ALSA: hda - Use generic parser for Cirrus codec driver
Subsystems (2)
Hello Takashi Iwai,
The patch 1077a024812d: "ALSA: hda - Use generic parser for Cirrus codec driver" from Dec 19, 2012, leads to the following static checker warning:
sound/pci/hda/patch_cirrus.c:1004 cs4210_spdif_automute() warn: we tested 'spdif_present' before and it was 'true'
sound/pci/hda/patch_cirrus.c
984 static void cs4210_spdif_automute(struct hda_codec *codec,
985 struct hda_jack_callback *tbl)
986 {
987 struct cs_spec *spec = codec->spec;
988 bool spdif_present = false;
989 hda_nid_t spdif_pin = spec->gen.autocfg.dig_out_pins[0];
990
991 /* detect on spdif is specific to CS4210 */
992 if (!spec->spdif_detect ||
993 spec->vendor_nid != CS4210_VENDOR_NID)
994 return;
995
996 spdif_present = snd_hda_jack_detect(codec, spdif_pin);
997 if (spdif_present == spec->spdif_present)
998 return;
999
1000 spec->spdif_present = spdif_present;
1001 /* SPDIF TX on/off */
1002 if (spdif_present)
^^^^^^^^^^^^^
1003 snd_hda_set_pin_ctl(codec, spdif_pin,
1004 spdif_present ? PIN_OUT : 0);
^^^^^^^^^^^^^I'm not positive but I think this is buggy. We should remove the if statement.
1005 1006 cs_automute(codec); 1007 }
regards,
dan carpenter
Quoted 35 lines · Dan Carpenter
Hello Takashi Iwai,
The patch
1077a024812d: "ALSA: hda - Use generic parser for Cirrus codec driver" from Dec 19, 2012, leads to the following static checker warning:sound/pci/hda/patch_cirrus.c:1004 cs4210_spdif_automute() warn: we tested 'spdif_present' before and it was 'true'
sound/pci/hda/patch_cirrus.c
984 static void cs4210_spdif_automute(structhda_codec*codec,
985 structhda_jack_callback*tbl)986 { 987 struct cs_spec *spec = codec->spec; 988 bool spdif_present = false; 989 hda_nid_t spdif_pin = spec->gen.autocfg.dig_out_pins[0]; 990 991 /* detect on spdif is specific to CS4210 */ 992 if (!spec->spdif_detect || 993 spec->vendor_nid != CS4210_VENDOR_NID) 994 return; 995 996 spdif_present = snd_hda_jack_detect(codec, spdif_pin); 997 if (spdif_present == spec->spdif_present) 998 return; 999 1000 spec->spdif_present = spdif_present; 1001 /* SPDIF TX on/off */ 1002 if (spdif_present) ^^^^^^^^^^^^^ 1003 snd_hda_set_pin_ctl(codec, spdif_pin, 1004 spdif_present ? PIN_OUT : 0); ^^^^^^^^^^^^^
Quoted 2 lines
I'm not positive but I think this is buggy. We should remove the if statement.
Yes, looks so. Could you submit a fix patch?
Thanks!
Takashi
PATCH 1/1 ALSA: hda - fix cs4210_spdif_automute()
Smatch complains that we have nested checks for "spdif_present". It turns out the current behavior isn't correct, we should remove the first check and keep the second.
sound/pci/hda/patch_cirrus.c +1 -3
@@ -999,9 +999,7 @@ static void cs4210_spdif_automute(struct hda_codec *codec, spec->spdif_present = spdif_present;/* SPDIF TX on/off */ -if (spdif_present) -snd_hda_set_pin_ctl (codec, spdif_pin, - spdif_present ?PIN_OUT :0 ); +snd_hda_set_pin_ctl (codec, spdif_pin, spdif_present ?PIN_OUT :0 );cs_automute (codec); }
Takashi Iwai
AppliedRe: [patch] ALSA: hda - fix cs4210_spdif_automute()
Quoted 7 lines · Dan Carpenter
Smatch complains that we have nested checks for "
spdif_present". It turns out the current behavior isn't correct, we should remove the first check and keep the second.Fixes 1077a024812d ('ALSA: hda - Use generic parser for Cirrus codec driver')Signed-off-by Dan Carpenter <[email protected]>
Applied, thanks.
Takashi
Quoted 17 lines
sound/pci/hda/patch_cirrus.c +1
-3@@ -999,9 +999,7 @@ static void cs4210_spdif_automute(struct hda_codec *codec, spec->spdif_present = spdif_present;/* SPDIF TX on/off */ -if (spdif_present) -snd_hda_set_pin_ctl (codec, spdif_pin, - spdif_present ?PIN_OUT :0 ); +snd_hda_set_pin_ctl (codec, spdif_pin, spdif_present ?PIN_OUT :0 );cs_automute (codec); }