[AMD] hipify torchaudio

Differential Revision: D64184710 Pull Request resolved: #3840
pytorch · Oct 16, 2024 · b4a286a · b4a286a
1 parent 3f05699
commit b4a286a
Show file tree

Hide file tree

Showing 2 changed files with 24 additions and 0 deletions.
diff --git a/src/libtorchaudio/rnnt/gpu/gpu_kernel_utils.cuh b/src/libtorchaudio/rnnt/gpu/gpu_kernel_utils.cuh
@@ -39,7 +39,11 @@ __global__ void ReduceMax2D(
 
   CAST_DTYPE shf;
   for (int stride = (WARP_SIZE >> 1); stride > 0; stride >>= 1) {
+#ifndef USE_ROCM
     shf = __shfl_down_sync(0xFFFFFFFF, val, stride);
+#else
+    shf = __shfl_down(val, stride);
+#endif
     if (threadIdx.x < stride && threadIdx.x + stride < dim) {
       if (shf > val) {
         val = shf;
@@ -81,7 +85,11 @@ __global__ void ReduceLogSumExpGivenMax2D(
 
   CAST_DTYPE shf;
   for (int stride = (WARP_SIZE >> 1); stride > 0; stride >>= 1) {
+#ifndef USE_ROCM
     shf = __shfl_down_sync(0xFFFFFFFF, val, stride);
+#else
+    shf = __shfl_down(val, stride);
+#endif
     if (threadIdx.x < stride && threadIdx.x + stride < dim) {
       val = val + shf;
     }

diff --git a/src/libtorchaudio/rnnt/gpu/gpu_kernels.cuh b/src/libtorchaudio/rnnt/gpu/gpu_kernels.cuh
@@ -126,7 +126,11 @@ __device__ void ComputeAlphas(
 
 #pragma unroll
     for (int i = 1; i < warpSize; i <<= 1) {
+#ifndef USE_ROCM
       val = __shfl_up_sync(0xffffffff, skip_prob, i);
+#else
+      val = __shfl_up(skip_prob, i);
+#endif
       if (i <= threadIdx.x) {
         skip_prob = skip_prob + val;
       }
@@ -150,7 +154,11 @@ __device__ void ComputeAlphas(
     CAST_DTYPE out = val;
 
     for (int i = 1; i < warpSize; ++i) {
+#ifndef USE_ROCM
       val = __shfl_up_sync(0xffffffff, val, 1);
+#else
+      val = __shfl_up(val, 1);
+#endif
       if (i == threadIdx.x) {
         val = math::lse(val + skip_prob, emit);
         out = val;
@@ -225,7 +233,11 @@ __device__ void ComputeBetasCosts(
 
 #pragma unroll
     for (int i = 1; i < warpSize; i <<= 1) {
+#ifndef USE_ROCM
       val = __shfl_up_sync(0xffffffff, skip_prob, i);
+#else
+      val = __shfl_up(skip_prob, i);
+#endif
       if (i <= threadIdx.x) {
         skip_prob = skip_prob + val;
       }
@@ -248,7 +260,11 @@ __device__ void ComputeBetasCosts(
     CAST_DTYPE out = val;
 
     for (int i = 1; i < warpSize; ++i) {
+#ifndef USE_ROCM
       val = __shfl_up_sync(0xffffffff, val, 1);
+#else
+      val = __shfl_up(val, 1);
+#endif
       if (i == threadIdx.x) {
         val = math::lse(val + skip_prob, emit);
         out = val;