| Message ID | 20220121081345.80320-1-songmuchun@bytedance.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | [1/2] mm: thp: fix wrong cache flush in remove_migration_pmd() | expand |
On 21 Jan 2022, at 3:13, Muchun Song wrote: > The flush_cache_page() is supposed to be justified only if the page > is already placed in process page table, and that is done right after > flush_cache_page(). So using this interface is wrong. And there is > no need to invalite cache since it was non-present before in > remove_migration_pmd(). So just to remove it. > > Signed-off-by: Muchun Song <songmuchun@bytedance.com> > --- > mm/huge_memory.c | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/mm/huge_memory.c b/mm/huge_memory.c > index f58524394dc1..45ede45b11f5 100644 > --- a/mm/huge_memory.c > +++ b/mm/huge_memory.c > @@ -3207,7 +3207,6 @@ void remove_migration_pmd(struct page_vma_mapped_walk *pvmw, struct page *new) > if (pmd_swp_uffd_wp(*pvmw->pmd)) > pmde = pmd_wrprotect(pmd_mkuffd_wp(pmde)); > > - flush_cache_range(vma, mmun_start, mmun_start + HPAGE_PMD_SIZE); > if (PageAnon(new)) > page_add_anon_rmap(new, vma, mmun_start, true); > else > @@ -3215,6 +3214,8 @@ void remove_migration_pmd(struct page_vma_mapped_walk *pvmw, struct page *new) > set_pmd_at(mm, mmun_start, pvmw->pmd, pmde); > if ((vma->vm_flags & VM_LOCKED) && !PageDoubleMap(new)) > mlock_vma_page(new); > + > + /* No need to invalidate - it was non-present before */ > update_mmu_cache_pmd(vma, address, pvmw->pmd); > } > #endif LGTM. Thanks. Reviewed-by: Zi Yan <ziy@nvidia.com> -- Best Regards, Yan, Zi
diff --git a/mm/huge_memory.c b/mm/huge_memory.c index f58524394dc1..45ede45b11f5 100644 --- a/mm/huge_memory.c +++ b/mm/huge_memory.c @@ -3207,7 +3207,6 @@ void remove_migration_pmd(struct page_vma_mapped_walk *pvmw, struct page *new) if (pmd_swp_uffd_wp(*pvmw->pmd)) pmde = pmd_wrprotect(pmd_mkuffd_wp(pmde)); - flush_cache_range(vma, mmun_start, mmun_start + HPAGE_PMD_SIZE); if (PageAnon(new)) page_add_anon_rmap(new, vma, mmun_start, true); else @@ -3215,6 +3214,8 @@ void remove_migration_pmd(struct page_vma_mapped_walk *pvmw, struct page *new) set_pmd_at(mm, mmun_start, pvmw->pmd, pmde); if ((vma->vm_flags & VM_LOCKED) && !PageDoubleMap(new)) mlock_vma_page(new); + + /* No need to invalidate - it was non-present before */ update_mmu_cache_pmd(vma, address, pvmw->pmd); } #endif
The flush_cache_page() is supposed to be justified only if the page is already placed in process page table, and that is done right after flush_cache_page(). So using this interface is wrong. And there is no need to invalite cache since it was non-present before in remove_migration_pmd(). So just to remove it. Signed-off-by: Muchun Song <songmuchun@bytedance.com> --- mm/huge_memory.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)