Re: [gentoo-dev] Re: Let's redesign the entire filesystem!

[Joshua Kinard <kumba@gentoo.org>]

[-- Attachment #1: Type: text/plain, Size: 2463 bytes --]

On 03/14/2012 11:04, Greg KH wrote:

> 
> Not always, no, it isn't obvious that something didn't start up
> correctly, or that it didn't fully load properly.  Some programs later
> on recover and handle things better.


I'm well aware of what I run on my own box, and when something isn't
running, I figure it out pretty quickly.  I tested udev-181 in a Gentoo VM
that I put together recently, giving it a separate /usr, and made sure that
CONFIG_DEVTMPFS was enbaled.  The only service that failed to load properly
at startup was udev, specifically because udevadm couldn't locate libkmod
(likely because kmod installs into /usr/lib*, which wasn't available yet
because I also don't use an initramfs in my kernel).  Everything else worked
fine, and udev later started properly once localmount was complete.

I even tried, out of curiosity, to tweak things and moved udev from sysinit
to boot and then to default runlevel.  In 'boot', udevadm still fails to
load, so no change.  In 'default', only net.lo failed because the 'lo'
device didn't yet exist until after udev was running.  udev itself loaded
fine, and the networking scripts restarted themselves.

So with the one exception of networking, which in Linux, has never created
/dev nodes (has to be some historical piece on why), one almost doesn't need
udev at boot to even get things working on a very simple setup like mine.
And since udev is the one service that didn't load correctly, one COULD put
forth the hypothesis that it is udev that is "broken".  But I doubt that
will get much traction, right?

This does lead me to wonder if a light-weight udev could exist that lacks
half or more of the functionality of the current udev.  I'll be honest, I've
only edited my udev rules file once, and that was only when I installed a
Sun Happy Meal quad ethernet card in which all four ports utilize the same
MAC address and udev doesn't handle this very gracefully (if I had Solaris,
I could edit the card's firmware and change this setting).

Devtmpfs quite literally handles 98% of my particular usage scenario.  Does
that apply to everyone?  Nope.  Just an interesting observation.

-- 
Joshua Kinard
Gentoo/MIPS
kumba@gentoo.org
4096R/D25D95E3 2011-03-28

"The past tempts us, the present confuses us, the future frightens us.  And
our lives slip away, moment by moment, lost in that vast, terrible in-between."

--Emperor Turhan, Centauri Republic


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 834 bytes --]