[Bug c/26508] New: 4.1.0 doesn't build in 64bit on PA-RISC

[Bug c/26508] New: 4.1.0 doesn't build in 64bit on PA-RISC

[h dot m dot brand at xs4all dot nl]

in 32bit mode all went fine, but in 64bit mode I get:



stage1/xgcc -Bstage1/ -B/usr/local/pa20_64/hppa64-hp-hpux11.11/bin/   -g -O2
-DI
N_GCC   -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes
-pedan
tic -Wno-long-long -Wno-variadic-macros -Wold-style-definition
-Wmissing-format-
attribute     -DHAVE_CONFIG_H -DGENERATOR_FILE  -o build/gencodes \
        build/gencodes.o build/rtl.o build/read-rtl.o build/ggc-none.o
build/min
-insn-modes.o build/gensupport.o build/insn-conditions.o build/print-rtl.o \
        build/errors.o ../build-hppa64-hp-hpux11.11/libiberty/libiberty.a
build/gencodes ../../src/gcc/config/pa/pa.md > tmp-codes.h
/usr/lib/pa20_64/dld.sl: Unsatisfied code symbol '__deregister_frame_info' in
lo
ad module 'build/gencodes'.
/usr/lib/pa20_64/dld.sl: Unsatisfied code symbol '__register_frame_info' in
load
 module 'build/gencodes'.
/bin/sh: 29400 Killed

--8<---
export CONFIG_SITE=
export CC=gcc64
export CFLAGS="-mpa-risc-2-0 -fPIC"
export CXXFLAGS="-mpa-risc-2-0 -fPIC"
export PATH=.:/usr/local/pa20_64/bin:/pro/local/bin
export PATH=$PATH"":/usr/bin:/opt/ansic/bin:/usr/ccs/bin:/opt/langtools/bin
export PATH=$PATH"":/opt/imake/bin

rm -rf obj-64
mkdir obj-64
cd obj-64
../src/configure \
    --enable-languages=c,c++ \
    --prefix=/usr/local/pa20_64 --with-local-prefix=/usr/local/pa20_64 \
    --with-gnu-as --with-as=/usr/local/pa20_64/bin/as \
    --with-gnu-ld --with-ld=/usr/local/pa20_64/bin/ld \
    --disable-shared \
    --disable-nls \
    --host=hppa64-hp-hpux11.11
make bootstrap-lean
-->8---


-- 
           Summary: 4.1.0 doesn't build in 64bit on PA-RISC
           Product: gcc
           Version: 4.1.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: h dot m dot brand at xs4all dot nl
 GCC build triplet: hppa64-hp-hpux11.11
  GCC host triplet: hppa64-hp-hpux11.11
GCC target triplet: hppa64-hp-hpux11.11


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[h dot m dot brand at xs4all dot nl]

------- Comment #1 from h dot m dot brand at xs4all dot nl  2006-03-01 12:56 -------
Same on HP-UX 11.00, where I used gcc-4.0.2/64 as initial compiler
The initial compiler for the original 11i (11.11v1) report was gcc-3.4.5/64

stage1/xgcc -Bstage1/ -B/usr/local/pa20_64/hppa64-hp-hpux11.00/bin/ -c   -g -O2
-DIN_GCC   -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes
-pedantic -Wno-long-long -Wno-variadic-macros -Wold-style-definition
-Wmissing-format-attribute     -DHAVE_CONFIG_H -DGENERATOR_FILE -I. -Ibuild
-I../../src/gcc -I../../src/gcc/build -I../../src/gcc/../include
-I../../src/gcc/../libcpp/include     -o build/errors.o ../../src/gcc/errors.c
stage1/xgcc -Bstage1/ -B/usr/local/pa20_64/hppa64-hp-hpux11.00/bin/   -g -O2
-DIN_GCC   -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes
-pedantic -Wno-long-long -Wno-variadic-macros -Wold-style-definition
-Wmissing-format-attribute     -DHAVE_CONFIG_H -DGENERATOR_FILE  -o
build/genmodes \
 build/genmodes.o build/errors.o
../build-hppa64-hp-hpux11.00/libiberty/libiberty.a
build/genmodes -h > tmp-modes.h
/usr/lib/pa20_64/dld.sl: Unsatisfied code symbol '__deregister_frame_info' in
load module 'build/genmodes'.
/usr/lib/pa20_64/dld.sl: Unsatisfied code symbol '__register_frame_info' in
load module 'build/genmodes'.
/bin/sh: 7042 Killed


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[pinskia at gcc dot gnu dot org]

------- Comment #2 from pinskia at gcc dot gnu dot org  2006-03-01 13:04 -------
http://www.cygwin.com/ml/binutils/2002-07/msg00702.html

What binutils version are you using?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[h dot m dot brand at xs4all dot nl]

------- Comment #3 from h dot m dot brand at xs4all dot nl  2006-03-01 13:15 -------
GNU assembler 2.16.1
Copyright 2005 Free Software Foundation, Inc.
This program is free software; you may redistribute it under the terms of
the GNU General Public License.  This program has absolutely no warranty.
This assembler was configured for a target of `hppa64-hp-hpux11.00'.

For the possitive news: All well on HP-UX 11.23 on Itanium2.
HP-UX 10.20 is still compiling


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[h dot m dot brand at xs4all dot nl]

------- Comment #4 from h dot m dot brand at xs4all dot nl  2006-03-01 13:45 -------
(temporary) workaround:

.../obj-64/gcc > cat xx.c
int _not_called_but_just_to_force_ref ()
{
    __deregister_frame_info ();
    __register_frame_info ();
    return (0);
    } /* ... */
.../obj-64/gcc > stage1/xgcc -Bstage1/
-B/usr/local/pa20_64/hppa64-hp-hpux11.11/bin/ -c -o xx.o xx.c
.../obj-64/gcc > stage1/xgcc \
    -Bstage1/ -B/usr/local/pa20_64/hppa64-hp-hpux11.11/bin/ \
    -g -O2 \
    -DIN_GCC \
    -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes \
    -pedantic \
    -Wno-long-long -Wno-variadic-macros -Wold-style-definition
-Wmissing-format-attribute \
    -DHAVE_CONFIG_H -DGENERATOR_FILE \
    -o build/gencodes \
    build/gencodes.o build/rtl.o build/read-rtl.o build/ggc-none.o \
    build/min-insn-modes.o build/gensupport.o build/insn-conditions.o \
    build/print-rtl.o build/errors.o xx.o \
    ../build-hppa64-hp-hpux11.11/libiberty/libiberty.a \
    stage1/libgcc.a
.../obj-64/gcc >

makes gencodes work again.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[pinskia at gcc dot gnu dot org]

------- Comment #5 from pinskia at gcc dot gnu dot org  2006-03-01 14:37 -------
Can you try without using the GNU binutils because this looks like a bug in
binutils and not in GCC?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[h dot m dot brand at xs4all dot nl]

------- Comment #6 from h dot m dot brand at xs4all dot nl  2006-03-01 15:01 -------
Continuing ...

.../obj-64/gcc > cat xx.c
int _not_called_but_just_to_force_ref ()
{
    __deregister_frame_info ();
    __register_frame_info ();
    return (0);
    } /* ... */
.../obj-64/gcc > stage1/xgcc -Bstage1/
-B/usr/local/pa20_64/hppa64-hp-hpux11.11/bin/ -c -o xx.o xx.c
.../obj-64/gcc > perl -pi -e'm/^BUILD_LDFLAGS=/ and s/$/ xx.o/' Makefile
.../obj-64/gcc >

makes it finish stage1 to here:

stage1/xgcc -Bstage1/ -B/usr/local/pa20_64/hppa64-hp-hpux11.11/bin/   -g -O2
-DIN_GCC   -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes
-pedantic -Wno-long-long -Wno-variadic-macros -Wold-style-definition
-Wmissing-format-attribute     -DHAVE_CONFIG_H  -o cpp gcc.o cppspec.o intl.o \
  prefix.o version.o  ../libcpp/libcpp.a   ../libiberty/libiberty.a
./xgcc -B./ -B/usr/local/pa20_64/hppa64-hp-hpux11.11/bin/ -isystem
/usr/local/pa20_64/hppa64-hp-hpux11.11/include -isystem
/usr/local/pa20_64/hppa64-hp-hpux11.11/sys-include
-L/pro/3gl/GNU/gcc/obj-64/gcc/../ld -dumpspecs > tmp-specs
/usr/lib/pa20_64/dld.sl: Unsatisfied code symbol '__deregister_frame_info' in
load module './xgcc'.
/usr/lib/pa20_64/dld.sl: Unsatisfied code symbol '__register_frame_info' in
load module './xgcc'.

changing $(LDFLAGS) to $(BUILD_LDFLAGS) for all targets brings me safely to
stage2.
There I have to compile xx.c again, because the cleanup removed xx.o

.../obj-64/gcc > stage2/xgcc -Bstage2/
-B/usr/local/pa20_64/hppa64-hp-hpux11.11/bin/ -c -o xx.o xx.c

If I now get to the end, I'll install it and try without binutils' as


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[h dot m dot brand at xs4all dot nl]

------- Comment #7 from h dot m dot brand at xs4all dot nl  2006-03-01 15:21 -------
echo stage3_build > stage_last

Bootstrap complete - make "quickstrap" to redo last build,
or "cleanstrap" to redo the bootstrap from scratch.
make[1]: Leaving directory `/pro/3gl/GNU/gcc/obj-64/gcc'
Comparing stage2 and stage3 of the compiler
make[1]: Entering directory `/pro/3gl/GNU/gcc/obj-64/gcc'
rm -f .bad_compare
:
:
  else true; fi; \
done
warning: ./cc1-checksum.o differs
warning: ./cc1plus-checksum.o differs
rm -f tmp-foo*
:
:
  esac; true; \
fi
Bootstrap comparison failure!
./attribs.o differs
./c-aux-info.o differs
./c-common.o differs
:
:
build/min-insn-modes.o differs
build/print-rtl.o differs
build/read-rtl.o differs
build/rtl.o differs
make[1]: *** [gnucompare-lean] Error 1

:P


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[h dot m dot brand at xs4all dot nl]

------- Comment #8 from h dot m dot brand at xs4all dot nl  2006-03-01 15:54 -------
Using HP's 'as' is not an option (which in turn might be called a gcc bug)

./xgcc -B./ -B/usr/local/pa20_64/hppa64-hp-hpux11.11/bin/ -isystem
/usr/local/pa20_64/hppa64-hp-hpux11.11/include -isystem
/usr/local/pa20_64/hppa64-hp-hpux11.11/sys-include
-L/pro/3gl/GNU/gcc/obj-64/gcc/../ld -O2 -O2 -mpa-risc-2-0 -fPIC   -DIN_GCC   
-W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes
-Wold-style-definition  -isystem ./include  -I. -I. -I../../src/gcc
-I../../src/gcc/. -I../../src/gcc/../include -I../../src/gcc/../libcpp/include 
 -g0 -finhibit-size-directive -fno-inline-functions -fno-exceptions
-fno-zero-initialized-in-bss -fno-unit-at-a-time   \
  -c ../../src/gcc/crtstuff.c -DCRT_BEGIN \
  -o crtbegin.o
../../src/gcc/crtstuff.c:1: warning: -g is only supported when using GAS on
this processor,
../../src/gcc/crtstuff.c:1: warning: -g option disabled
/tmp/ccPcUKIS.s: Assembler messages:
/tmp/ccPcUKIS.s:2: Error: unsupported variable size or fill value
/tmp/ccPcUKIS.s:2: Error: junk at end of line, first unrecognized character is
`='
/tmp/ccPcUKIS.s:3: Error: unknown pseudo-op: `.subspa'
/tmp/ccPcUKIS.s:4: Error: unknown pseudo-op: `.subspa'
/tmp/ccPcUKIS.s:5: Error: unsupported variable size or fill value
/tmp/ccPcUKIS.s:5: Error: junk at end of line, first unrecognized character is
`='
/tmp/ccPcUKIS.s:6: Error: unknown pseudo-op: `.subspa'
/tmp/ccPcUKIS.s:7: Error: unknown pseudo-op: `.subspa'
/tmp/ccPcUKIS.s:8: Error: unknown pseudo-op: `.subspa'
/tmp/ccPcUKIS.s:13: Error: unknown pseudo-op: `.subspa'
/tmp/ccPcUKIS.s:17: Error: unknown pseudo-op: `.subspa'
/tmp/ccPcUKIS.s:21: Error: unknown pseudo-op: `.subspa'
/tmp/ccPcUKIS.s:59: Error: unknown pseudo-op: `.subspa'
/tmp/ccPcUKIS.s:62: Error: unknown pseudo-op: `.subspa'
/tmp/ccPcUKIS.s:65: Error: unknown pseudo-op: `.subspa'
/tmp/ccPcUKIS.s:67: Error: unknown pseudo-op: `.subspa'
/tmp/ccPcUKIS.s:71: Error: unknown pseudo-op: `.subspa'
/tmp/ccPcUKIS.s:76: Error: unknown pseudo-op: `.subspa'
/tmp/ccPcUKIS.s:81: Error: unknown pseudo-op: `.subspa'
/tmp/ccPcUKIS.s:86: Error: unknown pseudo-op: `.subspa'
make[2]: *** [crtbegin.o] Error 1


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[steven at gcc dot gnu dot org]

------- Comment #9 from steven at gcc dot gnu dot org  2006-03-02 17:47 -------
Not being able to use the HP assembler is definitely not a GCC bug.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[pinskia at gcc dot gnu dot org]

------- Comment #10 from pinskia at gcc dot gnu dot org  2006-03-05 15:51 -------
(In reply to comment #7)
> Bootstrap comparison failure!
> ./attribs.o differs
> ./c-aux-info.o differs
> ./c-common.o differs

Yes this is because HPUX's as puts in timestamps into the .s file
Please read:
http://gcc.gnu.org/install/specific.html#hppa-hp-hpux11


Though it does not mention this issue but you should try a different version of
binutils first.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[h dot m dot brand at xs4all dot nl]

------- Comment #11 from h dot m dot brand at xs4all dot nl  2006-03-05 16:46 -------
To Steven: I'm not pushing, but IMHO, I think it could be said that it is a bug
in gcc if gcc does not generate valid HP PA-RISC assembly code *as understood
by the native compiler*.

To Andrew: For the last remark, what other version of binutils should/could I
try? 2.16.1 was the most recent as far as I can see, and
ftp://alpha.gnu.org/gnu/ doesn't have a binutils.
I'm very willing to help here, and can try out almost anything.

Would it help if I tried a more recent 4.2 snapshot?
ftp://ftp.fu-berlin.de/unix/languages/gcc/snapshots/4.2-20060304/


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[danglin at gcc dot gnu dot org]

------- Comment #12 from danglin at gcc dot gnu dot org  2006-03-11 15:14 -------
> /usr/lib/pa20_64/dld.sl: Unsatisfied code symbol '__deregister_frame_info' in
> load module 'build/gencodes'

Which HP linker patch are you using?  There is a problem with the most
recent couple of patch releases in that they don't appear to handle PLABEL
relocations correctly.  Try PHSS_30968 or the one before it.

The symbols should be in libgcc_s.sl or libgcc_eh.a.  You could add
'-Wl,-v' or '-Wl,-debug' to the link command for gencodes to see exactly
how gencodes has been linked.

> GNU assembler 2.16.1

Too old.  I know it's the last release but it contains a serious bug
handling 64-bit data relocations (symbol differences, etc) that was
fixed near the end of last November.  You have to use CVS GAS.

> Using HP's 'as' is not an option (which in turn might be called a gcc bug)

correct, it's not really an option.  Steve and I discussed removing it as an
option but we never got around to doing it.  HP 'as' doesn't have the features
needed for C++, etc.  If you will notice, all HP compilers other than C don't
use it.  They emit binary code directly.


-- 

danglin at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |danglin at gcc dot gnu dot
                   |                            |org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[h dot m dot brand at xs4all dot nl]

------- Comment #13 from h dot m dot brand at xs4all dot nl  2006-03-11 17:11 -------
HP-UX 11.00:
   8406 100555 -r-x    1      bin    864256  25 Jan 2005 06:52
        /usr/ccs/bin/ld:
        $Revision: 92453-07 linker linker crt0.o B.11.37 040218 $
        HP aC++ B3910B A.03.52 Classic Iostream Library
        HP aC++ B3910B A.03.52 Language Support Library
        ld_msgs.cat: $Revision: 1.85 $
        92453-07 linker command s800.sgs ld PA64 B.11.43 REL 050124

        /usr/ccs/bin/ld:
     $Revision: 92453-07 linker linker crt0.o B.11.37 040218 $
     $Revision: 1.85 $

# BUNDLE.PHSS_30969                                             1.0           
ld(1) and linker tools cumulative patch
# QPK1100.PHSS_28869                                            1.0           
ld(1) and linker tools cumulative patch

So I am running 30969, and 30968 doesn't seem to be a ld patch

HP-UX 11.11:

# BUNDLE.PHSS_30970                                             1.0           
ld(1) and linker tools cumulative patch

  15639 100555 -r-x    1      bin    864256  27 Jan 2005 06:21
        /usr/ccs/bin/ld:
        $Revision: 92453-07 linker linker crt0.o B.11.37 040218 $
        HP aC++ B3910B A.03.52 Classic Iostream Library
        HP aC++ B3910B A.03.52 Language Support Library
        ld_msgs.cat: $Revision: 1.85 $
        92453-07 linker command s800.sgs ld PA64 B.11.43 REL 050124

        /usr/ccs/bin/ld:
     $Revision: 92453-07 linker linker crt0.o B.11.37 040218 $
     $Revision: 1.85 $



After installing  PHSS_33033 & PHSS_33901  on 11.11, I still get


build/genmodes -h > tmp-modes.h
/usr/lib/pa20_64/dld.sl: Unsatisfied code symbol '__deregister_frame_info' in
load module 'build/genmodes'.
/usr/lib/pa20_64/dld.sl: Unsatisfied code symbol '__register_frame_info' in
load module 'build/genmodes'.

I have no time to install gnu binutils from cvs. If you have a snapshot url for
me, I can try it.

Thanks for still looking into it


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[h dot m dot brand at xs4all dot nl]

------- Comment #14 from h dot m dot brand at xs4all dot nl  2006-03-11 19:00 -------
found the binutils snapshots

r3:/pro/3gl/GNU/gcc 131 > gas --version
GNU assembler 060311 20060311
Copyright 2005 Free Software Foundation, Inc.
This program is free software; you may redistribute it under the terms of
the GNU General Public License.  This program has absolutely no warranty.
This assembler was configured for a target of `hppa64-hp-hpux11.11'.
r3:/pro/3gl/GNU/gcc 132 >

stage1/xgcc -Bstage1/ -B/usr/local/pa20_64/hppa64-hp-hpux11.11/bin/   -g -O2
-DIN_GCC   -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes
-pedantic -Wno-long-long -Wno-variadic-macros -Wold-style-definition
-Wmissing-format-attribute     -DHAVE_CONFIG_H -DGENERATOR_FILE  -o
build/genmodes \
 build/genmodes.o build/errors.o
../build-hppa64-hp-hpux11.11/libiberty/libiberty.a
build/genmodes -h > tmp-modes.h
/usr/lib/pa20_64/dld.sl: Unsatisfied code symbol '__deregister_frame_info' in
load module 'build/genmodes'.
/usr/lib/pa20_64/dld.sl: Unsatisfied code symbol '__register_frame_info' in
load module 'build/genmodes'.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[dave at hiauly1 dot hia dot nrc dot ca]

------- Comment #15 from dave at hiauly1 dot hia dot nrc dot ca  2006-03-11 23:57 -------
Subject: Re:  4.1.0 doesn't build in 64bit on PA-RISC

> found the binutils snapshots

Looking at your report, I see that you configured with GNU ld.
Don't use it, it needs work.  HP as isn't usable with C++, and
hasn't been tested in a long time.  Use GAS and HP ld.

You don't need "-mpa-risc-2-0 -fPIC".  This is the default for
this target.

Sample configure commands for successful builds can be found in
<http://gcc.gnu.org/ml/gcc-testresults/>.

Dave


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[dave at hiauly1 dot hia dot nrc dot ca]

------- Comment #16 from dave at hiauly1 dot hia dot nrc dot ca  2006-03-12 00:29 -------
Subject: Re:  4.1.0 doesn't build in 64bit on PA-RISC

> # BUNDLE.PHSS_30969                                             1.0           
> ld(1) and linker tools cumulative patch
> # QPK1100.PHSS_28869                                            1.0           
> ld(1) and linker tools cumulative patch
> 
> So I am running 30969, and 30968 doesn't seem to be a ld patch

30969 is for 11.00.  HP has issued a critical warning for it.  It is
the mate to 30970.

30968 is the 11.11 ld patch just before 30970.  It's been my experience
that 30969/30970 and all subsequent patches don't work correctly.  30968
works but it has a critical warning against it.  I believe that 30966
works and it doesn't have a critical warning against it.  It's 11.00
mate is 30965.

> build/genmodes -h > tmp-modes.h
> /usr/lib/pa20_64/dld.sl: Unsatisfied code symbol '__deregister_frame_info' in
> load module 'build/genmodes'.
> /usr/lib/pa20_64/dld.sl: Unsatisfied code symbol '__register_frame_info' in
> load module 'build/genmodes'.

Are you still using GNU ld?  These symbols are not present in genmodes
in my builds.

Dave


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[h dot m dot brand at xs4all dot nl]

------- Comment #17 from h dot m dot brand at xs4all dot nl  2006-03-12 12:26 -------
Super! I moved binutil's ld out of the way, removed those two default CCFLAGS,
Conf/Build/Install worked fine.

r3:/pro/3gl/CPAN/perl-current 129 > /usr/local/pa20_64/bin/gcc --version
gcc (GCC) 4.1.0
Copyright (C) 2006 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

r3:/pro/3gl/CPAN/perl-current 130 > /usr/local/pa20_64/bin/gas --version
GNU assembler 060311 20060311
Copyright 2005 Free Software Foundation, Inc.
This program is free software; you may redistribute it under the terms of
the GNU General Public License.  This program has absolutely no warranty.
This assembler was configured for a target of `hppa64-hp-hpux11.11'.
r3:/pro/3gl/CPAN/perl-current 131 > what `path ld`
/usr/bin/ld:
        $Revision: 92453-07 linker linker crt0.o B.11.37 040218 $
        HP aC++ B3910B A.03.52 Classic Iostream Library
        HP aC++ B3910B A.03.52 Language Support Library
        ld_msgs.cat: $Revision: 1.85 $
        92453-07 linker command s800.sgs ld PA64 B.11.47 REL 051005
r3:/pro/3gl/CPAN/perl-current 132 >

I will put these binaries on my site soon.
Thanks.

(I tried to close this bug, but I got re-assign errors. Using Opera 9)


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[dave at hiauly1 dot hia dot nrc dot ca]

------- Comment #18 from dave at hiauly1 dot hia dot nrc dot ca  2006-03-12 22:05 -------
Subject: Re:  4.1.0 doesn't build in 64bit on PA-RISC

> (I tried to close this bug, but I got re-assign errors. Using Opera 9)

Don't.  This is a regression.  It's caused by the addition of EH exception
support.  The issue is the handling of weak symbols by GNU ld.  The
approach taken previously was to add stubs.  I'm looking at this.

Dave


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[h dot m dot brand at xs4all dot nl]

------- Comment #19 from h dot m dot brand at xs4all dot nl  2006-03-13 18:03 -------
As development perl was unable to complete it's testsuite with the installed
4.1.0/64, I went back to 4.0.3, where all works well.

FYI pack and udp failures.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[dave at hiauly1 dot hia dot nrc dot ca]

------- Comment #20 from dave at hiauly1 dot hia dot nrc dot ca  2006-03-13 18:10 -------
Subject: Re:  4.1.0 doesn't build in 64bit on PA-RISC

> As development perl was unable to complete it's testsuite with the installed
> 4.1.0/64, I went back to 4.0.3, where all works well.
> 
> FYI pack and udp failures.

There may not be another 4.0 release, so it would be useful if you
could determine why these tests fail and submit PRs.

Dave


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[dave at hiauly1 dot hia dot nrc dot ca]

------- Comment #21 from dave at hiauly1 dot hia dot nrc dot ca  2006-03-23 17:49 -------
Subject: Re:  4.1.0 doesn't build in 64bit on PA-RISC

> stage1/xgcc -Bstage1/ -B/usr/local/pa20_64/hppa64-hp-hpux11.11/bin/   -g -O2
> -DIN_GCC   -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes
> -pedantic -Wno-long-long -Wno-variadic-macros -Wold-style-definition
> -Wmissing-format-attribute     -DHAVE_CONFIG_H -DGENERATOR_FILE  -o
> build/genmodes \
>  build/genmodes.o build/errors.o
> ../build-hppa64-hp-hpux11.11/libiberty/libiberty.a
> build/genmodes -h > tmp-modes.h
> /usr/lib/pa20_64/dld.sl: Unsatisfied code symbol '__deregister_frame_info' in
> load module 'build/genmodes'.
> /usr/lib/pa20_64/dld.sl: Unsatisfied code symbol '__register_frame_info' in
> load module 'build/genmodes'.

I've looked at this some more and come to the conclusion that it's a
GNU ld bug.  The problem seems to be that weak undefined references are
not being resolved by the linker.  The symbols are defined in libgcc_eh.a
and -lgcc_eh is provided by the gcc driver in linking genmodes.

Dave


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[h dot m dot brand at xs4all dot nl]

------- Comment #22 from h dot m dot brand at xs4all dot nl  2006-03-23 18:11 -------
http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2006-03/msg00524.html

shows that I've been trying to pin the perl5 failure related to this too, but
I'm kinda stuck for now.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[danglin at gcc dot gnu dot org]

------- Comment #23 from danglin at gcc dot gnu dot org  2006-04-09 18:32 -------
Subject: Bug 26508

Author: danglin
Date: Sun Apr  9 18:32:16 2006
New Revision: 112807

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=112807
Log:
        PR target/27034
        PR target/26508
        * config.gcc (hppa*64*-*-hpux11*): Add 'libgcc_stub.a' to extra_parts.
        * pa64-hpux.h (LIB_SPEC): Correct typo.  Don't append milli.a.
        (LINK_GCC_C_SEQUENCE_SPEC): Define.  Append milli.a here.
        (PA_CXA_FINALIZE_STUB, PA_JV_REGISTERCLASSES_STUB): Delete defines.
        (PA_INIT_FINI_HACK): Rename to PA_CRTBEGIN_HACK.
        (PA_CRTBEGIN_HACK): Ensure __do_global_ctors_aux is placed in text
        section.  Delete PA_CXA_FINALIZE_STUB and PA_JV_REGISTERCLASSES_STUB
        stubs.  Add ".align 8" directives before all ".dword" directives.
        (GTHREAD_USE_WEAK): Revise comment.
        (TARGET_ATTRIBUTE_WEAK): Define.
        * pa/t-pa64 (LIBGCCSTUB_OBJS): Add new rules for stublib.c, rfi-stub.o,
        dfi-stub.o, cxaf-stub.o and jvrc-stub.o.
        * stublib.c: New file.


Added:
    trunk/gcc/config/pa/stublib.c
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/config.gcc
    trunk/gcc/config/pa/pa64-hpux.h
    trunk/gcc/config/pa/t-pa64


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[danglin at gcc dot gnu dot org]

------- Comment #24 from danglin at gcc dot gnu dot org  2006-04-09 18:56 -------
Subject: Bug 26508

Author: danglin
Date: Sun Apr  9 18:56:23 2006
New Revision: 112808

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=112808
Log:
        PR target/26508
        * config.gcc (hppa*64*-*-hpux11*): Add 'libgcc_stub.a' to extra_parts.
        * pa64-hpux.h (LIB_SPEC): Don't append milli.a.
        (LINK_GCC_C_SEQUENCE_SPEC): Define.  Append milli.a here.
        (PA_CXA_FINALIZE_STUB, PA_JV_REGISTERCLASSES_STUB): Delete defines.
        (PA_INIT_FINI_HACK): Rename to PA_CRTBEGIN_HACK.
        (PA_CRTBEGIN_HACK): Ensure __do_global_ctors_aux is placed in text
        section.  Delete PA_CXA_FINALIZE_STUB and PA_JV_REGISTERCLASSES_STUB
        stubs.  Add ".align 8" directives before all ".dword" directives.
        (GTHREAD_USE_WEAK): Revise comment.
        (TARGET_ATTRIBUTE_WEAK): Define.
        * pa/t-pa64 (LIBGCCSTUB_OBJS): Add new rules for stublib.c, rfi-stub.o,
        dfi-stub.o, cxaf-stub.o and jvrc-stub.o.
        * stublib.c: New file.


Added:
    branches/gcc-4_1-branch/gcc/config/pa/stublib.c
Modified:
    branches/gcc-4_1-branch/gcc/ChangeLog
    branches/gcc-4_1-branch/gcc/config.gcc
    branches/gcc-4_1-branch/gcc/config/pa/pa64-hpux.h
    branches/gcc-4_1-branch/gcc/config/pa/t-pa64


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508

[Bug target/26508] 4.1.0 doesn't build in 64bit on PA-RISC

[danglin at gcc dot gnu dot org]

------- Comment #25 from danglin at gcc dot gnu dot org  2006-04-09 20:39 -------
This should now be fixed on trunk and 4.1.  Note however, that the
dwarf2 EH support is broken with GNU ld.  libgfortran doesn't build.
There are a few testsuite fails beyond that with HP ld.


-- 

danglin at gcc dot gnu dot org changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26508