Re: [PATCH v7 1/8] mm/memblock: Tag memblocks with crypto capabilities

[Martin Fernandez <martin.fernandez@eclypsium.com>]

On 4/26/22, Mike Rapoport <rppt@kernel.org> wrote:
> On Mon, Apr 25, 2022 at 02:15:19PM -0300, Martin Fernandez wrote:
>> Add the capability to mark regions of the memory memory_type able of
>> hardware memory encryption.
>>
>> Also add the capability to query if all regions of a memory node are
>> able to do hardware memory encryption to call it when initializing the
>> nodes. Warn the user if a node has both encryptable and
>> non-encryptable regions.
>>
>> Signed-off-by: Martin Fernandez <martin.fernandez@eclypsium.com>
>> ---
>>  include/linux/memblock.h |  5 ++++
>>  mm/memblock.c            | 62 ++++++++++++++++++++++++++++++++++++++++
>>  2 files changed, 67 insertions(+)
>>
>> diff --git a/include/linux/memblock.h b/include/linux/memblock.h
>> index 50ad19662a32..00c4f1a20335 100644
>> --- a/include/linux/memblock.h
>> +++ b/include/linux/memblock.h
>> @@ -40,6 +40,7 @@ extern unsigned long long max_possible_pfn;
>>   * via a driver, and never indicated in the firmware-provided memory map
>> as
>>   * system RAM. This corresponds to IORESOURCE_SYSRAM_DRIVER_MANAGED in
>> the
>>   * kernel resource tree.
>> + * @MEMBLOCK_CRYPTO_CAPABLE: capable of hardware encryption
>>   */
>>  enum memblock_flags {
>>  	MEMBLOCK_NONE		= 0x0,	/* No special request */
>> @@ -47,6 +48,7 @@ enum memblock_flags {
>>  	MEMBLOCK_MIRROR		= 0x2,	/* mirrored region */
>>  	MEMBLOCK_NOMAP		= 0x4,	/* don't add to kernel direct mapping */
>>  	MEMBLOCK_DRIVER_MANAGED = 0x8,	/* always detected via a driver */
>> +	MEMBLOCK_CRYPTO_CAPABLE = 0x10,	/* capable of hardware encryption */
>>  };
>>
>>  /**
>> @@ -120,6 +122,9 @@ int memblock_physmem_add(phys_addr_t base, phys_addr_t
>> size);
>>  void memblock_trim_memory(phys_addr_t align);
>>  bool memblock_overlaps_region(struct memblock_type *type,
>>  			      phys_addr_t base, phys_addr_t size);
>> +bool memblock_node_is_crypto_capable(int nid);
>> +int memblock_mark_crypto_capable(phys_addr_t base, phys_addr_t size);
>> +int memblock_clear_crypto_capable(phys_addr_t base, phys_addr_t size);
>>  int memblock_mark_hotplug(phys_addr_t base, phys_addr_t size);
>>  int memblock_clear_hotplug(phys_addr_t base, phys_addr_t size);
>>  int memblock_mark_mirror(phys_addr_t base, phys_addr_t size);
>> diff --git a/mm/memblock.c b/mm/memblock.c
>> index e4f03a6e8e56..fe62f81572e6 100644
>> --- a/mm/memblock.c
>> +++ b/mm/memblock.c
>> @@ -191,6 +191,40 @@ bool __init_memblock memblock_overlaps_region(struct
>> memblock_type *type,
>>  	return i < type->cnt;
>>  }
>>
>> +/**
>> + * memblock_node_is_crypto_capable - get if whole node is capable
>> + * of encryption
>> + * @nid: number of node
>> + *
>> + * Iterate over all memory memblock_type and find if all regions under
>> + * node @nid are capable of hardware encryption.
>> + *
>> + * Return:
>> + * true if every region in memory memblock_type is capable of
>
> I'd s/in memory memblock_type/in @nid
>

Good, thanks.

>> + * encryption, false otherwise.
>> + */
>> +bool __init_memblock memblock_node_is_crypto_capable(int nid)
>> +{
>> +	struct memblock_region *region;
>> +	int crypto_capables = 0;
>> +	int not_crypto_capables = 0;
>> +
>> +	for_each_mem_region(region) {
>> +		if (memblock_get_region_node(region) == nid) {
>> +			if (region->flags & MEMBLOCK_CRYPTO_CAPABLE)
>> +				crypto_capables++;
>> +			else
>> +				not_crypto_capables++;
>> +		}
>> +	}
>> +
>> +	if (crypto_capables > 0 && not_crypto_capables > 0)
>> +		pr_warn("Node %d has %d regions that are encryptable and %d regions
>> that aren't",
>> +			nid, not_crypto_capables, crypto_capables);
>> +
>> +	return not_crypto_capables == 0;
>
> This will return true for memoryless nodes as well. Do you mean to consider
> them as capable of encryption?
>

Not really, I didn't think about that to be honest. I don't think it's
a good idea to consider them as capable, right?