diff options
| author | tjg <tjg> | 2002-08-15 13:45:03 +0000 |
|---|---|---|
| committer | Toby Goodwin <toby@paccrat.org> | 2015-05-16 21:23:22 +0100 |
| commit | 8f26017bb9cb3f38424f50511263cf70e476d98b (patch) | |
| tree | ea09b06bf2fb9cc7a9dcfcf287f309ff1cd04dce /acinclude.m4 | |
| parent | 8f81d22c989808d3889a7ad6806903adec5ae27b (diff) | |
Portability: don't call sigaction() for SIGKILL or SIGCONT; don't
hand a garbage signal mask to sigaction() (thanks Jeremy
Fitzhardinge). Also, remove use of SA_INTERRUPT (SUSv3, BSD,
etc. have SA_RESTART with the inverted meaning).
Diffstat (limited to 'acinclude.m4')
| -rw-r--r-- | acinclude.m4 | 15 |
1 files changed, 0 insertions, 15 deletions
diff --git a/acinclude.m4 b/acinclude.m4 index 7fbc6aa..0cf3307 100644 --- a/acinclude.m4 +++ b/acinclude.m4 @@ -128,21 +128,6 @@ AC_DEFUN(RC_TYPE_SIG_ATOMIC_T, [ ]) -dnl Check for sigaction and SA_INTERRUPT -AC_DEFUN(RC_FUNC_SIGACTION, [ - AC_CACHE_CHECK(for sigaction and SA_INTERRUPT, rc_cv_sa_int, - AC_TRY_COMPILE([ -#include <signal.h> - ], [ -struct sigaction foo; -foo.sa_flags = SA_INTERRUPT; -sigaction(SIGINT, 0, 0); - ], rc_cv_sa_int=yes, rc_cv_sa_int=no - ) - ) -]) - - dnl Do we have SysV SIGCLD semantics? In other words, if we set the dnl action for SIGCLD to SIG_IGN does wait() always say ECHILD? Linux, dnl of course, is bizarre here. It basically implements the SysV |