On Sat, 12 Jan 2002, Greg Stein wrote: > > Duh! (/me slaps himself) Ofcourse. Who's patching, you, justin or me? > > Assign it to the guy not here! ... Cliff!! :-)
As of right about now I should be relatively "back". :) It's going to take me a while to catch up, though... > (since he broke DELONCLOSE Fixed, fixed I say! :-] > w.r.t. child forks...) Cliff: the issue is that both the parent and > child try to delete the file after a fork. Oh, crap. Except for that. :) So I guess the idea is to register the deletion as a separate cleanup, and choose whether to make it a regular cleanup or a child cleanup based on DELONCLOSE or CHILD_DELONCLOSE is used... plus if it's DELONCLOSE, there has to be a child cleanup that removes the DELONCLOSE flag from the apr_file_t in the child. Is that what I'm hearing? I guess that makes sense... actually, is there ever a reason for CHILD_DELONCLOSE? Or can I just register a child cleanup to remove the DELONCLOSE flag from the child unconditionally? --Cliff -------------------------------------------------------------- Cliff Woolley [EMAIL PROTECTED] Charlottesville, VA
