Re: 1.3.1 tarballs up for testing/signing (Again)
Philip Martin wrote on 3/28/2006 7:49 PM:
> Sviatoslav Sviridov <email@example.com> writes:
>> So with '--enable dso':
>> - all checks for FS_TYPE=fsfs passed
>> - checks for FS_TYPE=bdb passed except one: repos_test failed
>> - perl bindings tests passed except '6ra'
>> - all ruby tests passed
>> - javahl tests failed
> Is the repos_test error still the same:
> #0 0x3ae6bb67 in ?? ()
> #1 0x3ab57e8e in run_cleanups (cref=0x819ef30) at apr_pools.c:1952
> #2 0x3ab573d1 in apr_pool_destroy (pool=0x819ef20) at apr_pools.c:731
> #3 0x3ab572fb in apr_pool_clear (pool=0x80580f8) at apr_pools.c:691
> #4 0x3aac4b43 in main (argc=4, argv=0xaf9d7874) at
> #5 0x3abc949e in __libc_start_main () from /lib/libc.so.6
> #6 0x08049611 in _start ()
> That looks like a bug in the way pools are being used. Configuring
> APR with --enable-pool-debug=yes (and building with debug info) and
> then running repos_test under valgrind might help identify the
Yes, backtrace still points to apr_pools.c.
Ok, I'll try to rebuild APR with --enable-pool-debug=yes and debug info,
and then try to run it under valgring...
Can you show how to run in under valgrind (I'm not familiar with it).
Received on Tue Mar 28 18:56:27 2006
This is an archived mail posted to the Subversion Dev