From a10f58d076db56056d729664690a563b4627e85d Mon Sep 17 00:00:00 2001 From: Mark Johnston Date: Sun, 27 Oct 2013 16:01:11 +0000 Subject: Convert the lockstat(1) man page to mdoc and make sure that it gets installed. Additionally, remove Solaris-specific sections and references, and replace example outputs with output from lockstat on FreeBSD, since lockstat's output contains stack traces. This change also removes some examples that don't seem to work properly on FreeBSD. The examples should be re-added when lockstat is fixed. Reported by: avg MFC after: 1 week --- cddl/contrib/opensolaris/cmd/lockstat/lockstat.1 | 1178 +++++++--------------- cddl/usr.sbin/lockstat/Makefile | 1 - 2 files changed, 351 insertions(+), 828 deletions(-) diff --git a/cddl/contrib/opensolaris/cmd/lockstat/lockstat.1 b/cddl/contrib/opensolaris/cmd/lockstat/lockstat.1 index f3fc8e6b2d05..b634d45f9340 100644 --- a/cddl/contrib/opensolaris/cmd/lockstat/lockstat.1 +++ b/cddl/contrib/opensolaris/cmd/lockstat/lockstat.1 @@ -2,7 +2,7 @@ .\" CDDL HEADER START .\" .\" The contents of this file are subject to the terms of the -.\" Common Development and Distribution License (the "License"). +.\" 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 @@ -18,858 +18,382 @@ .\" .\" CDDL HEADER END .\" Copyright (c) 2008, Sun Microsystems, Inc. All Rights Reserved. -.TH lockstat 1M "28 Feb 2008" "SunOS 5.11" "System Administration Commands" -.SH NAME -lockstat \- report kernel lock and profiling statistics -.SH SYNOPSIS -.LP -.nf -\fBlockstat\fR [\fB-ACEHI\fR] [\fB-e\fR \fIevent_list\fR] [\fB-i\fR \fIrate\fR] - [\fB-b\fR | \fB-t\fR | \fB-h\fR | \fB-s\fR \fIdepth\fR] [\fB-n\fR \fInrecords\fR] - [\fB-l\fR \fIlock\fR [, \fIsize\fR]] [\fB-d\fR \fIduration\fR] - [\fB-f\fR \fIfunction\fR [, \fIsize\fR]] [\fB-T\fR] [\fB-ckgwWRpP\fR] [\fB-D\fR \fIcount\fR] - [\fB-o\fR \fIfilename\fR] [\fB-x\fR \fIopt\fR [=val]] \fIcommand\fR [\fIargs\fR] -.fi - -.SH DESCRIPTION -.sp -.LP -The \fBlockstat\fR utility gathers and displays kernel locking and profiling statistics. \fBlockstat\fR allows you to specify which events to watch (for example, spin on adaptive mutex, block on read access to rwlock due to waiting writers, and so forth) how much -data to gather for each event, and how to display the data. By default, \fBlockstat\fR monitors all lock contention events, gathers frequency and timing data about those events, and displays the data in decreasing frequency order, so that the most common events appear first. -.sp -.LP -\fBlockstat\fR gathers data until the specified command completes. For example, to gather statistics for a fixed-time interval, use \fBsleep\fR(1) as -the command, as follows: -.sp -.LP -\fBexample#\fR \fBlockstat\fR \fBsleep\fR \fB5\fR -.sp -.LP -When the \fB-I\fR option is specified, \fBlockstat\fR establishes a per-processor high-level periodic interrupt source to gather profiling data. The interrupt handler simply generates a \fBlockstat\fR event whose caller is the interrupted PC (program counter). -The profiling event is just like any other \fBlockstat\fR event, so all of the normal \fBlockstat\fR options are applicable. -.sp -.LP -\fBlockstat\fR relies on DTrace to modify the running kernel's text to intercept events of interest. This imposes a small but measurable overhead on all system activity, so access to \fBlockstat\fR is restricted to super-user by default. The system administrator -can permit other users to use \fBlockstat\fR by granting them additional DTrace privileges. Refer to the \fISolaris Dynamic Tracing Guide\fR for more information about DTrace security features. -.SH OPTIONS -.sp -.LP -The following options are supported: -.SS "Event Selection" -.sp -.LP -If no event selection options are specified, the default is \fB-C\fR. -.sp -.ne 2 -.mk -.na -\fB\fB-A\fR\fR -.ad -.sp .6 -.RS 4n -Watch all lock events. \fB-A\fR is equivalent to \fB-CH\fR. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB-C\fR\fR -.ad -.sp .6 -.RS 4n -Watch contention events. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB-E\fR\fR -.ad -.sp .6 -.RS 4n +.\" +.\" $FreeBSD$ +.\" +.Dd October 24, 2013 +.Dt LOCKSTAT 1 +.Os +.Sh NAME +.Nm lockstat +.Nd report kernel lock and profiling statistics +.Sh SYNOPSIS +.Nm +.Op Fl ACEHI +.Op Fl e Ar event-list +.Op Fl i Ar rate +.Op Fl b | t | h | s Ar depth +.Op Fl n Ar num-records +.Op Fl l Ar lock Oo Ns , Ns Ar size Oc +.Op Fl d Ar duration +.Op Fl f Ar function Oo Ns , Ns Ar size Oc +.Op Fl T +.Op Fl ckgwWRpP +.Op Fl D Ar count +.Op Fl o filename +.Op Fl x Ar opt Oo Ns = Ns Ar val Oc +.Ar command +.Op Oo Ar args Oc +.Sh DESCRIPTION +The +.Nm +utility gathers and displays kernel locking and profiling statistics. +.Nm +allows you to specify which events to watch (for example, spin on adaptive +mutex, block on read access to rwlock due to waiting writers, and so forth), how +much data to gather for each event, and how to display the data. +By default, +.Nm +monitors all lock contention events, gathers frequency and timing data about +those events, and displays the data in decreasing frequency order, so that the +most common events appear first. +.Pp +.Nm +gathers data until the specified command completes. +For example, to gather statistics for a fixed-time interval, use +.Xr sleep 1 +as the command, as follows: +.Pp +.Dl # lockstat sleep 5 +.Pp +When the +.Fl I +option is specified, +.Nm lockstat +establishes a per-processor high-level periodic interrupt source to gather +profiling data. +The interrupt handler simply generates a +.Nm +event whose caller is the interrupted PC (program counter). +The profiling event is just like any other +.Nm lockstat +event, so all of the normal +.Nm lockstat +options are applicable. +.Pp +.Nm +relies on DTrace to modify the running kernel's text to intercept events of +interest. +This imposes a small but measurable overhead on all system activity, so access +to +.Nm +is restricted to super-user by default. +.Sh OPTIONS +The following options are supported: +.Ss Event Selection +If no event selection options are specified, the default is +.Fl C . +.Bl -tag -width indent +.It Fl A +Watch all lock events. +.Fl A +is equivalent to +.Fl CH . +.It Fl C +Watch contention events. +.It Fl E Watch error events. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB\fR\fB-e\fR \fIevent_list\fR\fR -.ad -.sp .6 -.RS 4n -Only watch the specified events. \fIevent\fR \fIlist\fR is a comma-separated list of events or ranges of events such as 1,4-7,35. Run \fBlockstat\fR with no arguments to get a brief description of all events. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB-H\fR\fR -.ad -.sp .6 -.RS 4n +.It Fl e Ar event-list +Only watch the specified events. +.Ar event-list +is a comma-separated list of events or ranges of events such as 1,4-7,35. +Run +.Nm +with no arguments to get a brief description of all events. +.It Fl H Watch hold events. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB-I\fR\fR -.ad -.sp .6 -.RS 4n +.It Fl I Watch profiling interrupt events. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB\fR\fB-i\fR \fIrate\fR\fR -.ad -.sp .6 -.RS 4n -Interrupt rate (per second) for \fB-I\fR. The default is 97 Hz, so that profiling doesn't run in lockstep with the clock interrupt (which runs at 100 Hz). -.RE - -.SS "Data Gathering" -.sp -.ne 2 -.mk -.na -\fB\fB-x\fR \fIarg\fR[=\fIval\fR]\fR -.ad -.sp .6 -.RS 4n -Enable or modify a DTrace runtime option or D compiler option. The list of options is found in the \fI\fR. Boolean options are enabled by specifying their name. Options with values are set by separating the option name and -value with an equals sign (=). -.RE - -.SS "Data Gathering (Mutually Exclusive)" -.sp -.ne 2 -.mk -.na -\fB\fB-b\fR\fR -.ad -.sp .6 -.RS 4n +.It Fl i Ar rate +Interrupt rate (per second) for +.Fl I . +The default is 97 Hz, so that profiling doesn't run in lockstep with the clock +interrupt (which runs at 100 Hz). +.El +.Ss Data Gathering +.Bl -tag -width indent +.It Fl x Ar arg Oo Ns = Ns Ar val Oc +Enable or modify a +.Xr dtrace 1 +runtime option or D compiler option. +Boolean options are enabled by specifying their name. +Options with values are set by separating the option name and value with an +equals sign. +.El +.Ss "Data Gathering (Mutually Exclusive)" +.Bl -tag -width indent +.It Fl b Basic statistics: lock, caller, number of events. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB-h\fR\fR -.ad -.sp .6 -.RS 4n -Histogram: Timing plus time-distribution histograms. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB\fR\fB-s\fR \fIdepth\fR\fR -.ad -.sp .6 -.RS 4n -Stack trace: Histogram plus stack traces up to \fIdepth\fR frames deep. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB-t\fR\fR -.ad -.sp .6 -.RS 4n -Timing: Basic plus timing for all events [default]. -.RE - -.SS "Data Filtering" -.sp -.ne 2 -.mk -.na -\fB\fB\fR\fB-d\fR \fIduration\fR\fR -.ad -.sp .6 -.RS 4n -Only watch events longer than \fIduration\fR. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB\fR\fB-f\fR \fIfunc[,size]\fR\fR -.ad -.sp .6 -.RS 4n -Only watch events generated by \fIfunc\fR, which can be specified as a symbolic name or hex address. \fIsize\fR defaults to the \fBELF\fR symbol size if available, or \fB1\fR if not. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB\fR\fB-l\fR \fIlock[,size]\fR\fR -.ad -.sp .6 -.RS 4n -Only watch \fIlock\fR, which can be specified as a symbolic name or hex address. \fBsize\fR defaults to the \fBELF\fR symbol size or \fB1\fR if the symbol size is not available. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB\fR\fB-n\fR \fInrecords\fR\fR -.ad -.sp .6 -.RS 4n +.It Fl h +Histogram: timing plus time-distribution histograms. +.It Fl s Ar depth +Stack trace: histogram plus stack traces up to +.Ar depth +frames deep. +.It Fl t +Timing: Basic plus timing for all events (default). +.El +.Ss "Data Filtering" +.Bl -tag -width indent +.It Fl d Ar duration +Only watch events longer than +.Ar duration . +.It Fl f Ar func Ns Oo Ns , Ns Ar size Oc Ns +Only watch events generated by +.Ar func , +which can be specified as a symbolic name or hex address. +.Ar size +defaults to the ELF symbol size if available, or 1 if not. +.It Fl l Ar lock Ns Oo Ns , Ns Ar size Oc Ns +Only watch +.Ar lock , +which can be specified as a symbolic name or hex address. +.Ar size +defaults to the ELF symbol size or 1 if the symbol size is not available. +.It Fl n Ar num-records Maximum number of data records. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB-T\fR\fR -.ad -.sp .6 -.RS 4n -Trace (rather than sample) events [off by default]. -.RE - -.SS "Data Reporting" -.sp -.ne 2 -.mk -.na -\fB\fB-c\fR\fR -.ad -.sp .6 -.RS 4n -Coalesce lock data for lock arrays (for example, \fBpse_mutex[]\fR). -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB\fR\fB-D\fR \fIcount\fR\fR -.ad -.sp .6 -.RS 4n -Only display the top \fIcount\fR events of each type. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB-g\fR\fR -.ad -.sp .6 -.RS 4n -Show total events generated by function. For example, if \fBfoo()\fR calls \fBbar()\fR in a loop, the work done by \fBbar()\fR counts as work generated by \fBfoo()\fR (along with any work done by \fBfoo()\fR itself). -The \fB-g\fR option works by counting the total number of stack frames in which each function appears. This implies two things: (1) the data reported by \fB-g\fR can be misleading if the stack traces are not deep enough, and (2) functions that are called recursively might show -greater than 100% activity. In light of issue (1), the default data gathering mode when using \fB-g\fR is \fB-s\fR \fB50\fR. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB-k\fR\fR -.ad -.sp .6 -.RS 4n +.It Fl T +Trace (rather than sample) events. +This is off by default. +.El +.Ss Data Reporting +.Bl -tag -width indent +.It Fl c +Coalesce lock data for lock arrays. +.It Fl D Ar count +Only display the top +.Ar count +events of each type. +.It Fl g +Show total events generated by function. +For example, if +.Fn foo +calls +.Fn bar +in a loop, the work done by +.Fn bar +counts as work generated by +.Fn foo +(along with any work done by +.Fn foo +itself). +The +.Fl g +option works by counting the total number of stack frames in which each function +appears. +This implies two things: (1) the data reported by +.Fl g +can be misleading if the stack traces are not deep enough, and (2) functions +that are called recursively might show greater than 100% activity. +In light of issue (1), the default data gathering mode when using +.Fl g +is +.Fl s 50 . +.It Fl k Coalesce PCs within functions. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB\fR\fB-o\fR \fIfilename\fR\fR -.ad -.sp .6 -.RS 4n -Direct output to \fIfilename\fR. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB-P\fR\fR -.ad -.sp .6 -.RS 4n +.It Fl o Ar filename +Direct output to +.Ar filename . +.It Fl P Sort data by (\fIcount * time\fR) product. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB-p\fR\fR -.ad -.sp .6 -.RS 4n +.It Fl p Parsable output format. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB-R\fR\fR -.ad -.sp .6 -.RS 4n +.It Fl R Display rates (events per second) rather than counts. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB-W\fR\fR -.ad -.sp .6 -.RS 4n +.It Fl W Whichever: distinguish events only by caller, not by lock. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fB-w\fR\fR -.ad -.sp .6 -.RS 4n +.It Fl w Wherever: distinguish events only by lock, not by caller. -.RE - -.SH DISPLAY FORMATS -.sp -.LP +.El +.Sh DISPLAY FORMATS The following headers appear over various columns of data. -.sp -.ne 2 -.mk -.na -\fB\fBCount\fR or \fBops/s\fR\fR -.ad -.sp .6 -.RS 4n -Number of times this event occurred, or the rate (times per second) if \fB-R\fR was specified. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fBindv\fR\fR -.ad -.sp .6 -.RS 4n +.Bl -tag -width indent +.It Count or ops/s +Number of times this event occurred, or the rate (times per second) if +.Fl R +was specified. +.It indv Percentage of all events represented by this individual event. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fBgenr\fR\fR -.ad -.sp .6 -.RS 4n +.It genr Percentage of all events generated by this function. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fBcuml\fR\fR -.ad -.sp .6 -.RS 4n +.It cuml Cumulative percentage; a running total of the individuals. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fBrcnt\fR\fR -.ad -.sp .6 -.RS 4n -Average reference count. This will always be \fB1\fR for exclusive locks (mutexes, spin locks, rwlocks held as writer) but can be greater than \fB1\fR for shared locks (rwlocks held as reader). -.RE - -.sp -.ne 2 -.mk -.na -\fB\fBnsec\fR\fR -.ad -.sp .6 -.RS 4n -Average duration of the events in nanoseconds, as appropriate for the event. For the profiling event, duration means interrupt latency. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fBLock\fR\fR -.ad -.sp .6 -.RS 4n +.It rcnt +Average reference count. +This will always be 1 for exclusive locks (mutexes, +spin locks, rwlocks held as writer) but can be greater than 1 for shared locks +(rwlocks held as reader). +.It nsec +Average duration of the events in nanoseconds, as appropriate for the event. +For the profiling event, duration means interrupt latency. +.It Lock Address of the lock; displayed symbolically if possible. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fBCPU+PIL\fR\fR -.ad -.sp .6 -.RS 4n -\fBCPU\fR plus processor interrupt level (\fBPIL\fR). For example, if \fBCPU\fR 4 is interrupted while at \fBPIL\fR 6, this will be reported as \fBcpu[4]+6\fR. -.RE - -.sp -.ne 2 -.mk -.na -\fB\fBCaller\fR\fR -.ad -.sp .6 -.RS 4n +.It CPU+PIL +CPU plus processor interrupt level (PIL). +For example, if CPU 4 is interrupted while at PIL 6, this will be reported as +cpu[4]+6. +.It Caller Address of the caller; displayed symbolically if possible. -.RE - -.SH EXAMPLES -.LP -\fBExample 1 \fRMeasuring Kernel Lock Contention -.sp -.in +2 -.nf -example# \fBlockstat sleep 5\fR -Adaptive mutex spin: 2210 events in 5.055 seconds (437 events/sec) -.fi -.in -2 -.sp - -.sp -.in +2 -.nf -Count indv cuml rcnt nsec Lock Caller ------------------------------------------------------------------------- - 269 12% 12% 1.00 2160 service_queue background+0xdc - 249 11% 23% 1.00 86 service_queue qenable_locked+0x64 - 228 10% 34% 1.00 131 service_queue background+0x15c - 68 3% 37% 1.00 79 0x30000024070 untimeout+0x1c - 59 3% 40% 1.00 384 0x300066fa8e0 background+0xb0 - 43 2% 41% 1.00 30 rqcred_lock svc_getreq+0x3c - 42 2% 43% 1.00 341 0x30006834eb8 background+0xb0 - 41 2% 45% 1.00 135 0x30000021058 untimeout+0x1c - 40 2% 47% 1.00 39 rqcred_lock svc_getreq+0x260 - 37 2% 49% 1.00 2372 0x300068e83d0 hmestart+0x1c4 - 36 2% 50% 1.00 77 0x30000021058 timeout_common+0x4 - 36 2% 52% 1.00 354 0x300066fa120 background+0xb0 - 32 1% 53% 1.00 97 0x30000024070 timeout_common+0x4 - 31 1% 55% 1.00 2923 0x300069883d0 hmestart+0x1c4 - 29 1% 56% 1.00 366 0x300066fb290 background+0xb0 - 28 1% 57% 1.00 117 0x3000001e040 untimeout+0x1c - 25 1% 59% 1.00 93 0x3000001e040 timeout_common+0x4 - 22 1% 60% 1.00 25 0x30005161110 sync_stream_buf+0xdc - 21 1% 60% 1.00 291 0x30006834eb8 putq+0xa4 - 19 1% 61% 1.00 43 0x3000515dcb0 mdf_alloc+0xc - 18 1% 62% 1.00 456 0x30006834eb8 qenable+0x8 - 18 1% 63% 1.00 61 service_queue queuerun+0x168 - 17 1% 64% 1.00 268 0x30005418ee8 vmem_free+0x3c -[...] - -R/W reader blocked by writer: 76 events in 5.055 seconds (15 events/sec) +.El +.Sh EXAMPLES +.Bl -tag -width 0n +.It Example 1 Measuring Kernel Lock Contention +.Pp +.Li # lockstat sleep 5 +.Bd -literal +Adaptive mutex spin: 41411 events in 5.011 seconds (8263 events/sec) -Count indv cuml rcnt nsec Lock Caller ------------------------------------------------------------------------- - 23 30% 30% 1.00 22590137 0x300098ba358 ufs_dirlook+0xd0 - 17 22% 53% 1.00 5820995 0x3000ad815e8 find_bp+0x10 - 13 17% 70% 1.00 2639918 0x300098ba360 ufs_iget+0x198 - 4 5% 75% 1.00 3193015 0x300098ba360 ufs_getattr+0x54 - 3 4% 79% 1.00 7953418 0x3000ad817c0 find_bp+0x10 - 3 4% 83% 1.00 935211 0x3000ad815e8 find_read_lof+0x14 - 2 3% 86% 1.00 16357310 0x300073a4720 find_bp+0x10 - 2 3% 88% 1.00 2072433 0x300073a4720 find_read_lof+0x14 - 2 3% 91% 1.00 1606153 0x300073a4370 find_bp+0x10 - 1 1% 92% 1.00 2656909 0x300107e7400 ufs_iget+0x198 -[...] -.fi -.in -2 -.sp - -.LP -\fBExample 2 \fRMeasuring Hold Times -.sp -.in +2 -.nf -example# \fBlockstat -H -D 10 sleep 1\fR -Adaptive mutex spin: 513 events -.fi -.in -2 -.sp - -.sp -.in +2 -.nf -Count indv cuml rcnt nsec Lock Caller -------------------------------------------------------------------------- - 480 5% 5% 1.00 1136 0x300007718e8 putnext+0x40 - 286 3% 9% 1.00 666 0x3000077b430 getf+0xd8 - 271 3% 12% 1.00 537 0x3000077b430 msgio32+0x2fc - 270 3% 15% 1.00 3670 0x300007718e8 strgetmsg+0x3d4 - 270 3% 18% 1.00 1016 0x300007c38b0 getq_noenab+0x200 - 264 3% 20% 1.00 1649 0x300007718e8 strgetmsg+0xa70 - 216 2% 23% 1.00 6251 tcp_mi_lock tcp_snmp_get+0xfc - 206 2% 25% 1.00 602 thread_free_lock clock+0x250 - 138 2% 27% 1.00 485 0x300007c3998 putnext+0xb8 - 138 2% 28% 1.00 3706 0x300007718e8 strrput+0x5b8 -------------------------------------------------------------------------- +Count indv cuml rcnt nsec Lock Caller +------------------------------------------------------------------------------- +13750 33% 33% 0.00 72 vm_page_queue_free_mtx vm_page_free_toq+0x12e +13648 33% 66% 0.00 66 vm_page_queue_free_mtx vm_page_alloc+0x138 + 4023 10% 76% 0.00 51 vm_dom+0x80 vm_page_dequeue+0x68 + 2672 6% 82% 0.00 186 vm_dom+0x80 vm_page_enqueue+0x63 + 618 1% 84% 0.00 31 0xfffff8000cd83a88 qsyncvp+0x37 + 506 1% 85% 0.00 164 0xfffff8000cb3f098 vputx+0x5a + 477 1% 86% 0.00 69 0xfffff8000c7eb180 uma_dbg_getslab+0x5b + 288 1% 87% 0.00 77 0xfffff8000cd8b000 vn_finished_write+0x29 + 263 1% 88% 0.00 103 0xfffff8000cbad448 vinactive+0xdc + 259 1% 88% 0.00 53 0xfffff8000cd8b000 vfs_ref+0x24 + 237 1% 89% 0.00 20 0xfffff8000cbad448 vfs_hash_get+0xcc + 233 1% 89% 0.00 22 0xfffff8000bfd9480 uma_dbg_getslab+0x5b + 223 1% 90% 0.00 20 0xfffff8000cb3f098 cache_lookup+0x561 + 193 0% 90% 0.00 16 0xfffff8000cb40ba8 vref+0x27 + 175 0% 91% 0.00 34 0xfffff8000cbad448 vputx+0x5a + 169 0% 91% 0.00 51 0xfffff8000cd8b000 vfs_unbusy+0x27 + 164 0% 92% 0.00 31 0xfffff8000cb40ba8 vputx+0x5a [...] -.fi -.in -2 -.sp -.LP -\fBExample 3 \fRMeasuring Hold Times for Stack Traces Containing a Specific Function -.sp -.in +2 -.nf -example# \fBlockstat -H -f tcp_rput_data -s 50 -D 10 sleep 1\fR -Adaptive mutex spin: 11 events in 1.023 seconds (11 -events/sec) -.fi -.in -2 -.sp +Adaptive mutex block: 10 events in 5.011 seconds (2 events/sec) -.sp -.in +2 -.nf -------------------------------------------------------------------------- Count indv cuml rcnt nsec Lock Caller - 9 82% 82% 1.00 2540 0x30000031380 tcp_rput_data+0x2b90 - - nsec ------ Time Distribution ------ count Stack - 256 |@@@@@@@@@@@@@@@@ 5 tcp_rput_data+0x2b90 - 512 |@@@@@@ 2 putnext+0x78 - 1024 |@@@ 1 ip_rput+0xec4 - 2048 | 0 _c_putnext+0x148 - 4096 | 0 hmeread+0x31c - 8192 | 0 hmeintr+0x36c - 16384 |@@@ 1 -sbus_intr_wrapper+0x30 +------------------------------------------------------------------------------- + 3 30% 30% 0.00 17592 vm_page_queue_free_mtx vm_page_alloc+0x138 + 2 20% 50% 0.00 20528 vm_dom+0x80 vm_page_enqueue+0x63 + 2 20% 70% 0.00 55502 0xfffff8000cb40ba8 vputx+0x5a + 1 10% 80% 0.00 12007 vm_page_queue_free_mtx vm_page_free_toq+0x12e + 1 10% 90% 0.00 9125 0xfffff8000cbad448 vfs_hash_get+0xcc + 1 10% 100% 0.00 7864 0xfffff8000cd83a88 qsyncvp+0x37 +------------------------------------------------------------------------------- [...] +.Ed +.It Example 2 Measuring Hold Times +.Pp +.Li # lockstat -H -D 10 sleep 1 +.Bd -literal +Adaptive mutex hold: 109589 events in 1.039 seconds (105526 events/sec) Count indv cuml rcnt nsec Lock Caller - 1 9% 91% 1.00 1036 0x30000055380 freemsg+0x44 - - nsec ------ Time Distribution ------ count Stack - 1024 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ 1 freemsg+0x44 - tcp_rput_data+0x2fd0 - putnext+0x78 - ip_rput+0xec4 - _c_putnext+0x148 - hmeread+0x31c - hmeintr+0x36c - -sbus_intr_wrapper+0x30 -------------------------------------------------------------------------- -[...] -.fi -.in -2 -.sp - -.LP -\fBExample 4 \fRBasic Kernel Profiling -.sp -.LP -For basic profiling, we don't care whether the profiling interrupt sampled \fBfoo()\fR\fB+0x4c\fR or \fBfoo()\fR\fB+0x78\fR; we care only that it sampled somewhere in \fBfoo()\fR, so we use \fB-k\fR. The \fBCPU\fR and \fBPIL\fR aren't relevant to basic profiling because we are measuring the system as a whole, not a particular \fBCPU\fR or interrupt level, so we use \fB-W\fR. - -.sp -.in +2 -.nf -example# \fBlockstat -kIW -D 20 ./polltest\fR -Profiling interrupt: 82 events in 0.424 seconds (194 -events/sec) -.fi -.in -2 -.sp - -.sp -.in +2 -.nf -Count indv cuml rcnt nsec Hottest CPU+PIL Caller ------------------------------------------------------------------------ - 8 10% 10% 1.00 698 cpu[1] utl0 - 6 7% 17% 1.00 299 cpu[0] read - 5 6% 23% 1.00 124 cpu[1] getf - 4 5% 28% 1.00 327 cpu[0] fifo_read - 4 5% 33% 1.00 112 cpu[1] poll - 4 5% 38% 1.00 212 cpu[1] uiomove - 4 5% 43% 1.00 361 cpu[1] mutex_tryenter - 3 4% 46% 1.00 682 cpu[0] write - 3 4% 50% 1.00 89 cpu[0] pcache_poll - 3 4% 54% 1.00 118 cpu[1] set_active_fd - 3 4% 57% 1.00 105 cpu[0] syscall_trap32 - 3 4% 61% 1.00 640 cpu[1] (usermode) - 2 2% 63% 1.00 127 cpu[1] fifo_poll - 2 2% 66% 1.00 300 cpu[1] fifo_write - 2 2% 68% 1.00 669 cpu[0] releasef - 2 2% 71% 1.00 112 cpu[1] bt_getlowbit - 2 2% 73% 1.00 247 cpu[1] splx - 2 2% 76% 1.00 503 cpu[0] mutex_enter - 2 2% 78% 1.00 467 cpu[0]+10 disp_lock_enter - 2 2% 80% 1.00 139 cpu[1] default_copyin ------------------------------------------------------------------------ -[...] -.fi -.in -2 -.sp - -.LP -\fBExample 5 \fRGenerated-load Profiling -.sp -.LP -In the example above, 5% of the samples were in \fBpoll()\fR. This tells us how much time was spent inside \fBpoll()\fR itself, but tells us nothing about how much work was \fBgenerated\fR by \fBpoll()\fR; that is, how much time we spent -in functions called by \fBpoll()\fR. To determine that, we use the \fB-g\fR option. The example below shows that although \fBpolltest\fR spends only 5% of its time in \fBpoll()\fR itself, \fBpoll()\fR-induced work accounts for 34% of -the load. - -.sp -.LP -Note that the functions that generate the profiling interrupt (\fBlockstat_intr()\fR, \fBcyclic_fire()\fR, and so forth) appear in every stack trace, and therefore are considered to have generated 100% of the load. This illustrates an important point: the generated -load percentages do \fBnot\fR add up to 100% because they are not independent. If 72% of all stack traces contain both \fBfoo()\fR and \fBbar()\fR, then both \fBfoo()\fR and \fBbar()\fR are 72% load generators. - -.sp -.in +2 -.nf -example# \fBlockstat -kgIW -D 20 ./polltest\fR -Profiling interrupt: 80 events in 0.412 seconds (194 events/sec) -.fi -.in -2 -.sp - -.sp -.in +2 -.nf -Count genr cuml rcnt nsec Hottest CPU+PIL Caller -------------------------------------------------------------------------- - 80 100% ---- 1.00 310 cpu[1] lockstat_intr - 80 100% ---- 1.00 310 cpu[1] cyclic_fire - 80 100% ---- 1.00 310 cpu[1] cbe_level14 - 80 100% ---- 1.00 310 cpu[1] current_thread - 27 34% ---- 1.00 176 cpu[1] poll - 20 25% ---- 1.00 221 cpu[0] write - 19 24% ---- 1.00 249 cpu[1] read - 17 21% ---- 1.00 232 cpu[0] write32 - 17 21% ---- 1.00 207 cpu[1] pcache_poll - 14 18% ---- 1.00 319 cpu[0] fifo_write - 13 16% ---- 1.00 214 cpu[1] read32 - 10 12% ---- 1.00 208 cpu[1] fifo_read - 10 12% ---- 1.00 787 cpu[1] utl0 - 9 11% ---- 1.00 178 cpu[0] pcacheset_resolve - 9 11% ---- 1.00 262 cpu[0] uiomove - 7 9% ---- 1.00 506 cpu[1] (usermode) - 5 6% ---- 1.00 195 cpu[1] fifo_poll - 5 6% ---- 1.00 136 cpu[1] syscall_trap32 - 4 5% ---- 1.00 139 cpu[0] releasef - 3 4% ---- 1.00 277 cpu[1] polllock -------------------------------------------------------------------------- -[...] -.fi -.in -2 -.sp - -.LP -\fBExample 6 \fRGathering Lock Contention and Profiling Data for a Specific Module -.sp -.LP -In this example we use the \fB-f\fR option not to specify a single function, but rather to specify the entire text space of the \fBsbus\fR module. We gather both lock contention and profiling statistics so that contention can be correlated with overall load on the -module. - -.sp -.in +2 -.nf -example# \fBmodinfo | grep sbus\fR -24 102a8b6f b8b4 59 1 sbus (SBus (sysio) nexus driver) -.fi -.in -2 -.sp - -.sp -.in +2 -.nf -example# \fBlockstat -kICE -f 0x102a8b6f,0xb8b4 sleep 10\fR -Adaptive mutex spin: 39 events in 10.042 seconds (4 events/sec) -.fi -.in -2 -.sp - -.sp -.in +2 -.nf -Count indv cuml rcnt nsec Lock Caller -------------------------------------------------------------------------- - 15 38% 38% 1.00 206 0x30005160528 sync_stream_buf - 7 18% 56% 1.00 14 0x30005160d18 sync_stream_buf - 6 15% 72% 1.00 27 0x300060c3118 sync_stream_buf - 5 13% 85% 1.00 24 0x300060c3510 sync_stream_buf - 2 5% 90% 1.00 29 0x300060c2d20 sync_stream_buf - 2 5% 95% 1.00 24 0x30005161cf8 sync_stream_buf - 1 3% 97% 1.00 21 0x30005161110 sync_stream_buf - 1 3% 100% 1.00 23 0x30005160130 sync_stream_buf +------------------------------------------------------------------------------- + 8998 8% 8% 0.00 617 0xfffff8000c7eb180 uma_dbg_getslab+0xd4 + 5901 5% 14% 0.00 917 vm_page_queue_free_mtx vm_object_terminate+0x16a + 5040 5% 18% 0.00 902 vm_dom+0x80 vm_page_free_toq+0x88 + 4884 4% 23% 0.00 1056 vm_page_queue_free_mtx vm_page_alloc+0x44e + 4664 4% 27% 0.00 759 vm_dom+0x80 vm_fault_hold+0x1a13 + 4011 4% 31% 0.00 888 vm_dom vm_page_advise+0x11b + 4010 4% 34% 0.00 957 vm_dom+0x80 _vm_page_deactivate+0x5c + 3743 3% 38% 0.00 582 0xfffff8000cf04838 pmap_is_prefaultable+0x158 + 2254 2% 40% 0.00 952 vm_dom vm_page_free_toq+0x88 + 1639 1% 41% 0.00 591 0xfffff800d60065b8 trap_pfault+0x1f7 +------------------------------------------------------------------------------- [...] -Adaptive mutex block: 9 events in 10.042 seconds (1 events/sec) +R/W writer hold: 64314 events in 1.039 seconds (61929 events/sec) -Count indv cuml rcnt nsec Lock Caller -------------------------------------------------------------------------- - 4 44% 44% 1.00 156539 0x30005160528 sync_stream_buf - 2 22% 67% 1.00 763516 0x30005160d18 sync_stream_buf - 1 11% 78% 1.00 462130 0x300060c3510 sync_stream_buf - 1 11% 89% 1.00 288749 0x30005161110 sync_stream_buf - 1 11% 100% 1.00 1015374 0x30005160130 sync_stream_buf -[...] - -Profiling interrupt: 229 events in 10.042 seconds (23 events/sec) - -Count indv cuml rcnt nsec Hottest CPU+PIL Caller - -------------------------------------------------------------------------- - 89 39% 39% 1.00 426 cpu[0]+6 sync_stream_buf - 64 28% 67% 1.00 398 cpu[0]+6 sbus_intr_wrapper - 23 10% 77% 1.00 324 cpu[0]+6 iommu_dvma_kaddr_load - 21 9% 86% 1.00 512 cpu[0]+6 iommu_tlb_flush - 14 6% 92% 1.00 342 cpu[0]+6 iommu_dvma_unload - 13 6% 98% 1.00 306 cpu[1] iommu_dvma_sync - 5 2% 100% 1.00 389 cpu[1] iommu_dma_bindhdl -------------------------------------------------------------------------- -[...] -.fi -.in -2 -.sp - -.LP -\fBExample 7 \fRDetermining the Average PIL (processor interrupt level) for a CPU -.sp -.in +2 -.nf -example# \fBlockstat -Iw -l cpu[3] ./testprog\fR - -Profiling interrupt: 14791 events in 152.463 seconds (97 events/sec) - -Count indv cuml rcnt nsec CPU+PIL Hottest Caller - ------------------------------------------------------------------------ -13641 92% 92% 1.00 253 cpu[3] (usermode) - 579 4% 96% 1.00 325 cpu[3]+6 ip_ocsum+0xe8 - 375 3% 99% 1.00 411 cpu[3]+10 splx - 154 1% 100% 1.00 527 cpu[3]+4 fas_intr_svc+0x80 - 41 0% 100% 1.00 293 cpu[3]+13 send_mondo+0x18 - 1 0% 100% 1.00 266 cpu[3]+12 zsa_rxint+0x400 ------------------------------------------------------------------------ -[...] -.fi -.in -2 -.sp - -.LP -\fBExample 8 \fRDetermining which Subsystem is Causing the System to be Busy -.sp -.in +2 -.nf -example# \fBlockstat -s 10 -I sleep 20\fR - -Profiling interrupt: 4863 events in 47.375 seconds (103 events/sec) - -Count indv cuml rcnt nsec CPU+PIL Caller - ------------------------------------------------------------------------ -1929 40% 40% 0.00 3215 cpu[0] usec_delay+0x78 - nsec ------ Time Distribution ------ count Stack - 4096 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ 1872 ata_wait+0x90 - 8192 | 27 acersb_get_intr_status+0x34 -16384 | 29 ata_set_feature+0x124 -32768 | 1 ata_disk_start+0x15c - ata_hba_start+0xbc - ghd_waitq_process_and \e - _mutex_hold+0x70 - ghd_waitq_process_and \e - _mutex_exit+0x4 - ghd_transport+0x12c - ata_disk_tran_start+0x108 ------------------------------------------------------------------------ +Count indv cuml rcnt nsec Lock Caller +------------------------------------------------------------------------------- + 7421 12% 12% 0.00 2994 pvh_global_lock pmap_page_is_mapped+0xb6 + 4668 7% 19% 0.00 3313 pvh_global_lock pmap_enter+0x9ae + 1639 3% 21% 0.00 733 0xfffff80168d10200 vm_object_deallocate+0x683 + 1639 3% 24% 0.00 3061 0xfffff80168d10200 unlock_and_deallocate+0x2b + 1639 3% 26% 0.00 2966 0xfffff80168d10200 vm_fault_hold+0x16ee + 1567 2% 29% 0.00 733 0xfffff80168d10200 vm_fault_hold+0x19bc + 821 1% 30% 0.00 786 0xfffff801eb0cc000 vm_object_madvise+0x32d + 649 1% 31% 0.00 4918 0xfffff80191105300 vm_fault_hold+0x16ee + 648 1% 32% 0.00 8112 0xfffff80191105300 unlock_and_deallocate+0x2b + 647 1% 33% 0.00 1261 0xfffff80191105300 vm_object_deallocate+0x683 +------------------------------------------------------------------------------- +.Ed +.It Example 3 Measuring Hold Times for Stack Traces Containing a Specific Function +.Pp +.Li # lockstat -H -f tcp_input -s 50 -D 10 sleep 1 +.Bd -literal +Adaptive mutex hold: 68 events in 1.026 seconds (66 events/sec) + +------------------------------------------------------------------------------- +Count indv cuml rcnt nsec Lock Caller + 32 47% 47% 0.00 1631 0xfffff800686f50d8 tcp_do_segment+0x284b + + nsec ------ Time Distribution ------ count Stack + 1024 |@@@@@@@@@@ 11 tcp_input+0xf54 + 2048 |@@@@@@@@@@@@@ 14 ip_input+0xc8 + 4096 |@@@@@ 6 swi_net+0x192 + 8192 | 1 intr_event_execute_handlers+0x93 + ithread_loop+0xa6 + fork_exit+0x84 + 0xffffffff808cf9ee +------------------------------------------------------------------------------- +Count indv cuml rcnt nsec Lock Caller + 29 43% 90% 0.00 4851 0xfffff800686f50d8 sowakeup+0xf8 + + nsec ------ Time Distribution ------ count Stack + 4096 |@@@@@@@@@@@@@@@ 15 tcp_do_segment+0x2423 + 8192 |@@@@@@@@@@@@ 12 tcp_input+0xf54 + 16384 |@@ 2 ip_input+0xc8 + swi_net+0x192 + intr_event_execute_handlers+0x93 + ithread_loop+0xa6 + fork_exit+0x84 + 0xffffffff808cf9ee +------------------------------------------------------------------------------- [...] -.fi -.in -2 -.sp - -.SH ATTRIBUTES -.sp -.LP -See \fBattributes\fR(5) for descriptions of the following attributes: -.sp - -.sp -.TS -tab() box; -cw(2.75i) |cw(2.75i) -lw(2.75i) |lw(2.75i) -. -ATTRIBUTE TYPEATTRIBUTE VALUE -_ -AvailabilitySUNWdtrc -.TE - -.SH SEE ALSO -.sp -.LP -\fBdtrace\fR(1M), \fBplockstat\fR(1M), \fBattributes\fR(5), \fBlockstat\fR(7D), \fBmutex\fR(9F), \fBrwlock\fR(9F) -.sp -.LP -\fISolaris Dynamic Tracing Guide\fR -.SH NOTES -.sp -.LP -The profiling support provided by \fBlockstat\fR \fB-I\fR replaces the old (and undocumented) \fB/usr/bin/kgmon\fR and \fB/dev/profile\fR. -.sp -.LP -Tail-call elimination can affect call sites. For example, if \fBfoo()\fR\fB+0x50\fR calls \fBbar()\fR and the last thing \fBbar()\fR does is call \fBmutex_exit()\fR, the compiler can arrange for \fBbar()\fR to -branch to \fBmutex_exit()\fRwith a return address of \fBfoo()\fR\fB+0x58\fR. Thus, the \fBmutex_exit()\fR in \fBbar()\fR will appear as though it occurred at \fBfoo()\fR\fB+0x58\fR. -.sp -.LP -The \fBPC\fR in the stack frame in which an interrupt occurs can be bogus because, between function calls, the compiler is free to use the return address register for local storage. -.sp -.LP -When using the \fB-I\fR and \fB-s\fR options together, the interrupted PC will usually not appear anywhere in the stack since the interrupt handler is entered asynchronously, not by a function call from that \fBPC\fR. -.sp -.LP -The \fBlockstat\fR technology is provided on an as-is basis. The format and content of \fBlockstat\fR output reflect the current Solaris kernel implementation and are therefore subject to change in future releases. +.Ed +.El +.Sh SEE ALSO +.Xr dtrace 1 , +.Xr ksyms 4 , +.Xr locking 9 +.Sh NOTES +Tail-call elimination can affect call sites. +For example, if +.Fn foo Ns +0x50 +calls +.Fn bar +and the last thing +.Fn bar +does is call +.Fn mtx_unlock , +the compiler can arrange for +.Fn bar +to branch to +.Fn mtx_unlock +with a return address of +.Fn foo Ns +0x58. +Thus, the +.Fn mtx_unlock +in +.Fn bar +will appear as though it occurred at +.Fn foo Ns +0x58. +.Pp +The PC in the stack frame in which an interrupt occurs can be bogus because, +between function calls, the compiler is free to use the return address register +for local storage. +.Pp +When using the +.Fl I +and +.Fl s +options together, the interrupted PC will usually not appear anywhere in the +stack since the interrupt handler is entered asynchronously, not by a function +call from that PC. diff --git a/cddl/usr.sbin/lockstat/Makefile b/cddl/usr.sbin/lockstat/Makefile index ef3ac3c2d556..0668758c2f97 100644 --- a/cddl/usr.sbin/lockstat/Makefile +++ b/cddl/usr.sbin/lockstat/Makefile @@ -3,7 +3,6 @@ .PATH: ${.CURDIR}/../../../cddl/contrib/opensolaris/cmd/lockstat PROG= lockstat -NO_MAN= SRCS= lockstat.c sym.c BINDIR?= /usr/sbin -- cgit v1.2.3