Re: [RFC PATCH] epoll: Add synchronous wakeup support for ep_poll_callback

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Thu, Dec 12, 2024 at 12:14 PM Brian Geffon <bgeffon@xxxxxxxxxx> wrote:
>
> On Wed, Oct 16, 2024 at 03:05:38PM -0400, Brian Geffon wrote:
> > On Wed, Oct 16, 2024 at 03:10:34PM +0200, Christian Brauner wrote:
> > > On Fri, 26 Apr 2024 16:05:48 +0800, Xuewen Yan wrote:
> > > > Now, the epoll only use wake_up() interface to wake up task.
> > > > However, sometimes, there are epoll users which want to use
> > > > the synchronous wakeup flag to hint the scheduler, such as
> > > > Android binder driver.
> > > > So add a wake_up_sync() define, and use the wake_up_sync()
> > > > when the sync is true in ep_poll_callback().
> > > >
> > > > [...]
> > >
> > > Applied to the vfs.misc branch of the vfs/vfs.git tree.
> > > Patches in the vfs.misc branch should appear in linux-next soon.
> > >
> > > Please report any outstanding bugs that were missed during review in a
> > > new review to the original patch series allowing us to drop it.
> > >
> > > It's encouraged to provide Acked-bys and Reviewed-bys even though the
> > > patch has now been applied. If possible patch trailers will be updated.
> > >
> > > Note that commit hashes shown below are subject to change due to rebase,
> > > trailer updates or similar. If in doubt, please check the listed branch.
> > >
> > > tree:   https://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs.git
> > > branch: vfs.misc
> >
> > This is a bug that's been present for all of time, so I think we should:
> >
> > Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
> > Cc: stable@xxxxxxxxxxxxxxx
>
> This is in as 900bbaae ("epoll: Add synchronous wakeup support for
> ep_poll_callback"). How do maintainers feel about:
>
> Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
> Cc: stable@xxxxxxxxxxxxxxx

Dear stable maintainers, this fixes a bug goes all the way back and
beyond Linux 2.6.12-rc2. Can you please add this commit to the stable
releases?

commit 900bbaae67e980945dec74d36f8afe0de7556d5a upstream.

>
> >
> > I sent a patch which adds a benchmark for nonblocking pipes using epoll:
> > https://lore.kernel.org/lkml/20241016190009.866615-1-bgeffon@xxxxxxxxxx/
> >
> > Using this new benchmark I get the following results without this fix
> > and with this fix:
> >
> > $ tools/perf/perf bench sched pipe -n
> > # Running 'sched/pipe' benchmark:
> > # Executed 1000000 pipe operations between two processes
> >
> >      Total time: 12.194 [sec]
> >
> >       12.194376 usecs/op
> >           82005 ops/sec
> >
> >
> > $ tools/perf/perf bench sched pipe -n
> > # Running 'sched/pipe' benchmark:
> > # Executed 1000000 pipe operations between two processes
> >
> >      Total time: 9.229 [sec]
> >
> >        9.229738 usecs/op
> >          108345 ops/sec
> >
> > >
> > > [1/1] epoll: Add synchronous wakeup support for ep_poll_callback
> > >       https://git.kernel.org/vfs/vfs/c/2ce0e17660a7
>
> Thanks,
> Brian
>

 Thanks,
 Brian





[Index of Archives]     [Linux Kernel]     [Kernel Development Newbies]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite Hiking]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux