VUSMMLA
Widening 8-bit mixed integer matrix multiply-accumulate into 2x2 matrix
The widening integer matrix multiply-accumulate instruction multiplies the 2x8 matrix of unsigned 8-bit integer values held in the first source vector by the 8x2 matrix of signed 8-bit integer values in the second source vector. The resulting 2x2 32-bit integer matrix product is destructively added to the 32-bit integer matrix accumulator held in the destination vector. This is equivalent to performing an 8-way dot product per destination element.
From Armv8.2, this is an optional instruction. ID_ISAR6.I8MM indicates whether this instruction is supported in the T32 and A32 instruction sets.
It has encodings from the following instruction sets:
A32 (
A1
)
and
T32 (
T1
)
.
1
1
1
1
1
1
0
0
1
1
0
1
1
0
0
1
0
VUSMMLA{<q>}.S8 <Qd>, <Qn>, <Qm>
if !HaveAArch32Int8MatMulExt() then UNDEFINED;
boolean op1_unsigned;
boolean op2_unsigned;
case B:U of
when '00' op1_unsigned = FALSE; op2_unsigned = FALSE;
when '01' op1_unsigned = TRUE; op2_unsigned = TRUE;
when '10' op1_unsigned = TRUE; op2_unsigned = FALSE;
when '11' UNDEFINED;
if Vd<0> == '1' || Vn<0> == '1' || Vm<0> == '1' then UNDEFINED;
integer d = UInt(D:Vd);
integer n = UInt(N:Vn);
integer m = UInt(M:Vm);
1
1
1
1
1
1
0
0
1
1
0
1
1
0
0
1
0
VUSMMLA{<q>}.S8 <Qd>, <Qn>, <Qm>
if InITBlock() then UNPREDICTABLE;
if !HaveAArch32Int8MatMulExt() then UNDEFINED;
boolean op1_unsigned;
boolean op2_unsigned;
case B:U of
when '00' op1_unsigned = FALSE; op2_unsigned = FALSE;
when '01' op1_unsigned = TRUE; op2_unsigned = TRUE;
when '10' op1_unsigned = TRUE; op2_unsigned = FALSE;
when '11' UNDEFINED;
if Vd<0> == '1' || Vn<0> == '1' || Vm<0> == '1' then UNDEFINED;
integer d = UInt(D:Vd);
integer n = UInt(N:Vn);
integer m = UInt(M:Vm);
<q>
See Standard assembler syntax fields.
<Qd>
Is the 128-bit name of the SIMD&FP third source and destination register, encoded in the "D:Vd" field as <Qd>*2.
<Qn>
Is the 128-bit name of the first SIMD&FP source register, encoded in the "N:Vn" field as <Qn>*2.
<Qm>
Is the 128-bit name of the second SIMD&FP source register, encoded in the "M:Vm" field as <Qm>*2.
CheckAdvSIMDEnabled();
bits(128) operand1 = Q[n>>1];
bits(128) operand2 = Q[m>>1];
bits(128) addend = Q[d>>1];
Q[d>>1] = MatMulAdd(addend, operand1, operand2, op1_unsigned, op2_unsigned);