Date
1 - 1 of 1
Vector TG Minutes for 2020/7/24 meeting
Date: 2020/7/24
Task Group: Vector Extension
Chair: Krste Asanovic
Co-Chair: Roger Espasa
Number of Attendees: ~18
Current issues on github: https://github.com/riscv/riscv-v-spec
Issues discussed:
# Better figures in final documents
We review a proposal to add better figures to the ISA document outputs
(html and pdf). The group was in agreement this would be a great
improvement. A proposed reformatting will be pushed to the repo.
#365 should vsetvli x0, x0, imm set vill if vl changes?
There has been extensive discussion on the mailing list about this
issue, and also on behavior of vsetvl form when vtype.vill is set.
The consensus is that vill should be set if vl would change. Other
issues need more discussion.
#427 Make scalar input 2*SEW on integer widening operations
We discussed proposal to widen scalar input to widening operations to
2*SEW. For widening multiplies, this would double the size of
multiplier arrays required. The group discussed using a
microarchitectural check on scalar width to select a narrower
multiplier, but group consensus was that this information should be
supplied through a different opcode. Given that multiplies would
provide the larger benefit, and that adds would then have a
non-uniform format, the descision was made to stay with the PoR.
Task Group: Vector Extension
Chair: Krste Asanovic
Co-Chair: Roger Espasa
Number of Attendees: ~18
Current issues on github: https://github.com/riscv/riscv-v-spec
Issues discussed:
# Better figures in final documents
We review a proposal to add better figures to the ISA document outputs
(html and pdf). The group was in agreement this would be a great
improvement. A proposed reformatting will be pushed to the repo.
#365 should vsetvli x0, x0, imm set vill if vl changes?
There has been extensive discussion on the mailing list about this
issue, and also on behavior of vsetvl form when vtype.vill is set.
The consensus is that vill should be set if vl would change. Other
issues need more discussion.
#427 Make scalar input 2*SEW on integer widening operations
We discussed proposal to widen scalar input to widening operations to
2*SEW. For widening multiplies, this would double the size of
multiplier arrays required. The group discussed using a
microarchitectural check on scalar width to select a narrower
multiplier, but group consensus was that this information should be
supplied through a different opcode. Given that multiplies would
provide the larger benefit, and that adds would then have a
non-uniform format, the descision was made to stay with the PoR.