RFR: 8356760: VectorAPI: Optimize VectorMask.fromLong for all-true/all-false cases [v3]
erifan
duke at openjdk.org
Thu Jul 3 07:10:22 UTC 2025
> If the input long value `l` of `VectorMask.fromLong(SPECIES, l)` would set or unset all lanes, `VectorMask.fromLong(SPECIES, l)` is equivalent to `maskAll(true)` or `maskAll(false)`. But the cost of `maskAll` is relative smaller than that of `fromLong`. This patch does the conversion for these cases if `l` is a compile time constant.
>
> And this conversion also enables further optimizations that recognize maskAll patterns, see [1].
>
> Some JTReg test cases are added to ensure the optimization is effective.
>
> I tried many different ways to write a JMH benchmark, but failed. Since the input of `VectorMask.fromLong(SPECIES, l)` needs to be a specific compile-time constant, the statement will be hoisted out of the loop. If we don't use a loop, the hotspot will become other instructions, and no obvious performance change was observed. However, combined with the optimization of [1], we can observe a performance improvement of about 7% on both aarch64 and x64.
>
> The patch was tested on both aarch64 and x64, all of tier1 tier2 and tier3 tests passed.
>
> [1] https://github.com/openjdk/jdk/pull/24674
erifan has updated the pull request incrementally with one additional commit since the last revision:
Simplify the test code
-------------
Changes:
- all: https://git.openjdk.org/jdk/pull/25793/files
- new: https://git.openjdk.org/jdk/pull/25793/files/791e0ab7..9f07d5c7
Webrevs:
- full: https://webrevs.openjdk.org/?repo=jdk&pr=25793&range=02
- incr: https://webrevs.openjdk.org/?repo=jdk&pr=25793&range=01-02
Stats: 233 lines in 3 files changed: 40 ins; 180 del; 13 mod
Patch: https://git.openjdk.org/jdk/pull/25793.diff
Fetch: git fetch https://git.openjdk.org/jdk.git pull/25793/head:pull/25793
PR: https://git.openjdk.org/jdk/pull/25793
More information about the hotspot-compiler-dev
mailing list