1
0
mirror of https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git synced 2026-01-11 17:10:13 +00:00

Compare commits

...

5 Commits

Author SHA1 Message Date
Linus Torvalds
c53f467229 SCSI fixes on 20251226
Three HBA driver and one upper level driver (sg) fix.  The sg change
 is the largest, but that results mostly from moving code to avoid the
 described race condition.
 
 Signed-off-by: James E.J. Bottomley <James.Bottomley@HansenPartnership.com>
 -----BEGIN PGP SIGNATURE-----
 
 iLgEABMIAGAWIQTnYEDbdso9F2cI+arnQslM7pishQUCaU9UWRsUgAAAAAAEAA5t
 YW51MiwyLjUrMS4xMSwyLDImHGphbWVzLmJvdHRvbWxleUBoYW5zZW5wYXJ0bmVy
 c2hpcC5jb20ACgkQ50LJTO6YrIUsFQEA49RD3/43W3fNYU1RfqcjDFmaJgDeUpvk
 q+JF6RTzjaQA/j7t4sGNLnPq8kDBhacYhStIFUUMoeHkiS8Nq+s0TuWL
 =T0c1
 -----END PGP SIGNATURE-----

Merge tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi

Pull SCSI fixes from James Bottomley:
 "Three HBA driver and one upper level driver (sg) fix.

  The sg change is the largest, but that results mostly from moving code
  to avoid the described race condition"

* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
  scsi: ufs: core: Add ufshcd_update_evt_hist() for UFS suspend error
  scsi: sg: Fix occasional bogus elapsed time that exceeds timeout
  scsi: mpi3mr: Read missing IOCFacts flag for reply queue full overflow
  scsi: scsi_debug: Fix atomic write enable module param description
2025-12-26 19:44:08 -08:00
Seunghwan Baek
c9f36f04a8 scsi: ufs: core: Add ufshcd_update_evt_hist() for UFS suspend error
If UFS resume fails, the event history is updated in ufshcd_resume(), but
there is no code anywhere to record UFS suspend. Therefore, add code to
record UFS suspend error event history.

Fixes: dd11376b9f1b ("scsi: ufs: Split the drivers/scsi/ufs directory")
Cc: stable@vger.kernel.org
Signed-off-by: Seunghwan Baek <sh8267.baek@samsung.com>
Reviewed-by: Peter Wang <peter.wang@mediatek.com>
Link: https://patch.msgid.link/20251210063854.1483899-2-sh8267.baek@samsung.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2025-12-16 22:24:36 -05:00
Michal Rábek
0e16776542 scsi: sg: Fix occasional bogus elapsed time that exceeds timeout
A race condition was found in sg_proc_debug_helper(). It was observed on
a system using an IBM LTO-9 SAS Tape Drive (ULTRIUM-TD9) and monitoring
/proc/scsi/sg/debug every second. A very large elapsed time would
sometimes appear. This is caused by two race conditions.

We reproduced the issue with an IBM ULTRIUM-HH9 tape drive on an x86_64
architecture. A patched kernel was built, and the race condition could
not be observed anymore after the application of this patch. A
reproducer C program utilising the scsi_debug module was also built by
Changhui Zhong and can be viewed here:

https://github.com/MichaelRabek/linux-tests/blob/master/drivers/scsi/sg/sg_race_trigger.c

The first race happens between the reading of hp->duration in
sg_proc_debug_helper() and request completion in sg_rq_end_io().  The
hp->duration member variable may hold either of two types of
information:

 #1 - The start time of the request. This value is present while
      the request is not yet finished.

 #2 - The total execution time of the request (end_time - start_time).

If sg_proc_debug_helper() executes *after* the value of hp->duration was
changed from #1 to #2, but *before* srp->done is set to 1 in
sg_rq_end_io(), a fresh timestamp is taken in the else branch, and the
elapsed time (value type #2) is subtracted from a timestamp, which
cannot yield a valid elapsed time (which is a type #2 value as well).

To fix this issue, the value of hp->duration must change under the
protection of the sfp->rq_list_lock in sg_rq_end_io().  Since
sg_proc_debug_helper() takes this read lock, the change to srp->done and
srp->header.duration will happen atomically from the perspective of
sg_proc_debug_helper() and the race condition is thus eliminated.

The second race condition happens between sg_proc_debug_helper() and
sg_new_write(). Even though hp->duration is set to the current time
stamp in sg_add_request() under the write lock's protection, it gets
overwritten by a call to get_sg_io_hdr(), which calls copy_from_user()
to copy struct sg_io_hdr from userspace into kernel space. hp->duration
is set to the start time again in sg_common_write(). If
sg_proc_debug_helper() is called between these two calls, an arbitrary
value set by userspace (usually zero) is used to compute the elapsed
time.

To fix this issue, hp->duration must be set to the current timestamp
again after get_sg_io_hdr() returns successfully. A small race window
still exists between get_sg_io_hdr() and setting hp->duration, but this
window is only a few instructions wide and does not result in observable
issues in practice, as confirmed by testing.

Additionally, we fix the format specifier from %d to %u for printing
unsigned int values in sg_proc_debug_helper().

Signed-off-by: Michal Rábek <mrabek@redhat.com>
Suggested-by: Tomas Henzl <thenzl@redhat.com>
Tested-by: Changhui Zhong <czhong@redhat.com>
Reviewed-by: Ewan D. Milne <emilne@redhat.com>
Reviewed-by: John Meneghini <jmeneghi@redhat.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Link: https://patch.msgid.link/20251212160900.64924-1-mrabek@redhat.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2025-12-16 21:24:28 -05:00
Chandrakanth Patil
d373163194 scsi: mpi3mr: Read missing IOCFacts flag for reply queue full overflow
The driver was not reading the MAX_REQ_PER_REPLY_QUEUE_LIMIT IOCFacts
flag, so the reply-queue-full handling was never enabled, even on
firmware that supports it. Reading this flag enables the feature and
prevents reply queue overflow.

Fixes: f08b24d82749 ("scsi: mpi3mr: Avoid reply queue full condition")
Cc: stable@vger.kernel.org
Signed-off-by: Chandrakanth Patil <chandrakanth.patil@broadcom.com>
Link: https://patch.msgid.link/20251211002929.22071-1-chandrakanth.patil@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2025-12-16 21:19:26 -05:00
John Garry
1f7d6e2efe scsi: scsi_debug: Fix atomic write enable module param description
The atomic write enable module param is "atomic_wr", and not
"atomic_write", so fix the module param description.

Fixes: 84f3a3c01d70 ("scsi: scsi_debug: Atomic write support")
Signed-off-by: John Garry <john.g.garry@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://patch.msgid.link/20251211100651.9056-1-john.g.garry@oracle.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2025-12-16 21:09:44 -05:00
5 changed files with 21 additions and 9 deletions

View File

@ -166,6 +166,7 @@ struct mpi3_ioc_facts_data {
#define MPI3_IOCFACTS_FLAGS_SIGNED_NVDATA_REQUIRED (0x00010000)
#define MPI3_IOCFACTS_FLAGS_DMA_ADDRESS_WIDTH_MASK (0x0000ff00)
#define MPI3_IOCFACTS_FLAGS_DMA_ADDRESS_WIDTH_SHIFT (8)
#define MPI3_IOCFACTS_FLAGS_MAX_REQ_PER_REPLY_QUEUE_LIMIT (0x00000040)
#define MPI3_IOCFACTS_FLAGS_INITIAL_PORT_ENABLE_MASK (0x00000030)
#define MPI3_IOCFACTS_FLAGS_INITIAL_PORT_ENABLE_SHIFT (4)
#define MPI3_IOCFACTS_FLAGS_INITIAL_PORT_ENABLE_NOT_STARTED (0x00000000)

View File

@ -3158,6 +3158,8 @@ static void mpi3mr_process_factsdata(struct mpi3mr_ioc *mrioc,
mrioc->facts.dma_mask = (facts_flags &
MPI3_IOCFACTS_FLAGS_DMA_ADDRESS_WIDTH_MASK) >>
MPI3_IOCFACTS_FLAGS_DMA_ADDRESS_WIDTH_SHIFT;
mrioc->facts.max_req_limit = (facts_flags &
MPI3_IOCFACTS_FLAGS_MAX_REQ_PER_REPLY_QUEUE_LIMIT);
mrioc->facts.protocol_flags = facts_data->protocol_flags;
mrioc->facts.mpi_version = le32_to_cpu(facts_data->mpi_version.word);
mrioc->facts.max_reqs = le16_to_cpu(facts_data->max_outstanding_requests);

View File

@ -7459,7 +7459,7 @@ MODULE_PARM_DESC(lbprz,
MODULE_PARM_DESC(lbpu, "enable LBP, support UNMAP command (def=0)");
MODULE_PARM_DESC(lbpws, "enable LBP, support WRITE SAME(16) with UNMAP bit (def=0)");
MODULE_PARM_DESC(lbpws10, "enable LBP, support WRITE SAME(10) with UNMAP bit (def=0)");
MODULE_PARM_DESC(atomic_write, "enable ATOMIC WRITE support, support WRITE ATOMIC(16) (def=0)");
MODULE_PARM_DESC(atomic_wr, "enable ATOMIC WRITE support, support WRITE ATOMIC(16) (def=0)");
MODULE_PARM_DESC(lowest_aligned, "lowest aligned lba (def=0)");
MODULE_PARM_DESC(lun_format, "LUN format: 0->peripheral (def); 1 --> flat address method");
MODULE_PARM_DESC(max_luns, "number of LUNs per target to simulate(def=1)");

View File

@ -731,6 +731,8 @@ sg_new_write(Sg_fd *sfp, struct file *file, const char __user *buf,
sg_remove_request(sfp, srp);
return -EFAULT;
}
hp->duration = jiffies_to_msecs(jiffies);
if (hp->interface_id != 'S') {
sg_remove_request(sfp, srp);
return -ENOSYS;
@ -815,7 +817,6 @@ sg_common_write(Sg_fd * sfp, Sg_request * srp,
return -ENODEV;
}
hp->duration = jiffies_to_msecs(jiffies);
if (hp->interface_id != '\0' && /* v3 (or later) interface */
(SG_FLAG_Q_AT_TAIL & hp->flags))
at_head = 0;
@ -1338,9 +1339,6 @@ sg_rq_end_io(struct request *rq, blk_status_t status)
"sg_cmd_done: pack_id=%d, res=0x%x\n",
srp->header.pack_id, result));
srp->header.resid = resid;
ms = jiffies_to_msecs(jiffies);
srp->header.duration = (ms > srp->header.duration) ?
(ms - srp->header.duration) : 0;
if (0 != result) {
struct scsi_sense_hdr sshdr;
@ -1389,6 +1387,9 @@ sg_rq_end_io(struct request *rq, blk_status_t status)
done = 0;
}
srp->done = done;
ms = jiffies_to_msecs(jiffies);
srp->header.duration = (ms > srp->header.duration) ?
(ms - srp->header.duration) : 0;
write_unlock_irqrestore(&sfp->rq_list_lock, iflags);
if (likely(done)) {
@ -2533,6 +2534,7 @@ static void sg_proc_debug_helper(struct seq_file *s, Sg_device * sdp)
const sg_io_hdr_t *hp;
const char * cp;
unsigned int ms;
unsigned int duration;
k = 0;
list_for_each_entry(fp, &sdp->sfds, sfd_siblings) {
@ -2570,13 +2572,17 @@ static void sg_proc_debug_helper(struct seq_file *s, Sg_device * sdp)
seq_printf(s, " id=%d blen=%d",
srp->header.pack_id, blen);
if (srp->done)
seq_printf(s, " dur=%d", hp->duration);
seq_printf(s, " dur=%u", hp->duration);
else {
ms = jiffies_to_msecs(jiffies);
seq_printf(s, " t_o/elap=%d/%d",
duration = READ_ONCE(hp->duration);
if (duration)
duration = (ms > duration ?
ms - duration : 0);
seq_printf(s, " t_o/elap=%u/%u",
(new_interface ? hp->timeout :
jiffies_to_msecs(fp->timeout)),
(ms > hp->duration ? ms - hp->duration : 0));
duration);
}
seq_printf(s, "ms sgat=%d op=0x%02x\n", usg,
(int) srp->data.cmd_opcode);

View File

@ -10359,7 +10359,7 @@ static int ufshcd_suspend(struct ufs_hba *hba)
ret = ufshcd_setup_clocks(hba, false);
if (ret) {
ufshcd_enable_irq(hba);
return ret;
goto out;
}
if (ufshcd_is_clkgating_allowed(hba)) {
hba->clk_gating.state = CLKS_OFF;
@ -10371,6 +10371,9 @@ static int ufshcd_suspend(struct ufs_hba *hba)
/* Put the host controller in low power mode if possible */
ufshcd_hba_vreg_set_lpm(hba);
ufshcd_pm_qos_update(hba, false);
out:
if (ret)
ufshcd_update_evt_hist(hba, UFS_EVT_SUSPEND_ERR, (u32)ret);
return ret;
}