user/sven/linux.git/include/sound, branch v4.20.13

ASoC: soc-core: fix init platform memory handling

2019-02-27T09:09:56Z

commit 09ac6a817bd687e7f5dac00470262efdd72f9319 upstream. snd_soc_init_platform initializes pointers to snd_soc_dai_link which is statically allocated and it does this by devm_kzalloc. In the event of an EPROBE_DEFER the memory will be freed and the pointers are left dangling. snd_soc_init_platform sees the dangling pointers and assumes they are pointing to initialized memory and does not reallocate them on the second probe attempt which results in a use after free bug since devm has freed the memory from the first probe attempt. Since the intention for snd_soc_dai_link->platform is that it can be set statically by the machine driver we need to respect the pointer in the event we did not set it but still catch dangling pointers. The solution is to add a flag to track whether the pointer was dynamically allocated or not. Signed-off-by: Curtis Malainey Signed-off-by: Mark Brown Cc: Jon Hunter Signed-off-by: Greg Kroah-Hartman

ALSA: hda - Serialize codec registrations

2019-02-12T19:02:35Z

commit 305a0ade180981686eec1f92aa6252a7c6ebb1cf upstream. In the current code, the codec registration may happen both at the codec bind time and the end of the controller probe time. In a rare occasion, they race with each other, leading to Oops due to the still uninitialized card device. This patch introduces a simple flag to prevent the codec registration at the codec bind time as long as the controller probe is going on. The controller probe invokes snd_card_register() that does the whole registration task, and we don't need to register each piece beforehand. Cc: Signed-off-by: Takashi Iwai Signed-off-by: Greg Kroah-Hartman

ALSA: compress: Fix stop handling on compressed capture streams

2019-02-12T19:02:34Z

commit 4f2ab5e1d13d6aa77c55f4914659784efd776eb4 upstream. It is normal user behaviour to start, stop, then start a stream again without closing it. Currently this works for compressed playback streams but not capture ones. The states on a compressed capture stream go directly from OPEN to PREPARED, unlike a playback stream which moves to SETUP and waits for a write of data before moving to PREPARED. Currently however, when a stop is sent the state is set to SETUP for both types of streams. This leaves a capture stream in the situation where a new start can't be sent as that requires the state to be PREPARED and a new set_params can't be sent as that requires the state to be OPEN. The only option being to close the stream, and then reopen. Correct this issues by allowing snd_compr_drain_notify to set the state depending on the stream direction, as we already do in set_params. Fixes: 49bb6402f1aa ("ALSA: compress_core: Add support for capture streams") Signed-off-by: Charles Keepax Cc: Signed-off-by: Takashi Iwai Signed-off-by: Greg Kroah-Hartman

ASoC: wm97xx: fix uninitialized regmap pointer problem

2019-01-26T08:20:39Z

[ Upstream commit 576ce4075bfa0f03e0e91a89eecc539b3b828b08 ] gcc notices that without either the ac97 bus or the pdata, we never initialize the regmap pointer, which leads to an uninitialized variable access: sound/soc/codecs/wm9712.c: In function 'wm9712_soc_probe': sound/soc/codecs/wm9712.c:666:2: error: 'regmap' may be used uninitialized in this function [-Werror=maybe-uninitialized] Since that configuration is invalid, it's better to return an error here. I tried to avoid adding complexity to the conditions, and turned the #ifdef into a regular if(IS_ENABLED()) check for readability. This in turn requires moving some header file declarations out of an #ifdef. The same code is used in three drivers, all of which I'm changing the same way. Fixes: 2ed1a8e0ce8d ("ASoC: wm9712: add ac97 new bus support") Signed-off-by: Arnd Bergmann Signed-off-by: Mark Brown Signed-off-by: Sasha Levin

ALSA: pcm: Fix interval evaluation with openmin/max

2018-11-29T11:05:19Z

As addressed in alsa-lib (commit b420056604f0), we need to fix the case where the evaluation of PCM interval "(x x+1]" leading to -EINVAL. After applying rules, such an interval may be translated as "(x x+1)". Fixes: ff2d6acdf6f1 ("ALSA: pcm: Fix snd_interval_refine first/last with open min/max") Cc: Signed-off-by: Takashi Iwai

Merge tag 'asoc-v4.20-rc4' of https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus

2018-11-27T15:06:42Z

ASoC: Fixes for v4.20 Lots of fixes here, the majority of which are driver specific but there's a couple of core things and one notable driver specific one: - A core fix for a DAPM regression introduced during the component refactoring, we'd lost the code that forced a reevaluation of the DAPM graph after probe (which we suppress during init to save lots of recalcuation) and have now restored it. - A core fix for error handling using the newly added for_each_rtd_codec_dai_rollback() macro. - A fix for the names of widgets in the newly introduced pcm3060 driver, merged as a fix so we don't have a release with legacy names.

ASoC: fix oops w/ for_each_rtd_codec_dai_rollback() macro

2018-10-30T15:31:11Z

A kernel oops happens on an error case (usual missing BE mixer configuration required by Intel SST driver). Git bisect points to this macro and an operator precedence issue. for (; ((i--) >= 0) && ((dai) = rtd->codec_dais[i]);) The initial code replaced by this macro was while (--i >= 0) { codec_dai = rtd->codec_dais[i]; Fix the C operator precedence difference by reverting to pre-decrement Fixes: 0b7990e3897 ('ASoC: add for_each_rtd_codec_dai() macro') Cc: Kuninori Morimoto Signed-off-by: Pierre-Louis Bossart Signed-off-by: Mark Brown

Merge tag 'asoc-v5.0' of https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus

2018-10-22T21:26:37Z

ASoC: Updates for v5.0/v4.20 As ever there's a lot of small and driver specific changes going on here, but we do also have some relatively large changes in the core thanks to the hard work of Charles and Morimoto-san: - More component transitions from Morimoto-san, I think we're about finished with this. Thanks for all the hard work! - Morimoto-san also added a bunch of for_each_foo macros - A bunch of cleanups and fixes for DAPM from Charles. - MCLK support for several different devices, including CS42L51, STM32 SAI, and MAX98373. - Support for Allwinner A64 CODEC analog, Intel boards with DA7219 and MAX98927, Meson AXG PDM inputs, Nuvoton NAU8822, Renesas R8A7744 and TI PCM3060.

ALSA: rawmidi: A lightweight function to discard pending bytes

2018-10-04T18:13:17Z

For discarding the pending bytes on rawmidi, we process with a loop of snd_rawmidi_transmit() which is just a waste of CPU power. Implement a lightweight API function to discard the pending bytes and the proceed the ring buffer instantly, and use it instead of open codes. Signed-off-by: Takashi Iwai

Merge branch 'for-linus' into for-next

2018-10-04T05:53:06Z

Back-merge 4.19-devel branch into 4.20 for applying FireWire patches cleanly. Signed-off-by: Takashi Iwai