Re: [Qemu-devel] [PATCH] linux-aio: keep processing events if MAX_EVENTS reached

[Roman Penyaev]

On Jun 27, 2016 6:37 PM, "Stefan Hajnoczi" <address@hidden> wrote:
>
> Commit ccb9dc10129954d0bcd7814298ed445e684d5a2a ("linux-aio: Cancel BH
> if not needed") exposed a side-effect of scheduling the BH for nested
> event loops.  When io_getevents(2) fetches MAX_EVENTS events then it's
> necessary to call it again.  Failure to do so can lead to hung I/O
> because the eventfd has already been cleared and the completion BH will
> not run again.
>
> This patch fixes the hang by calling io_getevents(2) again if the events
> array was totally full.
>
> Reported-by: Roman Penyaev <address@hidden>
> Signed-off-by: Stefan Hajnoczi <address@hidden>
> ---
>  block/linux-aio.c | 5 +++++
>  1 file changed, 5 insertions(+)
>
> diff --git a/block/linux-aio.c b/block/linux-aio.c
> index e468960..af15f85 100644
> --- a/block/linux-aio.c
> +++ b/block/linux-aio.c
> @@ -117,6 +117,7 @@ static void qemu_laio_completion_bh(void *opaque)
>      LinuxAioState *s = opaque;
>
>      /* Fetch more completion events when empty */
> +more_events:
>      if (s->event_idx == s->event_max) {
>          do {
>              struct timespec ts = { 0 };
> @@ -146,6 +147,10 @@ static void qemu_laio_completion_bh(void *opaque)
>          qemu_laio_process_completion(laiocb);
>      }
>
> +    if (s->event_idx == MAX_EVENTS) {
> +        goto more_events; /* there might still be events waiting for us
*/
> +    }
> +

I am on vacation and can't check anything, but seems you also
could reproduce an issue and seems you are right: what I've
also noticed is that io hangs only if total number of requests
in guest > than 128 (what is defined in linux-aio.c).

But I do not understand why you have to repeat io_getevents in
case of return value == MAX_EVENTS. According to my
understanding you cant submit more than 128, so the first
io_getevents call returns you exactly what you've submitted.

--
Roman

>      if (!s->io_q.plugged && !QSIMPLEQ_EMPTY(&s->io_q.pending)) {
>          ioq_submit(s);
>      }
> --
> 2.7.4
>