[alsa-devel] [PATCH 1/5] ASoC: Blackfin AC97: add a few missing multichannel define handling
From: Cliff Cai cliff.cai@analog.com
Somewhere along the line, most of SND_BF5XX_MULTICHAN_SUPPORT handling was merged, but two places were missed (the probe/resume functions). Restore handling of this option so it gets initialized properly.
Signed-off-by: Cliff Cai cliff.cai@analog.com Signed-off-by: Mike Frysinger vapier@gentoo.org --- sound/soc/blackfin/bf5xx-ac97.c | 8 ++++++++ 1 files changed, 8 insertions(+), 0 deletions(-)
diff --git a/sound/soc/blackfin/bf5xx-ac97.c b/sound/soc/blackfin/bf5xx-ac97.c index 2758b90..e693229 100644 --- a/sound/soc/blackfin/bf5xx-ac97.c +++ b/sound/soc/blackfin/bf5xx-ac97.c @@ -277,7 +277,11 @@ static int bf5xx_ac97_resume(struct snd_soc_dai *dai) if (!dai->active) return 0;
+#if defined(CONFIG_SND_BF5XX_MULTICHAN_SUPPORT) + ret = sport_set_multichannel(sport, 16, 0x3FF, 1); +#else ret = sport_set_multichannel(sport, 16, 0x1F, 1); +#endif if (ret) { pr_err("SPORT is busy!\n"); return -EBUSY; @@ -334,7 +338,11 @@ static int bf5xx_ac97_probe(struct platform_device *pdev, goto sport_err; } /*SPORT works in TDM mode to simulate AC97 transfers*/ +#if defined(CONFIG_SND_BF5XX_MULTICHAN_SUPPORT) + ret = sport_set_multichannel(sport_handle, 16, 0x3FF, 1); +#else ret = sport_set_multichannel(sport_handle, 16, 0x1F, 1); +#endif if (ret) { pr_err("SPORT is busy!\n"); ret = -EBUSY;
Signed-off-by: Mike Frysinger vapier@gentoo.org --- sound/soc/blackfin/bf5xx-sport.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/sound/soc/blackfin/bf5xx-sport.c b/sound/soc/blackfin/bf5xx-sport.c index 469ce7f..99051ff 100644 --- a/sound/soc/blackfin/bf5xx-sport.c +++ b/sound/soc/blackfin/bf5xx-sport.c @@ -326,7 +326,7 @@ static inline int sport_hook_tx_dummy(struct sport_device *sport)
int sport_tx_start(struct sport_device *sport) { - unsigned flags; + unsigned long flags; pr_debug("%s: tx_run:%d, rx_run:%d\n", __func__, sport->tx_run, sport->rx_run); if (sport->tx_run)
From: Barry Song barry.song@analog.com
Signed-off-by: Barry Song barry.song@analog.com Signed-off-by: Mike Frysinger vapier@gentoo.org --- sound/soc/blackfin/bf5xx-ac97.h | 2 +- sound/soc/blackfin/bf5xx-i2s.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/sound/soc/blackfin/bf5xx-ac97.h b/sound/soc/blackfin/bf5xx-ac97.h index 3f2a911..a1f97dd 100644 --- a/sound/soc/blackfin/bf5xx-ac97.h +++ b/sound/soc/blackfin/bf5xx-ac97.h @@ -1,5 +1,5 @@ /* - * linux/sound/arm/bf5xx-ac97.h + * sound/soc/blackfin/bf5xx-ac97.h * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as diff --git a/sound/soc/blackfin/bf5xx-i2s.h b/sound/soc/blackfin/bf5xx-i2s.h index 7107d1a..264ecdc 100644 --- a/sound/soc/blackfin/bf5xx-i2s.h +++ b/sound/soc/blackfin/bf5xx-i2s.h @@ -1,5 +1,5 @@ /* - * linux/sound/arm/bf5xx-i2s.h + * sound/soc/blackfin/bf5xx-i2s.h * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as
From: Barry Song barry.song@analog.com
Commit dc7d7b830ee1 trimmed the platform_device parameter from all of the suspend functions, but it also accidentally removed it from the resume function in the Blackfin I2S driver. So restore it.
Signed-off-by: Barry Song barry.song@analog.com Signed-off-by: Mike Frysinger vapier@gentoo.org --- sound/soc/blackfin/bf5xx-i2s.c | 3 ++- 1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/sound/soc/blackfin/bf5xx-i2s.c b/sound/soc/blackfin/bf5xx-i2s.c index 876abad..19539c6 100644 --- a/sound/soc/blackfin/bf5xx-i2s.c +++ b/sound/soc/blackfin/bf5xx-i2s.c @@ -227,7 +227,8 @@ static int bf5xx_i2s_probe(struct platform_device *pdev, return 0; }
-static void bf5xx_i2s_remove(struct snd_soc_dai *dai) +static void bf5xx_i2s_remove(struct platform_device *pdev, + struct snd_soc_dai *dai) { pr_debug("%s enter\n", __func__); peripheral_free_list(&sport_req[sport_num][0]);
From: Cliff Cai cliff.cai@analog.com
If the sound system hasn't been utilized yet and we suspend, then we attempt to save/restore using state that doesn't exist. So use a global handle instead to reconfigure properly.
Signed-off-by: Cliff Cai cliff.cai@analog.com Signed-off-by: Mike Frysinger vapier@gentoo.org --- sound/soc/blackfin/bf5xx-i2s.c | 19 +++++++------------ 1 files changed, 7 insertions(+), 12 deletions(-)
diff --git a/sound/soc/blackfin/bf5xx-i2s.c b/sound/soc/blackfin/bf5xx-i2s.c index 19539c6..1e9d161 100644 --- a/sound/soc/blackfin/bf5xx-i2s.c +++ b/sound/soc/blackfin/bf5xx-i2s.c @@ -237,36 +237,31 @@ static void bf5xx_i2s_remove(struct platform_device *pdev, #ifdef CONFIG_PM static int bf5xx_i2s_suspend(struct snd_soc_dai *dai) { - struct sport_device *sport = - (struct sport_device *)dai->private_data;
pr_debug("%s : sport %d\n", __func__, dai->id); - if (!dai->active) - return 0; + if (dai->capture.active) - sport_rx_stop(sport); + sport_rx_stop(sport_handle); if (dai->playback.active) - sport_tx_stop(sport); + sport_tx_stop(sport_handle); return 0; }
static int bf5xx_i2s_resume(struct snd_soc_dai *dai) { int ret; - struct sport_device *sport = - (struct sport_device *)dai->private_data;
pr_debug("%s : sport %d\n", __func__, dai->id); - if (!dai->active) - return 0;
- ret = sport_config_rx(sport, RFSR | RCKFE, RSFSE|0x1f, 0, 0); + ret = sport_config_rx(sport_handle, bf5xx_i2s.rcr1, + bf5xx_i2s.rcr2, 0, 0); if (ret) { pr_err("SPORT is busy!\n"); return -EBUSY; }
- ret = sport_config_tx(sport, TFSR | TCKFE, TSFSE|0x1f, 0, 0); + ret = sport_config_tx(sport_handle, bf5xx_i2s.tcr1, + bf5xx_i2s.tcr2, 0, 0); if (ret) { pr_err("SPORT is busy!\n"); return -EBUSY;
On Wed, Sep 16, 2009 at 08:25:12PM -0400, Mike Frysinger wrote:
From: Cliff Cai cliff.cai@analog.com
If the sound system hasn't been utilized yet and we suspend, then we attempt to save/restore using state that doesn't exist. So use a global handle instead to reconfigure properly.
I expect that this will break if more than one DAI is in use, which can happen if people are using multiple DACs to do 5.1 or hardware mixing. Would it not make sense to allocate the SPORT when the DAI is probed?
I've applied the other four patches - thanks.
On Thu, Sep 17, 2009 at 05:52, Mark Brown wrote:
On Wed, Sep 16, 2009 at 08:25:12PM -0400, Mike Frysinger wrote:
From: Cliff Cai cliff.cai@analog.com
If the sound system hasn't been utilized yet and we suspend, then we attempt to save/restore using state that doesn't exist. So use a global handle instead to reconfigure properly.
I expect that this will break if more than one DAI is in use, which can happen if people are using multiple DACs to do 5.1 or hardware mixing. Would it not make sense to allocate the SPORT when the DAI is probed?
unfortunately, the system is already completely broken when more than one is used. ive opened an item to get this fixed, but it's going to require a few core pieces to be rewritten. so i'd like to have this merged since it improves the current situation. -mike
On Thu, Sep 17, 2009 at 10:02:26AM -0400, Mike Frysinger wrote:
unfortunately, the system is already completely broken when more than one is used. ive opened an item to get this fixed, but it's going to require a few core pieces to be rewritten. so i'd like to have this merged since it improves the current situation.
OK, I suspected something like that was going on. I'll apply the patch.
participants (3)
-
Mark Brown -
Mike Frysinger -
Mike Frysinger