Re: scsi: add support for a blk-mq based I/O path.

2014-09-16 Thread Webb Scales
On 9/16/14 10:41 AM, Daniel Gryniewicz wrote: Hi, Chris. I'm working with the OSD Initiator, which does bi-directional requests, and this commit (d285203 scsi: add support for a blk-mq based I/O path.) causes a use-after free panic for me. The panic itself follows, but the cause is that scsi

Re: scsi: add support for a blk-mq based I/O path.

2014-09-16 Thread Christoph Hellwig
On Tue, Sep 16, 2014 at 01:17:55PM -0400, Daniel Gryniewicz wrote: >> this should also be fixed with my patch "scsi: clean up S/G table freeing", >> can you test if that fixes the issue for you as well? >> > > > Yeah, that fixes it, thanks. Thanks. Looking at the impact of that I suspect we're bet

Re: scsi: add support for a blk-mq based I/O path.

2014-09-16 Thread Daniel Gryniewicz
On 09/16/2014 12:19 PM, Christoph Hellwig wrote: Hi Daniel, this should also be fixed with my patch "scsi: clean up S/G table freeing", can you test if that fixes the issue for you as well? Yeah, that fixes it, thanks. Daniel -- To unsubscribe from this list: send the line "unsubscribe linu

Re: scsi: add support for a blk-mq based I/O path.

2014-09-16 Thread Christoph Hellwig
Hi Daniel, this should also be fixed with my patch "scsi: clean up S/G table freeing", can you test if that fixes the issue for you as well? -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majord...@vger.kernel.org More majordomo info at http:

Re: scsi: add support for a blk-mq based I/O path.

2014-09-16 Thread Daniel Gryniewicz
Hi, Chris. I'm working with the OSD Initiator, which does bi-directional requests, and this commit (d285203 scsi: add support for a blk-mq based I/O path.) causes a use-after free panic for me. The panic itself follows, but the cause is that scsi_end_request() calls blk_finish_request(), whic