RFR: 8326139: C2 SuperWord: split packs (match use/def packs, implemented, mutual independence) [v2]

[Emanuel Peter]

On Mon, 26 Feb 2024 18:12:27 GMT, Vladimir Kozlov <kvn at openjdk.org> wrote:

>> Emanuel Peter has updated the pull request with a new target base due to a merge or a rebase. The pull request now contains 27 commits:
>> 
>>  - manual merge
>>  - more fixes for TestSplitPacks.java
>>  - fix some IR rules in TestSplitPacks.java
>>  - fix MulAddS2I
>>  - Fix data dependency IR rules in LoopArrayIndexComputeTest.java
>>  - fix populate index cases and test7
>>  - test7 WIP
>>  - WIP fix reductions
>>  - Update test/hotspot/jtreg/compiler/loopopts/superword/TestSplitPacks.java
>>    
>>    fix bug number
>>  - cleanup
>>  - ... and 17 more: https://git.openjdk.org/jdk/compare/490825fb...c39ace04
>
> src/hotspot/share/opto/superword.cpp line 1008:
> 
>> 1006: }
>> 1007: 
>> 1008: // Do we have pattern n1 = (iv + c) and n2 = (iv + c + 1)?
> 
> Why  `+1` is special case?  Does SuperWord handle case when loop's increment is not `1`?

Yes, currently it only seems to handle positive `1`.

Relevant line in `SuperWord::vector_opd`, the only place where the `PopulateIndexNode` is created:
`Node* vn = new PopulateIndexNode(iv(), igvn().intcon(1), vt);`

The verification code is at the beginning of `SuperWord::is_vector_use`:
`offset_t->get_con() != (jint) i` -> we expect `populate_index->at(i) = (iv + i)`

That is a limitation we can surely remove in a future RFE.

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

PR Review Comment: https://git.openjdk.org/jdk/pull/17848#discussion_r1503814212