[PATCH] HID: core: do not reject devices when they declare too many usages

[PATCH] HID: core: do not reject devices when they declare too many usages

[Benjamin Tissoires]

Some device present proprietary collections with a usage min of 0x00 and
a usage max of 0xffff. hid-core currently reject them while most of the
time this is harmless.
Let's ignore the exceeding usages, and hope for the best.

Reported-by: Simon Wörner <mail@simon-woerner.de>
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
---
 drivers/hid/hid-core.c | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)

diff --git a/drivers/hid/hid-core.c b/drivers/hid/hid-core.c
index 157c627..b403fe2 100644
--- a/drivers/hid/hid-core.c
+++ b/drivers/hid/hid-core.c
@@ -427,6 +427,7 @@ static int hid_parser_local(struct hid_parser *parser, struct hid_item *item)
 {
 	__u32 data;
 	unsigned n;
+	__u32 count;
 
 	data = item_udata(item);
 
@@ -490,6 +491,24 @@ static int hid_parser_local(struct hid_parser *parser, struct hid_item *item)
 		if (item->size <= 2)
 			data = (parser->global.usage_page << 16) + data;
 
+		count = data - parser->local.usage_minimum;
+		if (count + parser->local.usage_index >= HID_MAX_USAGES) {
+			/*
+			 * We do not warn if the name is not set, we are
+			 * actually pre-scanning the device.
+			 */
+			if (dev_name(&parser->device->dev))
+				hid_warn(parser->device,
+					 "ignoring exceeding usage max\n");
+			data = HID_MAX_USAGES - parser->local.usage_index +
+				parser->local.usage_minimum - 1;
+			if (data <= 0) {
+				hid_err(parser->device,
+					"no more usage index available\n");
+				return -1;
+			}
+		}
+
 		for (n = parser->local.usage_minimum; n <= data; n++)
 			if (hid_add_usage(parser, n)) {
 				dbg_hid("hid_add_usage failed\n");
-- 
2.4.3

Re: [PATCH] HID: core: do not reject devices when they declare too many usages

[Jiri Kosina]

On Fri, 24 Jul 2015, Benjamin Tissoires wrote:

> Some device present proprietary collections with a usage min of 0x00 and
> a usage max of 0xffff. hid-core currently reject them while most of the
> time this is harmless.
> Let's ignore the exceeding usages, and hope for the best.

Okay, why not, this should be really harmless.

Applied to for-4.3/upstream.

Thanks,

-- 
Jiri Kosina
SUSE Labs