Percpu memory provides access via offsets into the percpu address space.
Offsets are essentially fixed for the lifetime of a chunk and therefore
require all users be good samaritans. If a user improperly handles the
lifetime of the percpu object, it can result in corruption in a couple
of ways:
 - immediate double free - breaks percpu metadata accounting
 - free after subsequent allocation
    - corruption due to multiple owner problem (either prior owner still
      writes or future allocation happens)
    - potential for oops if the percpu pages are reclaimed as the
      subsequent allocation isn't pinning the pages down
    - can lead to page->private pointers pointing to freed chunks

Sebastian noticed that if this happens, none of the memory debugging
facilities add additional information [2].

This patch aims to catch invalid free scenarios within valid chunks. To
better guard free_percpu(), we can either add a magic number or some
tracking facility to the percpu subsystem in a separate patch.

The invalid free check in pcpu_free_area() validates that the
allocation’s starting bit is set in both alloc_map and bound_map. The
alloc_map bit test ensures the area is allocated while the bound_map bit
test checks we are freeing from the beginning of an allocation.  We
choose not to check the validity of the offset as that is encoded in
page->private being a valid chunk.

pcpu_stats_area_dealloc() is moved later to only be on the happy path so
stats are only updated on valid frees.

This is a respin of [1] adding the requested changes from me and
Christoph.

[1] https://lore.kernel.org/linux-mm/20250904143514.Yk6Ap-jy@linutronix.de/
[2] https://lore.kernel.org/lkml/20260119074813.ecAFsGaT@linutronix.de/

Signed-off-by: Dennis Zhou <dennis@kernel.org>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
---
v3:
- Removed bit_off in bounds check because it's derived from the address
  passed in. If bit_off is bad, it's because we are getting a bad chunk
  from page->private. Let's do a better check on chunk validity in a 
  future patch probably behind a Kconfig.
- Removed ratelimit in favor of WARN_ON_ONCE(1).

 mm/percpu.c | 15 ++++++++++++++-
 1 file changed, 14 insertions(+), 1 deletion(-)

diff --git a/mm/percpu.c b/mm/percpu.c
index 81462ce5866e..a2107bdebf0b 100644
--- a/mm/percpu.c
+++ b/mm/percpu.c
@@ -1279,12 +1279,16 @@ static int pcpu_free_area(struct pcpu_chunk *chunk, int off)
 	int bit_off, bits, end, oslot, freed;
 
 	lockdep_assert_held(&pcpu_lock);
-	pcpu_stats_area_dealloc(chunk);
 
 	oslot = pcpu_chunk_slot(chunk);
 
 	bit_off = off / PCPU_MIN_ALLOC_SIZE;
 
+	/* check invalid free */
+	if (!test_bit(bit_off, chunk->alloc_map) ||
+	    !test_bit(bit_off, chunk->bound_map))
+		return 0;
+
 	/* find end index */
 	end = find_next_bit(chunk->bound_map, pcpu_chunk_map_bits(chunk),
 			    bit_off + 1);
@@ -1303,6 +1307,8 @@ static int pcpu_free_area(struct pcpu_chunk *chunk, int off)
 
 	pcpu_chunk_relocate(chunk, oslot);
 
+	pcpu_stats_area_dealloc(chunk);
+
 	return freed;
 }
 
@@ -2242,6 +2248,13 @@ void free_percpu(void __percpu *ptr)
 
 	spin_lock_irqsave(&pcpu_lock, flags);
 	size = pcpu_free_area(chunk, off);
+	if (size == 0) {
+		spin_unlock_irqrestore(&pcpu_lock, flags);
+
+		/* invalid percpu free */
+		WARN_ON_ONCE(1);
+		return;
+	}
 
 	pcpu_alloc_tag_free_hook(chunk, off, size);
 
-- 
2.43.0