Re: [PATCH mm-unstable v4 5/5] mm/khugepaged: unify khugepaged and madv_collapse with collapse_single_pmd()

[Andrew Morton]

On Tue, 7 Apr 2026 09:38:19 +0100 "Lorenzo Stoakes (Oracle)" <ljs@xxxxxxxxxx> wrote:

> On Wed, Apr 01, 2026 at 01:31:45PM -0700, Andrew Morton wrote:
> > On Wed, 1 Apr 2026 09:14:35 +0100 "Lorenzo Stoakes (Oracle)" <ljs@xxxxxxxxxx> wrote:
> >
> > > From: "Lorenzo Stoakes (Oracle)" <ljs@xxxxxxxxxx>
> > > Date: Tue, 31 Mar 2026 13:11:18 +0100
> > > Subject: [PATCH] mm/khugepaged: fix issue with tracking lock
> > >
> > > We are incorrectly treating lock_dropped to track both whether the lock is
> > > currently held and whether or not the lock was ever dropped.
> > >
> > > Update this change to account for this.
> > >
> > > Signed-off-by: Lorenzo Stoakes (Oracle) <ljs@xxxxxxxxxx>
> >
> > Not sure if I should add this?
> >
> > Should it have
> >
> > Fixes: 330f3758a3bc ("mm/khugepaged: unify khugepaged and madv_collapse
> > with collapse_single_pmd()")?
> 
> (Sorry only just noticed this)
> 
> I mean we do need this patch, if you plan to do this as a fix follow up then
> should have a fixed tag, but I do worry about bisection hazard by leaving this
> as-is.
> 
> We definitely either need this patch folded in or as a fix that gets sent
> alongside the change.

I added this to mm-stable a couple of days ago as 5a62019807da
("mm/khugepaged: fix issue with tracking lock"), below.  It does have
the Fixes: so I think we're good.


commit 5a62019807da4a7add0717c962ac83b23dd12b2c
Author: Lorenzo Stoakes (Oracle) <ljs@xxxxxxxxxx>
Date:   Tue Mar 31 13:11:18 2026 +0100

    mm/khugepaged: fix issue with tracking lock
    
    We are incorrectly treating lock_dropped to track both whether the lock is
    currently held and whether or not the lock was ever dropped.
    
    Update this change to account for this.
    
    Link: https://lkml.kernel.org/r/7760c811-e100-4d40-9217-0813c28314be@lucifer.local
    Fixes: 330f3758a3bc ("mm/khugepaged: unify khugepaged and madv_collapse with collapse_single_pmd()")
    Signed-off-by: Lorenzo Stoakes (Oracle) <ljs@xxxxxxxxxx>
    Reviewed-by: Lance Yang <lance.yang@xxxxxxxxx>
    Reviewed-by: Nico Pache <npache@xxxxxxxxxx>
    Cc: Alistair Popple <apopple@xxxxxxxxxx>
    Cc: Andrea Arcangeli <aarcange@xxxxxxxxxx>
    Cc: Anshuman Khandual <anshuman.khandual@xxxxxxx>
    Cc: Baolin Wang <baolin.wang@xxxxxxxxxxxxxxxxx>
    Cc: Barry Song <baohua@xxxxxxxxxx>
    Cc: Brendan Jackman <jackmanb@xxxxxxxxxx>
    Cc: Byungchul Park <byungchul@xxxxxx>
    Cc: Catalin Marinas <catalin.marinas@xxxxxxx>
    Cc: David Hildenbrand <david@xxxxxxxxxx>
    Cc: David Rientjes <rientjes@xxxxxxxxxx>
    Cc: Dev Jain <dev.jain@xxxxxxx>
    Cc: Gregory Price <gourry@xxxxxxxxxx>
    Cc: "Huang, Ying" <ying.huang@xxxxxxxxxxxxxxxxx>
    Cc: Hugh Dickins <hughd@xxxxxxxxxx>
    Cc: Jan Kara <jack@xxxxxxx>
    Cc: Jann Horn <jannh@xxxxxxxxxx>
    Cc: Johannes Weiner <hannes@xxxxxxxxxxx>
    Cc: Jonathan Corbet <corbet@xxxxxxx>
    Cc: Joshua Hahn <joshua.hahnjy@xxxxxxxxx>
    Cc: Kefeng Wang <wangkefeng.wang@xxxxxxxxxx>
    Cc: Liam Howlett <liam.howlett@xxxxxxxxxx>
    Cc: "Masami Hiramatsu (Google)" <mhiramat@xxxxxxxxxx>
    Cc: Mathieu Desnoyers <mathieu.desnoyers@xxxxxxxxxxxx>
    Cc: Matthew Brost <matthew.brost@xxxxxxxxx>
    Cc: Matthew Wilcox (Oracle) <willy@xxxxxxxxxxxxx>
    Cc: Michal Hocko <mhocko@xxxxxxxx>
    Cc: Mike Rapoport <rppt@xxxxxxxxxx>
    Cc: Nanyong Sun <sunnanyong@xxxxxxxxxx>
    Cc: Pedro Falcato <pfalcato@xxxxxxx>
    Cc: Peter Xu <peterx@xxxxxxxxxx>
    Cc: Rafael Aquini <raquini@xxxxxxxxxx>
    Cc: Rakie Kim <rakie.kim@xxxxxx>
    Cc: Randy Dunlap <rdunlap@xxxxxxxxxxxxx>
    Cc: Ryan Roberts <ryan.roberts@xxxxxxx>
    Cc: Shivank Garg <shivankg@xxxxxxx>
    Cc: Steven Rostedt <rostedt@xxxxxxxxxxx>
    Cc: Suren Baghdasaryan <surenb@xxxxxxxxxx>
    Cc: Takashi Iwai (SUSE) <tiwai@xxxxxxx>
    Cc: Thomas Hellström <thomas.hellstrom@xxxxxxxxxxxxxxx>
    Cc: Usama Arif <usamaarif642@xxxxxxxxx>
    Cc: Vishal Moola (Oracle) <vishal.moola@xxxxxxxxx>
    Cc: Vlastimil Babka <vbabka@xxxxxxx>
    Cc: Wei Yang <richard.weiyang@xxxxxxxxx>
    Cc: Will Deacon <will@xxxxxxxxxx>
    Cc: Yang Shi <yang@xxxxxxxxxxxxxxxxxxxxxx>
    Cc: Zach O'Keefe <zokeefe@xxxxxxxxxx>
    Cc: Zi Yan <ziy@xxxxxxxxxx>
    Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>

diff --git a/mm/khugepaged.c b/mm/khugepaged.c
index d21348b85a59..b8452dbdb043 100644
--- a/mm/khugepaged.c
+++ b/mm/khugepaged.c
@@ -2828,6 +2828,7 @@ int madvise_collapse(struct vm_area_struct *vma, unsigned long start,
 	unsigned long hstart, hend, addr;
 	enum scan_result last_fail = SCAN_FAIL;
 	int thps = 0;
+	bool mmap_unlocked = false;
 
 	BUG_ON(vma->vm_start > start);
 	BUG_ON(vma->vm_end < end);
@@ -2850,10 +2851,11 @@ int madvise_collapse(struct vm_area_struct *vma, unsigned long start,
 	for (addr = hstart; addr < hend; addr += HPAGE_PMD_SIZE) {
 		enum scan_result result = SCAN_FAIL;
 
-		if (*lock_dropped) {
+		if (mmap_unlocked) {
 			cond_resched();
 			mmap_read_lock(mm);
-			*lock_dropped = false;
+			mmap_unlocked = false;
+			*lock_dropped = true;
 			result = hugepage_vma_revalidate(mm, addr, false, &vma,
 							 cc);
 			if (result  != SCAN_SUCCEED) {
@@ -2864,7 +2866,7 @@ int madvise_collapse(struct vm_area_struct *vma, unsigned long start,
 			hend = min(hend, vma->vm_end & HPAGE_PMD_MASK);
 		}
 
-		result = collapse_single_pmd(addr, vma, lock_dropped, cc);
+		result = collapse_single_pmd(addr, vma, &mmap_unlocked, cc);
 
 		switch (result) {
 		case SCAN_SUCCEED:
@@ -2893,8 +2895,10 @@ int madvise_collapse(struct vm_area_struct *vma, unsigned long start,
 
 out_maybelock:
 	/* Caller expects us to hold mmap_lock on return */
-	if (*lock_dropped)
+	if (mmap_unlocked) {
+		*lock_dropped = true;
 		mmap_read_lock(mm);
+	}
 out_nolock:
 	mmap_assert_locked(mm);
 	mmdrop(mm);