During H264 API overhaul subtle bug was introduced Cedrus driver.
Progressive references have both, top and bottom reference flags set.
Cedrus reference list expects only bottom reference flag and only when
interlaced frames are decoded. However, due to a bug in Cedrus check,
exclusivity is not tested and that flag is set also for progressive
references. That causes "jumpy" background with many videos.
Fix that by checking that only bottom reference flag is set in control
and nothing else.
Tested-by: Andre Heider <a.heider@gmail.com>
Fixes: cfc8c3ed53 ("media: cedrus: h264: Properly configure reference field")
Signed-off-by: Jernej Skrabec <jernej.skrabec@siol.net>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Cc: <stable@vger.kernel.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
The BIT macro is not available in userspace, so replace BIT(0) by
0x00000001.
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Fixes: 6446ec6cbf ("media: v4l2-subdev: add VIDIOC_SUBDEV_QUERYCAP ioctl")
Cc: <stable@vger.kernel.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Currently, the __is_lm_address() check just masks out the top 12 bits
of the address, but if they are 0, it still yields a true result.
This has as a side effect that virt_addr_valid() returns true even for
invalid virtual addresses (e.g. 0x0).
Fix the detection checking that it's actually a kernel address starting
at PAGE_OFFSET.
Fixes: 68dd8ef321 ("arm64: memory: Fix virt_addr_valid() using __is_lm_address()")
Cc: <stable@vger.kernel.org> # 5.4.x
Cc: Will Deacon <will@kernel.org>
Suggested-by: Catalin Marinas <catalin.marinas@arm.com>
Reviewed-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Vincenzo Frascino <vincenzo.frascino@arm.com>
Link: https://lore.kernel.org/r/20210126134056.45747-1-vincenzo.frascino@arm.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
During the driver refactor, a regression broke the logic inside
hi6421_spmi_regulator_get_optimum_mode(). Basically, if a LDO
has eco_uA == 0, it doesn't support economic mode. So, it should
return REGULATOR_MODE_NORMAL.
If economic mode is supported, it can return either
REGULATOR_MODE_IDLE or REGULATOR_MODE_NORMAL, depending on the
load current.
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Link: https://lore.kernel.org/r/f087981eb695eaab8c301c42977a4aa884affbbf.1611212783.git.mchehab+huawei@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove obvious comments and fix the comment for the
HI6421V600_LDO() macro.
While on it, use kernel-doc notation for HI6421V600_LDO(),
as kernel-doc can check if the arguments match its
description.
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Link: https://lore.kernel.org/r/d5e6dbdee5f7e143300249251ddbe09fdf64e669.1611212783.git.mchehab+huawei@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The original driver, which can be seen at
commit 42f24d9d44 ("staging: regulator: add a regulator driver for HiSilicon 6421v600 SPMI PMIC")
had a complex logic to ensure that there won't be multiple power
enable/disable commands running at the same time. At the original
logic, it were ensured that:
- a next power up/down would wait for at least the on/off period;
- an extra delay would be granted. It turns that such extra delay
has a value of zero, but it was relying on gettimeofday()
call, which can take some time.
This was later simplified, but there are still some possible
issues. In order to avoid that, let's simply add a delay
to wait for the power up line to stabilize after powering up
a device.
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Link: https://lore.kernel.org/r/6733dac9813ba6688def404142cb7b964accf758.1611212783.git.mchehab+huawei@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In preparation for de-staging, do some cleanups:
- Return error codes from hi6421_spmi_pmic_rmw();
- Remove a debug message;
- Change the module description;
- a few minor coding style adjustments.
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Link: https://lore.kernel.org/r/0bae0c05d997e4a5a0b3b86a65f3370dafb14596.1611212783.git.mchehab+huawei@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There are a few issues on this function:
1. Instead of using 1/0 for true/false, change the type to boolean;
2. there's a typo there:
seleted -> selected
3. It's logic is reversed.
Address them.
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Link: https://lore.kernel.org/r/a579004cfa0cb3cca55c2124a8574a7aeb4eacc3.1611052729.git.mchehab+huawei@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Instead of running a loop up to 100k times, add a small
delay inside it, running it up to 10 times, waiting up
to 100-200 us.
It should be noticed that I don't have the datasheet for
this PHY. So, not sure if this time will cover all
situations.
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Link: https://lore.kernel.org/r/b653d7d6073de176598a5026c41b1a845f360c9e.1611052729.git.mchehab+huawei@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The interrupt support for Advantech PCI-1730 currrently supports only
rising edge inputs for the trigger sources. Each of four interrupt
sources (each with its own Comedi subdevice) can be set to trigger on
either a rising edge or a falling edge. Add support for choosing the
edge during set-up of the asynchronous command for the subdevice, using
the `CR_INVERT` bit of `scan_begin_arg` to indicate falling edge when
set, or rising edge when clear. Also allow the `CR_EDGE` bit to be set,
but ignore it. All other bits of `scan_begin_arg` must be zero.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Link: https://lore.kernel.org/r/20210118144359.378730-3-abbotti@mev.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
On the Advantech PCI-1730, four digital inputs (DI0, DI1, IDI0 and IDI1)
can be used as external interrupt sources. Each input can be programmed
to latch an interrupt bit on either a rising edge or a falling edge (but
not both).
Add a new Comedi subdevice for each interrupt source, supporting the
asynchronous command interface. Subdevices 5, 6, 7 and 8 are for
interrupt sources DI0, DI1, IDI0 and IDI1. They each write the state of
16 digital inputs to the subdevice's data buffer each time the
corresponding interrupt occurs. (For DI0 and DI1, use the 16
non-isolated digital inputs. For IDI0 and IDI1, use the 16 isolated
digital inputs.)
Currently, only rising edge triggers are supported. Support could be
added for the PCI-1733 and PCI-1736.
Signed-off-by: Bernd Harries <bha@gmx.de>
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Link: https://lore.kernel.org/r/20210118144359.378730-2-abbotti@mev.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
On the ADLink PCI-7230, digital input channels 0 and 1 can be used as
external interrupt sources. A rising edge on each input latches a
corresponding local interrupt input of the PCI interface chip. Writing
a "clear IRQ" register clears both latches.
Add a new Comedi subdevice for each interrupt source, supporting the
asynchronous command interface. This writes the state of the 16 digital
inputs to the subdevice's data buffer each time the corresponding
interrupt occurs.
This could be adapted to support the PCI-7233, PCI-7432 and PCI-7433
boards too. They all have two interrupt sources, although for PCI-7233
each interrupt source is triggered by a change of state of 16 digital
inputs (0-15 and 16-31). The "clear IRQ" register is at a different
offset for some boards.
Signed-off-by: Bernd Harries <bha@gmx.de>
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Link: https://lore.kernel.org/r/20210118141829.376505-1-abbotti@mev.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This was for OpenWrt's swconfig driver, which never made it upstream,
and was also superseded by MT7530 DSA driver.
Reviewed-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Signed-off-by: DENG Qingfang <dqfext@gmail.com>
Link: https://lore.kernel.org/r/20210108025155.31556-1-dqfext@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
With some USB network adapters, such as DM96xx, the following message
is seen for each maximum size receive packet.
dwc2 ff540000.usb: dwc2_update_urb_state(): trimming xfer length
This happens because the packet size requested by the driver is 1522
bytes, wMaxPacketSize is 64, the dwc2 driver configures the chip to
receive 24*64 = 1536 bytes, and the chip does indeed send more than
1522 bytes of data. Since the event does not indicate an error condition,
the message is just noise. Demote it to debug level.
Fixes: 7359d482eb ("staging: HCD files for the DWC2 driver")
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Link: https://lore.kernel.org/r/20210113112052.17063-4-nsaenzjulienne@suse.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In some situations, the following error messages are reported.
dwc2 ff540000.usb: dwc2_hc_chhltd_intr_dma: Channel 1 - ChHltd set, but reason is unknown
dwc2 ff540000.usb: hcint 0x00000002, intsts 0x04000021
This is sometimes followed by:
dwc2 ff540000.usb: dwc2_update_urb_state_abn(): trimming xfer length
and then:
WARNING: CPU: 0 PID: 0 at kernel/v4.19/drivers/usb/dwc2/hcd.c:2913
dwc2_assign_and_init_hc+0x98c/0x990
The warning suggests that an odd buffer address is to be used for DMA.
After an error is observed, the receive buffer may be full
(urb->actual_length >= urb->length). However, the urb is still left in
the queue unless three errors were observed in a row. When it is queued
again, the dwc2 hcd code translates this into a 1-block transfer.
If urb->actual_length (ie the total expected receive length) is not
DMA-aligned, the buffer pointer programmed into the chip will be
unaligned. This results in the observed warning.
To solve the problem, abort input transactions after an error with
unknown cause if the entire packet was already received. This may be
a bit drastic, but we don't really know why the transfer was aborted
even though the entire packet was received. Aborting the transfer in
this situation is less risky than accepting a potentially corrupted
packet.
With this patch in place, the 'ChHltd set' and 'trimming xfer length'
messages are still observed, but there are no more transfer attempts
with odd buffer addresses.
Fixes: 151d0cbdbe ("usb: dwc2: make the scheduler handle excessive NAKs better")
Cc: Boris ARZUR <boris@konbu.org>
Cc: Douglas Anderson <dianders@chromium.org>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Link: https://lore.kernel.org/r/20210113112052.17063-3-nsaenzjulienne@suse.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The DWC2 documentation states that transfers with zero data length should
set the number of packets to 1 and the transfer length to 0. This is not
currently the case for inbound transfers: the transfer length is set to
the maximum packet length. This can have adverse effects if the chip
actually does transfer data as it is programmed to do. Follow chip
documentation and keep the transfer length set to 0 in that situation.
Fixes: 56f5b1cff2 ("staging: Core files for the DWC2 driver")
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Link: https://lore.kernel.org/r/20210113112052.17063-2-nsaenzjulienne@suse.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Tasklets have long been deprecated as being too heavy on the system
by running in irq context - and this is not a performance critical
path. If a higher priority process wants to run, it must wait for
the tasklet to finish before doing so.
c67x00_do_work() will now run in process context and have further
concurrency (tasklets being serialized among themselves), but this
is done holding the c67x00->lock, so it should be fine. Furthermore,
this patch fixes the usage of the lock in the callback as otherwise
it would need to be irq-safe.
Signed-off-by: Davidlohr Bueso <dbueso@suse.de>
Link: https://lore.kernel.org/r/20210113031537.79859-1-dave@stgolabs.net
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The OBJECT_FILES_NON_STANDARD annotation is used to tell objtool to
ignore a file. File-level ignores won't work when validating vmlinux.o.
Instead, convert restore_image() and core_restore_code() to be ELF
functions. Their code is conventional enough for objtool to be able to
understand them.
Cc: "Rafael J. Wysocki" <rjw@rjwysocki.net>
Cc: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Link: https://lore.kernel.org/r/974f8ceb5385e470f72e93974c70ab5c894bb0dc.1611263462.git.jpoimboe@redhat.com
Because restore_registers() is page-aligned, the assembler inexplicably
adds an unreachable jump from after the end of the previous function to
the beginning of restore_registers().
That confuses objtool, understandably. It also creates significant text
fragmentation. As a result, most of the object file is wasted text
(nops).
Move restore_registers() to the beginning of the file to both prevent
the text fragmentation and avoid the dead jump instruction.
$ size /tmp/hibernate_asm_64.before.o /tmp/hibernate_asm_64.after.o
text data bss dec hex filename
4415 0 0 4415 113f /tmp/hibernate_asm_64.before.o
524 0 0 524 20c /tmp/hibernate_asm_64.after.o
Cc: "Rafael J. Wysocki" <rjw@rjwysocki.net>
Cc: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Link: https://lore.kernel.org/r/8c7f634201d26453d73fe55032cbbdc05d004387.1611263462.git.jpoimboe@redhat.com
The OBJECT_FILES_NON_STANDARD annotation is used to tell objtool to
ignore a file. File-level ignores won't work when validating vmlinux.o.
Instead, tell objtool to ignore do_suspend_lowlevel() directly with the
STACK_FRAME_NON_STANDARD annotation.
Cc: "Rafael J. Wysocki" <rjw@rjwysocki.net>
Cc: Len Brown <len.brown@intel.com>
Cc: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Link: https://lore.kernel.org/r/269eda576c53bc9ecc8167c211989111013a67aa.1611263462.git.jpoimboe@redhat.com
With objtool vmlinux.o validation of return_to_handler(), now that
objtool has visibility inside the retpoline, jumping from EMPTY state to
a proper function state results in a stack state mismatch.
return_to_handler() is actually quite normal despite the underlying
magic. Just annotate it as a normal function.
Acked-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Link: https://lore.kernel.org/r/14f48e623f61dbdcd84cf27a56ed8ccae73199ef.1611263462.git.jpoimboe@redhat.com
The Xen hypercall page is filled with zeros, causing objtool to fall
through all the empty hypercall functions until it reaches a real
function, resulting in a stack state mismatch.
The build-time contents of the hypercall page don't matter because the
page gets rewritten by the hypervisor. Make it more palatable to
objtool by making each hypervisor function a true empty function, with
nops and a return.
Cc: Juergen Gross <jgross@suse.com>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Link: https://lore.kernel.org/r/0883bde1d7a1fb3b6a4c952bc0200e873752f609.1611263462.git.jpoimboe@redhat.com
The OBJECT_FILES_NON_STANDARD annotation is used to tell objtool to
ignore a file. File-level ignores won't work when validating vmlinux.o.
Tweak the ELF metadata and unwind hints to allow objtool to follow the
code.
Cc: Juergen Gross <jgross@suse.com>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Link: https://lore.kernel.org/r/8b042a09c69e8645f3b133ef6653ba28f896807d.1611263462.git.jpoimboe@redhat.com
In 'dma_pool_create()', we return -ENOMEM, but don't release the resources
already allocated, as in all the other error handling paths.
Go to 'err_res_free' instead of returning directly.
Fixes: 0177947397 ("dmaengine: ti: k3-udma: Initial support for K3 BCDMA")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Acked-by: Peter Ujfalusi <peter.ujfalusi@gmail.com>
Link: https://lore.kernel.org/r/20210124070923.724479-1-christophe.jaillet@wanadoo.fr
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Add a module parameter that overrides the SVA feature enabling. This keeps
the driver in legacy mode even when intel_iommu=sm_on is set. In this mode,
the descriptor fields must be programmed with dma_addr_t from the Linux DMA
API for source, destination, and completion descriptors.
Signed-off-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/161134110457.4005461.13171197785259115852.stgit@djiang5-desk3.ch.intel.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
