* [PATCH] Fix graphite configure w/o ppl/cloog
@ 2009-10-21 22:52 Richard Guenther
2009-10-22 6:19 ` Eric Botcazou
` (2 more replies)
0 siblings, 3 replies; 8+ messages in thread
From: Richard Guenther @ 2009-10-21 22:52 UTC (permalink / raw)
To: gcc-patches
The configure shouldn't leave LIBS set, so better not set it at all as
it's unused. Likewise the libelf LIBS saving is too late.
Bootstrapped on x86_64-unknown-linux-gnu, ok for trunk?
Thanks,
Richard.
2009-10-22 Richard Guenther <rguenther@suse.de>
* configure.ac: Do not set LIBS for ppl/cloog checks. Disable
cloog if the ppl version check failed. Move flags saving
before setting in libelf check.
* configure: Regenerate.
Index: configure.ac
===================================================================
*** configure.ac (revision 153055)
--- configure.ac (working copy)
*************** case $with_ppl in
*** 1523,1534 ****
ppllibs=
;;
yes)
- LIBS="$ppllibs $LIBS"
;;
*)
ppllibs="-L$with_ppl/lib -lppl_c -lppl -lgmpxx"
pplinc="-I$with_ppl/include $pplinc"
- LIBS="$ppllibs $LIBS"
;;
esac
if test "x$with_ppl_include" != x; then
--- 1523,1532 ----
*************** if test "x$with_ppl_include" != x; then
*** 1536,1547 ****
fi
if test "x$with_ppl_lib" != x; then
ppllibs="-L$with_ppl_lib -lppl_c -lppl -lgmpxx"
- LIBS="$ppllibs $LIBS"
fi
if test "x$with_ppl$with_ppl_include$with_ppl_lib" = x && test -d ${srcdir}/ppl; then
ppllibs='-L$$r/$(HOST_SUBDIR)/ppl/.libs -L$$r/$(HOST_SUBDIR)/ppl/_libs -lppl_c -lppl -lgmpxx '
pplinc='-I$$r/$(HOST_SUBDIR)/ppl/include -I$$s/ppl/include '
- LIBS="$ppllibs $LIBS"
fi
AC_ARG_ENABLE(ppl-version-check,
--- 1534,1543 ----
*************** if test "x$with_ppl" != "xno" -a "${ENAB
*** 1557,1563 ****
#if PPL_VERSION_MAJOR != $ppl_major_version || PPL_VERSION_MINOR != $ppl_minor_version
choke me
#endif
! ], [AC_MSG_RESULT([yes])], [AC_MSG_RESULT([no]); ppllibs= ; pplinc= ])
CFLAGS="$saved_CFLAGS"
fi
--- 1553,1559 ----
#if PPL_VERSION_MAJOR != $ppl_major_version || PPL_VERSION_MINOR != $ppl_minor_version
choke me
#endif
! ], [AC_MSG_RESULT([yes])], [AC_MSG_RESULT([no]); ppllibs= ; pplinc= ; with_ppl=no ])
CFLAGS="$saved_CFLAGS"
fi
*************** case $with_cloog in
*** 1586,1597 ****
clooginc=
;;
yes)
- LIBS="$clooglibs $LIBS"
;;
*)
clooglibs="-L$with_cloog/lib -lcloog"
clooginc="-I$with_cloog/include -DCLOOG_PPL_BACKEND "
- LIBS="$clooglibs $LIBS"
;;
esac
if test "x$with_cloog_include" != x; then
--- 1582,1591 ----
*************** if test "x$with_cloog_include" != x; the
*** 1599,1610 ****
fi
if test "x$with_cloog_lib" != x; then
clooglibs="-L$with_cloog_lib -lcloog"
- LIBS="$clooglibs $LIBS"
fi
if test "x$with_cloog$with_cloog_include$with_cloog_lib" = x && test -d ${srcdir}/cloog; then
clooglibs='-L$$r/$(HOST_SUBDIR)/cloog/.libs -L$$r/$(HOST_SUBDIR)/cloog/_libs -lcloog '
clooginc='-I$$r/$(HOST_SUBDIR)/cloog/include -I$$s/cloog/include -DCLOOG_PPL_BACKEND '
- LIBS="$clooglibs $LIBS"
fi
AC_ARG_ENABLE(cloog-version-check,
--- 1593,1602 ----
*************** if test x"$enable_lto" = x"yes" ; then
*** 1644,1649 ****
--- 1636,1645 ----
AC_ARG_WITH(libelf_lib, [ --with-libelf-lib=PATH Specify the directory for the installed libelf library])
+ saved_CFLAGS="$CFLAGS"
+ saved_CPPFLAGS="$CPPFLAGS"
+ saved_LIBS="$LIBS"
+
case $with_libelf in
"")
libelflibs="-lelf"
*************** if test x"$enable_lto" = x"yes" ; then
*** 1672,1681 ****
LIBS="$libelflibs $LIBS"
fi
- saved_CFLAGS="$CFLAGS"
- saved_CPPFLAGS="$CPPFLAGS"
- saved_LIBS="$LIBS"
-
CFLAGS="$CFLAGS $libelfinc"
CPPFLAGS="$CPPFLAGS $libelfinc"
LIBS="$LIBS $libelflibs"
--- 1668,1673 ----
Index: configure
===================================================================
*** configure (revision 153055)
--- configure (working copy)
*************** case $with_ppl in
*** 5771,5782 ****
ppllibs=
;;
yes)
- LIBS="$ppllibs $LIBS"
;;
*)
ppllibs="-L$with_ppl/lib -lppl_c -lppl -lgmpxx"
pplinc="-I$with_ppl/include $pplinc"
- LIBS="$ppllibs $LIBS"
;;
esac
if test "x$with_ppl_include" != x; then
--- 5771,5780 ----
*************** if test "x$with_ppl_include" != x; then
*** 5784,5795 ****
fi
if test "x$with_ppl_lib" != x; then
ppllibs="-L$with_ppl_lib -lppl_c -lppl -lgmpxx"
- LIBS="$ppllibs $LIBS"
fi
if test "x$with_ppl$with_ppl_include$with_ppl_lib" = x && test -d ${srcdir}/ppl; then
ppllibs='-L$$r/$(HOST_SUBDIR)/ppl/.libs -L$$r/$(HOST_SUBDIR)/ppl/_libs -lppl_c -lppl -lgmpxx '
pplinc='-I$$r/$(HOST_SUBDIR)/ppl/include -I$$s/ppl/include '
- LIBS="$ppllibs $LIBS"
fi
# Check whether --enable-ppl-version-check was given.
--- 5782,5791 ----
*************** if ac_fn_c_try_compile "$LINENO"; then :
*** 5825,5831 ****
$as_echo "yes" >&6; }
else
{ $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
! $as_echo "no" >&6; }; ppllibs= ; pplinc=
fi
rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
CFLAGS="$saved_CFLAGS"
--- 5821,5827 ----
$as_echo "yes" >&6; }
else
{ $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
! $as_echo "no" >&6; }; ppllibs= ; pplinc= ; with_ppl=no
fi
rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
CFLAGS="$saved_CFLAGS"
*************** case $with_cloog in
*** 5871,5882 ****
clooginc=
;;
yes)
- LIBS="$clooglibs $LIBS"
;;
*)
clooglibs="-L$with_cloog/lib -lcloog"
clooginc="-I$with_cloog/include -DCLOOG_PPL_BACKEND "
- LIBS="$clooglibs $LIBS"
;;
esac
if test "x$with_cloog_include" != x; then
--- 5867,5876 ----
*************** if test "x$with_cloog_include" != x; the
*** 5884,5895 ****
fi
if test "x$with_cloog_lib" != x; then
clooglibs="-L$with_cloog_lib -lcloog"
- LIBS="$clooglibs $LIBS"
fi
if test "x$with_cloog$with_cloog_include$with_cloog_lib" = x && test -d ${srcdir}/cloog; then
clooglibs='-L$$r/$(HOST_SUBDIR)/cloog/.libs -L$$r/$(HOST_SUBDIR)/cloog/_libs -lcloog '
clooginc='-I$$r/$(HOST_SUBDIR)/cloog/include -I$$s/cloog/include -DCLOOG_PPL_BACKEND '
- LIBS="$clooglibs $LIBS"
fi
# Check whether --enable-cloog-version-check was given.
--- 5878,5887 ----
*************** if test "${with_libelf_lib+set}" = set;
*** 5967,5972 ****
--- 5959,5968 ----
fi
+ saved_CFLAGS="$CFLAGS"
+ saved_CPPFLAGS="$CPPFLAGS"
+ saved_LIBS="$LIBS"
+
case $with_libelf in
"")
libelflibs="-lelf"
*************** fi
*** 5995,6004 ****
LIBS="$libelflibs $LIBS"
fi
- saved_CFLAGS="$CFLAGS"
- saved_CPPFLAGS="$CPPFLAGS"
- saved_LIBS="$LIBS"
-
CFLAGS="$CFLAGS $libelfinc"
CPPFLAGS="$CPPFLAGS $libelfinc"
LIBS="$LIBS $libelflibs"
--- 5991,5996 ----
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] Fix graphite configure w/o ppl/cloog
2009-10-21 22:52 [PATCH] Fix graphite configure w/o ppl/cloog Richard Guenther
@ 2009-10-22 6:19 ` Eric Botcazou
2009-10-22 7:41 ` Paolo Bonzini
2009-10-23 23:51 ` Jason Merrill
2 siblings, 0 replies; 8+ messages in thread
From: Eric Botcazou @ 2009-10-22 6:19 UTC (permalink / raw)
To: Richard Guenther; +Cc: gcc-patches
> 2009-10-22 Richard Guenther <rguenther@suse.de>
>
> * configure.ac: Do not set LIBS for ppl/cloog checks. Disable
> cloog if the ppl version check failed. Move flags saving
> before setting in libelf check.
> * configure: Regenerate.
This breaks LTO without cloog:
checking for libelf/gelf.h... yes
checking for the correct version of libelf... no
The following languages will be built: c,ada,c++
configure:6482: checking for the correct version of libelf
configure:6505: gcc32 -o
conftest -g -O2 -I/usr/include/libelf -I/usr/include/libelf
conftest.c -lcloog -lppl_c -lppl -lgmpxx -lelf >&5
/usr/bin/ld: cannot find -lcloog
collect2: ld returned 1 exit status
--
Eric Botcazou
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] Fix graphite configure w/o ppl/cloog
2009-10-21 22:52 [PATCH] Fix graphite configure w/o ppl/cloog Richard Guenther
2009-10-22 6:19 ` Eric Botcazou
@ 2009-10-22 7:41 ` Paolo Bonzini
2009-10-22 7:44 ` Paolo Bonzini
2009-10-23 23:51 ` Jason Merrill
2 siblings, 1 reply; 8+ messages in thread
From: Paolo Bonzini @ 2009-10-22 7:41 UTC (permalink / raw)
To: Richard Guenther; +Cc: gcc-patches
On 10/22/2009 12:46 AM, Richard Guenther wrote:
>
> The configure shouldn't leave LIBS set, so better not set it at all as
> it's unused. Likewise the libelf LIBS saving is too late.
>
> Bootstrapped on x86_64-unknown-linux-gnu, ok for trunk?
Ok.
Paolo
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] Fix graphite configure w/o ppl/cloog
2009-10-22 7:41 ` Paolo Bonzini
@ 2009-10-22 7:44 ` Paolo Bonzini
2009-10-22 9:09 ` Richard Guenther
0 siblings, 1 reply; 8+ messages in thread
From: Paolo Bonzini @ 2009-10-22 7:44 UTC (permalink / raw)
Cc: Richard Guenther, gcc-patches
On 10/22/2009 08:19 AM, Paolo Bonzini wrote:
> On 10/22/2009 12:46 AM, Richard Guenther wrote:
>>
>> The configure shouldn't leave LIBS set, so better not set it at all as
>> it's unused. Likewise the libelf LIBS saving is too late.
>>
>> Bootstrapped on x86_64-unknown-linux-gnu, ok for trunk?
>
> Ok.
Please verify Eric's case of course before committing.
Paolo
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] Fix graphite configure w/o ppl/cloog
2009-10-22 7:44 ` Paolo Bonzini
@ 2009-10-22 9:09 ` Richard Guenther
2009-10-22 12:17 ` Eric Botcazou
0 siblings, 1 reply; 8+ messages in thread
From: Richard Guenther @ 2009-10-22 9:09 UTC (permalink / raw)
To: Paolo Bonzini; +Cc: gcc-patches
On Thu, 22 Oct 2009, Paolo Bonzini wrote:
> On 10/22/2009 08:19 AM, Paolo Bonzini wrote:
> > On 10/22/2009 12:46 AM, Richard Guenther wrote:
> > >
> > > The configure shouldn't leave LIBS set, so better not set it at all as
> > > it's unused. Likewise the libelf LIBS saving is too late.
> > >
> > > Bootstrapped on x86_64-unknown-linux-gnu, ok for trunk?
> >
> > Ok.
>
> Please verify Eric's case of course before committing.
Andrew reported the same issue which the patch was for, nevertheless
I verified it works when I drop cloog from my installation.
Committed as rev. 153447.
Richard.
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] Fix graphite configure w/o ppl/cloog
2009-10-22 9:09 ` Richard Guenther
@ 2009-10-22 12:17 ` Eric Botcazou
0 siblings, 0 replies; 8+ messages in thread
From: Eric Botcazou @ 2009-10-22 12:17 UTC (permalink / raw)
To: Richard Guenther; +Cc: gcc-patches, Paolo Bonzini
> Andrew reported the same issue which the patch was for, nevertheless
> I verified it works when I drop cloog from my installation.
Yes, it works fine, pilot error, sorry.
--
Eric Botcazou
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] Fix graphite configure w/o ppl/cloog
2009-10-21 22:52 [PATCH] Fix graphite configure w/o ppl/cloog Richard Guenther
2009-10-22 6:19 ` Eric Botcazou
2009-10-22 7:41 ` Paolo Bonzini
@ 2009-10-23 23:51 ` Jason Merrill
2009-10-24 11:13 ` Richard Guenther
2 siblings, 1 reply; 8+ messages in thread
From: Jason Merrill @ 2009-10-23 23:51 UTC (permalink / raw)
To: Richard Guenther; +Cc: gcc-patches
Incidentally, due to the recent breakage with libstdc++ symbol versions,
I noticed that linking against PPL (which is linked against libstdc++)
means that now while running the C++ testsuite, cc1plus loads the
just-built libstdc++.
Jason
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] Fix graphite configure w/o ppl/cloog
2009-10-23 23:51 ` Jason Merrill
@ 2009-10-24 11:13 ` Richard Guenther
0 siblings, 0 replies; 8+ messages in thread
From: Richard Guenther @ 2009-10-24 11:13 UTC (permalink / raw)
To: Jason Merrill; +Cc: Richard Guenther, gcc-patches
On Sat, Oct 24, 2009 at 1:25 AM, Jason Merrill <jason@redhat.com> wrote:
> Incidentally, due to the recent breakage with libstdc++ symbol versions, I
> noticed that linking against PPL (which is linked against libstdc++) means
> that now while running the C++ testsuite, cc1plus loads the just-built
> libstdc++.
Which is nothing bad, right?
Richard.
^ permalink raw reply [flat|nested] 8+ messages in thread
end of thread, other threads:[~2009-10-24 11:12 UTC | newest]
Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2009-10-21 22:52 [PATCH] Fix graphite configure w/o ppl/cloog Richard Guenther
2009-10-22 6:19 ` Eric Botcazou
2009-10-22 7:41 ` Paolo Bonzini
2009-10-22 7:44 ` Paolo Bonzini
2009-10-22 9:09 ` Richard Guenther
2009-10-22 12:17 ` Eric Botcazou
2009-10-23 23:51 ` Jason Merrill
2009-10-24 11:13 ` Richard Guenther
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).