All of lore.kernel.org
 help / color / mirror / Atom feed
From: Philippe Gerum <rpm@xenomai.org>
To: ROSSIER Daniel <Daniel.Rossier@domain.hid>
Cc: xenomai@xenomai.org
Subject: Re: [Xenomai-help] Usage of sim with xenomai-2.0
Date: Tue, 25 Oct 2005 21:31:26 +0200	[thread overview]
Message-ID: <435E880E.8080301@domain.hid> (raw)
In-Reply-To: <FDBBB5CC70676540B3EF7CFE83FD94E004304B@EINT11.einet.ad.eivd.ch>

ROSSIER Daniel wrote:
> Well, I applied the patch and I was hopeful when it compiled the libio stuff. Unfortunately, the following error came out after a while:
> 
> /home/rossier/xenomai-2.0/sim/gcic/gcc-build/gcc/xgcc -B/home/rossier/xenomai-2.0/sim/gcic/gcc-build/gcc/ -B/usr/realtime/libexec/gcic/i686-pc-linux-gnu/bin/ -c -g -O2 -fvtable-thunks -D_GNU_SOURCE -fno-implicit-templates -I../../../gcc-2.95.3/libstdc++ -I../../../gcc-2.95.3/libstdc++/stl -I../libio -I../../../gcc-2.95.3/libstdc++/../libio -nostdinc++   ../../../gcc-2.95.3/libstdc++/stlinst.cc -o pic/stlinst.o
> /home/rossier/xenomai-2.0/sim/gcic/gcc-build/gcc/xgcc -B/home/rossier/xenomai-2.0/sim/gcic/gcc-build/gcc/ -B/usr/realtime/libexec/gcic/i686-pc-linux-gnu/bin/ -c -g -O2 -fvtable-thunks -D_GNU_SOURCE -fno-implicit-templates -I../../../gcc-2.95.3/libstdc++ -I../../../gcc-2.95.3/libstdc++/stl -I../libio -I../../../gcc-2.95.3/libstdc++/../libio -nostdinc++  ../../../gcc-2.95.3/libstdc++/stlinst.cc
> ../../../gcc-2.95.3/libstdc++/stl/stl_alloc.h: In instantiation of `__default_alloc_template<true,0>::_S_node_allocator_lock':
> ../../../gcc-2.95.3/libstdc++/stl/stl_alloc.h:394:   instantiated from `__default_alloc_template<true,0>::_Lock::~_Lock()'
> ../../../gcc-2.95.3/libstdc++/stl/stl_alloc.h:415:   instantiated from `__default_alloc_template<true,0>::allocate(unsigned int)'
> ../../../gcc-2.95.3/libstdc++/stlinst.cc:7:   instantiated from here
> ../../../gcc-2.95.3/libstdc++/stl/stl_alloc.h:581: Internal compiler error.
> ../../../gcc-2.95.3/libstdc++/stl/stl_alloc.h:581: Please submit a full bug report.
> ../../../gcc-2.95.3/libstdc++/stl/stl_alloc.h:581: See <URL:http://www.gnu.org/software/gcc/bugs.html> for instructions.
> make[2]: *** [stlinst.o] Error 1
> make[2]: Leaving directory `/home/rossier/xenomai-2.0/sim/gcic/gcc-build/i686-pc-linux-gnu/libstdc++'
> make[1]: *** [all-target-libstdc++] Error 2
> make[1]: Leaving directory `/home/rossier/xenomai-2.0/sim/gcic/gcc-build'
> make: *** [gcc-build/.gcic-built] Error 2
> root@domain.hid#  
> 
> 
> 'sounds bad :-(  No idea what it can be. Do you have an idea?
>

Nope. Unfortunately, whilst compiling the instrumenter is usually a no brainer 
on common distros, I have no clue wrt what your LFS setup is providing as libc 
headers, which seems to be at the root of the problem, since some of them are 
used to compile GCC against.

You could get rid of the C++ support if you don't plan to instrument C++ apps to 
run with the simulator, i.e.: in sim/gcic/GNUmakefile, just strip out "c++" from 
the LANG variable passed to the configure script, called from the 
gcc-build/.gcic-built rule. Just a work-around, but would perhaps allow you to 
build the simulator properly.

> 
> Daniel
> 
> 
> 
> -----Original Message-----
> From: Philippe Gerum [mailto:rpm@xenomai.org]
> Sent: Tue 10/25/2005 8:29 PM
> To: Philippe Gerum
> Cc: ROSSIER Daniel; xenomai@xenomai.org
> Subject: Re: [Xenomai-help] Usage of sim with xenomai-2.0
>  
> Philippe Gerum wrote:
> 
>>ROSSIER Daniel wrote:
>>
>>
>>>Hi Philippe,
>>>
>>>Here is the info about my env:
>>>
>>>- kernel 2.6.9
>>>- glibc 2.3.5
>>>- gcc 3.4.4
>>>- actually, I built my env. from scratch (no distro) - simply from 
>>>www.linuxfromscratch.org
>>>
>>>But everything else works fine :-) (xenomai works perfectly. I tested 
>>>it with the LinuxTraceToolkit, and it works as well)
>>>
>>>Here is the fragment of log, right before the failure (all the 
>>>compilation steps before are OK):
>>>
>>>
>>>make[4]: Leaving directory 
>>>`/home/rossier/xenomai-2.0/sim/gcic/gcc-build/i686-pc-linux-gnu/libiberty' 
>>>
>>>make[3]: Leaving directory 
>>>`/home/rossier/xenomai-2.0/sim/gcic/gcc-build/i686-pc-linux-gnu/libiberty' 
>>>
>>>make[3]: Entering directory 
>>>`/home/rossier/xenomai-2.0/sim/gcic/gcc-build/i686-pc-linux-gnu/libio'
>>>if [ x"no" = xyes ] && [ ! -d pic ]; then \
>>>  mkdir pic; \
>>>else true; fi
>>>touch stamp-picdir
>>>test x"no" != xyes ||\
>>>  /home/rossier/xenomai-2.0/sim/gcic/gcc-build/gcc/xgcc 
>>>-B/home/rossier/xenomai-2.0/sim/gcic/gcc-build/gcc/ 
>>>-B/usr/realtime/libexec/gcic/i686-pc-linux-gnu/bin/ -c -g -O2 -I. 
>>>-I../../../gcc-2.95.3/libio -D_IO_MTSAFE_IO  
>>>../../../gcc-2.95.3/libio/iogetline.c -o pic/iogetline.o 
>>>/home/rossier/xenomai-2.0/sim/gcic/gcc-build/gcc/xgcc 
>>>-B/home/rossier/xenomai-2.0/sim/gcic/gcc-build/gcc/ 
>>>-B/usr/realtime/libexec/gcic/i686-pc-linux-gnu/bin/ -c -g -O2 -I. 
>>>-I../../../gcc-2.95.3/libio -D_IO_MTSAFE_IO 
>>>../../../gcc-2.95.3/libio/iogetline.c
>>>In file included from ../../../gcc-2.95.3/libio/libio.h:167,
>>>                 from ../../../gcc-2.95.3/libio/iolibio.h:1,
>>>                 from ../../../gcc-2.95.3/libio/libioP.h:47,
>>>                 from ../../../gcc-2.95.3/libio/iogetline.c:26:
>>>/usr/include/bits/stdio-lock.h:24: lowlevellock.h: No such file or 
>>>directory
>>>make[3]: *** [iogetline.o] Error 1
>>>make[3]: Leaving directory 
>>>`/home/rossier/xenomai-2.0/sim/gcic/gcc-build/i686-pc-linux-gnu/libio'
>>>make[2]: *** [all-target-libio] Error 2
>>>make[2]: Leaving directory `/home/rossier/xenomai-2.0/sim/gcic/gcc-build'
>>>make[1]: *** [gcc-build/.gcic-built] Error 2
>>>make[1]: Leaving directory `/home/rossier/xenomai-2.0/sim/gcic'
>>>make: *** [all-recursive] Error 1
>>>root@domain.hid#
>>>
>>
>>It seems to be an incompatibility issue between the NPTL and gcc-2.95.x:
>>http://sources.redhat.com/ml/glibc-bugs/2004-06/msg00144.html
>>
>>The best way to solve this would be to get rid of the MT support for 
>>building the GCC-based instrumenter, since we don't need it, and libio/ 
>>compilation would not break. I'll look at some way to fix this.
>>
> 
> 
> Could you try the following steps and let me know if that works in your environment:
> 
> - unpack the gcc-2.95.3 archive
> - apply the attached patch against the gcc tree
> - pack back the tree
> - configure, build and install Xenomai again to use the patched gcc archive for 
> building the simulator.
> 
> In order to test the resulting stuff:
> 
> - export PATH=/usr/realtime/bin:$PATH
> - export LD_LIBRARY_PATH=/usr/realtime/lib:$LD_LIBRARY_PATH
> - cd xenomai-2.0/skins/vxworks/demos
> - make sim
> 
> You should have the "satch_s" executable available in the current dir at that 
> point; just start the following command then:
> $ xenoscope satch_s
> 
> PS: at Xenomai configuration time, make sure to have selected the VxWorks skin 
> among the APIs to build.
> 


-- 

Philippe.


  reply	other threads:[~2005-10-25 19:31 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2005-10-25 13:12 [Xenomai-help] Usage of sim with xenomai-2.0 ROSSIER Daniel
2005-10-25 14:00 ` Philippe Gerum
2005-10-25 18:12   ` ROSSIER Daniel
2005-10-25 19:23     ` Philippe Gerum
2005-10-25 18:29   ` Philippe Gerum
2005-10-25 19:12     ` ROSSIER Daniel
2005-10-25 19:31       ` Philippe Gerum [this message]
2005-10-25 19:52         ` ROSSIER Daniel
  -- strict thread matches above, loose matches on Subject: below --
2005-10-25 12:52 ROSSIER Daniel
2005-10-25 13:03 ` Philippe Gerum

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=435E880E.8080301@domain.hid \
    --to=rpm@xenomai.org \
    --cc=Daniel.Rossier@domain.hid \
    --cc=xenomai@xenomai.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.