Vector Reduction Instructions

vredsum

Vector single-width integer reduction instruction, vd[0] = sum( vs1[0] , vs2[*] ) where [*] denotes all active elements.

vredsum.vs vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
000000	vm	vs2	vs1	010	vd/rd	1010111

vredand

Vector single-width integer reduction instruction, vd[0] = and( vs1[0] , vs2[*] ) where [*] denotes all active elements.

vredand.vs vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
000001	vm	vs2	vs1	010	vd/rd	1010111

vfredusum

Vector single-width floating-point reduction instruction with unordered sum.

vfredusum.vs vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
000001	vm	vs2	vs1	001	vd/rd	1010111

vredor

Vector single-width integer reduction instruction, vd[0] = or( vs1[0] , vs2[*] ) where [*] denotes all active elements.

vredor.vs vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
000010	vm	vs2	vs1	010	vd/rd	1010111

vredxor

Vector single-width integer reduction instruction, vd[0] = xor( vs1[0] , vs2[*] ) where [*] denotes all active elements.

vredxor.vs vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
000011	vm	vs2	vs1	010	vd/rd	1010111

vfredosum

Vector single-width floating-point reduction instruction with ordered sum.

vfredosum.vs vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
000011	vm	vs2	vs1	001	vd/rd	1010111

vredminu

Vector single-width integer reduction instruction, vd[0] = minu( vs1[0] , vs2[*] ) where [*] denotes all active elements.

vredminu.vs vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
000100	vm	vs2	vs1	010	vd/rd	1010111

vredmin

Vector single-width integer reduction instruction, vd[0] = min( vs1[0] , vs2[*] ) where [*] denotes all active elements.

vredmin.vs vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
000101	vm	vs2	vs1	010	vd/rd	1010111

vfredmin

Vector single-width floating-point reduction instruction with minimum value.

vfredmin.vs vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
000101	vm	vs2	vs1	001	vd/rd	1010111

vredmaxu

Vector single-width integer reduction instruction, vd[0] = maxu( vs1[0] , vs2[*] ) where [*] denotes all active elements.

vredmaxu.vs vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
000110	vm	vs2	vs1	010	vd/rd	1010111

vredmax

Vector single-width integer reduction instruction, vd[0] = max( vs1[0] , vs2[*] ) where [*] denotes all active elements.

vredmax.vs vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
000111	vm	vs2	vs1	010	vd/rd	1010111

vfredmax

Vector single-width floating-point reduction instruction with maximum value.

vfredmax.vs vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
000111	vm	vs2	vs1	001	vd/rd	1010111

vwredsumu

Unsigned sum reduction into double-width accumulator.

vwredsumu.vs vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
110000	vm	vs2	vs1	000	vd	1010111

vwredsum

Signed sum reduction into double-width accumulator.

vwredsum.vs vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
110001	vm	vs2	vs1	000	vd	1010111

vfwredusum

Vector widening floating-point reduction instruction with unordered sum.

vfwredusum.vs vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
110001	vm	vs2	vs1	001	vd/rd	1010111

vfwredosum

Vector widening floating-point reduction instruction with ordered sum.

vfwredosum.vs vd, vs2, vs1, vm

31-26	25	24-20	19-15	14-12	11-7	6-0
110011	vm	vs2	vs1	001	vd/rd	1010111