bug-glibc
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: linuxthreads/malloc problem


From: Reto Baettig
Subject: Re: linuxthreads/malloc problem
Date: Thu, 02 Nov 2000 06:31:47 -0800

Hi!

I attached the test program.

As I said in my first email, I have the following configuration:

Linux reto0 2.2.14bapper #13 SMP Tue Sep 19 21:26:52 PDT 2000 alpha
unknown
Kernel modules         2.3.10-pre1
Gnu C                  egcs-2.91.66
Binutils               2.9.5.0.22
Linux C Library        2.1.3
Dynamic linker         ldd (GNU libc) 2.1.3

The program works fine on Intel UP. I could not test it on Intel SMP
yet.
Alpha SMP (Linux 2.2.14) shows the problems after less than one minute 
if you start 2 instances. One instance seems to work fine.

I try to test it with glibc2.2 now.

Thanks for your time,

Reto

Charles G Waldman wrote:
> 
> Wolfram Gloger writes:
> 
>  > <address@hidden> writes:
>  > > I have problems with malloc/free when I use it in my pthread programs. I
>  > > keep getting segfaults when I start 2 instances of the program.
>  >
>  > Huh?  That would indicate either a kernel or a hardware problem.
> 
> Is this on a Linux machine?  What kernel version?
> 
> I was part of the group that worked on Python2.0.  There was some code
> in the Python core which used pthread mutexes.  The test suite which
> exercised this code would exhibit misbehavior, but only on Linux.
> There were also anecodal reports that the behavior would differ if you
> had multiple instances of the program running. It was speculated, but
> not proven, that the problems we were seeing might be due to an
> underlying Linux kernel bug.  Note, this is all highly speculative,
> nobody managed to prove it!  We wound up simply changing the code in
> question, using a different approach which did not trigger the
> problem.  But it is a continuing task to try to better understand what
> was happening, so I am very interested in hearing about any problems
> with Pthreads which are really due to Linux kernel flaws.
> 
>  > > I wrote a little test program that demonstrates the problems on my
>  > > machine. When I remove the malloc/free's, everything runs fine. I can
>  > > send it to you when you think it helps.
> 
> I would like to see this program.
> 
> _______________________________________________
> Bug-glibc mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/bug-glibc

Attachment: malloctest.tgz
Description: Binary data


reply via email to

[Prev in Thread] Current Thread [Next in Thread]