Re: Hurd build system Makeconf

[Thomas Schwinge]

Hi!

On Wed, 25 Jul 2012 10:48:38 +0200, Richard Braun <rbraun@sceen.net> wrote:
> On Tue, Jul 24, 2012 at 11:12:11PM -0400, Barry deFreese wrote:
> > Now the question is, what to do with cancel-cond.c?  I'm thinking of just 
> > sticking it in
> > libshouldbeinlibc for now?  Is there a more appropriate place?
> 
> Why not in libpthread, near the pthread_cond_wait and
> pthread_cond_timedwait functions, considering how related they are ?

That's also what I suggested on IRC, I think.

But, it doesn't really matter for the time being, during development,
while this whole thing is being beaten into shape -- which I again want
to thank "the other Thomas" for picking it up again!  (Thomas, I'm
terribly busy this week with work and real life; let me try to answer
your remaining questions next week; sorry again for the delays... -- as
all our other contributors, I typically only have a few hours per week to
spend on GNU Hurd matters, and there's so many to chose from...)

Anyway, back on topic: put that function anywhere you like for the time
being -- libpthread in fact may not be the most convenient place since
its build process in Debian has been merged into glibc's, so any changes
there require something equivalent to a Debian glibc package rebuild.
So, yes, perhaps put it into libshouldbeinlibc for the time being.  Feel
free to hack it in, that is, use internal libpthread/libc interfaces as
you need; we'll work on integrating it properly afterwards, but now it is
important that you get it working at all.


Grüße,
 Thomas

pgp8qu8b5lbWc.pgp
Description: PGP signature