Re: [PATCH] Use spawn() in GNU Make on Cygwin, updated

[Chris Faylor]

On Mon, Sep 02, 2013 at 10:32:38AM -0400, David Boyce wrote:
>   Slight change of subject ...
>   All this arguing has been about the value of spawn as a speed
>   enhancement:
>   Pro: spawn is way faster!
>   Con: but it may destabilize things - we'll stick with slow-but-stable
>   fork.
>   Which ignores the fact that Cygwin fork is *not* that stable. In our
>   experience make -jN results in frequent failed builds due to fork
>   failures. Thus it seems to me that
>   Pro: spawn doesn't suffer from the documented problems which result
>   from trying to emulate fork on Windows.
>   should be part of the conversation too.

If "fork is *not* that stable" that is certainly something that needs to
be corrected for Cygwin*.

But then, maybe it's not needed.  I will just assert "fork is WICKED
STABLE!" and, lacking any data points, the two assertions cancel each
other out.  Actually, since I used exclamation points, I probably win
until someone counters with more exclamation points.

As the person who actually maintains fork and spawn, I'll state it
again: we don't spend a lot of time on the spawn code.  There are likely
to be bugs there.  And, I'm not really keenly interested in spending
time tracking them down since, as I said, I consider spawn() a
deprecated interface in Cygwin.  Yes, I know.  It's FASTER.  It also is
completely non-UNIX and goes against the whole point of Cygwin.  It is
available in Cygwin only grudgingly for backwards-compatibility.

To the OP: It is really pointless to keep arguing this without bothering
to respond to the concerns made by the people who maintain make.  Eli
suggested adding a command-line option.  Please try implementing that
and stop this back and forth.  It's getting really old.

cgf

*There is one know issue with fork where dlls in forked processes don't
load at the right address and cause cygwin errors.  That is usually
fixed by running the autorebase program.