VK_KHR_cooperative_matrix 😆

[nihui]

• conv1x1s1d1 pack4 16x16x16
• conv1x1s1d1 pack4 16x8x8
• conv gemm pack4 16x16x16
• conv gemm pack4 16x8x8
• conv 3x3 winograd pack4 16x16x16
• conv 3x3 winograd pack4 16x8x8
• deconv pack4 16x16x16
• deconv pack4 16x8x8
• update glslang Implement support for GL_KHR_cooperative_matrix extension KhronosGroup/glslang#3248
• khr and nv extension switch macro
• update test

[codecov-commenter]

Codecov Report

Merging #4823 (b4cf299) into master (07b8400) will decrease coverage by 5.00%.
The diff coverage is 55.31%.

@@             Coverage Diff             @@
##           master    #4823       +/-   ##
===========================================
- Coverage   94.80%   89.81%    -5.00%     
===========================================
  Files         776      306      -470     
  Lines      225898    86875   -139023     
===========================================
- Hits       214160    78024   -136136     
+ Misses      11738     8851     -2887     

Files Changed	Coverage Δ
src/layer/vulkan/deconvolution_vulkan.cpp	92.81% <50.00%> (-0.22%)	⬇️
src/layer/vulkan/convolution_vulkan.cpp	90.42% <51.72%> (-6.09%)	⬇️
src/gpu.cpp	79.74% <63.04%> (-2.46%)	⬇️

... and 634 files with indirect coverage changes

[nihui]

cpm 16 16 16  0 0 1 1  3
cpm 16 16 16  0 0 0 0  3
cpm 16 16 16  7 7 5 5  3
cpm 16 16 16  3 3 5 5  3
cpm 16 16 16  7 3 5 5  3
cpm 16 16 16  3 7 5 5  3
[0 AMD Radeon RX 7900 XTX]  queueC=1[2]  queueG=0[1]  queueT=2[2]
[0 AMD Radeon RX 7900 XTX]  bugsbn1=0  bugbilz=0  bugcopc=0  bugihfa=0
[0 AMD Radeon RX 7900 XTX]  fp16-p/s/a=1/1/1  int8-p/s/a=1/1/1
[0 AMD Radeon RX 7900 XTX]  subgroup=64  basic=1  vote=1  ballot=1  shuffle=1

cpm 16 16 16  0 0 0 0  3
cpm 16  8 16  0 0 0 0  3
cpm 16  8  8  0 0 0 0  3
cpm 16 16 16  0 0 1 1  3
cpm 16  8 16  0 0 1 1  3
cpm 16  8  8  0 0 1 1  3
cpm 16 16 32  7 7 9 9  3
cpm 16 16 32  3 3 5 5  3
cpm 16  8 32  7 7 9 9  3
cpm 16  8 32  3 3 5 5  3
cpm  8  8 32  7 7 9 9  3
cpm  8  8 32  3 3 5 5  3
[0 NVIDIA GeForce RTX 3060]  queueC=2[8]  queueG=0[16]  queueT=1[2]
[0 NVIDIA GeForce RTX 3060]  bugsbn1=0  bugbilz=0  bugcopc=0  bugihfa=0
[0 NVIDIA GeForce RTX 3060]  fp16-p/s/a=1/1/1  int8-p/s/a=1/1/1
[0 NVIDIA GeForce RTX 3060]  subgroup=32  basic=1  vote=1  ballot=1  shuffle=1

[nihui]

[whyb]

等呀等，熬呀熬，最终熬成了阿香婆~