diff options
| author | Zi Yan <ziy@nvidia.com> | 2024-07-24 15:01:14 +0200 |
|---|---|---|
| committer | Andrew Morton <akpm@linux-foundation.org> | 2024-09-02 05:25:47 +0200 |
| commit | 2a28713a67fd28eedc13b32300df6b9c5a2381f5 (patch) | |
| tree | b99b41c1317c28193490fe09ecbb5863edd34140 /mm/huge_memory.c | |
| parent | memory tiering: read last_cpupid correctly in do_huge_pmd_numa_page() (diff) | |
| download | linux-2a28713a67fd28eedc13b32300df6b9c5a2381f5.tar.xz linux-2a28713a67fd28eedc13b32300df6b9c5a2381f5.zip | |
memory tiering: introduce folio_use_access_time() check
If memory tiering mode is on and a folio is not in the top tier memory,
folio's cpupid field is repurposed to store page access time. Instead of
an open coded check, use a function to encapsulate the check.
Link: https://lkml.kernel.org/r/20240724130115.793641-3-ziy@nvidia.com
Signed-off-by: Zi Yan <ziy@nvidia.com>
Reviewed-by: "Huang, Ying" <ying.huang@intel.com>
Acked-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Cc: Baolin Wang <baolin.wang@linux.alibaba.com>
Cc: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Diffstat (limited to 'mm/huge_memory.c')
| -rw-r--r-- | mm/huge_memory.c | 6 |
1 files changed, 2 insertions, 4 deletions
diff --git a/mm/huge_memory.c b/mm/huge_memory.c index 632a92edec4e..29e76d285e4b 100644 --- a/mm/huge_memory.c +++ b/mm/huge_memory.c @@ -1707,8 +1707,7 @@ vm_fault_t do_huge_pmd_numa_page(struct vm_fault *vmf) * For memory tiering mode, cpupid of slow memory page is used * to record page access time. So use default value. */ - if (!(sysctl_numa_balancing_mode & NUMA_BALANCING_MEMORY_TIERING) || - node_is_toptier(nid)) + if (!folio_use_access_time(folio)) last_cpupid = folio_last_cpupid(folio); target_nid = numa_migrate_prep(folio, vmf, haddr, nid, &flags); if (target_nid == NUMA_NO_NODE) @@ -2058,8 +2057,7 @@ int change_huge_pmd(struct mmu_gather *tlb, struct vm_area_struct *vma, toptier) goto unlock; - if (sysctl_numa_balancing_mode & NUMA_BALANCING_MEMORY_TIERING && - !toptier) + if (folio_use_access_time(folio)) folio_xchg_access_time(folio, jiffies_to_msecs(jiffies)); } |