Re: [Bug 11817] New: SCSI: timing out command for dvd drive

Mike Christie <michaelc@xxxxxxxxxxx> · Fri, 24 Oct 2008 11:44:47 -0500

Mike Christie wrote:
bugme-daemon@xxxxxxxxxxxxxxxxxxx wrote:
http://bugzilla.kernel.org/show_bug.cgi?id=11817

           Summary: SCSI: timing out command for dvd drive
           Product: IO/Storage
           Version: 2.5
     KernelVersion: 2.6.28-rc1
          Platform: All
        OS/Version: Linux
              Tree: Mainline
            Status: NEW
          Severity: normal
          Priority: P1
         Component: SCSI
        AssignedTo: linux-scsi@xxxxxxxxxxxxxxx
        ReportedBy: bugzilla-kernel@xxxxxxxxxxxxxxx

Latest working kernel version: 2.6.27
Earliest failing kernel version: 2.6.28-rc1
Distribution: Gentoo
Hardware Environment: Dell Latitude E6500
Problem Description:
I tried 2.6.28-rc1 today and the boot is delayed by scsi timeouts. It 
does work
if I remove my dvd drive before booting.

dmesg:
Oct 24 12:07:12 jtb [    0.743466] Driver 'sd' needs updating - please 
use
bus_type methods
Oct 24 12:07:12 jtb [    0.743492] Driver 'sr' needs updating - please 
use
bus_type methods
Oct 24 12:07:12 jtb [    0.743555] ahci 0000:00:1f.2: version 3.0
Oct 24 12:07:12 jtb [    0.743564] ahci 0000:00:1f.2: PCI INT D -> GSI 19
(level, low) -> IRQ 19
Oct 24 12:07:12 jtb [    0.743596] ahci 0000:00:1f.2: irq 42 for 
MSI/MSI-X
Oct 24 12:07:12 jtb [    0.743666] ahci 0000:00:1f.2: AHCI 0001.0200 
32 slots 4
ports 3 Gbps 0x33 impl RAID mode
Oct 24 12:07:12 jtb [    0.743668] ahci 0000:00:1f.2: flags: 64bit ncq 
sntf
stag pm led clo pmp pio slum part ems
Oct 24 12:07:12 jtb [    0.743673] ahci 0000:00:1f.2: setting latency 
timer to
64
Oct 24 12:07:12 jtb [    0.743820] scsi0 : ahci
Oct 24 12:07:12 jtb [    0.743899] scsi1 : ahci
Oct 24 12:07:12 jtb [    0.743958] scsi2 : ahci
Oct 24 12:07:12 jtb [    0.744012] scsi3 : ahci
Oct 24 12:07:12 jtb [    0.744066] scsi4 : ahci
Oct 24 12:07:12 jtb [    0.744120] scsi5 : ahci
Oct 24 12:07:12 jtb [    0.744564] ata1: SATA max UDMA/133 abar
m2048@0xfed1c800 port 0xfed1c900 irq 42
Oct 24 12:07:12 jtb [    0.744568] ata2: SATA max UDMA/133 abar
m2048@0xfed1c800 port 0xfed1c980 irq 42
Oct 24 12:07:12 jtb [    0.744569] ata3: DUMMY
Oct 24 12:07:12 jtb [    0.744570] ata4: DUMMY
Oct 24 12:07:12 jtb [    0.744572] ata5: SATA max UDMA/133 abar
m2048@0xfed1c800 port 0xfed1cb00 irq 42
Oct 24 12:07:12 jtb [    0.744575] ata6: SATA max UDMA/133 abar
m2048@0xfed1c800 port 0xfed1cb80 irq 42
Oct 24 12:07:12 jtb [    1.063351] ata1: SATA link up 1.5 Gbps 
(SStatus 113
SControl 300)
Oct 24 12:07:12 jtb [    1.068536] ata1.00: ATA-8: SAMSUNG HM250JI, 
HS100-08,
max UDMA7
Oct 24 12:07:12 jtb [    1.068538] ata1.00: 488397168 sectors, multi 
0: LBA48
NCQ (depth 31/32)
Oct 24 12:07:12 jtb [    1.075477] ata1.00: configured for UDMA/133
Oct 24 12:07:12 jtb [    1.406682] ata2: SATA link up 1.5 Gbps 
(SStatus 113
SControl 300)
Oct 24 12:07:12 jtb [    1.409289] ata2.00: ATAPI: PLDS DVD+/-RW 
DU-8A2S, 4D12,
max UDMA/100
Oct 24 12:07:12 jtb [    1.412791] ata2.00: configured for UDMA/100
Oct 24 12:07:12 jtb [    1.746681] ata5: SATA link down (SStatus 0 
SControl
300)
Oct 24 12:07:12 jtb [    2.080014] ata6: SATA link down (SStatus 0 
SControl
300)
Oct 24 12:07:12 jtb [    2.093410] scsi 0:0:0:0: Direct-Access     
ATA     SAMSUNG HM250JI  HS10 PQ: 0 ANSI: 5
Oct 24 12:07:12 jtb [    2.093516] sd 0:0:0:0: [sda] 488397168 512-byte
hardware sectors: (250 GB/232 GiB)
Oct 24 12:07:12 jtb [    2.093527] sd 0:0:0:0: [sda] Write Protect is off
Oct 24 12:07:12 jtb [    2.093529] sd 0:0:0:0: [sda] Mode Sense: 00 3a 
00 00
Oct 24 12:07:12 jtb [    2.093547] sd 0:0:0:0: [sda] Write cache: 
enabled, read
cache: enabled, doesn't support DPO or FUA
Oct 24 12:07:12 jtb [    2.093592] sd 0:0:0:0: [sda] 488397168 512-byte
hardware sectors: (250 GB/232 GiB)
Oct 24 12:07:12 jtb [    2.093603] sd 0:0:0:0: [sda] Write Protect is off
Oct 24 12:07:12 jtb [    2.093605] sd 0:0:0:0: [sda] Mode Sense: 00 3a 
00 00
Oct 24 12:07:12 jtb [    2.093622] sd 0:0:0:0: [sda] Write cache: 
enabled, read
cache: enabled, doesn't support DPO or FUA
Oct 24 12:07:12 jtb [    2.093625]  sda: sda1 sda2 sda3
Oct 24 12:07:12 jtb [    2.132082] sd 0:0:0:0: [sda] Attached SCSI disk
Oct 24 12:07:12 jtb [    2.132157] sd 0:0:0:0: Attached scsi generic 
sg0 type 0
Oct 24 12:07:12 jtb [   24.136670] scsi 1:0:0:0: timing out command, 
waited 22s
Oct 24 12:07:12 jtb [   46.246667] scsi 1:0:0:0: timing out command, 
waited 22s
Oct 24 12:07:12 jtb [   68.356665] scsi 1:0:0:0: timing out command, 
waited 22s
Oct 24 12:07:12 jtb [   90.466663] scsi 1:0:0:0: timing out command, 
waited 22s
Oct 24 12:07:12 jtb [  112.576661] scsi 1:0:0:0: timing out command, 
waited 22s
Oct 24 12:07:12 jtb [  134.686658] scsi 1:0:0:0: timing out command, 
waited 22s

I think this caused due to my patch that adds the scsi_noretry_cmd check 
in scsi_decide_disposition. We saw this problem exposed in RHEL testing.

Basically what used to happen without my patch is that the command 
continuously timed out or we got an error, and we would eventually hit 
the cmd->retries and cmd->allowed check and the command would be failed 
fairly quietly in some code paths like during setup.

Now scsi_noretry_cmd is returning 0 and that causes us to retry the 
command until we hit the check:

        disposition = scsi_decide_disposition(cmd);
        if (disposition != SUCCESS &&
            time_before(cmd->jiffies_at_alloc + wait_for, jiffies)) {
                sdev_printk(KERN_ERR, cmd->device,
                            "timing out command, waited %lus\n",
                            wait_for/HZ);
                disposition = SUCCESS;
        }

in scsi_softirq_done so we now see this error message get printed out 
for the failure.

Ignore this. In RHEL it was slightly different problem.
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html