Markus Wichitill <[EMAIL PROTECTED]> writes: > Ville Skytt� wrote: >>>WAG: see if this patch helps any >> Nope, no difference at all :( I _did_ get the files in /tmp using your >> earlier debugging patch; was this supposed to change that somehow >> (didn't test with it now), or was it a WAG at fixing the original >> reported problem? > > Doesn't fix the original problem for me either.
Thanks. What's troubling me though is that while both you and Ville are seeing the leak, you're not reporting consistent behavior with the tempfile creation. So although we know it's the prefetch code that's gobbling up memory (since calling discard_request_body fixes the leak), I really don't know where to look for it. I'd really like to fix this sometime soon, so we can start the 2.06 candidates... -- Joe Schaefer
