[Bug target/44531] New: [SuperH] Multilib configuration does not work as expected

[Bug target/44531] New: [SuperH] Multilib configuration does not work as expected

[ljsebald at gmail dot com]

When building a clean GCC 4.5.0 targeting sh-elf, the multilib configuration of
the compiler does not appear to work at all. For instance, building GCC 4.5.0
with the following set of configure flags:

../gcc-4.5.0/configure --target=sh-elf --prefix=/usr/local/dc2/sh-elf
--with-newlib --enable-languages=c --disable-libssp --disable-tls
--without-headers --disable-nls --with-multilib-list=m4-single-only,m4-nofpu,m4
--with-endian=little --with-cpu=m4-single-only

Results in a compiler that can only compile with -m4-single-only. -m4 and
-m4-nofpu do not work at all, even though they were explicitly specified in the
--with-multilib-list.

Attempting to specify -m4 results in the following error (similar for
-m4-nofpu):
cc1: error: command line option '-m4' is not supported by this configuration

Configuring GCC 4.4.4 with the same set of configure flags results in a
compiler that does support all of -m4, -m4-single-only, and -m4-nofpu (and
defaults to -m4-single-only, as expected).


-- 
           Summary: [SuperH] Multilib configuration does not work as
                    expected
           Product: gcc
           Version: 4.5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: ljsebald at gmail dot com
 GCC build triplet: x86_64-apple-darwin10.3.1
  GCC host triplet: x86_64-apple-darwin10.3.1
GCC target triplet: sh-unknown-elf


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44531

[Bug target/44531] [SuperH] Multilib configuration does not work as expected

[kkojima at gcc dot gnu dot org]

------- Comment #1 from kkojima at gcc dot gnu dot org  2010-06-19 09:50 -------
I can't reproduce it on my i686-pc-linux-gnu host with
a similar configuration

/home/kkojima/kaz/xsh-elf-combined-450/combined/configure
--target=sh-unknown-elf --disable-libmudflap --disable-libssp --disable-libgomp
--with-gnu-as --with-gnu-ld --enable-languages=c --disable-shared --with-newlib
--disable-nls --prefix=/home/kkojima/kaz/xsh-elf-combined-450/install
--with-headers=yes --disable-gdbtk --with-mpfr=/opt2/i686-pc-linux-gnu
--with-gmp=/opt2/i686-pc-linux-gnu
--with-multilib-list=m4-single-only,m4-nofpu,m4 --with-endian=little
--with-cpu=m4-single-only

for vanilla gcc-4.5.0 under a usual combined tree setting.
It can be something configuration issue including a host
specific problem.  My gcc/tm.h starts like as:

#ifndef GCC_TM_H
#define GCC_TM_H
#define TARGET_CPU_DEFAULT (SELECT_SH4_SINGLE_ONLY)
#ifndef SH_MULTILIB_CPU_DEFAULT
# define SH_MULTILIB_CPU_DEFAULT "m4-single-only"
#endif
#ifndef SUPPORT_SH4_SINGLE_ONLY
# define SUPPORT_SH4_SINGLE_ONLY 1
#endif
#ifndef SUPPORT_SH4_SINGLE_ONLY
# define SUPPORT_SH4_SINGLE_ONLY 1
#endif
#ifndef SUPPORT_SH4_NOFPU
# define SUPPORT_SH4_NOFPU 1
#endif
#ifndef SUPPORT_SH4
# define SUPPORT_SH4 1
#endif

Could you look at how your gcc/tm.h is generated?


-- 

kkojima at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |kkojima at gcc dot gnu dot
                   |                            |org
             Status|UNCONFIRMED                 |WAITING


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44531

[Bug target/44531] [SuperH] Multilib configuration does not work as expected

[ljsebald at gmail dot com]

------- Comment #2 from ljsebald at gmail dot com  2010-06-19 15:58 -------
Judging by what I see from my gcc/tm.h file, I'm guessing it might well be a
host-specific problem (I'm on x86_64-apple-darwin10.3.1). Here's the start of
my gcc/tm.h file:

#ifndef GCC_TM_H
#define GCC_TM_H
#define TARGET_CPU_DEFAULT (SELECT_SH4_SINGLE_ONLY)
#ifndef SH_MULTILIB_CPU_DEFAULT
# define SH_MULTILIB_CPU_DEFAULT "m4-single-only"
#endif
#ifndef SUPPORT_SH4_SINGLE_ONLY
# define SUPPORT_SH4_SINGLE_ONLY 1
#endif

So, from the looks of it, it seems to be a Mac OS X/Darwin host specific issue.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44531

[Bug target/44531] [4.5/4.6 Regression] [SH] Multilib configuration does not work as expected on darwin

[kkojima at gcc dot gnu dot org]

------- Comment #3 from kkojima at gcc dot gnu dot org  2010-06-20 10:56 -------
I can reproduce it on an old ppc darwin.  It looks that this
darwin's sed doesn't support the 'i' modifier for the 's' command
and drops all multilibs list specified.
Does the patch below work for you?

--- gcc/config.gcc~     2010-04-07 19:34:00.000000000 +0900
+++ gcc/config.gcc      2010-06-20 18:08:46.000000000 +0900
@@ -2315,7 +2315,7 @@ sh-*-symbianelf* | sh[12346l]*-*-symbian
        target_cpu_default=SELECT_`echo ${sh_cpu_default}|tr
abcdefghijklmnopqrstuvwxyz- ABCDEFGHIJKLMNOPQRSTUVWXYZ_`
        tm_defines=${tm_defines}' SH_MULTILIB_CPU_DEFAULT=\"'`echo
$sh_cpu_default|sed s/sh/m/`'\"'
        tm_defines="$tm_defines SUPPORT_`echo $sh_cpu_default | sed 's/^m/sh/'
| tr abcdefghijklmnopqrstuvwxyz- ABCDEFGHIJKLMNOPQRSTUVWXYZ_`=1"
-       sh_multilibs=`echo $sh_multilibs | sed -e 's/,/ /g' -e 's/^sh/m/i' -e
's/ sh/ m/gi' | tr ABCDEFGHIJKLMNOPQRSTUVWXYZ_ abcdefghijklmnopqrstuvwxyz-`
+       sh_multilibs=`echo $sh_multilibs | sed -e 's/,/ /g' -e 's/^[Ss][Hh]/m/'
-e 's/ [Ss][Hh]/ m/g' | tr ABCDEFGHIJKLMNOPQRSTUVWXYZ_
abcdefghijklmnopqrstuvwxyz-`
        for sh_multilib in ${sh_multilibs}; do
                case ${sh_multilib} in
                m1 | m2 | m2e | m3 | m3e | \


-- 

kkojima at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|WAITING                     |NEW
     Ever Confirmed|0                           |1
      Known to fail|                            |4.5.0 4.6.0
      Known to work|                            |4.4.4
   Last reconfirmed|0000-00-00 00:00:00         |2010-06-20 10:56:53
               date|                            |
            Summary|[SuperH] Multilib           |[4.5/4.6 Regression] [SH]
                   |configuration does not work |Multilib configuration does
                   |as expected                 |not work as expected on
                   |                            |darwin


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44531

[Bug target/44531] [4.5/4.6 Regression] [SH] Multilib configuration does not work as expected on darwin

[ljsebald at gmail dot com]

------- Comment #4 from ljsebald at gmail dot com  2010-06-20 14:01 -------
The patch does indeed work, and everything builds as expected once patched.
Thanks.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44531

[Bug target/44531] [4.5/4.6 Regression] [SH] Multilib configuration does not work as expected on darwin

[kkojima at gcc dot gnu dot org]

------- Comment #5 from kkojima at gcc dot gnu dot org  2010-06-20 22:23 -------
Thanks for the confirmation.  I'll apply it when the usual
bootstrap&regression tests are done on 4.5/4.6.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44531

[Bug target/44531] [4.5/4.6 Regression] [SH] Multilib configuration does not work as expected on darwin

[rguenth at gcc dot gnu dot org]

-- 

rguenth at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.5.1


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44531

[Bug target/44531] [4.5/4.6 Regression] [SH] Multilib configuration does not work as expected on darwin

[rguenth at gcc dot gnu dot org]

-- 

rguenth at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |patch
           Priority|P3                          |P4


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44531

[Bug target/44531] [4.5/4.6 Regression] [SH] Multilib configuration does not work as expected on darwin

[kkojima at gcc dot gnu dot org]

------- Comment #6 from kkojima at gcc dot gnu dot org  2010-07-04 22:07 -------
Subject: Bug 44531

Author: kkojima
Date: Sun Jul  4 22:07:29 2010
New Revision: 161807

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=161807
Log:
        PR target/44531
        * config.gcc (sh*-*-*): Use regular expressions instead of
        the 'i' modifier for sed substitutions.


Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/config.gcc


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44531

[Bug target/44531] [4.5/4.6 Regression] [SH] Multilib configuration does not work as expected on darwin

[kkojima at gcc dot gnu dot org]

------- Comment #7 from kkojima at gcc dot gnu dot org  2010-07-05 22:34 -------
Subject: Bug 44531

Author: kkojima
Date: Mon Jul  5 22:33:58 2010
New Revision: 161857

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=161857
Log:
        Backport from mainline:
        PR target/44531
        * config.gcc (sh*-*-*): Use regular expressions instead of
        the 'i' modifier for sed substitutions.


Modified:
    branches/gcc-4_5-branch/gcc/ChangeLog
    branches/gcc-4_5-branch/gcc/config.gcc


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44531

[Bug target/44531] [4.5/4.6 Regression] [SH] Multilib configuration does not work as expected on darwin

[kkojima at gcc dot gnu dot org]

------- Comment #8 from kkojima at gcc dot gnu dot org  2010-07-05 22:39 -------
Fixed.


-- 

kkojima at gcc dot gnu dot org changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44531