__cxa_atexit and __cxa_finalize implementation

Stephen Warren swarren at wwwdotorg.org
Sun Sep 25 00:43:34 UTC 2005


Peter S. Mazinger wrote:
> On Fri, 23 Sep 2005, Stephen Warren wrote:
>>Here's another test case, that uses a DSO with global ctors/dtors loaded 
>>via dlopen().
>>
>>It works with the original completely unpatched uClibc, no doubt because 
>>global dtors on DSO unload are handled in some non-cxa_atexit fashion.
> 
> This should probably be adapted.
> When this patch goes upstream, looking at glibc
> libc-start.c: [rtld_]fini are registered w/ __cxa_atexit
> pthread.c: pthread_atexit_process(instead of pthread_onexit_process) 
> / pthread_atexit_retcode(havent found this or equiv. in uClibc) are 
> registered w/ __cxa_atexit (this is though a conditional HAVE_Z_NODELETE, 
> haven't looked further)

Hmm. I'll take a look at this once I've verified I can build everything!

-- 
Stephen Warren, Software Engineer, NVIDIA, Fort Collins, CO
swarren at wwwdotorg.org     http://www.wwwdotorg.org/pgp.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 251 bytes
Desc: OpenPGP digital signature
Url : http://lists.busybox.net/pipermail/uclibc/attachments/20050924/1a82dcf6/attachment.pgp 


More information about the uClibc mailing list