CUDA: no FP16 arithmetic for vector FA kernel (#17558)

ggml/src/ggml-cuda/common.cuh

@@ -558,8 +558,12 @@ static __device__ __forceinline__ void ggml_cuda_mad(float & acc, const float2 v
     acc += v.y*u.y;
 }
 
-static __device__ __forceinline__ void ggml_cuda_mad(float & acc, const half2 v, const half2 u) {
 #if defined(GGML_USE_HIP) && (defined(RDNA2) || defined(RDNA3) || defined(RDNA4) || defined(__gfx906__) || defined(CDNA))
+#define V_DOT2_F32_F16_AVAILABLE
+#endif // defined(GGML_USE_HIP) && (defined(RDNA2) || defined(RDNA3) || defined(RDNA4) || defined(__gfx906__) || defined(CDNA))
+
+static __device__ __forceinline__ void ggml_cuda_mad(float & acc, const half2 v, const half2 u) {
+#ifdef V_DOT2_F32_F16_AVAILABLE
     asm volatile("v_dot2_f32_f16 %0, %1, %2, %0" : "+v"(acc) : "v"(v), "v"(u));
 #else
 #ifdef FAST_FP16_AVAILABLE
@@ -571,7 +575,7 @@ static __device__ __forceinline__ void ggml_cuda_mad(float & acc, const half2 v,
     acc += tmpv.x * tmpu.x;
     acc += tmpv.y * tmpu.y;
 #endif // FAST_FP16_AVAILABLE
-#endif // defined(GGML_USE_HIP) && (defined(RDNA2)  || defined(RDNA3) || defined(RDNA4) || defined(GCN5) || defined(CDNA))
+#endif // V_DOT2_F32_F16_AVAILABLE
 }
 
 static __device__ __forceinline__ void ggml_cuda_mad(half2 & acc, const half2 v, const half2 u) {

ggml/src/ggml-cuda/fattn-common.cuh

@@ -55,11 +55,11 @@ static __device__ __forceinline__ float vec_dot_fattn_vec_KQ_f16(
         ggml_cuda_memcpy_1<sizeof(tmp)>(tmp, K_h2 + k_KQ_0 + (threadIdx.x % nthreads)*cpy_ne);
 #pragma unroll
         for (int k_KQ_1 = 0; k_KQ_1 < cpy_ne; ++k_KQ_1) {
-#ifdef FAST_FP16_AVAILABLE
+#ifdef V_DOT2_F32_F16_AVAILABLE
             ggml_cuda_mad(sum,                tmp[k_KQ_1] , ((const half2  *) Q_v)[k_KQ_0/nthreads + k_KQ_1]);
 #else
             ggml_cuda_mad(sum, __half22float2(tmp[k_KQ_1]), ((const float2 *) Q_v)[k_KQ_0/nthreads + k_KQ_1]);
-#endif // FP16_AVAILABLE
+#endif // V_DOT2_F32_F16_AVAILABLE
         }
     }
 

ggml/src/ggml-cuda/fattn-vec.cuh

@@ -86,11 +86,11 @@ static __global__ void flash_attn_ext_vec(
 
     constexpr vec_dot_KQ_t vec_dot_KQ = get_vec_dot_KQ<type_K, D, nthreads_KQ>();
     constexpr bool Q_q8_1 = type_K != GGML_TYPE_F16;
-#ifdef FAST_FP16_AVAILABLE
+#ifdef V_DOT2_F32_F16_AVAILABLE
     constexpr dequantize_V_t dequantize_V = get_dequantize_V<type_V, half,  V_rows_per_thread>();
 #else
     constexpr dequantize_V_t dequantize_V = get_dequantize_V<type_V, float, V_rows_per_thread>();
-#endif // FAST_FP16_AVAILABLE
+#endif // V_DOT2_F32_F16_AVAILABLE
 
     const int ic0 = blockIdx.x * ncols; // Index of the Q/QKV column to work on.
 
@@ -112,13 +112,13 @@ static __global__ void flash_attn_ext_vec(
 
     constexpr int ne_KQ      = ncols*D;
     constexpr int ne_combine = nwarps*V_cols_per_iter*D;
-#ifdef FAST_FP16_AVAILABLE
+#ifdef V_DOT2_F32_F16_AVAILABLE
     half2            VKQ[ncols][(D/2)/nthreads_V] = {{{0.0f, 0.0f}}};
     __shared__ half   KQ[ne_KQ > ne_combine ? ne_KQ : ne_combine];
 #else
     float2           VKQ[ncols][(D/2)/nthreads_V] = {{{0.0f, 0.0f}}};
     __shared__ float  KQ[ne_KQ > ne_combine ? ne_KQ : ne_combine];
-#endif // FAST_FP16_AVAILABLE
+#endif // V_DOT2_F32_F16_AVAILABLE
 
     float KQ_max[ncols];
     float KQ_sum[ncols];
@@ -129,11 +129,11 @@ static __global__ void flash_attn_ext_vec(
     }
 
     // Convert Q to float2 (f16 K) or q8_1 (quantized K) and store in registers:
-#ifdef FAST_FP16_AVAILABLE
+#ifdef V_DOT2_F32_F16_AVAILABLE
     half2  Q_reg[ncols][(D/2)/nthreads_KQ]; // Will be initialized completely.
 #else
     float2 Q_reg[ncols][(D/2)/nthreads_KQ] = {{{0.0f, 0.0f}}}; // May be only partially initialized.
-#endif // FAST_FP16_AVAILABLE
+#endif // V_DOT2_F32_F16_AVAILABLE
     int    Q_i32[ncols][1 > D/(sizeof(int)*nthreads_KQ) ? 1 : D/(sizeof(int)*nthreads_KQ)];
     float2  Q_ds[ncols][1 > D/(sizeof(int)*nthreads_KQ) ? 1 : D/(sizeof(int)*nthreads_KQ)];
     if constexpr (Q_q8_1) {
@@ -191,7 +191,7 @@ static __global__ void flash_attn_ext_vec(
 
         __syncthreads();
     } else {
-#ifdef FAST_FP16_AVAILABLE
+#ifdef V_DOT2_F32_F16_AVAILABLE
         const half2 scale_h2 = make_half2(scale, scale);
 #pragma unroll
         for (int j = 0; j < ncols; ++j) {
@@ -233,7 +233,7 @@ static __global__ void flash_attn_ext_vec(
                 Q_reg[j][k].y *= scale;
             }
         }
-#endif // FAST_FP16_AVAILABLE
+#endif // V_DOT2_F32_F16_AVAILABLE
     }
 
     const int k_VKQ_max = KV_max ? KV_max[sequence*gridDim.x + blockIdx.x] : ne11;
@@ -291,7 +291,7 @@ static __global__ void flash_attn_ext_vec(
             KQ_sum[j] = KQ_sum[j]*KQ_max_scale + KQ_reg[j];
             KQ[j*nthreads + tid] = KQ_reg[j];
 
-#ifdef FAST_FP16_AVAILABLE
+#ifdef V_DOT2_F32_F16_AVAILABLE
             const half2 KQ_max_scale_h2 = make_half2(KQ_max_scale, KQ_max_scale);
 #pragma unroll
             for (int i_VKQ_0 = 0; i_VKQ_0 < D/2; i_VKQ_0 += nthreads_V) {
@@ -303,7 +303,7 @@ static __global__ void flash_attn_ext_vec(
                 VKQ[j][i_VKQ_0/nthreads_V].x *= KQ_max_scale;
                 VKQ[j][i_VKQ_0/nthreads_V].y *= KQ_max_scale;
             }
-#endif // FAST_FP16_AVAILABLE
+#endif // V_DOT2_F32_F16_AVAILABLE
         }
 
 #ifndef GGML_USE_HIP
@@ -314,7 +314,7 @@ static __global__ void flash_attn_ext_vec(
         for (int k0 = 0; k0 < WARP_SIZE; k0 += V_cols_per_iter) {
             const int k = threadIdx.y*WARP_SIZE + k0 + (nthreads_V == WARP_SIZE ? 0 : threadIdx.x / nthreads_V);
 
-#ifdef FAST_FP16_AVAILABLE
+#ifdef V_DOT2_F32_F16_AVAILABLE
             half2 KQ_k[ncols];
 #pragma unroll
             for (int j = 0; j < ncols; ++j) {
@@ -353,7 +353,7 @@ static __global__ void flash_attn_ext_vec(
                     }
                 }
             }
-#endif // FAST_FP16_AVAILABLE
+#endif // V_DOT2_F32_F16_AVAILABLE
         }
     }
 
@@ -374,7 +374,7 @@ static __global__ void flash_attn_ext_vec(
 
             KQ_sum[j] = KQ_sum[j]*KQ_max_scale + (threadIdx.x == 0 ? expf(sink - KQ_max[j]) : 0.0f);
 
-#ifdef FAST_FP16_AVAILABLE
+#ifdef V_DOT2_F32_F16_AVAILABLE
             const half2 KQ_max_scale_h2 = make_half2(KQ_max_scale, KQ_max_scale);
 #pragma unroll
             for (int i_VKQ_0 = 0; i_VKQ_0 < D/2; i_VKQ_0 += nthreads_V) {
@@ -386,7 +386,7 @@ static __global__ void flash_attn_ext_vec(
                 VKQ[j][i_VKQ_0/nthreads_V].x *= KQ_max_scale;
                 VKQ[j][i_VKQ_0/nthreads_V].y *= KQ_max_scale;
             }
-#endif // FAST_FP16_AVAILABLE
+#endif // V_DOT2_F32_F16_AVAILABLE
         }
     }
 
@@ -421,7 +421,7 @@ static __global__ void flash_attn_ext_vec(
         const float kqmax_scale = expf(KQ_max[j_VKQ] - kqmax_new);
         KQ_max[j_VKQ] = kqmax_new;
 
-#ifdef FAST_FP16_AVAILABLE
+#ifdef V_DOT2_F32_F16_AVAILABLE
         half2 * VKQ_tmp = (half2 *) KQ + threadIdx.y*(V_cols_per_iter*D/2)
             + (nthreads_V == WARP_SIZE ? 0 : threadIdx.x / nthreads_V)*(D/2);
 
@@ -452,7 +452,7 @@ static __global__ void flash_attn_ext_vec(
             ggml_cuda_memcpy_1<V_rows_per_thread/2*sizeof(float)>(VKQ_tmp + i_VKQ,                       &VKQ[j_VKQ][i_VKQ_0/nthreads_V]);
             ggml_cuda_memcpy_1<V_rows_per_thread/2*sizeof(float)>(VKQ_tmp + i_VKQ + V_rows_per_thread/4, &VKQ[j_VKQ][i_VKQ_0/nthreads_V + V_rows_per_thread/4]);
         }
-#endif // FAST_FP16_AVAILABLE
+#endif // V_DOT2_F32_F16_AVAILABLE
 
         KQ_sum[j_VKQ] *= kqmax_scale;
         KQ_sum[j_VKQ] = warp_reduce_sum(KQ_sum[j_VKQ]);