Re: svn commit: r502649 - /tomcat/connectors/trunk/jk/native/common/jk_global.h

2007-02-05 Thread Jim Jagielski
It sounds as though the problem is that even though the header exists under HPUX, it is unusable. So we may still need to explicit HPUX test... I'll review the autoconf stuff to see if we can force undef'ing the HAVE_* for HPUX. On Feb 5, 2007, at 2:52 AM, Jean-Frederic wrote: Hi, I have test

Re: svn commit: r502649 - /tomcat/connectors/trunk/jk/native/common/jk_global.h

2007-02-05 Thread Jean-Frederic
Something like: +++ [[#ifdef HAVE_SYS_SOCKETVAR_H # include #endif ]]) AC_CHECK_HEADERS([sys/select.h], [], [], [[#ifdef HAVE_SYS_SELECT_H # include #endif ]]) +++ helps for sys/socketvar.h Should we have macros for that? Cheers Jean-Frederic On Fri, 2007-02-02 at 18:36 +0100, Rainer Jung wro

Re: svn commit: r502649 - /tomcat/connectors/trunk/jk/native/common/jk_global.h

2007-02-04 Thread Jean-Frederic
Hi, I have tested, AC_CHECK_HEADERS() does not help: +++ checking sys/socketvar.h usability... no checking sys/socketvar.h presence... yes configure: WARNING: sys/socketvar.h: present but cannot be compiled configure: WARNING: sys/socketvar.h: check for missing prerequisite headers? configure:

Re: svn commit: r502649 - /tomcat/connectors/trunk/jk/native/common/jk_global.h

2007-02-04 Thread Rainer Jung
You are right. Jim Jagielski wrote: On Feb 2, 2007, at 12:36 PM, Rainer Jung wrote: I can't test on HP-UX, but maybe you (jfc) could try: -#if !defined(_OSD_POSIX) && !defined(AS400) && !defined(CYGWIN) && !defined(HPUX11) +#if !defined(_OSD_POSIX) && !defined(AS400) && !defined(CYGWIN) &&

Re: svn commit: r502649 - /tomcat/connectors/trunk/jk/native/common/jk_global.h

2007-02-02 Thread Jean-frederic Clere
Rainer Jung wrote: I can't test on HP-UX, but maybe you (jfc) could try: The problem is that sys/socketvar.h can't be compiled using gcc and I think we don't need it HP-UX (for example APR doesn't include it). I will test the modifications in configure.in later. Cheers Jean-Frederic Ind

Re: svn commit: r502649 - /tomcat/connectors/trunk/jk/native/common/jk_global.h

2007-02-02 Thread Jim Jagielski
On Feb 2, 2007, at 12:36 PM, Rainer Jung wrote: I can't test on HP-UX, but maybe you (jfc) could try: -#if !defined(_OSD_POSIX) && !defined(AS400) && !defined(CYGWIN) && !defined(HPUX11) +#if !defined(_OSD_POSIX) && !defined(AS400) && !defined(CYGWIN) && !defined(HAVE_SYS_SOCKETVAR_H) #in

Re: svn commit: r502649 - /tomcat/connectors/trunk/jk/native/common/jk_global.h

2007-02-02 Thread Rainer Jung
I can't test on HP-UX, but maybe you (jfc) could try: Index: configure.in === --- configure.in(revision 502659) +++ configure.in(working copy) @@ -139,6 +139,10 @@ dnl check for filio.h used on Solaris to define FIONR

Re: svn commit: r502649 - /tomcat/connectors/trunk/jk/native/common/jk_global.h

2007-02-02 Thread Jim Jagielski
Don't we also have a HPUX11 specific check like the 2nd line after this one? Seems consistent to me :/ On Feb 2, 2007, at 12:14 PM, Rainer Jung wrote: Hi Jean-Frederic, shouldn't we be able to find out about the necessity to include it via configure? At least HP-UX should be able to use the

Re: svn commit: r502649 - /tomcat/connectors/trunk/jk/native/common/jk_global.h

2007-02-02 Thread Rainer Jung
Hi Jean-Frederic, shouldn't we be able to find out about the necessity to include it via configure? At least HP-UX should be able to use the configure mechanism. I think we mostly use the hard coded defines for the platforms, where we can't use the configure mechanism. Regards, Rainer [EMA

svn commit: r502649 - /tomcat/connectors/trunk/jk/native/common/jk_global.h

2007-02-02 Thread jfclere
Author: jfclere Date: Fri Feb 2 08:27:53 2007 New Revision: 502649 URL: http://svn.apache.org/viewvc?view=rev&rev=502649 Log: Otherwise it doesn't compile with gcc on HPUX. Modified: tomcat/connectors/trunk/jk/native/common/jk_global.h Modified: tomcat/connectors/trunk/jk/native/common/jk_g