Arm Community
Arm Community
  • Site
  • User
  • Site
  • Search
  • User
Arm Community blogs
Arm Community blogs
Operating Systems blog Arm Neon programming quick reference
  • Blogs
  • Mentions
  • Sub-Groups
  • Tags
  • Jump...
  • Cancel
More blogs in Arm Community blogs
  • AI blog

  • Announcements

  • Architectures and Processors blog

  • Automotive blog

  • Embedded and Microcontrollers blog

  • Internet of Things (IoT) blog

  • Laptops and Desktops blog

  • Mobile, Graphics, and Gaming blog

  • Operating Systems blog

  • Servers and Cloud Computing blog

  • SoC Design and Simulation blog

  • Tools, Software and IDEs blog

Tags
  • NEON
  • Cortex-A
  • simd
  • Tutorial
Actions
  • RSS
  • More
  • Cancel
Related blog posts
Related forum threads

Arm Neon programming quick reference

Yang Zhang 张洋
Yang Zhang 张洋
March 27, 2015
12 minute read time.

Welcome to the Arm Neon programming quick reference.

Introduction

This article aims to introduce Arm Neon technology. Hope that beginners can get started with Neon programming quickly after reading the article. The article will also inform users which documents can be consulted if more detailed information is needed.

Neon overview

This section describes the Neon technology and supplies some background knowledge.

What is Neon?

Neon technology is an advanced SIMD (Single Instruction, Multiple Data) architecture for the Arm Cortex-A series processors. It can accelerate multimedia and signal processing algorithms such as video encoder/decoder, 2D/3D graphics, gaming, audio and speech processing, image processing, telephony, and sound.

Neon instructions perform "Packed SIMD" processing:

  • Registers are considered as vectors of elements of the same data type
  • Data types can be: signed/unsigned 8-bit, 16-bit, 32-bit, 64-bit, single-precision floating-point on ARM 32-bit platform, both single-precision floating-point and double-precision floating-point on ARM 64-bit platform.
  • Instructions perform the same operation in all lanes

History of Arm Adv SIMD

Armv6
SIMD extension
Armv7-A
Neon
Armv8-A AArch64
Neon
•  Operates on 32-bit general purpose ARM registers
•  8-bit/16-bit integer
•  2x16-bit/4x8-bit operations per instruction
•  Separate register bank, 32x64-bit Neon registers
•  8/16/32/64-bit integer
•  Single precision floating point
•  Up to 16x8-bit operations per instruction
•  Separate register bank, 32x128-bit Neon registers
•  8/16/32/64-bit integer
•  Single precision floating point
•  Double precision floating point, both of them are IEEE compliance
•  Up to 16x8-bit operations per instruction

Why use Neon

Neon provides:

  • Support for both integer and floating point operations ensures the adaptability of a broad range of applications, from codecs to High Performance Computing to 3D graphics.
  • Tight coupling to the Arm processor provides a single instruction stream and a unified view of memory, presenting a single development platform target with a simpler tool flow

Armv7/v8 comparison

Armv8-A is a fundamental change to the Arm architecture. It supports the 64-bit Execution state called “AArch64”, and a new 64-bit instruction set “A64”. To provide compatibility with the Armv7-A (32-bit architecture) instruction set, a 32-bit variant of Armv8-A “AArch32” is provided. Most of existing Armv7-A code can be run in the AArch32 execution state of Armv8-A.

This section compares the Neon-related features of both the Armv7-A and Armv8-A architectures. In addition, general purpose Arm registers and Arm instructions, which are used often for Neon programming, will also be mentioned. However, the focus is still on the Neon technology.

Register

Armv7-A and AArch32 have the same general purpose Arm registers – 16 x 32-bit general purpose Arm registers (R0-R15).

Armv7-A and AArch32 have 32 x 64-bit Neon registers (D0-D31). These registers can also be viewed as 16x128-bit registers (Q0-Q15). Each of the Q0-Q15 registers maps to a pair of D registers, as shown in the following figure.

Register mapping Q0-Q15 D registers   

AArch64 by comparison, has 31 x 64-bit general purpose Arm registers and 1 special register having different names, depending on the context in which it is used. These registers can be viewed as either 31 x 64-bit registers (X0-X30) or as 31 x 32-bit registers (W0-W30).

AArch64 has 32 x 128-bit Neon registers (V0-V31). These registers can also be viewed as 32-bit Sn registers or 64-bit Dn registers.

 AArch64 registers

Instruction set

The following figure illustrates the relationship between Armv7-A, Armv8-A AArch32 and Armv8-A AArch64 instruction set.

 elationship between Armv7-A, Armv8-A AArch32 and Armv8-A AArch64 instruction set

The Armv8-A AArch32 instruction set consists of A32 (Arm instruction set, a 32-bit fixed length instruction set) and T32 (Thumb instruction set, a 16-bit fixed length instruction set; Thumb2 instruction set, 16 or 32-bit length instruction set). It is a superset of the Armv7-A instruction set, so that it retains the backwards compatibility necessary to run existing software. There are some additions to A32 and T32 to maintain alignment with the A64 instruction set, including Neon division, and the Cryptographic Extension instructions. Neon double precision floating point (IEEE compliance) is also supported.

Neon instruction format

This section describes the changes to the Neon instruction syntax.

Armv7-A/AArch32 instruction syntax

All mnemonics for Armv7-A/AAArch32 Neon instructions (as with VFP) begin with the letter “V”. Instructions are generally able to operate on different data types, with this being specified in the instruction encoding. The size is indicated with a suffix to the instruction. The number of elements is indicated by the specified register size and data type of operation. Instructions have the following general format:

V{<mod>}<op>{<shape>}{<cond>}{.<dt>}{<dest>}, src1, src2

Where:

<mod> - modifiers

  • Q: The instruction uses saturating arithmetic, so that the result is saturated within the range of the specified data type, such as VQABS, VQSHL etc.
  • H: The instruction will halve the result. It does this by shifting right by one place (effectively a divide by two with truncation), such as VHADD, VHSUB.
  • D: The instruction doubles the result, such as VQDMULL, VQDMLAL, VQDMLSL and VQ{R}DMULH
  • R: The instruction will perform rounding on the result, equivalent to adding 0.5 to the result before truncating, such as VRHADD, VRSHR.

<op> - the operation (for example, ADD, SUB, MUL).

<shape> - Shape.

Neon data processing instructions are typically available in Normal, Long, Wide and Narrow variants.

  • Long (L): instructions operate on double-word vector operands and produce a quad-word vector result. The result elements are twice the width of the operands, and of the same type. Lengthening instructions are specified using an L appended to the instruction.

Neon data processing instruction long

  • Wide (W): instructions operate on a double-word vector operand and a quad-word vector operand, producing a quad-word vector result. The result elements and the first operand are twice the width of the second operand elements. Widening instructions have a W appended to the instruction.

 Neon data processing instruction wide

  • Narrow (N): instructions operate on quad-word vector operands, and produce a double-word vector result. The result elements are half the width of the operand elements. Narrowing instructions are specified using an N appended to the instruction.

 Neon data processing instruction narrow

<cond> - Condition, used with IT instruction

<.dt> - Data type, such as s8, u8, f32 etc.

<dest> - Destination

<src1> - Source operand 1

<src2> - Source operand 2

Note: {} represents and optional parameter.

For example:

VADD.I8 D0, D1, D2

VMULL.S16 Q2, D8, D9

For more information, please refer to the documents listed in the Appendix.

AArch64 Neon instruction syntax

In the AArch64 execution state, the syntax of Neon instruction has changed. It can be described as follows:

{<prefix>}<op>{<suffix>}  Vd.<T>, Vn.<T>, Vm.<T>

Where:

<prefix> - prefix, such as using S/U/F/P to represent signed/unsigned/float/bool data type.

<op> – operation, such as ADD, AND etc.

<suffix> - suffix

  • P: “pairwise” operations, such as ADDP.
  • V: the new reduction (across-all-lanes) operations, such as FMAXV.
  • 2:new widening/narrowing “second part” instructions, such as ADDHN2, SADDL2.

ADDHN2: add two 128-bit vectors and produce a 64-bit vector result which is stored as high 64-bit part of Neon register.

SADDL2: add two high 64-bit vectors of Neon register and produce a 128-bit vector result.

<T> - data type, 8B/16B/4H/8H/2S/4S/2D. B represents byte (8-bit). H represents half-word (16-bit). S represents word (32-bit). D represents a double-word (64-bit).

For example:

UADDLP    V0.8H, V0.16B

FADD V0.4S, V0.4S, V0.4S

For more information, please refer to the documents listed in the Appendix.

Neon instructions

The following table compares the Armv7-A, AArch32 and AArch64 Neon instruction set.

“√” indicates that the AArch32 Neon instruction has the same format as Armv7-A Neon instruction.

“Y” indicates that the AArch64 Neon instruction has the same functionality as Armv7-A Neon instructions, but the format is different. Please check the Armv8-A ISA document.

If you are familiar with the Armv7-A Neon instructions, there is a simple way to map the Neon instructions of Armv7-A and AArch64. It is to check the Neon intrinsics document, so that you can find the AArch64 Neon instruction according to the intrinsics instruction.

New or changed functionality is highlighted.

Armv7-A AArch32 AArch64
logical and compare VAND, VBIC, VEOR, VORN, and VORR (register)

√

Y
VBIC and VORR (immediate)

√

Y
VBIF, VBIT, and VBSL

√

Y
VMOV, VMVN (register)

√

Y
VACGE and VACGT

√

Y
VCEQ, VCGE, VCGT, VCLE, and VCLT

√

Y
VTST

√

Y
general data processing VCVT (between fixed-point or integer, and floating-point)

√

Y
VCVT (between half-precision and single-precision floating-point)

√

Y
n/a       n/a FCVTXN(double)
to single-precision)
VDUP

√

Y
VEXT

√

Y
VMOV, VMVN (immediate)

√

Y
VMOVL, V{Q}MOVN, VQMOVUN

√

Y
VREV

√

Y
VSWP

√

n/a
VTBL, VTBX

√

Y
VTRN

√

TRN1, TRN2
VUZP, VZIP

√

UZP1,UZP2, ZIP, ZIP2
n/a n/a INS
n/a VRINTA, VRINM,
VRINTN, VRINTP,
VRINTR, VRINTX,
VRINTZ
FRINTA, FRINTI, FRINTM, FRINTN, FRINTP, FRINTX, FRINTZ
shift VSHL, VQSHL, VQSHLU, and VSHLL (by immediate)

√

Y
V{Q}{R}SHL (by signed variable)

√

Y
V{R}SHR

√

Y
V{R}SHRN

√

Y
V{R}SRA

√

Y
VQ{R}SHR{U}N

√

Y
VSLI and VSRI

√

Y
general arithmetic VABA{L} and VABD{L}

√

Y
V{Q}ABS and V{Q}NEG

√

Y
V{Q}ADD, VADDL, VADDW, V{Q}SUB, VSUBL, and VSUBW

√

Y
n/a n/a SUQADD, USQADD
V{R}ADDHN and V{R}SUBHN

√

Y
V{R}HADD and VHSUB

√

Y
VPADD{L}, VPADAL

√

Y
VMAX, VMIN, VPMAX, and VPMIN

√

Y
n/a n/a FMAXNMP, FMINNMP
VCLS, VCLZ, and VCNT

√

Y
VRECPE and VRSQRTE

√

Y
VRECPS and VRSQRTS

√

Y
n/a n/a FRECPX
RBIT
FSQRT
ADDV
SADDLV, UADDLV
SMAXV,UMAXV,FMAXV
FMAXNMV
SMINV,UMINV,FMINV
FMINNMV
multiply VMUL{L}, VMLA{L}, and VMLS{L}

√

There isn’t float MLA/MLS
VMUL{L}, VMLA{L}, and VMLS{L} (by scalar)

√

Y
VFMA, VFMS

√

Y
VQDMULL, VQDMLAL, and VQDMLSL (by vector or by scalar)

√

Y
VQ{R}DMULH (by vector or by scalar)

√

Y
n/a n/a FMULX
n/a n/a FDIV
load and store VLDn/VSTn(n=1, 2, 3, 4)

√

Y
VPUSH/VPOP

√

n/a
Crypto Extension n/a PMULL, PMULL2 PMULL, PMULL2
AESD, AESE AESD, AESE
AESIMC, AESMC AESIMC, AESMC
SHA1C, SHA1H, SHA1M, SHA1P SHA1C, SHA1H, SHA1M, SHA1P
SHA1SU0,
SHA1SU1
SHA1SU0,
SHA1SU1
SHA256H,
SHA256H2
SHA256H,
SHA256H2
SHA256SU0,
SHA256SU1
SHA256SU0,
SHA256SU1

Neon programming basics

There are four ways of using Neon

  • Neon optimized libraries
  • Vectorizing compilers
  • Neon intrinsics
  • Neon assembly

Libraries

The users can call the Neon optimized libraries directly in their program. Currently, you can use the following libraries:

  • OpenMax DL

This provides the recommended approach for accelerating AV codecs and supports signal processing and color space conversions.

  • Ne10

It is Arm’s open source project. Currently, the Ne10 library provides some math, image processing and FFT function. The FFT implementation is faster than other open source FFT implementations.

Vectorizing compilers

Adding vectorizing options in GCC can help C code to generate Neon code. GNU GCC gives you a wide range of options that aim to increase the speed, or reduce the size of the executable files they generate. For each line in your source code there are generally many possible choices of assembly instructions that could be used. The compiler must trade-off a number of resources, such as registers, stack and heap space, code size (number of instructions), compilation time, ease of debug, and number of cycles per instruction in order to produce an optimized image file.

Neon intrinsics

Neon intrinsics provides a C function call interface to Neon operations, and the compiler will automatically generate relevant Neon instructions allowing you to program once and run on either an Armv7-A or Armv8-A platform. If you intend to use the AArch64 specific Neon instructions, you can use the (__aarch64__) macro definition to separate these codes, as in the following example.

Neon intrinsics example:

//add for float array. assumed that count is multiple of 4

#include<arm_neon.h>

void add_float_c(float* dst, float* src1, float* src2, int count)

{

     int i;

     for (i = 0; i < count; i++)

         dst[i] = src1[i] + src2[i];

}

void add_float_neon1(float* dst, float* src1, float* src2, int count)

{

     int i;

     for (i = 0; i < count; i += 4)

     {

         float32x4_t in1, in2, out;

         in1 = vld1q_f32(src1);

         src1 += 4;

         in2 = vld1q_f32(src2);

         src2 += 4;

         out = vaddq_f32(in1, in2);

         vst1q_f32(dst, out);

         dst += 4;

         // The following is only an example describing how to use AArch64 specific NEON

         // instructions.                             

#if defined (__aarch64__)

         float32_t tmp = vaddvq_f32(in1);

#endif

     }

}

Checking disassembly, you can find vld1/vadd/vst1 Neon instruction on Armv7-A platform and ldr/fadd/str Neon instruction on Armv8-A platform.

Neon assembly

There are two ways to write Neon assembly:

  • Assembly files
  • Inline assembly

Assembly files

You can use ".S" or “.s” as the file suffix. The only difference is that C/C ++ preprocessor will process .S files first. C language features such as macro definitions can be used.

When writing Neon assembly in a separate file, you need to pay attention to saving the registers. For both Armv7 and Armv8, the following registers must be saved:

Armv7-A/AArch32 AArch64
General purpose registers R0-R3 parameters
R4-R11 need to be saved
R12 IP
R13(SP)
R14(LR) need to be saved
R0 for return value
X0-X7 parameters
X8-X18
X19-X28 need to be saved
X29(FP) need to be saved
X30(LR)
X0, X1  for return value
Neon registers D8-D15 need to be saved D part of V8-V15 need to be saved
Stack alignment 64-bit alignment 128-bit alignment
Stack push/pop PUSH/POP Rn list
VPUSH/VPOP Dn list
LDP/STP register pair

The following is an example of ARM v7-A and ARM v8-A Neon assembly.

//header
void add_float_Neon2(float* dst, float* src1, float* src2, int count);
//assembly code in .S file
Armv7-A/AArch32 AArch64
    .text     .syntax unified     .align 4     .global add_float_Neon2     .type add_float_Neon2, %function     .thumb .thumb_func add_float_Neon2: .L_loop:     vld1.32  {q0}, [r1]!
    vld1.32  {q1}, [r2]!
    vadd.f32 q0, q0, q1
    subs r3, r3, #4
    vst1.32  {q0}, [r0]!
    bgt .L_loop
    bx lr
    .text
    .align 4
    .global add_float_Neon2
    .type add_float_Neon2, %function
add_float_Neon2:
.L_loop:
    ld1     {v0.4s}, [x1], #16
    ld1     {v1.4s}, [x2], #16
    fadd    v0.4s, v0.4s, v1.4s
    subs x3, x3, #4
    st1  {v0.4s}, [x0], #16
    bgt .L_loop
    ret

For more examples, see GitHub.

Inline assembly

You can use Neon inline assembly directly in C/C++ code.

Pros:

  • The procedure call standard is simple. You do not need to save registers manually.
  • You can use C / C ++ variables and functions, so it can be easily integrated into C / C ++ code.

Cons:

  • Inline assembly has a complex syntax.
  • Neon assembly code is embedded in C/C ++ code, and it’s not easily ported to other platforms.

Example:


 

// Armv7-A/AArch32
void add_float_neon3(float* dst, float* src1, float* src2, int count)
{
asm volatile (
"1: \n"
"vld1.32 {q0}, [%[src1]]! \n"
"vld1.32 {q1}, [%[src2]]! \n"
"vadd.f32 q0, q0, q1 \n"
"subs %[count], %[count], #4 \n"
"vst1.32 {q0}, [%[dst]]! \n"
"bgt 1b \n"
: [dst] "+r" (dst)
: [src1] "r" (src1), [src2] "r" (src2), [count] "r" (count)
: "memory", "q0", "q1"
);
} // AArch64
void add_float_neon3(float* dst, float* src1, float* src2, int count)
{
asm volatile (
"1: \n"
"ld1 {v0.4s}, [%[src1]], #16 \n"
"ld1 {v1.4s}, [%[src2]], #16 \n"
"fadd v0.4s, v0.4s, v1.4s \n"
"subs %[count], %[count], #4 \n"
"st1 {v0.4s}, [%[dst]], #16 \n"
"bgt 1b \n"
: [dst] "+r" (dst)
: [src1] "r" (src1), [src2] "r" (src2), [count] "r" (count)
: "memory", "v0", "v1"
);
}

Neon intrinsics and assembly

Neon intrinsics and assembly are the commonly used Neon. The following table describes the pros and cons of these two approaches:

Neon assembly Neon intrinsic
Performance Always shows the best performance for the specified platform for an experienced developer. Depends heavily on the toolchain used
Portability The different ISAs (ARMv7-A/AArch32 and AArch64) have different assembly implementations. Even for the same ISA, the assembly might need to be fine-tuned to achieve ideal performance between different micro architectures. Program once and run on different ISA’s. The compiler may also grant performance fine-tuning for different micro-architectures.
Maintainability Hard to read/write compared to C. Similar to C code, it’s easy to read/write.

With the above information, you can choose a Neon implementation and start your Neon programming journey.

This is a simple summary. When applying Neon to more complex scenarios, there will be many special cases. This will be described in my next blog, which you can read by clicking on the link below.

Find out more about how Arm is transforming Edge Computing
Arm Neon optimization

Anonymous

Top Comments

  • vfar
    vfar over 10 years ago +1
    太好了   支持楼上
  • David.D
    David.D over 6 years ago +1
    That's really excellent, Thank you for sharing.
Parents
  • zaccur
    zaccur over 5 years ago

    Excellent! Mark! Thx!

    • Cancel
    • Up 0 Down
    • Reply
    • More
    • Cancel
Comment
  • zaccur
    zaccur over 5 years ago

    Excellent! Mark! Thx!

    • Cancel
    • Up 0 Down
    • Reply
    • More
    • Cancel
Children
No Data
Operating Systems blog
  • Enhancing Chromium’s Memory Safety with Armv9

    Richard Townsend
    Richard Townsend
    The Arm Open-source Software team is delighted to mark the release of Chromium M115, with experimental support for Arm’s Memory Tagging Extension (MTE).
    • August 7, 2023
  • New Memory Tagging Extension User Guide for Android OS Developers

    Roberto Lopez Mendez
    Roberto Lopez Mendez
    In this blog, read about what to expect with the new MTE User Guide for Android OS.
    • May 25, 2023
  • Enhancing Chromium's Control Flow Integrity with Armv9

    Richard Townsend
    Richard Townsend
    This blog explains how Control Flow Integrity, an Armv9 security feature, works on the newly launched Chromium M105.
    • October 11, 2022