RFR: 8294262: AArch64: compiler/vectorapi/TestReverseByteTransforms.java test failed on SVE machine
Andrew Haley
aph at openjdk.org
Tue Sep 27 10:04:04 UTC 2022
On Tue, 27 Sep 2022 09:31:43 GMT, Eric Liu <eliu at openjdk.org> wrote:
> This test failed at cases test_reversebytes_short/int/long_transform2, which expected the ReversBytesV node, but nothing was finally found. On SVE system, we have a specific optimization, `ReverseBytesV (ReverseBytesV X MASK) MASK => X`, which eliminates both ReverseBytesV nodes. This optimization rule is specifically on hardware with native predicate support. See https://github.com/openjdk/jdk/pull/9623 for more details.
>
> As there is an SVE specific case TestReverseByteTransformsSVE.java, this patch simply marks TestReverseByteTransforms.java as non-SVE only.
>
> [TEST]
> jdk/incubator/vector, hotspot/compiler/vectorapi pass on SVE machine
test/hotspot/jtreg/compiler/vectorapi/TestReverseByteTransforms.java line 42:
> 40: * @summary Test various reverse bytes ideal transforms on X86(AVX2, AVX512) and AARCH64(NEON)
> 41: * @requires vm.compiler2.enabled
> 42: * @requires !(vm.cpu.features ~= ".*sve.*")
Please add a comment such as:
"On SVE system, we have a specific optimization, `ReverseBytesV (ReverseBytesV X MASK) MASK => X`, which eliminates both ReverseBytesV nodes. This optimization rule is specifically on hardware with native predicate support.
There is an SVE specific case TestReverseByteTransformsSVE.java, so this patch simply marks TestReverseByteTransforms.java as non-SVE only.
-------------
PR: https://git.openjdk.org/jdk/pull/10442
More information about the hotspot-compiler-dev
mailing list