Introduction

This document is a quick reference of key ABI details of FreeBSD architecture ports. For full details consult the processor-specific ABI supplement documentation.

If not explicitly mentioned, sizes are in bytes. The architecture details in this document apply to FreeBSD 12.0 and later, unless otherwise noted.

FreeBSD uses a flat address space. Variables of types unsigned long, uintptr_t, and size_t and pointers all have the same representation.

In order to maximize compatibility with future pointer integrity mechanisms, manipulations of pointers as integers should be performed via uintptr_t or intptr_t and no other types. In particular, long and ptrdiff_t should be avoided.

On some architectures, e.g., sparc64, powerpc and AIM variants of powerpc64, the kernel uses a separate address space. On other architectures, kernel and a user mode process share a single address space. The kernel is located at the highest addresses.

On each architecture, the main user mode thread's stack starts near the highest user address and grows down.

FreeBSD architecture support varies by release. This table shows the first FreeBSD release to support each architecture, and, for discontinued architectures, the final release.

Type sizes

All FreeBSD architectures use some variant of the ELF (see elf(5)) Application Binary Interface (ABI) for the machine processor. All supported ABIs can be divided into two groups:

ILP32

int, long, void * types machine representations all have 4-byte size.

LP64

int type machine representation uses 4 bytes, while long and void * are 8 bytes.

Some machines support more than one FreeBSD ABI. Typically these are 64-bit machines, where the “native” LP64 execution environment is accompanied by the “legacy” ILP32 environment, which was the historical 32-bit predecessor for 64-bit evolution. Examples are:

aarch64 currently does not support execution of armv6 or armv7 binaries, even if the CPU implements AArch32 execution state.

On all supported architectures:

Integers are represented in two's complement. Alignment of integer and pointer types is natural, that is, the address of the variable must be congruent to zero modulo the type size. Most ILP32 ABIs, except arm, require only 4-byte alignment for 64-bit integers.

Machine-dependent type sizes:

time_t is 8 bytes on all supported architectures except i386.

Endianness and Char Signedness

Page Size

Floating Point

(1) Prior to FreeBSD 11.0, armv6 used the softfp ABI even though it supported only processors with a floating point unit.

Default Tool Chain

FreeBSD uses a variety of tool chain components for the supported CPU architectures: clang(1) and ld.lld(1) provided by the base system, GNU gcc(1) and Binutils ld(1), or an external toolchain compiler and linker provided by a port or package. This table shows the default tool chain for each architecture.

Note that GCC 4.2.1 is deprecated, and scheduled for removal on 2020-03-31. Any CPU architectures not migrated by then (to either base system Clang or external toolchain) may be removed from the tree after that date.

Predefined Macros

The compiler provides a number of predefined macros. Some of these provide architecture-specific details and are explained below. Other macros, including those required by the language standard, are not included here.

The full set of predefined macros can be obtained with this command:

cc -x c -dM -E /dev/null

Common type size and endianness macros:

Architecture-specific macros:

Compilers may define additional variants of architecture-specific macros. The macros above are preferred for use in FreeBSD.

Important make(1) variables

Most of the externally settable variables are defined in the build(7) man page. These variables are not otherwise documented and are used extensively in the build system.

MACHINE

Represents the hardware platform. This is the same as the native platform's uname(1) -m output. It defines both the userland / kernel interface, as well as the bootloader / kernel interface. It should only be used in these contexts. Each CPU architecture may have multiple hardware platforms it supports where MACHINE differs among them. It is used to collect together all the files from config(8) to build the kernel. It is often the same as MACHINE_ARCH just as one CPU architecture can be implemented by many different hardware platforms, one hardware platform may support multiple CPU architecture family members, though with different binaries. For example, MACHINE of i386 supported the IBM-AT hardware platform while the MACHINE of pc98 supported the Japanese company NEC's PC-9801 and PC-9821 hardware platforms. Both of these hardware platforms supported only the MACHINE_ARCH of i386 where they shared a common ABI, except for certain kernel / userland interfaces relating to underlying hardware platform differences in bus architecture, device enumeration and boot interface. Generally, MACHINE should only be used in src/sys and src/stand or in system imagers or installers.

MACHINE_ARCH

Represents the CPU processor architecture. This is the same as the native platforms uname(1) -p output. It defines the CPU instruction family supported. It may also encode a variation in the byte ordering of multi-byte integers (endian). It may also encode a variation in the size of the integer or pointer. It may also encode a ISA revision. It may also encode hard versus soft floating point ABI and usage. It may also encode a variant ABI when the other factors do not uniquely define the ABI (e.g., MIPS' n32 ABI). It, along with MACHINE, defines the ABI used by the system. For example, the MIPS CPU processor family supports 9 different combinations encoding pointer size, endian and hard versus soft float (for 8 combinations) as well as N32 (which only ever had one variation of all these). Generally, the plain CPU name specifies the most common (or at least first) variant of the CPU. This is why mips and mips64 imply 'big endian' while 'arm' and 'armv7' imply little endian. If we ever were to support the so-called x32 ABI (using 32-bit pointers on the amd64 architecture), it would most likely be encoded as amd64-x32. It is unfortunate that amd64 specifies the 64-bit evolution of the x86 platform (it matches the 'first rule') as everybody else uses x86_64. There is no standard name for the processor: each OS selects its own conventions.

MACHINE_CPUARCH

Represents the source location for a given MACHINE_ARCH. For example, MACHINE_CPUARCH is defined to be mips for all the flavors of mips that we support since we support them all with a shared set of sources. While amd64 and i386 are closely related, MACHINE_CPUARCH is not x86 for them. The FreeBSD source base supports amd64 and i386 with two distinct source bases living in subdirectories named amd64 and i386 (though behind the scenes there's some sharing that fits into this framework).

CPUTYPE

Sets the flavor of MACHINE_ARCH to build. It is used to optimize the build for a specific CPU / core that the binaries run on. Generally, this does not change the ABI, though it can be a fine line between optimization for specific cases.

TARGET

Used to set MACHINE in the top level Makefile for cross building. Unused outside of that scope. It is not passed down to the rest of the build. Makefiles outside of the top level should not use it at all (though some have their own private copy for hysterical raisons).

TARGET_ARCH

Used to set MACHINE_ARCH by the top level Makefile for cross building. Like TARGET, it is unused outside of that scope.