TPU v6e TEC 指令索引

TPU v6e TEC 指令索引#

由 libtpu 0.0.49 的 descriptor、encoder、LLVM TPU printer 与 TEC emitter 生成:登记 500 种形式、965 个槽与形式组合、2336 条展开签名、1054 个槽与助记符组合。encoder 拒绝的 125 个槽与形式组合不登记。这里的覆盖是编码、解码与重汇编覆盖,没有在设备上执行。

如何阅读#

s[0..31]、v[0..63] 表示寄存器编号,/ 表示可选的拼写,iN/uN/f32 是数值类型,其余字符照写。操作数顺序与 LLVM TPU printer 相同,目的在前。数值类型只限定输入位宽,立即数槽和整个指令包的资源是否够用由汇编器检查。写法见格式参考。

每个形式另有具名字段写法:助记符是 descriptor 中的形式名,操作数是按字段顺序排列的具名参数,用于 printer 写法表达不了的编码。表中只对没有 printer 写法的槽与形式组合列出这种写法,涉及以下形式:atomic_tile_add_set_done、atomic_tile_add_set_done_inverted、call_relative、call_sreg、descriptor_based_dma、eup_push、general_dma、indirect_vreg_stream、max_index_scan_bf16、max_index_scan_u16、min_index_scan_bf16、pop_rcmf、pop_xrf_write_partial_0、pop_xrf_write_partial_1、pop_xrf_write_partial_2、pop_xrf_write_partial_3、pop_xrf_write_partial_4、push_scmf、read_sync_state_done、read_sync_state_value、segmented_min_index_scan_bf16、segmented_min_scan_bf16、single_strided_dma、tile_spmem_store_indexed_circular_buffer_return_value_add_bf16、tile_spmem_store_indexed_circular_buffer_return_value_add_f32、tile_spmem_store_indexed_circular_buffer_return_value_add_s16、tile_spmem_store_indexed_circular_buffer_return_value_add_s32、unpack_interleaved_compressed_lower_s8_to_bf16、unpack_interleaved_compressed_lower_s8_to_s16、unpack_interleaved_compressed_lower_u8_to_bf16、unpack_interleaved_compressed_lower_u8_to_u16、unpack_interleaved_compressed_upper_s8_to_bf16、unpack_interleaved_compressed_upper_s8_to_s16、unpack_interleaved_compressed_upper_u8_to_bf16、unpack_interleaved_compressed_upper_u8_to_u16、vector_add_high_f32_bf16、vector_add_low_f32_bf16、vector_clamp_asymmetric_bf16、vector_clamp_asymmetric_f32、vector_convert_bf16_to_s4、vector_convert_bf16_to_s8、vector_convert_bf16_to_u4、vector_convert_bf16_to_u8、vector_convert_f32_to_bf16、vector_convert_s4_to_bf16、vector_convert_s8_to_bf16、vector_convert_u4_to_bf16、vector_convert_u8_to_bf16、vector_gt_u16、vector_gte_u16、vector_lane_left_shift_insert_b16、vector_lt_u16、vector_lte_u16、vector_remap_bf16、vector_remap_f32、yieldable_sync_done、yieldable_sync_equal、yieldable_sync_equal_or_done、yieldable_sync_greater、yieldable_sync_greater_or_done、yieldable_sync_greater_or_equal、yieldable_sync_greater_or_equal_or_done、yieldable_sync_less、yieldable_sync_less_or_done、yieldable_sync_not_done、yieldable_sync_not_equal、yieldable_sync_not_equal_or_done。

按槽查找#

s0 · s1 · dma · misc · va0 · va1 · va2 · vld · vst · stream · vr · vx

s0#

助记符

按顺序排列的操作数类型

call_relative

dest=(u5)

call_sreg

dest=(u5), x=(u5)

pop_rcmf

dest=(u5)

por

p[0..15], p[0..31], p[0..31]
p[0..15], @first_id_occurrence, p[0..31]
p[0..15], p[0..31], @first_id_occurrence

sadd.ov.s32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

sadd.s32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

sand.u32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

sbr.abs

label / u20

sbr.abs.clr

u20

sbr.ind

s[0..31]

sbr.rel

label / i20

sc.u32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

scall.abs

lr, label / u20

sceil.f32

s[0..31], s[0..31]

sclz.u32

s[0..31], s[0..31]

scvt.f32.s32

s[0..31], s[0..31]
s[0..31], f32

scvt.s32.f32

s[0..31], s[0..31]
s[0..31], f32

sdelay

u11

sdiv.u32

drf, s[0..31], u32
drf, s[0..31], s[0..31]

sdivrem.u32

drf, s[0..31], u32
drf, s[0..31], s[0..31]

seq.f32

p[0..15], s[0..31], s[0..31]
p[0..15], s[0..31], f32

seq.s32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sfence

无操作数

sfence.scmf

无操作数

sfence.sel

u32
s[0..31]

sfence.stream.hbm

无操作数

sfence.stream.spmem

无操作数

sfloor.f32

s[0..31], s[0..31]

sge.f32

p[0..15], s[0..31], s[0..31]
p[0..15], s[0..31], f32

sge.s32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sge.u32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sgt.f32

p[0..15], s[0..31], s[0..31]
p[0..15], s[0..31], f32

sgt.s32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sgt.u32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

shalt

无操作数

simm.f32

s[0..31], f32

simm.s32

s[0..31], u32

sle.f32

p[0..15], s[0..31], s[0..31]
p[0..15], s[0..31], f32

sle.s32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sle.u32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

slt.f32

p[0..15], s[0..31], s[0..31]
p[0..15], s[0..31], f32

slt.s32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

slt.u32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

smax.f32

s[0..31], s[0..31], s[0..31]
s[0..31], s[0..31], f32

smax.u32

s[0..31], s[0..31], u32
s[0..31], s[0..31], s[0..31]

smin.f32

s[0..31], s[0..31], s[0..31]
s[0..31], s[0..31], f32

smin.u32

s[0..31], s[0..31], u32
s[0..31], s[0..31], s[0..31]

smov.u32

s[0..31], s[0..31]

smul.f32

s[0..31], s[0..31], s[0..31]
s[0..31], f32, s[0..31]

smul.u32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

smulhi.u32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

sne.f32

p[0..15], s[0..31], s[0..31]
p[0..15], s[0..31], f32

sne.s32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sor.u32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

spop

s[0..31], drf
s[0..31], v2sf

srdreg.ddr

s[0..31]

srdreg.fsr

s[0..31]

srdreg.gtchi

s[0..31]

srdreg.gtclo

s[0..31]

srdreg.lcchi

s[0..31]

srdreg.lcclo

s[0..31]

srdreg.scid

s[0..31]

srdreg.tag

s[0..31]

srdreg.tbm

s[0..31]

srdreg.tm

s[0..31]

srdreg.yieldreq

s[0..31]

srem.u32

drf, s[0..31], u32
drf, s[0..31], s[0..31]

ssetifvalue

ifvalue, u32
ifvalue, s[0..31]

ssetpdepth

pdepth, u32
pdepth, s[0..31]

ssettag

u32
s[0..31]

sshla.ov.s32

s[0..31], s[0..31], u32
s[0..31], s[0..31], s[0..31]

sshll.u32

s[0..31], s[0..31], u32
s[0..31], s[0..31], s[0..31]

sshra.s32

s[0..31], s[0..31], u32
s[0..31], s[0..31], s[0..31]

sshrl.u32

s[0..31], s[0..31], u32
s[0..31], s[0..31], s[0..31]

ssub.ov.s32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

ssub.s32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

stileid.u32

s[0..31]

sweird.f32

p[0..15], s[0..31]

sxor.u32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

s1#

助记符

按顺序排列的操作数类型

cbreg.add

[cbreg:u5], u32
[cbreg:u5], s[0..31]
[cbreg:$0x4], u32
[cbreg:$0x4], s[0..31]

descriptor_based_dma

y=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32)

pop_rcmf

dest=(u5)

por

p[0..15], p[0..31], p[0..31]
p[0..15], @first_id_occurrence, p[0..31]
p[0..15], p[0..31], @first_id_occurrence

rdcbreg

s[0..31], [cbreg:u5 metadata:u2]

rddreg

s[0..31], [dreg:u6]

sadd.f32

s[0..31], s[0..31], s[0..31]
s[0..31], f32, s[0..31]

sadd.ov.s32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

sadd.s32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

sand.u32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

sc.u32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sceil.f32

s[0..31], s[0..31]

sclz.u32

s[0..31], s[0..31]

scvt.f32.s32

s[0..31], s[0..31]
s[0..31], f32

scvt.s32.f32

s[0..31], s[0..31]
s[0..31], f32

sdelay

u11

seq.f32

p[0..15], s[0..31], s[0..31]
p[0..15], s[0..31], f32

seq.s32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sfence

无操作数

sfence.scmf

无操作数

sfence.sel

u32
s[0..31]

sfence.stream.hbm

无操作数

sfence.stream.spmem

无操作数

sfloor.f32

s[0..31], s[0..31]

sge.f32

p[0..15], s[0..31], s[0..31]
p[0..15], s[0..31], f32

sge.s32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sge.u32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sgt.f32

p[0..15], s[0..31], s[0..31]
p[0..15], s[0..31], f32

sgt.s32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sgt.u32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

shalt

无操作数

simm.f32

s[0..31], f32

simm.s32

s[0..31], u32

sld

s[0..31], [smem:s[0..31]+u32]
s[0..31], [smem:s[0..31]+s[0..31]]
s[0..31], [smem:u32]

sld.cb

s[0..31], [smem:u32 cbreg:u5]
s[0..31], [smem:s[0..31] cbreg:u5]

sld.cb.upd

s[0..31], [smem:u32 cbreg:u5]
s[0..31], [smem:s[0..31] cbreg:u5]

sle.f32

p[0..15], s[0..31], s[0..31]
p[0..15], s[0..31], f32

sle.s32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sle.u32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

slt.f32

p[0..15], s[0..31], s[0..31]
p[0..15], s[0..31], f32

slt.s32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

slt.u32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

smax.f32

s[0..31], s[0..31], s[0..31]
s[0..31], s[0..31], f32

smax.u32

s[0..31], s[0..31], u32
s[0..31], s[0..31], s[0..31]

smin.f32

s[0..31], s[0..31], s[0..31]
s[0..31], s[0..31], f32

smin.u32

s[0..31], s[0..31], u32
s[0..31], s[0..31], s[0..31]

smov.u32

s[0..31], s[0..31]

sne.f32

p[0..15], s[0..31], s[0..31]
p[0..15], s[0..31], f32

sne.s32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sor.u32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

spop

s[0..31], drf
s[0..31], v2sf

srdreg.ddr

s[0..31]

srdreg.fsr

s[0..31]

srdreg.gtchi

s[0..31]

srdreg.gtclo

s[0..31]

srdreg.lcchi

s[0..31]

srdreg.lcclo

s[0..31]

srdreg.scid

s[0..31]

srdreg.tag

s[0..31]

srdreg.tbm

s[0..31]

srdreg.tm

s[0..31]

srdreg.yieldreq

s[0..31]

ssetifvalue

ifvalue, u32
ifvalue, s[0..31]

ssetpdepth

pdepth, u32
pdepth, s[0..31]

ssettag

u32
s[0..31]

sshla.ov.s32

s[0..31], s[0..31], u32
s[0..31], s[0..31], s[0..31]

sshll.u32

s[0..31], s[0..31], u32
s[0..31], s[0..31], s[0..31]

sshra.s32

s[0..31], s[0..31], u32
s[0..31], s[0..31], s[0..31]

sshrl.u32

s[0..31], s[0..31], u32
s[0..31], s[0..31], s[0..31]

sst

[smem:u32], s[0..31]
[smem:s[0..31]], s[0..31]

sst.cb

[smem:u32 cbreg:u5], s[0..31]
[smem:s[0..31] cbreg:u5], s[0..31]

sst.cb.upd

[smem:u32 cbreg:u5], s[0..31]
[smem:s[0..31] cbreg:u5], s[0..31]

ssub.f32

s[0..31], s[0..31], s[0..31]
s[0..31], f32, s[0..31]

ssub.ov.s32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

ssub.s32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

stileid.u32

s[0..31]

sweird.f32

p[0..15], s[0..31]

sxor.u32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

wrcbreg

[cbreg:0x6 metadata:u2], u32
[cbreg:0x6 metadata:u2], s[0..31]

wrdreg

[dreg:u5], u32
[dreg:u5], s[0..31]

dma#

助记符

按顺序排列的操作数类型

dma.local

[timem:s[0..31]], [sflag:u32], [hbm:s[0..31]], u32
[hbm:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32
[timem:s[0..31]], [sflag:u32], [hbm:s[0..31]], s[0..31]
[hbm:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31]
[timem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], u32
[hbm:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32
[timem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], s[0..31]
[hbm:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31]
[hbm:s[0..31]], [sflag:u32], [simem:s[0..31]], u32
[hbm:s[0..31]], [sflag:u32], [smem:s[0..31]], u32
[simem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32
[smem:s[0..31]], [sflag:u32], [hbm:s[0..31]], u32
[smem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32
[timem:s[0..31]], [sflag:u32], [simem:s[0..31]], u32
[timem:s[0..31]], [sflag:u32], [smem:s[0..31]], u32
[hbm:s[0..31]], [sflag:u32], [simem:s[0..31]], s[0..31]
[hbm:s[0..31]], [sflag:u32], [smem:s[0..31]], s[0..31]
[simem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31]
[smem:s[0..31]], [sflag:u32], [hbm:s[0..31]], s[0..31]
[smem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31]
[timem:s[0..31]], [sflag:u32], [simem:s[0..31]], s[0..31]
[timem:s[0..31]], [sflag:u32], [smem:s[0..31]], s[0..31]
[hbm:s[0..31]], [sflag:s[0..31]], [simem:s[0..31]], u32
[hbm:s[0..31]], [sflag:s[0..31]], [smem:s[0..31]], u32
[simem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32
[smem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], u32
[smem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32
[timem:s[0..31]], [sflag:s[0..31]], [simem:s[0..31]], u32
[timem:s[0..31]], [sflag:s[0..31]], [smem:s[0..31]], u32
[hbm:s[0..31]], [sflag:s[0..31]], [simem:s[0..31]], s[0..31]
[hbm:s[0..31]], [sflag:s[0..31]], [smem:s[0..31]], s[0..31]
[simem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31]
[smem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], s[0..31]
[smem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31]
[timem:s[0..31]], [sflag:s[0..31]], [simem:s[0..31]], s[0..31]
[timem:s[0..31]], [sflag:s[0..31]], [smem:s[0..31]], s[0..31]
[hbm:s[0..31]], [sflag:u32], [hbm:s[0..31]], u32
[hbm:s[0..31]], [sflag:u32], [timem:s[0..31]], u32
[simem:s[0..31]], [sflag:u32], [simem:s[0..31]], u32
[simem:s[0..31]], [sflag:u32], [smem:s[0..31]], u32
[smem:s[0..31]], [sflag:u32], [simem:s[0..31]], u32
[smem:s[0..31]], [sflag:u32], [smem:s[0..31]], u32
[spmem:s[0..31]], [sflag:u32], [hbm:s[0..31]], u32
[spmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32
[timem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32
[timem:s[0..31]], [sflag:u32], [timem:s[0..31]], u32
[hbm:s[0..31]], [sflag:u32], [hbm:s[0..31]], s[0..31]
[hbm:s[0..31]], [sflag:u32], [timem:s[0..31]], s[0..31]
[simem:s[0..31]], [sflag:u32], [simem:s[0..31]], s[0..31]
[simem:s[0..31]], [sflag:u32], [smem:s[0..31]], s[0..31]
[smem:s[0..31]], [sflag:u32], [simem:s[0..31]], s[0..31]
[smem:s[0..31]], [sflag:u32], [smem:s[0..31]], s[0..31]
[spmem:s[0..31]], [sflag:u32], [hbm:s[0..31]], s[0..31]
[spmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31]
[timem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31]
[timem:s[0..31]], [sflag:u32], [timem:s[0..31]], s[0..31]
[hbm:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], u32
[hbm:s[0..31]], [sflag:s[0..31]], [timem:s[0..31]], u32
[simem:s[0..31]], [sflag:s[0..31]], [simem:s[0..31]], u32
[simem:s[0..31]], [sflag:s[0..31]], [smem:s[0..31]], u32
[smem:s[0..31]], [sflag:s[0..31]], [simem:s[0..31]], u32
[smem:s[0..31]], [sflag:s[0..31]], [smem:s[0..31]], u32
[spmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], u32
[spmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32
[timem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32
[timem:s[0..31]], [sflag:s[0..31]], [timem:s[0..31]], u32
[hbm:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], s[0..31]
[hbm:s[0..31]], [sflag:s[0..31]], [timem:s[0..31]], s[0..31]
[simem:s[0..31]], [sflag:s[0..31]], [simem:s[0..31]], s[0..31]
[simem:s[0..31]], [sflag:s[0..31]], [smem:s[0..31]], s[0..31]
[smem:s[0..31]], [sflag:s[0..31]], [simem:s[0..31]], s[0..31]
[smem:s[0..31]], [sflag:s[0..31]], [smem:s[0..31]], s[0..31]
[spmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], s[0..31]
[spmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31]
[timem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31]
[timem:s[0..31]], [sflag:s[0..31]], [timem:s[0..31]], s[0..31]
[simem:s[0..31]], [sflag:u32], [hbm:s[0..31]], u32
[simem:s[0..31]], [sflag:u32], [timem:s[0..31]], u32
[smem:s[0..31]], [sflag:u32], [timem:s[0..31]], u32
[spmem:s[0..31]], [sflag:u32], [simem:s[0..31]], u32
[spmem:s[0..31]], [sflag:u32], [smem:s[0..31]], u32
[simem:s[0..31]], [sflag:u32], [hbm:s[0..31]], s[0..31]
[simem:s[0..31]], [sflag:u32], [timem:s[0..31]], s[0..31]
[smem:s[0..31]], [sflag:u32], [timem:s[0..31]], s[0..31]
[spmem:s[0..31]], [sflag:u32], [simem:s[0..31]], s[0..31]
[spmem:s[0..31]], [sflag:u32], [smem:s[0..31]], s[0..31]
[simem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], u32
[simem:s[0..31]], [sflag:s[0..31]], [timem:s[0..31]], u32
[smem:s[0..31]], [sflag:s[0..31]], [timem:s[0..31]], u32
[spmem:s[0..31]], [sflag:s[0..31]], [simem:s[0..31]], u32
[spmem:s[0..31]], [sflag:s[0..31]], [smem:s[0..31]], u32
[simem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], s[0..31]
[simem:s[0..31]], [sflag:s[0..31]], [timem:s[0..31]], s[0..31]
[smem:s[0..31]], [sflag:s[0..31]], [timem:s[0..31]], s[0..31]
[spmem:s[0..31]], [sflag:s[0..31]], [simem:s[0..31]], s[0..31]
[spmem:s[0..31]], [sflag:s[0..31]], [smem:s[0..31]], s[0..31]
[spmem:s[0..31]], [sflag:u32], [timem:s[0..31]], u32
[spmem:s[0..31]], [sflag:u32], [timem:s[0..31]], s[0..31]
[spmem:s[0..31]], [sflag:s[0..31]], [timem:s[0..31]], u32
[spmem:s[0..31]], [sflag:s[0..31]], [timem:s[0..31]], s[0..31]

general_dma

dma_vreg_dest_offset=(u6), dma_vreg_source_offset=(u6), dma_sreg_source_offset=(u5), dma_length=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32), thread_select=(u1), trace_en=(u1), dst_opcode=(write / write_4b / compare_and_swap / read_and_add), dst_mem_core_id=(reserved_self / non_core / tensor_core_0 / reserved_tc_1 / sparse_core_0 / sparse_core_1 / reserved_sc_2 / reserved_sc_3), dst_mem_mem_id=(hbm / host / imem / timem), sync_mode=(32b_word / descriptor), relaxed_ordering_or_thread_select_1=(u1), src_mem_core_id=(reserved_self / non_core / tensor_core_0 / reserved_tc_1 / sparse_core_0 / sparse_core_1 / reserved_sc_2 / reserved_sc_3), src_mem_mem_id=(hbm / host / imem / timem), dma_sreg_dest_offset=(u5), source_in_vreg=(u1), dest_in_vreg=(u1), nondefault_stride_dimensions=(u3), source_sync_flag_number_valid=(u1), source_sync_flag_number=(u5), dest_sync_flags_valid=(u1), dest_sync_flags_vs1=(u5), destination_id_valid=(u1), destination_id=(u5), upper_host_addr_valid=(u1), upper_host_addr=(u5)

single_strided_dma

dma_sreg_source_offset=(u5), dma_length=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32), thread_select=(u1), trace_en=(u1), dst_opcode=(write / write_4b / compare_and_swap / read_and_add), dst_mem_core_id=(reserved_self / non_core / tensor_core_0 / reserved_tc_1 / sparse_core_0 / sparse_core_1 / reserved_sc_2 / reserved_sc_3), dst_mem_mem_id=(hbm / host / imem / timem), sync_mode=(32b_word / descriptor), relaxed_ordering_or_thread_select_1=(u1), src_mem_core_id=(reserved_self / non_core / tensor_core_0 / reserved_tc_1 / sparse_core_0 / sparse_core_1 / reserved_sc_2 / reserved_sc_3), src_mem_mem_id=(hbm / host / imem / timem), dma_sreg_dest_offset=(u5), dest_sync_flags=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32), inner_vector_length=(u5), source_stride_valid=(u1), source_stride=(u5), destination_stride_valid=(u1), destination_stride=(u5), elements_per_stride_valid=(u1), elements_per_stride=(u5), upper_host_addr_valid=(u1), upper_host_addr=(u5)

misc#

助记符

按顺序排列的操作数类型

atomic_tile_add_set_done

x=(u5), y=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32)

atomic_tile_add_set_done_inverted

x=(u5), y=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32)

por

p[0..15], p[0..31], p[0..31]
p[0..15], @first_id_occurrence, p[0..31]
p[0..15], p[0..31], @first_id_occurrence

push_scmf

y=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32)

read_sync_state_done

pdst=(preg_0 / preg_1 / preg_2 / preg_3 / preg_4 / preg_5 / preg_6 / preg_7 / preg_8 / preg_9 / preg_10 / preg_11 / preg_12 / preg_13 / reserved_0 / reserved_1)

read_sync_state_value

dest=(u5)

sadd.ov.s32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

sadd.s32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

sand.u32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

sbarrier.arrive

[bflag:u2], u32
[bflag:u2], s[0..31]

sc.u32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sclz.u32

s[0..31], s[0..31]

seq.s32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sge.s32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sge.u32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sgt.s32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sgt.u32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

simm.f32

s[0..31], f32

simm.s32

s[0..31], u32

sint

u32
s[0..31]

sle.s32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sle.u32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

slt.s32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

slt.u32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

smax.u32

s[0..31], s[0..31], u32
s[0..31], s[0..31], s[0..31]

smem.add.s32

[smem:s[0..31]], [sflag:u5], u32

smin.u32

s[0..31], s[0..31], u32
s[0..31], s[0..31], s[0..31]

smov.u32

s[0..31], s[0..31]

sne.s32

p[0..15], s[0..31], u32
p[0..15], s[0..31], s[0..31]

sor.u32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

ssettm

u32
s[0..31]

sshla.ov.s32

s[0..31], s[0..31], u32
s[0..31], s[0..31], s[0..31]

sshll.u32

s[0..31], s[0..31], u32
s[0..31], s[0..31], s[0..31]

sshra.s32

s[0..31], s[0..31], u32
s[0..31], s[0..31], s[0..31]

sshrl.u32

s[0..31], s[0..31], u32
s[0..31], s[0..31], s[0..31]

ssub.ov.s32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

ssub.s32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

ssyncadd.both.s32

[sflag:s[0..31]], [sflag:s[0..31]], u32
[sflag:s[0..31]], [sflag:s[0..31]], s[0..31]

ssyncadd.other.s32

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

ssyncadd.remote.done.s32

[sflag:s[0..31]@s[0..31]], u32
[sflag:s[0..31]@s[0..31]], s[0..31]
[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

ssyncadd.remote.doneinv.s32

[sflag:s[0..31]@s[0..31]], u32
[sflag:s[0..31]@s[0..31]], s[0..31]
[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

ssyncadd.remote.s32

[sflag:s[0..31]@s[0..31]], u32
[sflag:s[0..31]@s[0..31]], s[0..31]
[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

ssyncadd.s32

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

ssyncadd.tile.s32

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

ssyncpa.u1

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

ssyncread

s[0..31], [sflag:u32]
s[0..31], [sflag:s[0..31]]

ssyncread.done

s[0..31], [sflag:u32]
s[0..31], [sflag:s[0..31]]

ssyncread.pa

s[0..31], [sflag:u32]
s[0..31], [sflag:s[0..31]]

ssyncset.both.done

[sflag:s[0..31]], [sflag:s[0..31]], u32
[sflag:s[0..31]], [sflag:s[0..31]], s[0..31]

ssyncset.both.s32

[sflag:s[0..31]], [sflag:s[0..31]], u32
[sflag:s[0..31]], [sflag:s[0..31]], s[0..31]

ssyncset.done

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

ssyncset.other.done

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

ssyncset.other.s32

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

ssyncset.remote.done

[sflag:s[0..31]@s[0..31]], u32
[sflag:s[0..31]@s[0..31]], s[0..31]
[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

ssyncset.remote.doneinv

[sflag:s[0..31]@s[0..31]], u32
[sflag:s[0..31]@s[0..31]], s[0..31]
[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

ssyncset.remote.s32

[sflag:s[0..31]@s[0..31]], u32
[sflag:s[0..31]@s[0..31]], s[0..31]
[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

ssyncset.s32

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

ssyncset.tile.done.s32

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

ssyncset.tile.doneinv.s32

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

ssyncset.tile.s32

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

strace

u32
s[0..31]

swait.done

[sflag:s[0..31]]

swait.eq

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swait.eqordone

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swait.ge

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swait.geordone

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swait.gt

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swait.gtordone

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swait.lt

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swait.ltordone

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swait.ne

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swait.neordone

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swait.notdone

[sflag:s[0..31]]

swatch.done

[sflag:s[0..31]]

swatch.end

无操作数

swatch.end.sel

u32
s[0..31]

swatch.eq

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swatch.eqordone

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swatch.ge

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swatch.geordone

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swatch.gt

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swatch.gtordone

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swatch.lt

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swatch.ltordone

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swatch.ne

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swatch.neordone

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

swatch.notdone

[sflag:s[0..31]]

swatch.wait

[sflag:s[0..31]]

swatch.wait.sel

[sflag:s[0..31]], u32
[sflag:s[0..31]], s[0..31]

sxor.u32

s[0..31], u32, s[0..31]
s[0..31], s[0..31], s[0..31]

yieldable_sync_done

x=(u5)

yieldable_sync_equal

x=(u5), y=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32)

yieldable_sync_equal_or_done

x=(u5), y=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32)

yieldable_sync_greater

x=(u5), y=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32)

yieldable_sync_greater_or_done

x=(u5), y=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32)

yieldable_sync_greater_or_equal

x=(u5), y=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32)

yieldable_sync_greater_or_equal_or_done

x=(u5), y=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32)

yieldable_sync_less

x=(u5), y=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32)

yieldable_sync_less_or_done

x=(u5), y=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32)

yieldable_sync_not_done

x=(u5)

yieldable_sync_not_equal

x=(u5), y=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32)

yieldable_sync_not_equal_or_done

x=(u5), y=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32)

va0#

助记符

按顺序排列的操作数类型

eup_push

x=(u6)

unpack_interleaved_compressed_lower_s8_to_bf16

dest=(u6), x=(u6)

unpack_interleaved_compressed_lower_s8_to_s16

dest=(u6), x=(u6)

unpack_interleaved_compressed_lower_u8_to_bf16

dest=(u6), x=(u6)

unpack_interleaved_compressed_lower_u8_to_u16

dest=(u6), x=(u6)

unpack_interleaved_compressed_upper_s8_to_bf16

dest=(u6), x=(u6)

unpack_interleaved_compressed_upper_s8_to_s16

dest=(u6), x=(u6)

unpack_interleaved_compressed_upper_u8_to_bf16

dest=(u6), x=(u6)

unpack_interleaved_compressed_upper_u8_to_u16

dest=(u6), x=(u6)

vadd.s16

v[0..63], v[0..63], v[0..63]

vadd.s32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]

vand.u32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]

vbcast.xlane.b16

v[0..63], v[0..63], u32
v[0..63], v[0..63], s[0..31]

vbroadcast

v[0..63], v[0..63], u32
v[0..63], v[0..63], s[0..31]

vc.u16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vc.u32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vceil.bf16

v[0..63], v[0..63]

vceil.f32

v[0..63], v[0..63]

vclamp.gez.bf16

v[0..63], v[0..63], v[0..63]

vclamp.gez.f32

v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]
v[0..63], v[0..63], f32

vclamps.bf16

v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]

vclamps.f32

v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]
v[0..63], v[0..63], f32

vclass.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vclass.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vclz

v[0..63], v[0..63]

vcmask

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, s[0..31]

vcombine.high

v[0..63], v[0..63], v[0..63]

vcombine.low

v[0..63], v[0..63], v[0..63]

vcosq.bf16

erf, v[0..63]

vcosq.f32

erf, v[0..63]

vcvt.f32.f8e5m2

v[0..63], v[0..63]

vcvt.f32.hf16

v[0..63], v[0..63]

vcvt.f32.if8

v[0..63], v[0..63]

vcvt.f32.s32

v[0..63], v[0..63]

vcvt.s32.f32

v[0..63], v[0..63]

vcvt.sr.f32.bf16

v[0..63], v[0..63], v[0..63]

vcvt.sr.f32.bf8

v[0..63], v[0..63], v[0..63]

vcvt.sr.f32.hf16

v[0..63], v[0..63], v[0..63]

vcvt.sr.f32.if8

v[0..63], v[0..63], v[0..63]

vector_clamp_asymmetric_bf16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_clamp_asymmetric_f32

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_convert_bf16_to_s4

dest=(u6), x=(u6)

vector_convert_bf16_to_s8

dest=(u6), x=(u6)

vector_convert_bf16_to_u4

dest=(u6), x=(u6)

vector_convert_bf16_to_u8

dest=(u6), x=(u6)

vector_convert_f32_to_bf16

dest=(u6), x=(u6)

vector_convert_s4_to_bf16

dest=(u6), x=(u6)

vector_convert_s8_to_bf16

dest=(u6), x=(u6)

vector_convert_u4_to_bf16

dest=(u6), x=(u6)

vector_convert_u8_to_bf16

dest=(u6), x=(u6)

vector_gt_u16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_gte_u16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_lane_left_shift_insert_b16

dest=(u6), x=(u6), y_vreg=(u6), shift_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30)

vector_lt_u16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_lte_u16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_remap_bf16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_remap_f32

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

veq.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

veq.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

veq.s16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

veq.s32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

verf.bf16

erf, v[0..63]

verf.f32

erf, v[0..63]

vfloor.bf16

v[0..63], v[0..63]

vfloor.f32

v[0..63], v[0..63]

vge.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vge.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vge.s16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vge.s32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vge.u32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vgt.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vgt.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vgt.s16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vgt.s32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vgt.u32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vimm.bf16

v[0..63], u32

vimm.f32

v[0..63], f32

vimm.hf16

v[0..63], u32

vimm.s32

v[0..63], u32

vlaneseq.c.b16

v[0..63]

vlaneseq.i.b16

v[0..63]

vlaneseq.u32

v[0..63]

vle.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vle.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vle.s16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vle.s32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vle.to.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vle.to.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vle.u32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vlog2.bf16

erf, v[0..63]

vlog2.f32

erf, v[0..63]

vlt.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vlt.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vlt.s16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vlt.s32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vlt.to.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vlt.to.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vlt.u32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vmand

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vmax.bf16

v[0..63], v[0..63], v[0..63]

vmax.f32

v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]
v[0..63], v[0..63], f32

vmax.u16

v[0..63], v[0..63], v[0..63]

vmax.u32

v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]

vmctz.xlane

v[0..63], vm[0..15]

vmctz.xlane.b16

v[0..63], vm[0..15]

vmin.bf16

v[0..63], v[0..63], v[0..63]

vmin.f32

v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]
v[0..63], v[0..63], f32

vmin.u16

v[0..63], v[0..63], v[0..63]

vmin.u32

v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]

vmmov

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vmneg

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vmor

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vmov

v[0..63], v[0..63]
v[0..63], s[0..31]

vmpackc.even

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vmpackc.low

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vmpcnt.ones.xlane

v[0..63], vm[0..15]

vmpcnt.ones.xlane.b16

v[0..63], vm[0..15]

vmperm.xlane.b16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15], s[0..31]

vmperm.xlane.b32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15], s[0..31]

vmperm.xlane.b8

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15], s[0..31]

vmperm.xlane.c2i.b16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15]

vmperm.xlane.c2i.b8

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15]

vmperm.xlane.i2c.b16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15]

vmperm.xlane.i2c.b8

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15]

vmprefix.xlane

v[0..63], vm[0..15]

vmprefix.xlane.b16

v[0..63], vm[0..15]

vmul.bf16

v[0..63], v[0..63], v[0..63]

vmul.f32

v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]
v[0..63], f32, v[0..63]

vmul.u16

v[0..63], v[0..63], v[0..63]

vmul.u32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]

vmulhi.u32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]

vmxor

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vne.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vne.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vne.s16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vne.s32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vnez.u8

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, s[0..31]

vnsel

v[0..63], vm[0..15], u32, v[0..63]
v[0..63], vm[0..15], s[0..31], v[0..63]

vor.u32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]

vpack.c.b16.b8

v[0..63], v[0..63], v[0..63]

vpack.c.b2.b1

v[0..63], v[0..63], v[0..63]

vpack.c.b32.b16

v[0..63], v[0..63], v[0..63]

vpack.c.b4.b2

v[0..63], v[0..63], v[0..63]

vpack.c.b8.b4

v[0..63], v[0..63], v[0..63]

vpack.c.bf16.s8

v[0..63], v[0..63], v[0..63]

vpack.c.bf16.u8

v[0..63], v[0..63], v[0..63]

vpack.c.f32.bf16

v[0..63], v[0..63], v[0..63]

vpack.i.b16.b8

v[0..63], v[0..63], v[0..63]

vpack.i.b32.b16

v[0..63], v[0..63], v[0..63]

vpack.i.b8.b4

v[0..63], v[0..63], v[0..63]

vpack.i.bf16.s8

v[0..63], v[0..63], v[0..63]

vpack.i.bf16.u8

v[0..63], v[0..63], v[0..63]

vpack.i.f32.bf16

v[0..63], v[0..63], v[0..63]

vpcnt

v[0..63], v[0..63]

vperm.xlane

v[0..63], v[0..63], v[0..63]

vperm.xlane.b16

v[0..63], v[0..63], v[0..63]

vperm.xlane.b8

v[0..63], v[0..63], v[0..63]

vperm.xlane.c2i.b16

v[0..63], v[0..63]

vperm.xlane.c2i.b8

v[0..63], v[0..63]

vperm.xlane.i2c.b16

v[0..63], v[0..63]

vperm.xlane.i2c.b8

v[0..63], v[0..63]

vpow2.bf16

erf, v[0..63]

vpow2.f32

erf, v[0..63]

vpsel

v[0..63], p5, u32, v[0..63]
v[0..63], p5, v[0..63], v[0..63]
v[0..63], !p5, s[0..31], v[0..63]

vrcp.bf16

erf, v[0..63]

vrcp.f32

erf, v[0..63]

vrot.xlane.b16

v[0..63], v[0..63], u32
v[0..63], v[0..63], s[0..31]

vrot.xlane.b32

v[0..63], v[0..63], u32
v[0..63], v[0..63], s[0..31]

vround.rtna.bf16

v[0..63], v[0..63]

vround.rtna.f32

v[0..63], v[0..63]

vrsqrt.bf16

erf, v[0..63]

vrsqrt.f32

erf, v[0..63]

vsel

v[0..63], vm[0..15], u32, v[0..63]
v[0..63], vm[0..15], v[0..63], v[0..63]
v[0..63], vm[0..15], s[0..31], v[0..63]

vshift.insert

v[0..63], v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63], s[0..31]

vshll.u32

v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]

vshra.s32

v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]

vshrl.u32

v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]

vsigshft.bf16

erf, v[0..63]

vsigshft.f32

erf, v[0..63]

vsinq.bf16

erf, v[0..63]

vsinq.f32

erf, v[0..63]

vsub.s16

v[0..63], v[0..63], v[0..63]

vsub.s32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]

vtanh.bf16

erf, v[0..63]

vtanh.f32

erf, v[0..63]

vtrunc.bf16

v[0..63], v[0..63]

vtrunc.f32

v[0..63], v[0..63]

vunpack.c.0.bf8.f32

v[0..63], v[0..63]

vunpack.c.0.if8.f32

v[0..63], v[0..63]

vunpack.c.0.s8.s32

v[0..63], v[0..63]

vunpack.c.1.bf8.f32

v[0..63], v[0..63]

vunpack.c.1.if8.f32

v[0..63], v[0..63]

vunpack.c.1.s8.s32

v[0..63], v[0..63]

vunpack.c.2.bf8.f32

v[0..63], v[0..63]

vunpack.c.2.if8.f32

v[0..63], v[0..63]

vunpack.c.2.s8.s32

v[0..63], v[0..63]

vunpack.c.3.bf8.f32

v[0..63], v[0..63]

vunpack.c.3.if8.f32

v[0..63], v[0..63]

vunpack.c.3.s8.s32

v[0..63], v[0..63]

vunpack.c.l.bf16.f32

v[0..63], v[0..63]

vunpack.c.l.hf16.f32

v[0..63], v[0..63]

vunpack.c.l.s1.s2

v[0..63], v[0..63]

vunpack.c.l.s16.s32

v[0..63], v[0..63]

vunpack.c.l.s2.s4

v[0..63], v[0..63]

vunpack.c.l.s4.s8

v[0..63], v[0..63]

vunpack.c.l.s8.bf16

v[0..63], v[0..63]

vunpack.c.l.s8.s16

v[0..63], v[0..63]

vunpack.c.l.u8.bf16

v[0..63], v[0..63]

vunpack.c.l.u8.u16

v[0..63], v[0..63]

vunpack.c.u.bf16.f32

v[0..63], v[0..63]

vunpack.c.u.hf16.f32

v[0..63], v[0..63]

vunpack.c.u.s1.s2

v[0..63], v[0..63]

vunpack.c.u.s16.s32

v[0..63], v[0..63]

vunpack.c.u.s2.s4

v[0..63], v[0..63]

vunpack.c.u.s4.s8

v[0..63], v[0..63]

vunpack.c.u.s8.bf16

v[0..63], v[0..63]

vunpack.c.u.s8.s16

v[0..63], v[0..63]

vunpack.c.u.u8.bf16

v[0..63], v[0..63]

vunpack.c.u.u8.u16

v[0..63], v[0..63]

vunpack.i.l.bf16.f32

v[0..63], v[0..63]

vunpack.i.l.s16.s32

v[0..63], v[0..63]

vunpack.i.l.s8.bf16

v[0..63], v[0..63]

vunpack.i.l.s8.s16

v[0..63], v[0..63]

vunpack.i.l.u8.bf16

v[0..63], v[0..63]

vunpack.i.l.u8.u16

v[0..63], v[0..63]

vunpack.i.u.bf16.f32

v[0..63], v[0..63]

vunpack.i.u.s16.s32

v[0..63], v[0..63]

vunpack.i.u.s8.bf16

v[0..63], v[0..63]

vunpack.i.u.s8.s16

v[0..63], v[0..63]

vunpack.i.u.u8.bf16

v[0..63], v[0..63]

vunpack.i.u.u8.u16

v[0..63], v[0..63]

vweird.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vweird.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vxor.u32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]

va1#

助记符

按顺序排列的操作数类型

eup_push

x=(u6)

unpack_interleaved_compressed_lower_s8_to_bf16

dest=(u6), x=(u6)

unpack_interleaved_compressed_lower_s8_to_s16

dest=(u6), x=(u6)

unpack_interleaved_compressed_lower_u8_to_bf16

dest=(u6), x=(u6)

unpack_interleaved_compressed_lower_u8_to_u16

dest=(u6), x=(u6)

unpack_interleaved_compressed_upper_s8_to_bf16

dest=(u6), x=(u6)

unpack_interleaved_compressed_upper_s8_to_s16

dest=(u6), x=(u6)

unpack_interleaved_compressed_upper_u8_to_bf16

dest=(u6), x=(u6)

unpack_interleaved_compressed_upper_u8_to_u16

dest=(u6), x=(u6)

vadd.bf16

v[0..63], v[0..63], v[0..63]

vadd.f32

v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]
v[0..63], f32, v[0..63]

vadd.s16

v[0..63], v[0..63], v[0..63]

vadd.s32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]

vand.u32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]

vbcast.xlane.b16

v[0..63], v[0..63], u32
v[0..63], v[0..63], s[0..31]

vbroadcast

v[0..63], v[0..63], u32
v[0..63], v[0..63], s[0..31]

vc.u16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vc.u32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vceil.bf16

v[0..63], v[0..63]

vceil.f32

v[0..63], v[0..63]

vclamp.gez.bf16

v[0..63], v[0..63], v[0..63]

vclamp.gez.f32

v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]
v[0..63], v[0..63], f32

vclamps.bf16

v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]

vclamps.f32

v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]
v[0..63], v[0..63], f32

vclass.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vclass.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vclz

v[0..63], v[0..63]

vcmask

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, s[0..31]

vcombine.high

v[0..63], v[0..63], v[0..63]

vcombine.low

v[0..63], v[0..63], v[0..63]

vcosq.bf16

erf, v[0..63]

vcosq.f32

erf, v[0..63]

vcvt.f32.f8e5m2

v[0..63], v[0..63]

vcvt.f32.hf16

v[0..63], v[0..63]

vcvt.f32.if8

v[0..63], v[0..63]

vcvt.f32.s32

v[0..63], v[0..63]

vcvt.s32.f32

v[0..63], v[0..63]

vcvt.sr.f32.bf16

v[0..63], v[0..63], v[0..63]

vcvt.sr.f32.bf8

v[0..63], v[0..63], v[0..63]

vcvt.sr.f32.hf16

v[0..63], v[0..63], v[0..63]

vcvt.sr.f32.if8

v[0..63], v[0..63], v[0..63]

vector_add_high_f32_bf16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_add_low_f32_bf16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_clamp_asymmetric_bf16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_clamp_asymmetric_f32

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_convert_bf16_to_s4

dest=(u6), x=(u6)

vector_convert_bf16_to_s8

dest=(u6), x=(u6)

vector_convert_bf16_to_u4

dest=(u6), x=(u6)

vector_convert_bf16_to_u8

dest=(u6), x=(u6)

vector_convert_f32_to_bf16

dest=(u6), x=(u6)

vector_convert_s4_to_bf16

dest=(u6), x=(u6)

vector_convert_s8_to_bf16

dest=(u6), x=(u6)

vector_convert_u4_to_bf16

dest=(u6), x=(u6)

vector_convert_u8_to_bf16

dest=(u6), x=(u6)

vector_gt_u16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_gte_u16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_lane_left_shift_insert_b16

dest=(u6), x=(u6), y_vreg=(u6), shift_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30)

vector_lt_u16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_lte_u16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_remap_bf16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_remap_f32

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

veq.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

veq.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

veq.s16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

veq.s32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

verf.bf16

erf, v[0..63]

verf.f32

erf, v[0..63]

vfloor.bf16

v[0..63], v[0..63]

vfloor.f32

v[0..63], v[0..63]

vge.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vge.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vge.s16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vge.s32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vge.u32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vgt.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vgt.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vgt.s16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vgt.s32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vgt.u32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vimm.bf16

v[0..63], u32

vimm.f32

v[0..63], f32

vimm.hf16

v[0..63], u32

vimm.s32

v[0..63], u32

vlaneseq.c.b16

v[0..63]

vlaneseq.i.b16

v[0..63]

vlaneseq.u32

v[0..63]

vle.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vle.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vle.s16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vle.s32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vle.to.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vle.to.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vle.u32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vlog2.bf16

erf, v[0..63]

vlog2.f32

erf, v[0..63]

vlt.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vlt.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vlt.s16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vlt.s32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vlt.to.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vlt.to.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vlt.u32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vmand

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vmax.bf16

v[0..63], v[0..63], v[0..63]

vmax.f32

v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]
v[0..63], v[0..63], f32

vmax.u16

v[0..63], v[0..63], v[0..63]

vmax.u32

v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]

vmctz.xlane

v[0..63], vm[0..15]

vmctz.xlane.b16

v[0..63], vm[0..15]

vmin.bf16

v[0..63], v[0..63], v[0..63]

vmin.f32

v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]
v[0..63], v[0..63], f32

vmin.u16

v[0..63], v[0..63], v[0..63]

vmin.u32

v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]

vmmov

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vmneg

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vmor

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vmov

v[0..63], v[0..63]
v[0..63], s[0..31]

vmpackc.even

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vmpackc.low

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vmpcnt.ones.xlane

v[0..63], vm[0..15]

vmpcnt.ones.xlane.b16

v[0..63], vm[0..15]

vmperm.xlane.b16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15], s[0..31]

vmperm.xlane.b32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15], s[0..31]

vmperm.xlane.b8

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15], s[0..31]

vmperm.xlane.c2i.b16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15]

vmperm.xlane.c2i.b8

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15]

vmperm.xlane.i2c.b16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15]

vmperm.xlane.i2c.b8

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm[0..15]

vmprefix.xlane

v[0..63], vm[0..15]

vmprefix.xlane.b16

v[0..63], vm[0..15]

vmul.bf16

v[0..63], v[0..63], v[0..63]

vmul.f32

v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]
v[0..63], f32, v[0..63]

vmul.u16

v[0..63], v[0..63], v[0..63]

vmul.u32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]

vmxor

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vne.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vne.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vne.s16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vne.s32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vnez.u8

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, s[0..31]

vnsel

v[0..63], vm[0..15], u32, v[0..63]
v[0..63], vm[0..15], s[0..31], v[0..63]

vor.u32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]

vpack.c.b16.b8

v[0..63], v[0..63], v[0..63]

vpack.c.b2.b1

v[0..63], v[0..63], v[0..63]

vpack.c.b32.b16

v[0..63], v[0..63], v[0..63]

vpack.c.b4.b2

v[0..63], v[0..63], v[0..63]

vpack.c.b8.b4

v[0..63], v[0..63], v[0..63]

vpack.c.bf16.s8

v[0..63], v[0..63], v[0..63]

vpack.c.bf16.u8

v[0..63], v[0..63], v[0..63]

vpack.c.f32.bf16

v[0..63], v[0..63], v[0..63]

vpack.i.b16.b8

v[0..63], v[0..63], v[0..63]

vpack.i.b32.b16

v[0..63], v[0..63], v[0..63]

vpack.i.b8.b4

v[0..63], v[0..63], v[0..63]

vpack.i.bf16.s8

v[0..63], v[0..63], v[0..63]

vpack.i.bf16.u8

v[0..63], v[0..63], v[0..63]

vpack.i.f32.bf16

v[0..63], v[0..63], v[0..63]

vpcnt

v[0..63], v[0..63]

vperm.xlane

v[0..63], v[0..63], v[0..63]

vperm.xlane.b16

v[0..63], v[0..63], v[0..63]

vperm.xlane.b8

v[0..63], v[0..63], v[0..63]

vperm.xlane.c2i.b16

v[0..63], v[0..63]

vperm.xlane.c2i.b8

v[0..63], v[0..63]

vperm.xlane.i2c.b16

v[0..63], v[0..63]

vperm.xlane.i2c.b8

v[0..63], v[0..63]

vpow2.bf16

erf, v[0..63]

vpow2.f32

erf, v[0..63]

vpsel

v[0..63], p5, u32, v[0..63]
v[0..63], p5, v[0..63], v[0..63]
v[0..63], !p5, s[0..31], v[0..63]

vrcp.bf16

erf, v[0..63]

vrcp.f32

erf, v[0..63]

vrot.xlane.b16

v[0..63], v[0..63], u32
v[0..63], v[0..63], s[0..31]

vrot.xlane.b32

v[0..63], v[0..63], u32
v[0..63], v[0..63], s[0..31]

vround.rtna.bf16

v[0..63], v[0..63]

vround.rtna.f32

v[0..63], v[0..63]

vrsqrt.bf16

erf, v[0..63]

vrsqrt.f32

erf, v[0..63]

vsel

v[0..63], vm[0..15], u32, v[0..63]
v[0..63], vm[0..15], v[0..63], v[0..63]
v[0..63], vm[0..15], s[0..31], v[0..63]

vshift.insert

v[0..63], v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63], s[0..31]

vshll.u32

v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]

vshra.s32

v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]

vshrl.u32

v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]

vsigshft.bf16

erf, v[0..63]

vsigshft.f32

erf, v[0..63]

vsinq.bf16

erf, v[0..63]

vsinq.f32

erf, v[0..63]

vsub.bf16

v[0..63], v[0..63], v[0..63]

vsub.f32

v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]
v[0..63], f32, v[0..63]

vsub.s16

v[0..63], v[0..63], v[0..63]

vsub.s32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]

vtanh.bf16

erf, v[0..63]

vtanh.f32

erf, v[0..63]

vtrunc.bf16

v[0..63], v[0..63]

vtrunc.f32

v[0..63], v[0..63]

vunpack.c.0.bf8.f32

v[0..63], v[0..63]

vunpack.c.0.if8.f32

v[0..63], v[0..63]

vunpack.c.0.s8.s32

v[0..63], v[0..63]

vunpack.c.1.bf8.f32

v[0..63], v[0..63]

vunpack.c.1.if8.f32

v[0..63], v[0..63]

vunpack.c.1.s8.s32

v[0..63], v[0..63]

vunpack.c.2.bf8.f32

v[0..63], v[0..63]

vunpack.c.2.if8.f32

v[0..63], v[0..63]

vunpack.c.2.s8.s32

v[0..63], v[0..63]

vunpack.c.3.bf8.f32

v[0..63], v[0..63]

vunpack.c.3.if8.f32

v[0..63], v[0..63]

vunpack.c.3.s8.s32

v[0..63], v[0..63]

vunpack.c.l.bf16.f32

v[0..63], v[0..63]

vunpack.c.l.hf16.f32

v[0..63], v[0..63]

vunpack.c.l.s1.s2

v[0..63], v[0..63]

vunpack.c.l.s16.s32

v[0..63], v[0..63]

vunpack.c.l.s2.s4

v[0..63], v[0..63]

vunpack.c.l.s4.s8

v[0..63], v[0..63]

vunpack.c.l.s8.bf16

v[0..63], v[0..63]

vunpack.c.l.s8.s16

v[0..63], v[0..63]

vunpack.c.l.u8.bf16

v[0..63], v[0..63]

vunpack.c.l.u8.u16

v[0..63], v[0..63]

vunpack.c.u.bf16.f32

v[0..63], v[0..63]

vunpack.c.u.hf16.f32

v[0..63], v[0..63]

vunpack.c.u.s1.s2

v[0..63], v[0..63]

vunpack.c.u.s16.s32

v[0..63], v[0..63]

vunpack.c.u.s2.s4

v[0..63], v[0..63]

vunpack.c.u.s4.s8

v[0..63], v[0..63]

vunpack.c.u.s8.bf16

v[0..63], v[0..63]

vunpack.c.u.s8.s16

v[0..63], v[0..63]

vunpack.c.u.u8.bf16

v[0..63], v[0..63]

vunpack.c.u.u8.u16

v[0..63], v[0..63]

vunpack.i.l.bf16.f32

v[0..63], v[0..63]

vunpack.i.l.s16.s32

v[0..63], v[0..63]

vunpack.i.l.s8.bf16

v[0..63], v[0..63]

vunpack.i.l.s8.s16

v[0..63], v[0..63]

vunpack.i.l.u8.bf16

v[0..63], v[0..63]

vunpack.i.l.u8.u16

v[0..63], v[0..63]

vunpack.i.u.bf16.f32

v[0..63], v[0..63]

vunpack.i.u.s16.s32

v[0..63], v[0..63]

vunpack.i.u.s8.bf16

v[0..63], v[0..63]

vunpack.i.u.s8.s16

v[0..63], v[0..63]

vunpack.i.u.u8.bf16

v[0..63], v[0..63]

vunpack.i.u.u8.u16

v[0..63], v[0..63]

vweird.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vweird.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vxor.u32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]

va2#

助记符

按顺序排列的操作数类型

unpack_interleaved_compressed_lower_s8_to_bf16

dest=(u6), x=(u6)

unpack_interleaved_compressed_lower_s8_to_s16

dest=(u6), x=(u6)

unpack_interleaved_compressed_lower_u8_to_bf16

dest=(u6), x=(u6)

unpack_interleaved_compressed_lower_u8_to_u16

dest=(u6), x=(u6)

unpack_interleaved_compressed_upper_s8_to_bf16

dest=(u6), x=(u6)

unpack_interleaved_compressed_upper_s8_to_s16

dest=(u6), x=(u6)

unpack_interleaved_compressed_upper_u8_to_bf16

dest=(u6), x=(u6)

unpack_interleaved_compressed_upper_u8_to_u16

dest=(u6), x=(u6)

vadd.bf16

v[0..63], v[0..63], v[0..63]

vadd.f32

v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]
v[0..63], f32, v[0..63]

vadd.s16

v[0..63], v[0..63], v[0..63]

vadd.s32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]

vand.u32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]

vc.u16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vc.u32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vclass.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vclass.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vclz

v[0..63], v[0..63]

vcmask

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, s[0..31]

vector_add_high_f32_bf16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_add_low_f32_bf16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_gt_u16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_gte_u16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_lt_u16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

vector_lte_u16

dest=(u6), x=(u6), y_src=(vreg / lane_id / u32 / vs0 / vs1 / vs2 / vs3 / compressed_to_interleaved / interleaved_to_compressed / reserved_9 / reserved_10 / reserved_11 / reserved_12 / reserved_13 / reserved_14 / reserved_15 / reserved_16 / reserved_17 / reserved_18 / reserved_19 / reserved_20 / reserved_21 / reserved_22 / reserved_23 / reserved_24 / reserved_25 / reserved_26 / reserved_27 / reserved_28 / reserved_29 / reserved_30), y_vreg=(u6)

veq.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

veq.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

veq.s16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

veq.s32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vge.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vge.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vge.s16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vge.s32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vge.u32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vgt.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vgt.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vgt.s16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vgt.s32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vgt.u32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vimm.bf16

v[0..63], u32

vimm.f32

v[0..63], f32

vimm.hf16

v[0..63], u32

vimm.s32

v[0..63], u32

vlaneseq.c.b16

v[0..63]

vlaneseq.i.b16

v[0..63]

vlaneseq.u32

v[0..63]

vle.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vle.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vle.s16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vle.s32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vle.to.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vle.to.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vle.u32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vlt.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vlt.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vlt.s16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vlt.s32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vlt.to.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vlt.to.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vlt.u32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vmand

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vmax.bf16

v[0..63], v[0..63], v[0..63]

vmax.f32

v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]
v[0..63], v[0..63], f32

vmax.u16

v[0..63], v[0..63], v[0..63]

vmax.u32

v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]

vmin.bf16

v[0..63], v[0..63], v[0..63]

vmin.f32

v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]
v[0..63], v[0..63], f32

vmin.u16

v[0..63], v[0..63], v[0..63]

vmin.u32

v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]

vmmov

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vmneg

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vmor

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vmov

v[0..63], v[0..63]
v[0..63], s[0..31]

vmpackc.even

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vmpackc.low

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vmul.u16

v[0..63], v[0..63], v[0..63]

vmul.u32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]

vmxor

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vne.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vne.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], f32

vne.s16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vne.s32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], u32
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], s[0..31]

vnez.u8

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, s[0..31]

vnsel

v[0..63], vm[0..15], u32, v[0..63]
v[0..63], vm[0..15], s[0..31], v[0..63]

vor.u32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]

vpack.c.b16.b8

v[0..63], v[0..63], v[0..63]

vpack.c.b2.b1

v[0..63], v[0..63], v[0..63]

vpack.c.b32.b16

v[0..63], v[0..63], v[0..63]

vpack.c.b4.b2

v[0..63], v[0..63], v[0..63]

vpack.c.b8.b4

v[0..63], v[0..63], v[0..63]

vpack.c.bf16.s8

v[0..63], v[0..63], v[0..63]

vpack.c.bf16.u8

v[0..63], v[0..63], v[0..63]

vpack.c.f32.bf16

v[0..63], v[0..63], v[0..63]

vpack.i.b16.b8

v[0..63], v[0..63], v[0..63]

vpack.i.b32.b16

v[0..63], v[0..63], v[0..63]

vpack.i.b8.b4

v[0..63], v[0..63], v[0..63]

vpack.i.bf16.s8

v[0..63], v[0..63], v[0..63]

vpack.i.bf16.u8

v[0..63], v[0..63], v[0..63]

vpack.i.f32.bf16

v[0..63], v[0..63], v[0..63]

vpcnt

v[0..63], v[0..63]

vpsel

v[0..63], p5, u32, v[0..63]
v[0..63], p5, v[0..63], v[0..63]
v[0..63], !p5, s[0..31], v[0..63]

vpush

v2sf, v[0..63], u32
v2sf, v[0..63], s[0..31]

vsel

v[0..63], vm[0..15], u32, v[0..63]
v[0..63], vm[0..15], v[0..63], v[0..63]
v[0..63], vm[0..15], s[0..31], v[0..63]

vshll.u32

v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]

vshra.s32

v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]

vshrl.u32

v[0..63], v[0..63], u32
v[0..63], v[0..63], v[0..63]
v[0..63], v[0..63], s[0..31]

vsub.bf16

v[0..63], v[0..63], v[0..63]

vsub.f32

v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]
v[0..63], f32, v[0..63]

vsub.s16

v[0..63], v[0..63], v[0..63]

vsub.s32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]

vunpack.c.0.bf8.f32

v[0..63], v[0..63]

vunpack.c.0.if8.f32

v[0..63], v[0..63]

vunpack.c.0.s8.s32

v[0..63], v[0..63]

vunpack.c.1.bf8.f32

v[0..63], v[0..63]

vunpack.c.1.if8.f32

v[0..63], v[0..63]

vunpack.c.1.s8.s32

v[0..63], v[0..63]

vunpack.c.2.bf8.f32

v[0..63], v[0..63]

vunpack.c.2.if8.f32

v[0..63], v[0..63]

vunpack.c.2.s8.s32

v[0..63], v[0..63]

vunpack.c.3.bf8.f32

v[0..63], v[0..63]

vunpack.c.3.if8.f32

v[0..63], v[0..63]

vunpack.c.3.s8.s32

v[0..63], v[0..63]

vunpack.c.l.bf16.f32

v[0..63], v[0..63]

vunpack.c.l.hf16.f32

v[0..63], v[0..63]

vunpack.c.l.s1.s2

v[0..63], v[0..63]

vunpack.c.l.s16.s32

v[0..63], v[0..63]

vunpack.c.l.s2.s4

v[0..63], v[0..63]

vunpack.c.l.s4.s8

v[0..63], v[0..63]

vunpack.c.l.s8.bf16

v[0..63], v[0..63]

vunpack.c.l.s8.s16

v[0..63], v[0..63]

vunpack.c.l.u8.bf16

v[0..63], v[0..63]

vunpack.c.l.u8.u16

v[0..63], v[0..63]

vunpack.c.u.bf16.f32

v[0..63], v[0..63]

vunpack.c.u.hf16.f32

v[0..63], v[0..63]

vunpack.c.u.s1.s2

v[0..63], v[0..63]

vunpack.c.u.s16.s32

v[0..63], v[0..63]

vunpack.c.u.s2.s4

v[0..63], v[0..63]

vunpack.c.u.s4.s8

v[0..63], v[0..63]

vunpack.c.u.s8.bf16

v[0..63], v[0..63]

vunpack.c.u.s8.s16

v[0..63], v[0..63]

vunpack.c.u.u8.bf16

v[0..63], v[0..63]

vunpack.c.u.u8.u16

v[0..63], v[0..63]

vunpack.i.l.bf16.f32

v[0..63], v[0..63]

vunpack.i.l.s16.s32

v[0..63], v[0..63]

vunpack.i.l.s8.bf16

v[0..63], v[0..63]

vunpack.i.l.s8.s16

v[0..63], v[0..63]

vunpack.i.l.u8.bf16

v[0..63], v[0..63]

vunpack.i.l.u8.u16

v[0..63], v[0..63]

vunpack.i.u.bf16.f32

v[0..63], v[0..63]

vunpack.i.u.s16.s32

v[0..63], v[0..63]

vunpack.i.u.s8.bf16

v[0..63], v[0..63]

vunpack.i.u.s8.s16

v[0..63], v[0..63]

vunpack.i.u.u8.bf16

v[0..63], v[0..63]

vunpack.i.u.u8.u16

v[0..63], v[0..63]

vweird.bf16

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vweird.f32

vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vxor.u32

v[0..63], u32, v[0..63]
v[0..63], v[0..63], v[0..63]
v[0..63], s[0..31], v[0..63]

vld#

助记符

按顺序排列的操作数类型

vld

v[0..63], [tilespmem:u32]
v[0..63], [tilespmem:s[0..31]+u32]

vld.cb.msk

v[0..63], [tilespmem:s[0..31]+u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15
v[0..63], [tilespmem:s[0..31]+u32 ss:u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15
v[0..63], [tilespmem:s[0..31]+u32 ss:s[0..31] cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vld.cb.upd.msk

v[0..63], [tilespmem:s[0..31]+u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15
v[0..63], [tilespmem:s[0..31]+u32 ss:u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15
v[0..63], [tilespmem:s[0..31]+u32 ss:s[0..31] cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vld.idx.cb.msk

v[0..63], [tilespmem:v[0..63]+s[0..31]+u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vld.idx.msk

v[0..63], [tilespmem:v[0..63]+s[0..31]+u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15
v[0..63], [tilespmem:v[0..63]+s[0..31]+u32 ss:u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15
v[0..63], [tilespmem:v[0..63]+s[0..31]+u32 ss:s[0..31]], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vld.msk

v[0..63], [tilespmem:u32 ss:u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15
v[0..63], [tilespmem:u32 ss:s[0..31]], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15
v[0..63], [tilespmem:s[0..31]+u32 ss:u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15
v[0..63], [tilespmem:s[0..31]+u32 ss:s[0..31]], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15
v[0..63], [tilespmem:u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15
v[0..63], [tilespmem:s[0..31]+u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15

vst#

助记符

按顺序排列的操作数类型

tile_spmem_store_indexed_circular_buffer_return_value_add_bf16

source=(u6), base_address=(0x0 / vs0 / vs1 / vs2 / vs3 / reserved_0 / reserved_1 / reserved_2), cbreg=(u4), offset=(u32), stride=(vs0 / vs1 / vs2 / vs3 / u32), index=(u6), mask=(vmsk0 / vmsk1 / vmsk2 / vmsk3 / vmsk4 / vmsk5 / vmsk6 / vmsk7 / vmsk8 / vmsk9 / vmsk10 / vmsk11 / vmsk12 / vmsk13 / vmsk14 / vmsk15 / all_ones / 0x7f / 0x3f / 0x1f / 0xf / 0x7 / 0x3 / 0x1 / reserved_0 / reserved_1 / reserved_2 / reserved_3 / reserved_4 / reserved_5 / reserved_6 / reserved_7), dest=(u6)

tile_spmem_store_indexed_circular_buffer_return_value_add_f32

source=(u6), base_address=(0x0 / vs0 / vs1 / vs2 / vs3 / reserved_0 / reserved_1 / reserved_2), cbreg=(u4), offset=(u32), stride=(vs0 / vs1 / vs2 / vs3 / u32), index=(u6), mask=(vmsk0 / vmsk1 / vmsk2 / vmsk3 / vmsk4 / vmsk5 / vmsk6 / vmsk7 / vmsk8 / vmsk9 / vmsk10 / vmsk11 / vmsk12 / vmsk13 / vmsk14 / vmsk15 / all_ones / 0x7f / 0x3f / 0x1f / 0xf / 0x7 / 0x3 / 0x1 / reserved_0 / reserved_1 / reserved_2 / reserved_3 / reserved_4 / reserved_5 / reserved_6 / reserved_7), dest=(u6)

tile_spmem_store_indexed_circular_buffer_return_value_add_s16

source=(u6), base_address=(0x0 / vs0 / vs1 / vs2 / vs3 / reserved_0 / reserved_1 / reserved_2), cbreg=(u4), offset=(u32), stride=(vs0 / vs1 / vs2 / vs3 / u32), index=(u6), mask=(vmsk0 / vmsk1 / vmsk2 / vmsk3 / vmsk4 / vmsk5 / vmsk6 / vmsk7 / vmsk8 / vmsk9 / vmsk10 / vmsk11 / vmsk12 / vmsk13 / vmsk14 / vmsk15 / all_ones / 0x7f / 0x3f / 0x1f / 0xf / 0x7 / 0x3 / 0x1 / reserved_0 / reserved_1 / reserved_2 / reserved_3 / reserved_4 / reserved_5 / reserved_6 / reserved_7), dest=(u6)

tile_spmem_store_indexed_circular_buffer_return_value_add_s32

source=(u6), base_address=(0x0 / vs0 / vs1 / vs2 / vs3 / reserved_0 / reserved_1 / reserved_2), cbreg=(u4), offset=(u32), stride=(vs0 / vs1 / vs2 / vs3 / u32), index=(u6), mask=(vmsk0 / vmsk1 / vmsk2 / vmsk3 / vmsk4 / vmsk5 / vmsk6 / vmsk7 / vmsk8 / vmsk9 / vmsk10 / vmsk11 / vmsk12 / vmsk13 / vmsk14 / vmsk15 / all_ones / 0x7f / 0x3f / 0x1f / 0xf / 0x7 / 0x3 / 0x1 / reserved_0 / reserved_1 / reserved_2 / reserved_3 / reserved_4 / reserved_5 / reserved_6 / reserved_7), dest=(u6)

vst

[tilespmem:u32], v[0..63]
[tilespmem:s[0..31]+u32], v[0..63]

vst.add.bf16.msk

[tilespmem:s[0..31]+u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:s[0..31]], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.add.f32.msk

[tilespmem:s[0..31]+f32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+f32 ss:f32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+f32 ss:s[0..31]], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.add.s16.msk

[tilespmem:s[0..31]+u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:s[0..31]], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.add.s32.msk

[tilespmem:s[0..31]+u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:s[0..31]], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.cb.add.bf16.msk

[tilespmem:s[0..31]+u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:s[0..31] cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.cb.add.f32.msk

[tilespmem:s[0..31]+f32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+f32 ss:f32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+f32 ss:s[0..31] cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.cb.add.s16.msk

[tilespmem:s[0..31]+u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:s[0..31] cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.cb.add.s32.msk

[tilespmem:s[0..31]+u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:s[0..31] cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.cb.msk

[tilespmem:s[0..31]+u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:s[0..31] cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.cb.upd.add.bf16.msk

[tilespmem:s[0..31]+u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:s[0..31] cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.cb.upd.add.f32.msk

[tilespmem:s[0..31]+f32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+f32 ss:f32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+f32 ss:s[0..31] cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.cb.upd.add.s16.msk

[tilespmem:s[0..31]+u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:s[0..31] cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.cb.upd.add.s32.msk

[tilespmem:s[0..31]+u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:s[0..31] cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.cb.upd.msk

[tilespmem:s[0..31]+u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:s[0..31] cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.idx.add.bf16.msk

[tilespmem:v[0..63]+s[0..31]+u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.idx.add.f32.msk

[tilespmem:v[0..63]+s[0..31]+f32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.idx.add.s16.msk

[tilespmem:v[0..63]+s[0..31]+u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.idx.add.s32.msk

[tilespmem:v[0..63]+s[0..31]+u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.idx.cb.add.bf16.msk

[tilespmem:v[0..63]+s[0..31]+u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.idx.cb.add.f32.msk

[tilespmem:v[0..63]+s[0..31]+f32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.idx.cb.add.s16.msk

[tilespmem:v[0..63]+s[0..31]+u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.idx.cb.add.s32.msk

[tilespmem:v[0..63]+s[0..31]+u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.idx.cb.msk

[tilespmem:v[0..63]+s[0..31]+u32 cbreg:u4], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.idx.msk

[tilespmem:v[0..63]+s[0..31]+u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:v[0..63]+s[0..31]+u32 ss:u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:v[0..63]+s[0..31]+u32 ss:s[0..31]], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.idx.ret.add.bf16.msk

[tilespmem:v[0..63]+s[0..31]+u32], v[0..63], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.idx.ret.add.f32.msk

[tilespmem:v[0..63]+s[0..31]+f32], v[0..63], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.idx.ret.add.s16.msk

[tilespmem:v[0..63]+s[0..31]+u32], v[0..63], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.idx.ret.add.s32.msk

[tilespmem:v[0..63]+s[0..31]+u32], v[0..63], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vst.msk

[tilespmem:s[0..31]+u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:u32], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]
[tilespmem:s[0..31]+u32 ss:s[0..31]], vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

stream#

助记符

按顺序排列的操作数类型

indirect_vreg_stream

indirect_offsets=(u6), valid_offset_mask=(u6), off_tile_start_offset_valid=(u1), off_tile_start_offset=(u5), sync_flag_core_type=(tec_scs / tac), sync_flag_count_type=(word_4b / descriptor), set_done_bit=(u1), tile_local_stride=(stride_32b / stride_64b / stride_128b / stride_256b / stride_512b / stride_1024b / stride_2048b / no_stride), post_update_circular_buffer=(u1), indirect_list_type=(word_offset / row_offset), indirect_list_stride=(u6), indirect_filter_en=(u1), trace_en=(u1), stream_opcode=(gather / gather_integer_add / gather_float_add / reserved_0 / scatter / scatter_integer_add / scatter_float_add / reserved_1), tile_local_memory_type=(smem / tile_spmem), tile_local_stream_type=(linear / circular_buffer), off_tile_memory_type=(spmem / tile_spmem_n / hbm / hbm_4b), gather_scatter_add_is_b16=(u1), s0_x=(u5), s0_y=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32), s1_x=(u5), s1_y=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32)

stream.indirect.gather

[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.indirect.gather.add.bf16

[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.indirect.gather.add.f32

[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [spmem:s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], f32, s[0..31], s[0..31], u32

stream.indirect.gather.add.s16

[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.indirect.gather.add.s32

[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.indirect.gather.cb

[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.indirect.gather.cb.add.bf16

[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.indirect.gather.cb.add.f32

[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:f32], [spmem:s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:f32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], f32, s[0..31], s[0..31], u32

stream.indirect.gather.cb.add.s16

[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.indirect.gather.cb.add.s32

[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.indirect.scatter

[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.indirect.scatter.add.bf16

[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.indirect.scatter.add.f32

[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:f32], f32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:f32], f32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:f32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:f32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], f32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], f32, s[0..31], s[0..31], u32

stream.indirect.scatter.add.s16

[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.indirect.scatter.add.s32

[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.indirect.scatter.cb

[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.indirect.scatter.cb.add.bf16

[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.indirect.scatter.cb.add.f32

[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:f32], f32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:f32], f32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:f32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:f32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], f32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], f32, s[0..31], s[0..31], u32

stream.indirect.scatter.cb.add.s16

[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.indirect.scatter.cb.add.s32

[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.linear.gather

[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], s[0..31], u32

stream.linear.gather.add.bf16

[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32

stream.linear.gather.add.f32

[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], f32, u32
[tilespmem:s[0..31]], [sflag:f32], [spmem:s[0..31]], f32, u32
[tilespmem:s[0..31]], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], f32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], f32, u32

stream.linear.gather.add.s16

[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32

stream.linear.gather.add.s32

[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32

stream.linear.gather.cb

[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]], s[0..31], u32

stream.linear.gather.cb.add.bf16

[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32

stream.linear.gather.cb.add.f32

[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], f32, u32
[tilespmem:u5], [sflag:f32], [spmem:s[0..31]], f32, u32
[tilespmem:u5], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:f32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], f32, u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], f32, u32

stream.linear.gather.cb.add.s16

[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32

stream.linear.gather.cb.add.s32

[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32

stream.linear.scatter

[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32

stream.linear.scatter.add.bf16

[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32

stream.linear.scatter.add.f32

[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:f32], f32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:f32], f32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:f32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:f32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], f32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], f32, u32

stream.linear.scatter.add.s16

[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32

stream.linear.scatter.add.s32

[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32

stream.linear.scatter.cb

[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32

stream.linear.scatter.cb.add.bf16

[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32

stream.linear.scatter.cb.add.f32

[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:f32], f32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:f32], f32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:f32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:f32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], f32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], f32, u32

stream.linear.scatter.cb.add.s16

[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32

stream.linear.scatter.cb.add.s32

[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32

stream.strided.gather

[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.strided.gather.add.bf16

[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.strided.gather.add.f32

[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [spmem:s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], f32, s[0..31], s[0..31], u32

stream.strided.gather.add.s16

[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.strided.gather.add.s32

[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.strided.gather.cb

[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.strided.gather.cb.add.bf16

[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.strided.gather.cb.add.f32

[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:f32], [spmem:s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:f32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], f32, s[0..31], s[0..31], u32

stream.strided.gather.cb.add.s16

[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.strided.gather.cb.add.s32

[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.strided.scatter

[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.strided.scatter.add.bf16

[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.strided.scatter.add.f32

[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:f32], f32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:f32], f32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:f32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:f32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], f32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], f32, s[0..31], s[0..31], u32

stream.strided.scatter.add.s16

[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.strided.scatter.add.s32

[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.strided.scatter.cb

[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.strided.scatter.cb.add.bf16

[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.strided.scatter.cb.add.f32

[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:f32], f32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:f32], f32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:f32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:f32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], f32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], f32, s[0..31], s[0..31], u32

stream.strided.scatter.cb.add.s16

[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32

stream.strided.scatter.cb.add.s32

[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32

vr#

助记符

按顺序排列的操作数类型

pop_xrf_write_partial_0

unit=(xrf0 / xrf1), dest_one=(u6)

pop_xrf_write_partial_1

unit=(xrf0 / xrf1), dest_one=(u6), dest_vm=(u4)

pop_xrf_write_partial_2

unit=(xrf0 / xrf1), dest_two=(u6)

pop_xrf_write_partial_3

unit=(xrf0 / xrf1), dest_two=(u6), dest_vm=(u4)

pop_xrf_write_partial_4

unit=(xrf0 / xrf1), dest_one=(u6), dest_two=(u6)

vmovc

v[0..63], v[0..63]

vpop

v[0..63], erf
v[0..63], v[0..63], vm[0..15], xrf

vx#

助记符

按顺序排列的操作数类型

max_index_scan_bf16

source_one=(vst_source / v0_y_vreg / v0_x / v1_y_vreg / v1_x / v2_y_vreg / v2_x / lane_id), vst_source=(u6), v0_y_vreg=(u6), v0_x=(u6), v1_y_vreg=(u6), v1_x=(u6), v2_y_vreg=(u6), v2_x=(u6), vmask=(vmsk0 / vmsk1 / vmsk2 / vmsk3 / vmsk4 / vmsk5 / vmsk6 / vmsk7 / vmsk8 / vmsk9 / vmsk10 / vmsk11 / vmsk12 / vmsk13 / vmsk14 / vmsk15 / all_ones / 0x7f / 0x3f / 0x1f / 0xf / 0x7 / 0x3 / 0x1 / reserved_0 / reserved_1 / reserved_2 / reserved_3 / reserved_4 / reserved_5 / reserved_6 / reserved_7)

max_index_scan_u16

source_one=(vst_source / v0_y_vreg / v0_x / v1_y_vreg / v1_x / v2_y_vreg / v2_x / lane_id), vst_source=(u6), v0_y_vreg=(u6), v0_x=(u6), v1_y_vreg=(u6), v1_x=(u6), v2_y_vreg=(u6), v2_x=(u6), vmask=(vmsk0 / vmsk1 / vmsk2 / vmsk3 / vmsk4 / vmsk5 / vmsk6 / vmsk7 / vmsk8 / vmsk9 / vmsk10 / vmsk11 / vmsk12 / vmsk13 / vmsk14 / vmsk15 / all_ones / 0x7f / 0x3f / 0x1f / 0xf / 0x7 / 0x3 / 0x1 / reserved_0 / reserved_1 / reserved_2 / reserved_3 / reserved_4 / reserved_5 / reserved_6 / reserved_7)

min_index_scan_bf16

source_one=(vst_source / v0_y_vreg / v0_x / v1_y_vreg / v1_x / v2_y_vreg / v2_x / lane_id), vst_source=(u6), v0_y_vreg=(u6), v0_x=(u6), v1_y_vreg=(u6), v1_x=(u6), v2_y_vreg=(u6), v2_x=(u6), vmask=(vmsk0 / vmsk1 / vmsk2 / vmsk3 / vmsk4 / vmsk5 / vmsk6 / vmsk7 / vmsk8 / vmsk9 / vmsk10 / vmsk11 / vmsk12 / vmsk13 / vmsk14 / vmsk15 / all_ones / 0x7f / 0x3f / 0x1f / 0xf / 0x7 / 0x3 / 0x1 / reserved_0 / reserved_1 / reserved_2 / reserved_3 / reserved_4 / reserved_5 / reserved_6 / reserved_7)

segmented_min_index_scan_bf16

source_one=(vst_source / v0_y_vreg / v0_x / v1_y_vreg / v1_x / v2_y_vreg / v2_x / lane_id), vst_source=(u6), v0_y_vreg=(u6), v0_x=(u6), v1_y_vreg=(u6), v1_x=(u6), v2_y_vreg=(u6), v2_x=(u6), vmask=(vmsk0 / vmsk1 / vmsk2 / vmsk3 / vmsk4 / vmsk5 / vmsk6 / vmsk7 / vmsk8 / vmsk9 / vmsk10 / vmsk11 / vmsk12 / vmsk13 / vmsk14 / vmsk15 / all_ones / 0x7f / 0x3f / 0x1f / 0xf / 0x7 / 0x3 / 0x1 / reserved_0 / reserved_1 / reserved_2 / reserved_3 / reserved_4 / reserved_5 / reserved_6 / reserved_7)

segmented_min_scan_bf16

source_one=(vst_source / v0_y_vreg / v0_x / v1_y_vreg / v1_x / v2_y_vreg / v2_x / lane_id), vst_source=(u6), v0_y_vreg=(u6), v0_x=(u6), v1_y_vreg=(u6), v1_x=(u6), v2_y_vreg=(u6), v2_x=(u6), vmask=(vmsk0 / vmsk1 / vmsk2 / vmsk3 / vmsk4 / vmsk5 / vmsk6 / vmsk7 / vmsk8 / vmsk9 / vmsk10 / vmsk11 / vmsk12 / vmsk13 / vmsk14 / vmsk15 / all_ones / 0x7f / 0x3f / 0x1f / 0xf / 0x7 / 0x3 / 0x1 / reserved_0 / reserved_1 / reserved_2 / reserved_3 / reserved_4 / reserved_5 / reserved_6 / reserved_7)

vadd.scan.msk.bf16

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vadd.scan.msk.bf16.f32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vadd.scan.msk.f32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vadd.scan.msk.s16

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vadd.scan.msk.s16.s32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vadd.scan.msk.s32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vadd.seg.scan.bf16

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vadd.seg.scan.bf16.f32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vadd.seg.scan.f32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vadd.seg.scan.s16

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vadd.seg.scan.s16.s32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vadd.seg.scan.s32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vdupcnt.msk.f32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vdupcnt.msk.u32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmax.index.scan.f32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmax.index.scan.u32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmax.scan.msk.bf16

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmax.scan.msk.f32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmax.scan.msk.u16

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmax.scan.msk.u32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmax.seg.index.scan.bf16

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmax.seg.index.scan.f32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmax.seg.index.scan.u16

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmax.seg.index.scan.u32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmax.seg.scan.bf16

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmax.seg.scan.f32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmax.seg.scan.u16

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmax.seg.scan.u32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmin.index.scan.f32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmin.index.scan.u16

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmin.index.scan.u32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmin.scan.msk.bf16

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmin.scan.msk.f32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmin.scan.msk.u16

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmin.scan.msk.u32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmin.seg.index.scan.f32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmin.seg.index.scan.u16

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmin.seg.index.scan.u32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmin.seg.scan.f32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmin.seg.scan.u16

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vmin.seg.scan.u32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vsort.ascd.msk.f32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]

vsort.ascd.msk.u32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]

vsort.dscd.msk.f32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]

vsort.dscd.msk.u32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63], v[0..63]

vunique.msk.f32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]

vunique.msk.u32

xrf, vm0 / vm1 / vm2 / vm3 / vm4 / vm5 / vm6 / vm7 / vm8 / vm9 / vm10 / vm11 / vm12 / vm13 / vm14 / vm15, v[0..63]