Karli,
I just wanted to update you that zlib had been empty. d'oh!
Apparently, using the GitHub GUI only cloned the benchmark directory and
libarchive but not zlip. Running the submodule update command from the
shell got me to where I needed to be!
Namik,
I am almost there. However, when I try to run the compiled program it
fails because archive.dll is missing. libarchive.dll was included with
libarchive, although I don't know if you are linking to it or not. Any
help would be greatly appreciated.
Thanks,
-Matt
On Sat, Aug 2, 2014 at 3:55 AM, Karl Rupp <r...@iue.tuwien.ac.at> wrote:
> Hi Matt,
>
>
> > Before I get into the latest CMake drama. I wanted to ask a question
> about:
>
>>
>> >> 1.) Boost: Namik, can you please remove all use of Boost (sources plus
>> >> build system files)? This is just required at very few locations,
>> mostly
>> >> sparse stuff, so it shouldn't be a big deal. I'll commit some matrix
>> >> generation code soon, which will allow us to get rid of all the IO and
>> >> the testdata files, which is the primary use of Boost.
>>
>> Does that mean that boost's compressed_matrix will not be supported by
>> the gui benchmark algorithm? Should I be using ell_matrix instead for
>> my sparse matrices? I ask, because the focus of all of my research and
>> development is on sparse matrices and I have relied heavily upon
>> compressed_matrix up till now. Which is why I was so distraught to find
>> out at ublas will be broken in boost 1.56. But that is a conversation
>> for another day.
>>
>
> this is mostly an internal thing, as the GUI is (at least in its first
> incarnation) a standalone executable and not a library. ViennaCL provides
> all the sparse matrix formats itself, so we can run all benchmarks using
> viennacl::compressed_matrix<>, viennacl::ell_matrix<> and so forth. There
> is no need to 'duplicate' these operations using uBLAS, as it is done in
> the tests for verification purposes.
>
> Btw: Is uBLAS broken in Boost 1.56? I saw your email on the uBLAS mailing
> list, but it didn't notice that things were (probably?) that bad...
>
>
>
>
> Now onto CMake builds. Historically, I was apparently incorrectly
>> downloading the zip file and making my own folder structure. Cloning
>> makes things much much easier and allowed me to download libarchive
>> automatically. I realize zlib is not working yet, but I thought I would
>> send you my CMake output with a clean build directory for your reference
>> in case you see other issues.
>>
>
> zlib *is* supposed to be working and should be included as a submodule
> just like libarchive. The following steps are sufficient both on my Linux
> Mint Maya machine and my Windows 7 machine using Git Bash:
>
> $> git clone https://github.com/viennacl/viennacl-benchmark-gui
> $> cd viennacl-benchmark-gui
>
> $> git submodule update --init
> $> mkdir build
> $> cd build
> $> cmake .. # will take a while
> $> make
>
> Notes:
> * The OpenCL library is assumed to be detectable through an environment
> variable.
> * The Boost-directories might have to be specified manually, but we try to
> get rid of this.
> * Qt should be made known to CMake by specifying CMAKE_PREFIX_PATH.
>
> I'll update the build instructions on the GitHub repository accordingly
> now, so you should have a complete step-by-step guide when you're up :-)
>
> Best regards,
> Karli
>
>
--
--------------------
Matthew Musto
matthew.mu...@gmail.com
------------------------------------------------------------------------------
Want fast and easy access to all the code in your enterprise? Index and
search up to 200,000 lines of code with a free copy of Black Duck
Code Sight - the same software that powers the world's largest code
search on Ohloh, the Black Duck Open Hub! Try it now.
http://p.sf.net/sfu/bds
_______________________________________________
ViennaCL-devel mailing list
ViennaCL-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/viennacl-devel
