pcp
[Top] [All Lists]

qa failures - valgrind oddity on SuSE

To: pcp@xxxxxxxxxxx
Subject: qa failures - valgrind oddity on SuSE
From: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Date: Tue, 29 Jan 2013 07:36:32 +1100
Delivered-to: pcp@xxxxxxxxxxx
Specifically on openSUSE 12.1 (i586) I am seeing a handful of valgrind
failures.

After some investigation I'm pretty sure these are valgrind failures not
PCP failures.  The installed valgrind is valgrind-3.6.1-16.11.2.i586.

The valgrind signature is ...

==23988== Conditional jump or move depends on uninitialised value(s)
==23988==    at 0x4076554: __pmLogFetchInterp (interp.c:1417)
==23988==    by 0x40713BB: __pmLogFetch (logutil.c:1653)
==23988==    by 0x50052601: ???
==23988==  Uninitialised value was created by a heap allocation
==23988==    at 0x402913D: malloc
(in /usr/lib/valgrind/vgpreload_memcheck-x86-linux.so)
==23988==    by 0x4076373: __pmLogFetchInterp (interp.c:1409)
==23988==    by 0x40713BB: __pmLogFetch (logutil.c:1653)
==23988==    by 0x50052601: ???

But the malloc block allocated at interp.c:1409 is correctly
initializaed and is not even referenced at interp.c:1417!

Does this ring bells for anyone?

<Prev in Thread] Current Thread [Next in Thread>