[Bug bootstrap/54642] New: bootstrap broken using clang++

[Bug bootstrap/54642] New: bootstrap broken using clang++

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54642

             Bug #: 54642
           Summary: bootstrap broken using clang++
    Classification: Unclassified
           Product: gcc
           Version: 4.8.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: bootstrap
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: howarth@nitro.med.uc.edu


At r191562, the bootstrap of gcc trunk is broken against clang++ from Xcode
4.5, llvm clang 3.1 and llvm clang svn with the build failure...

clang++ -c   -g -O2 -DIN_GCC   -fno-exceptions -fno-rtti -W -Wall
-Wno-narrowing -Wwrite-strings -Wcast-qual -Wmissing-format-attribute -pedantic
-Wno-long-long -Wno-variadic-macros -Wno-overlength-strings -Werror -fno-common
 -DHAVE_CONFIG_H -I. -I. -I../../gcc-4.8-20120920/gcc
-I../../gcc-4.8-20120920/gcc/. -I../../gcc-4.8-20120920/gcc/../include
-I../../gcc-4.8-20120920/gcc/../libcpp/include -I/sw/include -I/sw/include 
-I../../gcc-4.8-20120920/gcc/../libdecnumber
-I../../gcc-4.8-20120920/gcc/../libdecnumber/dpd -I../libdecnumber
-DCLOOG_INT_GMP -I/sw/include -I/sw/include -I/sw/include
../../gcc-4.8-20120920/gcc/lto-symtab.c -o lto-symtab.o
clang: error: treating 'c' input as 'c++' when in C++ mode, this behavior is
deprecated

[Bug bootstrap/54642] bootstrap broken using clang++

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54642

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jakub at gcc dot gnu.org

--- Comment #1 from Jakub Jelinek <jakub at gcc dot gnu.org> 2012-09-20 15:48:10 UTC ---
-Werror should only be used in stage2 and later, unless you force in configure
otherwise.  And stage2 and later are built by gcc, not by other compilers.

[Bug bootstrap/54642] bootstrap broken using clang++

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54642

--- Comment #2 from Jack Howarth <howarth at nitro dot med.uc.edu> 2012-09-21 00:36:53 UTC ---
Confirmed this on a second machine at r191596. as well. The full context of the
error is...


 cd ../darwin_objdir
 make -j 1 install DESTDIR=/sw/src/fink.build/root-gcc48-4.8.0-1000
/bin/sh ../gcc-4.8-20120920/mkinstalldirs
/sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8
/sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8
mkdir /sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib
mkdir /sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8
rm -rf
/sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8/libexec/gcc/x86_64-apple-darwin12.2.0/4.8.0/install-tools
/bin/sh ../../gcc-4.8-20120920/fixincludes/../mkinstalldirs
/sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8/libexec/gcc/x86_64-apple-darwin12.2.0/4.8.0/install-tools 
mkdir /sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8/libexec
mkdir /sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8/libexec/gcc
mkdir
/sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8/libexec/gcc/x86_64-apple-darwin12.2.0
mkdir
/sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8/libexec/gcc/x86_64-apple-darwin12.2.0/4.8.0
mkdir
/sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8/libexec/gcc/x86_64-apple-darwin12.2.0/4.8.0/install-tools
/bin/sh ../../gcc-4.8-20120920/fixincludes/../mkinstalldirs
/sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8/lib/gcc/x86_64-apple-darwin12.2.0/4.8.0/install-tools/include
mkdir /sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8/lib
mkdir /sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8/lib/gcc
mkdir
/sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8/lib/gcc/x86_64-apple-darwin12.2.0
mkdir
/sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8/lib/gcc/x86_64-apple-darwin12.2.0/4.8.0
mkdir
/sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8/lib/gcc/x86_64-apple-darwin12.2.0/4.8.0/install-tools
mkdir
/sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8/lib/gcc/x86_64-apple-darwin12.2.0/4.8.0/install-tools/include
/sw/bin/ginstall -c -m 644 ../../gcc-4.8-20120920/fixincludes/README-fixinc \
     
/sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8/lib/gcc/x86_64-apple-darwin12.2.0/4.8.0/install-tools/include/README
/sw/bin/ginstall -c fixinc.sh
/sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8/libexec/gcc/x86_64-apple-darwin12.2.0/4.8.0/install-tools/fixinc.sh
/sw/bin/ginstall -c fixincl \
     
/sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8/libexec/gcc/x86_64-apple-darwin12.2.0/4.8.0/install-tools/fixincl
/sw/bin/ginstall -c mkheaders
/sw/src/fink.build/root-gcc48-4.8.0-1000/sw/lib/gcc4.8/libexec/gcc/x86_64-apple-darwin12.2.0/4.8.0/install-tools/mkheaders
g++ -c   -g -O2 -DIN_GCC   -fno-exceptions -fno-rtti -W -Wall -Wno-narrowing
-Wwrite-strings -Wcast-qual -Wmissing-format-attribute -pedantic -Wno-long-long
-Wno-variadic-macros -Wno-overlength-strings -Werror -fno-common 
-DHAVE_CONFIG_H -I. -I. -I../../gcc-4.8-20120920/gcc
-I../../gcc-4.8-20120920/gcc/. -I../../gcc-4.8-20120920/gcc/../include
-I../../gcc-4.8-20120920/gcc/../libcpp/include -I/sw/include -I/sw/include 
-I../../gcc-4.8-20120920/gcc/../libdecnumber
-I../../gcc-4.8-20120920/gcc/../libdecnumber/dpd -I../libdecnumber
-DCLOOG_INT_GMP -I/sw/include -I/sw/include -I/sw/include
../../gcc-4.8-20120920/gcc/lto-symtab.c -o lto-symtab.o
clang: error: treating 'c' input as 'c++' when in C++ mode, this behavior is
deprecated
make[2]: *** [lto-symtab.o] Error 1
make[1]: *** [install-gcc] Error 2
make: *** [install] Error 2

[Bug bootstrap/54642] bootstrap broken using clang++

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54642

--- Comment #3 from Jack Howarth <howarth at nitro dot med.uc.edu> 2012-09-21 03:23:51 UTC ---
Regression hunt so far shows that r191448 was okay but r191493 fails. These are
for builds configured as...

 ../gcc-4.8-20120920/configure --prefix=/sw --prefix=/sw/lib/gcc4.8
--mandir=/sw/share/man --infodir=/sw/lib/gcc4.8/info
--enable-languages=c,c++,fortran,lto,objc,obj-c++,java --with-gmp=/sw
--with-libiconv-prefix=/sw --with-isl=/sw --with-cloog=/sw --with-mpc=/sw
--with-system-zlib --enable-checking=yes --x-includes=/usr/X11R6/include
--x-libraries=/usr/X11R6/lib --program-suffix=-fsf-4.8

[Bug bootstrap/54642] bootstrap broken using clang++

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54642

--- Comment #4 from Jack Howarth <howarth at nitro dot med.uc.edu> 2012-09-21 05:08:47 UTC ---
Further regression hunting shows r191463 is okay but r191474 fails.

[Bug bootstrap/54642] bootstrap broken using clang++

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54642

--- Comment #5 from Jack Howarth <howarth at nitro dot med.uc.edu> 2012-09-21 13:49:09 UTC ---
This failure is caused by the commit...

Author: hubicka
Date: Wed Sep 19 10:01:56 2012
New Revision: 191466

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=191466
Log:

    * symtab.c (insert_to_assembler_name_hash): Do not insert
    register vars.
    (unlink_from_assembler_name_hash): NULL out pointers of unlinked
    var.
    (symtab_prevail_in_asm_name_hash): New.
    (symtab_initialize_asm_name_hash): Break out from ...
    (symtab_node_for_asm): ... here.
    (dump_symtab_base): Dump LTO file data.
    (verify_symtab_base): Register vars are not in symtab.
    * cgraph.h (symtab_initialize_asm_name_hash,
    symtab_prevail_in_asm_name_hash): New functions.
    (symtab_real_symbol_p): New inline.
    * lto-symtab.c: Do not include gt-lto-symtab.h.
    (lto_symtab_entry_def): Remove.
    (lto_symtab_entry_t): Remove.
    (lto_symtab_identifiers): Remove.
    (lto_symtab_free): Remove.
    (lto_symtab_entry_hash): Remove.
    (lto_symtab_entry_eq): Remove.
    (lto_symtab_entry_marked_p): Remove.
    (lto_symtab_maybe_init_hash_table): Remove.
    (resolution_guessed_p, set_resolution_guessed): New functions.
    (lto_symtab_register_decl): Only set resolution info.
    (lto_symtab_get, lto_symtab_get_resolution): Remove.
    (lto_symtab_merge): Reorg to work across symtab; do nothing if decls
    are same.
    (lto_symtab_resolve_replaceable_p): Reorg to work on symtab.
    (lto_symtab_resolve_can_prevail_p): Likewise; only real symbols can
    prevail.
    (lto_symtab_resolve_symbols): Reorg to work on symtab.
    (lto_symtab_merge_decls_2): Likewise.
    (lto_symtab_merge_decls_1): Likewise; add debug dumps.
    (lto_symtab_merge_decls): Likewise; do not merge at ltrans stage.
    (lto_symtab_merge_cgraph_nodes_1): Reorg to work on symtab.
    (lto_symtab_merge_cgraph_nodes): Likewise; do not merge at ltrans stage.
    (lto_symtab_prevailing_decl): Rewrite to lookup into symtab.
    * lto-streaer.h (lto_symtab_free): Remove.
    * lto-cgraph.c (add_references): Cleanup.
    * varpool.c (varpool_assemble_decl): Skip hard regs.

    * lto.c (lto_materialize_function): Update confused comment.
    (read_cgraph_and_symbols): Do not free symtab.

At r191465, the mkheaders installation step doesn't trigger the recompilation
of lto-symtab.o.

[Bug bootstrap/54642] bootstrap broken using clang++

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54642

--- Comment #6 from Jack Howarth <howarth at nitro dot med.uc.edu> 2012-09-21 14:00:03 UTC ---
The recompilation of lto-symtab.o during 'make install', but not the -Werror
related build failure, can be reproduced on x86_64-apple-darwin12 with...

% setenv CC clang
% setenv CXX clang++
% ../gcc-4.8-20120920/configure --prefix=/Users/howarth/dist
--enable-languages=c,c++,lto --with-gmp=/sw --with-libiconv-prefix=/sw
--with-isl=/sw --with-cloog=/sw --with-mpc=/sw --with-system-zlib
--enable-checking=yes --x-includes=/usr/X11R6/include
--x-libraries=/usr/X11R6/lib --disable-bootstrap
% make
% make install

where lto-symtab.o is recompiled starting with the r191466 commit. I assume
that nothing should be recompiled during 'make install' since that appears to
use the system compiler set with the CC/CXX environmentals and not the newly
built compiler.

[Bug bootstrap/54642] bootstrap broken using clang++

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54642

--- Comment #7 from Jack Howarth <howarth at nitro dot med.uc.edu> 2012-09-21 14:02:21 UTC ---
Created attachment 28242
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=28242
compressed build log at r191466 as described in comment 6

[Bug bootstrap/54642] bootstrap broken using clang++

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54642

--- Comment #8 from Jack Howarth <howarth at nitro dot med.uc.edu> 2012-09-21 14:04:09 UTC ---
Created attachment 28243
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=28243
compressed build log at r191465 as described in comment 6

[Bug bootstrap/54642] bootstrap broken by r191466 using clang++

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54642

--- Comment #9 from Jack Howarth <howarth at nitro dot med.uc.edu> 2012-09-21 14:38:26 UTC ---
FYI, this failure isn't suppressed by removing lto from the configure language
list as lto-symtab.o continues to be recompiled during the step...

/sw/bin/ginstall -c mkheaders
/Users/howarth/dist/libexec/gcc/x86_64-apple-darwin12.2.0/4.8.0/install-tools/mkheaders

in 'make install'. Is there any way to have 'make install' be more verbose
during the call to mkheaders so we can understand why r191466 suddenly
triggered the recompilation of lto-symtab.o during 'make install' on darwin?
Also does any other targets show a recompilation of  lto-symtab.o after
r191466?

[Bug bootstrap/54642] [4.8 regression] make install broken by r191466

[ebotcazou at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54642

Eric Botcazou <ebotcazou at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Target|x86_64-apple-darwin12       |
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2012-09-21
                 CC|                            |ebotcazou at gcc dot
                   |                            |gnu.org
               Host|x86_64-apple-darwin12       |
     Ever Confirmed|0                           |1
            Summary|bootstrap broken by r191466 |[4.8 regression] make
                   |using clang++               |install broken by r191466
   Target Milestone|---                         |4.8.0
           Severity|normal                      |critical
              Build|x86_64-apple-darwin12       |

--- Comment #10 from Eric Botcazou <ebotcazou at gcc dot gnu.org> 2012-09-21 14:52:03 UTC ---
We have it on Linux as well:

Make install failed
make[1]: Entering directory
`/red.a/gnatmail/gcc-x/build-red/x86_64-linux/gnat/obj'
/bin/bash ../src/mkinstalldirs
/red.a/gnatmail/gcc-x/build-red/x86_64-linux/gnat/gnat-bin
/red.a/gnatmail/gcc-x/build-red/x86_64-linux/gnat/gnat-bin
[...]
libtool: install: /usr/bin/install -c .libs/liblto_plugin.lai
/red.a/gnatmail/gcc-x/build-red/x86_64-linux/gnat/gnat-bin/libexec/gcc/x86_64-pc-linux-gnu/4.8.0/liblto_plugin.la
libtool: install: warning: remember to run `libtool --finish
/usr/gnat/libexec/gcc/x86_64-pc-linux-gnu/4.8.0'
make[3]: Nothing to be done for `install-data-am'.
make[3]: Leaving directory
`/red.a/gnatmail/gcc-x/build-red/x86_64-linux/gnat/obj/lto-plugin'
make[2]: Leaving directory
`/red.a/gnatmail/gcc-x/build-red/x86_64-linux/gnat/obj/lto-plugin'
make[2]: Entering directory
`/red.a/gnatmail/gcc-x/build-red/x86_64-linux/gnat/obj/gcc'
g++ -c   -g -O2 -DIN_GCC   -fno-exceptions -fno-rtti -W -Wall -Wno-narrowing
-Wwrite-strings -Wcast-qual -Wmissing-format-attribute -pedantic -Wno-long-long
-Wno-variadic-macros -Wno-overlength-strings -Werror -fno-common 
-DHAVE_CONFIG_H -I. -I. -I../../src/gcc -I../../src/gcc/.
-I../../src/gcc/../include -I../../src/gcc/../libcpp/include
-I/red.a/gnatmail/gcc-x/build-red/x86_64-linux/libmpfr/install/include
-I/red.a/gnatmail/gcc-x/build-red/x86_64-linux/libmpfr/install/include
-I/red.a/gnatmail/gcc-x/build-red/x86_64-linux/libmpfr/install/include 
-I../../src/gcc/../libdecnumber -I../../src/gcc/../libdecnumber/bid
-I../libdecnumber    ../../src/gcc/lto-symtab.c -o lto-symtab.o
cc1plus: error: unrecognized command line option "-Wno-narrowing"
cc1plus: error: unrecognized command line option "-Wno-overlength-strings"
make[2]: *** [lto-symtab.o] Error 1
make[2]: Leaving directory
`/red.a/gnatmail/gcc-x/build-red/x86_64-linux/gnat/obj/gcc'
make[1]: *** [install-gcc] Error 2
make[1]: Leaving directory
`/red.a/gnatmail/gcc-x/build-red/x86_64-linux/gnat/obj'
make: *** [install] Error 2

[Bug bootstrap/54642] [4.8 regression] make install broken by r191466

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54642

--- Comment #11 from Jack Howarth <howarth at nitro dot med.uc.edu> 2012-09-21 18:33:14 UTC ---
Seems like this could be due to the change....

    * lto-symtab.c: Do not include gt-lto-symtab.h.

so I would think we need this change as well, no?

Index: gcc/Makefile.in
===================================================================
--- gcc/Makefile.in    (revision 191617)
+++ gcc/Makefile.in    (working copy)
@@ -2156,7 +2156,7 @@ lto-section-out.o : lto-section-out.c $(
    $(DATA_STREAMER_H)
 lto-symtab.o: lto-symtab.c $(CONFIG_H) $(SYSTEM_H) coretypes.h \
    $(TREE_H) $(GIMPLE_H) $(GGC_H) $(HASHTAB_H) \
-   $(LTO_STREAMER_H) $(LINKER_PLUGIN_API_H) gt-lto-symtab.h
+   $(LTO_STREAMER_H) $(LINKER_PLUGIN_API_H)
 lto-opts.o: lto-opts.c $(CONFIG_H) $(SYSTEM_H) coretypes.h $(TREE_H) \
    $(HASHTAB_H) $(GGC_H) $(BITMAP_H) $(FLAGS_H) $(OPTS_H) $(OPTIONS_H) \
    $(COMMON_TARGET_H) $(DIAGNOSTIC_H) $(LTO_STREAMER_H)

[Bug bootstrap/54642] [4.8 regression] make install broken by r191466

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54642

--- Comment #12 from Jack Howarth <howarth at nitro dot med.uc.edu> 2012-09-21 19:21:49 UTC ---
The patch proposed in Comment 11 eliminates the bootstrap failures under
clang/clang++ on x86_64-apple-darwin12.

[Bug bootstrap/54642] [4.8 regression] make install broken by r191466

[ebotcazou at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54642

--- Comment #13 from Eric Botcazou <ebotcazou at gcc dot gnu.org> 2012-09-21 19:30:34 UTC ---
> Seems like this could be due to the change....
> 
>     * lto-symtab.c: Do not include gt-lto-symtab.h.
> 
> so I would think we need this change as well, no?
> 
> Index: gcc/Makefile.in
> ===================================================================
> --- gcc/Makefile.in    (revision 191617)
> +++ gcc/Makefile.in    (working copy)
> @@ -2156,7 +2156,7 @@ lto-section-out.o : lto-section-out.c $(
>     $(DATA_STREAMER_H)
>  lto-symtab.o: lto-symtab.c $(CONFIG_H) $(SYSTEM_H) coretypes.h \
>     $(TREE_H) $(GIMPLE_H) $(GGC_H) $(HASHTAB_H) \
> -   $(LTO_STREAMER_H) $(LINKER_PLUGIN_API_H) gt-lto-symtab.h
> +   $(LTO_STREAMER_H) $(LINKER_PLUGIN_API_H)
>  lto-opts.o: lto-opts.c $(CONFIG_H) $(SYSTEM_H) coretypes.h $(TREE_H) \
>     $(HASHTAB_H) $(GGC_H) $(BITMAP_H) $(FLAGS_H) $(OPTS_H) $(OPTIONS_H) \
>     $(COMMON_TARGET_H) $(DIAGNOSTIC_H) $(LTO_STREAMER_H)

Good catch!  Please install it as obvious if not already done.

[Bug bootstrap/54642] [4.8 regression] make install broken by r191466

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54642

--- Comment #14 from Jack Howarth <howarth at nitro dot med.uc.edu> 2012-09-21 19:34:48 UTC ---
Patch posted http://gcc.gnu.org/ml/gcc-patches/2012-09/msg01559.html. Can
someone commit this for me?

[Bug bootstrap/54642] [4.8 regression] make install broken by r191466

[hjl at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54642

--- Comment #15 from hjl at gcc dot gnu.org <hjl at gcc dot gnu.org> 2012-09-21 20:07:42 UTC ---
Author: hjl
Date: Fri Sep 21 20:07:36 2012
New Revision: 191618

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=191618
Log:
Remove lto-symtab.o's dependency on gt-lto-symtab.h

    PR bootstrap/54642
    * Makefile.in (lto-symtab.o): Remove gt-lto-symtab.h dependency.

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/Makefile.in

[Bug bootstrap/54642] [4.8 regression] make install broken by r191466

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54642

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED

--- Comment #16 from H.J. Lu <hjl.tools at gmail dot com> 2012-09-21 20:09:05 UTC ---
Fixed.