[Clfs-dev] Sparc64-SVN-20080616

[wwh04660 at ucmo.edu]

On Jun 20, 2008, at 4:50 PM, Ken Moffat wrote:

> Interesting results.  I haven't checked the detail of your glibc
> against what I logged for ppc64 in the "clfs-1.1.0" ticket, but some
> of it certainly had a familiar feel to it.
>
> I've just uploaded my pure64 and powerpc64 patches for 4.2.4
> (pure64 had rejections, as did powerpc64 - the new pure64 works on
> x86_64-64, I haven't had time to start testing ppc64 (multilib) yet).
>
> Fingers crossed for your 2.6.24.7.
>
> ĸen

Ahoy Ken,

2.6.24 kernel series are a no go for my machine. Framebuffer gets  
garbled and only a few audio plays with aplay, unlike most all of them  
that are a valid format do in 2.6.25.7 or 2.6.23.1. Prom console works  
okay though. Something in the 2.6.24 kernel is messing with the audio  
and framebuffer ( I use permedia2 ) and the audio is the crystal audio  
device that comes with the Ultras.

I've asked Aaediwen to see what happens when he uses it with his Ultra  
60. He uses a UPA device (Elite3D) and am curious to see what happens.  
His audio isn't working for some reason. So I guess he can't test that  
part.

I've got people asking me about 2.6.25 headers, no clue on it as I  
haven't touched them.
Anyways, I think if 2.6.24.7 headers work, a note in the book about  
using the 2.6.24 series kernels with sparc might be needed heh. It's  
not usable with my machine for sure, even with a defconfig.

Also I have this with a 2.6.24 series kernel as well:

WARNING: vmlinux.o(.text+0x3f90c): Section mismatch: reference  
to .init.text:sun4v_ktsb_register (between 'smp_callin' and  
'smp_fill_in_sib_core_maps')
WARNING: vmlinux.o(.text+0x402ec): Section mismatch: reference  
to .init.text:sun4v_register_mondo_queues (between 'after_lock_tlb'  
and 'hv_cpu_startup')
WARNING: vmlinux.o(.text+0x4052c): Section mismatch: reference  
to .init.text:sun4v_register_mondo_queues (between 'hv_cpu_startup'  
and 'sys32_exit')

Looking around the LKML showed a few things, just saw comments saying  
it was bogus and people try this, and people giving up, etc.

William