Received: (qmail 26063 invoked by uid 2012); 13 Jul 1998 20:53:26 -0000 Message-Id: <19980713205326.26062.qmail@hyperreal.org> Date: 13 Jul 1998 20:53:26 -0000 From: Paul Williams Reply-To: paulw@uwex.edu To: apbugs@hyperreal.org Subject: Compile filed during make X-Send-Pr-Version: 3.2 >Number: 2605 >Category: os-aix >Synopsis: Compile filed during make >Confidential: no >Severity: critical >Priority: medium >Responsible: apache >State: closed >Class: sw-bug >Submitter-Id: apache >Arrival-Date: Mon Jul 13 15:10:00 PDT 1998 >Last-Modified: Wed Aug 12 10:37:06 PDT 1998 >Originator: paulw@uwex.edu >Organization: >Release: 1.3.0 >Environment: AIX 4.2 Output from uname -a is AIX popeye 2 4 000101657500 aix c compiler >Description: Following is the output from script: Script command is started on Fri Jul 3 11:50:56 CDT 1998.popeye% make^M ===> regex^M cc -I. -I../os/unix -I../include -DAIX=42 -U__STR__ -qnogenpcomp -qnousepcomp -DUSE_HSREGEX -DPOSIX_MISTAKE -c regcomp.c^M cc -I. -I../os/unix -I../include -DAIX=42 -U__STR__ -qnogenpcomp -qnousepcomp -DUSE_HSREGEX -DPOSIX_MISTAKE -c regexec.c^M cc -I. -I../os/unix -I../include -DAIX=42 -U__STR__ -qnogenpcomp -qnousepcomp -DUSE_HSREGEX -DPOSIX_MISTAKE -c regerror.c^M cc -I. -I../os/unix -I../include -DAIX=42 -U__STR__ -qnogenpcomp -qnousepcomp -DUSE_HSREGEX -DPOSIX_MISTAKE -c regfree.c^M rm -f libregex.a^M ar cr libregex.a regcomp.o regexec.o regerror.o regfree.o^M ranlib libregex.a^M Target "all" is up to date.^M Target "lib" is up to date.^M <=== regex^M ===> os/unix^M cc -c -I../../os/unix -I../../include -DAIX=42 -U__STR__ -qnogenpcomp -qnousepcomp -DUSE_HSR EGEX os.c^M "/usr/include/sys/cdefs.h", line 131.9: 1506-236 (W) Macro name __attribute__ has been redefined.^M "/usr/include/sys/cdefs.h", line 131.9: 1506-358 (I) "__attribute__" is defined on line 829 of ../../in clude/conf.h.^M "os-aix-dso.c", line 133.12: 1506-343 (S) Redeclaration of dlerror differs from previous declaration on line 111 of "../../os/unix/os.h".^M "os-aix-dso.c", line 133.12: 1506-050 (I) Return type "unsigned char*" in redeclaration is not compatib le with the previous return type "const unsigned char*".^M "os.c", line 77.13: 1506-343 (S) Redeclaration of ap_os_dso_error differs from previous declaration on line 130 of "../../os/unix/os.h".^M "os.c", line 77.13: 1506-050 (I) Return type "unsigned char*" in redeclaration is not compatible with t he previous return type "const unsigned char*".^M make: 1254-004 The error code from the last command is 1.^M ^M ^M Stop.^M make: 1254-004 The error code from the last command is 1.^M >How-To-Repeat: >Fix: >Audit-Trail: From: Marc Slemko To: Paul Williams Cc: apbugs@hyperreal.org Subject: Re: os-aix/2605: Compile filed during make Date: Tue, 14 Jul 1998 20:31:10 -0700 (PDT) On 13 Jul 1998, Paul Williams wrote: > "/usr/include/sys/cdefs.h", line 131.9: 1506-236 (W) Macro name __attribute__ has been redefined.^M > "/usr/include/sys/cdefs.h", line 131.9: 1506-358 (I) "__attribute__" is defined on line 829 of ../../in > clude/conf.h.^M What is the content of /usr/include/sys/cdefs.h around line 131? From: Marc Slemko To: apbugs@apache.org Cc: Subject: Re: os-aix/2605: Compile filed during make (fwd) Date: Wed, 15 Jul 1998 11:08:17 -0700 (PDT) Cute. Real cute. And we currently include conf.h first, so we can't ifndef in conf.h. ---------- Forwarded message ---------- Date: Wed, 15 Jul 1998 08:02:48 -0500 (CDT) From: "Paul F. Williams" To: Marc Slemko Subject: Re: os-aix/2605: Compile filed during make The lines in cdefs.h are +121 +122 /* +123 * GCC1 and some versions of GCC2 declare dead (non-returning) and +124 * pure (no side effects) functions using "volatile" and "const"; +125 * unfortunately, these then cause warnings under "-ansi -pedantic". +126 * GCC2 uses a new, peculiar __attribute__((attrs)) style. All of +127 * these work for GNU C++ (modulo a slight glitch in the C++ grammar +128 * in the distribution version of 2.5.5). +129 */ +130 #if !defined(__GNUC__) || __GNUC__ < 2 || __GNUC_MINOR__ < 5 +131 #define __attribute__(x) /* delete __attribute__ if non-gcc or gc c1 */ +132 #if defined(__GNUC__) && !defined(__STRICT_ANSI__) +133 #define __dead __volatile +134 #define __pure __const +135 #endif +136 #endif If I recall the variation here was one had a (x) and the other had a (__x). I don't know if that is a major problem. paulw -------------------------------------------------------------------- Paul F. Williams, UW-Extension, 432 N. Lake St., Madison, WI 53706 paulw@uwex.edu, 608-262-6011 "If you took all the economists in this country and laid them end to end, they'd point in every direction." - Will Rogers -------------------------------------------------------------------- On Tue, 14 Jul 1998, Marc Slemko wrote: > On 13 Jul 1998, Paul Williams wrote: > > > "/usr/include/sys/cdefs.h", line 131.9: 1506-236 (W) Macro name __attribute__ has been redefined.^M > > "/usr/include/sys/cdefs.h", line 131.9: 1506-358 (I) "__attribute__" is defined on line 829 of ../../in > > clude/conf.h.^M > > What is the content of /usr/include/sys/cdefs.h around line 131? > > From: Marc Slemko To: Apache bugs database Cc: Subject: Re: os-aix/2605: Compile filed during make (fwd) Date: Sun, 19 Jul 1998 00:09:39 -0700 (PDT) ---------- Forwarded message ---------- Date: Thu, 16 Jul 1998 08:35:52 -0500 (CDT) From: Paul F. Williams To: Marc Slemko Subject: Re: os-aix/2605: Compile filed during make On Wed, 15 Jul 1998, Marc Slemko wrote: > And if you remove the define of __attribute__ in conf.h what happens? > > [...] > > > On 13 Jul 1998, Paul Williams wrote: > > > > > > > "/usr/include/sys/cdefs.h", line 131.9: 1506-236 (W) Macro name __attribute__ has been redefined.^M > > > > "/usr/include/sys/cdefs.h", line 131.9: 1506-358 (I) "__attribute__" is defined on line 829 of ../../in > > > > clude/conf.h.^M > > > > > > What is the content of /usr/include/sys/cdefs.h around line 131? > > > > > > > > > Here is what I did. In conf.h the commented out section looks like /* #if !defined(__GNUC__) || __GNUC__ < 2 || __GNUC_MINOR__ < 7 || defined(NEXT) #define ap_inline #define __attribute__(x) #define ENUM_BITFIELD(e,n,w) signed int n : w #else #define ap_inline __inline__ #define USE_GNU_INLINE #define ENUM_BITFIELD(e,n,w) e n : w #endif */ The scpipt output is Script command is started on Thu Jul 16 08:28:53 CDT 1998.popeye% mv  make ===> regex cc -I. -I../os/unix -I../include -DAIX=42 -U__STR__ -qnogenpcomp -qnousepcomp -DUSE_HSREGEX -lm -DPOSIX_MISTAKE -c regcomp.c cc -I. -I../os/unix -I../include -DAIX=42 -U__STR__ -qnogenpcomp -qnousepcomp -DUSE_HSREGEX -lm -DPOSIX_MISTAKE -c regexec.c cc -I. -I../os/unix -I../include -DAIX=42 -U__STR__ -qnogenpcomp -qnousepcomp -DUSE_HSREGEX -lm -DPOSIX_MISTAKE -c regerror.c cc -I. -I../os/unix -I../include -DAIX=42 -U__STR__ -qnogenpcomp -qnousepcomp -DUSE_HSREGEX -lm -DPOSIX_MISTAKE -c regfree.c rm -f libregex.a ar cr libregex.a regcomp.o regexec.o regerror.o regfree.o ranlib libregex.a Target "all" is up to date. Target "lib" is up to date. <=== regex ===> os/unix cc -c -I../../os/unix -I../../include -DAIX=42 -U__STR__ -qnogenpcomp -qnousepcomp -DUSE_HSREGEX -lm os.c "os.c", line 8.12: 1506-343 (S) Redeclaration of ap_os_dso_error differs from previous declaration on line 130 of "../../os/unix/os.h". "os.c", line 8.12: 1506-050 (I) Return type "unsigned char*" in redeclaration is not compatible with the previous return type "const unsigned char*". "os-aix-dso.c", line 133.12: 1506-343 (S) Redeclaration of dlerror differs from previous declaration on line 111 of "../../os/unix/os.h". "os-aix-dso.c", line 133.12: 1506-050 (I) Return type "unsigned char*" in redeclaration is not compatible with the previous return type "const unsigned char*". make: 1254-004 The error code from the last command is 1. Stop. make: 1254-004 The error code from the last command is 1. paulw State-Changed-From-To: open-closed State-Changed-By: wgstodda@us.ibm.com State-Changed-When: Wed Aug 12 10:37:05 PDT 1998 State-Changed-Why: A patch has been committed to 1.3.2-dev to fix this AIX compile problem. >Unformatted: [In order for any reply to be added to the PR database, ] [you need to include in the Cc line ] [and leave the subject line UNCHANGED. This is not done] [automatically because of the potential for mail loops. ]