RFR: 8342103: C2 compiler support for Float16 type and associated scalar operations [v14]

[Sandhya Viswanathan]

On Tue, 28 Jan 2025 06:26:11 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:
> 
>   Updating typos in comments

Some more minor comments below. Rest of the PR looks good to me.

src/hotspot/share/opto/subnode.hpp line 549:

> 547:   SqrtHFNode(Compile* C, Node* c, Node* in1) : Node(c, in1) {
> 548:     init_flags(Flag_is_expensive);
> 549:     C->add_expensive_node(this);

Do we need to set SqrtHF as expensive node? It translates to a single instruction.

src/hotspot/share/opto/type.hpp line 2031:

> 2029: 
> 2030: inline const TypeH* Type::is_half_float_constant() const {
> 2031:   assert( _base == HalfFloatCon, "Not a Float" );

Should be "Not a HalfFloat" here.

test/hotspot/jtreg/compiler/c2/irTests/ConvF2HFIdealizationTests.java line 32:

> 30: /*
> 31:  * @test
> 32:  * @bug 8338061

This should now refer to bug 8342103.

test/hotspot/jtreg/compiler/c2/irTests/MulHFNodeIdealizationTests.java line 33:

> 31: /*
> 32:  * @test
> 33:  * @bug 8336406

This should now refer to bug 8342103.

test/hotspot/jtreg/compiler/c2/irTests/TestFloat16ScalarOperations.java line 27:

> 25: /**
> 26: * @test
> 27: * @bug 8308363 8336406

This should now refer to bug 8342103.

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

PR Review: https://git.openjdk.org/jdk/pull/22754#pullrequestreview-2579221309
PR Review Comment: https://git.openjdk.org/jdk/pull/22754#discussion_r1932741039
PR Review Comment: https://git.openjdk.org/jdk/pull/22754#discussion_r1932906990
PR Review Comment: https://git.openjdk.org/jdk/pull/22754#discussion_r1933045205
PR Review Comment: https://git.openjdk.org/jdk/pull/22754#discussion_r1933045649
PR Review Comment: https://git.openjdk.org/jdk/pull/22754#discussion_r1933045837