VSLIDEUP

Operation

Move old destination lanes toward higher indices and zero vacated lanes.

Assembler Syntax

VSLIDEUP.{B|W|L|Q}(z) Pn(p), imm6(i), Vn(v)

Privilege

Unprivileged

Required CPUID flags

VECTOR

For each active destination lane whose index is at least the unsigned count, VSLIDEUP reads the old destination lane at the index reduced by that count. It writes zero to an active lane with a lower index. The governing predicate retains the corresponding old destination value in each inactive lane.

At W width and count 2, each active result lane copies old Vd at an index two lower or becomes zero when its index is below the count; inactive lanes preserve their old destination values.

VSLIDEUP.W slide up by two lanes

2.75in

Encodings

VSLIDEUP.{B|W|L|Q}(z) Pn(p), imm6(i), Vn(v)

Format: Instruction format for VSLIDEUP.{B|W|L|Q}(z) Pn(p), imm6(i), Vn(v)

Format: Instruction format for VSLIDEUP.{B|W|L|Q}(z) Pn(p), imm6(i), Vn(v)