nexoniarz/SchOS
nexoniarz/SchOS
Watch
1
0
Fork
You've already forked SchOS
 0
Code Issues Pull requests Projects Releases 1 Packages Wiki Activity Actions

README: the logo, what works now, and measured numbers

Browse source 
The README led with a wallpaper and described the system of two months
ago: no desktop, no dynamic linking, a handful of drivers. It now leads
with Shift and covers Xorg and Window Maker, apk and Alpine packages,
SCFF, HTTPS, the storage, network and input drivers, and the machines it
has booted on. "Not there yet" lists what is actually missing today:
threads, udev, EHCI, touchpad gestures, an installer.

The numbers -- lines of code and of comments, syscalls, drivers, packages,
commits -- come from tools/python/stats.py, which rewrites them in the
README and on the website between stats markers instead of somebody
typing them once and letting them go stale.

.gitattributes keeps the generated font and logo tables and the vendored
Niveus out of the repository page's language bar.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01S1giNvrHTrPMPw7V6x1kQ4
This commit is contained in:
Nexoniarz 2026-10-24 22:02:44 +02:00
commit 4265794437
4 changed files with 358 additions and 47 deletions
Download patch file Download diff file Expand all files Collapse all files

8
.gitattributes vendored
View file

@ -27,3 +27,11 @@
*.webp binary *.webp binary
*.pdf binary *.pdf binary
*.wasm binary *.wasm binary
# The language bar on the repository page (Forgejo, GitHub) counts what
# linguist counts. Data tables are generated, not written, and Niveus is
# another project's design system carried for the website; neither is SchOS
# code. tools/python/stats.py leaves out the same files.
include/gfx/font.h linguist-generated
include/gfx/logo.h linguist-generated
docs/Website/niveus/** linguist-vendored

136
README.md
View file

@ -1,12 +1,12 @@
<div align="center"> <div align="center">
<img src="docs/branding/wallpapers/graphite.jpg" alt="SchOS — Shift, lit, on graphite" width="720"> <img src="docs/branding/shift.svg" alt="Shift, the SchOS mark" width="160">
# SchOS # SchOS
**A 64-bit operating system, written from the metal up.** **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 SMP kernel · copy-on-write filesystem · Linux syscall ABI · Alpine packages · X11 on real hardware
</div> </div>
@ -15,13 +15,46 @@ SMP · copy-on-write filesystem · TCP/IP · PCI, USB and virtio drivers · True
SchOS began as [xv6](https://github.com/mit-pdos/xv6-public) and is now mostly SchOS began as [xv6](https://github.com/mit-pdos/xv6-public) and is now mostly
something else. The scheduler, the memory manager and the trap path have been 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 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 replaced outright; the syscall table speaks Linux numbering, so real Linux
software — musl, toybox, mksh — runs unmodified. software runs on it unmodified — statically or dynamically linked, straight
out of Alpine's package repositories.
It is a hobby system built by one person, and it is honest about that: what 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 works below has been *run*, not inferred, and the gaps are listed as plainly as
the features. the features.
## In numbers
<!-- stats:begin -->
| | |
|---|---|
| **First-party code** | 61,367 lines, plus 21,128 lines of comments (26%) in 272 files |
| **Kernel** (`kernel/`, `include/`) | 45,207 lines of code, 16,111 of comments |
| **Syscalls** | 208, of which 184 on Linux numbering |
| **Drivers** | 15 in the kernel |
| **First-party programs** | 29 in `cmd/` |
| **Alpine packages in the root** | 88, from 11 asked for in `ports/alpine/world` |
| **History** | 253 commits since 2026-08-09 (76 days) |
<details>
<summary>Lines by language</summary>
| Language | Code | Comments |
|---|--:|--:|
| C | 56,580 | 19,427 |
| Assembly | 1,750 | 336 |
| PowerShell | 1,362 | 638 |
| Python | 759 | 147 |
| Shell | 716 | 469 |
| Perl | 129 | 45 |
| Linker script | 71 | 66 |
</details>
<sub>Measured by <code>tools/python/stats.py</code>; first-party code only, so
<code>ports/</code> and generated tables are not counted.</sub>
<!-- stats:end -->
## The pieces ## The pieces
| Layer | Name | What it is | | Layer | Name | What it is |
@ -29,6 +62,7 @@ the features.
| System | **SchOS** | The distribution — kernel, userland, tooling | | System | **SchOS** | The distribution — kernel, userland, tooling |
| Kernel | **SchIFT** | Multi-core x86-64 kernel with a modular VFS | | Kernel | **SchIFT** | Multi-core x86-64 kernel with a modular VFS |
| Filesystem | **SchFS** | Transactional, copy-on-write, the root filesystem | | Filesystem | **SchFS** | Transactional, copy-on-write, the root filesystem |
| Image format | **SCFF** | The SchOS Compressed File Format the live root ships in |
| API | **SchAPI** | System programming interface — the stable surface | | API | **SchAPI** | System programming interface — the stable surface |
| Compositor | **SchCom** | Display compositor *(planned)* | | Compositor | **SchCom** | Display compositor *(planned)* |
| Window manager | **SchWM** | *(planned)* | | Window manager | **SchWM** | *(planned)* |
@ -39,45 +73,61 @@ its distribution.
## What works ## What works
**Kernel** — Long mode, 4-level paging, preemptive multitasking. **SMP**: one, **Kernel** — Long mode, 4-level paging, preemptive multitasking. **SMP** up to
two and four processors all boot, with CPU enumeration from the ACPI MADT rather 64 processors, enumerated from the ACPI MADT. ACPI through
than the 1994 MP table. Memory sized from the E820 map, so it runs in tens of [uACPI](https://github.com/uACPI/uACPI): interrupt routing from `_PRT`, the
megabytes rather than demanding hundreds. power button, clean power-off and reset. Memory sized from the E820/UEFI map.
**Filesystem** — **SchFS** is the root: a copy-on-write B-tree with a **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 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 is nothing to repair. The live image ships as **[SCFF](docs/architecture/scff.md)**:
beside it, which removes a whole class of disagreement. `/proc/schfs` shows the the 358 MB root compresses to 102 MB with zstd and is unpacked a chunk at a
generation counter advancing as writes commit. time, the first time something reads it. `/proc` has `/proc/<pid>/fd`,
`/proc/self` and the rest that `ps`, `top` and Xorg read.
**Processes** — fork, exec, wait, pipes, signals, sessions and process groups, **Processes** — fork, exec, wait, pipes, signals, sessions and job control,
`mmap` (anonymous and file-backed), POSIX shared memory, Unix domain sockets, `mmap` (anonymous and file-backed), POSIX shared memory, `memfd`, Unix domain
`poll`. sockets with descriptor passing (`SCM_RIGHTS`), `poll`, `epoll`, `eventfd`,
`timerfd`, `signalfd`. **Dynamic linking**: `exec` loads `ld-musl`, so
binaries come from Alpine as they are.
**Devices** — PCI enumeration with a bus-agnostic device model at **Packages** — the root is managed by Alpine's **apk**. `ports/alpine/world`
`/proc/devices`. Drivers for **e1000** networking, **virtio-gpu** and stdvga says what is installed; the build resolves it once into `world.lock` and every
displays, **xHCI** USB with HID, hub and mass-storage class drivers, IDE, PS/2, image after that is reproducible. SchOS itself is one more package in it.
serial. `/dev/fb0` speaks the Linux fbdev interface.
**Networking** — ARP, IPv4, ICMP, UDP, TCP and DHCP, on lwIP behind a **Graphics** — **Xorg with Window Maker**, preinstalled, on the framebuffer
driver-agnostic interface class. Verified end to end from the host, both (`/dev/fb0`, Linux fbdev). Six virtual terminals, switched with
directions, rather than by inspection. Names resolve, so `wget http://host/` Ctrl+Alt+F1–F6 or Alt+F1–F6, with X on a terminal of its own; Ctrl+Alt+Backspace
works; `httpd` serves a directory back the other way. ends the session. Input as Linux evdev on `/dev/input/event*`.
**Userland** — musl, toybox, mksh and zsh, built from source. 268 applets: **Devices** — PCI with a bus-agnostic device model at `/proc/devices`.
`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](https://github.com/antirez/kilo) — arrow keys, Ctrl-S, Ctrl-Q, nothing |---|---|
to learn. | Storage | AHCI (SATA), NVMe, IDE, USB mass storage |
| Network | Intel e1000, Realtek r8169 and RTL8139 |
| USB | xHCI, hubs, HID keyboards and mice |
| Input | PS/2, I2C-HID laptop touchpads (DesignWare I2C, found through ACPI) |
| Display | UEFI GOP framebuffer, virtio-gpu, Bochs/stdvga |
| Audio | AC'97 |
**Networking** — ARP, IPv4, ICMP, UDP, TCP and DHCP on lwIP. DNS, `wget`
over **HTTPS**, `ping`, and an `httpd` serving a directory back.
**Userland** — musl, toybox, mksh and zsh, plus whatever Alpine has: `ls`,
`ps`, `top`, `awk`, `diff`, `xterm`, `fastfetch` behave the way you expect,
because they are the real programs.
**Text** — a TrueType rasteriser (`glyf` outlines, composite glyphs, cmap **Text** — a TrueType rasteriser (`glyf` outlines, composite glyphs, cmap
format 4) replacing the 8×16 cell grid. JetBrains Mono for the terminal, Roboto format 4). JetBrains Mono for the terminal, Roboto for the desktop to come.
for the desktop layer to come.
**When it dies** — PSOGM, the Purple Screen of Guru Meditation: stop code, **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 fault address, call chain, and a QR code containing a text copy of the whole
report. report.
**On real hardware** — booted on an AMD B550 desktop (Ryzen 5 4600G), an ASUS
TUF A15 laptop and an Ivy Bridge HP all-in-one, from a USB stick.
## Quick start ## Quick start
```sh ```sh
@ -86,6 +136,10 @@ make qemu # build it and boot it
make iso # a bootable ISO for VMware or real hardware make iso # a bootable ISO for VMware or real hardware
``` ```
The ISO is a hybrid image: write it to a USB stick with
`dd if=schos.iso of=/dev/sdX bs=4M` (or Rufus/Etcher in DD mode) and boot it
on any UEFI machine.
Builds on **Linux, macOS, FreeBSD/OpenBSD/NetBSD and Windows** (Git Bash). The 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 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 one LLVM install targets the bare-metal kernel, the 16-bit boot code and musl
@ -108,6 +162,9 @@ make schfs-check # SchFS B-tree, object and image suites
`bootcheck` is PowerShell because `pwsh` runs on every supported host. `bootcheck` is PowerShell because `pwsh` runs on every supported host.
The numbers at the top are measured, not typed: `python3 tools/python/stats.py`
rewrites them here and on the website.
## Layout ## Layout
``` ```
@ -136,16 +193,19 @@ Another: **syscalls are declared in exactly one place**,
## Not there yet ## Not there yet
- **No desktop.** Schcom composites and routes input; the plan is now to put it - **No threads.** `clone` makes processes, not threads (`CLONE_VM` is
on Wayland and delete the homegrown protocol. See refused), so multi-threaded programs run on one thread or not at all. Xorg
[docs/plan/wayland.md](docs/plan/wayland.md). runs with `-dumbSched` for this reason.
- **No descriptor passing.** No `sendmsg`/`recvmsg`, so no `SCM_RIGHTS` — which - **No udev.** Xorg is told its input devices by a config file instead of
is what the GUI work stops at. [docs/plan/foundations.md](docs/plan/foundations.md). finding them; `/sys` and the device-event socket are what is missing.
- **No dynamic linking.** `exec` handles static `ET_EXEC` only. - **USB 2.0 controllers (EHCI)** have no driver. On older machines the ports
wired only to EHCI do nothing; xHCI ports work.
- **Touchpads run in mouse mode**: pointer, click and tap, but no two-finger
scroll yet.
- **No page cache**, so writable `MAP_SHARED` of a file is refused rather than - **No page cache**, so writable `MAP_SHARED` of a file is refused rather than
faked. faked. SysV shared memory is missing too.
- **A USB disk cannot be booted from** — the root device is compiled in rather - **No installer.** SchOS runs live from the USB stick; nothing is written to
than selectable. the machine's disks.
- SchFS has no Linux driver yet; the plan is on-disk format version first, then - SchFS has no Linux driver yet; the plan is on-disk format version first, then
FUSE, then a kernel module. FUSE, then a kernel module.
- CFF/OpenType outlines and TrueType hinting are deliberately out of scope. - CFF/OpenType outlines and TrueType hinting are deliberately out of scope.

36
docs/Website/index.html
View file

@ -301,8 +301,8 @@
</div> </div>
<div class="nv-card__body nv-text-secondary"> <div class="nv-card__body nv-text-secondary">
Four-level paging, multiple cores, preemptive scheduling, a calibrated Four-level paging, multiple cores, preemptive scheduling, a calibrated
LAPIC timer, real signal delivery and job control. 207 syscalls on LAPIC timer, real signal delivery and job control. Close to two
Linux numbering. hundred syscalls on Linux numbering.
</div> </div>
</article> </article>
@ -545,20 +545,46 @@
<h2 class="nv-heading">Measured, not marketed</h2> <h2 class="nv-heading">Measured, not marketed</h2>
</div> </div>
<!-- stats:begin -->
<dl class="nv-stat-group"> <dl class="nv-stat-group">
<div class="nv-stat"> <div class="nv-stat">
<dt class="nv-stat__label">lines in <code class="nv-code">kernel/</code> and <code class="nv-code">include/</code></dt> <dt class="nv-stat__label">lines of code in <code class="nv-code">kernel/</code> and <code class="nv-code">include/</code></dt>
<dd class="nv-stat__value">46,961</dd> <dd class="nv-stat__value">45,207</dd>
</div>
<div class="nv-stat">
<dt class="nv-stat__label">lines of comments alongside them</dt>
<dd class="nv-stat__value">16,111</dd>
</div>
<div class="nv-stat">
<dt class="nv-stat__label">first-party lines in all</dt>
<dd class="nv-stat__value">61,367</dd>
</div> </div>
<div class="nv-stat"> <div class="nv-stat">
<dt class="nv-stat__label">syscalls, Linux numbering</dt> <dt class="nv-stat__label">syscalls, Linux numbering</dt>
<dd class="nv-stat__value">207</dd> <dd class="nv-stat__value">184</dd>
</div>
<div class="nv-stat">
<dt class="nv-stat__label">drivers in the kernel</dt>
<dd class="nv-stat__value">15</dd>
</div>
<div class="nv-stat">
<dt class="nv-stat__label">Alpine packages in the live root</dt>
<dd class="nv-stat__value">88</dd>
</div>
<div class="nv-stat">
<dt class="nv-stat__label">commits</dt>
<dd class="nv-stat__value">253</dd>
</div>
<div class="nv-stat">
<dt class="nv-stat__label">days since the first one</dt>
<dd class="nv-stat__value">76</dd>
</div> </div>
<div class="nv-stat"> <div class="nv-stat">
<dt class="nv-stat__label">patches carried against fastfetch</dt> <dt class="nv-stat__label">patches carried against fastfetch</dt>
<dd class="nv-stat__value">0</dd> <dd class="nv-stat__value">0</dd>
</div> </div>
</dl> </dl>
<!-- stats:end -->
</section> </section>
<!-- ====================================================================== <!-- ======================================================================

217
tools/python/stats.py Normal file
View file

@ -0,0 +1,217 @@
#!/usr/bin/env python3
# SPDX-License-Identifier: GPL-2.0-or-later
"""The project's numbers, measured from the tree and written where they are shown.
python3 tools/python/stats.py rewrite README.md and the website
python3 tools/python/stats.py --print only print them
Both places carry the numbers between markers -- <!-- stats:begin --> and
<!-- stats:end --> -- and everything between them is regenerated. A number in
the README that somebody typed is a number that was true once; these are true
on the commit that ran this.
First-party code only: ports/ is other people's software, build/ is output,
and docs/Website/niveus is a vendored design system. Comment lines are counted
apart from code, because in this tree they are a large part of the work.
"""
import datetime
import os
import re
import subprocess
import sys
ROOT = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
FIRST_PARTY = ["kernel", "include", "boot", "lib", "cmd", "tools"]
LANGS = {
".c": "C", ".h": "C", ".S": "Assembly", ".s": "Assembly", ".asm": "Assembly",
".py": "Python", ".sh": "Shell", ".ps1": "PowerShell", ".pl": "Perl",
".ld": "Linker script", ".inc": "C",
}
GENERATED = re.compile(r"include/gfx/(logo|font)[^/]*\.h$|_data\.h$")
def git(*args):
return subprocess.run(["git", "-C", ROOT, *args], capture_output=True,
text=True, check=True).stdout
def tracked():
return [p for p in git("ls-files").splitlines()
if p.split("/")[0] in FIRST_PARTY and not GENERATED.search(p)]
def count(path, lang):
"""(code, comment, blank) lines. Block comments are followed for C-family
files; for the scripting languages a line starting with # is a comment."""
code = comment = blank = 0
in_block = False
try:
with open(os.path.join(ROOT, path), encoding="utf-8", errors="replace") as f:
lines = f.read().splitlines()
except OSError:
return 0, 0, 0
hash_comments = lang in ("Python", "Shell", "PowerShell", "Perl")
for raw in lines:
s = raw.strip()
if not s:
blank += 1
continue
if hash_comments:
if s.startswith("#") and not s.startswith("#!"):
comment += 1
else:
code += 1
continue
if in_block:
comment += 1
if "*/" in s:
in_block = False
if s.split("*/", 1)[1].strip():
comment -= 1
code += 1
continue
if s.startswith("//") or (lang == "Assembly" and s.startswith("#") and
not re.match(r"#\s*(include|define|if|endif|else|elif|undef)", s)):
comment += 1
elif s.startswith("/*"):
comment += 1
if "*/" not in s[2:]:
in_block = True
else:
code += 1
return code, comment, blank
def measure():
by_lang = {}
kernel = [0, 0]
files = 0
for p in tracked():
ext = os.path.splitext(p)[1]
lang = LANGS.get(ext)
if lang is None:
continue
c, m, _ = count(p, lang)
files += 1
t = by_lang.setdefault(lang, [0, 0])
t[0] += c
t[1] += m
if p.startswith(("kernel/", "include/")):
kernel[0] += c
kernel[1] += m
table = open(os.path.join(ROOT, "kernel/sys/table.inc")).read()
calls = re.findall(r"^SYSCALL\(\s*\d+,\s*(\w+),", table, re.M)
linux = [n for n in calls if n != "NO_LINUX"]
drivers = set()
for p in tracked():
if p.startswith("kernel/drivers/") and p.endswith(".c"):
parts = p.split("/")
drivers.add("/".join(parts[2:4]) if len(parts) > 4 else parts[-1])
drivers.add("i2c-hid") # kernel/acpi/i2c_hid.c, found through ACPI
programs = {p.split("/")[1] for p in tracked() if p.startswith("cmd/") and p.count("/") >= 2}
world = [l for l in open(os.path.join(ROOT, "ports/alpine/world")).read().splitlines()
if l.strip() and not l.startswith("#")]
lock = [l for l in open(os.path.join(ROOT, "ports/alpine/world.lock")).read().splitlines()
if l.strip() and not l.startswith("#")]
commits = int(git("rev-list", "--count", "HEAD").strip())
first = git("log", "--reverse", "--format=%ad", "--date=short").splitlines()[0]
days = (datetime.date.today() - datetime.date.fromisoformat(first)).days
code = sum(v[0] for v in by_lang.values())
comments = sum(v[1] for v in by_lang.values())
return {
"code": code, "comments": comments, "files": files,
"by_lang": sorted(by_lang.items(), key=lambda kv: -kv[1][0]),
"kernel_code": kernel[0], "kernel_comments": kernel[1],
"syscalls": len(calls), "linux_syscalls": len(linux),
"drivers": len(drivers), "programs": len(programs),
"world": len(world), "packages": len(lock),
"commits": commits, "first": first, "days": days,
}
def n(v):
return f"{v:,}"
def readme_block(s):
ratio = 100 * s["comments"] / max(1, s["code"] + s["comments"])
rows = "\n".join(f"| {lang} | {n(c)} | {n(m)} |" for lang, (c, m) in s["by_lang"])
return f"""<!-- stats:begin -->
| | |
|---|---|
| **First-party code** | {n(s['code'])} lines, plus {n(s['comments'])} lines of comments ({ratio:.0f}%) in {n(s['files'])} files |
| **Kernel** (`kernel/`, `include/`) | {n(s['kernel_code'])} lines of code, {n(s['kernel_comments'])} of comments |
| **Syscalls** | {s['syscalls']}, of which {s['linux_syscalls']} on Linux numbering |
| **Drivers** | {s['drivers']} in the kernel |
| **First-party programs** | {s['programs']} in `cmd/` |
| **Alpine packages in the root** | {s['packages']}, from {s['world']} asked for in `ports/alpine/world` |
| **History** | {n(s['commits'])} commits since {s['first']} ({s['days']} days) |
<details>
<summary>Lines by language</summary>
| Language | Code | Comments |
|---|--:|--:|
{rows}
</details>
<sub>Measured by <code>tools/python/stats.py</code>; first-party code only, so
<code>ports/</code> and generated tables are not counted.</sub>
<!-- stats:end -->"""
def site_block(s):
items = [
(f"lines of code in <code class=\"nv-code\">kernel/</code> and <code class=\"nv-code\">include/</code>", n(s["kernel_code"])),
("lines of comments alongside them", n(s["kernel_comments"])),
("first-party lines in all", n(s["code"])),
("syscalls, Linux numbering", str(s["linux_syscalls"])),
("drivers in the kernel", str(s["drivers"])),
("Alpine packages in the live root", str(s["packages"])),
("commits", n(s["commits"])),
("days since the first one", str(s["days"])),
("patches carried against fastfetch", "0"),
]
out = ["<!-- stats:begin -->", " <dl class=\"nv-stat-group\">"]
for label, value in items:
out += [" <div class=\"nv-stat\">",
f" <dt class=\"nv-stat__label\">{label}</dt>",
f" <dd class=\"nv-stat__value\">{value}</dd>",
" </div>"]
out += [" </dl>", " <!-- stats:end -->"]
return "\n".join(out)
def rewrite(path, block):
full = os.path.join(ROOT, path)
text = open(full, encoding="utf-8").read()
new, k = re.subn(r"<!-- stats:begin -->.*?<!-- stats:end -->", lambda _: block,
text, flags=re.S)
if k != 1:
sys.exit(f"stats: {path} needs exactly one stats:begin/stats:end pair")
if new != text:
open(full, "w", encoding="utf-8").write(new)
print(f"stats: rewrote {path}")
def main():
s = measure()
if "--print" in sys.argv:
for k, v in s.items():
print(f"{k}: {v}")
return
rewrite("README.md", readme_block(s))
rewrite("docs/Website/index.html", site_block(s))
if __name__ == "__main__":
main()