diff options
author | Kleber Sacilotto de Souza <klebers@linux.vnet.ibm.com> | 2012-01-16 22:30:25 +0100 |
---|---|---|
committer | James Bottomley <JBottomley@Parallels.com> | 2012-02-18 15:33:13 +0100 |
commit | a92fa25c63a788758bd52e9123504d133210c8b7 (patch) | |
tree | bec201c46aa5ae21f2006bb87ad744741b35e241 /drivers/scsi | |
parent | [SCSI] mpt2sas: Fix mismatch in mpt2sas_base_hard_reset_handler() mutex lock-... (diff) | |
download | linux-a92fa25c63a788758bd52e9123504d133210c8b7.tar.xz linux-a92fa25c63a788758bd52e9123504d133210c8b7.zip |
[SCSI] ipr: fix eeh recovery for 64-bit adapters
In some scenarios, an EEH error can take a long time to be detected, since the
driver issues an MMIO read only after a device reset command times out and we
try to reset the adapter. This patch adds some code in ipr_cancel_op() to read
a hardware register so we detect the error earlier in case the op is being
aborted because of a timeout caused by a frozen adapter slot.
Another problem in such scenarios is that in __ipr_eh_host_reset() we change the
dump state flag from WAIT_FOR_DUMP to GET_DUMP, and the flag is later changed
from GET_DUMP to READ_DUMP in ipr_reset_restore_cfg_space(). However, if when
__ipr_eh_host_reset() is called by the SCSI error handling the function
ipr_reset_restore_cfg_space() has already been called by the PCI EEH code, we
end up with the flag in an inconsistent state. This patch also prevents this
problem.
Signed-off-by: Kleber Sacilotto de Souza <klebers@linux.vnet.ibm.com>
Acked-by: Brian King <brking@linux.vnet.ibm.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Diffstat (limited to 'drivers/scsi')
-rw-r--r-- | drivers/scsi/ipr.c | 24 |
1 files changed, 18 insertions, 6 deletions
diff --git a/drivers/scsi/ipr.c b/drivers/scsi/ipr.c index 67b169b7a5be..b538f0883fd2 100644 --- a/drivers/scsi/ipr.c +++ b/drivers/scsi/ipr.c @@ -4613,11 +4613,13 @@ static int __ipr_eh_host_reset(struct scsi_cmnd * scsi_cmd) ENTER; ioa_cfg = (struct ipr_ioa_cfg *) scsi_cmd->device->host->hostdata; - dev_err(&ioa_cfg->pdev->dev, - "Adapter being reset as a result of error recovery.\n"); + if (!ioa_cfg->in_reset_reload) { + dev_err(&ioa_cfg->pdev->dev, + "Adapter being reset as a result of error recovery.\n"); - if (WAIT_FOR_DUMP == ioa_cfg->sdt_state) - ioa_cfg->sdt_state = GET_DUMP; + if (WAIT_FOR_DUMP == ioa_cfg->sdt_state) + ioa_cfg->sdt_state = GET_DUMP; + } rc = ipr_reset_reload(ioa_cfg, IPR_SHUTDOWN_ABBREV); @@ -4907,7 +4909,7 @@ static int ipr_cancel_op(struct scsi_cmnd * scsi_cmd) struct ipr_ioa_cfg *ioa_cfg; struct ipr_resource_entry *res; struct ipr_cmd_pkt *cmd_pkt; - u32 ioasc; + u32 ioasc, int_reg; int op_found = 0; ENTER; @@ -4920,7 +4922,17 @@ static int ipr_cancel_op(struct scsi_cmnd * scsi_cmd) */ if (ioa_cfg->in_reset_reload || ioa_cfg->ioa_is_dead) return FAILED; - if (!res || !ipr_is_gscsi(res)) + if (!res) + return FAILED; + + /* + * If we are aborting a timed out op, chances are that the timeout was caused + * by a still not detected EEH error. In such cases, reading a register will + * trigger the EEH recovery infrastructure. + */ + int_reg = readl(ioa_cfg->regs.sense_interrupt_reg); + + if (!ipr_is_gscsi(res)) return FAILED; list_for_each_entry(ipr_cmd, &ioa_cfg->pending_q, queue) { |