3. AArch64 Specific Information

This chapter discusses the dependencies of the ARM AArch64 architecture (https://en.wikipedia.org/wiki/ARM_architecture#AArch64_features) in this port of RTEMS. The ARMv8-A versions are supported by RTEMS. Processors with a MMU use a static configuration which is set up during system start. SMP is supported.

Architecture Documents

For information on the ARM AArch64 architecture refer to the ARM Infocenter (http://infocenter.arm.com/).

3.1. CPU Model Dependent Features

This section presents the set of features which vary across ARM AArch64 implementations and are of importance to RTEMS. The set of CPU model feature macros are defined in the file cpukit/score/cpu/aarch64/rtems/score/aarch64.h based upon the particular CPU model flags specified on the compilation command line.

3.1.1. CPU Model Name

The macro CPU_MODEL_NAME is a string which designates the architectural level of this CPU model. See in cpukit/score/cpu/aarch64/rtems/score/aarch64.h for the values.

3.1.2. Floating Point Unit and SIMD

The Advanced SIMD (NEON) and Floating-point instruction set extension is supported and expected to be present since all ARMv8-A CPUs are expected to support it as per the ARMv8-A Programmer’s Guide Chapter 7 introduction (https://developer.arm.com/docs/den0024/a/aarch64-floating-point-and-neon). As such, CPU_HARDWARE_FP will always be set to TRUE.

3.2. Multilibs

The following multilib variants are available:

  1. ILP32: AArch64 instruction set and registers using 32bit long int and pointers

  2. LP64: AArch64 instruction set and registers using 64bit long int and pointers

Use for example the following GCC options:

-mcpu=cortex-a53 -mabi=ilp32

to build an application or BSP for the ARMv8-A architecture and tune the code for a Cortex-A53 processor. It is important to select the correct ABI.

3.3. Calling Conventions

Please refer to the Procedure Call Standard for the ARM 64-bit Architecture (https://github.com/ARM-software/abi-aa/releases/download/2019Q4/aapcs64.pdf).

3.4. Memory Model

A flat 64-bit or 32-bit memory model is supported depending on the selected multilib variant. All AArch64 CPU variants support a built-in MMU for which basic initialization for a flat memory model is handled.

Note that memcpy() and memset() must not be used on device memory as those functions are hand-optimized and will take advantage of unaligned accesses. *As per ARM*(https://developer.arm.com/documentation/ka004708/latest), unaligned accesses are not permitted for device memory.

3.5. Interrupt Processing

The Reset Vector is determined using RVBAR and is Read-Only. RVBAR is set using configuration signals only sampled at reset. The ARMv8 architecture has four exception types:

  • Synchronous Exception

  • Interrupt (IRQ)

  • Fast Interrupt (FIQ)

  • System Error Exception

Of these types only the synchronous and IRQ exceptions have explicit operating system support. It is intentional that the FIQ is not supported by the operating system. Without operating system support for the FIQ it is not necessary to disable them during critical sections of the system.

3.5.1. Interrupt Levels

There are exactly two interrupt levels on ARMv8 with respect to RTEMS. Level zero corresponds to interrupts enabled. Level one corresponds to interrupts disabled.

3.5.2. Interrupt Stack

The board support package must initialize the interrupt stack. The memory for the stacks is usually reserved in the linker script. The interrupt stack pointer is stored in the EL0 stack pointer and is accessed by switching to SP0 mode at the beginning of interrupt calls and back to SPx mode after completion of interrupt calls using the spsel instruction.

3.6. Symmetric Multiprocessing

SMP is supported on ARMv8-A. Available platforms are:

  • Xilinx ZynqMP (QEMU and hardware using PSCI via ARM Trusted Firmware)

3.7. Thread-Local Storage

Thread-local storage (TLS) is supported. AArch64 uses unmodified TLS variant I which is not explcitly stated, but can be inferred from the behavior of GCC and Addenda to, and Errata in, the ABI for the Arm® Architecture (https://developer.arm.com/documentation/ihi0045/g). This alters expectations for the size of the TLS Thread Control Block (TCB) such that, under the LP64 multilib variant, the TCB is 16 bytes in size instead of 8 bytes.