On Sun, Feb 20, 2011 at 5:08 AM, Robert Bielik robert.bie...@xponaut.se wrote:
Robert Bielik skrev 2011-02-20 10:54:
John Drescher skrev 2011-02-19 17:33:
#Use solution folders.
set_property(GLOBAL PROPERTY USE_FOLDERS ON)
Doesn't seem to work. Where is this documented ? Seems very hinky
In CTestTargets.cmake (included from CTest.cmake), there are lines that read:
SET_PROPERTY(TARGET ... PROPERTY FOLDER CTestDashboardTargets)
That's the thing that assigns that folder value to those targets. It's
explicitly separated from the CMake predefined targets folder because
they are two
On Mon, Feb 21, 2011 at 2:14 AM, Schmid Alexander a.sch...@de.ccv.eu wrote:
On Friday 18 February 2011, Schmid Alexander wrote:
Hi,
of course, I´d like to have you think further about it, so here you go...
;-)
This is the toolchain file I use.
The specialty about is that I want to use the
Use:
SET_PROPERTY(GLOBAL PROPERTY USE_FOLDERS ${BUILD_USE_SOLUTION_FOLDERS})
On Mon, Feb 21, 2011 at 11:20 AM, Robert Bielik
robert.bie...@xponaut.se wrote:
David Cole skrev 2011-02-20 17:55:
Those properties were first introduced in v2.8.3, that's why it didn't
do anything for 2.8.2
...@hdfgroup.org wrote:
One more detail, this seems to be a problem only on non windows (non-make) -
windows VS machines ran the tests in correct declaration order.
Allen
Bug reported:
http://public.kitware.com/Bug/view.php?id=11877
On Fri, Feb 18, 2011 at 10:30 AM, David Cole david.c
Allen,
I'm not sure what, if anything, changed in the ctest code, but I have
this observation to offer. Try:
ctest -D NightlyMemoryCheck
When I use NightlyMemoryCheck all of the dashboard steps execute,
and the submit at the end includes the dynamic analysis xml file in
the submission.
When
It looks to me from:
git gui blame -- Source/cmCTest.cxx
that this is the behavior that has always been -- ever since the -D
MemCheck -D MemoryCheck stuff has been added to ctest several years
ago.
Are you sure that nothing else changed (besides CMake from 2.8.2 to
2.8.4 in this
Please see the FAQ on the CMake Wiki:
http://www.cmake.org/Wiki/CMake_FAQ#Why_does_CMake_use_full_paths.2C_or_can_I_copy_my_build_tree.3F
No, you can't move your build tree to another computer, and no you
can't eliminate the full path references.
On Tue, Feb 22, 2011 at 1:59 PM, gekso
On Wed, Feb 23, 2011 at 11:28 AM, Alexander Neundorf
a.neundorf-w...@gmx.net wrote:
On Wednesday 23 February 2011, Schmid Alexander wrote:
This suffix should be set by the platform file.
Modules/CMakeSystemSpecificInformation.cmake includes
CMakeGenericSystem.cmake, which sets it to .
Can you send me (offlist) copies of the DynamicAnalysis.xml files from
each case? Do you still have them?
If not, no worried. I'll try to reproduce this on our CMake dashboard
by submitting a 2nd valgrind dashboard: same script, just drive it
with 2.8.4. It may take me a few days to get around to
What version of CDash is your server running?
I'm trying to view it in my web browser to find out for myself, but
when I try ( http://nei.hdfgroup.uiuc.edu/cdash ), I get a Problem
loading page message from Firefox.
I suspect we're compressing the test results, but CDash is not
de-compressing
On Thu, Feb 24, 2011 at 10:17 AM, Arnaud GELAS
arnaud_ge...@hms.harvard.edu wrote:
On 02/23/2011 07:35 PM, Clinton Stimpson wrote:
On Wednesday, February 23, 2011 02:38:38 pm Arnaud GELAS wrote:
On 02/23/2011 04:35 PM, Clinton Stimpson wrote:
On Wednesday, February 23, 2011 02:27:29 pm
On Thu, Feb 24, 2011 at 11:41 AM, Arnaud GELAS
arnaud_ge...@hms.harvard.edu wrote:
On 02/24/2011 11:35 AM, David Cole wrote:
On Thu, Feb 24, 2011 at 10:17 AM, Arnaud GELAS
arnaud_ge...@hms.harvard.edu wrote:
On 02/23/2011 07:35 PM, Clinton Stimpson wrote:
On Wednesday, February 23, 2011
Is C:/path/to/make a batch file?
What are its contents?
On Thu, Feb 24, 2011 at 1:23 PM, thehighhat thehighhat+...@gmail.comwrote:
using nsis 2.46. cmake 2.8.3 to config and build a 32 bit console app
on a 64bit windows 7 host using 32bit intel compilers and msvs 2010
express: all ok.
Are you using the cygwin cmake or the Windows cmake in C:\Program Files...
If you're using the Windows cmake, it's not going to be able to deal with
cygwin symlinks at all.
On Thu, Feb 24, 2011 at 1:49 PM, thehighhat thehighhat+...@gmail.comwrote:
On Thu, Feb 24, 2011 at 11:34 AM, David Cole
Do you have the same problem if you use a real binary for the 'make' instead
of this symlink?
On Thu, Feb 24, 2011 at 2:14 PM, thehighhat thehighhat+...@gmail.comwrote:
On Thu, Feb 24, 2011 at 12:05 PM, David Cole david.c...@kitware.com
wrote:
Are you using the cygwin cmake or the Windows
Just:
set( QT_QMAKE_EXECUTABLE ${QT_BINARY_DIR}/qmake.exe )
should be sufficient here. Everything else that FindQt does switches off of
the qmake setting... If you change the qmake executable, all the other
variables will adapt correctly.
On Fri, Feb 25, 2011 at 10:21 AM, Thomas Wolf
On Mon, Feb 28, 2011 at 5:25 AM, Michael Wild them...@gmail.com wrote:
On 02/28/2011 11:17 AM, Jochen Wilhelmy wrote:
Hi!
after looking into the source of cmake I have seen that
MACOSX_BUNDLEs are only supported for executables and
I have not enough insight yet to port it to module
The mandatory function arguments are the ones that you name. Optional
ones go after the named args and are referenced via ARGV0, ARGV1,
ARGV2, ... or the ARGV and ARGN variables.
See function documentation for more info:
http://www.cmake.org/cmake/help/cmake-2-8-docs.html#command:function
On Wed, Mar 2, 2011 at 3:31 AM, Rolf Eike Beer e...@sf-mail.de wrote:
I am defining separate output paths for different configurations:
CMAKE_RUNTIME_OUTPUT_DIRECTORY_DEBUG = bin/win32/dbg
CMAKE_RUN TIME_OUTPUT_DIRECTORY_RELEASE = bin/win32/opt
In CMake 2.8.2 and later, you can:
set(CTEST_CONFIGURATION_TYPE Release) # or Debug
in the ctest -S script.
That will influence both the ctest_build and ctest_test commands.
HTH,
David
On Thu, Mar 3, 2011 at 2:09 PM, Macumber, Daniel
daniel.macum...@nrel.govwrote:
Is it possible to set
add_definitions( -DSTORE=XYZ )
should do what you want.
If you have a cmake_minimum_required of at least version 2.6, then the
cmake_policy call is unnecessary -- it is simply implied by the minimum
required version of CMake for 2.6 and later.
What version of CMake are you using?
What OS and
One other thing. You may try using the NMake Makefiles generator if you
must use the VS10 compiler: the dependencies definitely work via the
makefiles. I just tried it out...
On Fri, Mar 4, 2011 at 10:15 AM, David Cole david.c...@kitware.com wrote:
Using:
UPDATE_COMMAND
is perfectly
Are you calling ctest_start(Experimental) or ctest_start(Nightly) before
callng ctest_submit...?
On Mon, Mar 7, 2011 at 4:32 PM, Zou, Di (Cont, ARL/CISD) di@us.army.mil
wrote:
I have been looking at this webpage:
http://www.kitware.com/products/html/CDashSubprojects.html. I have created
Is it in a public repo? Can we just try to repro the complicated case
here?
On Tue, Mar 8, 2011 at 11:01 AM, kent williams nkwmailingli...@gmail.comwrote:
Thanks for getting back to me on this. After my initial post I tried
to come up with some 'cut-down' examples, and the simple cases work.
How did you configure it? With a previous successful build of CMake or with
./bootstrap or ./configure?
On Wed, Mar 9, 2011 at 12:38 PM, Olivier Cessenat cesse...@free.fr wrote:
Hello,
I downloaded cmake-2.8.4 and also cmake-2.8.3 Unix sources and tried
to compile it for Mac OS-X 10.4.
I
Rather than building it from source, you could simply download the pre-built
binaries for the Mac. They work on 10.4.
On Wed, Mar 9, 2011 at 12:56 PM, David Cole david.c...@kitware.com wrote:
How did you configure it? With a previous successful build of CMake or with
./bootstrap
This is already at least partially fixed in CMake 2.8.4...
The chunk you reference:
IF(CMAKE_CL_64)
SET(CMAKE_MSVC_ARCH amd64)
ELSE(CMAKE_CL_64)
SET(CMAKE_MSVC_ARCH x86)
ENDIF(CMAKE_CL_64)
Now looks like this:
IF(CMAKE_CL_64)
IF(MSVC_VERSION GREATER 1599)
# VS 10 and later:
I only know about problems with this w.r.t. Visual Studio 10, as reported
just the other day here:
http://www.cmake.org/pipermail/cmake/2011-March/043205.html
http://public.kitware.com/Bug/view.php?id=11927
Are you using VS 10? Or something else?
You can always write a custom
Good point, Dave P...
Kent, are you changing GIT_TAG argument values when this happens? If so,
this explains it entirely, and this is the known bug that Dave P. has
pointed to.
If not, we need to figure out why the download step is re-running when it
shouldn't.
Thanks,
David C.
On Wed, Mar
with:
${msvc_install_path}/../../VC/redist
Should make it into 2.8.5. (Barring unforeseen madness, of course.)
Thanks,
David
On Wed, Mar 9, 2011 at 3:42 PM, J Decker d3c...@gmail.com wrote:
On Wed, Mar 9, 2011 at 11:56 AM, David Cole david.c...@kitware.com
wrote:
This is already at least partially fixed
For reference, here's the commit that I just pushed to 'next' :
http://cmake.org/gitweb?p=cmake.git;a=commitdiff;h=971a735ba2b4d9e4ac846722fdf280dbd0feb0d4
On Wed, Mar 9, 2011 at 6:01 PM, David Cole david.c...@kitware.com wrote:
I'm going to commit this as a fix in CMake 'next' -- but I'm
Well if the build directory is *in* the source directory, then make
package_source *should* include it.
This is not a problem, but the expected behavior.
If you don't want the build tree in the source tree, then don't put it
there.
On Wed, Mar 9, 2011 at 9:11 PM, Pierre Abbat
Tyler's right. Use install(SCRIPT instead.
And use execute_process instead of exec_program. It's better.
From the help at
http://cmake.org/cmake/help/cmake-2-8-docs.html#command:execute_process
:
The execute_process command is a newer more powerful version of
exec_program, but the old command
You also need to add an install rule for your executable with
INSTALL(TARGETS ...
make install should work first, then make package will work too.
HTH,
David
On Thu, Jul 31, 2008 at 12:00 AM, Michael Masters [EMAIL PROTECTED]wrote:
I'm trying to get cpack working with our build and I'm a
Rename your directory to get rid of the ++ in the name...
It's confusing some regular expression matching rules in the CMake scripts.
On Fri, Aug 1, 2008 at 11:43 AM, Jan Dinger [EMAIL PROTECTED]wrote:
Hello folks,
I've wrote a test program for testing cmake in Qt4. Iam using cmake-2.6
What does cmake --version report? There was no CMake 2.2.6.
On Fri, Aug 8, 2008 at 7:02 AM, Marianne B. Wiese [EMAIL PROTECTED] wrote:
Hello
I am trying to build FLTK in order to build the Orfeo Toolbox later.
Apparently CMake cannot find the headers for openGL and/or GLU
I use:
DESTDIR will not work with Windows style paths including a C:. In fact, it
may not work at all on Windows. In general, it's a bad idea to force files
into a C:/work (or any other non-standard) directory. Windows installers
typically default to putting their programs and all of their files into the
In CMake/Source/CTest/cmCTestBuildHandler.cxx, the line from the build
output has to match one of the regular expressions in cmCTestErrorMatches
*and* not match any of the regular expressions in cmCTestErrorExceptions.
I do not see at first glance why those lines would be reported as an error,
EXECUTE_PROCESS with cmake -E copy
(or copy_directory)
On Tue, Aug 12, 2008 at 1:39 PM, [EMAIL PROTECTED] wrote:
Hi,
I need to duplicate (Python) files from a source directory into a
destination (build) directory, then build a CRAMFS volume image from the
destination dir.
How can I
I use:
SET_TESTS_PROPERTIES(${name} PROPERTIES
FAIL_REGULAR_EXPRESSION (ERROR|Warning): In
)
which will cause the test to fail if the string ERROR: In or Warning: In
occurs in the test output.
${name} should be the name of the test as passed to ADD_TEST, and the
argument after
Make that cmake -P (not cmake -S) -S is the run this script arg to
ctest...
On Fri, Aug 15, 2008 at 11:45 AM, Bill Hoffman [EMAIL PROTECTED]wrote:
Gregory C. Sharp wrote:
Bill Hoffman wrote:
Gregory C. Sharp wrote:
I found this little tidbit in the FAQ:
[E]nvironment variables
If you are using a new style command-based ctest -S script, you can:
SET(CTEST_BUILD_FLAGS -j8)
before using the CTEST_BUILD command.
On Mon, Aug 18, 2008 at 4:02 PM, Moreland, Kenneth [EMAIL PROTECTED]wrote:
Is there a way to add arguments to the invocation of make? In particular,
I
want
It looks like it must be a bug of some sort if it works with one name, but
it does not work with another name...
However, you do have overlapping regular expressions. All of your files that
match the 2nd regular expression also match the 1st regular expression.
A better 1st regular expression
think of an easier way to do it. Maybe this is a good question for
the NSIS community? If there's a way to do it via NSIS script, we could add
it into CPack's default NSIS script, or at least tell you how to hook it
in... Let us know if the NSIS community gives you any other info...
HTH,
David
Was there ever a bug filed on this? If so, what's the issue number in the
bug tracker?
Thx,
David Cole
On Mon, Jul 28, 2008 at 7:15 AM, Doug Gregor [EMAIL PROTECTED] wrote:
On Fri, Jul 25, 2008 at 7:56 PM, Gregory Peele ARA/CFD [EMAIL PROTECTED]
wrote:
I'm really pleased with the component
Probably not. It's probably just that there is no real code in those
particular files on your platform due to ifdef conditionals...
On Fri, Aug 22, 2008 at 2:12 PM, Darren Weber
[EMAIL PROTECTED]wrote:
Should I be concerned about this note of no symbols in libcmcurl?
Scanning dependencies
There is not a pre-built Mac installer available, but you can download the
zip file of the source tree from here and build it:
http://public.kitware.com/KWStyle/download.htm
Or you can get the source from CVS and build it. It builds with CMake and
it's a very quick build.
Let us know if you have
the details. The tutorial in Chpt 12 is
fairly light. Is there anything else I should take a look at?
Thanks!
On Fri, Aug 22, 2008 at 12:38 PM, David Cole [EMAIL PROTECTED]wrote:
There is not a pre-built Mac installer available, but you can download the
zip file of the source tree from here
On Fri, Aug 22, 2008 at 6:54 PM, Darren Weber
[EMAIL PROTECTED]wrote:
When building CMake, there are options to enable CPACK_BINARY_* and
CPACK_SOURCE_*; are these options switching on functionality within CMake or
are they creating packages of the current CMake build? I'm guessing they do
Make sure the headers are part of the library, too, like
this:add_library(${EBML}
${SOURCES} ${HEADERS})
HTH,
David
On Sat, Aug 23, 2008 at 11:10 AM, Boris Dušek [EMAIL PROTECTED] wrote:
Hi,
I tried to use CMake to generate a framework with one library file and
bunch of public header files,
On Thu, Aug 28, 2008 at 5:12 AM, Vandenbroucke Sander
[EMAIL PROTECTED] wrote:
-Original Message-
From: Bill Hoffman [mailto:[EMAIL PROTECTED]
Sent: woensdag 27 augustus 2008 14:37
To: Vandenbroucke Sander
Cc: cmake@cmake.org; Brad King
Subject: Re: [CMake] re-config on make
Now we're getting somewhere...
Which 2.4 are you on?
To force caching of -D variables with older cmake versions, specify a type
for each one (the :STRING in the modified version below). Also, I would
put the -D values before naming the source directory:
cmake -G MSYS Makefiles
prog.1 is a file, not a target.
Use INSTALL(FILES instead.
TARGETS is for CMake ADD_LIBRARY and ADD_EXECUTABLE targets.
HTH,
David
On Thu, Aug 28, 2008 at 5:16 PM, Matthew Gates [EMAIL PROTECTED]wrote:
Hello,
I have a sub-directory in my project, doc, which contains pod markup of a
And yet, we will probably not add ADD_HEADER... (I don't think. Other CMake
devs can chime in here if they think I'm wrong.)
The problem is more than just sources and headers. What about resource
files? Should .bmp and .ico resources show up in the IDE? Yet, they're not
compiled. Are they headers?
in it, and it
should work.
Let us know if it works, and we will commit that change to CVS HEAD in VTK.
Thanks,
David Cole
On Fri, Aug 29, 2008 at 3:51 PM, Darren Weber
[EMAIL PROTECTED]wrote:
I had some choices on how to install MySQL and PostgreSQL on OS X (Leopard
Server, in my case
, Aug 29, 2008 at 12:57 PM, David Cole [EMAIL PROTECTED]wrote:
As David Thompson rightfully pointed out, we should not be using
LINK_LIBRARIES in VTK's CMakeLists.txt files.
Change the LINK_LIBRARIES statement in VTK/IO/CMakeLists.txt to:
SET(KIT_LIBS ${KIT_LIBS} ${POSTGRES_LIBRARIES
Including the libpq.dylib at the end. It's a full path to a file.
On Fri, Aug 29, 2008 at 4:45 PM, David Cole [EMAIL PROTECTED] wrote:
Set POSTGRES_LIBRARIES to the full path to the pq dylib...
On Fri, Aug 29, 2008 at 4:39 PM, Darren Weber
[EMAIL PROTECTED] wrote:
For my install
You'll want to write a new C++ subclass of cmCPackGenerator, naming it
cmCPackWIXGenerator to follow the existing naming convention of the other
generators.
You'll need to add code the CMake/Source/CPack/cmCPackGeneratorFactory.cxx
to expose it to the world.
And you'll need to add code to
What CMake generator are you using?
On Wed, Sep 3, 2008 at 4:10 AM, Anders Moe [EMAIL PROTECTED] wrote:
Hi all
I'm trying to compile a 32 bit version of VTK using the 32 bit MSVC 2005
command shell. Note that I'm running XP64.
This works fine for Qt, but cmake gives an error during the
${PLINK_EXECUTABLE} contains spaces...
Maybe you need escaped quotes:
SET(CMAKE_CXX_LINK_EXECUTABLE \${PLINK_EXECUTABLE}\
\ around the executable variable reference...?
On Wed, Sep 3, 2008 at 6:04 PM, Alexander Neundorf
[EMAIL PROTECTED]wrote:
On Monday 25 August 2008, Phil Smith wrote:
Maybe save and restore the CMAKE_SHARED_LINKER_FLAGS around your reset of it
in the sub-CMakeLists.txt...?
On Wed, Sep 3, 2008 at 10:41 AM, Martin Apel [EMAIL PROTECTED] wrote:
Hi,
I have defined the following in a top-level CMakeLists.txt:
SET (CMAKE_SHARED_LINKER_FLAGS
VTK_USE_RPATH should be sufficient...
On Fri, Sep 5, 2008 at 5:51 PM, Darren Weber
[EMAIL PROTECTED]wrote:
On Wed, Sep 3, 2008 at 3:28 PM, Mike Jackson
[EMAIL PROTECTED] wrote:
Cross posting to vtkusers list as that may be more appropriate:
Can you rebuild VTK with these two settings:
(not only for shared libraries, but in general).
Regards,
Martin
David Cole wrote:
Maybe save and restore the CMAKE_SHARED_LINKER_FLAGS around your reset
of it in the sub-CMakeLists.txt...?
On Wed, Sep 3, 2008 at 10:41 AM, Martin Apel [EMAIL PROTECTED]
mailto:[EMAIL PROTECTED] wrote
There is no occurrence of Develeper in the CMake code base right now
(either in CVS HEAD or the CMake-2-6 branch...)
It must be coming from someplace else on your system.
(Check the files ~/.bash_profile and/or ~/.MacOSX/environment.plist ...
?)
See
are in
different CMakeLists files?
On Mon, Sep 8, 2008 at 3:02 AM, Martin Apel [EMAIL PROTECTED] wrote:
I am using ADD_SUBDIRECTORY. I did'nt know, that anything like SUBDIRS
exists and haven't found anything in the documentation about it.
Martin
David Cole wrote:
Are you using SUBDIRS
Is there an issue on this in the bug tracker?If not, could you open one and
attach the patch there?
Thx,
Dave
On Mon, Sep 8, 2008 at 4:42 AM, Mathieu Malaterre
[EMAIL PROTECTED] wrote:
Patch works for me on debian stable. Please consider.
Credit goes to 0xd34df00d.
Thanks
-Mathieu
Add this line after the list LENGTH call:math(EXPR list_max_index
${list_count}-1)
and then use list_max_index instead of list_count in the foreach loop.
The documentation (cmake --help-command foreach) is quite clear that the one
arg form of the range command loops from 0 to total including
There's a cdash mailing list, too:
http://public.kitware.com/cgi-bin/mailman/listinfo/cdash
What do you mean by connect -- there is a link from a project's CDash page
to its bug tracker. Did you need more connection than that?
On Fri, Sep 12, 2008 at 9:42 AM, [EMAIL PROTECTED] wrote:
Hi
With a new-style ctest script like this, you should set CTEST_BUILD_COMMAND
and CTEST_BUILD_FLAGS to specify the make executable and arguments to it if
you need values different than those ctest uses by default. If you set
those, there is no need to set the MAKECOMMAND in the cache.
HTH,
David
If you run EXECUTE_PROCESS to do a checkout before ADD_SUBDIRECTORY it
should work fine. Do it conditionally though, or else you'll be doing
checkouts constantly. You probably just want to do an update if the
directory already exists...
HTH,
David
On Thu, Oct 2, 2008 at 10:16 AM, kent williams
to reproduce your problem if I have misread something...
Thanks,
David Cole
On Thu, Sep 18, 2008 at 3:21 PM, Bill Hoffman [EMAIL PROTECTED]wrote:
Jesper Eskilson wrote:
Hi all,
When CMake (2.6) discovers that a CMakeLists.txt file has changed, and
that one or more Visual Studio projects
From your same command prompt that you ran cmake in (at
N:\dsci\repo\WMI\trunk\wxWMI\src\nmake) what is the output of: cl
where cl
echo %PATH%
?
On Thu, Oct 9, 2008 at 11:51 AM, Luke Kucalaba [EMAIL PROTECTED] wrote:
Our software team is having problems upgrading to CMake 2.6.2 from
No, it doesn't make sense. Can you construct a minimal example that
demonstrates this?
If not, there must be something in your project files that differentiates
between in-source and out-of-source builds...?
Are you changing compilers based on in-source/out-of-source?
The in-source/out-of-source
at the same time that
you upgraded to CMake 2.6.2.??
Let me know if you still think this is a CMake issue after you verify that
you really have different behavior between 2.6.0 and 2.6.2 on the same
machine.
Thanks,
David Cole
Kitware, Inc.
On Mon, Oct 13, 2008 at 6:05 AM, Petri Hodju [EMAIL
will work on the fix
:-)
Sorry for the pilot error,
David
On Tue, Oct 14, 2008 at 4:24 PM, David Cole [EMAIL PROTECTED] wrote:
OK so when I reproduce this issue, I see that I do not have directory
creation permissions in /opt on my Linux machine. So... the error message
is not really
issues. (That way,
when I do a make pack on your project, I get the correct cpack being used
rather than whichever one is available in my environment...)
Sorry for the hassle,
David Cole
On Wed, Oct 15, 2008 at 12:46 AM, Petri Hodju [EMAIL PROTECTED] wrote:
Great news!
I started to browse through
that this resolves my problem. Thanks again for a quick cure
and for the tip! Could it be that this solves also issues 7000 7410? They
seem to have the same problem with temporary directory creation.
Br,
Petri
--- On *Wed, 10/15/08, David Cole [EMAIL PROTECTED]* wrote:
From: David Cole
That depends. Do you want it to be easy or hard to diagnose the failure...
:-) (I suppose you could concoct a case where leaving the file would make it
easier to diagnose and another case where deleting it would make it
easier...)
CMake didn't write the file, why should it delete it?
Shouldn't
I use ADD_CUSTOM_COMMAND to specify the exact C# compiler command line that
I want...
There is much more to full C# language support than just getting it to
compile/link successfully. (Involving .csproj files vs. .vcproj files,
resources, references, strong naming, blah, blah, blah.)
HTH,
. Otherwise, I will try to put these into the bug tracker and/or fix
them in the CMake code base.
Thanks for the report,
David Cole
___
CMake mailing list
CMake@cmake.org
http://www.cmake.org/mailman/listinfo/cmake
=7828
And a feature request for the new CPACK_PACKAGE_EXECUTABLES behavior:
http://public.kitware.com/Bug/view.php?id=7829
I don't see a way to edit the issues and assign them to David.
Thanks,
Sam
--
*From:* David Cole [mailto:[EMAIL PROTECTED]
*Sent
Another common source of apparently random results is uninitialized
variables. I don't think it is in ctest itself, as this is the first we've
heard of something like this on this mailing list (although anything's
possible...)
Have you run your test code through valgrind to see if it catches any
How are you configuring? With a CMake GUI or a command line tool?
If with a gui, set the CMAKE_INSTALL_PREFIX in the gui.
If with a command line tool, pass -DCMAKE_INSTALL_PREFIX:STRING=/blah on the
command line.
To workaround this entirely, invent your own CMake variable that CMake knows
The test framework can be found at CMake/Tests/Framework. Follow the
CMakeLists.txt in that directory to get started.
Let us know if you think anything is missing...
David
On Tue, Oct 21, 2008 at 7:56 AM, Michael Jackson
[EMAIL PROTECTED] wrote:
OS X Frameworks are pretty well supported. I
From: David Cole [EMAIL PROTECTED]
Subject: Re: [CMake] Mac OS X frameworks
To: Michael Jackson [EMAIL PROTECTED]
Cc: cmake@cmake.org
Message-ID:
[EMAIL PROTECTED]
Content-Type: text/plain; charset=iso-8859-1
The test framework can be found at CMake/Tests/Framework. Follow
bias in its initial functionality...
Do you have a suggestion for a regex that would work with all ldd's? (Or at
least with yours...?)
Thanks,
David Cole
On Fri, Oct 24, 2008 at 8:36 AM, Mike Arthur [EMAIL PROTECTED] wrote:
I'm really real problems with GetPrequisites, anyone got it working
Ah ha. Thanks for that. I will commit this today to CVS HEAD...
David
On Fri, Oct 24, 2008 at 9:19 AM, Mike Arthur [EMAIL PROTECTED] wrote:
in get_prerequisites:
set(gp_regex ^\t([\t ]+)[\t ].*${eol_char}$)
which is missing a '^' before the second \t so should be:
set(gp_regex ^\t([^\t
for get_prerequisites purposes.
I will wait to hear a little more from you before I do any commits to CVS
HEAD... :-)
Thanks for the attention on this,
David
On Fri, Oct 24, 2008 at 9:59 AM, Mike Arthur [EMAIL PROTECTED] wrote:
On Friday 24 October 2008 14:49:49 David Cole wrote:
Ah ha. Thanks for that. I
Just give ADD_SUBDIRECTORY a second argument that tells it where to put the
corresponding binary directory structure.
Something like:
ADD_SUBDIRECTORY(../blah ${CMAKE_CURRENT_BINARY_DIR}/blah)
(Assumes you do not also have an ADD_SUBDIRECTORY(blah) call and a
./blah directory in addition to the
You should be able to configure a file into your binary directory that
references all the option variables of interest and then make a dependency
on that file for your add_custom_command.
That way, the file will change only if one of the options changes in cmake
and the dependency on it should
of all the custom commands that exist in a project and then compare them
upon regeneration. I imagine that would not be an easy task.
James
On Tue, Oct 28, 2008 at 1:52 PM, David Cole [EMAIL PROTECTED]wrote:
You should be able to configure a file into your binary directory
You are not being stupid. It does need to be used at installation time to be
useful.
The intent is to analyze a given built executable to determine the required
set of (possibly 3rd-party, possibly built elsewhere on the build machine)
libraries that it needs to run. Then, given that set, you need
Well it works with your smaller examples, but not your large project..
What (besides small vs. large) is different between the projects?
The message Cannot find any GCov coverage files means that there are no
*.da or *.gcda gcov output files in the entire binary tree. We do a
recursive find
This line:INSTALL(SCRIPT ${CMAKE_MODULE_PATH}/GetPrerequisites.cmake)
doesn't really do anything at make install time.
GetPrerequisites.cmake just defines a bunch of functions. You have to
include it and then *call* some of the functions for it to do anything
useful.
So, you should write your
There is a way to do so, but it's not as easy as you'll want it to be... :-)
You would have to provide your own NSIS.template.in file that does NSIS
stuff directly. You can start out with a copy of the one in the CMake
Modules directory, and then put it in a place in your own source tree, and
In CVS HEAD of ParaView. It was committed just after the 3.4.0 release was
created...
On Thu, Nov 6, 2008 at 10:00 AM, Stephen Collyer
[EMAIL PROTECTED]wrote:
David Cole wrote:
It is actively used in CVS HEAD of ParaView on the Mac to build the
standalone paraview client .app bundle
In a ${CMAKE_BINARY_DIR}/CTestCustom.cmake file, use something like this
expression copied from VTK's code:
SET(CTEST_CUSTOM_COVERAGE_EXCLUDE
${CTEST_CUSTOM_COVERAGE_EXCLUDE}
vtk.*TCLInit.cxx
vtk[^\\.]+(Java|Python|Tcl).cxx
Utilities.kwsys
Utilities.vtksys
)
The variable CTEST_INITIAL_CACHE is ignored in new-style (CTEST_BUILD()
command based) ctest scripts.
Instead, you should use:
FILE(WRITE ${CTEST_BINARY_DIRECTORY}/CMakeCache.txt
MAKECOMMAND:STRING=nmake -i
CMAKE_MAKE_PROGRAM:FILEPATH=nmake
CMAKE_GENERATOR:INTERNAL=NMake Makefiles
You have discovered the magic of Windows. Again.
CMake is a 32-bit process. When Windows launches a 32-bit process on a Win64
machine, it gives the 32-bit process C:\ProgramFiles (x86) as the value of
the env var ProgramFiles. But you are testing it from a 64-bit program,
the default cmd.exe on
901 - 1000 of 3692 matches
Mail list logo