or the
CUDAHOSTCXX environment variable.
--
Chuck Atkins
Staff R Engineer, Scientific Computing
Kitware, Inc.
On Wed, May 8, 2019 at 7:27 AM JR Cary <mailto:c...@txcorp.com>> wrote:
Is there a standard way to deal with 2 C++ compilers? Getting both
there versions, etc
Is there a standard way to deal with 2 C++ compilers? Getting both
there versions, etc.?
I need one compiler for compiling ordinary C++ code and a different
one to use as the host compiler for CUDA.
Thx..John Cary
--
Powered by www.kitware.com
Please keep messages on-topic and check the
Regarding the previous email, below
The method of library construction is to first create an
object library with
add_library(${libname}obj OBJECT ${${LIBNAME}_SOURCES}
${${LIBNAME}_CUDASRCS})
then to create the actual library with
add_library(${libname} $)
and I have tried setting:
Porting our cmake build system to use enable_language(CUDA). This is a
complex
computational application that make use of many libraries. Upon doing this,
add_executable ended up generating a link line of 9455 chars, causing
Windows
cmd to fail to execute it. In addition, the final link used
Re the thread below,
Does the newer, enable_languagedo a better job of not having the
repeated reconfigurations and full project rebuilds as described at
https://cmake.org/pipermail/cmake/2011-January/042173.html
and many times after?
John
On 3/30/19 6:45 AM, cmake-requ...@cmake.org wrote:
Thanks. All now configures with 3.9.4.
On 12/12/17 5:50 AM, Brad King wrote:
> On 12/11/2017 04:34 PM, John Cary wrote:
>> Bill Hoffman wrote:
>>> LINK: command "xilink ... /out:FortranCInterface.exe ..." failed (exit code
>>> 1169)
>>> LIBCMT.lib(winapisupp.obj) : error LNK2005:
>>>
at the preprocessor is not being run. It requires -fpp to be added to the
Fortran compiler
flags. I do not see how to do that for CMakeFortranCompilerId.F. Tried
set(CMAKE_Fortran_FLAGS "${CMAKE_Fortran_FLAGS} -fpp")
but that did not help.
On 11/26/17 4:29 PM, JR Cary wrote:
>
Works with 3.8.2.
Compilers specified by
-DCMAKE_C_COMPILER:FILEPATH='C:/Program Files (x86)/Intel/Composer XE
2015/bin/intel64/icl.exe' -DCMAKE_CXX_COMPILER:FILEPATH='C:/Program
Files (x86)/Intel/Composer XE 2015/bin/intel64/icl.exe'
-DCMAKE_Fortran_COMPILER:FILEPATH='C:/Program Files
This is partially a continuation of the issue described at
http://public.kitware.com/pipermail/cmake/2015-June/060937.html,
where the hope was to be able to use object libraries with cuda.
There was no solution at that time, so we dropped this path,
but now this has become more important.
I
Working on a just delivered Cray I am again challenged by the way that
CMake adds libraries to the end of the link line that I did not specify.
The Cray requires use of the compiler wrappers. After adding the
appropriate libraries (hdf5, z) to my target, the link line is
ilto:nilsglad...@gmail.com>> wrote:
On 14.10.2015 14:51, JR Cary wrote:
Thanks for your response. This is a big headache for us, as
it slows down
our builds tremendously.
Is there any way to prevent or control this behavior?
I don't see how. You
Thanks for your response. This is a big headache for us, as it slows down
our builds tremendously.
Is there any way to prevent or control this behavior?
Thx..John
On 10/14/15 6:45 AM, Nils Gladitz wrote:
On 14.10.2015 02:01, John Cary wrote:
For future reference, one can determine what
In our use of cmake on a particular project, we noticed that
after building the project, upon invoking "make install", the
build starts all over again. Hence we are building the project
twice.
It seems that this is occurring because nearly everything depends
on the target,
I would like to use OBJECT libraries with CUDA.
If I try
cuda_add_library(foo OBJECT some .cpp files some .cu files)
I get
CMake Error at
/opt/contrib-clangcxx11/cmake-3.2.2-ser/share/cmake-3.2/Modules/FindCUDA.cmake:1518
(add_library):
OBJECT library txsbaseobj contains:
How can one use ctest/cdash with a target that is not all?
I have a project with two top-level targets, say all and other, where
make all does a standard build of executables that one can test, while
make other does a distinct set of actions, in particular not
creating any executables, so one
Oops, found it: CTEST_BUILD_TARGET
Sorry using up bandwidth.JC
On 1/11/14 6:45 AM, JR Cary wrote:
How can one use ctest/cdash with a target that is not all?
I have a project with two top-level targets, say all and other, where
make all does a standard build of executables that one can
16 matches
Mail list logo