The same logic is used for clock stop prepare and deprepare, and
having different logs for the two steps helps identify problems.
In addition, when the "NotFinished" bit remains set, the error
handling is not quite right:
a) for the clock stop prepare, the error is handled at the caller
level, and the error is ignored: there's no good reason to prevent the
pm_runtime suspend from happening. Throwing an error that is later
ignored is confusing.
b) for the clock stop deprepare, the error is ignored in bus.c and a
dev_warn() log shown. Throwing an error is also alarming users for no
good reason.
For both cases, demoting the error to dev_dbg() makes more sense.
Link: https://github.com/thesofproject/linux/issues/4619
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Link: https://lore.kernel.org/r/20231013010812.114216-1-yung-chuan.liao@linux.intel.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
In case of the prep descriptor while the channel is already running, the
CCR register value stored into the channel could already have its EN bit
set. This would lead to a bad transfer since, at start transfer time,
enabling the channel while other registers aren't yet properly set.
To avoid this, ensure to mask the CCR_EN bit when storing the ccr value
into the mdma channel structure.
Fixes: a4ffb13c89 ("dmaengine: Add STM32 MDMA driver")
Signed-off-by: Alain Volmat <alain.volmat@foss.st.com>
Signed-off-by: Amelie Delaunay <amelie.delaunay@foss.st.com>
Cc: stable@vger.kernel.org
Tested-by: Alain Volmat <alain.volmat@foss.st.com>
Link: https://lore.kernel.org/r/20231009082450.452877-1-amelie.delaunay@foss.st.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Implement load_unaligned_zeropad() and enable DCACHE_WORD_ACCESS to
speed up string operations in fs/dcache.c and fs/namei.c.
Reviewed-by: Sven Schnelle <svens@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Provide an s390 specific word-at-a-time implementation. Compared to the
generic variant the generated code for has_zero() is slightly
better. However find_zero() is much simpler since it reuses the result
of __fls() aka flogr() and now comes without any conditional branches,
while the generic variant has three of them.
Reviewed-by: Sven Schnelle <svens@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Get rid of __EX_TABLE() macro, rename __EX_TABLE_UA() to __EX_TABLE()
and convert users of old __EX_TABLE() macro so they pass more parameters
to the changed __EX_TABLE() semantics.
Reviewed-by: Sven Schnelle <svens@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Fix virtual vs physical address confusion (which currently are the same).
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Funciton loads_init() does not use loads_offset parameter.
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
With Secure Execution the error reply RX 0x8B now carries
an APFS value indicating why the request has been filtered
by a lower layer of the firmware. So display this value
as a warning line in the s390 debug feature trace.
Signed-off-by: Harald Freudenberger <freude@linux.ibm.com>
Reviewed-by: Holger Dengler <dengler@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
This patch introduces a new AP queue internal attribute
se_bound which reflects the bound state of an APQN within
a Secure Execution environment.
With introduction of Secure Execution guests now an
AP firmware queue needs to be bound to the guest before
usage. This patch introduces a new internal attribute
reflecting this bound state and some glue code to handle
this new field during lifetime of an AP queue device.
Together with that now the zcrypt scheduler considers
the state of the AP queues when a message is about to be
distributed among the existing queues. There is a new
function ap_queue_usable() which returns true only when
all conditions for using this AP queue device are fulfilled.
In details this means: the AP queue needs to be configured,
not checkstopped and within an SE environment it needs
to be bound. So the new function gives and indication
if the AP queue device is ready to serve requests or not.
Signed-off-by: Harald Freudenberger <freude@linux.ibm.com>
Reviewed-by: Holger Dengler <dengler@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
On a state toggle from config off to config on and on the
state toggle from checkstop to not checkstop the queue's
internal states was set but the state machine was not
nudged. This did not care as on the first enqueue of a
request the state machine kick ran.
However, within an Secure Execution guest a queue is
only chosen by the scheduler when it has been bound.
But to bind a queue, it needs to run through the initial
states (reset, enable interrupts, ...). So this is like
a chicken-and-egg problem and the result was in fact
that a queue was unusable after a config off/on toggle.
With some slight rework of the handling of these states
now the new function _ap_queue_init_state() is called
which is the core of the ap_queue_init_state() function
but without locking handling. This has the benefit that
it can be called on all the places where a (re-)init
of the AP queue's state machine is needed.
Fixes: 2d72eaf036 ("s390/ap: implement SE AP bind, unbind and associate")
Signed-off-by: Harald Freudenberger <freude@linux.ibm.com>
Reviewed-by: Holger Dengler <dengler@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
When physical memory is defined under z/VM using DEF STOR CONFIG, there
may be memory holes that are not hotpluggable memory. In such cases,
DCSS mapping could be placed in one of these memory holes. Subsequently,
attempting memory access to such DCSS mapping would result in a kasan
failure because there is no shadow memory mapping for it.
To maintain consistency with cases where DCSS mapping is positioned after
the kernel identity mapping, which is then covered by kasan zero shadow
mapping, handle the scenario above by populating zero shadow mapping
for memory holes where DCSS mapping could potentially be placed.
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Reviewed-by: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
When dma_set_coherent_mask() fails, sch->lock has not been
freed, which is allocated in css_sch_create_locks(), leading
to a memleak.
Fixes: 4520a91a97 ("s390/cio: use dma helpers for setting masks")
Signed-off-by: Dinghao Liu <dinghao.liu@zju.edu.cn>
Message-Id: <20230921071412.13806-1-dinghao.liu@zju.edu.cn>
Link: https://lore.kernel.org/linux-s390/bd38baa8-7b9d-4d89-9422-7e943d626d6e@linux.ibm.com/
Reviewed-by: Halil Pasic <pasic@linux.ibm.com>
Reviewed-by: Peter Oberparleiter <oberpar@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Add DT support, by:
1. Creating a panel bridge from DT, and attaching it to the encoder,
2. Replacing the custom connector with a bridge connector,
3. Obtaining clock configuration based on the compatible value.
Note that for now the driver uses a fixed clock configuration selecting
the bus clock, as the current code to select other clock inputs needs
changes to support any other SoCs than SH7724.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/6185ab76aa300fa402e4f6610b2109665f2d8a1c.1694767209.git.geert+renesas@glider.be
Now the suspend/resume methods no longer need to look at internal driver
state, the dpms and started fields in the shmob_drm_crtc structure can
be removed, as well as the shmob_drm_crtc_dpms() wrapper. After this,
shmob_drm_crtc_atomic_{en,dis}able() became just wrappers around
shmob_drm_crtc_st{art,op}(), so inline the latter.
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/905b9ca72f43d40438c8cb1199cde140eb123204.1694767209.git.geert+renesas@glider.be
Implement atomic mode setting for both the primary and overlay planes.
This involves:
- Moving the primary plane handling code from CRTC mode setting to
plane handling shared by primary and overlay planes,
- Adding basic CRTC and mode config atomic mode setting ops, which
don't do much yet.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/a6b698a1c8c1eaf631c2f928f5888ab5fe7aa4c3.1694767209.git.geert+renesas@glider.be
The DRM core vblank handling mechanism requires drivers to forcefully
turn vblank reporting off when disabling the CRTC, and to restore the
vblank reporting status when enabling the CRTC.
Implement this using the drm_crtc_vblank_{on,off}() helpers.
Note that drm_crtc_vblank_off() must be called at startup to synchronize
the state of the vblank core code with the hardware, which is initially
disabled. This is performed at CRTC creation time, requiring vertical
blank initialization to be moved before creating CRTCs.
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/e5833e5706b7792bfca8e6e56fc154a7c3e0574f.1694767209.git.geert+renesas@glider.be
Turning a CRTC off will prevent a queued page flip from ever completing,
potentially confusing userspace. Wait for queued page flips to complete
before turning the CRTC off to avoid this.
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/c97d5859c43fa36043c61de28d67688ebe345092.1694767209.git.geert+renesas@glider.be
Move legacy interface handling to the connector setup code.
Set up bus_flags and bus_formats in display_info according to the
bus format and panel information from platform data, to make it more
similar with DT-based connector/bridge/panel setup.
This will allow us to use the same LCD interface setup code for both
legacy and DT-based systems.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/912f615eb87c847804a268200ab57c63453c65d4.1694767209.git.geert+renesas@glider.be
Replace the drm_mode_modeinfo field with videomode that includes more
signal polarity flags. This simplifies driver handling of panel modes
and prepares for DT support.
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
[geert: Simplify]
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/4312e56de424d94399c6105e7159317eae86c9d5.1694767209.git.geert+renesas@glider.be
Now that drm_device is embedded in shmob_drm_device, we can use
a container_of()-based helper to get the shmob_drm_device pointer from
the drm_device, instead of using the deprecated drm_device.dev_private
field.
While at it, restore reverse Xmas tree ordering of local variable
declarations.
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Reviewed-by: Sui Jingfeng <suijingfeng@loongson.cn>
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/8ef4be8bffe75efc7f4b66f3732ec357f7d43e0f.1694767209.git.geert+renesas@glider.be
There is no need to call drm_plane_force_disable() from the plane's
.destroy() callback, as the plane should have been disabled already
before. See also commit 3c858a3385 ("drm/plane_helper: don't
disable plane in destroy function") for the generic plane helper case.
After removing this call, shmob_drm_plane_destroy() becomes a simple
wrapper around shmob_drm_plane_destroy(), hence replace it by the
latter.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/8d7a3f250612085fdf4e06d377843e8f874b22d9.1694767209.git.geert+renesas@glider.be
Prepare for DT conversion, where panel probe can be deferred, by
streamlining error propagation and handling:
- Use dev_err_probe() to avoid printing error messages in case of
probe deferral,
- Propagate errors where needed.
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/6dfac76e5b1c7dda3f96801ce83845a4235e2ccd.1694767209.git.geert+renesas@glider.be
Prepare for DT bindings by using more appropriate names for the input
clocks.
Note that all LDDCKR_ICKSEL_* definitions but the one for the bus clock
are valid only for SH7724, so the clock selection code needs to be
updated when extending clock support to other SoCs.
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
[geert: Add note]
Reviewed-by: Sui Jingfeng <suijingfeng@loongson.cn>
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/d4b4d2164541ae055d10064103db3c2d6540e846.1694767209.git.geert+renesas@glider.be
The display info width_mm and height_mm fields are set at init time and
never overwritten, don't set them a second time when getting modes.
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Reviewed-by: Sui Jingfeng <suijingfeng@loongson.cn>
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/edd49fcb92af83d81df655b3db1685e8ed14380a.1694767209.git.geert+renesas@glider.be
Backlight support should be implemented by panels, not by the LCDC
driver. As the feature is currently unused anyway, remove it.
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
[geert: Cleanups]
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/69707650245bc2193d072f24723d4d5482ea590b.1694767209.git.geert+renesas@glider.be
Improve the table containing hardware information related to the
supported plane formats:
1. Move (part of) the overlay format register settings from multiple
switch() statements spread across the code into the table, like is
already done for the primary plane register settings,
2. Remove the .yuv field, as that information can easily be extracted
from the register settings using a new helper macro,
3. Shrink and move the .bpp field to reduce table size.
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/b41f755e80ffe5fb4adbb2d8f96f2073de5c33bc.1694767209.git.geert+renesas@glider.be
The primary plane supports various YCbCr formats, and the CRTC code
already knows how to handle them. Enable support for the missing
formats by adding them to the table of supported modes.
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/e5a5d5ab2083280be5fcdc428e8b8ca198b4448d.1694767209.git.geert+renesas@glider.be
