[Mlir-commits] [mlir] [mlir][math] Add vector support for math-to-apfloat (PR #172715)

[Maksim Levental]

makslevental wrote:

> My one concern is that you're storing ArrayRef, which will have lifetime issues - I'd copy those into a SmallVector instead

I don't think this is a big deal - the backing for the `ArrayRef` outlives all the Patterns because the patterns are added to a `PatternSet` "owned" by the pass (ie created in `runOnOperation`). If we expose the patterns via "populators" it'll still be the same thing (whichever pass adds the patterns will have its backing which outlives its own `PatternSet`).

https://github.com/llvm/llvm-project/pull/172715