ASoC: remove zte zx dangling kconfig

In commit dc98f1d we removed the zte zx sound drivers but there
was a dangling Kconfig left around for the codec so fix this.

Fixes: dc98f1d655ca ("ASoC: remove zte zx drivers")
Signed-off-by: Peter Robinson <pbrobinson@gmail.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Mark Brown <broonie@kernel.org>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/r/20210627105955.3410015-1-pbrobinson@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: qcom: lpass-cpu: mark IRQ_CLEAR register as volatile and readable

Currently IRQ_CLEAR register is marked as write-only, however using
regmap_update_bits on this register will have some side effects.
so mark IRQ_CLEAR register appropriately as readable and volatile.

Fixes: da0363f7bfd3 ("ASoC: qcom: Fix for DMA interrupt clear reg overwriting")
Reported-by: Marek Szyprowski <m.szyprowski@samsung.com>
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Tested-by: Srinivasa Rao Mandadapu <srivasam@codeaurora.org>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Link: https://lore.kernel.org/r/20210624092153.5771-1-srinivas.kandagatla@linaro.org
Signed-off-by: Mark Brown <broonie@kernel.org>

Merge series "ASoC: tlv320aic32x4: Add support for TAS2505" from Claudius Heine <ch@denx.de>:

Hi,

this is v2 from my patchset that add support for the TAS2505 to the tlv320aic32x4 driver.

kind regards,
Claudius

Changes from v1:
- clarified commit message of first patch, which add the type value to the struct
- removed unnecessary code to put and get speaker volume
- removed 'Gain' from 'HP Driver Playback Volume' control
- fixed rebase issues

Claudius Heine (3):
  ASoC: tlv320aic32x4: add type to device private data struct
  ASoC: tlv320aic32x4: add support for TAS2505
  ASoC: tlv320aic32x4: dt-bindings: add TAS2505 to compatible

 .../bindings/sound/tlv320aic32x4.txt          |   1 +
 sound/soc/codecs/tlv320aic32x4-i2c.c          |  22 ++-
 sound/soc/codecs/tlv320aic32x4-spi.c          |  23 ++-
 sound/soc/codecs/tlv320aic32x4.c              | 139 +++++++++++++++++-
 sound/soc/codecs/tlv320aic32x4.h              |  10 ++
 5 files changed, 186 insertions(+), 9 deletions(-)

base-commit: 70585216fe7730d9fb5453d3e2804e149d0fe201
--
2.32.0

Merge series "ASoC: tegra: Use devm_platform_get_and_ioremap_resource()" from Yang Yingliang <yangyingliang@huawei.com>:

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Yang Yingliang (4):
  ASoC: tegra20: i2s: Use devm_platform_get_and_ioremap_resource()
  ASoC: tegra20: spdif: Use devm_platform_get_and_ioremap_resource()
  ASoC: tegra: tegra210_admaif: Use
    devm_platform_get_and_ioremap_resource()
  ASoC: tegra30: ahub: Use devm_platform_get_and_ioremap_resource()

 sound/soc/tegra/tegra20_i2s.c     | 3 +--
 sound/soc/tegra/tegra20_spdif.c   | 3 +--
 sound/soc/tegra/tegra210_admaif.c | 4 +---
 sound/soc/tegra/tegra30_ahub.c    | 3 +--
 4 files changed, 4 insertions(+), 9 deletions(-)

--
2.25.1

ASoC: fsl: remove unnecessary oom message

Fixes scripts/checkpatch.pl warning:
WARNING: Possible unnecessary 'out of memory' message

Remove it can help us save a bit of memory.

Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
Link: https://lore.kernel.org/r/20210617103141.1765-1-thunder.leizhen@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: tlv320aic32x4: dt-bindings: add TAS2505 to compatible

This adds 'ti,tas2505' for TAS2505 to the list of allowed compatible
strings.

Signed-off-by: Claudius Heine <ch@denx.de>
Link: https://lore.kernel.org/r/20210617085230.1851503-4-ch@denx.de
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: tlv320aic32x4: add support for TAS2505

This adds support for TAS2505 and TAS2521 to the tlv320aic32x4 driver.

The TAS2505 seems to be a stripped down version of the TLV320AIC32X4 so
it makes sense to handle them in the same driver.

Signed-off-by: Claudius Heine <ch@denx.de>
Link: https://lore.kernel.org/r/20210617085230.1851503-3-ch@denx.de
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: tlv320aic32x4: add type to device private data struct

While this driver can already handle different device variants, the
variant information cannot be used in the driver code and therefor
cannot have different code paths depending on the device variant.

This change adds a `type` value into the `aic32x4_priv` structure, that
contains a device variant identifier, which was set when the driver was
bound to the device.

Signed-off-by: Claudius Heine <ch@denx.de>
Link: https://lore.kernel.org/r/20210617085230.1851503-2-ch@denx.de
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: tegra30: ahub: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210618024722.2618842-5-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: tegra: tegra210_admaif: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210618024722.2618842-4-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: tegra20: spdif: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210618024722.2618842-3-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: tegra20: i2s: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210618024722.2618842-2-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

Merge series "ASoC: Intel: machine driver corrections" from Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>:

Some of these patches dealing with Sparse warnings were submitted
earlier but not merged. I grouped them with a couple of fixes from
Kai.

v2:
added Mark Brown in CC, was missed in v1
added Richard Fitzgerald Tested-by

Kai Vehmanen (3):
  ASoC: Intel: sof_sdw: use mach data for ADL RVP DMIC count
  ASoC: Intel: sof_sdw: remove hdac-hdmi support
  ASoC: Intel: skl_hda_dsp_generic: Update Kconfig documentation

Pierre-Louis Bossart (7):
  ASoC: Intel: sof_rt5682: shrink platform_id names below 20 characters
  ASoC: Intel: glk_rt5682_max98357a: shrink platform_id below 20
    characters
  ASoC: Intel: kbl_da7219_max98357a: shrink platform_id below 20
    characters
  ASoC: Intel: sof_da7219_max98373: shrink platform_id below 20
    characters
  ASoC: Intel: sof_sdw: fix signed/unsigned warning
  ASoC: Intel: soc-acpi: add ull suffix for SoundWire _ADR values
  ASoC: Intel: use MODULE_DEVICE_TABLE with platform_device_id tables

 sound/soc/intel/boards/Kconfig                |  7 ++--
 sound/soc/intel/boards/bxt_da7219_max98357a.c |  4 +-
 sound/soc/intel/boards/bxt_rt298.c            |  3 +-
 sound/soc/intel/boards/ehl_rt5660.c           |  2 +-
 sound/soc/intel/boards/glk_rt5682_max98357a.c |  4 +-
 sound/soc/intel/boards/kbl_da7219_max98357a.c |  4 +-
 sound/soc/intel/boards/kbl_da7219_max98927.c  |  5 +--
 sound/soc/intel/boards/kbl_rt5660.c           |  2 +-
 sound/soc/intel/boards/kbl_rt5663_max98927.c  |  3 +-
 .../intel/boards/kbl_rt5663_rt5514_max98927.c |  2 +-
 .../soc/intel/boards/skl_nau88l25_max98357a.c |  3 +-
 sound/soc/intel/boards/skl_nau88l25_ssm4567.c |  3 +-
 sound/soc/intel/boards/skl_rt286.c            |  3 +-
 sound/soc/intel/boards/sof_cs42l42.c          |  3 +-
 sound/soc/intel/boards/sof_da7219_max98373.c  |  6 +--
 sound/soc/intel/boards/sof_rt5682.c           | 20 +++-------
 sound/soc/intel/boards/sof_sdw.c              |  5 +--
 sound/soc/intel/boards/sof_sdw_common.h       |  1 -
 sound/soc/intel/boards/sof_sdw_hdmi.c         | 37 +-----------------
 .../intel/common/soc-acpi-intel-adl-match.c   | 28 +++++++-------
 .../intel/common/soc-acpi-intel-cml-match.c   | 20 +++++-----
 .../intel/common/soc-acpi-intel-cnl-match.c   |  2 +-
 .../intel/common/soc-acpi-intel-glk-match.c   |  2 +-
 .../intel/common/soc-acpi-intel-icl-match.c   | 12 +++---
 .../intel/common/soc-acpi-intel-jsl-match.c   |  6 +--
 .../intel/common/soc-acpi-intel-kbl-match.c   |  2 +-
 .../intel/common/soc-acpi-intel-tgl-match.c   | 38 +++++++++----------
 27 files changed, 83 insertions(+), 144 deletions(-)

--
2.25.1

ASoC: fsl-asoc-card: change dev_err to dev_dbg for defer probe

Don't need to print error message for defer probe

Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Link: https://lore.kernel.org/r/1622616132-10391-1-git-send-email-shengjiu.wang@nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: simple-card: Fill in driver name

alsa-ucm groups by driver name so fill that in as well. Otherwise the
presented information is redundant and doesn't reflect the used
driver. We can't just use 'asoc-simple-card' since the driver name is
restricted to 15 characters.

Before:

 # cat /proc/asound/cards
 0 [Devkit         ]: Librem_5_Devkit - Librem 5 Devkit
                      Librem 5 Devkit
After:

 0 [Devkit         ]: simple-card - Librem 5 Devkit
                      Librem 5 Devkit

Signed-off-by: Guido Günther <agx@sigxcpu.org>
Link: https://lore.kernel.org/r/YNGe3akAntQi8qJD@qwark.sigxcpu.org
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: imx: Add missing of_node_put() in imx8_probe()

This node pointer is returned by of_parse_phandle() with refcount
incremented in this function. of_node_put() on it before exiting
this function.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210617032756.599359-1-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: codecs: wcd938x: fix uninitialized symbol warnings

This patch fixes below two uninitialized symbol warnings

warning:
sound/soc/codecs/wcd938x.c:2092 wcd938x_tx_swr_ctrl()
error: uninitialized symbol 'rate'

sound/soc/codecs/wcd938x.c:2189 wcd938x_tx_channel_config()
error: uninitialized symbol 'reg'.

First one my brining in check to already existing if condition and
second one by adding a default switch case to avoid any access to reg.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Link: https://lore.kernel.org/r/20210621134502.19537-2-srinivas.kandagatla@linaro.org
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: codecs: wcd938x: fix unused variable warning

This patch fixes below

warning: unused variable wcd938x_dt_match

by placing device match table under CONFIG_OF

Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Link: https://lore.kernel.org/r/20210621134502.19537-1-srinivas.kandagatla@linaro.org
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: skl_hda_dsp_generic: Update Kconfig documentation

The Kconfig documentation for SND_SOC_INTEL_SKL_HDA_DSP_GENERIC_MACH
is a bit misleading as it refers to a set of older platforms,
while in practise this machine driver supports all modern
Intel systems with Smart Sound Technology based DSP and HDA codecs.

Modify the Kconfig text to reflect current state.

Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20210621194057.21711-11-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: use MODULE_DEVICE_TABLE with platform_device_id tables

When we have a platform_device_id table, we can use
MODULE_DEVICE_TABLE to automatically generate the modalias. As a
result we can remove the manual insertion of MODULE_ALIAS.

Reported-by: Hulk Robot <hulkci@huawei.com>
Suggested-by: Zou Wei <zou_wei@huawei.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Link: https://lore.kernel.org/r/20210621194057.21711-10-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: soc-acpi: add ull suffix for SoundWire _ADR values

Sparse throws the following type of warnings:

sound/soc/intel/common/soc-acpi-intel-adl-match.c:34:24: error:
constant 0x000020025D071100 is so big it is long

Let's add the 'ull' suffix to make this go away and find real issues.

Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Paul Olaru <paul.olaru@oss.nxp.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Link: https://lore.kernel.org/r/20210621194057.21711-9-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: sof_sdw: fix signed/unsigned warning

Sparse throws the following warning:

sound/soc/intel/boards/sof_sdw.c:796:31: error: incorrect type in argument 6 (different signedness)
sound/soc/intel/boards/sof_sdw.c:796:31:    expected int *group_id
sound/soc/intel/boards/sof_sdw.c:796:31:    got unsigned int *

The group_id cannot be negative, use unsigned int.

Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Paul Olaru <paul.olaru@oss.nxp.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Link: https://lore.kernel.org/r/20210621194057.21711-8-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: sof_da7219_max98373: shrink platform_id below 20 characters

Sparse throws the following warning:

sound/soc/intel/boards/sof_da7219_max98373.c:438:25: error: too long
initializer-string for array of char(no space for nul char)

Fix by using 'mx' acronym for Maxim.

Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Paul Olaru <paul.olaru@oss.nxp.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Link: https://lore.kernel.org/r/20210621194057.21711-7-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: kbl_da7219_max98357a: shrink platform_id below 20 characters

Sparse throws the following warnings:

sound/soc/intel/boards/kbl_da7219_max98357a.c:647:25: error: too long
initializer-string for array of char(no space for nul char)

Fix by using the 'mx' acronym for Maxim.

Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Paul Olaru <paul.olaru@oss.nxp.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Link: https://lore.kernel.org/r/20210621194057.21711-6-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: glk_rt5682_max98357a: shrink platform_id below 20 characters

Sparse throws the following warning:

sound/soc/intel/boards/glk_rt5682_max98357a.c:622:25: error: too long
initializer-string for array of char(no space for nul char)

Fix by using the 'mx' acronym for Maxim

Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Paul Olaru <paul.olaru@oss.nxp.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Link: https://lore.kernel.org/r/20210621194057.21711-5-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: sof_rt5682: shrink platform_id names below 20 characters

Some Chromebooks machine driver aliases exceed 20 characters, which
leads to sparse warnings:

sound/soc/intel/boards/sof_rt5682.c:959:25: error: too long
initializer-string for array of char(no space for nul char)

sound/soc/intel/boards/sof_rt5682.c:989:25: error: too long
initializer-string for array of char(no space for nul char)

sound/soc/intel/boards/sof_rt5682.c:1039:25: error: too long
initializer-string for array of char(no space for nul char)

Fix by using the 'mx' shortcut for Maxim platforms (already used in
platform firmware)

Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Paul Olaru <paul.olaru@oss.nxp.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Link: https://lore.kernel.org/r/20210621194057.21711-4-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: sof_sdw: remove hdac-hdmi support

Remove support for using hdac_hdmi codec driver. No known products use
this configuration and hdac_hdmi cannot support all the platforms
sof_sdw does.

This change also fixes a bug in Kconfig rules.
SND_SOC_INTEL_SOUNDWIRE_SOF_MACH did not have a select SND_SOC_HDAC_HDMI
and this could cause build failures.

Reported-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Tested-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20210621194057.21711-3-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: sof_sdw: use mach data for ADL RVP DMIC count

On the reference boards, number of PCH dmics may vary and the number
should be taken from driver machine data. Remove the SOF_SDW_PCH_DMIC
quirk to make DMIC number configurable.

Fixes:d25bbe80485f8 ("ASoC: Intel: sof_sdw: add quirk for new ADL-P Rvp")

BugLink: https://github.com/thesofproject/sof/issues/4185
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20210621194057.21711-2-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

Merge series "ASoC: fsl: Use devm_platform_get_and_ioremap_resource()" from Yang Yingliang <yangyingliang@huawei.com>:

patch #1 ~ #8:
  Use devm_platform_get_and_ioremap_resource()

patch #9
  check return value of platform_get_resource_byname()

v2:
  change error message in patch #9

Yang Yingliang (9):
  ASoC: fsl_asrc: Use devm_platform_get_and_ioremap_resource()
  ASoC: fsl_aud2htx: Use devm_platform_get_and_ioremap_resource()
  ASoC: fsl_easrc: Use devm_platform_get_and_ioremap_resource()
  ASoC: fsl_esai: Use devm_platform_get_and_ioremap_resource()
  ASoC: fsl_micfil: Use devm_platform_get_and_ioremap_resource()
  ASoC: fsl_sai: Use devm_platform_get_and_ioremap_resource()
  ASoC: fsl_spdif: Use devm_platform_get_and_ioremap_resource()
  ASoC: fsl_ssi: Use devm_platform_get_and_ioremap_resource()
  ASoC: fsl_xcvr: check return value after calling
    platform_get_resource_byname()

 sound/soc/fsl/fsl_asrc.c    | 3 +--
 sound/soc/fsl/fsl_aud2htx.c | 3 +--
 sound/soc/fsl/fsl_easrc.c   | 3 +--
 sound/soc/fsl/fsl_esai.c    | 3 +--
 sound/soc/fsl/fsl_micfil.c  | 3 +--
 sound/soc/fsl/fsl_sai.c     | 3 +--
 sound/soc/fsl/fsl_spdif.c   | 3 +--
 sound/soc/fsl/fsl_ssi.c     | 3 +--
 sound/soc/fsl/fsl_xcvr.c    | 4 ++++
 9 files changed, 12 insertions(+), 16 deletions(-)

--
2.25.1

Merge series "ASoC: tidyup snd_soc_of_parse_daifmt()" from Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>:

Hi Mark

These are v3 of parsing for daifmt.

I want to add new audio-graph-card2 sound card driver,
and this is last part of necessary soc-core cleanup for it.

Current some drivers are using DT, and then,
snd_soc_of_parse_daifmt() parses daifmt, but bitclock/frame provider
parsing part is one of headache, because we are assuming below both cases.

A) node {
bitclock-master;
frame-master;
...
};

B) link {
bitclock-master = <&xxx>;
frame-master = <&xxx>;
...
};

The original was style A), and style B) was added later.

snd_soc_of_parse_daifmt() parses A) style as original style,
and user need to update to B) style for clock_provider part if needed.

To handle it more flexibile, this patch-set adds new functions
which separates snd_soc_of_parse_daifmt() helper function.

snd_soc_daifmt_parse_format() : format part
snd_soc_daifmt_parse_clock_provider_as_flag() : clock part for style A)
snd_soc_daifmt_parse_clock_provider_as_phandl() : clock part for style B)
snd_soc_daifmt_parse_clock_provider_as_bitmap() : clock part use with _from_bitmap

v1 -> v2
- tidyup parse_clock_provider functions to _as_flag/phandle/bitmap()
- don't exchange code style on each drivers.

v2 -> v3
- use daifmt as much as possible (don't use daiclk) on each driver.

Link: https://lore.kernel.org/r/875yypdxlm.wl-kuninori.morimoto.gx@renesas.com
Link: https://lore.kernel.org/r/87czsvdc4o.wl-kuninori.morimoto.gx@renesas.com
Kuninori Morimoto (8):
  ASoC: soc-core: add snd_soc_daifmt_clock_provider_from_bitmap()
  ASoC: soc-core: add snd_soc_daifmt_clock_provider_fliped()
  ASoC: soc-core: add snd_soc_daifmt_parse_format/clock_provider()
  ASoC: atmel: switch to use snd_soc_daifmt_parse_format/clock_provider()
  ASoC: fsl: switch to use snd_soc_daifmt_parse_format/clock_provider()
  ASoC: meson: switch to use snd_soc_daifmt_parse_format/clock_provider()
  ASoC: simple-card-utils: switch to use snd_soc_daifmt_parse_format/clock_provider()
  ASoC: soc-core: remove snd_soc_of_parse_daifmt()

 include/sound/soc.h                   |  21 ++++-
 sound/soc/atmel/mikroe-proto.c        |   9 ++-
 sound/soc/fsl/fsl-asoc-card.c         |   7 +-
 sound/soc/generic/simple-card-utils.c |  16 ++--
 sound/soc/meson/meson-card-utils.c    |   6 +-
 sound/soc/soc-core.c                  | 109 +++++++++++++++++---------
 6 files changed, 106 insertions(+), 62 deletions(-)

--
2.25.1

Merge series "ASoC: sunxi: Use devm_platform_get_and_ioremap_resource()" from Yang Yingliang <yangyingliang@huawei.com>:

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Yang Yingliang (3):
  ASoC: sunxi: sun4i-codec: Use devm_platform_get_and_ioremap_resource()
  ASoC: sun4i-i2s: Use devm_platform_get_and_ioremap_resource()
  ASoC: sunxi: sun4i-spdif: Use devm_platform_get_and_ioremap_resource()

 sound/soc/sunxi/sun4i-codec.c | 3 +--
 sound/soc/sunxi/sun4i-i2s.c   | 3 +--
 sound/soc/sunxi/sun4i-spdif.c | 3 +--
 3 files changed, 3 insertions(+), 6 deletions(-)

--
2.25.1

Merge series "ASoC: samsung: Use devm_platform_get_and_ioremap_resource()" from Yang Yingliang <yangyingliang@huawei.com>:

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Yang Yingliang (4):
  ASoC: samsung: i2s: Use devm_platform_get_and_ioremap_resource()
  ASoC: samsung: pcm: Use devm_platform_get_and_ioremap_resource()
  ASoC: samsung: s3c2412-i2s: Use
    devm_platform_get_and_ioremap_resource()
  ASoC: samsung: s3c24xx-i2s: Use
    devm_platform_get_and_ioremap_resource()

 sound/soc/samsung/i2s.c         | 3 +--
 sound/soc/samsung/pcm.c         | 3 +--
 sound/soc/samsung/s3c2412-i2s.c | 3 +--
 sound/soc/samsung/s3c24xx-i2s.c | 3 +--
 4 files changed, 4 insertions(+), 8 deletions(-)

--
2.25.1

ASoC: fsl_xcvr: disable all interrupts when suspend happens

There is an unhandled interrupt after suspend, which cause endless
interrupt when system resume, so system may hang.

Disable all interrupts in runtime suspend callback to avoid above
issue.

Fixes: 28564486866f ("ASoC: fsl_xcvr: Add XCVR ASoC CPU DAI driver")
Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Link: https://lore.kernel.org/r/1624019913-3380-1-git-send-email-shengjiu.wang@nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: tegra: Fix a NULL vs IS_ERR() check

The tegra_machine_parse_phandle() function doesn't return NULL, it returns
error pointers.

Fixes: cc8f70f56039 ("ASoC: tegra: Unify ASoC machine drivers")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Link: https://lore.kernel.org/r/YMyjOKFsPe9SietU@mwanda
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: qcom: lpass-cpu: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210615140711.1676704-2-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: qcom: apq8016_sbc: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210615140711.1676704-1-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rt711: add two jack detection modes

Some boards use different circuits for jack detection.
This patch adds two modes as below
1. JD2/2 ports/external resister 100k
2. JD2/1 port/JD voltage 1.8V

Signed-off-by: Shuming Fan <shumingf@realtek.com>
Link: https://lore.kernel.org/r/20210617090822.16960-1-shumingf@realtek.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: samsung: s3c24xx-i2s: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Link: https://lore.kernel.org/r/20210616091652.2552927-5-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: samsung: s3c2412-i2s: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Link: https://lore.kernel.org/r/20210616091652.2552927-4-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: samsung: pcm: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Link: https://lore.kernel.org/r/20210616091652.2552927-3-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: samsung: i2s: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Link: https://lore.kernel.org/r/20210616091652.2552927-2-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: soc-core: remove snd_soc_of_parse_daifmt()

No driver is using snd_soc_of_parse_daifmt().
This patch removes it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87zgvtuuro.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: simple-card-utils: switch to use snd_soc_daifmt_parse_format/clock_provider()

This patch switch to use snd_soc_daifmt_parse_format/clock_provider() from
snd_soc_of_parse_daifmt().

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/871r95w9ch.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: meson: switch to use snd_soc_daifmt_parse_format/clock_provider()

This patch switch to use snd_soc_daifmt_parse_format/clock_provider() from
snd_soc_of_parse_daifmt().

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: Jerome Brunet <jbrunet@baylibre.com>
Link: https://lore.kernel.org/r/8735tlw9cy.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: fsl: switch to use snd_soc_daifmt_parse_format/clock_provider()

This patch switch to use snd_soc_daifmt_parse_format/clock_provider() from
snd_soc_of_parse_daifmt().

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/874ke1w9dc.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: atmel: switch to use snd_soc_daifmt_parse_format/clock_provider()

This patch switch to use snd_soc_daifmt_parse_format/clock_provider() from
snd_soc_of_parse_daifmt().

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: Codrin Ciubotariu <codrin.ciubotariu@microchip.com>
Link: https://lore.kernel.org/r/875yyhw9e8.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: soc-core: add snd_soc_daifmt_parse_format/clock_provider()

snd_soc_of_parse_daifmt() parses daifmt, but bitclock/frame provider
parsing part is one of headacke, because we are assuming below both cases.

A) node {
bitclock-master;
frame-master;
...
};

B) link {
bitclock-master = <&xxx>;
frame-master = <&xxx>;
...
};

The original was style A), and style B) was added later
by commit b3ca11ff59bc ("ASoC: simple-card: Move dai-link level
properties away from dai subnodes").

snd_soc_of_parse_daifmt() parses it as style A),
and user need to update it to style B) if needed.

To handle it more flexibile, this patch adds new functions
which separates snd_soc_of_parse_daifmt() helper function.

 snd_soc_daifmt_parse_format()  :for DAI format
 snd_soc_daifmt_parse_clock_provider_as_flag()  :for style A)
 snd_soc_daifmt_parse_clock_provider_as_phandl() :for style B)
 snd_soc_daifmt_parse_clock_provider_as_bitmap() :use with _from_bitmap

This means

 snd_soc_of_parse_daifmt() ==
snd_soc_daifmt_parse_format() |
snd_soc_daifmt_parse_clock_provider_as_flag()

This patch also indicate relatesionship comment for
snd_soc_daifmt_clock_provider_from_bitmap().

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/877dixw9ej.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: soc-core: add snd_soc_daifmt_clock_provider_fliped()

Sometimes we want to get CLOCK_PROVIDER fliped dai_fmt.
This patch adds new snd_soc_daifmt_clock_provider_fliped() for it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/878s3dw9ex.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: soc-core: add snd_soc_daifmt_clock_provider_from_bitmap()

This patch adds snd_soc_daifmt_clock_provider_from_bitmap() function
to judge clock/frame master from its bitmap.
This is prepare for snd_soc_of_parse_daifmt() cleanup.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87a6ntw9f5.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: sunxi: sun4i-spdif: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210617045012.1119650-4-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: sun4i-i2s: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210617045012.1119650-3-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: sunxi: sun4i-codec: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210617045012.1119650-2-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: fsl_xcvr: check return value after calling platform_get_resource_byname()

It will cause null-ptr-deref if platform_get_resource_byname() returns NULL,
we need check the return value.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210615013922.784296-10-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: fsl_ssi: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210615013922.784296-9-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: fsl_spdif: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210615013922.784296-8-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: fsl_sai: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210615013922.784296-7-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: fsl_micfil: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210615013922.784296-6-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: fsl_esai: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210615013922.784296-5-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: fsl_easrc: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210615013922.784296-4-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: fsl_aud2htx: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210615013922.784296-3-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: fsl_asrc: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210615013922.784296-2-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rk817: Constify static struct snd_soc_dai_ops

The snd_soc_dai_ops structures is only stored in the ops field of a
snd_soc_dai_driver structure, so make the snd_soc_dai_ops structure
const to allow the compiler to put it in read-only memory.

Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Link: https://lore.kernel.org/r/20210602113643.3037374-1-weiyongjun1@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: fsi: fix spelling mistake

Fix "thse" --> "these" in struct fsi_stream declaration.

Signed-off-by: Flavio Suligoi <f.suligoi@asem.it>
Link: https://lore.kernel.org/r/20210618085324.1038524-1-f.suligoi@asem.it
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: bdw-rt5677: remove unnecessary oom message

Fixes scripts/checkpatch.pl warning:
WARNING: Possible unnecessary 'out of memory' message

Remove it can help us save a bit of memory.

Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20210617103538.1818-1-thunder.leizhen@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

Merge series "ASoC: stm32: Use devm_platform_get_and_ioremap_resource()" from Yang Yingliang <yangyingliang@huawei.com>:

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Yang Yingliang (3):
  ASoC: stm32: i2s: Use devm_platform_get_and_ioremap_resource()
  ASoC: stm32: sai: Use devm_platform_get_and_ioremap_resource()
  ASoC: stm32: spdifrx: Use devm_platform_get_and_ioremap_resource()

 sound/soc/stm/stm32_i2s.c     | 3 +--
 sound/soc/stm/stm32_sai_sub.c | 3 +--
 sound/soc/stm/stm32_spdifrx.c | 3 +--
 3 files changed, 3 insertions(+), 6 deletions(-)

--
2.25.1

ASoC: spear: spdif_out: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210617032900.600124-1-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: sprd: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210617033237.605808-1-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: soc-core: Fix the error return code in snd_soc_of_parse_audio_routing()

When devm_kcalloc() fails, the error code -ENOMEM should be returned
instead of -EINVAL.

Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
Link: https://lore.kernel.org/r/20210617103729.1918-1-thunder.leizhen@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: sti: sti_uniperif: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210617033903.613727-1-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: stm32: spdifrx: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210617043847.1113092-4-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: stm32: sai: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210617043847.1113092-3-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: stm32: i2s: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210617043847.1113092-2-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

Merge branch 'akpm' (patches from Andrew)

Merge misc fixes from Andrew Morton:
 "18 patches.

  Subsystems affected by this patch series: mm (memory-failure, swap,
  slub, hugetlb, memory-failure, slub, thp, sparsemem), and coredump"

* emailed patches from Andrew Morton <akpm@linux-foundation.org>:
  mm/sparse: fix check_usemap_section_nr warnings
  mm: thp: replace DEBUG_VM BUG with VM_WARN when unmap fails for split
  mm/thp: unmap_mapping_page() to fix THP truncate_cleanup_page()
  mm/thp: fix page_address_in_vma() on file THP tails
  mm/thp: fix vma_address() if virtual address below file offset
  mm/thp: try_to_unmap() use TTU_SYNC for safe splitting
  mm/thp: make is_huge_zero_pmd() safe and quicker
  mm/thp: fix __split_huge_pmd_locked() on shmem migration entry
  mm, thp: use head page in __migration_entry_wait()
  mm/slub.c: include swab.h
  crash_core, vmcoreinfo: append 'SECTION_SIZE_BITS' to vmcoreinfo
  mm/memory-failure: make sure wait for page writeback in memory_failure
  mm/hugetlb: expand restore_reserve_on_error functionality
  mm/slub: actually fix freelist pointer vs redzoning
  mm/slub: fix redzoning for small allocations
  mm/slub: clarify verification reporting
  mm/swap: fix pte_same_as_swp() not removing uffd-wp bit when compare
  mm,hwpoison: fix race with hugetlb page allocation

mm/sparse: fix check_usemap_section_nr warnings

I see a "virt_to_phys used for non-linear address" warning from
check_usemap_section_nr() on arm64 platforms.

In current implementation of NODE_DATA, if CONFIG_NEED_MULTIPLE_NODES=y,
pglist_data is dynamically allocated and assigned to node_data[].

For example, in arch/arm64/include/asm/mmzone.h:

  extern struct pglist_data *node_data[];
  #define NODE_DATA(nid)          (node_data[(nid)])

If CONFIG_NEED_MULTIPLE_NODES=n, pglist_data is defined as a global
variable named "contig_page_data".

For example, in include/linux/mmzone.h:

  extern struct pglist_data contig_page_data;
  #define NODE_DATA(nid)          (&contig_page_data)

If CONFIG_DEBUG_VIRTUAL is not enabled, __pa() can handle both
dynamically allocated linear addresses and symbol addresses.  However,
if (CONFIG_DEBUG_VIRTUAL=y && CONFIG_NEED_MULTIPLE_NODES=n) we can see
the "virt_to_phys used for non-linear address" warning because that
&contig_page_data is not a linear address on arm64.

Warning message:

  virt_to_phys used for non-linear address: (contig_page_data+0x0/0x1c00)
  WARNING: CPU: 0 PID: 0 at arch/arm64/mm/physaddr.c:15 __virt_to_phys+0x58/0x68
  Modules linked in:
  CPU: 0 PID: 0 Comm: swapper Tainted: G        W         5.13.0-rc1-00074-g1140ab592e2e #3
  Hardware name: linux,dummy-virt (DT)
  pstate: 600000c5 (nZCv daIF -PAN -UAO -TCO BTYPE=--)
  Call trace:
     __virt_to_phys+0x58/0x68
     check_usemap_section_nr+0x50/0xfc
     sparse_init_nid+0x1ac/0x28c
     sparse_init+0x1c4/0x1e0
     bootmem_init+0x60/0x90
     setup_arch+0x184/0x1f0
     start_kernel+0x78/0x488

To fix it, create a small function to handle both translation.

Link: https://lkml.kernel.org/r/1623058729-27264-1-git-send-email-miles.chen@mediatek.com
Signed-off-by: Miles Chen <miles.chen@mediatek.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Baoquan He <bhe@redhat.com>
Cc: Kazu <k-hagio-ab@nec.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm: thp: replace DEBUG_VM BUG with VM_WARN when unmap fails for split

When debugging the bug reported by Wang Yugui [1], try_to_unmap() may
fail, but the first VM_BUG_ON_PAGE() just checks page_mapcount() however
it may miss the failure when head page is unmapped but other subpage is
mapped.  Then the second DEBUG_VM BUG() that check total mapcount would
catch it.  This may incur some confusion.

As this is not a fatal issue, so consolidate the two DEBUG_VM checks
into one VM_WARN_ON_ONCE_PAGE().

[1] https://lore.kernel.org/linux-mm/20210412180659.B9E3.409509F4@e16-tech.com/

Link: https://lkml.kernel.org/r/d0f0db68-98b8-ebfb-16dc-f29df24cf012@google.com
Signed-off-by: Yang Shi <shy828301@gmail.com>
Reviewed-by: Zi Yan <ziy@nvidia.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Hugh Dickins <hughd@google.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jue Wang <juew@google.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Wang Yugui <wangyugui@e16-tech.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/thp: unmap_mapping_page() to fix THP truncate_cleanup_page()

There is a race between THP unmapping and truncation, when truncate sees
pmd_none() and skips the entry, after munmap's zap_huge_pmd() cleared
it, but before its page_remove_rmap() gets to decrement
compound_mapcount: generating false "BUG: Bad page cache" reports that
the page is still mapped when deleted.  This commit fixes that, but not
in the way I hoped.

The first attempt used try_to_unmap(page, TTU_SYNC|TTU_IGNORE_MLOCK)
instead of unmap_mapping_range() in truncate_cleanup_page(): it has
often been an annoyance that we usually call unmap_mapping_range() with
no pages locked, but there apply it to a single locked page.
try_to_unmap() looks more suitable for a single locked page.

However, try_to_unmap_one() contains a VM_BUG_ON_PAGE(!pvmw.pte,page):
it is used to insert THP migration entries, but not used to unmap THPs.
Copy zap_huge_pmd() and add THP handling now? Perhaps, but their TLB
needs are different, I'm too ignorant of the DAX cases, and couldn't
decide how far to go for anon+swap.  Set that aside.

The second attempt took a different tack: make no change in truncate.c,
but modify zap_huge_pmd() to insert an invalidated huge pmd instead of
clearing it initially, then pmd_clear() between page_remove_rmap() and
unlocking at the end.  Nice.  But powerpc blows that approach out of the
water, with its serialize_against_pte_lookup(), and interesting pgtable
usage.  It would need serious help to get working on powerpc (with a
minor optimization issue on s390 too).  Set that aside.

Just add an "if (page_mapped(page)) synchronize_rcu();" or other such
delay, after unmapping in truncate_cleanup_page()? Perhaps, but though
that's likely to reduce or eliminate the number of incidents, it would
give less assurance of whether we had identified the problem correctly.

This successful iteration introduces "unmap_mapping_page(page)" instead
of try_to_unmap(), and goes the usual unmap_mapping_range_tree() route,
with an addition to details.  Then zap_pmd_range() watches for this
case, and does spin_unlock(pmd_lock) if so - just like
page_vma_mapped_walk() now does in the PVMW_SYNC case.  Not pretty, but
safe.

Note that unmap_mapping_page() is doing a VM_BUG_ON(!PageLocked) to
assert its interface; but currently that's only used to make sure that
page->mapping is stable, and zap_pmd_range() doesn't care if the page is
locked or not.  Along these lines, in invalidate_inode_pages2_range()
move the initial unmap_mapping_range() out from under page lock, before
then calling unmap_mapping_page() under page lock if still mapped.

Link: https://lkml.kernel.org/r/a2a4a148-cdd8-942c-4ef8-51b77f643dbe@google.com
Fixes: fc127da085c2 ("truncate: handle file thp")
Signed-off-by: Hugh Dickins <hughd@google.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Reviewed-by: Yang Shi <shy828301@gmail.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jue Wang <juew@google.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Wang Yugui <wangyugui@e16-tech.com>
Cc: Zi Yan <ziy@nvidia.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/thp: fix page_address_in_vma() on file THP tails

Anon THP tails were already supported, but memory-failure may need to
use page_address_in_vma() on file THP tails, which its page->mapping
check did not permit: fix it.

hughd adds: no current usage is known to hit the issue, but this does
fix a subtle trap in a general helper: best fixed in stable sooner than
later.

Link: https://lkml.kernel.org/r/a0d9b53-bf5d-8bab-ac5-759dc61819c1@google.com
Fixes: 800d8c63b2e9 ("shmem: add huge pages support")
Signed-off-by: Jue Wang <juew@google.com>
Signed-off-by: Hugh Dickins <hughd@google.com>
Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Yang Shi <shy828301@gmail.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Wang Yugui <wangyugui@e16-tech.com>
Cc: Zi Yan <ziy@nvidia.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/thp: fix vma_address() if virtual address below file offset

Running certain tests with a DEBUG_VM kernel would crash within hours,
on the total_mapcount BUG() in split_huge_page_to_list(), while trying
to free up some memory by punching a hole in a shmem huge page: split's
try_to_unmap() was unable to find all the mappings of the page (which,
on a !DEBUG_VM kernel, would then keep the huge page pinned in memory).

When that BUG() was changed to a WARN(), it would later crash on the
VM_BUG_ON_VMA(end < vma->vm_start || start >= vma->vm_end, vma) in
mm/internal.h:vma_address(), used by rmap_walk_file() for
try_to_unmap().

vma_address() is usually correct, but there's a wraparound case when the
vm_start address is unusually low, but vm_pgoff not so low:
vma_address() chooses max(start, vma->vm_start), but that decides on the
wrong address, because start has become almost ULONG_MAX.

Rewrite vma_address() to be more careful about vm_pgoff; move the
VM_BUG_ON_VMA() out of it, returning -EFAULT for errors, so that it can
be safely used from page_mapped_in_vma() and page_address_in_vma() too.

Add vma_address_end() to apply similar care to end address calculation,
in page_vma_mapped_walk() and page_mkclean_one() and try_to_unmap_one();
though it raises a question of whether callers would do better to supply
pvmw->end to page_vma_mapped_walk() - I chose not, for a smaller patch.

An irritation is that their apparent generality breaks down on KSM
pages, which cannot be located by the page->index that page_to_pgoff()
uses: as commit 4b0ece6fa016 ("mm: migrate: fix remove_migration_pte()
for ksm pages") once discovered.  I dithered over the best thing to do
about that, and have ended up with a VM_BUG_ON_PAGE(PageKsm) in both
vma_address() and vma_address_end(); though the only place in danger of
using it on them was try_to_unmap_one().

Sidenote: vma_address() and vma_address_end() now use compound_nr() on a
head page, instead of thp_size(): to make the right calculation on a
hugetlbfs page, whether or not THPs are configured.  try_to_unmap() is
used on hugetlbfs pages, but perhaps the wrong calculation never
mattered.

Link: https://lkml.kernel.org/r/caf1c1a3-7cfb-7f8f-1beb-ba816e932825@google.com
Fixes: a8fa41ad2f6f ("mm, rmap: check all VMAs that PTE-mapped THP can be part of")
Signed-off-by: Hugh Dickins <hughd@google.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jue Wang <juew@google.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Wang Yugui <wangyugui@e16-tech.com>
Cc: Yang Shi <shy828301@gmail.com>
Cc: Zi Yan <ziy@nvidia.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/thp: try_to_unmap() use TTU_SYNC for safe splitting

Stressing huge tmpfs often crashed on unmap_page()'s VM_BUG_ON_PAGE
(!unmap_success): with dump_page() showing mapcount:1, but then its raw
struct page output showing _mapcount ffffffff i.e.  mapcount 0.

And even if that particular VM_BUG_ON_PAGE(!unmap_success) is removed,
it is immediately followed by a VM_BUG_ON_PAGE(compound_mapcount(head)),
and further down an IS_ENABLED(CONFIG_DEBUG_VM) total_mapcount BUG():
all indicative of some mapcount difficulty in development here perhaps.
But the !CONFIG_DEBUG_VM path handles the failures correctly and
silently.

I believe the problem is that once a racing unmap has cleared pte or
pmd, try_to_unmap_one() may skip taking the page table lock, and emerge
from try_to_unmap() before the racing task has reached decrementing
mapcount.

Instead of abandoning the unsafe VM_BUG_ON_PAGE(), and the ones that
follow, use PVMW_SYNC in try_to_unmap_one() in this case: adding
TTU_SYNC to the options, and passing that from unmap_page().

When CONFIG_DEBUG_VM, or for non-debug too? Consensus is to do the same
for both: the slight overhead added should rarely matter, except perhaps
if splitting sparsely-populated multiply-mapped shmem.  Once confident
that bugs are fixed, TTU_SYNC here can be removed, and the race
tolerated.

Link: https://lkml.kernel.org/r/c1e95853-8bcd-d8fd-55fa-e7f2488e78f@google.com
Fixes: fec89c109f3a ("thp: rewrite freeze_page()/unfreeze_page() with generic rmap walkers")
Signed-off-by: Hugh Dickins <hughd@google.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jue Wang <juew@google.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Wang Yugui <wangyugui@e16-tech.com>
Cc: Yang Shi <shy828301@gmail.com>
Cc: Zi Yan <ziy@nvidia.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/thp: make is_huge_zero_pmd() safe and quicker

Most callers of is_huge_zero_pmd() supply a pmd already verified
present; but a few (notably zap_huge_pmd()) do not - it might be a pmd
migration entry, in which the pfn is encoded differently from a present
pmd: which might pass the is_huge_zero_pmd() test (though not on x86,
since L1TF forced us to protect against that); or perhaps even crash in
pmd_page() applied to a swap-like entry.

Make it safe by adding pmd_present() check into is_huge_zero_pmd()
itself; and make it quicker by saving huge_zero_pfn, so that
is_huge_zero_pmd() will not need to do that pmd_page() lookup each time.

__split_huge_pmd_locked() checked pmd_trans_huge() before: that worked,
but is unnecessary now that is_huge_zero_pmd() checks present.

Link: https://lkml.kernel.org/r/21ea9ca-a1f5-8b90-5e88-95fb1c49bbfa@google.com
Fixes: e71769ae5260 ("mm: enable thp migration for shmem thp")
Signed-off-by: Hugh Dickins <hughd@google.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Reviewed-by: Yang Shi <shy828301@gmail.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jue Wang <juew@google.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Wang Yugui <wangyugui@e16-tech.com>
Cc: Zi Yan <ziy@nvidia.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/thp: fix __split_huge_pmd_locked() on shmem migration entry

Patch series "mm/thp: fix THP splitting unmap BUGs and related", v10.

Here is v2 batch of long-standing THP bug fixes that I had not got
around to sending before, but prompted now by Wang Yugui's report
https://lore.kernel.org/linux-mm/20210412180659.B9E3.409509F4@e16-tech.com/

Wang Yugui has tested a rollup of these fixes applied to 5.10.39, and
they have done no harm, but have *not* fixed that issue: something more
is needed and I have no idea of what.

This patch (of 7):

Stressing huge tmpfs page migration racing hole punch often crashed on
the VM_BUG_ON(!pmd_present) in pmdp_huge_clear_flush(), with DEBUG_VM=y
kernel; or shortly afterwards, on a bad dereference in
__split_huge_pmd_locked() when DEBUG_VM=n.  They forgot to allow for pmd
migration entries in the non-anonymous case.

Full disclosure: those particular experiments were on a kernel with more
relaxed mmap_lock and i_mmap_rwsem locking, and were not repeated on the
vanilla kernel: it is conceivable that stricter locking happens to avoid
those cases, or makes them less likely; but __split_huge_pmd_locked()
already allowed for pmd migration entries when handling anonymous THPs,
so this commit brings the shmem and file THP handling into line.

And while there: use old_pmd rather than _pmd, as in the following
blocks; and make it clearer to the eye that the !vma_is_anonymous()
block is self-contained, making an early return after accounting for
unmapping.

Link: https://lkml.kernel.org/r/af88612-1473-2eaa-903-8d1a448b26@google.com
Link: https://lkml.kernel.org/r/dd221a99-efb3-cd1d-6256-7e646af29314@google.com
Fixes: e71769ae5260 ("mm: enable thp migration for shmem thp")
Signed-off-by: Hugh Dickins <hughd@google.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Yang Shi <shy828301@gmail.com>
Cc: Wang Yugui <wangyugui@e16-tech.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: Zi Yan <ziy@nvidia.com>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Jue Wang <juew@google.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm, thp: use head page in __migration_entry_wait()

We notice that hung task happens in a corner but practical scenario when
CONFIG_PREEMPT_NONE is enabled, as follows.

Process 0                       Process 1                     Process 2..Inf
split_huge_page_to_list
    unmap_page
        split_huge_pmd_address
                                __migration_entry_wait(head)
                                                              __migration_entry_wait(tail)
    remap_page (roll back)
        remove_migration_ptes
            rmap_walk_anon
                cond_resched

Where __migration_entry_wait(tail) is occurred in kernel space, e.g.,
copy_to_user in fstat, which will immediately fault again without
rescheduling, and thus occupy the cpu fully.

When there are too many processes performing __migration_entry_wait on
tail page, remap_page will never be done after cond_resched.

This makes __migration_entry_wait operate on the compound head page,
thus waits for remap_page to complete, whether the THP is split
successfully or roll back.

Note that put_and_wait_on_page_locked helps to drop the page reference
acquired with get_page_unless_zero, as soon as the page is on the wait
queue, before actually waiting.  So splitting the THP is only prevented
for a brief interval.

Link: https://lkml.kernel.org/r/b9836c1dd522e903891760af9f0c86a2cce987eb.1623144009.git.xuyu@linux.alibaba.com
Fixes: ba98828088ad ("thp: add option to setup migration entries during PMD split")
Suggested-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Gang Deng <gavin.dg@linux.alibaba.com>
Signed-off-by: Xu Yu <xuyu@linux.alibaba.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Hugh Dickins <hughd@google.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/slub.c: include swab.h

Fixes build with CONFIG_SLAB_FREELIST_HARDENED=y.

Hopefully.  But it's the right thing to do anwyay.

Fixes: 1ad53d9fa3f61 ("slub: improve bit diffusion for freelist ptr obfuscation")
Link: https://bugzilla.kernel.org/show_bug.cgi?id=213417
Reported-by: <vannguye@cisco.com>
Acked-by: Kees Cook <keescook@chromium.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

crash_core, vmcoreinfo: append 'SECTION_SIZE_BITS' to vmcoreinfo

As mentioned in kernel commit 1d50e5d0c505 ("crash_core, vmcoreinfo:
Append 'MAX_PHYSMEM_BITS' to vmcoreinfo"), SECTION_SIZE_BITS in the
formula:

    #define SECTIONS_SHIFT    (MAX_PHYSMEM_BITS - SECTION_SIZE_BITS)

Besides SECTIONS_SHIFT, SECTION_SIZE_BITS is also used to calculate
PAGES_PER_SECTION in makedumpfile just like kernel.

Unfortunately, this arch-dependent macro SECTION_SIZE_BITS changes, e.g.
recently in kernel commit f0b13ee23241 ("arm64/sparsemem: reduce
SECTION_SIZE_BITS").  But user space wants a stable interface to get
this info.  Such info is impossible to be deduced from a crashdump
vmcore.  Hence append SECTION_SIZE_BITS to vmcoreinfo.

Link: https://lkml.kernel.org/r/20210608103359.84907-1-kernelfans@gmail.com
Link: http://lists.infradead.org/pipermail/kexec/2021-June/022676.html
Signed-off-by: Pingfan Liu <kernelfans@gmail.com>
Acked-by: Baoquan He <bhe@redhat.com>
Cc: Bhupesh Sharma <bhupesh.sharma@linaro.org>
Cc: Kazuhito Hagio <k-hagio@ab.jp.nec.com>
Cc: Dave Young <dyoung@redhat.com>
Cc: Boris Petkov <bp@alien8.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: James Morse <james.morse@arm.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Dave Anderson <anderson@redhat.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/memory-failure: make sure wait for page writeback in memory_failure

Our syzkaller trigger the "BUG_ON(!list_empty(&inode->i_wb_list))" in
clear_inode:

  kernel BUG at fs/inode.c:519!
  Internal error: Oops - BUG: 0 [#1] SMP
  Modules linked in:
  Process syz-executor.0 (pid: 249, stack limit = 0x00000000a12409d7)
  CPU: 1 PID: 249 Comm: syz-executor.0 Not tainted 4.19.95
  Hardware name: linux,dummy-virt (DT)
  pstate: 80000005 (Nzcv daif -PAN -UAO)
  pc : clear_inode+0x280/0x2a8
  lr : clear_inode+0x280/0x2a8
  Call trace:
    clear_inode+0x280/0x2a8
    ext4_clear_inode+0x38/0xe8
    ext4_free_inode+0x130/0xc68
    ext4_evict_inode+0xb20/0xcb8
    evict+0x1a8/0x3c0
    iput+0x344/0x460
    do_unlinkat+0x260/0x410
    __arm64_sys_unlinkat+0x6c/0xc0
    el0_svc_common+0xdc/0x3b0
    el0_svc_handler+0xf8/0x160
    el0_svc+0x10/0x218
  Kernel panic - not syncing: Fatal exception

A crash dump of this problem show that someone called __munlock_pagevec
to clear page LRU without lock_page: do_mmap -> mmap_region -> do_munmap
-> munlock_vma_pages_range -> __munlock_pagevec.

As a result memory_failure will call identify_page_state without
wait_on_page_writeback.  And after truncate_error_page clear the mapping
of this page.  end_page_writeback won't call sb_clear_inode_writeback to
clear inode->i_wb_list.  That will trigger BUG_ON in clear_inode!

Fix it by checking PageWriteback too to help determine should we skip
wait_on_page_writeback.

Link: https://lkml.kernel.org/r/20210604084705.3729204-1-yangerkun@huawei.com
Fixes: 0bc1f8b0682c ("hwpoison: fix the handling path of the victimized page frame that belong to non-LRU")
Signed-off-by: yangerkun <yangerkun@huawei.com>
Acked-by: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Yu Kuai <yukuai3@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/hugetlb: expand restore_reserve_on_error functionality

The routine restore_reserve_on_error is called to restore reservation
information when an error occurs after page allocation.  The routine
alloc_huge_page modifies the mapping reserve map and potentially the
reserve count during allocation.  If code calling alloc_huge_page
encounters an error after allocation and needs to free the page, the
reservation information needs to be adjusted.

Currently, restore_reserve_on_error only takes action on pages for which
the reserve count was adjusted(HPageRestoreReserve flag).  There is
nothing wrong with these adjustments.  However, alloc_huge_page ALWAYS
modifies the reserve map during allocation even if the reserve count is
not adjusted.  This can cause issues as observed during development of
this patch [1].

One specific series of operations causing an issue is:

 - Create a shared hugetlb mapping
   Reservations for all pages created by default

 - Fault in a page in the mapping
   Reservation exists so reservation count is decremented

 - Punch a hole in the file/mapping at index previously faulted
   Reservation and any associated pages will be removed

 - Allocate a page to fill the hole
   No reservation entry, so reserve count unmodified
   Reservation entry added to map by alloc_huge_page

 - Error after allocation and before instantiating the page
   Reservation entry remains in map

 - Allocate a page to fill the hole
   Reservation entry exists, so decrement reservation count

This will cause a reservation count underflow as the reservation count
was decremented twice for the same index.

A user would observe a very large number for HugePages_Rsvd in
/proc/meminfo.  This would also likely cause subsequent allocations of
hugetlb pages to fail as it would 'appear' that all pages are reserved.

This sequence of operations is unlikely to happen, however they were
easily reproduced and observed using hacked up code as described in [1].

Address the issue by having the routine restore_reserve_on_error take
action on pages where HPageRestoreReserve is not set.  In this case, we
need to remove any reserve map entry created by alloc_huge_page.  A new
helper routine vma_del_reservation assists with this operation.

There are three callers of alloc_huge_page which do not currently call
restore_reserve_on error before freeing a page on error paths.  Add
those missing calls.

[1] https://lore.kernel.org/linux-mm/20210528005029.88088-1-almasrymina@google.com/

Link: https://lkml.kernel.org/r/20210607204510.22617-1-mike.kravetz@oracle.com
Fixes: 96b96a96ddee ("mm/hugetlb: fix huge page reservation leak in private mapping error paths"
Signed-off-by: Mike Kravetz <mike.kravetz@oracle.com>
Reviewed-by: Mina Almasry <almasrymina@google.com>
Cc: Axel Rasmussen <axelrasmussen@google.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/slub: actually fix freelist pointer vs redzoning

It turns out that SLUB redzoning ("slub_debug=Z") checks from
s->object_size rather than from s->inuse (which is normally bumped to
make room for the freelist pointer), so a cache created with an object
size less than 24 would have the freelist pointer written beyond
s->object_size, causing the redzone to be corrupted by the freelist
pointer.  This was very visible with "slub_debug=ZF":

  BUG test (Tainted: G    B            ): Right Redzone overwritten
  -----------------------------------------------------------------------------

  INFO: 0xffff957ead1c05de-0xffff957ead1c05df @offset=1502. First byte 0x1a instead of 0xbb
  INFO: Slab 0xffffef3950b47000 objects=170 used=170 fp=0x0000000000000000 flags=0x8000000000000200
  INFO: Object 0xffff957ead1c05d8 @offset=1496 fp=0xffff957ead1c0620

  Redzone  (____ptrval____): bb bb bb bb bb bb bb bb               ........
  Object   (____ptrval____): 00 00 00 00 00 f6 f4 a5               ........
  Redzone  (____ptrval____): 40 1d e8 1a aa                        @....
  Padding  (____ptrval____): 00 00 00 00 00 00 00 00               ........

Adjust the offset to stay within s->object_size.

(Note that no caches of in this size range are known to exist in the
kernel currently.)

Link: https://lkml.kernel.org/r/20210608183955.280836-4-keescook@chromium.org
Link: https://lore.kernel.org/linux-mm/20200807160627.GA1420741@elver.google.com/
Link: https://lore.kernel.org/lkml/0f7dd7b2-7496-5e2d-9488-2ec9f8e90441@suse.cz/Fixes:
Link: https://lore.kernel.org/lkml/CANpmjNOwZ5VpKQn+SYWovTkFB4VsT-RPwyENBmaK0dLcpqStkA@mail.gmail.com
Signed-off-by: Kees Cook <keescook@chromium.org>
Reported-by: Marco Elver <elver@google.com>
Reported-by: "Lin, Zhenpeng" <zplin@psu.edu>
Tested-by: Marco Elver <elver@google.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Christoph Lameter <cl@linux.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/slub: fix redzoning for small allocations

The redzone area for SLUB exists between s->object_size and s->inuse
(which is at least the word-aligned object_size).  If a cache were
created with an object_size smaller than sizeof(void *), the in-object
stored freelist pointer would overwrite the redzone (e.g.  with boot
param "slub_debug=ZF"):

  BUG test (Tainted: G    B            ): Right Redzone overwritten
  -----------------------------------------------------------------------------

  INFO: 0xffff957ead1c05de-0xffff957ead1c05df @offset=1502. First byte 0x1a instead of 0xbb
  INFO: Slab 0xffffef3950b47000 objects=170 used=170 fp=0x0000000000000000 flags=0x8000000000000200
  INFO: Object 0xffff957ead1c05d8 @offset=1496 fp=0xffff957ead1c0620

  Redzone  (____ptrval____): bb bb bb bb bb bb bb bb    ........
  Object   (____ptrval____): f6 f4 a5 40 1d e8          ...@..
  Redzone  (____ptrval____): 1a aa                      ..
  Padding  (____ptrval____): 00 00 00 00 00 00 00 00    ........

Store the freelist pointer out of line when object_size is smaller than
sizeof(void *) and redzoning is enabled.

Additionally remove the "smaller than sizeof(void *)" check under
CONFIG_DEBUG_VM in kmem_cache_sanity_check() as it is now redundant:
SLAB and SLOB both handle small sizes.

(Note that no caches within this size range are known to exist in the
kernel currently.)

Link: https://lkml.kernel.org/r/20210608183955.280836-3-keescook@chromium.org
Fixes: 81819f0fc828 ("SLUB core")
Signed-off-by: Kees Cook <keescook@chromium.org>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Christoph Lameter <cl@linux.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: "Lin, Zhenpeng" <zplin@psu.edu>
Cc: Marco Elver <elver@google.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/slub: clarify verification reporting

Patch series "Actually fix freelist pointer vs redzoning", v4.

This fixes redzoning vs the freelist pointer (both for middle-position
and very small caches).  Both are "theoretical" fixes, in that I see no
evidence of such small-sized caches actually be used in the kernel, but
that's no reason to let the bugs continue to exist, especially since
people doing local development keep tripping over it.  :)

This patch (of 3):

Instead of repeating "Redzone" and "Poison", clarify which sides of
those zones got tripped.  Additionally fix column alignment in the
trailer.

Before:

  BUG test (Tainted: G    B            ): Redzone overwritten
  ...
  Redzone (____ptrval____): bb bb bb bb bb bb bb bb      ........
  Object (____ptrval____): f6 f4 a5 40 1d e8            ...@..
  Redzone (____ptrval____): 1a aa                        ..
  Padding (____ptrval____): 00 00 00 00 00 00 00 00      ........

After:

  BUG test (Tainted: G    B            ): Right Redzone overwritten
  ...
  Redzone  (____ptrval____): bb bb bb bb bb bb bb bb      ........
  Object   (____ptrval____): f6 f4 a5 40 1d e8            ...@..
  Redzone  (____ptrval____): 1a aa                        ..
  Padding  (____ptrval____): 00 00 00 00 00 00 00 00      ........

The earlier commits that slowly resulted in the "Before" reporting were:

  d86bd1bece6f ("mm/slub: support left redzone")
  ffc79d288000 ("slub: use print_hex_dump")
  2492268472e7 ("SLUB: change error reporting format to follow lockdep loosely")

Link: https://lkml.kernel.org/r/20210608183955.280836-1-keescook@chromium.org
Link: https://lkml.kernel.org/r/20210608183955.280836-2-keescook@chromium.org
Link: https://lore.kernel.org/lkml/cfdb11d7-fb8e-e578-c939-f7f5fb69a6bd@suse.cz/
Signed-off-by: Kees Cook <keescook@chromium.org>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Marco Elver <elver@google.com>
Cc: "Lin, Zhenpeng" <zplin@psu.edu>
Cc: Christoph Lameter <cl@linux.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: David Rientjes <rientjes@google.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Roman Gushchin <guro@fb.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/swap: fix pte_same_as_swp() not removing uffd-wp bit when compare

I found it by pure code review, that pte_same_as_swp() of unuse_vma()
didn't take uffd-wp bit into account when comparing ptes.
pte_same_as_swp() returning false negative could cause failure to
swapoff swap ptes that was wr-protected by userfaultfd.

Link: https://lkml.kernel.org/r/20210603180546.9083-1-peterx@redhat.com
Fixes: f45ec5ff16a7 ("userfaultfd: wp: support swap and page migration")
Signed-off-by: Peter Xu <peterx@redhat.com>
Acked-by: Hugh Dickins <hughd@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: <stable@vger.kernel.org> [5.7+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm,hwpoison: fix race with hugetlb page allocation

When hugetlb page fault (under overcommitting situation) and
memory_failure() race, VM_BUG_ON_PAGE() is triggered by the following
race:

    CPU0:                           CPU1:

                                    gather_surplus_pages()
                                      page = alloc_surplus_huge_page()
    memory_failure_hugetlb()
      get_hwpoison_page(page)
        __get_hwpoison_page(page)
          get_page_unless_zero(page)
                                      zero = put_page_testzero(page)
                                      VM_BUG_ON_PAGE(!zero, page)
                                      enqueue_huge_page(h, page)
      put_page(page)

__get_hwpoison_page() only checks the page refcount before taking an
additional one for memory error handling, which is not enough because
there's a time window where compound pages have non-zero refcount during
hugetlb page initialization.

So make __get_hwpoison_page() check page status a bit more for hugetlb
pages with get_hwpoison_huge_page().  Checking hugetlb-specific flags
under hugetlb_lock makes sure that the hugetlb page is not transitive.
It's notable that another new function, HWPoisonHandlable(), is helpful
to prevent a race against other transitive page states (like a generic
compound page just before PageHuge becomes true).

Link: https://lkml.kernel.org/r/20210603233632.2964832-2-nao.horiguchi@gmail.com
Fixes: ead07f6a867b ("mm/memory-failure: introduce get_hwpoison_page() for consistent refcount handling")
Signed-off-by: Naoya Horiguchi <naoya.horiguchi@nec.com>
Reported-by: Muchun Song <songmuchun@bytedance.com>
Acked-by: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: <stable@vger.kernel.org> [5.12+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Merge tag 'dmaengine-fix-5.13' of git://git./linux/kernel/git/vkoul/dmaengine

Pull dmaengine fixes from Vinod Koul:
 "A bunch of driver fixes, notably:

   - More idxd fixes for driver unregister, error handling and bus
     assignment

   - HAS_IOMEM depends fix for few drivers

   - lock fix in pl330 driver

   - xilinx drivers fixes for initialize registers, missing dependencies
     and limiting descriptor IDs

   - mediatek descriptor management fixes"

* tag 'dmaengine-fix-5.13' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine:
  dmaengine: mediatek: use GFP_NOWAIT instead of GFP_ATOMIC in prep_dma
  dmaengine: mediatek: do not issue a new desc if one is still current
  dmaengine: mediatek: free the proper desc in desc_free handler
  dmaengine: ipu: fix doc warning in ipu_irq.c
  dmaengine: rcar-dmac: Fix PM reference leak in rcar_dmac_probe()
  dmaengine: idxd: Fix missing error code in idxd_cdev_open()
  dmaengine: stedma40: add missing iounmap() on error in d40_probe()
  dmaengine: SF_PDMA depends on HAS_IOMEM
  dmaengine: QCOM_HIDMA_MGMT depends on HAS_IOMEM
  dmaengine: ALTERA_MSGDMA depends on HAS_IOMEM
  dmaengine: idxd: Add missing cleanup for early error out in probe call
  dmaengine: xilinx: dpdma: Limit descriptor IDs to 16 bits
  dmaengine: xilinx: dpdma: Add missing dependencies to Kconfig
  dmaengine: stm32-mdma: fix PM reference leak in stm32_mdma_alloc_chan_resourc()
  dmaengine: zynqmp_dma: Fix PM reference leak in zynqmp_dma_alloc_chan_resourc()
  dmaengine: xilinx: dpdma: initialize registers before request_irq
  dmaengine: pl330: fix wrong usage of spinlock flags in dma_cyclc
  dmaengine: fsl-dpaa2-qdma: Fix error return code in two functions
  dmaengine: idxd: add missing dsa driver unregister
  dmaengine: idxd: add engine 'struct device' missing bus type assignment

Merge tag 'clang-features-v5.13-rc7' of git://git./linux/kernel/git/kees/linux

Pull clang LTO fix from Kees Cook:
 "It seems Clang has been scrubbing through the missing LTO IR flags for
  Clang 13, and the last of these 'only with LTO' flags is fixed now.

  I've asked that they please consider making these changes in a less
  'break all the Clang kernel builds' kind of way in the future. :P

  Summary:

   - The '-warn-stack-size' option under LTO has moved in Clang 13 (Tor
     Vic)"

* tag 'clang-features-v5.13-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux:
  Makefile: lto: Pass -warn-stack-size only on LLD < 13.0.0

ASoC: hdmi-codec: make hdmi_codec_controls static

This symbol is not used outside of hdmi-codec.c, so marks it static.

Fix the following sparse warning:

sound/soc/codecs/hdmi-codec.c:750:25: warning: symbol
'hdmi_codec_controls' was not declared. Should it be static?

Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Signed-off-by: Jiapeng Chong <jiapeng.chong@linux.alibaba.com>
Link: https://lore.kernel.org/r/1623822941-3077-1-git-send-email-jiapeng.chong@linux.alibaba.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: codecs: wcd938x: fix boolreturn.cocci warning

Return statements in functions returning bool should use true/false
instead of 1/0.

Fix the following coccicheck warning:
./sound/soc/codecs/wcd938x.c:1190:9-10: WARNING: return of 0/1 in
function 'wcd938x_volatile_register' with return type bool.

Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Signed-off-by: Yang Li <yang.lee@linux.alibaba.com>
Link: https://lore.kernel.org/r/1623811535-15841-1-git-send-email-yang.lee@linux.alibaba.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rockchip: spdif: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210615141502.1683686-3-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rockchip: pdm: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210615141502.1683686-2-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rockchip: i2s: Use devm_platform_get_and_ioremap_resource()

Use devm_platform_get_and_ioremap_resource() to simplify
code.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20210615141502.1683686-1-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: codecs: Fix duplicate included sound/soc.h

Clean up the following includecheck warnings:

./sound/soc/codecs/wcd938x.c: sound/soc.h is included more than once.
./sound/soc/codecs/wcd938x-sdw.c: sound/soc.h is included more than
once.

No functional change.

Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Signed-off-by: Jiapeng Chong <jiapeng.chong@linux.alibaba.com>
Link: https://lore.kernel.org/r/1623822667-130511-1-git-send-email-jiapeng.chong@linux.alibaba.com
Signed-off-by: Mark Brown <broonie@kernel.org>

proc: only require mm_struct for writing

Commit 591a22c14d3f ("proc: Track /proc/$pid/attr/ opener mm_struct") we
started using __mem_open() to track the mm_struct at open-time, so that
we could then check it for writes.

But that also ended up making the permission checks at open time much
stricter - and not just for writes, but for reads too.  And that in turn
caused a regression for at least Fedora 29, where NIC interfaces fail to
start when using NetworkManager.

Since only the write side wanted the mm_struct test, ignore any failures
by __mem_open() at open time, leaving reads unaffected.  The write()
time verification of the mm_struct pointer will then catch the failure
case because a NULL pointer will not match a valid 'current->mm'.

Link: https://lore.kernel.org/netdev/YMjTlp2FSJYvoyFa@unreal/
Fixes: 591a22c14d3f ("proc: Track /proc/$pid/attr/ opener mm_struct")
Reported-and-tested-by: Leon Romanovsky <leon@kernel.org>
Cc: Kees Cook <keescook@chromium.org>
Cc: Christian Brauner <christian.brauner@ubuntu.com>
Cc: Andrea Righi <andrea.righi@canonical.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>