[PATCH] drivers/mt: restrict length of kzalloc...

[PATCH] drivers/mt: restrict length of kzalloc in input_mt_init_slots()

From: Sabyrzhan Tasbolatov <hidden>
Date: 2021-02-02 12:10:06

syzbot found WARNING in input_mt_init_slots [1] when
struct_size(mt, slots, num_slots)=0x40006 where num_slots=0x10001,
which exceeds KMALLOC_MAX_SIZE (0x40000) and causes
order >= MAX_ORDER condition.

[1]
Call Trace:
 alloc_pages_current+0x18c/0x2a0 mm/mempolicy.c:2267
 alloc_pages include/linux/gfp.h:547 [inline]
 kmalloc_order+0x2e/0xb0 mm/slab_common.c:837
 kmalloc_order_trace+0x14/0x120 mm/slab_common.c:853
 kmalloc include/linux/slab.h:557 [inline]
 kzalloc include/linux/slab.h:682 [inline]
 input_mt_init_slots drivers/input/input-mt.c:49 [inline]

Reported-by: syzbot+0122fa359a69694395d5@syzkaller.appspotmail.com
Signed-off-by: Sabyrzhan Tasbolatov <redacted>
---
 drivers/input/input-mt.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/input/input-mt.c b/drivers/input/input-mt.c
index 44fe6f2f063c..e542f45a45ab 100644
--- a/drivers/input/input-mt.c
+++ b/drivers/input/input-mt.c

@@ -40,13 +40,18 @@ int input_mt_init_slots(struct input_dev *dev, unsigned int num_slots,
 {
 	struct input_mt *mt = dev->mt;
 	int i;
+	size_t mt_size = 0;
 
 	if (!num_slots)
 		return 0;
 	if (mt)
 		return mt->num_slots != num_slots ? -EINVAL : 0;
 
-	mt = kzalloc(struct_size(mt, slots, num_slots), GFP_KERNEL);
+	mt_size = struct_size(mt, slots, num_slots);
+	if (mt_size > KMALLOC_MAX_SIZE)
+		return -ENOMEM;
+
+	mt = kzalloc(mt_size, GFP_KERNEL);
 	if (!mt)
 		goto err_mem;

-- 
2.25.1

Re: [PATCH] drivers/mt: restrict num_slots in input_mt_init_slots()

From: Sabyrzhan Tasbolatov <hidden>
Date: 2021-02-26 08:43:58

On Tue,  2 Feb 2021 18:08:07 +0600, Sabyrzhan Tasbolatov wrote:

quoted hunk

syzbot found WARNING in input_mt_init_slots [1] when
struct_size(mt, slots, num_slots)=0x40006 where num_slots=0x10001,
which exceeds KMALLOC_MAX_SIZE (0x40000) and causes
order >= MAX_ORDER condition.

[1]
Call Trace:
 alloc_pages_current+0x18c/0x2a0 mm/mempolicy.c:2267
 alloc_pages include/linux/gfp.h:547 [inline]
 kmalloc_order+0x2e/0xb0 mm/slab_common.c:837
 kmalloc_order_trace+0x14/0x120 mm/slab_common.c:853
 kmalloc include/linux/slab.h:557 [inline]
 kzalloc include/linux/slab.h:682 [inline]
 input_mt_init_slots drivers/input/input-mt.c:49 [inline]

Reported-by: syzbot+0122fa359a69694395d5@syzkaller.appspotmail.com
Signed-off-by: Sabyrzhan Tasbolatov <redacted>
---
 drivers/input/input-mt.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/input/input-mt.c b/drivers/input/input-mt.c
index 44fe6f2f063c..e542f45a45ab 100644
--- a/drivers/input/input-mt.c
+++ b/drivers/input/input-mt.c

@@ -40,13 +40,18 @@ int input_mt_init_slots(struct input_dev *dev, unsigned int num_slots,
 {
 	struct input_mt *mt = dev->mt;
 	int i;
+	size_t mt_size = 0;
 
 	if (!num_slots)
 		return 0;
 	if (mt)
 		return mt->num_slots != num_slots ? -EINVAL : 0;
 
-	mt = kzalloc(struct_size(mt, slots, num_slots), GFP_KERNEL);
+	mt_size = struct_size(mt, slots, num_slots);
+	if (mt_size > KMALLOC_MAX_SIZE)
+		return -ENOMEM;
+
+	mt = kzalloc(mt_size, GFP_KERNEL);
 	if (!mt)
 		goto err_mem;

--

Following-up. I've also just found out that in this function, there is another
allocation with num_slots length:

int input_mt_init_slots(..)
{
	..
	if (flags & INPUT_MT_TRACK) {
		unsigned int n2 = num_slots * num_slots;
		mt->red = kcalloc(n2, sizeof(*mt->red), GFP_KERNEL);
	..

I've checked HID vendors' xrefs for input_mt_init_slots(), most of them
pass >= 5 value to num_slots parameter. So either we should choose some
optimal limit of num_slots or just restrict it with big KMALLOC_MAX_SIZE.

Comments?

`h`	back out one level
`j`	next message in thread
`k`	previous message in thread
`l`	drill in
`Esc`	close help / fold thread tree
`?`	toggle this help