On a centos 4.6 box
rpm -qa | grep glib
glib-1.2.10-15 glibc-2.3.4-2.39 glibc-headers-2.3.4-2.39 glibc-common-2.3.4-2.39 dbus-glib-0.22-12.EL.9 glibc-kernheaders-2.4-9.1.100.EL glibc-devel-2.3.4-2.39 glib2-2.4.7-1
I am getting this type of error at times on a mail server
*** glibc detected *** free(): invalid pointer: 0xbffa11b8 ***
While I am researching this more, I am wondering if the rpm-qa | grep glib info above showing different version for kernheaders makes a difference in any way.
There are lots of search hits out there for this yet I am wondering if I should be focusing on a certain glib area or in the application providing the error in the error logs
- rh
Robert - elists wrote:
On a centos 4.6 box
rpm -qa | grep glib
glib-1.2.10-15 glibc-2.3.4-2.39 glibc-headers-2.3.4-2.39 glibc-common-2.3.4-2.39 dbus-glib-0.22-12.EL.9 glibc-kernheaders-2.4-9.1.100.EL glibc-devel-2.3.4-2.39 glib2-2.4.7-1
I am getting this type of error at times on a mail server
*** glibc detected *** free(): invalid pointer: 0xbffa11b8 ***
While I am researching this more, I am wondering if the rpm-qa | grep glib info above showing different version for kernheaders makes a difference in any way.
There are lots of search hits out there for this yet I am wondering if I should be focusing on a certain glib area or in the application providing the error in the error logs
I'd be looking at the application that triggered that error. Apparently, it tried to 'free()' a memory block that wasn't malloc()'d in the first place.
I'd be looking at the application that triggered that error. Apparently, it tried to 'free()' a memory block that wasn't malloc()'d in the first place. _______________________________________________
Thank you John and Ignacio and others,
So, since this box has been online for like 2 or 3 years, since like centos 4.0 or 4.1 load with multiple yum updates, I am guessing a recompile of that app should make the difference?
;->
Or is that way off base?
I really couldn't imagine finding the bad code yet ill try if necessary
- rh
On Fri, 2008-03-21 at 11:27 -0700, Robert - elists wrote:
So, since this box has been online for like 2 or 3 years, since like centos 4.0 or 4.1 load with multiple yum updates, I am guessing a recompile of that app should make the difference?
No. You need to find out where the memory is being double-freed and remove the appropriate free() call. That or set it to NULL once freed.
On Fri, 2008-03-21 at 11:11 -0700, Robert - elists wrote:
*** glibc detected *** free(): invalid pointer: 0xbffa11b8 ***
... I am wondering if I should be focusing on a certain glib area or in the application providing the error in the error logs
Application. This error happens when an app does a double free.
-
Ignacio Vazquez-Abrams -
John R Pierce -
Robert - elists