# RUN: llvm-mc -arch=amdgcn -mcpu=gfx940 -disassemble -show-encoding %s | FileCheck -strict-whitespace --check-prefix=GFX940 %s

# GFX940: global_load_dword v2, v[2:3], off sc0   ; encoding: [0x00,0x80,0x51,0xdc,0x02,0x00,0x7f,0x02]
0x00,0x80,0x51,0xdc,0x02,0x00,0x7f,0x02

# GFX940: global_load_dword v2, v[2:3], off sc1   ; encoding: [0x00,0x80,0x50,0xde,0x02,0x00,0x7f,0x02]
0x00,0x80,0x50,0xde,0x02,0x00,0x7f,0x02

# GFX940: global_load_dword v2, v[2:3], off nt    ; encoding: [0x00,0x80,0x52,0xdc,0x02,0x00,0x7f,0x02]
0x00,0x80,0x52,0xdc,0x02,0x00,0x7f,0x02

# GFX940: s_load_dword s2, s[2:3], 0x0 glc        ; encoding: [0x81,0x00,0x03,0xc0,0x00,0x00,0x00,0x00]
0x81,0x00,0x03,0xc0,0x00,0x00,0x00,0x00

# GFX940: buffer_load_dword v5, off, s[8:11], s3 sc0 nt sc1 ; encoding: [0x00,0xc0,0x52,0xe0,0x00,0x05,0x02,0x03]
0x00,0xc0,0x52,0xe0,0x00,0x05,0x02,0x03

# GFX940: flat_atomic_add_f32 v[2:3], v1          ; encoding: [0x00,0x00,0x34,0xdd,0x02,0x01,0x00,0x00]
0x00,0x00,0x34,0xdd,0x02,0x01,0x00,0x00

# GFX940: flat_atomic_add_f32 v[2:3], a1          ; encoding: [0x00,0x00,0x34,0xdd,0x02,0x01,0x80,0x00]
0x00,0x00,0x34,0xdd,0x02,0x01,0x80,0x00

# GFX940: flat_atomic_add_f32 v4, v[2:3], v1 sc0  ; encoding: [0x00,0x00,0x35,0xdd,0x02,0x01,0x00,0x04]
0x00,0x00,0x35,0xdd,0x02,0x01,0x00,0x04

# GFX940: flat_atomic_add_f32 a4, v[2:3], a1 sc0  ; encoding: [0x00,0x00,0x35,0xdd,0x02,0x01,0x80,0x04]
0x00,0x00,0x35,0xdd,0x02,0x01,0x80,0x04

# GFX940: flat_atomic_pk_add_f16 v4, v[2:3], v1 sc0 ; encoding: [0x00,0x00,0x39,0xdd,0x02,0x01,0x00,0x04]
0x00,0x00,0x39,0xdd,0x02,0x01,0x00,0x04

# GFX940: flat_atomic_pk_add_f16 a4, v[2:3], a1 sc0 ; encoding: [0x00,0x00,0x39,0xdd,0x02,0x01,0x80,0x04]
0x00,0x00,0x39,0xdd,0x02,0x01,0x80,0x04

# GFX940: flat_atomic_pk_add_f16 v[2:3], v1       ; encoding: [0x00,0x00,0x38,0xdd,0x02,0x01,0x00,0x00]
0x00,0x00,0x38,0xdd,0x02,0x01,0x00,0x00

# GFX940: flat_atomic_pk_add_f16 v[2:3], a1       ; encoding: [0x00,0x00,0x38,0xdd,0x02,0x01,0x80,0x00]
0x00,0x00,0x38,0xdd,0x02,0x01,0x80,0x00

# GFX940: flat_atomic_pk_add_bf16 v4, v[2:3], v1 sc0 ; encoding: [0x00,0x00,0x49,0xdd,0x02,0x01,0x00,0x04]
0x00,0x00,0x49,0xdd,0x02,0x01,0x00,0x04

# GFX940: flat_atomic_pk_add_bf16 a4, v[2:3], a1 sc0 ; encoding: [0x00,0x00,0x49,0xdd,0x02,0x01,0x80,0x04]
0x00,0x00,0x49,0xdd,0x02,0x01,0x80,0x04

# GFX940: flat_atomic_pk_add_bf16 v[2:3], v1      ; encoding: [0x00,0x00,0x48,0xdd,0x02,0x01,0x00,0x00]
0x00,0x00,0x48,0xdd,0x02,0x01,0x00,0x00

# GFX940: flat_atomic_pk_add_bf16 v[2:3], a1      ; encoding: [0x00,0x00,0x48,0xdd,0x02,0x01,0x80,0x00]
0x00,0x00,0x48,0xdd,0x02,0x01,0x80,0x00

# GFX940: global_atomic_pk_add_bf16 v4, v[2:3], v1, off sc0 ; encoding: [0x00,0x80,0x49,0xdd,0x02,0x01,0x7f,0x04]
0x00,0x80,0x49,0xdd,0x02,0x01,0x7f,0x04

# GFX940: global_atomic_pk_add_bf16 a4, v[2:3], a1, off sc0 ; encoding: [0x00,0x80,0x49,0xdd,0x02,0x01,0xff,0x04]
0x00,0x80,0x49,0xdd,0x02,0x01,0xff,0x04

# GFX940: global_atomic_pk_add_bf16 v[2:3], v1, off ; encoding: [0x00,0x80,0x48,0xdd,0x02,0x01,0x7f,0x00]
0x00,0x80,0x48,0xdd,0x02,0x01,0x7f,0x00

# GFX940: global_atomic_pk_add_bf16 v[2:3], a1, off ; encoding: [0x00,0x80,0x48,0xdd,0x02,0x01,0xff,0x00]
0x00,0x80,0x48,0xdd,0x02,0x01,0xff,0x00

# GFX940: ds_pk_add_f16 v2, v1                    ; encoding: [0x00,0x00,0x2e,0xd8,0x02,0x01,0x00,0x00]
0x00,0x00,0x2e,0xd8,0x02,0x01,0x00,0x00

# GFX940: ds_pk_add_f16 v2, a1                    ; encoding: [0x00,0x00,0x2e,0xda,0x02,0x01,0x00,0x00]
0x00,0x00,0x2e,0xda,0x02,0x01,0x00,0x00

# GFX940: ds_pk_add_rtn_f16 v3, v2, v1            ; encoding: [0x00,0x00,0x6e,0xd9,0x02,0x01,0x00,0x03]
0x00,0x00,0x6e,0xd9,0x02,0x01,0x00,0x03

# GFX940: ds_pk_add_rtn_f16 a3, v2, a1            ; encoding: [0x00,0x00,0x6e,0xdb,0x02,0x01,0x00,0x03]
0x00,0x00,0x6e,0xdb,0x02,0x01,0x00,0x03

# GFX940: ds_pk_add_bf16 v2, v1                   ; encoding: [0x00,0x00,0x30,0xd8,0x02,0x01,0x00,0x00]
0x00,0x00,0x30,0xd8,0x02,0x01,0x00,0x00

# GFX940: ds_pk_add_bf16 v2, a1                   ; encoding: [0x00,0x00,0x30,0xda,0x02,0x01,0x00,0x00]
0x00,0x00,0x30,0xda,0x02,0x01,0x00,0x00

# GFX940: ds_pk_add_rtn_bf16 v3, v2, v1           ; encoding: [0x00,0x00,0x70,0xd9,0x02,0x01,0x00,0x03]
0x00,0x00,0x70,0xd9,0x02,0x01,0x00,0x03

# GFX940: ds_pk_add_rtn_bf16 a3, v2, a1           ; encoding: [0x00,0x00,0x70,0xdb,0x02,0x01,0x00,0x03]
0x00,0x00,0x70,0xdb,0x02,0x01,0x00,0x03

# GFX940: global_load_lds_dword v[2:3], off       ; encoding: [0x00,0x80,0xa8,0xdc,0x02,0x00,0x7f,0x00]
0x00,0x80,0xa8,0xdc,0x02,0x00,0x7f,0x00

# GFX940: global_load_lds_dword v[2:3], off sc0 nt sc1 ; encoding: [0x00,0x80,0xab,0xde,0x02,0x00,0x7f,0x00]
0x00,0x80,0xab,0xde,0x02,0x00,0x7f,0x00

# GFX940: global_load_lds_dword v[2:3], off offset:4 ; encoding: [0x04,0x80,0xa8,0xdc,0x02,0x00,0x7f,0x00]
0x04,0x80,0xa8,0xdc,0x02,0x00,0x7f,0x00

# GFX940: global_load_lds_dword v2, s[4:5] offset:4 ; encoding: [0x04,0x80,0xa8,0xdc,0x02,0x00,0x04,0x00]
0x04,0x80,0xa8,0xdc,0x02,0x00,0x04,0x00

# GFX940: global_load_lds_ubyte v[2:3], off       ; encoding: [0x00,0x80,0x98,0xdc,0x02,0x00,0x7f,0x00]
0x00,0x80,0x98,0xdc,0x02,0x00,0x7f,0x00

# GFX940: global_load_lds_sbyte v[2:3], off       ; encoding: [0x00,0x80,0x9c,0xdc,0x02,0x00,0x7f,0x00]
0x00,0x80,0x9c,0xdc,0x02,0x00,0x7f,0x00

# GFX940: global_load_lds_sshort v[2:3], off      ; encoding: [0x00,0x80,0xa4,0xdc,0x02,0x00,0x7f,0x00]
0x00,0x80,0xa4,0xdc,0x02,0x00,0x7f,0x00

# GFX940: global_load_lds_ushort v[2:3], off      ; encoding: [0x00,0x80,0xa0,0xdc,0x02,0x00,0x7f,0x00]
0x00,0x80,0xa0,0xdc,0x02,0x00,0x7f,0x00

# GFX940: scratch_load_lds_dword v2, off          ; encoding: [0x00,0x60,0xa8,0xdc,0x02,0x00,0x7f,0x00]
0x00,0x60,0xa8,0xdc,0x02,0x00,0x7f,0x00

# GFX940: scratch_load_lds_dword v2, s4           ; encoding: [0x00,0x60,0xa8,0xdc,0x02,0x00,0x04,0x00]
0x00,0x60,0xa8,0xdc,0x02,0x00,0x04,0x00

# GFX940: scratch_load_lds_dword v2, s4 offset:4  ; encoding: [0x04,0x60,0xa8,0xdc,0x02,0x00,0x04,0x00]
0x04,0x60,0xa8,0xdc,0x02,0x00,0x04,0x00

# GFX940: scratch_load_lds_dword off, s4 offset:4 ; encoding: [0x04,0x40,0xa8,0xdc,0x00,0x00,0x04,0x00]
0x04,0x40,0xa8,0xdc,0x00,0x00,0x04,0x00

# GFX940: scratch_load_lds_dword off, off offset:4 ; encoding: [0x04,0x40,0xa8,0xdc,0x00,0x00,0x7f,0x00]
0x04,0x40,0xa8,0xdc,0x00,0x00,0x7f,0x00

# GFX940: scratch_load_lds_dword off, off offset:4 ; encoding: [0x04,0x40,0xa8,0xdc,0x00,0x00,0x7f,0x00]
0x04,0x40,0xa8,0xdc,0x00,0x00,0x7f,0x00

# GFX940: scratch_load_lds_ubyte v2, off          ; encoding: [0x00,0x60,0x98,0xdc,0x02,0x00,0x7f,0x00]
0x00,0x60,0x98,0xdc,0x02,0x00,0x7f,0x00

# GFX940: scratch_load_lds_sbyte v2, off          ; encoding: [0x00,0x60,0x9c,0xdc,0x02,0x00,0x7f,0x00]
0x00,0x60,0x9c,0xdc,0x02,0x00,0x7f,0x00

# GFX940: scratch_load_lds_ushort v2, off         ; encoding: [0x00,0x60,0xa0,0xdc,0x02,0x00,0x7f,0x00]
0x00,0x60,0xa0,0xdc,0x02,0x00,0x7f,0x00

# GFX940: scratch_load_lds_sshort v2, off         ; encoding: [0x00,0x60,0xa4,0xdc,0x02,0x00,0x7f,0x00]
0x00,0x60,0xa4,0xdc,0x02,0x00,0x7f,0x00

# GFX940: s_getreg_b32 s1, hwreg(HW_REG_XCC_ID)   ; encoding: [0x14,0xf8,0x81,0xb8]
0x14,0xf8,0x81,0xb8

# GFX940: s_getreg_b32 s1, hwreg(HW_REG_SQ_PERF_SNAPSHOT_DATA) ; encoding: [0x15,0xf8,0x81,0xb8]
0x15,0xf8,0x81,0xb8

# GFX940: s_getreg_b32 s1, hwreg(HW_REG_SQ_PERF_SNAPSHOT_DATA1) ; encoding: [0x16,0xf8,0x81,0xb8]
0x16,0xf8,0x81,0xb8

# GFX940: s_getreg_b32 s1, hwreg(HW_REG_SQ_PERF_SNAPSHOT_PC_LO) ; encoding: [0x17,0xf8,0x81,0xb8]
0x17,0xf8,0x81,0xb8

# GFX940: s_getreg_b32 s1, hwreg(HW_REG_SQ_PERF_SNAPSHOT_PC_HI) ; encoding: [0x18,0xf8,0x81,0xb8]
0x18,0xf8,0x81,0xb8

# GFX940: v_mov_b64_e32 v[2:3], v[4:5]            ; encoding: [0x04,0x71,0x04,0x7e]
0x04,0x71,0x04,0x7e

# GFX940: v_mov_b64_dpp v[2:3], v[4:5] row_newbcast:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0x70,0x04,0x7e,0x04,0x51,0x01,0xff]
0xfa,0x70,0x04,0x7e,0x04,0x51,0x01,0xff

# GFX940: v_mov_b64_e32 v[2:3], s[4:5]            ; encoding: [0x04,0x70,0x04,0x7e]
0x04,0x70,0x04,0x7e

# GFX940: v_mov_b64_e32 v[2:3], 1                 ; encoding: [0x81,0x70,0x04,0x7e]
0x81,0x70,0x04,0x7e

# GFX940: v_mov_b64_e32 v[2:3], 0x64              ; encoding: [0xff,0x70,0x04,0x7e,0x64,0x00,0x00,0x00]
0xff,0x70,0x04,0x7e,0x64,0x00,0x00,0x00

# GFX940: v_lshl_add_u64 v[2:3], s[4:5], v7, v[8:9] ; encoding: [0x02,0x00,0x08,0xd2,0x04,0x0e,0x22,0x04]
0x02,0x00,0x08,0xd2,0x04,0x0e,0x22,0x04

# GFX940: v_lshl_add_u64 v[2:3], v[4:5], 0, 1     ; encoding: [0x02,0x00,0x08,0xd2,0x04,0x01,0x05,0x02]
0x02,0x00,0x08,0xd2,0x04,0x01,0x05,0x02

# GFX940: v_lshl_add_u64 v[2:3], v[4:5], 3, s[2:3] ; encoding: [0x02,0x00,0x08,0xd2,0x04,0x07,0x09,0x00]
0x02,0x00,0x08,0xd2,0x04,0x07,0x09,0x00

# GFX940: v_lshl_add_u64 v[2:3], s[4:5], 4, v[2:3] ; encoding: [0x02,0x00,0x08,0xd2,0x04,0x08,0x09,0x04]
0x02,0x00,0x08,0xd2,0x04,0x08,0x09,0x04

# GFX940: buffer_wbl2 sc1                         ; encoding: [0x00,0x80,0xa0,0xe0,0x00,0x00,0x00,0x00]
0x00,0x80,0xa0,0xe0,0x00,0x00,0x00,0x00

# GFX940: buffer_wbl2 sc0                         ; encoding: [0x00,0x40,0xa0,0xe0,0x00,0x00,0x00,0x00]
0x00,0x40,0xa0,0xe0,0x00,0x00,0x00,0x00

# GFX940: buffer_wbl2 sc0 sc1                     ; encoding: [0x00,0xc0,0xa0,0xe0,0x00,0x00,0x00,0x00]
0x00,0xc0,0xa0,0xe0,0x00,0x00,0x00,0x00

# GFX940: buffer_inv sc0                          ; encoding: [0x00,0x40,0xa4,0xe0,0x00,0x00,0x00,0x00]
0x00,0x40,0xa4,0xe0,0x00,0x00,0x00,0x00

# GFX940: buffer_inv sc1                          ; encoding: [0x00,0x80,0xa4,0xe0,0x00,0x00,0x00,0x00]
0x00,0x80,0xa4,0xe0,0x00,0x00,0x00,0x00

# GFX940: buffer_inv sc0 sc1                      ; encoding: [0x00,0xc0,0xa4,0xe0,0x00,0x00,0x00,0x00]
0x00,0xc0,0xa4,0xe0,0x00,0x00,0x00,0x00

# GFX940: buffer_atomic_swap v5, off, s[8:11], s3 sc0 ; encoding: [0x00,0x40,0x00,0xe1,0x00,0x05,0x02,0x03]
0x00,0x40,0x00,0xe1,0x00,0x05,0x02,0x03

# GFX940: buffer_atomic_swap v5, off, s[8:11], s3 nt ; encoding: [0x00,0x00,0x02,0xe1,0x00,0x05,0x02,0x03]
0x00,0x00,0x02,0xe1,0x00,0x05,0x02,0x03

# GFX940: v_fmamk_f32 v0, v2, 0x42c80000, v3      ; encoding: [0x02,0x07,0x00,0x2e,0x00,0x00,0xc8,0x42]
0x02,0x07,0x00,0x2e,0x00,0x00,0xc8,0x42

# GFX940: v_fmaak_f32 v0, v2, v3, 0x42c80000      ; encoding: [0x02,0x07,0x00,0x30,0x00,0x00,0xc8,0x42]
0x02,0x07,0x00,0x30,0x00,0x00,0xc8,0x42

# GFX940: global_atomic_add_f32 v0, v[0:1], v2, off sc0 sc1 ; encoding: [0x00,0x80,0x35,0xdf,0x00,0x02,0x7f,0x00]
0x00,0x80,0x35,0xdf,0x00,0x02,0x7f,0x00

# GFX940: global_atomic_add_f32 v[0:1], v2, off sc1 ; encoding: [0x00,0x80,0x34,0xdf,0x00,0x02,0x7f,0x00]
0x00,0x80,0x34,0xdf,0x00,0x02,0x7f,0x00

# GFX940: global_atomic_add_f32 v0, v2, s[0:1] sc1 ; encoding: [0x00,0x80,0x34,0xdf,0x00,0x02,0x00,0x00]
0x00,0x80,0x34,0xdf,0x00,0x02,0x00,0x00

# GFX940: global_atomic_add_f32 v1, v0, v2, s[0:1] sc0 sc1 ; encoding: [0x00,0x80,0x35,0xdf,0x00,0x02,0x00,0x01]
0x00,0x80,0x35,0xdf,0x00,0x02,0x00,0x01

# GFX940: global_atomic_pk_add_f16 v0, v[0:1], v2, off sc0 sc1 ; encoding: [0x00,0x80,0x39,0xdf,0x00,0x02,0x7f,0x00]
0x00,0x80,0x39,0xdf,0x00,0x02,0x7f,0x00

# GFX940: flat_atomic_add_f64 v[0:1], v[0:1], v[2:3] sc0 sc1 ; encoding: [0x00,0x00,0x3d,0xdf,0x00,0x02,0x00,0x00]
0x00,0x00,0x3d,0xdf,0x00,0x02,0x00,0x00

# GFX940: flat_atomic_add_f64 v[0:1], v[2:3] sc1  ; encoding: [0x00,0x00,0x3c,0xdf,0x00,0x02,0x00,0x00]
0x00,0x00,0x3c,0xdf,0x00,0x02,0x00,0x00

# GFX940: flat_atomic_min_f64 v[0:1], v[2:3] sc1  ; encoding: [0x00,0x00,0x40,0xdf,0x00,0x02,0x00,0x00]
0x00,0x00,0x40,0xdf,0x00,0x02,0x00,0x00

# GFX940: flat_atomic_max_f64 v[0:1], v[2:3] sc1  ; encoding: [0x00,0x00,0x44,0xdf,0x00,0x02,0x00,0x00]
0x00,0x00,0x44,0xdf,0x00,0x02,0x00,0x00

# GFX940: global_atomic_add_f64 v[0:1], v[2:3], off sc1 ; encoding: [0x00,0x80,0x3c,0xdf,0x00,0x02,0x7f,0x00]
0x00,0x80,0x3c,0xdf,0x00,0x02,0x7f,0x00

# GFX940: global_atomic_min_f64 v[0:1], v[2:3], off sc1 ; encoding: [0x00,0x80,0x40,0xdf,0x00,0x02,0x7f,0x00]
0x00,0x80,0x40,0xdf,0x00,0x02,0x7f,0x00

# GFX940: global_atomic_max_f64 v[0:1], v[2:3], off sc1 ; encoding: [0x00,0x80,0x44,0xdf,0x00,0x02,0x7f,0x00]
0x00,0x80,0x44,0xdf,0x00,0x02,0x7f,0x00

# GFX940: buffer_atomic_add_f32 v4, off, s[8:11], s3 sc1 ; encoding: [0x00,0x80,0x34,0xe1,0x00,0x04,0x02,0x03]
0x00,0x80,0x34,0xe1,0x00,0x04,0x02,0x03

# GFX940: buffer_atomic_pk_add_f16 v4, off, s[8:11], s3 sc1 ; encoding: [0x00,0x80,0x38,0xe1,0x00,0x04,0x02,0x03]
0x00,0x80,0x38,0xe1,0x00,0x04,0x02,0x03

# GFX940: buffer_atomic_add_f64 v[4:5], off, s[8:11], s3 sc1 ; encoding: [0x00,0x80,0x3c,0xe1,0x00,0x04,0x02,0x03]
0x00,0x80,0x3c,0xe1,0x00,0x04,0x02,0x03

# GFX940: buffer_atomic_max_f64 v[4:5], off, s[8:11], s3 sc1 ; encoding: [0x00,0x80,0x44,0xe1,0x00,0x04,0x02,0x03]
0x00,0x80,0x44,0xe1,0x00,0x04,0x02,0x03

# GFX940: buffer_atomic_min_f64 v[4:5], off, s[8:11], s3 sc1 ; encoding: [0x00,0x80,0x40,0xe1,0x00,0x04,0x02,0x03]
0x00,0x80,0x40,0xe1,0x00,0x04,0x02,0x03

# GFX940: v_cvt_f32_bf8_e32 v1, s3                ; encoding: [0x03,0xaa,0x02,0x7e]
0x03,0xaa,0x02,0x7e

# GFX940: v_cvt_f32_bf8_e32 v1, 3                 ; encoding: [0x83,0xaa,0x02,0x7e]
0x83,0xaa,0x02,0x7e

# GFX940: v_cvt_f32_bf8_e32 v1, v3                ; encoding: [0x03,0xab,0x02,0x7e]
0x03,0xab,0x02,0x7e

# GFX940: v_cvt_f32_bf8_sdwa v1, s3 src0_sel:BYTE_1 ; encoding: [0xf9,0xaa,0x02,0x7e,0x03,0x06,0x81,0x00]
0xf9,0xaa,0x02,0x7e,0x03,0x06,0x81,0x00

# GFX940: v_cvt_f32_bf8_dpp v1, v3 quad_perm:[0,2,1,1] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xaa,0x02,0x7e,0x03,0x58,0x00,0xff]
0xfa,0xaa,0x02,0x7e,0x03,0x58,0x00,0xff

# GFX940: v_cvt_f32_bf8_e64 v1, s3 mul:2          ; encoding: [0x01,0x00,0x95,0xd1,0x03,0x00,0x00,0x08]
0x01,0x00,0x95,0xd1,0x03,0x00,0x00,0x08

# GFX940: v_cvt_f32_bf8_sdwa v1, s3 clamp mul:2 src0_sel:BYTE_1 ; encoding: [0xf9,0xaa,0x02,0x7e,0x03,0x66,0x81,0x00]
0xf9,0xaa,0x02,0x7e,0x03,0x66,0x81,0x00

# GFX940: v_cvt_f32_bf8_e64 v1, s3 clamp          ; encoding: [0x01,0x80,0x95,0xd1,0x03,0x00,0x00,0x00]
0x01,0x80,0x95,0xd1,0x03,0x00,0x00,0x00

# GFX940: v_cvt_f32_fp8_e32 v1, s3                ; encoding: [0x03,0xa8,0x02,0x7e]
0x03,0xa8,0x02,0x7e

# GFX940: v_cvt_f32_fp8_e32 v1, 3                 ; encoding: [0x83,0xa8,0x02,0x7e]
0x83,0xa8,0x02,0x7e

# GFX940: v_cvt_f32_fp8_e32 v1, v3                ; encoding: [0x03,0xa9,0x02,0x7e]
0x03,0xa9,0x02,0x7e

# GFX940: v_cvt_f32_fp8_sdwa v1, s3 src0_sel:BYTE_1 ; encoding: [0xf9,0xa8,0x02,0x7e,0x03,0x06,0x81,0x00]
0xf9,0xa8,0x02,0x7e,0x03,0x06,0x81,0x00

# GFX940: v_cvt_f32_fp8_dpp v1, v3 quad_perm:[0,2,1,1] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xa8,0x02,0x7e,0x03,0x58,0x00,0xff]
0xfa,0xa8,0x02,0x7e,0x03,0x58,0x00,0xff

# GFX940: v_cvt_f32_fp8_e64 v1, s3 mul:2          ; encoding: [0x01,0x00,0x94,0xd1,0x03,0x00,0x00,0x08]
0x01,0x00,0x94,0xd1,0x03,0x00,0x00,0x08

# GFX940: v_cvt_f32_fp8_sdwa v1, s3 clamp mul:2 src0_sel:BYTE_1 ; encoding: [0xf9,0xa8,0x02,0x7e,0x03,0x66,0x81,0x00]
0xf9,0xa8,0x02,0x7e,0x03,0x66,0x81,0x00

# GFX940: v_cvt_f32_fp8_e64 v1, s3 clamp          ; encoding: [0x01,0x80,0x94,0xd1,0x03,0x00,0x00,0x00]
0x01,0x80,0x94,0xd1,0x03,0x00,0x00,0x00

# GFX940: v_cvt_f32_fp8_sdwa v1, 3 src0_sel:BYTE_1 ; encoding: [0xf9,0xa8,0x02,0x7e,0x83,0x06,0x81,0x00]
0xf9,0xa8,0x02,0x7e,0x83,0x06,0x81,0x00

# GFX940: v_cvt_pk_f32_bf8_e32 v[2:3], s3         ; encoding: [0x03,0xae,0x04,0x7e]
0x03,0xae,0x04,0x7e

# GFX940: v_cvt_pk_f32_bf8_e32 v[2:3], 3          ; encoding: [0x83,0xae,0x04,0x7e]
0x83,0xae,0x04,0x7e

# GFX940: v_cvt_pk_f32_bf8_e32 v[2:3], v3         ; encoding: [0x03,0xaf,0x04,0x7e]
0x03,0xaf,0x04,0x7e

# GFX940: v_cvt_pk_f32_bf8_sdwa v[2:3], s3 src0_sel:WORD_1 ; encoding: [0xf9,0xae,0x04,0x7e,0x03,0x06,0x85,0x00]
0xf9,0xae,0x04,0x7e,0x03,0x06,0x85,0x00

# GFX940: v_cvt_pk_f32_bf8_dpp v[0:1], v3 quad_perm:[0,2,1,1] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xae,0x00,0x7e,0x03,0x58,0x00,0xff]
0xfa,0xae,0x00,0x7e,0x03,0x58,0x00,0xff

# GFX940: v_cvt_pk_f32_bf8_e64 v[2:3], s3 mul:2   ; encoding: [0x02,0x00,0x97,0xd1,0x03,0x00,0x00,0x08]
0x02,0x00,0x97,0xd1,0x03,0x00,0x00,0x08

# GFX940: v_cvt_pk_f32_bf8_sdwa v[2:3], s3 clamp mul:2 src0_sel:WORD_1 ; encoding: [0xf9,0xae,0x04,0x7e,0x03,0x66,0x85,0x00]
0xf9,0xae,0x04,0x7e,0x03,0x66,0x85,0x00

# GFX940: v_cvt_pk_f32_bf8_e64 v[2:3], s3 clamp   ; encoding: [0x02,0x80,0x97,0xd1,0x03,0x00,0x00,0x00]
0x02,0x80,0x97,0xd1,0x03,0x00,0x00,0x00

# GFX940: v_cvt_pk_f32_fp8_e32 v[2:3], s3         ; encoding: [0x03,0xac,0x04,0x7e]
0x03,0xac,0x04,0x7e

# GFX940: v_cvt_pk_f32_fp8_e32 v[2:3], 3          ; encoding: [0x83,0xac,0x04,0x7e]
0x83,0xac,0x04,0x7e

# GFX940: v_cvt_pk_f32_fp8_e32 v[2:3], v3         ; encoding: [0x03,0xad,0x04,0x7e]
0x03,0xad,0x04,0x7e

# GFX940: v_cvt_pk_f32_fp8_sdwa v[2:3], s3 src0_sel:WORD_1 ; encoding: [0xf9,0xac,0x04,0x7e,0x03,0x06,0x85,0x00]
0xf9,0xac,0x04,0x7e,0x03,0x06,0x85,0x00

# GFX940: v_cvt_pk_f32_fp8_sdwa v[2:3], 3 src0_sel:WORD_1 ; encoding: [0xf9,0xac,0x04,0x7e,0x83,0x06,0x85,0x00]
0xf9,0xac,0x04,0x7e,0x83,0x06,0x85,0x00

# GFX940: v_cvt_pk_f32_fp8_dpp v[0:1], v3 quad_perm:[0,2,1,1] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xac,0x00,0x7e,0x03,0x58,0x00,0xff]
0xfa,0xac,0x00,0x7e,0x03,0x58,0x00,0xff

# GFX940: v_cvt_pk_f32_fp8_e64 v[2:3], s3 mul:2   ; encoding: [0x02,0x00,0x96,0xd1,0x03,0x00,0x00,0x08]
0x02,0x00,0x96,0xd1,0x03,0x00,0x00,0x08

# GFX940: v_cvt_pk_f32_fp8_sdwa v[2:3], s3 clamp mul:2 src0_sel:WORD_1 ; encoding: [0xf9,0xac,0x04,0x7e,0x03,0x66,0x85,0x00]
0xf9,0xac,0x04,0x7e,0x03,0x66,0x85,0x00

# GFX940: v_cvt_pk_f32_fp8_e64 v[2:3], s3 clamp   ; encoding: [0x02,0x80,0x96,0xd1,0x03,0x00,0x00,0x00]
0x02,0x80,0x96,0xd1,0x03,0x00,0x00,0x00

# GFX940: v_cvt_pk_bf8_f32 v1, v2, v3             ; encoding: [0x01,0x00,0xa3,0xd2,0x02,0x07,0x02,0x00]
0x01,0x00,0xa3,0xd2,0x02,0x07,0x02,0x00

# GFX940: v_cvt_pk_bf8_f32 v1, -v2, |v3|          ; encoding: [0x01,0x02,0xa3,0xd2,0x02,0x07,0x02,0x20]
0x01,0x02,0xa3,0xd2,0x02,0x07,0x02,0x20

# GFX940: v_cvt_pk_bf8_f32 v1, s2, 3              ; encoding: [0x01,0x00,0xa3,0xd2,0x02,0x06,0x01,0x00]
0x01,0x00,0xa3,0xd2,0x02,0x06,0x01,0x00

# GFX940: v_cvt_pk_bf8_f32 v1, v2, v3 op_sel:[0,0,1] ; encoding: [0x01,0x40,0xa3,0xd2,0x02,0x07,0x02,0x00]
0x01,0x40,0xa3,0xd2,0x02,0x07,0x02,0x00

# GFX940: v_cvt_pk_fp8_f32 v1, v2, v3             ; encoding: [0x01,0x00,0xa2,0xd2,0x02,0x07,0x02,0x00]
0x01,0x00,0xa2,0xd2,0x02,0x07,0x02,0x00

# GFX940: v_cvt_pk_fp8_f32 v1, -v2, |v3|          ; encoding: [0x01,0x02,0xa2,0xd2,0x02,0x07,0x02,0x20]
0x01,0x02,0xa2,0xd2,0x02,0x07,0x02,0x20

# GFX940: v_cvt_pk_fp8_f32 v1, s2, 3              ; encoding: [0x01,0x00,0xa2,0xd2,0x02,0x06,0x01,0x00]
0x01,0x00,0xa2,0xd2,0x02,0x06,0x01,0x00

# GFX940: v_cvt_pk_fp8_f32 v1, v2, v3 op_sel:[0,0,1] ; encoding: [0x01,0x40,0xa2,0xd2,0x02,0x07,0x02,0x00]
0x01,0x40,0xa2,0xd2,0x02,0x07,0x02,0x00

# GFX940: v_cvt_sr_bf8_f32 v1, v2, v3             ; encoding: [0x01,0x00,0xa5,0xd2,0x02,0x07,0x02,0x00]
0x01,0x00,0xa5,0xd2,0x02,0x07,0x02,0x00

# GFX940: v_cvt_sr_bf8_f32 v1, s2, 3              ; encoding: [0x01,0x00,0xa5,0xd2,0x02,0x06,0x01,0x00]
0x01,0x00,0xa5,0xd2,0x02,0x06,0x01,0x00

# GFX940: v_cvt_sr_bf8_f32 v1, v2, v3 op_sel:[0,0,1,1] ; encoding: [0x01,0x60,0xa5,0xd2,0x02,0x07,0x02,0x00]
0x01,0x60,0xa5,0xd2,0x02,0x07,0x02,0x00

# GFX940: v_cvt_sr_bf8_f32 v1, v2, v3 op_sel:[0,0,0,1] ; encoding: [0x01,0x40,0xa5,0xd2,0x02,0x07,0x02,0x00]
0x01,0x40,0xa5,0xd2,0x02,0x07,0x02,0x00

# GFX940: v_cvt_sr_bf8_f32 v1, -|s2|, v3          ; encoding: [0x01,0x01,0xa5,0xd2,0x02,0x06,0x02,0x20]
0x01,0x01,0xa5,0xd2,0x02,0x06,0x02,0x20

# GFX940: v_cvt_sr_fp8_f32 v1, v2, v3             ; encoding: [0x01,0x00,0xa4,0xd2,0x02,0x07,0x02,0x00]
0x01,0x00,0xa4,0xd2,0x02,0x07,0x02,0x00

# GFX940: v_cvt_sr_fp8_f32 v1, s2, 3              ; encoding: [0x01,0x00,0xa4,0xd2,0x02,0x06,0x01,0x00]
0x01,0x00,0xa4,0xd2,0x02,0x06,0x01,0x00

# GFX940: v_cvt_sr_fp8_f32 v1, v2, v3 op_sel:[0,0,1,1] ; encoding: [0x01,0x60,0xa4,0xd2,0x02,0x07,0x02,0x00]
0x01,0x60,0xa4,0xd2,0x02,0x07,0x02,0x00

# GFX940: v_cvt_sr_fp8_f32 v1, v2, v3 op_sel:[0,0,0,1] ; encoding: [0x01,0x40,0xa4,0xd2,0x02,0x07,0x02,0x00]
0x01,0x40,0xa4,0xd2,0x02,0x07,0x02,0x00

# GFX940: v_cvt_sr_fp8_f32 v1, -|s2|, v3          ; encoding: [0x01,0x01,0xa4,0xd2,0x02,0x06,0x02,0x20]
0x01,0x01,0xa4,0xd2,0x02,0x06,0x02,0x20