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>

drm: Call drm_atomic_helper_shutdown() at shutdown/remove time for misc drivers

[ Upstream commit 3c4babae3c4a1ae05f8f3f5f3d50c440ead7ca6a ]

Based on grepping through the source code these drivers appear to be
missing a call to drm_atomic_helper_shutdown() at system shutdown time
and at driver remove (or unbind) time. Among other things, this means
that if a panel is in use that it won't be cleanly powered off at
system shutdown time.

The fact that we should call drm_atomic_helper_shutdown() in the case
of OS shutdown/restart and at driver remove (or unbind) time comes
straight out of the kernel doc "driver instance overview" in
drm_drv.c.

A few notes about these fixes:
- I confirmed that these drivers were all DRIVER_MODESET type drivers,
  which I believe makes this relevant.
- I confirmed that these drivers were all DRIVER_ATOMIC.
- When adding drm_atomic_helper_shutdown() to the remove/unbind path,
  I added it after drm_kms_helper_poll_fini() when the driver had
  it. This seemed to be what other drivers did. If
  drm_kms_helper_poll_fini() wasn't there I added it straight after
  drm_dev_unregister().
- This patch deals with drivers using the component model in similar
  ways as the patch ("drm: Call drm_atomic_helper_shutdown() at
  shutdown time for misc drivers")
- These fixes rely on the patch ("drm/atomic-helper:
  drm_atomic_helper_shutdown(NULL) should be a noop") to simplify
  shutdown.

Suggested-by: Maxime Ripard <mripard@kernel.org>
Reviewed-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Tested-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com> # tilcdc
Acked-by: Maxime Ripard <mripard@kernel.org>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230901163944.RFT.5.I771eb4bd03d8772b19e7dcfaef3e2c167bce5846@changeid
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: tc358768: Fix tc358768_ns_to_cnt()

[ Upstream commit f1dabbe645065d20ca863c8d446c74c59ca1ca9d ]

The tc358768_ns_to_cnt() is, most likely, supposed to do a div-round-up
operation, but it misses subtracting one from the dividend.

Fix this by just using DIV_ROUND_UP().

Fixes: ff1ca6397b1d ("drm/bridge: Add tc358768 driver")
Reviewed-by: Peter Ujfalusi <peter.ujfalusi@gmail.com>
Tested-by: Maxim Schwalm <maxim.schwalm@gmail.com> # Asus TF700T
Tested-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230906-tc358768-v4-11-31725f008a50@ideasonboard.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: tc358768: Clean up clock period code

[ Upstream commit b3aa7b34924a9ed64cf96899cac4d8ea08cd829e ]

The driver defines TC358768_PRECISION as 1000, and uses "nsk" to refer
to clock periods. The original author does not remember where all this
came from. Effectively the driver is using picoseconds as the unit for
clock periods, yet referring to them by "nsk".

Clean this up by just saying the periods are in picoseconds.

Reviewed-by: Peter Ujfalusi <peter.ujfalusi@gmail.com>
Tested-by: Maxim Schwalm <maxim.schwalm@gmail.com> # Asus TF700T
Tested-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230906-tc358768-v4-10-31725f008a50@ideasonboard.com
Stable-dep-of: f1dabbe64506 ("drm/bridge: tc358768: Fix tc358768_ns_to_cnt()")
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: tc358768: Rename dsibclk to hsbyteclk

[ Upstream commit 699cf62a7d4550759f4a50e614b1952f93de4783 ]

The Toshiba documentation talks about HSByteClk when referring to the
DSI HS byte clock, whereas the driver uses 'dsibclk' name. Also, in a
few places the driver calculates the byte clock from the DSI clock, even
if the byte clock is already available in a variable.

To align the driver with the documentation, change the 'dsibclk'
variable to 'hsbyteclk'. This also make it easier to visually separate
'dsibclk' and 'dsiclk' variables.

Reviewed-by: Peter Ujfalusi <peter.ujfalusi@gmail.com>
Tested-by: Maxim Schwalm <maxim.schwalm@gmail.com> # Asus TF700T
Tested-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230906-tc358768-v4-9-31725f008a50@ideasonboard.com
Stable-dep-of: f1dabbe64506 ("drm/bridge: tc358768: Fix tc358768_ns_to_cnt()")
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: tc358768: Use dev for dbg prints, not priv->dev

[ Upstream commit 89cfd50e13f1bead4350998a3a77422bef1ee0a5 ]

Simplify the code by capturing the priv->dev value to dev variable, and
use it.

Reviewed-by: Peter Ujfalusi <peter.ujfalusi@gmail.com>
Tested-by: Maxim Schwalm <maxim.schwalm@gmail.com> # Asus TF700T
Tested-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230906-tc358768-v4-8-31725f008a50@ideasonboard.com
Stable-dep-of: f1dabbe64506 ("drm/bridge: tc358768: Fix tc358768_ns_to_cnt()")
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: tc358768: Print logical values, not raw register values

[ Upstream commit 013ea98cdfccef3b7c38b087c1f629488d2ef683 ]

The driver debug prints DSI related timings as raw register values in
hex. It is much more useful to see the "logical" value of the timing,
not the register value.

Change the prints to print the values separately, in case a single
register contains multiple values, and use %u to have it in a more human
consumable form.

Reviewed-by: Peter Ujfalusi <peter.ujfalusi@gmail.com>
Tested-by: Maxim Schwalm <maxim.schwalm@gmail.com> # Asus TF700T
Tested-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230906-tc358768-v4-7-31725f008a50@ideasonboard.com
Stable-dep-of: f1dabbe64506 ("drm/bridge: tc358768: Fix tc358768_ns_to_cnt()")
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: tc358768: Use struct videomode

[ Upstream commit e5fb21678136a9d009d5c43821881eb4c34fae97 ]

The TC358768 documentation uses HFP, HBP, etc. values to deal with the
video mode, while the driver currently uses the DRM display mode
(htotal, hsync_start, etc).

Change the driver to convert the DRM display mode to struct videomode,
which then allows us to use the same units the documentation uses. This
makes it much easier to work on the code when using the TC358768
documentation as a reference.

Reviewed-by: Peter Ujfalusi <peter.ujfalusi@gmail.com>
Tested-by: Maxim Schwalm <maxim.schwalm@gmail.com> # Asus TF700T
Tested-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230906-tc358768-v4-6-31725f008a50@ideasonboard.com
Stable-dep-of: f1dabbe64506 ("drm/bridge: tc358768: Fix tc358768_ns_to_cnt()")
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: tc358768: Fix bit updates

[ Upstream commit 66962d5c3c51377b9b90cae35b7e038950438e02 ]

The driver has a few places where it does:

if (thing_is_enabled_in_config)
update_thing_bit_in_hw()

This means that if the thing is _not_ enabled, the bit never gets
cleared. This affects the h/vsyncs and continuous DSI clock bits.

Fix the driver to always update the bit.

Fixes: ff1ca6397b1d ("drm/bridge: Add tc358768 driver")
Reviewed-by: Peter Ujfalusi <peter.ujfalusi@gmail.com>
Tested-by: Maxim Schwalm <maxim.schwalm@gmail.com> # Asus TF700T
Tested-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230906-tc358768-v4-4-31725f008a50@ideasonboard.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: tc358768: Fix use of uninitialized variable

[ Upstream commit a2d9036615f0adfa5b0a46bb2ce42ef1d9a04fbe ]

smatch reports:

drivers/gpu/drm/bridge/tc358768.c:223 tc358768_update_bits() error: uninitialized symbol 'orig'.

Fix this by bailing out from tc358768_update_bits() if the
tc358768_read() produces an error.

Fixes: ff1ca6397b1d ("drm/bridge: Add tc358768 driver")
Reviewed-by: Peter Ujfalusi <peter.ujfalusi@gmail.com>
Tested-by: Maxim Schwalm <maxim.schwalm@gmail.com> # Asus TF700T
Tested-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230906-tc358768-v4-2-31725f008a50@ideasonboard.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

x86/tdx: Zero out the missing RSI in TDX_HYPERCALL macro

[ Upstream commit 5d092b66119d774853cc9308522620299048a662 ]

In the TDX_HYPERCALL asm, after the TDCALL instruction returns from the
untrusted VMM, the registers that the TDX guest shares to the VMM need
to be cleared to avoid speculative execution of VMM-provided values.

RSI is specified in the bitmap of those registers, but it is missing
when zeroing out those registers in the current TDX_HYPERCALL.

It was there when it was originally added in commit 752d13305c78
("x86/tdx: Expand __tdx_hypercall() to handle more arguments"), but was
later removed in commit 1e70c680375a ("x86/tdx: Do not corrupt
frame-pointer in __tdx_hypercall()"), which was correct because %rsi is
later restored in the "pop %rsi".  However a later commit 7a3a401874be
("x86/tdx: Drop flags from __tdx_hypercall()") removed that "pop %rsi"
but forgot to add the "xor %rsi, %rsi" back.

Fix by adding it back.

Fixes: 7a3a401874be ("x86/tdx: Drop flags from __tdx_hypercall()")
Signed-off-by: Kai Huang <kai.huang@intel.com>
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/all/e7d1157074a0b45d34564d5f17f3e0ffee8115e9.1692096753.git.kai.huang%40intel.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/mediatek: Fix coverity issue with unintentional integer overflow

[ Upstream commit b0b0d811eac6b4c52cb9ad632fa6384cf48869e7 ]

1. Instead of multiplying 2 variable of different types. Change to
assign a value of one variable and then multiply the other variable.

2. Add a int variable for multiplier calculation instead of calculating
different types multiplier with dma_addr_t variable directly.

Fixes: 1a64a7aff8da ("drm/mediatek: Fix cursor plane no update")
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>
Link: https://patchwork.kernel.org/project/dri-devel/patch/20230907091425.9526-1-jason-jh.lin@mediatek.com/
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/ssd130x: Fix screen clearing

[ Upstream commit 4dbce3d6fea59e1df1d1a35aacea0c186f72107a ]

Due to the reuse of buffers, ssd130x_clear_screen() no longers clears
the screen, but merely redraws the last image that is residing in the
intermediate buffer.

As there is no point in clearing the intermediate buffer and transposing
an all-black image, fix this by just clearing the HW format buffer, and
writing it to the panel.

Fixes: 49d7d581ceaf4cf8 ("drm/ssd130x: Don't allocate buffers on each plane update")
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Reviewed-by: Javier Martinez Canillas <javierm@redhat.com>
Tested-by: Javier Martinez Canillas <javierm@redhat.com>
Signed-off-by: Javier Martinez Canillas <javierm@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/c19cd5a57205597bb38a446c3871092993498f01.1692888745.git.geert@linux-m68k.org
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: lt8912b: Add missing drm_bridge_attach call

[ Upstream commit f45acf7acf75921c0409d452f0165f51a19a74fd ]

The driver does not call drm_bridge_attach(), which causes the next
bridge to not be added to the bridge chain. This causes the pipeline
init to fail when DRM_BRIDGE_ATTACH_NO_CONNECTOR is used.

Add the call to drm_bridge_attach().

Fixes: 30e2ae943c26 ("drm/bridge: Introduce LT8912B DSI to HDMI bridge")
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Reviewed-by: Robert Foss <rfoss@kernel.org>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230804-lt8912b-v1-4-c542692c6a2f@ideasonboard.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: lt8912b: Manually disable HPD only if it was enabled

[ Upstream commit 6985c5efc4057bc79137807295d84ada3123d051 ]

lt8912b only calls drm_bridge_hpd_enable() if it creates a connector and
the next bridge has DRM_BRIDGE_OP_HPD set. However, when calling
drm_bridge_hpd_disable() it misses checking if a connector was created,
calling drm_bridge_hpd_disable() even if HPD was never enabled. I don't
see any issues caused by this wrong call, though.

Add the check to avoid wrongly calling drm_bridge_hpd_disable().

Fixes: 3b0a01a6a522 ("drm/bridge: lt8912b: Add hot plug detection")
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Tested-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Reviewed-by: Robert Foss <rfoss@kernel.org>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230804-lt8912b-v1-3-c542692c6a2f@ideasonboard.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: lt8912b: Fix crash on bridge detach

[ Upstream commit 44283993144a03af9df31934d6c32bbd42d1a347 ]

The lt8912b driver, in its bridge detach function, calls
drm_connector_unregister() and drm_connector_cleanup().

drm_connector_unregister() should be called only for connectors
explicitly registered with drm_connector_register(), which is not the
case in lt8912b.

The driver's drm_connector_funcs.destroy hook is set to
drm_connector_cleanup().

Thus the driver should not call either drm_connector_unregister() nor
drm_connector_cleanup() in its lt8912_bridge_detach(), as they cause a
crash on bridge detach:

Unable to handle kernel NULL pointer dereference at virtual address 0000000000000000
Mem abort info:
  ESR = 0x0000000096000006
  EC = 0x25: DABT (current EL), IL = 32 bits
  SET = 0, FnV = 0
  EA = 0, S1PTW = 0
  FSC = 0x06: level 2 translation fault
Data abort info:
  ISV = 0, ISS = 0x00000006, ISS2 = 0x00000000
  CM = 0, WnR = 0, TnD = 0, TagAccess = 0
  GCS = 0, Overlay = 0, DirtyBit = 0, Xs = 0
user pgtable: 4k pages, 48-bit VAs, pgdp=00000000858f3000
[0000000000000000] pgd=0800000085918003, p4d=0800000085918003, pud=0800000085431003, pmd=0000000000000000
Internal error: Oops: 0000000096000006 [#1] PREEMPT SMP
Modules linked in: tidss(-) display_connector lontium_lt8912b tc358768 panel_lvds panel_simple drm_dma_helper drm_kms_helper drm drm_panel_orientation_quirks
CPU: 3 PID: 462 Comm: rmmod Tainted: G        W          6.5.0-rc2+ #2
Hardware name: Toradex Verdin AM62 on Verdin Development Board (DT)
pstate: 80000005 (Nzcv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
pc : drm_connector_cleanup+0x78/0x2d4 [drm]
lr : lt8912_bridge_detach+0x54/0x6c [lontium_lt8912b]
sp : ffff800082ed3a90
x29: ffff800082ed3a90 x28: ffff0000040c1940 x27: 0000000000000000
x26: 0000000000000000 x25: dead000000000122 x24: dead000000000122
x23: dead000000000100 x22: ffff000003fb6388 x21: 0000000000000000
x20: 0000000000000000 x19: ffff000003fb6260 x18: fffffffffffe56e8
x17: 0000000000000000 x16: 0010000000000000 x15: 0000000000000038
x14: 0000000000000000 x13: ffff800081914b48 x12: 000000000000040e
x11: 000000000000015a x10: ffff80008196ebb8 x9 : ffff800081914b48
x8 : 00000000ffffefff x7 : ffff0000040c1940 x6 : ffff80007aa649d0
x5 : 0000000000000000 x4 : 0000000000000001 x3 : ffff80008159e008
x2 : 0000000000000000 x1 : 0000000000000000 x0 : 0000000000000000
Call trace:
 drm_connector_cleanup+0x78/0x2d4 [drm]
 lt8912_bridge_detach+0x54/0x6c [lontium_lt8912b]
 drm_bridge_detach+0x44/0x84 [drm]
 drm_encoder_cleanup+0x40/0xb8 [drm]
 drmm_encoder_alloc_release+0x1c/0x30 [drm]
 drm_managed_release+0xac/0x148 [drm]
 drm_dev_put.part.0+0x88/0xb8 [drm]
 devm_drm_dev_init_release+0x14/0x24 [drm]
 devm_action_release+0x14/0x20
 release_nodes+0x5c/0x90
 devres_release_all+0x8c/0xe0
 device_unbind_cleanup+0x18/0x68
 device_release_driver_internal+0x208/0x23c
 driver_detach+0x4c/0x94
 bus_remove_driver+0x70/0xf4
 driver_unregister+0x30/0x60
 platform_driver_unregister+0x14/0x20
 tidss_platform_driver_exit+0x18/0xb2c [tidss]
 __arm64_sys_delete_module+0x1a0/0x2b4
 invoke_syscall+0x48/0x110
 el0_svc_common.constprop.0+0x60/0x10c
 do_el0_svc_compat+0x1c/0x40
 el0_svc_compat+0x40/0xac
 el0t_32_sync_handler+0xb0/0x138
 el0t_32_sync+0x194/0x198
Code: 9104a276 f2fbd5b7 aa0203e1 91008af8 (f85c0420)

Fixes: 30e2ae943c26 ("drm/bridge: Introduce LT8912B DSI to HDMI bridge")
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Reviewed-by: Robert Foss <rfoss@kernel.org>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230804-lt8912b-v1-2-c542692c6a2f@ideasonboard.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/bridge: lt8912b: Fix bridge_detach

[ Upstream commit 941882a0e96d245f38116e940912b404b6a93c6f ]

The driver calls lt8912_bridge_detach() from its lt8912_remove()
function. As the DRM core detaches bridges automatically, this leads to
calling lt8912_bridge_detach() twice. The code probably has tried to
manage the double-call with the 'is_attached' variable, but the driver
never sets the variable to false, so its of no help.

Fix the issue by dropping the call to lt8912_bridge_detach() from
lt8912_remove(), as the DRM core will handle the detach call for us,
and also drop the useless is_attached field.

Fixes: 30e2ae943c26 ("drm/bridge: Introduce LT8912B DSI to HDMI bridge")
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Reviewed-by: Robert Foss <rfoss@kernel.org>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230804-lt8912b-v1-1-c542692c6a2f@ideasonboard.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm: bridge: it66121: Fix invalid connector dereference

[ Upstream commit d0375f6858c4ff7244b62b02eb5e93428e1916cd ]

Fix the NULL pointer dereference when no monitor is connected, and the
sound card is opened from userspace.

Instead return an empty buffer (of zeroes) as the EDID information to
the sound framework if there is no connector attached.

Fixes: e0fd83dbe924 ("drm: bridge: it66121: Add audio support")
Reported-by: Nishanth Menon <nm@ti.com>
Closes: https://lore.kernel.org/all/20230825105849.crhon42qndxqif4i@gondola/
Reviewed-by: Helen Koike <helen.koike@collabora.com>
Signed-off-by: Jai Luthra <j-luthra@ti.com>
Tested-by: Nishanth Menon <nm@ti.com>
Reviewed-by: Aradhya Bhatia <a-bhatia1@ti.com>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230901-it66121_edid-v2-1-aa59605336b9@ti.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/radeon: Remove the references of radeon_gem_ pread & pwrite ioctls

[ Upstream commit 52548038496fd58b762067b946f943c9bbcbd01e ]

Removing the functions of pread & pwrite & IOCTL defines, as their
existence allows an authorized client to spam the system logs.

Fixes: db996e64b293 ("drm/radeon: Fix ENOSYS with better fitting error codes in radeon_gem.c")
Suggested-by: Christian König <christian.koenig@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/radeon: possible buffer overflow

[ Upstream commit dd05484f99d16715a88eedfca363828ef9a4c2d4 ]

Buffer 'afmt_status' of size 6 could overflow, since index 'afmt_idx' is
checked after access.

Fixes: 5cc4e5fc293b ("drm/radeon: Cleanup HDMI audio interrupt handling for evergreen")
Co-developed-by: Ivanov Mikhail <ivanov.mikhail1@huawei-partners.com>
Signed-off-by: Konstantin Meskhidze <konstantin.meskhidze@huawei.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/rockchip: vop2: Add missing call to crtc reset helper

[ Upstream commit 4d49d87b3606369c6e29b9d051892ee1a6fc4e75 ]

Add missing call to crtc reset helper to properly vblank reset.

Also move vop2_crtc_reset and call vop2_crtc_destroy_state to simplify
and remove duplicated code.

Fixes: 604be85547ce ("drm/rockchip: Add VOP2 driver")
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20230621223311.2239547-6-jonas@kwiboo.se
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/rockchip: vop2: Don't crash for invalid duplicate_state

[ Upstream commit 342f7e4967d02b0ec263b15916304fc54841b608 ]

It's possible for users to try to duplicate the CRTC state even when the
state doesn't exist. drm_atomic_helper_crtc_duplicate_state() (and other
users of __drm_atomic_helper_crtc_duplicate_state()) already guard this
with a WARN_ON() instead of crashing, so let's do that here too.

Fixes: 604be85547ce ("drm/rockchip: Add VOP2 driver")
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20230621223311.2239547-5-jonas@kwiboo.se
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/rockchip: vop: Fix call to crtc reset helper

[ Upstream commit 5aacd290837828c089a83ac9795c74c4c9e2c923 ]

Allocation of crtc_state may fail in vop_crtc_reset, causing an invalid
pointer to be passed to __drm_atomic_helper_crtc_reset.

Fix this by adding a NULL check of crtc_state, similar to other drivers.

Fixes: 01e2eaf40c9d ("drm/rockchip: Convert to using __drm_atomic_helper_crtc_reset() for reset.")
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20230621223311.2239547-4-jonas@kwiboo.se
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/rockchip: vop: Fix reset of state in duplicate state crtc funcs

[ Upstream commit 13fc28804bf10ca0b7bce3efbba95c534836d7ca ]

struct rockchip_crtc_state members such as output_type, output_bpc and
enable_afbc is always reset to zero in the atomic_duplicate_state crtc
funcs.

Fix this by using kmemdup on the subclass rockchip_crtc_state struct.

Fixes: 4e257d9eee23 ("drm/rockchip: get rid of rockchip_drm_crtc_mode_config")
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20230621223311.2239547-2-jonas@kwiboo.se
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/loongson: Fix error handling in lsdc_pixel_pll_setup()

[ Upstream commit 5976a28b344ecb6810882a01b76a320cac21d307 ]

There are two problems in lsdc_pixel_pll_setup()
1. If kzalloc() fails then call iounmap() to release the resources.
2. Both kzalloc and ioremap does not return error pointers on failure, so
   using IS_ERR_OR_NULL() checks is a bit confusing and not very right,
   fix this by changing those to NULL checks instead.

Fixes: f39db26c5428 ("drm: Add kms driver for loongson display controller")
Signed-off-by: Harshit Mogalapalli <harshit.m.mogalapalli@oracle.com>
Reviewed-by: Sui Jingfeng <suijingfeng@loongson.cn>
Signed-off-by: Sui Jingfeng <sui.jingfeng@linux.dev>
Link: https://patchwork.freedesktop.org/patch/msgid/20230720123950.543082-1-harshit.m.mogalapalli@oracle.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm: bridge: samsung-dsim: Fix waiting for empty cmd transfer FIFO on older Exynos

[ Upstream commit 15f389da11257b806da75a070cfa41ca0cc15aae ]

Samsung DSIM used in older Exynos SoCs (like Exynos 4210, 4x12, 3250)
doesn't report empty level of packer header FIFO. In case of those SoCs,
use the old way of waiting for empty command tranfsfer FIFO, removed
recently by commit 14806c641582 ("drm: bridge: samsung-dsim: Drain command transfer FIFO before transfer").

Fixes: 14806c641582 ("drm: bridge: samsung-dsim: Drain command transfer FIFO before transfer")
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230809145641.3213210-1-m.szyprowski@samsung.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm: bridge: for GENERIC_PHY_MIPI_DPHY also select GENERIC_PHY

[ Upstream commit 96413b355a49fd684430a230479bd231d977894f ]

Three DRM bridge drivers select GENERIC_PHY_MIPI_DPHY when GENERIC_PHY
might not be set.  This causes Kconfig warnings and a build error.

WARNING: unmet direct dependencies detected for GENERIC_PHY_MIPI_DPHY
  Depends on [n]: GENERIC_PHY [=n]
  Selected by [y]:
  - DRM_NWL_MIPI_DSI [=y] && DRM_BRIDGE [=y] && DRM [=y] && COMMON_CLK [=y] && OF [=y] && HAS_IOMEM [=y]
  - DRM_SAMSUNG_DSIM [=y] && DRM [=y] && DRM_BRIDGE [=y] && COMMON_CLK [=y] && OF [=y] && HAS_IOMEM [=y]

(drm/bridge/cadence/Kconfig was found by inspection.)

aarch64-linux-ld: drivers/gpu/drm/bridge/samsung-dsim.o: in function `samsung_dsim_set_phy_ctrl':
drivers/gpu/drm/bridge/samsung-dsim.c:731: undefined reference to `phy_mipi_dphy_get_default_config_for_hsclk'

Prevent these warnings and build error by also selecting GENERIC_PHY
whenever selecting GENERIC_PHY_MIPI_DPHY.

Fixes: fced5a364dee ("drm/bridge: cdns: Convert to phy framework")
Fixes: 44cfc6233447 ("drm/bridge: Add NWL MIPI DSI host controller support")
Fixes: 171b3b1e0f8b ("drm: bridge: samsung-dsim: Select GENERIC_PHY_MIPI_DPHY")
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Reported-by: Aleksandr Nogikh <nogikh@google.com>
Link: lore.kernel.org/r/20230803144227.2187749-1-nogikh@google.com
Cc: Adam Ford <aford173@gmail.com>
Cc: Maxime Ripard <maxime.ripard@bootlin.com>
Cc: Guido Günther <agx@sigxcpu.org>
Cc: Robert Chiras <robert.chiras@nxp.com>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: Neil Armstrong <neil.armstrong@linaro.org>
Cc: Andrzej Hajda <andrzej.hajda@intel.com>
Cc: Robert Foss <rfoss@kernel.org>
Cc: David Airlie <airlied@gmail.com>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: dri-devel@lists.freedesktop.org
Reviewed-by: Adam Ford <aford173@gmail.com>
Tested-by: Aleksandr Nogikh <nogikh@google.com>
Reviewed-by: Guido Günther <agx@sigxcpu.org>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230804030140.21395-1-rdunlap@infradead.org
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm: bridge: samsung-dsim: Initialize ULPS EXIT for i.MX8M DSIM

[ Upstream commit 192948f6a923bedf461b4aa09e70a25cfb8a6041 ]

The ULPS EXIT is initialized to 0xaf in downstream BSP as well as older
revisions of this patchset, in newer revisions of the DSIM patchset it
was left out and set to 0. Fix it.

Fixes: 4d562c70c4dc ("drm: bridge: samsung-dsim: Add i.MX8M Mini/Nano support")
Signed-off-by: Marek Vasut <marex@denx.de>
Reviewed-by: Robert Foss <rfoss@kernel.org>
Signed-off-by: Robert Foss <rfoss@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230709134827.449185-1-marex@denx.de
Signed-off-by: Sasha Levin <sashal@kernel.org>

spi: omap2-mcspi: Fix hardcoded reference clock

[ Upstream commit 2d9f4877988f64f0f336983de65c365b6a7debfb ]

A hardcoded reference clock of 48 MHz is used to calculate the
clock divisor values, but the reference clock frequency can be
different across devices and can be configured which can cause
a mismatch between the reported frequency and actual SPI clock
frequency observed. Fix this by fetching the clock rate from
the clock provider and falling back to hardcoded reference only
if the clock is not supplied.

Fixes: 2cd7d393f461 ("arm64: dts: ti: k3-am654: Add McSPI DT nodes")

Signed-off-by: Vaishnav Achath <vaishnav.a@ti.com>
Reviewed-by: Dhruva Gole <d-gole@ti.com>
Link: https://lore.kernel.org/r/20230926113812.30692-1-vaishnav.a@ti.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

spi: omap2-mcspi: switch to use modern name

[ Upstream commit ee0f793cc1881225dee04216f1080201603b46bb ]

Change legacy name master/slave to modern name host/target or controller.

No functional changed.

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Link: https://lore.kernel.org/r/20230823033003.3407403-22-yangyingliang@huawei.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Stable-dep-of: 2d9f4877988f ("spi: omap2-mcspi: Fix hardcoded reference clock")
Signed-off-by: Sasha Levin <sashal@kernel.org>

platform/chrome: cros_ec_lpc: Separate host command and irq disable

[ Upstream commit 47ea0ddb1f5604ba3496baa19110aec6a3151f2e ]

Both cros host command and irq disable were moved to suspend
prepare stage from late suspend recently. This is causing EC
to report MKBP event timeouts during suspend stress testing.
When the MKBP event timeouts happen during suspend, subsequent
wakeup of AP by EC using MKBP doesn't happen properly. Move the
irq disabling part back to late suspend stage which is a general
suggestion from the suspend kernel documentaiton to do irq
disable as late as possible.

Fixes: 4b9abbc132b8 ("platform/chrome: cros_ec_lpc: Move host command to prepare/complete")
Signed-off-by: Lalith Rajendran <lalithkraj@chromium.org>
Link: https://lore.kernel.org/r/20231027160221.v4.1.I1725c3ed27eb7cd9836904e49e8bfa9fb0200a97@changeid
Signed-off-by: Tzung-Bi Shih <tzungbi@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

hte: tegra: Fix missing error code in tegra_hte_test_probe()

[ Upstream commit b7c3ca3553d1de5e86c85636828e186d30cd0628 ]

The value of 'ret' is zero when of_hte_req_count() fails to get number
of entitties to timestamp. And returning success(zero) on this failure
path is incorrect.

Fixes: 9a75a7cd03c9 ("hte: Add Tegra HTE test driver")
Signed-off-by: Harshit Mogalapalli <harshit.m.mogalapalli@oracle.com>
Reviewed-by: Dipen Patel <dipenp@nvidia.com>
Signed-off-by: Dipen Patel <dipenp@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

hwmon: (sch5627) Disallow write access if virtual registers are locked

[ Upstream commit 7da8a635436029957c5350da3acf51d78ed64071 ]

When the lock bit inside SCH5627_REG_CTRL is set, then the virtual
registers become read-only until the next power cycle.
Disallow write access to those registers in such a case.

Tested on a Fujitsu Esprimo P720.

Fixes: aa9f833dfc12 ("hwmon: (sch5627) Add pwmX_auto_channels_temp support")
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://lore.kernel.org/r/20230907052639.16491-3-W_Armin@gmx.de
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

hwmon: (sch5627) Use bit macros when accessing the control register

[ Upstream commit 7f0b28e0653f36b51542d25dd54ed312c397ecfc ]

Use bit macros then accessing SCH5627_REG_CTRL, so that people
do not need to look at the datasheet to find out what each bit
does.

Tested on a Fujitsu Esprimo P720.

Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://lore.kernel.org/r/20230907052639.16491-2-W_Armin@gmx.de
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Stable-dep-of: 7da8a6354360 ("hwmon: (sch5627) Disallow write access if virtual registers are locked")
Signed-off-by: Sasha Levin <sashal@kernel.org>

hwmon: (pmbus/mp2975) Move PGOOD fix

[ Upstream commit 9da2901c47332b030ea4d2a2302bc7c0b83fc67c ]

The PGOOD fix was intended for MP2973 & MP2971 & not for MP2975.

Fixes: acda945afb46 ("hwmon: (pmbus/mp2975) Fix PGOOD in READ_STATUS_WORD")
Signed-off-by: Naresh Solanki <naresh.solanki@9elements.com>
Link: https://lore.kernel.org/r/20231027103352.918895-1-naresh.solanki@9elements.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

Revert "hwmon: (sch56xx-common) Add automatic module loading on supported devices"

[ Upstream commit d621a46d05107f4e510383d6a38f2160c62d28f7 ]

This reverts commit 393935baa45e5ccb9603cf7f9f020ed1bc0915f7.

As reported by Ian Nartowicz, this and the next patch
result in a failure to load the driver on Celsius W280.
While the alternative would be to add the board to the DMI
override table, it is quite likely that other systems are
also affected. Revert the offending patches to avoid future
problems.

Fixes: 393935baa45e ("hwmon: (sch56xx-common) Add automatic module loading on supported devices")
Reported-by: Ian Nartowicz <deadbeef@nartowicz.co.uk>
Closes: https://lore.kernel.org/linux-hwmon/20231025192239.3c5389ae@debian.org/T/#t
Cc: Armin Wolf <W_Armin@gmx.de>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

Revert "hwmon: (sch56xx-common) Add DMI override table"

[ Upstream commit 28da9dee3594423534f3ea1e1f61e6bb2d2fa651 ]

This reverts commit fd2d53c367ae9983c2100ac733a834e0c79d7537.

As reported by Ian Nartowicz, this and the preceding patch
result in a failure to load the driver on Celsius W280.
While the alternative would be to add the board to the DMI
override table, it is quite likely that other systems are
also affected. Revert the offending patches to avoid future
problems.

Fixes: fd2d53c367ae ("hwmon: (sch56xx-common) Add DMI override table")
Reported-by: Ian Nartowicz <deadbeef@nartowicz.co.uk>
Closes: https://lore.kernel.org/linux-hwmon/20231025192239.3c5389ae@debian.org/T/#t
Cc: Armin Wolf <W_Armin@gmx.de>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

hwmon: (coretemp) Fix potentially truncated sysfs attribute name

[ Upstream commit bbfff736d30e5283ad09e748caff979d75ddef7f ]

When build with W=1 and "-Werror=format-truncation", below error is
observed in coretemp driver,

   drivers/hwmon/coretemp.c: In function 'create_core_data':
>> drivers/hwmon/coretemp.c:393:34: error: '%s' directive output may be truncated writing likely 5 or more bytes into a region of size between 3 and 13 [-Werror=format-truncation=]
     393 |                          "temp%d_%s", attr_no, suffixes[i]);
         |                                  ^~
   drivers/hwmon/coretemp.c:393:26: note: assuming directive output of 5 bytes
     393 |                          "temp%d_%s", attr_no, suffixes[i]);
         |                          ^~~~~~~~~~~
   drivers/hwmon/coretemp.c:392:17: note: 'snprintf' output 7 or more bytes (assuming 22) into a destination of size 19
     392 |                 snprintf(tdata->attr_name[i], CORETEMP_NAME_LENGTH,
         |                 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
     393 |                          "temp%d_%s", attr_no, suffixes[i]);
         |                          ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   cc1: all warnings being treated as errors

Given that
1. '%d' could take 10 charactors,
2. '%s' could take 10 charactors ("crit_alarm"),
3. "temp", "_" and the NULL terminator take 6 charactors,
fix the problem by increasing CORETEMP_NAME_LENGTH to 28.

Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Fixes: 7108b80a542b ("hwmon/coretemp: Handle large core ID value")
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202310200443.iD3tUbbK-lkp@intel.com/
Link: https://lore.kernel.org/r/20231025122316.836400-1-rui.zhang@intel.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

hwmon: (axi-fan-control) Fix possible NULL pointer dereference

[ Upstream commit 2a5b3370a1d9750eca325292e291c8c7cb8cf2e0 ]

axi_fan_control_irq_handler(), dependent on the private
axi_fan_control_data structure, might be called before the hwmon
device is registered. That will cause an "Unable to handle kernel
NULL pointer dereference" error.

Fixes: 8412b410fa5e ("hwmon: Support ADI Fan Control IP")
Signed-off-by: Dragos Bogdan <dragos.bogdan@analog.com>
Signed-off-by: Nuno Sa <nuno.sa@analog.com>
Link: https://lore.kernel.org/r/20231025132100.649499-1-nuno.sa@analog.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

regulator: qcom-rpmh: Fix smps4 regulator for pm8550ve

[ Upstream commit bc00d9f3813a40bc2d854ae0edab14c6b43a3219 ]

The type of the smps4 regulator from pm8550ve is actually FTSMPS525
medium voltage. So fix it accordingly.

Fixes: e6e3776d682d ("regulator: qcom-rpmh: Add support for PM8550 regulators")
Signed-off-by: Abel Vesa <abel.vesa@linaro.org>
Link: https://lore.kernel.org/r/20231024134626.2364426-1-abel.vesa@linaro.org
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

platform/x86: wmi: Fix opening of char device

[ Upstream commit eba9ac7abab91c8f6d351460239108bef5e7a0b6 ]

Since commit fa1f68db6ca7 ("drivers: misc: pass miscdevice pointer via
file private data"), the miscdevice stores a pointer to itself inside
filp->private_data, which means that private_data will not be NULL when
wmi_char_open() is called. This might cause memory corruption should
wmi_char_open() be unable to find its driver, something which can
happen when the associated WMI device is deleted in wmi_free_devices().

Fix the problem by using the miscdevice pointer to retrieve the WMI
device data associated with a char device using container_of(). This
also avoids wmi_char_open() picking a wrong WMI device bound to a
driver with the same name as the original driver.

Fixes: 44b6b7661132 ("platform/x86: wmi: create userspace interface for drivers")
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://lore.kernel.org/r/20231020211005.38216-5-W_Armin@gmx.de
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

platform/x86: wmi: Fix probe failure when failing to register WMI devices

[ Upstream commit ed85891a276edaf7a867de0e9acd0837bc3008f2 ]

When a WMI device besides the first one somehow fails to register,
retval is returned while still containing a negative error code. This
causes the ACPI device fail to probe, leaving behind zombie WMI devices
leading to various errors later.

Handle the single error path separately and return 0 unconditionally
after trying to register all WMI devices to solve the issue. Also
continue to register WMI devices even if some fail to allocate memory.

Fixes: 6ee50aaa9a20 ("platform/x86: wmi: Instantiate all devices before adding them")
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://lore.kernel.org/r/20231020211005.38216-4-W_Armin@gmx.de
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: mediatek: fix double free in mtk_clk_register_pllfh()

[ Upstream commit bd54ccc0f147019dac38e7841876a7415459b875 ]

The mtk_clk_register_pll_ops() currently frees the "pll" parameter.
The function has two callers, mtk_clk_register_pll() and
mtk_clk_register_pllfh().  The first one, the _pll() function relies on
the free, but for the second _pllfh() function it causes a double free
bug.

Really the frees should be done in the caller because that's where
the allocation is.

Fixes: d7964de8a8ea ("clk: mediatek: Add new clock driver to handle FHCTL hardware")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Link: https://lore.kernel.org/r/cd7fa365-28cc-4c34-ac64-6da57c98baa6@moroto.mountain
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: qcom: ipq5332: drop the CLK_SET_RATE_PARENT flag from GPLL clocks

[ Upstream commit 5635ef0bd1052420bc659a00be6fd0c60cec5cb9 ]

GPLL clock rates are fixed and shouldn't be scaled based on the
request from dependent clocks. Doing so will result in the unexpected
behaviour. So drop the CLK_SET_RATE_PARENT flag from the GPLL clocks.

Fixes: 3d89d52970fd ("clk: qcom: add Global Clock controller (GCC) driver for IPQ5332 SoC")
Signed-off-by: Kathiravan Thirumoorthy <quic_kathirav@quicinc.com>
Reviewed-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Link: https://lore.kernel.org/r/20230913-gpll_cleanup-v2-5-c8ceb1a37680@quicinc.com
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: qcom: ipq9574: drop the CLK_SET_RATE_PARENT flag from GPLL clocks

[ Upstream commit 99a8f8764b70158a712992640a6be46a8fd79d15 ]

GPLL clock rates are fixed and shouldn't be scaled based on the request
from dependent clocks. Doing so will result in the unexpected behaviour.
So drop the CLK_SET_RATE_PARENT flag from the GPLL clocks.

----
Changes in V2:
- No changes

Fixes: d75b82cff488 ("clk: qcom: Add Global Clock Controller driver for IPQ9574")
Signed-off-by: Kathiravan Thirumoorthy <quic_kathirav@quicinc.com>
Reviewed-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Link: https://lore.kernel.org/r/20230913-gpll_cleanup-v2-4-c8ceb1a37680@quicinc.com
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: qcom: ipq5018: drop the CLK_SET_RATE_PARENT flag from GPLL clocks

[ Upstream commit 01a5e4c6731ab6b4b74822661d296f8893fc1230 ]

GPLL clock rates are fixed and shouldn't be scaled based on the
request from dependent clocks. Doing so will result in the unexpected
behaviour. So drop the CLK_SET_RATE_PARENT flag from the GPLL clocks.

----
Changes in V2:
- New patch

Fixes: e3fdbef1bab8 ("clk: qcom: Add Global Clock controller (GCC) driver for IPQ5018")
Signed-off-by: Kathiravan Thirumoorthy <quic_kathirav@quicinc.com>
Reviewed-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Link: https://lore.kernel.org/r/20230913-gpll_cleanup-v2-3-c8ceb1a37680@quicinc.com
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: qcom: apss-ipq-pll: Fix 'l' value for ipq5332_pll_config

[ Upstream commit 5b7a4d3d2b33398330aef69e0ff5656273483587 ]

The earlier 'l' value of 0x3e is for 1.5GHz. Not all SKUs support
this frequency. Hence set it to 0x2d to get 1.1GHz which is
supported in all SKUs.

The frequency can still increase above this initial configuration
made here when the cpufreq driver picks a different OPP.

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Fixes: c7ef7fbb1ccf ("clk: qcom: apss-ipq-pll: add support for IPQ5332")
Signed-off-by: Kathiravan T <quic_kathirav@quicinc.com>
Signed-off-by: Varadarajan Narayanan <quic_varada@quicinc.com>
Link: https://lore.kernel.org/r/00e6be6cb9cee56628123a64ade118d0a752018b.1697781921.git.quic_varada@quicinc.com
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: qcom: apss-ipq-pll: Use stromer plus ops for stromer plus pll

[ Upstream commit 267e29198436a8cb6770213471f72502c895096a ]

The set rate and determine rate operations are different between
Stromer and Stromer Plus PLLs. Since the programming sequence is
different, the PLLs dont get configured properly and random,
inexplicable crash/freeze is seen. Hence, use stromer plus ops
for ipq_pll_stromer_plus.

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Acked-by: Stephen Boyd <sboyd@kernel.org>
Fixes: c7ef7fbb1ccf ("clk: qcom: apss-ipq-pll: add support for IPQ5332")
Signed-off-by: Kathiravan T <quic_kathirav@quicinc.com>
Signed-off-by: Varadarajan Narayanan <quic_varada@quicinc.com>
Link: https://lore.kernel.org/r/c86ecaa23dc4f39650bcf4a3bd54a617a932e4fd.1697781921.git.quic_varada@quicinc.com
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: qcom: clk-alpha-pll: introduce stromer plus ops

[ Upstream commit 84da48921a97cee3dd1391659e93ee01d122b78b ]

Stromer plus APSS PLL does not support dynamic frequency scaling.
To switch between frequencies, we have to shut down the PLL,
configure the L and ALPHA values and turn on again. So introduce the
separate set of ops for Stromer Plus PLL.

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Kathiravan T <quic_kathirav@quicinc.com>
Signed-off-by: Varadarajan Narayanan <quic_varada@quicinc.com>
Link: https://lore.kernel.org/r/2affa6c63ff0c4342230623a7d4eef02ec7c02d4.1697781921.git.quic_varada@quicinc.com
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Stable-dep-of: 267e29198436 ("clk: qcom: apss-ipq-pll: Use stromer plus ops for stromer plus pll")
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: qcom: config IPQ_APSS_6018 should depend on QCOM_SMEM

[ Upstream commit 6a15647d0adc686226045e8046369f34d6ab03ed ]

The config IPQ_APSS_6018 should depend on QCOM_SMEM, to
avoid the following error reported by 'kernel test robot'

loongarch64-linux-ld: drivers/clk/qcom/apss-ipq6018.o: in function `apss_ipq6018_probe':
>> apss-ipq6018.c:(.text+0xd0): undefined reference to `qcom_smem_get_soc_id'

Fixes: 5e77b4ef1b19 ("clk: qcom: Add ipq6018 apss clock controller")
Reported-by: kernel test robot <yujie.liu@intel.com>
Closes: https://lore.kernel.org/r/202310181650.g8THtfsm-lkp@intel.com/
Signed-off-by: Varadarajan Narayanan <quic_varada@quicinc.com>
Link: https://lore.kernel.org/r/f4c4d65a7cb71e807d6d472c63c7718408c8f5f0.1697781921.git.quic_varada@quicinc.com
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: mediatek: clk-mt2701: Add check for mtk_alloc_clk_data

[ Upstream commit 0d6e24b422a2166a9297a8286ff2e6ab9a5e8cd3 ]

Add the check for the return value of mtk_alloc_clk_data() in order to
avoid NULL pointer dereference.

Fixes: e9862118272a ("clk: mediatek: Add MT2701 clock support")
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Link: https://lore.kernel.org/r/20230901024658.23405-1-jiasheng@iscas.ac.cn
Reviewed-by: Markus Schneider-Pargmann <msp@baylibre.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: mediatek: clk-mt7629: Add check for mtk_alloc_clk_data

[ Upstream commit 2befa515c1bb6cdd33c262b909d93d1973a219aa ]

Add the check for the return value of mtk_alloc_clk_data() in order to
avoid NULL pointer dereference.

Fixes: 3b5e748615e7 ("clk: mediatek: add clock support for MT7629 SoC")
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Link: https://lore.kernel.org/r/20230912093407.21505-5-jiasheng@iscas.ac.cn
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: mediatek: clk-mt7629-eth: Add check for mtk_alloc_clk_data

[ Upstream commit 0884393c63cc9a1772f7121a6645ba7bd76feeb9 ]

Add the check for the return value of mtk_alloc_clk_data() in order to
avoid NULL pointer dereference.

Fixes: 3b5e748615e7 ("clk: mediatek: add clock support for MT7629 SoC")
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Link: https://lore.kernel.org/r/20230912093407.21505-4-jiasheng@iscas.ac.cn
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: mediatek: clk-mt6797: Add check for mtk_alloc_clk_data

[ Upstream commit 606f6366a35a3329545e38129804d65ef26ed7d2 ]

Add the check for the return value of mtk_alloc_clk_data() in order to
avoid NULL pointer dereference.

Fixes: 96596aa06628 ("clk: mediatek: add clk support for MT6797")
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Link: https://lore.kernel.org/r/20230912093407.21505-3-jiasheng@iscas.ac.cn
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: mediatek: clk-mt6779: Add check for mtk_alloc_clk_data

[ Upstream commit 1f57f78fbacf630430bf954e5a84caafdfea30c0 ]

Add the check for the return value of mtk_alloc_clk_data() in order to
avoid NULL pointer dereference.

Fixes: 710774e04861 ("clk: mediatek: Add MT6779 clock support")
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Link: https://lore.kernel.org/r/20230912093407.21505-2-jiasheng@iscas.ac.cn
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: mediatek: clk-mt6765: Add check for mtk_alloc_clk_data

[ Upstream commit b82681042724924ae3ba0f2f2eeec217fa31e830 ]

Add the check for the return value of mtk_alloc_clk_data() in order to
avoid NULL pointer dereference.

Fixes: 1aca9939bf72 ("clk: mediatek: Add MT6765 clock support")
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Link: https://lore.kernel.org/r/20230912093407.21505-1-jiasheng@iscas.ac.cn
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: npcm7xx: Fix incorrect kfree

[ Upstream commit bbc5080bef4a245106aa8e8d424ba8847ca7c0ca ]

The corresponding allocation is:

> npcm7xx_clk_data = kzalloc(struct_size(npcm7xx_clk_data, hws,
>       NPCM7XX_NUM_CLOCKS), GFP_KERNEL);

... so, kfree should be applied to npcm7xx_clk_data, not
npcm7xx_clk_data->hws.

Fixes: fcfd14369856 ("clk: npcm7xx: add clock controller")
Signed-off-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
Link: https://lore.kernel.org/r/20230923133127.1815621-1-j.neuschaefer@gmx.net
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: ti: fix double free in of_ti_divider_clk_setup()

[ Upstream commit 7af5b9eadd64c9e02a71f97c45bcdf3b64841f6b ]

The "div" pointer is freed in _register_divider() and again in
of_ti_divider_clk_setup().  Delete the free in _register_divider()

Fixes: fbbc18591585 ("clk: ti: divider: cleanup _register_divider and ti_clk_get_div_table")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Link: https://lore.kernel.org/r/6d36eeec-6c8a-4f11-a579-aa3cd7c38749@moroto.mountain
Reviewed-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: keystone: pll: fix a couple NULL vs IS_ERR() checks

[ Upstream commit a5d14f8b551eb1551c10053653ee8e27f19672fa ]

The clk_register_divider() and clk_register_mux() functions returns
error pointers on error but this code checks for NULL.  Fix that.

Fixes: b9e0d40c0d83 ("clk: keystone: add Keystone PLL clock driver")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Link: https://lore.kernel.org/r/d9da4c97-0da9-499f-9a21-1f8e3f148dc1@moroto.mountain
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: ralink: mtmips: quiet unused variable warning

[ Upstream commit 619102313466eaf8a6ac188e711f5df749dac6d4 ]

When CONFIG_OF is disabled then the matching table is not referenced and
the following warning appears:

drivers/clk/ralink/clk-mtmips.c:821:34: warning: unused variable 'mtmips_of_match' [-Wunused-const-variable]
821 |   static const struct of_device_id mtmips_of_match[] = {
    |                          ^

There are two match tables in the driver: one for the clock driver and the
other for the reset driver. The only difference between them is that the
clock driver uses 'data' and does not have 'ralink,rt2880-reset' compatible.
Both just can be merged into a single one just by adding the compatible
'ralink,rt2880-reset' entry to 'mtmips_of_match[]', which will allow it to
be used for 'mtmips_clk_driver' (which doesn't use the data) as well as for
'mtmips_clk_init()' (which doesn't need get called for 'ralink,rt2880-reset').

Doing in this way ensures that 'CONFIG_OF' is not disabled anymore so the
above warning disapears.

Fixes: 6f3b15586eef ("clk: ralink: add clock and reset driver for MTMIPS SoCs")
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202307242310.CdOnd2py-lkp@intel.com/
Suggested-by: Arnd Bergmann <arnd@kernel.org>
Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Link: https://lore.kernel.org/r/20230827023932.501102-1-sergio.paracuellos@gmail.com
Reviewed-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

spi: nxp-fspi: use the correct ioremap function

[ Upstream commit c3aa5cb264a38ae9bbcce32abca4c155af0456df ]

AHB memory as MMIO should be mapped with ioremap rather than ioremap_wc,
which should have been used initially just to handle unaligned access as
a workaround.

Fixes: d166a73503ef ("spi: fspi: dynamically alloc AHB memory")
Signed-off-by: Han Xu <han.xu@nxp.com>
Link: https://lore.kernel.org/r/20231010201524.2021340-1-han.xu@nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: linux/clk-provider.h: fix kernel-doc warnings and typos

[ Upstream commit 84aefafe6b294041b7fa0757414c4a29c1bdeea2 ]

Fix spelling of "Structure".

Fix multiple kernel-doc warnings:

clk-provider.h:269: warning: Function parameter or member 'recalc_rate' not described in 'clk_ops'
clk-provider.h:468: warning: Function parameter or member 'parent_data' not described in 'clk_hw_register_fixed_rate_with_accuracy_parent_data'
clk-provider.h:468: warning: Excess function parameter 'parent_name' description in 'clk_hw_register_fixed_rate_with_accuracy_parent_data'
clk-provider.h:482: warning: Function parameter or member 'parent_data' not described in 'clk_hw_register_fixed_rate_parent_accuracy'
clk-provider.h:482: warning: Excess function parameter 'parent_name' description in 'clk_hw_register_fixed_rate_parent_accuracy'
clk-provider.h:687: warning: Function parameter or member 'flags' not described in 'clk_divider'
clk-provider.h:1164: warning: Function parameter or member 'flags' not described in 'clk_fractional_divider'
clk-provider.h:1164: warning: Function parameter or member 'approximation' not described in 'clk_fractional_divider'
clk-provider.h:1213: warning: Function parameter or member 'flags' not described in 'clk_multiplier'

Fixes: 9fba738a53dd ("clk: add duty cycle support")
Fixes: b2476490ef11 ("clk: introduce the common clock framework")
Fixes: 2d34f09e79c9 ("clk: fixed-rate: Add support for specifying parents via DT/pointers")
Fixes: f5290d8e4f0c ("clk: asm9260: use parent index to link the reference clock")
Fixes: 9d9f78ed9af0 ("clk: basic clock hardware types")
Fixes: e2d0e90fae82 ("clk: new basic clk type for fractional divider")
Fixes: f2e0a53271a4 ("clk: Add a basic multiplier clock")

Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Michael Turquette <mturquette@baylibre.com>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: linux-clk@vger.kernel.org
Link: https://lore.kernel.org/r/20230930221428.18463-1-rdunlap@infradead.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: renesas: rzg2l: Fix computation formula

[ Upstream commit a2b23159499efd36b2d63b3c4534075d12ddc97a ]

According to the hardware manual for RZ/G2L
(r01uh0914ej0130-rzg2l-rzg2lc.pdf), the computation formula for PLL rate
is as follows:

    Fout = ((m + k/65536) * Fin) / (p * 2^s)

and k has values in the range [-32768, 32767].  Dividing k by 65536 with
integer arithmetic gives zero all the time, causing slight differences
b/w what has been set vs. what is displayed.  Thus, get rid of this and
decompose the formula before dividing k by 65536.

Fixes: ef3c613ccd68a ("clk: renesas: Add CPG core wrapper for RZ/G2L SoC")
Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/20230929053915.1530607-6-claudiu.beznea@bp.renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: renesas: rzg2l: Use FIELD_GET() for PLL register fields

[ Upstream commit 72977f07b035e488c3f1928832a1616c6cae7278 ]

Use FIELD_GET() for PLL register fields.  This is its purpose.

Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/20230912045157.177966-14-claudiu.beznea.uj@bp.renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Stable-dep-of: a2b23159499e ("clk: renesas: rzg2l: Fix computation formula")
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: renesas: rzg2l: Trust value returned by hardware

[ Upstream commit bf51d3b2d048c312764a55d91d67a85ee5535e31 ]

The onitial value of the CPG_PL2SDHI_DSEL bits 0..1 or 4..6 is 01b.  The
hardware user's manual (r01uh0914ej0130-rzg2l-rzg2lc.pdf) specifies that
setting 0 is prohibited.  Hence rzg2l_cpg_sd_clk_mux_get_parent() should
just read CPG_PL2SDHI_DSEL, trust the value, and return the proper clock
parent index based on the value read.

Fixes: eaff33646f4cb ("clk: renesas: rzg2l: Add SDHI clk mux support")
Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/20230929053915.1530607-5-claudiu.beznea@bp.renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: renesas: rzg2l: Lock around writes to mux register

[ Upstream commit d2692ed490e680a41401cef879adebcfafb4298f ]

The SD MUX output (SD0) is further divided by 4 in G2{L,UL}.  The
divided clock is SD0_DIV4. SD0_DIV4 is registered with
CLK_SET_RATE_PARENT which means a rate request for it is propagated to
the MUX and could reach rzg2l_cpg_sd_clk_mux_set_parent() concurrently
with the users of SD0.
Add proper locking to avoid concurrent accesses on SD MUX set rate
registers.

Fixes: eaff33646f4cb ("clk: renesas: rzg2l: Add SDHI clk mux support")
Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/20230929053915.1530607-4-claudiu.beznea@bp.renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: renesas: rzg2l: Wait for status bit of SD mux before continuing

[ Upstream commit 549f4ae2601f968e2474c6031fb4799468882f64 ]

The hardware user manual for RZ/G2L (r01uh0914ej0130-rzg2l-rzg2lc.pdf,
chapter 7.4.7 Procedure for Switching Clocks by the Dynamic Switching
Frequency Selectors) specifies that we need to check CPG_PL2SDHI_DSEL
for SD clock switching status.

Fixes: eaff33646f4cb ("clk: renesas: rzg2l: Add SDHI clk mux support")
Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/20230929053915.1530607-3-claudiu.beznea@bp.renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: renesas: rcar-gen3: Extend SDnH divider table

[ Upstream commit d5252d9697a3e7007c741e9c103073868955a304 ]

The clock dividers might be used with clock stop bit enabled or not.
Current tables only support recommended values from the datasheet.  This
might result in warnings like below because no valid clock divider is
found. Resulting in a 0 divider.

There are Renesas ARM Trusted Firmware version out there which e.g.
configure 0x201 (shifted logical right by 2: 0x80) and with this match
the added { STPnHCK | 0, 1 }:

https://github.com/renesas-rcar/arm-trusted-firmware/blob/rcar_gen3_v2.3/drivers/renesas/rcar/emmc/emmc_init.c#L108

------------[ cut here ]------------
sd1h: Zero divisor and CLK_DIVIDER_ALLOW_ZERO not set
WARNING: CPU: 1 PID: 1 at drivers/clk/clk-divider.c:141 divider_recalc_rate+0x48/0x70
Modules linked in:
CPU: 1 PID: 1 Comm: swapper/0 Not tainted 6.1.52 #1
Hardware name: Custom board based on r8a7796 (DT)
pstate: 40000005 (nZcv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
pc : divider_recalc_rate+0x48/0x70
...
------------[ cut here ]------------

Fixes: bb6d3fa98a41 ("clk: renesas: rcar-gen3: Switch to new SD clock handling")
Signed-off-by: Dirk Behme <dirk.behme@de.bosch.com>
[wsa: extended the table to 5 entries, added comments, reword commit message a little]
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Tested-by: Dirk Behme <dirk.behme@de.bosch.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/20230928080317.28224-1-wsa+renesas@sang-engineering.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: imx: imx8qxp: Fix elcdif_pll clock

[ Upstream commit 15cee75dacb82ade710d61bfd536011933ef9bf2 ]

Move the elcdif_pll clock initialization before the lcd_clk, since the
elcdif_clk needs to be initialized ahead of lcd_clk, being its parent.
This change fixes issues with the LCD clocks during suspend/resume.

Fixes: babfaa9556d7 ("clk: imx: scu: add more scu clocks")
Suggested-by: Ranjani Vaidyanathan <ranjani.vaidyanathan@nxp.com>
Acked-by: Laurentiu Palcu <laurentiu.palcu@nxp.com>
Signed-off-by: Robert Chiras <robert.chiras@nxp.com>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Abel Vesa <abel.vesa@linaro.org>
Link: https://lore.kernel.org/r/20230912-imx8-clk-v1-v1-2-69a34bcfcae1@nxp.com
Signed-off-by: Abel Vesa <abel.vesa@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: imx: imx8mq: correct error handling path

[ Upstream commit 577ad169966e6e75b10e004389a3f79813e84b5d ]

Avoid memory leak in error handling path. It does not make
much sense for the SoC without clk driver, to make program behavior
correct, let's fix it.

Fixes: b80522040cd3 ("clk: imx: Add clock driver for i.MX8MQ CCM")
Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Dan Carpenter <error27@gmail.com>
Closes: https://lore.kernel.org/r/202309240551.e46NllPa-lkp@intel.com/
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Link: https://lore.kernel.org/r/20231001122618.194498-1-peng.fan@oss.nxp.com
Signed-off-by: Abel Vesa <abel.vesa@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: imx: imx8: Fix an error handling path in imx8_acm_clk_probe()

[ Upstream commit e9a164e367f039629fd5466a79b9f495646e1261 ]

If an error occurs after a successful clk_imx_acm_attach_pm_domains() call,
it must be undone.

Add an explicit error handling path, re-order the code and add the missing
clk_imx_acm_detach_pm_domains() call.

Fixes: d3a0946d7ac9 ("clk: imx: imx8: add audio clock mux driver")
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Abel Vesa <abel.vesa@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>