|
From: Tom H. <to...@co...> - 2007-12-21 09:05:01
|
In message <e2e...@ma...>
Bart Van Assche <bar...@gm...> wrote:
> On Dec 20, 2007 10:31 PM, Tom Hughes <to...@co...> wrote:
>
>> Are you sure that's everything? My memory was that the problems
>> were with signal handlers and things, but maybe those have all
>> been fixed now?
>
> What I had noticed for signal handlers is that VG_(running_tid) was correct,
> but that VG_(running_tid) was changed before VG_TRACK(start_client_code)()
> was called. Instead of caching VG_(running_tid) upon
> VG_TRACK(start_client_code)(), exp-drd now queries VG_(running_tid) upon
> every memory access. This solved the issues I encountered with memory
> accesses from within signal handlers.
Surely if it was changed before VG_TRACK(start_client_code)() then it
would have been fine for it to cache it?
Did you mean that it was changed after that tracking call but before
the code in the handler was executed?
Tom
--
Tom Hughes (to...@co...)
http://www.compton.nu/
|