Hi Ken,
On Jun 21 14:47, Ken Brown wrote:
> On 6/20/2015 4:55 PM, Corinna Vinschen wrote:
> >- First cut of an implementation to allow signal handlers running on an
> > alternate signal stack.
> >
> >- New API sigaltstack, plus definitions for SA_ONSTACK, SS_ONSTACK,
> >SS_DISABLE,
> > MINSIGSTKSZ, SIGSTKSZ.
>
> I must be doing something wrong. Shouldn't including signal.h make the new
> API available?
>
> $ uname -a
> CYGWIN_NT-6.1-WOW fiona 2.1.0(0.287/5/3) 2015-06-20 21:44 i686 Cygwin
>
> $ cygcheck -cd cygwin-devel
> Cygwin Package Information
> Package Version
> cygwin-devel 2.1.0-0.1
>
> $ cat test.c
> #include <signal.h>
> int
> main()
> {
> int foo = SIGSTKSZ;
> return 0;
> }
>
> $ gcc test.c
> test.c: In function ‘main’:
> test.c:6:13: error: ‘SIGSTKSZ’ undeclared (first use in this function)
> int foo = SIGSTKSZ;
> ^You're not doing anything wrong. The relevant definitions in sys/signal.h were originally only available for RTEMS. I just made them availbale for all platforms. The problem was that the orignal code missed to include sys/cdefs.h, which is required to get the macros guarding the definitions. I fixed that in the git repo. I also made a bigger change to code setting up the alternate stack when calling the signal handler function. It turned out that my code moving to the new stack missed to safe all potentially clobbered volatile registers on both platforms. I'll upload new snapshots and 2.1.0-0.2 test releases shortly. Thanks, Corinna -- Corinna Vinschen Please, send mails regarding Cygwin to Cygwin Maintainer cygwin AT cygwin DOT com Red Hat
pgps0yhI9DWOG.pgp
Description: PGP signature
