[llvm] [AMDGPU][True16][MC] update a few mc test for true16 (PR #135816)

[Joe Nash via llvm-commits]

================
@@ -0,0 +1,89 @@
+// NOTE: Assertions have been autogenerated by utils/update_mc_test_checks.py UTC_ARGS: --unique --version 5
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1100 -mattr=+wavefrontsize32,+real-true16 -show-encoding %s | FileCheck --check-prefixes=GFX11,W32 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1100 -mattr=+wavefrontsize64,+real-true16 -show-encoding %s | FileCheck --check-prefixes=GFX11,W64 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1100 -mattr=+wavefrontsize32,+real-true16 -filetype=null %s 2>&1 | FileCheck --check-prefixes=W32-ERR,GFX11-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1100 -mattr=+wavefrontsize64,+real-true16 -filetype=null %s 2>&1 | FileCheck --check-prefixes=W64-ERR,GFX11-ERR --implicit-check-not=error: %s
+
+//===----------------------------------------------------------------------===//
+// FIXME-TRUE16
+// It is unknown whether fp inline constants are allowed by hardware. Upstream
----------------
Sisyph wrote:

I don't know what the intended hardware design was, but whatever we are currently emitting seems to work. If you like the test here, that's fine with me, but it should at least stop talking about 'Upstream' 

https://github.com/llvm/llvm-project/pull/135816