[PATCH 6/6] arm64/efi: remove idmap manipulations from UEFI code

[Grant Likely]

On Fri, Oct 24, 2014 at 2:53 PM, Ard Biesheuvel
<ard.biesheuvel at linaro.org> wrote:
> On 24 October 2014 15:41, Grant Likely <grant.likely at linaro.org> wrote:
>> On Fri, Oct 24, 2014 at 1:39 PM, Ard Biesheuvel
>> <ard.biesheuvel at linaro.org> wrote:
>>> Now that we have moved the call to SetVirtualAddressMap() to the stub,
>>> UEFI has no use for the ID map, so we can drop the code that installs
>>> ID mappings for UEFI memory regions.
>>>
>>> Signed-off-by: Ard Biesheuvel <ard.biesheuvel at linaro.org>
>>
>> I have to say, this series makes me happy. :-)
>>
>> This method will go a long way to catching UEFI implementations that
>> do incorrect things after exitbootservices is called. I'm assuming
>> that any attempt to access a region that boot services has not
>> requested will get trapped by the kernel, correct?
>>
>
> If we really want to catch firmware problems, we should probably wipe
> all boot services regions between the calls to ExitBootServices() and
> SetVirtualAddressMap(). Mark Salter's original approach here was
> fairly cautious here, i.e., reserving boot services regions until
> after the call to SetVirtualAddressMap(), but there is no point in
> doing that for kexec, that's why I removed it.

I quite like that idea. Let's do that and see if anyone screams in agony.

g.