This is the mail archive of the
cygwin-developers@cygwin.com
mailing list for the Cygwin project.
RE: muto object.
- To: <cygwin-developers at cygwin dot com>
- Subject: RE: muto object.
- From: "Robert Collins" <robert dot collins at itdomain dot com dot au>
- Date: Mon, 17 Sep 2001 12:24:53 +1000
- Cc: <cygwin-patches at cygwin dot com>
I'll try finishing the email this time.
What I meant to say was, if this looks ok, it makes muto's a potential
replacement for critical sections on 95 for pthreads, which would be
very good speed wise.
Anyway, I'll draw up a change log and the rest if you want this
included.
Rob
> -----Original Message-----
> From: Robert Collins
> Sent: Monday, September 17, 2001 12:23 PM
> To: cygwin-developers@cygwin.com
> Cc: cygwin-patches@cygwin.com
> Subject: muto object.
>
>
> Chris,
> This update to muto handles threads exiting spontaneously without
> releasing the muto properly. I think it fixes the FIXME you have in
> ::release, but as I can't see how release can check for other thread
> activity, it may not have fixed that.
>
> The logic it uses is:
> if we fail to wait for the event,
> protect ourselves with recover
> check for the thread having died (should be fast - noop basically) and
> if it has aquire the muto anyway.
>
> There was also a typo in the destructor that could be causing memory
> leaks within process.
>
> Rob
>