[PATCH] os: make timers signal-safe

[Chase Douglas]

On 04/26/2012 07:47 PM, Peter Hutterer wrote:
> Block signals for all list manipulations in the timers.
> 
> If TimerSet() is called from a signal handler (synaptics tap handling code)
> may result in list corruption if we're currently inside TimerSet().
> 
> See backtrace in
> https://bugzilla.redhat.com/show_bug.cgi?id=814869
> 
> Signed-off-by: Peter Hutterer <peter.hutterer at who-t.net>
> ---
>  os/WaitFor.c |   18 ++++++++++++++++--
>  1 file changed, 16 insertions(+), 2 deletions(-)
> 
> diff --git a/os/WaitFor.c b/os/WaitFor.c
> index 95e64ba..393890f 100644
> --- a/os/WaitFor.c
> +++ b/os/WaitFor.c
> @@ -382,6 +382,7 @@ CheckAllTimers(void)
>      OsTimerPtr timer;
>      CARD32 now;
>  
> +    OsBlockSignals();
>   start:
>      now = GetTimeInMillis();
>  
> @@ -391,6 +392,7 @@ CheckAllTimers(void)
>              goto start;
>          }
>      }
> +    OsReleaseSignals();
>  }
>  
>  static void
> @@ -398,11 +400,13 @@ DoTimer(OsTimerPtr timer, CARD32 now, OsTimerPtr *prev)
>  {
>      CARD32 newTime;
>  
> +    OsBlockSignals();
>      *prev = timer->next;
>      timer->next = NULL;
>      newTime = (*timer->callback) (timer, now, timer->arg);
>      if (newTime)
>          TimerSet(timer, 0, newTime, timer->callback, timer->arg);
> +    OsReleaseSignals();
>  }
>  
>  OsTimerPtr
> @@ -418,6 +422,7 @@ TimerSet(OsTimerPtr timer, int flags, CARD32 millis,
>              return NULL;
>      }
>      else {
> +        OsBlockSignals();
>          for (prev = &timers; *prev; prev = &(*prev)->next) {
>              if (*prev == timer) {
>                  *prev = timer->next;
> @@ -426,6 +431,7 @@ TimerSet(OsTimerPtr timer, int flags, CARD32 millis,
>                  break;
>              }
>          }
> +        OsReleaseSignals();
>      }
>      if (!millis)
>          return timer;
> @@ -445,26 +451,32 @@ TimerSet(OsTimerPtr timer, int flags, CARD32 millis,
>          if (!millis)
>              return timer;
>      }
> +    OsBlockSignals();
>      for (prev = &timers;
>           *prev && (int) ((*prev)->expires - millis) <= 0;
>           prev = &(*prev)->next);
>      timer->next = *prev;
>      *prev = timer;
> +    OsReleaseSignals();
>      return timer;
>  }
>  
>  Bool
>  TimerForce(OsTimerPtr timer)
>  {
> +    int rc = FALSE;
>      OsTimerPtr *prev;
>  
> +    OsBlockSignals();
>      for (prev = &timers; *prev; prev = &(*prev)->next) {
>          if (*prev == timer) {
>              DoTimer(timer, GetTimeInMillis(), prev);
> -            return TRUE;
> +            rc = TRUE;
> +            break;
>          }
>      }
> -    return FALSE;
> +    OsReleaseSignals();
> +    return rc;
>  }
>  
>  void
> @@ -474,12 +486,14 @@ TimerCancel(OsTimerPtr timer)
>  
>      if (!timer)
>          return;
> +    OsBlockSignals();
>      for (prev = &timers; *prev; prev = &(*prev)->next) {
>          if (*prev == timer) {
>              *prev = timer->next;
>              break;
>          }
>      }
> +    OsReleaseSignals();
>  }
>  
>  void

Ewwww..... We've been seeming what looks like random corruption in the X
server. Hopefully this fixes it all.

Reviewed-by; Chase Douglas <chase.douglas at canonical.com>