More optimizations of Sin and Cos

[pleroy]

1. Cleanup leftovers from the previous version of OSACA macros.
2. Simplify the FMA wrappers.
3. Change all the algorithms except Sin near 0 to operate on the absolute value of their operand and restore the sign at the end if needed.
4. Extract a deeply nested FMA and replace it with a multiplication. This makes a difference on Zen.

Comparison on Zen 3. Before:

RAW TSC:                         min      1‰      1%      5%     10%     25%     50%
            identity            4.94   +0.00   +0.00   +0.00   +0.38   +0.38   +0.38
    sqrtps_xmm0_xmm0           16.72   +0.38   +0.38   +0.38   +0.38   +0.38   +0.38
     mulsd_xmm0_xmm0            7.60   +0.00   +0.00   +0.00   +0.00   +0.00   +0.00
  mulsd_xmm0_xmm0_4x           15.20   +0.00   +0.00   +0.00   +0.00   +0.00   +0.00
Slope: 1.186186 cycle/TSC
Correlation coefficient: 0.999887
Cycles:             expected     min      1‰      1%      5%     10%     25%     50%
R           identity       0   -0.07   +0.00   +0.00   +0.00   +0.45   +0.45   +0.45
R    mulsd_xmm0_xmm0       3    3.08   +0.00   +0.00   +0.00   +0.00   +0.00   +0.00
R mulsd_xmm0_xmm0_4x      12   11.64   +0.45   +0.45   +0.45   +0.45   +0.45   +0.45
       principia_cos           52.21   +0.45   +0.90   +0.90   +0.90   +0.90   +1.35
R   sqrtps_xmm0_xmm0      14   13.90   +0.45   +0.45   +0.45   +0.45   +0.45   +0.45
             std_cos           30.58   +0.45   +0.45   +0.45   +0.45   +0.45   +0.45

After:

RAW TSC:                         min      1‰      1%      5%     10%     25%     50%
            identity            4.94   +0.00   +0.00   +0.00   +0.38   +0.38   +0.38
    sqrtps_xmm0_xmm0           16.72   +0.00   +0.38   +0.38   +0.38   +0.38   +0.38
     mulsd_xmm0_xmm0            7.60   +0.00   +0.00   +0.00   +0.00   +0.00   +0.00
  mulsd_xmm0_xmm0_4x           15.20   +0.00   +0.00   +0.00   +0.00   +0.00   +0.00
Slope: 1.186186 cycle/TSC
Correlation coefficient: 0.999887
Cycles:             expected     min      1‰      1%      5%     10%     25%     50%
R           identity       0   -0.07   +0.00   +0.00   +0.00   +0.45   +0.45   +0.45
R    mulsd_xmm0_xmm0       3    3.08   +0.00   +0.00   +0.00   +0.00   +0.00   +0.00
R mulsd_xmm0_xmm0_4x      12   12.10   +0.00   +0.00   +0.00   +0.00   +0.00   +0.00
       principia_cos           48.16   +0.90   +0.90   +0.90   +1.35   +1.35   +1.35
R   sqrtps_xmm0_xmm0      14   13.90   +0.00   +0.45   +0.45   +0.45   +0.45   +0.45
             std_cos           30.58   +0.45   +0.45   +0.45   +0.45   +0.45   +0.45

Comparison on Golden Cove. Before:

RAW TSC:                         min      1‰      1%      5%     10%     25%     50%
            identity            2.40   +0.06   +0.06   +0.08   +0.08   +0.10   +0.10
    sqrtps_xmm0_xmm0            9.62   +0.02   +0.04   +0.06   +0.06   +0.08   +0.10
     mulsd_xmm0_xmm0            5.16   +0.02   +0.02   +0.04   +0.06   +0.76   +0.80
  mulsd_xmm0_xmm0_4x           11.84   +0.04   +0.06   +0.08   +0.08   +0.10   +0.12
Slope: 1.710658 cycle/TSC
Correlation coefficient: 0.999084
Cycles:             expected     min      1‰      1%      5%     10%     25%     50%
R           identity       0   -0.27   +0.07   +0.07   +0.10   +0.10   +0.14   +0.17
R    mulsd_xmm0_xmm0       4    4.38   +0.07   +0.07   +0.10   +0.10   +0.14   +1.33
R mulsd_xmm0_xmm0_4x      16   15.88   +0.03   +0.07   +0.10   +0.10   +0.14   +0.17
       principia_cos           52.35   +0.48   +0.72   +0.96   +1.13   +1.78   +2.39
R   sqrtps_xmm0_xmm0      12   11.94   +0.14   +0.17   +0.21   +0.21   +0.24   +0.24
             std_cos           31.58   +0.17   +0.24   +0.31   +0.34   +0.41   +0.51

After:

RAW TSC:                         min      1‰      1%      5%     10%     25%     50%
            identity            2.40   +0.06   +0.06   +0.08   +0.08   +0.10   +0.12
    sqrtps_xmm0_xmm0            9.62   +0.02   +0.04   +0.04   +0.06   +0.08   +0.08
     mulsd_xmm0_xmm0            5.16   +0.02   +0.02   +0.04   +0.04   +0.06   +0.78
  mulsd_xmm0_xmm0_4x           11.86   +0.02   +0.04   +0.06   +0.06   +0.08   +0.10
Slope: 1.707841 cycle/TSC
Correlation coefficient: 0.999116
Cycles:             expected     min      1‰      1%      5%     10%     25%     50%
R           identity       0   -0.23   +0.03   +0.07   +0.07   +0.07   +0.10   +0.68
R    mulsd_xmm0_xmm0       4    4.41   +0.03   +0.03   +0.07   +0.07   +0.10   +0.14
R mulsd_xmm0_xmm0_4x      16   15.86   +0.03   +0.07   +0.10   +0.14   +0.17   +3.01
       principia_cos           46.05   +0.41   +0.58   +0.72   +0.82   +0.99   +1.26
R   sqrtps_xmm0_xmm0      12   11.96   +0.10   +0.14   +0.17   +0.17   +0.20   +0.24
             std_cos           31.50   +0.17   +0.27   +0.34   +0.38   +0.44   +0.55

[eggrobin]

Suggested change

	double const abs_e = _mm_cvtsd_f64(_mm_xor_pd(_mm_set_sd(e), sign));
	double const e_abs = _mm_cvtsd_f64(_mm_xor_pd(_mm_set_sd(e), sign));

(meaning e_abs, short for something like e_{abs θ}) otherwise it looks like it is abs e.