I'm trying to install the cuda-accelerated library found on this website : http://mklab.iti.gr/project/GPU-LIBSVM
I downloaded the linux makefile and executable and tried to follow the readme file :
Instructions to compile Linux GPU-Accelerated LIBSVM
Install the NVIDIA drivers, CUDA toolkit and GPU Computing SDK code samples. You can find them in:
http://developer.nvidia.com/object/cuda_3_2_downloads.html (January
2011)
You may need some additional packets to be installed in order to
complete the installations above. Please refer to the web for more
details.
Copy this folder to "/NVIDIA_GPU_Computing_SDK/C/src"
Use the Makefile found in "/NVIDIA_GPU_Computing_SDK/C"
Find the "svm-train-gpu" executable in /NVIDIA_GPU_Computing_SDK/C/bin/linux/release
But i could not find the "NVIDIA_GPU_Computing_SDK" folder, even after upgrading cuda to 5.5 (wich should contain the SDK). It seems that this instructions are outdated. So I copied the folder into "opt/cuda/src/" and used this makefile :
EXECUTABLE := svm-train-gpu
CUFILES :=
CU_DEPS :=
CCFILES := svm-train.c svm.cpp
CUDACCFLAGS := -po maxrregcount=16
USECUBLAS := 1
include ../../common/common.mk
It first told me it could not find "common.mk", so I found in my disk an other common.mk file from an other CUDA project implementation called "common-gcc-cuda-4.0.mk" wich is described in the header as a "Common build script for CUDA source projects for Linux and Mac platforms" and tried again.
Now I have this error :
$ make
make: *** No rule to make target `makedirectories', needed by `bin/linux/release/svm-train-gpu'. Stop.
Can anyone help me please ?
It sounds like the library is expecting an old version of CUDA. The structure of the sample codes (GPU_Computing_SDK) changed substantially after CUDA 4.2
Try starting over with an older version of CUDA which you can find here. I would recommend choosing a cuda version from 3.2 to 4.2, and if you install that completely (driver, toolkit and SDK) you will find the folders and makefiles being referenced.
The version you were trying was created with the CUDA SDK 3.0. As Robert pointed out the structure of sample codes and makefiles changed substantially after CUDA 4.2.
A new version (1.2) is online on the mklab website http://mklab.iti.gr/project/GPU-LIBSVM
which supports the latest LIBSVM v3.17 and CUDA SDK v5.5.
Also you can see a detailed changelog here: https://github.com/MKLab-ITI/CUDA/releases
Related
I have freetype2 installed on my system but I am also trying to install it from source, as well.
My installed version is here: /usr/lib64/pkgconfig/freetype2.pc. That file is listed as version 20.0.14.
To compile another tool, namely fontconfig, I need freetype2 >= 21.0.15.
So I downloaded freetype2 from its source git repo.
The trouble I'm having is that repository doesn't list any version even close to 21.0.15+. All of the tags in that repository are listed as VER-2-6-2, VER-2-9, VER-2-9-1, etc. And the generated freetype.pc file which is created after compilation are also listed as 2.6.2, 2.9, 2.9.1, etc. The version number is not 21.0.15 or even 20.0.14, which my CentOS 7.6 already has.
What determines the listed version in the freetype.pc file? Where can I find the version of freetype that will create a .pc file of >= 21.0.15+? Any clarification would be greatly appreciated.
Apparently the reason is because I was building freetype using CMake. And in the CMakeLists.txt, it's defining a version based on the version of freetype.
# . `CMakeLists.txt' is provided as-is since it is normally not used by the
# developer team.
So it's unsupported. That's why the version number doesn't match.
In reality, the version number actually comes from libtool. See this table. e.g. I was building 2.10.0 freetype, so the freetype.pc should have been written with a version of 23.0.17
And if you build using configure, instead of cmake, you can find that exact same version 23.0.17 listed as "23:0:17" in the builds/unix/configure.raw file.
I don't mean the version(s) provided by the various distributions but the binary from the official website.
I have an old VM running 32bit OpenSUSE 12.1 that is configured for a project I'm working on at work. I need to install WebKitGTK. The problem is that the cmake in the repositories is ancient 2.x, while WebKitGTK at least 3.6 (or similar). So I went to the official website and (my fault) without looking too much into it downloaded the 3.10 installation for Linux.
Upon executing the binary that was installed I got the error that the file could not be run. I checked the execution rights and it was fine. Then it struck me...I ran file cmake and got 64 instead of the required 32bit.
I went back to the website and all I could find were 32bit versions for Windows but none for Linux.
I can build it from source but just out of curiousity would like to know if support has been dropped. I was unable to find any information so far.
32-bit support for CMake hasn't been dropped. They just don't provide binaries for it on their website as of CMake 3.7.0
When I create a new javaFX project the javaFX variables are unresolved. I use java 8 (121) and eclipse Neon on which I installed e(fx)clipse. OS is Linux Mint 18.1. What I did:
I uninstalled and installed e(fx)clipse again, that didn't help
In another post I read that someone had solved the problem by adding
jfxrt.jar to the build path. I searched for this file but it is not present on my file system
Anybody any idea what is wrong in my setup?
Edit
James_D suggested adding openjfx. I read documentation of openjfx and it says that only a subset is implemented. I decided to install the oracle JDK and adjust the building path with a reference to the oracle version of jfxrt. That worked. I wonder whether it wouldn't be better to switch to the oracle version of JDK?
The release notes for RedHawk 2.0 say that the GPP device previously written in Python has been replaced with one written in "Written in C++, so it is more responsive". But I find it still running in Python (according to ps command python is running GPP.py, and the $SDRROOT/dev/devices/GPP/GPP.spd.xml which also has softpkg version="1.10.0". Was my installation defective and I still have parts of the 1.10 runtime system? My IDE says 2.0.
It sounds like REDHAWK 2.0 was not properly installed on your system, the IDE and the framework/assets are separate and it is possible to get into a situation with conflicting versions depending on the installation steps taken.
Determining what version of REDHAWK you have installed can be determined in a handful of ways. If you installed via yum or rpm you can check the versions of the rpms installed with:
rpm -qa | grep -i redhawk
The redhawk package, and redhawk-ide package should both be at 2.0. Note that the REDHAWK assets are versioned independently.
If you installed via source, you can use the package config files to obtain version information. The framework keeps it's pc files in $OSSIEHOME/lib64/pkgconfig:
cat $OSSIEHOME/lib64/pkgconfig/ossie.pc
Will print out version information for the core framework installed. Depending on what is installed, there are pc files for the framework, bulkio, frontend, and burstio.
I am sorry. The GPP-2.0.0-3.el6.x86_64 DOES contain an ELF executable for GPP device. But the rpm does not install unless I manually erase the GPP-1.10 pkg. Until erased yum says "nothing to do" for some reason. I saw the source code in GPP-debuginfo but did not notice the executable in GPP-2.0.0 since it was all caps and looked like the directory.
I am using z3-osx-4.1-x64 running on the max os, currently I need it to run under the linux system. I tried to download the linux version http://research.microsoft.com/en-us/um/redmond/projects/z3/ , but it seems that there is no linux version.
Could you please tell me where I can download it?
Thank you very much.
You should get the source from their current site and follow the instructions in their README file.
You can simply append z3-4.1.tar.gz to http://research.microsoft.com/en-us/um/redmond/projects/z3/ to download Z3 4.1 binaries for Linux.
A more complete list of old versions could be found in Previous version of Z3 for linux.
For newest versions, it is recommended to download the master branch, compile and install on your Linux system.