- Index
- December 2023
KSHIFTLW/KSHIFTLB/KSHIFTLQ/KSHIFTLD — Shift Left Mask Registers
Opcode/Instruction | Op/En | 64/32 bit Mode Support | CPUID Feature Flag | Description |
---|---|---|---|---|
VEX.L0.66.0F3A.W1 32 /r KSHIFTLW k1, k2, imm8 | RRI | V/V | AVX512F | Shift left 16 bits in k2 by immediate and write result in k1. |
VEX.L0.66.0F3A.W0 32 /r KSHIFTLB k1, k2, imm8 | RRI | V/V | AVX512DQ | Shift left 8 bits in k2 by immediate and write result in k1. |
VEX.L0.66.0F3A.W1 33 /r KSHIFTLQ k1, k2, imm8 | RRI | V/V | AVX512BW | Shift left 64 bits in k2 by immediate and write result in k1. |
VEX.L0.66.0F3A.W0 33 /r KSHIFTLD k1, k2, imm8 | RRI | V/V | AVX512BW | Shift left 32 bits in k2 by immediate and write result in k1. |
Instruction Operand Encoding ¶
Op/En | Operand 1 | Operand 2 | Operand 3 |
---|---|---|---|
RRI | ModRM:reg (w) | ModRM:r/m (r, ModRM:[7:6] must be 11b) | imm8 |
Description ¶
Shifts 8/16/32/64 bits in the second operand (source operand) left by the count specified in immediate byte and place the least significant 8/16/32/64 bits of the result in the destination operand. The higher bits of the destination are zero-extended. The destination is set to zero if the count value is greater than 7 (for byte shift), 15 (for word shift), 31 (for doubleword shift) or 63 (for quadword shift).
Operation ¶
KSHIFTLW ¶
COUNT := imm8[7:0] DEST[MAX_KL-1:0] := 0 IF COUNT <=15 THEN DEST[15:0] := SRC1[15:0] << COUNT; FI;
KSHIFTLB ¶
COUNT := imm8[7:0] DEST[MAX_KL-1:0] := 0 IF COUNT <=7 THEN DEST[7:0] := SRC1[7:0] << COUNT; FI;
KSHIFTLQ ¶
COUNT := imm8[7:0] DEST[MAX_KL-1:0] := 0 IF COUNT <=63 THEN DEST[63:0] := SRC1[63:0] << COUNT; FI;
KSHIFTLD ¶
COUNT := imm8[7:0] DEST[MAX_KL-1:0] := 0 IF COUNT <=31 THEN DEST[31:0] := SRC1[31:0] << COUNT; FI;
Intel C/C++ Compiler Intrinsic Equivalent ¶
Compiler auto generates KSHIFTLW when needed.
Flags Affected ¶
None.
SIMD Floating-Point Exceptions ¶
None.
Other Exceptions ¶
See Table 2-63, “TYPE K20 Exception Definition (VEX-Encoded OpMask Instructions w/o Memory Arg).”