December 2015 - Alsa-devel - mailman.alsa-project.org

[alsa-devel] [PATCH v2 MC Next Gen] sound/usb: Fix out of bounds access in media_entity_init()
by Shuah Khan 07 Dec '15

07 Dec '15

Fix the out of bounds access in media_entity_init() found by KASan. This is a result of media_mixer_init() failing to allocate memory for all 3 of its pads before calling media_entity_init(). Fix it to allocate memory for the right struct media_mixer_ctl instead of struct media_ctl. Signed-off-by: Shuah Khan <shuahkh(a)osg.samsung.com> --- Changes since v1: Change to address review comment from Takashi Iwai This patch fixes the mixer patch below: https://patchwork.linuxtv.org/patch/31827/ sound/usb/media.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/usb/media.c b/sound/usb/media.c index bebe27b..b0d283f 100644 --- a/sound/usb/media.c +++ b/sound/usb/media.c @@ -233,8 +233,8 @@ int media_mixer_init(struct snd_usb_audio *chip) if (mixer->media_mixer_ctl) continue; - /* allocate media_ctl */ - mctl = kzalloc(sizeof(struct media_ctl), GFP_KERNEL); + /* allocate media_mixer_ctl */ + mctl = kzalloc(sizeof(*mctl), GFP_KERNEL); if (!mctl) return -ENOMEM; -- 2.5.0

1 0

[alsa-devel] Sound UsbPre : minimal patch for support
by shal＠free.fr 07 Dec '15

07 Dec '15

Hello, I have a Sound UsbPre (the first not the UsbPRe2) sound card. There is no support for this card in alsa. I do a usb packet sniffing on Windows for detect problem. Some control quirk should be added at boot in order to use the card. It works to play and record in 44100 khz. It works only with 44100khz (for example, if you play a 88200 file, it will be played too slow). This patch is not well tested, work on my desktop : Linux Ubuntu with 4.2.19-lowlatency kernel. The button emulation is not done (by a control packet, you can change the card configuration, like phantom power activation). If somebody is interested, I can provide needed information. If Alsa's developer is interested, I can provide support for test patches or develop a second version of this patch in order to integrate the support for this card in the linux kernel Regards, Olivier linux-source-4.2.0/sound/usb$ diff -rub quirks.c.ORIG quirks.c --- quirks.c.ORIG 2015-11-24 22:43:16.878635155 +0100 +++ quirks.c 2015-11-26 20:56:09.492126460 +0100 @@ -601,6 +601,40 @@ return 0; } +/* Add some control messages needed by Sound Devices UsbPre. + * Values here are chosen based on sniffing USB traffic + * under Windows.*/ +static int snd_usb_usbpre_boot_quirk(struct usb_device *dev) +{ + u8 buf = 1; + u8 data1[1] = {0x81}; + u8 data2[1] = {0xc1}; + u8 data3[3] = {0x44,0xac,0x00}; + + snd_usb_ctl_msg(dev, usb_sndctrlpipe(dev, 0), 1, + USB_DIR_OUT | USB_TYPE_VENDOR | USB_RECIP_DEVICE, + 0x00f0, 496, data1, 1); + snd_usb_ctl_msg(dev, usb_rcvctrlpipe(dev, 0), 0x81 , + USB_DIR_IN | USB_TYPE_VENDOR | USB_RECIP_DEVICE, + 0x00f0, 0, &buf, 1); + if (buf == 0) { + return -ENODEV; + } + snd_usb_ctl_msg(dev, usb_sndctrlpipe(dev, 0), 1, + USB_DIR_OUT | USB_TYPE_VENDOR | USB_RECIP_DEVICE, + 0x00f0, 496, data2, 1); + snd_usb_ctl_msg(dev, usb_rcvctrlpipe(dev, 0), 0x81 , + USB_DIR_IN | USB_TYPE_VENDOR | USB_RECIP_DEVICE, + 0x00f0, 0, &buf, 1); + if (buf == 0) { + return -ENODEV; + } + snd_usb_ctl_msg(dev, usb_sndctrlpipe(dev, 0), 1, + USB_DIR_OUT | USB_TYPE_VENDOR | USB_RECIP_ENDPOINT, + 0x0100, 133, data3, 3); + return 0; +} + static int snd_usb_fasttrackpro_boot_quirk(struct usb_device *dev) { int err; @@ -985,6 +1019,10 @@ /* SB Audigy 2 NX needs its own boot-up magic, too */ return snd_usb_audigy2nx_boot_quirk(dev); + case USB_ID(0x0926, 0x0100): + /* Sound Devices UsbPre needs its own boot-up magic, too */ + return snd_usb_usbpre_boot_quirk(dev); + case USB_ID(0x10f5, 0x0200): /* C-Media CM106 / Turtle Beach Audio Advantage Roadie */ return snd_usb_cm106_boot_quirk(dev);

2 1

[alsa-devel] [PATCH MC Next Gen] sound/usb: Fix out of bounds access in media_entity_init()
by Shuah Khan 07 Dec '15

07 Dec '15

Fix the out of bounds access in media_entity_init() found by KASan. This is a result of media_mixer_init() failing to allocate memory for all 3 of its pads before calling media_entity_init(). Fix it to allocate memory for the right struct media_mixer_ctl instead of struct media_ctl. Signed-off-by: Shuah Khan <shuahkh(a)osg.samsung.com> --- This patch fixes the mixer patch below: https://patchwork.linuxtv.org/patch/31827/ sound/usb/media.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/sound/usb/media.c b/sound/usb/media.c index bebe27b..0cb44b9 100644 --- a/sound/usb/media.c +++ b/sound/usb/media.c @@ -233,8 +233,8 @@ int media_mixer_init(struct snd_usb_audio *chip) if (mixer->media_mixer_ctl) continue; - /* allocate media_ctl */ - mctl = kzalloc(sizeof(struct media_ctl), GFP_KERNEL); + /* allocate media_mixer_ctl */ + mctl = kzalloc(sizeof(struct media_mixer_ctl), GFP_KERNEL); if (!mctl) return -ENOMEM; @@ -244,6 +244,7 @@ int media_mixer_init(struct snd_usb_audio *chip) mctl->media_pad[0].flags = MEDIA_PAD_FL_SINK; mctl->media_pad[1].flags = MEDIA_PAD_FL_SOURCE; mctl->media_pad[2].flags = MEDIA_PAD_FL_SOURCE; + media_entity_init(&mctl->media_entity, MEDIA_MIXER_PAD_MAX, mctl->media_pad); ret = media_device_register_entity(mctl->media_dev, -- 2.5.0

2 2

[alsa-devel] [GIT PULL] AC'97 gpiochip
by Mark Brown 07 Dec '15

07 Dec '15

The following changes since commit 8005c49d9aea74d382f474ce11afbbc7d7130bec: Linux 4.4-rc1 (2015-11-15 17:00:27 -0800) are available in the git repository at: git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git tags/asoc-ac97-gpio for you to fetch changes up to 9bf5c3d11f1fbaf43399d189f05fb20ceb46ee5d: ASoC: ac97: add gpio chip (2015-11-18 18:08:54 +0000) ---------------------------------------------------------------- ASoC: Add a GPIO chip for AC'97 GPIOs are part of the AC'97 spec, enable their use on embedded platforms using AC'97. ---------------------------------------------------------------- Robert Jarzmik (1): ASoC: ac97: add gpio chip include/sound/ac97_codec.h | 3 ++ sound/soc/soc-ac97.c | 125 +++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 128 insertions(+)

2 1

[alsa-devel] [PATCH] ALSA: hda - Add inverted dmic for Packard Bell DOTS
by David Henningsson 07 Dec '15

07 Dec '15

On the internal mic of the Packard Bell DOTS, one channel has an inverted signal. Add a quirk to fix this up. Cc: stable(a)vger.kernel.org BugLink: https://bugs.launchpad.net/bugs/1523232 Signed-off-by: David Henningsson <david.henningsson(a)canonical.com> --- sound/pci/hda/patch_realtek.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 9bedf7c..ebc5362 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -6409,6 +6409,7 @@ static const struct hda_fixup alc662_fixups[] = { static const struct snd_pci_quirk alc662_fixup_tbl[] = { SND_PCI_QUIRK(0x1019, 0x9087, "ECS", ALC662_FIXUP_ASUS_MODE2), SND_PCI_QUIRK(0x1025, 0x022f, "Acer Aspire One", ALC662_FIXUP_INV_DMIC), + SND_PCI_QUIRK(0x1025, 0x0241, "Packard Bell DOTS", ALC662_FIXUP_INV_DMIC), SND_PCI_QUIRK(0x1025, 0x0308, "Acer Aspire 8942G", ALC662_FIXUP_ASPIRE), SND_PCI_QUIRK(0x1025, 0x031c, "Gateway NV79", ALC662_FIXUP_SKU_IGNORE), SND_PCI_QUIRK(0x1025, 0x0349, "eMachines eM250", ALC662_FIXUP_INV_DMIC), -- 1.9.1

2 1

[alsa-devel] [PATCH v3] ALSA: compress: add support for 32bit calls in a 64bit kernel
by Ravindra Lokhande 07 Dec '15

07 Dec '15

Compress offload does not support ioctl calls from a 32bit userspace in a 64 bit kernel. This patch adds support for ioctls from a 32bit userspace in a 64bit kernel Signed-off-by: Ravindra Lokhande <rlokhande(a)nvidia.com> --- v1 -> v2: use compat_ptr to convert 32 bit address to 64 bit pointer v2 -> v3: add header file compat.h sound/core/compress_offload.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/sound/core/compress_offload.c b/sound/core/compress_offload.c index b123c42..4597820 100644 --- a/sound/core/compress_offload.c +++ b/sound/core/compress_offload.c @@ -38,6 +38,7 @@ #include <linux/uio.h> #include <linux/uaccess.h> #include <linux/module.h> +#include <linux/compat.h> #include <sound/core.h> #include <sound/initval.h> #include <sound/compress_params.h> @@ -847,6 +848,15 @@ static long snd_compr_ioctl(struct file *f, unsigned int cmd, unsigned long arg) return retval; } +/* support of 32bit userspace on 64bit platforms */ +#ifdef CONFIG_COMPAT +static long snd_compr_ioctl_compat(struct file *file, unsigned int cmd, + unsigned long arg) +{ + return snd_compr_ioctl(file, cmd, (unsigned long)compat_ptr(arg)); +} +#endif + static const struct file_operations snd_compr_file_ops = { .owner = THIS_MODULE, .open = snd_compr_open, @@ -854,6 +864,9 @@ static const struct file_operations snd_compr_file_ops = { .write = snd_compr_write, .read = snd_compr_read, .unlocked_ioctl = snd_compr_ioctl, +#ifdef CONFIG_COMPAT + .compat_ioctl = snd_compr_ioctl_compat, +#endif .mmap = snd_compr_mmap, .poll = snd_compr_poll, }; -- 2.1.4

3 3

[alsa-devel] [PATCH 1/1] audio: Fix playback noise with 24/32 bit sample size on BXT
by han.lu＠intel.com 07 Dec '15

07 Dec '15

From: "Lu, Han" <han.lu(a)intel.com> In BXT-P A0, HD-Audio DMA requests is later than expected, and makes an audio stream sensitive to system latencies when 24/32 bits are playing. Adjusting threshold of DMA fifo to force the DMA request sooner to improve latency tolerance at the expense of power. Signed-off-by: Lu, Han <han.lu(a)intel.com> diff --git a/include/sound/hda_register.h b/include/sound/hda_register.h index 2ae8812..94dc6a9 100644 --- a/include/sound/hda_register.h +++ b/include/sound/hda_register.h @@ -93,6 +93,9 @@ enum { SDI0, SDI1, SDI2, SDI3, SDO0, SDO1, SDO2, SDO3 }; #define AZX_REG_HSW_EM4 0x100c #define AZX_REG_HSW_EM5 0x1010 +/* Skylake/Broxton display HD-A controller Extended Mode registers */ +#define AZX_REG_SKL_EM4L 0x1040 + /* PCI space */ #define AZX_PCIREG_TCSEL 0x44 diff --git a/sound/pci/hda/hda_controller.c b/sound/pci/hda/hda_controller.c index 22dbfa5..6c2aa4c 100644 --- a/sound/pci/hda/hda_controller.c +++ b/sound/pci/hda/hda_controller.c @@ -27,6 +27,7 @@ #include <linux/module.h> #include <linux/pm_runtime.h> #include <linux/slab.h> +#include <linux/pci.h> #include <sound/core.h> #include <sound/initval.h> #include "hda_controller.h" @@ -34,6 +35,8 @@ #define CREATE_TRACE_POINTS #include "hda_controller_trace.h" +#define is_broxton(pci) ((pci)->device == 0x5a98) + /* DSP lock helpers */ #define dsp_lock(dev) snd_hdac_dsp_lock(azx_stream(dev)) #define dsp_unlock(dev) snd_hdac_dsp_unlock(azx_stream(dev)) @@ -880,14 +883,35 @@ EXPORT_SYMBOL_GPL(snd_hda_codec_load_dsp_cleanup); #endif /* CONFIG_SND_HDA_DSP_LOADER */ /* + * In BXT-P A0, HD-Audio DMA requests is later than expected, + * and makes an audio stream sensitive to system latencies when + * 24/32 bits are playing. + * Adjusting threshold of DMA fifo to force the DMA request + * sooner to improve latency tolerance at the expense of power. + */ +static void bxt_reduce_dma_latency(struct azx *chip) +{ + u32 val; + + val = azx_readl(chip, SKL_EM4L); + val &= (0x3 << 20); + azx_writel(chip, SKL_EM4L, val); +} + +/* * reset and start the controller registers */ void azx_init_chip(struct azx *chip, bool full_reset) { + struct pci_dev *pci = chip->pci; + if (snd_hdac_bus_init_chip(azx_bus(chip), full_reset)) { /* correct RINTCNT for CXT */ if (chip->driver_caps & AZX_DCAPS_CTX_WORKAROUND) azx_writew(chip, RINTCNT, 0xc0); + /* reduce dma latency to avoid noise */ + if (is_broxton(pci)) + bxt_reduce_dma_latency(chip); } } EXPORT_SYMBOL_GPL(azx_init_chip); -- 2.5.0

3 2

[alsa-devel] [PATCH] crec: Always read whole number of fragments when length limit specified
by Richard Fitzgerald 07 Dec '15

07 Dec '15

This patch removes the lines that adjusted the read size to exactly read to the end of the calculated length in bytes. Doing this could lead to invalid fragment sizes being requested which may then cause a read error on some codecs. Instead we keep reading whole fragments until we have >= length bytes of data captured. Signed-off-by: Richard Fitzgerald <rf(a)opensource.wolfsonmicro.com> --- src/utils/crec.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/src/utils/crec.c b/src/utils/crec.c index 258fade..e61e620 100644 --- a/src/utils/crec.c +++ b/src/utils/crec.c @@ -332,9 +332,6 @@ void capture_samples(char *name, unsigned int card, unsigned int device, fprintf(finfo, "%s: Capturing audio NOW!!!\n", __func__); do { - if (length && size > length - total_read) - size = length - total_read; - read = compress_read(compress, buffer, size); if (read < 0) { fprintf(stderr, "Error reading sample\n"); -- 1.9.1

2 1

[alsa-devel] M-audio 410 : complex procedure to get a sound
by ROUGE ETIENNE 06 Dec '15

06 Dec '15

Hello ! I've just bought a M-audio firewire 410 and I'm very thankfull to takashi's work that allows us to use the device almost like an USB one ! I can have sound and have access to the routing controls. However 2 bugs makes it a bit tedious : NOTE : I use an up to date freshly installed ubuntustudio 15.10. The only modifications I made was 1) deleting pulse 2) select the soundcard as default in asoundrc 1) I do not have sound out of the box, but I can manage to have sound by starting ffado-mixer then force restart alsa. After that everything works fine.... but I gotta do it at every restart which isn't nice 2) ffado-mixer suffers several bugs : - if I close the window I can't restart it again unless I kill ffado-dbus-server (it says : "another instance of ffado-mixer is already running). - I got few errors on startup "invalid samplerate for input"... don't know it it has any impact - when I tweak too much ffado-mixer (AFAIK, it's random, but happens within few minuts) the mouse shows a 'hand closed' cursor and I can't click anything. The system remains fully functionnal, but I can't close anything nor click... If I start an emergency terminal I can kill whatever I want, but the problem remains. I'll try with a fedora this evening to see if it's not related with ubuntustudio, but I'd be glad to have some help with those 2 issues (especially the stalling ffado-mixer) thanks ! Etienne

2 2

[alsa-devel] [RFC][PATCH 00/17 v2] ALSA: oxfw: refactoring and merging scs1x module
by Takashi Sakamoto 06 Dec '15

06 Dec '15

Hi, This patchset update my previous post for merging ALSA scs1x module to oxfw module: [alsa-devel] [RFC][PATCH 0/9] ALSA: oxfw: merge scs1x module http://mailman.alsa-project.org/pipermail/alsa-devel/2015-November/100534.h… This patchset consists of three parts: * 01-05: code refactoring for OXFW module * 06-09: code refactoring for old firewire-speaker functionalities in OXFW module * 10-17: merge scs1x module Changes: * split some patches for a fine granularity * drop a small framework I proposed in previous patch * rename functions for AV/C Audio Subunit command After this patchset is applied, OXFW module is loaded instead of scs1x module. Then, some names except for rawmidi ID are changed, see patch 16. Currently, I find no advantage to keep them what they were. I'm welcome to receive your oppositions about it. Takashi Sakamoto (17): ALSA: oxfw: rename a file for control elements so that it's for model-specific ALSA: oxfw: rename local functions for control elements so that they represent as local ALSA: oxfw: change function prototype for AV/C Audio Subunit command ALSA: oxfw: reuse driver entry to detect quirk ALSA: oxfw: gather model-dependent conditions to a function ALSA: oxfw: add memory allocation for model-specific structure ALSA: oxfw: move model-specific members from common structure ALSA: oxfw: move model-specific parameters from common structure ALSA: oxfw: rename a structure so that it means backward compatibility to old drivers ALSA: oxfw: add scs1x layer ALSA: oxfw: allocate own address region for SCS.1 series ALSA: oxfw: copy handlers of asynchronous transaction for MIDI capture ALSA: oxfw: add MIDI capture port for SCS.1 models ALSA: oxfw: copy handlers of asynchronous transaction for MIDI playback ALSA: oxfw: add MIDI playback port for SCS.1 models ALSA: oxfw: obsolete scs1x module ALSA: oxfw: add stream format quirk for SCS.1 models sound/firewire/Kconfig | 12 +- sound/firewire/Makefile | 2 - sound/firewire/oxfw/Makefile | 4 +- sound/firewire/oxfw/oxfw-scs1x.c | 409 +++++++++++++++++++++ .../firewire/oxfw/{oxfw-control.c => oxfw-spkr.c} | 142 ++++--- sound/firewire/oxfw/oxfw.c | 110 ++++-- sound/firewire/oxfw/oxfw.h | 23 +- 7 files changed, 586 insertions(+), 116 deletions(-) create mode 100644 sound/firewire/oxfw/oxfw-scs1x.c rename sound/firewire/oxfw/{oxfw-control.c => oxfw-spkr.c} (55%) -- 2.5.0

1 17