[Gc] Reproducible bug report for thread registration

Juan Jose Garcia-Ripoll juanjose.garciaripoll at gmail.com
Wed Oct 31 11:48:00 PST 2012


On Wed, Oct 31, 2012 at 4:47 PM, Ivan Maidanski <ivmai at mail.ru> wrote:

>
> To my understanding, it should crash in case pthread_create has been
> intercepted (because key_buffer is NULL).
>

The makefile I sent you has been designed such that pthread_create is not
intercepted (the only real place where thread registration makes sense),
because that is what is causing trouble: people with third-party software
that uses standard thread creation routines and invokes the ECL library
which itself registers the calling threads.

On Wed, Oct 31, 2012 at 5:13 PM, Ivan Maidanski <ivmai at mail.ru> wrote:

> Your usage pattern of GC_register_my_thread is wrong (leading to unstable
> app behavior),
>

Thanks for the clarification. I obviously did not understand the gc.h
header and has been doing so for a long time. Definitely the collector
needs a manual.

Juanjo

-- 
Instituto de Física Fundamental, CSIC
c/ Serrano, 113b, Madrid 28006 (Spain)
http://juanjose.garciaripoll.googlepages.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://napali.hpl.hp.com/pipermail/gc/attachments/20121031/79fb77ea/attachment.htm


More information about the Gc mailing list