Re: [Qemu-devel] QEMU/KVM SCSI lock up

[Avi Kivity <avi@qumranet.com>]

Matteo Frigo wrote:
> kvm-64 hangs under heavy disk I/O with scsi disks.  To reproduce,
> create a fresh qcow2 disk, boot linux, and execute
>
>   dd if=/dev/sdX of=/dev/null bs=1M
>
> on the fresh disk.  See also https://sourceforge.net/tracker/?func=detail&atid=893831&aid=1895893&group_id=180599
>
> I have attached a patch that appears to fix the problem.  The bug
> seems to be the following.  scsi_read_data() does the following
>
>     bdrv_aio_read()
>     r->sector += n;
>     r->sector_count -= n;
>
> For reasons that I do not fully understand, bdrv_aio_read() does
> not return immediately, but instead it calls scsi_read_data()
> recursively.

What happens (I think) is that bdrv_aio_read() completes immediately, 
calls the completion callback, which starts a read for the next batch of 
sectors.


>   Since ``r->sector += n;'' has not been executed
> yet, the re-entrant call triggers a read of the same sector, which
> breaks the producer-consumer lockstep.  The fix is to swap the operations
> as follows:
>
>     r->sector += n;
>     r->sector_count -= n;
>     bdrv_aio_read()
>
> A similar fix applies to scsi_write_data().
>
>   

Will that not issue the read for the wrong sector?

I think the correct fix is to move r->sector and r->sector_count 
adjustment into scsi_read_complete() and scsi_write_complete().

Long term we want to replace the recursion by queuing.

-- 
error compiling committee.c: too many arguments to function