The primary changes in this patch are cleanups of __check_input_term()
and move to a non-nested switch-case block by evaluating the pair of
UAC version and the unit type, as we've done for parse_audio_unit().
Also each parser is split into the function for readability.
Now, a slight behavior change by this cleanup is the handling of
processing and extension units. Formerly we've dealt with them
differently between UAC1/2 and UAC3; the latter returns an error if no
input sources are available, while the former continues to parse.
In this patch, unify the behavior in all cases: when input sources are
available, it parses recursively, then override the type and the id,
as well as channel information if not provided yet.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The input pool of a client might be deleted via the resize ioctl, the
the access to it should be covered by the proper locks. Currently the
only missing place is the call in snd_seq_ioctl_get_client_pool(), and
this patch papers over it.
Reported-by: syzbot+4a75454b9ca2777f35c7@syzkaller.appspotmail.com
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
It is easy to read code if it is cleanly using paired function/naming,
like start <-> stop, register <-> unregister, etc, etc.
But, current ALSA SoC code is very random, unbalance, not paired, etc.
It is easy to create bug at the such code, and it will be difficult to
debug.
From function name point of view, "soc_post_component_init()" sounds
like "component initialize function".
But in reality it is rtd setup function.
This patch renames soc_post_component_init() to soc_rtd_init()
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87v9uo7lc3.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Initialize component related list at random place is very difficult
to read. This patch initialize it at snd_soc_component_initialize().
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/87y2zozazp.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
sof_get_control_data returns negative values even though the return
value is defined unsigned (size_t). So change the return value type to
int and add the data size as pointer argument to sof_get_control_data to
avoid ambiquity in the meaning of the return type.
Fixes: cac974a51ebb ("ASoC: SOF: topology: use set_get_data in process load")
Reported by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Jaska Uimonen <jaska.uimonen@intel.com>
Link: https://lore.kernel.org/r/20190821211138.14618-1-jaska.uimonen@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
It might return without initializing in error case.
In such case, uninitialized variable might be used at error handler.
This patch initializes all necessary variable before return.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/87zhk4zazt.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
commit 52abe6cc1866a ("ASoC: topology: fix oops/use-after-free case
with dai driver") fixups remove_dai() error, but it is using
list_for_each_entry() for component->dai_list.
We already have for_each_component_dais() macro for it.
Let's use exising method.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87tvaczazd.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
There is no need to check return value for
soc_cleanup_card_resources(). Let't makes it as void.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/871rxg1lda.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
We can find specified name component via snd_soc_rtdcom_lookup().
But, it is not enough under multi CPU/Codec/Platform, because many
components which have same driver name might be connected to same rtd.
Not using this function as much as possible is best solution,
but some drivers are already deeply depended to it.
We can expand this function, for example having "num" which specifies
found order at parameter, etc (In such case, it need to have fixed
probing order).
Or, use different driver name in such component, etc.
We will have such issue if multi CPU/Codec/Platform were supported.
To indicate it, this patch adds NOTE to this function.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/874l2c1ldi.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
It is easy to read code if it is cleanly using paired function/naming,
like start <-> stop, register <-> unregister, etc, etc.
But, current ALSA SoC code is very random, unbalance, not paired, etc.
soc-core.c is using device_unregiser(), but there is no its paired
device_regiser(). We can find its code at soc_post_component_init()
which is using device_initialize() and device_add().
Here, device_initialize() + device_add() = device_register().
-- linux/drivers/base/core.c --
int device_register(struct device *dev)
{
device_initialize(dev);
return device_add(dev);
}
device_initialize() is doing each dev member's initialization only,
not related to device parent/release/groups.
Thus, we can postpone it.
let's use device_register() instead of device_initialize()/device_add().
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/878sro1ldw.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
sound/soc/mediatek/mt2701/mt2701-afe-common.h:66:27: warning:
mt2701_afe_backup_list defined but not used [-Wunused-const-variable=]
mt2701_afe_backup_list is only used in mt2701-afe-pcm.c,
so just move the definition over there.
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Link: https://lore.kernel.org/r/20190822143747.20944-1-yuehaibing@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>
sound/soc/codecs/wm8988.c:270:30: warning:
wm8988_rline_enum defined but not used [-Wunused-const-variable=]
wm8988_rline_enum should be used in wm8988_right_line_controls.
Reported-by: Hulk Robot <hulkci@huawei.com>
Fixes: 5409fb4e327a ("ASoC: Add WM8988 CODEC driver")
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Link: https://lore.kernel.org/r/20190822143608.59824-1-yuehaibing@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Pull USB validation patches. It's based on the latest 5.3 development
branch, so we shall catch up the whole things.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Now that we got the more comprehensive validation code for USB-audio
descriptors, the check of overflow in each descriptor unit parser
became superfluous. Drop some of the obvious cases.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Instead of the direct kfree() calls, introduce a new local helper to
release the usb_mixer_elem_info object. This will be extended to do
more than a single kfree() in the later patches.
Also, use the standard goto instead of multiple calls in
parse_audio_selector_unit() error paths.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Minor code refactoring by combining the UAC version and the type in
the switch-case flow, so that we reduce the indentation and
redundancy. One good bonus is that the duplicated definition of the
same type value (e.g. UAC2_EFFECT_UNIT) can be handled more cleanly.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Introduce a new helper to validate each audio descriptor unit before
and check the unit before actually accessing it. This should harden
against the OOB access cases with malformed descriptors that have been
recently frequently reported by fuzzers.
The existing descriptor checks are still kept although they become
superfluous after this patch. They'll be cleaned up eventually
later.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The bmControls (for UAC1) or bmMixerControls (for UAC2/3) bitmap has a
variable size depending on both input and output pins. Its size is to
fit with input * output bits. The problem is that the input size
can't be determined simply from the unit descriptor itself but it
needs to parse the whole connected sources. Although the
uac_mixer_unit_get_channels() tries to check some possible overflow of
this bitmap, it's incomplete due to the lack of the evaluation of
input pins.
For covering possible overflows, this patch adds the bitmap overflow
check in the loop of input pins in parse_audio_mixer_unit().
Fixes: 0bfe5e434e66 ("ALSA: usb-audio: Check mixer unit descriptors more strictly")
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Lenovo ThinkCentre M73 and M93 don't seem to have a proper beep
although the driver tries to probe and set up blindly.
Blacklist these machines for suppressing the beep creation.
BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=204635
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Some Realtek codec parsers didn't check the availability of PC beep.
Add has_cdefine_beep() check appropriately.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
I forgot to release the allocated object at the early error path in
line6_init_pcm(). For addressing it, slightly shuffle the code so
that the PCM destructor (pcm->private_free) is assigned properly
before all error paths.
Fixes: 3450121997ce ("ALSA: line6: Fix write on zero-sized buffer")
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
In addition to the I2S format, the controller also supports the DSP_*
formats.
This requires some extra care on the LRCK period calculation, since the
controller, with the PCM formats, require that the value set is no longer
the periods of LRCK for a single channel, but for all of them.
Let's add the code to deal with this, and support the DSP_A and DSP_B
formats.
Signed-off-by: Maxime Ripard <maxime.ripard@bootlin.com>
Link: https://lore.kernel.org/r/5562db1ac8759f12b1b87c3258223eed629ef771.1566392800.git-series.maxime.ripard@bootlin.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The LRCK period field in the FMT0 register holds the number of LRCK period
for one channel in I2S mode.
This has been hardcoded to 32, while it really should be the physical width
of the format, which creates an improper clock when using a 16bit format,
with the i2s controller as LRCK master.
Fixes: 7d2993811a1e ("ASoC: sun4i-i2s: Add support for H3")
Signed-off-by: Maxime Ripard <maxime.ripard@bootlin.com>
Link: https://lore.kernel.org/r/f08a0c3605cd1d79752b38d704690190183f7865.1566392800.git-series.maxime.ripard@bootlin.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The clock dividers function has been using the word size to compute the
clock rate at which it's supposed to be running, but the proper formula
would be to use the physical width and / or slot width in TDM.
It doesn't make any difference at the moment since all the formats
supported have the same sample width and physical width, but it's not going
to last forever.
Fixes: 7d2993811a1e ("ASoC: sun4i-i2s: Add support for H3")
Signed-off-by: Maxime Ripard <maxime.ripard@bootlin.com>
Link: https://lore.kernel.org/r/41a359d9885f397e066816961e5e3236afcbe0a1.1566392800.git-series.maxime.ripard@bootlin.com
Signed-off-by: Mark Brown <broonie@kernel.org>
snd_soc_remove_dai_link() has card connected dai_link check. but
1) we need to call list_del() anyway,
because it is "remove" function,
2) It is doing many thing for this card / dai_link already
before checking dai_link.
This patch removes poinless dai_link check
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/875zms1ldm.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
snd_soc_initialize_card_lists() is doing card related
INIT_LIST_HEAD(), but, it is already doing at
snd_soc_register_card(). We don't need to do it separately.
This patch merges these.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/877e781ldq.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
SAI module on imx8qm features a register map similar with imx6 series
(it doesn't have VERID and PARAM registers at the beginning
of address spece).
Also, it has one FIFO which can help up to 64 * 32 bit samples.
Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Link: https://lore.kernel.org/r/20190814082911.665-2-daniel.baluta@nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The quirk function snd_emuusb_set_samplerate() has a NULL check for
the mixer element, but this is useless in the current code. It used
to be a check against mixer->id_elems[unitid] but it was changed later
to the value after mixer_eleme_list_to_info() which is always non-NULL
due to the container_of() usage.
This patch fixes the check before the conversion.
While we're at it, correct a typo in the comment in the function,
too.
Fixes: 8c558076c740 ("ALSA: usb-audio: Clean up mixer element list traverse")
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This controller supports capture and playback running at the same time,
with the limitation that both capture and playback must be configured the
same way (sample rate, sample format, number of channels, etc). For this,
we have to assure that the configuration registers look the same when
capture and playback are initiated.
This patch fixes a bug in which the controller is in master mode and the
hw_params() callback fails for the second audio stream. The fail occurs
because the divisors are calculated after comparing the configuration
registers for capture and playback. The fix consists in calculating the
divisors before comparing the configuration registers. BCLK and LRC are
then configured and started only if the controller is not already running.
Fixes: 7e0cdf545a55 ("ASoC: mchp-i2s-mcc: add driver for I2SC Multi-Channel Controller")
Signed-off-by: Codrin Ciubotariu <codrin.ciubotariu@microchip.com>
Link: https://lore.kernel.org/r/20190820162411.24836-4-codrin.ciubotariu@microchip.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Since hw_free() can be called multiple times and not just after a stop
trigger command, we should check whether the RX or TX ready interrupt was
truly enabled previously. For this, we assure that the condition of the
wait event is always true, except when RX/TX interrupts are enabled.
Fixes: 7e0cdf545a55 ("ASoC: mchp-i2s-mcc: add driver for I2SC Multi-Channel Controller")
Signed-off-by: Codrin Ciubotariu <codrin.ciubotariu@microchip.com>
Link: https://lore.kernel.org/r/20190820162411.24836-3-codrin.ciubotariu@microchip.com
Signed-off-by: Mark Brown <broonie@kernel.org>
If hw_free() gets called after hw_params(), GCLK remains prepared,
preventing further use of it. This patch fixes this by unpreparing the
clock in hw_free() or if hw_params() gets an error.
Fixes: 7e0cdf545a55 ("ASoC: mchp-i2s-mcc: add driver for I2SC Multi-Channel Controller")
Signed-off-by: Codrin Ciubotariu <codrin.ciubotariu@microchip.com>
Link: https://lore.kernel.org/r/20190820162411.24836-2-codrin.ciubotariu@microchip.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The channels number have been hardcoded to 2 so far, while the controller
supports more than that.
Remove the instance where it has been hardcoded to compute the BCLK
divider, and pass it through as an argument to ease further support of more
channels.
Signed-off-by: Maxime Ripard <maxime.ripard@bootlin.com>
Link: https://lore.kernel.org/r/48887cf7abfaab6597db233b24d7a088a913e48a.1566242458.git-series.maxime.ripard@bootlin.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The LRCK polarity "normal" polarity in the I2S/TDM specs and in the
Allwinner datasheet are not the same. In the case where the i2s controller
is being used as the LRCK master, it's pretty clear when looked at under a
scope.
Let's fix this, and add a comment to clear up as much the confusion as
possible.
Fixes: 7d2993811a1e ("ASoC: sun4i-i2s: Add support for H3")
Fixes: 21faaea1343f ("ASoC: sun4i-i2s: Add support for A83T")
Signed-off-by: Maxime Ripard <maxime.ripard@bootlin.com>
Link: https://lore.kernel.org/r/e03fb6b2a916223070b9f18405b0ef117a452ff4.1566242458.git-series.maxime.ripard@bootlin.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The LRCK and BCLK polarity offsets on newer SoCs has been
changed, yet the driver didn't take it into account for all of them.
This was taken into account for the H3, but not the A83t. This was handled
using a reg_field for the H3.
However, the value in that field will not be the same, so reg_field is not
adapted in that case. Let's change for proper calls with the regular
values.
Fixes: 21faaea1343f ("ASoC: sun4i-i2s: Add support for A83T")
Signed-off-by: Maxime Ripard <maxime.ripard@bootlin.com>
Link: https://lore.kernel.org/r/9cbdde80a299288878e58225df4d7884e0301348.1566242458.git-series.maxime.ripard@bootlin.com
Signed-off-by: Mark Brown <broonie@kernel.org>
From: Marcus Cooper <codekipper@gmail.com>
The clock division dividers have changed between the older (A10/A31) and
newer (H3, A64, etc) SoCs.
While this was addressed through an offset on some SoCs, it was missing
some dividers as well, so the support wasn't perfect. Let's introduce a
pointer in the quirk structure for the divider calculation functions to use
so we can have the proper range now.
Signed-off-by: Marcus Cooper <codekipper@gmail.com>
[Maxime: Fix the commit log, use a field in the quirk structure]
Signed-off-by: Maxime Ripard <maxime.ripard@bootlin.com>
Link: https://lore.kernel.org/r/0e5b4abf06cd3202354315201c6af44caeb20236.1566242458.git-series.maxime.ripard@bootlin.com
Signed-off-by: Mark Brown <broonie@kernel.org>
On the first generation of Allwinner SoCs (A10-A31), the i2s controller was
using the MCLK as BCLK parent. However, this changed since the introduction
of the A83t and BCLK now uses the module clock as its parent.
Let's introduce a hook to get the parent rate and use that in our divider
calculations.
Fixes: 7d2993811a1e ("ASoC: sun4i-i2s: Add support for H3")
Fixes: 21faaea1343f ("ASoC: sun4i-i2s: Add support for A83T")
Signed-off-by: Maxime Ripard <maxime.ripard@bootlin.com>
Link: https://lore.kernel.org/r/0b6665be216b3bd0e7bc43724818f05f3f8ee881.1566242458.git-series.maxime.ripard@bootlin.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The two main generations of our I2S controller require a slightly different
format configuration, mostly because of a quite different register layout
and some additional registers being needed on the newer generation.
This used to be controlled through a bunch of booleans, however this proved
to be quite impractical, especially since a bunch of SoCs forgot to set
those parameters and therefore were broken from that point of view.
Fixes: 21faaea1343f ("ASoC: sun4i-i2s: Add support for A83T")
Signed-off-by: Maxime Ripard <maxime.ripard@bootlin.com>
Link: https://lore.kernel.org/r/dc818644c3e40734e7a97247c994b1fca1c3c047.1566242458.git-series.maxime.ripard@bootlin.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The two main generations of our I2S controller require a slightly different
channel configuration, mostly because of a quite different register layout
and some additional registers being needed on the newer generation.
This used to be controlled through a bunch of booleans, however this proved
to be quite impractical, especially since a bunch of SoCs forgot to set
those parameters and therefore were broken from that point of view.
Fixes: 21faaea1343f ("ASoC: sun4i-i2s: Add support for A83T")
Signed-off-by: Maxime Ripard <maxime.ripard@bootlin.com>
Link: https://lore.kernel.org/r/6414463de69584e8227fa495b13aa5f4798e1f0e.1566242458.git-series.maxime.ripard@bootlin.com
Signed-off-by: Mark Brown <broonie@kernel.org>
