Re: [Qemu-devel] [PATCH] timer: fix qemu_poll_ns early timeout on windows

2014-04-28 Thread Paolo Bonzini
Il 18/04/2014 10:29, Alex Bligh ha scritto: I suspect the easiest fix (certainly for debug purposes) would be to make all polls for non-zero time at least 10ms long. IE after the assignment of 'ms', do something like #ifdef BROKEN_POLL if (ms < 10) { ms = 10; #endif } If you're

Re: [Qemu-devel] [PATCH] timer: fix qemu_poll_ns early timeout on windows

2014-04-18 Thread Alex Bligh
On 18 Apr 2014, at 10:26, Stanislav Vorobiov wrote: > Yes, it's possible to work around like this, but if we look at this: > > if (ms < 10) { > ms = 10; > } > > the question arises: where did 10 come from ? It looks like a magic number > and in fact > it is, it was taken from g

Re: [Qemu-devel] [PATCH] timer: fix qemu_poll_ns early timeout on windows

2014-04-18 Thread Stanislav Vorobiov
Hi, see below On 04/18/2014 12:03 PM, Stefan Weil wrote: > Am 18.04.2014 09:34, schrieb Stanislav Vorobiov: >> Hi, >> >> Please see below >> >> On 04/18/2014 10:46 AM, Stefan Weil wrote: >>> Hi, > [...] >> >>> >>> Would it be sufficient to round any timeout > 0 and < 10 to 10 for >>> Windows hosts

Re: [Qemu-devel] [PATCH] timer: fix qemu_poll_ns early timeout on windows

2014-04-18 Thread Stanislav Vorobiov
Hi, Please see below On 04/18/2014 12:29 PM, Alex Bligh wrote: > > On 18 Apr 2014, at 03:11, Sangho Park wrote: > >>> >>> g_poll has a problem on windows when using timeouts < 10ms, in >>> glib/gpoll.c: >>> >>> /* If not, and we have a significant timeout, poll again with >>> * timeout then. N

Re: [Qemu-devel] [PATCH] timer: fix qemu_poll_ns early timeout on windows

2014-04-18 Thread Alex Bligh
On 18 Apr 2014, at 03:11, Sangho Park wrote: >> >> g_poll has a problem on windows when using timeouts < 10ms, in >> glib/gpoll.c: >> >> /* If not, and we have a significant timeout, poll again with >> * timeout then. Note that this will return indication for only >> * one event, or only for m

Re: [Qemu-devel] [PATCH] timer: fix qemu_poll_ns early timeout on windows

2014-04-18 Thread Stefan Weil
Am 18.04.2014 09:34, schrieb Stanislav Vorobiov: > Hi, > > Please see below > > On 04/18/2014 10:46 AM, Stefan Weil wrote: >> Hi, [...] > >> >> Would it be sufficient to round any timeout > 0 and < 10 to 10 for >> Windows hosts? Maybe this could be done in qemu_timeout_ns_to_ms. > We tried that,

Re: [Qemu-devel] [PATCH] timer: fix qemu_poll_ns early timeout on windows

2014-04-18 Thread Stanislav Vorobiov
biov [mailto:s.vorob...@samsung.com] >> Sent: Thursday, April 17, 2014 6:08 PM >> To: qemu-devel@nongnu.org >> Cc: syeon.hw...@samsung.com; sangho1206.p...@samsung.com >> Subject: Re: [Qemu-devel] [PATCH] timer: fix qemu_poll_ns early timeout on >> windows >> >

Re: [Qemu-devel] [PATCH] timer: fix qemu_poll_ns early timeout on windows

2014-04-17 Thread Stefan Weil
tion or mistake, let us know. We will appreciate your favor. > > -Original Message- > From: Stanislav Vorobiov [mailto:s.vorob...@samsung.com] > Sent: Thursday, April 17, 2014 6:08 PM > To: qemu-devel@nongnu.org > Cc: syeon.hw...@samsung.com; sangho1206.p...@samsung.com >

Re: [Qemu-devel] [PATCH] timer: fix qemu_poll_ns early timeout on windows

2014-04-17 Thread Sangho Park
ssage- From: Stanislav Vorobiov [mailto:s.vorob...@samsung.com] Sent: Thursday, April 17, 2014 6:08 PM To: qemu-devel@nongnu.org Cc: syeon.hw...@samsung.com; sangho1206.p...@samsung.com Subject: Re: [Qemu-devel] [PATCH] timer: fix qemu_poll_ns early timeout on windows Hi, everyone Any commen

Re: [Qemu-devel] [PATCH] timer: fix qemu_poll_ns early timeout on windows

2014-04-17 Thread Stanislav Vorobiov
Hi, everyone Any comments on this one ? This patch fixes pretty serious performance issues on windows, it would be great to have this in 2.0.0 On 04/15/2014 12:41 PM, Stanislav Vorobiov wrote: > From: Sangho Park > > g_poll has a problem on windows when using > timeouts < 10ms, in glib/gpoll.c:

[Qemu-devel] [PATCH] timer: fix qemu_poll_ns early timeout on windows

2014-04-15 Thread Stanislav Vorobiov
From: Sangho Park g_poll has a problem on windows when using timeouts < 10ms, in glib/gpoll.c: /* If not, and we have a significant timeout, poll again with * timeout then. Note that this will return indication for only * one event, or only for messages. We ignore timeouts less than * ten mil