Hi, Le 16/12/2009 19:50:32, Peter Bloomfield a écrit : .... >This looks like the same bug as in ><URL:http://mail.gnome.org/archives/balsa-list/2009-July/ > msg00025.html>. > >I see it quite often myself--in gdb, it's always somewhere in malloc, >and >always results from a corrupt struct malloc_chunk, pointed to by >victim. > >I've used MALLOC_CHECK_ to try to find the corrupter, and also >Valgrind, >with no success. However it happens, it's below their respective > radar screens. > Hi, Yes, I had reported myself this problem in the past. It is a bit strange because depending (maybe) on the svn progression it is there or not. I suspected for a time my hardware. But in the meantime I've changed about evrything (motherboard / processor / memory)… If I run balsa from alleyoop (a gnome frontend from valgrind), I can see things like this (while it seems to me everything is right): ==23735== Thread 3: ==23735== Invalid write of size 8 ==23735== at 0x09a1e90f: AES_cbc_encrypt (in /usr/lib/ libcrypto.so.0.9.8) ==23735== by 0x00000003: ??? ==23735== Address 0x1f631400 is not stack'd, malloc'd or (recently) free'd ==23735== Is it the kind of things you are looking for? Regards Jean-Luc
Attachment:
pgpgTn0vDoRzM.pgp
Description: PGP signature