Cygwin "execvp: Argument list too long" errors

[Jeremiah Willcock]

In the test "bronf1 - graph - gursoy_atun_layout_test / gcc-cygwin_4.3.2",
there is an error "execvp: Argument list too long".  I understand that
those tests have not run recently, but is there a way around that length
problem?  I believe Cygwin has a response file feature that allows longer
command lines.  Is this problem important enough to suggest using that
feature?  Also, will that platform start to have tests run again?

-- Jeremiah Willcock
_______________________________________________
Boost-Testing mailing list
Boost-Testing@...
http://lists.boost.org/mailman/listinfo.cgi/boost-testing

[frederic.bron]

> In the test "bronf1 - graph - gursoy_atun_layout_test / gcc-cygwin_4.3.2",
> there is an error "execvp: Argument list too long".  I understand that
> those tests have not run recently, but is there a way around that length
> problem?  I believe Cygwin has a response file feature that allows longer
> command lines.  Is this problem important enough to suggest using that
> feature?  Also, will that platform start to have tests run again?

Hi!
I was running the tests on cygwin but I encountered some difficulties:

1) suddenly, the test procedure stopped working (since somewhere between the 2nd of April 16h53 +0000 and the 3rd of April 8h45 +0000); I get the message below when building process_jam_log (I suspect
something related to the definition of BOOST_FILESYSTEM_NARROW_ONLY),

2) I was running the tests on a USB FAT32 external disk and each time I wanted to perform a full test (i.e. not incremental), it took so much time to download the trunk that I posted on the list and
was answered that it was not supposed to run on that type of disk...

so I decided to stop running the tests. If it is possible to find what's wrong in the test procedure with cygwin, I can restart testing but still with my USB FAT32 disk.

F. Bron



    "/usr/bin/g++"  -ftemplate-depth-128 -O3 -finline-functions -Wno-inline -Wall -DBOOST_POSIX_API -DBOOST_POSIX_PATH  -DBOOST_ALL_NO_LIB=1 -DBOOST_FILESYSTEM_STATIC_LINK=1
-DBOOST_SYSTEM_STATIC_LINK=1 -DNDEBUG -D_CRT_SECURE_NO_WARNINGS  -I"/cygdrive/i/boost_regression_cygwin/boost" -c -o "bin\gcc-cygwin_3.4.4\release\process_jam_log.o"
"i:\boost_regression_cygwin\tools_regression\src\process_jam_log.cpp"

In file included from /cygdrive/i/boost_regression_cygwin/boost/boost/filesystem/operations.hpp:17,
                 from i:\boost_regression_cygwin\tools_regression\src\process_jam_log.cpp:12:
/cygdrive/i/boost_regression_cygwin/boost/boost/filesystem/path.hpp:65: error: `wstring' is not a member of `std'
/cygdrive/i/boost_regression_cygwin/boost/boost/filesystem/path.hpp:65: error: `wstring' is not a member of `std'
/cygdrive/i/boost_regression_cygwin/boost/boost/filesystem/path.hpp:65: error: template argument 1 is invalid
/cygdrive/i/boost_regression_cygwin/boost/boost/filesystem/path.hpp:65: error: ISO C++ forbids declaration of `wpath' with no type
/cygdrive/i/boost_regression_cygwin/boost/boost/filesystem/path.hpp:69: error: ISO C++ forbids declaration of `wstring' with no type
/cygdrive/i/boost_regression_cygwin/boost/boost/filesystem/path.hpp:69: error: expected `;' before "internal_string_type"
/cygdrive/i/boost_regression_cygwin/boost/boost/filesystem/path.hpp:79: error: expected `,' or `...' before '&' token
/cygdrive/i/boost_regression_cygwin/boost/boost/filesystem/path.hpp:79: error: ISO C++ forbids declaration of `internal_string_type' with no type
/cygdrive/i/boost_regression_cygwin/boost/boost/filesystem/path.hpp:80: error: `internal_string_type' does not name a type
...
...skipped <pbin/gcc-cygwin_3.4.4/release>process_jam_log for lack of <pbin/gcc-cygwin_3.4.4/release>process_jam_log.o...
...failed updating 4 targets...
...skipped 3 targets...
...updated 20 targets...
# Searching for "process_jam_log.exe" in "/cygdrive/i/boost_regression_cygwin/tools_regression/build"...
Traceback (most recent call last):
  File "run.py", line 60, in <module>
    runner(root)
  File "/cygdrive/i/boost_regression_cygwin/tools_regression_src/regression.py", line 193, in __init__
    self.main()
  File "/cygdrive/i/boost_regression_cygwin/tools_regression_src/regression.py", line 551, in main
    getattr(self,action_m)()
  File "/cygdrive/i/boost_regression_cygwin/tools_regression_src/regression.py", line 508, in command_regression
    self.command_setup()
  File "/cygdrive/i/boost_regression_cygwin/tools_regression_src/regression.py", line 300, in command_setup
    self.build_if_needed(self.process_jam_log,self.pjl_toolset)
  File "/cygdrive/i/boost_regression_cygwin/tools_regression_src/regression.py", line 650, in build_if_needed
    tool[ 'build_path' ] = self.tool_path( tool )
  File "/cygdrive/i/boost_regression_cygwin/tools_regression_src/regression.py", line 675, in tool_path
    , '\n'.join( [ name_or_spec[ 'path' ], build_dir ] )
Exception: Cannot find "process_jam_log.exe" in any of the following locations:
/cygdrive/i/boost_regression_cygwin/process_jam_log.exe
/cygdrive/i/boost_regression_cygwin/tools_regression/build

Avis :
Ce message et toute pièce jointe sont la propriété d'Alcan et sont destinés seulement aux personnes ou à l'entité à qui le message est adressé. Si vous avez reçu ce message par erreur, veuillez le
détruire et en aviser l'expéditeur par courriel. Si vous n'êtes pas le destinataire du message, vous n'êtes pas autorisé à utiliser, à copier ou à divulguer le contenu du message ou ses pièces jointes
en tout ou en partie.

Notice:
This message and any attachments are the property of Alcan and are intended solely for the named recipients or entity to whom this message is addressed. If you have received this message in error
please inform the sender via e-mail and destroy the message. If you are not the intended recipient you are not allowed to use, copy or disclose the contents or attachments in whole or in part.

_______________________________________________
Boost-Testing mailing list
Boost-Testing@...
http://lists.boost.org/mailman/listinfo.cgi/boost-testing

[Juergen Hunold-2]

Hi Frederic !

On Friday 17 April 2009, frederic.bron@... wrote:
> Hi!
> I was running the tests on cygwin but I encountered some
> difficulties:
>
> 1) suddenly, the test procedure stopped working (since somewhere
> between the 2nd of April 16h53 +0000 and the 3rd of April 8h45
> +0000); I get the message below when building process_jam_log (I
> suspect something related to the definition of
> BOOST_FILESYSTEM_NARROW_ONLY),

mmh. Looks like the "lack of wide charakter support in cygwin" issue.
There are several threads about that one on the devel mailing list.

> 2) I was running the tests on a USB FAT32 external disk and each time
> I wanted to perform a full test (i.e. not incremental), it took so
> much time to download the trunk that I posted on the list and was
> answered that it was not supposed to run on that type of disk...

Why don't you use "--force-update" option for run.py
>From run.py --help:

 --force-update   do an SVN update (if applicable) instead of a clean
                        checkout, even when performing a full run

That would prevent a new checkout and simply update the existing working
copy. Using a fresh checkout each time is simply wasting bandwith...

> so I decided to stop running the tests. If it is possible to find
> what's wrong in the test procedure with cygwin, I can restart testing
> but still with my USB FAT32 disk.

Well, there have been some changes to Boost.Config recently.
>From boost/filesystem/config.hpp I get the impression that
BOOST_NO_STD_WSTRING is/was not defined for cygwin...

And using "update" instead  of  a full checkout should bring some
speedup. Maybe upgrading to svn 1.6.1 might speed things up, too.

Yours,

Jürgen
--
* Dipl.-Math. Jürgen Hunold       ! Ingenieurgesellschaft für
* voice: ++49 511 262926 57       ! Verkehrs- und Eisenbahnwesen mbH  
* fax  : ++49 511 262926 99       ! Lister Straße 15
* juergen.hunold@...        ! www.ivembh.de
*
* Geschäftsführer:                ! Sitz des Unternehmens: Hannover
* Prof. Dr.-Ing. Thomas Siefer    ! Amtsgericht Hannover, HRB 56965
* PD Dr.-Ing. Alfons Radtke       !


_______________________________________________
Boost-Testing mailing list
Boost-Testing@...
http://lists.boost.org/mailman/listinfo.cgi/boost-testing

[Frédéric Bron]

> mmh. Looks like the "lack of wide charakter support in cygwin" issue.
> There are several threads about that one on the devel mailing list.

Yes but it worked for a long time! so something must have changed
somewhere (not in boost itself: version 3 of filesystem has not been
merge to the trunk yet) but maybe in boost build or regression or ...)

> Why don't you use "--force-update" option for run.py

This is not a very big issue as I heard that I can use this option and
also download tarballs. I do not know why people said it was a
surprise that it worked before on FAT32 disk.

> Well, there have been some changes to Boost.Config recently.
> >From boost/filesystem/config.hpp I get the impression that
> BOOST_NO_STD_WSTRING is/was not defined for cygwin...

OK, I will try to set this macro in my user-config.jam. I hope this
will revive the regression process.

Frédéric Bron
_______________________________________________
Boost-Testing mailing list
Boost-Testing@...
http://lists.boost.org/mailman/listinfo.cgi/boost-testing

[Frédéric Bron]

OK, I have rerun the tests for cygwin 1.5.
For some reason, bjam stalled at some point and I had to kill it to
finish the tests...
I also had reports of multiple definition of BOOST_NO_STD_WSTRING.

Nevertheless, the tests for graph have run but
gursoy_atun_layout_test/gcc-cygwin_4.3.2 still does not pass at link
time but I do not see "argument list too long now".

Frédéric
_______________________________________________
Boost-Testing mailing list
Boost-Testing@...
http://lists.boost.org/mailman/listinfo.cgi/boost-testing