Randy Dunlap [Fri, 15 Jul 2022 01:59:48 +0000 (18:59 -0700)]
leds: clevo-mail: fix Kconfig "its" grammar
Use the possessive "its" instead of the contraction "it's"
where appropriate.
Signed-off-by: Randy Dunlap <rdunlap@infradead.org> Cc: Pavel Machek <pavel@ucw.cz> Cc: Lee Jones <lee.jones@linaro.org> Cc: linux-leds@vger.kernel.org Signed-off-by: Pavel Machek <pavel@ucw.cz>
rpmsg: qcom: glink: replace strncpy() with strscpy_pad()
The use of strncpy() is considered deprecated for NUL-terminated
strings[1]. Replace strncpy() with strscpy_pad(), to keep existing
pad-behavior of strncpy, similarly to commit 08de420a8014 ("rpmsg:
glink: Replace strncpy() with strscpy_pad()"). This fixes W=1 warning:
In function ‘qcom_glink_rx_close’,
inlined from ‘qcom_glink_work’ at ../drivers/rpmsg/qcom_glink_native.c:1638:4:
drivers/rpmsg/qcom_glink_native.c:1549:17: warning: ‘strncpy’ specified bound 32 equals destination size [-Wstringop-truncation]
1549 | strncpy(chinfo.name, channel->name, sizeof(chinfo.name));
Stephan Gerhold [Tue, 12 Jul 2022 12:44:18 +0000 (14:44 +0200)]
ARM: dts: qcom: msm8974: Disable remoteprocs by default
The remoteproc configuration in qcom-msm8974.dtsi is incomplete because
it lacks the regulator supplies that should be added in the board DT
files. Some of the msm8974 boards are currently missing the regulator
supplies and should have the remoteprocs disabled to avoid making use
of the incomplete configuration.
This also fixes dtbs_check warnings after moving "qcom,msm8974-mss-pil"
to DT schema, which rightfully complains that the -supply properties
are missing for some boards:
qcom-apq8074-dragonboard.dtb:
remoteproc@fc880000: 'pll-supply' is a required property
From schema: remoteproc/qcom,msm8916-mss-pil.yaml
remoteproc@fc880000: 'mss-supply' is a required property
From schema: remoteproc/qcom,msm8916-mss-pil.yaml
remoteproc@fc880000: 'oneOf' conditional failed, one must be fixed:
'power-domains' is a required property
'power-domain-names' is a required property, or
'cx-supply' is a required property
'mx-supply' is a required property
Cc: Luca Weiss <luca@z3ntu.xyz> Cc: Konrad Dybcio <konrad.dybcio@somainline.org> Fixes: f300826d27be ("ARM: dts: qcom-msm8974: Sort and clean up nodes") Signed-off-by: Stephan Gerhold <stephan.gerhold@kernkonzept.com> Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org> Link: https://lore.kernel.org/r/20220712124421.3129206-4-stephan.gerhold@kernkonzept.com
Add missing hwlock for ipq8064 dtsi provided by qcom,sfpb-mutex.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com> Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
[bjorn: Moved the node inside /soc] Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org> Link: https://lore.kernel.org/r/20220707102040.1859-2-ansuelsmth@gmail.com
ARM: dts: qcom: ipq8064: reduce pci IO size to 64K
The current value for pci IO is problematic for ath10k wifi card
commonly connected to ipq8064 SoC.
The current value is probably a typo and is actually uncommon to find
1MB IO space even on a x86 arch. Also with recent changes to the pci
driver, pci1 and pci2 now fails to function as any connected device
fails any reg read/write. Reduce this to 64K as it should be more than
enough and 3 * 64K of total IO space doesn't exceed the IO_SPACE_LIMIT
hardcoded for the ARM arch.
ARM: dts: qcom: ipq8064: disable usb phy by default
Disable usb phy by default. When the usb phy were pushed, half of them
were flagged as disabled by mistake.
Correctly disable all usb phy and enable them only if a device actually
use them.
ARM: dts: qcom: ipq8064: add specific dtsi with smb208 rpm regulators
Add specific ipq8064 dtsi with smb208 rpm regulators.
Qcom advise to use this configuration but it's not mandatory and OEM
can decide to implement their own regulators.
smb208 regulators are used to scale CPU voltage, L2 cache voltage and
Ubi32 cores.
There regulators are controlled by rpm and to correctly works gsbi4-i2c
require to be NEVER disabled or rpm will reject any regulator change
request.
clk: qcom: fix build error initializer element is not constant
CONFIG_CC_VERSION_TEXT="x86_64-linux-gnu-gcc (Ubuntu 7.4.0-1ubuntu1~18.04.1) 7.4.0"
make ARCH=x86_64 CROSS_COMPILE=x86_64-linux-gnu-, will be failed, like this:
drivers/clk/qcom/gpucc-sm8350.c:111:2: error: initializer element is not constant
gpu_cc_parent,
^~~~~~~~~~~~~
drivers/clk/qcom/gpucc-sm8350.c:111:2: note: (near initialization for ‘gpu_cc_parent_data_0[0]’)
drivers/clk/qcom/gpucc-sm8350.c:126:2: error: initializer element is not constant
gpu_cc_parent,
^~~~~~~~~~~~~
drivers/clk/qcom/gpucc-sm8350.c:126:2: note: (near initialization for ‘gpu_cc_parent_data_1[0]’)
make[3]: *** [drivers/clk/qcom/gpucc-sm8350.o] Error 1
It seems that nested constant initializer is not supported in GCC 7.4.0. For portability resons, we should fix it.
Reported-by: Hulk Robot <hulkci@huawei.com> Fixes: 160758b05ab1 ("clk: qcom: add support for SM8350 GPUCC") Signed-off-by: Ren Zhijie <renzhijie2@huawei.com> Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org> Link: https://lore.kernel.org/r/20220714121144.71062-1-renzhijie2@huawei.com
Johan Hovold [Fri, 15 Jul 2022 07:02:48 +0000 (09:02 +0200)]
ARM: dts: qcom: sdx65: reorder USB interrupts
Three SoCs did not follow the interrupt order specified by the USB
controller binding.
While keeping the non-SuperSpeed interrupts together seems natural,
reorder the interrupts to match the binding.
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
[bjorn: Split out from arm64 patch] Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org> Link: https://lore.kernel.org/r/20220715070248.19078-5-johan+linaro@kernel.org
Johan Hovold [Fri, 15 Jul 2022 07:02:48 +0000 (09:02 +0200)]
arm64: dts: qcom: reorder USB interrupts
Three SoCs did not follow the interrupt order specified by the USB
controller binding.
While keeping the non-SuperSpeed interrupts together seems natural,
reorder the interrupts to match the binding.
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
[bjorn: Omitted sdx65 part from this patch] Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org> Link: https://lore.kernel.org/r/20220715070248.19078-5-johan+linaro@kernel.org
Johan Hovold [Fri, 15 Jul 2022 07:02:46 +0000 (09:02 +0200)]
arm64: dts: qcom: sc8280xp: fix USB interrupts
The two single-port SC8280XP USB controllers do not have an hs_phy_irq
interrupt. Instead they have a pwr_event interrupt which is distinct
from the former and not yet supported by the driver.
Fix the USB node interrupt names so that they match the devicetree
binding.
Also fix the pwr_event interrupt of the second controller which should
be 811 as noticed by Andrew Halaney.
Fixes: 152d1faf1e2f ("arm64: dts: qcom: add SC8280XP platform") Reviewed-by: Andrew Halaney <ahalaney@redhat.com> Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Signed-off-by: Johan Hovold <johan+linaro@kernel.org> Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org> Link: https://lore.kernel.org/r/20220715070248.19078-3-johan+linaro@kernel.org
Johan Hovold [Fri, 15 Jul 2022 07:02:45 +0000 (09:02 +0200)]
arm64: dts: qcom: sc8280xp: fix USB clock order and naming
Fix the USB controller clock order and naming so that they match the
devicetree binding.
Note that the driver currently simply enables all clocks in the order
that they are specified in the devicetree. Reordering the clocks as per
the binding means that the only explicit ordering constraint found in
the vendor driver, that cfg_noc should be enabled before the core_clk,
is now honoured.
It seems to me that ab8500 driver is using dedicated workqueues and
is not calling schedule{,_delayed}_work{,_on}(). Then, there will be
no work to flush using flush_scheduled_work().
Linus Walleij [Thu, 23 Jun 2022 14:25:45 +0000 (16:25 +0200)]
power: supply: ab8500: Drop external charger leftovers
Some leftover code for external chargers only used with
unreleased ASIC revisions and the header file for the
unsupported PM2301 was left behind in an earlier cleanup,
fix it by deleting the remnants.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.com>
dt-bindings: power: reset: qcom,pshold: convert to dtschema
Convert the Qualcomm Power Supply Hold Reset bindings to DT schema.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Reviewed-by: Rob Herring <robh@kernel.org> Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.com>
Merge tag 'for-5.19-rc7-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux
Pull btrfs reverts from David Sterba:
"Due to a recent report [1] we need to revert the radix tree to xarray
conversion patches.
There's a problem with sleeping under spinlock, when xa_insert could
allocate memory under pressure. We use GFP_NOFS so this is a real
problem that we unfortunately did not discover during review.
I'm sorry to do such change at rc6 time but the revert is IMO the
safer option, there are patches to use mutex instead of the spin locks
but that would need more testing. The revert branch has been tested on
a few setups, all seem ok.
The conversion to xarray will be revisited in the future"
Link: https://lore.kernel.org/linux-btrfs/cover.1657097693.git.fdmanana@suse.com/
* tag 'for-5.19-rc7-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:
Revert "btrfs: turn delayed_nodes_tree into an XArray"
Revert "btrfs: turn name_cache radix tree into XArray in send_ctx"
Revert "btrfs: turn fs_info member buffer_radix into XArray"
Revert "btrfs: turn fs_roots_radix in btrfs_fs_info into an XArray"
Merge tag 'arm-soc/for-5.20/soc-part2' of https://github.com/Broadcom/stblinux into arm/soc
This pull request contains Broadcom ARM-based SoCs Kconfig updates for
5.20, second part, please pull the following:
- Rafal spells out what the BCA acronym means
- Florian removes the forced thermal subsystem selection done by
ARCH_BCM_NSP
* tag 'arm-soc/for-5.20/soc-part2' of https://github.com/Broadcom/stblinux:
ARM: bcmbca: Include full family name in Kconfig
ARM: bcm: NSP: Removed forced thermal selection
Merge tag 'arm-soc/for-5.20/devicetree-arm64-part2' of https://github.com/Broadcom/stblinux into arm/dt
This pull request contains Broadcom ARM64-based SoCs Device Tree changes
for 5.20, second part, please pull the following:
- Rafal adds basic support for the BCM4912-based router Asus GT-AX6000,
he also has ARCH_BCM4908 select ARCH_BCMBCA such that in the next cycle
we can remove ARCH_BCM4908 and continue to have make oldconfig work.
Finally he spells out what BCA means to make it clearer
* tag 'arm-soc/for-5.20/devicetree-arm64-part2' of https://github.com/Broadcom/stblinux:
arm64: bcmbca: Include full family name in Kconfig
arm64: make ARCH_BCM4908 select ARCH_BCMBCA
arm64: dts: Add base DTS file for bcmbca device Asus GT-AX6000
Merge tag 'tty-5.19-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty
Pull tty and serial driver fixes from Greg KH:
"Here are some TTY and Serial driver fixes for 5.19-rc7. They resolve a
number of reported problems including:
- longtime bug in pty_write() that has been reported in the past.
- 8250 driver fixes
- new serial device ids
- vt overlapping data copy bugfix
- other tiny serial driver bugfixes
All of these have been in linux-next for a while with no reported
problems"
* tag 'tty-5.19-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty:
tty: use new tty_insert_flip_string_and_push_buffer() in pty_write()
tty: extract tty_flip_buffer_commit() from tty_flip_buffer_push()
serial: 8250: dw: Fix the macro RZN1_UART_xDMACR_8_WORD_BURST
vt: fix memory overlapping when deleting chars in the buffer
serial: mvebu-uart: correctly report configured baudrate value
serial: 8250: Fix PM usage_count for console handover
serial: 8250: fix return error code in serial8250_request_std_resource()
serial: stm32: Clear prev values before setting RTS delays
tty: Add N_CAN327 line discipline ID for ELM327 based CAN driver
serial: 8250: Fix __stop_tx() & DMA Tx restart races
serial: pl011: UPSTAT_AUTORTS requires .throttle/unthrottle
tty: serial: samsung_tty: set dma burst_size to 1
serial: 8250: dw: enable using pdata with ACPI
Rework the IIO test Kconfig to support building KUnit tests as modules.
This lets users execute tests at runtime in addition to the usual
tools/testing/kunit/kunit.py script.
The KUnit documentation indicates that all KUnit Kconfig entries must
match CONFIG_<name>_KUNIT_TEST: where <name> is the name of the test
suite. Rename the IIO_TEST_FORMAT configuration accordingly.
Merge tag 's390-5.19-6' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux
Pull s390 fixes from Alexander Gordeev:
- Fix building of out-of-tree kernel modules without a pre-built kernel
in case CONFIG_EXPOLINE_EXTERN=y.
- Fix a reference counting error that could prevent unloading of zcrypt
modules.
* tag 's390-5.19-6' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:
s390/ap: fix error handling in __verify_queue_reservations()
s390/nospec: remove unneeded header includes
s390/nospec: build expoline.o for modules_prepare target
Merge tag 'pm-5.19-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull power management fix from Rafael Wysocki
"Fix recent regression in the cpufreq mediatek driver related to
incorrect handling of regulator_get_optional() return value
(AngeloGioacchino Del Regno)"
* tag 'pm-5.19-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
cpufreq: mediatek: Handle sram regulator probe deferral
Merge tag 'acpi-5.19-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull ACPI fix from Rafael Wysocki:
"Fix more fallout from recent changes of the ACPI CPPC handling on AMD
platforms (Mario Limonciello)"
* tag 'acpi-5.19-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
ACPI: CPPC: Fix enabling CPPC on AMD systems with shared memory
iio: dac: cio-dac: Cleanup indexing for DAC writes
Simplify DAC write code by defining base member as u16 __iomem *; DAC
registers are 16-bit so this allows us to index each DAC channel
directly in a loop rather than calculating the offsets by multipling by
2 each time.
random: cap jitter samples per bit to factor of HZ
Currently the jitter mechanism will require two timer ticks per
iteration, and it requires N iterations per bit. This N is determined
with a small measurement, and if it's too big, it won't waste time with
jitter entropy because it'd take too long or not have sufficient entropy
anyway.
With the current max N of 32, there are large timeouts on systems with a
small CONFIG_HZ. Rather than set that maximum to 32, instead choose a
factor of CONFIG_HZ. In this case, 1/30 seems to yield sane values for
different configurations of CONFIG_HZ.
Reported-by: Vladimir Murzin <vladimir.murzin@arm.com> Fixes: 78c768e619fb ("random: vary jitter iterations based on cycle counter speed") Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> Tested-by: Vladimir Murzin <vladimir.murzin@arm.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Andy Shevchenko [Wed, 15 Jun 2022 11:47:46 +0000 (14:47 +0300)]
iio: proximity: sx_common: Allow IIO core to take care of firmware node
IIO core correctly will take care of firmware node if it's not set in
the driver. Drop ACPI and OF specifics from the driver and allow IIO
core to handle this.
Andy Shevchenko [Wed, 15 Jun 2022 11:47:45 +0000 (14:47 +0300)]
iio: proximity: sx_common: Don't use IIO device for properties
It's not correct to use artificial device created by IIO core to
retrieve device properties. Even ->get_default_reg() callback
takes a simple struct device pointer which suggests it wants to
operate on the real device.
Correct this by replacing pointer to IIO device by a real device
pointer in the caller of ->get_default_reg().
Dmitry Rokosov [Wed, 1 Jun 2022 17:48:32 +0000 (17:48 +0000)]
iio: trigger: move trig->owner init to trigger allocate() stage
To provide a new IIO trigger to the IIO core, usually driver executes the
following pipeline: allocate()/register()/get(). Before, IIO core assigned
trig->owner as a pointer to the module which registered this trigger at
the register() stage. But actually the trigger object is owned by the
module earlier, on the allocate() stage, when trigger object is
successfully allocated for the driver.
This patch moves trig->owner initialization from register()
stage of trigger initialization pipeline to allocate() stage to
eliminate all misunderstandings and time gaps between trigger object
creation and owner acquiring.
Currently the DT for QCS404 SoC has setup for 2 USB2 PHYs with one each
assigned to USB3 controller and USB2 controller. This assignment is
incorrect which only works by luck: as when each USB HCI comes up it
configures the *other* controllers PHY which is enough to make them
happy. If, for any reason, we were to disable one of the controllers then
both would stop working.
This was a difficult inconsistency to be caught which was found while
trying to enable USB support in u-boot. So with all the required drivers
ported to u-boot, I couldn't get the same USB storage device enumerated
in u-boot which was being enumerated fine by the kernel.
The root cause of the problem came out to be that I wasn't enabling USB2
PHY: "usb2_phy_prim" in u-boot. Then I realised that via simply disabling
the same USB2 PHY currently assigned to USB2 host controller in the
kernel disabled enumeration for USB3 host controller as well.
So fix this inconsistency by correctly assigning USB2 PHYs.
Fixes: 9375e7d719b3 ("arm64: dts: qcom: qcs404: Add USB devices and PHYs") Signed-off-by: Sumit Garg <sumit.garg@linaro.org> Reviewed-by: Daniel Thompson <daniel.thompson@linaro.org> Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org> Link: https://lore.kernel.org/r/20220711083038.1518529-1-sumit.garg@linaro.org
arm64: dts: qcom: sc7280: drop unused clocks from eDP node
The eDP node includes two clocks which are used by the eDP PHY rather
than eDP controller itself. Drop these clocks to remove extra difference
between eDP and DP controllers.
arm64: dts: qcom: sc7280: drop address/size-cells from eDP node
Drop #address/#size-cells from eDP device node. For eDP the panels are
not described directly under the controller node. They are either
present under aux-bus child node, or they are declared separately (e.g.
in a /soc node).
arm64: dts: qcom: sc7280: drop #clock-cells from displayport-controller
Drop #clock-cells from DP device node. It is a leftover from the times
before splitting the deviice into controller and PHY devices. Now the
clocks are provided by the PHY, while the controller doesn't provide any
clocks.
arm64: dts: qcom: sc7280: split register block for DP controller
Follow the schema for the DP controller and declare 5 register regions
instead of using a single region for all the registers. Note, this
extends the dts by adding p1 region to the DP node (to be used for DP
MST).
arm64: dts: qcom: sc7180: drop #clock-cells from displayport-controller
Drop #clock-cells from DP device node. It is a leftover from the times
before splitting the device into controller and PHY devices. Now the
clocks are provided by the PHY, while the controller doesn't provide any
clocks.
arm64: dts: qcom: sc7180: split register block for DP controller
Follow the schema for the DP controller and declare 5 register regions
instead of using a single region for all the registers. Note, this
extends the dts by adding p1 region to the DP node (to be used for DP
MST).
Marijn Suijten [Sat, 7 May 2022 15:36:27 +0000 (17:36 +0200)]
arm64: dts: qcom: msm8998: Make regulator voltages multiple of step-size
These voltages are not a multiple of the given step-size 8000 (with base
voltage 1664000) in pm8998_pldo, resulting in PLDO regulators l18 and
l22 failing to validate and in turn not probing the rpm-pm8998-regulator
driver:
Marijn Suijten [Sat, 7 May 2022 22:46:45 +0000 (00:46 +0200)]
arm64: dts: qcom: sm6350: Replace literal rpmhpd indices with constants
It seems the SM6350_CX definition was temporarily replaced with its
literal value 0 in 1797e1c9a95c ("arm64: dts: qcom: sm6350: Add SDHCI1/2
nodes") to prevent a dependency on the qcom-rpmpd.h header patch being
available prior to this DT patch being applied, similar to c23f1b77358c
("arm64: dts: qcom: sm6125: Avoid using missing SM6125_VDDCX").
However, unlike the revert of that in the sm6125 tree the next merge
window around in a90b8adfa2dd ("Revert "arm64: dts: qcom: sm6125: Avoid
using missing SM6125_VDDCX""), this has not yet happened for sm6350:
replace them back now that the definitions are definitely available.
Robert Marko [Thu, 7 Jul 2022 17:37:33 +0000 (19:37 +0200)]
arm64: dts: qcom: ipq8074: add APCS node
APCS now has support for providing the APSS clocks as the child device
for IPQ8074.
So, add the required DT node for it as it will later be used as the CPU
clocksource.
Signed-off-by: Robert Marko <robimarko@gmail.com> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
[bjorn: Sorted node based on address] Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org> Link: https://lore.kernel.org/r/20220707173733.404947-4-robimarko@gmail.com
Konrad Dybcio [Thu, 7 Jul 2022 21:21:58 +0000 (23:21 +0200)]
soc: qcom: Make QCOM_RPMPD depend on PM
QCOM_RPMPD requires PM_GENERIC_DOMAINS/_OF, which in turns requires
CONFIG_PM. I forgot about the latter in my earlier patch (it's still
in -next as of the time of committing, hence no Fixes: tag). Fix it.
coccinelle: Remove script that checks replacing 0/1 with false/true in functions returning bool
There is nothing wrong with current code that returns 0 or 1 for a
function returning bool. It is perfectly acceptable by the C standard.
To avoid churn of unwanted patches that are constantly sent to maintainers
who do not care about this change, remove the script that flags it as an
issue. This issue is not worth the burden on maintainers to accept
useless patches.
Now that all callers of ->llseek are going through vfs_llseek(), we
don't gain anything by keeping no_llseek around. Nothing actually calls
it and setting ->llseek to no_lseek is completely equivalent to
leaving it NULL.
Longer term (== by the end of merge window) we want to remove all such
intializations. To simplify the merge window this commit does *not*
touch initializers - it only defines no_llseek as NULL (and simplifies
the tests on file opening).
At -rc1 we'll need do a mechanical removal of no_llseek -
git grep -l -w no_llseek | grep -v porting.rst | while read i; do
sed -i '/\<no_llseek\>/d' $i
done
would do it.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
fs: check FMODE_LSEEK to control internal pipe splicing
The original direct splicing mechanism from Jens required the input to
be a regular file because it was avoiding the special socket case. It
also recognized blkdevs as being close enough to a regular file. But it
forgot about chardevs, which behave the same way and work fine here.
This is an okayish heuristic, but it doesn't totally work. For example,
a few chardevs should be spliceable here. And a few regular files
shouldn't. This patch fixes this by instead checking whether FMODE_LSEEK
is set, which represents decently enough what we need rewinding for when
splicing to internal pipes.
Fixes: b92ce5589374 ("[PATCH] splice: add direct fd <-> fd splicing support") Cc: Jens Axboe <axboe@kernel.dk> Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
This file does not support llseek, so don't set the flag advertising it.
Acked-by: Alex Williamson <alex.williamson@redhat.com> Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
This is already set by anon_inode_getfile(), since dma_buf_fops has
non-NULL ->llseek, so we don't need to set it here too.
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch> Acked-by: Sumit Semwal <sumit.semwal@linaro.org> Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Now vfs_llseek() can simply check for FMODE_LSEEK; if it's set,
we know that ->llseek() won't be NULL and if it's not we should
just fail with -ESPIPE.
A couple of other places where we used to check for special
values of ->llseek() (somewhat inconsistently) switched to
checking FMODE_LSEEK.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
fs: clear or set FMODE_LSEEK based on llseek function
Pipe-like behaviour on llseek(2) (i.e. unconditionally failing with
-ESPIPE) can be expresses in 3 ways:
1) ->llseek set to NULL in file_operations
2) ->llseek set to no_llseek in file_operations
3) FMODE_LSEEK *not* set in ->f_mode.
Enforce (3) in cases (1) and (2); that will allow to simplify the
checks and eventually get rid of no_llseek boilerplate.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
i2c: mlxcpld: Add callback to notify probing completion
Add notification to inform caller that driver probing has been
completed. It allows to user, invoked platform device registration for
"i2c-mlxcpld" driver, to be notified that bus adapter is available, and
thus some devices could be connected to this bus.
Signed-off-by: Vadim Pasternak <vadimp@nvidia.com> Signed-off-by: Wolfram Sang <wsa@kernel.org>
Robert Hancock [Tue, 14 Jun 2022 23:29:19 +0000 (17:29 -0600)]
i2c: cadence: Change large transfer count reset logic to be unconditional
Problems were observed on the Xilinx ZynqMP platform with large I2C reads.
When a read of 277 bytes was performed, the controller NAKed the transfer
after only 252 bytes were transferred and returned an ENXIO error on the
transfer.
There is some code in cdns_i2c_master_isr to handle this case by resetting
the transfer count in the controller before it reaches 0, to allow larger
transfers to work, but it was conditional on the CDNS_I2C_BROKEN_HOLD_BIT
quirk being set on the controller, and ZynqMP uses the r1p14 version of
the core where this quirk is not being set. The requirement to do this to
support larger reads seems like an inherently required workaround due to
the core only having an 8-bit transfer size register, so it does not
appear that this should be conditional on the broken HOLD bit quirk which
is used elsewhere in the driver.
Remove the dependency on the CDNS_I2C_BROKEN_HOLD_BIT for this transfer
size reset logic to fix this problem.
Fixes: 63cab195bf49 ("i2c: removed work arounds in i2c driver for Zynq Ultrascale+ MPSoC") Signed-off-by: Robert Hancock <robert.hancock@calian.com> Reviewed-by: Shubhrajyoti Datta <Shubhrajyoti.datta@amd.com> Acked-by: Michal Simek <michal.simek@amd.com> Signed-off-by: Wolfram Sang <wsa@kernel.org>
Ming Lei [Sat, 16 Jul 2022 09:53:44 +0000 (17:53 +0800)]
ublk_drv: fix build warning with -Wmaybe-uninitialized and one sparse warning
After applying -Wmaybe-uninitialized manually, two build warnings are
triggered:
drivers/block/ublk_drv.c:940:11: warning: ‘io’ may be used uninitialized [-Wmaybe-uninitialized]
940 | io->flags &= ~UBLK_IO_FLAG_ACTIVE;
drivers/block/ublk_drv.c: In function ‘ublk_ctrl_uring_cmd’:
drivers/block/ublk_drv.c:1531:9: warning: ‘ret’ may be used uninitialized [-Wmaybe-uninitialized]
Fix the 1st one by removing 'io->flags &= ~UBLK_IO_FLAG_ACTIVE;' which
isn't needed since the function always return successfully after setting
this flag.
Fix the 2nd one by always initializing 'ret'.
Also fix another sparse warning of 'sparse: sparse: incorrect type in return
expression' by changing return type of ublk_setup_iod().
projects / people / ms / linux.git / log
