mSL/NABI

macOS Subsystem for Linux / Native ABI — Linux binaries running as ordinary macOS processes.

This is the main module of mSL/XNU, a macOS Subsystem for Linux.

The larger project

mSL/XNU — macOS Subsystem for Linux / X is Now UNIX — is a set of compatibility layers and utilities for macOS that aim to improve system compatibility with ported Linux, BSD and other Unix-like code.

It also aims at native, seamless execution of Linux ELF binaries on macOS: not in a container and not in a virtual machine, but as ordinary processes on the running system.

For this we have several independent pieces, which is why the project is modular rather than one monolith. Each is useful on its own, and each can be installed, replaced or omitted:

Piece	What it does	Where
Syscall translation	Linux system calls onto Darwin's, over Hypervisor.framework	this repository
Filesystem Hierarchy Standard	Native Linux-style filesystem layout	mSL/FHS
procfs	/proc, as a native pseudo-filesystem	mSL/ProcFS

What is mSL/NABI?

A Linux program is a file of AArch64 instructions that expects a Linux kernel underneath it. On an Apple Silicon Mac the instructions are already native — it is the same CPU — and the kernel is not. NABI supplies the missing kernel interface and nothing else.

It runs the guest's own instructions on the real CPU inside Hypervisor.framework, catches each svc where a Linux kernel would have been, and answers it from Darwin. There is no instruction emulation, no second kernel, and no disk image. A guest process is one host process; its files are host files; its memory is host memory mapped into the guest's address space.

That shape is what distinguishes it. NABI is not a virtual machine that happens to be small, and not a container: it is a system-call translator that borrows the hypervisor purely as a privilege boundary, so that svc has somewhere to land.

NABI is experimental. It is a fork of Noah, which is unmaintained and last targeted macOS Sierra on Intel; the AArch64 port, the mSL integration and everything below are new work. For the original design, see the academic paper.

Tested on:

- macOS 26.5.2 (Tahoe), Darwin 25.5.0, Apple Silicon (arm64) — primary target

Design

Where a syscall goes

The guest runs at EL0 with the host's stage-2 translation underneath it. A guest svc enters a small EL1 trampoline NABI maps into the guest, which immediately issues hvc #1 — trapping to the VMM, which reads the guest's registers, decides what the call meant, and performs it against Darwin.

hvc #1 rather than hvc #0, for a reason worth knowing: an hvc #0 whose x0 falls in the SMCCC range is answered by firmware and never reaches the VMM at all. That produced a guest which worked for most syscalls and silently failed across a 16-million-value band of arguments.

Two-stage translation means two page tables to keep in step — the guest's own (4KiB granule) and the host's (hv_vm_map, fixed 16KiB). Every protection change has to touch both: stage 1 cannot grant what stage 2 withholds, and an mprotect that updated only stage 1 left the guest faulting forever on memory that both of NABI's tables described as writable.

fork is fork plus exec

Hypervisor.framework will not let a forked child create a vCPU once the parent has ever had one — it crashes inside the framework rather than returning an error. Measured standalone, that is not something NABI can work around in place.

So a guest fork is a host fork followed immediately by an exec of NABI itself, with the parent's state handed over through a checkpoint: memory regions, descriptor table, credentials, signal dispositions, the vCPU snapshot. The child rebuilds and resumes where the guest expects to be.

This is invisible to the guest and has one consequence worth stating, because it shapes every investigation here: a forked child is a fresh process, so anything not written into the checkpoint does not travel. Credentials, the supplementary group list and the debug sinks each had to be added after a bug showed they were missing.

The rootfs must be case-sensitive

macOS formats the boot volume case-insensitively. No Linux distribution can be unpacked there: Debian ships _exit.2.gz beside _Exit.2.gz, and xt_connmark.h beside xt_CONNMARK.h. One such pair is enough.

The failure does not resemble its cause. dpkg clears a stale .dpkg-new before each unpack, which for the second name deletes the first name's freshly written file; the second is a symlink, so it is created in the gap, pointing at a name that does not exist yet; and the deferred fsync pass then reports ENOENT against the file that unpacked correctly.

Holding both names would mean mangling them in the VFS and unmangling in readdir, in readlink, and in every path handed back — and any host tool looking at the tree would see the mangled form. So NABI detects the filesystem and says so, and msl keeps the trees on a case-sensitive sparse image it creates on demand. That needs no administrator rights, and the image is sparse, so its size is a ceiling rather than an allocation.

Credentials are emulated, not delegated

The host account NABI runs as is what the guest sees as root. setuid and friends move a struct cred inside NABI and never touch the host process, so a guest can be root, drop to an unprivileged user, and use sudo, while the host process remains the ordinary account that started it.

That is the honest boundary: the guest's root is a fiction, and the host account is the real limit. Nothing a guest does with its own credentials widens what the host process may touch.

How it compares

	Approach	Guest processes	Filesystem	Kernel
mSL/NABI	Syscall translation; hypervisor as a privilege boundary	Host processes, one to one	Host files, directly	Darwin
hyper-linux	The same: syscall translation, Hypervisor.framework, EL1 shim	Host processes, one to one	Host files, directly	Darwin
hylyx	Full VM (Virtualization.framework)	Inside the guest	Guest disk image	Linux
Docker / OrbStack / Lima	Full VM, containers inside it	Inside the guest	Guest disk image, shared over virtiofs	Linux
QEMU user-mode	Instruction emulation + syscall translation	Host processes	Host files	Host
WSL 1	Syscall translation in the NT kernel	Host processes	Host files, via a driver	NT
WSL 2	Full VM	Inside the guest	Guest disk image	Linux

The closest comparison is WSL 1, and the ambition is the same: Linux programs as first-class citizens of the host rather than guests behind a boundary. WSL 1 had the advantage of doing it inside the NT kernel; NABI does it from userspace, which is why Hypervisor.framework appears at all — it is the cheapest way to get a trap on svc without a kernel extension.

Against a VM manager such as hylyx, the difference is not size or speed but where things live. A VM gives you a complete, correct Linux — a real kernel, so everything works — reached over SSH, with its own disk, its own memory and its own process table. NABI gives you no kernel at all, and therefore an incomplete Linux, but the processes are yours: they appear in ps, they exit into your shell's exit status, they read and write your files with no sharing layer in between, and kill from the host works on them. Both are reasonable. They answer different questions — if you want to run Linux, run a VM; NABI is for wanting to run a Linux program.

The nearest peer by approach is hyper-linux, which does the same thing the same way: a per-process VM under Hypervisor.framework, an EL1 shim trapping svc, and Linux syscalls translated to Darwin's. Where the two differ is coverage. It reports 172 syscalls translated against NABI's 316, and states that it lacks namespaces and cgroups, treats MAP_SHARED as MAP_PRIVATE, and has no clone3 or robust futexes. NABI has all eight namespaces, a cgroup hierarchy, real shared file mappings, System V IPC, mount(2) with bind mounts and propagation, inotify and fanotify. It has clone3 and the robust futex list as well, which it did not when this comparison was first written.

Where hyper-linux is ahead: it runs x86-64 Linux binaries on Apple Silicon through Rosetta, which NABI does not — NABI's x86 backend is a separate build for Intel Macs, and is compile-tested rather than run (see PORTING-arm64.md). It also describes demand-paged memory over a 1TB address space, where NABI maps eagerly.

Against QEMU user-mode, architecturally the nearest thing: QEMU interprets or JITs the guest's instructions, because it exists to cross architectures. Running AArch64 Linux binaries on Apple Silicon there is nothing to cross, so NABI executes them directly and spends nothing on translation.

What NABI gives up for that is real, and it is what WSL 1 gave up too: every kernel interface has to be implemented by hand, and the ones that are not are simply absent. A VM never has that problem.

Requirements

• Apple Silicon Mac, macOS 15 or later
• Xcode Command Line Tools
• A case-sensitive volume for the rootfs — msl makes one for you
• No kernel extension, no Reduced Security, no SIP changes

Installing

make                # build into out/
sudo make install   # /usr/local/bin/msl, and the helpers under libexec

Quick start

$ msl install debian        # or ubuntu
$ msl login debian

msl install downloads the base system from the distribution's own archive, verifies every package against the checksums in the signed index, unpacks it, and then finishes the job inside NABI by running dpkg --configure -a — the same second stage debootstrap performs. That is why the maintainer scripts run, and why the tree comes out with /etc/shadow, an ld.so cache and working update-alternatives. It also creates an account named after you, with sudo.

msl on its own shows what is installed and what to type next; msl help lists everything.

user@host:~$ sudo apt install gcc
user@host:~$ gcc -O2 -o hello hello.c && ./hello

Status

Apple Silicon (arm64) is the primary target and is working. A Debian 13 or Ubuntu 24.04 rootfs installs, configures, updates over the network with OpenPGP signature verification, and installs and runs a C toolchain.

Area	Status
ELF loading, static and dynamic	Working
Syscall translation	Working for what a base system exercises
Memory: mmap, mprotect, mremap, brk	Working, including two-stage protection changes
fork / clone / execve	Working, via checkpointed fork-plus-exec
Threads, futexes	Working
Signals	Working, including tgkill/tkill
Terminals	Working — /dev/ptmx and the /dev/pts translation
Filesystem	Working — host-backed, with passthrough prefixes
/proc/self/{maps,cmdline,comm,exe,fd}	Working — served by NABI
Credentials, su, sudo	Working — emulated in software, with per-file owner and mode
Unix socket peer credentials	Working — SO_PEERCRED answers in the guest's own ids, which is what D-Bus authenticates on
Real-time signals	Working, not queued — delivered by number; several sends of one number before it is handled arrive once
Network interfaces	Read-only — AF_NETLINK serves link and address dumps, so ip, getifaddrs and if_nametoindex work; anything that would change the host's network is EPERM
Disk images	Read-only — mount of an ext2/3/4 image, performed by the host and reached through emulated loop devices
apt, dpkg, pacman, gcc	Working
Sibling modules (/proc, /sys)	Detected, optional; NABI runs without them
Debian, Ubuntu	Working — resolved from the archive, apt works inside
Arch	Working — from the published tarball; pacman installs, signatures and all
Fedora	Working — from the published container image; dnf installs and its rpm triggers run
x86-64	Builds, unverified — see below

The x86-64 backend is the original Noah VT-x code. It compiles, and is kept as the reference implementation of what each syscall path is meant to do — the arm64 code was written by reading it — but it cannot be executed on Apple Silicon (hv_vm_create returns HV_UNSUPPORTED under Rosetta). It is a compile baseline and a specification, not a runnable test.

PORTING-arm64.md is the working record of the port: the design, and every bug worth remembering, with what the symptom looked like and why it did not resemble the cause. INTEGRATION.md covers how NABI fits with the other modules, and HACKING.md is the developer guide.

Syscalls

330 of 375 implemented.

✅	implemented
⊘	answered deliberately, and the answer is ENOSYS (36 calls) — most are ones Linux does not implement either; rseq is the one that could have been and is not, for the reason in src/proc/rseq.c
·	no handler; the unknown-syscall path, which is also ENOSYS

A dash in a number column means that architecture has no such call.

The full table (375 calls)

Syscall	aarch64	x86-64	NABI
_sysctl	—	156	⊘
accept	202	43	✅
accept4	242	288	✅
access	—	21	✅
acct	89	163	⊘
add_key	217	248	⊘
adjtimex	171	159	✅
afs_syscall	—	183	⊘
alarm	—	37	✅
arch_prctl	—	158	✅
bind	200	49	✅
bpf	280	321	⊘
brk	214	12	✅
cachestat	451	451	✅
capget	90	125	✅
capset	91	126	✅
chdir	49	80	✅
chmod	—	90	✅
chown	—	92	✅
chroot	51	161	✅
clock_adjtime	266	305	✅
clock_getres	114	229	✅
clock_gettime	113	228	✅
clock_nanosleep	115	230	✅
clock_settime	112	227	✅
clone	220	56	✅
clone3	435	435	✅
close	57	3	✅
close_range	436	436	✅
connect	203	42	✅
copy_file_range	285	326	✅
creat	—	85	✅
create_module	—	174	⊘
delete_module	106	176	⊘
dup	23	32	✅
dup2	—	33	✅
dup3	24	292	✅
epoll_create	—	213	✅
epoll_create1	20	291	✅
epoll_ctl	21	233	✅
epoll_ctl_old	—	214	⊘
epoll_pwait	22	281	✅
epoll_pwait2	441	441	✅
epoll_wait	—	232	✅
epoll_wait_old	—	215	⊘
eventfd	—	284	✅
eventfd2	19	290	✅
execve	221	59	✅
execveat	281	322	✅
exit	93	60	✅
exit_group	94	231	✅
faccessat	48	269	✅
faccessat2	439	439	✅
fadvise64	223	221	✅
fallocate	47	285	✅
fanotify_init	262	300	✅
fanotify_mark	263	301	✅
fchdir	50	81	✅
fchmod	52	91	✅
fchmodat	53	268	✅
fchmodat2	452	452	✅
fchown	55	93	✅
fchownat	54	260	✅
fcntl	25	72	✅
fdatasync	83	75	✅
fgetxattr	10	193	✅
finit_module	273	313	⊘
flistxattr	13	196	✅
flock	32	73	✅
fork	—	57	✅
fremovexattr	16	199	✅
fsconfig	431	431	✅
fsetxattr	7	190	✅
fsmount	432	432	✅
fsopen	430	430	✅
fspick	433	433	✅
fstat	80	5	✅
fstatfs	44	138	✅
fsync	82	74	✅
ftruncate	46	77	✅
futex	98	202	✅
futex_requeue	456	456	✅
futex_wait	455	455	✅
futex_waitv	449	449	✅
futex_wake	454	454	✅
futimesat	—	261	✅
get_kernel_syms	—	177	⊘
get_mempolicy	236	239	✅
get_robust_list	100	274	✅
get_thread_area	—	211	⊘
getcpu	168	309	✅
getcwd	17	79	✅
getdents	—	78	✅
getdents64	61	217	✅
getegid	177	108	✅
geteuid	175	107	✅
getgid	176	104	✅
getgroups	158	115	✅
getitimer	102	36	✅
getpeername	205	52	✅
getpgid	155	121	✅
getpgrp	—	111	✅
getpid	172	39	✅
getpmsg	—	181	⊘
getppid	173	110	✅
getpriority	141	140	✅
getrandom	278	318	✅
getresgid	150	120	✅
getresuid	148	118	✅
getrlimit	163	97	✅
getrusage	165	98	✅
getsid	156	124	✅
getsockname	204	51	✅
getsockopt	209	55	✅
gettid	178	186	✅
gettimeofday	169	96	✅
getuid	174	102	✅
getxattr	8	191	✅
init_module	105	175	⊘
inotify_add_watch	27	254	✅
inotify_init	—	253	✅
inotify_init1	26	294	✅
inotify_rm_watch	28	255	✅
io_cancel	3	210	✅
io_destroy	1	207	✅
io_getevents	4	208	✅
io_pgetevents	292	333	✅
io_setup	0	206	✅
io_submit	2	209	✅
io_uring_enter	426	426	✅
io_uring_register	427	427	✅
io_uring_setup	425	425	✅
ioctl	29	16	✅
ioperm	—	173	⊘
iopl	—	172	⊘
ioprio_get	31	252	✅
ioprio_set	30	251	✅
kcmp	272	312	✅
kexec_file_load	294	320	⊘
kexec_load	104	246	⊘
keyctl	219	250	⊘
kill	129	62	✅
landlock_add_rule	445	445	⊘
landlock_create_ruleset	444	444	⊘
landlock_restrict_self	446	446	⊘
lchown	—	94	✅
lgetxattr	9	192	✅
link	—	86	✅
linkat	37	265	✅
listen	201	50	✅
listmount	458	458	✅
listxattr	11	194	✅
llistxattr	12	195	✅
lookup_dcookie	18	212	⊘
lremovexattr	15	198	✅
lseek	62	8	✅
lsetxattr	6	189	✅
lsm_get_self_attr	459	459	✅
lsm_list_modules	461	461	✅
lsm_set_self_attr	460	460	✅
lstat	—	6	✅
madvise	233	28	✅
map_shadow_stack	453	453	⊘
mbind	235	237	✅
membarrier	283	324	✅
memfd_create	279	319	✅
memfd_secret	447	447	⊘
migrate_pages	238	256	✅
mincore	232	27	✅
mkdir	—	83	✅
mkdirat	34	258	✅
mknod	—	133	✅
mknodat	33	259	✅
mlock	228	149	✅
mlock2	284	325	✅
mlockall	230	151	✅
mmap	222	9	✅
modify_ldt	—	154	⊘
mount	40	165	✅
mount_setattr	442	442	✅
move_mount	429	429	✅
move_pages	239	279	✅
mprotect	226	10	✅
mq_getsetattr	185	245	✅
mq_notify	184	244	✅
mq_open	180	240	✅
mq_timedreceive	183	243	✅
mq_timedsend	182	242	✅
mq_unlink	181	241	✅
mremap	216	25	✅
mseal	462	462	✅
msgctl	187	71	✅
msgget	186	68	✅
msgrcv	188	70	✅
msgsnd	189	69	✅
msync	227	26	✅
munlock	229	150	✅
munlockall	231	152	✅
munmap	215	11	✅
name_to_handle_at	264	303	✅
nanosleep	101	35	✅
newfstatat	79	262	✅
nfsservctl	42	180	⊘
open	—	2	✅
open_by_handle_at	265	304	✅
open_tree	428	428	✅
openat	56	257	✅
openat2	437	437	✅
pause	—	34	✅
perf_event_open	241	298	⊘
personality	92	135	✅
pidfd_getfd	438	438	✅
pidfd_open	434	434	✅
pidfd_send_signal	424	424	✅
pipe	—	22	✅
pipe2	59	293	✅
pivot_root	41	155	✅
pkey_alloc	289	330	⊘
pkey_free	290	331	⊘
pkey_mprotect	288	329	✅
poll	—	7	✅
ppoll	73	271	✅
prctl	167	157	✅
pread64	67	17	✅
preadv	69	295	✅
preadv2	286	327	✅
prlimit64	261	302	✅
process_madvise	440	440	✅
process_mrelease	448	448	✅
process_vm_readv	270	310	✅
process_vm_writev	271	311	✅
pselect6	72	270	✅
ptrace	117	101	✅
putpmsg	—	182	⊘
pwrite64	68	18	✅
pwritev	70	296	✅
pwritev2	287	328	✅
query_module	1078	178	⊘
quotactl	60	179	✅
quotactl_fd	443	443	✅
read	63	0	✅
readahead	213	187	✅
readlink	—	89	✅
readlinkat	78	267	✅
readv	65	19	✅
reboot	142	169	✅
recvfrom	207	45	✅
recvmmsg	243	299	✅
recvmsg	212	47	✅
remap_file_pages	234	216	✅
removexattr	14	197	✅
rename	—	82	✅
renameat	38	264	✅
renameat2	276	316	✅
request_key	218	249	⊘
restart_syscall	128	219	⊘
rmdir	—	84	✅
rseq	293	334	⊘
rt_sigaction	134	13	✅
rt_sigpending	136	127	✅
rt_sigprocmask	135	14	✅
rt_sigqueueinfo	138	129	✅
rt_sigreturn	139	15	✅
rt_sigsuspend	133	130	✅
rt_sigtimedwait	137	128	✅
rt_tgsigqueueinfo	240	297	✅
sched_get_priority_max	125	146	✅
sched_get_priority_min	126	147	✅
sched_getaffinity	123	204	✅
sched_getattr	275	315	✅
sched_getparam	121	143	✅
sched_getscheduler	120	145	✅
sched_rr_get_interval	127	148	✅
sched_setaffinity	122	203	✅
sched_setattr	274	314	✅
sched_setparam	118	142	✅
sched_setscheduler	119	144	✅
sched_yield	124	24	✅
seccomp	277	317	✅
security	—	185	⊘
select	—	23	✅
semctl	191	66	✅
semget	190	64	✅
semop	193	65	✅
semtimedop	192	220	✅
sendfile	71	40	✅
sendmmsg	269	307	✅
sendmsg	211	46	✅
sendto	206	44	✅
set_mempolicy	237	238	✅
set_mempolicy_home_node	450	450	✅
set_robust_list	99	273	✅
set_thread_area	—	205	⊘
set_tid_address	96	218	✅
setdomainname	162	171	✅
setfsgid	152	123	✅
setfsuid	151	122	✅
setgid	144	106	✅
setgroups	159	116	✅
sethostname	161	170	✅
setitimer	103	38	✅
setns	268	308	✅
setpgid	154	109	✅
setpriority	140	141	✅
setregid	143	114	✅
setresgid	149	119	✅
setresuid	147	117	✅
setreuid	145	113	✅
setrlimit	164	160	✅
setsid	157	112	✅
setsockopt	208	54	✅
settimeofday	170	164	⊘
setuid	146	105	✅
setxattr	5	188	✅
shmat	196	30	✅
shmctl	195	31	✅
shmdt	197	67	✅
shmget	194	29	✅
shutdown	210	48	✅
sigaltstack	132	131	✅
signalfd	—	282	✅
signalfd4	74	289	✅
socket	198	41	✅
socketpair	199	53	✅
splice	76	275	✅
stat	—	4	✅
statfs	43	137	✅
statmount	457	457	✅
statx	291	332	✅
swapoff	225	168	⊘
swapon	224	167	⊘
symlink	—	88	✅
symlinkat	36	266	✅
sync	81	162	✅
sync_file_range	84	277	✅
syncfs	267	306	✅
sysfs	—	139	✅
sysinfo	179	99	✅
syslog	116	103	✅
tee	77	276	✅
tgkill	131	234	✅
time	—	201	✅
timer_create	107	222	✅
timer_delete	111	226	✅
timer_getoverrun	109	225	✅
timer_gettime	108	224	✅
timer_settime	110	223	✅
timerfd_create	85	283	✅
timerfd_gettime	87	287	✅
timerfd_settime	86	286	✅
times	153	100	✅
tkill	130	200	✅
truncate	45	76	✅
tuxcall	—	184	⊘
umask	166	95	✅
umount2	39	166	✅
uname	160	63	✅
unlink	—	87	✅
unlinkat	35	263	✅
unshare	97	272	✅
uretprobe	1079	335	⊘
uselib	—	134	⊘
userfaultfd	282	323	✅
ustat	—	136	✅
utime	—	132	✅
utimensat	88	280	✅
utimes	—	235	✅
vfork	—	58	✅
vhangup	58	153	⊘
vmsplice	75	278	✅
vserver	—	236	⊘
wait4	260	61	✅
waitid	95	247	✅
write	64	1	✅
writev	66	20	✅

Regenerate with:

make syscalls UNISTD=/usr/include/asm-generic/unistd.h \
              UNISTD_X86=/usr/include/x86_64-linux-gnu/asm/unistd_64.h

That regenerates both dispatch tables from the same headers and writes this table out beside them, so the three cannot disagree.

The numbers and names come from the kernel's own headers and the status from NABI's dispatch tables, so this says what is true of the tree it was generated against rather than what someone remembered to update.

Testing

make check          # unit tests: instruction decode, page tables, checkpoint
make check-smoke    # freestanding guest binaries, run under NABI

The smoke suite is the interesting half. Each test is a small static AArch64 Linux binary that exercises one thing NABI got wrong at some point, and each is verified to fail without its fix — a test that passes either way is not evidence. They cover pty allocation, O_* flag values, AT_EMPTY_PATH, futex timeouts and wakeups, stage-2 protection, signal delivery, and the process-entry stack alignment.

One lesson is built into them: a test that shares a constant with the code under test cannot catch that constant being wrong. ptytest passed for some time while every real guest failed to allocate a pty, because the test and NABI used the same incorrect ioctl number. The tests now spell ABI constants out from the kernel headers.

Repository layout

Path	What
src/	The VMM, syscall translation, memory, process and filesystem code
include/linux/	The Linux ABI as NABI understands it — structures and constants
lib/	Hypervisor.framework glue
util/	msl and its helpers: rootfs builder, volume maker, shell launcher
test/	Unit tests and the freestanding smoke suite
spike/	Standalone experiments that settled design questions

Credits

NABI is a fork of Noah by Yuichi Nishiwaki and contributors. The Homebrew and MacPorts packages install the original noah, not this fork, and are not updated for it.

License

MIT — see LICENSE. Files carrying their own notice keep it; the inherited Noah sources are dual MIT/GPL.