The most interesting thing is to be able to set the relative RPATH.
Is there a way to do this for a normal build, and not the install target? I
never use the install target for my project, but I'd like to be able to set
the relative path to my own copy of libstdc++. All of the RPATH settings in
Mark Moll wrote:
Is there any mechanism in CMake that would make it easier to create
Linux binaries that run on more than one version of Linux / glibc? I
am thinking of something along the lines of the support for OS X
Universal Binaries in CMake or the functionality provided by apgcc in
the auto
Thanks to all the suggestions. There are now a few working CDash
submissions for HDF5 Version 1.8.4. So far we have an OS X 10.5.8
Debug build and a Windows VS 2008 Debug build. There is a third
submission working on the Fortran support under windows.
Is there any mechanism in CMake that would make it easier to create Linux
binaries that run on more than one version of Linux / glibc? I am thinking of
something along the lines of the support for OS X Universal Binaries in CMake
or the functionality provided by apgcc in the autopackage tool
(h
Thanks. I'll make those changes. I had switched to "NMake Makefiles"
but I accidentally had "ADD_DEFINITIONS(-Wall)" in my CMakeLists.txt
file and that simply overwhelmed the console. I think the compile
output was like 25 MB in size. No sense in having that sent to CDash.
It is building/te
Hi Michael,
Sorry for coming into this conversation late, I just caught this thread
on the mailing list. I rigged a CMake build around HDF5 for one of my
projects. It's really tailored to a couple of my systems (Linux/GCC 32
and 64 bit cluster and IBM/VisualAge 64 bit cluster), so I'm not su
We are moving pretty good with the CMakeified HDF5 V 1.8.4. You can
find it on Gitorious under the hdf5/hdf5-v18 project. I have most of
the autoconf stuff transferred over. Currently I am struggling trying
to get an HDF5/Visual Studio based build to actually, configure,
build, test and sub
You might need a
SET(CTEST_PROJECT_NAME foo)
if you have a foo.sln file you want it to build.
For makefiles I set the project name to "all"
I think ctest -VV will show the command used to build the visual studio
project.
And for the CMakeCache.txt, you should need to set the generator, or set
What is the "accepted" setup to run a CTest with CDash submission on
Windows using Visual Studio tools?
I tried writing a "Hdf5Dashboard.cmake" file with some basics in it:
cmake_minimum_required ( VERSION 2.6 )
set (CTEST_SITE "vs2...@bluequartz.net")
set (CTEST_BUILD_NAME "Debug-Visual Studi
Perhaps you need a "*" before .xpm and .gif?
Perhaps the input files do not exist in "${SGICOMMONLIB_SOURCE_DIR}/"?
Perhaps you don't have write permission in
"${CMAKE_RUNTIME_OUTPUT_DIRECTORY}/${DIRNAME}/" ?
For best results, I'd recommend using something like:
> FILE(GLOB DEFAULT_CONFIG_FILES "$
2009/11/23 kent williams :
> CMake Error at common/CMakeLists.txt:83 (CONFIGURE_FILE):
> configure_file Problem configuring file
>
> The issue being the CONFIGURE_FILE command:
>
> PROJECT(SGICOMMONLIB)
> #
> # other stuff removed for brevity.
> #
> SET(DIRNAME pixmaps)
> FILE(GLOB DEFAULT_CONFIG
CMake Error at common/CMakeLists.txt:83 (CONFIGURE_FILE):
configure_file Problem configuring file
The issue being the CONFIGURE_FILE command:
PROJECT(SGICOMMONLIB)
#
# other stuff removed for brevity.
#
SET(DIRNAME pixmaps)
FILE(GLOB DEFAULT_CONFIG_FILES "*.xbm" ".xpm" ".gif" "*.tif" "*.pal" )
On Monday 23 November 2009 10:03:35 Mark Moll wrote:
> Between cmake 2.6.4 and cmake 2.8 the following lines were removed from the
> FIND_LIBRARY(PYTHON_LIBRARY ...) command in FindPythonLibs.cmake:
>
> -PATH_SUFFIXES
> - python${_CURRENT_VERSION}/config
>
> I don’t understand why this
Between cmake 2.6.4 and cmake 2.8 the following lines were removed from the
FIND_LIBRARY(PYTHON_LIBRARY ...) command in FindPythonLibs.cmake:
-PATH_SUFFIXES
- python${_CURRENT_VERSION}/config
I don’t understand why this was done, because this is where python libraries
are found with th
if I have the following code:
EXECUTE_PROCESS(COMMAND ${CMAKE_COMMAND}
-E copy_if_different ${TEMP_CONFIG_FILE} ${HDF5_BINARY_DIR}/
H5pubconf.h)
How exactly is "Different" determined? By date modified, by content,
by some sort of hash?
_
Hi all,
The FindOpenMP has some broken logic. For the Intel compiler, the
fopenmp is accepted because ICC is based off of the GCC source code.
However, this flag just causes a warning. The order of checking flags
should have the Intel flag (-openmp) listed before the GCC flag (-fopenmp).
I
Ahh. did not see that the CDash provided a CTestConfig.cmake file.
I'll try to work on that. I have old ParaView dashboard scripts that
can get me started.
_
Mike Jackson mike.jack...@bluequartz.net
On Nov 23, 2009, a
SET(CPACK_POSTFLIGHT_SCRIPT "/installed/script/here.sh")
should do what you want...
"grep -rn POSTFLIGHT_SCRIPT *" in the CMake source tree for more
info/example use...
HTH,
David
On Sat, Nov 21, 2009 at 7:37 AM, Gustavo Gutierrez <
gustavo.ggutier...@gmail.com> wrote:
> Hello,
>
> Is it pos
Well your question was how to set up the drop site. You did it. You're done.
Now to use it, clients must run dashbaords and submit them to the new site.
Presumably you will download the CTestConfig.cmake from cdash since you set
up the project and add it to the HDF5 source tree next to the top lev
19 matches
Mail list logo
