public inbox for linux-scsi@vger.kernel.org
 help / color / mirror / Atom feed
From: "Darrick J. Wong" <djwong@us.ibm.com>
To: linux-scsi@vger.kernel.org, linux-kernel@vger.kernel.org
Cc: alexisb@us.ibm.com
Subject: [PATCH 10/15] aic94xx: Don't call pci_map_sg for already-mapped scatterlists
Date: Fri, 17 Nov 2006 13:08:17 -0800	[thread overview]
Message-ID: <20061117210817.17052.27478.stgit@localhost.localdomain> (raw)
In-Reply-To: <20061117210737.17052.67041.stgit@localhost.localdomain>


It turns out that libata has already dma_map_sg'd the scatterlist
entries that go with an ata_queued_cmd by the time it calls
sas_ata_qc_issue.  sas_ata_qc_issue passes this scatterlist to aic94xx. 
Unfortunately, aic94xx assumes that any scatterlist passed to it needs
to be pci_map_sg'd... which blows away the mapping that libata created! 
This causes (on a x260) Calgary IOMMU table leaks and duplicate frees
when aic94xx and libata try to {pci,dma}_unmap_sg the scatterlist. 

Since dma_map_sg and pci_map_sg are fed the same struct device, I think
it's safe to add a flag to sas_task that tells aic94xx that it need
not map the scatterlist.  It didn't break anything on the x260, though
I don't have any SATAPI devices handy for testing.

Signed-off-by: Darrick J. Wong <djwong@us.ibm.com>
---

 drivers/scsi/aic94xx/aic94xx_task.c |   17 +++++++++++------
 drivers/scsi/libsas/sas_ata.c       |    1 +
 include/scsi/libsas.h               |    1 +
 3 files changed, 13 insertions(+), 6 deletions(-)

diff --git a/drivers/scsi/aic94xx/aic94xx_task.c b/drivers/scsi/aic94xx/aic94xx_task.c
index 466b492..f801c64 100644
--- a/drivers/scsi/aic94xx/aic94xx_task.c
+++ b/drivers/scsi/aic94xx/aic94xx_task.c
@@ -74,8 +74,11 @@ static inline int asd_map_scatterlist(st
 		return 0;
 	}
 
-	num_sg = pci_map_sg(asd_ha->pcidev, task->scatter, task->num_scatter,
-			    task->data_dir);
+	if (task->external_sg)
+		num_sg = task->num_scatter;
+	else
+		num_sg = pci_map_sg(asd_ha->pcidev, task->scatter,
+				    task->num_scatter, task->data_dir);
 	if (num_sg == 0)
 		return -ENOMEM;
 
@@ -120,8 +123,9 @@ static inline int asd_map_scatterlist(st
 
 	return 0;
 err_unmap:
-	pci_unmap_sg(asd_ha->pcidev, task->scatter, task->num_scatter,
-		     task->data_dir);
+	if (!task->external_sg)
+		pci_unmap_sg(asd_ha->pcidev, task->scatter, task->num_scatter,
+			     task->data_dir);
 	return res;
 }
 
@@ -142,8 +146,9 @@ static inline void asd_unmap_scatterlist
 	}
 
 	asd_free_coherent(asd_ha, ascb->sg_arr);
-	pci_unmap_sg(asd_ha->pcidev, task->scatter, task->num_scatter,
-		     task->data_dir);
+	if (!task->external_sg)
+		pci_unmap_sg(asd_ha->pcidev, task->scatter, task->num_scatter,
+			     task->data_dir);
 }
 
 /* ---------- Task complete tasklet ---------- */
diff --git a/drivers/scsi/libsas/sas_ata.c b/drivers/scsi/libsas/sas_ata.c
index 209f402..77860ab 100644
--- a/drivers/scsi/libsas/sas_ata.c
+++ b/drivers/scsi/libsas/sas_ata.c
@@ -169,6 +169,7 @@ static unsigned int sas_ata_qc_issue(str
 		task->num_scatter = num;
 	}
 
+	task->external_sg = 1;
 	task->data_dir = qc->dma_dir;
 	task->scatter = qc->__sg;
 	task->ata_task.retry_count = 1;
diff --git a/include/scsi/libsas.h b/include/scsi/libsas.h
index a06cbde..7dcf593 100644
--- a/include/scsi/libsas.h
+++ b/include/scsi/libsas.h
@@ -538,6 +538,7 @@ struct sas_task {
 	void   *uldd_task;
 
 	struct work_struct abort_work;
+	int    external_sg;
 };
 
 

  parent reply	other threads:[~2006-11-17 21:01 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-11-17 21:07 [PATCH 00/15] Roll-up of my libsas, aic94xx and sas_ata patches Darrick J. Wong
2006-11-17 21:07 ` [PATCH 01/15] aic94xx: Handle REQ_DEVICE_RESET Darrick J. Wong
2006-11-17 21:07 ` [PATCH 02/15] libsas: Clean up rphys/port dev list after a discovery error Darrick J. Wong
2006-11-17 21:07 ` [PATCH 03/15] aic94xx: Delete ascb timers when freeing queues Darrick J. Wong
2006-11-17 21:07 ` [PATCH 04/15] libsas: Don't give scsi_cmnds to the EH if they never made it to the SAS LLDD or have already returned Darrick J. Wong
2006-11-25  6:58   ` [PATCH v2] " Darrick J. Wong
2006-11-25 19:02     ` James Bottomley
2006-11-17 21:07 ` [PATCH 05/15] libsas: Add a sysfs knob to enable/disable a phy Darrick J. Wong
2006-11-17 21:07 ` [PATCH 06/15] sas_ata: Require CONFIG_ATA in Kconfig Darrick J. Wong
2006-11-17 21:08 ` [PATCH 07/15] sas_ata: Satisfy libata qc function locking requirements Darrick J. Wong
2006-11-17 21:08 ` [PATCH 08/15] sas_ata: sas_ata_qc_issue should return AC_ERR_* Darrick J. Wong
2006-11-17 21:08 ` [PATCH 09/15] sas_ata: ata_post_internal should abort the sas_task Darrick J. Wong
2006-11-17 21:08 ` Darrick J. Wong [this message]
2006-11-17 21:08 ` [PATCH 11/15] sas_ata: Don't copy aic94xx's sactive to ata_port Darrick J. Wong
2006-11-17 21:08 ` [PATCH 12/15] sas_ata: Implement sata phy control Darrick J. Wong
2006-11-17 21:08 ` [PATCH 13/15] sas_ata: Implement a libata error handler Darrick J. Wong
2006-11-17 21:08 ` [PATCH 14/15] libsas: Provide a generic SATL registration function Darrick J. Wong
2006-11-17 21:08 ` [PATCH 15/15] sas_ata: Make this a module separate from libsas Darrick J. Wong
2006-11-21  1:16 ` [PATCH 00/15] Roll-up of my libsas, aic94xx and sas_ata patches Darrick J. Wong

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=20061117210817.17052.27478.stgit@localhost.localdomain \
    --to=djwong@us.ibm.com \
    --cc=alexisb@us.ibm.com \
    --cc=linux-kernel@vger.kernel.org \
    --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