linux-input.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH] HID: core: detect and skip invalid inputs to snto32()
@ 2020-12-17  1:12 Randy Dunlap
  2021-01-08 13:52 ` Jiri Kosina
  0 siblings, 1 reply; 2+ messages in thread
From: Randy Dunlap @ 2020-12-17  1:12 UTC (permalink / raw)
  To: linux-kernel
  Cc: Randy Dunlap, syzbot+1e911ad71dd4ea72e04a, Jiri Kosina,
	Benjamin Tissoires, linux-input

Prevent invalid (0, 0) inputs to hid-core's snto32() function.

Maybe it is just the dummy device here that is causing this, but
there are hundreds of calls to snto32(0, 0). Having n (bits count)
of 0 is causing the current UBSAN trap with a shift value of
0xffffffff (-1, or n - 1 in this function).

Either of the value to shift being 0 or the bits count being 0 can be
handled by just returning 0 to the caller, avoiding the following
complex shift + OR operations:

	return value & (1 << (n - 1)) ? value | (~0U << n) : value;

Fixes: dde5845a529f ("[PATCH] Generic HID layer - code split")
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Reported-by: syzbot+1e911ad71dd4ea72e04a@syzkaller.appspotmail.com
Cc: Jiri Kosina <jikos@kernel.org>
Cc: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Cc: linux-input@vger.kernel.org
---
 drivers/hid/hid-core.c |    3 +++
 1 file changed, 3 insertions(+)

--- lnx-510.orig/drivers/hid/hid-core.c
+++ lnx-510/drivers/hid/hid-core.c
@@ -1307,6 +1307,9 @@ EXPORT_SYMBOL_GPL(hid_open_report);
 
 static s32 snto32(__u32 value, unsigned n)
 {
+	if (!value || !n)
+		return 0;
+
 	switch (n) {
 	case 8:  return ((__s8)value);
 	case 16: return ((__s16)value);

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [PATCH] HID: core: detect and skip invalid inputs to snto32()
  2020-12-17  1:12 [PATCH] HID: core: detect and skip invalid inputs to snto32() Randy Dunlap
@ 2021-01-08 13:52 ` Jiri Kosina
  0 siblings, 0 replies; 2+ messages in thread
From: Jiri Kosina @ 2021-01-08 13:52 UTC (permalink / raw)
  To: Randy Dunlap
  Cc: linux-kernel, syzbot+1e911ad71dd4ea72e04a, Benjamin Tissoires,
	linux-input

On Wed, 16 Dec 2020, Randy Dunlap wrote:

> Prevent invalid (0, 0) inputs to hid-core's snto32() function.
> 
> Maybe it is just the dummy device here that is causing this, but
> there are hundreds of calls to snto32(0, 0). Having n (bits count)
> of 0 is causing the current UBSAN trap with a shift value of
> 0xffffffff (-1, or n - 1 in this function).
> 
> Either of the value to shift being 0 or the bits count being 0 can be
> handled by just returning 0 to the caller, avoiding the following
> complex shift + OR operations:
> 
> 	return value & (1 << (n - 1)) ? value | (~0U << n) : value;
> 
> Fixes: dde5845a529f ("[PATCH] Generic HID layer - code split")
> Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
> Reported-by: syzbot+1e911ad71dd4ea72e04a@syzkaller.appspotmail.com
> Cc: Jiri Kosina <jikos@kernel.org>
> Cc: Benjamin Tissoires <benjamin.tissoires@redhat.com>
> Cc: linux-input@vger.kernel.org
> ---
>  drivers/hid/hid-core.c |    3 +++
>  1 file changed, 3 insertions(+)
> 
> --- lnx-510.orig/drivers/hid/hid-core.c
> +++ lnx-510/drivers/hid/hid-core.c
> @@ -1307,6 +1307,9 @@ EXPORT_SYMBOL_GPL(hid_open_report);
>  
>  static s32 snto32(__u32 value, unsigned n)
>  {
> +	if (!value || !n)
> +		return 0;
> +

Given the fact that this has been in the code basically since ever, we're 
probably fine, but it's good to have this fixed nevertheless. Applied 
conservatively to hid.git#for-5.12/core.

Thanks,

-- 
Jiri Kosina
SUSE Labs


^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2021-01-08 13:52 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-12-17  1:12 [PATCH] HID: core: detect and skip invalid inputs to snto32() Randy Dunlap
2021-01-08 13:52 ` Jiri Kosina

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).