Re: Lots of kernel problems with U30
In article <[🔎] 42664959.90208@yannos.com> you write:
>I've two U30 Creator 3D and debians kernel are pretty buggy on them.
I have one, with no problems recently.
>=> 2.4.18 : Works great but no support for SATA (U30 loaded with 2x50 Go
>SCSI SCA + 300 Go SATA + Silicon Image controleur + Adaptec 39160 +
>DLT7000 + DDS2)
I Never tried SATA.
>=> 2.4.27-2 : No network support
Works fine for me.
>=> 2.6.8-2 : Keyboard problems
As mentioned by others, 2.6 uses different keymaps. I'm running 2.6.8
(smp) on an ultra 2, but have not tried 2.6 on the ultra 30.
>I know U30 Creator3D is not the most common Sun station, but it's the
>first on which i've such number of kernel problems.
It's not much differnt than the more common Ultra 60.
>I wanted to know how can i help to test/debug Sparc port.
Testing rc3 or current dailys of the debian-installer and submitting
install reports wouldn't hurt. Note that some problems are related to
memory (silo might not find the initrd) or disk (partconf gets disk
size wrong on 2nd and later drives) configurations.
>I've a large number of Sun stations:
sun4u: (should mostly be fine)
>2 x U30 3D
>2 x U1 3D
>1 x U1
sun4m: (2.6 kernel doesn't work last I heard, I think rc3 has known bug)
>1 x SS20 DualRoss 200 Mhz
>1 x SS20
>1 x SS10
>1 x SS5
>6 x LX
>3 x Classic
sun4c: (all d-i attempts I know of failed, not much interst in supporting)
(upgrading from woody may work if you install a backport 2.4 kernel first)
>1 x SS2 + weitek powerUp
>5 x SS2
>25 x SS1+
>4 x IPX
>12 x IPC
sun4d with sbus and vme: (or is it sun4m?) (vme won't work, sun4d is mp sun4c)
(may not work mp)
>2 x 670 MP
sun4: (never supported by linux)
>2 x 4/360
sun3: (ask on a 68k mailing list if they are supported)
>2 x 3/60
I sorted the list more or less in order of usability. 32 megs is
considered low memory by d-i, 24 megs might not work.
I've tested d-i on all the sun4u/sun4m types you mention except the
ss10, but not rc3.
--
Blars Blarson blarson@blars.org
http://www.blars.org/blars.html
With Microsoft, failure is not an option. It is a standard feature.
Reply to: