Puppy Linux Discussion Forum

gyro · Posted: Tue 12 Jul 2016, 08:33 Post subject:

A question:
Given "huge" kernels, is the zdrv critical to a boot?
By critical, I mean don't continue on with the boot without it.

My current version considers puppy...sfs as critical, everything else is optional.

gyro

starhawk · Posted: Tue 12 Jul 2016, 10:10 Post subject:

zdrv is your drivers, gyro. Not just for your network card... Wink

Sailor Enceladus · Joined: 22 Feb 2016 Posts: 1560

When I boot Slacko 6.3 without the zdrv it makes it to desktop, but my mouse/trackpad doesn't work, even with using the dead mouse trick, and neither my wireless or wired ethernet show up. I did manage to hit Alt-F1 to get to Pmount to mount a drive that had the zdrv in it and put it back in with terminal commands though. Interestingly, "depmod" runs when I remove the zdrv but not when I keep it, and the pcmcia part says 1 2 3 4 5 6 7 8 9 10 before continuing. Maybe this info will be useful, or not...

gyro · Posted: Tue 12 Jul 2016, 13:59 Post subject:

I have accidently booted without a zdrv, and it did boot to a desktop, but of course without any kernel modules, lots of hardware simply did not work.

I guess my question really is:
If the init script detects that there is no zdrv, should it abandon the boot or continue?

My current thinking is that it should continue, since it will most likely still boot to a desktop (although it will most likely be a crippled one).

Whereas, if the main puppy...sfs is missing the boot is abandoned.

gyro

learnhow2code · Joined: 12 Jun 2016 Posts: 1015

01micko · Joined: 11 Oct 2008 Posts: 8694 Location: qld

Hi gyro

Maybe someone could put the zdrv in the main sfs as part of a manual remaster? In that case boot should continue if zdrv is not found.

2 c

gyro · Posted: Wed 13 Jul 2016, 13:45 Post subject:

Hi 01micko,

I think you are agreeing with me, only the main puppy...sfs should be considered critical.

gyro

gyro · Posted: Thu 14 Jul 2016, 13:40 Post subject:

Another question:
The "upgrade" code in the current init script, is it really needed?
Does any one know?
gyro

jlst · Joined: 23 Nov 2012 Posts: 571

I triggered involuntarily one day... when i replaced only the sfs files but forgot to delete the savefile, apparently it did detect a new version.

That's the only time it happened to me since 2012

gcmartin · Joined: 14 Oct 2005 Posts: 6730 Location: Earth

Yes, as I remember, when booting a newer version of a PUP from CD/DVD, I am often presented with an option to upgrade prior version session data.

Never knew if the init code was doing this.

gyro · Posted: Fri 15 Jul 2016, 05:15 Post subject:

Yes, I know it's there, and it's triggered if it thinks the version being booted is newer than the one in the save.
But, does it do anything that is not done by "rc.update"?
If nobody knows, then I'll have to sort it out the hard way.
(I never do updrades, always start with new savefolder.)

gyro

gyro · Posted: Fri 15 Jul 2016, 06:32 Post subject:

The current version does boot ok from hd and usb.
It contains support for pupmodes 5, 6, 7, 12, 13, and 77. Although 6, 7 and 77 are untested.

If I don't post any thing in the next week, it's because I do have a life apart from puppy.

gyro

jlst · Joined: 23 Nov 2012 Posts: 571

I think rc.update should take care of this.

But there is something to take into account.

The initrd DISTRO_SPECS is copied to /etc on each boot... rc.shutdown should copy that file to /var/log/DISTRO_SPECS .. so rc.update can try to determine if a newer version is being booted.

Regarding testing for other pupmodes.. well, this is the difficult part.. actually the reality is discouraging.

Moose On The Loose · Joined: 24 Feb 2011 Posts: 778

gyro · Posted: Sun 24 Jul 2016, 04:58 Post subject:

@Moose On The Loose,

I assume that you are referring to the code processing the "pfix=" boot parameter?

gyro