[CI] Fix gfx1100 pipeline test to not require rocm backend plugin (#24151)

Fixes arm64 clang CI regression introduced by #24073.

ci-extra: linux_arm64_clang
Assisted-by: Cursor (Claude)

Signed-off-by: Yu-Zhewen <zhewenyu@amd.com>

compiler/src/iree/compiler/Codegen/LLVMGPU/test/ROCDL/pipeline_tile_and_fuse_gfx1100.mlir

diff --git a/compiler/src/iree/compiler/Codegen/LLVMGPU/test/ROCDL/pipeline_tile_and_fuse_gfx1100.mlir b/compiler/src/iree/compiler/Codegen/LLVMGPU/test/ROCDL/pipeline_tile_and_fuse_gfx1100.mlir
index ab44b53..a323709 100644
--- a/compiler/src/iree/compiler/Codegen/LLVMGPU/test/ROCDL/pipeline_tile_and_fuse_gfx1100.mlir
+++ b/compiler/src/iree/compiler/Codegen/LLVMGPU/test/ROCDL/pipeline_tile_and_fuse_gfx1100.mlir
@@ -1,5 +1,5 @@
 // RUN: iree-opt --split-input-file --iree-gpu-test-target=gfx1100 \
-// RUN:   --pass-pipeline="builtin.module(hal.executable(hal.executable.variant(iree-hal-configure-target-executable-variants{target=rocm}, builtin.module(iree-llvmgpu-select-lowering-strategy, func.func(iree-llvmgpu-lower-executable-target, iree-codegen-gpu-check-resource-usage)))))" \
+// RUN:   --pass-pipeline="builtin.module(hal.executable(hal.executable.variant(builtin.module(iree-codegen-llvmgpu-configuration-pipeline, iree-codegen-llvmgpu-rocdl-lowering-pipeline{include-llvm-lowering=false}, func.func(iree-codegen-gpu-check-resource-usage)))))" \
 // RUN:   %s | FileCheck %s
 
 // Regression test for dynamic batch_matmul with WMMAR3 on gfx1100.