Re: HMM (Heterogeneous Memory Management) v10

[Pavel Machek <pavel@ucw.cz>]

Hi!

> Minor fixes since last post (1), apply on top of 4.2-rc6 done
> that because conflict in infiniband are harder to solve then

is.

> conflict with mm tree.
> 
> Tree with the patchset:
> git://people.freedesktop.org/~glisse/linux hmm-v10 branch
> 
> Previous cover letter :
> 
> 
> 
> HMM (Heterogeneous Memory Management) is an helper layer for device
> that want to mirror a process address space into their own mmu. Main
> target is GPU but other hardware, like network device can take also

GPU, ... devices

> use HMM.
> 
> There is two side to HMM, first one is mirroring of process address

There are

> space on behalf of a device. HMM will manage a secondary page table
> for the device and keep it synchronize with the CPU page table. HMM

synchronized

> also do DMA mapping on behalf of the device (which would allow new

also does

> kind of optimization further down the road (2)).

kinds of optimalizations

> Second side is allowing to migrate process memory to device memory

,

> where device memory is unmappable by the CPU. Any CPU access will
> trigger special fault that will migrate memory back. This patchset
> does not deal with remote memory migration.
> 
> 
> Why doing this ?
> 
> Mirroring a process address space is mandatory with OpenCL 2.0 and
> with other GPU compute API. OpenCL 2.0 allow different level of

APIs... allows

> implementation and currently only the lowest 2 are supported on
> Linux. To implement the highest level, where CPU and GPU access
> can happen concurently and are cache coherent, HMM is needed, or
> something providing same functionality, for instance through
> platform hardware.
> 
> Hardware solution such as PCIE ATS/PASID is limited to mirroring
> system memory and does not provide way to migrate memory to device
> memory (which offer significantly more bandwidth up to 10 times

, up to

> faster than regular system memory with discret GPU, also have

discrete? 

> lower latency than PCIE transaction).

> Current CPU with GPU on same die (AMD or Intel) use the ATS/PASID
> and for Intel a special level of cache (backed by a large pool of
> fast memory).
> 
> For foreeseeable futur, discrete GPU will remain releveant as they

future .. GPUs

> can have a large quantity of faster memory than integrated GPU.
> 
> Thus we believe HMM will allow to leverage discret GPU memory in

allow us... discrete

> a transparent fashion to the application, with minimum disruption
> to the linux kernel mm code. Also HMM can work along hardware
> solution such as PCIE ATS/PASID (leaving regular case to ATS/PASID
> while HMM handles the migrated memory case).

Best regards,
									Pavel

-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>