[llvm] [AMDGPU] Rework pseudo-scalar transcendental inline constant restriction handling (PR #208459)

[Frederik Harwath via llvm-commits]

================
@@ -1,121 +1,329 @@
-# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py UTC_ARGS: --version 5
-# RUN: llc -mtriple=amdgpu11.70-amd-amdpal -run-pass=si-fold-operands -verify-machineinstrs -o - %s | FileCheck --check-prefixes=GCN %s
-# RUN: llc -mtriple=amdgpu12.00-amd-amdpal -run-pass=si-fold-operands -verify-machineinstrs -o - %s | FileCheck --check-prefixes=GCN %s
+# NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 6
+# RUN: llc -mtriple=amdgpu12.00-amd-amdpal -x mir --show-mc-encoding < %s | FileCheck --check-prefixes=GFX12 %s
----------------
frederik-h wrote:

I reset the test to the baseline version and updated the checks and the comment. It should be enough to test the encoding in the MC tests. There are existing tests that show how the encoding has changed for all of the instructions covered by this test. Do you think the existing assembler tests should be more comprehensive?

https://github.com/llvm/llvm-project/pull/208459