Re: [PATCH +stable] block: don't attempt to merge overlapping requests

2010-05-19 Thread Christoph Hellwig
On Wed, May 19, 2010 at 10:23:44AM +0100, Stefan Hajnoczi wrote: > On Wed, May 19, 2010 at 10:06 AM, Avi Kivity wrote: > >> In the cache=writeback case the virtio-blk guest driver does: > >> > >> blk_queue_ordered(q, QUEUE_ORDERED_DRAIN_FLUSH, ...) > >> > > > > I don't follow. ?What's the implicat

Re: [PATCH +stable] block: don't attempt to merge overlapping requests

2010-05-19 Thread Stefan Hajnoczi
On Wed, May 19, 2010 at 10:06 AM, Avi Kivity wrote: >> In the cache=writeback case the virtio-blk guest driver does: >> >> blk_queue_ordered(q, QUEUE_ORDERED_DRAIN_FLUSH, ...) >> > > I don't follow.  What's the implication? I was wondering whether the queue is incorrectly set to a mode where over

Re: [PATCH +stable] block: don't attempt to merge overlapping requests

2010-05-19 Thread Avi Kivity
On 05/19/2010 12:01 PM, Stefan Hajnoczi wrote: On Wed, May 19, 2010 at 9:09 AM, Avi Kivity wrote: On 05/18/2010 10:22 PM, Stefan Hajnoczi wrote: What cache= mode are you running? writeback. In the cache=writeback case the virtio-blk guest driver does: blk_queue_order

Re: [PATCH +stable] block: don't attempt to merge overlapping requests

2010-05-19 Thread Stefan Hajnoczi
On Wed, May 19, 2010 at 9:09 AM, Avi Kivity wrote: > On 05/18/2010 10:22 PM, Stefan Hajnoczi wrote: >> What cache= mode are you running? > > writeback. In the cache=writeback case the virtio-blk guest driver does: blk_queue_ordered(q, QUEUE_ORDERED_DRAIN_FLUSH, ...) Stefan -- To unsubscribe fro

Re: [PATCH +stable] block: don't attempt to merge overlapping requests

2010-05-19 Thread Avi Kivity
On 05/18/2010 10:22 PM, Stefan Hajnoczi wrote: On Tue, May 18, 2010 at 6:18 PM, Avi Kivity wrote: The block multiwrite code pretends to be able to merge overlapping requests, but doesn't do so in fact. This leads to I/O errors (for example on mkfs of a large virtio disk). Are overla

Re: [PATCH +stable] block: don't attempt to merge overlapping requests

2010-05-18 Thread Stefan Hajnoczi
On Tue, May 18, 2010 at 8:41 PM, Michael Tokarev wrote: > But actually I don't quite see where that dependency is: guest > does not know how its data is cached on host... I was thinking of this bit in linux-2.6:drivers/block/virtio_blk.c: /* If barriers are supported, tell block layer that queue

Re: [PATCH +stable] block: don't attempt to merge overlapping requests

2010-05-18 Thread Michael Tokarev
18.05.2010 23:37, Stefan Hajnoczi wrote: I just caught up on mails and saw you had already mentioned that overlapping writes from the guest look fishy in the "the>1Tb block issue". Cache mode might still be interesting because it affects how guest virtio-blk chooses queue ordering mode. What I

Re: [PATCH +stable] block: don't attempt to merge overlapping requests

2010-05-18 Thread Stefan Hajnoczi
I just caught up on mails and saw you had already mentioned that overlapping writes from the guest look fishy in the "the >1Tb block issue". Cache mode might still be interesting because it affects how guest virtio-blk chooses queue ordering mode. Stefan -- To unsubscribe from this list: send the

Re: [PATCH +stable] block: don't attempt to merge overlapping requests

2010-05-18 Thread Stefan Hajnoczi
On Tue, May 18, 2010 at 6:18 PM, Avi Kivity wrote: > The block multiwrite code pretends to be able to merge overlapping requests, > but doesn't do so in fact.  This leads to I/O errors (for example on mkfs > of a large virtio disk). Are overlapping write requests correct guest behavior? I though

[PATCH +stable] block: don't attempt to merge overlapping requests

2010-05-18 Thread Avi Kivity
The block multiwrite code pretends to be able to merge overlapping requests, but doesn't do so in fact. This leads to I/O errors (for example on mkfs of a large virtio disk). Signed-off-by: Avi Kivity --- block.c |4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/block.