RFR: 8261671: X86 I2L conversion can be skipped for certain masked positive values [v3]

[Vladimir Kozlov]

On Wed, 17 Feb 2021 17:51:18 GMT, Marcus G K Williams <github.com+168222+mgkwill at openjdk.org> wrote:

>> 8261671: X86 I2L conversion can be skipped for certain masked positive values
>> 
>> For the following expression:
>>  (long)(value & mask)
>> Where value is of int type and mask is constant (power of two – 1), we can directly generate bzhi instruction to zero the upper bits instead of doing an andl, followed by movslq
>> 
>> Before:
>> Benchmark Mode Cnt Score Error Units
>> SkipIntToLongCast.skipMaskedSmallPositiveCast avgt 15 10.679 ± 1.496 ns/op
>> 
>> 
>> After:
>> Benchmark Mode Cnt Score Error Units
>> SkipIntToLongCast.skipMaskedSmallPositiveCast avgt 15 7.870 ± 0.067 ns/op
>> 
>> Signed-off-by: Marcus G K Williams <marcus.williams at intel.com>
>
> Marcus G K Williams has updated the pull request incrementally with one additional commit since the last revision:
> 
>   Revert to predicate supports_bmi2
>   
>   Signed-off-by: Marcus G K Williams <marcus.williams at intel.com>

Please, add a test which verifies result of generated code for different ranges of mask (especially corner cases). See `compiler//codegen/BMI1.java`
There are also example of verification of generated assembler in `compiler//intrinsics/bmi/`

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

Changes requested by kvn (Reviewer).

PR: https://git.openjdk.java.net/jdk/pull/2590