vector,decode: fix vector insts' src type* lsrc(2) is assigned to vd if the inst is vector instruction* set src type of no-used src of vector inst to SrcType.no
bump & rm useless code
vfcvt rtl: fixed cvt fu
backend: refactor FuType* use OHEnumeration to represent FuType
Backend: fix vtype's bug in DefaultConfig
VecDecoder: support vfredosum vfwredosum, add some vf oldVd srcType
backend: support vfredosum vfwredosum
Bump rocket-chip (#2353)
top-down: fix uncounted bubbles from decode and rename
backend,perf: enhance pmc implementation
backend: support unordered vfreduction
Backend, Fusion: enable fused_lui_load
Backend, Fusion: another implementation for instruction fusion case 'lui + addi(w)' without widening imm bits
rv64v: fix vmask instructions' tail elements*pass: vmand.mm, vmnand.mm, vmandn.mm, vmxor.mm, vmor.mm, vmnor.mm, vmorn.mm, vmxnor.mm
Backend, Fusion: support instruction fusion case 'lui + addiw'
Backend, Fusion: support instruction fusion case 'lui + addi'
exu: vfalu support vfmv_f_s vfmv_s_f
exu: vector float units(vfalu,vfma,vfdivsqrt) execute scalar float instructions
Rob: support ROB compression (#2192)For consecutive instructions that do not raise exceptions,they can share a same rob entry and reduce rob consumption.Only scalar instructions are supported no
Rob: support ROB compression (#2192)For consecutive instructions that do not raise exceptions,they can share a same rob entry and reduce rob consumption.Only scalar instructions are supported now.---------Co-authored-by: fdy <[email protected]>
show more ...
merge master into new-backendTodo: fix error
exu: vfdivsqrt support all instructions
Predecode: fix ebreak predecoded as jalr (#2186)
exu: vfma support all instructions
exu: vfalu support vfcmp vfmerge vfclass
VecDecoder: vfalu support vfsub vfmin vfmax vfwsub vfsgnj
12345678910>>...23