hckrnws

Some related stuff:

In 2004, Gavin Barraclough’s mini-OS [0] won the IOCCC, packing a 32-bit multitasking operating system for x86 computers, with GUI and filesystem, support for loading and executing user applications in ELF binary format, with PS/2 mouse and keyboard drivers, VESA graphics, a command shell, and an application into 3.5 KB of highly obfuscated C code.

In 2021, Justine Tunney wrote SectorLISP [1], a Lisp implementation that fits into a bootsector and is able to run McCarthy’s metacircular evaluator.

[0]: https://www.ioccc.org/2004/gavin/index.html [1]: https://github.com/jart/sectorlisp

Two questions:

1. I just saw how str_print is implemented. It's so short even though it's asm. Is this why nul-terminated strings were so popular and became the default in C? Would pascal strings be much longer/uglier/harder in asm?

2. Why is str_print repeated in multiple files? How would you do code sharing in asm? I assume str_print is currently not "static" and you'd have to make it so via linking or something, and then be able to get its address using an asm macro or something?

1. If you look through the commit history, you'll see that the first implementation was actually with Pascal strings.

Printing with Pascal strings is actually shorter (you skip the null test, basically), but constructing Pascal strings to pass as an argument when they are not constants yielded much more code to prepare for that call. Had I had more leeway, I would have used Pascal strings, it much less headache.

2. Files in `/bin` all include from the SDK. You can pretty much do the same for utility functions.

The includes, at least in nasm, are very much like copy-pasted code (or includes in C for that matter), and then you can just jump/call to the label.

I did not do it because I haven't been able to get nasm to optimize away the code that I don't use, and I didn't want to bloat the binaries or make a file for a 5LOC function.

All in all not good reasons in general, but it made sense to me in this context.

Thanks for answering my questions. Your project is really really interesting.

Two more questions if you find some spare time:

3. Why does it use tty for interrupts instead of directly calling int 10?

4. How does this even print to the screen or use a tty in the first place? Is it just something inherent in bios api?

Hey, thanks for your interest in this project!

3. The tty interrupt advances the cursor along with printing. So, once again, I do it to save on some instructions. In the first iterations I wanted to retain more control (by printing and moving as separate operations) so that I could reuse this across the board, but eventually I ran out of space.

4. I am relying heavily on BIOS interrupts, which are criminally underdocumented. The most reliable source is Ralph Brown's documentation[1] which is very far from what I was expecting to be authoritative documentation. Turns out this collection is really good and basically _the_ source of truth for BIOS interrupts.

To answer your question, yes, this is basically calling the BIOS API.

[1]: https://wiki.osdev.org/Ralf_Brown's_Interrupt_List

THIS is the bible for BIOS APIs"

https://bitsavers.trailing-edge.com/pdf/ibm/pc/ps2/PS2_and_P...

Complete with reference assembler source code.

Oh boy, this is amazing! Thanks for the reference

The linker probably compacts all of the code blocks and generally futher optimizes the final binary size.

I would have assumed the same, but I haven't managed. On the other hand, I did not tinker too much with all these toggles; it's such a little amount of shared code (which is also partially different in some cases) that didn't particularly make sense to me.

If you know how to make it happen and/or want to contribute, hit me up (:

Very cool! I have to ask: what would the total size be if the package included the bios functions?

Also: what could be done if the size limit were 8kbyte like the mask-rom bios days?

Thanks for pointing me towards the bosh emulator.

Hey, thanks for taking a look!

On the former, I have no idea how to estimate BIOS functions size. Maybe I could just peek into an image and get a sense for it...

On the latter, with a 16x increase in available space, I guess I would do a much more thorough work in putting guardrails in place.

The API currently comes with a couple of traps (e.g., file names can be duplicated, processes are cooperative, all file operations perform disk I/O...) and it essentially requires guest applications to know about BIOS services in order to function.

Another sticky point I wish I had the space to address better are calling conventions, which I had to get rid of almost immediately to save on instructions.

> Thanks for pointing me towards the bosh emulator.

You're welcome! Bochs is such a nice tool which I discovered only for this project as well. It was a no-brainer, since I got no way to debug 16-bit assembly from QEMU (unless you go off and fork it[1])

[1]: https://gist.github.com/Theldus/4e1efc07ec13fb84fa10c2f3d054...

what would the total size be if the package included the bios functions?

Probably a few dozen to over a hundred KB, maybe even over a MB, depending on the era of machine and what it has installed; e.g. the GPU option ROM would be included if you use int 10h, int 13h might be hooked by a disk adapter, and if you use int 16h to read from a USB keyboard, that'll go through the BIOS' USB stack which normally includes some code in SMM too.

I have a more general question - what is the minimum that is needed to qualify as an operating system? Is there something agreed on generally? Searching operating system minimum requirements leads to the wrong kind of info for me....

Honestly, I made it up :)

I thought about what would be the minimum I have to build in order to run some userland software that does "something". That to me looked like: spawn guest applications, make them persist something.

With slightly more leeway, I would probably do memory management as the next thing (besides what I mentioned in another thread here)

BIOS is underrated. Basically the driver portion of a DOS- (or CP/M)-like operating system. As demonstrated, you don't need to add too much to it (program loader, simple file system, maybe a command shell and system utilities if you are ambitious) to get a functional mini-DOS.

Well that's cool. Does the name stand for something?

The -le suffix is used in south of Germany for the small version of something. So OSle stands for small OS.

I'm not a native speaker, so maybe somebody else can paint a better picture. I used it just because part of my extended family comes from there (:

EDIT: s/prefix/suffix/

Similar in English, the ie suffix is used to create a diminutive. Sweet -> sweetie. You can make cute cuter by saying cutie.

I live in Alsace, which is in France but has a German-like dialect (Alemannic)

https://en.m.wikipedia.org/wiki/Alsatian_dialect

-ele is used a lot to denote something small, cute, adorable; maybe think of it as kind of like ちび (chibi) or -ちゃん (-chan) in Japanese.

Mann (man) => Mannele https://cookingwithbrendagantt.net/mannele-st-nicholas-bread...

Katz (cat) => katzele (kitty)

The suffix can be liberally (ab)used with any - native or foreign - word or (sur)name to emphatic or comedic effect.

Here I kinda guessed the -le use was such but around here I would have said "OSele" (oh-ess-uh-luh)

as seen also in Spätzle, Müsli, or, to pick something more relevant on HN, the words Brötli (or Zöpfli)

-li is a different version of the same ending

*suffix.

A prefix goes before something.

Indeed. Thanks for the correction; I edited the original message

On projects like this, where the IMG is small enough, I would think it was ideal to include osle.img with the zip.

cool stuff, like you still fit quite a bit in there too, 510 bytes can be tricky.

if you want an ahci controller to 'see' it, it will need partition table too, which will make it even less bytes (or maybe cleverly encoded)

I went back and forth about the file system and disk stuff a fair bunch, to be honest. Most of it, as you say, was mostly due to wrestling the space constraints.

If one day I'll give in and take the shell out or go multi-stage, I will definitely look at that.

Maybe it's worth blogging about the journey; it's been a few weeks of merciless trade-offs to reach a usable API. It can make for a fun read (:

Thanks for taking a look!

haha, well all the best! its a cool project. i am happy i can forgot about BIOS and went UEFI haha. remember so many tedious nights trying to get an mbr to load an elf file and init x64 mode in one go :'). uefi (edk2) is a blessing if you come from BIOS land (tho mybe less fun in a way!)

What sectors contain is irrelevant to AHCI. As long as the BIOS contains the appropriate interface to a block device, it can be used.

the BIOS will recognize block devices as being of certain type and present them to controllers.

if you do not put partition table, qemu AHCI controller will not recognize disk as bootable and u cant use SATA. with only the magic footer at the end of mbr, it will only work on IDE controller.

try it.

the BIOS will recognize block devices as being of certain type and present them to controllers.

What exactly do you mean by that? Device discovery proceeds from the root (usually PCIe bus, after CPU-specific init) to the leaves, not the other way around.

qemu AHCI controller

That's its problem then. This isn't a problem on real hardware.

All professors should be doing this decades ago right ?

Why? That is to say: it's a really cool project, and clearly a labor of love, but from an academic perspective it's a collection of x86-specific commands.

I must be honest. Professors are not doing their good job here.