[PATCH v2 0/2] ASoC: wm_adsp: Report when a control write changes the value
Writing a firmware control should be returning 1 if the control value changed, so these two patches add that.
Though this is an ALSA requirement it is also useful for non-ALSA clients of cs_dsp to know if the control value changed, so the main handling is implemented in cs_dsp. TLV controls are specifically an ALSA thing so they are handled specially in wm_adsp.
Simon Trimmer (2): firmware: cs_dsp: cs_dsp_coeff_write_ctrl() should report changed ASoC: wm_adsp: Return whether changed when writing controls
drivers/firmware/cirrus/cs_dsp.c | 17 ++++++++++++----- sound/soc/codecs/wm_adsp.c | 27 ++++++++++++++++++--------- 2 files changed, 30 insertions(+), 14 deletions(-)
From: Simon Trimmer simont@opensource.cirrus.com
ALSA callers need to know whether there was a change to the value so that they can report a control write change correctly.
Signed-off-by: Simon Trimmer simont@opensource.cirrus.com Signed-off-by: Richard Fitzgerald rf@opensource.cirrus.com --- drivers/firmware/cirrus/cs_dsp.c | 17 ++++++++++++----- 1 file changed, 12 insertions(+), 5 deletions(-)
diff --git a/drivers/firmware/cirrus/cs_dsp.c b/drivers/firmware/cirrus/cs_dsp.c index 81cc3d0f6eec..17b7198b87dc 100644 --- a/drivers/firmware/cirrus/cs_dsp.c +++ b/drivers/firmware/cirrus/cs_dsp.c @@ -752,7 +752,7 @@ static int cs_dsp_coeff_write_ctrl_raw(struct cs_dsp_coeff_ctl *ctl, * * Must be called with pwr_lock held. * - * Return: Zero for success, a negative number on error. + * Return: < 0 on error, 1 when the control value changed and 0 when it has not. */ int cs_dsp_coeff_write_ctrl(struct cs_dsp_coeff_ctl *ctl, unsigned int off, const void *buf, size_t len) @@ -767,16 +767,23 @@ int cs_dsp_coeff_write_ctrl(struct cs_dsp_coeff_ctl *ctl, if (len + off * sizeof(u32) > ctl->len) return -EINVAL;
- if (ctl->flags & WMFW_CTL_FLAG_VOLATILE) + if (ctl->flags & WMFW_CTL_FLAG_VOLATILE) { ret = -EPERM; - else if (buf != ctl->cache) - memcpy(ctl->cache + off * sizeof(u32), buf, len); + } else if (buf != ctl->cache) { + if (memcmp(ctl->cache + off * sizeof(u32), buf, len)) + memcpy(ctl->cache + off * sizeof(u32), buf, len); + else + return 0; + }
ctl->set = 1; if (ctl->enabled && ctl->dsp->running) ret = cs_dsp_coeff_write_ctrl_raw(ctl, off, buf, len);
- return ret; + if (ret < 0) + return ret; + + return 1; } EXPORT_SYMBOL_GPL(cs_dsp_coeff_write_ctrl);
From: Simon Trimmer simont@opensource.cirrus.com
Functions that update cs_dsp controls need to handle return codes that indicate whether the control value changed. A return code of 1 indicates a change, 0 indicates no-change and a negative value is an error condition.
Acked controls implicitly change value when written so a successful write shall always report that the value changed.
Signed-off-by: Simon Trimmer simont@opensource.cirrus.com Signed-off-by: Richard Fitzgerald rf@opensource.cirrus.com --- sound/soc/codecs/wm_adsp.c | 27 ++++++++++++++++++--------- 1 file changed, 18 insertions(+), 9 deletions(-)
diff --git a/sound/soc/codecs/wm_adsp.c b/sound/soc/codecs/wm_adsp.c index 34a94b011518..02fbffd73853 100644 --- a/sound/soc/codecs/wm_adsp.c +++ b/sound/soc/codecs/wm_adsp.c @@ -19,6 +19,7 @@ #include <linux/regmap.h> #include <linux/regulator/consumer.h> #include <linux/slab.h> +#include <linux/vmalloc.h> #include <linux/workqueue.h> #include <linux/debugfs.h> #include <sound/core.h> @@ -419,16 +420,21 @@ static int wm_coeff_tlv_put(struct snd_kcontrol *kctl, (struct soc_bytes_ext *)kctl->private_value; struct wm_coeff_ctl *ctl = bytes_ext_to_ctl(bytes_ext); struct cs_dsp_coeff_ctl *cs_ctl = ctl->cs_ctl; + void *scratch; int ret = 0;
- mutex_lock(&cs_ctl->dsp->pwr_lock); + scratch = vmalloc(size); + if (!scratch) + return -ENOMEM;
- if (copy_from_user(cs_ctl->cache, bytes, size)) + if (copy_from_user(scratch, bytes, size)) { ret = -EFAULT; - else - ret = cs_dsp_coeff_write_ctrl(cs_ctl, 0, cs_ctl->cache, size); - - mutex_unlock(&cs_ctl->dsp->pwr_lock); + } else { + mutex_lock(&cs_ctl->dsp->pwr_lock); + ret = cs_dsp_coeff_write_ctrl(cs_ctl, 0, scratch, size); + mutex_unlock(&cs_ctl->dsp->pwr_lock); + } + vfree(scratch);
return ret; } @@ -455,7 +461,10 @@ static int wm_coeff_put_acked(struct snd_kcontrol *kctl,
mutex_unlock(&cs_ctl->dsp->pwr_lock);
- return ret; + if (ret < 0) + return ret; + + return 1; }
static int wm_coeff_get(struct snd_kcontrol *kctl, @@ -682,10 +691,10 @@ int wm_adsp_write_ctl(struct wm_adsp *dsp, const char *name, int type, int ret;
ret = cs_dsp_coeff_write_ctrl(cs_ctl, 0, buf, len); - if (ret) + if (ret < 0) return ret;
- if (cs_ctl->flags & WMFW_CTL_FLAG_SYS) + if (ret == 0 || (cs_ctl->flags & WMFW_CTL_FLAG_SYS)) return 0;
ctl = cs_ctl->priv;
On Wed, 23 Nov 2022 16:58:09 +0000, Richard Fitzgerald wrote:
Writing a firmware control should be returning 1 if the control value changed, so these two patches add that.
Though this is an ALSA requirement it is also useful for non-ALSA clients of cs_dsp to know if the control value changed, so the main handling is implemented in cs_dsp. TLV controls are specifically an ALSA thing so they are handled specially in wm_adsp.
[...]
Applied to
https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git for-next
Thanks!
[1/2] firmware: cs_dsp: cs_dsp_coeff_write_ctrl() should report changed commit: c56f4b2442d33bd94c418697f753271099384bee [2/2] ASoC: wm_adsp: Return whether changed when writing controls commit: 7406bdbc4fb8b99cf0150cb2056a585c95ceafe7
All being well this means that it will be integrated into the linux-next tree (usually sometime in the next 24 hours) and sent to Linus during the next merge window (or sooner if it is a bug fix), however if problems are discovered then the patch may be dropped or reverted.
You may get further e-mails resulting from automated or manual testing and review of the tree, please engage with people reporting problems and send followup patches addressing any issues that are reported if needed.
If any updates are required or you are submitting further changes they should be sent as incremental updates against current git, existing patches will not be replaced.
Please add any relevant lists and maintainers to the CCs when replying to this mail.
Thanks, Mark
participants (2)
-
Mark Brown -
Richard Fitzgerald