Re: Support for ITIMER_PROF?

[Corinna Vinschen <corinna-cygwin at cygwin dot com>]

On Jun 22 17:07, Ken Brown wrote:
> On 6/22/2015 9:25 AM, Corinna Vinschen wrote:
> >On Jun 22 08:05, Ken Brown wrote:
> >>According to the Cygwin API documentation
> >>(https://cygwin.com/cygwin-api/std-notes.html), "getitimer and setitimer
> >>only support ITIMER_REAL for now."  I'm wondering whether there's any chance
> >>that support for ITIMER_PROF might be added.  I have no idea what the
> >>obstacles are.
> >>
> >>My reason for asking is that emacs has a CPU profiling function, which
> >>doesn't work on Cygwin because it relies on ITIMER_PROF.  (There's an
> >>alternative implementation, but it requires timer_getoverrun, also not
> >>available in Cygwin.)
> >
> >Both very tricky.  Given the description of timer_getoverrun and how
> >this stuff is implemented in Cygwin, you might get away with
> >
> >   #ifdef __CYGWIN__
> >   #define timer_getoverrun(x)	0
> >   #endif
> >
> >indepedently of using SIGEV_SIGNAL or SIGEV_THREAD.
> 
> This seems to work.  Thanks!
> 
> One question: It seems that SIGEV_SIGNAL and friends are defined as macros
> on Linux (or at least on the one Linux distribution I looked at), but on
> Cygwin they're only enum constants.  And emacs expects them to be defined as
> macros. Would something like the following be reasonable for the sake of
> compatibility?
> 
> --- a/winsup/cygwin/include/cygwin/signal.h
> +++ b/winsup/cygwin/include/cygwin/signal.h
> @@ -311,6 +311,10 @@ enum
>                                            perform notification */
>  };
> 
> +#define SIGEV_SIGNAL SIGEV_SIGNAL
> +#define SIGEV_NONE   SIGEV_NONE
> +#define SIGEV_THREAD SIGEV_THREAD
> +
>  #if __WORDSIZE == 64
>  typedef __uint64_t sigset_t;
>  #else

Sure.  I applied the patch.


Thanks,
Corinna

-- 
Corinna Vinschen                  Please, send mails regarding Cygwin to
Cygwin Maintainer                 cygwin AT cygwin DOT com
Red Hat

Attachment: pgpr30ry3TGD3.pgp
Description: PGP signature