media: pvrusb2: fix array-index-out-of-bounds in pvr2_i2c_core_init

[ Upstream commit 471bec6 ]

Syzbot reported that -1 is used as array index. The problem was in
missing validation check.

hdw->unit_number is initialized with -1 and then if init table walk fails
this value remains unchanged. Since code blindly uses this member for
array indexing adding sanity check is the easiest fix for that.

hdw->workpoll initialization moved upper to prevent warning in
__flush_work.

Reported-and-tested-by: syzbot+1a247e36149ffd709a9b@syzkaller.appspotmail.com

Fixes: d855497 ("V4L/DVB (4228a): pvrusb2 to kernel 2.6.18")
Signed-off-by: Pavel Skripkin <paskripkin@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drivers/media/usb/pvrusb2/pvrusb2-hdw.c

@@ -2569,6 +2569,11 @@ struct pvr2_hdw *pvr2_hdw_create(struct usb_interface *intf,
 	} while (0);
 	mutex_unlock(&pvr2_unit_mtx);
 
+	INIT_WORK(&hdw->workpoll, pvr2_hdw_worker_poll);
+
+	if (hdw->unit_number == -1)
+		goto fail;
+
 	cnt1 = 0;
 	cnt2 = scnprintf(hdw->name+cnt1,sizeof(hdw->name)-cnt1,"pvrusb2");
 	cnt1 += cnt2;
@@ -2580,8 +2585,6 @@ struct pvr2_hdw *pvr2_hdw_create(struct usb_interface *intf,
 	if (cnt1 >= sizeof(hdw->name)) cnt1 = sizeof(hdw->name)-1;
 	hdw->name[cnt1] = 0;
 
-	INIT_WORK(&hdw->workpoll,pvr2_hdw_worker_poll);
-
 	pvr2_trace(PVR2_TRACE_INIT,"Driver unit number is %d, name is %s",
 		   hdw->unit_number,hdw->name);