[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-bugs-dist
Subject:    [Bug 91153] Possible new leak to suppress in glibc-2.?.supp
From:       "Daniel Richard G." <skunk () iskunk ! org>
Date:       2004-11-01 1:54:35
Message-ID: 20041101015435.32719.qmail () ktown ! kde ! org
[Download RAW message or body]

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.
        
http://bugs.kde.org/show_bug.cgi?id=91153        




------- Additional Comments From skunk iskunk org  2004-11-01 02:54 -------
Sorry, I should have given a bit more background on this.

My company tests their software nightly through Valgrind, with --show-reachable=yes, \
and (till recently) everything would run without any loss records at all. We \
hair-trigger it such that even one program losing one byte of memory effects a \
failure (and we keep enough programming discipline to make this a relatively rare \
occurrence).

After we upgraded the build system to use Libtool (thereby compiling/linking \
everything a bit differently), a large number of logs started showing these 16-byte \
loss records, causing the nightly tests to fail. The last thing I wanted to do at \
that point was make an exception to the zero-tolerance criterion, which had already \
been well-established in our shop; let alone remove --show-reachable=yes from our \
Valgrind invocations, as that has flagged a great many programming errors for us.

Anyway, the fact that everything went through fine beforehand suggested that the \
programs themselves were not at fault, but it could be that they were missing e.g. \
pthread_exit() calls, and for whatever reason this had not caused any trouble until \
now. I wil investigate this during the week.


[prev in list] [next in list] [prev in thread] [next in thread]