Re: [Valgrind-users] Problem running valgrind stored in Rational ClearCase

[Dan K. <da...@ke...>]

Dominik Stadler wrote:
> I don't think valgrind is responsible, but I thought I'd report it anyway,
> maybe you can add sort of a "Known Issue" somewhere.
> 
> When I try to run the valgrind-binaries stored in the Rational/IBM
> ClearCase version control-system, valgrind does not work any more:
> ...
> ==24043== Warning: segment-override prefix encountered, but thread has no LDT
> ==24043== Warning: segment access: virtual addr 0 exceeds segment limit of 0
> ==24043== Invalid read of size 4
> ==24043==    at 0x40249EF4: __pthread_alt_lock (spinlock.c:446)
> ==24043==    by 0x40246D25: __pthread_mutex_lock (mutex.c:120)
> ==24043==    by 0x403AE6D0: __register_frame_info (in /usr/lib/libstdc++-3-libc6.2-2-2.10.0.so)
> ==24043==    by 0x40021A61: (within /.automount/dcaclust2/root/dcaclearcase/VOBS/eParty.vbs/c/cdft/11/41/ee7ef167450311d89c82000180e4a89e)
> ==24043==    by 0x400217D0: (within /.automount/dcaclust2/root/dcaclearcase/VOBS/eParty.vbs/c/cdft/11/41/ee7ef167450311d89c82000180e4a89e)
> ==24043==    by 0x4000DAA6: _dl_init (dl-init.c:70)
> ==24043==    by 0x40001ED0: (within /lib/ld-2.2.4.so)
> ==24043==  Address 0x0 is not stack'd, malloc'd or free'd
> ...
> when the binaries are not stored in ClearCase, valgrind works perfectly.

ClearCase is extremely intrusive.  You have just run into a case
where ClearCase's strange architecture has bitten you.  Because
it is closed source, there's not a lot the valgrind team can
do about it.  Feel free to report this as a bug to ClearCase.
And please consider using a less insane version control system,
e.g. perforce or bitkeeper.
- Dan