RFR: 8341976: C2: use_mem_state != load->find_exact_control(load->in(0)) assert failure [v2]

Roland Westrelin roland at openjdk.org
Tue Apr 1 07:32:47 UTC 2025


On Mon, 31 Mar 2025 12:26:42 GMT, Christian Hagedorn <chagedorn at openjdk.org> wrote:

>> Right. So maybe, we could treat that `Opaque` node the way we do for `OpaqueZeroTripGuard` and have it constant fold when the backedge is never taken.
>> 
>> So I should revert the change to the `IdealLoopTree::dump_head()` and the test run with `TraceLoopOpts`?
>
>> So maybe, we could treat that Opaque node the way we do for OpaqueZeroTripGuard and have it constant fold when the backedge is never taken.
> 
> Right, that sounds like a good solution.
> 
>> So I should revert the change to the IdealLoopTree::dump_head() and the test run with TraceLoopOpts?
> 
> Yes, that would be great. We can make a comment in [JDK-8297752](https://bugs.openjdk.org/browse/JDK-8297752) to add `-XX:+TraceLoopOpts` as additional run to this test when we fix it.

Done.

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

PR Review Comment: https://git.openjdk.org/jdk/pull/23465#discussion_r2022305281


More information about the hotspot-compiler-dev mailing list