[Mlir-commits] [mlir] [MLIR][Transform] Don't error when a structurally inlinable call exists (PR #195770)

[Mehdi Amini]

joker-eph wrote:

> I don’t think transform-library merging should impose an invariant on arbitrary non-transform dialect calls under the same symbol table. 

That is what I thought.

> So this PR fixes the immediate regression, but I agree the underlying check is broader than the invariant it is meant to protect. 

I think the fix in this PR is completely wrong and based on a lack of understanding of the invariants at play: adding more complexity without actually addressing correctly the problem. Please revert and fix appropriately.

https://github.com/llvm/llvm-project/pull/195770