vrshl
Vector Rounding Shift Left
VRSHL<c>.<dt> <Qd>, <Qm>, <Qn>
Shifts left with rounding based on a register value.
Details
Shifts each element in Qm left or right by the amount specified in the corresponding element of Qn with rounding applied when shifting right. Positive shift amounts shift left; negative amounts shift right with rounding. Results are stored in Qd. No condition flags are modified. This is a NEON instruction available in both A32 and T32 instruction sets.
Pseudocode Operation
for i = 0 to elements_in_Qd - 1
shift_amount ← signed(Qn[i])
if shift_amount >= 0 then
Qd[i] ← Qm[i] << shift_amount
else
if shift_amount <= -element_size then
Qd[i] ← 0
else
rounded ← Qm[i] + (1 << (-shift_amount - 1))
Qd[i] ← rounded >> (-shift_amount)
endif
endif
endfor
Example
VRSHL.dt q0, q2, q1
Encoding
Binary Layout
1111001
U
0
D
size
Vn
Vd
0101
N
0
M
0
Vm
Operands
-
Qd
Destination 128-bit SIMD register -
Qm
Second source 128-bit SIMD register -
Qn
Shift Reg
Reference (Arm AArch32 ISA)
Instruction Forms
| Encoding | Instruction | ISA | Bit pattern | ||
|---|---|---|---|---|---|
| 0xF2000500 | VRSHL{<c>}{<q>}.<dt> {<Dd>,} <Dm>, <Dn> | A32 | 1111001 | U | 0 | D | size | Vn | Vd | 0101 | N | 0 | M | 0 | Vm | ||
| 0xF2000540 | VRSHL{<c>}{<q>}.<dt> {<Qd>,} <Qm>, <Qn> | A32 | 1111001 | U | 0 | D | size | Vn | Vd | 0101 | N | 1 | M | 0 | Vm | ||
| 0xEF000500 | VRSHL{<c>}{<q>}.<dt> {<Dd>,} <Dm>, <Dn> | T32 | 111 | U | 11110 | D | size | Vn | Vd | 0101 | N | 0 | M | 0 | Vm | ||
| 0xEF000540 | VRSHL{<c>}{<q>}.<dt> {<Qd>,} <Qm>, <Qn> | T32 | 111 | U | 11110 | D | size | Vn | Vd | 0101 | N | 1 | M | 0 | Vm |
Description
Vector Rounding Shift Left takes each element in a vector, shifts them by a value from the least significant byte of the corresponding element of a second vector, and places the results in the destination vector. If the shift value is positive, the operation is a left shift. If the shift value is negative, it is a rounding right shift. For a truncating shift, see VSHL.
The first operand and result elements are the same data type, and can be any one of:
The second operand is always a signed integer of the same size.
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.
Operation
if ConditionPassed() then
EncodingSpecificOperations(); CheckAdvSIMDEnabled();
integer result;
for r = 0 to regs-1
for e = 0 to elements-1
integer element = Int(Elem[D[m+r], e, esize], unsigned);
integer shift = SInt(Elem[D[n+r], e, esize]<7:0>);
if shift >= 0 then // left shift
result = element << shift;
else // rounding right shift
shift = -shift;
result = (element + (1 << (shift - 1))) >> shift;
Elem[D[d+r], e, esize] = result<esize-1:0>;