[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-ha-dev
Subject:    Re: [Linux-ha-dev] segfaults when defined MARK_PRISTINE
From:       Alan Robertson <alanr () unix ! sh>
Date:       2007-02-20 23:34:08
Message-ID: 45DB8570.9000404 () unix ! sh
[Download RAW message or body]

Pavol Gono wrote:
> Hi
> 
> I wanted to play with memory allocation a little.
> I added #define HA_MALLOC_TRACK 1 in include/clplumbing/cl_malloc.h
> and removed #undef  MARK_PRISTINE in lib/clplumbing/cl_malloc.c
> 
> After compilation the heartbeat processes segfaulted very soon with
> this backtrace:
> (gdb) bt
> 
>                             #0  0xa7c8a547 in memset () from
> /lib/tls/i686/cmov/libc.so.6
>                                                                   #1
> 0x00001478 in ?? ()
> 
>                         #2  0xa7f38fb8 in cl_mark_pristine
> (v=0x80e9458, size=64) at cl_malloc.c:1020
>                                                                 #3
> 0xa7f38249 in cl_free (ptr=0x80e9458) at cl_malloc.c:604
> 
>                         #4  0xa7f38eb8 in cl_free_glib
> (mem=0x80e9458) at cl_malloc.c:970
> 
> #5  0xa7ec0b31 in g_free () from /usr/lib/libglib-2.0.so.0
> 
>                             #6  0xa7f539cf in NewPILPluginUniv
> (basepluginpath=0x807265c "/usr/local/lib/heartbeat/plugins") at
> pils.c:484
> #7  0x0806775f in module_init () at module.c:117
> 
>                             #8  0x08056087 in main (argc=1,
> argv=0xafa4cb84, envp=0xafa4cb8c) at heartbeat.c:4527
> 
> Is the macro MARK_PRISTINE considered to be unusable? Or shall I use
> other configure options?
> 
> revision 10091, my configure options:
> --with-group-name=haclient --with-ccmuser-name=hacluster
> CFLAGS='-fno-unit-at-a-time -g -O0' --sysconfdir=/etc
> --localstatedir=/var --disable-bundled_ltdl --disable-fatal-warnings
> --disable-tipc --disable-ldirectord --disable-snmp --disable-mgmt
> --disable-quorumd --disable-dopd --disable-static --enable-crm-dev
> 
> When I let the MARK_PRISTINE undefined, no segfaults.

It has worked correctly in the last year or two.  I haven't tried it
recently though.  Lars did a lot of work in the malloc code (the
MALLOC_TRACK stuff), and I haven't tried it since then.

Does it work if MALLOC_TRACK is disabled?


-- 
    Alan Robertson <alanr@unix.sh>

"Openness is the foundation and preservative of friendship...  Let me
claim from you at all times your undisguised opinions." - William
Wilberforce
_______________________________________________________
Linux-HA-Dev: Linux-HA-Dev@lists.linux-ha.org
http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
Home Page: http://linux-ha.org/
[prev in list] [next in list] [prev in thread] [next in thread]