Nitpick: s/callbac/callback and s/returning/return in subject
On 2014-09-11 16:19, Takashi Iwai wrote:
STAC/IDT driver calls snd_hda_jack_tbl_get() again after calling snd_hda_jack_detect_enable_callback(). For simplifying this, let's make snd_hda_jack_detect_enable_callback() returning the pointer while handling the error with the standard IS_ERR() & co.
Signed-off-by: Takashi Iwai tiwai@suse.de
sound/pci/hda/hda_jack.c | 20 +++++++++++++------- sound/pci/hda/hda_jack.h | 5 +++-- sound/pci/hda/patch_sigmatel.c | 14 ++++++-------- 3 files changed, 22 insertions(+), 17 deletions(-)
diff --git a/sound/pci/hda/hda_jack.c b/sound/pci/hda/hda_jack.c index 7f332794993f..27fccd2c8d41 100644 --- a/sound/pci/hda/hda_jack.c +++ b/sound/pci/hda/hda_jack.c @@ -215,28 +215,34 @@ EXPORT_SYMBOL_GPL(snd_hda_jack_detect_state); /**
- snd_hda_jack_detect_enable - enable the jack-detection
*/ -int snd_hda_jack_detect_enable_callback(struct hda_codec *codec, hda_nid_t nid,
hda_jack_callback cb)+struct hda_jack_tbl * +snd_hda_jack_detect_enable_callback(struct hda_codec *codec, hda_nid_t nid,
hda_jack_callback cb)
I have not seen the ERR_PTR usage before and it's not used commonly in the hda driver, so I think it's important to document that it can return a non-null pointer that is not a valid hda_jack_tbl. Looks dangerous otherwise.