[bug report] mt76: mt7615: wake device before accessing regmap in debugfs

[Lorenzo Bianconi]

>
> Hello Lorenzo Bianconi,
>
> The patch ea4906c4be49: "mt76: mt7615: wake device before accessing
> regmap in debugfs" from Jul 3, 2020, leads to the following static
> checker warning:

Hi Dan,

thx for the report. I will post a fix soon.

Regards,
Lorenzo

>
>         drivers/net/wireless/mediatek/mt76/mt7615/debugfs.c:179 mt7615_reset_test_set()
>         warn: inconsistent returns 'dev->mt76.mutex'.
>
> drivers/net/wireless/mediatek/mt76/mt7615/debugfs.c
>    159  static int
>    160  mt7615_reset_test_set(void *data, u64 val)
>    161  {
>    162          struct mt7615_dev *dev = data;
>    163          struct sk_buff *skb;
>    164
>    165          if (!mt7615_wait_for_mcu_init(dev))
>    166                  return 0;
>    167
>    168          mt7615_mutex_acquire(dev);
>    169
>    170          skb = alloc_skb(1, GFP_KERNEL);
>    171          if (!skb)
>    172                  return -ENOMEM;
>                         ^^^^^^^^^^^^^^
> This is returning with the mutex held.  What is the mutex protecting?
> We could move the allocation and the skb_put() before the
> mt7615_mutex_acquire().
>
>    173
>    174          skb_put(skb, 1);
>    175          mt76_tx_queue_skb_raw(dev, 0, skb, 0);
>    176
>    177          mt7615_mutex_release(dev);
>    178
>    179          return 0;
>    180  }
>
> regards,
> dan carpenter



-- 
UNIX is Sexy: who | grep -i blonde | talk; cd ~; wine; talk; touch;
unzip; touch; strip; gasp; finger; gasp; mount; fsck; more; yes; gasp;
umount; make clean; sleep