uhid_maxrepid() is currently not able to observe all report IDs as it

passes hid_none as the wanted hid kind to hid_start_parse(). This
becomes a problem for descriptors with multiple report IDs per
collection. Therefore change the semantics of the hid_none kind to let
it represents all possible kinds.

Gets rid of "uhidev_intr: bad repid" console spam as reported by sthen@
on bugs@.

Have been in snaps for a week.

diff --git a/sys/dev/hid/hid.c b/sys/dev/hid/hid.c
index 1c4d5fa..8611d92 100644 (file)
--- a/sys/dev/hid/hid.c
+++ b/sys/dev/hid/hid.c
@@ -1,4 +1,4 @@
-/*     $OpenBSD: hid.c,v 1.3 2020/06/04 23:03:43 deraadt Exp $ */
+/*     $OpenBSD: hid.c,v 1.4 2022/05/20 05:02:47 anton Exp $ */
 /*     $NetBSD: hid.c,v 1.23 2002/07/11 21:14:25 augustss Exp $        */
 /*     $FreeBSD: src/sys/dev/usb/hid.c,v 1.11 1999/11/17 22:33:39 n_hibma Exp $ */
 
@@ -229,7 +229,7 @@ hid_get_item(struct hid_data *s, struct hid_item *h)
                 * Only copy HID item, increment position and return
                 * if correct kind!
                 */
-               if (s->kind == c->kind) {
+               if (s->kind == hid_none || s->kind == c->kind) {
                        *h = *c;
                        DPRINTF("%u,%u,%u\n", h->loc.pos,
                            h->loc.size, h->loc.count);