RFR: 8342103: C2 compiler support for Float16 type and associated scalar operations [v13]

[Sandhya Viswanathan]

On Mon, 27 Jan 2025 08:35:44 GMT, Jatin Bhateja <jbhateja at openjdk.org> wrote:

>> Hi All,
>> 
>> This patch adds C2 compiler support for various Float16 operations added by [PR#22128](https://github.com/openjdk/jdk/pull/22128)
>> 
>> Following is the summary of changes included with this patch:-
>> 
>> 1. Detection of various Float16 operations through inline expansion or pattern folding idealizations.
>> 2. Float16 operations like add, sub, mul, div, max, and min are inferred through pattern folding idealization.
>> 3. Float16 SQRT and FMA operation are inferred through inline expansion and their corresponding entry points are defined in the newly added Float16Math class.
>>       -    These intrinsics receive unwrapped short arguments encoding IEEE 754 binary16 values.
>> 5. New specialized IR nodes for Float16 operations, associated idealizations, and constant folding routines.
>> 6. New Ideal type for constant and non-constant Float16 IR nodes. Please refer to [FAQs ](https://github.com/openjdk/jdk/pull/22754#issuecomment-2543982577)for more details.
>> 7. Since Float16 uses short as its storage type, hence raw FP16 values are always loaded into general purpose register, but FP16 ISA generally operates over floating point registers, thus the compiler injects reinterpretation IR before and after Float16 operation nodes to move short value to floating point register and vice versa.
>> 8. New idealization routines to optimize redundant reinterpretation chains. HF2S + S2HF = HF
>> 9. X86  backend implementation for all supported intrinsics.
>> 10. Functional and Performance validation tests.
>> 
>> Kindly review the patch and share your feedback.
>> 
>> Best Regards,
>> Jatin
>
> Jatin Bhateja has updated the pull request incrementally with one additional commit since the last revision:
> 
>   Copyright header fix

Some more minor comments.

src/hotspot/share/opto/addnode.cpp line 1546:

> 1544: 
> 1545:   // As per IEEE 754 specification, floating point comparison consider +ve and -ve
> 1546:   // zeros as equals. Thus, performing signed integral comparison for max value

Should be "min value detection".

src/hotspot/share/opto/addnode.cpp line 1624:

> 1622:   // As per IEEE 754 specification, floating point comparison consider +ve and -ve
> 1623:   // zeros as equals. Thus, performing signed integral comparison for min value
> 1624:   // detection.

Should be "max value detection".

src/hotspot/share/opto/divnode.cpp line 848:

> 846:   // If the dividend is a constant zero
> 847:   // Note: if t1 and t2 are zero then result is NaN (JVMS page 213)
> 848:   // Test TypeF::ZERO is not sufficient as it could be negative zero

Comment should be TypeH:ZERO is not sufficient

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

PR Review: https://git.openjdk.org/jdk/pull/22754#pullrequestreview-2576801437
PR Review Comment: https://git.openjdk.org/jdk/pull/22754#discussion_r1931347845
PR Review Comment: https://git.openjdk.org/jdk/pull/22754#discussion_r1931347430
PR Review Comment: https://git.openjdk.org/jdk/pull/22754#discussion_r1931376932