Crash at ippsCalcSF_16s32F in MP3 Floating Point Encoder 7.0 Update 5

[pgvildys]

Hi All,

I am still gathering information, but I am experiencing a crash in the MP3 Floating Point Encoder method at the ippsCalcSF_16s32F function when running it on a 32-bit Windows platform using IPP 7.0 Update 5 (both UMC and IPP libraries have been updated to IPP 7.0 Update 5).

I noticed that in Update 3, there's a fix listed as:

DPD200207003 ippsCalcSF_16s32F method crashes on 64 bit systems.

Has the above fix been regression tested on 32-bit platforms using the MP3 encoder or similar? Is this crash a known issue?

I will update if I get more information, but it would be great to hear if there is some information known internally to Intel.

NOTE: This crash was not seen when running IPP 6.0 Update 1.

Thanks!

Cheers,
Paul

[Sergey_K_Intel]

Hi Paul,

Yes, there was some bug in this function fixed on base of bug requestDPD200207003 "The ippsCalcSF_16s32F method crashes on 64 bit systems". But, that problem was only for 64-bit version and there was a small test case worked fine on 32 bit.

It could be different problem from what you are observing. If you can provide us with additional information regarding the issue, it will be great.

Thank you in advance,

Sergey