diff options
| author | Jan Kiszka <jan.kiszka@siemens.com> | 2011-06-17 11:25:49 +0200 |
|---|---|---|
| committer | Anthony Liguori <aliguori@us.ibm.com> | 2011-07-23 11:26:12 -0500 |
| commit | d25f89c9e91d6c46b85969922411a211a6347a7d (patch) | |
| tree | 6ef37246ff23527e8bc62fcf4f271bded16619d3 /qemu-timer.c | |
| parent | 17604dac28b2410c021a4a52dcfa58e8803dfb24 (diff) | |
Register Linux dyntick timer as per-thread signal
Derived from kvm-tool patch
http://thread.gmane.org/gmane.comp.emulators.kvm.devel/74309
Ingo Molnar pointed out that sending the timer signal to the whole
process, just blocking it everywhere, is suboptimal with an increasing
number of threads. QEMU is also using this pattern so far.
Linux provides a (non-portable) way to restrict the signal to a single
thread: We can use SIGEV_THREAD_ID unless we are forced to emulate
signalfd via an additional thread. That case could theoretically be
optimized as well, but it doesn't look worth bothering.
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Diffstat (limited to 'qemu-timer.c')
| -rw-r--r-- | qemu-timer.c | 8 |
1 files changed, 8 insertions, 0 deletions
diff --git a/qemu-timer.c b/qemu-timer.c index f95374c7c1..30e8f1272e 100644 --- a/qemu-timer.c +++ b/qemu-timer.c @@ -831,6 +831,8 @@ static int64_t qemu_next_alarm_deadline(void) #if defined(__linux__) +#include "compatfd.h" + static int dynticks_start_timer(struct qemu_alarm_timer *t) { struct sigevent ev; @@ -850,6 +852,12 @@ static int dynticks_start_timer(struct qemu_alarm_timer *t) memset(&ev, 0, sizeof(ev)); ev.sigev_value.sival_int = 0; ev.sigev_notify = SIGEV_SIGNAL; +#ifdef SIGEV_THREAD_ID + if (qemu_signalfd_available()) { + ev.sigev_notify = SIGEV_THREAD_ID; + ev._sigev_un._tid = qemu_get_thread_id(); + } +#endif /* SIGEV_THREAD_ID */ ev.sigev_signo = SIGALRM; if (timer_create(CLOCK_REALTIME, &ev, &host_timer)) { |