On Thu, Jul 28, 2011 at 1:15 PM, Christoph Hellwig wrote:
> On Thu, Jul 28, 2011 at 09:47:05AM +0200, Kevin Wolf wrote:
>> > Indeed. This has come up a few times, and actually is a mostly trivial
>> > task. Maybe we should give up waiting for -blockdev and separate cache
>> > mode settings and a
On Thu, Jul 28, 2011 at 02:41:02PM +0200, Kevin Wolf wrote:
> > More or less. There's one corner case for all Linux I/O, and that is
> > only writes up to INT_MAX are supported, and larger writes (and reads)
> > get truncated to it. It's pretty nasty, but Linux has been vocally
> > opposed to fix
Am 28.07.2011 14:15, schrieb Christoph Hellwig:
>> Christoph, on another note: Can we rely on Linux AIO never returning
>> short writes except on EOF? Currently we return -EINVAL in this case, so
"short reads" I meant, of course.
>> I hope it's true or we wouldn't return the correct error code.
>
On Thu, Jul 28, 2011 at 09:47:05AM +0200, Kevin Wolf wrote:
> > Indeed. This has come up a few times, and actually is a mostly trivial
> > task. Maybe we should give up waiting for -blockdev and separate cache
> > mode settings and allow a nocache-writethrough or similar mode now? It's
> > going
Am 27.07.2011 21:57, schrieb Christoph Hellwig:
> On Wed, Jul 27, 2011 at 09:52:51PM +0200, Frediano Ziglio wrote:
>> Also I notice that combining XFS, Linux AIO, O_DIRECT and O_DSYNC give
>> impressive performance but currently there is no way to specify all that
>> flags together cause nocache
On Wed, Jul 27, 2011 at 09:52:51PM +0200, Frediano Ziglio wrote:
> >
>
> Yes! Not at kernel level :-)
In that case we have a bad error handling problem somewhere in qemu.
the IOCB_CMD_FDSYNC aio opcode will always return EINVAL currently,
and we really should have cought that somewhere in qemu.
Il giorno 27/lug/2011, alle ore 20:31, Christoph Hellwig ha
scritto:
> Did you test this at all?
>
Yes! Not at kernel level :-)
Usually I trust documentation and man pages.
> On Wed, Jul 27, 2011 at 08:25:25PM +0200, Frediano Ziglio wrote:
>> +case QEMU_AIO_FLUSH:
>> +io_prep_fds
Did you test this at all?
On Wed, Jul 27, 2011 at 08:25:25PM +0200, Frediano Ziglio wrote:
> +case QEMU_AIO_FLUSH:
> +io_prep_fdsync(iocbs, fd);
> +break;
Looks great, but doesn't work as expected.
Hint: grep for aio_fsync in the linux kernel.
Signed-off-by: Frediano Ziglio
---
block/raw-posix.c |7 +++
linux-aio.c |3 +++
2 files changed, 10 insertions(+), 0 deletions(-)
diff --git a/block/raw-posix.c b/block/raw-posix.c
index 3c6bd4b..27ae81e 100644
--- a/block/raw-posix.c
+++ b/block/raw-posix.c
@@ -628,6 +628,13
