[PATCH RESEND] scsi: return correct blkprep status code in case scsi_init_io() fails.

[PATCH RESEND] scsi: return correct blkprep status code in case scsi_init_io() fails.

[Johannes Thumshirn]

When instrumenting the SCSI layer to run into the
!blk_rq_nr_phys_segments(rq) case the following  warning emitted from the
block layer:

blk_peek_request: bad return=-22

This happens because since commit fd3fc0b4d730 ('scsi: don't BUG_ON()
empty DMA transfers') we return the wrong error value from iscsi_prep_fn()
back to the block layer.

Signed-off-by: Johannes Thumshirn <jthumshirn@suse.de>
Fixes: fd3fc0b4d730 scsi: don't BUG_ON() empty DMA transfers
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 drivers/scsi/scsi_lib.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

 Re-send as in the initial send I forgot to Cc linux-scsi@

diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
index 19125d72f322..5558e212368b 100644
--- a/drivers/scsi/scsi_lib.c
+++ b/drivers/scsi/scsi_lib.c
@@ -1061,10 +1061,10 @@ int scsi_init_io(struct scsi_cmnd *cmd)
 	struct scsi_device *sdev = cmd->device;
 	struct request *rq = cmd->request;
 	bool is_mq = (rq->mq_ctx != NULL);
-	int error;
+	int error = BLKPREP_KILL;
 
 	if (WARN_ON_ONCE(!blk_rq_nr_phys_segments(rq)))
-		return -EINVAL;
+		goto err_exit;
 
 	error = scsi_init_sgtable(rq, &cmd->sdb);
 	if (error)
-- 
2.12.0

Re: [PATCH RESEND] scsi: return correct blkprep status code in case scsi_init_io() fails.

[Bart Van Assche]

On Tue, 2017-04-11 at 09:46 +0200, Johannes Thumshirn wrote:
> When instrumenting the SCSI layer to run into the
> !blk_rq_nr_phys_segments(rq) case the following  warning emitted from the
> block layer:
> 
> blk_peek_request: bad return=-22
> 
> This happens because since commit fd3fc0b4d730 ('scsi: don't BUG_ON()
> empty DMA transfers') we return the wrong error value from iscsi_prep_fn()
                                                             ^^^^^^^^^^^^^^^
There is no iscsi_prep_fn() in the kernel tree. Did you perhaps mean
scsi_prep_fn()?

> back to the block layer.
> 
> Signed-off-by: Johannes Thumshirn <jthumshirn@suse.de>
> Fixes: fd3fc0b4d730 scsi: don't BUG_ON() empty DMA transfers
> Reviewed-by: Christoph Hellwig <hch@lst.de>

Commit fd3fc0b4d730 has a "Cc: stable" tag. I assume that that means that
this patch also needs that tag?

Otherwise this patch looks fine to me.

Bart.

Re: [PATCH RESEND] scsi: return correct blkprep status code in case scsi_init_io() fails.

[Johannes Thumshirn]

On Tue, Apr 11, 2017 at 06:43:22PM +0000, Bart Van Assche wrote:
> On Tue, 2017-04-11 at 09:46 +0200, Johannes Thumshirn wrote:
> > When instrumenting the SCSI layer to run into the
> > !blk_rq_nr_phys_segments(rq) case the following  warning emitted from the
> > block layer:
> > 
> > blk_peek_request: bad return=-22
> > 
> > This happens because since commit fd3fc0b4d730 ('scsi: don't BUG_ON()
> > empty DMA transfers') we return the wrong error value from iscsi_prep_fn()
>                                                              ^^^^^^^^^^^^^^^
> There is no iscsi_prep_fn() in the kernel tree. Did you perhaps mean
> scsi_prep_fn()?

*doh* Yup, you're right.

> 
> > back to the block layer.
> > 
> > Signed-off-by: Johannes Thumshirn <jthumshirn@suse.de>
> > Fixes: fd3fc0b4d730 scsi: don't BUG_ON() empty DMA transfers
> > Reviewed-by: Christoph Hellwig <hch@lst.de>
> 
> Commit fd3fc0b4d730 has a "Cc: stable" tag. I assume that that means that
> this patch also needs that tag?

Probably yes. I'll include both "fixes" in my resend.

-- 
Johannes Thumshirn                                          Storage
jthumshirn@suse.de                                +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850