This is the mail archive of the
cygwin-developers@cygwin.com
mailing list for the Cygwin project.
Thread problems fixed - on XP at least
- From: Christopher Faylor <cgf at redhat dot com>
- To: cygwin-developers at cygwin dot com
- Date: Wed, 2 Oct 2002 00:41:31 -0400
- Subject: Thread problems fixed - on XP at least
- References: <20020930151227.GA10898@redhat.com> <1033419006.30057.210.camel@lifelesswks> <20020930210309.GD4372@redhat.com> <1033422095.11275.216.camel@lifelesswks>
- Reply-to: cygwin-developers at cygwin dot com
On Tue, Oct 01, 2002 at 07:41:34AM +1000, Robert Collins wrote:
>On Tue, 2002-10-01 at 07:03, Christopher Faylor wrote:
>> On Tue, Oct 01, 2002 at 06:50:06AM +1000, Robert Collins wrote:
>> >On Tue, 2002-10-01 at 01:12, Christopher Faylor wrote:
>> >> I've changed the static declarations in thread.cc to use
>> >> NO_COPY and things seem to be working better in my build.
>> >
>> >I've got some confusion going on I think.
>> >
>> >Whats the call order during fork of dll_crt0_1, and
>> >MTinterface::fixup_after_fork ()?
>>
>> dll_crt0_1 calls fork.
>>
>> It seemed like it was possible that the static mutex value would be
>> overwritten when fork is called. That's not handled in
>> MTinterface::fixup_after_fork, is it?
>
>No, because I am handling it in MTinterface::init, which is called from
>dll_crt0_1. I suspect it was handling it too early though.
>
>Anyway, it's NO_COPY now, so that should be fixed.
It does seem to be working fine. Test suite works. "make all" works.
Did we ever figure out Pierre's problem, though? I never saw a
resolution for that.
cgf