英特尔® 集成性能原件

Invalid read of size 8 when srcRoiSize is for full source image i.e same as first argument pSrc in ippiCopyConstBorder_16u_C1R

Invalid read of size 8 when using ippiCopyConstBorder_16u_C1R(const Ipp<datatype>* pSrc, int srcStep, IppiSize srcRoiSize, Ipp<datatype>* pDst, int dstStep, IppiSizedstRoiSize, int topBorderHeight, int leftBorderWidth, const Ipp<datatype> value[3] for full source image ROI as highlighted.

When pSrc image is changed( increasing its width and height by 1) then there is no issue.

Benchmark IPP loss less compression

Dear all,

 

I have an request of an customer for compressing some video and control data in real time.

The data shall received as a data stream with 600 - 700 MB/s via PCIe. The system contains a Intel Core I7, 32 GB RAM and Raid 5 with SSD.

My open points are:

- Which data througput is possible with the loss less compressin in IPP?

- Where I can find some measurements?

- Who has some experiences with compression of high data rates? 

Thanks and Regards

 

RHa123

Getting Started with Intel® Integrated Performance Primitives for Microcontrollers Examples

Intel® Integrated Performance Primitives (Intel® IPP) for Microcontrollers are a set of the performance building blocks for the signal and data processing applications for x86-compatible microcontrollers. The article provide demonstrate how to add Intel IPP functions into the application with the Intel IPP example code.   It uses the QMSI based application as the example. For the zephyr application code, it can use the similar steps to add Intel IPP functions:

Building an Intel IPP Example Code

  • 专业版
  • 学生
  • Linux*
  • 英特尔® 集成性能原件
  • signal processing
  • 媒体处理
  • 微型控制器
  • 许可协议: 

    using ippiAlphaComp with separate RGB and A buffers

    Hi - I'm trying to composite images where the RGB and Alpha channels are held in different image buffers.

    My app uses windows GDI bitmaps (BGR order) for to hold color information, and a home-built image format to hold a one-byte-per-pixel alpha channel. The alpha is independent from the RGB and needs to be kept separate.

    I can pass pointers to my buffers to IPP as Ipp8u for functions expecting 8u_C3R and 8u_C1R.

    However ippiAlphaComp seems to work only with 8u_AC4R 

    Bit exact issue with ippsLPCToLSP_GSMAMR_16s function which leads to crash

     

    Hi,

    We got a crash issue reported from the deployment in the field in GSM AMR encoder using Intel IPP.

    We are able to easily reproduce this issue with a 30 frame voice clip.

    The crash is happening in function ippsAlgebraicCodebookSearch_GSMAMR_16s. After back tracing the crash, we found that the issue happened because of invalid T0 and gain_pit values. So we analyzed the difference in bit exactness, and followed the trail, and found that it is originating in the function ippsLPCToLSP_GSMAMR_16s.

    Is it possible to use IPP by command-line parameter to compiler or in source code pragma?

    Hi everyone!

    We are using CMake to manage our project generation, and we generates VS projects in Windows platform. However, to use IPP library, we have to manually enable it in VS project configuration dialog, which is very inconvenient. So is it possible to declare IPP usage in source code pragma like this:

    #ifdef __INTEL_COMPILER
    #some_pragma_that_declare_use_ipp
    #endif

    If that is not possible, how to use IPP by command-line parameters to icl program? We can pass these parameters by force in CMake.

    Thanks for a lot!

    订阅 英特尔® 集成性能原件