-
Notifications
You must be signed in to change notification settings - Fork 25
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
balloons: support for L2 cache cluster allocation #384
Conversation
This PR depends on #383 |
745e5b1
to
d4c7412
Compare
10357d5
to
cd8e43a
Compare
When two caches of the same level and kind share the same id, they collide in saved caches. This could lead to assigning cpu.caches caches from different package than the CPU itself. This change fixes the issue by ensuring unique cache ids to all caches of the same level and kind. Signed-off-by: Antti Kervinen <[email protected]>
Cache level functions, for instance CPU.GetCachesByLevel(), expect cpu.caches to be sorted by level, the lowest first. This change ensures that overridden caches follow this ordering, too, no matter in which order user has spesified them. Signed-off-by: Antti Kervinen <[email protected]>
This implementation expects CPUs sharing the same L2 cache also belonging into the same NUMA node, and that hyperthreads of the same CPU core always share L2 cache. Signed-off-by: Antti Kervinen <[email protected]>
Signed-off-by: Antti Kervinen <[email protected]>
Signed-off-by: Antti Kervinen <[email protected]>
cd8e43a
to
f9b7a24
Compare
Signed-off-by: Antti Kervinen <[email protected]>
(Documentation added as the last commit.) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM 👍
{"cpusets": ["0-7", "8-15"], "level": 3, "size": "128M"}, | ||
|
||
{"cpusets": ["0-1", "2-3", "4-5", "6-7"], "level": 2, "size": "4M", "kind": "u"}, | ||
{"cpusets": ["8-9", "10-11", "12-13", "14-15"], "level": 2, "size": "4M", "kind": "u"}, | ||
|
||
{"cpusets": ["0-1", "2-3", "4-5", "6-7"], "level": 1, "size": "256k", "kind": "d"}, | ||
{"cpusets": ["0-1", "2-3", "4-5", "6-7"], "level": 1, "size": "128k", "kind": "i"}, | ||
|
||
{"cpusets": ["8-9", "10-11", "12-13", "14-15"], "level": 1, "size": "256k", "kind": "d"}, | ||
{"cpusets": ["8-9", "10-11", "12-13", "14-15"], "level": 1, "size": "128k", "kind": "i"} | ||
]''' |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Is this cache configuration override interesting/useful ? Doesn't it provide the same grouping as the socket and the physical cores ?
No description provided.