Re: [PATCH 6/6] arm64/mm: Enable ZONE_DEVICE

[Anshuman Khandual <anshuman.khandual@arm.com>]

On 04/03/2019 09:37 PM, Jerome Glisse wrote:
> On Wed, Apr 03, 2019 at 02:58:28PM +0100, Robin Murphy wrote:
>> [ +Dan, Jerome ]
>>
>> On 03/04/2019 05:30, Anshuman Khandual wrote:
>>> Arch implementation for functions which create or destroy vmemmap mapping
>>> (vmemmap_populate, vmemmap_free) can comprehend and allocate from inside
>>> device memory range through driver provided vmem_altmap structure which
>>> fulfils all requirements to enable ZONE_DEVICE on the platform. Hence just
>>
>> ZONE_DEVICE is about more than just altmap support, no?
>>
>>> enable ZONE_DEVICE by subscribing to ARCH_HAS_ZONE_DEVICE. But this is only
>>> applicable for ARM64_4K_PAGES (ARM64_SWAPPER_USES_SECTION_MAPS) only which
>>> creates vmemmap section mappings and utilize vmem_altmap structure.
>>
>> What prevents it from working with other page sizes? One of the foremost
>> use-cases for our 52-bit VA/PA support is to enable mapping large quantities
>> of persistent memory, so we really do need this for 64K pages too. FWIW, it
>> appears not to be an issue for PowerPC.
>>
>>> Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
>>> ---
>>>   arch/arm64/Kconfig | 1 +
>>>   1 file changed, 1 insertion(+)
>>>
>>> diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig
>>> index db3e625..b5d8cf5 100644
>>> --- a/arch/arm64/Kconfig
>>> +++ b/arch/arm64/Kconfig
>>> @@ -31,6 +31,7 @@ config ARM64
>>>   	select ARCH_HAS_SYSCALL_WRAPPER
>>>   	select ARCH_HAS_TEARDOWN_DMA_OPS if IOMMU_SUPPORT
>>>   	select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
>>> +	select ARCH_HAS_ZONE_DEVICE if ARM64_4K_PAGES
>>
>> IIRC certain configurations (HMM?) don't even build if you just turn this on
>> alone (although of course things may have changed elsewhere in the meantime)
>> - crucially, though, from previous discussions[1] it seems fundamentally
>> unsafe, since I don't think we can guarantee that nobody will touch the
>> corners of ZONE_DEVICE that also require pte_devmap in order not to go
>> subtly wrong. I did get as far as cooking up some patches to sort that out
>> [2][3] which I never got round to posting for their own sake, so please
>> consider picking those up as part of this series.
> 
> Correct _do not_ enable ZONE_DEVICE without support for pte_devmap detection.

Driver managed ZONE_DEVICE memory which never maps into user page table is not
a valid use case for ZONE_DEVICE ? Also what about MEMORY_DEVICE_PRIVATE ? That
can never be mapped into user page table. A driver can still manage these non
coherent memory through it's struct pages (which will be allocated inside RAM)

> If you want some feature of ZONE_DEVICE. Like HMM as while DAX does require
> pte_devmap, HMM device private does not. So you would first have to split
> ZONE_DEVICE into more sub-features kconfig option.

CONFIG_ZONE_DEVICE does not do that already ! All it says is that a device
memory range can be plugged into ZONE_DEVICE either as PRIVATE (non-coherent)
or PUBLIC/PCI_P2PDMA (coherent) memory without mandating anything about how
these memory will be further used.

> 
> What is the end use case you are looking for ? Persistent memory ?

Persistent memory is one of the primary use cases.