Vulkan Flash Attention Coopmat1 Refactor (#19075)

* vulkan: use coopmat for flash attention p*v matrix multiplication

* fix P loading issue

* fix barrier position

* remove reduction that is no longer needed

* move max thread reduction into loop

* remove osh padding

* add bounds checks and padding

* remove unused code

* fix shmem sizes, loop duration and accesses

* don't overwrite Qf, add new shared psh buffer instead

* add missing bounds checks

* use subgroup reductions

* optimize

* move bounds check, reduce barriers

* support other Bc values and other subgroup sizes

* remove D_split

* replace Of register array with shared memory Ofsh array

* parallelize HSV across the rowgroups

* go back to Of in registers, not shmem

* vectorize sfsh

* don't store entire K tile in shmem

* fixes

* load large k tiles to shmem on Nvidia

* adapt shared memory host check function to shader changes

* remove Bc 32 case

* remove unused variable

* fix missing mask reduction tmspsh barrier

* fix mask bounds check

* fix rowmax f16 under/overflow to inf

* fix flash_attn_cm2 BLOCK_SIZE preprocessor directives

ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_cm2.comp

@@ -55,7 +55,7 @@ ACC_TYPE Max(const in uint32_t row, const in uint32_t col, const in ACC_TYPE ele
     return max(elem0, elem1);
 }
 
-#if defined(BLOCK_SIZE)
+#if BLOCK_SIZE > 1
 #define DECODEFUNC , DEQUANTFUNC
 #else
 #define DECODEFUNC
@@ -85,7 +85,7 @@ void main() {
 
     tensorViewNV<2, false, 1, 0> tensorViewTranspose = createTensorViewNV(2, false, 1, 0);
 
-#if defined(BLOCK_SIZE)
+#if BLOCK_SIZE > 1
     tensorLayoutK = setTensorLayoutBlockSizeNV(tensorLayoutK, 1, BLOCK_SIZE);
     tensorLayoutV = setTensorLayoutBlockSizeNV(tensorLayoutV, 1, BLOCK_SIZE);
 #endif
@@ -98,7 +98,7 @@ void main() {
     if (Clamp != gl_CooperativeMatrixClampModeConstantNV)
     {
         q_stride &= ~7;
-#if !defined(BLOCK_SIZE)
+#if BLOCK_SIZE == 1
         k_stride &= ~7;
         v_stride &= ~7;
 #endif