gm2
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Gm2] gm2 (20180320) on gcc-4.1.2 build on Solaris 10/sparc successful


From: john o goyo
Subject: [Gm2] gm2 (20180320) on gcc-4.1.2 build on Solaris 10/sparc successful
Date: Sat, 07 Apr 2018 17:58:33 -0400
User-agent: Mozilla/5.0 (Macintosh; U; PPC Mac OS X 10.5; en-GB; rv:1.9.2.23) Gecko/20110920 Thunderbird/3.1.15

Successfully built gm2 using the tarball gcc-4.1.2+gm2-20180320.tar.gz with gcc-5.4.0; some modifications needed (below). Thank you, Gaius, for assembling this tarball!

The following configuration script was used.

GCCVER=5.4.0
DATE=20180320

CC=/home/build/gcc/$GCCVER/bin/gcc \
CXX=/home/build/gcc/$GCCVER/bin/g++ \
CFLAGS="-D_XPG6 --std=gnu99" \
LDFLAGS="-lnsl -lsocket -lpthread -lrt" \
SHELL=/bin/bash \
CONFIG_SHELL=/bin/bash \
../../src/gcc-4.1.2+gm2-$DATE/gm2/gcc-versionno/configure \
--prefix=$HOME/gm2/4.1.2/$DATE \
--without-gnu-ld --with-ld=/usr/sfw/bin/gld \
--with-gnu-as --with-as=/usr/sfw/bin/gas  \
--enable-languages=c,c++,gm2 --disable-bootstrap --enable-checking \
--disable-libitm

------------------------------------
Gmake check-gm2 gave me the following.

                === gm2 Summary ===

# of expected passes            9236
# of unexpected failures        126
# of unresolved testcases       24

(Full log to be sent to Gaius off list.)

--------------------------------------
Modifications:

1. The script above put all languages in the Makefile and required manual pruning.

2. Removed the "inline" definitions of floor_log2() and exact_log2() from gm2/gcc-versionno/gcc/toplev.h into the toplev.c file. (Gcc kept reporting an error: redefinition.)

3. Moved the code for libc_name_p() from gm2/gcc-versionno/gcc/cp/cfns.h to gm2/gcc-versionno/gcc/cp/except.c as the code was not being put into the object file.

4. Gcc claimed that PATH_MAX was undefined in gm2/gcc-versionno/gcc/config/sparc/gmon-sol2.c. I added <limits.h> to no avail, so I manually defined PATH_MAX to 1024. Also changed asm() to __asm__() as gcc complained about the former.

Sincerely,
john







reply via email to

[Prev in Thread] Current Thread [Next in Thread]