- C 90.2%
- Makefile 3.2%
- PowerShell 2.3%
- Shell 1.7%
- Python 1.3%
- Other 1.2%
| Filename | Latest commit message | Latest commit date |
|---|---|---|
c98bdf0019
Gate 0's last structural row. kernel/core/kmem.c: slabs are pages from
kalloc() with a header at the front, and free objects hold the free list in
their own first eight bytes.
That last detail is the whole design. **An object carries no header and
freeing performs no search**: the page base of any object *is* its slab, and
the slab names its cache. It is also exactly why an object cannot exceed a
page, which matters below.
Two interfaces. kmem_cache_create/alloc/free for a subsystem with one kind of
thing -- it gets its own line in /proc/slab, which is most of why naming it is
worth the argument -- and kmalloc/kmfree over power-of-two classes from 16 to
2048 for code with no single type to build a cache around. Each cache keeps
partial and full lists; a slab that empties goes straight back to kalloc()
rather than being kept, so /proc/slab never overstates what is held.
`make kmem-test` asserts the bookkeeping on the build machine. Same arrangement
as schfs-objtest, and for the same reason: slab lists, free-list threading, the
page-to-cache lookup and whether pages actually come back are all things that
can be wrong without a kernel to be wrong in. 26 checks, including one that a
permissive allocator cannot pass -- write a distinct pattern through 512
objects and read them all back, because two objects sharing storage shows up
there and nowhere else.
One bug the test found immediately: kmem_max_obj() returned the raw page
remainder, which kmem_cache_create() then rounded *up* past its own limit. The
largest legal size was reported as illegal.
**It does not do what the register said it would.** The claim was that every
fixed table exists to work around kalloc. Measured:
struct proc 10656 NPROC 128 larger than a page
struct socket 4928 NSOCK 32 larger than a page
struct buf 4272 NBUF 256 larger than a page
struct inode 280 NINODE 200 a cache, bounded on purpose
struct device 120 NDEVICE 32 3.8 KB, never hit
struct file 48 NFILE 512 24 KB, never hit
The three largest cannot go in a page-backed slab whatever it does. NINODE is
a cache and is meant to be bounded. NFILE and NDEVICE are small enough that
their tables are rounding errors and their ceilings have never been reached.
Freeing the big ones needs a contiguous multi-page allocator, which is a
different thing and is not written. That is now in the register and in
docs/architecture/kernel.md rather than left as a claim nobody checked.
What it does fix is real, and was measured rather than assumed:
- **A pipe is 640 bytes and came from a whole 4096-byte page.** Every `|` in
a shell command burned three and a half kilobytes it never touched. Six
pipes per page now, and /proc/slab shows the page going back when they
close.
- **procread() built its answer in a 2048-byte array on the kernel stack**,
under a comment that had warned about exactly that since it was written --
it stayed a local because a whole page for 2 KB felt worse, and kmalloc is
the answer that was missing. Its frame is 64 bytes now instead of ~2100.
- procfs's other generator took a page from kalloc for the same 2 KB.
Verified on a boot: /proc/slab tracks pipe and kmalloc-2048 through real
pipelines, allocs and frees balance, and $(), ps, time, readlink, sigtest and
rename are all still green with no traps.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
|
||
| .claude/memory | ||
| boot | ||
| cmd | ||
| docs | ||
| etc | ||
| include | ||
| kernel | ||
| ldls | ||
| lib | ||
| ports | ||
| projects | ||
| srv | ||
| tools | ||
| .gdbinit | ||
| .gitattributes | ||
| .gitignore | ||
| LICENSE | ||
| LICENSE.MIT | ||
| Makefile | ||
| NOTICE | ||
| README.md | ||
| run.ps1 | ||
| shell.nix | ||
A 64-bit operating system, written from the metal up.
SMP · copy-on-write filesystem · TCP/IP · PCI, USB and virtio drivers · TrueType text · Linux syscall ABI
SchOS began as xv6 and is now mostly something else. The scheduler, the memory manager and the trap path have been rewritten for x86-64 and for more than one processor; the filesystem has been replaced outright; the syscall table speaks Linux numbering so that real software — musl, toybox, mksh — runs unmodified.
It is a hobby system built by one person, and it is honest about that: what works below has been run, not inferred, and the gaps are listed as plainly as the features.
The pieces
| Layer | Name | What it is |
|---|---|---|
| System | SchOS | The distribution — kernel, userland, tooling |
| Kernel | SchIFT | Multi-core x86-64 kernel with a modular VFS |
| Filesystem | SchFS | Transactional, copy-on-write, the root filesystem |
| API | SchAPI | System programming interface — the stable surface |
| Compositor | SchCom | Display compositor (planned) |
| Window manager | SchWM | (planned) |
| Desktop | SchDE | (planned) |
uname -s reports SchIFT, the way every Unix reports its kernel rather than
its distribution.
What works
Kernel — Long mode, 4-level paging, preemptive multitasking. SMP: one, two and four processors all boot, with CPU enumeration from the ACPI MADT rather than the 1994 MP table. Memory sized from the E820 map, so it runs in tens of megabytes rather than demanding hundreds.
Filesystem — SchFS is the root: a copy-on-write B-tree with a
uberblock ring, so a crash leaves either the old tree or the new one and there
is nothing to repair. Free space is derived from the tree rather than stored
beside it, which removes a whole class of disagreement. /proc/schfs shows the
generation counter advancing as writes commit.
Processes — fork, exec, wait, pipes, signals, sessions and process groups,
mmap (anonymous and file-backed), POSIX shared memory, Unix domain sockets,
poll.
Devices — PCI enumeration with a bus-agnostic device model at
/proc/devices. Drivers for e1000 networking, virtio-gpu and stdvga
displays, xHCI USB with HID, hub and mass-storage class drivers, IDE, PS/2,
serial. /dev/fb0 speaks the Linux fbdev interface.
Networking — ARP, IPv4, ICMP, UDP, TCP and DHCP, on lwIP behind a
driver-agnostic interface class. Verified end to end from the host, both
directions, rather than by inspection. Names resolve, so wget http://host/
works; httpd serves a directory back the other way.
Userland — musl, toybox, mksh and zsh, built from source. 268 applets:
ls, cat, ps, top, wc, cp, mount, awk, bc, diff and the rest
behave the way you expect, because they are the real programs. edit is
kilo — arrow keys, Ctrl-S, Ctrl-Q, nothing
to learn.
Text — a TrueType rasteriser (glyf outlines, composite glyphs, cmap
format 4) replacing the 8×16 cell grid. JetBrains Mono for the terminal, Roboto
for the desktop layer to come.
When it dies — PSOGM, the Purple Screen of Guru Meditation: stop code, fault address, call chain, and a QR code containing a text copy of the whole report.
Quick start
make # kernel + UEFI loader + root filesystem
make qemu # build it and boot it
make iso # a bootable ISO for VMware or real hardware
Builds on Linux, macOS, FreeBSD/OpenBSD/NetBSD and Windows (Git Bash). The toolchain is clang + lld everywhere — it is a cross-compiler by construction, so one LLVM install targets the bare-metal kernel, the 16-bit boot code and musl userland from any host. See docs/guides/building.md for prerequisites per platform, and docs/history/moving-to-linux.md for moving a checkout between machines.
Testing
Booting to a prompt proves nothing: a filesystem can mount and still fail the first real command, and that has happened here. So the test harness drives the serial console and asserts on what comes back.
tools/windows/bootcheck.ps1 # boot, run real commands, check the answers
tools/windows/bootcheck.ps1 -Tty -Ps # terminal behaviour, process enumeration
make schfs-check # SchFS B-tree, object and image suites
bootcheck is PowerShell because pwsh runs on every supported host.
Layout
kernel/ arch/ core/ dev/ drivers/ fs/ net/ proc/ sys/
fs/schfs/ the filesystem -- same source the host tools compile
include/ kernel and userland headers
boot/ uefi/ (PE loader) common/ (shared loader UI)
lib/ cmd/ first-party userland runtime and programs
ports/ third-party software built from source
tools/ windows/ (.ps1) posix/ (.sh) python/ (.py) and .c/.pl
docs/ architecture/ (what it is) plan/ (what's next)
guides/ (how to work on it) history/ (why it's like this)
docs/README.md is the index. If you are reading the tree for the first time, docs/architecture/README.md is the system layer by layer.
One rule worth knowing: the kernel and the host tools compile the same SchFS
source. mkschfs cannot disagree with the kernel about the on-disk format,
because there is only one implementation — found and fixed on the host, where a
test takes a second instead of a boot.
Another: syscalls are declared in exactly one place,
kernel/sys/table.inc. Everything else is generated from it.
Not there yet
- No desktop. Schcom composites and routes input; the plan is now to put it on Wayland and delete the homegrown protocol. See docs/plan/wayland.md.
- No descriptor passing. No
sendmsg/recvmsg, so noSCM_RIGHTS— which is what the GUI work stops at. docs/plan/foundations.md. - No dynamic linking.
exechandles staticET_EXEConly. - No page cache, so writable
MAP_SHAREDof a file is refused rather than faked. - A USB disk cannot be booted from — the root device is compiled in rather than selectable.
- SchFS has no Linux driver yet; the plan is on-disk format version first, then FUSE, then a kernel module.
- CFF/OpenType outlines and TrueType hinting are deliberately out of scope.
Licensing
Apache-2.0, copyright Nexoniarz. See LICENSE.
The kernel derives from xv6, which is MIT — see LICENSE.MIT. Third-party code redistributed in a built image (musl, toybox, mksh, zsh, kilo, the fonts) keeps its own licence; NOTICE carries the attributions. All of it is permissive — there is no GPL in an image this builds, which is why the userland is toybox and not BusyBox and the editor is kilo and not nano. See docs/licensing.md.
Drivers may be adapted from BSD sources, which are compatible. Linux drivers are not: GPL-2.0-only and Apache-2.0 cannot be combined and redistributed. Reading Linux to understand hardware is fine; translating a driver is not. Best of all is the public specification — xHCI, virtio, AHCI, e1000 — which carries no licence on what you write from it.