You copied the Doc URL to your clipboard.

VST4 (single 4-element structure from one lane)

Store single 4-element structure from one lane of four registers stores one 4-element structure to memory from corresponding elements of four registers. For details of the addressing mode see Advanced SIMD addressing mode.

Depending on settings in the CPACR, NSACR, and HCPTR registers, and the Security state and PE mode in which the instruction is executed, an attempt to execute the instruction might be undefined, or trapped to Hyp mode. For more information see Enabling Advanced SIMD and floating-point support.

A3

313029282726252423222120191817161514131211109876543210
111101001D00RnVd1011index_align1111
sizeRm
if size == '11' then UNDEFINED;
if size != '10' then SEE "Related encodings";
if index_align<1:0> == '11' then UNDEFINED;
ebytes = 4;  index = UInt(index_align<3>);
inc = if index_align<2> == '0' then 1 else 2;
alignment = if index_align<1:0> == '00' then 1 else 4 << UInt(index_align<1:0>);
d = UInt(D:Vd);  d2 = d + inc;  d3 = d2 + inc;  d4 = d3 + inc;  n = UInt(Rn);  m = UInt(Rm);
wback = (m != 15);  register_index = (m != 15 && m != 13);
if n == 15 || d4 > 31 then UNPREDICTABLE;

For more information about the constrained unpredictable behavior of this instruction, see Architectural Constraints on UNPREDICTABLE behaviors, and particularly VST4 (single 4-element structure from one lane).

Assembler Symbols

<c>

See Standard assembler syntax fields. This encoding must be unconditional.

<q>

See Standard assembler syntax fields.

<size> Is the data size, encoded in size:
size <size>
00 8
01 16
10 32
<list>

Is a list containing the 64-bit names of the four SIMD&FP registers holding the element.

The list must be one of:

{ <Dd>[<index>], <Dd+1>[<index>], <Dd+2>[<index>], <Dd+3>[<index>] }
Single-spaced registers, encoded as "spacing" = 0.
{ <Dd>[<index>], <Dd+2>[<index>], <Dd+4>[<index>], <Dd+6>[<index>] }
Double-spaced registers, encoded as "spacing" = 1. Not permitted when <size> == 8.

The encoding of "spacing" depends on <size>:

<size> == 16
"spacing" is encoded in the "index_align<1>" field.
<size> == 32
"spacing" is encoded in the "index_align<2>" field.

The register <Dd> is encoded in the "D:Vd" field.

The permitted values and encoding of <index> depend on <size>:

<size> == 8
<index> is in the range 0 to 7, encoded in the "index_align<3:1>" field.
<size> == 16
<index> is in the range 0 to 3, encoded in the "index_align<3:2>" field.
<size> == 32
<index> is 0 or 1, encoded in the "index_align<3>" field.
<Rn>

Is the general-purpose base register, encoded in the "Rn" field.

<align>

Is the optional alignment.

Whenever <align> is omitted, the standard alignment is used, see Unaligned data access, and the encoding depends on <size>:

<size> == 8
Encoded in the "index_align<0>" field as 0.
<size> == 16
Encoded in the "index_align<0>" field as 0.
<size> == 32
Encoded in the "index_align<1:0>" field as 0b00.

Whenever <align> is present, the permitted values and encoding depend on <size>:

<size> == 8
<align> is 32, meaning 32-bit alignment, encoded in the "index_align<0>" field as 1.
<size> == 16
<align> is 64, meaning 64-bit alignment, encoded in the "index_align<0>" field as 1.
<size> == 32
<align> can be 64 or 128. 64-bit alignment is encoded in the "index_align<1:0>" field as 0b01, and 128-bit alignment is encoded in the "index_align<1:0>" field as 0b10.

: is the preferred separator before the <align> value, but the alignment can be specified as @<align>, see Advanced SIMD addressing mode.

For more information about the variants of this instruction, see Advanced SIMD addressing mode.

Operation

if ConditionPassed() then
    EncodingSpecificOperations();  CheckAdvSIMDEnabled();
    address = R[n];  iswrite = TRUE;
    - = AArch32.CheckAlignment(address, alignment, AccType_VEC, iswrite);
    MemU[address,         ebytes] = Elem[D[d], index];
    MemU[address+ebytes,  ebytes] = Elem[D[d2],index];
    MemU[address+2*ebytes,ebytes] = Elem[D[d3],index];
    MemU[address+3*ebytes,ebytes] = Elem[D[d4],index];
    if wback then
        if register_index then
            R[n] = R[n] + R[m];
        else
            R[n] = R[n] + 4*ebytes;
Was this page helpful? Yes No