vcvtb
Vector Convert Half-Precision (Bottom)
VCVTB<c>.F16.F32 <Sd>, <Sm>
Converts single-precision to half-precision (Bottom half).
Details
Converts a single-precision floating-point value to half-precision and stores it in the bottom half of the destination single-precision register. The top half of the destination is unchanged. The instruction is conditional and sets VFP flags (FPSCR) according to the conversion result.
Pseudocode Operation
Sd[15:0] ← ConvertToHalf(Sm); Sd[31:16] unchanged
Example
VCVTB.F16.F32 s0, s2
Encoding
Binary Layout
cond
11101
D
11
0
01
0
Vd
10
1
0
0
1
M
0
Vm
Operands
-
Sd
Dest (Half) -
Sm
Src (Single)
Reference (Arm AArch32 ISA)
Instruction Forms
| Encoding | Instruction | ISA | Bit pattern | ||
|---|---|---|---|---|---|
| 0x0EB20A40 | VCVTB{<c>}{<q>}.F32.F16 <Sd>, <Sm> | A32 | cond | 11101 | D | 11 | 0 | 01 | 0 | Vd | 10 | 1 | 0 | 0 | 1 | M | 0 | Vm | ||
| 0x0EB20B40 | VCVTB{<c>}{<q>}.F64.F16 <Dd>, <Sm> | A32 | cond | 11101 | D | 11 | 0 | 01 | 0 | Vd | 10 | 1 | 1 | 0 | 1 | M | 0 | Vm | ||
| 0x0EB30A40 | VCVTB{<c>}{<q>}.F16.F32 <Sd>, <Sm> | A32 | cond | 11101 | D | 11 | 0 | 01 | 1 | Vd | 10 | 1 | 0 | 0 | 1 | M | 0 | Vm | ||
| 0x0EB30B40 | VCVTB{<c>}{<q>}.F16.F64 <Sd>, <Dm> | A32 | cond | 11101 | D | 11 | 0 | 01 | 1 | Vd | 10 | 1 | 1 | 0 | 1 | M | 0 | Vm | ||
| 0xEEB20A40 | VCVTB{<c>}{<q>}.F32.F16 <Sd>, <Sm> | T32 | 111011101 | D | 11 | 0 | 01 | 0 | Vd | 10 | 1 | 0 | 0 | 1 | M | 0 | Vm | ||
| 0xEEB20B40 | VCVTB{<c>}{<q>}.F64.F16 <Dd>, <Sm> | T32 | 111011101 | D | 11 | 0 | 01 | 0 | Vd | 10 | 1 | 1 | 0 | 1 | M | 0 | Vm | ||
| 0xEEB30A40 | VCVTB{<c>}{<q>}.F16.F32 <Sd>, <Sm> | T32 | 111011101 | D | 11 | 0 | 01 | 1 | Vd | 10 | 1 | 0 | 0 | 1 | M | 0 | Vm | ||
| 0xEEB30B40 | VCVTB{<c>}{<q>}.F16.F64 <Sd>, <Dm> | T32 | 111011101 | D | 11 | 0 | 01 | 1 | Vd | 10 | 1 | 1 | 0 | 1 | M | 0 | Vm | ||
| 0x0EB30940 | VCVTB{<c>}{<q>}.BF16.F32 <Sd>, <Sm> | A32 | cond | 11101 | D | 11 | 0 | 011 | Vd | 10 | 01 | 0 | 1 | M | 0 | Vm | ||
| 0xEEB30940 | VCVTB{<c>}{<q>}.BF16.F32 <Sd>, <Sm> | T32 | 111011101 | D | 11 | 0 | 011 | Vd | 10 | 01 | 0 | 1 | M | 0 | Vm |
Description
Convert to or from a half-precision value in the bottom half of a single-precision register does one of the following:
Depending on settings in the CPACR, NSACR, HCPTR, and FPEXC 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(); CheckVFPEnabled(TRUE);
bits(16) hp;
if convert_from_half then
hp = S[m]<lowbit+15:lowbit>;
if uses_double then
D[d] = FPConvert(hp, FPSCR[], 64);
else
S[d] = FPConvert(hp, FPSCR[], 32);
else
if uses_double then
hp = FPConvert(D[m], FPSCR[], 16);
else
hp = FPConvert(S[m], FPSCR[], 16);
S[d]<lowbit+15:lowbit> = hp;