Bug#726009: Insufficient RAM on build-machines (was Bug#726009: yade: FTBFS on i386 (and others))

Bernd Zeimetz bernd at bzed.de
Sun Oct 13 17:45:18 UTC 2013


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

On 10/12/2013 07:43 PM, Anton Gladky wrote:
> Thanks all, guys, for recommendations and explanations. I will try to split
> some large cpp-files as Dmitrijs mentioned (there are a couple of them,
> which are failing reliably on weak archs), remove some optimization options
> as Mathieu advised and we will see, whether it helps.

As I have a similar problem with gmic, here is my workaround in debian/rules,
maybe its helpful for you:

BUILDD_MEM := $(shell awk '/^MemTotal:/ {print $$2}' /proc/meminfo)
BUILDD_MEM_OK := $(strip $(shell test $(BUILDD_MEM) -gt 1800000 && echo yes))

BUILDD_SLOW_ARCH := $(strip $(shell dpkg-architecture -qDEB_BUILD_ARCH_CPU | \
                           grep -q -E '^(arm|mips|sh4|m68k).*' && echo yes))

ifneq (yes,$(BUILDD_MEM_OK))
  GMIC_CFLAGS = -O0 -g
else
  ifneq (yes,$(BUILDD_SLOW_ARCH))
    GMIC_CFLAGS = -O3 -g
  else
    GMIC_CFLAGS = -O1 -g -fno-tree-pre
  endif
endif

Good luck,

Bernd


> 
> I would also try to build Yade with clang on failing archs, if upper 
> mentioned actions will not help to solve the problem. Though 
> openmp-parallelisation will be disabled in this case, ant it is not very 
> convenient.
> 
> Thanks again,
> 
> Anton
> 
> On 10/12/2013 02:34 PM, Mathieu Malaterre wrote:
>> Anton,
>> 
>> On Fri, Oct 11, 2013 at 9:00 PM, Anton Gladky <gladk at debian.org> wrote: 
>> [...]
>>>> /usr/bin/c++   -Dyade_EXPORTS -g -O2 -fstack-protector
>>>> --param=ssp-buffer-size=4 -Wformat -Werror=format-security
>>>> -D_FORTIFY_SOURCE=2  -DYADE_PTR_CAST=static_pointer_cast
>>>> -DYADE_CAST=static_cast -fPIC -DYADE_VTK -DYADE_OPENMP -fopenmp
>>>> -DYADE_GTS  -DQGLVIEWER_FOUND -DYADE_OPENGL -frounding-math
>>>> -DYADE_CGAL -DFLOW_ENGINE -frounding-math -DLINSOLV -DFLOW_ENGINE
>>>> -DYADE_GL2PS -O3 -DNDEBUG -fPIC -
>> 
>> You are using cmake in Release mode, which is bad IMHO. So having -O2 and
>> -O3 (in that order) compiles code in -O3, which is way too much AFAIK.
>> 
>> Ref: http://gcc.gnu.org/onlinedocs/gcc/Optimize-Options.html ... If you
>> use multiple -O options, with or without level numbers, the last such
>> option is the one that is effective. ...
>> 
>> See also here why CMAKE_BUILD_TYPE should be required to be set to "" on
>> debian: http://bugs.debian.org/701231#12
>> 
>> 2cts
> 
> 


- -- 
 Bernd Zeimetz                            Debian GNU/Linux Developer
 http://bzed.de                                http://www.debian.org
 GPG Fingerprint: ECA1 E3F2 8E11 2432 D485  DD95 EB36 171A 6FF9 435F
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.14 (GNU/Linux)
Comment: Using GnuPG with Icedove - http://www.enigmail.net/
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=47Fr
-----END PGP SIGNATURE-----



More information about the debian-science-maintainers mailing list