httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jeff Trawick <trawi...@bellsouth.net>
Subject signal() vs. ap_signal()
Date Tue, 04 Apr 2000 20:25:28 GMT
This post started out as a patch to ap_ac_config.h to see if anyone
could think of a situation where it would break on some system (see
below for the patch).  The patch fixes a compile failure on OS/390.

I think the big picture is that we/I/whomever need to change Apache
code to call ap_signal() instead of signal() and then the macro can go
away.  Right?  I'd rather spend time on that than with a work-around
for a compile failure caused by another work-around which was intended
to disappear almost immediately.

[PATCH]

At least one compiler needs to see signal.h before this macro.
Can anyone think of a reason not to add the include of signal.h before
the macro definition, as shown below?  I think the macro is temporary
anyway according to the cvs log entry for when this macro was added.

Index: src/include/ap_ac_config.h
===================================================================
RCS file: /cvs/apache/apache-2.0/src/include/ap_ac_config.h,v
retrieving revision 1.12
diff -u -r1.12 ap_ac_config.h
--- ap_ac_config.h      2000/04/03 18:40:36     1.12
+++ ap_ac_config.h      2000/04/04 20:26:27
@@ -227,6 +227,10 @@
 #undef USE_MMAP_FILES
 #endif

+#if APR_HAVE_SIGNAL_H
+#include <signal.h>
+#endif
+
 #define signal(s, f)    ap_signal(s, f)

 #endif /* AP_AC_CONFIG_H */              

-- 
Jeff Trawick | trawick@ibm.net | PGP public key at web site:
     http://www.geocities.com/SiliconValley/Park/9289/
          Born in Roswell... married an alien...

Mime
View raw message