[llvm] [AMDGPU][True16] Support V_CEIL_F16. (PR #73108)

[Joe Nash via llvm-commits]

================
@@ -2303,7 +2303,17 @@ bool AMDGPUOperand::isInlineValue() const {
 // AsmParser
 //===----------------------------------------------------------------------===//
 
-static int getRegClass(RegisterKind Is, unsigned RegWidth) {
+static int getRegClass(RegisterKind Is, unsigned RegWidth,
+                       RegisterSuffix Suffix) {
+  if (Suffix != RegisterSuffix::None) {
+    if (Is == IS_VGPR && RegWidth == 16) {
+      if (Suffix == RegisterSuffix::Lo)
+        return AMDGPU::VGPR_LO16RegClassID;
----------------
Sisyph wrote:

VGPR_LO16 and VGPR_HI16  are scarcely used and can be replaced with VGPR_16 at some point in the True16 support timeline. So is it necessary to use them here for the transition ? If so, please put a comment to replace the register class with VGPR_16RegClassID and remove the Suffix parameter. If not, please switch it to VGPR_16RegClassID now 

https://github.com/llvm/llvm-project/pull/73108