[llvm] [AMDGPU] Improved Lowering of abs(i16) and -abs(i16) (PR #165626)

[Patrick Simmons via llvm-commits]

================
@@ -0,0 +1,168 @@
+//===-- SISAbs16Fixup.cpp - Lower I1 Copies -----------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+//
+// This pass matches the pattern for 16-bit ABS instructions after they have
+// been lowered to for execution on the Scalar Unit.
+//
----------------
linuxrocks123 wrote:

Yeah I can merge it with the SII1Copies one (but first let's see if you can help me fix the earlier DAG-level attempt).

https://github.com/llvm/llvm-project/pull/165626