From: David Hildenbrand <david@redhat.com>
To: linux-kernel@vger.kernel.org
Cc: linux-mm@kvack.org, David Hildenbrand <david@redhat.com>,
Andrew Morton <akpm@linux-foundation.org>,
Linus Torvalds <torvalds@linux-foundation.org>,
Ryan Roberts <ryan.roberts@arm.com>,
Matthew Wilcox <willy@infradead.org>,
Hugh Dickins <hughd@google.com>,
Yin Fengwei <fengwei.yin@intel.com>,
Yang Shi <shy828301@gmail.com>, Ying Huang <ying.huang@intel.com>,
Zi Yan <ziy@nvidia.com>, Peter Zijlstra <peterz@infradead.org>,
Ingo Molnar <mingo@redhat.com>, Will Deacon <will@kernel.org>,
Waiman Long <longman@redhat.com>,
"Paul E. McKenney" <paulmck@kernel.org>
Subject: [PATCH WIP v1 13/20] mm/huge_memory: batch rmap operations in __split_huge_pmd_locked()
Date: Fri, 24 Nov 2023 14:26:18 +0100 [thread overview]
Message-ID: <20231124132626.235350-14-david@redhat.com> (raw)
In-Reply-To: <20231124132626.235350-1-david@redhat.com>
Let's batch the rmap operations, as a preparation to making individual
page_add_anon_rmap() calls more expensive.
While at it, use more folio operations (but only in the code branch we're
touching), use VM_WARN_ON_FOLIO(), and pass RMAP_COMPOUND instead of
manually setting PageAnonExclusive.
We should never see non-anon pages on that branch: otherwise, the
existing page_add_anon_rmap() call would have been flawed already.
Signed-off-by: David Hildenbrand <david@redhat.com>
---
mm/huge_memory.c | 23 +++++++++++++++--------
1 file changed, 15 insertions(+), 8 deletions(-)
diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index fd7251923557..f47971d1afbf 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -2100,6 +2100,7 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd,
unsigned long haddr, bool freeze)
{
struct mm_struct *mm = vma->vm_mm;
+ struct folio *folio;
struct page *page;
pgtable_t pgtable;
pmd_t old_pmd, _pmd;
@@ -2195,16 +2196,18 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd,
uffd_wp = pmd_swp_uffd_wp(old_pmd);
} else {
page = pmd_page(old_pmd);
+ folio = page_folio(page);
if (pmd_dirty(old_pmd)) {
dirty = true;
- SetPageDirty(page);
+ folio_set_dirty(folio);
}
write = pmd_write(old_pmd);
young = pmd_young(old_pmd);
soft_dirty = pmd_soft_dirty(old_pmd);
uffd_wp = pmd_uffd_wp(old_pmd);
- VM_BUG_ON_PAGE(!page_count(page), page);
+ VM_WARN_ON_FOLIO(!folio_ref_count(folio), folio);
+ VM_WARN_ON_FOLIO(!folio_test_anon(folio), folio);
/*
* Without "freeze", we'll simply split the PMD, propagating the
@@ -2221,11 +2224,18 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd,
*
* See page_try_share_anon_rmap(): invalidate PMD first.
*/
- anon_exclusive = PageAnon(page) && PageAnonExclusive(page);
+ anon_exclusive = PageAnonExclusive(page);
if (freeze && anon_exclusive && page_try_share_anon_rmap(page))
freeze = false;
- if (!freeze)
- page_ref_add(page, HPAGE_PMD_NR - 1);
+ if (!freeze) {
+ rmap_t rmap_flags = RMAP_NONE;
+
+ folio_ref_add(folio, HPAGE_PMD_NR - 1);
+ if (anon_exclusive)
+ rmap_flags = RMAP_EXCLUSIVE;
+ folio_add_anon_rmap_range(folio, page, HPAGE_PMD_NR,
+ vma, haddr, rmap_flags);
+ }
}
/*
@@ -2268,8 +2278,6 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd,
entry = mk_pte(page + i, READ_ONCE(vma->vm_page_prot));
if (write)
entry = pte_mkwrite(entry, vma);
- if (anon_exclusive)
- SetPageAnonExclusive(page + i);
if (!young)
entry = pte_mkold(entry);
/* NOTE: this may set soft-dirty too on some archs */
@@ -2279,7 +2287,6 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd,
entry = pte_mksoft_dirty(entry);
if (uffd_wp)
entry = pte_mkuffd_wp(entry);
- page_add_anon_rmap(page + i, vma, addr, RMAP_NONE);
}
VM_BUG_ON(!pte_none(ptep_get(pte)));
set_pte_at(mm, addr, pte, entry);
--
2.41.0
next prev parent reply other threads:[~2023-11-24 13:27 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-11-24 13:26 [PATCH WIP v1 00/20] mm: precise "mapped shared" vs. "mapped exclusively" detection for PTE-mapped THP / partially-mappable folios David Hildenbrand
2023-11-24 13:26 ` [PATCH WIP v1 01/20] mm/rmap: factor out adding folio range into __folio_add_rmap_range() David Hildenbrand
2023-11-24 13:26 ` [PATCH WIP v1 02/20] mm: add a total mapcount for large folios David Hildenbrand
2023-11-24 13:26 ` [PATCH WIP v1 03/20] mm: convert folio_estimated_sharers() to folio_mapped_shared() and improve it David Hildenbrand
2023-11-24 13:26 ` [PATCH WIP v1 04/20] mm/rmap: pass dst_vma to page_try_dup_anon_rmap() and page_dup_file_rmap() David Hildenbrand
2023-11-24 13:26 ` [PATCH WIP v1 05/20] mm/rmap: abstract total mapcount operations for partially-mappable folios David Hildenbrand
2023-11-24 13:26 ` [PATCH WIP v1 06/20] atomic_seqcount: new (raw) seqcount variant to support concurrent writers David Hildenbrand
2023-11-24 13:26 ` [PATCH WIP v1 07/20] mm/rmap_id: track if one ore multiple MMs map a partially-mappable folio David Hildenbrand
2023-12-17 19:13 ` Nadav Amit
2023-12-18 14:04 ` David Hildenbrand
2023-12-18 14:34 ` Nadav Amit
2023-11-24 13:26 ` [PATCH WIP v1 08/20] mm: pass MM to folio_mapped_shared() David Hildenbrand
2023-11-24 13:26 ` [PATCH WIP v1 09/20] mm: improve folio_mapped_shared() for partially-mappable folios using rmap IDs David Hildenbrand
2023-11-24 13:26 ` [PATCH WIP v1 10/20] mm/memory: COW reuse support for PTE-mapped THP with " David Hildenbrand
2023-11-24 13:26 ` [PATCH WIP v1 11/20] mm/rmap_id: support for 1, 2 and 3 values by manual calculation David Hildenbrand
2023-11-24 13:26 ` [PATCH WIP v1 12/20] mm/rmap: introduce folio_add_anon_rmap_range() David Hildenbrand
2023-11-24 13:26 ` David Hildenbrand [this message]
2023-11-24 13:26 ` [PATCH WIP v1 14/20] mm/huge_memory: avoid folio_refcount() < folio_mapcount() in __split_huge_pmd_locked() David Hildenbrand
2023-11-24 13:26 ` [PATCH WIP v1 15/20] mm/rmap_id: verify precalculated subids with CONFIG_DEBUG_VM David Hildenbrand
2023-11-24 13:26 ` [PATCH WIP v1 16/20] atomic_seqcount: support a single exclusive writer in the absence of other writers David Hildenbrand
2023-11-24 13:26 ` [PATCH WIP v1 17/20] mm/rmap_id: reduce atomic RMW operations when we are the exclusive writer David Hildenbrand
2023-11-24 13:26 ` [PATCH WIP v1 18/20] atomic_seqcount: use atomic add-return instead of atomic cmpxchg on 64bit David Hildenbrand
2023-11-24 13:26 ` [PATCH WIP v1 19/20] mm/rmap: factor out removing folio range into __folio_remove_rmap_range() David Hildenbrand
2023-11-24 13:26 ` [PATCH WIP v1 20/20] mm/rmap: perform all mapcount operations of large folios under the rmap seqcount David Hildenbrand
2023-11-24 20:55 ` [PATCH WIP v1 00/20] mm: precise "mapped shared" vs. "mapped exclusively" detection for PTE-mapped THP / partially-mappable folios Linus Torvalds
2023-11-25 17:02 ` David Hildenbrand
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20231124132626.235350-14-david@redhat.com \
--to=david@redhat.com \
--cc=akpm@linux-foundation.org \
--cc=fengwei.yin@intel.com \
--cc=hughd@google.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=longman@redhat.com \
--cc=mingo@redhat.com \
--cc=paulmck@kernel.org \
--cc=peterz@infradead.org \
--cc=ryan.roberts@arm.com \
--cc=shy828301@gmail.com \
--cc=torvalds@linux-foundation.org \
--cc=will@kernel.org \
--cc=willy@infradead.org \
--cc=ying.huang@intel.com \
--cc=ziy@nvidia.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).