6.6-stable review patch. If anyone has any objections, please let me know. ------------------ From: Andrew Ballance <andrewjballance@xxxxxxxxx> commit 89e1ee118d6f0ee6bd6e80d8fe08839875daa241 upstream. syzbot reported a potential read out of bounds in asus_report_fixup. this patch adds checks so that a read out of bounds will not occur Signed-off-by: Andrew Ballance <andrewjballance@xxxxxxxxx> Reported-by: <syzbot+07762f019fd03d01f04c@xxxxxxxxxxxxxxxxxxxxxxxxx> Closes: https://syzkaller.appspot.com/bug?extid=07762f019fd03d01f04c Fixes: 59d2f5b7392e ("HID: asus: fix more n-key report descriptors if n-key quirked") Link: https://lore.kernel.org/r/20240602085023.1720492-1-andrewjballance@xxxxxxxxx Signed-off-by: Benjamin Tissoires <bentiss@xxxxxxxxxx> Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> --- drivers/hid/hid-asus.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) --- a/drivers/hid/hid-asus.c +++ b/drivers/hid/hid-asus.c @@ -1235,8 +1235,8 @@ static __u8 *asus_report_fixup(struct hi } /* match many more n-key devices */ - if (drvdata->quirks & QUIRK_ROG_NKEY_KEYBOARD) { - for (int i = 0; i < *rsize + 1; i++) { + if (drvdata->quirks & QUIRK_ROG_NKEY_KEYBOARD && *rsize > 15) { + for (int i = 0; i < *rsize - 15; i++) { /* offset to the count from 0x5a report part always 14 */ if (rdesc[i] == 0x85 && rdesc[i + 1] == 0x5a && rdesc[i + 14] == 0x95 && rdesc[i + 15] == 0x05) {
