diff options
| author | Jaeyoon Choi <jaeyoon@FreeBSD.org> | 2026-08-10 01:53:27 +0000 |
|---|---|---|
| committer | Jaeyoon Choi <jaeyoon@FreeBSD.org> | 2026-08-10 02:28:54 +0000 |
| commit | a33860b0a2b98caf32c2ff62707f254ca92773f4 (patch) | |
| tree | 1e6448ca997a807f010978ea1f44b83e443b35bd | |
| parent | 2d32961316248fdff54a2a9fc24ac8b712fee9b0 (diff) | |
ufshci: do not reset the device in the XPT_RESET_DEV handler
CAM calls the SIM action callback with the SIM lock and the CAM
device lock held. The XPT_RESET_DEV handler called
ufshci_dev_reset(), which sleeps on device commands. Sleeping there
panics when another thread contends for the lock: "panic: sleeping
thread holds CAM device lock".
Report success without touching the device, as nvme_sim(4) does.
A real device reset needs the controller reset path. That rework is
planned together with in-flight request recovery.
Sponsored by: Samsung Electronics
Reviewed by: imp (mentor)
Differential Revision: https://reviews.freebsd.org/D58671
| -rw-r--r-- | sys/dev/ufshci/ufshci_sim.c | 13 |
1 files changed, 7 insertions, 6 deletions
diff --git a/sys/dev/ufshci/ufshci_sim.c b/sys/dev/ufshci/ufshci_sim.c index e40079b23354..f969d9c8311b 100644 --- a/sys/dev/ufshci/ufshci_sim.c +++ b/sys/dev/ufshci/ufshci_sim.c @@ -288,13 +288,14 @@ ufshci_cam_action(struct cam_sim *sim, union ccb *ccb) break; } case XPT_RESET_BUS: - ccb->ccb_h.status = CAM_REQ_CMP; - break; case XPT_RESET_DEV: - if (ufshci_dev_reset(ctrlr)) - ccb->ccb_h.status = CAM_REQ_CMP_ERR; - else - ccb->ccb_h.status = CAM_REQ_CMP; + /* + * This callback cannot sleep: CAM calls it with the SIM + * lock and the CAM device lock held. It cannot reset the + * device here. Report success so CAM keeps going, like + * nvme_sim(4) does. + */ + ccb->ccb_h.status = CAM_REQ_CMP; break; case XPT_ABORT: ccb->ccb_h.status = CAM_FUNC_NOTAVAIL; |