[Bug libgcc/62097] New: mipsel-sde-elf build fails on OS X 10.7 host

[Bug libgcc/62097] New: mipsel-sde-elf build fails on OS X 10.7 host

[Anders.Montonen at iki dot fi]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=62097

            Bug ID: 62097
           Summary: mipsel-sde-elf build fails on OS X 10.7 host
           Product: gcc
           Version: 4.9.1
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: libgcc
          Assignee: unassigned at gcc dot gnu.org
          Reporter: Anders.Montonen at iki dot fi

Created attachment 33294
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=33294&action=edit
Make all log

Trying to build a crosscompiler targeting mipsel-sde-elf on an OS X 10.7 host
fails with the following error:

# If this is the top-level multilib, build all the other
# multilibs.
DEFINES='' HEADERS='' \
        ../../../../gcc-4.9.1/libgcc/mkheader.sh > tmp-libgcc_tm.h
/bin/sh ../../../../gcc-4.9.1/libgcc/../move-if-change tmp-libgcc_tm.h
libgcc_tm.h
echo timestamp > libgcc_tm.stamp
Mode = sf\|df
Suffix = si\|2\|3
/Users/anders/work/toolchain/gcc/4.9.1/mipsel-sde-elf/./gcc/xgcc
-B/Users/anders/work/toolchain/gcc/4.9.1/mipsel-sde-elf/./gcc/ -nostdinc
-B/Users/anders/work/toolchain/gcc/4.9.1/mipsel-sde-elf/mipsel-sde-elf/newlib/
-isystem
/Users/anders/work/toolchain/gcc/4.9.1/mipsel-sde-elf/mipsel-sde-elf/newlib/targ-include
-isystem /Users/anders/work/toolchain/gcc/4.9.1/gcc-4.9.1/newlib/libc/include
-B/Users/anders/work/toolchain/gcc/4.9.1/mipsel-sde-elf/mipsel-sde-elf/libgloss/mips
-L/Users/anders/work/toolchain/gcc/4.9.1/mipsel-sde-elf/mipsel-sde-elf/libgloss/libnosys
-L/Users/anders/work/toolchain/gcc/4.9.1/gcc-4.9.1/libgloss/mips
-B/Users/anders/local/mipsel-sde-elf/bin/
-B/Users/anders/local/mipsel-sde-elf/lib/ -isystem
/Users/anders/local/mipsel-sde-elf/include -isystem
/Users/anders/local/mipsel-sde-elf/sys-include    -g -O2 -Os -minterlink-mips16
-mcode-readable=pcrel -mno-gpopt -EB -O2  -g -O2 -Os -minterlink-mips16
-mcode-readable=pcrel -mno-gpopt -DIN_GCC  -DCROSS_DIRECTORY_STRUCTURE  -W
-Wall -Wwrite-strings -Wcast-qual -Wstrict-prototypes -Wmissing-prototypes
-Wold-style-definition  -isystem ./include   -g -DIN_LIBGCC2 -fbuilding-libgcc
-fno-stack-protector -Dinhibit_libc  -I. -I. -I../../.././gcc
-I../../../../gcc-4.9.1/libgcc -I../../../../gcc-4.9.1/libgcc/.
-I../../../../gcc-4.9.1/libgcc/../gcc -I../../../../gcc-4.9.1/libgcc/../include
 -DHAVE_CC_TLS  -o addsf3.o -MT addsf3.o -MD -MP -MF addsf3.dep addsf3 -c
../../../../gcc-4.9.1/libgcc/config/hardfp.c -fvisibility=hidden -DHIDE_EXPORTS
-Wno-missing-prototypes
xgcc: error: addsf3: No such file or directory
make[4]: *** [addsf3.o] Error 1
make[3]: *** [multi-do] Error 1
make[2]: *** [all-multi] Error 2
make[1]: *** [all-target-libgcc] Error 2
make: *** [all] Error 2

The configure command used:
--prefix=/Users/anders/local --target=mipsel-sde-elf --disable-nls
--enable-languages="c,c++" --enable-lto --with-newlib --with-arch-32=mips32r2
--with-arch-64=mips64r2

Building 4.8.3 with the same configuration succeeds without problems, and
building 4.9.1 for the m68k-unknown-elf and sh-elf targets also succeeds
without issues.
Building 4.9.1 for a the mipsel-sde-elf target on a Linux host also works fine.

[Bug libgcc/62097] mipsel-sde-elf build fails on OS X 10.7 host

[Anders.Montonen at iki dot fi]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=62097

--- Comment #1 from Anders Montonen <Anders.Montonen at iki dot fi> ---
Created attachment 33295
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=33295&action=edit
Config log

[Bug libgcc/62097] mipsel-sde-elf build fails on OS X 10.7 host

[Anders.Montonen at iki dot fi]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=62097

--- Comment #2 from Anders Montonen <Anders.Montonen at iki dot fi> ---
Behaviour is the same with GCC 4.9.2 and GCC 5.0.1 RC-20150418, on OS X 10.10
with Xcode 6.3 tools (clang --version: Apple LLVM version 6.1.0
(clang-602.0.49) (based on LLVM 3.6.0svn))

[Bug libgcc/62097] mipsel-sde-elf build fails on OS X 10.7 host

[Anders.Montonen at iki dot fi]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=62097

Anders Montonen <Anders.Montonen at iki dot fi> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Version|4.9.1                       |5.2.0

--- Comment #4 from Anders Montonen <Anders.Montonen at iki dot fi> ---
The problem lies in libgcc/config/t-hardfp, the sed script that is supposed to
generate the -DOP parameter requires GNU sed:

hardfp_defines_for = \
  $(shell echo $1 | \
    sed
's/\(.*\)\($(hardfp_mode_regexp)\)\($(hardfp_suffix_regexp)\|\)$$/-DFUNC=__&
-DOP_\1\3 -DTYPE=\2/')

[Bug libgcc/62097] t-hardfp requires GNU sed

[joel at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=62097

--- Comment #5 from Joel Sherrill <joel at gcc dot gnu.org> ---
MIPS tools for RTEMS are unbuildable on *BSD due to this GNU sed dependency.

[Bug libgcc/62097] t-hardfp requires GNU sed

[joel at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=62097

--- Comment #6 from Joel Sherrill <joel at gcc dot gnu.org> ---
git blame pins us down a bit on the change:

4b366ca9 (rsandifo 2014-02-07 07:46:34 +0000 74) hardfp_defines_for = \
4b366ca9 (rsandifo 2014-02-07 07:46:34 +0000 75)   $(shell echo $1 | \
4b366ca9 (rsandifo 2014-02-07 07:46:34 +0000 76)     sed
's/\(.*\)\($(hardfp_mode_regexp)\)\($(hardfp_suffix_regexp)\|\)$$/-DFUNC=__&
-DOP_\1\3 -DTYPE=\2/')
4b366ca9 (rsandifo 2014-02-07 07:46:34 +0000 77) 

commit 4b366ca99e51b2e80ceb5f2b7ece5189f43bd661
Author: rsandifo <rsandifo@138bc75d-0d04-0410-961f-82ee72b054a4>
Date:   Fri Feb 7 07:46:34 2014 +0000

    libgcc/
        * configure.ac (libgcc_cv_mips_hard_float): New.
        * configure: Regenerate.
        * config.host (mips*-*-*): Use t-hardfp-sfdf rather than
        t-softfp-sfdf for hard-float targets.
        * config/mips/t-mips (LIB2_SIDITI_CONV_FUNCS): Reinstate.
        (softfp_float_modes, softfp_int_modes, softfp_extensions)
        (softfp_truncations, softfp_exclude_libgcc2): New.
        * config/t-hardfp: New file.
        * config/t-hardfp-sfdf: Likewise.
        * config/hardfp.c: Likewise.


    git-svn-id: svn+ssh://gcc.gnu.org/svn/gcc/trunk@207593
138bc75d-0d04-0410-961f-82ee72b054a4

[Bug libgcc/62097] t-hardfp requires GNU sed

[rsandifo at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=62097

rsandifo at gcc dot gnu.org <rsandifo at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Assignee|rsandifo at gcc dot gnu.org        |unassigned at gcc dot gnu.org

--- Comment #7 from rsandifo at gcc dot gnu.org <rsandifo at gcc dot gnu.org> ---
AIUI usual practice is not to assign stuff to people
without asking them first, otherwise it gives the
impression that someone is actively working on it
when they're not.

[Bug libgcc/62097] t-hardfp requires GNU sed

[joel at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=62097

--- Comment #8 from Joel Sherrill <joel at gcc dot gnu.org> ---
(In reply to rsandifo@gcc.gnu.org from comment #7)
> AIUI usual practice is not to assign stuff to people
> without asking them first, otherwise it gives the
> impression that someone is actively working on it
> when they're not.

Sorry Richard. I didn't mean to imply that you were 
working on it. Just that it looked very likely you
wrote the code causing the problem.

[Bug libgcc/62097] t-hardfp requires GNU sed

[rsandifo at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=62097

rsandifo at gcc dot gnu.org <rsandifo at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2015-09-03
                 CC|                            |rsandifo at gcc dot gnu.org
     Ever confirmed|0                           |1

--- Comment #9 from rsandifo at gcc dot gnu.org <rsandifo at gcc dot gnu.org> ---
Oops, didn't mean to remove myself from the bug completely.
I'd assumed bugzilla would keep me on cc:.