Mailing-List: contact cygwin-developers-help AT sourceware DOT cygnus DOT com; run by ezmlm List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-developers-owner AT sourceware DOT cygnus DOT com Delivered-To: mailing list cygwin-developers AT sourceware DOT cygnus DOT com Message-Id: <200003021905.NAA09040@hp2.xraylith.wisc.edu> To: cygwin-developers AT sourceware DOT cygnus DOT com Subject: Re: problem with fork/exec in Cygwin DLL called from non-Cygwin E XE In-reply-to: Your message of "Thu, 02 Mar 2000 13:32:41 EST." <20000302133241 DOT F20207 AT cygnus DOT com> Date: Thu, 02 Mar 2000 13:05:46 -0600 From: Mumit Khan Chris Faylor writes: > > No, not at all. I don't want to add to the current ugly dynamic DLL > handling code though. It seems incredibly complicated (to me) for > something that AFAICT should be relatively simple. > > Both Mumit and I have banged on the code but neither of us have been > brave enough to streamline or simplify it. The last time we went through this code, the consensus (then) was that we won't be able to support subprocess handling via fork/exec in Cygwin DLLs that are dynamically loaded, which is the case when it's loaded via Excel. Note that this non-working set includes exec* family as well, and what works is the spawn* family of routines, which can replace *some* instances of fork/exec pair. Regards, Mumit ps: I'd love to be proven grossly wrong in this case!