rpms/python-torch
3
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Update hipblaslt patch

Browse source 
Upstream add some more things to change

Signed-off-by: Tom Rix <trix@redhat.com>
This commit is contained in:
Tom Rix 2024-03-03 07:31:08 -05:00
commit b2eaedd0c2
2 changed files with 65 additions and 20 deletions
Download patch file Download diff file Expand all files Collapse all files

63
next/0001-Optionally-use-hipblaslt.patch
View file

@ -1,4 +1,4 @@
From 186728c9f4de720547fa3d1b7951c7d5563ad3c7 Mon Sep 17 00:00:00 2001
From d77e05d90df006322cda021f1a8affdcc2c7eaef Mon Sep 17 00:00:00 2001
From: Tom Rix <trix@redhat.com>
Date: Fri, 23 Feb 2024 08:27:30 -0500
Subject: [PATCH] Optionally use hipblaslt
@ -10,21 +10,30 @@ Convert the checks for ROCM_VERSION >= 507000 to HIPBLASLT checks
Signed-off-by: Tom Rix <trix@redhat.com>
---
aten/src/ATen/cuda/CUDABlas.cpp | 5 +++--
aten/src/ATen/cuda/CUDABlas.cpp | 7 ++++---
aten/src/ATen/cuda/CUDABlas.h | 2 +-
aten/src/ATen/cuda/CUDAContextLight.h | 4 ++--
aten/src/ATen/cuda/CublasHandlePool.cpp | 4 ++--
aten/src/ATen/cuda/tunable/TunableGemm.h | 6 +++---
aten/src/ATen/native/cuda/Blas.cpp | 10 ++++++----
aten/src/ATen/native/cuda/Blas.cpp | 14 ++++++++------
cmake/Dependencies.cmake | 3 +++
cmake/public/LoadHIP.cmake | 4 ++--
8 files changed, 22 insertions(+), 16 deletions(-)
8 files changed, 25 insertions(+), 19 deletions(-)
diff --git a/aten/src/ATen/cuda/CUDABlas.cpp b/aten/src/ATen/cuda/CUDABlas.cpp
index 0a3de5f9d77..b67031e88a7 100644
index d534ec5a178..e815463f630 100644
--- a/aten/src/ATen/cuda/CUDABlas.cpp
+++ b/aten/src/ATen/cuda/CUDABlas.cpp
@@ -778,7 +778,7 @@ void gemm<at::BFloat16>(CUDABLAS_GEMM_ARGTYPES(at::BFloat16)) {
@@ -14,7 +14,7 @@
#include <c10/util/irange.h>
#ifdef USE_ROCM
-#if ROCM_VERSION >= 60000
+#ifdef HIPBLASLT
#include <hipblaslt/hipblaslt-ext.hpp>
#endif
// until hipblas has an API to accept flags, we must use rocblas here
@@ -781,7 +781,7 @@ void gemm<at::BFloat16>(CUDABLAS_GEMM_ARGTYPES(at::BFloat16)) {
}
}
@ -33,7 +42,7 @@ index 0a3de5f9d77..b67031e88a7 100644
#if defined(USE_ROCM) && ROCM_VERSION >= 50700 && ROCM_VERSION < 60000
// only for rocm 5.7 where we first supported hipblaslt, it was difficult
@@ -909,6 +909,7 @@ class CuBlasLtMatmulPreference : public CuBlasLtDescriptor<
@@ -912,6 +912,7 @@ class CuBlasLtMatmulPreference : public CuBlasLtDescriptor<
};
} // namespace
@ -41,7 +50,7 @@ index 0a3de5f9d77..b67031e88a7 100644
template <typename Dtype>
void gemm_and_bias(
bool transpose_mat1,
@@ -1121,7 +1122,7 @@ template void gemm_and_bias(
@@ -1124,7 +1125,7 @@ template void gemm_and_bias(
at::BFloat16* result_ptr,
int64_t result_ld,
GEMMAndBiasActivationEpilogue activation);
@ -86,10 +95,10 @@ index 4ec35f59a21..e28dc42034f 100644
#endif
diff --git a/aten/src/ATen/cuda/CublasHandlePool.cpp b/aten/src/ATen/cuda/CublasHandlePool.cpp
index d8ee09b1486..4510193a5cc 100644
index 6913d2cd95e..3d4276be372 100644
--- a/aten/src/ATen/cuda/CublasHandlePool.cpp
+++ b/aten/src/ATen/cuda/CublasHandlePool.cpp
@@ -28,7 +28,7 @@ namespace at::cuda {
@@ -29,7 +29,7 @@ namespace at::cuda {
namespace {
@ -98,7 +107,7 @@ index d8ee09b1486..4510193a5cc 100644
void createCublasLtHandle(cublasLtHandle_t *handle) {
TORCH_CUDABLAS_CHECK(cublasLtCreate(handle));
}
@@ -177,7 +177,7 @@ cublasHandle_t getCurrentCUDABlasHandle() {
@@ -190,7 +190,7 @@ cublasHandle_t getCurrentCUDABlasHandle() {
return handle;
}
@ -139,10 +148,10 @@ index 3ba0d761277..dde1870cfbf 100644
if (env == nullptr || strcmp(env, "1") == 0) {
// disallow tuning of hipblaslt with c10::complex
diff --git a/aten/src/ATen/native/cuda/Blas.cpp b/aten/src/ATen/native/cuda/Blas.cpp
index 2b7ff1b502b..9ae2ccbd16d 100644
index 29e5c5e3cf1..df56f3d7f1d 100644
--- a/aten/src/ATen/native/cuda/Blas.cpp
+++ b/aten/src/ATen/native/cuda/Blas.cpp
@@ -153,7 +153,7 @@ enum class Activation {
@@ -155,7 +155,7 @@ enum class Activation {
GELU,
};
@ -151,7 +160,7 @@ index 2b7ff1b502b..9ae2ccbd16d 100644
cuda::blas::GEMMAndBiasActivationEpilogue activation_to_gemm_and_blas_arg(Activation a) {
switch (a) {
case Activation::None:
@@ -191,6 +191,7 @@ static bool getDisableAddmmCudaLt() {
@@ -193,6 +193,7 @@ static bool getDisableAddmmCudaLt() {
#ifdef USE_ROCM
static bool isSupportedHipLtROCmArch(int index) {
@ -159,7 +168,7 @@ index 2b7ff1b502b..9ae2ccbd16d 100644
hipDeviceProp_t* prop = at::cuda::getDeviceProperties(index);
std::string device_arch = prop->gcnArchName;
static const std::vector<std::string> archs = {"gfx90a", "gfx940", "gfx941", "gfx942"};
@@ -201,6 +202,7 @@ static bool isSupportedHipLtROCmArch(int index) {
@@ -203,6 +204,7 @@ static bool isSupportedHipLtROCmArch(int index) {
}
}
TORCH_CHECK(false, "Attempting to use hipBLASLt on a unsupported architecture!");
@ -167,7 +176,7 @@ index 2b7ff1b502b..9ae2ccbd16d 100644
return false;
}
#endif
@@ -226,7 +228,7 @@ Tensor& addmm_out_cuda_impl(Tensor& result, const Tensor& self, const Tensor& ma
@@ -228,7 +230,7 @@ Tensor& addmm_out_cuda_impl(Tensor& result, const Tensor& self, const Tensor& ma
at::ScalarType scalar_type = self.scalar_type();
c10::MaybeOwned<Tensor> self_;
if (&result != &self) {
@ -176,7 +185,7 @@ index 2b7ff1b502b..9ae2ccbd16d 100644
// Strangely, if mat2 has only 1 row or column, we get
// CUBLAS_STATUS_INVALID_VALUE error from cublasLtMatmulAlgoGetHeuristic.
// self.dim() == 1 && result.dim() == 2 && self.sizes()[0] == mat2_sizes[1]
@@ -269,7 +271,7 @@ Tensor& addmm_out_cuda_impl(Tensor& result, const Tensor& self, const Tensor& ma
@@ -271,7 +273,7 @@ Tensor& addmm_out_cuda_impl(Tensor& result, const Tensor& self, const Tensor& ma
}
self__sizes = self_->sizes();
} else {
@ -185,7 +194,7 @@ index 2b7ff1b502b..9ae2ccbd16d 100644
useLtInterface = !disable_addmm_cuda_lt &&
result.dim() == 2 && result.is_contiguous() &&
isSupportedHipLtROCmArch(self.device().index()) &&
@@ -320,7 +322,7 @@ Tensor& addmm_out_cuda_impl(Tensor& result, const Tensor& self, const Tensor& ma
@@ -322,7 +324,7 @@ Tensor& addmm_out_cuda_impl(Tensor& result, const Tensor& self, const Tensor& ma
TORCH_INTERNAL_ASSERT_DEBUG_ONLY(!args.result->is_conj());
@ -194,6 +203,24 @@ index 2b7ff1b502b..9ae2ccbd16d 100644
if (useLtInterface) {
AT_DISPATCH_FLOATING_TYPES_AND2(
at::ScalarType::Half,
@@ -876,7 +878,7 @@ _scaled_mm_out_cuda(const Tensor& mat1, const Tensor& mat2,
at::native::resize_output(out, {mat1_sizes[0], mat2_sizes[1]});
at::native::resize_output(amax, {});
-#if !defined(USE_ROCM) && !defined(_MSC_VER) || (defined(USE_ROCM) && ROCM_VERSION >= 60000)
+#if !defined(USE_ROCM) && !defined(_MSC_VER) || (defined(USE_ROCM) && defined(HIPBLASLT))
cublasCommonArgs args(mat1, mat2, out);
const auto out_dtype_ = args.result->scalar_type();
TORCH_CHECK(args.transa == 't' && args.transb == 'n', "Only multiplication of row-major and column-major matrices is supported by cuBLASLt");
@@ -906,7 +908,7 @@ _scaled_mm_out_cuda(const Tensor& mat1, const Tensor& mat2,
TORCH_CHECK(false, "_scaled_mm_out_cuda is not compiled for this platform.");
#endif
-#if defined(USE_ROCM) && ROCM_VERSION >= 60000
+#if defined(USE_ROCM) && defined(HIPBLASLT)
// rocm's hipblaslt does not yet support amax, so calculate separately
auto out_float32 = out.to(kFloat);
out_float32.abs_();
diff --git a/cmake/Dependencies.cmake b/cmake/Dependencies.cmake
index b7ffbeb07dc..2b6c3678984 100644
--- a/cmake/Dependencies.cmake