On Thu, Jul 06, 2017 at 10:30:46PM -0700, John Hubbard wrote:
On 07/06/2017 02:52 PM, Ross Zwisler wrote:
> The naming collision between Jerome's "Heterogeneous Memory Management
> (HMM)" and this "Heterogeneous Memory (HMEM)" series is unfortunate,
> was trying to stick with the word "Heterogeneous" because of the naming
> the ACPI 6.2 Heterogeneous Memory Attribute Table table. Suggestions for
> better naming are welcome.
Say, most of the places (file names, function and variable names, and even
print statements) where this patchset uses hmem or HMEM, it really seems to
mean, the Heterogeneous Memory Attribute Table. That's not *always* true, but
given that it's a pretty severe naming conflict, how about just changing:
hmem --> hmat
HMEM --> HMAT
...everywhere? Then you still have Heterogeneous Memory in the name, but
there is enough lexical distance (is that a thing? haha) between HMM and HMAT
to keep us all sane. :)
With or without the above suggestion, there are a few places (Kconfig, comments,
prints) where we can more easily make it clear that HMM != HMEM (or HMAT),
so for those I can just comment on them separately in the individual patches.
Sure, that change makes sense to me. I had initially tried to make this
enabling more generic so that other, non-ACPI systems could use the same sysfs
representation even if they got their performance numbers from some other
source, but while implementing it pretty quickly became very tightly tied to
the ACPI HMAT.