Documentation of kernel sources

Adam Mckee (
Mon, 14 Jul 1997 16:58:18 -0600 (CST)

Hello all.

I am a kernel-hacker wanna-be myself. I have been trying to understand
the basics of how control of the machine is transferred between the kernel
and processes. To this end, I am working on a patch to the scheduler.
There's a lot to be said for "learn-by-doing"... identify a simple project
and just start hacking... a lot of gaps in your knowledge will be revealed
in the (painful) debugging process. However, the ideal thing in terms of
attracting new talent to the effort would be a very well-written reference
to the core kernel code, device driver interfaces, etc. Some walk-thrus
of drivers for different types of devices (block,char,scsi,etc.) would be
nice also.

I think there is a commercial opportunity here... if O'Reilly actually
released a good book that documents the core kernel code, I'd buy it in a
second and I'm sure a lot of others would also. Documentation like this
could bring a lot of new talent to the effort. I.e. maybe some people
could actually stop whining and start coding!

-- Adam

On Mon, 14 Jul 1997, Steven N. Hirsch wrote:

> On Mon, 14 Jul 1997, Geert Uytterhoeven wrote:
> > On Mon, 14 Jul 1997, Steven N. Hirsch wrote:
> >
> > > I almost guarantee that 10% more effort put into documentation and
> > > commenting would reap great rewards in terms of contributed input from the
> > > community at large. Then, you CAN go off and design and implement the
> > > "latest and greatest" - AFTER having paved the way for others to maintain
> > > and fine-tune the previous "latest and greatest".
> >
> > IMHO, it takes more effort to document the code than to write it. That's my
> > experience with kernel hacking (and I touched a lot of the Linux/m68k kernel
> > code). So your 10% rule is not correct (and as a professional developer,
> > shouldn't you know that documentation is the hardest part? :-)
> Oh absolutely true. I never meant to imply that 10% more time would
> result in complete documentation <g>. Just trying to get a point
> across..
> > And the `real' kernel hackers have so much to do, and so many new ideas, that
> > they'd rather do what they like most: i.e. turning ideas into new code...
> And they do that well. I would be glad to document kernel source -
> problem is that I don't know what the h*ll is going on in most places!
> In another message, Alan Cox implies that something might be coming down
> the pike from O'Reilly. I bought a copy of Linux Kernel Internals, and it
> was of some value (if a bit dated). Perhaps it's time for a re-read of
> that and "Modern Unix Internals".
> Just so hard to get that first foot in the door!
> Steve