Not all ACPI-devices have a HID + UID, allow specifying quirks for
acpi_device_override_status() by path too.
Note this moves the path/HID+UID check to after the CPU + DMI checks
since the path lookup is somewhat costly.
This way this lookup is only done on devices where the other checks
match.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Currently, acpi_bus_get_status() calls acpi_device_always_present() to
allow platform quirks to override the _STA return to report that a
device is present (status = ACPI_STA_DEFAULT) independent of the _STA
return.
In some cases it might also be useful to have the opposite functionality
and have a platform quirk which marks a device as not present (status = 0)
to work around ACPI table bugs.
Change acpi_device_always_present() into a more generic
acpi_device_override_status() function to allow this.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
It turns out that there is a WMI object which controls the PWM2 device
used for the keyboard backlight and that WMI object also provides some
other useful functionality.
The upcoming lenovo-yogabook-wmi driver will offer both backlight
control and the other functionality, so there no longer is a need
to have the lpss-pwm driver binding to PWM2 for backlight control;
and this is now actually undesirable because this will cause both
the WMI code and the lpss-pwm driver to poke at the same PWM
controller.
Drop the always-present quirk for the PWM2 ACPI-device, so that the
lpss-pwm controller will no longer bind to it.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
For larger (bigger than a page) and noncontiguous mobs we have
to create page tables that allow the host to find the memory.
Those page tables just used regular system memory. Unfortunately
in TTM those BO's are not allowed to be busy thus can't be
fenced and we have to fence those bo's because we don't want
to destroy the page tables while the host is still executing
the command buffers which might be accessing them.
To solve it we introduce a new placement VMW_PL_SYSTEM which
is very similar to TTM_PL_SYSTEM except that it allows
fencing. This fixes kernel oops'es during unloading of the driver
(and pci hot remove/add) which were caused by busy BO's in
TTM_PL_SYSTEM being present in the delayed deletion list in
TTM (TTM_PL_SYSTEM manager is destroyed before the delayed
deletions are executed)
In existing implementation, core_clk_setrate() is getting called
concurrently in concurrent video sessions. Before the previous call to
core_clk_setrate returns, new call to core_clk_setrate is invoked from
another video session running concurrently. This results in latest
calculated frequency being set (higher/lower) instead of actual frequency
required for that video session. It also results in stability crashes
mention below. These resources are specific to video core, hence keeping
under core lock would ensure that they are estimated for all running video
sessions and called once for the video core.
Currently 'ar' reference is not added in skb_cb during
WMI mgmt tx. Though this is generally not used during tx completion
callbacks, on interface removal the remaining idr cleanup callback
uses the ar ptr from skb_cb from mgmt txmgmt_idr. Hence
fill them during tx call for proper usage.
Also free the skb which is missing currently in these
callbacks.
Crash_info:
[19282.489476] Unable to handle kernel NULL pointer dereference at virtual address 00000000
[19282.489515] pgd = 91eb8000
[19282.496702] [00000000] *pgd=00000000
[19282.502524] Internal error: Oops: 5 [#1] PREEMPT SMP ARM
[19282.783728] PC is at ath11k_mac_vif_txmgmt_idr_remove+0x28/0xd8 [ath11k]
[19282.789170] LR is at idr_for_each+0xa0/0xc8
rsi_get_* functions rely on an offset variable from usb
input. The size of usb input is RSI_MAX_RX_USB_PKT_SIZE(3000),
while 2-byte offset can be up to 0xFFFF. Thus a large offset
can cause out-of-bounds read.
The patch adds a bound checking condition when rcv_pkt_len is 0,
indicating it's USB. It's unclear whether this is triggerable
from other type of bus. The following check might help in that case.
offset > rcv_pkt_len - FRAME_DESC_SZ
The bug is trigerrable with conpromised/malfunctioning USB devices.
I tested the patch with the crashing input and got no more bug report.
Attached is the KASAN report from fuzzing.
BUG: KASAN: slab-out-of-bounds in rsi_read_pkt+0x42e/0x500 [rsi_91x]
Read of size 2 at addr ffff888019439fdb by task RX-Thread/227
When freeing rx_cb->rx_skb, the pointer is not set to NULL,
a later rsi_rx_done_handler call will try to read the freed
address.
This bug will very likley lead to double free, although
detected early as use-after-free bug.
The bug is triggerable with a compromised/malfunctional usb
device. After applying the patch, the same input no longer
triggers the use-after-free.
Currently, with an unknown recv_type, mwifiex_usb_recv
just return -1 without restoring the skb. Next time
mwifiex_usb_rx_complete is invoked with the same skb,
calling skb_put causes skb_over_panic.
The bug is triggerable with a compromised/malfunctioning
usb device. After applying the patch, skb_over_panic
no longer shows up with the same input.
The APT compares the current time stamp with a pre-set value. The
current code only considered the 4 LSB only. Yet, after reviews by
mathematicians of the user space Jitter RNG version >= 3.1.0, it was
concluded that the APT can be calculated on the 32 LSB of the time
delta. Thi change is applied to the kernel.
This fixes a bug where an AMD EPYC fails this test as its RDTSC value
contains zeros in the LSB. The most appropriate fix would have been to
apply a GCD calculation and divide the time stamp by the GCD. Yet, this
is a significant code change that will be considered for a future
update. Note, tests showed that constantly the GCD always was 32 on
these systems, i.e. the 5 LSB were always zero (thus failing the APT
since it only considered the 4 LSB for its calculation).
Signed-off-by: Stephan Mueller <smueller@chronox.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Sasha Levin <sashal@kernel.org>
If the IRQ is already in use, then acpi_dev_gpio_irq_get_by() really
should not change the type underneath the current owner.
I specifically hit an issue with this an a Chuwi Hi8 Super (CWI509) Bay
Trail tablet, when the Boot OS selection in the BIOS is set to Android.
In this case _STA for a MAX17047 ACPI I2C device wrongly returns 0xf and
the _CRS resources for this device include a GpioInt pointing to a GPIO
already in use by an _AEI handler, with a different type then specified
in the _CRS for the MAX17047 device. Leading to the acpi_dev_gpio_irq_get()
call done by the i2c-core-acpi.c code changing the type breaking the
_AEI handler.
Now this clearly is a bug in the DSDT of this tablet (in Android mode),
but in general calling irq_set_irq_type() on an IRQ which already is
in use seems like a bad idea.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
The UCR4_OREN should be disabled before disabling the uart receiver in
.stop_rx() instead of in the .shutdown().
Otherwise, if we have the overrun error during the receiver disable
process, the overrun interrupt will keep trigging until we disable the
OREN interrupt in the .shutdown(), because the ORE status can only be
cleared when read the rx FIFO or reset the controller. Although the
called time between the receiver disable and OREN disable in .shutdown()
is very short, there is still the risk of endless interrupt during this
short period of time. So here change to disable OREN before the receiver
been disabled in .stop_rx().
The others are superfluous with tty refcounting in place now. And they
are racy in fact:
* tty_port_initialized() reports false for a small moment after
interrupts are enabled.
* closing is 1 while the port is still alive.
The queues are flushed later during close anyway. So there is no need
for this special handling. Actually, the ISR should not flush the
queues. It should behave as every other driver, just queue the chars
into tty buffer and go on. But this will be changed later. There is
still a lot code depending on having tty in ISR (and not only tty_port).
In the configuration used by the b850v3, the STDP2690 is used to read EDID
data whilst it's the STDP4028 which can detect when monitors are connected.
This can result in problems at boot with monitors connected when the
STDP4028 is probed first, a monitor is detected and an attempt is made to
read the EDID data before the STDP2690 has probed:
On an arm64 platform with the Spectrum ASIC, after loading and executing
a new kernel via kexec, the following trace [1] is observed. This seems
to be caused by the fact that the device is not properly shutdown before
executing the new kernel.
Fix this by implementing a shutdown method which mirrors the remove
method, as recommended by the kexec maintainer [2][3].
Since commit 4b563a066611 ("ARM: imx: Remove imx21 support"), the config
DEBUG_IMX21_IMX27_UART is really only debug support for IMX27.
So, rename this option to DEBUG_IMX27_UART and adjust dependencies in
Kconfig and rename the definitions to IMX27 as further clean-up.
This issue was discovered with ./scripts/checkkconfigsymbols.py, which
reported that DEBUG_IMX21_IMX27_UART depends on the non-existing config
SOC_IMX21.
In case the following power domain sequence happens, iMX8M Mini always hangs:
gpumix:on -> gpu:on -> gpu:off -> gpu:on
This is likely due to another quirk of the GPC block. This situation can be
prevented by always synchronously powering off both the domain and MIX domain.
Make it so. This turns the aforementioned sequence into:
gpumix:on -> gpu:on -> gpu:off -> gpumix:off -> gpumix:on -> gpu:on
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Frieder Schrempf <frieder.schrempf@kontron.de> Cc: Lucas Stach <l.stach@pengutronix.de> Cc: NXP Linux Team <linux-imx@nxp.com> Cc: Peng Fan <peng.fan@nxp.com> Cc: Shawn Guo <shawnguo@kernel.org> Acked-by: Lucas Stach <l.stach@pengutronix.de> Signed-off-by: Shawn Guo <shawnguo@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Thermal zone names should not be longer than 20 names, which is indicated by
a message at boot. Change "camera-thermal-bottom" to "cam-thermal-bottom" to
fix it.
do_div() does a 64-by-32 division. Here the divisor is an
unsigned long which on some platforms is 64 bit wide. So use
div64_ul instead of do_div to avoid a possible truncation.
Eliminate the following coccicheck warning:
./drivers/net/ethernet/renesas/ravb_main.c:2492:1-7: WARNING:
do_div() does a 64-by-32 division, please consider using div64_ul
instead.
Reported-by: Abaci Robot <abaci@linux.alibaba.com> Signed-off-by: Yang Li <yang.lee@linux.alibaba.com> Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be> Reviewed-by: Sergey Shtylyov <s.shtylyov@omp.ru> Link: https://lore.kernel.org/r/1637228883-100100-1-git-send-email-yang.lee@linux.alibaba.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
According to [0], compilers sometimes might produce duplicate DWARF
definitions for exactly the same struct/union within the same
compilation unit (CU). We've had similar issues with identical arrays
and handled them with a similar workaround in 6b6e6b1d09aa ("libbpf:
Accomodate DWARF/compiler bug with duplicated identical arrays"). Do the
same for struct/union by ensuring that two structs/unions are exactly
the same, down to the integer values of field referenced type IDs.
Solving this more generically (allowing referenced types to be
equivalent, but using different type IDs, all within a single CU)
requires a huge complexity increase to handle many-to-many mappings
between canonidal and candidate type graphs. Before we invest in that,
let's see if this approach handles all the instances of this issue in
practice. Thankfully it's pretty rare, it seems.
A out-of-bounds bug can be triggered by an interrupt, the reason for
this bug is the lack of checking of register values.
In flexcop_pci_isr, the driver reads value from a register and uses it as
a dma address. Finally, this address will be passed to the count parameter
of find_next_packet. If this value is larger than the size of dma, the
index of buffer will be out-of-bounds.
Fix this by adding a check after reading the value of the register.
The following KASAN report reveals it:
BUG: KASAN: slab-out-of-bounds in find_next_packet
drivers/media/dvb-core/dvb_demux.c:528 [inline]
BUG: KASAN: slab-out-of-bounds in _dvb_dmx_swfilter
drivers/media/dvb-core/dvb_demux.c:572 [inline]
BUG: KASAN: slab-out-of-bounds in dvb_dmx_swfilter+0x3fa/0x420
drivers/media/dvb-core/dvb_demux.c:603
Read of size 1 at addr ffff8880608c00a0 by task swapper/2/0
Memory state around the buggy address: ffff8880608bff80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ffff8880608c0000: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>ffff8880608c0080: 00 00 00 00 fc fc fc fc fc fc fc fc 00 00 00 00
^ ffff8880608c0100: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ffff8880608c0180: fc fc fc fc fc fc fc fc 00 00 00 00 00 00 00 00
==================================================================
When connected over USB, the Apple Magic Mouse 2 and the Apple Magic
Trackpad 2 register multiple interfaces, one of them is used to report
the battery level.
However, unlike when connected over Bluetooth, the battery level is not
reported automatically and it is required to fetch it manually.
Fix the battery report descriptor and add a timer to fetch the battery
level.
We've noticed cases where tasks in a cgroup are stalled on memory but
there is little memory FULL pressure since tasks stay on the runqueue
in reclaim.
A simple example involves a single threaded program that keeps leaking
and touching large amounts of memory. It runs in a cgroup with swap
enabled, memory.high set at 10M and cpu.max ratio set at 5%. Though
there is significant CPU pressure and memory SOME, there is barely any
memory FULL since the task enters reclaim and stays on the runqueue.
However, this memory-bound task is effectively stalled on memory and
we expect memory FULL to match memory SOME in this scenario.
The code is confused about memstall && running, thinking there is a
stalled task and a productive task when there's only one task: a
reclaimer that's counted as both. To fix this, we redefine the
condition for PSI_MEM_FULL to check that all running tasks are in an
active memstall instead of checking that there are no running tasks.
Function fs endpoint file operations are synchronized via an interruptible
mutex wait. However we see threads that do ep file operations concurrently
are getting blocked for the mutex lock in __fdget_pos(). This is an
uninterruptible wait and we see hung task warnings and kernel panic
if hung_task_panic systcl is enabled if host does not send/receive
the data for long time.
The reason for threads getting blocked in __fdget_pos() is due to
the file position protection introduced by the commit 9c225f2655e3
("vfs: atomic f_pos accesses as per POSIX"). Since function fs
endpoint files does not have the notion of the file position, switch
to the stream mode. This will bypass the file position mutex and
threads will be blocked in interruptible state for the function fs
mutex.
It should not affects user space as we are only changing the task state
changes the task state from UNINTERRUPTIBLE to INTERRUPTIBLE while waiting
for the USB transfers to be finished. However there is a slight change to
the O_NONBLOCK behavior. Earlier threads that are using O_NONBLOCK are also
getting blocked inside fdget_pos(). Now they reach to function fs and error
code is returned. The non blocking behavior is actually honoured now.
Due to (wIndex & 0xff) - 1 can get an integer greater than 15, this
can cause array index to be out of bounds since the size of array
port_status is 15. This change prevents a possible out-of-bounds
pointer computation by forcing the use of a valid port number.
reset_control_(de)assert() calls are called on a shared reset line when
reset_control_reset has been used. This is not allowed by the reset
framework.
Use reset_control_rearm() call in suspend() and remove() as a way to state
that the resource is no longer used, hence the shared reset line
may be triggered again by other devices. Use reset_control_rearm() also in
case probe fails after reset() has been called.
reset_control_rearm() keeps use of triggered_count sane in the reset
framework, use of reset_control_reset() on shared reset line should be
balanced with reset_control_rearm().
Commit 31582373a4a8 ("ath11k: Change number of TCL rings to one for
QCA6390") avoids initializing the other entries of dp->tx_ring cause
the corresponding TX rings on QCA6390/WCN6855 are not used, but leaves
those ring masks in ath11k_hw_ring_mask_qca6390.tx unchanged. Normally
this is OK because we will only get interrupts from the first TX ring
on these chips and thus only the first entry of dp->tx_ring is involved.
In case of one MSI vector, all DP rings share the same IRQ. For each
interrupt, all rings have to be checked, which means the other entries
of dp->tx_ring are involved. However since they are not initialized,
system crashes.
Fix this issue by simply removing those ring masks.
The succ var tracks memory allocation erros on this function.
Fix it, in order to stop this W=1 Werror in clang:
drivers/staging/media/atomisp/pci/sh_css_params.c:2430:7: error: variable 'succ' set but not used [-Werror,-Wunused-but-set-variable]
bool succ = true;
^
Currently, creating a batman-adv interface in an unprivileged LXD
container and attaching secondary interfaces to it with "ip" or "batctl"
works fine. However all batctl debug and configuration commands
fail:
To fix this change the generic netlink permissions from GENL_ADMIN_PERM
to GENL_UNS_ADMIN_PERM. This way a batman-adv interface is fully
maintainable as root from within a user namespace, from an unprivileged
container.
All except one batman-adv netlink setting are per interface and do not
leak information or change settings from the host system and are
therefore save to retrieve or modify as root from within an unprivileged
container.
"batctl routing_algo" / BATADV_CMD_GET_ROUTING_ALGOS is the only
exception: It provides the batman-adv kernel module wide default routing
algorithm. However it is read-only from netlink and an unprivileged
container is still not allowed to modify
/sys/module/batman_adv/parameters/routing_algo. Instead it is advised to
use the newly introduced "batctl if create routing_algo RA_NAME" /
IFLA_BATADV_ALGO_NAME to set the routing algorithm on interface
creation, which already works fine in an unprivileged container.
Cc: Tycho Andersen <tycho@tycho.pizza> Signed-off-by: Linus Lüssing <linus.luessing@c0d3.blue> Signed-off-by: Sven Eckelmann <sven@narfation.org> Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de> Signed-off-by: Sasha Levin <sashal@kernel.org>
Fix following coccicheck warning:
./arch/arm/mach-shmobile/regulator-quirk-rcar-gen2.c:156:1-33: Function
for_each_matching_node_and_match should have of_node_put() before break
and goto.
Early exits from for_each_matching_node_and_match() should decrement the
node reference counter.
Now that we restore the default or last user set exposure setting on
power_up() there is no need for the registers written by ov2680_set_fmt()
to write to the exposure register.
Not doing so fixes the exposure always being reset to the value from
the res->regs array after a set_fmt().
The atomisp driver originally used the s_parm command to
initialize the run_mode type to the driver. So, before start
setting up the streaming, s_parm should be called.
So, even having 5 "normal" video devices, one meant to be used
for each type, the run_mode was actually selected when
s_parm is called.
Without setting the run mode, applications that don't call
VIDIOC_SET_PARM with a custom atomisp parameters won't work, as
the pipeline won't be set:
atomisp-isp2 0000:00:03.0: can't create streams
atomisp-isp2 0000:00:03.0: __get_frame_info 1600x1200 (padded to 0) returned -22
However, commit 8a7c5594c020 ("media: v4l2-ioctl: clear fields in s_parm")
broke support for it, with a good reason, as drivers shoudn't be
extending the API for their own purposes.
So, as an step to allow generic apps to use this driver, put
the device's run_mode in preview after open.
After this patch, using v4l2grab starts to work on preview
mode (/dev/video2):
The internal try_fmt logic is not meant to provide everything
that the V4L2 API should provide. Also, it doesn't decrement
the pads that are used only internally by the driver, but aren't
part of the device's output.
There have been reports of the WFI timing out on some boards, and a
patch was proposed to just remove it. This stuff is rather fragile,
and I believe the WFI might be needed with our FW prior to GM200.
However, we probably should not be touching PMU during init on GPUs
where we depend on NVIDIA FW, outside of limited circumstances, so
this should be a somewhat safer change that achieves the desired
result.
Reported-by: Diego Viola <diego.viola@gmail.com> Signed-off-by: Ben Skeggs <bskeggs@redhat.com> Reviewed-by: Karol Herbst <kherbst@redhat.com> Signed-off-by: Karol Herbst <kherbst@redhat.com> Link: https://gitlab.freedesktop.org/drm/nouveau/-/merge_requests/10 Signed-off-by: Sasha Levin <sashal@kernel.org>
The current ELD handling takes the internal connector ELD buffer and
shares it to the I2S and AHB sub-driver.
But with DRM_BRIDGE_ATTACH_NO_CONNECTOR, the connector is created
elsewhere (or not), and an eventual connector is known only
if the bridge chain up to a connector is enabled.
The current dw-hdmi code gets the current connector from
atomic_enable() so use the already stored connector pointer and
replace the buffer pointer with a callback returning the current
connector ELD buffer.
Since a connector is not always available, either pass an empty
ELD to the alsa HDMI driver or don't call snd_pcm_hw_constraint_eld()
in AHB driver.
I found the bug using a custome USBFuzz port. It's a research work
to fuzz USB stack/drivers. I modified it to fuzz ath9k driver only,
providing hand-crafted usb descriptors to QEMU.
After fixing the code (fourth byte in usb packet) to WDCMSG_TARGET_START,
I got the null-ptr-deref bug. I believe the bug is triggerable whenever
cmd->odata is NULL. After patching, I tested with the same input and no
longer see the KASAN report.
skb_ctx selftest didn't close bpf_object implicitly allocated by
bpf_prog_test_load() helper. Fix the problem by explicitly calling
bpf_object__close() at the end of the test.
The issue is that we received a DLM message for a user lock but the
destination lock is a kernel lock. Note that the address which is trying
to derefence is 00000000deadbeef, which is in a kernel lock
lkb->lkb_astparam, this field should never be derefenced by the DLM
kernel stack. In case of a user lock lkb->lkb_astparam is lkb->lkb_ua
(memory is shared by a union field). The struct lkb_ua will be handled
by the DLM kernel stack but on a kernel lock it will contain invalid
data and ends in most likely crashing the kernel.
If we remove one instance of adv using Set Extended Adv Enable, there
is a possibility of issue occurs when processing the Command Complete
event. Especially, the adv_info might not be found since we already
remove it in hci_req_clear_adv_instance() -> hci_remove_adv_instance().
If that's the case, we will mistakenly proceed to remove all adv
instances instead of just one single instance.
This patch fixes the issue by checking the content of the HCI command
instead of checking whether the adv_info is found.
Signed-off-by: Archie Pusaka <apusaka@chromium.org> Reviewed-by: Sonny Sasaka <sonnysasaka@chromium.org> Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Many DSI panel drivers fail to clean up their panel references on
mipi_dsi_attach() failure, so we're leaving a dangling drm_panel
reference to freed memory. Clean that up on failure.
Noticed by inspection, after seeing similar problems on other drivers.
Therefore, I'm not marking Fixes/stable.
hci_alloc_dev() do not init the device's flag. And hci_free_dev()
using put_device() to free the memory allocated for this device,
but it calls just put_device(dev) only in case of HCI_UNREGISTER
flag is set, So any error handing before hci_register_dev() success
will cause memory leak.
To avoid this behaviour we can using kfree() to release dev before
hci_register_dev() success.
Commit a5d3d1adc95f ("leds: lp55xx: Initialize enable GPIO direction to
output") attempts to fix this, but the fix did not work since at least
for the Nokia N900 the value needs to be set to HIGH, per the device
tree. So rather than hardcoding the value to a potentially invalid value
for some devices, let's set direction in lp55xx_init_device.
Fixes: a5d3d1adc95f ("leds: lp55xx: Initialize enable GPIO direction to output") Fixes: 92a81562e695 ("leds: lp55xx: Add multicolor framework support to lp55xx") Fixes: ac219bf3c9bd ("leds: lp55xx: Convert to use GPIO descriptors") Signed-off-by: Merlijn Wajer <merlijn@wizzup.org> Reviewed-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Pavel Machek <pavel@ucw.cz> Signed-off-by: Sasha Levin <sashal@kernel.org>
Since the LED multicolor framework support was added in commit 92a81562e695 ("leds: lp55xx: Add multicolor framework support to lp55xx")
LEDs on this platform stopped working.
Fixes: 92a81562e695 ("leds: lp55xx: Add multicolor framework support to lp55xx") Fixes: ac219bf3c9bd ("leds: lp55xx: Convert to use GPIO descriptors") Signed-off-by: Merlijn Wajer <merlijn@wizzup.org> Signed-off-by: Sicelo A. Mhlongo <absicsz@gmail.com> Signed-off-by: Pavel Machek <pavel@ucw.cz> Signed-off-by: Sasha Levin <sashal@kernel.org>
mips64el-linux-ld: arch/mips/boot/compressed/decompress.o: in function `FSE_buildDTable_internal':
decompress.c:(.text.FSE_buildDTable_internal+0x2cc): undefined reference to `__clzdi2'
mips64el-linux-ld: arch/mips/boot/compressed/decompress.o: in function `BIT_initDStream':
decompress.c:(.text.BIT_initDStream+0x7c): undefined reference to `__clzdi2'
mips64el-linux-ld: decompress.c:(.text.BIT_initDStream+0x158): undefined reference to `__clzdi2'
mips64el-linux-ld: arch/mips/boot/compressed/decompress.o: in function `ZSTD_buildFSETable_body_default.constprop.0':
decompress.c:(.text.ZSTD_buildFSETable_body_default.constprop.0+0x2a8): undefined reference to `__clzdi2'
mips64el-linux-ld: arch/mips/boot/compressed/decompress.o: in function `FSE_readNCount_body_default':
decompress.c:(.text.FSE_readNCount_body_default+0x130): undefined reference to `__ctzdi2'
mips64el-linux-ld: decompress.c:(.text.FSE_readNCount_body_default+0x1a4): undefined reference to `__ctzdi2'
mips64el-linux-ld: decompress.c:(.text.FSE_readNCount_body_default+0x2e4): undefined reference to `__clzdi2'
mips64el-linux-ld: arch/mips/boot/compressed/decompress.o: in function `HUF_readStats_body_default':
decompress.c:(.text.HUF_readStats_body_default+0x184): undefined reference to `__clzdi2'
mips64el-linux-ld: decompress.c:(.text.HUF_readStats_body_default+0x1b4): undefined reference to `__clzdi2'
mips64el-linux-ld: arch/mips/boot/compressed/decompress.o: in function `ZSTD_DCtx_getParameter':
decompress.c:(.text.ZSTD_DCtx_getParameter+0x60): undefined reference to `__clzdi2'
Fixes: a510b616131f ("MIPS: Add support for ZSTD-compressed kernels") Reported-by: kernel test robot <lkp@intel.com> Reported-by: Nick Terrell <terrelln@fb.com> Signed-off-by: Paul Cercueil <paul@crapouillou.net> Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Sasha Levin <sashal@kernel.org>
Just like before with __bswapdi2(), for MIPS pre-boot when
CONFIG_KERNEL_ZSTD=y the decompressor function will use __ashldi3(), so
the object file should be added to the target object file.
Fixes these build errors:
mipsel-linux-ld: arch/mips/boot/compressed/decompress.o: in function `FSE_buildDTable_internal':
decompress.c:(.text.FSE_buildDTable_internal+0x48): undefined reference to `__ashldi3'
mipsel-linux-ld: arch/mips/boot/compressed/decompress.o: in function `FSE_decompress_wksp_body_default':
decompress.c:(.text.FSE_decompress_wksp_body_default+0xa8): undefined reference to `__ashldi3'
mipsel-linux-ld: arch/mips/boot/compressed/decompress.o: in function `ZSTD_getFrameHeader_advanced':
decompress.c:(.text.ZSTD_getFrameHeader_advanced+0x134): undefined reference to `__ashldi3'
Signed-off-by: Paul Cercueil <paul@crapouillou.net> Reviewed-by: Randy Dunlap <rdunlap@infradead.org> Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Sasha Levin <sashal@kernel.org>
In commit 8a5a75e5e9e5 ("of/fdt: Make sure no-map does not remove
already reserved regions") we returned -EBUSY when trying to mark
regions as no-map when they intersect with reserved memory. The goal was
to find bad no-map reserved memory DT nodes that would unmap the kernel
text/data sections.
The problem is the reserved memory check will still trigger if the DT
has a /memreserve/ that completely subsumes the no-map memory carveouts
in the reserved memory node _and_ that region is also not part of the
memory reg property. For example in sc7180.dtsi we have the following
reserved-memory and memory node:
memory@80000000 {
/* We expect the bootloader to fill in the size */
reg = <0 0x80000000 0 0>;
};
memory@80000000 {
/* The bootloader fills in the size, and adds another region */
reg = <0 0x80000000 0 0x00800000>,
<0 0x80c00000 0 0x7f200000>;
};
The smem region is doubly reserved via /memreserve/ and by not being
part of the /memory reg property. This leads to the following warning
printed at boot.
OF: fdt: Reserved memory: failed to reserve memory for node 'memory@80900000': base 0x0000000080900000, size 2 MiB
Otherwise nothing really goes wrong because the smem region is not going
to be mapped by the kernel's direct linear mapping given that it isn't
part of the memory node. Therefore, let's only consider this to be a
problem if we're trying to mark a region as no-map and it is actually
memory that we're intending to keep out of the kernel's direct mapping
but it's already been reserved.
Acked-by: Mike Rapoport <rppt@kernel.org> Cc: Douglas Anderson <dianders@chromium.org> Cc: Nicolas Boichat <drinkcat@chromium.org> Cc: Quentin Perret <qperret@google.com> Cc: Jan Kiszka <jan.kiszka@siemens.com> Fixes: 8a5a75e5e9e5 ("of/fdt: Make sure no-map does not remove already reserved regions") Signed-off-by: Stephen Boyd <swboyd@chromium.org> Signed-off-by: Rob Herring <robh@kernel.org> Link: https://lore.kernel.org/r/20220107194233.2793146-1-swboyd@chromium.org Signed-off-by: Sasha Levin <sashal@kernel.org>
The cell_count field of of_phandle_iterator is the number of cells we
expect in the phandle arguments list when cells_name is missing. The
error message should show the number of cells we actually see.
bm1880_clk_unregister_pll & bm1880_clk_unregister_div both try to
free statically allocated variables, so remove those kfrees.
For example, if we take L703 kfree(div_hw):
- div_hw is a bm1880_div_hw_clock pointer
- in bm1880_clk_register_plls this is pointed to an element of arg1:
struct bm1880_div_hw_clock *clks
- in the probe, where bm1880_clk_register_plls is called arg1 is
bm1880_div_clks, defined on L371:
static struct bm1880_div_hw_clock bm1880_div_clks[]
According to RM, the clock divider range is from 1 to 8, clock
prescaling ratio may be any power of 2 from 1 to 128.
So the supported divider is not all the value between
1 and 1024, just limited value in that range.
Create table for the supported divder and add function to
check the clock divider is available by comparing with
the table.
According to RM, on auto mode:
For codec AK4458 and AK4497, the lowest ratio of MLCK/FS is 256
if sample rate is 8kHz-48kHz,
For codec AK5558, the lowest ratio of MLCK/FS is 512 if sample
rate is 8kHz-48kHz.
With these setting the sound quality for 8kHz-48kHz can be improved.
Transfer the refined slots and slot_width to akcodec_get_mclk_rate()
for mclk calculation, otherwise the mclk frequency does not match
with the slots and slot_width for S16_LE format, because the default
slot_width is 32.
The SAI on i.MX8MQ don't support one2one ratio for mclk:bclk, so
the mclk frequency exceeds the supported range of codec for
the case that sample rate is larger than 705kHZ and format is
S32_LE. Update the supported width for such case.
The suspend code unconditionally sets ->hp_jack_in and ->mic_jack_in
to zero but without reporting this status change to the HDA core.
To compensate for this, always assume a status change on the
first unsol event after boot or resume.
Fixes: 424e531b47f8 ("ALSA: hda/cs8409: Ensure Type Detection is only run on startup when necessary") Signed-off-by: Christian A. Ehrhardt <lk@c--e.de> Link: https://lore.kernel.org/r/20211231134432.atwmuzeceqiklcoa@cae.in-ulm.de Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Sasha Levin <sashal@kernel.org>
Commit c8b4f0865e82 reduced delays related to cs42l42 jack
detection. However, the change was too aggressive. As a result
internal speakers on DELL Inspirion 3501 are not detected.
Increase the delay in cs42l42_run_jack_detect() a bit.
The MIPS BMC63XX subarch does not provide/support clk_set_parent().
This causes build errors in a few drivers, so add a simple implementation
of that function so that callers of it will build without errors.
Fixes: e7300d04bd08 ("MIPS: BCM63xx: Add support for the Broadcom BCM63xx family of SOCs." ) Signed-off-by: Randy Dunlap <rdunlap@infradead.org> Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com> Acked-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Sasha Levin <sashal@kernel.org>
Originally, the conditions for preventing reentry are not correct.
dai->component->active is not the state specifically for pcmif dai, so it
is not a correct condition to indicate the status of pcmif dai.
On the other hand, snd_soc_dai_stream_actvie() in prepare ops for both
playback and capture possibly return true at the first entry when these
two streams are opened at the same time.
In the patch, I refer to the implementation in mt8192-dai-pcm.c.
Clock and enabling bit for PCMIF are managed by DAPM, and the condition
for prepare ops is replaced by the status of dai widget.
Fixes: 1f95c019115c ("ASoC: mediatek: mt8195: support pcm in platform driver") Signed-off-by: Trevor Wu <trevor.wu@mediatek.com> Link: https://lore.kernel.org/r/20211230084731.31372-2-trevor.wu@mediatek.com Signed-off-by: Mark Brown <broonie@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
When firmware load failed, kernel report task hung as follows:
INFO: task xrun:5191 blocked for more than 147 seconds.
Tainted: G W 5.16.0-rc5-next-20211220+ #11
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
task:xrun state:D stack: 0 pid: 5191 ppid: 270 flags:0x00000004
Call Trace:
__schedule+0xc12/0x4b50 kernel/sched/core.c:4986
schedule+0xd7/0x260 kernel/sched/core.c:6369 (discriminator 1)
schedule_timeout+0x7aa/0xa80 kernel/time/timer.c:1857
wait_for_completion+0x181/0x290 kernel/sched/completion.c:85
lattice_ecp3_remove+0x32/0x40 drivers/misc/lattice-ecp3-config.c:221
spi_remove+0x72/0xb0 drivers/spi/spi.c:409
lattice_ecp3_remove() wait for signals from firmware loading, but when
load failed, firmware_load() does not send this signal. This cause
device remove hung. Fix it by sending signal even if load failed.
Because of the potential failure of the ioremap(), the buf->area could
be NULL.
Therefore, we need to check it and return -ENOMEM in order to transfer
the error.
Fixes: f09aecd50f39 ("ASoC: SAMSUNG: Add I2S0 internal dma driver") Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com> Link: https://lore.kernel.org/r/20211228034026.1659385-1-jiasheng@iscas.ac.cn Signed-off-by: Mark Brown <broonie@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Fix get_parent() callback to return the correct index of the parent for
PLL_CMNLC1 clock. Add a separate table of register values corresponding
to the parent index for PLL_CMNLC1. Update set_parent() callback
accordingly.
When CONFIG_PPC_RFI_SRR_DEBUG=y we check the SRR values before returning
from interrupts. This is done in asm using EMIT_BUG_ENTRY, and passing
BUGFLAG_WARNING.
However that fails to create an exception table entry for the warning,
and so do_program_check() fails the exception table search and proceeds
to call _exception(), resulting in an oops like:
We should instead use EMIT_WARN_ENTRY, which creates an exception table
entry for the warning, allowing the warning to be correctly recognised,
and the code to resume after printing the warning.
Note however that because this warning is buried deep in the interrupt
return path, we are not able to recover from it (due to MSR_RI being
clear), so we still end up in die() with an unrecoverable exception.
When CONFIG_PPC_RFI_SRR_DEBUG=y we check that NIP and SRR0 match when
returning from interrupts. This can trigger falsely if NIP has either of
its two low bits set via sigreturn or ptrace, while SRR0 has its low two
bits masked in hardware.
As a quick fix make sure to mask the low bits before doing the check.
Fixes: 59dc5bfca0cb ("powerpc/64s: avoid reloading (H)SRR registers if they are still valid") Reported-by: Sachin Sant <sachinp@linux.vnet.ibm.com> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Tested-by: Sachin Sant <sachinp@linux.vnet.ibm.com> Link: https://lore.kernel.org/r/20211221135101.2085547-1-mpe@ellerman.id.au Signed-off-by: Sasha Levin <sashal@kernel.org>
Similar to commit 4a90bbb478db ("phy: uniphier-pcie: Fix updating phy
parameters"), in function uniphier_u3ssphy_set_param(), unintentionally
write zeros to other fields when writing PHY registers.
John Garry reported a deadlock that occurs when trying to access a
runtime-suspended SATA device. For obscure reasons, the rescan procedure
causes the link to be hard-reset, which disconnects the device.
The rescan tries to carry out a runtime resume when accessing the device.
scsi_rescan_device() holds the SCSI device lock and won't release it until
it can put commands onto the device's block queue. This can't happen until
the queue is successfully runtime-resumed or the device is unregistered.
But the runtime resume fails because the device is disconnected, and
__scsi_remove_device() can't do the unregistration because it can't get the
device lock.
The best way to resolve this deadlock appears to be to allow the block
queue to start running again even after an unsuccessful runtime resume.
The idea is that the driver or the SCSI error handler will need to be able
to use the queue to resolve the runtime resume failure.
This patch removes the err argument to blk_post_runtime_resume() and makes
the routine act as though the resume was successful always. This fixes the
deadlock.
Link: https://lore.kernel.org/r/1639999298-244569-4-git-send-email-chenxiang66@hisilicon.com Fixes: e27829dc92e5 ("scsi: serialize ->rescan against ->remove") Reported-and-tested-by: John Garry <john.garry@huawei.com> Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Xiang Chen <chenxiang66@hisilicon.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
This bug introduced by commit b261dba2fdb2 ("arm64: kdump: Remove custom
linux,usable-memory-range handling"), which moves
memblock_cap_memory_range() to fdt, but it breaches the rules that
memblock_cap_memory_range() should come after memblock_add() etc as said
in commit e888fa7bb882 ("memblock: Check memory add/cap ordering").
As a consequence, the virtual address set up by copy_oldmem_page() does
not bail out from the test of virt_addr_valid() in check_heap_object(),
and finally hits the BUG_ON().
Since memblock allocator has no idea about when the memblock is fully
populated, while efi_init() is aware, so tackling this issue by calling the
interface early_init_dt_check_for_usable_mem_range() exposed by of/fdt.
Fixes: b261dba2fdb2 ("arm64: kdump: Remove custom linux,usable-memory-range handling") Signed-off-by: Pingfan Liu <kernelfans@gmail.com> Cc: Rob Herring <robh+dt@kernel.org> Cc: Zhen Lei <thunder.leizhen@huawei.com> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Will Deacon <will@kernel.org> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Mike Rapoport <rppt@kernel.org> Cc: Geert Uytterhoeven <geert+renesas@glider.be> Cc: Frank Rowand <frowand.list@gmail.com> Cc: Ard Biesheuvel <ardb@kernel.org> Cc: Nick Terrell <terrelln@fb.com> Cc: linux-arm-kernel@lists.infradead.org
To: devicetree@vger.kernel.org
To: linux-efi@vger.kernel.org Acked-by: Ard Biesheuvel <ardb@kernel.org> Signed-off-by: Rob Herring <robh@kernel.org> Link: https://lore.kernel.org/r/20211215021348.8766-1-kernelfans@gmail.com Signed-off-by: Sasha Levin <sashal@kernel.org>
Currently, we parse the "linux,usable-memory-range" property in
early_init_dt_scan_chosen(), to obtain the specified memory range of the
crash kernel. We then reserve the required memory after
early_init_dt_scan_memory() has identified all available physical memory.
Because the two pieces of code are separated far, the readability and
maintainability are reduced. So bring them together.
Suggested-by: Rob Herring <robh@kernel.org> Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
(change the prototype of early_init_dt_check_for_usable_mem_range(), in
order to use it outside) Signed-off-by: Pingfan Liu <kernelfans@gmail.com> Tested-by: Dave Kleikamp <dave.kleikamp@oracle.com> Acked-by: John Donnelly <john.p.donnelly@oracle.com> Reviewed-by: Rob Herring <robh@kernel.org> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Will Deacon <will@kernel.org> Cc: linux-arm-kernel@lists.infradead.org
To: devicetree@vger.kernel.org
To: linux-efi@vger.kernel.org Signed-off-by: Rob Herring <robh@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
mt8195_cg_patch is used to reset the default value of audio cg, so the
register value could be consistent with CCF reference count.
Nevertheless, AUDIO_TOP_CON1[1:0] is used to control an internal mux,
and it's expected to keep the default value 0.
This patch corrects the default value in case an unexpected behavior
happens in the future.
It turns out to be possible for hotplugging out a device to reach the
stage of tearing down the device's group and default domain before the
domain's flush queue has drained naturally. At this point, it is then
possible for the timeout to expire just before the del_timer() call
in free_iova_flush_queue(), such that we then proceed to free the FQ
resources while fq_flush_timeout() is still accessing them on another
CPU. Crashes due to this have been observed in the wild while removing
NVMe devices.
Close the race window by using del_timer_sync() to safely wait for any
active timeout handler to finish before we start to free things. We
already avoid any locking in free_iova_flush_queue() since the FQ is
supposed to be inactive anyway, so the potential deadlock scenario does
not apply.
After calling dmaengine_submit(), the submitted transfer descriptor
belongs to the DMA engine. Pointer to that descriptor may no longer be
valid after the call and should be tested before awaiting transfer
completion.
Reported-by: Kevin Tian <kevin.tian@intel.com> Suggested-by: Dave Jiang <dave.jiang@intel.com> Fixes: 4fac9b31d0b9 ("ASoC: Intel: Add catpt base members") Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Link: https://lore.kernel.org/r/20211216115743.2130622-2-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Use IRQCHIP_MASK_ON_SUSPEND to make the core irq code to
mask the iommu interrupt on suspend and unmask it on the resume.
Since now the unmask function updates the INTX settings,
that will restore them on resume from s3/s4.
Since IRQCHIP_MASK_ON_SUSPEND is only effective for interrupts
which are not wakeup sources, remove IRQCHIP_SKIP_SET_WAKE flag
and instead implement a dummy .irq_set_wake which doesn't allow
the interrupt to become a wakeup source.
The last driver referencing the slave_id on Marvell PXA and MMP platforms
was the SPI driver, but this stopped doing so a long time ago, so the
TODO from the earlier patch can no be removed.
Fixes: b729bf34535e ("spi/pxa2xx: Don't use slave_id of dma_slave_config") Fixes: 13b3006b8ebd ("dma: mmp_pdma: add filter function") Signed-off-by: Arnd Bergmann <arnd@arndb.de> Acked-by: Mark Brown <broonie@kernel.org> Link: https://lore.kernel.org/r/20211122222203.4103644-7-arnd@kernel.org Signed-off-by: Vinod Koul <vkoul@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Commit d4a451d5fc84 ("arch: remove the ARCH_PHYS_ADDR_T_64BIT config
symbol") removes config ARCH_PHYS_ADDR_T_64BIT with all instances of that
config refactored appropriately. Since then, it is recommended to use the
config PHYS_ADDR_T_64BIT instead.
Commit 171543e75272 ("MIPS: Disallow CPU_SUPPORTS_HUGEPAGES for XPA,EVA")
introduces the expression "!(32BIT && (ARCH_PHYS_ADDR_T_64BIT || EVA))"
for config CPU_SUPPORTS_HUGEPAGES, which unintentionally refers to the
non-existing symbol ARCH_PHYS_ADDR_T_64BIT instead of the intended
PHYS_ADDR_T_64BIT.
Fix this Kconfig reference to the intended PHYS_ADDR_T_64BIT.
This issue was identified with the script ./scripts/checkkconfigsymbols.py.
I then reported it on the mailing list and Paul confirmed the mistake in
the linked email thread.
Link: https://lore.kernel.org/lkml/H8IU3R.H5QVNRA077PT@crapouillou.net/ Suggested-by: Paul Cercueil <paul@crapouillou.net> Fixes: 171543e75272 ("MIPS: Disallow CPU_SUPPORTS_HUGEPAGES for XPA,EVA") Signed-off-by: Lukas Bulwahn <lukas.bulwahn@gmail.com> Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Sasha Levin <sashal@kernel.org>
Commit ab7c01fdc3cf ("mips: Add MIPS Release 5 support") adds the two
configs CPU_MIPS32_R5 and CPU_MIPS64_R5, which depend on the corresponding
SYS_HAS_CPU_MIPS32_R5 and SYS_HAS_CPU_MIPS64_R5, respectively.
The config SYS_HAS_CPU_MIPS32_R5 was already introduced with commit c5b367835cfc ("MIPS: Add support for XPA."); the config
SYS_HAS_CPU_MIPS64_R5, however, was never introduced.
Add the definition for config SYS_HAS_CPU_MIPS64_R5 under the assumption
that SYS_HAS_CPU_MIPS64_R5 follows the same pattern as the existing
SYS_HAS_CPU_MIPS32_R5 and SYS_HAS_CPU_MIPS64_R6.
stm32's clk driver register two ltdc gate clk to clk core by
clk_hw_register_gate() and clk_hw_register_composite()
first: 'stm32f429_gates[]', clk name is 'ltdc', which no user to use.
second: 'stm32f429_aux_clk[]', clk name is 'lcd-tft', used by ltdc driver
both of them point to the same offset of stm32's RCC register. after
kernel enter console, clk core turn off ltdc's clk as 'stm32f429_gates[]'
is no one to use. but, actually 'stm32f429_aux_clk[]' is in use.
projects / thirdparty / kernel / stable.git / log
