Skip to content

Successes and Failures of Building LibTomMath

Jump to bottom Edit New page
Christoph Zurnieden edited this page Jan 5, 2021 · 2 revisions

Successes and Failures of Building LibTomMath

LibTomMath builds with a lot of compilers on a lot of architectures but not in every combination possible and not always out of the box. Test-base is LibTomMath version 1.1.0 with some floating point operations on top (get/set float, double, and long double) to evaluate the usability of the FPU, if available, on the tested platform. The architectures were emulated with qemu version 3.1.0. unless stated otherwise.

Architecture

  • amd64 (hardware)

    • Debian Versions 8 (jessie) to 10 (buster)
      Out of the box.
    • NetBSD 8.0
      Static library builds out of the box but building the shared library needs GNU make and libtool.

  • ARM926EJ-S rev 5 (v5l)

    • Debian 8.11 (jessie)
      Out of the box. This is a system with `double == long double`. 
      
$ cat dbl_print.c
#include 
#include 
#include 

      int main (void) {
printf("%d == %d  %d == %d\n",
LDBL_MANT_DIG, DBL_MANT_DIG,
LDBL_MAX_EXP,  DBL_MAX_EXP);
exit(EXIT_SUCCESS);
}
$ ./dbl_print
53 == 53  1024 == 1024
$ uname -a
Linux debian 3.16.0-6-versatile #1 Debian 3.16.57-2 (2018-07-14) armv5tejl GNU/Linux
$ lsb_release -a
No LSB modules are available.
Distributor ID: Debian
Description:    Debian GNU/Linux 8.11 (jessie)
Release:        8.11
Codename:       jessie
$ cat /proc/cpuinfo
processor       : 0
model name      : ARM926EJ-S rev 5 (v5l)
BogoMIPS        : 500.22
Features        : swp half thumb fastmult vfp edsp java
CPU implementer : 0x41
CPU architecture: 5TEJ
CPU variant     : 0x0
CPU part        : 0x926
CPU revision    : 5
      

      Hardware        : ARM-Versatile PB
Revision        : 0000
Serial          : 0000000000000000

  • Sparc64

    • Debian sid (buster) 
      
cpu		: TI UltraSparc IIi (Sabre)
fpu		: UltraSparc IIi integrated FPU
pmu		: ultra12
prom		: OBP 3.10.24 1999/01/01 01:01
type		: sun4u
ncpus probed	: 1
ncpus active	: 1
D$ parity tl1	: 0
I$ parity tl1	: 0
Cpu0ClkTck	: 0000000005f5e100
cpucaps		: flush,stbar,swap,muldiv,v9,mul32,div32,v8plus,vis
MMU Type	: Spitfire
MMU PGSZs	: 8K,64K,512K,4MB
$ uname -a
Linux debiansparc64 4.15.0-2-sparc64 #1 Debian 4.15.11-1 (2018-03-20) sparc64 GNU/Linux
$ gcc -dM -E -x c /dev/null | grep BYTE_ORDER
#define __BYTE_ORDER__ __ORDER_BIG_ENDIAN__
$ gcc -dM -E -x c /dev/null | grep LDBL
#define __LDBL_MAX__ 1.18973149535723176508575932662800702e+4932L
#define __LDBL_MAX_EXP__ 16384
#define __LDBL_HAS_INFINITY__ 1
#define __LDBL_MIN__ 3.36210314311209350626267781732175260e-4932L
#define __LDBL_HAS_QUIET_NAN__ 1
#define __LDBL_HAS_DENORM__ 1
#define __LDBL_DECIMAL_DIG__ 36
#define __LDBL_EPSILON__ 1.92592994438723585305597794258492732e-34L
#define __LDBL_MANT_DIG__ 113
#define __LDBL_MIN_EXP__ (-16381)
#define __LDBL_MAX_10_EXP__ 4932
#define __LDBL_DENORM_MIN__ 6.47517511943802511092443895822764655e-4966L
#define __LDBL_MIN_10_EXP__ (-4931)
#define __LDBL_DIG__
    • OpenBSD 6.4
      The static library builds but the shared library needs both `gmake` and the GNU `libtool` installed. The wrong `make` will result in a lot of errors of the kind: 
      
*** Parse error in /home/user/GITHUB/libtommath: Need an operator in 'LIBNAME' )
*** Parse error: Need an operator in 'endif' (makefile.shared:8)
*** Parse error: Need an operator in 'CROSS_COMPILE' (makefile_include.mk:16)
*** Parse error: Need an operator in 'endif' (makefile_include.mk:18)
*** Parse error: Missing dependency operator (makefile_include.mk:22)
*** Parse error: Missing dependency operator (makefile_include.mk:23)

      The wrong libtool gives this error at linking time:

      
cc -I./ -Wall -Wsign-compare -Wextra -Wshadow -Wsystem-headers -Wdeclaration-afo...
cc -I./ -Wall -Wsign-compare -Wextra -Wshadow -Wsystem-headers -Wdeclaration-afo...
cc -I./ -Wall -Wsign-compare -Wextra -Wshadow -Wsystem-headers -Wdeclaration-afo...
libtool --mode=link --tag=CC cc  bn_error.lo bn_fast_mp_invmod.lo bn_fast_mp_mo 
libtool: link: cc bn_error.lo bn_fast_mp_invmod.lo bn_fast_mp_montgomery_reduce0
bn_error.lo: file not recognized: File format not recognized
cc: error: linker command failed with exit code 1 (use -v to see invocation)
Error while executing cc bn_error.lo bn_fast_mp_invmod.lo bn_fast_mp_montgomery0
gmake: *** [makefile.shared:64: libtommath.la] Error 1

      As a little extra gift the GNU libtool and the build-in libtool are both named libtool. If those two packets have been installed by sudo pkg_add gmake libtool the GNU libtool will be placed in /usr/local/bin/ and the native version in /usr/bin/. To be sure ask the programs (actual version numbers may differ, of course):

      
$ /usr/local/bin/libtool --version                              
libtool (GNU libtool) 2.4.2
Written by Gordon Matzigkeit , 1996

Copyright (C) 2011 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.

$ libtool --version
libtool (not (GNU libtool)) 1.5.26

      The shared library should build now with the command LIBTOOL="/usr/local/bin/libtool" gmake -f makefile.shared

  • M68K Atari Falcon (Aranym 1.0.2 emulating a M68040 with MMU and FPU)

  • MIPS (32-bit big-endian)

    • Debian 9.8.0 (stretch) 
      
$ cat /proc/cpuinfo 
system type		: MIPS Malta
machine			: mti,malta
processor		: 0
cpu model		: MIPS 24Kc V0.0  FPU V0.0
BogoMIPS		: 688.12
wait instruction	: yes
microsecond timers	: yes
tlb_entries		: 16
extra interrupt vector	: yes
hardware watchpoint	: yes, count: 1, address/irw mask: [0x0ff8]
isa			: mips1 mips2 mips32r1 mips32r2
ASEs implemented	: mips16
shadow register sets	: 1
kscratch registers	: 0
package			: 0
core			: 0
VCED exceptions		: not available
VCEI exceptions		: not available
$  uname -a
Linux debianmips32 4.9.0-8-4kc-malta #1 Debian 4.9.144-3.1 (2019-02-19) mips GNU/Linux
$ gcc -dM -E -x c /dev/null | grep BYTE_ORDER
#define __BYTE_ORDER__ __ORDER_BIG_ENDIAN__
$ gcc -dM -E -x c /dev/null | grep LDBL
#define __LDBL_MAX__ 1.7976931348623157e+308L
#define __LDBL_MAX_EXP__ 1024
#define __LDBL_HAS_INFINITY__ 1
#define __LDBL_MIN__ 2.2250738585072014e-308L
#define __LDBL_HAS_QUIET_NAN__ 1
#define __LDBL_HAS_DENORM__ 1
#define __LDBL_EPSILON__ 2.2204460492503131e-16L
#define __LDBL_MANT_DIG__ 53
#define __LDBL_MIN_EXP__ (-1021)
#define __LDBL_MAX_10_EXP__ 308
#define __LDBL_DENORM_MIN__ 4.9406564584124654e-324L
#define __LDBL_MIN_10_EXP__ (-307)
#define __LDBL_DIG__ 15

  • S390x

    • Debian 9.8.0 (stretch) 
      
user@debians390x:~/GITHUB/libtommath$ cat /proc/cpuinfo
vendor_id       : IBM/S390
# processors    : 1
bogomips per cpu: 13370.00
max thread id   : 0
features	: esan3 zarch stfle msa ldisp eimm etf3eh highgprs 
processor 0: version = 00,  identification = 000000,  machine = 2827
user@debians390x:~/GITHUB/libtommath$ uname -a
Linux debians390x 4.9.0-8-s390x #1 SMP Debian 4.9.144-3.1 (2019-02-19) s390x GNU/Linux
user@debians390x:~/GITHUB/libtommath$ gcc -dM -E -x c /dev/null | grep BYTE_ORDER
#define __BYTE_ORDER__ __ORDER_BIG_ENDIAN__
user@debians390x:~/GITHUB/libtommath$ gcc -dM -E -x c /dev/null | grep LDBL
#define __LDBL_MAX__ 1.18973149535723176508575932662800702e+4932L
#define __LDBL_MAX_EXP__ 16384
#define __LDBL_HAS_INFINITY__ 1
#define __LDBL_MIN__ 3.36210314311209350626267781732175260e-4932L
#define __LDBL_HAS_QUIET_NAN__ 1
#define __LDBL_HAS_DENORM__ 1
#define __LDBL_EPSILON__ 1.92592994438723585305597794258492732e-34L
#define __LDBL_MANT_DIG__ 113
#define __LDBL_MIN_EXP__ (-16381)
#define __LDBL_MAX_10_EXP__ 4932
#define __LDBL_DENORM_MIN__ 6.47517511943802511092443895822764655e-4966L
#define __LDBL_MIN_10_EXP__ (-4931)
#define __LDBL_DIG__ 33

  • x86 (hardware)

    • Debian Versions 8 (jessie) to 10 (buster)
      Out of the box.

  • x86_64 (hardware)

    • Debian Versions 8 (jessie) to 10 (buster)
      Out of the box

Compilers

  • GCC
  • clang
    Versions greater than 7 with optimizations greater than -O0 need a newer version of Valgrind. Version 3.14.0 works, newer ones are recommended. Version 3.13.0, which does not work is in Ubuntu "bionic", 3.15.0 in Ubuntu "focal" and 3.16.1 is in Ubuntu "groovy".
  • Intel compiler ICC version used was 19.0.2.187 20190117
    The makefiles assume a GCC compatible compiler which ICC is. Mostly. Just do a `CC=/path/to/bin/icc make` to build the static and CC=/path/to/bin/icc make -f makefile.shared` to build the shared library.

    It does not support two of GCC's diagnostic options used to build LibTomMath and complains with

    
icc: command line warning #10148: option '-Wbad-function-cast' not supported
icc: command line warning #10148: option '-Wcast-align' not supported

    To suppress this warnings use the option -diag-disable=10006. Although it is not really the best advise to switch warnings off it does no harm here, the GCC diagnostics just get ignored.

Add a custom footer
Add a custom sidebar
Clone this wiki locally