when invoked with option -pthread, gcc-3.x does not add the pre-processor definition -D_REENTRANT, as explained in the documentation & as gcc-2.x was previously doing. This behaviour is duplicated on the old gcc-3.0.4 on RedHat 7.2 / ia64, then on the same compiler on RedHat Advanced Server 2.1, then on gcc-3.2.3 for Debian Sarge / ia64. Ref: <http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=84589> RedHat did not take this bug into account. Release: gcc-3.x Environment: Linux ia64 How-To-Repeat: Steps to Reproduce: 1. get the attached program pthbug.c 2. compile it with gcc -v -pthread -Wall -Wstrict-prototypes pthbug.c 3. note that -D_REENTRANT is not defined
State-Changed-From-To: open->feedback State-Changed-Why: [Before responding, I checked the gcc documentation that shipped with both FSF gcc 3.2.2 and FSF gcc 3.3.] Question: To which documentation are you referring? To my knowledge, there was no documentation shipped with FSF gcc 3.2.X or FSF gcc 3.3 which claims that -D_REENTRANT is added when -pthread is present. It is true that some ports do indeed add it under that condition however I can find no documentation on the matter.
From: Francois-Xavier Kowalski <francois-xavier.kowalski@hp.com> To: ljrittle@gcc.gnu.org, francois-xavier.kowalski@hp.com, gcc-bugs@gcc.gnu.org, gcc-prs@gcc.gnu.org, jean-pierre.joannin@hp.com, nobody@gcc.gnu.org, gcc-gnats@gcc.gnu.org Cc: Subject: Re: target/10865: [ia64] gcc -pthread does not add -D_REENTRANT Date: Tue, 20 May 2003 09:38:50 +0200 Hello, [sorry for the direct replay, but I cannot find any "Edit PR" button on the page below & on the entire gants/gcc web. Thanks to point me to a link, so that I can copy/paste the below in the gnats database]. "-pthread" is documented as an RS6000/PowerPC option only: <http://gcc.gnu.org/onlinedocs/gcc-3.2.2/gcc/RS-6000-and-PowerPC-Options.html#RS%2f6000%20and%20PowerPC%20Options> Quoting: |-pthread| Adds support for multithreading with the pthreads library. This option sets flags for both the preprocessor and linker. However, this option exists for every architectures I worked with (HP-UX, Linux, ia32, ia64...). It is even reported as part of the gcc specs: RedHat 8.0 / i386: gcc -v Reading specs from /usr/lib/gcc-lib/i386-redhat-linux/3.2/specs Configured with: ../configure --prefix=/usr --mandir=/usr/share/man --infodir=/usr/share/info --enable-shared --enable-threads=posix --disable-checking --host=i386-redhat-linux --with-system-zlib --enable-__cxa_atexit Thread model: *posix* gcc version 3.2 20020903 (Red Hat Linux 8.0 3.2-7) Debian Sarge / ia64: gcc-3.2 -v Reading specs from /usr/lib/gcc-lib/ia64-linux/3.2.3/specs Configured with: ../src/configure -v --enable-languages=c,c++,java,f77,proto,objc,ada --prefix=/usr --mandir=/usr/share/man --infodir=/usr/share/info --with-gxx-include-dir=/usr/include/c++/3.2 --enable-shared --with-system-zlib --enable-nls --without-included-gettext --enable-__cxa_atexit --enable-clocale=gnu --enable-java-gc=boehm --enable-objc-gc ia64-linux Thread model: *posix* gcc version 3.2.3 20030309 (Debian prerelease) So I have always assumed (as it always worked) that the description for RS6000/PowerPC was Ok for every supported architecture. What I can observe in the current GCC I am using is that the linker option (-lpthread) is set, but not the pre-compiler option -- FiX ljrittle@gcc.gnu.org wrote: >Synopsis: [ia64] gcc -pthread does not add -D_REENTRANT > >State-Changed-From-To: open->feedback >State-Changed-By: ljrittle >State-Changed-When: Tue May 20 00:19:36 2003 >State-Changed-Why: > [Before responding, I checked the gcc documentation that shipped > with both FSF gcc 3.2.2 and FSF gcc 3.3.] > > Question: To which documentation are you referring? To my knowledge, there was no documentation shipped with FSF gcc > 3.2.X or FSF gcc 3.3 which claims that -D_REENTRANT is added > when -pthread is present. It is true that some ports do indeed > add it under that condition however I can find no documentation > on the matter. > >http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=10865 > > -- Francois-Xavier "FiX" KOWALSKI /_ __ Tel: +33 (0)4 76 14 63 27 OpenCall Business Unit -- OCBU / //_/ Fax: +33 (0)4 76 14 43 23 Signalling Products / HP Telnet: 779-6327 http://www.hp.com/go/opencall i n v e n t
Loren, is there a reason this report is still in WAITING? Thanks, Dara
1. Do I have to populate any kind of information so that this DR can move forward to a non-WAITING state? 2. BTW, the pre-processor #define that should be added for GCC on HP-UX 11.x are "-D_REENTRANT -D_PTHREADS_DRAFT4", not only "-D_REENTRANT". The link statement remain "-lpthread".
I have no further opinion on this matter (my position is that this is an inconsistency between ports not a bug with any one of them). It appears that the only way to move out of WAITING is to supply a patch to an HP maintainer or excite one to fix it. Please note that I'm not such a person.
I really want to say this is documentation error rather than an errror due to GCC so changing back to NEW. The documentation should talk about -pthread in a more non-target specfic way and then add something about it for AIX unless some else (maintainers) agrues otherwise.
I am going to resolve this as invalid since I can't find any documentation to fix. All the existing documentation I checked, even 3.2.2, is generic about what flags pthread sets. None of them mention _REENTRENT.
This bug is not invalid; it is a special case of bug 20705.