crypto: ccp - Fix DBC sample application error handling

[ Upstream commit 70f242c1933e9e881c13c31640bb6d56e8b7e738 ]

The sample application was taking values from ioctl() and treating
those as the error codes to present to a user.

This is incorrect when ret is non-zero, the error is stored to `errno`.
Use this value instead.

Fixes: f40d42f116cf ("crypto: ccp - Add a sample python script for Dynamic Boost Control")
Fixes: febe3ed3222f ("crypto: ccp - Add a sample library for ioctl use")
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Sasha Levin <sashal@kernel.org>

crypto: ccp - Fix ioctl unit tests

[ Upstream commit 7f71c3e033824e1da237916a1885e3c0699f86b2 ]

A local environment change was importing ioctl_opt which is required
for ioctl tests to pass.  Add the missing import for it.

Fixes: 15f8aa7bb3e5 ("crypto: ccp - Add unit tests for dynamic boost control")
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Sasha Levin <sashal@kernel.org>

crypto: ccp - Get a free page to use while fetching initial nonce

[ Upstream commit 53f7f779f45cbe1771bc4ae05f0320e204a18611 ]

dbc_dev_init() gets a free page from `GFP_KERNEL`, but if that page has
any data in it the first nonce request will fail.
This prevents dynamic boost control from probing. To fix this, explicitly
request a zeroed page with `__GFP_ZERO` to ensure first nonce fetch works.

Fixes: c04cf9e14f10 ("crypto: ccp - Add support for fetching a nonce for dynamic boost control")
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Sasha Levin <sashal@kernel.org>

KEYS: Include linux/errno.h in linux/verification.h

[ Upstream commit 0a596b0682a7ce37e26c36629816f105c6459d06 ]

Add inclusion of linux/errno.h as otherwise the reference to EINVAL
may be invalid.

Fixes: f3cf4134c5c6 ("bpf: Add bpf_lookup_*_key() and bpf_key_put() kfuncs")
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202308261414.HKw1Mrip-lkp@intel.com/
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ALSA: hda: cs35l41: Undo runtime PM changes at driver exit time

[ Upstream commit 85a1bf86fac0c195929768b4e92c78cad107523b ]

According to the documentation, drivers are responsible for undoing at
removal time all runtime PM changes done during probing.

Hence, add the missing calls to pm_runtime_dont_use_autosuspend(), which
are necessary for undoing pm_runtime_use_autosuspend().

Fixes: 1873ebd30cc8 ("ALSA: hda: cs35l41: Support Hibernation during Suspend")
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Link: https://lore.kernel.org/r/20230907171010.1447274-11-cristian.ciocaltea@collabora.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ALSA: hda: cs35l41: Fix unbalanced pm_runtime_get()

[ Upstream commit 486465508f8a5fe441939a7d97607f4460a60891 ]

If component_add() fails, probe() returns without calling
pm_runtime_put(), which leaves the runtime PM usage counter incremented.

Fix the issue by jumping to err_pm label and drop the now unnecessary
pm_runtime_disable() call.

Fixes: 7b2f3eb492da ("ALSA: hda: cs35l41: Add support for CS35L41 in HDA systems")
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Link: https://lore.kernel.org/r/20230907171010.1447274-10-cristian.ciocaltea@collabora.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ASoC: cs35l41: Undo runtime PM changes at driver exit time

[ Upstream commit 2d5661e6008ae1a1cd6df7cc844908fb8b982c58 ]

According to the documentation, drivers are responsible for undoing at
removal time all runtime PM changes done during probing.

Hence, add the missing calls to pm_runtime_dont_use_autosuspend(), which
are necessary for undoing pm_runtime_use_autosuspend().

Note this would have been handled implicitly by
devm_pm_runtime_enable(), but there is a need to continue using
pm_runtime_enable()/pm_runtime_disable() in order to ensure the runtime
PM is disabled as soon as the remove() callback is entered.

Fixes: f517ba4924ad ("ASoC: cs35l41: Add support for hibernate memory retention mode")
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Link: https://lore.kernel.org/r/20230907171010.1447274-7-cristian.ciocaltea@collabora.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ASoC: cs35l41: Verify PM runtime resume errors in IRQ handler

[ Upstream commit 9f8948db9849d202dee3570507d3a0642f92d632 ]

The interrupt handler invokes pm_runtime_get_sync() without checking the
returned error code.

Add a proper verification and switch to pm_runtime_resume_and_get(), to
avoid the need to call pm_runtime_put_noidle() for decrementing the PM
usage counter before returning from the error condition.

Fixes: f517ba4924ad ("ASoC: cs35l41: Add support for hibernate memory retention mode")
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
Acked-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Link: https://lore.kernel.org/r/20230907171010.1447274-6-cristian.ciocaltea@collabora.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ASoC: cs35l41: Fix broken shared boost activation

[ Upstream commit 77bf613f0bf08c021309cdb5f84b5f630b829261 ]

Enabling the active/passive shared boosts requires setting SYNC_EN, but
*not* before receiving the PLL Lock signal.

Due to improper error handling, it was not obvious that waiting for the
completion operation times out and, consequently, the shared boost is
never activated.

Further investigations revealed the signal is triggered while
snd_pcm_start() is executed, right after receiving the
SNDRV_PCM_TRIGGER_START command, which happens long after the
SND_SOC_DAPM_PRE_PMU event handler is invoked as part of
snd_pcm_prepare().  That is where cs35l41_global_enable() is called
from.

Increasing the wait duration doesn't help, as it only causes an
unnecessary delay in the invocation of snd_pcm_start().  Moving the wait
and the subsequent regmap operations to the SNDRV_PCM_TRIGGER_START
callback is not a solution either, since they would be executed in an
IRQ-off atomic context.

Solve the issue by setting the SYNC_EN bit in PWR_CTRL3 register right
after receiving the PLL Lock interrupt.

Additionally, drop the unnecessary writes to PWR_CTRL1 register, part of
the original mdsync_up_seq, which would have toggled GLOBAL_EN with
unwanted consequences on PLL locking behavior.

Fixes: f5030564938b ("ALSA: cs35l41: Add shared boost feature")
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
Reviewed-by: David Rhodes <david.rhodes@cirrus.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Link: https://lore.kernel.org/r/20230907171010.1447274-5-cristian.ciocaltea@collabora.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ASoC: cs35l41: Initialize completion object before requesting IRQ

[ Upstream commit 5ad668a9ce83d819701fb7abc1c2236049ec15c2 ]

Technically, an interrupt handler can be called before probe() finishes
its execution, hence ensure the pll_lock completion object is always
initialized before being accessed in cs35l41_irq().

Fixes: f5030564938b ("ALSA: cs35l41: Add shared boost feature")
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
Acked-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Link: https://lore.kernel.org/r/20230907171010.1447274-4-cristian.ciocaltea@collabora.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ASoC: cs35l41: Handle mdsync_up reg write errors

[ Upstream commit 4bb5870ab60abca6ad18196090831b5e4cf82d93 ]

The return code of regmap_multi_reg_write() call related to "MDSYNC up"
sequence is shadowed by the subsequent regmap_read_poll_timeout()
invocation, which will hit a timeout in case the write operation above
fails.

Make sure cs35l41_global_enable() returns the correct error code instead
of -ETIMEDOUT.

Additionally, to be able to distinguish between the timeouts of
wait_for_completion_timeout() and regmap_read_poll_timeout(), print an
error message for the former and return immediately.  This also avoids
having to wait unnecessarily for the second time.

Fixes: f8264c759208 ("ALSA: cs35l41: Poll for Power Up/Down rather than waiting a fixed delay")
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
Acked-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Link: https://lore.kernel.org/r/20230907171010.1447274-3-cristian.ciocaltea@collabora.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ASoC: cs35l41: Handle mdsync_down reg write errors

[ Upstream commit a9a3f54a23d844971c274f352500dddeadb4412c ]

The return code of regmap_multi_reg_write() call related to "MDSYNC
down" sequence is shadowed by the subsequent
wait_for_completion_timeout() invocation, which is expected to time
timeout in case the write operation failed.

Let cs35l41_global_enable() return the correct error code instead of
-ETIMEDOUT.

Fixes: f5030564938b ("ALSA: cs35l41: Add shared boost feature")
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
Acked-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Link: https://lore.kernel.org/r/20230907171010.1447274-2-cristian.ciocaltea@collabora.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

module/decompress: use vmalloc() for gzip decompression workspace

[ Upstream commit 3737df782c740b944912ed93420c57344b1cf864 ]

Use a similar approach as commit a419beac4a07 ("module/decompress: use
vmalloc() for zstd decompression workspace") and replace kmalloc() with
vmalloc() also for the gzip module decompression workspace.

In this case the workspace is represented by struct inflate_workspace
that can be fairly large for kmalloc() and it can potentially lead to
allocation errors on certain systems:

$ pahole inflate_workspace
struct inflate_workspace {
struct inflate_state       inflate_state;        /*     0  9544 */
/* --- cacheline 149 boundary (9536 bytes) was 8 bytes ago --- */
unsigned char              working_window[32768]; /*  9544 32768 */

/* size: 42312, cachelines: 662, members: 2 */
/* last cacheline: 8 bytes */
};

Considering that there is no need to use continuous physical memory,
simply switch to vmalloc() to provide a more reliable in-kernel module
decompression.

Fixes: b1ae6dc41eaa ("module: add in-kernel support for decompressing")
Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
Signed-off-by: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

iommufd: Add iopt_area_alloc()

[ Upstream commit 361d744ddd61de065fbeb042aaed590d32dd92ec ]

We never initialize the two interval tree nodes, and zero fill is not the
same as RB_CLEAR_NODE. This can hide issues where we missed adding the
area to the trees. Factor out the allocation and clear the two nodes.

Fixes: 51fe6141f0f6 ("iommufd: Data structure to provide IOVA to PFN mapping")
Link: https://lore.kernel.org/r/20231030145035.GG691768@ziepe.ca
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ARM: dts: BCM5301X: Explicitly disable unused switch CPU ports

[ Upstream commit 473baeab929444295b0530f8766e4becb6a08973 ]

When redescribing ports I assumed that missing "label" (like "cpu")
means switch port isn't used. That was incorrect and I realized my
change made Linux always use the first (5) CPU port (there are 3 of
them).

While above should technically be possible it often isn't correct:
1. Non-default switch ports are often connected to Ethernet interfaces
   not fully covered by vendor setup (they may miss MACs)
2. On some devices non-default ports require specifying fixed link

This fixes network connectivity for some devices. It was reported &
tested for Netgear R8000. It also affects Linksys EA9200 with its
downstream DTS.

Fixes: ba4aebce23b2 ("ARM: dts: BCM5301X: Describe switch ports in the main DTS")
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Link: https://lore.kernel.org/r/20231013103314.10306-1-zajec5@gmail.com
Signed-off-by: Florian Fainelli <florian.fainelli@broadcom.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

soc: qcom: pmic_glink: fix connector type to be DisplayPort

[ Upstream commit f86955f2b1ff9fbc7ae4f6595112b2f896885366 ]

As it was pointed out by Simon Ser, the DRM_MODE_CONNECTOR_USB connector
is reserved for the GUD devices. Other drivers (i915, amdgpu) use
DRM_MODE_CONNECTOR_DisplayPort even if the DP stream is handled by the
USB-C altmode. While we are still working on implementing the proper way
to let userspace know that the DP is wrapped into USB-C, change
connector type to be DRM_MODE_CONNECTOR_DisplayPort.

Fixes: 080b4e24852b ("soc: qcom: pmic_glink: Introduce altmode support")
Cc: Simon Ser <contact@emersion.fr>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Acked-by: Simon Ser <contact@emersion.fr>
Link: https://lore.kernel.org/r/20231010225229.77027-1-dmitry.baryshkov@linaro.org
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

selftests/resctrl: Ensure the benchmark commands fits to its array

[ Upstream commit 4a28c7665c2a1ac0400864eabb0c641e135f61aa ]

Benchmark command is copied into an array in the stack. The array is
BENCHMARK_ARGS items long but the command line could try to provide a
longer command. Argument size is also fixed by BENCHMARK_ARG_SIZE (63
bytes of space after fitting the terminating \0 character) and user
could have inputted argument longer than that.

Return error in case the benchmark command does not fit to the space
allocated for it.

Fixes: ecdbb911f22d ("selftests/resctrl: Add MBM test")
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Tested-by: Shaopeng Tan <tan.shaopeng@jp.fujitsu.com>
Reviewed-by: Shaopeng Tan <tan.shaopeng@jp.fujitsu.com>
Reviewed-by: "Wieczor-Retman, Maciej" <maciej.wieczor-retman@intel.com>
Reviewed-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

selftests/pidfd: Fix ksft print formats

[ Upstream commit 4d7f4e8158b62f63031510cdc24acc520956c091 ]

Compiling pidfd selftest after adding a __printf() attribute to
ksft_print_msg() and ksft_test_result_pass() exposes -Wformat warnings
in error_report(), test_pidfd_poll_exec_thread(),
child_poll_exec_test(), test_pidfd_poll_leader_exit_thread(),
child_poll_leader_exit_test().

The ksft_test_result_pass() in error_report() expects a string but
doesn't provide any argument after the format string. All the other
calls to ksft_print_msg() in the functions mentioned above have format
strings that don't match with other passed arguments.

Fix format specifiers so they match the passed variables.

Add a missing variable to ksft_test_result_pass() inside
error_report() so it matches other cases in the switch statement.

Fixes: 2def297ec7fb ("pidfd: add tests for NSpid info in fdinfo")

Signed-off-by: Maciej Wieczor-Retman <maciej.wieczor-retman@intel.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: tegra: Use correct interrupts for Tegra234 TKE

[ Upstream commit c0b80988eb78d6423249ab530bfbc6b238790a26 ]

The shared interrupts 0-9 of the TKE are mapped to interrupts 0-9, but
shared interrupts 10-15 are mapped to 256-261. Correct the mapping for
the final 6 interrupts. This prevents the TKE from requesting the RTC
interrupt (along with several GTE and watchdog interrupts).

Reported-by: Shubhi Garg <shgarg@nvidia.com>
Fixes: 28d860ed02c2 ("arm64: tegra: Enable native timers on Tegra234")
Reviewed-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

memory: tegra: Set BPMP msg flags to reset IPC channels

[ Upstream commit f344675a34383ae26a8230f4b1cd99cbd0defebd ]

Set the 'TEGRA_BPMP_MESSAGE_RESET' bit in newly added 'flags' field
of 'struct tegra_bpmp_message' to request for the reset of BPMP IPC
channels. This is used along with the 'suspended' check in BPMP driver
for handling early bandwidth requests due to the hotplug of CPU's
during system resume before the driver gets resumed.

Fixes: f41e1442ac5b ("cpufreq: tegra194: add OPP support and set bandwidth")
Co-developed-by: Sumit Gupta <sumitg@nvidia.com>
Signed-off-by: Sumit Gupta <sumitg@nvidia.com>
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

firmware: tegra: Add suspend hook and reset BPMP IPC early on resume

[ Upstream commit ea608a01d4ee66f8b51070e623f9adb8684c0dd4 ]

Add suspend hook and a 'suspended' field in the 'struct tegra_bpmp'
to mark if BPMP is suspended. Also, add a 'flags' field in the
'struct tegra_bpmp_message' whose 'TEGRA_BPMP_MESSAGE_RESET' bit can be
set from the Tegra MC driver to signal that the reset of BPMP IPC
channels is required before sending MRQ to the BPMP FW. Together both
the fields allow us to handle any requests that might be sent too soon
as they can cause hang during system resume.

One case where we see BPMP requests being sent before the BPMP driver
has resumed is the memory bandwidth requests which are triggered by
onlining the CPUs during system resume. The CPUs are onlined before the
BPMP has resumed and we need to reset the BPMP IPC channels to handle
these requests.

The additional check for 'flags' is done to avoid any un-intended BPMP
IPC reset if the tegra_bpmp_transfer*() API gets called during suspend
sequence after the BPMP driver is suspended.

Fixes: f41e1442ac5b ("cpufreq: tegra194: add OPP support and set bandwidth")
Co-developed-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Sumit Gupta <sumitg@nvidia.com>
Acked-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: tegra: Fix P3767 QSPI speed

[ Upstream commit 57ea99ba176913c325fc8324a24a1b5e8a6cf520 ]

The QSPI device used on Jetson Orin NX and Nano modules (p3767) is
the same as Jetson AGX Orin (p3701) and should have a maximum speed of
102 MHz.

Fixes: 13b0aca303e9 ("arm64: tegra: Support Jetson Orin NX")
Signed-off-by: Brad Griffis <bgriffis@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: tegra: Fix P3767 card detect polarity

[ Upstream commit c6b7a1d11d0fa6333078141251908f48042016e1 ]

The SD card detect pin is active-low on all Orin Nano and NX SKUs that
have an SD card slot.

Fixes: 13b0aca303e9 ("arm64: tegra: Support Jetson Orin NX")
Signed-off-by: Brad Griffis <bgriffis@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: imx8mn: Add sound-dai-cells to micfil node

[ Upstream commit db1925454a2e7cadcac8756442ca7c3198332336 ]

Per the DT bindings, the micfil node should have a sound-dai-cells
entry.

Fixes: cca69ef6eba5 ("arm64: dts: imx8mn: Add support for micfil")
Signed-off-by: Adam Ford <aford173@gmail.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: imx8mm: Add sound-dai-cells to micfil node

[ Upstream commit 0e6cc2b8bb7d67733f4a47720787eff1ce2666f2 ]

Per the DT bindings, the micfil node should have a sound-dai-cells
entry.

Fixes: 3bd0788c43d9 ("arm64: dts: imx8mm: Add support for micfil")
Signed-off-by: Adam Ford <aford173@gmail.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: imx8mp-debix-model-a: Remove USB hub reset-gpios

[ Upstream commit 0ce9a2c121e3ab354cf66aeecd3ed0758f3c5067 ]

The SAI2_TXC pin is left unconnected per the imx8mp-debix-model-a
schematics:

https://debix.io/Uploads/Temp/file/20230331/DEBIX%20Model%20A%20Schematics.pdf

Also, the RTS5411E USB hub chip does not have a reset pin.

Remove this pin description to properly describe the hardware.

This also fixes the following schema warning:

hub@1: 'reset-gpios' does not match any of the regexes: 'pinctrl-[0-9]+'
from schema $id: http://devicetree.org/schemas/usb/realtek,rts5411.yaml#

Fixes: 0253e1cb6300 ("arm64: dts: imx8mp-debix: add USB host support")
Signed-off-by: Fabio Estevam <festevam@denx.de>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: imx8qm-ss-img: Fix jpegenc compatible entry

[ Upstream commit 1d33cd614d89b0ec024d25ec45acf4632211b5a7 ]

The first compatible entry for the jpegenc should be 'nxp,imx8qm-jpgenc'.

Change it accordingly to fix the following schema warning:

imx8qm-apalis-eval.dtb: jpegenc@58450000: compatible: 'oneOf' conditional failed, one must be fixed:
'nxp,imx8qm-jpgdec' is not one of ['nxp,imx8qxp-jpgdec', 'nxp,imx8qxp-jpgenc']
'nxp,imx8qm-jpgenc' was expected
'nxp,imx8qxp-jpgdec' was expected

Fixes: 5bb279171afc ("arm64: dts: imx8: Add jpeg encoder/decoder nodes")
Signed-off-by: Fabio Estevam <festevam@denx.de>
Reviewed-by: Mirela Rabulea <mirela.rabulea@nxp.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: scmi: Free scmi_clk allocated when the clocks with invalid info are skipped

[ Upstream commit 3537a75e73f3420614a358d0c8b390ea483cc87d ]

Add the missing devm_kfree() when we skip the clocks with invalid or
missing information from the firmware.

Cc: Cristian Marussi <cristian.marussi@arm.com>
Cc: Michael Turquette <mturquette@baylibre.com>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: linux-clk@vger.kernel.org
Fixes: 6d6a1d82eaef ("clk: add support for clocks provided by SCMI")
Link: https://lore.kernel.org/r/20231004193600.66232-1-sudeep.holla@arm.com
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ARM: dts: am3517-evm: Fix LED3/4 pinmux

[ Upstream commit 2ab6b437c65233f06bdd2988fd5913baeca5f159 ]

The pinmux for LED3 and LED4 are incorrectly attached to the
omap3_pmx_core when they should be connected to the omap3_pmx_wkup
pin mux.  This was likely masked by the fact that the bootloader
used to do all the pinmuxing.

Fixes: 0dbf99542caf ("ARM: dts: am3517-evm: Add User LEDs and Pushbutton")
Signed-off-by: Adam Ford <aford173@gmail.com>
Message-ID: <20231005000402.50879-1-aford173@gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

firmware: arm_ffa: Allow the FF-A drivers to use 32bit mode of messaging

[ Upstream commit 2d698e8b4fd22374dac0a2d5150ab24d57a222ab ]

An FF-A ABI could support both the SMC32 and SMC64 conventions.
A callee that runs in the AArch64 execution state and implements such
an ABI must implement both SMC32 and SMC64 conventions of the ABI.

So the FF-A drivers will need the option to choose the mode irrespective
of FF-A version and the partition execution mode flag in the partition
information.

Let us remove the check on the FF-A version for allowing the selection
of 32bit mode of messaging. The driver will continue to set the 32-bit
mode if the partition execution mode flag specified that the partition
supports only 32-bit execution.

Fixes: 106b11b1ccd5 ("firmware: arm_ffa: Set up 32bit execution mode flag using partiion property")
Link: https://lore.kernel.org/r/20231005142823.278121-1-sudeep.holla@arm.com
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

firmware: arm_ffa: Assign the missing IDR allocation ID to the FFA device

[ Upstream commit 7d0bc6360f17ea323ab25939a34857123d7d87e5 ]

Commit 19b8766459c4 ("firmware: arm_ffa: Fix FFA device names for logical
partitions") added an ID to the FFA device using ida_alloc() and append
the same to "arm-ffa" to make up a unique device name. However it missed
to stash the id value in ffa_dev to help freeing the ID later when the
device is destroyed.

Due to the missing/unassigned ID in FFA device, we get the following
warning when the FF-A device is unregistered.

  |   ida_free called for id=0 which is not allocated.
  |   WARNING: CPU: 7 PID: 1 at lib/idr.c:525 ida_free+0x114/0x164
  |   CPU: 7 PID: 1 Comm: swapper/0 Not tainted 6.6.0-rc4 #209
  |   pstate: 61400009 (nZCv daif +PAN -UAO -TCO +DIT -SSBS BTYPE=--)
  |   pc : ida_free+0x114/0x164
  |   lr : ida_free+0x114/0x164
  |   Call trace:
  |    ida_free+0x114/0x164
  |    ffa_release_device+0x24/0x3c
  |    device_release+0x34/0x8c
  |    kobject_put+0x94/0xf8
  |    put_device+0x18/0x24
  |    klist_devices_put+0x14/0x20
  |    klist_next+0xc8/0x114
  |    bus_for_each_dev+0xd8/0x144
  |    arm_ffa_bus_exit+0x30/0x54
  |    ffa_init+0x68/0x330
  |    do_one_initcall+0xdc/0x250
  |    do_initcall_level+0x8c/0xac
  |    do_initcalls+0x54/0x94
  |    do_basic_setup+0x1c/0x28
  |    kernel_init_freeable+0x104/0x170
  |    kernel_init+0x20/0x1a0
  |    ret_from_fork+0x10/0x20

Fix the same by actually assigning the ID in the FFA device this time
for real.

Fixes: 19b8766459c4 ("firmware: arm_ffa: Fix FFA device names for logical partitions")
Link: https://lore.kernel.org/r/20231003085932.3553985-1-sudeep.holla@arm.com
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: ti: Fix HDMI Audio overlay in Makefile

[ Upstream commit 69c570ebc3964534c19dc4438d3b96f55d489fc3 ]

Apply HDMI audio overlay to AM625 and AM62-LP SK-EVMs DT binaries,
instead of leaving it in a floating state.

Fixes: b50ccab9e07c ("arm64: dts: ti: am62x-sk: Add overlay for HDMI audio")
Reported-by: Rob Herring <robh@kernel.org>
Signed-off-by: Aradhya Bhatia <a-bhatia1@ti.com>
Link: https://lore.kernel.org/r/20231003092259.28103-1-a-bhatia1@ti.com
Signed-off-by: Vignesh Raghavendra <vigneshr@ti.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: ti: k3-am62a7-sk: Drop i2c-1 to 100Khz

[ Upstream commit 63e5aa69b821472a3203a29e17c025329c1b151f ]

The TLV320AIC3106 audio codec is interfaced on the i2c-1 bus. With the
default rate of 400Khz the i2c register writes fail to sync:

[   36.026387] tlv320aic3x 1-001b: Unable to sync registers 0x16-0x16. -110
[   38.101130] omap_i2c 20010000.i2c: controller timed out

Dropping the rate to 100Khz fixes the issue.

Fixes: 38c4a08c820c ("arm64: dts: ti: Add support for AM62A7-SK")
Reviewed-by: Devarsh Thakkar <devarsht@ti.com>
Reviewed-by: Aradhya Bhatia <a-bhatia1@ti.com>
Signed-off-by: Jai Luthra <j-luthra@ti.com>
Link: https://lore.kernel.org/r/20231003-mcasp_am62a-v3-3-2b631ff319ca@ti.com
Signed-off-by: Vignesh Raghavendra <vigneshr@ti.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: ti: k3-am625-beagleplay: Fix typo in ramoops reg

[ Upstream commit 33269ac0b768b07da017df173d52952625c57870 ]

Seems like the address value of the reg property was mistyped.
Update reg to 0x9ca00000 to match node's definition.

Fixes: f5a731f0787f ("arm64: dts: ti: Add k3-am625-beagleplay")
Signed-off-by: Wadim Egorov <w.egorov@phytec.de>
Reviewed-by: Nishanth Menon <nm@ti.com>
Link: https://lore.kernel.org/r/20230925151444.1856852-1-w.egorov@phytec.de
Signed-off-by: Vignesh Raghavendra <vigneshr@ti.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: ti: verdin-am62: disable MIPI DSI bridge

[ Upstream commit 664e2852aa9142193c2e241327631f032b966742 ]

Keep the DPI to MIPI-DSI bridge disabled in the SoM dtsi file.

The display chain is not wholly described in the device tree file, on
Verdin product family the displays are additional accessories that are
configured/enabled using DT overlays.

With this enabled we have issues when a display is enabled on
TIDSS port1 (LVDS) and port0 (DSI) is not used.

Fixes: 9e77200356ba ("arm64: dts: ti: verdin-am62: Add DSI display support")
Signed-off-by: Francesco Dolcini <francesco.dolcini@toradex.com>
Link: https://lore.kernel.org/r/20230922123003.25002-1-francesco@dolcini.it
Signed-off-by: Vignesh Raghavendra <vigneshr@ti.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: ti: k3-j721s2-evm-gesi: Specify base dtb for overlay file

[ Upstream commit 35be6ac964450687ab39b846d65ee1cb2a352280 ]

Specify the base dtb file k3-j721s2-common-proc-board.dtb on which the
k3-j721s2-evm-gesi-exp-board.dtbo overlay has to be applied. Name the
resulting dtb as k3-j721s2-evm.dtb.

Fixes: cac04e27f093 ("arm64: dts: ti: k3-j721s2: Add overlay to enable main CPSW2G with GESI")
Reported-by: Rob Herring <robh+dt@kernel.org>
Signed-off-by: Siddharth Vadapalli <s-vadapalli@ti.com>
Link: https://lore.kernel.org/r/20230912043308.20629-1-s-vadapalli@ti.com
Signed-off-by: Vignesh Raghavendra <vigneshr@ti.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

firmware: ti_sci: Mark driver as non removable

[ Upstream commit 7b7a224b1ba1703583b25a3641ad9798f34d832a ]

The TI-SCI message protocol provides a way to communicate between
various compute processors with a central system controller entity. It
provides the fundamental device management capability and clock control
in the SOCs that it's used in.

The remove function failed to do all the necessary cleanup if
there are registered users. Some things are freed however which
likely results in an oops later on.

Ensure that the driver isn't unbound by suppressing its bind and unbind
sysfs attributes. As the driver is built-in there is no way to remove
device once bound.

We can also remove the ti_sci_remove call along with the
ti_sci_debugfs_destroy as there are no callers for it any longer.

Fixes: aa276781a64a ("firmware: Add basic support for TI System Control Interface (TI-SCI) protocol")
Reported-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Closes: https://lore.kernel.org/linux-arm-kernel/20230216083908.mvmydic5lpi3ogo7@pengutronix.de/
Suggested-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Acked-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Dhruva Gole <d-gole@ti.com>
Link: https://lore.kernel.org/r/20230921091025.133130-1-d-gole@ti.com
Signed-off-by: Nishanth Menon <nm@ti.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ARM: dts: stm32: stm32f7-pinctrl: don't use multiple blank lines

[ Upstream commit 88bb50edb61068c4416df2e55677fb3159f647f1 ]

The patch fixes the following warning:

arch/arm/dts/stm32f7-pinctrl.dtsi:380: check: Please don't use multiple blank lines

Fixes: ba287d1a0137 ("ARM: dts: stm32: add pin map for LTDC on stm32f7")
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Reviewed-by: Raphaël Gallais-Pou <raphael.gallais-pou@foss.st.com>
Signed-off-by: Alexandre Torgue <alexandre.torgue@foss.st.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

kunit: test: Fix the possible memory leak in executor_test

[ Upstream commit 8040345fdae4cb256c5d981f91ae0f22bea8adcc ]

When CONFIG_KUNIT_ALL_TESTS=y, making CONFIG_DEBUG_KMEMLEAK=y and
CONFIG_DEBUG_KMEMLEAK_AUTO_SCAN=y, the below memory leak is detected.

If kunit_filter_suites() succeeds, not only copy but also filtered_suite
and filtered_suite->test_cases should be freed.

So as Rae suggested, to avoid the suite set never be freed when
KUNIT_ASSERT_EQ() fails and exits after kunit_filter_suites() succeeds,
update kfree_at_end() func to free_suite_set_at_end() to use
kunit_free_suite_set() to free them as kunit_module_exit() and
kunit_run_all_tests() do it. As the second arg got of
free_suite_set_at_end() is a local variable, copy it for free to avoid
wild-memory-access. After applying this patch, the following memory leak
is never detected.

unreferenced object 0xffff8881001de400 (size 1024):
  comm "kunit_try_catch", pid 1396, jiffies 4294720452 (age 932.801s)
  hex dump (first 32 bytes):
    73 75 69 74 65 32 00 00 00 00 00 00 00 00 00 00  suite2..........
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<ffffffff817db753>] __kmalloc_node_track_caller+0x53/0x150
    [<ffffffff817bd242>] kmemdup+0x22/0x50
    [<ffffffff829e961d>] kunit_filter_suites+0x44d/0xcc0
    [<ffffffff829eb69f>] filter_suites_test+0x12f/0x360
    [<ffffffff829e802a>] kunit_generic_run_threadfn_adapter+0x4a/0x90
    [<ffffffff81236fc6>] kthread+0x2b6/0x380
    [<ffffffff81096afd>] ret_from_fork+0x2d/0x70
    [<ffffffff81003511>] ret_from_fork_asm+0x11/0x20
unreferenced object 0xffff8881052cd388 (size 192):
  comm "kunit_try_catch", pid 1396, jiffies 4294720452 (age 932.801s)
  hex dump (first 32 bytes):
    a0 85 9e 82 ff ff ff ff 80 cd 7c 84 ff ff ff ff  ..........|.....
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<ffffffff817dbad2>] __kmalloc+0x52/0x150
    [<ffffffff829e9651>] kunit_filter_suites+0x481/0xcc0
    [<ffffffff829eb69f>] filter_suites_test+0x12f/0x360
    [<ffffffff829e802a>] kunit_generic_run_threadfn_adapter+0x4a/0x90
    [<ffffffff81236fc6>] kthread+0x2b6/0x380
    [<ffffffff81096afd>] ret_from_fork+0x2d/0x70
    [<ffffffff81003511>] ret_from_fork_asm+0x11/0x20

unreferenced object 0xffff888100da8400 (size 1024):
  comm "kunit_try_catch", pid 1398, jiffies 4294720454 (age 781.945s)
  hex dump (first 32 bytes):
    73 75 69 74 65 32 00 00 00 00 00 00 00 00 00 00  suite2..........
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<ffffffff817db753>] __kmalloc_node_track_caller+0x53/0x150
    [<ffffffff817bd242>] kmemdup+0x22/0x50
    [<ffffffff829e961d>] kunit_filter_suites+0x44d/0xcc0
    [<ffffffff829eb13f>] filter_suites_test_glob_test+0x12f/0x560
    [<ffffffff829e802a>] kunit_generic_run_threadfn_adapter+0x4a/0x90
    [<ffffffff81236fc6>] kthread+0x2b6/0x380
    [<ffffffff81096afd>] ret_from_fork+0x2d/0x70
    [<ffffffff81003511>] ret_from_fork_asm+0x11/0x20
unreferenced object 0xffff888105117878 (size 96):
  comm "kunit_try_catch", pid 1398, jiffies 4294720454 (age 781.945s)
  hex dump (first 32 bytes):
    a0 85 9e 82 ff ff ff ff a0 ac 7c 84 ff ff ff ff  ..........|.....
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<ffffffff817dbad2>] __kmalloc+0x52/0x150
    [<ffffffff829e9651>] kunit_filter_suites+0x481/0xcc0
    [<ffffffff829eb13f>] filter_suites_test_glob_test+0x12f/0x560
    [<ffffffff829e802a>] kunit_generic_run_threadfn_adapter+0x4a/0x90
    [<ffffffff81236fc6>] kthread+0x2b6/0x380
    [<ffffffff81096afd>] ret_from_fork+0x2d/0x70
    [<ffffffff81003511>] ret_from_fork_asm+0x11/0x20
unreferenced object 0xffff888102c31c00 (size 1024):
  comm "kunit_try_catch", pid 1404, jiffies 4294720460 (age 781.948s)
  hex dump (first 32 bytes):
    6e 6f 72 6d 61 6c 5f 73 75 69 74 65 00 00 00 00  normal_suite....
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<ffffffff817db753>] __kmalloc_node_track_caller+0x53/0x150
    [<ffffffff817bd242>] kmemdup+0x22/0x50
    [<ffffffff829ecf17>] kunit_filter_attr_tests+0xf7/0x860
    [<ffffffff829e99ff>] kunit_filter_suites+0x82f/0xcc0
    [<ffffffff829ea975>] filter_attr_test+0x195/0x5f0
    [<ffffffff829e802a>] kunit_generic_run_threadfn_adapter+0x4a/0x90
    [<ffffffff81236fc6>] kthread+0x2b6/0x380
    [<ffffffff81096afd>] ret_from_fork+0x2d/0x70
    [<ffffffff81003511>] ret_from_fork_asm+0x11/0x20
unreferenced object 0xffff8881052cd250 (size 192):
  comm "kunit_try_catch", pid 1404, jiffies 4294720460 (age 781.948s)
  hex dump (first 32 bytes):
    a0 85 9e 82 ff ff ff ff 00 a9 7c 84 ff ff ff ff  ..........|.....
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<ffffffff817dbad2>] __kmalloc+0x52/0x150
    [<ffffffff829ecfc1>] kunit_filter_attr_tests+0x1a1/0x860
    [<ffffffff829e99ff>] kunit_filter_suites+0x82f/0xcc0
    [<ffffffff829ea975>] filter_attr_test+0x195/0x5f0
    [<ffffffff829e802a>] kunit_generic_run_threadfn_adapter+0x4a/0x90
    [<ffffffff81236fc6>] kthread+0x2b6/0x380
    [<ffffffff81096afd>] ret_from_fork+0x2d/0x70
    [<ffffffff81003511>] ret_from_fork_asm+0x11/0x20
unreferenced object 0xffff888104f4e400 (size 1024):
  comm "kunit_try_catch", pid 1408, jiffies 4294720464 (age 781.944s)
  hex dump (first 32 bytes):
    73 75 69 74 65 00 00 00 00 00 00 00 00 00 00 00  suite...........
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<ffffffff817db753>] __kmalloc_node_track_caller+0x53/0x150
    [<ffffffff817bd242>] kmemdup+0x22/0x50
    [<ffffffff829ecf17>] kunit_filter_attr_tests+0xf7/0x860
    [<ffffffff829e99ff>] kunit_filter_suites+0x82f/0xcc0
    [<ffffffff829e9fc3>] filter_attr_skip_test+0x133/0x6e0
    [<ffffffff829e802a>] kunit_generic_run_threadfn_adapter+0x4a/0x90
    [<ffffffff81236fc6>] kthread+0x2b6/0x380
    [<ffffffff81096afd>] ret_from_fork+0x2d/0x70
    [<ffffffff81003511>] ret_from_fork_asm+0x11/0x20
unreferenced object 0xffff8881052cc620 (size 192):
  comm "kunit_try_catch", pid 1408, jiffies 4294720464 (age 781.944s)
  hex dump (first 32 bytes):
    a0 85 9e 82 ff ff ff ff c0 a8 7c 84 ff ff ff ff  ..........|.....
    00 00 00 00 00 00 00 00 02 00 00 00 02 00 00 00  ................
  backtrace:
    [<ffffffff817dbad2>] __kmalloc+0x52/0x150
    [<ffffffff829ecfc1>] kunit_filter_attr_tests+0x1a1/0x860
    [<ffffffff829e99ff>] kunit_filter_suites+0x82f/0xcc0
    [<ffffffff829e9fc3>] filter_attr_skip_test+0x133/0x6e0
    [<ffffffff829e802a>] kunit_generic_run_threadfn_adapter+0x4a/0x90
    [<ffffffff81236fc6>] kthread+0x2b6/0x380
    [<ffffffff81096afd>] ret_from_fork+0x2d/0x70
    [<ffffffff81003511>] ret_from_fork_asm+0x11/0x20

Fixes: e5857d396f35 ("kunit: flatten kunit_suite*** to kunit_suite** in .kunit_test_suites")
Fixes: 76066f93f1df ("kunit: add tests for filtering attributes")
Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
Suggested-by: Rae Moar <rmoar@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Suggested-by: David Gow <davidgow@google.com>
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202309142251.uJ8saAZv-lkp@intel.com/
Closes: https://lore.kernel.org/oe-kbuild-all/202309270433.wGmFRGjd-lkp@intel.com/
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

kunit: Fix possible memory leak in kunit_filter_suites()

[ Upstream commit 24de14c98b37ea40a7e493dfd0d93b400b6efbca ]

If the outer layer for loop is iterated more than once and it fails not
in the first iteration, the filtered_suite and filtered_suite->test_cases
allocated in the last kunit_filter_attr_tests() in last inner for loop
is leaked.

So add a new free_filtered_suite err label and free the filtered_suite
and filtered_suite->test_cases so far. And change kmalloc_array of copy
to kcalloc to Clear the copy to make the kfree safe.

Fixes: 529534e8cba3 ("kunit: Add ability to filter attributes")
Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
Reviewed-by: Rae Moar <rmoar@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

kunit: Fix the wrong kfree of copy for kunit_filter_suites()

[ Upstream commit e44679515a7b803cf0143dc9de3d2ecbe907f939 ]

If the outer layer for loop is iterated more than once and it fails not
in the first iteration, the copy pointer has been moved. So it should free
the original copy's backup copy_start.

Fixes: abbf73816b6f ("kunit: fix possible memory leak in kunit_filter_suites()")
Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
Reviewed-by: Rae Moar <rmoar@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

kunit: Fix missed memory release in kunit_free_suite_set()

[ Upstream commit a6074cf0126b0bee51ab77a15930dc24a4d5db90 ]

modprobe cpumask_kunit and rmmod cpumask_kunit, kmemleak detect
a suspected memory leak as below.

If kunit_filter_suites() in kunit_module_init() succeeds, the
suite_set.start will not be NULL and the kunit_free_suite_set() in
kunit_module_exit() should free all the memory which has not
been freed. However the test_cases in suites is left out.

unreferenced object 0xffff54ac47e83200 (size 512):
  comm "modprobe", pid 592, jiffies 4294913238 (age 1367.612s)
  hex dump (first 32 bytes):
    84 13 1a f0 d3 b6 ff ff 30 68 1a f0 d3 b6 ff ff  ........0h......
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<000000008dec63a2>] slab_post_alloc_hook+0xb8/0x368
    [<00000000ec280d8e>] __kmem_cache_alloc_node+0x174/0x290
    [<00000000896c7740>] __kmalloc+0x60/0x2c0
    [<000000007a50fa06>] kunit_filter_suites+0x254/0x5b8
    [<0000000078cc98e2>] kunit_module_notify+0xf4/0x240
    [<0000000033cea952>] notifier_call_chain+0x98/0x17c
    [<00000000973d05cc>] notifier_call_chain_robust+0x4c/0xa4
    [<000000005f95895f>] blocking_notifier_call_chain_robust+0x4c/0x74
    [<0000000048e36fa7>] load_module+0x1a2c/0x1c40
    [<0000000004eb8a91>] init_module_from_file+0x94/0xcc
    [<0000000037dbba28>] idempotent_init_module+0x184/0x278
    [<00000000161b75cb>] __arm64_sys_finit_module+0x68/0xa8
    [<000000006dc1669b>] invoke_syscall+0x44/0x100
    [<00000000fa87e304>] el0_svc_common.constprop.1+0x68/0xe0
    [<000000009d8ad866>] do_el0_svc+0x1c/0x28
    [<000000005b83c607>] el0_svc+0x3c/0xc4

Fixes: a127b154a8f2 ("kunit: tool: allow filtering test cases via glob")
Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
Reviewed-by: Rae Moar <rmoar@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

soc: qcom: llcc: Handle a second device without data corruption

[ Upstream commit f1a1bc8775b26345aba2be278118999e7f661d3d ]

Usually there is only one llcc device. But if there were a second, even
a failed probe call would modify the global drv_data pointer. So check
if drv_data is valid before overwriting it.

Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Fixes: a3134fb09e0b ("drivers: soc: Add LLCC driver")
Link: https://lore.kernel.org/r/20230926083229.2073890-1-u.kleine-koenig@pengutronix.de
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ARM: dts: qcom: mdm9615: populate vsdcc fixed regulator

[ Upstream commit 09f8ee81b6da5f76de8b83c8bfc4475b54e101e0 ]

Fixed regulator put under "regulators" node will not be populated,
unless simple-bus or something similar is used.  Drop the "regulators"
wrapper node to fix this.

Fixes: 2c5e596524e7 ("ARM: dts: Add MDM9615 dtsi")
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20230924183914.51414-3-krzysztof.kozlowski@linaro.org
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ARM: dts: qcom: apq8026-samsung-matisse-wifi: Fix inverted hall sensor

[ Upstream commit 0b73519790d29e4bc71afc4882a9aa9ea649bcf7 ]

Fix hall sensor GPIO polarity and also allow disabling the sensor.
Remove unneeded interrupt.

Fixes: f15623bda1dc ("ARM: dts: qcom: Add support for Samsung Galaxy Tab 4 10.1 (SM-T530)")
Signed-off-by: Matti Lehtimäki <matti.lehtimaki@gmail.com>
Reviewed-by: Luca Weiss <luca@z3ntu.xyz>
Link: https://lore.kernel.org/r/20230922011211.115234-1-matti.lehtimaki@gmail.com
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: qcom: apq8016-sbc: Add missing ADV7533 regulators

[ Upstream commit 33e9032a1875bb1aee3c68a4540f5a577ff44130 ]

Add the missing regulator supplies to the ADV7533 HDMI bridge to fix
the following dtbs_check warnings. They are all also supplied by
pm8916_l6 so there is no functional difference.

apq8016-sbc.dtb: bridge@39: 'dvdd-supply' is a required property
apq8016-sbc.dtb: bridge@39: 'pvdd-supply' is a required property
apq8016-sbc.dtb: bridge@39: 'a2vdd-supply' is a required property
        from schema display/bridge/adi,adv7533.yaml

Fixes: 28546b095511 ("arm64: dts: apq8016-sbc: Add HDMI display support")
Signed-off-by: Stephan Gerhold <stephan@gerhold.net>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20230922-db410c-adv7533-regulators-v1-1-68aba71e529b@gerhold.net
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

riscv: dts: allwinner: remove address-cells from intc node

[ Upstream commit 267860b10c67dd396c73a9e6e8103670d78a4c01 ]

A recent submission [1] from Rob has added additionalProperties: false
to the interrupt-controller child node of RISC-V cpus, highlighting that
the D1 DT has been incorrectly using #address-cells since its
introduction. It has no child nodes, so #address-cells is not needed.
Remove it.

Fixes: 077e5f4f5528 ("riscv: dts: allwinner: Add the D1/D1s SoC devicetree")
Link: https://patchwork.kernel.org/project/linux-riscv/patch/20230915201946.4184468-1-robh@kernel.org/
Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
Acked-by: Jernej Skrabec <jernej.skrabec@gmail.com>
Link: https://lore.kernel.org/r/20230916-saddling-dastardly-8cf6d1263c24@spud
Signed-off-by: Jernej Skrabec <jernej.skrabec@gmail.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: qcom: msm8939: Fix iommu local address range

[ Upstream commit d40291e52d5ac4d0ff18ca433e84e6ddccc13427 ]

Fix the apps iommu local address space range as per data sheet.

Fixes: 61550c6c156c ("arm64: dts: qcom: Add msm8939 SoC")
Signed-off-by: Gaurav Kohli <quic_gkohli@quicinc.com>
Reviewed-by: Stephan Gerhold <stephan@gerhold.net>
Link: https://lore.kernel.org/r/20230917140039.25283-1-quic_gkohli@quicinc.com
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: qcom: msm8976: Fix ipc bit shifts

[ Upstream commit 684277525c70f329300cc687e27248e405a4ff9e ]

Update bits to match downstream irq-bitmask values.

Fixes: 0484d3ce0902 ("arm64: dts: qcom: Add DTS for MSM8976 and MSM8956 SoCs")
Signed-off-by: Adam Skladowski <a39.skl@gmail.com>
Link: https://lore.kernel.org/r/20230812112534.8610-8-a39.skl@gmail.com
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ARM64: dts: marvell: cn9310: Use appropriate label for spi1 pins

[ Upstream commit 0878fd86f554ab98aa493996c7e0c72dff58437f ]

Both the CN9130-CRB and CN9130-DB use the SPI1 interface but had the
pinctrl node labelled as "cp0_spi0_pins". Use the label "cp0_spi1_pins"
and update the node name to "cp0-spi-pins-1" to avoid confusion with the
pinctrl options for SPI0.

Fixes: 4c43a41e5b8c ("arm64: dts: cn913x: add device trees for topology B boards")
Fixes: 5c0ee54723f3 ("arm64: dts: add support for Marvell cn9130-crb platform")
Signed-off-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Gregory CLEMENT <gregory.clement@bootlin.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: qcom: sdx75-idp: align RPMh regulator nodes with bindings

[ Upstream commit 815ea491460766dbd4b39a3c9904b44b5880c41c ]

Device node names should be generic and bindings expect certain pattern
for RPMh regulator nodes:

  sdx75-idp.dtb: rsc@17a00000: 'pmx75-rpmh-regulators' does not match any of the regexes: '^regulators(-[0-9])?$', 'pinctrl-[0-9]+'

Fixes: 8a2dc39d1043 ("arm64: dts: qcom: sdx75-idp: Add regulator nodes")
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Link: https://lore.kernel.org/r/20230905163103.257412-1-krzysztof.kozlowski@linaro.org
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: qcom: sdm845-mtp: fix WiFi configuration

[ Upstream commit b33868a52f342d9b1f20aa5bffe40cbd69bd0a4b ]

Enable the host-cap-8bit quirk on this device. It is required for the
WiFi to function properly.

Fixes: 022bccb840b7 ("arm64: dts: sdm845: Add WCN3990 WLAN module device node")
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20230826221915.846937-2-dmitry.baryshkov@linaro.org
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: qcom: sm8350: fix pinctrl for UART18

[ Upstream commit c1efa960114f743924b884da098298512a7e9983 ]

On sm8350 QUP18 uses GPIO 68/69, not 58/59. Fix correponding UART18
pinconf configuraion.

Fixes: 98374e6925b8 ("arm64: dts: qcom: sm8350: Set up WRAP2 QUPs")
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Link: https://lore.kernel.org/r/20230825214550.1650938-1-dmitry.baryshkov@linaro.org
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: qcom: sm8150: add ref clock to PCIe PHYs

[ Upstream commit c204b3709409279ac019f3d374e444bb0b1424f0 ]

Follow the rest of the platforms and add "ref" clocks to both PCIe PHYs
found on the Qualcomm SM8150 platform.

Fixes: a1c86c680533 ("arm64: dts: qcom: sm8150: Add PCIe nodes")
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20230820142035.89903-15-dmitry.baryshkov@linaro.org
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: qcom: sc7280: drop incorrect EUD port on SoC side

[ Upstream commit 39c8af78cbefb8c71a5ad1fa088e761ef418f0a0 ]

Qualcomm Embedded USB Debugger (EUD) second port should point to Type-C
USB connector.  Such connector was defined directly in root node of
sc7280.dtsi which is clearly wrong.  SC7280 is a chip, so physically it
does not have USB Type-C port.  The connector is usually accessible
through some USB switch or controller.

Doug Anderson said that he wasn't ever able to use EUD on Herobrine
boards, probably because of invalid or missing DTS description - DTS is
saying EUD is on usb_2 node, which is connected to a USB Hub, not to the
Type-C port.

Correct the EUD/USB connector topology by removing the top-level fake
USB connector and EUD port pointing to it, and disabling the incomplete
EUD device node.

This fixes also dtbs_check warnings:

  sc7280-herobrine-crd.dtb: connector: ports:port@0: 'reg' is a required property

Link: https://lore.kernel.org/all/CAD=FV=Xt26=rBf99mzkAuwwtb2f-jnKtnHaEhXnthz0a5zke4Q@mail.gmail.com/
Fixes: 9ee402ccfeb1 ("arm64: dts: qcom: sc7280: Fix EUD dt node syntax")
Cc: Souradeep Chowdhury <quic_schowdhu@quicinc.com>
Cc: Bhupesh Sharma <bhupesh.sharma@linaro.org>
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Link: https://lore.kernel.org/r/20230820075626.22600-1-krzysztof.kozlowski@linaro.org
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: qcom: sdm670: Fix pdc mapping

[ Upstream commit ad75cda991f7b335d3b2417f82db07680f92648a ]

As pointed out by Richard, I missed a non-continuity in one of the ranges.
Fix it.

Reported-by: Richard Acayan <mailingradian@gmail.com>
Fixes: b51ee205dc4f ("arm64: dts: qcom: sdm670: Add PDC")
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Acked-by: Richard Acayan <mailingradian@gmail.com>
Link: https://lore.kernel.org/r/20230818-topic-670_pdc_fix-v1-1-1ba025041de7@linaro.org
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: qcom: qrb2210-rb1: Fix regulators

[ Upstream commit 31bee70793b67f4b428825434542afc72ddb2b3b ]

Commit b4fe47d12f1f ("arm64: dts: qcom: qrb2210-rb1: Add regulators")
introduced regulator settings that were never put in place, as all of the
properties ended 'microvolts' instead of 'microvolt' (which dt schema did
not check for back then).

Fix the microvolts-microvolt typo and adjust voltage ranges where it's
necessary to fit within the volt = base + n*step formula.

Reported-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
Reported-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Fixes: b4fe47d12f1f ("arm64: dts: qcom: qrb2210-rb1: Add regulators")
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20230906-topic-rb1_features_sans_icc-v1-2-e92ce6fbde16@linaro.org
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: qcom: qrb2210-rb1: Swap UART index

[ Upstream commit 973c015facabcbd320063648010942c51992c1a1 ]

Newer RB1 board revisions have a debug UART on QUP0. Sadly, it looks
like even when ordering one in retail, customers receive prototype
boards with "Enginering Sample" written on them.

Use QUP4 for UART to make all known RB1 boards boot.

Fixes: e18771961336 ("arm64: dts: qcom: Add initial QTI RB1 device tree")
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reported-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20230906-topic-rb1_features_sans_icc-v1-1-e92ce6fbde16@linaro.org
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: qcom: sc7280: Add missing LMH interrupts

[ Upstream commit 3f93d119c9d6e1744d55cd48af764160a1a3aca3 ]

Hook up the interrupts that signal the Limits Management Hardware has
started some sort of throttling action.

Fixes: 7dbd121a2c58 ("arm64: dts: qcom: sc7280: Add cpufreq hw node")
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Link: https://lore.kernel.org/r/20230811-topic-7280_lmhirq-v1-1-c262b6a25c8f@linaro.org
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: qcom: sm6125: Pad APPS IOMMU address to 8 characters

[ Upstream commit 310cdafc4a56827d1aeda7cc297939034adb8f99 ]

APPS IOMMU is the only node in sm6125.dtsi that doesn't have its
address padded to 8 hexadecimals; fix this by prepending a 0.

Fixes: 8ddb4bc3d3b5 ("arm64: dts: qcom: sm6125: Configure APPS SMMU")
Signed-off-by: Marijn Suijten <marijn.suijten@somainline.org>
Link: https://lore.kernel.org/r/20230723-sm6125-dpu-v4-2-a3f287dd6c07@somainline.org
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: qcom: msm8992-libra: drop duplicated reserved memory

[ Upstream commit f32096602c19e68fb9bf04b494d13f1190602554 ]

There are two entries for similar reserved memory: qseecom@cb400000 and
audio@cb400000.  Keep the qseecom as it is longer.

  Warning (unique_unit_address_if_enabled): /reserved-memory/audio@cb400000: duplicate unit-address (also used in node /reserved-memory/qseecom@cb400000)

Fixes: 69876bc6fd4d ("arm64: dts: qcom: msm8992-libra: Fix the memory map")
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20230720072048.10093-2-krzysztof.kozlowski@linaro.org
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: qcom: msm8916: Fix iommu local address range

[ Upstream commit 2de8ee9f58fa51f707c71f8fbcd8470ab0078102 ]

Fix the apps iommu local address space range as per data sheet.

Fixes: 6a6729f38436 ("arm64: dts: qcom: msm8916: Add IOMMU support")
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Tested-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Gaurav Kohli <quic_gkohli@quicinc.com>
Reviewed-by: Stephan Gerhold <stephan@gerhold.net>
Acked-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Link: https://lore.kernel.org/r/20230915143304.477-1-quic_gkohli@quicinc.com
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: qcom: sc7280: link usb3_phy_wrapper_gcc_usb30_pipe_clk

[ Upstream commit 70c4a1ca13b333b00e01266d299605fa1041b0d5 ]

Use usb_1_ssphy's clock as gcc's usb3_phy_wrapper_gcc_usb30_pipe_clk
clock source.

Suggested-by: Neil Armstrong <neil.armstrong@linaro.org>
Fixes: 1c39e6f9b534 ("arm64: dts: qcom: sc7280: Add USB related nodes")
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20230711120916.4165894-7-dmitry.baryshkov@linaro.org
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: qcom: sdm845: cheza doesn't support LMh node

[ Upstream commit 197ae69d1caedb3203e0b189a39efb820675fd5c ]

Cheza firmware doesn't allow controlling LMh from the operating system.

Fixes: 36c6581214c4 ("arm64: dts: qcom: sdm845: Add support for LMh node")
Suggested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: David Heidelberg <david@ixit.cz>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Link: https://lore.kernel.org/r/20230912071205.11502-2-david@ixit.cz
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: qcom: sdm845: Fix PSCI power domain names

[ Upstream commit a5f01673d3946e424091e6b8ff274716f9c21454 ]

The original commit hasn't been updated according to
refactoring done in sdm845.dtsi.

Fixes: a1ade6cac5a2 ("arm64: dts: qcom: sdm845: Switch PSCI cpu idle states from PC to OSI")
Suggested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: David Heidelberg <david@ixit.cz>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Reviewed-by: Abel Vesa <abel.vesa@linaro.org>
Link: https://lore.kernel.org/r/20230912071205.11502-1-david@ixit.cz
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ARM: dts: renesas: blanche: Fix typo in GP_11_2 pin name

[ Upstream commit edc6ef026fe69154bb6b70dd6e7f278cfd7d6919 ]

On blanche, the GPIO keyboard fails to probe with:

    sh-pfc e6060000.pinctrl: could not map pin config for "GP_11_02"

Fix this by correcting the name for this pin to "GP_11_2".

Fixes: 1f27fedead91eb60 ("ARM: dts: blanche: Configure pull-up for SOFT_SW and SW25 GPIO keys")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/203128eca2261ffc33b83637818dd39c488f42b0.1693408326.git.geert+renesas@glider.be
Signed-off-by: Sasha Levin <sashal@kernel.org>

perf: hisi: Fix use-after-free when register pmu fails

[ Upstream commit b805cafc604bfdb671fae7347a57f51154afa735 ]

When we fail to register the uncore pmu, the pmu context may not been
allocated. The error handing will call cpuhp_state_remove_instance()
to call uncore pmu offline callback, which migrate the pmu context.
Since that's liable to lead to some kind of use-after-free.

Use cpuhp_state_remove_instance_nocalls() instead of
cpuhp_state_remove_instance() so that the notifiers don't execute after
the PMU device has been failed to register.

Fixes: a0ab25cd82ee ("drivers/perf: hisi: Add support for HiSilicon PA PMU driver")
FIxes: 3bf30882c3c7 ("drivers/perf: hisi: Add support for HiSilicon SLLC PMU driver")
Signed-off-by: Junhao He <hejunhao3@huawei.com>
Link: https://lore.kernel.org/r/20231024113630.13472-1-hejunhao3@huawei.com
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drivers/perf: hisi_pcie: Check the type first in pmu::event_init()

[ Upstream commit 6d7d51e88e21c0af1ca96a3617afef334bfeffcf ]

Check whether the event type matches the PMU type firstly in
pmu::event_init() before touching the event. Otherwise we'll
change the events of others and lead to incorrect results.
Since in perf_init_event() we may call every pmu's event_init()
in a certain case, we should not modify the event if it's not
ours.

Fixes: 8404b0fbc7fb ("drivers/perf: hisi: Add driver for HiSilicon PCIe PMU")
Signed-off-by: Yicong Yang <yangyicong@hisilicon.com>
Link: https://lore.kernel.org/r/20231024092954.42297-2-yangyicong@huawei.com
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

perf/arm-cmn: Fix DTC domain detection

[ Upstream commit e3e73f511c49c741f6309862c2248958ad77bbaa ]

It transpires that dtm_unit_info is another register which got shuffled
in CMN-700 without me noticing. Fix that in a way which also proactively
fixes the fragile laziness of its consumer, just in case any further
fields ever get added alongside dtc_domain.

Fixes: 23760a014417 ("perf/arm-cmn: Add CMN-700 support")
Signed-off-by: Robin Murphy <robin.murphy@arm.com>
Reviewed-by: Ilkka Koskinen <ilkka@os.amperecomputing.com>
Link: https://lore.kernel.org/r/3076ee83d0554f6939fbb6ee49ab2bdb28d8c7ee.1697824215.git.robin.murphy@arm.com
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/amd/pm: Fix a memory leak on an error path

[ Upstream commit 828f8e31379b28fe7f07fb5865b8ed099d223fca ]

Add missing free on an error path.

Fixes: 511a95552ec8 ("drm/amd/pm: Add SMU 13.0.6 support")
Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Kunwu.Chan <chentao@kylinos.cn>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drivers/perf: hisi: use cpuhp_state_remove_instance_nocalls() for hisi_hns3_pmu uninit process

[ Upstream commit 50b560783f7f71790bcf70e9e9855155fb0af8c1 ]

When tearing down a 'hisi_hns3' PMU, we mistakenly run the CPU hotplug
callbacks after the device has been unregistered, leading to fireworks
when we try to execute empty function callbacks within the driver:

  | Unable to handle kernel NULL pointer dereference at virtual address 0000000000000000
  | CPU: 0 PID: 15 Comm: cpuhp/0 Tainted: G        W  O      5.12.0-rc4+ #1
  | Hardware name:  , BIOS KpxxxFPGA 1P B600 V143 04/22/2021
  | pstate: 80400009 (Nzcv daif +PAN -UAO -TCO BTYPE=--)
  | pc : perf_pmu_migrate_context+0x98/0x38c
  | lr : perf_pmu_migrate_context+0x94/0x38c
  |
  | Call trace:
  |  perf_pmu_migrate_context+0x98/0x38c
  |  hisi_hns3_pmu_offline_cpu+0x104/0x12c [hisi_hns3_pmu]

Use cpuhp_state_remove_instance_nocalls() instead of
cpuhp_state_remove_instance() so that the notifiers don't execute after
the PMU device has been unregistered.

Fixes: 66637ab137b4 ("drivers/perf: hisi: add driver for HNS3 PMU")
Signed-off-by: Hao Chen <chenhao418@huawei.com>
Signed-off-by: Jijie Shao <shaojijie@huawei.com>
Reviewed-by: Yicong Yang <yangyicong@hisilicon.com>
Link: https://lore.kernel.org/r/20231019091352.998964-1-shaojijie@huawei.com
[will: Rewrote commit message]
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm: mediatek: mtk_dsi: Fix NO_EOT_PACKET settings/handling

[ Upstream commit 5855d422a6f250f3518f43b49092c8e87a5e42be ]

Due to the initial confusion about MIPI_DSI_MODE_EOT_PACKET, properly
renamed to MIPI_DSI_MODE_NO_EOT_PACKET, reflecting its actual meaning,
both the DSI_TXRX_CON register setting for bit (HSTX_)DIS_EOT and the
later calculation for horizontal sync-active (HSA), back (HBP) and
front (HFP) porches got incorrect due to the logic being inverted.

This means that a number of settings were wrong because....:
 - DSI_TXRX_CON register setting: bit (HSTX_)DIS_EOT should be
   set in order to disable the End of Transmission packet;
 - Horizontal Sync and Back/Front porches: The delta used to
   calculate all of HSA, HBP and HFP should account for the
   additional EOT packet.

Before this change...
 - Bit (HSTX_)DIS_EOT was being set when EOT packet was enabled;
 - For HSA/HBP/HFP delta... all three were wrong, as words were
   added when EOT disabled, instead of when EOT packet enabled!

Invert the logic around flag MIPI_DSI_MODE_NO_EOT_PACKET in the
MediaTek DSI driver to fix the aforementioned issues.

Fixes: 8b2b99fd7931 ("drm/mediatek: dsi: Fine tune the line time caused by EOTp")
Fixes: c87d1c4b5b9a ("drm/mediatek: dsi: Use symbolized register definition")
Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Reviewed-by: Alexandre Mergnat <amergnat@baylibre.com>
Tested-by: Michael Walle <mwalle@kernel.org>
Link: https://patchwork.kernel.org/project/dri-devel/patch/20230523104234.7849-1-angelogioacchino.delregno@collabora.com/
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clocksource/drivers/arm_arch_timer: limit XGene-1 workaround

[ Upstream commit 851354cbd12bb9500909733c3d4054306f61df87 ]

The AppliedMicro XGene-1 CPU has an erratum where the timer condition
would only consider TVAL, not CVAL. We currently apply a workaround when
seeing the PartNum field of MIDR_EL1 being 0x000, under the assumption
that this would match only the XGene-1 CPU model.
However even the Ampere eMAG (aka XGene-3) uses that same part number, and
only differs in the "Variant" and "Revision" fields: XGene-1's MIDR is
0x500f0000, our eMAG reports 0x503f0002. Experiments show the latter
doesn't show the faulty behaviour.

Increase the specificity of the check to only consider partnum 0x000 and
variant 0x00, to exclude the Ampere eMAG.

Fixes: 012f18850452 ("clocksource/drivers/arm_arch_timer: Work around broken CVAL implementations")
Reported-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Acked-by: Marc Zyngier <maz@kernel.org>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Link: https://lore.kernel.org/r/20231016153127.116101-1-andre.przywara@arm.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/msm/dsi: free TX buffer in unbind

[ Upstream commit 5e05be78264594634860087953649487f486ffcc ]

If the drm/msm init code gets an error during output modeset
initialisation, the kernel will report an error regarding DRM memory
manager not being clean during shutdown. This is because
msm_dsi_modeset_init() allocates a piece of GEM memory for the TX
buffer, but destruction of the buffer happens only at
msm_dsi_host_destroy(), which is called during DSI driver's remove()
time, much later than the DRM MM shutdown.

To solve this issue, move the TX buffer destruction to dsi_unbind(), so
that the buffer is destructed at the correct time. Note, we also have to
store a reference to the address space, because priv->kms->aspace is
cleared before components are unbound.

Reported-by: Bjorn Andersson <andersson@kernel.org>
Fixes: 8f59ee9a570c ("drm/msm/dsi: Adjust probe order")
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/562238/
Signed-off-by: Rob Clark <robdclark@chromium.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/msm/dsi: use msm_gem_kernel_put to free TX buffer

[ Upstream commit 69b321b2c3df4f7e51a9de587e41f324b0b717b0 ]

Use exiting function to free the allocated GEM object instead of
open-coding it. This has a bonus of internally calling
msm_gem_put_vaddr() to compensate for msm_gem_get_vaddr() in
msm_get_kernel_new().

Fixes: 1e29dff00400 ("drm/msm: Add a common function to free kernel buffer objects")
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Patchwork: https://patchwork.freedesktop.org/patch/562239/
Signed-off-by: Rob Clark <robdclark@chromium.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

xen-pciback: Consider INTx disabled when MSI/MSI-X is enabled

[ Upstream commit 2c269f42d0f382743ab230308b836ffe5ae9b2ae ]

Linux enables MSI-X before disabling INTx, but keeps MSI-X masked until
the table is filled. Then it disables INTx just before clearing MASKALL
bit. Currently this approach is rejected by xen-pciback.
According to the PCIe spec, device cannot use INTx when MSI/MSI-X is
enabled (in other words: enabling MSI/MSI-X implicitly disables INTx).

Change the logic to consider INTx disabled if MSI/MSI-X is enabled. This
applies to three places:
 - checking currently enabled interrupts type,
 - transition to MSI/MSI-X - where INTx would be implicitly disabled,
 - clearing INTx disable bit - which can be allowed even if MSI/MSI-X is
   enabled, as device should consider INTx disabled anyway in that case

Fixes: 5e29500eba2a ("xen-pciback: Allow setting PCI_MSIX_FLAGS_MASKALL too")
Signed-off-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
Acked-by: Juergen Gross <jgross@suse.com>
Link: https://lore.kernel.org/r/20231016131348.1734721-1-marmarek@invisiblethingslab.com
Signed-off-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

xen: irqfd: Use _IOW instead of the internal _IOC() macro

[ Upstream commit 767e33ca47dd8ace7769e0b0c19d7b0c38b2f72d ]

_IOC() an internal helper that we should not use in driver code.  In
particular, we got the data direction wrong here, which breaks a number
of tools, as having "_IOC_NONE" should never be paired with a nonzero
size.

Use _IOW() instead.

Fixes: f8941e6c4c71 ("xen: privcmd: Add support for irqfd")
Reported-by: Arnd Bergmann <arnd@kernel.org>
Closes: https://lore.kernel.org/all/268a2031-63b8-4c7d-b1e5-8ab83ca80b4a@app.fastmail.com/
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Juergen Gross <jgross@suse.com>
Link: https://lore.kernel.org/r/599ca6f1b9dd2f0e6247ea37bee3ea6827404b6d.1697439990.git.viresh.kumar@linaro.org
Signed-off-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

xen: Make struct privcmd_irqfd's layout architecture independent

[ Upstream commit 8dd765a5d769c521d73931850d1c8708fbc490cb ]

Using indirect pointers in an ioctl command argument means that the
layout is architecture specific, in particular we can't use the same one
from 32-bit compat tasks. The general recommendation is to have __u64
members and use u64_to_user_ptr() to access it from the kernel if we are
unable to avoid the pointers altogether.

Fixes: f8941e6c4c71 ("xen: privcmd: Add support for irqfd")
Reported-by: Arnd Bergmann <arnd@kernel.org>
Closes: https://lore.kernel.org/all/268a2031-63b8-4c7d-b1e5-8ab83ca80b4a@app.fastmail.com/
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Juergen Gross <jgross@suse.com>
Link: https://lore.kernel.org/r/a4ef0d4a68fc858b34a81fd3f9877d9b6898eb77.1697439990.git.viresh.kumar@linaro.org
Signed-off-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

xenbus: fix error exit in xenbus_init()

[ Upstream commit 44961b81a9e9059b5c0443643915386db7035227 ]

In case an error occurs in xenbus_init(), xen_store_domain_type should
be set to XS_UNKNOWN.

Fix one instance where this action is missing.

Fixes: 5b3353949e89 ("xen: add support for initializing xenstore later as HVM domain")
Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Dan Carpenter <error27@gmail.com>
Link: https://lore.kernel.org/r/202304200845.w7m4kXZr-lkp@intel.com/
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Oleksandr Tyshchenko <oleksandr_tyshchenko@epam.com>
Link: https://lore.kernel.org/r/20230822091138.4765-1-jgross@suse.com
Signed-off-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/rockchip: Fix type promotion bug in rockchip_gem_iommu_map()

[ Upstream commit 6471da5ee311d53ef46eebcb7725bc94266cc0cf ]

The "ret" variable is declared as ssize_t and it can hold negative error
codes but the "rk_obj->base.size" variable is type size_t.  This means
that when we compare them, they are both type promoted to size_t and the
negative error code becomes a high unsigned value and is treated as
success.  Add a cast to fix this.

Fixes: 38f993b7c59e ("drm/rockchip: Do not use DMA mapping API if attached to IOMMU domain")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Link: https://patchwork.freedesktop.org/patch/msgid/2bfa28b5-145d-4b9e-a18a-98819dd686ce@moroto.mountain
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64/arm: xen: enlighten: Fix KPTI checks

[ Upstream commit 20f3b8eafe0ba5d3c69d5011a9b07739e9645132 ]

When KPTI is in use, we cannot register a runstate region as XEN
requires that this is always a valid VA, which we cannot guarantee. Due
to this, xen_starting_cpu() must avoid registering each CPU's runstate
region, and xen_guest_init() must avoid setting up features that depend
upon it.

We tried to ensure that in commit:

  f88af7229f6f22ce (" xen/arm: do not setup the runstate info page if kpti is enabled")

... where we added checks for xen_kernel_unmapped_at_usr(), which wraps
arm64_kernel_unmapped_at_el0() on arm64 and is always false on 32-bit
arm.

Unfortunately, as xen_guest_init() is an early_initcall, this happens
before secondary CPUs are booted and arm64 has finalized the
ARM64_UNMAP_KERNEL_AT_EL0 cpucap which backs
arm64_kernel_unmapped_at_el0(), and so this can subsequently be set as
secondary CPUs are onlined. On a big.LITTLE system where the boot CPU
does not require KPTI but some secondary CPUs do, this will result in
xen_guest_init() intializing features that depend on the runstate
region, and xen_starting_cpu() registering the runstate region on some
CPUs before KPTI is subsequent enabled, resulting the the problems the
aforementioned commit tried to avoid.

Handle this more robsutly by deferring the initialization of the
runstate region until secondary CPUs have been initialized and the
ARM64_UNMAP_KERNEL_AT_EL0 cpucap has been finalized. The per-cpu work is
moved into a new hotplug starting function which is registered later
when we're certain that KPTI will not be used.

Fixes: f88af7229f6f ("xen/arm: do not setup the runstate info page if kpti is enabled")
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Cc: Bertrand Marquis <bertrand.marquis@arm.com>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: Juergen Gross <jgross@suse.com>
Cc: Stefano Stabellini <sstabellini@kernel.org>
Cc: Suzuki K Poulose <suzuki.poulose@arm.com>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: lt9611uxc: fix the race in the error path

[ Upstream commit 15fe53be46eaf4f6339cd433972ecc90513e3076 ]

If DSI host attachment fails, the LT9611UXC driver will remove the
bridge without ensuring that there is no outstanding HPD work being
done. In rare cases this can result in the warnings regarding the mutex
being incorrect. Fix this by forcebly freing IRQ and flushing the work.

DEBUG_LOCKS_WARN_ON(lock->magic != lock)
WARNING: CPU: 0 PID: 10 at kernel/locking/mutex.c:582 __mutex_lock+0x468/0x77c
Modules linked in:
CPU: 0 PID: 10 Comm: kworker/0:1 Tainted: G     U             6.6.0-rc5-next-20231011-gd81f81c2b682-dirty #1206
Hardware name: Qualcomm Technologies, Inc. Robotics RB5 (DT)
Workqueue: events lt9611uxc_hpd_work
pstate: 60400005 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
pc : __mutex_lock+0x468/0x77c
lr : __mutex_lock+0x468/0x77c
sp : ffff8000800a3c70
x29: ffff8000800a3c70 x28: 0000000000000000 x27: ffffd595fe333000
x26: ffff7c2f0002c005 x25: ffffd595ff1b3000 x24: ffffd595fccda5a0
x23: 0000000000000000 x22: 0000000000000002 x21: ffff7c2f056d91c8
x20: 0000000000000000 x19: ffff7c2f056d91c8 x18: fffffffffffe8db0
x17: 000000040044ffff x16: 005000f2b5503510 x15: 0000000000000000
x14: 000000000006efb8 x13: 0000000000000000 x12: 0000000000000037
x11: 0000000000000001 x10: 0000000000001470 x9 : ffff8000800a3ae0
x8 : ffff7c2f0027f8d0 x7 : ffff7c2f0027e400 x6 : ffffd595fc702b54
x5 : 0000000000000000 x4 : ffff8000800a0000 x3 : 0000000000000000
x2 : 0000000000000000 x1 : 0000000000000000 x0 : ffff7c2f0027e400
Call trace:
 __mutex_lock+0x468/0x77c
 mutex_lock_nested+0x24/0x30
 drm_bridge_hpd_notify+0x2c/0x5c
 lt9611uxc_hpd_work+0x6c/0x80
 process_one_work+0x1ec/0x51c
 worker_thread+0x1ec/0x3e4
 kthread+0x120/0x124
 ret_from_fork+0x10/0x20
irq event stamp: 15799
hardirqs last  enabled at (15799): [<ffffd595fc702ba4>] finish_task_switch.isra.0+0xa8/0x278
hardirqs last disabled at (15798): [<ffffd595fd5a1580>] __schedule+0x7b8/0xbd8
softirqs last  enabled at (15794): [<ffffd595fc690698>] __do_softirq+0x498/0x4e0
softirqs last disabled at (15771): [<ffffd595fc69615c>] ____do_softirq+0x10/0x1c

Fixes: bc6fa8676ebb ("drm/bridge/lontium-lt9611uxc: move HPD notification out of IRQ handler")
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Robert Foss <rfoss@kernel.org>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20231011220002.382422-1-dmitry.baryshkov@linaro.org
Signed-off-by: Sasha Levin <sashal@kernel.org>

gpu: host1x: Correct allocated size for contexts

[ Upstream commit e889a311f74f4ae8bd40755a2c58d02e1c684fef ]

Original implementation over allocates the memory size for the
contexts list. The size of memory for the contexts list is based
on the number of iommu groups specified in the device tree.

Fixes: 8aa5bcb61612 ("gpu: host1x: Add context device management code")
Signed-off-by: Johnny Liu <johnliu@nvidia.com>
Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230901115910.701518-1-cyndis@kapsi.fi
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/rockchip: cdn-dp: Fix some error handling paths in cdn_dp_probe()

[ Upstream commit 44b968d0d0868b7a9b7a5c64464ada464ff4d532 ]

cdn_dp_audio_codec_init() can fail. So add some error handling.

If component_add() fails, the previous cdn_dp_audio_codec_init() call
should be undone, as already done in the remove function.

Fixes: 88582f564692 ("drm/rockchip: cdn-dp: Don't unregister audio dev when unbinding")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Link: https://patchwork.freedesktop.org/patch/msgid/8494a41602fadb7439630921a9779640698f2f9f.1693676045.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/msm/a6xx: Fix unknown speedbin case

[ Upstream commit 75cb60d4f5f762b12643b67cbefefcf05ecfd7eb ]

When opp-supported-hw is present under an OPP node, but no form of
opp_set_supported_hw() has been called, that OPP is ignored by the API
and marked as unsupported.

Before Commit c928a05e4415 ("drm/msm/adreno: Move speedbin mapping to
device table"), an unknown speedbin would result in marking all OPPs
as available, but it's better to avoid potentially overclocking the
silicon - the GMU will simply refuse to power up the chip.

Currently, the Adreno speedbin code does just that (AND returns an
invalid error, (int)UINT_MAX). Fix that by defaulting to speedbin 0
(which is conveniently always bound to fuseval == 0).

Fixes: c928a05e4415 ("drm/msm/adreno: Move speedbin mapping to device table")
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/559604/
Signed-off-by: Rob Clark <robdclark@chromium.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/msm/adreno: Fix SM6375 GPU ID

[ Upstream commit beb3542320479cd59a08273be0b19dfea0b36042 ]

SM6375 comes with a patchlevel=1. Fix the chipid up to reflect that.

Fixes: 90b593ce1c9e ("drm/msm/adreno: Switch to chip-id for identifying GPU")
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/554527/
Signed-off-by: Rob Clark <robdclark@chromium.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

accel/habanalabs/gaudi2: Fix incorrect string length computation in gaudi2_psoc_razwi_get_engines()

[ Upstream commit 90f3de616259cbb5666f00f8daf5420cd7d71d18 ]

snprintf() returns the "number of characters which *would* be generated for
the given input", not the size *really* generated.

In order to avoid too large values for 'str_size' (and potential negative
values for "PSOC_RAZWI_ENG_STR_SIZE - str_size") use scnprintf()
instead of snprintf().

Fixes: c0e6df916050 ("accel/habanalabs: fix address decode RAZWI handling")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/mediatek: Fix iommu fault during crtc enabling

[ Upstream commit 53412dc2905401207f264dc30890f6b9e41524a6 ]

The difference between drm_atomic_helper_commit_tail() and
drm_atomic_helper_commit_tail_rpm() is
drm_atomic_helper_commit_tail() will commit plane first and
then enable crtc, drm_atomic_helper_commit_tail_rpm() will
enable crtc first and then commit plane.

Before mediatek-drm enables crtc, the power and clk required
by OVL have not been turned on, so the commit plane cannot be
committed before crtc is enabled. That means OVL layer should
not be enabled before crtc is enabled.
Therefore, the atomic_commit_tail of mediatek-drm is hooked with
drm_atomic_helper_commit_tail_rpm().

Another reason is that the plane_state of drm_atomic_state is not
synchronized with the plane_state stored in mtk_crtc during crtc enablng,
so just set all planes to disabled.

Fixes: 119f5173628a ("drm/mediatek: Add DRM Driver for Mediatek SoC MT8173.")
Signed-off-by: Jason-JH.Lin <jason-jh.lin@mediatek.com>
Reviewed-by: Alexandre Mergnat <amergnat@baylibre.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Reviewed-by: CK Hu <ck.hu@mediatek.com>
Link: https://patchwork.kernel.org/project/linux-mediatek/patch/20230809125722.24112-3-jason-jh.lin@mediatek.com/
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/mediatek: Fix iommu fault by swapping FBs after updating plane state

[ Upstream commit 3ec71e05ae6e7f46512e568ed81c92be589003dd ]

According to the comment in drm_atomic_helper_async_commit(),
we should make sure FBs have been swapped, so that cleanups in the
new_state performs a cleanup in the old FB.

So we should move swapping FBs after calling mtk_plane_update_new_state(),
to avoid using the old FB which could be freed.

Fixes: 1a64a7aff8da ("drm/mediatek: Fix cursor plane no update")
Signed-off-by: Jason-JH.Lin <jason-jh.lin@mediatek.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Reviewed-by: CK Hu <ck.hu@mediatek.com>
Reviewed-by: Alexandre Mergnat <amergnat@baylibre.com>
Link: https://patchwork.kernel.org/project/linux-mediatek/patch/20230809125722.24112-2-jason-jh.lin@mediatek.com/
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/mediatek: Add mmsys_dev_num to mt8188 vdosys0 driver data

[ Upstream commit ff64e4c31d969cdba20a41969edb3def15f3aaa0 ]

Add missing mmsys_dev_num to mt8188 vdosys0 driver data.

Fixes: 54b48080278a ("drm/mediatek: Add mediatek-drm of vdosys0 support for mt8188")
Signed-off-by: Jason-JH.Lin <jason-jh.lin@mediatek.com>
Reviewed-by: CK Hu <ck.hu@mediatek.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Reviewed-by: Fei Shao <fshao@chromium.org>
Tested-by: Fei Shao <fshao@chromium.org>
Link: https://patchwork.kernel.org/project/dri-devel/patch/20231004024013.18956-2-jason-jh.lin@mediatek.com/
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

io_uring/kbuf: Allow the full buffer id space for provided buffers

[ Upstream commit f74c746e476b9dad51448b9a9421aae72b60e25f ]

nbufs tracks the number of buffers and not the last bgid. In 16-bit, we
have 2^16 valid buffers, but the check mistakenly rejects the last
bid. Let's fix it to make the interface consistent with the
documentation.

Fixes: ddf0322db79c ("io_uring: add IORING_OP_PROVIDE_BUFFERS")
Signed-off-by: Gabriel Krisman Bertazi <krisman@suse.de>
Link: https://lore.kernel.org/r/20231005000531.30800-3-krisman@suse.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>

io_uring/kbuf: Fix check of BID wrapping in provided buffers

[ Upstream commit ab69838e7c75b0edb699c1a8f42752b30333c46f ]

Commit 3851d25c75ed0 ("io_uring: check for rollover of buffer ID when
providing buffers") introduced a check to prevent wrapping the BID
counter when sqe->off is provided, but it's off-by-one too
restrictive, rejecting the last possible BID (65534).

i.e., the following fails with -EINVAL.

     io_uring_prep_provide_buffers(sqe, addr, size, 0xFFFF, 0, 0);

Fixes: 3851d25c75ed ("io_uring: check for rollover of buffer ID when providing buffers")
Signed-off-by: Gabriel Krisman Bertazi <krisman@suse.de>
Link: https://lore.kernel.org/r/20231005000531.30800-2-krisman@suse.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/amd/display: Bail from dm_check_crtc_cursor if no relevant change

[ Upstream commit bc0b79ce2050aa523c38c96b6d26340a96bfbdca ]

If no plane was newly enabled or changed scaling, there can be no new
scaling mismatch with the cursor plane.

By not pulling non-cursor plane states into all atomic commits while
the cursor plane is enabled, this avoids synchronizing all cursor plane
changes to vertical blank, which caused the following IGT tests to fail:

kms_cursor_legacy@cursor-vs-flip.*
kms_cursor_legacy@flip-vs-cursor.*

Fixes: 003048ddf44b ("drm/amd/display: Check all enabled planes in dm_check_crtc_cursor")
Signed-off-by: Michel Dänzer <mdaenzer@redhat.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/amd/display: Refactor dm_get_plane_scale helper

[ Upstream commit ec4d770bbb155674c2497f255f4199bdc42287a9 ]

Cleanup, no functional change intended.

Signed-off-by: Michel Dänzer <mdaenzer@redhat.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Stable-dep-of: bc0b79ce2050 ("drm/amd/display: Bail from dm_check_crtc_cursor if no relevant change")
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/amd/display: Check all enabled planes in dm_check_crtc_cursor

[ Upstream commit 003048ddf44b1a6cfa57afa5a0cf40673e13f1ba ]

It was only checking planes which had any state changes in the same
commit. However, it also needs to check other enabled planes.

Not doing this meant that a commit might spuriously "succeed", resulting
in the cursor plane displaying with incorrect scaling. See
https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3177#note_1824263
for an example.

Fixes: d1bfbe8a3202 ("amd/display: check cursor plane matches underlying plane")
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Michel Dänzer <mdaenzer@redhat.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/amd/display: Fix null pointer dereference in error message

[ Upstream commit 0c3601a2fbfb265ce283651480e30c8e60459112 ]

This patch fixes a null pointer dereference in the error message that is
printed when the Display Core (DC) fails to initialize. The original
message includes the DC version number, which is undefined if the DC is
not initialized.

Fixes: 9788d087caff ("drm/amd/display: improve the message printed when loading DC")
Signed-off-by: Cong Liu <liucong2@kylinos.cn>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/amdkfd: Handle errors from svm validate and map

[ Upstream commit eb3c357bcb286e89386e89302061fe717fe4e562 ]

If new range is splited to multiple pranges with max_svm_range_pages
alignment and added to update_list, svm validate and map should keep
going after error to make sure prange->mapped_to_gpu flag is up to date
for the whole range.

svm validate and map update set prange->mapped_to_gpu after mapping to
GPUs successfully, otherwise clear prange->mapped_to_gpu flag (for
update mapping case) instead of setting error flag, we can remove
the redundant error flag to simpliy code.

Refactor to remove goto and update prange->mapped_to_gpu flag inside
svm_range_lock, to guarant we always evict queues or unmap from GPUs if
there are invalid ranges.

After svm validate and map return error -EAGIN, the caller retry will
update the mapping for the whole range again.

Fixes: c22b04407097 ("drm/amdkfd: flag added to handle errors from svm validate and map")
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Tested-by: James Zhu <james.zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/amdkfd: Remove svm range validated_once flag

[ Upstream commit c99b16128082de519975aa147d9da3e40380de67 ]

The validated_once flag is not used after the prefault was removed, The
prefault was needed to ensure validate all system memory pages at least
once before mapping or migrating the range to GPU.

Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Stable-dep-of: eb3c357bcb28 ("drm/amdkfd: Handle errors from svm validate and map")
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/amdkfd: fix some race conditions in vram buffer alloc/free of svm code

[ Upstream commit 7bfaa160caed8192f8262c4638f552cad94bcf5a ]

This patch fixes:
1: ref number of prange's svm_bo got decreased by an async call from hmm. When
wait svm_bo of prange got released we shoul also wait prang->svm_bo become NULL,
otherwise prange->svm_bo may be set to null after allocate new vram buffer.

2: During waiting svm_bo of prange got released in a while loop should reschedule
current task to give other tasks oppotunity to run, specially the the workque
task that handles svm_bo ref release, otherwise we may enter to softlock.

Signed-off-by: Xiaogang.Chen <xiaogang.chen@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/amdgpu: Increase IH soft ring size for GFX v9.4.3 dGPU

[ Upstream commit bcfb9cee61207b80f37663ffa08c135657a27ad5 ]

On GFX v9.4.3 dGPU, applications have random timeout failure when XNACK
on, dmesg log has "amdgpu: IH soft ring buffer overflow 0x900, 0x900",
because dGPU mode has 272 cam entries. After increasing IH soft ring
to 512 entries, no more IH soft ring overflow message and application
passed.

Fixes: bf80d34b6c58 ("drm/amdgpu: Increase soft IH ring size")
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>