This is the mail archive of the
mailing list for the Cygwin project.
Re: mumit: gcc advice please
----- Original Message -----
From: "Christopher Faylor" <email@example.com>
Sent: Thursday, April 12, 2001 9:31 AM
Subject: Re: mumit: gcc advice please
> On Thu, Apr 12, 2001 at 09:11:17AM +1000, Robert Collins wrote:
> >Thanks Mumit, I caught that myself immediately after posting... I
> >repost as there were more questions there.
> >The good news is:
> >It seems to work.
> >The test case that was dying is no longer dying, and it uses
> >pthread_key's and pthread_once both of which are fully operation
> >(well, barring cancellation points as previously mentioned.
> >So I'm going to clean up my code a little, and then post an interim
> >patch to cygwin-patches for any adventure seeking folk.
> Once you do this, I'll give your stuff a couple of days to settle down
> and then I'll create a brand-new 1.3.0 release.
> I'll be willing to lay bets right now that it will be followed by a
> 1.3.1 release very soon thereafter because I'm sure there is something
> stupid lurking in the wings (like the recent "I can't run a .bat file
> problem") but I think that enough people are expressing satisfaction
> with recent snapshots that we should be in *relatively* good shape.
Don't be too hopeful! I'm not claiming that this initial code is
complete, bug free or race free. All I'm claiming is that the ABI should
be static from here on in, and that bugs as and when they are reported
will be looked at. As a matter of fact there will be some _known races_
in my initial code. However I feel that getting the support out there,
and compile problems etc reported is better than shooting for a perfect
implementation from day 1.
So I'm betting on early releases too :].
Want to unsubscribe from this list?
Check out: http://cygwin.com/ml/#unsubscribe-simple