mbox series

[git,pull] epoll fixes

Message ID 20201002172025.GJ3421308@ZenIV.linux.org.uk (mailing list archive)
State New
Headers show
Series [git,pull] epoll fixes | expand


git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs.git work.epoll


Al Viro Oct. 2, 2020, 5:20 p.m. UTC
Several race fixes in epoll.

The following changes since commit 77f4689de17c0887775bb77896f4cc11a39bf848:

  fix regression in "epoll: Keep a reference on files added to the check list" (2020-09-02 11:30:48 -0400)

are available in the git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs.git work.epoll

for you to fetch changes up to 3701cb59d892b88d569427586f01491552f377b1:

  ep_create_wakeup_source(): dentry name can change under you... (2020-09-24 19:41:58 -0400)

Al Viro (4):
      epoll: do not insert into poll queues until all sanity checks are done
      epoll: replace ->visited/visited_list with generation count
      epoll: EPOLL_CTL_ADD: close the race in decision to take fast path
      ep_create_wakeup_source(): dentry name can change under you...

 fs/eventpoll.c | 72 +++++++++++++++++++++++++---------------------------------
 1 file changed, 31 insertions(+), 41 deletions(-)


pr-tracker-bot@kernel.org Oct. 2, 2020, 5:36 p.m. UTC | #1
The pull request you sent on Fri, 2 Oct 2020 18:20:25 +0100:

> git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs.git work.epoll

has been merged into torvalds/linux.git:

Thank you!
Linus Torvalds Oct. 2, 2020, 5:44 p.m. UTC | #2
On Fri, Oct 2, 2020 at 10:20 AM Al Viro <viro@zeniv.linux.org.uk> wrote:
>         Several race fixes in epoll.

Fudge. I screwed up the commit message due to a cut-and-paste error
(don't ask - sometimes google chrome and gnome-terminal seem to stop
agreeing about the normal X paste buffer)

And I extra stupidly pushed the thing out after the build succeeded,
not having noticed how I screwed up the trivial commit message.

I've force-updated the public sites, and I really hope nobody pulled
in that (very short) time when my tree had a bogus commit message.

(In case anybody cares, the commit message said "SEQCNT_MUTEX_ZERO"
instead of "Several race fixes in epoll" because that's what I had
looked at in another terminal. So it was a very WTF message)

I think this was only the second time I had a forced push to fix some
stupidity of mine. So it's not exactly _common_, but it's