make check failures on debian/MIPS QEMU (Mali)

Guhl Peter peter.guhl at
Fri Aug 9 11:06:05 CEST 2013


Well... I got ask by "make check" to report bugs... uhm... so I do ;-)
If it's not helpfull feel free to ignore.

What I really try to do is to compile a toolchain wich -msoft-fpu
because I want to create MIPS binaries able to be run on a standard
OpenWRT which happens to have FPU Kernel emulation turned off. Sounds
easy, seems to be tricky. The currently used howto is (which I do not follow exactly, but

The output of make check (excerpt, but I left the last few hard
float/soft float Warnings intact... maybe that's the problem after all):
/usr/bin/ld: Warning: .libs/t-bdiv uses hard float, t-bdiv.o uses soft
/usr/bin/ld: Warning: .libs/t-bdiv uses hard float,
../../tests/.libs/libtests.a(misc.o) uses soft float
/usr/bin/ld: Warning: .libs/t-bdiv uses hard float,
../../tests/.libs/libtests.a(memory.o) uses soft float
make[4]: Leaving directory
make  check-TESTS
make[4]: Entering directory
PASS: t-asmtype
PASS: t-aors_1
PASS: t-divrem_1
PASS: t-mod_1
PASS: t-fat
mpn_get_d wrong on 2^n
   bit      0
   exp      0
   want_bit 0
   sign     0
   n        =0xGNU MP: Cannot allocate memory (size=2147483666)
/bin/bash: line 5: 30464 Aborted                 ${dir}$tst
FAIL: t-get_d
PASS: t-instrument
PASS: t-iord_u
PASS: t-mp_bases
PASS: t-perfsqr
PASS: t-scan
PASS: t-toom22
PASS: t-toom32
PASS: t-toom33
PASS: t-toom42
PASS: t-toom43
PASS: t-toom44
PASS: t-toom52
PASS: t-toom53
PASS: t-toom62
PASS: t-toom63
PASS: t-toom6h
PASS: t-toom8h
PASS: t-mul
PASS: t-mullo
PASS: t-mulmod_bnm1
PASS: t-sqrmod_bnm1
*** glibc detected ***
/root/soft-fpu-toolchain/build/gmp/tests/mpn/.libs/lt-t-hgcd: realloc():
invalid next size: 0x00784bf8 ***
/bin/bash: line 5: 30931 Aborted                 ${dir}$tst
FAIL: t-hgcd
PASS: t-matrix22
PASS: t-invert
PASS: t-div
PASS: t-bdiv
2 of 32 tests failed
Please report to gmp-bugs at, see
----- end of make check output -----------------------------

The contents of my $CFLAGS:
#echo $CFLAGS

# gcc -v
Using built-in specs.
Target: mips-linux-gnu
Configured with: ../src/configure -v --with-pkgversion='Debian 4.6.3-14'
--enable-languages=c,c++,fortran,objc,obj-c++ --prefix=/usr
--program-suffix=-4.6 --enable-shared --enable-linker-build-id
--with-system-zlib --libexecdir=/usr/lib --without-included-gettext
--enable-threads=posix --with-gxx-include-dir=/usr/include/c++/4.6
--libdir=/usr/lib --enable-nls --with-sysroot=/ --enable-clocale=gnu
--enable-libstdcxx-debug --enable-libstdcxx-time=yes
--enable-gnu-unique-object --enable-plugin --with-mips-plt
--with-arch-32=mips2 --with-tune-32=mips32 --enable-targets=all
--with-arch-64=mips3 --with-tune-64=mips64 --enable-checking=release
--build=mips-linux-gnu --host=mips-linux-gnu --target=mips-linux-gnu
Thread model: posix
gcc version 4.6.3 (Debian 4.6.3-14)

(This is the gcc which came with Debian/MIPS)

# uname -a
Linux EW-MIPS 3.2.0-4-4kc-malta #1 Debian 3.2.46-1 mips GNU/Linux

The configure command (telling you which GMP version I got too):
../../sources/gmp-5.0.4/configure --prefix=$PROJECT_DIR/output/

Well... probably this is not the latest GMP... right?

Anyway; hth.

Debian Version: 7.1

Last but not least: the QEMU start command (Windows 7): 
.\qemu-014-original\qemu-system-mips.exe -m 256 -kernel
vmlinux-3.2.0-4-4kc-malta -append "console=ttyS0 root=/dev/sda1" -hda
hda.img -localtime -net nic -net user -redir tcp:80::80 -redir

I'll continue trying to geht something working... we'll see.


Peter Guhl

More information about the gmp-bugs mailing list