[PATCH] retry HARDWARE ERROR for IBM ESS

[PATCH] retry HARDWARE ERROR for IBM ESS

[Martin Peschke3]

Under certain conditions IBM's Enterprise Storage Server returns sense data
HARDWARE ERROR - INTERNAL TARGET FAILURE, while it wants the OS
to retry the very command.
Although this is apparently a firmware issue (ABORTED COMMAND would be
an appropriate choice), a Linux workaround would be needed to handle it.
It doesn't look like we'll ever get hold of a firmware fix for the model
addressed
by the patch below.

Martin

Signed-off-by: Martin Peschke <mpeschke@de.ibm.com>

--- linux-2.6/drivers/scsi/scsi_error.c   2004-11-19 18:38:45.000000000 +0000
+++ linux-2.6/drivers/scsi/scsi_error.c   2004-11-19 18:39:04.000000000 +0000
@@ -325,10 +325,16 @@
      case MEDIUM_ERROR:
            return NEEDS_RETRY;

+     case HARDWARE_ERROR:
+           if ((strncmp(SCpnt->device->vendor, "IBM", 3) == 0) &&
+               (strncmp(SCpnt->device->model, "2105", 4) == 0))
+                 return NEEDS_RETRY;
+           else
+                 return SUCCESS;
+
      case ILLEGAL_REQUEST:
      case BLANK_CHECK:
      case DATA_PROTECT:
-     case HARDWARE_ERROR:
      default:
            return SUCCESS;
      }

Re: [PATCH] retry HARDWARE ERROR for IBM ESS

[Christoph Hellwig]

> +     case HARDWARE_ERROR:
> +           if ((strncmp(SCpnt->device->vendor, "IBM", 3) == 0) &&
> +               (strncmp(SCpnt->device->model, "2105", 4) == 0))
> +                 return NEEDS_RETRY;
> +           else
> +                 return SUCCESS;
> +

please don't hardcode models deep inside the scsu layer.  Add a new
blacklist flags and set it from the table in scsi_devinfo.c

[PATCH] retry HARDWARE ERROR for IBM ESS

[Martin Peschke3]

James,

I did not see objections to the patch below. Please apply.

Martin


Signed-off-by: Martin Peschke <mpeschke@de.ibm.com>


diff -ur scsi-rc-fixes-2.6-ref/drivers/scsi/scsi_devinfo.c scsi-rc-fixes-2.6/drivers/scsi/scsi_devinfo.c
--- scsi-rc-fixes-2.6-ref/drivers/scsi/scsi_devinfo.c 2004-12-10 21:50:00.000000000 +0000
+++ scsi-rc-fixes-2.6/drivers/scsi/scsi_devinfo.c     2004-12-10 22:03:01.000000000 +0000
@@ -159,6 +159,7 @@
      {"HP", "C3323-300", "4269", BLIST_NOTQ},
      {"IBM", "AuSaV1S2", NULL, BLIST_FORCELUN},
      {"IBM", "ProFibre 4000R", "*", BLIST_SPARSELUN | BLIST_LARGELUN},
+     {"IBM", "2105", NULL, BLIST_RETRY_HWERROR},
      {"iomega", "jaz 1GB", "J.86", BLIST_NOTQ | BLIST_NOLUN},
      {"IOMEGA", "Io20S         *F", NULL, BLIST_KEY},
      {"INSITE", "Floptical   F*8I", NULL, BLIST_KEY},
diff -ur scsi-rc-fixes-2.6-ref/drivers/scsi/scsi_error.c scsi-rc-fixes-2.6/drivers/scsi/scsi_error.c
--- scsi-rc-fixes-2.6-ref/drivers/scsi/scsi_error.c   2004-12-10 21:50:00.000000000 +0000
+++ scsi-rc-fixes-2.6/drivers/scsi/scsi_error.c 2004-12-10 22:03:03.000000000 +0000
@@ -31,6 +31,7 @@
 #include <scsi/scsi_host.h>
 #include <scsi/scsi_ioctl.h>
 #include <scsi/scsi_request.h>
+#include <scsi/scsi_devinfo.h>

 #include "scsi_priv.h"
 #include "scsi_logging.h"
@@ -325,10 +326,18 @@
      case MEDIUM_ERROR:
            return NEEDS_RETRY;

+     case HARDWARE_ERROR:
+           if (scsi_get_device_flags(scmd->device,
+                               scmd->device->vendor,
+                               scmd->device->model)
+               & BLIST_RETRY_HWERROR)
+                 return NEEDS_RETRY;
+           else
+                 return SUCCESS;
+
      case ILLEGAL_REQUEST:
      case BLANK_CHECK:
      case DATA_PROTECT:
-     case HARDWARE_ERROR:
      default:
            return SUCCESS;
      }
diff -ur scsi-rc-fixes-2.6-ref/include/scsi/scsi_devinfo.h scsi-rc-fixes-2.6/include/scsi/scsi_devinfo.h
--- scsi-rc-fixes-2.6-ref/include/scsi/scsi_devinfo.h 2004-12-10 21:50:32.000000000 +0000
+++ scsi-rc-fixes-2.6/include/scsi/scsi_devinfo.h     2004-12-10 22:03:04.000000000 +0000
@@ -27,4 +27,5 @@
 #define BLIST_NOT_LOCKABLE   0x80000     /* don't use PREVENT-ALLOW commands */
 #define BLIST_NO_ULD_ATTACH  0x100000 /* device is actually for RAID config */
 #define BLIST_SELECT_NO_ATN  0x200000 /* select without ATN */
+#define BLIST_RETRY_HWERROR  0x400000 /* retry HARDWARE_ERROR */
 #endif

Re: [PATCH] retry HARDWARE ERROR for IBM ESS

[James Bottomley]

On Fri, 2004-12-10 at 22:15 +0100, Martin Peschke3 wrote:
> I did not see objections to the patch below. Please apply.

It looks like your mailer converted tabs to spaces.  Could you resend
unmangled with a nice change log?

Thanks,

James