On 03/09/2010 01:06 PM, Sven Joachim wrote:
On 2010-03-09 10:29 +0100, eric2.vale...@orange-ftgroup.com wrote:
On 03/09/2010 10:18 AM, VALETTE Eric RD-MAPS-REN wrote:
Something along symlinks:malloc assert failed and a absolutely
unreadable formula. Note that compiling the same tree with a level
link less (I now have a symlink for /scratch were previously I had
a real mounted fs was working.
Note that if there is a malloc/free problem, using valgrind could
potentially highlight where in the debian version...
This might be useful, but it is you who has to do it since I suspect
that the problem is not easily reproducible by me or anyone else.
That's why I suggested to ignore the description and simply take the new
version.
Here is the alway reproducible error (assert)
symlinks -r -d
/home/ceva6380/scratch/sahtree1/sah/gwmiddleware/REL/2009-11-27_V3.29.0/cm4/fsn
/dev/null
symlinks: malloc.c:3096: sYSMALLOc: Assertion `(old_top == (((mbinptr)
(((char *)&((av)->bins[((1) - 1) * 2])) - __builtin_offsetof (struct
malloc_chunk, fd))))&& old_size == 0) || ((unsigned long) (old_size)
= (unsigned long)((((__builtin_offsetof (struct malloc_chunk,
fd_nextsize))+((2 * (sizeof(size_t))) - 1))& ~((2 * (sizeof(size_t)))
-
1)))&& ((old_top)->size& 0x1)&& ((unsigned long)old_end& pagemask)
== 0)' failed.
/bin/sh: line 1: 3824 Aborted symlinks -r -d
/home/ceva6380/scratch/sahtree1/sah/gwmiddleware/REL/2009-11-27_V3.29.0/cm4/fsn
/dev/null
Can you please run an unstripped version of symlinks under GDB and get a
backtrace? Either rebuild the package with DEB_BUILD_OPTS=nostrip or
run the symlinks binary directly from the source tree. Please also
install the libc6-dbg package.
No. I have a working and already proposed a way to fix it. If you want
to know if the bug is in the original 1.2 or in the debian specific
patches, You can packages 1.4 with all the fixes and I will retest it.
--eric
