diff options
| author | Jake Freeland <jfree@FreeBSD.org> | 2026-03-20 06:34:15 +0000 |
|---|---|---|
| committer | Jake Freeland <jfree@FreeBSD.org> | 2026-03-20 07:15:09 +0000 |
| commit | 33345b8834859cf1ad7f1256f3a34b4a7c83e8e8 (patch) | |
| tree | a2794324ce3e943f069c6274e87947f35c736b82 | |
| parent | b11fba41265ae132acfb05587a6dbf60334d3279 (diff) | |
timerfd: Suppress kqueue readability after jump read
Do not report EVFILT_READ after reading a discontinuous clock jump.
This makes the kqueue filter consistent with Linux epoll behavior
and timerfd_poll(), which already checks tfd_jumped != TFD_READ before
reporting POLLIN.
MFC after: 2 weeks
| -rw-r--r-- | sys/kern/sys_timerfd.c | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/sys/kern/sys_timerfd.c b/sys/kern/sys_timerfd.c index e4d00f224505..22e6a30faa21 100644 --- a/sys/kern/sys_timerfd.c +++ b/sys/kern/sys_timerfd.c @@ -290,7 +290,7 @@ filt_timerfdread(struct knote *kn, long hint) mtx_assert(&tfd->tfd_lock, MA_OWNED); kn->kn_data = (int64_t)tfd->tfd_count; - return (tfd->tfd_count > 0); + return (tfd->tfd_count > 0 && tfd->tfd_jumped != TFD_READ); } static const struct filterops timerfd_rfiltops = { |