ara
ara copied to clipboard
[hardware] Floating-Point classify, division, sqrt
This PR depends on https://github.com/pulp-platform/ara/pull/133 and, for ease of merging, on https://github.com/pulp-platform/ara/pull/131 as well
Add floating-point classify, division, sqrt instructions, delayed because of the FPU. I updated the CVA6 submodule, which now points to the modified FPU, able to classify as expected. Moreover, the bug related to the synchronization of the FPU lanes is now solved.
Changelog
Added
- Vector floating-point classify instruction (
vfclass
) - Vector floating-point divide instructions (
vfdiv
,vfrdiv
) - Vector floating-point square-root instruction (
vfsqrt
)
Checklist
- [x] Automated tests pass
- [x] Changelog updated
- [x] Code style guideline is observed
- [x] No frequency degradation
Ready for review when the automated tests pass! ;-)
Delaying the merge until the backend issues with the divider are solved.
Any chance we can replace FPNew's divider with the new divider everyone is talking about? :smirk:
Any chance we can replace FPNew's divider with the new divider everyone is talking about? smirk
Indeed, it's in the pipeline! ;-)
Any chance we can replace FPNew's divider with the new divider everyone is talking about? smirk
Unluckily, it is 32-bits only :-/
Backend problems are solved now, let's merge this as soon as the tests pass ;-)
@suehtamacv, we can merge this now, right?