Skip to content

memory tiering: calculate abstract distance based on ACPI HMAT

Huang Ying linux-mm lkml 1 version In review

Get series with b4
b4 am [email protected]

Versions newest first

  1. v1 Latest4 patches 9 repliesIn review

Changes in v1 from the cover letter

- Added some comments per Aneesh's comments, Thanks!

Patches in v1 No applies recorded yet

#Subject and filesSizeRepliesApplied
1/4memory tiering: add abstract distance calculation algorithms managementinclude/linux/memory-tiers.h mm/memory-tiers.c+78 -02 replies
2/4acpi, hmat: refactor hmat_register_target_initiators()drivers/acpi/numa/hmat.c+30 -511 reply
3/4acpi, hmat: calculate abstract distance with HMATdrivers/acpi/numa/hmat.c include/linux/memory-tiers.h mm/memory-tiers.c+181 -23 replies
4/4dax, kmem: calculate abstract distance with general interfacedrivers/dax/kmem.c include/linux/memory-tiers.h mm/memory-tiers.c+52 -143 replies

Reviews trailers per patch

All 4 patches have a Reviewed-by or Acked-by.

PatchBharata B RaoAlistair PoppleDave Jiang
1/4T-b R-b R-b
2/4T-b R-b T-b R-b
3/4T-b R-b
4/4T-b R-b

Trailers given in replies are filled; those carried in the patches from earlier versions are outlined with dashes. Scroll the table sideways for all 3 reviewers.

Read the v1 thread

Cover letter [PATCH -V3 0/4] memory tiering: calculate abstract distance based on ACPI HMAT

We have the explicit memory tiers framework to manage systems with multiple types of memory, e.g., DRAM in DIMM slots and CXL memory devices. Where, same kind of memory devices will be grouped into memory types, then put into memory tiers. To describe the performance of a memory type, abstract distance is defined. Which is in direct proportion to the memory latency and inversely proportional to the memory bandwidth. To keep the code as simple as possible, fixed abstract distance is used in dax/kmem to describe slow memory such as Optane DCPMM.

To support more memory types, in this series, we added the abstract distance calculation algorithm management mechanism, provided a algorithm implementation based on ACPI HMAT, and used the general abstract distance calculation interface in dax/kmem driver. So, dax/kmem can support HBM (high bandwidth memory) in addition to the original Optane DCPMM.

Changelog:

v3:

- Move algorithm to calculate abstract distance from read/write latency/bandwidth from hmat.c to memory-ters.c per Alistair's comments.

- Fix memory types putting in kmem.c for error path.

V2:

- Fix a typo in 4/4.

- Collected reviewed-by and tested-by.

V1 (from RFC):

- Added some comments per Aneesh's comments, Thanks!

Best Regards,
Huang, Ying