Re: [PATCH 1/2] Support TIF_RESTORE_SIGMASK on x86_64

[David Woodhouse]

On Mon, 2006-07-03 at 03:19 -0700, Andrew Morton wrote:
> Could you please describe the signal mask fix? 

@@ -583,7 +583,7 @@
        if (!user_mode(regs))
                return;

-       if (!test_thread_flag(TIF_RESTORE_SIGMASK))
+       if (test_thread_flag(TIF_RESTORE_SIGMASK))
                oldset = &current->saved_sigmask;
        else
                oldset = &current->blocked;


>  Is that likely to have caused the above symptoms?

Yeah, it'll screw up the signal mask all over the place. 
cf. https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=180567
also https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=179228

> Should we be setting TASK_INTERRUPTIBLE before releasing that lock?

Before releasing current->sighand->siglock? Nah, schedule() will check
for pending signals -- it's not like racing with a wake_up() 

-- 
dwmw2

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/