On Fri, 23 Aug 2024 05:46:29 GMT, Jatin Bhateja <[email protected]> wrote:
>> Hi All,
>>
>> As per the discussion on panama-dev mailing list[1], patch adds the support
>> for following new two vector permutation APIs.
>>
>>
>> Declaration:-
>> Vector<E>.selectFrom(Vector<E> v1, Vector<E> v2)
>>
>>
>> Semantics:-
>> Using index values stored in the lanes of "this" vector, assemble the
>> values stored in first (v1) and second (v2) vector arguments. Thus, first
>> and second vector serves as a table, whose elements are selected based on
>> index value vector. API is applicable to all integral and floating-point
>> types. The result of this operation is semantically equivalent to
>> expression v1.rearrange(this.toShuffle(), v2). Values held in index vector
>> lanes must lie within valid two vector index range [0, 2*VLEN) else an
>> IndexOutOfBoundException is thrown.
>>
>> Summary of changes:
>> - Java side implementation of new selectFrom API.
>> - C2 compiler IR and inline expander changes.
>> - In absence of direct two vector permutation instruction in target ISA, a
>> lowering transformation dismantles new IR into constituent IR supported by
>> target platforms.
>> - Optimized x86 backend implementation for AVX512 and legacy target.
>> - Function tests covering new API.
>>
>> JMH micro included with this patch shows around 10-15x gain over existing
>> rearrange API :-
>> Test System: Intel(R) Xeon(R) Platinum 8480+ [ Sapphire Rapids Server]
>>
>>
>> Benchmark (size) Mode Cnt
>> Score Error Units
>> SelectFromBenchmark.rearrangeFromByteVector 1024 thrpt 2 2041.762
>> ops/ms
>> SelectFromBenchmark.rearrangeFromByteVector 2048 thrpt 2 1028.550
>> ops/ms
>> SelectFromBenchmark.rearrangeFromIntVector 1024 thrpt 2 962.605
>> ops/ms
>> SelectFromBenchmark.rearrangeFromIntVector 2048 thrpt 2 479.004
>> ops/ms
>> SelectFromBenchmark.rearrangeFromLongVector 1024 thrpt 2 359.758
>> ops/ms
>> SelectFromBenchmark.rearrangeFromLongVector 2048 thrpt 2 178.192
>> ops/ms
>> SelectFromBenchmark.rearrangeFromShortVector 1024 thrpt 2 1463.459
>> ops/ms
>> SelectFromBenchmark.rearrangeFromShortVector 2048 thrpt 2 727.556
>> ops/ms
>> SelectFromBenchmark.selectFromByteVector 1024 thrpt 2 33254.830
>> ops/ms
>> SelectFromBenchmark.selectFromByteVector 2048 thrpt 2 17313.174
>> ops/ms
>> SelectFromBenchmark.selectFromIntVector 1024 thrpt 2 10756.804
>> ops/ms
>> S...
>
> Jatin Bhateja has updated the pull request incrementally with one additional
> commit since the last revision:
>
> Defaulting to index wrapping scheme.
Hi @rose00 , @PaulSandoz , @sviswa7,
Latest patch removed explicit wrap argument passed to selectFrom API, instead
uses wrapping scheme as a mitigation strategy to handle OOB partially wrapped
indexes.
Summarizing the new scheme for index wrapping:-
- Shuffle always holds indexes in valid vector index range or partially wraps
OOB indexes.
- Following are the shuffle creation intercepts
- VectorShuffle.fromArray - Partially wraps OOB indexes
- iotaShuffle - Accepts explicit wrap argument to
chooses b/w wrapping vs partial wrapping of OOB indexes.
- Vector.toShuffle - Partially wraps OOB indexes.
- Partial wrapping generate -ve indexes for OOB indices after wrapping them
into valid index range.
- Objective is to delegate mitigation strategy to subsequent APIs which
can either generate a IndexOutOfBounds exception or create valid index by
adding vector length.
- An important point to mention here is that partially wrapped indexing
schemes first wraps OOB index ( index < 0 OR index >= VECLEN) into valid
index range and then subtracts VECLEN from wrapped index to generate a -ve
number in [-VECLEN: -1] range.
- With new scheme we are choosing wrapping as a default mitigation strategy
hence only client which make effective use of a partially wrapped indexes is
two vector re-arrange, which uses it to compute the mask for blending two
permuted vectors.
- Two vector re-arrange and selectFrom API differ in terms of acceptable index
range, while former accepts shuffle indices in single vector index range
[0:VECLEN) latter operates on two vector index range [0:2*VECLEN).
Best Regards,
Jatin
-------------
PR Comment: https://git.openjdk.org/jdk/pull/20508#issuecomment-2306344606
