scsi: aic94xx: Add missing check for dma_map_single()

[ Upstream commit 32fe45274edb5926abc0fac7263d9f889d02d9cf ]

Add check for dma_map_single() and return error if it fails in order to
avoid invalid DMA address.

Fixes: 2908d778ab3e ("[SCSI] aic94xx: new driver")
Link: https://lore.kernel.org/r/20230128110832.6792-1-jiasheng@iscas.ac.cn
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Reviewed-by: Jason Yan <yanaijie@huawei.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

scsi: mpt3sas: Fix a memory leak

[ Upstream commit 54dd96015e8d7a2a07359e2dfebf05b529d1780c ]

Add a forgotten kfree().

Fixes: dbec4c9040ed ("scsi: mpt3sas: lockless command submission")
Link: https://lore.kernel.org/r/20230207152159.18627-1-thenzl@redhat.com
Signed-off-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/amdgpu: fix enum odm_combine_mode mismatch

[ Upstream commit 087bad7eb1f6945f8232f132953ecc2bda8bd38d ]

A conversion from 'bool' to 'enum odm_combine_mode' was incomplete,
and gcc warns about this with many instances of

display/dc/dml/dcn20/display_mode_vba_20.c:3899:44: warning: implicit conversion from 'enum <anonymous>' to 'enum
odm_combine_mode' [-Wenum-conversion]
 3899 |     locals->ODMCombineEnablePerState[i][k] = false;

Change the ones that we get a warning for, using the same numerical
values to leave the behavior unchanged.

Fixes: 5fc11598166d ("drm/amd/display: expand dml structs")
Link: https://lore.kernel.org/all/20201026210039.3884312-3-arnd@kernel.org/
Link: https://lore.kernel.org/all/20210927100659.1431744-1-arnd@kernel.org/
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ALSA: hda: Fix the control element identification for multiple codecs

[ Upstream commit d045bceff5a904bd79d71dede9f927c00ce4906f ]

Some motherboards have multiple HDA codecs connected to the serial bus.
The current code may create multiple mixer controls with the almost
identical identification.

The current code use id.device field from the control element structure
to store the codec address to avoid such clashes for multiple codecs.
Unfortunately, the user space do not handle this correctly. For mixer
controls, only name and index are used for the identifiers.

This patch fixes this problem to compose the index using the codec
address as an offset in case, when the control already exists. It is
really unlikely that one codec will create 10 similar controls.

This patch adds new kernel module parameter 'ctl_dev_id' to allow
select the old behaviour, too. The CONFIG_SND_HDA_CTL_DEV_ID Kconfig
option sets the default value.

BugLink: https://github.com/alsa-project/alsa-lib/issues/294
BugLink: https://github.com/alsa-project/alsa-lib/issues/205
Fixes: 54d174031576 ("[ALSA] hda-codec - Fix connection list parsing")
Fixes: 1afe206ab699 ("ALSA: hda - Try to find an empty control index when it's occupied")
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Link: https://lore.kernel.org/r/20230202092013.4066998-1-perex@perex.cz
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>

hwmon: (ltc2945) Handle error case in ltc2945_value_store

[ Upstream commit 178b01eccfb0b8149682f61388400bd3d903dddc ]

ltc2945_val_to_reg errors were not being handled
which would have resulted in register being set to
0 (clamped) instead of being left alone.

Fixes: 6700ce035f83 ("hwmon: Driver for Linear Technologies LTC2945")

Signed-off-by: Jonathan Cormier <jcormier@criticallink.com>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

hwmon: (asus-ec-sensors) add missing mutex path

[ Upstream commit e2de0e6abd91b05411cb1f0953115dbbbe9b11ce ]

Add missing mutex path for ProArt X570-CREATOR WIFI.

Fixes: de8fbac5e59e (hwmon: (asus-ec-sensors) implement locking via the ACPI global lock)
Signed-off-by: Eugene Shalygin <eugene.shalygin@gmail.com>
Link: https://lore.kernel.org/r/20230121111728.168514-2-eugene.shalygin@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

regulator: tps65219: use generic set_bypass()

[ Upstream commit 0365df81145a4cfaae5f4da896160de512256e6d ]

Due to wrong interpretation of the specification,
custom implementation was used instead of standard regmap helper.
LINK: https://lore.kernel.org/all/c2014039-f1e8-6976-33d6-52e2dd4e7b66@baylibre.com/
Fixes: c12ac5fc3e0a ("regulator: drivers: Add TI TPS65219 PMIC regulators support")

Regulator does NOT require to be off to be switched to bypass mode.

Signed-off-by: Jerome Neanne <jneanne@baylibre.com>
Link: https://lore.kernel.org/r/20230203140119.13029-1-jneanne@baylibre.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ASoC: dt-bindings: meson: fix gx-card codec node regex

[ Upstream commit 480b26226873c88e482575ceb0d0a38d76e1be57 ]

'codec' is a valid node name when there is a single codec
in the link. Fix the node regular expression to apply this.

Fixes: fd00366b8e41 ("ASoC: meson: gx: add sound card dt-binding documentation")
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20230202183653.486216-3-jbrunet@baylibre.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ASoC: mchp-spdifrx: Fix uninitialized use of mr in mchp_spdifrx_hw_params()

[ Upstream commit 218674a45930c700486d27b765bf2f1b43f8cbf7 ]

Clang warns:

  ../sound/soc/atmel/mchp-spdifrx.c:455:3: error: variable 'mr' is uninitialized when used here [-Werror,-Wuninitialized]
                  mr |= SPDIFRX_MR_ENDIAN_BIG;
                  ^~
  ../sound/soc/atmel/mchp-spdifrx.c:432:8: note: initialize the variable 'mr' to silence this warning
          u32 mr;
                ^
                 = 0
  1 error generated.

Zero initialize mr so that these bitwise OR and assignment operation
works unconditionally.

Fixes: fa09fa60385a ("ASoC: mchp-spdifrx: fix controls which rely on rsr register")
Link: https://github.com/ClangBuiltLinux/linux/issues/1797
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Reviewed-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Link: https://lore.kernel.org/r/20230202-mchp-spdifrx-fix-uninit-mr-v1-1-629a045d7a2f@kernel.org
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ASoC: rsnd: fixup #endif position

[ Upstream commit 49123b51cd896e00b256a27c2ce9e6bfe1bbc22f ]

commit 1f9c82b5ab83ff2 ("ASoC: rsnd: add debugfs support") added
CONFIG_DEBUG_FS related definitions on rsnd.h, but it should be
added inside of RSND_H. This patch fixup it.

Fixes: 1f9c82b5ab83 ("ASoC: rsnd: add debugfs support")
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/877cx26t7r.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

regmap: apply reg_base and reg_downshift for single register ops

[ Upstream commit 697c3892d825fb78f42ec8e53bed065dd728db3e ]

reg_base and reg_downshift currently don't have any effect if used with
a regmap_bus or regmap_config which only offers single register
operations (ie. reg_read, reg_write and optionally reg_update_bits).

Fix that and take them into account also for regmap_bus with only
reg_read and read_write operations by applying reg_base and
reg_downshift in _regmap_bus_reg_write, _regmap_bus_reg_read.

Also apply reg_base and reg_downshift in _regmap_update_bits, but only
in case the operation is carried out with a reg_update_bits call
defined in either regmap_bus or regmap_config.

Fixes: 0074f3f2b1e43d ("regmap: allow a defined reg_base to be added to every address")
Fixes: 86fc59ef818beb ("regmap: add configurable downshift for addresses")
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Tested-by: Colin Foster <colin.foster@in-advantage.com>
Link: https://lore.kernel.org/r/Y9clyVS3tQEHlUhA@makrotopia.org
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

dm: improve shrinker debug names

[ Upstream commit c87791bcc455a91e51ca9800faaacc21c8d67785 ]

Commit e33c267ab70d ("mm: shrinkers: provide shrinkers with names")
chose some fairly bad names for DM's shrinkers.

Fixes: e33c267ab70d ("mm: shrinkers: provide shrinkers with names")
Signed-off-by : Mike Snitzer <snitzer@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ASoC: mchp-spdifrx: disable all interrupts in mchp_spdifrx_dai_remove()

[ Upstream commit aaecdc32b7e35b4f9b457fb3509414aa9a932589 ]

CSC interrupts which might be used in controls are on bits 8 and 9 of
SPDIFRX_IDR register. Thus disable all the interrupts that are exported
by driver.

Fixes: ef265c55c1ac ("ASoC: mchp-spdifrx: add driver for SPDIF RX")
Signed-off-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Link: https://lore.kernel.org/r/20230130120647.638049-5-claudiu.beznea@microchip.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ASoC: mchp-spdifrx: fix controls that works with completion mechanism

[ Upstream commit d3681df44e856aab523a6eb7ba15b5e41efcbb1c ]

Channel status get and channel subcode get controls relies on data
returned by controls when certain IRQs are raised. To achieve that
completions are used b/w controls and interrupt service routine. The
concurrent accesses to these controls are protected by
struct snd_card::controls_rwsem.

Issues identified:
- reinit_completion() may be called while waiting for completion
  which should be avoided
- in case of multiple threads waiting, the complete() call in interrupt
  will signal only one waiting thread per interrupt which may lead to
  timeout for the others
- in case of channel status get as the CSC interrupt is not refcounted
  ISR may disable interrupt for threads that were just enabled it.

To solve these the access to controls were protected by a mutex. Along
with this there is no need for spinlock to protect the software cache
reads/updates b/w controls and ISR as the update is happening only when
requested from control, and only one reader can reach the control.

Fixes: ef265c55c1ac ("ASoC: mchp-spdifrx: add driver for SPDIF RX")
Signed-off-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Link: https://lore.kernel.org/r/20230130120647.638049-4-claudiu.beznea@microchip.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ASoC: mchp-spdifrx: fix return value in case completion times out

[ Upstream commit a4c4161d6eae3ef5f486d1638ef452d9bc1376b0 ]

wait_for_completion_interruptible_timeout() returns 0 in case of
timeout. Check this into account when returning from function.

Fixes: ef265c55c1ac ("ASoC: mchp-spdifrx: add driver for SPDIF RX")
Signed-off-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Link: https://lore.kernel.org/r/20230130120647.638049-3-claudiu.beznea@microchip.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ASoC: mchp-spdifrx: fix controls which rely on rsr register

[ Upstream commit fa09fa60385abbf99342494b280da8b4aebbc0e9 ]

The SPDIFRX block is clocked by 2 clocks: peripheral and generic clocks.
Peripheral clock feeds user interface (registers) and generic clock feeds
the receiver.

To enable the receiver the generic clock needs to be enabled and also the
ENABLE bit of MCHP_SPDIFRX_MR register need to be set.

The signal control exported by mchp-spdifrx driver reports wrong status
when the receiver is disabled. This can happen when requesting the signal
and the capture was not previously started. To solve this the receiver
needs to be enabled (by enabling generic clock and setting ENABLE bit of
MR register) before reading the signal status.

As with this fix there are 2 paths now that need to control the generic
clock and ENABLE bit of SPDIFRX_MR register (one path though controls, one
path though configuration) a mutex has been introduced. We can't rely on
subsystem locking as the controls are protected by
struct snd_card::controls_rwsem semaphore and configuration is protected
by a different lock (embedded in snd_pcm_stream_lock_irq()).

The introduction of mutex is also extended to other controls which rely on
SPDIFRX_RSR.ULOCK bit as it has been discovered experimentally that having
both clocks enabled but not the receiver (through ENABLE bit of SPDIFRX.MR)
leads to inconsistent values of SPDIFRX_RSR.ULOCK. Thus on some controls we
rely on software state (dev->trigger_enabled protected by mutex) to
retrieve proper values.

Fixes: ef265c55c1ac ("ASoC: mchp-spdifrx: add driver for SPDIF RX")
Signed-off-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Link: https://lore.kernel.org/r/20230130120647.638049-2-claudiu.beznea@microchip.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

spi: dw_bt1: fix MUX_MMIO dependencies

[ Upstream commit d4bde04318c0d33705e9a77d4c7df72f262011e0 ]

Selecting a symbol with additional dependencies requires
adding the same dependency here:

WARNING: unmet direct dependencies detected for MUX_MMIO
  Depends on [n]: MULTIPLEXER [=y] && OF [=n]
  Selected by [y]:
  - SPI_DW_BT1 [=y] && SPI [=y] && SPI_MASTER [=y] && SPI_DESIGNWARE [=y] && (MIPS_BAIKAL_T1 || COMPILE_TEST [=y])

Drop the 'select' here to avoid the problem. Anyone using
the dw-bt1 SPI driver should make sure they include the
mux driver as well now.

Fixes: 7218838109fe ("spi: dw-bt1: Fix undefined devm_mux_control_get symbol")
Fixes: abf00907538e ("spi: dw: Add Baikal-T1 SPI Controller glue driver")
Link: https://lore.kernel.org/all/20221218192523.c6vnfo26ua6xqf26@mobilestation/
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Serge Semin <fancer.lancer@gmail.com>
Link: https://lore.kernel.org/r/20230130140156.3620863-1-arnd@kernel.org
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ASoC: topology: Properly access value coming from topology file

[ Upstream commit c5d184c92df2b631fb81fe2ce6e96bfc5ba720e5 ]

When accessing values coming from topology, le32_to_cpu should be used.
One of recent commits missed that.

Fixes: 86e2d14b6d1a ("ASoC: topology: Add header payload_size verification")
Reviewed-by: Cezary Rojewski <cezary.rojewski@intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski@linux.intel.com>
Link: https://lore.kernel.org/r/20230127231111.937721-2-amadeuszx.slawinski@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

gpio: vf610: connect GPIO label to dev name

[ Upstream commit 6f8ecb7f85f441eb7d78ba2a4df45ee8a821934e ]

Current GPIO label is fixed, so can't distinguish different GPIO
controllers through labels. Use dev name instead.

Fixes: 7f2691a19627 ("gpio: vf610: add gpiolib/IRQ chip driver for Vybrid")
Signed-off-by: Clark Wang <xiaoning.wang@nxp.com>
Signed-off-by: Haibo Chen <haibo.chen@nxp.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

dt-bindings: display: mediatek: Fix the fallback for mediatek,mt8186-disp-ccorr

[ Upstream commit 137272ef1b0f17a1815fec00d583515a0163f85e ]

The mt8186-disp-ccorr is not fully compatible with the mt8183-disp-ccorr
implementation. It causes a crash when system resumes if it binds to the
device.

We should use mt8192-disp-ccorr as fallback of mt8186-disp-ccorr.

Fixes: 8a26ea19d4dc ("dt-bindings: display: mediatek: add MT8186 SoC binding")
Signed-off-by: Allen-KH Cheng <allen-kh.cheng@mediatek.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Reviewed-by: Matthias Brugger <matthias.bgg@gmail.com>
Link: https://patchwork.kernel.org/project/linux-mediatek/patch/20230118091829.755-10-allen-kh.cheng@mediatek.com/
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ASoC: soc-compress.c: fixup private_data on snd_soc_new_compress()

[ Upstream commit ffe4c0f0bfaa571a676a0e946d4a6a0607f94294 ]

commit d3268a40d4b19f ("ASoC: soc-compress.c: fix NULL dereference")
enables DPCM capture, but it should independent from playback.
This patch fixup it.

Fixes: d3268a40d4b1 ("ASoC: soc-compress.c: fix NULL dereference")
Link: https://lore.kernel.org/r/87tu0i6j7j.wl-kuninori.morimoto.gx@renesas.com
Acked-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/871qnkvo1s.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/mediatek: Clean dangling pointer on bind error path

[ Upstream commit 36aa8c61af55675ed967900fbe5deb32d776f051 ]

mtk_drm_bind() can fail, in which case drm_dev_put() is called,
destroying the drm_device object. However a pointer to it was still
being held in the private object, and that pointer would be passed along
to DRM in mtk_drm_sys_prepare() if a suspend were triggered at that
point, resulting in a panic. Clean the pointer when destroying the
object in the error path to prevent this from happening.

Fixes: 119f5173628a ("drm/mediatek: Add DRM Driver for Mediatek SoC MT8173.")
Signed-off-by: Nícolas F. R. A. Prado <nfraprado@collabora.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://patchwork.kernel.org/project/linux-mediatek/patch/20221122143949.3493104-1-nfraprado@collabora.com/
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/mediatek: mtk_drm_crtc: Add checks for devm_kcalloc

[ Upstream commit 5bf1e3bd7da625ccf9a22c8cb7d65271e6e47f4c ]

As the devm_kcalloc may return NULL, the return value needs to be checked
to avoid NULL poineter dereference.

Fixes: 31c5558dae05 ("drm/mediatek: Refactor plane init")
Signed-off-by: ruanjinjie <ruanjinjie@huawei.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://patchwork.kernel.org/project/linux-mediatek/patch/20221205095115.2905090-1-ruanjinjie@huawei.com/
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/mediatek: Drop unbalanced obj unref

[ Upstream commit 4deef811828e87e26a978d5d6433b261d4713849 ]

In the error path, mtk_drm_gem_object_mmap() is dropping an obj
reference that it doesn't own.

Fixes: 119f5173628a ("drm/mediatek: Add DRM Driver for Mediatek SoC MT8173.")
Signed-off-by: Rob Clark <robdclark@chromium.org>
Link: https://patchwork.kernel.org/project/linux-mediatek/patch/20230119231255.2883365-1-robdclark@gmail.com/
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/mediatek: Use NULL instead of 0 for NULL pointer

[ Upstream commit 4744cde06f57dd6fbaac468663b1fe2f653eaa16 ]

Use NULL for NULL pointer to fix the following sparse warning:
drivers/gpu/drm/mediatek/mtk_drm_gem.c:265:27: sparse: warning: Using plain integer as NULL pointer

Fixes: 3df64d7b0a4f ("drm/mediatek: Implement gem prime vmap/vunmap function")
Signed-off-by: Miles Chen <miles.chen@mediatek.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://patchwork.kernel.org/project/linux-mediatek/patch/20230111024443.24559-1-miles.chen@mediatek.com/
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/mediatek: dsi: Reduce the time of dsi from LP11 to sending cmd

[ Upstream commit 91aeaed2c1147e3b1157dc084d23f190856a6c23 ]

According to Figure 16 Turnaround Procedure on page 36 in [1], you
can see the status of LP-00 -> LP10 -> LP11. This state can correspond
to the state of DSI from LP00 -> LP11 in mtk_dsi_lane_ready function
in mtk_dsi.c.

LP-00 -> LP10 -> LP11 takes about 2*TLPX time (refer to [1] page 51
to see that TLPX is 50ns)

The delay at the end of the mtk_dsi_lane_ready function should be
greater than the 2*TLPX specified by the DSI spec, and less than
the time specified by the DSI_RX (generally 6ms to 40ms), to avoid
problems caused by the RX specification

[1]:mipi_D-PHY_specification_v1-1

Fixes: 39e8d062b03c ("drm/mediatek: Keep dsi as LP00 before dcs cmds transfer")
Signed-off-by: Xinlei Lee <xinlei.lee@mediatek.com>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://patchwork.kernel.org/project/linux-mediatek/patch/1673330093-6771-2-git-send-email-xinlei.lee@mediatek.com/
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/msm/dpu: set pdpu->is_rt_pipe early in dpu_plane_sspp_atomic_update()

[ Upstream commit 1d233b1cb149ec78c20fac58331b27bb460f9558 ]

The function dpu_plane_sspp_atomic_update() updates pdpu->is_rt_pipe
flag, but after the commit 854f6f1c653b ("drm/msm/dpu: update the qos
remap only if the client type changes") it sets the flag late, after all
the qos functions have updated QoS programming. Move the flag update
back to the place where it happened before the mentioned commit to let
the pipe be programmed according to its current RT/non-RT state.

Fixes: 854f6f1c653b ("drm/msm/dpu: update the qos remap only if the client type changes")
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Patchwork: https://patchwork.freedesktop.org/patch/516239/
Link: https://lore.kernel.org/r/20221229191856.3508092-2-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

pinctrl: renesas: rzg2l: Fix configuring the GPIO pins as interrupts

[ Upstream commit 00dfe29887761405ccd23cc0aa07cb86623bb2b7 ]

On the RZ/G2UL SoC we have less number of pins compared to RZ/G2L and also
the pin configs are completely different. This patch makes sure we use the
appropriate pin configs for each SoC (which is passed as part of the OF
data) while configuring the GPIO pin as interrupts instead of using
rzg2l_gpio_configs[] for all the SoCs.

Fixes: bfc69bdbaad1 ("pinctrl: renesas: rzg2l: Add RZ/G2UL support")
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20230102221815.273719-3-prabhakar.mahadev-lad.rj@bp.renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/tegra: firewall: Check for is_addr_reg existence in IMM check

[ Upstream commit 1b5c09de25e8c08655c270a70e5e74e93b6bad1f ]

In the IMM opcode check, don't call is_addr_reg if it's not set.

Fixes: 8cc95f3fd35e ("drm/tegra: Add job firewall")
Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

gpu: host1x: Don't skip assigning syncpoints to channels

[ Upstream commit eb258cc1fd458e584082be987dbc6ec42668c05e ]

The code to write the syncpoint channel assignment register
incorrectly skips the write if hypervisor registers are not available.

The register, however, is within the guest aperture so remove the
check and assign syncpoints properly even on virtualized systems.

Fixes: c3f52220f276 ("gpu: host1x: Enable Tegra186 syncpoint protection")
Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

gpu: host1x: Fix mask for syncpoint increment register

[ Upstream commit 79aad29c7d2d2cd64790115d3a6ebac28c00a8ec ]

On Tegra186+, the syncpoint ID has 10 bits of space. To allow
using more than 256 syncpoints, fix the mask.

Fixes: 9abdd497cd0a ("gpu: host1x: Tegra234 device data and headers")
Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

pinctrl: mediatek: Initialize variable *buf to zero

[ Upstream commit 2e34f82ba214134ecf590fbe0cdbd87401645a8a ]

Coverity spotted that *buf is not initialized to zero in
mtk_pctrl_dbg_show. Using uninitialized variable *buf as argument to %s
when calling seq_printf. Fix this coverity by initializing *buf as zero.

Fixes: 184d8e13f9b1 ("pinctrl: mediatek: Add support for pin configuration dump via debugfs.")
Signed-off-by: Guodong Liu <Guodong.Liu@mediatek.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://lore.kernel.org/r/20230118062036.26258-3-Guodong.Liu@mediatek.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

pinctrl: mediatek: Initialize variable pullen and pullup to zero

[ Upstream commit a298c70a10c604a6b3df5a0aa56597b705ba0f6b ]

Coverity spotted that pullen and pullup is not initialized to zero in
mtk_pctrl_show_one_pin. The uninitialized variable pullen is used in
assignment statement "rsel = pullen;" in mtk_pctrl_show_one_pin, and
Uninitialized variable pullup is used when calling scnprintf. Fix this
coverity by initializing pullen and pullup as zero.

Fixes: 184d8e13f9b1 ("pinctrl: mediatek: Add support for pin configuration dump via debugfs.")
Signed-off-by: Guodong Liu <Guodong.Liu@mediatek.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://lore.kernel.org/r/20230118062036.26258-2-Guodong.Liu@mediatek.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

pinctrl: bcm2835: Remove of_node_put() in bcm2835_of_gpio_ranges_fallback()

[ Upstream commit 2d578dd27871372f7159dd3206149ec616700d87 ]

Remove wrong of_node_put() in bcm2835_of_gpio_ranges_fallback(),
there is no counterpart of_node_get() for it.

Fixes: d2b67744fd99 ("pinctrl: bcm2835: implement hook for missing gpio-ranges")
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Stefan Wahren <stefan.wahren@i2se.com>
Tested-by: Stefan Wahren <stefan.wahren@i2se.com>
Tested-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Acked-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Link: https://lore.kernel.org/r/20230113215352.44272-3-andriy.shevchenko@linux.intel.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

habanalabs: bugs fixes in timestamps buff alloc

[ Upstream commit 1693fef9e95dbe8ab767d208a02328fff13fbb94 ]

use argument instead of fixed GFP value for allocation
in Timestamps buffers alloc function.
change data type of size to size_t.

Fixes: 9158bf69e74f ("habanalabs: Timestamps buffers registration")
Signed-off-by: farah kassabri <fkassabri@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/msm/mdp5: Add check for kzalloc

[ Upstream commit 13fcfcb2a9a4787fe4e49841d728f6f2e9fa6911 ]

As kzalloc may fail and return NULL pointer,
it should be better to check the return value
in order to avoid the NULL pointer dereference.

Fixes: 1cff7440a86e ("drm/msm: Convert to using __drm_atomic_helper_crtc_reset() for reset.")
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/514154/
Link: https://lore.kernel.org/r/20221206074819.18134-1-jiasheng@iscas.ac.cn
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/msm/dpu: Add check for pstates

[ Upstream commit 93340e10b9c5fc86730d149636e0aa8b47bb5a34 ]

As kzalloc may fail and return NULL pointer,
it should be better to check pstates
in order to avoid the NULL pointer dereference.

Fixes: 25fdd5933e4c ("drm/msm: Add SDM845 DPU support")
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Patchwork: https://patchwork.freedesktop.org/patch/514160/
Link: https://lore.kernel.org/r/20221206080236.43687-1-jiasheng@iscas.ac.cn
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/msm/dpu: Add check for cstate

[ Upstream commit c96988b7d99327bb08bd9efd29a203b22cd88ace ]

As kzalloc may fail and return NULL pointer,
it should be better to check cstate
in order to avoid the NULL pointer dereference
in __drm_atomic_helper_crtc_reset.

Fixes: 1cff7440a86e ("drm/msm: Convert to using __drm_atomic_helper_crtc_reset() for reset.")
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/514163/
Link: https://lore.kernel.org/r/20221206080517.43786-1-jiasheng@iscas.ac.cn
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/msm: use strscpy instead of strncpy

[ Upstream commit d7fd8634f48d76aa799ed57beb7d87dab91bde80 ]

Using strncpy can result in non-NULL-terminated destination string. Use
strscpy instead. This fixes following warning:

drivers/gpu/drm/msm/msm_fence.c: In function ‘msm_fence_context_alloc’:
drivers/gpu/drm/msm/msm_fence.c:25:9: warning: ‘strncpy’ specified bound 32 equals destination size [-Wstringop-truncation]
   25 |         strncpy(fctx->name, name, sizeof(fctx->name));
      |         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Fixes: f97decac5f4c ("drm/msm: Support multiple ringbuffers")
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Patchwork: https://patchwork.freedesktop.org/patch/518787/
Link: https://lore.kernel.org/r/20230118020152.1689213-1-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/msm/dpu: sc7180: add missing WB2 clock control

[ Upstream commit 255f056181ac69278dcd8778f0be32a21b2d2a6a ]

Add missing DPU_CLK_CTRL_WB2 to sc7180_mdp clocks array.

Fixes: 51e4d60e6ba5 ("drm/msm/dpu: add writeback support for sc7180")
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Tested-by: Jessica Zhang <quic_jesszhan@quicinc.com> # Trogdor (sc7180)
Patchwork: https://patchwork.freedesktop.org/patch/518504/
Link: https://lore.kernel.org/r/20230116103055.780767-1-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

scsi: ufs: exynos: Fix DMA alignment for PAGE_SIZE != 4096

[ Upstream commit 86bd0c4a2a5dc4265884648cb92c681646509692 ]

The Exynos UFS controller only supports scatter/gather list elements that
are aligned on a 4 KiB boundary. Fix DMA alignment in case PAGE_SIZE !=
4096. Rename UFSHCD_QUIRK_ALIGN_SG_WITH_PAGE_SIZE into
UFSHCD_QUIRK_4KB_DMA_ALIGNMENT.

Cc: Kiwoong Kim <kwmad.kim@samsung.com>
Fixes: 2b2bfc8aa519 ("scsi: ufs: Introduce a quirk to allow only page-aligned sg entries")
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Alim Akhtar <alim.akhtar@samsung.com>
Tested-by: Alim Akhtar <alim.akhtar@samsung.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/msm/dsi: Allow 2 CTRLs on v2.5.0

[ Upstream commit 1ae654ded7c5a19dc13f57a4fe4434fef879b6f9 ]

v2.5.0 support was originally added for SC7280, but this hw is also
present on SM8350, which has one more DSI host. Bump up the dsi count
and fill in the register of the secondary host to allow it to probe.

This should not have any adverse effects on SC7280, as the secondary
CTRL will only be touched if it's defined, anyway.

Fixes: 65c391b31994 ("drm/msm/dsi: Add DSI support for SC7280")
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/519513/
Link: https://lore.kernel.org/r/20230120210101.2146852-1-konrad.dybcio@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm: exynos: dsi: Fix MIPI_DSI*_NO_* mode flags

[ Upstream commit 996e1defca34485dd2bd70b173f069aab5f21a65 ]

HFP/HBP/HSA/EOT_PACKET modes in Exynos DSI host specifies
0 = Enable and 1 = Disable.

The logic for checking these mode flags was correct before
the MIPI_DSI*_NO_* mode flag conversion.

This patch is trying to fix this MIPI_DSI*_NO_* mode flags handling
Exynos DSI host and update the mode_flags in relevant panel drivers.

Fixes: 0f3b68b66a6d ("drm/dsi: Add _NO_ to MIPI_DSI_* flags disabling features")
Reviewed-by: Marek Vasut <marex@denx.de>
Reviewed-by: Nicolas Boichat <drinkcat@chromium.org>
Reported-by: Sébastien Szymanski <sebastien.szymanski@armadeus.com>
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Signed-off-by: Marek Vasut <marex@denx.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20221212145745.15387-1-jagan@amarulasolutions.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/mipi-dsi: Fix byte order of 16-bit DCS set/get brightness

[ Upstream commit c9d27c6be518b4ef2966d9564654ef99292ea1b3 ]

The MIPI DCS specification demands that brightness values are sent in
big endian byte order. It also states that one parameter (i.e. one byte)
shall be sent/received for 8 bit wide values, and two parameters shall
be used for values that are between 9 and 16 bits wide.

Add new functions to properly handle 16-bit brightness in big endian,
since the two 8- and 16-bit cases are distinct from each other.

[richard: use separate functions instead of switch/case]
[richard: split into 16-bit component]

Fixes: 1a9d759331b8 ("drm/dsi: Implement DCS set/get display brightness")
Signed-off-by: Daniel Mentz <danielmentz@google.com>
Link: https://android.googlesource.com/kernel/msm/+/754affd62d0ee268c686c53169b1dbb7deac8550
[richard: fix 16-bit brightness_get]
Signed-off-by: Richard Acayan <mailingradian@gmail.com>
Tested-by: Caleb Connolly <caleb@connolly.tech>
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Reviewed-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230116224909.23884-2-mailingradian@gmail.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

regulator: tps65219: use IS_ERR() to detect an error pointer

[ Upstream commit 2bbba115c3c9a647bcb3201b014fcc3728fe75c8 ]

Fix pointer comparison to integer warning from gcc & sparse:

GCC:
../drivers/regulator/tps65219-regulator.c:370:26: warning: ordered comparison of pointer with integer zero [-Wextra]
  370 |                 if (rdev < 0) {
      |                          ^

sparse warning:
drivers/regulator/tps65219-regulator.c:370:26: sparse: error: incompatible types for operation (<):
drivers/regulator/tps65219-regulator.c:370:26: sparse:    struct regulator_dev *[assigned] rdev
drivers/regulator/tps65219-regulator.c:370:26: sparse:    int

Fixes: c12ac5fc3e0a ("regulator: drivers: Add TI TPS65219 PMIC regulators support")
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Jerome Neanne <jneanne@baylibre.com>
Cc: Tony Lindgren <tony@atomide.com>
Cc: linux-omap@vger.kernel.org
Cc: Liam Girdwood <lgirdwood@gmail.com>
Cc: Mark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/20230114185736.2076-1-rdunlap@infradead.org
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: lt9611: pass a pointer to the of node

[ Upstream commit b0a7f8736789935f62d6df32d441cdf05a5c05d2 ]

Pass a pointer to the OF node while registering lt9611 MIPI device.

Fixes: 23278bf54afe ("drm/bridge: Introduce LT9611 DSI to HDMI bridge")
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230118081658.2198520-7-dmitry.baryshkov@linaro.org
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: lt9611: fix clock calculation

[ Upstream commit 2576eb26494eb0509dd9ceb0cd27771a7a5e3674 ]

Instead of having several fixed values for the pcr register, calculate
it before programming. This allows the bridge to support most of the
display modes.

Fixes: 23278bf54afe ("drm/bridge: Introduce LT9611 DSI to HDMI bridge")
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230118081658.2198520-6-dmitry.baryshkov@linaro.org
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: lt9611: fix programming of video modes

[ Upstream commit ad188aa47edaa033a270e1a3efae43836ff47569 ]

Program the upper part of the hfront_porch into the proper register.

Fixes: 23278bf54afe ("drm/bridge: Introduce LT9611 DSI to HDMI bridge")
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230118081658.2198520-5-dmitry.baryshkov@linaro.org
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: lt9611: fix polarity programming

[ Upstream commit 0b157efa384ea417304b1da284ee2f603c607fc3 ]

Fix programming of hsync and vsync polarities

Fixes: 23278bf54afe ("drm/bridge: Introduce LT9611 DSI to HDMI bridge")
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230118081658.2198520-4-dmitry.baryshkov@linaro.org
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: lt9611: fix HPD reenablement

[ Upstream commit a7790f6bd38f3642b60ae3504a2c749135b89451 ]

The driver will reset the bridge in the atomic_pre_enable(). However
this will also drop the HPD interrupt state. Instead of resetting the
bridge, properly wake it up. This fixes the HPD interrupt delivery after
the disable/enable cycle.

Fixes: 23278bf54afe ("drm/bridge: Introduce LT9611 DSI to HDMI bridge")
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230118081658.2198520-3-dmitry.baryshkov@linaro.org
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: lt9611: fix sleep mode setup

[ Upstream commit ae2d329f104b75a0a78dcaded29fe6283289cdf9 ]

On atomic_post_disable the bridge goes to the low power state. However
the code disables too much of the chip, so the HPD event is not being
detected and delivered to the host. Reduce the power saving in order to
get the HPD event.

Fixes: 23278bf54afe ("drm/bridge: Introduce LT9611 DSI to HDMI bridge")
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230118081658.2198520-2-dmitry.baryshkov@linaro.org
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/msm/dpu: Disallow unallocated resources to be returned

[ Upstream commit abc40122d9a69f56c04efb5a7485795f5ac799d1 ]

In the event that the topology requests resources that have not been
created by the system (because they are typically not represented in
dpu_mdss_cfg ^1), the resource(s) in global_state (in this case DSC
blocks, until their allocation/assignment is being sanity-checked in
"drm/msm/dpu: Reject topologies for which no DSC blocks are available")
remain NULL but will still be returned out of
dpu_rm_get_assigned_resources, where the caller expects to get an array
containing num_blks valid pointers (but instead gets these NULLs).

To prevent this from happening, where null-pointer dereferences
typically result in a hard-to-debug platform lockup, num_blks shouldn't
increase past NULL blocks and will print an error and break instead.
After all, max_blks represents the static size of the maximum number of
blocks whereas the actual amount varies per platform.

^1: which can happen after a git rebase ended up moving additions to
_dpu_cfg to a different struct which has the same patch context.

Fixes: bb00a452d6f7 ("drm/msm/dpu: Refactor resource manager")
Signed-off-by: Marijn Suijten <marijn.suijten@somainline.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/517636/
Link: https://lore.kernel.org/r/20230109231556.344977-1-marijn.suijten@somainline.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/msm/gem: Add check for kmalloc

[ Upstream commit d839f0811a31322c087a859c2b181e2383daa7be ]

Add the check for the return value of kmalloc in order to avoid
NULL pointer dereference in copy_from_user.

Fixes: 20224d715a88 ("drm/msm/submit: Move copy_from_user ahead of locking bos")
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/514678/
Link: https://lore.kernel.org/r/20221212091117.43511-1-jiasheng@iscas.ac.cn
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/amdgpu: Use the sched from entity for amdgpu_cs trace

[ Upstream commit cf22ef78f22ce4df4757472c5dbd33c430c5b659 ]

The problem is that base sched hasn't been assigned yet at this moment,
causing something like "ring=0" all the time from trace.

mpv:cs0-3473    [002] ..... 129.047431: amdgpu_cs: ring=0, dw=48, fences=0
mpv:cs0-3473    [002] ..... 129.089125: amdgpu_cs: ring=0, dw=48, fences=0
mpv:cs0-3473    [002] ..... 129.130987: amdgpu_cs: ring=0, dw=48, fences=0
mpv:cs0-3473    [002] ..... 129.172478: amdgpu_cs: ring=0, dw=48, fences=0

Fixes: 4624459c84d7 ("drm/amdgpu: add gang submit frontend v6")
Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ALSA: hda/ca0132: minor fix for allocation size

[ Upstream commit 3ee0fe7fa39b14d1cea455b7041f2df933bd97d2 ]

Although the "dma_chan" pointer occupies more or equal space compared
to "*dma_chan", the allocation size should use the size of variable
itself.

Found by Linux Verification Center (linuxtesting.org) with SVACE.

Fixes: 01ef7dbffb41 ("ALSA: hda - Update CA0132 codec to load DSP firmware binary")
Signed-off-by: Alexey V. Vissarionov <gremlin@altlinux.org>
Link: https://lore.kernel.org/r/20230117111522.GA15213@altlinux.org
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/msm/adreno: Fix null ptr access in adreno_gpu_cleanup()

[ Upstream commit dbeedbcb268d055d8895aceca427f897e12c2b50 ]

Fix the below kernel panic due to null pointer access:
[   18.504431] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000048
[   18.513464] Mem abort info:
[   18.516346]   ESR = 0x0000000096000005
[   18.520204]   EC = 0x25: DABT (current EL), IL = 32 bits
[   18.525706]   SET = 0, FnV = 0
[   18.528878]   EA = 0, S1PTW = 0
[   18.532117]   FSC = 0x05: level 1 translation fault
[   18.537138] Data abort info:
[   18.540110]   ISV = 0, ISS = 0x00000005
[   18.544060]   CM = 0, WnR = 0
[   18.547109] user pgtable: 4k pages, 39-bit VAs, pgdp=0000000112826000
[   18.553738] [0000000000000048] pgd=0000000000000000, p4d=0000000000000000, pud=0000000000000000
[   18.562690] Internal error: Oops: 0000000096000005 [#1] PREEMPT SMP
**Snip**
[   18.696758] Call trace:
[   18.699278]  adreno_gpu_cleanup+0x30/0x88
[   18.703396]  a6xx_destroy+0xc0/0x130
[   18.707066]  a6xx_gpu_init+0x308/0x424
[   18.710921]  adreno_bind+0x178/0x288
[   18.714590]  component_bind_all+0xe0/0x214
[   18.718797]  msm_drm_bind+0x1d4/0x614
[   18.722566]  try_to_bring_up_aggregate_device+0x16c/0x1b8
[   18.728105]  __component_add+0xa0/0x158
[   18.732048]  component_add+0x20/0x2c
[   18.735719]  adreno_probe+0x40/0xc0
[   18.739300]  platform_probe+0xb4/0xd4
[   18.743068]  really_probe+0xfc/0x284
[   18.746738]  __driver_probe_device+0xc0/0xec
[   18.751129]  driver_probe_device+0x48/0x110
[   18.755421]  __device_attach_driver+0xa8/0xd0
[   18.759900]  bus_for_each_drv+0x90/0xdc
[   18.763843]  __device_attach+0xfc/0x174
[   18.767786]  device_initial_probe+0x20/0x2c
[   18.772090]  bus_probe_device+0x40/0xa0
[   18.776032]  deferred_probe_work_func+0x94/0xd0
[   18.780686]  process_one_work+0x190/0x3d0
[   18.784805]  worker_thread+0x280/0x3d4
[   18.788659]  kthread+0x104/0x1c0
[   18.791981]  ret_from_fork+0x10/0x20
[   18.795654] Code: f9400408 aa0003f3 aa1f03f4 91142015 (f9402516)
[   18.801913] ---[ end trace 0000000000000000 ]---
[   18.809039] Kernel panic - not syncing: Oops: Fatal exception

Fixes: 17e822f7591f ("drm/msm: fix unbalanced pm_runtime_enable in adreno_gpu_{init, cleanup}")
Signed-off-by: Akhil P Oommen <quic_akhilpo@quicinc.com>
Patchwork: https://patchwork.freedesktop.org/patch/515605/
Link: https://lore.kernel.org/r/20221221203925.v2.1.Ib978de92c4bd000b515486aad72e96c2481f84d0@changeid
Signed-off-by: Rob Clark <robdclark@chromium.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: tc358767: Set default CLRSIPO count

[ Upstream commit 01338bb82fed40a6a234c2b36a92367c8671adf0 ]

The current CLRSIPO count is still marginal and does not work with high
DSI clock rates in burst mode. Increase it further to allow the DSI link
to work at up to 1Gbps lane speed. This returns the counts to defaults
as provided by datasheet.

Fixes: ea6490b02240b ("drm/bridge: tc358767: increase CLRSIPO count")
Signed-off-by: Marek Vasut <marex@denx.de>
Acked-by: Maxime Ripard <maxime@cerno.tech>
Link: https://patchwork.freedesktop.org/patch/msgid/20221016003556.406441-1-marex@denx.de
Signed-off-by: Sasha Levin <sashal@kernel.org>

ASoC: fsl_sai: initialize is_dsp_mode flag

[ Upstream commit a23924b7dd7b748fff8e305e1daf590fed2af21b ]

Initialize is_dsp_mode flag in the beginning of function
fsl_sai_set_dai_fmt_tr().

When the DAIFMT is DAIFMT_DSP_B the first time, is_dsp_mode is
true, then the second time DAIFMT is DAIFMT_I2S, is_dsp_mode
still true, which is a wrong state. So need to initialize
is_dsp_mode flag every time.

Fixes: a3f7dcc9cc03 ("ASoC: fsl-sai: Add SND_SOC_DAIFMT_DSP_A/B support.")
Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
Reviewed-by: Iuliana Prodan <iuliana.prodan@nxp.com>
Link: https://lore.kernel.org/r/1673852874-32200-1-git-send-email-shengjiu.wang@nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

scsi: qla2xxx: edif: Fix clang warning

[ Upstream commit 2f5fab1b6c3a8efc93ba52c28539c45a8d0142ad ]

clang warning:

  drivers/scsi/qla2xxx/qla_edif_bsg.h:93:12: warning: field remote_pid
  within 'struct app_pinfo_req' is less aligned than 'port_id_t' and is
  usually due to 'struct app_pinfo_req' being packed, which can lead to
  unaligned accesses [-Wunaligned-access]
  port_id_t remote_pid;
        ^
  2 warnings generated.

Remove u32 field in remote_pid to silence warning.

Reported-by: kernel test robot <lkp@intel.com>
Fixes: 7ebb336e45ef ("scsi: qla2xxx: edif: Add start + stop bsgs")
Signed-off-by: Quinn Tran <qutran@marvell.com>
Signed-off-by: Nilesh Javali <njavali@marvell.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

scsi: qla2xxx: Fix exchange oversubscription for management commands

[ Upstream commit 5f63a163ed2f12c34dd4ae9b2757962ec7bb86e5 ]

Add resource checking for management (non-I/O) commands.

Fixes: 89c72f4245a8 ("scsi: qla2xxx: Add IOCB resource tracking")
Signed-off-by: Quinn Tran <qutran@marvell.com>
Signed-off-by: Nilesh Javali <njavali@marvell.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

scsi: qla2xxx: Fix exchange oversubscription

[ Upstream commit 41e5afe51f75f2858f5563145348f6c26d307b8f ]

In large environment, it is possible to experience command timeout and
escalation of path recovery. Currently the driver does not track the number
of exchanges/commands sent to FW. If there is a delay for commands at the
head of the queue, then this will create back pressure for commands at the
back of the queue.

Check for exchange availability before command submission.

Fixes: 89c72f4245a8 ("scsi: qla2xxx: Add IOCB resource tracking")
Signed-off-by: Quinn Tran <qutran@marvell.com>
Signed-off-by: Nilesh Javali <njavali@marvell.com>
Reviewed-by: Himanshu Madhani <himanshu.madhani@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/panel-edp: fix name for IVO product id 854b

[ Upstream commit 9cce08cadc6ce8670280d0a042cf0b6d2987d9f9 ]

The actual name is R133NW4K-R0.

Fixes: 0f9fa5f58c78 ("drm/panel-edp: add IVO M133NW4J-R3 panel entry")
Signed-off-by: Abel Vesa <abel.vesa@linaro.org>
Reviewed-by: Johan Hovold <johan+linaro@kernel.org>
Tested-by: Steev Klimaszewski <steev@kali.org>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20221231142721.338643-1-abel.vesa@linaro.org
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/msm: clean event_thread->worker in case of an error

[ Upstream commit c79bb6b92defdcb834ceeeed9c1cf591beb1b71a ]

If worker creation fails, nullify the event_thread->worker, so that
msm_drm_uninit() doesn't try accessing invalid memory location. While we
are at it, remove duplicate assignment to the ret variable.

Fixes: 1041dee2178f ("drm/msm: use kthread_create_worker instead of kthread_run")
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Patchwork: https://patchwork.freedesktop.org/patch/490106/
Link: https://lore.kernel.org/r/20220617233328.1143665-2-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/vc4: hdmi: Correct interlaced timings again

[ Upstream commit 771d6539f27bd55f43d8a95d53a7eeaaffa2681c ]

The back porch timings were correct, only the sync offset was wrong.
Correct timing is now reported for 1080i and 576i, but the h offset is
incorrect for 480i for non-obvious reasons.

Fixes: fb10dc451c0f ("drm/vc4: hdmi: Correct HDMI timing registers for interlaced modes")
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Link: https://lore.kernel.org/r/20221207-rpi-hvs-crtc-misc-v1-14-1f8e0770798b@cerno.tech
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/vc4: hvs: Fix colour order for xRGB1555 on HVS5

[ Upstream commit 902973dc1a049c0d7bf0c222b8f2b3876f01b4a2 ]

Same as the xRGB8888 formats, HVS5 has managed to swap the colour
channels for the xRGB1555 formats as well. Add the relevant
config for pixel_order_hvs5.

Fixes: c54619b0bfb3 ("drm/vc4: Add support for the BCM2711 HVS5")
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Link: https://lore.kernel.org/r/20221207-rpi-hvs-crtc-misc-v1-6-1f8e0770798b@cerno.tech
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/vc4: hvs: Correct interrupt masking bit assignment for HVS5

[ Upstream commit 87551ec650bb87d35f1b29bba6a2430896e08da0 ]

HVS5 has moved the interrupt enable bits around within the
DISPCTRL register, therefore the configuration has to be updated
to account for this.

Fixes: c54619b0bfb3 ("drm/vc4: Add support for the BCM2711 HVS5")
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Link: https://lore.kernel.org/r/20221207-rpi-hvs-crtc-misc-v1-4-1f8e0770798b@cerno.tech
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/vc4: hvs: SCALER_DISPBKGND_AUTOHS is only valid on HVS4

[ Upstream commit 982ee94486863a41c6af9f2ab3f6681f72bc5c48 ]

The bit used for SCALER_DISPBKGND_AUTOHS in SCALER_DISPBKGNDX
has been repurposed on HVS5 to configure whether a display can
win back-to-back arbitration wins for the COB.

This is not desirable, therefore only select this bit on HVS4,
and explicitly clear it on HVS5.

Fixes: c54619b0bfb3 ("drm/vc4: Add support for the BCM2711 HVS5")
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Link: https://lore.kernel.org/r/20221207-rpi-hvs-crtc-misc-v1-3-1f8e0770798b@cerno.tech
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/vc4: hvs: Set AXI panic modes

[ Upstream commit df993fced230daa8452892406f3180c93ebf7e7b ]

The HVS can change AXI request mode based on how full the COB
FIFOs are.
Until now the vc4 driver has been relying on the firmware to
have set these to sensible values.

With HVS channel 2 now being used for live video, change the
panic mode for all channels to be explicitly set by the driver,
and the same for all channels.

Fixes: c54619b0bfb3 ("drm/vc4: Add support for the BCM2711 HVS5")
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Link: https://lore.kernel.org/r/20221207-rpi-hvs-crtc-misc-v1-2-1f8e0770798b@cerno.tech
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Signed-off-by: Sasha Levin <sashal@kernel.org>

pinctrl: rockchip: Fix refcount leak in rockchip_pinctrl_parse_groups

[ Upstream commit c818ae563bf99457f02e8170aabd6b174f629f65 ]

of_find_node_by_phandle() returns a node pointer with refcount incremented,
We should use of_node_put() on it when not needed anymore.
Add missing of_node_put() to avoid refcount leak.

Fixes: d3e5116119bd ("pinctrl: add pinctrl driver for Rockchip SoCs")
Signed-off-by: Miaoqian Lin <linmq006@gmail.com>
Link: https://lore.kernel.org/r/20230102112845.3982407-1-linmq006@gmail.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

pinctrl: stm32: Fix refcount leak in stm32_pctrl_get_irq_domain

[ Upstream commit dcef18c8ac40aa85bb339f64c1dd31dd458b06fb ]

of_irq_find_parent() returns a node pointer with refcount incremented,
We should use of_node_put() on it when not needed anymore.
Add missing of_node_put() to avoid refcount leak.

Fixes: d86f4d71e42a ("pinctrl: stm32: check irq controller availability at probe")
Signed-off-by: Miaoqian Lin <linmq006@gmail.com>
Link: https://lore.kernel.org/r/20230102082503.3944927-1-linmq006@gmail.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

pinctrl: qcom: pinctrl-msm8976: Correct function names for wcss pins

[ Upstream commit a7cc0e2685082a0d79baec02df184dfa83cbfac3 ]

Adjust names of function for wcss pins, also fix third gpio in bt group.

Fixes: bcd11493f0ab ("pinctrl: qcom: Add a pinctrl driver for MSM8976 and 8956")
Signed-off-by: Adam Skladowski <a39.skl@gmail.com>
Reviewed-by: Marijn Suijten <marijn.suijten@somainline.org>
Link: https://lore.kernel.org/r/20221231164250.74550-1-a39.skl@gmail.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/msm/hdmi: Add missing check for alloc_ordered_workqueue

[ Upstream commit afe4cb96153a0d8003e4e4ebd91b5c543e10df84 ]

Add check for the return value of alloc_ordered_workqueue as it may return
NULL pointer and cause NULL pointer dereference in `hdmi_hdcp.c` and
`hdmi_hpd.c`.

Fixes: c6a57a50ad56 ("drm/msm/hdmi: add hdmi hdcp support (V3)")
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/517211/
Link: https://lore.kernel.org/r/20230106023011.3985-1-jiasheng@iscas.ac.cn
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/msm/dpu: check for null return of devm_kzalloc() in dpu_writeback_init()

[ Upstream commit 21e9a838f505178e109ccb3bf19d7808eb0326f4 ]

Because of the possilble failure of devm_kzalloc(), dpu_wb_conn might
be NULL and will cause null pointer dereference later.

Therefore, it might be better to check it and directly return -ENOMEM.

Fixes: 77b001acdcfe ("drm/msm/dpu: add the writeback connector layer")
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Patchwork: https://patchwork.freedesktop.org/patch/512277/
Link: https://lore.kernel.org/r/20221119055518.179937-1-tanghui20@huawei.com
[DB: fixed typo in commit message]
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

hwmon: (ftsteutates) Fix scaling of measurements

[ Upstream commit ca8fd8c16a8b77dfcf7f6ce52d2c863220693a78 ]

A user complained that the ftsteutates driver was displaying
bogus values since its introduction. This happens because the
sensor measurements need to be scaled in order to produce
meaningful results:
- the fan speed needs to be multiplied by 60 since its in RPS
- the temperature is in degrees celsius and needs an offset of 64
- the voltage is in 1/256 of 3.3V

The offical datasheet says the voltage needs to be divided by 256,
but this is likely an off-by-one-error, since even the BIOS
devides by 255 (otherwise 3.3V could not be measured).

The voltage channels additionally need a board-specific multiplier,
however this can be done by the driver since its board-specific.

The reason the missing scaling of measurements is the way Fujitsu
used this driver when it was still out-of-tree. Back then, all
scaling was done in userspace by libsensors, even the generic one.

Tested on a Fujitsu DS3401-B1.

Fixes: 08426eda58e0 ("hwmon: Add driver for FTS BMC chip "Teutates"")
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://lore.kernel.org/r/20221224041855.83981-2-W_Armin@gmx.de
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/vc4: drop all currently held locks if deadlock happens

[ Upstream commit 479d4f0be4237ba33bc9432787aeb62c90e30f95 ]

If vc4_hdmi_reset_link() returns -EDEADLK, it means that a deadlock
happened in the locking context. This situation should be addressed by
dropping all currently held locks and block until the contended lock
becomes available. Currently, vc4 is not dealing with the deadlock
properly, producing the following output when PROVE_LOCKING is enabled:

[  825.612809] ------------[ cut here ]------------
[  825.612852] WARNING: CPU: 1 PID: 116 at drivers/gpu/drm/drm_modeset_lock.c:276 drm_modeset_drop_locks+0x60/0x68 [drm]
[  825.613458] Modules linked in: 8021q mrp garp stp llc
raspberrypi_cpufreq brcmfmac brcmutil crct10dif_ce hci_uart cfg80211
btqca btbcm bluetooth vc4 raspberrypi_hwmon snd_soc_hdmi_codec cec
clk_raspberrypi ecdh_generic drm_display_helper ecc rfkill
drm_dma_helper drm_kms_helper pwm_bcm2835 bcm2835_thermal bcm2835_rng
rng_core i2c_bcm2835 drm fuse ip_tables x_tables ipv6
[  825.613735] CPU: 1 PID: 116 Comm: kworker/1:2 Tainted: G        W 6.1.0-rc6-01399-g941aae326315 #3
[  825.613759] Hardware name: Raspberry Pi 3 Model B Rev 1.2 (DT)
[  825.613777] Workqueue: events output_poll_execute [drm_kms_helper]
[  825.614038] pstate: 60000005 (nZCv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[  825.614063] pc : drm_modeset_drop_locks+0x60/0x68 [drm]
[  825.614603] lr : drm_helper_probe_detect+0x120/0x1b4 [drm_kms_helper]
[  825.614829] sp : ffff800008313bf0
[  825.614844] x29: ffff800008313bf0 x28: ffffcd7778b8b000 x27: 0000000000000000
[  825.614883] x26: 0000000000000001 x25: 0000000000000001 x24: ffff677cc35c2758
[  825.614920] x23: ffffcd7707d01430 x22: ffffcd7707c3edc7 x21: 0000000000000001
[  825.614958] x20: 0000000000000000 x19: ffff800008313c10 x18: 000000000000b6d3
[  825.614995] x17: ffffcd777835e214 x16: ffffcd7777cef870 x15: fffff81000000000
[  825.615033] x14: 0000000000000000 x13: 0000000000000099 x12: 0000000000000002
[  825.615070] x11: 72917988020af800 x10: 72917988020af800 x9 : 72917988020af800
[  825.615108] x8 : ffff677cc665e0a8 x7 : d00a8c180000110c x6 : ffffcd77774c0054
[  825.615145] x5 : 0000000000000000 x4 : 0000000000000001 x3 : 0000000000000000
[  825.615181] x2 : ffff677cc55e1880 x1 : ffffcd7777cef8ec x0 : ffff800008313c10
[  825.615219] Call trace:
[  825.615232]  drm_modeset_drop_locks+0x60/0x68 [drm]
[  825.615773]  drm_helper_probe_detect+0x120/0x1b4 [drm_kms_helper]
[  825.616003]  output_poll_execute+0xe4/0x224 [drm_kms_helper]
[  825.616233]  process_one_work+0x2b4/0x618
[  825.616264]  worker_thread+0x24c/0x464
[  825.616288]  kthread+0xec/0x110
[  825.616310]  ret_from_fork+0x10/0x20
[  825.616335] irq event stamp: 7634
[  825.616349] hardirqs last  enabled at (7633): [<ffffcd777831ee90>] _raw_spin_unlock_irq+0x3c/0x78
[  825.616384] hardirqs last disabled at (7634): [<ffffcd7778315a78>] __schedule+0x134/0x9f0
[  825.616411] softirqs last  enabled at (7630): [<ffffcd7707aacea0>] local_bh_enable+0x4/0x30 [ipv6]
[  825.617019] softirqs last disabled at (7618): [<ffffcd7707aace70>] local_bh_disable+0x4/0x30 [ipv6]
[  825.617586] ---[ end trace 0000000000000000 ]---

Therefore, deal with the deadlock as suggested by [1], using the
function drm_modeset_backoff().

[1] https://docs.kernel.org/gpu/drm-kms.html?highlight=kms#kms-locking

Fixes: 6bed2ea3cb38 ("drm/vc4: hdmi: Reset link on hotplug")
Reported-by: Stefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: Maíra Canal <mcanal@igalia.com>
Tested-by: Stefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://patchwork.freedesktop.org/patch/msgid/20221229194638.178712-1-mcanal@igalia.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

gpu: ipu-v3: common: Add of_node_put() for reference returned by of_graph_get_port_by_id()

[ Upstream commit 9afdf98cfdfa2ba8ec068cf08c5fcdc1ed8daf3f ]

In ipu_add_client_devices(), we need to call of_node_put() for
reference returned by of_graph_get_port_by_id() in fail path.

Fixes: 17e052175039 ("gpu: ipu-v3: Do not bail out on missing optional port nodes")
Signed-off-by: Liang He <windhl@126.com>
Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de>
Link: https://lore.kernel.org/r/20220720152227.1288413-1-windhl@126.com
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20220720152227.1288413-1-windhl@126.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm: tidss: Fix pixel format definition

[ Upstream commit 2df0433b18f2735a49d2c3a968b40fa2881137c0 ]

There was a long-standing bug from a typo that created 2 ARGB1555 and
ABGR1555 pixel format entries. Weston 10 has a sanity check that alerted
me to this issue.

According to the Supported Pixel Data formats table we have the later
entries should have been for Alpha-X instead.

Signed-off-by: Randolph Sapp <rs@ti.com>
Fixes: 32a1795f57eecc ("drm/tidss: New driver for TI Keystone platform Display SubSystem")
Reviewed-by: Aradhya Bhatia <a-bhatia1@ti.com>
Acked-by: Andrew Davis <afd@ti.com>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20221202001803.1765805-1-rs@ti.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/vc4: dpi: Fix format mapping for RGB565

[ Upstream commit 0870d86eac8a9abd89a0be1b719d5dc5bac936f0 ]

The mapping is incorrect for RGB565_1X16 as it should be
DPI_FORMAT_18BIT_666_RGB_1 instead of DPI_FORMAT_18BIT_666_RGB_3.

Fixes: 08302c35b59d ("drm/vc4: Add DPI driver")
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Link: https://lore.kernel.org/r/20221013-rpi-dpi-improvements-v3-7-eb76e26a772d@cerno.tech
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/vkms: Fix null-ptr-deref in vkms_release()

[ Upstream commit 2fe2a8f40c21161ffe7653cc234e7934db5b7cc5 ]

A null-ptr-deref is triggered when it tries to destroy the workqueue in
vkms->output.composer_workq in vkms_release().

 KASAN: null-ptr-deref in range [0x0000000000000118-0x000000000000011f]
 CPU: 5 PID: 17193 Comm: modprobe Not tainted 6.0.0-11331-gd465bff130bf #24
 RIP: 0010:destroy_workqueue+0x2f/0x710
 ...
 Call Trace:
  <TASK>
  ? vkms_config_debugfs_init+0x50/0x50 [vkms]
  __devm_drm_dev_alloc+0x15a/0x1c0 [drm]
  vkms_init+0x245/0x1000 [vkms]
  do_one_initcall+0xd0/0x4f0
  do_init_module+0x1a4/0x680
  load_module+0x6249/0x7110
  __do_sys_finit_module+0x140/0x200
  do_syscall_64+0x35/0x80
  entry_SYSCALL_64_after_hwframe+0x46/0xb0

The reason is that an OOM happened which triggers the destroy of the
workqueue, however, the workqueue is alloced in the later process,
thus a null-ptr-deref happened. A simple call graph is shown as below:

 vkms_init()
  vkms_create()
    devm_drm_dev_alloc()
      __devm_drm_dev_alloc()
        devm_drm_dev_init()
          devm_add_action_or_reset()
            devm_add_action() # an error happened
            devm_drm_dev_init_release()
              drm_dev_put()
                kref_put()
                  drm_dev_release()
                    vkms_release()
                      destroy_workqueue() # null-ptr-deref happened
    vkms_modeset_init()
      vkms_output_init()
        vkms_crtc_init() # where the workqueue get allocated

Fix this by checking if composer_workq is NULL before passing it to
the destroy_workqueue() in vkms_release().

Fixes: 6c234fe37c57 ("drm/vkms: Implement CRC debugfs API")
Signed-off-by: Yuan Can <yuancan@huawei.com>
Reviewed-by: Melissa Wen <mwen@igalia.com>
Signed-off-by: Melissa Wen <melissa.srw@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20221101065156.41584-3-yuancan@huawei.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/vkms: Fix memory leak in vkms_init()

[ Upstream commit 0d0b368b9d104b437e1f4850ae94bdb9a3601e89 ]

A memory leak was reported after the vkms module install failed.

unreferenced object 0xffff88810bc28520 (size 16):
  comm "modprobe", pid 9662, jiffies 4298009455 (age 42.590s)
  hex dump (first 16 bytes):
    01 01 00 64 81 88 ff ff 00 00 dc 0a 81 88 ff ff  ...d............
  backtrace:
    [<00000000e7561ff8>] kmalloc_trace+0x27/0x60
    [<000000000b1954a0>] 0xffffffffc45200a9
    [<00000000abbf1da0>] do_one_initcall+0xd0/0x4f0
    [<000000001505ee87>] do_init_module+0x1a4/0x680
    [<00000000958079ad>] load_module+0x6249/0x7110
    [<00000000117e4696>] __do_sys_finit_module+0x140/0x200
    [<00000000f74b12d2>] do_syscall_64+0x35/0x80
    [<000000008fc6fcde>] entry_SYSCALL_64_after_hwframe+0x46/0xb0

The reason is that the vkms_init() returns without checking the return
value of vkms_create(), and if the vkms_create() failed, the config
allocated at the beginning of vkms_init() is leaked.

 vkms_init()
   config = kmalloc(...) # config allocated
   ...
   return vkms_create() # vkms_create failed and config is leaked

Fix this problem by checking return value of vkms_create() and free the
config if error happened.

Fixes: 2df7af93fdad ("drm/vkms: Add vkms_config type")
Signed-off-by: Yuan Can <yuancan@huawei.com>
Reviewed-by: Melissa Wen <mwen@igalia.com>
Signed-off-by: Melissa Wen <melissa.srw@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20221101065156.41584-2-yuancan@huawei.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: megachips: Fix error handling in i2c_register_driver()

[ Upstream commit 4ecff954c370b82bce45bdca2846c5c5563e8a8a ]

A problem about insmod megachips-stdpxxxx-ge-b850v3-fw.ko failed is
triggered with the following log given:

[ 4497.981497] Error: Driver 'stdp4028-ge-b850v3-fw' is already registered, aborting...
insmod: ERROR: could not insert module megachips-stdpxxxx-ge-b850v3-fw.ko: Device or resource busy

The reason is that stdpxxxx_ge_b850v3_init() returns i2c_add_driver()
directly without checking its return value, if i2c_add_driver() failed,
it returns without calling i2c_del_driver() on the previous i2c driver,
resulting the megachips-stdpxxxx-ge-b850v3-fw can never be installed
later.
A simple call graph is shown as below:

 stdpxxxx_ge_b850v3_init()
   i2c_add_driver(&stdp4028_ge_b850v3_fw_driver)
   i2c_add_driver(&stdp2690_ge_b850v3_fw_driver)
     i2c_register_driver()
       driver_register()
         bus_add_driver()
           priv = kzalloc(...) # OOM happened
   # return without delete stdp4028_ge_b850v3_fw_driver

Fix by calling i2c_del_driver() on stdp4028_ge_b850v3_fw_driver when
i2c_add_driver() returns error.

Fixes: fcfa0ddc18ed ("drm/bridge: Drivers for megachips-stdpxxxx-ge-b850v3-fw (LVDS-DP++)")
Signed-off-by: Yuan Can <yuancan@huawei.com>
Reviewed-by: Andrzej Hajda <andrzej.hajda@intel.com>
Tested-by: Ian Ray <ian.ray@ge.com>
Signed-off-by: Robert Foss <robert.foss@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20221108091226.114524-1-yuancan@huawei.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm: mxsfb: DRM_MXSFB should depend on ARCH_MXS || ARCH_MXC

[ Upstream commit 7783cc67862f9166c901bfa0f80b717aa8d354dd ]

Freescale/NXP i.MX LCDIF and eLCDIF LCD controllers are only present on
Freescale/NXP i.MX SoCs.  Hence add a dependency on ARCH_MXS ||
ARCH_MXC, to prevent asking the user about this driver when configuring
a kernel without Freescale/NXP i.MX support.

Fixes: 45d59d704080cc0c ("drm: Add new driver for MXSFB controller")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Marek Vasut <marex@denx.de>
Link: https://patchwork.freedesktop.org/patch/msgid/98e74779ca2bc575d91afff03369e86b080c01ac.1669046358.git.geert+renesas@glider.be
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm: mxsfb: DRM_IMX_LCDIF should depend on ARCH_MXC

[ Upstream commit 10ef5f2992006720318b9886961820155b3750fd ]

The Freescale/NXP i.MX LCDIFv3 LCD controller is only present on
Freescale/NXP i.MX SoCs.  Hence add a dependency on ARCH_MXC, to prevent
asking the user about this driver when configuring a kernel without
Freescale/NXP i.MX support.

Fixes: 9db35bb349a0ef32 ("drm: lcdif: Add support for i.MX8MP LCDIF variant")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Marek Vasut <marex@denx.de>
Link: https://patchwork.freedesktop.org/patch/msgid/6103c1aa65a7888c12d351ae63f29850f29f42b9.1669046403.git.geert+renesas@glider.be
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: ti-sn65dsi83: Fix delay after reset deassert to match spec

[ Upstream commit 4b03d5e0d3e86ee492d54254927d020dc0fe8acf ]

The datasheet specifies a delay of 10 milliseconds, but the current
driver only waits for 1 ms. Fix this to make sure the initialization
sequence meets the spec.

Fixes: ceb515ba29ba ("drm/bridge: ti-sn65dsi83: Add TI SN65DSI83 and SN65DSI84 driver")
Signed-off-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Reviewed-by: Alexander Stein <alexander.stein@ew.tq-group.com>
Signed-off-by: Robert Foss <robert.foss@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20221122081219.20143-1-frieder@fris.de
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/fourcc: Add missing big-endian XRGB1555 and RGB565 formats

[ Upstream commit 6fb6c979ca628583d4d0c59a0f8ff977e581ecc0 ]

As of commit eae06120f1974e1a ("drm: refuse ADDFB2 ioctl for broken
bigendian drivers"), drivers must set the
quirk_addfb_prefer_host_byte_order quirk to make the drm_mode_addfb()
compat code work correctly on big-endian machines.

While that works fine for big-endian XRGB8888 and ARGB8888, which are
mapped to the existing little-endian BGRX8888 and BGRA8888 formats, it
does not work for big-endian XRGB1555 and RGB565, as the latter are not
listed in the format database.

Fix this by adding the missing formats.  Limit this to big-endian
platforms, as there is currently no need to support these formats on
little-endian platforms.

Fixes: 6960e6da9cec3f66 ("drm: fix drm_mode_addfb() on big endian machines.")
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/3ee1f8144feb96c28742b22384189f1f83bcfc1a.1669221671.git.geert@linux-m68k.org
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm: Fix potential null-ptr-deref due to drmm_mode_config_init()

[ Upstream commit 834c23e4f798dcdc8af251b3c428ceef94741991 ]

drmm_mode_config_init() will call drm_mode_create_standard_properties()
and won't check the ret value. When drm_mode_create_standard_properties()
failed due to alloc, property will be a NULL pointer and may causes the
null-ptr-deref. Fix the null-ptr-deref by adding the ret value check.

Found null-ptr-deref while testing insert module bochs:
general protection fault, probably for non-canonical address
    0xdffffc000000000c: 0000 [#1] SMP KASAN PTI
KASAN: null-ptr-deref in range [0x0000000000000060-0x0000000000000067]
CPU: 3 PID: 249 Comm: modprobe Not tainted 6.1.0-rc1+ #364
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS
rel-1.15.0-0-g2dd4b9b3f840-prebuilt.qemu.org 04/01/2014
RIP: 0010:drm_object_attach_property+0x73/0x3c0 [drm]
Call Trace:
 <TASK>
 __drm_connector_init+0xb6c/0x1100 [drm]
 bochs_pci_probe.cold.11+0x4cb/0x7fe [bochs]
 pci_device_probe+0x17d/0x340
 really_probe+0x1db/0x5d0
 __driver_probe_device+0x1e7/0x250
 driver_probe_device+0x4a/0x120
 __driver_attach+0xcd/0x2c0
 bus_for_each_dev+0x11a/0x1b0
 bus_add_driver+0x3d7/0x500
 driver_register+0x18e/0x320
 do_one_initcall+0xc4/0x3e0
 do_init_module+0x1b4/0x630
 load_module+0x5dca/0x7230
 __do_sys_finit_module+0x100/0x170
 do_syscall_64+0x3f/0x90
 entry_SYSCALL_64_after_hwframe+0x63/0xcd
RIP: 0033:0x7ff65af9f839

Fixes: 6b4959f43a04 ("drm/atomic: atomic plane properties")
Signed-off-by: Shang XiaoJing <shangxiaojing@huawei.com>
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20221118021651.2460-1-shangxiaojing@huawei.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

sefltests: netdevsim: wait for devlink instance after netns removal

[ Upstream commit f922c7b1c1c45740d329bf248936fdb78c0cff6e ]

When devlink instance is put into network namespace and that network
namespace gets deleted, devlink instance is moved back into init_ns.
This is done as a part of cleanup_net() routine. Since cleanup_net()
is called asynchronously from workqueue, there is no guarantee that
the devlink instance move is done after "ip netns del" returns.

So fix this race by making sure that the devlink instance is present
before any other operation.

Reported-by: Amir Tzin <amirtz@nvidia.com>
Fixes: b74c37fd35a2 ("selftests: netdevsim: add tests for devlink reload with resources")
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Link: https://lore.kernel.org/r/20230220132336.198597-1-jiri@resnulli.us
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

selftest: fib_tests: Always cleanup before exit

[ Upstream commit b60417a9f2b890a8094477b2204d4f73c535725e ]

Usage of `set -e` before executing a command causes immediate exit
on failure, without cleanup up the resources allocated at setup.
This can affect the next tests that use the same resources,
leading to a chain of failures.

A simple fix is to always call cleanup function when the script exists.
This approach is already used by other existing tests.

Fixes: 1056691b2680 ("selftests: fib_tests: Make test results more verbose")
Signed-off-by: Roxana Nicolescu <roxana.nicolescu@canonical.com>
Link: https://lore.kernel.org/r/20230220110400.26737-2-roxana.nicolescu@canonical.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net/mlx4_en: Introduce flexible array to silence overflow warning

[ Upstream commit f8f185e39b4de91bc5235e5be0d829bea69d9b06 ]

The call "skb_copy_from_linear_data(skb, inl + 1, spc)" triggers a FORTIFY
memcpy() warning on ppc64 platform:

In function ‘fortify_memcpy_chk’,
    inlined from ‘skb_copy_from_linear_data’ at ./include/linux/skbuff.h:4029:2,
    inlined from ‘build_inline_wqe’ at drivers/net/ethernet/mellanox/mlx4/en_tx.c:722:4,
    inlined from ‘mlx4_en_xmit’ at drivers/net/ethernet/mellanox/mlx4/en_tx.c:1066:3:
./include/linux/fortify-string.h:513:25: error: call to ‘__write_overflow_field’ declared with
attribute warning: detected write beyond size of field (1st parameter); maybe use struct_group()?
[-Werror=attribute-warning]
  513 |                         __write_overflow_field(p_size_field, size);
      |                         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Same behaviour on x86 you can get if you use "__always_inline" instead of
"inline" for skb_copy_from_linear_data() in skbuff.h

The call here copies data into inlined tx destricptor, which has 104
bytes (MAX_INLINE) space for data payload. In this case "spc" is known
in compile-time but the destination is used with hidden knowledge
(real structure of destination is different from that the compiler
can see). That cause the fortify warning because compiler can check
bounds, but the real bounds are different.  "spc" can't be bigger than
64 bytes (MLX4_INLINE_ALIGN), so the data can always fit into inlined
tx descriptor. The fact that "inl" points into inlined tx descriptor is
determined earlier in mlx4_en_xmit().

Avoid confusing the compiler with "inl + 1" constructions to get to past
the inl header by introducing a flexible array "data" to the struct so
that the compiler can see that we are not dealing with an array of inl
structs, but rather, arbitrary data following the structure. There are
no changes to the structure layout reported by pahole, and the resulting
machine code is actually smaller.

Reported-by: Josef Oskera <joskera@redhat.com>
Link: https://lore.kernel.org/lkml/20230217094541.2362873-1-joskera@redhat.com
Fixes: f68f2ff91512 ("fortify: Detect struct member overflows in memcpy() at compile-time")
Cc: Yishai Hadas <yishaih@nvidia.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Link: https://lore.kernel.org/r/20230218183842.never.954-kees@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net: lan966x: Fix possible deadlock inside PTP

[ Upstream commit 3a70e0d4c9d74cb00f7c0ec022f5599f9f7ba07d ]

When doing timestamping in lan966x and having PROVE_LOCKING
enabled the following warning is shown.

========================================================
WARNING: possible irq lock inversion dependency detected
6.2.0-rc7-01749-gc54e1f7f7e36 #2786 Tainted: G                 N
--------------------------------------------------------
swapper/0/0 just changed the state of lock:
c2609f50 (_xmit_ETHER#2){+.-.}-{2:2}, at: sch_direct_xmit+0x16c/0x2e8
but this lock took another, SOFTIRQ-unsafe lock in the past:
 (&lan966x->ptp_ts_id_lock){+.+.}-{2:2}

and interrupts could create inverse lock ordering between them.

other info that might help us debug this:
 Possible interrupt unsafe locking scenario:

       CPU0                    CPU1
       ----                    ----
  lock(&lan966x->ptp_ts_id_lock);
                               local_irq_disable();
                               lock(_xmit_ETHER#2);
                               lock(&lan966x->ptp_ts_id_lock);
  <Interrupt>
    lock(_xmit_ETHER#2);

 *** DEADLOCK ***

5 locks held by swapper/0/0:
 #0: c1001e18 ((&ndev->rs_timer)){+.-.}-{0:0}, at: call_timer_fn+0x0/0x33c
 #1: c105e7c4 (rcu_read_lock){....}-{1:2}, at: ndisc_send_skb+0x134/0x81c
 #2: c105e7d8 (rcu_read_lock_bh){....}-{1:2}, at: ip6_finish_output2+0x17c/0xc64
 #3: c105e7d8 (rcu_read_lock_bh){....}-{1:2}, at: __dev_queue_xmit+0x4c/0x1224
 #4: c3056174 (dev->qdisc_tx_busylock ?: &qdisc_tx_busylock){+...}-{2:2}, at: __dev_queue_xmit+0x354/0x1224

the shortest dependencies between 2nd lock and 1st lock:
 -> (&lan966x->ptp_ts_id_lock){+.+.}-{2:2} {
    HARDIRQ-ON-W at:
                      lock_acquire.part.0+0xb0/0x248
                      _raw_spin_lock+0x38/0x48
                      lan966x_ptp_irq_handler+0x164/0x2a8
                      irq_thread_fn+0x1c/0x78
                      irq_thread+0x130/0x278
                      kthread+0xec/0x110
                      ret_from_fork+0x14/0x28
    SOFTIRQ-ON-W at:
                      lock_acquire.part.0+0xb0/0x248
                      _raw_spin_lock+0x38/0x48
                      lan966x_ptp_irq_handler+0x164/0x2a8
                      irq_thread_fn+0x1c/0x78
                      irq_thread+0x130/0x278
                      kthread+0xec/0x110
                      ret_from_fork+0x14/0x28
    INITIAL USE at:
                     lock_acquire.part.0+0xb0/0x248
                     _raw_spin_lock_irqsave+0x4c/0x68
                     lan966x_ptp_txtstamp_request+0x128/0x1cc
                     lan966x_port_xmit+0x224/0x43c
                     dev_hard_start_xmit+0xa8/0x2f0
                     sch_direct_xmit+0x108/0x2e8
                     __dev_queue_xmit+0x41c/0x1224
                     packet_sendmsg+0xdb4/0x134c
                     __sys_sendto+0xd0/0x154
                     sys_send+0x18/0x20
                     ret_fast_syscall+0x0/0x1c
  }
  ... key      at: [<c174ba0c>] __key.2+0x0/0x8
  ... acquired at:
   _raw_spin_lock_irqsave+0x4c/0x68
   lan966x_ptp_txtstamp_request+0x128/0x1cc
   lan966x_port_xmit+0x224/0x43c
   dev_hard_start_xmit+0xa8/0x2f0
   sch_direct_xmit+0x108/0x2e8
   __dev_queue_xmit+0x41c/0x1224
   packet_sendmsg+0xdb4/0x134c
   __sys_sendto+0xd0/0x154
   sys_send+0x18/0x20
   ret_fast_syscall+0x0/0x1c

-> (_xmit_ETHER#2){+.-.}-{2:2} {
   HARDIRQ-ON-W at:
                    lock_acquire.part.0+0xb0/0x248
                    _raw_spin_lock+0x38/0x48
                    netif_freeze_queues+0x38/0x68
                    dev_deactivate_many+0xac/0x388
                    dev_deactivate+0x38/0x6c
                    linkwatch_do_dev+0x70/0x8c
                    __linkwatch_run_queue+0xd4/0x1e8
                    linkwatch_event+0x24/0x34
                    process_one_work+0x284/0x744
                    worker_thread+0x28/0x4bc
                    kthread+0xec/0x110
                    ret_from_fork+0x14/0x28
   IN-SOFTIRQ-W at:
                    lock_acquire.part.0+0xb0/0x248
                    _raw_spin_lock+0x38/0x48
                    sch_direct_xmit+0x16c/0x2e8
                    __dev_queue_xmit+0x41c/0x1224
                    ip6_finish_output2+0x5f4/0xc64
                    ndisc_send_skb+0x4cc/0x81c
                    addrconf_rs_timer+0xb0/0x2f8
                    call_timer_fn+0xb4/0x33c
                    expire_timers+0xb4/0x10c
                    run_timer_softirq+0xf8/0x2a8
                    __do_softirq+0xd4/0x5fc
                    __irq_exit_rcu+0x138/0x17c
                    irq_exit+0x8/0x28
                    __irq_svc+0x90/0xbc
                    arch_cpu_idle+0x30/0x3c
                    default_idle_call+0x44/0xac
                    do_idle+0xc8/0x138
                    cpu_startup_entry+0x18/0x1c
                    rest_init+0xcc/0x168
                    arch_post_acpi_subsys_init+0x0/0x8
   INITIAL USE at:
                   lock_acquire.part.0+0xb0/0x248
                   _raw_spin_lock+0x38/0x48
                   netif_freeze_queues+0x38/0x68
                   dev_deactivate_many+0xac/0x388
                   dev_deactivate+0x38/0x6c
                   linkwatch_do_dev+0x70/0x8c
                   __linkwatch_run_queue+0xd4/0x1e8
                   linkwatch_event+0x24/0x34
                   process_one_work+0x284/0x744
                   worker_thread+0x28/0x4bc
                   kthread+0xec/0x110
                   ret_from_fork+0x14/0x28
 }
 ... key      at: [<c175974c>] netdev_xmit_lock_key+0x8/0x1c8
 ... acquired at:
   __lock_acquire+0x978/0x2978
   lock_acquire.part.0+0xb0/0x248
   _raw_spin_lock+0x38/0x48
   sch_direct_xmit+0x16c/0x2e8
   __dev_queue_xmit+0x41c/0x1224
   ip6_finish_output2+0x5f4/0xc64
   ndisc_send_skb+0x4cc/0x81c
   addrconf_rs_timer+0xb0/0x2f8
   call_timer_fn+0xb4/0x33c
   expire_timers+0xb4/0x10c
   run_timer_softirq+0xf8/0x2a8
   __do_softirq+0xd4/0x5fc
   __irq_exit_rcu+0x138/0x17c
   irq_exit+0x8/0x28
   __irq_svc+0x90/0xbc
   arch_cpu_idle+0x30/0x3c
   default_idle_call+0x44/0xac
   do_idle+0xc8/0x138
   cpu_startup_entry+0x18/0x1c
   rest_init+0xcc/0x168
   arch_post_acpi_subsys_init+0x0/0x8

stack backtrace:
CPU: 0 PID: 0 Comm: swapper/0 Tainted: G                 N 6.2.0-rc7-01749-gc54e1f7f7e36 #2786
Hardware name: Generic DT based system
 unwind_backtrace from show_stack+0x10/0x14
 show_stack from dump_stack_lvl+0x58/0x70
 dump_stack_lvl from mark_lock.part.0+0x59c/0x93c
 mark_lock.part.0 from __lock_acquire+0x978/0x2978
 __lock_acquire from lock_acquire.part.0+0xb0/0x248
 lock_acquire.part.0 from _raw_spin_lock+0x38/0x48
 _raw_spin_lock from sch_direct_xmit+0x16c/0x2e8
 sch_direct_xmit from __dev_queue_xmit+0x41c/0x1224
 __dev_queue_xmit from ip6_finish_output2+0x5f4/0xc64
 ip6_finish_output2 from ndisc_send_skb+0x4cc/0x81c
 ndisc_send_skb from addrconf_rs_timer+0xb0/0x2f8
 addrconf_rs_timer from call_timer_fn+0xb4/0x33c
 call_timer_fn from expire_timers+0xb4/0x10c
 expire_timers from run_timer_softirq+0xf8/0x2a8
 run_timer_softirq from __do_softirq+0xd4/0x5fc
 __do_softirq from __irq_exit_rcu+0x138/0x17c
 __irq_exit_rcu from irq_exit+0x8/0x28
 irq_exit from __irq_svc+0x90/0xbc
Exception stack(0xc1001f20 to 0xc1001f68)
1f20: ffffffff ffffffff 00000001 c011f840 c100e000 c100e000 c1009314 c1009370
1f40: c10f0c1a c0d5e564 c0f5da8c 00000000 00000000 c1001f70 c010f0bc c010f0c0
1f60: 600f0013 ffffffff
 __irq_svc from arch_cpu_idle+0x30/0x3c
 arch_cpu_idle from default_idle_call+0x44/0xac
 default_idle_call from do_idle+0xc8/0x138
 do_idle from cpu_startup_entry+0x18/0x1c
 cpu_startup_entry from rest_init+0xcc/0x168
 rest_init from arch_post_acpi_subsys_init+0x0/0x8

Fix this by using spin_lock_irqsave/spin_lock_irqrestore also
inside lan966x_ptp_irq_handler.

Fixes: e85a96e48e33 ("net: lan966x: Add support for ptp interrupts")
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Link: https://lore.kernel.org/r/20230217210917.2649365-1-horatiu.vultur@microchip.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net: bcmgenet: fix MoCA LED control

[ Upstream commit a7515af9fb8f0890fe540b108def4a86b9e8330a ]

When the bcmgenet_mii_config() code was refactored it was missed
that the LED control for the MoCA interface got overwritten by
the port_ctrl value. Its previous programming is restored here.

Fixes: 4f8d81b77e66 ("net: bcmgenet: Refactor register access in bcmgenet_mii_config")
Signed-off-by: Doug Berger <opendmb@gmail.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

l2tp: Avoid possible recursive deadlock in l2tp_tunnel_register()

[ Upstream commit 9ca5e7ecab064f1f47da07f7c1ddf40e4bc0e5ac ]

When a file descriptor of pppol2tp socket is passed as file descriptor
of UDP socket, a recursive deadlock occurs in l2tp_tunnel_register().
This situation is reproduced by the following program:

int main(void)
{
int sock;
struct sockaddr_pppol2tp addr;

sock = socket(AF_PPPOX, SOCK_DGRAM, PX_PROTO_OL2TP);
if (sock < 0) {
perror("socket");
return 1;
}

addr.sa_family = AF_PPPOX;
addr.sa_protocol = PX_PROTO_OL2TP;
addr.pppol2tp.pid = 0;
addr.pppol2tp.fd = sock;
addr.pppol2tp.addr.sin_family = PF_INET;
addr.pppol2tp.addr.sin_port = htons(0);
addr.pppol2tp.addr.sin_addr.s_addr = inet_addr("192.168.0.1");
addr.pppol2tp.s_tunnel = 1;
addr.pppol2tp.s_session = 0;
addr.pppol2tp.d_tunnel = 0;
addr.pppol2tp.d_session = 0;

if (connect(sock, (const struct sockaddr *)&addr, sizeof(addr)) < 0) {
perror("connect");
return 1;
}

return 0;
}

This program causes the following lockdep warning:

 ============================================
 WARNING: possible recursive locking detected
 6.2.0-rc5-00205-gc96618275234 #56 Not tainted
 --------------------------------------------
 repro/8607 is trying to acquire lock:
 ffff8880213c8130 (sk_lock-AF_PPPOX){+.+.}-{0:0}, at: l2tp_tunnel_register+0x2b7/0x11c0

 but task is already holding lock:
 ffff8880213c8130 (sk_lock-AF_PPPOX){+.+.}-{0:0}, at: pppol2tp_connect+0xa82/0x1a30

 other info that might help us debug this:
  Possible unsafe locking scenario:

        CPU0
        ----
   lock(sk_lock-AF_PPPOX);
   lock(sk_lock-AF_PPPOX);

  *** DEADLOCK ***

  May be due to missing lock nesting notation

 1 lock held by repro/8607:
  #0: ffff8880213c8130 (sk_lock-AF_PPPOX){+.+.}-{0:0}, at: pppol2tp_connect+0xa82/0x1a30

 stack backtrace:
 CPU: 0 PID: 8607 Comm: repro Not tainted 6.2.0-rc5-00205-gc96618275234 #56
 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.1-2.fc37 04/01/2014
 Call Trace:
  <TASK>
  dump_stack_lvl+0x100/0x178
  __lock_acquire.cold+0x119/0x3b9
  ? lockdep_hardirqs_on_prepare+0x410/0x410
  lock_acquire+0x1e0/0x610
  ? l2tp_tunnel_register+0x2b7/0x11c0
  ? lock_downgrade+0x710/0x710
  ? __fget_files+0x283/0x3e0
  lock_sock_nested+0x3a/0xf0
  ? l2tp_tunnel_register+0x2b7/0x11c0
  l2tp_tunnel_register+0x2b7/0x11c0
  ? sprintf+0xc4/0x100
  ? l2tp_tunnel_del_work+0x6b0/0x6b0
  ? debug_object_deactivate+0x320/0x320
  ? lockdep_init_map_type+0x16d/0x7a0
  ? lockdep_init_map_type+0x16d/0x7a0
  ? l2tp_tunnel_create+0x2bf/0x4b0
  ? l2tp_tunnel_create+0x3c6/0x4b0
  pppol2tp_connect+0x14e1/0x1a30
  ? pppol2tp_put_sk+0xd0/0xd0
  ? aa_sk_perm+0x2b7/0xa80
  ? aa_af_perm+0x260/0x260
  ? bpf_lsm_socket_connect+0x9/0x10
  ? pppol2tp_put_sk+0xd0/0xd0
  __sys_connect_file+0x14f/0x190
  __sys_connect+0x133/0x160
  ? __sys_connect_file+0x190/0x190
  ? lockdep_hardirqs_on+0x7d/0x100
  ? ktime_get_coarse_real_ts64+0x1b7/0x200
  ? ktime_get_coarse_real_ts64+0x147/0x200
  ? __audit_syscall_entry+0x396/0x500
  __x64_sys_connect+0x72/0xb0
  do_syscall_64+0x38/0xb0
  entry_SYSCALL_64_after_hwframe+0x63/0xcd

This patch fixes the issue by getting/creating the tunnel before
locking the pppol2tp socket.

Fixes: 0b2c59720e65 ("l2tp: close all race conditions in l2tp_tunnel_register()")
Cc: Cong Wang <cong.wang@bytedance.com>
Signed-off-by: Shigeru Yoshida <syoshida@redhat.com>
Reviewed-by: Guillaume Nault <gnault@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

selftests/net: Interpret UDP_GRO cmsg data as an int value

[ Upstream commit 436864095a95fcc611c20c44a111985fa9848730 ]

Data passed to user-space with a (SOL_UDP, UDP_GRO) cmsg carries an
int (see udp_cmsg_recv), not a u16 value, as strace confirms:

  recvmsg(8, {msg_name=...,
              msg_iov=[{iov_base="\0\0..."..., iov_len=96000}],
              msg_iovlen=1,
              msg_control=[{cmsg_len=20,         <-- sizeof(cmsghdr) + 4
                            cmsg_level=SOL_UDP,
                            cmsg_type=0x68}],    <-- UDP_GRO
                            msg_controllen=24,
                            msg_flags=0}, 0) = 11200

Interpreting the data as an u16 value won't work on big-endian platforms.
Since it is too late to back out of this API decision [1], fix the test.

[1]: https://lore.kernel.org/netdev/20230131174601.203127-1-jakub@cloudflare.com/

Fixes: 3327a9c46352 ("selftests: add functionals test for UDP GRO")
Suggested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Sitnicki <jakub@cloudflare.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net/smc: fix application data exception

[ Upstream commit 475f9ff63ee8c296aa46c6e9e9ad9bdd301c6bdf ]

There is a certain probability that following
exceptions will occur in the wrk benchmark test:

Running 10s test @ http://11.213.45.6:80
  8 threads and 64 connections
  Thread Stats   Avg      Stdev     Max   +/- Stdev
    Latency     3.72ms   13.94ms 245.33ms   94.17%
    Req/Sec     1.96k   713.67     5.41k    75.16%
  155262 requests in 10.10s, 23.10MB read
Non-2xx or 3xx responses: 3

We will find that the error is HTTP 400 error, which is a serious
exception in our test, which means the application data was
corrupted.

Consider the following scenarios:

CPU0                            CPU1

buf_desc->used = 0;
                                cmpxchg(buf_desc->used, 0, 1)
                                deal_with(buf_desc)

memset(buf_desc->cpu_addr,0);

This will cause the data received by a victim connection to be cleared,
thus triggering an HTTP 400 error in the server.

This patch exchange the order between clear used and memset, add
barrier to ensure memory consistency.

Fixes: 1c5526968e27 ("net/smc: Clear memory when release and reuse buffer")
Signed-off-by: D. Wythe <alibuda@linux.alibaba.com>
Reviewed-by: Wenjia Zhang <wenjia@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net/smc: fix potential panic dues to unprotected smc_llc_srv_add_link()

[ Upstream commit e40b801b3603a8f90b46acbacdea3505c27f01c0 ]

There is a certain chance to trigger the following panic:

PID: 5900   TASK: ffff88c1c8af4100  CPU: 1   COMMAND: "kworker/1:48"
 #0 [ffff9456c1cc79a0] machine_kexec at ffffffff870665b7
 #1 [ffff9456c1cc79f0] __crash_kexec at ffffffff871b4c7a
 #2 [ffff9456c1cc7ab0] crash_kexec at ffffffff871b5b60
 #3 [ffff9456c1cc7ac0] oops_end at ffffffff87026ce7
 #4 [ffff9456c1cc7ae0] page_fault_oops at ffffffff87075715
 #5 [ffff9456c1cc7b58] exc_page_fault at ffffffff87ad0654
 #6 [ffff9456c1cc7b80] asm_exc_page_fault at ffffffff87c00b62
    [exception RIP: ib_alloc_mr+19]
    RIP: ffffffffc0c9cce3  RSP: ffff9456c1cc7c38  RFLAGS: 00010202
    RAX: 0000000000000000  RBX: 0000000000000002  RCX: 0000000000000004
    RDX: 0000000000000010  RSI: 0000000000000000  RDI: 0000000000000000
    RBP: ffff88c1ea281d00   R8: 000000020a34ffff   R9: ffff88c1350bbb20
    R10: 0000000000000000  R11: 0000000000000001  R12: 0000000000000000
    R13: 0000000000000010  R14: ffff88c1ab040a50  R15: ffff88c1ea281d00
    ORIG_RAX: ffffffffffffffff  CS: 0010  SS: 0018
 #7 [ffff9456c1cc7c60] smc_ib_get_memory_region at ffffffffc0aff6df [smc]
 #8 [ffff9456c1cc7c88] smcr_buf_map_link at ffffffffc0b0278c [smc]
 #9 [ffff9456c1cc7ce0] __smc_buf_create at ffffffffc0b03586 [smc]

The reason here is that when the server tries to create a second link,
smc_llc_srv_add_link() has no protection and may add a new link to
link group. This breaks the security environment protected by
llc_conf_mutex.

Fixes: 2d2209f20189 ("net/smc: first part of add link processing as SMC server")
Signed-off-by: D. Wythe <alibuda@linux.alibaba.com>
Reviewed-by: Larysa Zaremba <larysa.zaremba@intel.com>
Reviewed-by: Wenjia Zhang <wenjia@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

irqchip/irq-bcm7120-l2: Set IRQ_LEVEL for level triggered interrupts

[ Upstream commit 13a157b38ca5b4f9eed81442b8821db293755961 ]

When support for the interrupt controller was added with a5042de2688d,
we forgot to update the flags to be set to contain IRQ_LEVEL. While the
flow handler is correct, the output from /proc/interrupts does not show
such interrupts as being level triggered when they are, correct that.

Fixes: a5042de2688d ("irqchip: bcm7120-l2: Add Broadcom BCM7120-style Level 2 interrupt controller")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221216230934.2478345-3-f.fainelli@gmail.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

irqchip/irq-brcmstb-l2: Set IRQ_LEVEL for level triggered interrupts

[ Upstream commit 94debe03e8afa1267f95a9001786a6aa506b9ff3 ]

When support for the level triggered interrupt controller flavor was
added with c0ca7262088e, we forgot to update the flags to be set to
contain IRQ_LEVEL. While the flow handler is correct, the output from
/proc/interrupts does not show such interrupts as being level triggered
when they are, correct that.

Fixes: c0ca7262088e ("irqchip/brcmstb-l2: Add support for the BCM7271 L2 controller")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221216230934.2478345-2-f.fainelli@gmail.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

bpf: Fix global subprog context argument resolution logic

[ Upstream commit d384dce281ed1b504fae2e279507827638d56fa3 ]

KPROBE program's user-facing context type is defined as typedef
bpf_user_pt_regs_t. This leads to a problem when trying to passing
kprobe/uprobe/usdt context argument into global subprog, as kernel
always strip away mods and typedefs of user-supplied type, but takes
expected type from bpf_ctx_convert as is, which causes mismatch.

Current way to work around this is to define a fake struct with the same
name as expected typedef:

  struct bpf_user_pt_regs_t {};

  __noinline my_global_subprog(struct bpf_user_pt_regs_t *ctx) { ... }

This patch fixes the issue by resolving expected type, if it's not
a struct. It still leaves the above work-around working for backwards
compatibility.

Fixes: 91cc1a99740e ("bpf: Annotate context types")
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/bpf/20230216045954.3002473-2-andrii@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>

LoongArch, bpf: Use 4 instructions for function address in JIT

[ Upstream commit 64f50f6575721ef03d001e907455cbe3baa2a5b1 ]

This patch fixes the following issue of function calls in JIT, like:

  [   29.346981] multi-func JIT bug 105 != 103

The issus can be reproduced by running the "inline simple bpf_loop call"
verifier test.

This is because we are emiting 2-4 instructions for 64-bit immediate moves.
During the first pass of JIT, the placeholder address is zero, emiting two
instructions for it. In the extra pass, the function address is in XKVRANGE,
emiting four instructions for it. This change the instruction index in
JIT context. Let's always use 4 instructions for function address in JIT.
So that the instruction sequences don't change between the first pass and
the extra pass for function calls.

Fixes: 5dc615520c4d ("LoongArch: Add BPF JIT support")
Signed-off-by: Hengqi Chen <hengqi.chen@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Tested-by: Tiezhu Yang <yangtiezhu@loongson.cn>
Link: https://lore.kernel.org/bpf/20230214152633.2265699-1-hengqi.chen@gmail.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

xsk: check IFF_UP earlier in Tx path

[ Upstream commit 1596dae2f17ec5c6e8c8f0e3fec78c5ae55c1e0b ]

Xsk Tx can be triggered via either sendmsg() or poll() syscalls. These
two paths share a call to common function xsk_xmit() which has two
sanity checks within. A pseudo code example to show the two paths:

__xsk_sendmsg() :                       xsk_poll():
if (unlikely(!xsk_is_bound(xs)))        if (unlikely(!xsk_is_bound(xs)))
    return -ENXIO;                          return mask;
if (unlikely(need_wait))                (...)
    return -EOPNOTSUPP;                 xsk_xmit()
mark napi id
(...)
xsk_xmit()

xsk_xmit():
if (unlikely(!(xs->dev->flags & IFF_UP)))
return -ENETDOWN;
if (unlikely(!xs->tx))
return -ENOBUFS;

As it can be observed above, in sendmsg() napi id can be marked on
interface that was not brought up and this causes a NULL ptr
dereference:

[31757.505631] BUG: kernel NULL pointer dereference, address: 0000000000000018
[31757.512710] #PF: supervisor read access in kernel mode
[31757.517936] #PF: error_code(0x0000) - not-present page
[31757.523149] PGD 0 P4D 0
[31757.525726] Oops: 0000 [#1] PREEMPT SMP NOPTI
[31757.530154] CPU: 26 PID: 95641 Comm: xdpsock Not tainted 6.2.0-rc5+ #40
[31757.536871] Hardware name: Intel Corporation S2600WFT/S2600WFT, BIOS SE5C620.86B.02.01.0008.031920191559 03/19/2019
[31757.547457] RIP: 0010:xsk_sendmsg+0xde/0x180
[31757.551799] Code: 00 75 a2 48 8b 00 a8 04 75 9b 84 d2 74 69 8b 85 14 01 00 00 85 c0 75 1b 48 8b 85 28 03 00 00 48 8b 80 98 00 00 00 48 8b 40 20 <8b> 40 18 89 85 14 01 00 00 8b bd 14 01 00 00 81 ff 00 01 00 00 0f
[31757.570840] RSP: 0018:ffffc90034f27dc0 EFLAGS: 00010246
[31757.576143] RAX: 0000000000000000 RBX: ffffc90034f27e18 RCX: 0000000000000000
[31757.583389] RDX: 0000000000000001 RSI: ffffc90034f27e18 RDI: ffff88984cf3c100
[31757.590631] RBP: ffff88984714a800 R08: ffff88984714a800 R09: 0000000000000000
[31757.597877] R10: 0000000000000001 R11: 0000000000000000 R12: 00000000fffffffa
[31757.605123] R13: 0000000000000000 R14: 0000000000000003 R15: 0000000000000000
[31757.612364] FS:  00007fb4c5931180(0000) GS:ffff88afdfa00000(0000) knlGS:0000000000000000
[31757.620571] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[31757.626406] CR2: 0000000000000018 CR3: 000000184b41c003 CR4: 00000000007706e0
[31757.633648] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[31757.640894] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[31757.648139] PKRU: 55555554
[31757.650894] Call Trace:
[31757.653385]  <TASK>
[31757.655524]  sock_sendmsg+0x8f/0xa0
[31757.659077]  ? sockfd_lookup_light+0x12/0x70
[31757.663416]  __sys_sendto+0xfc/0x170
[31757.667051]  ? do_sched_setscheduler+0xdb/0x1b0
[31757.671658]  __x64_sys_sendto+0x20/0x30
[31757.675557]  do_syscall_64+0x38/0x90
[31757.679197]  entry_SYSCALL_64_after_hwframe+0x72/0xdc
[31757.687969] Code: 8e f6 ff 44 8b 4c 24 2c 4c 8b 44 24 20 41 89 c4 44 8b 54 24 28 48 8b 54 24 18 b8 2c 00 00 00 48 8b 74 24 10 8b 7c 24 08 0f 05 <48> 3d 00 f0 ff ff 77 3a 44 89 e7 48 89 44 24 08 e8 b5 8e f6 ff 48
[31757.707007] RSP: 002b:00007ffd49c73c70 EFLAGS: 00000293 ORIG_RAX: 000000000000002c
[31757.714694] RAX: ffffffffffffffda RBX: 000055a996565380 RCX: 00007fb4c5727c16
[31757.721939] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000003
[31757.729184] RBP: 0000000000000040 R08: 0000000000000000 R09: 0000000000000000
[31757.736429] R10: 0000000000000040 R11: 0000000000000293 R12: 0000000000000000
[31757.743673] R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000000
[31757.754940]  </TASK>

To fix this, let's make xsk_xmit a function that will be responsible for
generic Tx, where RCU is handled accordingly and pull out sanity checks
and xs->zc handling. Populate sanity checks to __xsk_sendmsg() and
xsk_poll().

Fixes: ca2e1a627035 ("xsk: Mark napi_id on sendmsg()")
Fixes: 18b1ab7aa76b ("xsk: Fix race at socket teardown")
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Reviewed-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Link: https://lore.kernel.org/r/20230215143309.13145-1-maciej.fijalkowski@intel.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>