[LayerNorm Optimize x86] AVX512/AVX/SSE intrinsic

[LRY89757]

• Add the AVX512/AVX/SSE intrinsic for layernorm
• Add some test samples for elempack == 16

[codecov-commenter]

Codecov Report

Merging #4060 (cf053cc) into master (4f414c1) will decrease coverage by 1.43%.
The diff coverage is 97.35%.

@@            Coverage Diff             @@
##           master    #4060      +/-   ##
==========================================
- Coverage   94.41%   92.98%   -1.44%     
==========================================
  Files         745      743       -2     
  Lines      178496   178516      +20     
==========================================
- Hits       168533   165990    -2543     
- Misses       9963    12526    +2563     

Impacted Files	Coverage Δ
src/layer/x86/layernorm_x86.cpp	97.35% <97.35%> (ø)
src/layer/x86/convolution_2x2_pack8.h	2.75% <0.00%> (-97.25%)	⬇️
src/layer/x86/deconvolution_pack8.h	10.76% <0.00%> (-89.24%)	⬇️
src/layer/x86/convolution_sgemm_pack8.h	14.24% <0.00%> (-85.24%)	⬇️
src/layer/x86/convolution_sgemm_pack4to8.h	29.16% <0.00%> (-70.84%)	⬇️
src/layer/x86/convolution_pack8.h	34.42% <0.00%> (-65.58%)	⬇️
src/layer/x86/convolution_pack4to8.h	42.85% <0.00%> (-55.11%)	⬇️
...c/layer/x86/convolution_winograd_transform_pack8.h	54.90% <0.00%> (-45.10%)	⬇️
src/layer/x86/convolution_3x3_pack1to8.h	39.95% <0.00%> (-40.04%)	⬇️
src/layer/x86/convolution_winograd_dot_pack8.h	60.24% <0.00%> (-39.16%)	⬇️
... and 83 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 4158e63...cf053cc. Read the comment docs.

[LRY89757]

I will try to finish the merge of instancenorm and the batchnorm then. At the same time, I will also finish the layernorm in my own way

[nihui]

close for 03f2ad3