Convert DEVICE_ATTR uses to DEVICE_ATTR_RO where possible.

Done with perl script:

$ git grep -w --name-only DEVICE_ATTR | \
  xargs perl -i -e 'local $/; while (<>) { s/\bDEVICE_ATTR\s*\(\s*(\w+)\s*,\s*\(?(?:\s*S_IRUGO\s*|\s*0444\s*)\)?\s*,\s*\1_show\s*,\s*NULL\s*\)/DEVICE_ATTR_RO(\1)/g; print;}'
Signed-off-by: NJoe Perches <joe@perches.com>
Acked-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NRobert Jarzmik <robert.jarzmik@free.fr>
Acked-by: NSagi Grimberg <sagi@grimberg.me>
Acked-by: NZhang Rui <rui.zhang@intel.com>
Acked-by: NHarald Freudenberger <freude@linux.vnet.ibm.com>
Acked-by: NJani Nikula <jani.nikula@intel.com>
Acked-by: NCorey Minyard <cminyard@mvista.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Fix kernel-doc build error. A symbol that ends with an underscore
character ('_') has special meaning in reST (reStructuredText), so add
a '*' to prevent this error and to indicate that there are several of
these values to choose from.

../sound/soc/soc-core.c:2799: ERROR: Unknown target name: "snd_soc_daifmt".
Signed-off-by: NRandy Dunlap <rdunlap@infradead.org>
Signed-off-by: NMark Brown <broonie@kernel.org>

Codec will be replaced into Component, then Codec side
doesn't use legacy_dai_naming on snd_soc_register_dais().

This patch adds new non_legacy_dai_naming flag on Component driver
and use converted its value for snd_soc_register_dais().

When Codec is replaced into Component, Codec driver needs
to have non_legacy_dai_naming = 1 flags.
Existing CPU side of course doesn't have this flag, thus CPU calls
it as true.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Codec will be replaced into Component, then Codec side only
needs to call fixup_codec_formats() at this point.

This patch adds new endianness flag on Component driver
and call convert_endianness_formats() (= was fixup_codec_format())
if endianness was true.

When Codec is replaced into Component, Codec driver needs
to have endianness = 1 flags.
Existing CPU side of course doesn't have this flag, thus CPU doesn't
call it.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

No need to print an error message if kzalloc fails.
The core will print it.
Reported-by: NDaniel Baluta <daniel.baluta@gmail.com>
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

ALSA SoC platform/codec will be replaced to component soon.
But, some function exist in "platform" doesn't exist in "component".
Current soc-core has snd_soc_register_component(), but
doesn't have snd_soc_add_component() like snd_soc_add_platform().
This patch adds it.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

ALSA SoC platform/codec will be replaced to component soon.
This means 1 device might have multiple components. But current
unregister component function only checks "dev" to find it.
This means, unexpected component might be unregistered by current
function.
But, it is no problem if driver registered only 1 component.

To prepare avoid this issue, this patch adds new component
lookup function. it finds component by "dev" and "driver name".

Here, the reason why it uses "driver name" is that "component name"
was created by fmt_single_name() and difficult to use it from driver.
Driver of course knows its "driver name", thus, using it is more easy.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

In current ALSA SoC, Codec only has set_bias_level feature.
Codec will be merged into Component in next generation ALSA SoC,
thus current Codec specific feature need to be merged into it.
This is glue patch for it.

Codec driver has .idle_bias_off for dapm bias. But Component
driver doesn't have it, and dapm->idle_bias_off is set as "true".
To keep compatibility, this patch adds "idle_bias_on" instead of
".idle_bias_off" on Component driver.
dapm->idle_bias_off will be set by inverted idle_bias_on.

When we replace Codec to Component, the driver which has
".idle_bias_off = true" is just remove it,
and the driver which doesn't have it will have new
".idle_bias_on = true".
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

In current ALSA SoC, Platform only has pcm_new/pcm_free feature,
but it should be supported on Component level. This patch adds it.

The v1 was added commit 99b04f4c ("ASoC: add Component level
pcm_new/pcm_free") but it called all "card" connected component's
pcm_new/free, it was wrong.
This patch calls "rtd" connected component.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Double NULL pointer check for ops and ops->func is difficult to read
and might be forget to check it if new func was add.
This patch adds new null_snd_soc_ops and use it if rtd->dai_link didn't
have it to avoid NULL ops, and reduces ops NULL check.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

On soc_add_dai(), it is checking dai->driver->xxx,
This means, dai->driver is mandatory, never been NULL.
dai->driver check is not needed.
This patch removes it.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

static soc_free_pcm_runtime() is never called with rtd == NULL.
This patch removes unnecessary rtd == NULL check from
soc_free_pcm_runtime().

Furthermore, NULL check before kfree() is not needed.
This patch removes such ckeck too.
Reported-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NFengguang Wu <fengguang.wu@intel.com>
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

In current ALSA SoC, Codec only has set_jack feature.
Codec will be merged into Component in next generation ALSA SoC,
thus current Codec specific feature need to be merged into it.
This is glue patch for it.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

In current ALSA SoC, Codec only has set_pll feature.
Codec will be merged into Component in next generation ALSA SoC,
thus current Codec specific feature need to be merged into it.
This is glue patch for it.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

In current ALSA SoC, Codec only has set_sysclk feature.
Codec will be merged into Component in next generation ALSA SoC,
thus current Codec specific feature need to be merged into it.
This is glue patch for it.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Currently we're taking power_lock at each card component for assuring
the power-up sequence, but it doesn't help anything in the
implementation at the moment: it just serializes unnecessarily the
callers, but it doesn't protect about the power state change itself.
It used to have some usefulness in the early days where we managed the
PM manually.  But now the suspend/resume core procedure is beyond our
hands, and power_lock lost its meaning.

This patch drops the power_lock from allover the places.
There shouldn't be any issues by this change, as it's no helper
regarding the power state change.  Rather we'll get better performance
by removing the serialization; which is the only slight concern of any
behavior change, but it can't be a showstopper, after all.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

snd_soc_component and snd_soc_component_driver both have
dapm_routes/num_dapm_routes, but these are duplicated.
Let's remove duplicated definition.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

snd_soc_component and snd_soc_component_driver both have
dapm_widgets/num_dapm_widgets, but these are duplicated.
Let's remove duplicated definition.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

snd_soc_component and snd_soc_component_driver both have
controls/num_controls, but these are duplicated.
Let's remove duplicated definition.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Now we have snd_soc_component_get_dapm(),
and as soc.h say below, let's use it.

/* Don't use these, use snd_soc_component_get_dapm() */
struct snd_soc_dapm_context dapm;
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

The dai driver's name is allowed to be NULL. So add a sanity check for
that.
Signed-off-by: NJeffy Chen <jeffy.chen@rock-chips.com>
Reported-by: NDonglin Peng <dolinux.peng@gmail.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Currently we are searching dai name in snd_soc_find_dai, which could
either be dai driver name or component device name(for legacy naming).

Allow searching dai driver name in snd_soc_find_dai too, so that we can
use dai driver name to find legacy naming dais.
Signed-off-by: NJeffy Chen <jeffy.chen@rock-chips.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Current snd_soc_unregister_component() unregisters first found
component only which was specified by dev.
This style can't specify concrete component if system registered some
component with same dev. And system need to call this function many
times.
This patch unregister all related component by 1 call.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Basically, current ALSA SoC framework is based on CPU/Codec/Platform,
but its operation doesn't have consistent.
Thus, source code was unreadable, and difficult to understand.
This patch connects each component (= CPU/Codec/Platform) to rtd by
using snd_soc_rtdcom_add(), and convert uneven operations to consistent
operation.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Tested-by: NArnaud Pouliquen <arnaud.pouliquen@st.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Current snd_soc_pcm_runtime has platform / codec pointers, and we could
use these specific pointer. But these will be replaced to more generic
"component" soon, and will need more generic method to get each
connected component pointer from rtd.

This patch adds new snd_soc_rtdcom_xxx() to connect/disconnect
component to rtd. It means same as previous "platform" / "codec"
pointer style, but more generic.
We can find necessary component pointer from rtd by using component
driver name on snd_soc_rtdcom_lookup().

Here, the reason why it uses "driver name" is that "component name"
was created by fmt_single_name() and difficult to use it from driver.
Driver of course knows its "driver name", thus, using it is more easy.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Tested-by: NArnaud Pouliquen <arnaud.pouliquen@st.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Current snd_soc_unregister_component() is using multiple mutex_unlock()
for found/non-found cases. But it is unreadable and confusable code.
This patch tidyup current code to be readable.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

To unify notation, to readable.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

The check of IS_ERR or NULL pointer can be replaced by IS_ERR_OR_NULL(),
which helps readability.
Signed-off-by: NFabio Estevam <festevam@gmail.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

There is no need to have the *debugfs_reg dentry member as part of
the snd_soc_codec structure as its only usage is inside
soc_init_codec_debugfs().

Use a local dentry variable instead.
Signed-off-by: NFabio Estevam <festevam@gmail.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Fix inconsistent use of of_graph_get_port_parent() where
asoc_simple_card_parse_graph_dai() does of_node_get() before
calling it while other callers do not. We can fix this by
not trashing the node passed to of_graph_get_port_parent().

Let's also make sure the callers have correct refcounts and remove
related incorrect of_node_put() calls for of_for_each_phandle
as that's done by of_phandle_iterator_next() except when
we break out of the loop early.

Let's fix both issues with a single patch to avoid kobject
refcounts getting messed up more if two patches are merged
separately.

Otherwise strange issues can happen caused by memory corruption
caused by too many kobject_del() calls such as:

BUG: sleeping function called from invalid context at
kernel/locking/mutex.c:747
...
(___might_sleep)
(__mutex_lock)
(mutex_lock_nested)
(kernfs_remove)
(kobject_del)
(kobject_put)
(of_get_next_parent)
(of_graph_get_port_parent)
(asoc_simple_card_parse_graph_dai [snd_soc_simple_card_utils])
(asoc_graph_card_probe [snd_soc_audio_graph_card])

Fixes: 0ef472a9 ("of_graph: add of_graph_get_port_parent()")
Fixes: 2692c1c6 ("ASoC: add audio-graph-card support")
Fixes: 1689333f ("ASoC: simple-card-utils: add asoc_simple_card_parse_graph_dai()")
Signed-off-by: NTony Lindgren <tony@atomide.com>
Reviewed-by: NRob Herring <robh@kernel.org>
Tested-by: NAntonio Borneo <borneo.antonio@gmail.com>
Tested-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

This reverts commit 99b04f4c ("ASoC: add Component level
pcm_new/pcm_free"), which started calling the pcm_new callback for every
component in a *card* when creating a new pcm, something which does not
seem to make any sense.

This specifically led to memory leaks in systems with more than one
platform component and where DMA memory is allocated in the
platform-driver callback. For example, when both mcasp devices are being
used on an am335x board, DMA memory would be allocated twice for every
DAI link during probe.

When CONFIG_SND_VERBOSE_PROCFS was set this fortunately also led to
warnings such as:

WARNING: CPU: 0 PID: 565 at ../fs/proc/generic.c:346 proc_register+0x110/0x154
proc_dir_entry 'sub0/prealloc' already registered

Since there seems to be no users of the new component callbacks, and the
current implementation introduced a regression, let's revert the
offending commit for now.

Fixes: 99b04f4c ("ASoC: add Component level pcm_new/pcm_free")
Signed-off-by: NJohan Hovold <johan@kernel.org>
Reviewed-by: NLinus Walleij <linus.walleij@linaro.org>
Tested-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NMark Brown <broonie@kernel.org>
Cc: stable <stable@vger.kernel.org>	# 4.10

sound/soc/soc-core.c:1961:2-3: Unneeded semicolon

 Remove unneeded semicolon.

Generated by: scripts/coccinelle/misc/semicolon.cocci

Fixes: 98faf436 ("ASoC: Drop invalid DMI fields when setting card long name from DMI info")
CC: Mengdong Lin <mengdong.lin@linux.intel.com>
Signed-off-by: NFengguang Wu <fengguang.wu@intel.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Sometimes DMI fields may be invalid and so can't give useful vendor,
product or board info, such as "Type2 - Board Manufacturer" or
"Type1 - TBD by OEM". Including such invalid DMI fileds may create silly
card long name. So this patch creates a black list of invalid strings.
And if a DMI field contains any string in this list, it will be excluded
from the card long name.
Signed-off-by: NMengdong Lin <mengdong.lin@linux.intel.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

ALSA SoC needs to know connected DAI ID for detecting.
It is not a big problem if device/driver was only for sound,
but getting DAI ID will be difficult if device includes both
Video/Sound, like HDMI.
To solve this issue, this patch adds new snd_soc_get_dai_id() and
its related .of_xlate_dai_id callback on component driver.
In below case, we can handle Sound port (= port@2) as ID = 0
if .of_xlate_dai_id has its support.

	hdmi {
		port@0 { /* VIDEO */ };
		port@1 { /* VIDEO */ };
		port@2 { /* SOUND */ };
	};
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

soc_cleanup_card_resources() call snd_card_free() at the last of its
procedure.  This turned out to lead to a use-after-free.
PCM runtimes have been already removed via soc_remove_pcm_runtimes(),
while it's dereferenced later in soc_pcm_free() called via
snd_card_free().

The fix is simple: just move the snd_card_free() call to the beginning
of the whole procedure.  This also gives another benefit: it
guarantees that all operations have been shut down before actually
releasing the resources, which was racy until now.
Reported-and-tested-by: NRobert Jarzmik <robert.jarzmik@free.fr>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>
Signed-off-by: NMark Brown <broonie@kernel.org>
Cc: <stable@vger.kernel.org>

Current snd_soc_of_parse_daifmt() detects [prefix]format, but
"format" was unclear in some case. This patch checks "dai-format"
first, and try to check "[prefix]format" if "dai-format" was not
exist.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

This typo is quite common.  Fix it and add it to the spelling file so
that checkpatch catches it earlier.

Link: http://lkml.kernel.org/r/20170317011131.6881-2-sboyd@codeaurora.orgSigned-off-by: NStephen Boyd <sboyd@codeaurora.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Since recently UCM can pick up a configuration specific to the board
via card longname field, and we introduced a helper function
snd_soc_set_dmi_name() for that.  So far, it was used only in one
place (sound/soc/intel/boards/broadwell.c), but it should be more
widely applied.

This patch puts a big hammer for that: it lets snd_soc_register_card()
calling snd_soc_set_dmi_name() unconditionally, so that all x86
devices get the better longname string.  This would have no impact for
other systems without DMI support, as snd_soc_set_dmi_name() is no-op
on them.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>
Acked-by: NLiam Girdwood <liam.r.girdwood@linux.intel.com>
Acked-by: NVinod Koul <vinod.koul@intel.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

For systems without DMI, it makes no sense to have the code.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>
Acked-by: NLiam Girdwood <liam.r.girdwood@linux.intel.com>
Acked-by: NVinod Koul <vinod.koul@intel.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Current ALSA SoC Sound Card basically consists of CPU/Codec/Platform
components. If system uses Kernel modules, we can disable these drivers
by using rmmod command. In such case, we can't disable
CPU/Codec/Platform driver without disabling Sound Card driver.

But on the other hand, we can disable these drivers by using unbind
command. In such case, we can disable these drivers randomly.
In this case, we can create dirty Sound Card which is missing necessary
components.

(1) If user disabled Sound Card first, but did nothing to other drivers,
user can't use Sound because Sound Card is no longer exists.
(2) If user disabled CPU/Codec/Platform driver randomly, but did nothing
to Sound Card, user still be able to use Sound Card, because dirty Sound
Card still exists. In this case, Sound system will be crashed if user
started sound playback/capture. But we can't block such random unbind
now.

To avoid Sound Card crash in (2) case, we need to unregister Sound Card
whenever CPU/Codec/Platform component were unregistered.
This patch solves this issue.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>