On Tue, 2001-09-18 at 18:03, Greg Stein wrote:
> On Tue, Sep 18, 2001 at 09:38:03PM -0000, [EMAIL PROTECTED] wrote:
> > rbb 01/09/18 14:38:03
> >
> > Modified: include/arch/win32 thread_cond.h
> > locks/win32 thread_cond.c
> > Log:
> > An initial thread_condition variable implementation for Windows.
> > This is kind of ugly, but I believe it works. It does pass all of
> > the APR tests, and I don't see any race conditions, but other
> > eyes are more than welcome on this code now.
> >...
> > --- thread_cond.c 2001/09/13 01:04:23 1.1
> > +++ thread_cond.c 2001/09/18 21:38:03 1.2
> > @@ -60,29 +60,92 @@
> > #include "win32/thread_cond.h"
> > #include "apr_portable.h"
> >
> > +static apr_status_t thread_cond_cleanup(void *data)
> > +{
> > + apr_thread_cond_t *cond = data;
> > + if (cond->num_waiting != 0) {
> > + printf("somebody's waiting, but I'm closing it anyway.\n");
>
> Can't have a printf() in there. Not sure what the right answer is; maybe
> call the pool's abort function.
you >could< use the apr_file_printf function
and use stderr from the apr_open_stderr call.
this gets put into the log files, the std fprintf doesnt.
>
> Cheers,
> -g
>
> --
> Greg Stein, http://www.lyra.org/
--
Ian Holsman
Performance Measurement & Analysis
CNET Networks - 415 364-8608
