Re: developer accessible machines
On Tue, May 13, 2008 at 04:55:22PM -0700, Matt Taggart wrote:
>
> The problem isn't lack of machines, it's lack of a known good kernel
> package for the existing machines. My understanding is that it needs to
> stand up to a buildd, have all security fixes, and be installable on etch.
> Nobody seems to be working on making one though :( In the past I was able
> to make newer kernels from unstable install on paer with some backported
> tools and working around stuff in postinst, but I haven't tried lately. If
> someone could come up with a nice recipe on how to do this we could use it
> for the existing machines.
The only kernel I ever successfully used on my J5600 is Debian's
2.6.22-3-parisc-smp/2.6.22-3-parisc64-smp.
Anyway I know a secret stroke with rtorrent which completely kills
the box. No SysRq is available though configured and tested when the
system is still alive. I hope it is not an hardware failure although
the system display leaves much less hopes.
Debian packages for 2.6.23 had never been built. 2.6.24 and 2.6.25 do
not boot, probably for different reasons I do not remember right now.
Vanilla 2.6.25 and 2.6.26-rc1 both boot at 32 bit, instead at 64 bit
one CPU gets stuck.
Kyle has been informed [0], I hope he finds few more minutes to dig
into this. I am available to test patches. I am sure I would find the
fix sooner or later, say in few months... :)
thanks,
Domenico
[0] http://lkml.org/lkml/2008/4/22/593
-----[ Domenico Andreoli, aka cavok
--[ http://www.dandreoli.com/gpgkey.asc
---[ 3A0F 2F80 F79C 678A 8936 4FEE 0677 9033 A20E BC50
Reply to: