Heterogeneous Memory Management Work Is Being Renewed For The Linux Kernel

Back in 2014 is when longtime open-source graphics driver developer Jerome Glisse began pushing his patches for Heterogeneous Memory Management in the Linux kernel while that work is still ongoing but has now been renewed.

We last heard an update on Heterogeneous Memory Management (HMM) last summer when Glisse sent out his latest patches. This big patch-set is a helper layer for a device wanting to mirror a process address space into their own MMU. HMM is designed for GPUs and others in needing to support OpenCL 2.0+ for mirroring a process address space. HMM also makes it possible for using the discrete GPU memory in a transparent manner to the application/game and other possible use-cases.

Jerome wrote on Tuesday to the kernel mailing list:

Last time I spoke with Linus and Andrew, the requirement for getting HMM upstream was having real hardware working with it beside Mellanox (as Mellanox does not use all HMM features), both with closed source driver and open source driver. Work on open source driver is underway, and I anticipate we will get update from NVIDIA and other parties on their efforts and plans shortly.

I am re-posting now because I want people to have time to look at HMM again. The open source driver will stay behind close doors until hardware is released. I can however have the upstream maintainer share his progress here if anyone feels the need for that.

Other parties such as IBM and Mediatek are also interested in HMM. I expect they will comment on their respective hardware when they can.

I hope that HMM can be considered for inclusion upstream soon.

Find out more about Linux HMM via the kernel mailing list with the code now up to its 12th revision.