[x265] [PATCH] quant.cpp: use 'rdoQuant_c' primitive to optimize rdoQuant path

[Ashok Kumar Mishra]

On Tue, Nov 28, 2017 at 3:36 PM, <praveen at multicorewareinc.com> wrote:

> # HG changeset patch
> # User Praveen Tiwari <praveen at multicorewareinc.com>
> # Date 1511851222 -19800
> #      Tue Nov 28 12:10:22 2017 +0530
> # Node ID d732ca2095defdbf42748327006083befb30a89e
> # Parent  4d242c555d14ca8214d9da89cef41c4418af4dca
> quant.cpp: use 'rdoQuant_c' primitive to optimize rdoQuant path
>
> diff -r 4d242c555d14 -r d732ca2095de source/common/quant.cpp
> --- a/source/common/quant.cpp   Tue Nov 28 11:43:00 2017 +0530
> +++ b/source/common/quant.cpp   Tue Nov 28 12:10:22 2017 +0530
> @@ -803,20 +803,14 @@
>
>              if (usePsyMask)
>              {
> -                // TODO: we can't SIMD optimize because PSYVALUE need
> 64-bits multiplication, convert to Double can work faster by FMA
> +                // Expected to work faster by FMA SIMD
> +                primitives.rdoQuant(m_resiDctCoeff, m_fencDctCoeff,
> costUncoded, &totalUncodedCost, &totalRdCost, psyScale, blkPos, log2TrSize);
> +                blkPos = codeParams.scan[scanPosBase];
> +
>                  for (int y = 0; y < MLS_CG_SIZE; y++)
>                  {
>                      for (int x = 0; x < MLS_CG_SIZE; x++)
>                      {
> -                        int signCoef         = m_resiDctCoeff[blkPos +
> x];            /* pre-quantization DCT coeff */
> -                        int predictedCoef    = m_fencDctCoeff[blkPos + x]
> - signCoef; /* predicted DCT = source DCT - residual DCT*/
> -                        costUncoded[blkPos + x] =
> static_cast<double>(((int64_t)signCoef * signCoef) << scaleBits);
> -                        /* when no residual coefficient is coded,
> predicted coef == recon coef */
> -                        costUncoded[blkPos + x] -=
> PSYVALUE(predictedCoef);
> -
> -                        totalUncodedCost += costUncoded[blkPos + x];
> -                        totalRdCost += costUncoded[blkPos + x];
> -
>
Is it not possible to insert the following code in primitive, so that it
can be written in assembly?

                         const uint32_t scanPosOffset =  y * MLS_CG_SIZE +
> x;
>                          const uint32_t ctxSig =
> table_cnt[patternSigCtx][g_scan4x4[codeParams.scanType][scanPosOffset]] +
> ctxSigOffset;
>                          X265_CHECK(trSize > 4, "trSize check failure\n");
> _______________________________________________
> x265-devel mailing list
> x265-devel at videolan.org
> https://mailman.videolan.org/listinfo/x265-devel
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.videolan.org/pipermail/x265-devel/attachments/20171129/a51462dd/attachment.html>