qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Paolo Bonzini <pbonzini@redhat.com>
To: Jan Kiszka <jan.kiszka@siemens.com>
Cc: Kevin Wolf <kwolf@redhat.com>,
	Anthony Liguori <aliguori@us.ibm.com>,
	Alex Bligh <alex@alex.org.uk>,
	Stefan Hajnoczi <stefanha@gmail.com>,
	qemu-devel@nongnu.org, liu ping fan <qemulist@gmail.com>,
	Stefan Hajnoczi <stefanha@redhat.com>,
	MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>,
	rth@twiddle.net
Subject: Re: [Qemu-devel] [RFC] [PATCHv10 00/31] aio / timers: Add AioContext timers and use ppoll
Date: Tue, 17 Sep 2013 19:04:42 +0200	[thread overview]
Message-ID: <52388BAA.1060207@redhat.com> (raw)
In-Reply-To: <5238883A.90607@siemens.com>

Il 17/09/2013 18:50, Jan Kiszka ha scritto:
> On 2013-09-17 18:38, Alex Bligh wrote:
>>
>> On 17 Sep 2013, at 17:19, Paolo Bonzini wrote:
>>
>>> Il 17/09/2013 18:09, Jan Kiszka ha scritto:
>>>> On 2013-08-13 16:22, Stefan Hajnoczi wrote:
>>>>> On Tue, Aug 13, 2013 at 03:45:44PM +0200, Jan Kiszka wrote:
>>>>>> Yeah:
>>>>>>
>>>>>> -    /* No AIO operations?  Get us out of here */
>>>>>> -    if (!busy) {
>>>>>> +    /* early return if we only have the aio_notify() fd */
>>>>>> +    if (ctx->pollfds->len == 1) {
>>>>>>         return progress;
>>>>>>     }
>>>>>>
>>>>>> So this is even worse for my use case.
>>>>>
>>>>> We can change the semantics of aio_poll() so long as we don't break
>>>>> existing callers and tests.  It would make sense to do that after
>>>>> merging the io_flush and AioContext timers series.
>>>>
>>>> Need to pick up this topic again because above change is now mainline
>>>> and breaks aio_poll-based timer threads:
>>>>
>>>> How can we make progress with overcoming that check, at least for the
>>>> timer thread use case? Additional argument "truly_block" for aio_poll?
>>>
>>> I wonder if we still need that "if" at all.  Guys, do you remember what
>>> it is good for? O:-)
>>
>> I can't remember whether the code above was in my patch or Stefan's
>> subsequent ones, but I had suggested
>>    if (blocking && (ctx->pollfds->len <= 1))
>>
>> On reflection, I don't think the 'if' line should be there at all.
>>
>> I think the argument was that it was meant to be for stupidity
>> protection, i.e. where someone calls with blocking set to 1 and no
>> fds, it would block indefinitely (if there were no timers)
>> as it would select with no timeout and no fds.
>>
>> Personally I think if you call things this way you deserve all you
>> get. I'm not sure attempting to rescue the user by returning
>> immediately is a great plan. If there are truly no fds at all
>> (not even the notify fd) and an infinite timeout perhaps we should
>> set the timeout to a second and log an error?
>>
>> I presume the reason it's breaking aio_poll based timer threads is
>> because there is only one fd (the aio_notify_fd), there are
>> no other fd's, but there is a timeout (from the timers)?
> 
> Right.
> 
>> If
>> that's true, I think we /shouldn't/ return. Equally if there
>> are no timers but something is genuinely attempting to wait
>> on an aio_notify, I don't think we should return.
>>
> 
> In any case, test-aio seems to stress that if clause. If we remove it,
> the test case hangs infinitely. But I'm more worried about understanding
> if there are actual users depending on the current behavior
> (bdrv_drain_all?).

Yes, bdrv_drain_all comes to mind.  But now we should be able to fix
this old remark:

        /* FIXME: We do not have timer support here, so this is effectively
         * a busy wait.
         */
        QTAILQ_FOREACH(bs, &bdrv_states, list) {
            if (bdrv_start_throttled_reqs(bs)) {
                busy = true;
            }
        }

        busy = bdrv_requests_pending_all();
        busy |= aio_poll(qemu_get_aio_context(), busy);

Alex, what's missing before block.c and QED can use aio_timer_new on
the main AioContext, instead of timer_new?

Paolo

  parent reply	other threads:[~2013-09-17 17:04 UTC|newest]

Thread overview: 70+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-08-11 16:42 [Qemu-devel] [RFC] [PATCHv10 00/31] aio / timers: Add AioContext timers and use ppoll Alex Bligh
2013-08-11 16:42 ` [Qemu-devel] [RFC] [PATCHv10 01/31] aio / timers: Rename qemu_timer_* functions Alex Bligh
2013-08-11 16:42 ` [Qemu-devel] [RFC] [PATCHv10 02/31] aio / timers: Rename qemu_new_clock and expose clock types Alex Bligh
2013-08-11 16:42 ` [Qemu-devel] [RFC] [PATCHv10 03/31] aio / timers: add qemu-timer.c utility functions Alex Bligh
2013-08-11 16:42 ` [Qemu-devel] [RFC] [PATCHv10 04/31] aio / timers: Consistent treatment of disabled clocks for deadlines Alex Bligh
2013-08-11 16:42 ` [Qemu-devel] [RFC] [PATCHv10 05/31] aio / timers: add ppoll support with qemu_poll_ns Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 06/31] aio / timers: Add prctl(PR_SET_TIMERSLACK, 1, ...) to reduce timer slack Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 07/31] aio / timers: Make qemu_run_timers and qemu_run_all_timers return progress Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 08/31] aio / timers: Split QEMUClock into QEMUClock and QEMUTimerList Alex Bligh
2013-08-12 16:14   ` Jan Kiszka
2013-08-12 16:25     ` Alex Bligh
2013-08-12 16:36       ` Jan Kiszka
2013-08-12 17:04       ` Richard Henderson
2013-08-12 17:25         ` Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 09/31] aio / timers: Untangle include files Alex Bligh
2013-08-12 16:40   ` Jan Kiszka
2013-08-12 17:04     ` Alex Bligh
2013-08-12 17:12       ` Jan Kiszka
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 10/31] aio / timers: Add QEMUTimerListGroup and helper functions Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 11/31] aio / timers: Add QEMUTimerListGroup to AioContext Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 12/31] aio / timers: Add a notify callback to QEMUTimerList Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 13/31] aio / timers: aio_ctx_prepare sets timeout from AioContext timers Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 14/31] aio / timers: Add aio_timer_init & aio_timer_new wrappers Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 15/31] aio / timers: Convert aio_poll to use AioContext timers' deadline Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 16/31] aio / timers: Convert mainloop to use timeout Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 17/31] aio / timers: On timer modification, qemu_notify or aio_notify Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 18/31] aio / timers: Introduce new API timer_new and friends Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 19/31] aio / timers: Use all timerlists in icount warp calculations Alex Bligh
2013-08-15 12:30   ` Stefan Hajnoczi
2013-08-15 12:37     ` Alex Bligh
2013-08-15 18:31       ` Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 20/31] aio / timers: Add documentation and new format calls Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 21/31] aio / timers: Remove alarm timers Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 22/31] aio / timers: Remove legacy qemu_clock_deadline & qemu_timerlist_deadline Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 23/31] aio / timers: Add qemu_clock_get_ms and qemu_clock_get_ms Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 24/31] aio / timers: Rearrange timer.h & make legacy functions call non-legacy Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 25/31] aio / timers: Remove main_loop_timerlist Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 26/31] aio / timers: Convert rtc_clock to be a QEMUClockType Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 27/31] aio / timers: convert block_job_sleep_ns and co_sleep_ns to new API Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 28/31] aio / timers: Add test harness for AioContext timers Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 29/31] aio / timers: Add scripts/switch-timer-api Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 30/31] aio / timers: Switch entire codebase to the new timer API Alex Bligh
2013-08-11 16:43 ` [Qemu-devel] [RFC] [PATCHv10 31/31] aio / timers: Remove legacy interface Alex Bligh
2013-08-13 12:22 ` [Qemu-devel] [RFC] [PATCHv10 00/31] aio / timers: Add AioContext timers and use ppoll Jan Kiszka
2013-08-13 12:44   ` Alex Bligh
2013-08-13 12:57     ` Jan Kiszka
2013-08-13 13:12   ` Alex Bligh
2013-08-13 13:25     ` Jan Kiszka
2013-08-13 13:39       ` Alex Bligh
2013-08-13 13:45         ` Jan Kiszka
2013-08-13 13:56           ` Alex Bligh
2013-08-13 14:22           ` Stefan Hajnoczi
2013-08-13 14:26             ` Alex Bligh
2013-08-13 14:34               ` Jan Kiszka
2013-08-14 13:09               ` Stefan Hajnoczi
2013-09-17 16:09             ` Jan Kiszka
2013-09-17 16:19               ` Paolo Bonzini
2013-09-17 16:38                 ` Alex Bligh
2013-09-17 16:50                   ` Jan Kiszka
2013-09-17 17:03                     ` Alex Bligh
2013-09-17 17:04                     ` Paolo Bonzini [this message]
2013-09-17 17:32                       ` Alex Bligh
2013-09-18  7:57                         ` Paolo Bonzini
2013-09-18  8:23                           ` Alex Bligh
2013-09-18  9:02                             ` Alex Bligh
2013-09-18  9:25                               ` Paolo Bonzini
2013-09-24 13:47                                 ` Stefan Hajnoczi
2013-09-24 13:48                             ` Stefan Hajnoczi
2013-08-15 12:40 ` Stefan Hajnoczi
2013-08-15 13:05   ` Alex Bligh

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=52388BAA.1060207@redhat.com \
    --to=pbonzini@redhat.com \
    --cc=alex@alex.org.uk \
    --cc=aliguori@us.ibm.com \
    --cc=jan.kiszka@siemens.com \
    --cc=kwolf@redhat.com \
    --cc=morita.kazutaka@lab.ntt.co.jp \
    --cc=qemu-devel@nongnu.org \
    --cc=qemulist@gmail.com \
    --cc=rth@twiddle.net \
    --cc=stefanha@gmail.com \
    --cc=stefanha@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).