[llvm] [AMDGPU] Be more careful about using expandDivRem24 (PR #201186)

[via llvm-commits]

================
@@ -1062,7 +1062,18 @@ Value *AMDGPUCodeGenPrepareImpl::expandDivRem24(IRBuilder<> &Builder,
                                                 Value *Den, bool IsDiv,
                                                 bool IsSigned) const {
   unsigned DivBits = getDivNumBits(I, Num, Den, 24, IsSigned);
-  if (DivBits > 24)
+
+  // v_rcp_f32(float(X)) can have an error of 1 ulp.
+  // This can cause expandDivRem24Impl to sometimes calculate Y/X incorrectly
+  // when abs(Y)>0x800000.
+  // For example,
+  // (0xbf2758/0xbf2759) erroneously produces 1 instead of 0.
+  // (0xe3170d/0x000c32) erroneously produces 4767 instead of 4766.
+  //
+  // Note that for DivBits==24 && IsSigned, Y is in the range
+  // [-0x800000:0x7FFFFF]. abs(Y) is at most
+  // 0x800000 so it cannot hit this issue.
----------------
LU-JOHN wrote:

Updated PR description.

https://github.com/llvm/llvm-project/pull/201186