On Wed, 15 Nov 2006 11:23:58 -0800, Trevor Hamm <Trevor_Hamm@pmc-sierra.com>
wrote:
> > Just for confirm:
> > 1. This can happen on latest lmo git tree or 2.6.19-rc5.
> > 2. UP kernel.
> > 3. No L2 cache.
> > 4. icache and dcache are both virtually indexed and physically tagged.
> > All correct?
> >
>
> Except for #1 which I haven't tried (we need to get this working
> with 2.6.18), all correct. The caches are 64k, 4-way.
Thanks. I re-read your report and wonder why this happens only on
power-up time.
If flush_data_cache_page or update_mmu_cache were broken, the problem
can happen on every restart.
The entire icache/dcache should be cleared at end of r4k_cache_init()
at least. MIPS34K could have some magical state (lock bit or
something) which is not initialized by local_r4k___flush_cache_all()?
---
Atsushi Nemoto
|