ksm: use range-walk function to jump over holes in scan_get_next_rmap_item
PATCH v4 1/1 ksm: use range-walk function to jump over holes in scan_get_next_rmap_item
Currently, scan_get_next_rmap_item() walks every page address in a VMA to locate mergeable pages. This becomes highly inefficient when scanning large virtual memory areas that contain mostly unmapped regions.
This patch replaces the per-address lookup with a range walk using walk_page_range(). The range walker allows KSM to skip over entire unmapped holes in a VMA, avoiding unnecessary lookups. This problem was previously discussed in [1].
[1] https://lore.kernel.org/linux-mm/[email protected]/
---
v4:
- Make minimal changes to replace folio_walk by walk_page_range_vma
v3: https://lore.kernel.org/all/[email protected]/
- Treat THPs in ksm_pmd_entry
- Update ksm_scan.address outside walk_page_range
- Change goto to while loop
v2: https://lore.kernel.org/all/[email protected]/
- Use pmd_entry to walk page range
- Use cond_resched inside pmd_entry()
- walk_page_range returns page+folio
v1: https://lore.kernel.org/all/[email protected]/
--- mm/ksm.c | 113 ++++++++++++++++++++++++++++++++++++++++++++++++++----- 1 file changed, 104 insertions(+), 9 deletions(-)
mm/ksm.c +104 -9
@@ -2455,6 +2455,95 @@ static bool should_skip_rmap_item(struct folio *folio,return true; } +struct ksm_next_page_arg { +struct folio *folio; +struct page *page; +unsigned long addr; +}; + +static int ksm_next_page_pmd_entry (pmd_t *pmdp,unsigned long addr,unsigned long end, +struct mm_walk *walk) +{ +struct ksm_next_page_arg *private = walk->private; +struct vm_area_struct *vma = walk->vma; +pte_t *start_ptep =NULL , *ptep, pte; +struct mm_struct *mm = walk->mm; +struct folio *folio; +struct page *page; +spinlock_t *ptl; +pmd_t pmd; + +if (ksm_test_exit (mm)) +return 0 ; + +cond_resched (); + + pmd =pmdp_get_lockless (pmdp); +if (!pmd_present (pmd)) +return 0 ; + +if (IS_ENABLED (CONFIG_TRANSPARENT_HUGEPAGE ) &&pmd_leaf (pmd)) { + ptl =pmd_lock (mm, pmdp); + pmd =pmdp_get (pmdp); + +if (!pmd_present (pmd)) { + goto not_found_unlock; + }else if (pmd_leaf (pmd)) { + page =vm_normal_page_pmd (vma, addr, pmd); +if (!page) + goto not_found_unlock; + folio =page_folio (page); + +if (folio_is_zone_device (folio) || !folio_test_anon (folio)) + goto not_found_unlock; + + page += ((addr & (PMD_SIZE -1 )) >>PAGE_SHIFT ); + goto found_unlock; + } +spin_unlock (ptl); + } + + start_ptep =pte_offset_map_lock (mm, pmdp, addr, &ptl); +if (!start_ptep) +return 0 ; + +for (ptep = start_ptep; addr < end; ptep++, addr +=PAGE_SIZE ) { + pte =ptep_get (ptep); + +if (!pte_present (pte)) +continue ; + + page =vm_normal_page (vma, addr, pte); +if (!page) +continue ; + folio =page_folio (page); + +if (folio_is_zone_device (folio) || !folio_test_anon (folio)) +continue ; + goto found_unlock; + } + +not_found_unlock: +spin_unlock (ptl); +if (start_ptep) +pte_unmap (start_ptep); +return 0 ; +found_unlock: +folio_get (folio); +spin_unlock (ptl); +if (start_ptep) +pte_unmap (start_ptep); + private->page = page; + private->folio = folio; + private->addr = addr; +return 1 ; +} + +static struct mm_walk_ops ksm_next_page_ops = { + .pmd_entry = ksm_next_page_pmd_entry, + .walk_lock =PGWALK_RDLOCK , +}; +static struct ksm_rmap_item *scan_get_next_rmap_item (struct page **page) {struct mm_struct *mm;
@@ -2542,21 +2631,27 @@ static struct ksm_rmap_item *scan_get_next_rmap_item(struct page **page) ksm_scan.address = vma->vm_end;while (ksm_scan.address < vma->vm_end) { +struct ksm_next_page_arg ksm_next_page_arg;struct page *tmp_page =NULL ; -struct folio_walk fw;struct folio *folio;if (ksm_test_exit (mm))break ; - folio =folio_walk_start (&fw, vma, ksm_scan.address,0 ); -if (folio) { -if (!folio_is_zone_device (folio) && -folio_test_anon (folio)) { -folio_get (folio); - tmp_page = fw.page; - } -folio_walk_end (&fw, vma); +int found; + + found =walk_page_range_vma (vma, ksm_scan.address, + vma->vm_end, + &ksm_next_page_ops, + &ksm_next_page_arg); + +if (found >0 ) { + folio = ksm_next_page_arg.folio; + tmp_page = ksm_next_page_arg.page; + ksm_scan.address = ksm_next_page_arg.addr; + }else { +VM_WARN_ON_ONCE (found <0 ); + ksm_scan.address = vma->vm_end -PAGE_SIZE ; }if (tmp_page) {
Signature · 1 line
2.43.0
Andrew Morton
maintainerQuoted 11 lines · On Wed, 22 Oct 2025 12:30:59 -0300 Pedro Demarchi Gomes…
Currently,
scan_get_next_rmap_item()walks every page address in a VMA to locate mergeable pages. This becomes highly inefficient when scanning large virtual memory areas that contain mostly unmapped regions.This patch replaces the per-address lookup with a range walk using
walk_page_range(). The range walker allows KSM to skip over entire unmapped holes in a VMA, avoiding unnecessary lookups. This problem was previously discussed in [1].
Thanks. It would be helpful of the changelog were to tell people how significant this change is for our users.
Quoted 2 lines
Buried in here is a claim that large amount of CPU are being used, but nothing quantitative.
So is there something we can tell people who are looking at this patch in Feb 2026 and wondering "hm, should I add that to our kernel"?
Quoted 4 lines
Suggested-by David Hildenbrand <[email protected]>Co-developed-by David Hildenbrand <[email protected]>Signed-off-by David Hildenbrand <[email protected]>Fixes 31dbd01f3143 ("ksm: Kernel SamePage Merging")
If the observed runtime problem is bad enough then a cc:stable might be justified. But a description of that observed runtime behavior would be needed for that, please.
Quoted 26 lines · Andrew Morton
Quoted 11 lines · On Wed, 22 Oct 2025 12:30:59 -0300 Pedro Demarchi Gomes…
Currently,
scan_get_next_rmap_item()walks every page address in a VMA to locate mergeable pages. This becomes highly inefficient when scanning large virtual memory areas that contain mostly unmapped regions.This patch replaces the per-address lookup with a range walk using
walk_page_range(). The range walker allows KSM to skip over entire unmapped holes in a VMA, avoiding unnecessary lookups. This problem was previously discussed in [1].Thanks. It would be helpful of the changelog were to tell people how significant this change is for our users.
Quoted 2 lines
Buried in here is a claim that large amount of CPU are being used, but nothing quantitative.
So is there something we can tell people who are looking at this patch in Feb 2026 and wondering "hm, should I add that to our kernel"?
Quoted 8 lines
Quoted 4 lines
Suggested-by David Hildenbrand <[email protected]>Co-developed-by David Hildenbrand <[email protected]>Signed-off-by David Hildenbrand <[email protected]>Fixes 31dbd01f3143 ("ksm: Kernel SamePage Merging")If the observed runtime problem is bad enough then a cc:stable might be justified. But a description of that observed runtime behavior would be needed for that, please.
Agreed.
With the following simple program
#include <unistd.h>
#include <stdio.h>
#include <sys/mman.h>
/* 32 TiB */
const size_t size = 32ul * 1024 * 1024 * 1024 * 1024;
int main() {
char *area = mmap(NULL, size, PROT_READ | PROT_WRITE,
MAP_NORESERVE | MAP_PRIVATE | MAP_ANON, -1, 0);
if (area == MAP_FAILED) {
perror("mmap() failed\n");
return -1;
}
/* Populate a single page such that we get an anon_vma. */
*area = 0;
/* Enable KSM. */
madvise(area, size, MADV_MERGEABLE);
pause();
return 0;
}$ ./ksm-sparse & $ echo 1 > /sys/kernel/mm/ksm/run
ksmd goes to 100% for quite a long time.
Now imagine if a cloud user spins up a couple of these programs.
KSM in the system is essentially deadlocked not able to deduplicate anything of value.
@Pedro, can you incorporate all that in the patch description?
Signature · 3 lines
Cheers David / dhildenb
Quoted 68 lines · David Hildenbrand
Quoted 26 lines · Andrew Morton
Quoted 11 lines · On Wed, 22 Oct 2025 12:30:59 -0300 Pedro Demarchi Gomes…
Currently,
scan_get_next_rmap_item()walks every page address in a VMA to locate mergeable pages. This becomes highly inefficient when scanning large virtual memory areas that contain mostly unmapped regions.This patch replaces the per-address lookup with a range walk using
walk_page_range(). The range walker allows KSM to skip over entire unmapped holes in a VMA, avoiding unnecessary lookups. This problem was previously discussed in [1].Thanks. It would be helpful of the changelog were to tell people how significant this change is for our users.
Quoted 2 lines
Buried in here is a claim that large amount of CPU are being used, but nothing quantitative.
So is there something we can tell people who are looking at this patch in Feb 2026 and wondering "hm, should I add that to our kernel"?
Quoted 8 lines
Quoted 4 lines
Suggested-by David Hildenbrand <[email protected]>Co-developed-by David Hildenbrand <[email protected]>Signed-off-by David Hildenbrand <[email protected]>Fixes 31dbd01f3143 ("ksm: Kernel SamePage Merging")If the observed runtime problem is bad enough then a cc:stable might be justified. But a description of that observed runtime behavior would be needed for that, please.
Agreed.
With the following simple program
#include <
unistd.h>
#include <stdio.h>
#include <sys/mman.h>/* 32 TiB */
constsize_tsize = 32ul * 1024 * 1024 * 1024 * 1024;int
main(){
char *area =mmap(NULL, size, PROT_READ | PROT_WRITE, MAP_NORESERVE | MAP_PRIVATE | MAP_ANON, -1, 0);if (area ==
MAP_FAILED) {
perror("mmap() failed\n");
return -1;
}/* Populate a single page such that we get an
anon_vma. */ *area = 0;/* Enable KSM. */
madvise(area, size, MADV_MERGEABLE);
pause();
return 0;
}$ ./ksm-sparse & $ echo 1 > /sys/kernel/mm/ksm/run
Quoted 8 lines
ksmd goes to 100% for quite a long time.
Now imagine if a cloud user spins up a couple of these programs.
KSM in the system is essentially deadlocked not able to deduplicate anything of value.
@Pedro, can you incorporate all that in the patch description?
Ok, I will send a v5 with a better changelog and patch description. Thanks!
Quoted 68 lines · David Hildenbrand
Quoted 26 lines · Andrew Morton
Quoted 11 lines · On Wed, 22 Oct 2025 12:30:59 -0300 Pedro Demarchi Gomes…
Currently,
scan_get_next_rmap_item()walks every page address in a VMA to locate mergeable pages. This becomes highly inefficient when scanning large virtual memory areas that contain mostly unmapped regions.This patch replaces the per-address lookup with a range walk using
walk_page_range(). The range walker allows KSM to skip over entire unmapped holes in a VMA, avoiding unnecessary lookups. This problem was previously discussed in [1].Thanks. It would be helpful of the changelog were to tell people how significant this change is for our users.
Quoted 2 lines
Reported-by craftfever <[email protected]>Buried in here is a claim that large amount of CPU are being used, but nothing quantitative.
So is there something we can tell people who are looking at this patch in Feb 2026 and wondering "hm, should I add that to our kernel"?
Quoted 8 lines
Quoted 4 lines
Suggested-by David Hildenbrand <[email protected]>Co-developed-by David Hildenbrand <[email protected]>Signed-off-by David Hildenbrand <[email protected]>Fixes 31dbd01f3143 ("ksm: Kernel SamePage Merging")If the observed runtime problem is bad enough then a cc:stable might be justified. But a description of that observed runtime behavior would be needed for that, please.
Agreed.
With the following simple program
#include <
unistd.h>
#include <stdio.h>
#include <sys/mman.h>/* 32 TiB */
constsize_tsize = 32ul * 1024 * 1024 * 1024 * 1024;int
main(){
char *area =mmap(NULL, size, PROT_READ | PROT_WRITE, MAP_NORESERVE | MAP_PRIVATE | MAP_ANON, -1, 0);if (area ==
MAP_FAILED) {
perror("mmap() failed\n");
return -1;
}/* Populate a single page such that we get an
anon_vma. */ *area = 0;/* Enable KSM. */
madvise(area, size, MADV_MERGEABLE);
pause();
return 0;
}$ ./ksm-sparse & $ echo 1 > /sys/kernel/mm/ksm/run
Quoted 9 lines
ksmd goes to 100% for quite a long time.
Now imagine if a cloud user spins up a couple of these programs.
KSM in the system is essentially deadlocked not able to deduplicate anything of value.
@Pedro, can you incorporate all that in the patch description?
Thanks for example and explanation, that's exactly what I meant. Big datacenters and servers are primary use cases with Linux< for example, when many VMs are using and KSM operation have to be very robust to deal with so huge amount of memory. And, as being said, that bug is happened with consumers apps, like Chromkum/Electron (and based on it, like VS Code), when user decides to apply KSM for his apps. This patch is really necessary to apply in master branch and very preferably backport 6.17-stable branch, it has high importance. I've tested by the way v4, and it's fine, stable, effective and very lite. Is v5 version with more comprehensive description is final version?