Re: [PAE support] patch review

[Jean-Yves Migeon <jeanyves.migeon%free.fr@localhost>]

On 17.07.2010 19:32, David Young wrote:
> On Sat, Jul 17, 2010 at 03:52:19PM +0200, Jean-Yves Migeon wrote:
>> Hi lists,
>>
>> You will find attached the patch for PAE support [1]; diff is against
>> latest current. This one is for a last review before commit.
> 
> On x86 the bus_addr_t width follows the paddr_t width (we have "typedef
> paddr_t bus_addr_t" in x86/include/bus.h).  That makes sense on non-PAE
> i386 and on amd64, but on PAE i386?

Yes, for memory above 4GiB, it does make sense (provided there is
support for it).

FWIW, this is the biggest reason why I expect troubles when mixing PAE
and !PAE in kernel and modules, the paddr_t/bus_addr_t ABI breakage (drm
modules are affected, for example).

And it cannot be solved easily (at least, from my PoV), as that would
force 64 paddr_t even in i386 normal case, which could get things messy
if not done right (consider pd_entry_t/pt_entry_t 32 bits with 64 bits
paddr_t, for example). But still possible. I don't have time for it
right now, though :/

-- 
Jean-Yves Migeon
jeanyves.migeon%free.fr@localhost