[Kernels] refactor: Use Int thread_idx etc. in layout/ (Part 48/N)

MODULAR_ORIG_COMMIT_REV_ID: 57dd1bdb19420adc1662f593b478e2d261319fd0

Author: JoeLoser

max/kernels/src/layout/layout_tensor.mojo

@@ -34,10 +34,10 @@ from std.builtin.dtype import _unsigned_integral_type_of
 from std.gpu.host import DeviceBuffer, HostBuffer, DeviceContext
 from std.gpu.host.nvidia.tma import TensorMapSwizzle
 from std.gpu import (
-    block_dim_int as block_dim,
+    block_dim,
     block_idx,
-    lane_id_int as lane_id,
-    thread_idx_int as thread_idx,
+    lane_id,
+    thread_idx,
 )
 from std.gpu.intrinsics import AMDBufferResource
 from std.gpu.memory import CacheEviction, CacheOperation, Fill, async_copy

max/kernels/src/layout/tensor_core.mojo

@@ -67,8 +67,8 @@ from std.sys.info import (
 
 from std.gpu import (
     WARP_SIZE,
-    lane_id_int as lane_id,
-    thread_idx_int as thread_idx,
+    lane_id,
+    thread_idx,
 )
 from std.gpu.intrinsics import lop, ds_read_tr16_b64
 from std.gpu.compute.mma import (