How do you use ippsDeflateHuff_8u with ippsDeflateLZ77_8u?

[dfati1]

Hello,

Subject is the question,  How do you use ippsDeflateHuff_8u with ippsDeflateLZ77_8u?

I presume you use ippsDeflateLZ77_8u first, then ippsDeflateHuff_8u  but I don't see any in-between function to build the Huffman Tables?

Is there a clean example of just compressing one buffer of data using LZ77 with Huffman Encoding?

Thanks.

[dfati1]

I found ippsDeflateLZ77FastestGenHuffTable_8u which seems to work.

Although I finally figured out how to use the latest version, it turns out it's much slower than Version 7.   The tests I did with the old ipp_zlib took 8 seconds, with the new one it took 18 seconds, also the raw compression not using zlib, the old took 6 seconds (and compressed more) vs the new which takes 8 seconds (and compressed less).    It was consistent so I'll just stick with the old stuff.

[Ruqiu_C_Intel]

Please share out your reproducers for the performance gaps, Including share the Software, OS, and Hardware information