]> git.ipfire.org Git - thirdparty/kernel/stable.git/commitdiff
ata: libata-scsi: Fix delayed scsi_rescan_device() execution
authorDamien Le Moal <dlemoal@kernel.org>
Tue, 5 Sep 2023 00:06:23 +0000 (09:06 +0900)
committerDamien Le Moal <dlemoal@kernel.org>
Thu, 28 Sep 2023 12:23:10 +0000 (21:23 +0900)
Commit 6aa0365a3c85 ("ata: libata-scsi: Avoid deadlock on rescan after
device resume") modified ata_scsi_dev_rescan() to check the scsi device
"is_suspended" power field to ensure that the scsi device associated
with an ATA device is fully resumed when scsi_rescan_device() is
executed. However, this fix is problematic as:
1) It relies on a PM internal field that should not be used without PM
   device locking protection.
2) The check for is_suspended and the call to scsi_rescan_device() are
   not atomic and a suspend PM event may be triggered between them,
   casuing scsi_rescan_device() to be called on a suspended device and
   in that function blocking while holding the scsi device lock. This
   would deadlock a following resume operation.
These problems can trigger PM deadlocks on resume, especially with
resume operations triggered quickly after or during suspend operations.
E.g., a simple bash script like:

for (( i=0; i<10; i++ )); do
echo "+2 > /sys/class/rtc/rtc0/wakealarm
echo mem > /sys/power/state
done

that triggers a resume 2 seconds after starting suspending a system can
quickly lead to a PM deadlock preventing the system from correctly
resuming.

Fix this by replacing the check on is_suspended with a check on the
return value given by scsi_rescan_device() as that function will fail if
called against a suspended device. Also make sure rescan tasks already
scheduled are first cancelled before suspending an ata port.

Fixes: 6aa0365a3c85 ("ata: libata-scsi: Avoid deadlock on rescan after device resume")
Cc: stable@vger.kernel.org
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Niklas Cassel <niklas.cassel@wdc.com>
Tested-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
drivers/ata/libata-core.c
drivers/ata/libata-scsi.c

index a0bc01606b307cac6353910c4635039435115925..092372334e9253d4cd1bf88d734f860ee917ee0c 100644 (file)
@@ -5168,11 +5168,27 @@ static const unsigned int ata_port_suspend_ehi = ATA_EHI_QUIET
 
 static void ata_port_suspend(struct ata_port *ap, pm_message_t mesg)
 {
+       /*
+        * We are about to suspend the port, so we do not care about
+        * scsi_rescan_device() calls scheduled by previous resume operations.
+        * The next resume will schedule the rescan again. So cancel any rescan
+        * that is not done yet.
+        */
+       cancel_delayed_work_sync(&ap->scsi_rescan_task);
+
        ata_port_request_pm(ap, mesg, 0, ata_port_suspend_ehi, false);
 }
 
 static void ata_port_suspend_async(struct ata_port *ap, pm_message_t mesg)
 {
+       /*
+        * We are about to suspend the port, so we do not care about
+        * scsi_rescan_device() calls scheduled by previous resume operations.
+        * The next resume will schedule the rescan again. So cancel any rescan
+        * that is not done yet.
+        */
+       cancel_delayed_work_sync(&ap->scsi_rescan_task);
+
        ata_port_request_pm(ap, mesg, 0, ata_port_suspend_ehi, true);
 }
 
index a0e58d22d22252cb35c0ff3cd0da8092fdf5ef9f..6850cac803c1ffc5dcab659ca50eb7cd2feaa4af 100644 (file)
@@ -4756,7 +4756,7 @@ void ata_scsi_dev_rescan(struct work_struct *work)
        struct ata_link *link;
        struct ata_device *dev;
        unsigned long flags;
-       bool delay_rescan = false;
+       int ret = 0;
 
        mutex_lock(&ap->scsi_scan_mutex);
        spin_lock_irqsave(ap->lock, flags);
@@ -4765,37 +4765,34 @@ void ata_scsi_dev_rescan(struct work_struct *work)
                ata_for_each_dev(dev, link, ENABLED) {
                        struct scsi_device *sdev = dev->sdev;
 
+                       /*
+                        * If the port was suspended before this was scheduled,
+                        * bail out.
+                        */
+                       if (ap->pflags & ATA_PFLAG_SUSPENDED)
+                               goto unlock;
+
                        if (!sdev)
                                continue;
                        if (scsi_device_get(sdev))
                                continue;
 
-                       /*
-                        * If the rescan work was scheduled because of a resume
-                        * event, the port is already fully resumed, but the
-                        * SCSI device may not yet be fully resumed. In such
-                        * case, executing scsi_rescan_device() may cause a
-                        * deadlock with the PM code on device_lock(). Prevent
-                        * this by giving up and retrying rescan after a short
-                        * delay.
-                        */
-                       delay_rescan = sdev->sdev_gendev.power.is_suspended;
-                       if (delay_rescan) {
-                               scsi_device_put(sdev);
-                               break;
-                       }
-
                        spin_unlock_irqrestore(ap->lock, flags);
-                       scsi_rescan_device(sdev);
+                       ret = scsi_rescan_device(sdev);
                        scsi_device_put(sdev);
                        spin_lock_irqsave(ap->lock, flags);
+
+                       if (ret)
+                               goto unlock;
                }
        }
 
+unlock:
        spin_unlock_irqrestore(ap->lock, flags);
        mutex_unlock(&ap->scsi_scan_mutex);
 
-       if (delay_rescan)
+       /* Reschedule with a delay if scsi_rescan_device() returned an error */
+       if (ret)
                schedule_delayed_work(&ap->scsi_rescan_task,
                                      msecs_to_jiffies(5));
 }