TiEmu Carsh en changeant de ROM - Page 1

J'étais sous AMS, puis j'ai voulu repasser sous PedroM :
F12 -> sélection de ROM.
Je double-clic sur ma ROM de PedroM, et là ça a crashé directement :

[Thread debugging using libthread_db enabled]
[New Thread 0xb5d376c0 (LWP 4239)]
[KCrash handler]
#6  0xb63d4727 in memset () from /lib/i686/cmov/libc.so.6
#7  0xb7ebfe28 in dbus_decomp (
    filename=0x994eaa8 "/home/folco/.ticables/ticables-log", resync=0)
    at hex2dbus.c:148
#8  0xb7ebf486 in log_dbus_stop () at log_dbus.c:69
#9  0xb7eb98da in log_stop (h=0x996da28) at data_log.c:116
#10 0xb7ebf298 in ticables_cable_close (handle=0x996da28) at link_xxx.c:138
#11 0xb7efcb35 in ticalcs_cable_detach (handle=0x99b1c80) at ticalcs.c:294
#12 0x0811b69b in hw_dbus_exit () at core/ti_hw/dbus.c:194
#13 0x0811ce25 in hw_exit () at core/ti_hw/hw.c:140
#14 0x0811877c in ti68k_exit () at core/interface.c:170
#15 0x0812e815 in main (argc=1275081518, argv=0x206b6e69) at gui/main.c:395

Je suis sous une Debian Lenny, je file toutes les infos dont ya besoin à la demande.

Je n'arrive pas à reproduire. D'avantage de détails pourrait aider.

C'est la première fois que ça m'arrive. Jamais eu de souci auparavant...
J'ai installé ce TiEmu avec le script de Lionel.

Ca m'arrive au deuxième changement AMS->PedroM->AMS apparemment.
Le log en console n'est pas très explicite, mais c'est un segfault.

tiemu-INFO: Scrutation des images/màj...
tiemu-INFO: Fait.
KCrash: Application 'tiemu' crashing...
Erreur de segmentation
folco@desktop:~$

Pour que ça crashe dans memset, ça doit être qu'au moins un des deux pointeurs pointe vers n'importe quoi, et/ou que la taille est trop grande et permet de déborder sur une zone protégée... mais pourquoi

Valgrind me fait la gueule :

==4568==
==4568== More than 10000000 total errors detected. I'm not reporting any more.
==4568== Final error counts will be inaccurate. Go fix your program!
==4568== Rerun with --error-limit=no to disable this cutoff. Note
==4568== that errors may occur in your program without prior warning from
==4568== Valgrind, because errors are no longer being displayed. ==4568==

Ceci dit, ya visiblement des trucs rigolos un peu partout :

==4568== Conditional jump or move depends on uninitialised value(s)
==4568==    at 0x5AD18C2: vfprintf (in /lib/i686/cmov/libc-2.7.so)
==4568==    by 0x5AF7FFB: vasprintf (in /lib/i686/cmov/libc-2.7.so)
==4568==    by 0x48E9276: g_vasprintf (in /usr/lib/libglib-2.0.so.0.1600.6)
==4568==    by 0x48D55D5: g_strdup_printf (in /usr/lib/libglib-2.0.so.0.1600.6)
==4568==    by 0x8143FFF: clist_populate (dbg_pclog.c:103)
==4568==    by 0x81443DD: dbgpclog_create_window (dbg_pclog.c:147)
==4568==    by 0x81378B8: gtk_debugger_preload (dbg_all.c:76)
==4568==    by 0x812E7A2: main (main.c:343)
==4568==
==4568== Invalid read of size 1
==4568==    at 0x8122AC6: iodefs_load (iodefs.c:147)
==4568==    by 0x814051B: ctree_populate (dbg_iop.c:336)
==4568==    by 0x814107A: dbgiop_create_window (dbg_iop.c:439)
==4568==    by 0x81378D6: gtk_debugger_preload (dbg_all.c:79)
==4568==    by 0x812E7A2: main (main.c:343)
==4568==  Address 0x78de724 is 0 bytes after a block of size 12 alloc'd
==4568==    at 0x4023D6E: malloc (vg_replace_malloc.c:207)
==4568==    by 0x48BB5A3: g_malloc (in /usr/lib/libglib-2.0.so.0.1600.6)
==4568==    by 0x48D5980: g_strndup (in /usr/lib/libglib-2.0.so.0.1600.6)
==4568==    by 0x48D5CAF: g_strsplit (in /usr/lib/libglib-2.0.so.0.1600.6)
==4568==    by 0x8122773: iodefs_load (iodefs.c:315)
==4568==    by 0x814051B: ctree_populate (dbg_iop.c:336)
==4568==    by 0x814107A: dbgiop_create_window (dbg_iop.c:439)
==4568==    by 0x81378D6: gtk_debugger_preload (dbg_all.c:79)
==4568==    by 0x812E7A2: main (main.c:343)

hmmm... y'aurait de la mémoire qui partirait en fumée ?

==4568==
==4568== ERROR SUMMARY: 10000000 errors from 14 contexts (suppressed: 213 from 3)
==4568== malloc/free: in use at exit: 201,356,478 bytes in 138,883 blocks.
==4568== malloc/free: 7,302,761 allocs, 7,163,878 frees, 735,786,915 bytes allocated.
==4568== For counts of detected errors, rerun with: -v
==4568== searching for pointers to 138,883 not-freed blocks.
==4568== checked 202,598,884 bytes.
==4568==
==4568== LEAK SUMMARY:
==4568==    definitely lost: 148,974 bytes in 8,629 blocks.
==4568==      possibly lost: 183,650,721 bytes in 2,330 blocks.
==4568==    still reachable: 17,556,783 bytes in 127,924 blocks.
==4568==         suppressed: 0 bytes in 0 blocks.
==4568== Rerun with --leak-check=full to see details of leaked memory.

Tout ça aurait-il pu arriver parce que mon ~/.libticables/ était interdit en lecture/écriture ?