[llvm] [InstCombine] Transform high latency, dependent FSQRT/FDIV into FMUL (PR #87474)

[Sushant Gokhale via llvm-commits]

================
@@ -0,0 +1,631 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 4
+; RUN: opt -S -passes='instcombine<no-verify-fixpoint>' < %s | FileCheck %s
+
+ at x = global double 0.000000e+00
+ at r1 = global double 0.000000e+00
+ at r2 = global double 0.000000e+00
+ at r3 = global double 0.000000e+00
+ at v = global [2 x double] zeroinitializer
+ at v1 = global [2 x double] zeroinitializer
+ at v2 = global [2 x double] zeroinitializer
+
+; div/mul/div1 in the same block.
+define void @bb_constraint_case1(double %a) {
+; CHECK-LABEL: define void @bb_constraint_case1(
+; CHECK-SAME: double [[A:%.*]]) {
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[SQRT1:%.*]] = call reassoc double @llvm.sqrt.f64(double [[A]])
+; CHECK-NEXT:    [[TMP0:%.*]] = fdiv reassoc double 1.000000e+00, [[A]]
+; CHECK-NEXT:    [[DIV:%.*]] = fmul reassoc ninf arcp double [[TMP0]], [[SQRT1]]
----------------
sushgokh wrote:

In the context of the `arcp` flag, just trying to clarify the last sentence of the definition from langref:
```
arcp
Allows division to be treated as a multiplication by a reciprocal. Specifically, this permits a / b to be considered equivalent to a * (1.0 / b) (which may subsequently be susceptible to code motion), and it also permits a / (b / c) to be considered equivalent to a * (c / b). Both of these rewrites can be applied in either direction: a * (c / b) can be rewritten into a / (b / c).
```
This means that `arcp` can be used to switch back and forth.

For the concerned example, if the transformed operation is
`[[DIV:%.*]] = fmul reassoc ninf arcp double [[TMP0]], [[SQRT1]]       --> div = (1/a) * sqrt(a) '

then I should have `arcp` flag on this operation because as per the last sentence of arcp definition, I should be able to rewrite it as:
`div = sqrt(a)/a`

If we remove the arcp flag, we wont be able to do so. Am I misinterpreting anything?

https://github.com/llvm/llvm-project/pull/87474