April 29, 2003

Mmmm, memory mismanagement!

Finally got around to downloading and installing valgrind to run against parrot, to try and track down some weird memory problems that show up on OS X, but not on Linux. (OS X is my primary platform, the patch that triggers the bug came from someone on Linux) Turns out that the bug does manifest on linux, just nothing whines about it. OS X's C library's much pickier about things by default. This, I think, is a good thing.

Definitely a cool and useful toy, though. Almost enough to get me back to Linux as a development system...

Posted by Dan at April 29, 2003 05:10 PM | TrackBack (0)
Comments

What, an actual poll() implementation is not enough? ;-)

You might want to also checkout cachegrind (aka kcachegrind). Its a neat hack that adds a bit of profiling information to valgrind, and makes a superb profiler.

Posted by: Sterling Hughes at April 29, 2003 09:22 PM

A working poll() is not enough to draw me in, alas. :-P

I saw the other *grind bits, and I think I'll give them a whirl. (I'd really love to have a tool like this on other CPU architectures and OSes, though I expect there's something similar for at least Solaris, and VMS/Tru64 had a nice set of tools last I knew. I just need to get a VMS and/or Tru64 system....)

First, though, I need to convince valgrind that the system stack really is walkable, and figure out if the arena walking GC code is busted or not, since it has some complaints there. Hard to tell if they're correct complaints, or just noticing that we're doing Evil Things that are, nonetheless, safe.

Posted by: Dan at April 30, 2003 10:15 AM