linux-scsi.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jeff Skirvin <jeffrey.d.skirvin@intel.com>
To: jbottomley@parallels.com
Cc: linux-scsi@vger.kernel.org, jeffrey.d.skirvin@intel.com,
	dave.jiang@intel.com
Subject: [PATCH] isci: Allow SSP tasks into the task management path.
Date: Thu, 06 Sep 2012 10:56:49 -0700	[thread overview]
Message-ID: <20120906175608.30998.36898.stgit@jdskirvi-devbox.ch.intel.com> (raw)

This commit fixes a driver bug for SSP tasks that require task management
in the target after they complete in the SCU hardware.  The problem was
manifested in the function "isci_task_abort_task", which tests
to see if the sas_task.lldd_task is non-NULL before allowing task
management; this bug would always NULL lldd_task in the SCU I/O completion
path even if target management was required, which would prevent
task / target manangement from happening.

Note that in the case of SATA/STP targets, error recovery is provided by
the libata error handler which is why SATA/STP device recovery worked
correctly even though SSP handling did not.

Signed-off-by: Jeff Skirvin <jeffrey.d.skirvin@intel.com>
---
Resubmitting to include James...

 drivers/scsi/isci/host.c |   10 ++++++----
 1 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/scsi/isci/host.c b/drivers/scsi/isci/host.c
index 45385f5..446fdd8 100644
--- a/drivers/scsi/isci/host.c
+++ b/drivers/scsi/isci/host.c
@@ -1079,7 +1079,6 @@ static void sci_controller_completion_handler(struct isci_host *ihost)
 
 void ireq_done(struct isci_host *ihost, struct isci_request *ireq, struct sas_task *task)
 {
-	task->lldd_task = NULL;
 	if (!test_bit(IREQ_ABORT_PATH_ACTIVE, &ireq->flags) &&
 	    !(task->task_state_flags & SAS_TASK_STATE_ABORTED)) {
 		if (test_bit(IREQ_COMPLETE_IN_TARGET, &ireq->flags)) {
@@ -1087,16 +1086,19 @@ void ireq_done(struct isci_host *ihost, struct isci_request *ireq, struct sas_ta
 			dev_dbg(&ihost->pdev->dev,
 				"%s: Normal - ireq/task = %p/%p\n",
 				__func__, ireq, task);
-
+			task->lldd_task = NULL;
 			task->task_done(task);
 		} else {
 			dev_dbg(&ihost->pdev->dev,
 				"%s: Error - ireq/task = %p/%p\n",
 				__func__, ireq, task);
-
+			if (sas_protocol_ata(task->task_proto))
+				task->lldd_task = NULL;
 			sas_task_abort(task);
 		}
-	}
+	} else
+		task->lldd_task = NULL;
+
 	if (test_and_clear_bit(IREQ_ABORT_PATH_ACTIVE, &ireq->flags))
 		wake_up_all(&ihost->eventq);
 


             reply	other threads:[~2012-09-06 17:37 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-09-06 17:56 Jeff Skirvin [this message]
  -- strict thread matches above, loose matches on Subject: below --
2012-09-06  4:36 [PATCH] isci: Allow SSP tasks into the task management path Jeff Skirvin

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20120906175608.30998.36898.stgit@jdskirvi-devbox.ch.intel.com \
    --to=jeffrey.d.skirvin@intel.com \
    --cc=dave.jiang@intel.com \
    --cc=jbottomley@parallels.com \
    --cc=linux-scsi@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).