Fix lit tests on Windows by using env command (#5203)

Signed-off-by: Anatoly Myachev <[email protected]>

Author: anmyachev

test/TritonIntelGPU/blockptr_load.mlir

@@ -1,5 +1,5 @@
-// RUN: TRITON_INTEL_ONE_MATRIX_PER_LOAD_BT=0 triton-opt %s -split-input-file --intel-allocate-shared-memory --convert-triton-intel-gpu-to-llvm | FileCheck %s --implicit-check-not=llvm.inline_asm --check-prefixes=CHECK,LARGE-BLOCK-SIZE-TRANS-B
-// RUN: TRITON_INTEL_ONE_MATRIX_PER_LOAD_BT=1 triton-opt %s -split-input-file --intel-allocate-shared-memory --convert-triton-intel-gpu-to-llvm | FileCheck %s --implicit-check-not=llvm.inline_asm --check-prefixes=CHECK,SMALL-BLOCK-SIZE-TRANS-B
+// RUN: env TRITON_INTEL_ONE_MATRIX_PER_LOAD_BT=0 triton-opt %s -split-input-file --intel-allocate-shared-memory --convert-triton-intel-gpu-to-llvm | FileCheck %s --implicit-check-not=llvm.inline_asm --check-prefixes=CHECK,LARGE-BLOCK-SIZE-TRANS-B
+// RUN: env TRITON_INTEL_ONE_MATRIX_PER_LOAD_BT=1 triton-opt %s -split-input-file --intel-allocate-shared-memory --convert-triton-intel-gpu-to-llvm | FileCheck %s --implicit-check-not=llvm.inline_asm --check-prefixes=CHECK,SMALL-BLOCK-SIZE-TRANS-B
 
 #blocked = #ttg.blocked<{sizePerThread = [1, 1], threadsPerWarp = [1, 16], warpsPerCTA = [2, 4], order = [1, 0]}>
 #dpas = #ttig.dpas<{repeatCount = 8, systolicDepth = 8, executionSize = 16, opsPerChan = 2, threadsPerWarp = 16, warpsPerCTA = [4, 2], repCluster = [1, 1], A = [8, 16], B = [16, 16], C = [8, 16]}>

test/TritonIntelGPU/subgroup-2d-block-io.mlir

@@ -1,5 +1,5 @@
-// RUN: TRITON_INTEL_ONE_MATRIX_PER_LOAD_BT=0 triton-opt %s -split-input-file --allocate-shared-memory --convert-triton-intel-gpu-to-llvm | FileCheck %s --check-prefixes=STD-CHECK,CHECK
-// RUN: TRITON_INTEL_ONE_MATRIX_PER_LOAD_BT=1 triton-opt %s -split-input-file --allocate-shared-memory --convert-triton-intel-gpu-to-llvm | FileCheck %s --check-prefixes=ONE-MATRIX-CHECK
+// RUN: env TRITON_INTEL_ONE_MATRIX_PER_LOAD_BT=0 triton-opt %s -split-input-file --allocate-shared-memory --convert-triton-intel-gpu-to-llvm | FileCheck %s --check-prefixes=STD-CHECK,CHECK
+// RUN: env TRITON_INTEL_ONE_MATRIX_PER_LOAD_BT=1 triton-opt %s -split-input-file --allocate-shared-memory --convert-triton-intel-gpu-to-llvm | FileCheck %s --check-prefixes=ONE-MATRIX-CHECK
 
 // COM: A matrix, 16x16 block size, 1 warp w/ repCluster=1
 #dpas = #ttig.dpas<{repeatCount = 8, systolicDepth = 8, executionSize = 16, opsPerChan = 2, threadsPerWarp = 16, warpsPerCTA = [1, 1], repCluster = [1, 1]}>