Commit f485702
committed
x86: fix cache ids on Intel
Use the official way to compute the cache id on Intel
(same one as in the Linux kernel) by rounding-up
nbthreads_sharing and masking lower bits of the APIC ids
(instead of dividing the APIC id by nbthreads_sharing).
This should not affect any existing platform topology,
but it's more future-proof.
Thanks to Jonathan Peyton for the help.
Signed-off-by: Brice Goglin <Brice.Goglin@inria.fr>1 parent fe18673 commit f485702
1 file changed
+8
-2
lines changed| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
1 | 1 | | |
2 | | - | |
| 2 | + | |
3 | 3 | | |
4 | 4 | | |
5 | 5 | | |
| |||
751 | 751 | | |
752 | 752 | | |
753 | 753 | | |
754 | | - | |
| 754 | + | |
| 755 | + | |
| 756 | + | |
| 757 | + | |
| 758 | + | |
| 759 | + | |
| 760 | + | |
755 | 761 | | |
756 | 762 | | |
757 | 763 | | |
| |||
0 commit comments