audit.36 kernel

[David Woodhouse]

On Mon, 2005-05-16 at 10:36 -0500, Timothy R. Chavez wrote:
> interesting... i'm not seeing these problems (not at least, with the latest 
> update patch I replied to the #7U5 thread with)... let me look into it deeper 

Please make sure you're testing on the current RPM kernel -- nothing
else. However, you can ignore audit.38; it had the auditfs patch
disabled because I suspected it of causing memory corruption. As it
happens, it was merely the victim of memory corruption which persists
for me even when I revert to about what was in audit.36. I'm still
debugging it, but the currently lack of responsiveness of my test box
seems to be hinting that it's time to call it a night...
 
audit.37 in the yum repo has a first attempt to log sys_socketcall()
arguments and any sockaddr which is passed as a syscall argument. (And
no, if I fix the obvious bug in that first version, it doesn't fix my
memory corruption).

-- 
dwmw2