From 3c9496f1467af5b8c7e2c3236dd04526d87d719c Mon Sep 17 00:00:00 2001
From: khluu <khluu000@gmail.com>
Date: Tue, 24 Feb 2026 20:28:45 -0800
Subject: [PATCH] Revert "[Bugfix][ROCm][GPT-OSS] Use old triton_kernels
 implementation on ROCm if the new API is not available (#34153)"

This reverts commit 55a1baebc5c00dfafe5a2ae7b010bb0ebc959487.
---
 .../fused_moe/gpt_oss_triton_kernels_moe.py   | 61 +++----------------
 1 file changed, 9 insertions(+), 52 deletions(-)

diff --git a/vllm/model_executor/layers/fused_moe/gpt_oss_triton_kernels_moe.py b/vllm/model_executor/layers/fused_moe/gpt_oss_triton_kernels_moe.py
index eafdf97a9..3801814d9 100644
--- a/vllm/model_executor/layers/fused_moe/gpt_oss_triton_kernels_moe.py
+++ b/vllm/model_executor/layers/fused_moe/gpt_oss_triton_kernels_moe.py
@@ -19,14 +19,11 @@ from vllm.model_executor.layers.fused_moe.utils import _resize_cache
 from vllm.model_executor.layers.quantization.utils.quant_utils import (
     QuantKey,
 )
-from vllm.platforms import current_platform
 from vllm.triton_utils import tl, triton
 from vllm.utils.import_utils import has_triton_kernels
 
 logger = init_logger(__name__)
 
-use_legacy_triton_kernels = False
-
 if has_triton_kernels():
     try:
         import triton_kernels.swiglu
@@ -41,20 +38,10 @@ if has_triton_kernels():
         from triton_kernels.tensor import (
             BIT,
             Bitmatrix,
+            SparseMatrix,
+            make_ragged_tensor_metadata,
         )
         from triton_kernels.topk import topk
-
-        try:
-            from triton_kernels.tensor import (
-                SparseMatrix,
-                make_ragged_tensor_metadata,
-            )
-        except ImportError:
-            if current_platform.is_rocm():
-                logger.warning_once("Using legacy triton_kernels on ROCm")
-                use_legacy_triton_kernels = True
-            else:
-                raise
     except (AttributeError, ImportError) as e:
         logger.error(
             "Failed to import Triton kernels. Please make sure your triton "
@@ -114,12 +101,6 @@ def legacy_routing_from_bitmatrix(
     Replacement for the removed triton_kernels.routing.routing_from_bitmatrix.
     Creates routing data from a bitmatrix representation.
     """
-    if use_legacy_triton_kernels:
-        from triton_kernels.routing import routing_from_bitmatrix
-
-        return routing_from_bitmatrix(
-            bitmatrix, expt_scal, expt_indx, n_expts_tot, n_expts_act
-        )
     sparse_logits = SparseMatrix(indx=expt_indx, vals=expt_scal, mask=bitmatrix)
     dispatch_indx = sparse_logits.mask_metadata.row_sorted_indx
     combine_indx = sparse_logits.mask_metadata.col_sorted_indx
@@ -149,10 +130,6 @@ def legacy_routing(
     Replacement for the removed triton_kernels.routing.routing function.
     Computes routing data from gating logits.
     """
-    if use_legacy_triton_kernels:
-        from triton_kernels.routing import routing
-
-        return routing(logits, n_expts_act, sm_first=sm_first)
     if sm_first:
         logits = torch.softmax(logits, dim=-1)
     sparse_logits = topk(logits, n_expts_act, apply_softmax=not sm_first)
@@ -254,22 +231,11 @@ def triton_kernel_fused_experts(
     )
     output_tensor = _resize_cache(output_tensor, (batch_dim, M, K))
 
-    act = (
-        FusedActivation(
-            FnSpecs(
-                "swiglu",
-                triton_kernels.swiglu.swiglu_fn,
-                ("alpha", "limit"),
-                reduction_n=2,
-            ),
-            (swiglu_alpha, swiglu_limit),
-        )
-        if not use_legacy_triton_kernels
-        else FusedActivation(
-            FnSpecs("swiglu", triton_kernels.swiglu.swiglu_fn, ("alpha", "limit")),
-            (swiglu_alpha, swiglu_limit),
-            2,
-        )
+    act = FusedActivation(
+        FnSpecs(
+            "swiglu", triton_kernels.swiglu.swiglu_fn, ("alpha", "limit"), reduction_n=2
+        ),
+        (swiglu_alpha, swiglu_limit),
     )
     gammas = routing_data.gate_scal if routing_data else None
 
@@ -330,17 +296,8 @@ def make_routing_data(
 
     bitmatrix_shape = [n_rows, bm_cols * 32]
     bitmatrix_shape_max = [n_rows, None]
-    bitmatrix = (
-        Bitmatrix(
-            bitmatrix, dtype=BIT, shape=bitmatrix_shape, shape_max=bitmatrix_shape_max
-        )
-        if not use_legacy_triton_kernels
-        else Bitmatrix(
-            bitmatrix,
-            shape=bitmatrix_shape,
-            shape_max=bitmatrix_shape_max,
-            scratchpad=None,
-        )
+    bitmatrix = Bitmatrix(
+        bitmatrix, dtype=BIT, shape=bitmatrix_shape, shape_max=bitmatrix_shape_max
     )
 
     # matmul_ogs expects invalid topk_weights to be -1s