src: cpu: aarch64: Enable JIT fp16 -> fp16 reorder

[karmeh01]

Description

This PR adds support for fp16 JIT reorder for data type f16->f16.

General

• Do all unit and benchdnn tests (make test and make test_benchdnn_*) pass locally for each commit?
• Have you formatted the code using clang-format?
• Have you submitted performance data that demonstrates performance improvements?

Performance improvements

Below are small test case logs that demonstrate performance numbers before and after.

./tests/benchdnn/benchdnn --reorder --sdt=f16 --ddt=f16 --stag=abc --dtag=aBc8b --mode=p 1x4096x4096

(new) jit:uni: total perf: min(ms):0.564941 avg(ms):1.01874
(old) simple:any: total perf: min(ms):25.6047 avg(ms):27.257

[aditew01]

Thanks for the PR. changes LGTM!
Are the relevant tests already present in benchdnn for this config: --sdt=f16 --ddt=f16 ?

[karmeh01]

No they are not. I will add them and update the PR

[karmeh01]

82 tests added to check the --sdt=f16 --ddt=f16 case
Passes all tests
Command: ./tests/benchdnn/benchdnn --reorder --batch=tests/benchdnn/inputs/reorder/test_reorder_float16
Result: tests:566 passed:446 skipped:120 mistrusted:0 unimplemented:0 invalid_arguments:0 failed:0 listed:0

[Ryo-not-rio]

lgtm