; i < nfiles; i += 2, h++) { + j = i+1; + + /* seek h, write i */ + rfd = _open_file(h, 0); + wfd = _open_file(i, 1); + count += _seek_file(h, rfd); + _write_file(i, wfd); + close(rfd); + close(wfd); + + /* seek i, write j */ + rfd = _open_file(i, 0); + wfd = _open_file(j, 1); + count += _seek_file(i, rfd); + _write_file(j, wfd); + close(rfd); + close(wfd); + } + + /* return count of failed seeks to parent */ + exit(count < 256 ? count : 255); + } + } + + /* wait for threads, take their seek fail counts from exit code */ + int count = 0, crashed = 0; + for (int i = 0; i < nthreads; i++) { + int wstatus; + wait(&wstatus); + if (WIFEXITED(wstatus)) + count += WEXITSTATUS(wstatus); + else + crashed++; + } + + if (crashed) { + fprintf(stderr, "Error: child crashed; test failed\n"); + exit(1); + } + + if (count) { + fprintf(stderr, "Error: %d seek failures; test failed\n", + count); + exit(1); + } + + exit(0); +} diff --git a/tests/zfs-tests/tests/functional/ctime/ctime.c b/tests/zfs-tests/tests/functional/ctime/ctime.c index d01fa0d4ed3e..01a08dd3c7e7 100644 --- a/tests/zfs-tests/tests/functional/ctime/ctime.c +++ b/tests/zfs-tests/tests/functional/ctime/ctime.c @@ -362,12 +362,20 @@ main(int argc, char *argv[]) return (1); } - if (t1 == t2) { - (void) fprintf(stderr, "%s: t1(%ld) == t2(%ld)\n", + + /* + * Ideally, time change would be exactly two seconds, but allow + * a little slack in case of scheduling delays or similar. + */ + long delta = (long)t2 - (long)t1; + if (delta < 2 || delta > 4) { + (void) fprintf(stderr, + "%s: BAD time change: t1(%ld), t2(%ld)\n", timetest_table[i].name, (long)t1, (long)t2); return (1); } else { - (void) fprintf(stderr, "%s: t1(%ld) != t2(%ld)\n", + (void) fprintf(stderr, + "%s: good time change: t1(%ld), t2(%ld)\n", timetest_table[i].name, (long)t1, (long)t2); } } diff --git a/tests/zfs-tests/tests/functional/io/io_uring.ksh b/tests/zfs-tests/tests/functional/io/io_uring.ksh index 189c11f0d64f..fd8df69676db 100755 --- a/tests/zfs-tests/tests/functional/io/io_uring.ksh +++ b/tests/zfs-tests/tests/functional/io/io_uring.ksh @@ -44,6 +44,13 @@ if ! $(grep -q "CONFIG_IO_URING=y" /boot/config-$(uname -r)); then log_unsupported "Requires io_uring support" fi +if [ -e /etc/os-release ] ; then + source /etc/os-release + if [ -n "$REDHAT_SUPPORT_PRODUCT_VERSION" ] && ((floor($REDHAT_SUPPORT_PRODUCT_VERSION) == 9)) ; then + log_unsupported "Disabled on CentOS 9, fails with 'Operation not permitted'" + fi +fi + fio --ioengine=io_uring --parse-only || log_unsupported "io_uring support required" function cleanup diff --git a/tests/zfs-tests/tests/functional/l2arc/persist_l2arc_001_pos.ksh b/tests/zfs-tests/tests/functional/l2arc/persist_l2arc_001_pos.ksh index 0a9049490c71..8963803f6c0b 100755 --- a/tests/zfs-tests/tests/functional/l2arc/persist_l2arc_001_pos.ksh +++ b/tests/zfs-tests/tests/functional/l2arc/persist_l2arc_001_pos.ksh @@ -27,15 +27,14 @@ # # STRATEGY: # 1. Create pool with a cache device. -# 2. Export and re-import pool without writing any data. -# 3. Create a random file in that pool and random read for 10 sec. -# 4. Export pool. -# 5. Read the amount of log blocks written from the header of the +# 2. Create a random file in that pool and random read for 10 sec. +# 3. Export pool. +# 4. Read the amount of log blocks written from the header of the # L2ARC device. -# 6. Import pool. -# 7. Read the amount of log blocks rebuilt in arcstats and compare to +# 5. Import pool. +# 6. Read the amount of log blocks rebuilt in arcstats and compare to # (5). -# 8. Check if the labels of the L2ARC device are intact. +# 7. Check if the labels of the L2ARC device are intact. # # * We can predict the minimum bytes of L2ARC restored if we subtract # from the effective size of the cache device the bytes l2arc_evict() @@ -75,10 +74,8 @@ export FILE_SIZE=$(( floor($fill_mb / $NUMJOBS) ))M log_must truncate -s ${cache_sz}M $VDEV_CACHE -log_must zpool create -f $TESTPOOL $VDEV cache $VDEV_CACHE - -log_must zpool export $TESTPOOL -log_must zpool import -d $VDIR $TESTPOOL +log_must zpool create -f -o ashift=12 $TESTPOOL $VDEV +log_must zpool add $TESTPOOL cache $VDEV_CACHE log_must fio $FIO_SCRIPTS/mkfiles.fio log_must fio $FIO_SCRIPTS/random_reads.fio diff --git a/tests/zfs-tests/tests/functional/mmap/Makefile.am b/tests/zfs-tests/tests/functional/mmap/Makefile.am index b26791ee7ce0..526405954c08 100644 --- a/tests/zfs-tests/tests/functional/mmap/Makefile.am +++ b/tests/zfs-tests/tests/functional/mmap/Makefile.am @@ -2,10 +2,12 @@ pkgdatadir = $(datadir)/@PACKAGE@/zfs-tests/tests/functional/mmap dist_pkgdata_SCRIPTS = \ setup.ksh \ cleanup.ksh \ + mmap_mixed.ksh \ mmap_read_001_pos.ksh \ mmap_write_001_pos.ksh \ mmap_libaio_001_pos.ksh \ - mmap_seek_001_pos.ksh + mmap_seek_001_pos.ksh \ + mmap_sync_001_pos.ksh dist_pkgdata_DATA = \ mmap.cfg diff --git a/tests/zfs-tests/tests/functional/mmap/mmap_mixed.ksh b/tests/zfs-tests/tests/functional/mmap/mmap_mixed.ksh new file mode 100755 index 000000000000..6c8246d48acf --- /dev/null +++ b/tests/zfs-tests/tests/functional/mmap/mmap_mixed.ksh @@ -0,0 +1,86 @@ +#!/bin/ksh -p +# +# CDDL HEADER START +# +# The contents of this file are subject to the terms of the +# Common Development and Distribution License (the "License"). +# You may not use this file except in compliance with the License. +# +# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE +# or https://opensource.org/licenses/CDDL-1.0. +# See the License for the specific language governing permissions +# and limitations under the License. +# +# When distributing Covered Code, include this CDDL HEADER in each +# file and include the License file at usr/src/OPENSOLARIS.LICENSE. +# If applicable, add the following below this CDDL HEADER, with the +# fields enclosed by brackets "[]" replaced with your own identifying +# information: Portions Copyright [yyyy] [name of copyright owner] +# +# CDDL HEADER END +# + +# +# Copyright (c) 2023 by Lawrence Livermore National Security, LLC. +# + +. $STF_SUITE/include/libtest.shlib +. $STF_SUITE/tests/functional/mmap/mmap.cfg + +# +# DESCRIPTION: +# Verify mixed buffered and mmap IO. +# +# STRATEGY: +# 1. Create an empty file. +# 2. Start a background buffered read/write fio to the file. +# 3. Start a background mmap read/write fio to the file. +# + +verify_runnable "global" + +function cleanup +{ + log_must rm -f "$tmp_file" +} + +log_assert "Verify mixed buffered and mmap IO" + +log_onexit cleanup + +mntpnt=$(get_prop mountpoint $TESTPOOL/$TESTFS) +tmp_file=$mntpnt/file +bs=$((128 * 1024)) +blocks=64 +size=$((bs * blocks)) +runtime=60 + +log_must dd if=/dev/zero of=$tmp_file bs=$bs count=$blocks + +# Buffered IO writes +log_must eval "fio --filename=$tmp_file --name=buffer-write \ + --rw=randwrite --size=$size --bs=$bs --direct=0 --numjobs=1 \ + --ioengine=sync --fallocate=none --group_reporting --minimal \ + --runtime=$runtime --time_based --norandommap &" + +# Buffered IO reads +log_must eval "fio --filename=$tmp_file --name=buffer-read \ + --rw=randread --size=$size --bs=$bs --direct=0 --numjobs=1 \ + --ioengine=sync --fallocate=none --group_reporting --minimal \ + --runtime=$runtime --time_based --norandommap &" + +# mmap IO writes +log_must eval "fio --filename=$tmp_file --name=mmap-write \ + --rw=randwrite --size=$size --bs=$bs --numjobs=1 \ + --ioengine=mmap --fallocate=none --group_reporting --minimal \ + --runtime=$runtime --time_based --norandommap &" + +# mmap IO reads +log_must eval "fio --filename=$tmp_file --name=mmap-read \ + --rw=randread --size=$size --bs=$bs --numjobs=1 \ + --ioengine=mmap --fallocate=none --group_reporting --minimal \ + --runtime=$runtime --time_based --norandommap &" + +log_must wait + +log_pass "Verfied mixed buffered and mmap IO" diff --git a/tests/zfs-tests/tests/functional/mmap/mmap_sync_001_pos.ksh b/tests/zfs-tests/tests/functional/mmap/mmap_sync_001_pos.ksh new file mode 100755 index 000000000000..b764d6607ba6 --- /dev/null +++ b/tests/zfs-tests/tests/functional/mmap/mmap_sync_001_pos.ksh @@ -0,0 +1,63 @@ +#!/bin/ksh -p + +# +# This file and its contents are supplied under the terms of the +# Common Development and Distribution License ("CDDL"), version 1.0. +# You may only use this file in accordance with the terms of version +# 1.0 of the CDDL. +# +# A full copy of the text of the CDDL should have accompanied this +# source. A copy of the CDDL is also available via the Internet at +# http://www.illumos.org/license/CDDL. +# + +# +# Copyright (c) 2015, 2016 by Delphix. All rights reserved. +# + +. $STF_SUITE/include/libtest.shlib + +# +# DESCRIPTION: +# msync()s of mmap()'ed file should complete quickly during +# background dirty page writebacks by the kernel. +# + +function cleanup +{ + log_must eval "echo $saved_vm_dirty_expire_centisecs > /proc/sys/vm/dirty_expire_centisecs" + log_must eval "echo $saved_vm_dirty_background_ratio > /proc/sys/vm/dirty_background_ratio" + log_must eval "echo $saved_vm_dirty_writeback_centisecs > /proc/sys/vm/dirty_writeback_centisecs" + + # revert to some sensible defaults if the values we saved + # were incorrect due to a previous run being interrupted + if [ $(</proc/sys/vm/dirty_expire_centisecs) -eq 1 ]; then + log_must eval "echo 3000 > /proc/sys/vm/dirty_expire_centisecs" + fi + + if [ $(</proc/sys/vm/dirty_background_ratio) -eq 0 ]; then + log_must eval "echo 10 > /proc/sys/vm/dirty_background_ratio" + fi + + if [ $(</proc/sys/vm/dirty_writeback_centisecs) -eq 1 ]; then + log_must eval "echo 500 > /proc/sys/vm/dirty_writeback_centisecs" + fi +} + +if ! is_linux; then + log_unsupported "Only supported on Linux, requires /proc/sys/vm/ tunables" +fi + +log_onexit cleanup +log_assert "Run the tests for mmap_sync" + +read -r saved_vm_dirty_expire_centisecs < /proc/sys/vm/dirty_expire_centisecs +read -r saved_vm_dirty_background_ratio < /proc/sys/vm/dirty_background_ratio +read -r saved_vm_dirty_writeback_centisecs < /proc/sys/vm/dirty_writeback_centisecs + +log_must eval "echo 1 > /proc/sys/vm/dirty_expire_centisecs" +log_must eval "echo 1 > /proc/sys/vm/dirty_background_bytes" +log_must eval "echo 1 > /proc/sys/vm/dirty_writeback_centisecs" + +log_must mmap_sync +log_pass "mmap_sync tests passed." diff --git a/tests/zfs-tests/tests/functional/rsend/Makefile.am b/tests/zfs-tests/tests/functional/rsend/Makefile.am index d80d2124e291..2cedf03d3d69 100644 --- a/tests/zfs-tests/tests/functional/rsend/Makefile.am +++ b/tests/zfs-tests/tests/functional/rsend/Makefile.am @@ -25,6 +25,7 @@ dist_pkgdata_SCRIPTS = \ rsend_022_pos.ksh \ rsend_024_pos.ksh \ send_encrypted_files.ksh \ + send_encrypted_freeobjects.ksh \ send_encrypted_hierarchy.ksh \ send_encrypted_props.ksh \ send_encrypted_truncated_files.ksh \ diff --git a/tests/zfs-tests/tests/functional/rsend/send-c_volume.ksh b/tests/zfs-tests/tests/functional/rsend/send-c_volume.ksh index 988ed91b9918..1bf234823459 100755 --- a/tests/zfs-tests/tests/functional/rsend/send-c_volume.ksh +++ b/tests/zfs-tests/tests/functional/rsend/send-c_volume.ksh @@ -29,6 +29,7 @@ function cleanup { + rm $BACKDIR/copy log_must_busy zfs destroy -r $vol cleanup_pool $POOL2 } @@ -60,7 +61,9 @@ log_must eval "zfs recv -d $POOL2 <$BACKDIR/full" verify_stream_size $BACKDIR/full $vol verify_stream_size $BACKDIR/full $vol2 -md5=$(dd if=$voldev2 bs=1024k count=$megs 2>/dev/null | md5digest) +block_device_wait $voldev2 +log_must dd if=$voldev2 of=$BACKDIR/copy bs=1024k count=$megs +md5=$(md5digest $BACKDIR/copy) [[ $md5 = $md5_1 ]] || log_fail "md5 mismatch: $md5 != $md5_1" # Repeat, for an incremental send @@ -72,7 +75,9 @@ log_must eval "zfs recv -d $POOL2 <$BACKDIR/inc" verify_stream_size $BACKDIR/inc $vol 90 $vol@snap verify_stream_size $BACKDIR/inc $vol2 90 $vol2@snap -md5=$(dd skip=$megs if=$voldev2 bs=1024k count=$megs 2>/dev/null | md5digest) +block_device_wait $voldev2 +log_must dd skip=$megs if=$voldev2 of=$BACKDIR/copy bs=1024k count=$megs +md5=$(md5digest $BACKDIR/copy) [[ $md5 = $md5_2 ]] || log_fail "md5 mismatch: $md5 != $md5_2" log_pass "Verify compressed send works with volumes" diff --git a/tests/zfs-tests/tests/functional/rsend/send_encrypted_freeobjects.ksh b/tests/zfs-tests/tests/functional/rsend/send_encrypted_freeobjects.ksh new file mode 100755 index 000000000000..92451bd1ab6f --- /dev/null +++ b/tests/zfs-tests/tests/functional/rsend/send_encrypted_freeobjects.ksh @@ -0,0 +1,87 @@ +#!/bin/ksh + +# +# This file and its contents are supplied under the terms of the +# Common Development and Distribution License ("CDDL"), version 1.0. +# You may only use this file in accordance with the terms of version +# 1.0 of the CDDL. +# +# A full copy of the text of the CDDL should have accompanied this +# source. A copy of the CDDL is also available via the Internet at +# http://www.illumos.org/license/CDDL. +# + +# +# Copyright (c) 2017 by Lawrence Livermore National Security, LLC. +# Copyright (c) 2023 by Findity AB +# + +. $STF_SUITE/tests/functional/rsend/rsend.kshlib + +# +# Description: +# Verify that receiving a raw encrypted stream, with a FREEOBJECTS +# removing all existing objects in a block followed by an OBJECT write +# to the same block, does not result in a panic. +# +# Strategy: +# 1. Create a new encrypted filesystem +# 2. Create file f1 as the first object in some block (here object 128) +# 3. Take snapshot A +# 4. Create file f2 as the second object in the same block (here object 129) +# 5. Delete f1 +# 6. Take snapshot B +# 7. Receive a full raw encrypted send of A +# 8. Receive an incremental raw send of B +# +verify_runnable "both" + +function create_object_with_num +{ + file=$1 + num=$2 + + tries=100 + for ((i=0; i<$tries; i++)); do + touch $file + onum=$(ls -li $file | awk '{print $1}') + + if [[ $onum -ne $num ]] ; then + rm -f $file + else + break + fi + done + if [[ $i -eq $tries ]]; then + log_fail "Failed to create object with number $num" + fi +} + +log_assert "FREEOBJECTS followed by OBJECT in encrypted stream does not crash" + +sendds=sendencfods +recvds=recvencfods +keyfile=/$POOL/keyencfods +f1=/$POOL/$sendds/f1 +f2=/$POOL/$sendds/f2 + +log_must eval "echo 'password' > $keyfile" + +# +# xattr=sa and dnodesize=legacy for sequential object numbers, see +# note in send_freeobjects.ksh. +# +log_must zfs create -o xattr=sa -o dnodesize=legacy -o encryption=on \ + -o keyformat=passphrase -o keylocation=file://$keyfile $POOL/$sendds + +create_object_with_num $f1 128 +log_must zfs snap $POOL/$sendds@A +create_object_with_num $f2 129 +log_must rm $f1 +log_must zfs snap $POOL/$sendds@B + +log_must eval "zfs send -w $POOL/$sendds@A | zfs recv $POOL/$recvds" +log_must eval "zfs send -w -i $POOL/$sendds@A $POOL/$sendds@B |" \ + "zfs recv $POOL/$recvds" + +log_pass "FREEOBJECTS followed by OBJECT in encrypted stream did not crash" diff --git a/tests/zfs-tests/tests/functional/trim/trim_l2arc.ksh b/tests/zfs-tests/tests/functional/trim/trim_l2arc.ksh index ecf9f3424eb5..04dbf5d2e2a8 100755 --- a/tests/zfs-tests/tests/functional/trim/trim_l2arc.ksh +++ b/tests/zfs-tests/tests/functional/trim/trim_l2arc.ksh @@ -65,7 +65,7 @@ typeset VDEV_MIN_MB=$((MINVDEVSIZE * 0.30 / 1024 / 1024)) log_must zpool create -f $TESTPOOL $TRIM_VDEV1 cache $TRIM_VDEV2 verify_vdevs "-le" "$VDEV_MIN_MB" $TRIM_VDEV2 -typeset fill_mb=$(( floor(2 * MINVDEVSIZE) )) +typeset fill_mb=$(( floor(3 * MINVDEVSIZE) )) export DIRECTORY=/$TESTPOOL export NUMJOBS=1 export FILE_SIZE=${fill_mb} |