The patch titled
Subject: mm: keep nid around during hot-remove
has been added to the -mm mm-hotfixes-unstable branch. Its filename is
mm-keep-nid-around-during-hot-remove.patch
This patch will shortly appear at
https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patches/mm-keep-nid-around-during-hot-remove.patch
This patch will later appear in the mm-hotfixes-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: Pasha Tatashin <pasha.tatashin@xxxxxxxxxx>
Subject: mm: keep nid around during hot-remove
Date: Tue, 6 Aug 2024 22:14:54 +0000
nid is needed during memory hot-remove in order to account the information
about the memmap overhead that is being removed.
In addition, we cannot use page_pgdat(pfn_to_page(pfn)) during hotremove
after remove_pfn_range_from_zone().
We also cannot determine nid from walking through memblocks after
remove_memory_block_devices() is called.
Therefore, pass nid down from the beginning of hotremove to where it is
used for the accounting purposes.
Link: https://lkml.kernel.org/r/20240806221454.1971755-2-pasha.tatashin@xxxxxxxxxx
Fixes: 15995a352474 ("mm: report per-page metadata information")
Signed-off-by: Pasha Tatashin <pasha.tatashin@xxxxxxxxxx>
Reported-by: Yi Zhang <yi.zhang@xxxxxxxxxx>
Closes: https://lore.kernel.org/linux-cxl/CAHj4cs9Ax1=CoJkgBGP_+sNu6-6=6v=_L-ZBZY0bVLD3wUWZQg@xxxxxxxxxxxxxx
Reported-by: Alison Schofield <alison.schofield@xxxxxxxxx>
Closes: https://lore.kernel.org/linux-mm/Zq0tPd2h6alFz8XF@aschofie-mobl2/#t
Cc: Albert Ou <aou@xxxxxxxxxxxxxxxxx>
Cc: Alexander Gordeev <agordeev@xxxxxxxxxxxxx>
Cc: Alexandre Ghiti <alexghiti@xxxxxxxxxxxx>
Cc: Andy Lutomirski <luto@xxxxxxxxxx>
Cc: Ard Biesheuvel <ardb@xxxxxxxxxx>
Cc: Arnd Bergmann <arnd@xxxxxxxx>
Cc: Baoquan He <bhe@xxxxxxxxxx>
Cc: Björn Töpel <bjorn@xxxxxxxxxxxx>
Cc: Borislav Petkov <bp@xxxxxxxxx>
Cc: Catalin Marinas <catalin.marinas@xxxxxxx>
Cc: Chen Jiahao <chenjiahao16@xxxxxxxxxx>
Cc: Christian Borntraeger <borntraeger@xxxxxxxxxxxxx>
Cc: Christophe Leroy <christophe.leroy@xxxxxxxxxx>
Cc: Dave Hansen <dave.hansen@xxxxxxxxxxxxxxx>
Cc: David Hildenbrand <david@xxxxxxxxxx>
Cc: David Rientjes <rientjes@xxxxxxxxxx>
Cc: Dawei Li <dawei.li@xxxxxxxxxxxx>
Cc: Gerald Schaefer <gerald.schaefer@xxxxxxxxxxxxx>
Cc: Heiko Carstens <hca@xxxxxxxxxxxxx>
Cc: "H. Peter Anvin" <hpa@xxxxxxxxx>
Cc: Huacai Chen <chenhuacai@xxxxxxxxxx>
Cc: Ingo Molnar <mingo@xxxxxxxxxx>
Cc: Kent Overstreet <kent.overstreet@xxxxxxxxx>
Cc: Luis Chamberlain <mcgrof@xxxxxxxxxx>
Cc: Mark Rutland <mark.rutland@xxxxxxx>
Cc: Michael Ellerman <mpe@xxxxxxxxxxxxxx>
Cc: Mike Rapoport <rppt@xxxxxxxxxx>
Cc: Muchun Song <muchun.song@xxxxxxxxx>
Cc: Nam Cao <namcao@xxxxxxxxxxxxx>
Cc: Naveen N Rao <naveen@xxxxxxxxxx>
Cc: Nicholas Piggin <npiggin@xxxxxxxxx>
Cc: Oscar Salvador <osalvador@xxxxxxx>
Cc: Palmer Dabbelt <palmer@xxxxxxxxxxx>
Cc: Paul Walmsley <paul.walmsley@xxxxxxxxxx>
Cc: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>
Cc: Philippe Mathieu-Daudé <philmd@xxxxxxxxxx>
Cc: Randy Dunlap <rdunlap@xxxxxxxxxxxxx>
Cc: Ryan Roberts <ryan.roberts@xxxxxxx>
Cc: Sourav Panda <souravpanda@xxxxxxxxxx>
Cc: Sven Schnelle <svens@xxxxxxxxxxxxx>
Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Cc: Thomas Zimmermann <tzimmermann@xxxxxxx>
Cc: Vasily Gorbik <gor@xxxxxxxxxxxxx>
Cc: WANG Xuerui <kernel@xxxxxxxxxx>
Cc: Will Deacon <will@xxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
arch/arm64/mm/mmu.c | 5 +++--
arch/loongarch/mm/init.c | 5 +++--
arch/powerpc/mm/mem.c | 5 +++--
arch/riscv/mm/init.c | 5 +++--
arch/s390/mm/init.c | 5 +++--
arch/x86/mm/init_64.c | 5 +++--
include/linux/memory_hotplug.h | 7 ++++---
mm/memory_hotplug.c | 18 +++++++++---------
mm/memremap.c | 6 ++++--
mm/sparse-vmemmap.c | 14 ++++++++------
mm/sparse.c | 20 +++++++++++---------
11 files changed, 54 insertions(+), 41 deletions(-)
--- a/arch/arm64/mm/mmu.c~mm-keep-nid-around-during-hot-remove
+++ a/arch/arm64/mm/mmu.c
@@ -1363,12 +1363,13 @@ int arch_add_memory(int nid, u64 start,
return ret;
}
-void arch_remove_memory(u64 start, u64 size, struct vmem_altmap *altmap)
+void arch_remove_memory(u64 start, u64 size, struct vmem_altmap *altmap,
+ int nid)
{
unsigned long start_pfn = start >> PAGE_SHIFT;
unsigned long nr_pages = size >> PAGE_SHIFT;
- __remove_pages(start_pfn, nr_pages, altmap);
+ __remove_pages(start_pfn, nr_pages, altmap, nid);
__remove_pgd_mapping(swapper_pg_dir, __phys_to_virt(start), size);
}
--- a/arch/loongarch/mm/init.c~mm-keep-nid-around-during-hot-remove
+++ a/arch/loongarch/mm/init.c
@@ -106,7 +106,8 @@ int arch_add_memory(int nid, u64 start,
return ret;
}
-void arch_remove_memory(u64 start, u64 size, struct vmem_altmap *altmap)
+void arch_remove_memory(u64 start, u64 size, struct vmem_altmap *altmap,
+ int nid)
{
unsigned long start_pfn = start >> PAGE_SHIFT;
unsigned long nr_pages = size >> PAGE_SHIFT;
@@ -115,7 +116,7 @@ void arch_remove_memory(u64 start, u64 s
/* With altmap the first mapped page is offset from @start */
if (altmap)
page += vmem_altmap_offset(altmap);
- __remove_pages(start_pfn, nr_pages, altmap);
+ __remove_pages(start_pfn, nr_pages, altmap, nid);
}
#ifdef CONFIG_NUMA
--- a/arch/powerpc/mm/mem.c~mm-keep-nid-around-during-hot-remove
+++ a/arch/powerpc/mm/mem.c
@@ -157,12 +157,13 @@ int __ref arch_add_memory(int nid, u64 s
return rc;
}
-void __ref arch_remove_memory(u64 start, u64 size, struct vmem_altmap *altmap)
+void __ref arch_remove_memory(u64 start, u64 size, struct vmem_altmap *altmap,
+ int nid)
{
unsigned long start_pfn = start >> PAGE_SHIFT;
unsigned long nr_pages = size >> PAGE_SHIFT;
- __remove_pages(start_pfn, nr_pages, altmap);
+ __remove_pages(start_pfn, nr_pages, altmap, nid);
arch_remove_linear_mapping(start, size);
}
#endif
--- a/arch/riscv/mm/init.c~mm-keep-nid-around-during-hot-remove
+++ a/arch/riscv/mm/init.c
@@ -1789,9 +1789,10 @@ int __ref arch_add_memory(int nid, u64 s
return ret;
}
-void __ref arch_remove_memory(u64 start, u64 size, struct vmem_altmap *altmap)
+void __ref arch_remove_memory(u64 start, u64 size, struct vmem_altmap *altmap,
+ int nid)
{
- __remove_pages(start >> PAGE_SHIFT, size >> PAGE_SHIFT, altmap);
+ __remove_pages(start >> PAGE_SHIFT, size >> PAGE_SHIFT, altmap, nid);
remove_linear_mapping(start, size);
flush_tlb_all();
}
--- a/arch/s390/mm/init.c~mm-keep-nid-around-during-hot-remove
+++ a/arch/s390/mm/init.c
@@ -295,12 +295,13 @@ int arch_add_memory(int nid, u64 start,
return rc;
}
-void arch_remove_memory(u64 start, u64 size, struct vmem_altmap *altmap)
+void arch_remove_memory(u64 start, u64 size, struct vmem_altmap *altmap,
+ int nid)
{
unsigned long start_pfn = start >> PAGE_SHIFT;
unsigned long nr_pages = size >> PAGE_SHIFT;
- __remove_pages(start_pfn, nr_pages, altmap);
+ __remove_pages(start_pfn, nr_pages, altmap, nid);
vmem_remove_mapping(start, size);
}
#endif /* CONFIG_MEMORY_HOTPLUG */
--- a/arch/x86/mm/init_64.c~mm-keep-nid-around-during-hot-remove
+++ a/arch/x86/mm/init_64.c
@@ -1262,12 +1262,13 @@ kernel_physical_mapping_remove(unsigned
remove_pagetable(start, end, true, NULL);
}
-void __ref arch_remove_memory(u64 start, u64 size, struct vmem_altmap *altmap)
+void __ref arch_remove_memory(u64 start, u64 size, struct vmem_altmap *altmap,
+ int nid)
{
unsigned long start_pfn = start >> PAGE_SHIFT;
unsigned long nr_pages = size >> PAGE_SHIFT;
- __remove_pages(start_pfn, nr_pages, altmap);
+ __remove_pages(start_pfn, nr_pages, altmap, nid);
kernel_physical_mapping_remove(start, start + size);
}
#endif /* CONFIG_MEMORY_HOTPLUG */
--- a/include/linux/memory_hotplug.h~mm-keep-nid-around-during-hot-remove
+++ a/include/linux/memory_hotplug.h
@@ -201,9 +201,10 @@ static inline bool movable_node_is_enabl
return movable_node_enabled;
}
-extern void arch_remove_memory(u64 start, u64 size, struct vmem_altmap *altmap);
+extern void arch_remove_memory(u64 start, u64 size, struct vmem_altmap *altmap,
+ int nid);
extern void __remove_pages(unsigned long start_pfn, unsigned long nr_pages,
- struct vmem_altmap *altmap);
+ struct vmem_altmap *altmap, int nid);
/* reasonably generic interface to expand the physical pages */
extern int __add_pages(int nid, unsigned long start_pfn, unsigned long nr_pages,
@@ -369,7 +370,7 @@ extern int sparse_add_section(int nid, u
unsigned long nr_pages, struct vmem_altmap *altmap,
struct dev_pagemap *pgmap);
extern void sparse_remove_section(unsigned long pfn, unsigned long nr_pages,
- struct vmem_altmap *altmap);
+ struct vmem_altmap *altmap, int nid);
extern struct page *sparse_decode_mem_map(unsigned long coded_mem_map,
unsigned long pnum);
extern struct zone *zone_for_pfn_range(int online_type, int nid,
--- a/mm/memory_hotplug.c~mm-keep-nid-around-during-hot-remove
+++ a/mm/memory_hotplug.c
@@ -571,7 +571,7 @@ void __ref remove_pfn_range_from_zone(st
* calling offline_pages().
*/
void __remove_pages(unsigned long pfn, unsigned long nr_pages,
- struct vmem_altmap *altmap)
+ struct vmem_altmap *altmap, int nid)
{
const unsigned long end_pfn = pfn + nr_pages;
unsigned long cur_nr_pages;
@@ -586,7 +586,7 @@ void __remove_pages(unsigned long pfn, u
/* Select all remaining pages up to the next section boundary */
cur_nr_pages = min(end_pfn - pfn,
SECTION_ALIGN_UP(pfn + 1) - pfn);
- sparse_remove_section(pfn, cur_nr_pages, altmap);
+ sparse_remove_section(pfn, cur_nr_pages, altmap, nid);
}
}
@@ -1386,7 +1386,7 @@ bool mhp_supports_memmap_on_memory(void)
}
EXPORT_SYMBOL_GPL(mhp_supports_memmap_on_memory);
-static void __ref remove_memory_blocks_and_altmaps(u64 start, u64 size)
+static void __ref remove_memory_blocks_and_altmaps(u64 start, u64 size, int nid)
{
unsigned long memblock_size = memory_block_size_bytes();
u64 cur_start;
@@ -1409,7 +1409,7 @@ static void __ref remove_memory_blocks_a
remove_memory_block_devices(cur_start, memblock_size);
- arch_remove_memory(cur_start, memblock_size, altmap);
+ arch_remove_memory(cur_start, memblock_size, altmap, nid);
/* Verify that all vmemmap pages have actually been freed. */
WARN(altmap->alloc, "Altmap not fully unmapped");
@@ -1454,7 +1454,7 @@ static int create_altmaps_and_memory_blo
ret = create_memory_block_devices(cur_start, memblock_size,
params.altmap, group);
if (ret) {
- arch_remove_memory(cur_start, memblock_size, NULL);
+ arch_remove_memory(cur_start, memblock_size, NULL, nid);
kfree(params.altmap);
goto out;
}
@@ -1463,7 +1463,7 @@ static int create_altmaps_and_memory_blo
return 0;
out:
if (ret && cur_start != start)
- remove_memory_blocks_and_altmaps(start, cur_start - start);
+ remove_memory_blocks_and_altmaps(start, cur_start - start, nid);
return ret;
}
@@ -1532,7 +1532,7 @@ int __ref add_memory_resource(int nid, s
/* create memory block devices after memory was added */
ret = create_memory_block_devices(start, size, NULL, group);
if (ret) {
- arch_remove_memory(start, size, params.altmap);
+ arch_remove_memory(start, size, params.altmap, nid);
goto error;
}
}
@@ -2275,10 +2275,10 @@ static int __ref try_remove_memory(u64 s
* No altmaps present, do the removal directly
*/
remove_memory_block_devices(start, size);
- arch_remove_memory(start, size, NULL);
+ arch_remove_memory(start, size, NULL, nid);
} else {
/* all memblocks in the range have altmaps */
- remove_memory_blocks_and_altmaps(start, size);
+ remove_memory_blocks_and_altmaps(start, size, nid);
}
if (IS_ENABLED(CONFIG_ARCH_KEEP_MEMBLOCK))
--- a/mm/memremap.c~mm-keep-nid-around-during-hot-remove
+++ a/mm/memremap.c
@@ -112,9 +112,11 @@ static void pageunmap_range(struct dev_p
{
struct range *range = &pgmap->ranges[range_id];
struct page *first_page;
+ int nid;
/* make sure to access a memmap that was actually initialized */
first_page = pfn_to_page(pfn_first(pgmap, range_id));
+ nid = page_to_nid(first_page);
/* pages are dead and unused, undo the arch mapping */
mem_hotplug_begin();
@@ -122,10 +124,10 @@ static void pageunmap_range(struct dev_p
PHYS_PFN(range_len(range)));
if (pgmap->type == MEMORY_DEVICE_PRIVATE) {
__remove_pages(PHYS_PFN(range->start),
- PHYS_PFN(range_len(range)), NULL);
+ PHYS_PFN(range_len(range)), NULL, nid);
} else {
arch_remove_memory(range->start, range_len(range),
- pgmap_altmap(pgmap));
+ pgmap_altmap(pgmap), nid);
kasan_remove_zero_shadow(__va(range->start), range_len(range));
}
mem_hotplug_done();
--- a/mm/sparse.c~mm-keep-nid-around-during-hot-remove
+++ a/mm/sparse.c
@@ -638,13 +638,15 @@ static struct page * __meminit populate_
}
static void depopulate_section_memmap(unsigned long pfn, unsigned long nr_pages,
- struct vmem_altmap *altmap)
+ struct vmem_altmap *altmap, int nid)
{
unsigned long start = (unsigned long) pfn_to_page(pfn);
unsigned long end = start + nr_pages * sizeof(struct page);
- mod_node_page_state(page_pgdat(pfn_to_page(pfn)), NR_MEMMAP,
- -1L * (DIV_ROUND_UP(end - start, PAGE_SIZE)));
+ if (nid != NUMA_NO_NODE) {
+ mod_node_page_state(NODE_DATA(nid), NR_MEMMAP,
+ -1L * (DIV_ROUND_UP(end - start, PAGE_SIZE)));
+ }
vmemmap_free(start, end, altmap);
}
static void free_map_bootmem(struct page *memmap)
@@ -713,7 +715,7 @@ static struct page * __meminit populate_
}
static void depopulate_section_memmap(unsigned long pfn, unsigned long nr_pages,
- struct vmem_altmap *altmap)
+ struct vmem_altmap *altmap, int nid)
{
kvfree(pfn_to_page(pfn));
}
@@ -781,7 +783,7 @@ static int fill_subsection_map(unsigned
* For 2 and 3, the SPARSEMEM_VMEMMAP={y,n} cases are unified
*/
static void section_deactivate(unsigned long pfn, unsigned long nr_pages,
- struct vmem_altmap *altmap)
+ struct vmem_altmap *altmap, int nid)
{
struct mem_section *ms = __pfn_to_section(pfn);
bool section_is_early = early_section(ms);
@@ -821,7 +823,7 @@ static void section_deactivate(unsigned
* section_activate() and pfn_valid() .
*/
if (!section_is_early)
- depopulate_section_memmap(pfn, nr_pages, altmap);
+ depopulate_section_memmap(pfn, nr_pages, altmap, nid);
else if (memmap)
free_map_bootmem(memmap);
@@ -865,7 +867,7 @@ static struct page * __meminit section_a
memmap = populate_section_memmap(pfn, nr_pages, nid, altmap, pgmap);
if (!memmap) {
- section_deactivate(pfn, nr_pages, altmap);
+ section_deactivate(pfn, nr_pages, altmap, nid);
return ERR_PTR(-ENOMEM);
}
@@ -928,13 +930,13 @@ int __meminit sparse_add_section(int nid
}
void sparse_remove_section(unsigned long pfn, unsigned long nr_pages,
- struct vmem_altmap *altmap)
+ struct vmem_altmap *altmap, int nid)
{
struct mem_section *ms = __pfn_to_section(pfn);
if (WARN_ON_ONCE(!valid_section(ms)))
return;
- section_deactivate(pfn, nr_pages, altmap);
+ section_deactivate(pfn, nr_pages, altmap, nid);
}
#endif /* CONFIG_MEMORY_HOTPLUG */
--- a/mm/sparse-vmemmap.c~mm-keep-nid-around-during-hot-remove
+++ a/mm/sparse-vmemmap.c
@@ -469,12 +469,14 @@ struct page * __meminit __populate_secti
if (r < 0)
return NULL;
- if (system_state == SYSTEM_BOOTING) {
- mod_node_early_perpage_metadata(nid, DIV_ROUND_UP(end - start,
- PAGE_SIZE));
- } else {
- mod_node_page_state(NODE_DATA(nid), NR_MEMMAP,
- DIV_ROUND_UP(end - start, PAGE_SIZE));
+ if (nid != NUMA_NO_NODE) {
+ if (system_state == SYSTEM_BOOTING) {
+ mod_node_early_perpage_metadata(nid, DIV_ROUND_UP(end - start,
+ PAGE_SIZE));
+ } else {
+ mod_node_page_state(NODE_DATA(nid), NR_MEMMAP,
+ DIV_ROUND_UP(end - start, PAGE_SIZE));
+ }
}
return pfn_to_page(pfn);
_
Patches currently in -mm which might be from pasha.tatashin@xxxxxxxxxx are
mm-update-the-memmap-stat-before-page-is-freed.patch
mm-keep-nid-around-during-hot-remove.patch
memcg-increase-the-valid-index-range-for-memcg-stats-v5.patch
vmstat-kernel-stack-usage-histogram.patch
task_stack-uninline-stack_not_used.patch
