This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: libc core dump



Olenin,

Have you looked for patches from Sun for solaris 2.6 that hit libc? Sun has
been upgrading it's threads concept in the various releases of solaris. I
have run across several vendors that REQUIRE solaris 2.8 because they say,
"earlier threads code in solaris DOES NOT WORK CORRECTLY in all cases".

Good luck.

George...





"Olenin, Vladimir" <vladimir.olenin@bgminteractive.com>@gcc.gnu.org on
01/23/2002 12:00:54 PM

Sent by:  gcc-owner@gcc.gnu.org


To:   "'gcc@gcc.gnu.org'" <gcc@gcc.gnu.org>
cc:

Subject:  libc core dump


Hi,

I'm having some really weird problems using shared library from multiple
Java threads. The deal is that from time to time during stress test I'm
getting core dump from libc (it's always either "strlen" or "atoi" call).
Partly this problem was eliminated by upgrading from 2.8 to 3.0.3 compiler,
but it still occurs in some situations.

Does anyone has an idea what might be causing this error? I've tried
various
flags to compile (-pthread, etc, etc), but the problems hasn't dissappeared
completely. Might it be because I'm using precompiled gcc (is it compiled
with --enable-threads flag)? Or some particular flags should be used
beyound
-pthread? Could it be because of X Window libraries we r linking with? Or
it
can be the problem of calling native library from MT Java application
(though various docs say it's completely safe providing that native library
is MT-safe)?

System configuration is Solaris 2.6, gcc 3.0.3.

Any help would be greately appreciated! Thanks in advance,

vladimir




Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]