On Sat, Jan 30, 2010 at 7:41 AM, Jeff Trawick <[email protected]> wrote: > On Fri, Jan 29, 2010 at 3:25 AM, Tatsuki Sugiura <[email protected]> wrote: >> Hello all, >> >> I found a problem that mod_fcgi makes segmentation fault >> when FastCGI process returns empty outout. It's caused >> with missing to check end of brigade when parsing CGI >> response header. >> >> I worte a patch to fix this, and attach on this mail. >> Please review. > > Thanks for the patch. I've reproduced the error and confirmed that > your patch solves the segfault. I hope to look at a couple of other > aspects: > > a) is it better (and practical) to ensure that the brigade has an EOS > bucket to start with so this special case doesn't have to be handled
I was tricked by the existing check for EOS; there never is one AFAICT (the "done" flag is the usual escape from the loop). So your new check stays and the EOS check goes. > b) how to get a 500 error to the client in this situation (the FastCGI > app returns no headers) I still need to look at this (cgid returns 500 in this case)
