Gentoo Websites Logo
Go to: Gentoo Home Documentation Forums Lists Bugs Planet Store Wiki Get Gentoo!
Bug 143719 - OldWorld Beige G3 fails to BootX boot with 2006.1 Universal CD
Summary: OldWorld Beige G3 fails to BootX boot with 2006.1 Universal CD
Status: RESOLVED FIXED
Alias: None
Product: Gentoo Release Media
Classification: Unclassified
Component: LiveCD/DVD/USB (show other bugs)
Hardware: PPC Linux
: High normal
Assignee: Gentoo Release Team
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2006-08-12 14:44 UTC by michael@smith-li.com
Modified: 2007-05-13 22:45 UTC (History)
1 user (show)

See Also:
Package list:
Runtime testing required: ---


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description michael@smith-li.com 2006-08-12 14:44:37 UTC
Using the instructions from the handbook, I booted into OS 9.2.2 and tried to use BootX to start the 2006.1 LiveCD. The only way I deviated from the handbook was to use "apple" and "apple.igz" instead of ppc32 and ppc32.igz.

When I clicked "Linux" in BootX, an initial screen appeared that said "Welcome to linux kernel <version>" and described the very initial part of the boot process. That screen didn't change for several minutes. Finally after a long while (3-10 minutes) the machine rebooted itself.

For the record, this machine does boot properly into the 2006.0 LiveCD using BootX and the ppc32 kernel and image.
Comment 1 Lars Weiler (RETIRED) gentoo-dev 2006-08-12 14:54:25 UTC
(In reply to comment #0)
> Using the instructions from the handbook, I booted into OS 9.2.2 and tried to
> use BootX to start the 2006.1 LiveCD. The only way I deviated from the handbook
> was to use "apple" and "apple.igz" instead of ppc32 and ppc32.igz.

That is already changed in the 2006.1-handbook.  But thanks for the reminder.

> When I clicked "Linux" in BootX, an initial screen appeared that said "Welcome
> to linux kernel <version>" and described the very initial part of the boot
> process. That screen didn't change for several minutes. Finally after a long
> while (3-10 minutes) the machine rebooted itself.

Then you are further than I am...  On my OldWorld beige G3 I see only a scrambled screen, but the machine reboots after several minutes.

> For the record, this machine does boot properly into the 2006.0 LiveCD using
> BootX and the ppc32 kernel and image.

That looks really bad for OldWorld-Support in 2006.1.  Probably it's caused by 2.6.17 or the new toolchain?

Comments from the rest of the ppc-team are appreciated.

Comment 2 Chris Gianelloni (RETIRED) gentoo-dev 2006-08-23 10:11:58 UTC
Since this likely isn't fixed for 2006.1, I'm removing the RTppc from it.
Comment 3 Joe Jezak (RETIRED) gentoo-dev 2006-08-23 10:49:28 UTC
It actually appears that there is BootX memory corruption with newer (2.6.16 and newer) kernels.  I think what I'll end up doing a seperate OW kernel like I did for 2005.0 once this issue is worked out. I just got a new MB for my OW machine, so once I get that replaced, I'll see if I can work something out.

See this thread (but the bug report is not exact, we certainly had OW machines booting with 2.6.15):
http://ozlabs.org/pipermail/linuxppc-dev/2006-August/024944.html
Comment 4 Rory 2006-09-23 08:52:15 UTC
Hi guys,

If confirmation helps I'm in the same boat. 

Powermac 5500/250 96mb ram clean install of MacOS 9 on a 200m partition and BootX 1.2.2 Using apple and apple.igz as per the handbook exactly and the disc image "install-ppc-minimal-2006.1.iso" (gentoo torrent).

My run ends about 15 odd lines in as follows:


Blah   - - - Kernel 2.6.17   blah, 

Total space used by parameters & ramdisk: 002f8000
Prep boot parameters ...
prep BAT ...
idmach(): done
MMU:enter
MMU:hwinit
MMU:mapin
MMU:setio
MMU:exit 

<<DIES HERE after intitial cd and HDD drive activty which stops, then the cd just spins down. It then sits like this for 2-4 min then re-boots>>

Extra attempts made with other included kernel parameters were: "nosmp" then "nodma", "acpi=off" and finaly "video=atyfb", all unsucessfull. 

The only other thing I've noticed is that BootX consistently fails to hold my Ramdisk size Option after boot. It must be entered every single time. Either way the line: "Total space used by parameters & ramdisk: 002f8000", never alters regardless of what size ramdisk i set.

Hope this helps.












> It actually appears that there is BootX memory corruption with newer (2.6.16
> and newer) kernels.  I think what I'll end up doing a seperate OW kernel like I
> did for 2005.0 once this issue is worked out. I just got a new MB for my OW
> machine, so once I get that replaced, I'll see if I can work something out.
> 
> See this thread (but the bug report is not exact, we certainly had OW machines
> booting with 2.6.15):
> http://ozlabs.org/pipermail/linuxppc-dev/2006-August/024944.html
> 

Comment 5 Lars Weiler (RETIRED) gentoo-dev 2006-09-23 14:06:34 UTC
It's fixed in 2.6.18.  I created a new CD for testing purposes: http://gentooexperimental.org/~pylon/ISOs/ (ask me in IRC for login)

I tested it on a beige G3 and it works! :-)

Let's see that I can push that CD into experimental-space on gentoo-mirrors, when it has been tested by some more persons.
Comment 6 Rory 2006-09-23 20:58:09 UTC
excellent.

I'd love to help test it but I am bandwith poor. Took me a week to get the other ISO. is it just the kernel or was other stuff changed too?. I'd appreciate an email when the fix is publicly available tho. :)



> It's fixed in 2.6.18.  I created a new CD for testing purposes:
> http://gentooexperimental.org/~pylon/ISOs/ (ask me in IRC for login)
> 
> I tested it on a beige G3 and it works! :-)
> 
> Let's see that I can push that CD into experimental-space on gentoo-mirrors,
> when it has been tested by some more persons.
> 
Comment 7 Lars Weiler (RETIRED) gentoo-dev 2007-05-10 15:02:27 UTC
2007.0 is out and works on OldWorld Macs (just verified).
Comment 8 simon 2007-05-13 08:18:48 UTC
2007.0 Boot CD works fine, but then, after having compiled and installed the kernel on the HD, and rebooting, same problem:
boot process dies after MMU:exit
Comment 9 Andrew Gaffney (RETIRED) gentoo-dev 2007-05-13 22:45:54 UTC
That's a completely different bug. Please file a bug for that or join the #gentoo-ppc channel on freenode to get help for it.