May 2016 - Alsa-devel - mailman.alsa-project.org

[alsa-devel] New codecs support for ALC234/ALC274/ALC294
by Kailang 06 May '16

06 May '16

Hi Takashi, We have new codecs for ALC234/ALC274/ALC294. Attach file is the support patch. BR, Kailang

1 0

[alsa-devel] Errors iterating over PCM stream hints and opening based on NAME/IOID
by Trent Reed 06 May '16

06 May '16

Hey All, I'm trying to understand ALSA, having a little trouble with some errors that I can't ration over. Gist Repro: https://gist.github.com/TReed0803/353ce27d37bc1ebe965854ea88c8b97b Whenever I run the above simple code (iterate over PCM stream hints, open for Playback/Capture depending on IOID; output also provided at link above) I notice that I sometimes hit errors. I'm having trouble understanding these errors, because as I understand I'm using the ALSA library correctly. For example, dmix is telling me that it "Only supports playback streams", well then, why does the hint IOID return that it supports capture as well? (same for dsnoop). More confusing are the "ALSA lib pcm_route.c:867:(find_matching_chmap) Found no matching channel map" errors. Can someone help me understand these issues, and perhaps what I'm doing wrong here? Would this point to something wrong with an installed package which perhaps configures these hints incorrectly? (Assuming that is probably the case, but I'd like to know for sure). I'm on a fairly new install of Ubuntu 16.04 LTS. Thanks, - Trent Reed

1 0

[alsa-devel] Dice packet sequence quirk and ALSA firewire stack in Linux 4.6
by Takashi Sakamoto 06 May '16

06 May '16

Hi all, In this developing cycle for Linux 4.7, I concentrate on working to check packet sequence of ALSA firewire stack, because I ensure that the stack causes Dice-based models to generate sound including periodical noise or to generate no sound in the worst case. As of the end of developing cycle for Linux 4.6, I assumed that this is a bug of ALSA firewire stack[1]. However, the assumption is denied by my analysis with added tracepoints of Linux tracing framework[2]. Then, I concluded that the issue comes from quirks of Dice in an aspect of processing presentation timestamp. I'm under further investigation and have seen some of the quirks. In this message, I describe the summary of superficial issues from the quirks, the status of ALSA firewire stack in Linux 4.6 for Dice-based models and direction of my work for ALSA firewire stack in this developing cycle. Besides, resources and my memo for this work is available in this URL. I wish to upload it to alsa-project.org, not in my server: http://sakamocchi.jp/media/20160505-dice-packet-dump.tar.xz (233MB archive file compressed by xz, expands 6.7 GB text files.) Summary - Dice ASICs/firmwares/drivers communicate with packet sequences which is partly not compliant to IEC 61883-6. - Dice ASICs/firmwares/drivers doesn't sometimes transfer data blocks as the same as sampling transfer frequency. - Calculation of presentation timestamp is not based on duration of data blocks in a packet. This way is different from implementation of ALSA firewire stack. - The value of estimated 'transfer delay' is against IEEE 1394 spec. As a result: - ALSA dice driver causes these issues. ALSA dice driver in Linux kernels since 3.13 have included this issue. - To fix this issue, ALSA firewire stack require timestamp generator and data block calculator specific for Dice-based models. At least, three functions should be changed; amdtp_stream_start(), calculate_data_blocks(), calculate_syt() in 'sound/firewire/amdtp-stream.c'. - The work requires much changes, thus it's impossible for Linux 4.6. ALSA firewire stack in Linux 4.6 will be shipped with these issues. - We have a short time till merge window of Linux 4.7 (Just one week). I'll work for merging my proposed patchset[2][3][4], without fixing the issues. [1] http://mailman.alsa-project.org/pipermail/alsa-devel/2016-March/105608.html [2] http://mailman.alsa-project.org/pipermail/alsa-devel/2016-April/106950.html [3] http://mailman.alsa-project.org/pipermail/alsa-devel/2016-April/107290.html [4] http://mailman.alsa-project.org/pipermail/alsa-devel/2016-April/107531.html Regards Takashi Sakamoto

1 0

[alsa-devel] I-MX6 & I2s output
by nick83ola 05 May '16

05 May '16

Hello to All I have an IMX6 board and I need a simple way to capture/put an I2s master signal I'm working with yocto and kernel version 3.14.28-1.0.0 I have on my development board an sgtl5000 codec but it work only with I2S slave mode I've found the driver fsl-asoc-card.c from more recent kernel and I've modified it But no matter what I do i cannot set I2S to master mode Please Help -- P.S. Le informazioni trasmesse attraverso la presente comunicazione sono di esclusiva spettanza dell'effettivo destinatario. Nel caso in cui le stesse raggiungessero, per qualunque motivo, soggetti non interessati, questi ultimi vorranno darne immediata notizia al mittente. In ogni caso, eventuali soggetti diversi dai legittimi destinatari della presente comunicazione e dei dati contenuti negli allegati, possono essere sanzionati ai sensi del T.U. sul trattamento dei dati personali d.lgs. 196/2003, sia ai sensi dell'art. 616 del Codice Penale che disciplina la violazione del segreto sulla corrispondenza.

1 0

[alsa-devel] [PATCH 1/2] ASoC: wm_adsp: Move compr_attach/attached functions
by Charles Keepax 05 May '16

05 May '16

Move wm_adsp_compr_attach and wm_adsp_compr_attached functions so they will stay logically grouped with similar functions after some additional changes. Signed-off-by: Charles Keepax <ckeepax(a)opensource.wolfsonmicro.com> --- sound/soc/codecs/wm_adsp.c | 38 +++++++++++++++++++------------------- 1 file changed, 19 insertions(+), 19 deletions(-) diff --git a/sound/soc/codecs/wm_adsp.c b/sound/soc/codecs/wm_adsp.c index 630ebcd..42fc469 100644 --- a/sound/soc/codecs/wm_adsp.c +++ b/sound/soc/codecs/wm_adsp.c @@ -2452,6 +2452,25 @@ void wm_adsp2_remove(struct wm_adsp *dsp) } EXPORT_SYMBOL_GPL(wm_adsp2_remove); +static inline int wm_adsp_compr_attached(struct wm_adsp_compr *compr) +{ + return compr->buf != NULL; +} + +static int wm_adsp_compr_attach(struct wm_adsp_compr *compr) +{ + /* + * Note this will be more complex once each DSP can support multiple + * streams + */ + if (!compr->dsp->buffer) + return -EINVAL; + + compr->buf = compr->dsp->buffer; + + return 0; +} + int wm_adsp_compr_open(struct wm_adsp *dsp, struct snd_compr_stream *stream) { struct wm_adsp_compr *compr; @@ -2810,25 +2829,6 @@ static int wm_adsp_buffer_free(struct wm_adsp *dsp) return 0; } -static inline int wm_adsp_compr_attached(struct wm_adsp_compr *compr) -{ - return compr->buf != NULL; -} - -static int wm_adsp_compr_attach(struct wm_adsp_compr *compr) -{ - /* - * Note this will be more complex once each DSP can support multiple - * streams - */ - if (!compr->dsp->buffer) - return -EINVAL; - - compr->buf = compr->dsp->buffer; - - return 0; -} - int wm_adsp_compr_trigger(struct snd_compr_stream *stream, int cmd) { struct wm_adsp_compr *compr = stream->runtime->private_data; -- 2.1.4

2 3

[alsa-devel] [PATCH 1/2] Revert "ASoC: rockchip: i2s: remove unused variables"
by John Keeping 05 May '16

05 May '16

This reverts commit 5938448b99275cba95167c3f9d39ca9225fdad38. It turns out that the commit that made these variables unused is wrong so we're about to revert it. Bring back the variables in prepration. Signed-off-by: John Keeping <john(a)metanate.com> --- sound/soc/rockchip/rockchip_i2s.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) diff --git a/sound/soc/rockchip/rockchip_i2s.c b/sound/soc/rockchip/rockchip_i2s.c index 2f8e20416bd3..34743ec61c49 100644 --- a/sound/soc/rockchip/rockchip_i2s.c +++ b/sound/soc/rockchip/rockchip_i2s.c @@ -34,6 +34,13 @@ struct rk_i2s_dev { struct regmap *regmap; +/* + * Used to indicate the tx/rx status. + * I2S controller hopes to start the tx and rx together, + * also to stop them when they are both try to stop. +*/ + bool tx_start; + bool rx_start; bool is_master_mode; }; @@ -77,7 +84,11 @@ static void rockchip_snd_txctrl(struct rk_i2s_dev *i2s, int on) regmap_update_bits(i2s->regmap, I2S_XFER, I2S_XFER_TXS_START, I2S_XFER_TXS_START); + + i2s->tx_start = true; } else { + i2s->tx_start = false; + regmap_update_bits(i2s->regmap, I2S_DMACR, I2S_DMACR_TDE_ENABLE, I2S_DMACR_TDE_DISABLE); @@ -115,7 +126,11 @@ static void rockchip_snd_rxctrl(struct rk_i2s_dev *i2s, int on) regmap_update_bits(i2s->regmap, I2S_XFER, I2S_XFER_RXS_START, I2S_XFER_RXS_START); + + i2s->rx_start = true; } else { + i2s->rx_start = false; + regmap_update_bits(i2s->regmap, I2S_DMACR, I2S_DMACR_RDE_ENABLE, I2S_DMACR_RDE_DISABLE); -- 2.8.0.rc4.238.g874082a

3 5

[alsa-devel] [PATCH] ASoC: rt5645: polling jd status in all conditions
by Bard Liao 05 May '16

05 May '16

We only polling jd status when rt5645->pdata.jd_invert is true. However, it should be done at all time since there will be no interrupt for jd if we press a headset button and remove the headset at the same time. Signed-off-by: Bard Liao <bardliao(a)realtek.com> --- sound/soc/codecs/rt5645.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) diff --git a/sound/soc/codecs/rt5645.c b/sound/soc/codecs/rt5645.c index dff706a..3c6594d 100644 --- a/sound/soc/codecs/rt5645.c +++ b/sound/soc/codecs/rt5645.c @@ -3286,10 +3286,8 @@ static void rt5645_jack_detect_work(struct work_struct *work) if (btn_type == 0)/* button release */ report = rt5645->jack_type; else { - if (rt5645->pdata.jd_invert) { - mod_timer(&rt5645->btn_check_timer, - msecs_to_jiffies(100)); - } + mod_timer(&rt5645->btn_check_timer, + msecs_to_jiffies(100)); } break; @@ -3816,9 +3814,9 @@ static int rt5645_i2c_probe(struct i2c_client *i2c, if (rt5645->pdata.jd_invert) { regmap_update_bits(rt5645->regmap, RT5645_IRQ_CTRL2, RT5645_JD_1_1_MASK, RT5645_JD_1_1_INV); - setup_timer(&rt5645->btn_check_timer, - rt5645_btn_check_callback, (unsigned long)rt5645); } + setup_timer(&rt5645->btn_check_timer, + rt5645_btn_check_callback, (unsigned long)rt5645); INIT_DELAYED_WORK(&rt5645->jack_detect_work, rt5645_jack_detect_work); INIT_DELAYED_WORK(&rt5645->rcclock_work, rt5645_rcclock_work); -- 1.8.1.1.439.g50a6b54

2 1

[alsa-devel] [PATCH v3 1/2] ASoC: fsl_sai: Introduce a compatible string for MX6UL
by Fabio Estevam 05 May '16

05 May '16

From: Fabio Estevam <fabio.estevam(a)nxp.com> MX6UL may need to configure the General Purpose Register 1 (GPR1), so it is better to add a new compatible string to differentiate. Signed-off-by: Fabio Estevam <fabio.estevam(a)nxp.com> --- Changes since v2: - Add binding doc entry for mx6ul (Nicolin) Documentation/devicetree/bindings/sound/fsl-sai.txt | 4 ++-- sound/soc/fsl/fsl_sai.c | 4 +++- 2 files changed, 5 insertions(+), 3 deletions(-) diff --git a/Documentation/devicetree/bindings/sound/fsl-sai.txt b/Documentation/devicetree/bindings/sound/fsl-sai.txt index 044e5d7..777b941 100644 --- a/Documentation/devicetree/bindings/sound/fsl-sai.txt +++ b/Documentation/devicetree/bindings/sound/fsl-sai.txt @@ -7,8 +7,8 @@ codec/DSP interfaces. Required properties: - - compatible : Compatible list, contains "fsl,vf610-sai" or - "fsl,imx6sx-sai". + - compatible : Compatible list, contains "fsl,vf610-sai", + "fsl,imx6sx-sai" or "fsl,imx6ul-sai" - reg : Offset and length of the register set for the device. diff --git a/sound/soc/fsl/fsl_sai.c b/sound/soc/fsl/fsl_sai.c index 0754df7..d8b673f 100644 --- a/sound/soc/fsl/fsl_sai.c +++ b/sound/soc/fsl/fsl_sai.c @@ -797,7 +797,8 @@ static int fsl_sai_probe(struct platform_device *pdev) sai->pdev = pdev; - if (of_device_is_compatible(pdev->dev.of_node, "fsl,imx6sx-sai")) + if (of_device_is_compatible(pdev->dev.of_node, "fsl,imx6sx-sai") || + of_device_is_compatible(pdev->dev.of_node, "fsl,imx6ul-sai")) sai->sai_on_imx = true; sai->is_lsb_first = of_property_read_bool(np, "lsb-first"); @@ -898,6 +899,7 @@ static int fsl_sai_probe(struct platform_device *pdev) static const struct of_device_id fsl_sai_ids[] = { { .compatible = "fsl,vf610-sai", }, { .compatible = "fsl,imx6sx-sai", }, + { .compatible = "fsl,imx6ul-sai", }, { /* sentinel */ } }; MODULE_DEVICE_TABLE(of, fsl_sai_ids); -- 1.9.1

3 5

[alsa-devel] [PATCH 0/2] ASoC: Memeory leak fixes
by Vinod Koul 05 May '16

05 May '16

This series fixes two memeory leaks, first in topology core to freeup the objects in widget creation and second one on Skylake driver for NHLT table cleanup Jeeja KP (2): ASoC: topology: Fix memory leak in widget creation ASoC: Intel: Skylake: Fix memory leak in nhlt init sound/soc/intel/skylake/skl-nhlt.c | 15 +++++++++------ sound/soc/intel/skylake/skl.c | 5 ++++- sound/soc/intel/skylake/skl.h | 6 +++--- sound/soc/soc-topology.c | 2 ++ 4 files changed, 18 insertions(+), 10 deletions(-) -- 1.9.1

2 4

[alsa-devel] [PATCH v2 1/2] ASoC: fsl_sai: Introduce a compatible string for MX6UL
by Fabio Estevam 05 May '16

05 May '16

From: Fabio Estevam <fabio.estevam(a)nxp.com> MX6UL may need to configure the General Purpose Register 1 (GPR1), so it is better to add a new compatible string to differentiate. Signed-off-by: Fabio Estevam <fabio.estevam(a)nxp.com> --- Changes since v1: - Newly introduced in this series. sound/soc/fsl/fsl_sai.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/sound/soc/fsl/fsl_sai.c b/sound/soc/fsl/fsl_sai.c index 0754df7..d8b673f 100644 --- a/sound/soc/fsl/fsl_sai.c +++ b/sound/soc/fsl/fsl_sai.c @@ -797,7 +797,8 @@ static int fsl_sai_probe(struct platform_device *pdev) sai->pdev = pdev; - if (of_device_is_compatible(pdev->dev.of_node, "fsl,imx6sx-sai")) + if (of_device_is_compatible(pdev->dev.of_node, "fsl,imx6sx-sai") || + of_device_is_compatible(pdev->dev.of_node, "fsl,imx6ul-sai")) sai->sai_on_imx = true; sai->is_lsb_first = of_property_read_bool(np, "lsb-first"); @@ -898,6 +899,7 @@ static int fsl_sai_probe(struct platform_device *pdev) static const struct of_device_id fsl_sai_ids[] = { { .compatible = "fsl,vf610-sai", }, { .compatible = "fsl,imx6sx-sai", }, + { .compatible = "fsl,imx6ul-sai", }, { /* sentinel */ } }; MODULE_DEVICE_TABLE(of, fsl_sai_ids); -- 1.9.1

2 4