------------------------------------------------------------------------------ NVIDIA CUDA Visual Profiler Linux Release Notes Version 1.0 ------------------------------------------------------------------------------ FILES IN THE RELEASE: -------------------- * CudaVisualProfiler/bin/cudaprof : CUDA Visual Profiler Executable * CudaVisualProfiler/bin/libQt*.so.4 : Qt core shared libraries * CudaVisualProfiler/projects : Directory containing sample profiler projects SUPPORTED LINUX DISTRIBUTIONS ----------------------------- NVIDIA CUDA Visual Profiler is supported and tested on the following Linux distributions: 32-bit operating systems * Red Hat Enterprise Linux 4.3 * Red Hat Enterprise Linux 5.0 64-bit operating systems * Red Hat Enterprise Linux 4.3 * Red Hat Enterprise Linux 5.0 CUDA Visual Profiler is compiled with gcc 3.4.6 and the same tools should be used on other distributions. Alternatively, compatibility packages containing libstdc++.so.6 may be used. These are available for many Linux distributions. For example, on recent Fedora Core releases, the compat-libstdc++-34 is needed. SYSTEM REQUIREMENTS ------------------- . CUDA compatible NVIDIA GPU Card . NVIDIA CUDA Toolkit INSTALLATION AND SETUP --------------------- Extract all files from the release tar zip into a directory (referred to as ) > cd > gunzip tar xvf Setup LD_LIBRARY PATH to include the CudaVisualProfiler bin directory: > export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/CudaVisualProfiler/bin RUNNING CUDA Visual Profiler ---------------------------- > /CudaVisualProfiler/bin/cudaprof & Refer to the CUDA Visual Profiler readme for more information. KNOWN ISSUES ------------ 1) Note that timestamps and profiler counters are not supported in CUDA v1.0. The CUDA Visual Profiler currenly does not detect the CUDA version and does not disable any incompatible features. 2) You need to use the command line argument "--noprompt" for running most of the CUDA SDK samples. You can enable the "Run in separate window" checkbox in the Session settings dialog to open a separate window. Only with this option you can give some keyboard input for console-based CUDA programs.