RFR: 8356708: C2: loop strip mining expansion doesn't take sunk stores into account

[Roberto Castañeda Lozano]

On Tue, 10 Jun 2025 10:17:11 GMT, Roland Westrelin <roland at openjdk.org> wrote:

> `test1()` has a counted loop with a `Store` to `field`. That `Store`
> is sunk out of loop. When the `OuterStripMinedLoop` is expanded, only
> `Phi`s that exist at the inner loop are added to the outer
> loop. There's no `Phi` for the slice of the sunk `Store` (because
> there's no `Store` left in the inner loop) so no `Phi` is added for
> that slice to the outer loop. As a result, there's a missing anti
> dependency for `Load` of `field` that's before the loop and it can be
> scheduled inside the outer strip mined loop which is incorrect.
> 
> `test2()` is the same as `test1()` but with a chain of 2 `Store`s.
> 
> `test3()` is another variant where a `Store` is left in the inner loop
> after one is sunk out of it so the inner loop still has a `Phi`. As a
> result, the outer loop also gets a `Phi` but it's incorrectly wired as
> the sunk `Store` should be the input along the backedge but is
> not. That one doesn't cause any failure AFAICT.
> 
> The fix I propose is some extra logic at expansion of the
> `OuterStripMinedLoop` to handle these corner cases.

Looks good otherwise!

Another question: could `PhaseIdealLoop::try_move_store_before_loop` cause similar issues on strip-mined loops?

src/hotspot/share/opto/loopnode.cpp line 3026:

> 3024:         // Do we already have a memory Phi for that slice on the outer loop? If that is the case, that Phi was created
> 3025:         // by cloning an inner loop Phi. The inner loop Phi should have mem, the memory state of the first Store out of
> 3026:         // the inner loop as input on the backedge. So does the outer loop Phi given it's a clone.

Suggestion:

        // the inner loop, as input on the backedge. So does the outer loop Phi given it's a clone.

src/hotspot/share/opto/loopnode.cpp line 3043:

> 3041:           igvn->replace_input_of(first, MemNode::Memory, phi);
> 3042:         } else {
> 3043:           // Fix memory state along the backedge: it should be the last sunk Stores of the chain

Suggestion:

          // Fix memory state along the backedge: it should be the last sunk Store of the chain

test/hotspot/jtreg/compiler/loopstripmining/TestStoresSunkInOuterStripMinedLoop.java line 1:

> 1: /*

It would be good, for completeness, to add a "Couple stores sunk in outer loop, store in inner loop" test.

-------------

Changes requested by rcastanedalo (Reviewer).

PR Review: https://git.openjdk.org/jdk/pull/25717#pullrequestreview-2917408313
PR Review Comment: https://git.openjdk.org/jdk/pull/25717#discussion_r2140301451
PR Review Comment: https://git.openjdk.org/jdk/pull/25717#discussion_r2140302546
PR Review Comment: https://git.openjdk.org/jdk/pull/25717#discussion_r2140308882