November 2017 - Alsa-devel - mailman.alsa-project.org

[alsa-devel] [PATCH 0/7] ALSA: Fine-tuning for several function implementations
by SF Markus Elfring 14 Nov '17

14 Nov '17

From: Markus Elfring <elfring(a)users.sourceforge.net> Date: Tue, 24 Jan 2017 11:02:34 +0100 Some update suggestions were taken into account from static source code analysis. Markus Elfring (7): seq: Delete unnecessary checks in snd_seq_ioctl_query_subs() seq: Delete unnecessary checks in snd_seq_ioctl_get_subscription() seq: Delete unnecessary checks in snd_seq_ioctl_unsubscribe_port() seq: Delete unnecessary checks in snd_seq_ioctl_subscribe_port() seq: Adjust 16 function calls together with a variable assignment dmasound_core: Move two assignments for the variable "ret" in state_open() dmasound_core: Adjust six function calls together with a variable assignment sound/core/seq/seq_clientmgr.c | 163 ++++++++++++++++++++----------------- sound/oss/dmasound/dmasound_core.c | 33 +++++--- 2 files changed, 109 insertions(+), 87 deletions(-) -- 2.11.0

2 9

[alsa-devel] [PATCH 0/5] ASoC: cleanup random snd_soc_read/write - step [4/6]
by Kuninori Morimoto 14 Nov '17

14 Nov '17

Hi Mark Previous step [3/6] will be applied after merge window, thus, I try to post next step [4/6]. Current Codec framework is supporting its original reg_cache, but, we have "regmap" now, and common snd_soc_component_read/write are using it. This "legacy reg_cache" user driver is using its own xxx_read/write which is using reg_cache inside. If we can copy "Codec framework reg_cache" into its driver, we can remove original reg_cache from Codec framework. No problem, the user driver is very few. Let's do it (and remove original reg_cache in the future) (= for [1/5][2/5][3/5]). About [3/5] case, cx20442 driver is using reg_cache, but it is only for 1byte at this point. and this driver doesn't use snd_soc_write/read (which is using reg_cache). The reg_cache user is only debugfs. Let's just clean reg_cache for now, but keep cx20442_write. It seems Lars-Peter has some plan/patch for this driver, and it will update read/write related things which will use this cx20442_write. Our common snd_soc_read/write is wrapper of snd_soc_component_read/write now. snd_soc_component_read/write will try to use regmap first. And will try Codec framework's original .read/.write if driver doesn't use regmap. But, almost all driver which doesn't use regmap is using their original xxx_read/write instead of snd_soc_component_read/write. We can remove Codec framework's .read/.write if all these driver uses its own xxx_read/write instead of snd_soc_read/write (= for [4/5][5/5]). Kuninori Morimoto (5): ASoC: use internal reg_cache on uda1380 ASoC: use internal reg_cache on tlv320dac33 ASoC: cx20442: don't use reg_cache ASoC: don't use snd_soc_write/read on twl6040 ASoC: don't use snd_soc_write/read on twl4030 sound/soc/codecs/cx20442.c | 21 ++++----------------- sound/soc/codecs/tlv320dac33.c | 33 +++++++++++++-------------------- sound/soc/codecs/twl4030.c | 10 ++++------ sound/soc/codecs/twl6040.c | 4 +--- sound/soc/codecs/uda1380.c | 24 ++++++++++++++---------- 5 files changed, 36 insertions(+), 56 deletions(-) -- 1.9.1

1 5

[alsa-devel] [PATCH 0/9] Remove use of "gpio-reset" from DT
by Andrew F. Davis 13 Nov '17

13 Nov '17

Hello all, I was working on fixing up the tlv320aic31xx driver when I noticed its gpio reset was not working, it was due to this driver looking for "gpio-reset" instead of the usual "reset-gpio". A quick check shows this mistake is rare and only copied by one other audio CODECs: $ git grep "reset-gpio" | wc -l 630 $ git grep "gpio-reset" | wc -l 6 Luckliy the two effected drivers only use this reset line when power has been cut to the device, so not only were these optional properties but they had no real functional effect anyway. Lets just fix this typo before is spreads to drivers were it matters. I've also added fixes tags to each patch so it can be individually back-ported to where this bug was introduced if one wanted to. I'm hoping this can sit on next for a while to get the most testing, just in case I'm wrong about this not breaking anything :) Thanks, Andrew Andrew F. Davis (9): ASoC: tlv320aic31xx: Fix typo in DT binding documentation ASoC: tlv320aic3x: Fix typo in DT binding documentation ASoC: cs42l56: bindings: sound: Fix reset GPIO name in example DT binding ARM: dts: am335x-pepper: Fix the audio CODEC's reset pin ARM: dts: imx6: RDU2: Fix the audio CODEC's reset pin ARM: dts: imx: Fix the audio CODEC's reset pin ARM: dts: omap3-n900: Fix the audio CODEC's reset pin ASoC: tlv320aic31xx: Fix the reset GPIO OF name ASoC: tlv320aic3x: Fix the reset GPIO OF name Documentation/devicetree/bindings/sound/cs42l56.txt | 2 +- Documentation/devicetree/bindings/sound/tlv320aic31xx.txt | 5 ++++- Documentation/devicetree/bindings/sound/tlv320aic3x.txt | 6 +++++- arch/arm/boot/dts/am335x-pepper.dts | 2 +- arch/arm/boot/dts/imx6qdl-gw5903.dtsi | 2 +- arch/arm/boot/dts/imx6qdl-zii-rdu2.dtsi | 4 ++-- arch/arm/boot/dts/omap3-n900.dts | 4 ++-- sound/soc/codecs/tlv320aic31xx.c | 2 +- sound/soc/codecs/tlv320aic3x.c | 2 +- 9 files changed, 18 insertions(+), 11 deletions(-) -- 2.15.0

5 21

[alsa-devel] [PATCH] ALSA: echoaudio: Use common error handling code in two functions
by SF Markus Elfring 13 Nov '17

13 Nov '17

From: Markus Elfring <elfring(a)users.sourceforge.net> Date: Mon, 13 Nov 2017 21:57:57 +0100 * Add jump targets so that a bit of exception handling can be better reused at the end of these functions. This issue was detected by using the Coccinelle software. * The script "checkpatch.pl" pointed information out like the following. ERROR: do not use assignment in if condition Thus fix six affected source code places. Signed-off-by: Markus Elfring <elfring(a)users.sourceforge.net> --- sound/pci/echoaudio/echoaudio.c | 55 ++++++++++++++++++++++------------------- 1 file changed, 30 insertions(+), 25 deletions(-) diff --git a/sound/pci/echoaudio/echoaudio.c b/sound/pci/echoaudio/echoaudio.c index d68f99e076a8..ae58c48aaaf9 100644 --- a/sound/pci/echoaudio/echoaudio.c +++ b/sound/pci/echoaudio/echoaudio.c @@ -1913,7 +1913,8 @@ static int snd_echo_create(struct snd_card *card, pci_write_config_byte(pci, PCI_LATENCY_TIMER, 0xC0); - if ((err = pci_enable_device(pci)) < 0) + err = pci_enable_device(pci); + if (err < 0) return err; pci_set_master(pci); @@ -1945,11 +1946,11 @@ static int snd_echo_create(struct snd_card *card, if (sz > PAGE_SIZE) sz = PAGE_SIZE; /* We map only the required part */ - if ((chip->iores = request_mem_region(chip->dsp_registers_phys, sz, - ECHOCARD_NAME)) == NULL) { + chip->iores = request_mem_region(chip->dsp_registers_phys, sz, + ECHOCARD_NAME); + if (!chip->iores) { dev_err(chip->card->dev, "cannot get memory region\n"); - snd_echo_free(chip); - return -EBUSY; + goto e_busy; } chip->dsp_registers = (volatile u32 __iomem *) ioremap_nocache(chip->dsp_registers_phys, sz); @@ -1957,8 +1958,7 @@ static int snd_echo_create(struct snd_card *card, if (request_irq(pci->irq, snd_echo_interrupt, IRQF_SHARED, KBUILD_MODNAME, chip)) { dev_err(chip->card->dev, "cannot grab irq\n"); - snd_echo_free(chip); - return -EBUSY; + goto e_busy; } chip->irq = pci->irq; dev_dbg(card->dev, "pci=%p irq=%d subdev=%04x Init hardware...\n", @@ -1970,8 +1970,8 @@ static int snd_echo_create(struct snd_card *card, sizeof(struct comm_page), &chip->commpage_dma_buf) < 0) { dev_err(chip->card->dev, "cannot allocate the comm page\n"); - snd_echo_free(chip); - return -ENOMEM; + err = -ENOMEM; + goto free_sound_chip; } chip->comm_page_phys = chip->commpage_dma_buf.addr; chip->comm_page = (struct comm_page *)chip->commpage_dma_buf.area; @@ -1981,17 +1981,22 @@ static int snd_echo_create(struct snd_card *card, err = set_mixer_defaults(chip); if (err < 0) { dev_err(card->dev, "init_hw err=%d\n", err); - snd_echo_free(chip); - return err; + goto free_sound_chip; } - if ((err = snd_device_new(card, SNDRV_DEV_LOWLEVEL, chip, &ops)) < 0) { - snd_echo_free(chip); - return err; - } + err = snd_device_new(card, SNDRV_DEV_LOWLEVEL, chip, &ops); + if (err < 0) + goto free_sound_chip; + *rchip = chip; /* Init done ! */ return 0; + +e_busy: + err = -EBUSY; +free_sound_chip: + snd_echo_free(chip); + return err; } @@ -2020,10 +2025,9 @@ static int snd_echo_probe(struct pci_dev *pci, return err; chip = NULL; /* Tells snd_echo_create to allocate chip */ - if ((err = snd_echo_create(card, pci, &chip)) < 0) { - snd_card_free(card); - return err; - } + err = snd_echo_create(card, pci, &chip); + if (err < 0) + goto free_card; strcpy(card->driver, "Echo_" ECHOCARD_NAME); strcpy(card->shortname, chip->card_name); @@ -2036,18 +2040,18 @@ static int snd_echo_probe(struct pci_dev *pci, card->shortname, pci_id->subdevice & 0x000f, dsp, chip->dsp_registers_phys, chip->irq); - if ((err = snd_echo_new_pcm(chip)) < 0) { + err = snd_echo_new_pcm(chip); + if (err < 0) { dev_err(chip->card->dev, "new pcm error %d\n", err); - snd_card_free(card); - return err; + goto free_card; } #ifdef ECHOCARD_HAS_MIDI if (chip->has_midi) { /* Some Mia's do not have midi */ - if ((err = snd_echo_midi_create(card, chip)) < 0) { + err = snd_echo_midi_create(card, chip); + if (err < 0) { dev_err(chip->card->dev, "new midi error %d\n", err); - snd_card_free(card); - return err; + goto free_card; } } #endif @@ -2152,6 +2156,7 @@ static int snd_echo_probe(struct pci_dev *pci, ctl_error: dev_err(card->dev, "new control error %d\n", err); +free_card: snd_card_free(card); return err; } -- 2.15.0

1 0

[alsa-devel] [PATCH] ALSA: cs5530: Use common error handling code in snd_cs5530_create()
by SF Markus Elfring 13 Nov '17

13 Nov '17

From: Markus Elfring <elfring(a)users.sourceforge.net> Date: Mon, 13 Nov 2017 20:33:07 +0100 Add jump targets so that a bit of exception handling can be better reused at the end of this function. This issue was detected by using the Coccinelle software. Signed-off-by: Markus Elfring <elfring(a)users.sourceforge.net> --- sound/pci/cs5530.c | 48 ++++++++++++++++++++++++------------------------ 1 file changed, 24 insertions(+), 24 deletions(-) diff --git a/sound/pci/cs5530.c b/sound/pci/cs5530.c index 0a8cf94c4858..88f253489075 100644 --- a/sound/pci/cs5530.c +++ b/sound/pci/cs5530.c @@ -124,8 +124,8 @@ static int snd_cs5530_create(struct snd_card *card, chip = kzalloc(sizeof(*chip), GFP_KERNEL); if (chip == NULL) { - pci_disable_device(pci); - return -ENOMEM; + err = -ENOMEM; + goto disable_device; } chip->card = card; @@ -134,15 +134,14 @@ static int snd_cs5530_create(struct snd_card *card, err = pci_request_regions(pci, "CS5530"); if (err < 0) { kfree(chip); - pci_disable_device(pci); - return err; + goto disable_device; } chip->pci_base = pci_resource_start(pci, 0); mem = pci_ioremap_bar(pci, 0); if (mem == NULL) { - snd_cs5530_free(chip); - return -EBUSY; + err = -EBUSY; + goto free_sound_chip; } map = readw(mem + 0x18); @@ -163,8 +162,7 @@ static int snd_cs5530_create(struct snd_card *card, dev_info(card->dev, "XpressAudio at 0x%lx\n", sb_base); else { dev_err(card->dev, "Could not find XpressAudio!\n"); - snd_cs5530_free(chip); - return -ENODEV; + goto e_nodev; } if (map & (1<<5)) @@ -183,8 +181,7 @@ static int snd_cs5530_create(struct snd_card *card, dma16 = 7; else { dev_err(card->dev, "No 16bit DMA enabled\n"); - snd_cs5530_free(chip); - return -ENODEV; + goto e_nodev; } if (dma8 & 0x01) @@ -195,8 +192,7 @@ static int snd_cs5530_create(struct snd_card *card, dma8 = 3; else { dev_err(card->dev, "No 8bit DMA enabled\n"); - snd_cs5530_free(chip); - return -ENODEV; + goto e_nodev; } if (irq & 1) @@ -209,8 +205,7 @@ static int snd_cs5530_create(struct snd_card *card, irq = 10; else { dev_err(card->dev, "SoundBlaster IRQ not set\n"); - snd_cs5530_free(chip); - return -ENODEV; + goto e_nodev; } dev_info(card->dev, "IRQ: %d DMA8: %d DMA16: %d\n", irq, dma8, dma16); @@ -219,32 +214,37 @@ static int snd_cs5530_create(struct snd_card *card, dma16, SB_HW_CS5530, &chip->sb); if (err < 0) { dev_err(card->dev, "Could not create SoundBlaster\n"); - snd_cs5530_free(chip); - return err; + goto free_sound_chip; } err = snd_sb16dsp_pcm(chip->sb, 0); if (err < 0) { dev_err(card->dev, "Could not create PCM\n"); - snd_cs5530_free(chip); - return err; + goto free_sound_chip; } err = snd_sbmixer_new(chip->sb); if (err < 0) { dev_err(card->dev, "Could not create Mixer\n"); - snd_cs5530_free(chip); - return err; + goto free_sound_chip; } err = snd_device_new(card, SNDRV_DEV_LOWLEVEL, chip, &ops); - if (err < 0) { - snd_cs5530_free(chip); - return err; - } + if (err < 0) + goto free_sound_chip; *rchip = chip; return 0; + +disable_device: + pci_disable_device(pci); + return err; + +e_nodev: + err = -ENODEV; +free_sound_chip: + snd_cs5530_free(chip); + return err; } static int snd_cs5530_probe(struct pci_dev *pci, -- 2.15.0

1 0

[alsa-devel] [PATCH 0/4] ALSA: cs46xx: Adjustments for some function implementations
by SF Markus Elfring 13 Nov '17

13 Nov '17

From: Markus Elfring <elfring(a)users.sourceforge.net> Date: Mon, 13 Nov 2017 19:14:56 +0100 A few update suggestions were taken into account from static source code analysis. Markus Elfring (4): Adjust 33 function calls together with a variable assignment Use common error handling code in two functions Improve a size determination in cs46xx_dsp_proc_register_scb_desc() Adjust 35 checks for null pointers sound/pci/cs46xx/cs46xx.c | 75 ++++++++++---------- sound/pci/cs46xx/cs46xx_lib.c | 134 +++++++++++++++++++++--------------- sound/pci/cs46xx/dsp_spos.c | 60 +++++++++------- sound/pci/cs46xx/dsp_spos_scb_lib.c | 36 +++++----- 4 files changed, 166 insertions(+), 139 deletions(-) -- 2.15.0

2 5

[alsa-devel] [PATCH 0/1] USB Audio Device Class 3.0 support
by Ruslan Bilovol 13 Nov '17

13 Nov '17

Hi Takashi, This patch adds initial USB Audio Device Class 3.0 [1] support to the ALSA that we discussed at ELCE. The patch was tested with UAC3 gadget [2] that I posted to USB mailing list before. It is good for working with BADD device which implements such topologies like BAIF (Basic Audio Input Functions), BAOF (Basic Audio Output Functions) or compbination of both. UAC3 spec has changed descriptors laout and many other codes comparing to UAC2 spec, thus making reuse of existing sources impossible or quite complex. There are still many areas of improvement, as this patch doesn't have UAC3 Mixer Unit support nor some new features like Power Management I tested this with BeagleBone Black as UAC3 gadget device. [1] http://www.usb.org/developers/docs/devclass_docs/USB_Audio_v3.0.zip [2] http://www.spinics.net/lists/linux-usb/msg162482.html Ruslan Bilovol (1): ALSA: usb: initial USB Audio Device Class 3.0 support include/linux/usb/audio-v2.h | 4 +- include/linux/usb/audio-v3.h | 343 +++++++++++++++++++++++++++++++++++++++++ include/uapi/linux/usb/audio.h | 1 + sound/usb/card.c | 7 +- sound/usb/card.h | 2 +- sound/usb/clock.c | 228 ++++++++++++++++++++++++--- sound/usb/clock.h | 4 +- sound/usb/format.c | 87 +++++++++-- sound/usb/format.h | 6 +- sound/usb/mixer.c | 332 ++++++++++++++++++++++++++------------- sound/usb/stream.c | 296 ++++++++++++++++++++++++++++++----- 11 files changed, 1117 insertions(+), 193 deletions(-) create mode 100644 include/linux/usb/audio-v3.h -- 1.9.1

5 13

[alsa-devel] [alsa-lib][PATCH v2 0/2] ctl: deprecate APIs of dimension information
by Takashi Sakamoto 13 Nov '17

13 Nov '17

Hi, This patchset is to update my previous one: [alsa-devel] [alsa-lib][PATCH 0/2] ctl: deprecate APIs of dimension information http://mailman.alsa-project.org/pipermail/alsa-devel/2017-November/126972.h… In ALSA control interface of asound.h, 'struct snd_ctl_elem_info' has 'dimen' member to deliver information for multi-dimensional array, however there's no common way to handle the member. As a result, drivers can force userspace applications to handle the information by inconsistent ways. A series of echoaudio drivers is an actual example of the inconsistent usage. This issue was addressed in a commit 51db452df07b ('Revert "ALSA: echoaudio: purge contradictions between dimension matrix members and total number of members"') to Linux kernel[1]. Fortunately, at present, this feature of interface is just used by the drivers, and there's a way to obsolete usage of the feature. We can obsolete it without large impacts to userland. As a result of discussion at Linux miniconference 2017, usage of 'dimen' member of 'struct snd_ctl_elem_info' is going to be deprecated for future removal. This patchset is for proposed tasks at a development period of Linux kernel v4.15[3]. Some APIs of dimension information are deprecated. In a short discussion in my previous patchset, it's clear that actual removal of kernel interface is uncertain depending on several conditions. Thus the timing to remove the library APIs is not decided yet. This updated version reflects the issue in API documentation (but simply removes statement of the schedule...). [1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/s… [2] [alsa-devel] [ANNOUNCE] Audio Mini Summit 2017 at Prague, Oct 27 http://mailman.alsa-project.org/pipermail/alsa-devel/2017-July/123110.html [3] https://github.com/takaswie/presentations/blob/master/20171027/contents.md Takashi Sakamoto (2): ctl: deprecate APIs of dimensional information test: obsolete usage of APIs of dimensional information src/control/control.c | 9 +++++++++ test/user-ctl-element-set.c | 31 ------------------------------- 2 files changed, 9 insertions(+), 31 deletions(-) -- 2.14.1

2 4

[alsa-devel] [PATCH 0/3] ALSA: cs4281: Fine-tuning for five function implementations
by SF Markus Elfring 13 Nov '17

13 Nov '17

From: Markus Elfring <elfring(a)users.sourceforge.net> Date: Mon, 13 Nov 2017 13:46:23 +0100 A few update suggestions were taken into account from static source code analysis. Markus Elfring (3): Adjust 18 function calls together with a variable assignment Use common error handling code in snd_cs4281_probe() Use common error handling code in snd_cs4281_create() sound/pci/cs4281.c | 136 ++++++++++++++++++++++++++++++----------------------- 1 file changed, 77 insertions(+), 59 deletions(-) -- 2.15.0

1 3

[alsa-devel] Broken child-node lookup in sound/soc/mediatek/mt8173
by Johan Hovold 13 Nov '17

13 Nov '17

Hi, I'm trying to fix up incorrect usage of of_find_node_by_name() to lookup child nodes, and found the following code in sound/soc/mediatek/mt8173/mt8173-rt5650.c: static int mt8173_rt5650_dev_probe(struct platform_device *pdev) { ... platform_node = of_parse_phandle(pdev->dev.of_node, "mediatek,platform", 0); ... if (of_find_node_by_name(platform_node, "codec-capture")) { np = of_get_child_by_name(pdev->dev.of_node, "codec-capture"); if (!np) { dev_err(&pdev->dev, "%s: Can't find codec-capture DT node\n", __func__); return -EINVAL; } ret = snd_soc_of_get_dai_name(np, &codec_capture_dai); if (ret < 0) { dev_err(&pdev->dev, "%s codec_capture_dai name fail %d\n", __func__, ret); return ret; } mt8173_rt5650_codecs[1].dai_name = codec_capture_dai; } added by commit d349caeb0510 ("ASoC: mediatek: Add second I2S on mt8173-rt5650 machine driver"). First of all the "codec-capture" node is indeed documented as a child node of the sound node, so the tree-wide depth-first search from the platform_node looks entirely bogus. Note that of_find_node_by_name() also drops a reference to its first argument, in this case the sound node, which could end up being prematurely freed. And then the reference to any returned codec-capture node (from either lookup) is never dropped. And since support for this second codec was added retrospectively and is documented as optional, that -EINVAL in case the node is missing looks broken too. I figured I better just report this one to the author of the patch and the maintainers to be straightened out. Thanks, Johan

1 0