The patch titled
Subject: mm: pass meminit_context to __free_pages_core()
has been added to the -mm mm-unstable branch. Its filename is
mm-pass-meminit_context-to-__free_pages_core.patch
This patch will shortly appear at
https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patches/mm-pass-meminit_context-to-__free_pages_core.patch
This patch will later appear in the mm-unstable branch at
git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next via the mm-everything
branch at git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
and is updated there every 2-3 working days
------------------------------------------------------
From: David Hildenbrand <david@xxxxxxxxxx>
Subject: mm: pass meminit_context to __free_pages_core()
Date: Fri, 7 Jun 2024 11:09:36 +0200
Patch series "mm/memory_hotplug: use PageOffline() instead of
PageReserved() for !ZONE_DEVICE".
This can be a considered a long-overdue follow-up to some parts of [1].
The patches are based on [2], but they are not strictly required -- just
makes it clearer why we can use adjust_managed_page_count() for memory
hotplug without going into details about highmem.
We stop initializing pages with PageReserved() in memory hotplug code --
except when dealing with ZONE_DEVICE for now. Instead, we use
PageOffline(): all pages are initialized to PageOffline() when onlining a
memory section, and only the ones actually getting exposed to the
system/page allocator will get PageOffline cleared.
This way, we enlighten memory hotplug more about PageOffline() pages and
can cleanup some hacks we have in virtio-mem code.
What about ZONE_DEVICE? PageOffline() is wrong, but we might just stop
using PageReserved() for them later by simply checking for
is_zone_device_page() at suitable places. That will be a separate patch
set / proposal.
This primarily affects virtio-mem, HV-balloon and XEN balloon. I only
briefly tested with virtio-mem, which benefits most from these cleanups.
[1] https://lore.kernel.org/all/20191024120938.11237-1-david@xxxxxxxxxx/
[2] https://lkml.kernel.org/r/20240607083711.62833-1-david@xxxxxxxxxx
This patch (of 3):
In preparation for further changes, let's teach __free_pages_core() about
the differences of memory hotplug handling.
Move the memory hotplug specific handling from generic_online_page() to
__free_pages_core(), use adjust_managed_page_count() on the memory hotplug
path, and spell out why memory freed via memblock cannot currently use
adjust_managed_page_count().
Link: https://lkml.kernel.org/r/20240607090939.89524-1-david@xxxxxxxxxx
Link: https://lkml.kernel.org/r/20240607090939.89524-2-david@xxxxxxxxxx
Signed-off-by: David Hildenbrand <david@xxxxxxxxxx>
Cc: Alexander Potapenko <glider@xxxxxxxxxx>
Cc: Dexuan Cui <decui@xxxxxxxxxxxxx>
Cc: Dmitry Vyukov <dvyukov@xxxxxxxxxx>
Cc: Eugenio Pérez <eperezma@xxxxxxxxxx>
Cc: Haiyang Zhang <haiyangz@xxxxxxxxxxxxx>
Cc: Jason Wang <jasowang@xxxxxxxxxx>
Cc: Juergen Gross <jgross@xxxxxxxx>
Cc: "K. Y. Srinivasan" <kys@xxxxxxxxxxxxx>
Cc: Marco Elver <elver@xxxxxxxxxx>
Cc: Michael S. Tsirkin <mst@xxxxxxxxxx>
Cc: Mike Rapoport (IBM) <rppt@xxxxxxxxxx>
Cc: Oleksandr Tyshchenko <oleksandr_tyshchenko@xxxxxxxx>
Cc: Oscar Salvador <osalvador@xxxxxxx>
Cc: Stefano Stabellini <sstabellini@xxxxxxxxxx>
Cc: Wei Liu <wei.liu@xxxxxxxxxx>
Cc: Xuan Zhuo <xuanzhuo@xxxxxxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
mm/internal.h | 3 ++-
mm/kmsan/init.c | 2 +-
mm/memory_hotplug.c | 9 +--------
mm/mm_init.c | 4 ++--
mm/page_alloc.c | 17 +++++++++++++++--
5 files changed, 21 insertions(+), 14 deletions(-)
--- a/mm/internal.h~mm-pass-meminit_context-to-__free_pages_core
+++ a/mm/internal.h
@@ -604,7 +604,8 @@ extern void __putback_isolated_page(stru
int mt);
extern void memblock_free_pages(struct page *page, unsigned long pfn,
unsigned int order);
-extern void __free_pages_core(struct page *page, unsigned int order);
+extern void __free_pages_core(struct page *page, unsigned int order,
+ enum meminit_context);
/*
* This will have no effect, other than possibly generating a warning, if the
--- a/mm/kmsan/init.c~mm-pass-meminit_context-to-__free_pages_core
+++ a/mm/kmsan/init.c
@@ -172,7 +172,7 @@ static void do_collection(void)
shadow = smallstack_pop(&collect);
origin = smallstack_pop(&collect);
kmsan_setup_meta(page, shadow, origin, collect.order);
- __free_pages_core(page, collect.order);
+ __free_pages_core(page, collect.order, MEMINIT_EARLY);
}
}
--- a/mm/memory_hotplug.c~mm-pass-meminit_context-to-__free_pages_core
+++ a/mm/memory_hotplug.c
@@ -630,14 +630,7 @@ EXPORT_SYMBOL_GPL(restore_online_page_ca
void generic_online_page(struct page *page, unsigned int order)
{
- /*
- * Freeing the page with debug_pagealloc enabled will try to unmap it,
- * so we should map it first. This is better than introducing a special
- * case in page freeing fast path.
- */
- debug_pagealloc_map_pages(page, 1 << order);
- __free_pages_core(page, order);
- totalram_pages_add(1UL << order);
+ __free_pages_core(page, order, MEMINIT_HOTPLUG);
}
EXPORT_SYMBOL_GPL(generic_online_page);
--- a/mm/mm_init.c~mm-pass-meminit_context-to-__free_pages_core
+++ a/mm/mm_init.c
@@ -1938,7 +1938,7 @@ static void __init deferred_free_range(u
for (i = 0; i < nr_pages; i++, page++, pfn++) {
if (pageblock_aligned(pfn))
set_pageblock_migratetype(page, MIGRATE_MOVABLE);
- __free_pages_core(page, 0);
+ __free_pages_core(page, 0, MEMINIT_EARLY);
}
}
@@ -2513,7 +2513,7 @@ void __init memblock_free_pages(struct p
}
}
- __free_pages_core(page, order);
+ __free_pages_core(page, order, MEMINIT_EARLY);
}
DEFINE_STATIC_KEY_MAYBE(CONFIG_INIT_ON_ALLOC_DEFAULT_ON, init_on_alloc);
--- a/mm/page_alloc.c~mm-pass-meminit_context-to-__free_pages_core
+++ a/mm/page_alloc.c
@@ -1214,7 +1214,8 @@ static void __free_pages_ok(struct page
__count_vm_events(PGFREE, 1 << order);
}
-void __free_pages_core(struct page *page, unsigned int order)
+void __free_pages_core(struct page *page, unsigned int order,
+ enum meminit_context context)
{
unsigned int nr_pages = 1 << order;
struct page *p = page;
@@ -1234,7 +1235,19 @@ void __free_pages_core(struct page *page
__ClearPageReserved(p);
set_page_count(p, 0);
- atomic_long_add(nr_pages, &page_zone(page)->managed_pages);
+ if (IS_ENABLED(CONFIG_MEMORY_HOTPLUG) &&
+ unlikely(context == MEMINIT_HOTPLUG)) {
+ /*
+ * Freeing the page with debug_pagealloc enabled will try to
+ * unmap it; some archs don't like double-unmappings, so
+ * map it first.
+ */
+ debug_pagealloc_map_pages(page, nr_pages);
+ adjust_managed_page_count(page, nr_pages);
+ } else {
+ /* memblock adjusts totalram_pages() ahead of time. */
+ atomic_long_add(nr_pages, &page_zone(page)->managed_pages);
+ }
if (page_contains_unaccepted(page, order)) {
if (order == MAX_PAGE_ORDER && __free_unaccepted(page))
_
Patches currently in -mm which might be from david@xxxxxxxxxx are
revert-mm-init_mlocked_on_free_v3.patch
mm-memory-move-page_count-check-into-validate_page_before_insert.patch
mm-memory-cleanly-support-zeropage-in-vm_insert_page-vm_map_pages-and-vmf_insert_mixed.patch
mm-rmap-sanity-check-that-zeropages-are-not-passed-to-rmap.patch
mm-update-_mapcount-and-page_type-documentation.patch
mm-allow-reuse-of-the-lower-16-bit-of-the-page-type-with-an-actual-type.patch
mm-zsmalloc-use-a-proper-page-type.patch
mm-page_alloc-clear-pagebuddy-using-__clearpagebuddy-for-bad-pages.patch
mm-filemap-reinitialize-folio-_mapcount-directly.patch
mm-mm_init-initialize-page-_mapcount-directly-in-__init_single_page.patch
fs-proc-task_mmu-indicate-pm_file-for-pmd-mapped-file-thp.patch
fs-proc-task_mmu-dont-indicate-pm_mmap_exclusive-without-pm_present.patch
fs-proc-task_mmu-properly-detect-pm_mmap_exclusive-per-page-of-pmd-mapped-thps.patch
fs-proc-task_mmu-account-non-present-entries-as-maybe-shared-but-no-idea-how-often.patch
fs-proc-move-page_mapcount-to-fs-proc-internalh.patch
documentation-admin-guide-mm-pagemaprst-drop-using-pagemap-to-do-something-useful.patch
mm-pass-meminit_context-to-__free_pages_core.patch
mm-pass-meminit_context-to-__free_pages_core-fix.patch
mm-memory_hotplug-initialize-memmap-of-zone_device-with-pageoffline-instead-of-pagereserved.patch
mm-memory_hotplug-skip-adjust_managed_page_count-for-pageoffline-pages-when-offlining.patch
