The patch titled
Subject: mm/mmu_notifier: contextual information for event triggering invalidation
has been added to the -mm tree. Its filename is
mm-mmu_notifier-contextual-information-for-event-triggering-invalidation-v2.patch
This patch should soon appear at
http://ozlabs.org/~akpm/mmots/broken-out/mm-mmu_notifier-contextual-information-for-event-triggering-invalidation-v2.patch
and later at
http://ozlabs.org/~akpm/mmotm/broken-out/mm-mmu_notifier-contextual-information-for-event-triggering-invalidation-v2.patch
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 and is updated
there every 3-4 working days
------------------------------------------------------
From: Jérôme Glisse <jglisse@xxxxxxxxxx>
Subject: mm/mmu_notifier: contextual information for event triggering invalidation
CPU page table update can happens for many reasons, not only as a result
of a syscall (munmap(), mprotect(), mremap(), madvise(), ...) but also as
a result of kernel activities (memory compression, reclaim, migration,
...).
Users of mmu notifier API track changes to the CPU page table and take
specific action for them. While current API only provide range of virtual
address affected by the change, not why the changes is happening.
This patchset do the initial mechanical convertion of all the places that
calls mmu_notifier_range_init to also provide the default MMU_NOTIFY_UNMAP
event as well as the vma if it is know (most invalidation happens against
a given vma). Passing down the vma allows the users of mmu notifier to
inspect the new vma page protection.
The MMU_NOTIFY_UNMAP is always the safe default as users of mmu notifier
should assume that every for the range is going away when that event
happens. A latter patch do convert mm call path to use a more appropriate
events for each call.
This is done as 2 patches so that no call site is forgotten especialy
as it uses this following coccinelle patch:
%<----------------------------------------------------------------------
@@
identifier I1, I2, I3, I4;
@@
static inline void mmu_notifier_range_init(struct mmu_notifier_range *I1,
+enum mmu_notifier_event event,
+unsigned flags,
+struct vm_area_struct *vma,
struct mm_struct *I2, unsigned long I3, unsigned long I4) { ... }
@@
@@
-#define mmu_notifier_range_init(range, mm, start, end)
+#define mmu_notifier_range_init(range, event, flags, vma, mm, start, end)
@@
expression E1, E3, E4;
identifier I1;
@@
<...
mmu_notifier_range_init(E1,
+MMU_NOTIFY_UNMAP, 0, I1,
I1->vm_mm, E3, E4)
...>
@@
expression E1, E2, E3, E4;
identifier FN, VMA;
@@
FN(..., struct vm_area_struct *VMA, ...) {
<...
mmu_notifier_range_init(E1,
+MMU_NOTIFY_UNMAP, 0, VMA,
E2, E3, E4)
...> }
@@
expression E1, E2, E3, E4;
identifier FN, VMA;
@@
FN(...) {
struct vm_area_struct *VMA;
<...
mmu_notifier_range_init(E1,
+MMU_NOTIFY_UNMAP, 0, VMA,
E2, E3, E4)
...> }
@@
expression E1, E2, E3, E4;
identifier FN;
@@
FN(...) {
<...
mmu_notifier_range_init(E1,
+MMU_NOTIFY_UNMAP, 0, NULL,
E2, E3, E4)
...> }
---------------------------------------------------------------------->%
Applied with:
spatch --all-includes --sp-file mmu-notifier.spatch fs/proc/task_mmu.c --in-place
spatch --sp-file mmu-notifier.spatch --dir kernel/events/ --in-place
spatch --sp-file mmu-notifier.spatch --dir mm --in-place
Link: http://lkml.kernel.org/r/20190326164747.24405-6-jglisse@xxxxxxxxxx
Signed-off-by: Jérôme Glisse <jglisse@xxxxxxxxxx>
Cc: Christian König <christian.koenig@xxxxxxx>
Cc: Joonas Lahtinen <joonas.lahtinen@xxxxxxxxxxxxxxx>
Cc: Jani Nikula <jani.nikula@xxxxxxxxxxxxxxx>
Cc: Rodrigo Vivi <rodrigo.vivi@xxxxxxxxx>
Cc: Jan Kara <jack@xxxxxxx>
Cc: Andrea Arcangeli <aarcange@xxxxxxxxxx>
Cc: Peter Xu <peterx@xxxxxxxxxx>
Cc: Felix Kuehling <Felix.Kuehling@xxxxxxx>
Cc: Jason Gunthorpe <jgg@xxxxxxxxxxxx>
Cc: Ross Zwisler <zwisler@xxxxxxxxxx>
Cc: Dan Williams <dan.j.williams@xxxxxxxxx>
Cc: Paolo Bonzini <pbonzini@xxxxxxxxxx>
Cc: Radim Krcmar <rkrcmar@xxxxxxxxxx>
Cc: Michal Hocko <mhocko@xxxxxxxxxx>
Cc: Christian Koenig <christian.koenig@xxxxxxx>
Cc: Ralph Campbell <rcampbell@xxxxxxxxxx>
Cc: John Hubbard <jhubbard@xxxxxxxxxx>
Cc: Arnd Bergmann <arnd@xxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
fs/proc/task_mmu.c | 3 ++-
include/linux/mmu_notifier.h | 5 ++++-
kernel/events/uprobes.c | 3 ++-
mm/huge_memory.c | 12 ++++++++----
mm/hugetlb.c | 12 ++++++++----
mm/khugepaged.c | 3 ++-
mm/ksm.c | 6 ++++--
mm/madvise.c | 3 ++-
mm/memory.c | 25 ++++++++++++++++---------
mm/migrate.c | 5 ++++-
mm/mprotect.c | 3 ++-
mm/mremap.c | 3 ++-
mm/oom_kill.c | 3 ++-
mm/rmap.c | 6 ++++--
14 files changed, 62 insertions(+), 30 deletions(-)
--- a/fs/proc/task_mmu.c~mm-mmu_notifier-contextual-information-for-event-triggering-invalidation-v2
+++ a/fs/proc/task_mmu.c
@@ -1169,7 +1169,8 @@ static ssize_t clear_refs_write(struct f
break;
}
- mmu_notifier_range_init(&range, mm, 0, -1UL);
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0,
+ NULL, mm, 0, -1UL);
mmu_notifier_invalidate_range_start(&range);
}
walk_page_range(0, mm->highest_vm_end, &clear_refs_walk);
--- a/include/linux/mmu_notifier.h~mm-mmu_notifier-contextual-information-for-event-triggering-invalidation-v2
+++ a/include/linux/mmu_notifier.h
@@ -356,6 +356,9 @@ static inline void mmu_notifier_mm_destr
static inline void mmu_notifier_range_init(struct mmu_notifier_range *range,
+ enum mmu_notifier_event event,
+ unsigned flags,
+ struct vm_area_struct *vma,
struct mm_struct *mm,
unsigned long start,
unsigned long end)
@@ -491,7 +494,7 @@ static inline void _mmu_notifier_range_i
range->end = end;
}
-#define mmu_notifier_range_init(range, mm, start, end) \
+#define mmu_notifier_range_init(range,event,flags,vma,mm,start,end) \
_mmu_notifier_range_init(range, start, end)
static inline bool
--- a/kernel/events/uprobes.c~mm-mmu_notifier-contextual-information-for-event-triggering-invalidation-v2
+++ a/kernel/events/uprobes.c
@@ -161,7 +161,8 @@ static int __replace_page(struct vm_area
struct mmu_notifier_range range;
struct mem_cgroup *memcg;
- mmu_notifier_range_init(&range, mm, addr, addr + PAGE_SIZE);
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, mm, addr,
+ addr + PAGE_SIZE);
VM_BUG_ON_PAGE(PageTransHuge(old_page), old_page);
--- a/mm/huge_memory.c~mm-mmu_notifier-contextual-information-for-event-triggering-invalidation-v2
+++ a/mm/huge_memory.c
@@ -1220,7 +1220,8 @@ static vm_fault_t do_huge_pmd_wp_page_fa
cond_resched();
}
- mmu_notifier_range_init(&range, vma->vm_mm, haddr,
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, vma->vm_mm,
+ haddr,
haddr + HPAGE_PMD_SIZE);
mmu_notifier_invalidate_range_start(&range);
@@ -1384,7 +1385,8 @@ alloc:
vma, HPAGE_PMD_NR);
__SetPageUptodate(new_page);
- mmu_notifier_range_init(&range, vma->vm_mm, haddr,
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, vma->vm_mm,
+ haddr,
haddr + HPAGE_PMD_SIZE);
mmu_notifier_invalidate_range_start(&range);
@@ -2060,7 +2062,8 @@ void __split_huge_pud(struct vm_area_str
spinlock_t *ptl;
struct mmu_notifier_range range;
- mmu_notifier_range_init(&range, vma->vm_mm, address & HPAGE_PUD_MASK,
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, vma->vm_mm,
+ address & HPAGE_PUD_MASK,
(address & HPAGE_PUD_MASK) + HPAGE_PUD_SIZE);
mmu_notifier_invalidate_range_start(&range);
ptl = pud_lock(vma->vm_mm, pud);
@@ -2278,7 +2281,8 @@ void __split_huge_pmd(struct vm_area_str
spinlock_t *ptl;
struct mmu_notifier_range range;
- mmu_notifier_range_init(&range, vma->vm_mm, address & HPAGE_PMD_MASK,
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, vma->vm_mm,
+ address & HPAGE_PMD_MASK,
(address & HPAGE_PMD_MASK) + HPAGE_PMD_SIZE);
mmu_notifier_invalidate_range_start(&range);
ptl = pmd_lock(vma->vm_mm, pmd);
--- a/mm/hugetlb.c~mm-mmu_notifier-contextual-information-for-event-triggering-invalidation-v2
+++ a/mm/hugetlb.c
@@ -3293,7 +3293,8 @@ int copy_hugetlb_page_range(struct mm_st
cow = (vma->vm_flags & (VM_SHARED | VM_MAYWRITE)) == VM_MAYWRITE;
if (cow) {
- mmu_notifier_range_init(&range, src, vma->vm_start,
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, src,
+ vma->vm_start,
vma->vm_end);
mmu_notifier_invalidate_range_start(&range);
}
@@ -3405,7 +3406,8 @@ void __unmap_hugepage_range(struct mmu_g
/*
* If sharing possible, alert mmu notifiers of worst case.
*/
- mmu_notifier_range_init(&range, mm, start, end);
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, mm, start,
+ end);
adjust_range_if_pmd_sharing_possible(vma, &range.start, &range.end);
mmu_notifier_invalidate_range_start(&range);
address = start;
@@ -3672,7 +3674,8 @@ retry_avoidcopy:
pages_per_huge_page(h));
__SetPageUptodate(new_page);
- mmu_notifier_range_init(&range, mm, haddr, haddr + huge_page_size(h));
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, mm, haddr,
+ haddr + huge_page_size(h));
mmu_notifier_invalidate_range_start(&range);
/*
@@ -4394,7 +4397,8 @@ unsigned long hugetlb_change_protection(
* start/end. Set range.start/range.end to cover the maximum possible
* range if PMD sharing is possible.
*/
- mmu_notifier_range_init(&range, mm, start, end);
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, mm, start,
+ end);
adjust_range_if_pmd_sharing_possible(vma, &range.start, &range.end);
BUG_ON(address >= end);
--- a/mm/khugepaged.c~mm-mmu_notifier-contextual-information-for-event-triggering-invalidation-v2
+++ a/mm/khugepaged.c
@@ -1016,7 +1016,8 @@ static void collapse_huge_page(struct mm
pte = pte_offset_map(pmd, address);
pte_ptl = pte_lockptr(mm, pmd);
- mmu_notifier_range_init(&range, mm, address, address + HPAGE_PMD_SIZE);
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, NULL, mm,
+ address, address + HPAGE_PMD_SIZE);
mmu_notifier_invalidate_range_start(&range);
pmd_ptl = pmd_lock(mm, pmd); /* probably unnecessary */
/*
--- a/mm/ksm.c~mm-mmu_notifier-contextual-information-for-event-triggering-invalidation-v2
+++ a/mm/ksm.c
@@ -1066,7 +1066,8 @@ static int write_protect_page(struct vm_
BUG_ON(PageTransCompound(page));
- mmu_notifier_range_init(&range, mm, pvmw.address,
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, mm,
+ pvmw.address,
pvmw.address + PAGE_SIZE);
mmu_notifier_invalidate_range_start(&range);
@@ -1154,7 +1155,8 @@ static int replace_page(struct vm_area_s
if (!pmd)
goto out;
- mmu_notifier_range_init(&range, mm, addr, addr + PAGE_SIZE);
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, mm, addr,
+ addr + PAGE_SIZE);
mmu_notifier_invalidate_range_start(&range);
ptep = pte_offset_map_lock(mm, pmd, addr, &ptl);
--- a/mm/madvise.c~mm-mmu_notifier-contextual-information-for-event-triggering-invalidation-v2
+++ a/mm/madvise.c
@@ -472,7 +472,8 @@ static int madvise_free_single_vma(struc
range.end = min(vma->vm_end, end_addr);
if (range.end <= vma->vm_start)
return -EINVAL;
- mmu_notifier_range_init(&range, mm, range.start, range.end);
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, mm,
+ range.start, range.end);
lru_add_drain();
tlb_gather_mmu(&tlb, mm, range.start, range.end);
--- a/mm/memory.c~mm-mmu_notifier-contextual-information-for-event-triggering-invalidation-v2
+++ a/mm/memory.c
@@ -1010,7 +1010,8 @@ int copy_page_range(struct mm_struct *ds
is_cow = is_cow_mapping(vma->vm_flags);
if (is_cow) {
- mmu_notifier_range_init(&range, src_mm, addr, end);
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma,
+ src_mm, addr, end);
mmu_notifier_invalidate_range_start(&range);
}
@@ -1334,7 +1335,8 @@ void unmap_vmas(struct mmu_gather *tlb,
{
struct mmu_notifier_range range;
- mmu_notifier_range_init(&range, vma->vm_mm, start_addr, end_addr);
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, vma->vm_mm,
+ start_addr, end_addr);
mmu_notifier_invalidate_range_start(&range);
for ( ; vma && vma->vm_start < end_addr; vma = vma->vm_next)
unmap_single_vma(tlb, vma, start_addr, end_addr, NULL);
@@ -1356,7 +1358,8 @@ void zap_page_range(struct vm_area_struc
struct mmu_gather tlb;
lru_add_drain();
- mmu_notifier_range_init(&range, vma->vm_mm, start, start + size);
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, vma->vm_mm,
+ start, start + size);
tlb_gather_mmu(&tlb, vma->vm_mm, start, range.end);
update_hiwater_rss(vma->vm_mm);
mmu_notifier_invalidate_range_start(&range);
@@ -1382,7 +1385,8 @@ static void zap_page_range_single(struct
struct mmu_gather tlb;
lru_add_drain();
- mmu_notifier_range_init(&range, vma->vm_mm, address, address + size);
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, vma->vm_mm,
+ address, address + size);
tlb_gather_mmu(&tlb, vma->vm_mm, address, range.end);
update_hiwater_rss(vma->vm_mm);
mmu_notifier_invalidate_range_start(&range);
@@ -2279,7 +2283,8 @@ static vm_fault_t wp_page_copy(struct vm
__SetPageUptodate(new_page);
- mmu_notifier_range_init(&range, mm, vmf->address & PAGE_MASK,
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, mm,
+ vmf->address & PAGE_MASK,
(vmf->address & PAGE_MASK) + PAGE_SIZE);
mmu_notifier_invalidate_range_start(&range);
@@ -4104,8 +4109,9 @@ static int __follow_pte_pmd(struct mm_st
goto out;
if (range) {
- mmu_notifier_range_init(range, mm, address & PMD_MASK,
- (address & PMD_MASK) + PMD_SIZE);
+ mmu_notifier_range_init(range, MMU_NOTIFY_UNMAP, 0,
+ NULL, mm, address & PMD_MASK,
+ (address & PMD_MASK) + PMD_SIZE);
mmu_notifier_invalidate_range_start(range);
}
*ptlp = pmd_lock(mm, pmd);
@@ -4122,8 +4128,9 @@ static int __follow_pte_pmd(struct mm_st
goto out;
if (range) {
- mmu_notifier_range_init(range, mm, address & PAGE_MASK,
- (address & PAGE_MASK) + PAGE_SIZE);
+ mmu_notifier_range_init(range, MMU_NOTIFY_UNMAP, 0, NULL, mm,
+ address & PAGE_MASK,
+ (address & PAGE_MASK) + PAGE_SIZE);
mmu_notifier_invalidate_range_start(range);
}
ptep = pte_offset_map_lock(mm, pmd, address, ptlp);
--- a/mm/migrate.c~mm-mmu_notifier-contextual-information-for-event-triggering-invalidation-v2
+++ a/mm/migrate.c
@@ -2356,7 +2356,8 @@ static void migrate_vma_collect(struct m
mm_walk.mm = migrate->vma->vm_mm;
mm_walk.private = migrate;
- mmu_notifier_range_init(&range, mm_walk.mm, migrate->start,
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, NULL, mm_walk.mm,
+ migrate->start,
migrate->end);
mmu_notifier_invalidate_range_start(&range);
walk_page_range(migrate->start, migrate->end, &mm_walk);
@@ -2764,6 +2765,8 @@ static void migrate_vma_pages(struct mig
notified = true;
mmu_notifier_range_init(&range,
+ MMU_NOTIFY_UNMAP, 0,
+ NULL,
migrate->vma->vm_mm,
addr, migrate->end);
mmu_notifier_invalidate_range_start(&range);
--- a/mm/mprotect.c~mm-mmu_notifier-contextual-information-for-event-triggering-invalidation-v2
+++ a/mm/mprotect.c
@@ -185,7 +185,8 @@ static inline unsigned long change_pmd_r
/* invoke the mmu notifier if the pmd is populated */
if (!range.start) {
- mmu_notifier_range_init(&range, vma->vm_mm, addr, end);
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0,
+ vma, vma->vm_mm, addr, end);
mmu_notifier_invalidate_range_start(&range);
}
--- a/mm/mremap.c~mm-mmu_notifier-contextual-information-for-event-triggering-invalidation-v2
+++ a/mm/mremap.c
@@ -249,7 +249,8 @@ unsigned long move_page_tables(struct vm
old_end = old_addr + len;
flush_cache_range(vma, old_addr, old_end);
- mmu_notifier_range_init(&range, vma->vm_mm, old_addr, old_end);
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, vma->vm_mm,
+ old_addr, old_end);
mmu_notifier_invalidate_range_start(&range);
for (; old_addr < old_end; old_addr += extent, new_addr += extent) {
--- a/mm/oom_kill.c~mm-mmu_notifier-contextual-information-for-event-triggering-invalidation-v2
+++ a/mm/oom_kill.c
@@ -531,7 +531,8 @@ bool __oom_reap_task_mm(struct mm_struct
struct mmu_notifier_range range;
struct mmu_gather tlb;
- mmu_notifier_range_init(&range, mm, vma->vm_start,
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0,
+ vma, mm, vma->vm_start,
vma->vm_end);
tlb_gather_mmu(&tlb, mm, range.start, range.end);
if (mmu_notifier_invalidate_range_start_nonblock(&range)) {
--- a/mm/rmap.c~mm-mmu_notifier-contextual-information-for-event-triggering-invalidation-v2
+++ a/mm/rmap.c
@@ -896,7 +896,8 @@ static bool page_mkclean_one(struct page
* We have to assume the worse case ie pmd for invalidation. Note that
* the page can not be free from this function.
*/
- mmu_notifier_range_init(&range, vma->vm_mm, address,
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, vma->vm_mm,
+ address,
min(vma->vm_end, address +
(PAGE_SIZE << compound_order(page))));
mmu_notifier_invalidate_range_start(&range);
@@ -1371,7 +1372,8 @@ static bool try_to_unmap_one(struct page
* Note that the page can not be free in this function as call of
* try_to_unmap() must hold a reference on the page.
*/
- mmu_notifier_range_init(&range, vma->vm_mm, address,
+ mmu_notifier_range_init(&range, MMU_NOTIFY_UNMAP, 0, vma, vma->vm_mm,
+ address,
min(vma->vm_end, address +
(PAGE_SIZE << compound_order(page))));
if (PageHuge(page)) {
_
Patches currently in -mm which might be from jglisse@xxxxxxxxxx are
mm-hmm-select-mmu-notifier-when-selecting-hmm-v2.patch
mm-hmm-use-reference-counting-for-hmm-struct-v3.patch
mm-hmm-do-not-erase-snapshot-when-a-range-is-invalidated.patch
mm-hmm-improve-and-rename-hmm_vma_get_pfns-to-hmm_range_snapshot-v2.patch
mm-hmm-improve-and-rename-hmm_vma_fault-to-hmm_range_fault-v3.patch
mm-hmm-improve-driver-api-to-work-and-wait-over-a-range-v3.patch
mm-hmm-add-default-fault-flags-to-avoid-the-need-to-pre-fill-pfns-arrays-v2.patch
mm-hmm-mirror-hugetlbfs-snapshoting-faulting-and-dma-mapping-v3.patch
mm-hmm-allow-to-mirror-vma-of-a-file-on-a-dax-backed-filesystem-v3.patch
mm-hmm-add-helpers-to-test-if-mm-is-still-alive-or-not.patch
mm-hmm-add-an-helper-function-that-fault-pages-and-map-them-to-a-device-v3.patch
mm-hmm-add-an-helper-function-that-fault-pages-and-map-them-to-a-device-v3-fix.patch
mm-hmm-convert-various-hmm_pfn_-to-device_entry-which-is-a-better-name.patch
mm-mmu_notifier-helper-to-test-if-a-range-invalidation-is-blockable.patch
mm-mmu_notifier-convert-user-range-blockable-to-helper-function.patch
mm-mmu_notifier-convert-mmu_notifier_range-blockable-to-a-flags.patch
mm-mmu_notifier-contextual-information-for-event-enums.patch
mm-mmu_notifier-contextual-information-for-event-triggering-invalidation-v2.patch
mm-mmu_notifier-use-correct-mmu_notifier-events-for-each-invalidation.patch
mm-mmu_notifier-pass-down-vma-and-reasons-why-mmu-notifier-is-happening-v2.patch
mm-mmu_notifier-mmu_notifier_range_update_to_read_only-helper.patch
