Is it the same as vand.vv/vor.vv?
No. Mask logical instructions operate on mask bits; vand/vor operate on normal vector data elements.
VMXOR.MM
RISC-V VMXOR.MM Instruction Details
Instruction ManualV-typeVector mask XOR: vd.mask[i] = vs2.mask[i] ^ vs1.mask[i]
Instruction Syntax
vmxor.mm vd, vs2, vs1
Operand Breakdown
This is a V-type-format instruction. Confirm operand positions based on the assembly syntax.
VVector OperationsMask Logic
Instruction Behavior
VMXOR.MM computes mask XOR: vd.mask[i] = vs2.mask[i] ^ vs1.mask[i], commonly used to identify differing mask-condition bits.
Quick Understanding & Search Notes
VMXOR.MM is the RVV vector mask XOR instruction. It computes on one-bit mask elements with assembly operand order vd, vs2, vs1; mask logical instructions are always unmasked.
The operation is independent of data SEW/LMUL element width; sources and destination are mask values held in vector registers.
VMXOR.MM's Boolean function must be read in vs2, vs1 order; this especially matters for ANDN/ORN.
Common Usage Scenarios
Mask Difference
e.g., sub t0, a0, a1 — compute the difference a0 - a1.
Mask Toggle
Understand this scenario with real code like «vmxor.mm v1, v2, v3».
Pre-Use Checklist
Syntax Check
- Confirm the current instruction format is V-type.
- Confirm the operand order matches the example.
Semantic Check
- Ensure the destination register usage is compatible with the calling convention.
- Confirm this is not the lower-level form of a pseudo-instruction expansion.
Pitfalls / Common Confusions
Assembly operand order is vd, vs2, vs1; ANDN/ORN invert vs1.
These mask logical instructions are always unmasked; vm=0 encodings are reserved.
They operate on mask bits, not normal vector data elements; do not confuse them with vand/vor/vxor.
FAQ
Is it controlled by the v0 mask?
No. Vector mask logical instructions are always unmasked, with tail elements updated tail-agnostically.