[Gc] Segfaults in GC_core_malloc

Andrei Golovizin andrei.golovizin at tigress.ru
Tue Feb 13 00:28:32 PST 2007


Hello,

I am currently porting some old (and apparently broken) program to libgc.
I am using libgc.so built with --enable-redirect-malloc. Unfortunately the 
program segfaults quite often. Here is the example backtrace:

-------8<----------------------------------------------------------------------
#0  GC_core_malloc (lb=64) at malloc.c:280
#1  0x080ebe08 in GC_malloc (bytes=64) at thread_local_alloc.c:150
#2  0x080e6da6 in GC_realloc (p=0x90ea320, lb=64) at mallocx.c:130
#3  0x001a92a2 in XShrinkRegion () from /usr/X11R6/lib/libX11.so.6
#4  0x001a8ad9 in XOffsetRegion () from /usr/X11R6/lib/libX11.so.6
#5  0x001a8c7b in XUnionRegion () from /usr/X11R6/lib/libX11.so.6
#6  0x001a8dd4 in XUnionRectWithRegion () from /usr/X11R6/lib/libX11.so.6
#7  0x00137047 in XtAddExposureToRegion () from /usr/X11R6/lib/libXt.so.6
#8  0x001376a6 in XtDispatchEventToWidget () from /usr/X11R6/lib/libXt.so.6
#9  0x00137d5f in _XtOnGrabList () from /usr/X11R6/lib/libXt.so.6
#10 0x001380c9 in XtDispatchEvent () from /usr/X11R6/lib/libXt.so.6
#11 0x08213375 in pxinpt_ ()
#12 0x0806fb7e in egw_app_shdn (subnam=0xbff2ea70 "pxinpt_", ier=0xbff2ea2c) 
at egw_app_subs.c:395
#13 0x0806f20a in egw_app_strt (argc=1, argv=0xbff2eb34, 
subnam=0xbff2ea70 "pxinpt_", ier=0xbff2ea2c) at egw_app_subs.c:109
#14 0x08051fd3 in main (argc=1, argv=0xbff2eb34) at ebh001.c:111
-------8<----------------------------------------------------------------------

I've also tried emulating libgc behaviour by simply replacing free() with a 
do-nothing function instead of using real libgc. In this case segfaults never 
occur (although the memory leaks badly).

Is such behaviour of GC expected? If yes, is it possible to know, what is 
causing it? I'm completely stuck now and will appreciate any help. :)

Andrey


More information about the Gc mailing list