Optimización

Parallel build with prof-gen=srcpos option not working?

I am working on code coverage for a large project. We build our project which contains hundreds of cpp/h files spread in different directories. 

I build our project using latest icpc with options -prof-gen=srcpos before I run profmerge and codecov. I build with parallel mode : make -j

I found the sizes of generated spi file varies a lot and the final code coverage report is incorrect. I suspect the spi file is generated incorrectly.

Is it because of the parallel build? Do I have to do sequential build with prof-gen=srcpos option?

Thank.

64-ia-32 instruction reference manual --- FXAM description incorrect?

On page 3-378 of the 64-ia-32 documentation set, volume 2A, order number 325383 (April 2016), the instruction FXAM is described as "FXAM --- Examine ModR/M".  The instruction's documentation does not show the instruction takes ModR/M bytes.  In fact, the instruction is described as operating on a fixed register, i.e. ST(0).

Hard-wire 1 thread/core into an OpenMP application

I would like to to hard-wire the number of threads per core (along with affinity) into an executable using OpenMP, so that 1 thread per core would be used

  1. Regardless of how many physical cores the processor has
  2. Regardless of whether hyper-threading is enabled or not
  3. Regardless of whether the user has set OMP_NUM_THREADS or not

Intel C++ compiler supports -qopt-threads-per-core for the Intel MIC architecture. It would have been the ideal solution for me, but for some reason this argument is not supported for general-purpose CPU architecture.

How to get integer add and sub count from intel Vtune

I am using intel VTune to profile my program.

The CPU I am using is IVY Bridge.

All the hardware instruction event can be found here:

https://software.intel.com/en-us/node/589933

FP_COMP_OPS_EXE.X87

Number of FP Computational Uops Executed this cycle. The number of FADD, FSUB, FCOM, FMULs, integer MULsand IMULs, FDIVs, FPREMs, FSQRTS, integer DIVs, and IDIVs. This event does not distinguish an FADD used in the middle of a transcendental flow from a s

FFMPEG Live versus File Priorities

Hi,

We are using Intel Media Server Studio with ffmpeg to transcode 10 live streams each to 2 HLS bitrates.  This is using about 80-90% of the GPU capacity.  Each transcode is run as a separate process and the decoding is done using the cpu.  The processor is a Core-i7 5775C.

If a file is also transcoded then the GPU processing continuously hits 100%.

Is it known or is there a way to make sure that the transcoding for the live streams has priority over file transcoding?  Is it possible to be able set the equivalent of "nice" value for the GPU?

Thanks,

An internal error detected during analysis

Unfortunately I can't use inspector for any of the projects that I am working on. I am not sure if it is related to a Windows update, as it used to work before, but now for any project, level 2 (Detect Memory Problems) and level 3 (Locate Memory Problems) of Inspector fail to work. I use Windows 7 Professional, SP1, If needed I can provide a list of installed updates.

I am using Inspector_XE_2016_update2 integrated with MSVS2010. I also tried Inspector 2017 beta with MSVS2012 and got the same error. Here is a part of log in Collector Messages in Inspector 2016:

Issue installing Parallel Studio 2016 update 3 Cluster Edition

Hi, I have an issue installing Parallel Studio 2016 update 3 on windows 10. I have Visual Studio 2015 installed in my PC. The error is.

Intel(R) Parallel Studio XE 2016 Update 3 Cluster Edition for Windows* Setup Wizard ended prematurely because of an error(s).
  Windows Installer failed to configure product:
    Error message: An error occured during the installation of assembly component 844EFBA7-1C24-93B2-A01F-C8B3B9A1E18E. HRESULT: 0x80073715.  
    Error code: 1603
    File name: msvcrt_x64.msi

Using newer version of gcc compilers on Centos 6.x (devtoolset-3) with icc

Our primary development platform on Linux is Centos 6.7  with Composer XE 2015 - the default gcc version is 4.4.7. We need to use C++11 and STL features as supported by gcc 4.9.2. This gcc version can be installed and activated by installing "devtoolset-3" and "enabled"  via "source /opt/rh/devtoolset-3/enable".

I am found two confusing issues. If I compile a C++ file that uses , for example, C++11 <regex> features  this is what I find

Using gcc 4.4.7 : gcc sample.cpp ->does not compile as 4.4.7 does not support std::regex properly

Suscribirse a Optimización