[Cbe-oss-dev] [PATCH] fix reboot fail on kexec for CBE

[Benjamin Herrenschmidt]

> That is how it works. On the way down, the kexec/kdump code calls
> ppc_md.hpte_clear_all() from kexec_sequence (kernel/misc_64.S).
> 
> On cell hpte_clear_all() calls into native_hpte_clear(), there are known
> TODOs in there, Luke was working on those. I think we can just adapt
> native_hpte_clear() to work on cell, but if we need platform specific
> hacks, then cell needs its own separate clear function.

Note that if the "magic SPRs" method above is actually working (though
we also need to handle SPEs I beleive, which this doesn't do, but it
might be better to have that done when initializing spufs instead), it's
probably worth using it.

That's a common problem with our native hpte clear. It needs a per-CPU
function to flush the TLB, though we could default to a generic one
based on reading the hash table (what's Luke is working on). Having a
CPU specific hook would allow significant performance improvements.

Maybe worth adding a flush_tlb() hook in cputable ?

Ben.