[jdk11u-dev] RFR: 8297730: C2: Arraycopy intrinsic throws incorrect exception

[Martin Doerr]

On Fri, 2 Jun 2023 15:09:12 GMT, Goetz Lindenmaier <goetz at openjdk.org> wrote:

> I backport this for parity with 11.0.20-oracle.
> 
> I had to resolve library_call.cpp:
> 
> 11 has an additional check:
>       if (slow_region != NULL && slow_region->find_edge(not_ctl) >= 1) {
>         ctl = iff->in(0);       // This test feeds the known slow_region.
>         continue;
>       }
> which was removed in 17 by "8263615: Cleanup tightly_coupled_allocation"
> I reworked the code to keep this.
> 
> The changes to library_call.hpp need to be done in library_call.cpp
> as the header is only created later.
> 
> I had to replace nullptr.

Looks correct, but I'd prefer to remove the dead code snippet and maybe assert `slow_region == NULL`. (Or backport JDK-8263615 which would make it almost clean.)

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

Marked as reviewed by mdoerr (Reviewer).

PR Review: https://git.openjdk.org/jdk11u-dev/pull/1926#pullrequestreview-1462494176