From db5f9339de4d4fad2d449afc48bc355b9f459e02 Mon Sep 17 00:00:00 2001
From: Dmitry Nikolaev <139769634+dnikolaev-amd@users.noreply.github.com>
Date: Mon, 21 Jul 2025 19:01:08 +0200
Subject: [PATCH] [rocm7.0_internal_testing] skip test_transformer_req_grad on
 Navi32/Navi4x (#2385)

Skip
`distributed/tensor/parallel/test_tp_examples.py::DistTensorParallelExampleTest::test_transformer_req_grad_*
` on Navi32/Navi4x

Fixes SWDEV-510742
---
 test/distributed/tensor/parallel/test_tp_examples.py | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/test/distributed/tensor/parallel/test_tp_examples.py b/test/distributed/tensor/parallel/test_tp_examples.py
index 9b412f88440f1..194735f912368 100644
--- a/test/distributed/tensor/parallel/test_tp_examples.py
+++ b/test/distributed/tensor/parallel/test_tp_examples.py
@@ -27,6 +27,8 @@
     RowwiseParallel,
 )
 from torch.distributed.tensor.parallel.input_reshard import input_reshard
+from torch.testing._internal.common_cuda import PLATFORM_SUPPORTS_FUSED_ATTENTION
+from torch.testing._internal.common_device_type import skipIf
 from torch.testing._internal.common_utils import (
     instantiate_parametrized_tests,
     parametrize,
@@ -412,6 +414,7 @@ def test_transformer_training(self, is_seq_parallel, dtype: torch.dtype):
         + f"{str(dtype).split('.')[-1]}_"
         + f"thaw_{'__'.join(sorted({n.rpartition('.')[0].replace('.', '_') for n in thaw})) if thaw else 'all'}",
     )
+    @skipIf(not PLATFORM_SUPPORTS_FUSED_ATTENTION, "Does not support fused scaled dot product attention")
     def test_transformer_req_grad(self, thaw_params, is_seq_parallel, dtype, exp_cnts):
         # Sample a subset of `requires_grad` patterns