[llvm] [AMDGPU] Rework getDivNumBits API (PR #119768)

[Jay Foad via llvm-commits]

================
@@ -254,9 +254,8 @@ class AMDGPUCodeGenPrepareImpl
 
   bool divHasSpecialOptimization(BinaryOperator &I,
                                  Value *Num, Value *Den) const;
-  int getDivNumBits(BinaryOperator &I,
-                    Value *Num, Value *Den,
-                    unsigned AtLeast, bool Signed) const;
+  unsigned getDivNumBits(BinaryOperator &I, Value *Num, Value *Den,
+                         unsigned AtLeast, bool Signed) const;
----------------
jayfoad wrote:

Change argument name to MaxDivBits

https://github.com/llvm/llvm-project/pull/119768