submodule(yunsuan): bump yunsuan
area(VirtualLoadQueue): remove useless regs (#4061)* remove datavalid, addrvalid, veccommitted* add committed
submodule(yunsuan): bump yunsuan to fix VFMA/FMA area (#4069)This pr fix VFMA/FMA area with merging partial pipeline registers: VFMA save about 3000*2*2=12000 bit reg FMA save about 1300*3=
submodule(yunsuan): bump yunsuan to fix VFMA/FMA area (#4069)This pr fix VFMA/FMA area with merging partial pipeline registers: VFMA save about 3000*2*2=12000 bit reg FMA save about 1300*3=3900 bit reg
show more ...
Bump yunsuan
submodule(ready-to-run,yunsuan): bump ready-to-run and yunsuan
feat(zvfh,zfh): add F16 support
submodule(yunsuan): bump yunsuan* Including:* fix(zfa, fround): fix fround/froundnx inst implement* fix(zfa, fcvtmod.w.d): fix fcvtmod.w.d inst implement* fix(InToFp): fix the bit-width no
submodule(yunsuan): bump yunsuan* Including:* fix(zfa, fround): fix fround/froundnx inst implement* fix(zfa, fcvtmod.w.d): fix fcvtmod.w.d inst implement* fix(InToFp): fix the bit-width non-parameterization problem of fp_exp in the postnorm
submodule(YunSuan): bump yunsuan to fix neg of condition for f32toi16 (#3494)
fix(CVT64): fix unsynchronized selection signals for result and fflags (#3473)* Fix the bug where the selection signals for result and fflags were not synchronized and modified the form of the cod
fix(CVT64): fix unsynchronized selection signals for result and fflags (#3473)* Fix the bug where the selection signals for result and fflags were not synchronized and modified the form of the code to enhance readability. * Due to the fact that the zfa extension has not yet been merged into the master, in order to compile without errors, values have been assigned to the signals related to the zfa extension.
bump yunsuan* support zfa extension
bump yunsuan * FCVT:add conversion for FP16 and modified CVT64 module to parameterize it
Bump yunsuan: VIdiv fix state-machine, prioritize flush (#3391)
bump yunsuan: fix fflags update (#3384)
Bump yunsuan: VFALU fix vfredusum (#3297)
bump yunsuan (#3283)
bump yunsuan (#3256)vfcvt: fix condition of exp overflow for (u)i32Tof16 in CVT32
chore: bump chisel 6.5.0 (#3210)
Bump Yunsuan: fix vfredmin/max fflags result (#3185)when one src is active SNAN , the NV of fflags must be raised regardless of whether the other element is active.
vperm: bump yunsuan, fix vl compute for vslideup (#3150)
Vfalu: add fflagsEn default value and fix vfredmin/max (#3137)fix vfredmin/max when only one active src and the src is sNAN, the result should be sNAN not cNAN set default result of fflags mask ,
Vfalu: add fflagsEn default value and fix vfredmin/max (#3137)fix vfredmin/max when only one active src and the src is sNAN, the result should be sNAN not cNAN set default result of fflags mask ,when not fold set fflagsen of last uop to 0x00000001
Vfalu: fix ffagsEn logic for vfredunction (#3116)when fold , only some bits of fflags from vfalu0 are valid fflags vfredmin/max has to be masked
bump yunsuan (#3102)
VPU: new vcompress to fit v0&vl split; fix vfredsum/min/max (#3053)fix vfredsum/max/min: When the vector vfredusum/max/min consists entirely of inactive elements and vs1[0] is NaN, the result sho
VPU: new vcompress to fit v0&vl split; fix vfredsum/min/max (#3053)fix vfredsum/max/min: When the vector vfredusum/max/min consists entirely of inactive elements and vs1[0] is NaN, the result should be vs1[0] When both elements of vfredusum are inactive, the temporary result changes from positive zero to negative zero. nes vcompress to fit v0/vl split: The vcompress calculation combines the ones_sum result with vs1 using a temporary register to reduce one read operation. Additionally, other uops, except ones_sum, reduce the basemask calculation and the right shift basemask operation. fix vpermtest to fit new vcompress
bump yunsuan : Fpu,fsqrt: fix error when computing square roots for powers of 2 and sNan qNan error (#3047)
1234