[U-Boot] [PATCH v2 06/11] efi_loader: Allow SMBIOS tables in highmem

[Alexander Graf]

On 14.06.18 21:26, Heinrich Schuchardt wrote:
> On 06/14/2018 09:13 PM, Alexander Graf wrote:
>>
>>
>> On 14.06.18 21:01, Simon Glass wrote:
>>> Hi Alex,
>>>
>>> On 14 June 2018 at 12:22, Alexander Graf <agraf at suse.de> wrote:
>>>> We try hard to make sure that SMBIOS tables live in the lower 32bit.
>>>> However, when we can not find any space at all there, we should not
>>>> error out but instead just fall back to map them in the full address
>>>> space instead.
>>>>
>>>> Signed-off-by: Alexander Graf <agraf at suse.de>
>>>> ---
>>>>  lib/efi_loader/efi_smbios.c | 11 +++++++++--
>>>>  1 file changed, 9 insertions(+), 2 deletions(-)
>>>
>>> Does that actually work? I thought the addresses in the table were
>>> always 32-bit?
>>
>>
>> There is only a single table reference which indeed is 32bit:
>> se->struct_table_address.
>>
>> That address however is unused in the EFI case usually, because the
>> SMBIOS information is already encapsulated in a table, so there's no
>> need to search through address space for a _DMI_ entry:
>>
>>   https://github.com/mirror/dmidecode/blob/master/dmidecode.c#L5122
>>
>>
>> Alex
>>
> We still have an open issue with E820 memory reservations not observed
> by Linux because they are not mirrored in the memory map. Could the same
> happen with smbios?

The SMBIOS tables get allocated from the EFI pool in
lib/efi_loader/efi_smbios.c.

So the e820 patch fell through the cracks?


Alex