[AMD] fix(rocm): support flydsl 0.3.0 in the FlyDSL fused norm kernel (#34536)

Co-authored-by: Bingxu Chen <195740905+bingxche@users.noreply.github.com>
Co-authored-by: thomawan <thomawan@amd.com>

python/sglang/kernels/ops/diffusion/norm/fused_residual_norm_flydsl.py

@@ -26,10 +26,17 @@ from flydsl._mlir.dialects import (
 )
 from flydsl._mlir.dialects import vector as _vector
 from flydsl.compiler.kernel_function import CompilationContext
-from flydsl.expr import arith, buffer_ops, const_expr, range_constexpr
+from flydsl.expr import arith, const_expr, range_constexpr
 from flydsl.expr.arith import ArithValue, CmpIPredicate
 from flydsl.expr.typing import Int32, T
 
+try:
+    from flydsl.expr import buffer_ops
+except ImportError:
+    # flydsl 0.3.0 removed flydsl.expr.buffer_ops and pushed the buffer-resource
+    # layer down to its consumers; AITER carries the same module, same API.
+    from aiter.ops.flydsl.kernels import buffer_ops
+
 WARP_SIZE = 64
 _VEC = 8
 _NUM_WAVES = 10