This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH] Memory fencing problem in pthread cancellation
On Thu, Jan 17, 2013 at 04:39:26PM +0000, Joseph S. Myers wrote:
> On Wed, 16 Jan 2013, Rich Felker wrote:
>
> > I've seen that the unwinder goes to great lengths (I would call it
> > "hideous hacks") for stuff related to signal trampolines, and as far
> > as I can tell, the only reason it would ever be needed is for
> > unwinding out of signal handlers, which is DEEP in the realm of UB and
> > not something I can see any justification for trying to support. Is
> > there another reason the signal trampolines need special consideration
> > that I'm not seeing?
>
> Unwinding is not just about things defined at language level such as C++
> exceptions (and thread cancellation). Accurate unwinding from signal
> handlers can be used for other things such as profiling (see the Froyd et
> al paper in the 2006 GCC Summit proceedings).
Thanks. Yes, I see how it can be useful for profiling and debugging.
But I'm not sure I see how it matters whether the program's own
unwinder is capable of doing this; as long as the profiling/debugging
tool's is capable of unwinding across signal handlers, that should
suffice.
Rich