Vector Mask Instructions

vpopc

Vector count population in mask. Counts the number of mask elements of the active elements of the vector source mask register. ‘vpopc’ is an alias of ‘vcpop’.

vcpop.m rd, vs2, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
010000	vm	vs2	10000	010	vd/rd	1010111

vfirst

Find-rst-set mask bit. Finds the lowest-numbered active element of the source mask vector that has the value 1 and writes that element’s index to a GPR. If no active element has the value 1, -1 is written to the GPR.

vfirst.m rd, vs2, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
010000	vm	vs2	10001	010	vd/rd	1010111

vmsbf

Set-before-rst mask bit. Takes a mask register as input and writes results to a mask register.

vmsbf.m vd, vs2, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
010100	vm	vs2	00001	010	vd/rd	1010111

vmsof

Set-only-rst mask bit. It is similar to ‘vmsbf’, except it only sets the rst element with a bit set, if any.

vmsof.m vd, vs2, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
010100	vm	vs2	00010	010	vd/rd	1010111

vmsif

Set-including-rst mask bit. It is similar to ‘vmsbf’, except it also includes the element with a set bit.

vmsif.m vd, vs2, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
010100	vm	vs2	00011	010	vd/rd	1010111

viota

Reads a source vector mask register and writes to each element of the destination vector register

viota.m vd, vs2, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
010100	vm	vs2	10000	010	vd/rd	1010111

vid

Vector element index instruction. Writes each element’s index to the destination vector register group, from 0 to vl-1.

vid.v vd, vm # Write element ID to destination.

31-26	25	24-20	19-15	14-12	11-7	6-0
010100	vm	vs2	10001	010	vd/rd	1010111

vmandnot

Vector mask-register logical operation, vd.mask[i] = vs2.mask[i] && !vs1.mask[i]. The previous assembler mnemonic ‘vmandnot’ have been changed to ‘vmandn’ to be consistent with the equivalent scalar instructions. The old mnemonic can be retained as assembler aliases for compatibility.

vmandnot.mm vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
011000	vm	vs2	vs1	010	vd/rd	1010111

vmand

Vector mask-register logical operation, vd.mask[i] = vs2.mask[i] && vs1.mask[i].

vmand.mm vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
011001	vm	vs2	vs1	010	vd/rd	1010111

vmor

Vector mask-register logical operation, vd.mask[i] = vs2.mask[i] || vs1.mask[i].

vmor.mm vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
011010	vm	vs2	vs1	010	vd/rd	1010111

vmxor

Vector mask-register logical operation, vd.mask[i] = vs2.mask[i] ^^ vs1.mask[i].

vmxor.mm vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
011011	vm	vs2	vs1	010	vd/rd	1010111

vmornot

Vector mask-register logical operation, vd.mask[i] = vs2.mask[i] || !vs1.mask[i]. The previous assembler mnemonic ‘vmornot’ have been changed to ‘vmorn’ to be consistent with the equivalent scalar instructions. The old mnemonic can be retained as assembler aliases for compatibility.

vmornot.mm vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
011100	vm	vs2	vs1	010	vd/rd	1010111

vmnand

Vector mask-register logical operation, vd.mask[i] = !(vs2.mask[i] && vs1.mask[i]).

vmnand.mm vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
011101	vm	vs2	vs1	010	vd/rd	1010111

vmnor

Vector mask-register logical operation, vd.mask[i] = !(vs2.mask[i] || vs1.mask[i]).

vmnor.mm vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
011110	vm	vs2	vs1	010	vd/rd	1010111

vmxnor

Vector mask-register logical operation, vd.mask[i] = !(vs2.mask[i] ^^ vs1.mask[i]).

vmxnor.mm vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
011111	vm	vs2	vs1	010	vd/rd	1010111