[valgrind] [Bug 492678] SIGALRM race condition (sighandler called after timer disarmed)

[Paul Floyd]

https://bugs.kde.org/show_bug.cgi?id=492678

--- Comment #4 from Paul Floyd <pjfl...@wanadoo.fr> ---
This works for me on FreeBSD

diff --git a/coregrind/m_syswrap/syswrap-freebsd.c
b/coregrind/m_syswrap/syswrap-freebsd.c
index d358c90eb..339724938 100644
--- a/coregrind/m_syswrap/syswrap-freebsd.c
+++ b/coregrind/m_syswrap/syswrap-freebsd.c
@@ -2411,6 +2411,7 @@ PRE(sys_timer_delete)
 {
    PRINT("sys_timer_delete( %#" FMT_REGWORD "x )", ARG1);
    PRE_REG_READ1(long, "timer_delete", vki_timer_t, timerid);
+   *flags |= SfPollAfter;
 }

 // SYS_ktimer_settime   237
diff --git a/coregrind/m_syswrap/syswrap-generic.c
b/coregrind/m_syswrap/syswrap-generic.c
index bfe4c6fe0..b76013dc8 100644
--- a/coregrind/m_syswrap/syswrap-generic.c
+++ b/coregrind/m_syswrap/syswrap-generic.c
@@ -2967,6 +2967,7 @@ PRE(sys_setitimer)
       PRE_timeval_WRITE( "setitimer(&ovalue->it_value)",
                          &(ovalue->it_value));
    }
+   *flags |= SfPollAfter;
 }

 POST(sys_setitimer)

For this testcase, it's the setitimer that seems to matter. Not sure if that
should be for all values or just for a new time of 0 to turn off the itimer.

-- 
You are receiving this mail because:
You are watching all bug changes.