rhythmbox crashes on startup if you run it with MALLOC_PERTURB_="149" (gdb) bt #0 0x0000003aaa2728c0 in strcmp () from /lib64/libc.so.6 #1 0x00002b871c64929d in g_str_equal () from /usr/lib64/libglib-2.0.so.0 #2 0x00002b871c6268f2 in g_hash_table_lookup () from /usr/lib64/libglib-2.0.so.0 #3 0x000000000045b862 in rhythmdb_idle_poll_events (db=0x926000) at rhythmdb.c:1409 #4 0x00002b871c630e91 in g_main_context_dispatch () from /usr/lib64/libglib-2.0.so.0 #5 0x00002b871c633cad in g_main_context_check () from /usr/lib64/libglib-2.0.so.0 #6 0x00002b871c633ffe in g_main_loop_run () from /usr/lib64/libglib-2.0.so.0 #7 0x00002b871b9c7904 in bonobo_main () from /usr/lib64/libbonobo-2.so.0 #8 0x00000000004276a5 in main (argc=1, argv=0x7fffffb8de68) at main.c:398
I've been tracking rhythmbox lately, will have a look in the morning.
Which version are you using? I just did: export MALLOC_PERTURB_="149" rhythmbox using rhythmbox-0.9.2.cvs20060123-1 both my i386 laptop and my x86_64 workstation (64bit) and neither of them crashed...
Also, I just built a newer CVS snapshot (0.9.2.cvs20060201) which had a 64bit crash fix in the changelog since the last snapshot i pulled. See if that fixes it.
I was also running rhythmbox-0.9.2.cvs20060123-1 It crashed whilst it was scanning new oggs that I had added, so it may have only trigger if you have awesome tunes. I'll try out the new build when I get home.
actually, it was my work machine that died. Just tested latest build, and it seems to work. Whee, I have noises again.