problem with CAVLC Decoding IPP func

[arunjohnson]

I was developing an H.264 decoder using IPP func. I found that the func ippiDecodeCAVLCCoeffs_H264_1u16s() is crashing for input value 0x39123800, . which is to be decoded as(0,0,-1,0, 0,0,0,0, 0,-1,0,0, 0,0,0,0).

i made the tables for the func using huffman IPP functions.

Message Edited by arunjohnson on 06-20-2005 04:15 AM

Message Edited by arunjohnson on 06-20-2005 04:17 AM

[Ying_S_Intel]

Have you tried the latest media procssing sample including H.264 coding component from IPP 5.0 beta ?

Thanks,
Ying

[arunjohnson]

I downloaded the ipp 5.0 beta from here.

http://www.intel.com/software/products/ipp/beta/

If you have latest version Let me know.

I am using the func ippiDecodeCAVLCCoeffs_H264_1u16s(). I is crashing for input 0x39100000 , to be decoded as (0,0,-1,0,0,0,0,0,0,-1,0,0,0,0,0,0) with

nC = 0 and maxCoeff = 16.

This input specifies a Tot Coeff = 2 Tot Zeros = 8. I found the func is working for Tot Coeff = 2 Tot Zeros = 9, Tot Coeff = 2 Tot Zeros = 7.

The func ippiDecodeCAVLCChromaDcCoeffs_H264_1u16s() is found to be working fine.

[Ying_S_Intel]

This is the latest version.

Could you please submit the details on the crash for this funciton ippiDecodeCAVLCCoeffs_H264_1u16s and system configuration to product "Beta Intel IPP for Windows*" at Intel Premier Support ?

Thanks,
Ying

[arunjohnson]

I am not able to log into Intel Premier Support.

How do I become a registered Premier support user ?

I have told the problem before. I was unit testing the func and found this problem.

My configuration:

Intel 2.2 GHz Pentium Processor

512 MB RAM,

Windows 2000 service pack 4.

VC++ 6 with IPP version 5.0 beta.