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>,
Matthew Wilcox <willy@infradead.org>,
Ryan Roberts <ryan.roberts@arm.com>,
Catalin Marinas <catalin.marinas@arm.com>,
Yin Fengwei <fengwei.yin@intel.com>,
Michal Hocko <mhocko@suse.com>, Will Deacon <will@kernel.org>,
"Aneesh Kumar K.V" <aneesh.kumar@linux.ibm.com>,
Nick Piggin <npiggin@gmail.com>,
Peter Zijlstra <peterz@infradead.org>,
Michael Ellerman <mpe@ellerman.id.au>,
Christophe Leroy <christophe.leroy@csgroup.eu>,
"Naveen N. Rao" <naveen.n.rao@linux.ibm.com>,
Heiko Carstens <hca@linux.ibm.com>,
Vasily Gorbik <gor@linux.ibm.com>,
Alexander Gordeev <agordeev@linux.ibm.com>,
Christian Borntraeger <borntraeger@linux.ibm.com>,
Sven Schnelle <svens@linux.ibm.com>,
Arnd Bergmann <arnd@arndb.de>,
linux-arch@vger.kernel.org, linuxppc-dev@lists.ozlabs.org,
linux-s390@vger.kernel.org
Subject: [PATCH v3 10/10] mm/memory: optimize unmap/zap with PTE-mapped THP
Date: Wed, 14 Feb 2024 21:44:35 +0100 [thread overview]
Message-ID: <20240214204435.167852-11-david@redhat.com> (raw)
In-Reply-To: <20240214204435.167852-1-david@redhat.com>
Similar to how we optimized fork(), let's implement PTE batching when
consecutive (present) PTEs map consecutive pages of the same large
folio.
Most infrastructure we need for batching (mmu gather, rmap) is already
there. We only have to add get_and_clear_full_ptes() and
clear_full_ptes(). Similarly, extend zap_install_uffd_wp_if_needed() to
process a PTE range.
We won't bother sanity-checking the mapcount of all subpages, but only
check the mapcount of the first subpage we process. If there is a real
problem hiding somewhere, we can trigger it simply by using small
folios, or when we zap single pages of a large folio. Ideally, we had
that check in rmap code (including for delayed rmap), but then we cannot
print the PTE. Let's keep it simple for now. If we ever have a cheap
folio_mapcount(), we might just want to check for underflows there.
To keep small folios as fast as possible force inlining of a specialized
variant using __always_inline with nr=1.
Reviewed-by: Ryan Roberts <ryan.roberts@arm.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
---
include/linux/pgtable.h | 70 +++++++++++++++++++++++++++++++
mm/memory.c | 92 +++++++++++++++++++++++++++++------------
2 files changed, 136 insertions(+), 26 deletions(-)
diff --git a/include/linux/pgtable.h b/include/linux/pgtable.h
index aab227e12493..49ab1f73b5c2 100644
--- a/include/linux/pgtable.h
+++ b/include/linux/pgtable.h
@@ -580,6 +580,76 @@ static inline pte_t ptep_get_and_clear_full(struct mm_struct *mm,
}
#endif
+#ifndef get_and_clear_full_ptes
+/**
+ * get_and_clear_full_ptes - Clear present PTEs that map consecutive pages of
+ * the same folio, collecting dirty/accessed bits.
+ * @mm: Address space the pages are mapped into.
+ * @addr: Address the first page is mapped at.
+ * @ptep: Page table pointer for the first entry.
+ * @nr: Number of entries to clear.
+ * @full: Whether we are clearing a full mm.
+ *
+ * May be overridden by the architecture; otherwise, implemented as a simple
+ * loop over ptep_get_and_clear_full(), merging dirty/accessed bits into the
+ * returned PTE.
+ *
+ * Note that PTE bits in the PTE range besides the PFN can differ. For example,
+ * some PTEs might be write-protected.
+ *
+ * Context: The caller holds the page table lock. The PTEs map consecutive
+ * pages that belong to the same folio. The PTEs are all in the same PMD.
+ */
+static inline pte_t get_and_clear_full_ptes(struct mm_struct *mm,
+ unsigned long addr, pte_t *ptep, unsigned int nr, int full)
+{
+ pte_t pte, tmp_pte;
+
+ pte = ptep_get_and_clear_full(mm, addr, ptep, full);
+ while (--nr) {
+ ptep++;
+ addr += PAGE_SIZE;
+ tmp_pte = ptep_get_and_clear_full(mm, addr, ptep, full);
+ if (pte_dirty(tmp_pte))
+ pte = pte_mkdirty(pte);
+ if (pte_young(tmp_pte))
+ pte = pte_mkyoung(pte);
+ }
+ return pte;
+}
+#endif
+
+#ifndef clear_full_ptes
+/**
+ * clear_full_ptes - Clear present PTEs that map consecutive pages of the same
+ * folio.
+ * @mm: Address space the pages are mapped into.
+ * @addr: Address the first page is mapped at.
+ * @ptep: Page table pointer for the first entry.
+ * @nr: Number of entries to clear.
+ * @full: Whether we are clearing a full mm.
+ *
+ * May be overridden by the architecture; otherwise, implemented as a simple
+ * loop over ptep_get_and_clear_full().
+ *
+ * Note that PTE bits in the PTE range besides the PFN can differ. For example,
+ * some PTEs might be write-protected.
+ *
+ * Context: The caller holds the page table lock. The PTEs map consecutive
+ * pages that belong to the same folio. The PTEs are all in the same PMD.
+ */
+static inline void clear_full_ptes(struct mm_struct *mm, unsigned long addr,
+ pte_t *ptep, unsigned int nr, int full)
+{
+ for (;;) {
+ ptep_get_and_clear_full(mm, addr, ptep, full);
+ if (--nr == 0)
+ break;
+ ptep++;
+ addr += PAGE_SIZE;
+ }
+}
+#endif
/*
* If two threads concurrently fault at the same page, the thread that
diff --git a/mm/memory.c b/mm/memory.c
index a3efc4da258a..3b8e56eb08a3 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -1515,7 +1515,7 @@ static inline bool zap_drop_file_uffd_wp(struct zap_details *details)
*/
static inline void
zap_install_uffd_wp_if_needed(struct vm_area_struct *vma,
- unsigned long addr, pte_t *pte,
+ unsigned long addr, pte_t *pte, int nr,
struct zap_details *details, pte_t pteval)
{
/* Zap on anonymous always means dropping everything */
@@ -1525,20 +1525,27 @@ zap_install_uffd_wp_if_needed(struct vm_area_struct *vma,
if (zap_drop_file_uffd_wp(details))
return;
- pte_install_uffd_wp_if_needed(vma, addr, pte, pteval);
+ for (;;) {
+ /* the PFN in the PTE is irrelevant. */
+ pte_install_uffd_wp_if_needed(vma, addr, pte, pteval);
+ if (--nr == 0)
+ break;
+ pte++;
+ addr += PAGE_SIZE;
+ }
}
-static inline void zap_present_folio_pte(struct mmu_gather *tlb,
+static __always_inline void zap_present_folio_ptes(struct mmu_gather *tlb,
struct vm_area_struct *vma, struct folio *folio,
- struct page *page, pte_t *pte, pte_t ptent, unsigned long addr,
- struct zap_details *details, int *rss, bool *force_flush,
- bool *force_break)
+ struct page *page, pte_t *pte, pte_t ptent, unsigned int nr,
+ unsigned long addr, struct zap_details *details, int *rss,
+ bool *force_flush, bool *force_break)
{
struct mm_struct *mm = tlb->mm;
bool delay_rmap = false;
if (!folio_test_anon(folio)) {
- ptent = ptep_get_and_clear_full(mm, addr, pte, tlb->fullmm);
+ ptent = get_and_clear_full_ptes(mm, addr, pte, nr, tlb->fullmm);
if (pte_dirty(ptent)) {
folio_mark_dirty(folio);
if (tlb_delay_rmap(tlb)) {
@@ -1548,36 +1555,49 @@ static inline void zap_present_folio_pte(struct mmu_gather *tlb,
}
if (pte_young(ptent) && likely(vma_has_recency(vma)))
folio_mark_accessed(folio);
- rss[mm_counter(folio)]--;
+ rss[mm_counter(folio)] -= nr;
} else {
/* We don't need up-to-date accessed/dirty bits. */
- ptep_get_and_clear_full(mm, addr, pte, tlb->fullmm);
- rss[MM_ANONPAGES]--;
+ clear_full_ptes(mm, addr, pte, nr, tlb->fullmm);
+ rss[MM_ANONPAGES] -= nr;
}
+ /* Checking a single PTE in a batch is sufficient. */
arch_check_zapped_pte(vma, ptent);
- tlb_remove_tlb_entry(tlb, pte, addr);
+ tlb_remove_tlb_entries(tlb, pte, nr, addr);
if (unlikely(userfaultfd_pte_wp(vma, ptent)))
- zap_install_uffd_wp_if_needed(vma, addr, pte, details, ptent);
+ zap_install_uffd_wp_if_needed(vma, addr, pte, nr, details,
+ ptent);
if (!delay_rmap) {
- folio_remove_rmap_pte(folio, page, vma);
+ folio_remove_rmap_ptes(folio, page, nr, vma);
+
+ /* Only sanity-check the first page in a batch. */
if (unlikely(page_mapcount(page) < 0))
print_bad_pte(vma, addr, ptent, page);
}
- if (unlikely(__tlb_remove_page(tlb, page, delay_rmap))) {
+ if (unlikely(__tlb_remove_folio_pages(tlb, page, nr, delay_rmap))) {
*force_flush = true;
*force_break = true;
}
}
-static inline void zap_present_pte(struct mmu_gather *tlb,
+/*
+ * Zap or skip at least one present PTE, trying to batch-process subsequent
+ * PTEs that map consecutive pages of the same folio.
+ *
+ * Returns the number of processed (skipped or zapped) PTEs (at least 1).
+ */
+static inline int zap_present_ptes(struct mmu_gather *tlb,
struct vm_area_struct *vma, pte_t *pte, pte_t ptent,
- unsigned long addr, struct zap_details *details,
- int *rss, bool *force_flush, bool *force_break)
+ unsigned int max_nr, unsigned long addr,
+ struct zap_details *details, int *rss, bool *force_flush,
+ bool *force_break)
{
+ const fpb_t fpb_flags = FPB_IGNORE_DIRTY | FPB_IGNORE_SOFT_DIRTY;
struct mm_struct *mm = tlb->mm;
struct folio *folio;
struct page *page;
+ int nr;
page = vm_normal_page(vma, addr, ptent);
if (!page) {
@@ -1587,14 +1607,29 @@ static inline void zap_present_pte(struct mmu_gather *tlb,
tlb_remove_tlb_entry(tlb, pte, addr);
VM_WARN_ON_ONCE(userfaultfd_wp(vma));
ksm_might_unmap_zero_page(mm, ptent);
- return;
+ return 1;
}
folio = page_folio(page);
if (unlikely(!should_zap_folio(details, folio)))
- return;
- zap_present_folio_pte(tlb, vma, folio, page, pte, ptent, addr, details,
- rss, force_flush, force_break);
+ return 1;
+
+ /*
+ * Make sure that the common "small folio" case is as fast as possible
+ * by keeping the batching logic separate.
+ */
+ if (unlikely(folio_test_large(folio) && max_nr != 1)) {
+ nr = folio_pte_batch(folio, addr, pte, ptent, max_nr, fpb_flags,
+ NULL);
+
+ zap_present_folio_ptes(tlb, vma, folio, page, pte, ptent, nr,
+ addr, details, rss, force_flush,
+ force_break);
+ return nr;
+ }
+ zap_present_folio_ptes(tlb, vma, folio, page, pte, ptent, 1, addr,
+ details, rss, force_flush, force_break);
+ return 1;
}
static unsigned long zap_pte_range(struct mmu_gather *tlb,
@@ -1609,6 +1644,7 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
pte_t *start_pte;
pte_t *pte;
swp_entry_t entry;
+ int nr;
tlb_change_page_size(tlb, PAGE_SIZE);
init_rss_vec(rss);
@@ -1622,7 +1658,9 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
pte_t ptent = ptep_get(pte);
struct folio *folio;
struct page *page;
+ int max_nr;
+ nr = 1;
if (pte_none(ptent))
continue;
@@ -1630,10 +1668,12 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
break;
if (pte_present(ptent)) {
- zap_present_pte(tlb, vma, pte, ptent, addr, details,
- rss, &force_flush, &force_break);
+ max_nr = (end - addr) / PAGE_SIZE;
+ nr = zap_present_ptes(tlb, vma, pte, ptent, max_nr,
+ addr, details, rss, &force_flush,
+ &force_break);
if (unlikely(force_break)) {
- addr += PAGE_SIZE;
+ addr += nr * PAGE_SIZE;
break;
}
continue;
@@ -1687,8 +1727,8 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
WARN_ON_ONCE(1);
}
pte_clear_not_present_full(mm, addr, pte, tlb->fullmm);
- zap_install_uffd_wp_if_needed(vma, addr, pte, details, ptent);
- } while (pte++, addr += PAGE_SIZE, addr != end);
+ zap_install_uffd_wp_if_needed(vma, addr, pte, 1, details, ptent);
+ } while (pte += nr, addr += PAGE_SIZE * nr, addr != end);
add_mm_rss_vec(mm, rss);
arch_leave_lazy_mmu_mode();
--
2.43.0
prev parent reply other threads:[~2024-02-14 20:45 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-02-14 20:44 [PATCH v3 00/10] mm/memory: optimize unmap/zap with PTE-mapped THP David Hildenbrand
2024-02-14 20:44 ` [PATCH v3 01/10] mm/memory: factor out zapping of present pte into zap_present_pte() David Hildenbrand
2024-02-14 20:44 ` [PATCH v3 02/10] mm/memory: handle !page case in zap_present_pte() separately David Hildenbrand
2024-02-14 20:44 ` [PATCH v3 03/10] mm/memory: further separate anon and pagecache folio handling in zap_present_pte() David Hildenbrand
2024-02-14 20:44 ` [PATCH v3 04/10] mm/memory: factor out zapping folio pte into zap_present_folio_pte() David Hildenbrand
2024-02-14 20:44 ` [PATCH v3 05/10] mm/mmu_gather: pass "delay_rmap" instead of encoded page to __tlb_remove_page_size() David Hildenbrand
2024-02-14 20:44 ` [PATCH v3 06/10] mm/mmu_gather: define ENCODED_PAGE_FLAG_DELAY_RMAP David Hildenbrand
2024-02-14 20:44 ` [PATCH v3 07/10] mm/mmu_gather: add tlb_remove_tlb_entries() David Hildenbrand
2024-02-14 20:44 ` [PATCH v3 08/10] mm/mmu_gather: add __tlb_remove_folio_pages() David Hildenbrand
2024-02-14 20:44 ` [PATCH v3 09/10] mm/mmu_gather: improve cond_resched() handling with large folios and expensive page freeing David Hildenbrand
2024-02-14 20:44 ` David Hildenbrand [this message]
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=20240214204435.167852-11-david@redhat.com \
--to=david@redhat.com \
--cc=agordeev@linux.ibm.com \
--cc=akpm@linux-foundation.org \
--cc=aneesh.kumar@linux.ibm.com \
--cc=arnd@arndb.de \
--cc=borntraeger@linux.ibm.com \
--cc=catalin.marinas@arm.com \
--cc=christophe.leroy@csgroup.eu \
--cc=fengwei.yin@intel.com \
--cc=gor@linux.ibm.com \
--cc=hca@linux.ibm.com \
--cc=linux-arch@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux-s390@vger.kernel.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=mhocko@suse.com \
--cc=mpe@ellerman.id.au \
--cc=naveen.n.rao@linux.ibm.com \
--cc=npiggin@gmail.com \
--cc=peterz@infradead.org \
--cc=ryan.roberts@arm.com \
--cc=svens@linux.ibm.com \
--cc=will@kernel.org \
--cc=willy@infradead.org \
/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 read-only IMAP folder(s) and NNTP newsgroup(s).