RFR: 8325821: [REDO] use "dmb.ishst+dmb.ishld" for release barrier [v8]

[Tobias Hartmann]

On Wed, 5 Jun 2024 03:52:24 GMT, kuaiwei <duke at openjdk.org> wrote:

>> he origin patch for https://bugs.openjdk.org/browse/JDK-8324186 has 2 issues:
>> 1 It show regression in some platform, like Apple silicon in mac os
>> 2 Can not handle instruction sequence like "dmb.ishld; dmb.ishst; dmb.ishld; dmb.ishld"
>> 
>> It can be fixed by:
>> 1 Enable AlwaysMergeDMB by default, only disable it in architecture we can see performance improvement (N1 or N2)
>> 2 Check the special pattern and merge the subsequent dmb.
>> 
>> It also fix a bug when code buffer is expanding, st/ld/dmb can not be merged. I added unit tests for these.
>> 
>> This patch still has a unhandled case. Insts like "dmb.ishld; dmb.ishst; dmb.ish", it will merge the last 2 instructions and can not merge all three. Because when emitting dmb.ish, if merge all previous dmbs, the code buffer will shrink the size. I think it may break some resumption and think it's not a common pattern.
>> 
>> In previous PR https://github.com/openjdk/jdk/pull/18467 , I tried an implementation to use state machine for merging. But it looks risky to pending instruction during emitting.
>
> kuaiwei has updated the pull request with a new target base due to a merge or a rebase. The pull request now contains eight commits:
> 
>  - Merge master
>  - Use constexpr for test encoding
>  - Add comment in aarch64.ad
>  - Remove tailing white space
>  - Refine merge dmb test cases
>  - Add more unit tests
>  - Make MacroAssembler::merge more clear
>  - 8325821: [REDO] use "dmb.ishst+dmb.ishld" for release barrier

Sorry for the delay, I had to re-run a subset of the benchmarks due to high variance. All green now.

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

PR Comment: https://git.openjdk.org/jdk/pull/19278#issuecomment-2157834049