| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
struct platform_driver::remove returning an integer made driver authors
expect that returning an error code was proper error handling. However
the driver core ignores the error and continues to remove the device
because there is nothing the core could do anyhow and reentering the
remove callback again is only calling for trouble.
To prevent such wrong assumptions, change the return type of the remove
callback to void. This was prepared by introducing an alternative remove
callback returning void and converting all drivers to that. So .remove()
can be changed without further changes in drivers.
This corresponds to step b) of the plan outlined in commit
5c5a7680e67b ("platform: Provide a remove callback that returns no value").
Signed-off-by: Uwe Kleine-König <[email protected]>
|
|
The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is ignored (apart
from emitting a warning) and this typically results in resource leaks.
To improve here there is a quest to make the remove callback return
void. In the first step of this quest all drivers are converted to
.remove_new(), which already returns void. Eventually after all drivers
are converted, .remove_new() will be renamed to .remove().
Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.
Reviewed-by: Dave Jiang <[email protected]>
Link: https://lore.kernel.org/r/8de0900f8c9f40648295fd9e2f445c85b2593d26.1712756722.git.u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <[email protected]>
|
|
The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is ignored (apart
from emitting a warning) and this typically results in resource leaks.
To improve here there is a quest to make the remove callback return
void. In the first step of this quest all drivers are converted to
.remove_new(), which already returns void. Eventually after all drivers
are converted, .remove_new() will be renamed to .remove().
Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.
Reviewed-by: Dave Jiang <[email protected]>
Link: https://lore.kernel.org/r/fcb5545d45cf31caee31e0c66ed3521ead12c9b4.1712756722.git.u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <[email protected]>
|
|
The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is ignored (apart
from emitting a warning) and this typically results in resource leaks.
To improve here there is a quest to make the remove callback return
void. In the first step of this quest all drivers are converted to
.remove_new(), which already returns void. Eventually after all drivers
are converted, .remove_new() will be renamed to .remove().
Trivially convert the ipu-v3 platform drivers from always returning zero
in the remove callback to the void returning variant.
Reviewed-by: Philipp Zabel <[email protected]>
Reviewed-by: Thomas Zimmermann <[email protected]>
Link: https://lore.kernel.org/r/dee3c0e1c8c6bd1027a91c65be55ac1d6ba9e099.1712681770.git.u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <[email protected]>
|
|
The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is ignored (apart
from emitting a warning) and this typically results in resource leaks.
To improve here there is a quest to make the remove callback return
void. In the first step of this quest all drivers are converted to
.remove_new(), which already returns void. Eventually after all drivers
are converted, .remove_new() will be renamed to .remove().
Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.
Acked-by: Thierry Reding <[email protected]>
Reviewed-by: Thomas Zimmermann <[email protected]>
Link: https://lore.kernel.org/r/7e31909b1e536f0ddbb060b1aaa0a9e943687c8a.1712681770.git.u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <[email protected]>
|
|
The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is ignored (apart
from emitting a warning) and this typically results in resource leaks.
To improve here there is a quest to make the remove callback return
void. In the first step of this quest all drivers are converted to
.remove_new(), which already returns void. Eventually after all drivers
are converted, .remove_new() will be renamed to .remove().
Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.
Reviewed-by: Thomas Zimmermann <[email protected]>
Link: https://lore.kernel.org/r/4a64dfbfbcfdf9b7cd46bc8026223e69a4b453b4.1712681770.git.u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <[email protected]>
|
|
The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is ignored (apart
from emitting a warning) and this typically results in resource leaks.
To improve here there is a quest to make the remove callback return
void. In the first step of this quest all drivers are converted to
.remove_new(), which already returns void. Eventually after all drivers
are converted, .remove_new() will be renamed to .remove().
Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.
Reviewed-by: Matt Coster <[email protected]>
Acked-by: Matt Coster <[email protected]>
Reviewed-by: Thomas Zimmermann <[email protected]>
Link: https://lore.kernel.org/r/4cf10e420863f40a268f26b9bdb0c4b53dbf3406.1712681770.git.u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <[email protected]>
|
|
When using devm_clk_get_prepared() instead of devm_clk_get() the clock
is already returned prepared. So probe doesn't need to call
clk_prepare() and at remove time the call to clk_unprepare() can be
dropped. The latter makes the remove callback empty, so it can be
dropped, too.
Acked-by: Thierry Reding <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Uwe Kleine-König <[email protected]>
|
|
The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is ignored (apart
from emitting a warning) and this typically results in resource leaks.
To improve here there is a quest to make the remove callback return
void. In the first step of this quest all drivers are converted to
.remove_new(), which already returns void. Eventually after all drivers
are converted, .remove_new() will be renamed to .remove().
Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.
Acked-by: Rodolfo Giometti <[email protected]>
Link: https://lore.kernel.org/r/f4b9402af72e5f285c8b0f068076a76418f653f5.1709886922.git.u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <[email protected]>
|
|
The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is ignored (apart
from emitting a warning) and this typically results in resource leaks.
To improve here there is a quest to make the remove callback return
void. In the first step of this quest all drivers are converted to
.remove_new(), which already returns void. Eventually after all drivers
are converted, .remove_new() will be renamed to .remove().
Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.
Link: https://lists.ozlabs.org/pipermail/linux-fsi/2024-March/000613.html
Acked-by: Jeremy Kerr <[email protected]>
Signed-off-by: Uwe Kleine-König <[email protected]>
|
|
The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is ignored (apart
from emitting a warning) and this typically results in resource leaks.
To improve here there is a quest to make the remove callback return
void. In the first step of this quest all drivers are converted to
.remove_new(), which already returns void. Eventually after all drivers
are converted, .remove_new() will be renamed to .remove().
Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.
Acked-by: Jeremy Kerr <[email protected]>
Link: https://lists.ozlabs.org/pipermail/linux-fsi/2024-March/000612.html
Signed-off-by: Uwe Kleine-König <[email protected]>
|
|
The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is ignored (apart
from emitting a warning) and this typically results in resource leaks.
To improve here there is a quest to make the remove callback return
void. In the first step of this quest all drivers are converted to
.remove_new(), which already returns void. Eventually after all drivers
are converted, .remove_new() will be renamed to .remove().
Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.
Acked-by: Jeremy Kerr <[email protected]>
Link: https://lists.ozlabs.org/pipermail/linux-fsi/2024-March/000614.html
Signed-off-by: Uwe Kleine-König <[email protected]>
|
|
The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is ignored (apart
from emitting a warning) and this typically results in resource leaks.
To improve here there is a quest to make the remove callback return
void. In the first step of this quest all drivers are converted to
.remove_new(), which already returns void. Eventually after all drivers
are converted, .remove_new() will be renamed to .remove().
Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.
Acked-by: Jeremy Kerr <[email protected]>
Link: https://lore.kernel.org/r/de0f2d4cb529a433d4620ca0e8fda0dfb1e950db.1709673414.git.u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <[email protected]>
|
|
Philipp didn't come around to send a pull request in the last
development cycle. His next branch only contains patches converting
three platform drivers to stop using .remove().
They are merged with his consent to be able to change the prototype of
.remove() in the upcoming development cycle.
Link: https://lore.kernel.org/all/[email protected]/
|
|
Commit b8b8b4e0c052 ("ata: ahci: Add Intel Alder Lake-P AHCI controller
to low power chipsets list") added Intel Alder Lake to the ahci_pci_tbl.
Because of the way that the Intel PCS quirk was implemented, having
an explicit entry in the ahci_pci_tbl caused the Intel PCS quirk to
be applied. (The quirk was not being applied if there was no explict
entry.)
Thus, entries that were added to the ahci_pci_tbl also got the Intel
PCS quirk applied.
The quirk was cleaned up in commit 7edbb6059274 ("ahci: clean up
intel_pcs_quirk"), such that it is clear which entries that actually
applies the Intel PCS quirk.
Newer Intel AHCI controllers do not need the Intel PCS quirk,
and applying it when not needed actually breaks some platforms.
Do not apply the Intel PCS quirk for Intel Alder Lake.
This is in line with how things worked before commit b8b8b4e0c052 ("ata:
ahci: Add Intel Alder Lake-P AHCI controller to low power chipsets list"),
such that certain platforms using Intel Alder Lake will work once again.
Cc: [email protected] # 6.7
Fixes: b8b8b4e0c052 ("ata: ahci: Add Intel Alder Lake-P AHCI controller to low power chipsets list")
Signed-off-by: Jason Nader <[email protected]>
Signed-off-by: Niklas Cassel <[email protected]>
|
|
This is an effort to get rid of all multiplications from allocation
functions in order to prevent integer overflows [1][2].
The "struct dma_fence_array" can be refactored to add a flex array in order
to have the "callback structures allocated behind the array" be more
explicit.
Do so:
- makes the code more readable and safer.
- allows using __counted_by() for additional checks
- avoids some pointer arithmetic in dma_fence_array_enable_signaling()
Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#open-coded-arithmetic-in-allocator-arguments [1]
Link: https://github.com/KSPP/linux/issues/160 [2]
Signed-off-by: Christophe JAILLET <[email protected]>
Reviewed-by: Kees Cook <[email protected]>
Reviewed-by: Christian König <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/8b4e556e07b5dd78bb8a39b67ea0a43b199083c8.1716652811.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Christian König <[email protected]>
|
|
https://gitlab.freedesktop.org/drm/misc/kernel into drm-fixes
Short summary of fixes pull:
buddy:
- stop using PAGE_SIZE
shmem-helper:
- avoid kernel panic in mmap()
tests:
- buddy: fix PAGE_SIZE dependency
Signed-off-by: Dave Airlie <[email protected]>
From: Thomas Zimmermann <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
|
|
Fix the following from 'make W=1' with allmodconfig:
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/clk/qcom/clk-qcom.o
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/clk/qcom/gcc-msm8976.o
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/clk/qcom/lpass-gfm-sm8250.o
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/clk/qcom/videocc-sdm845.o
Signed-off-by: Jeff Johnson <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Bjorn Andersson <[email protected]>
|
|
After commit 2fbe479c0024 ("platform/chrome: cros_ec: Handle events
during suspend after resume completion"), cros_ec_resume() starts to
call cros_ec_resume_early() and cros_ec_resume_complete().
To be neat, let cros_ec_suspend() call cros_ec_suspend_prepare() and
cros_ec_suspend_late().
Reviewed-by: Prashant Malani <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Tzung-Bi Shih <[email protected]>
|
|
Mark the documentation block as kernel doc and drop the documentation of
the non-existing parameter "@msg".
Signed-off-by: Thomas Weißschuh <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Tzung-Bi Shih <[email protected]>
|
|
Mark the documentation block as kernel doc and drop the documentation of
the non-existing parameter "@msg".
Signed-off-by: Thomas Weißschuh <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Tzung-Bi Shih <[email protected]>
|
|
The refactoring makes code less verbose and easier to read.
Besides that the binary size is also reduced, which sounds
like a win-win case:
add/remove: 0/1 grow/shrink: 2/2 up/down: 210/-226 (-16)
Function old new delta
spi_destroy_queue 42 156 +114
spi_controller_suspend 101 197 +96
spi_unregister_controller 346 319 -27
spi_register_controller 1834 1794 -40
spi_stop_queue 159 - -159
Total: Before=49230, After=49214, chg -0.03%
Signed-off-by: Andy Shevchenko <[email protected]>
Link: https://msgid.link/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
Since driver core provides a generic device_match_acpi_handle()
we may replace the custom code with it.
Signed-off-by: Andy Shevchenko <[email protected]>
Link: https://msgid.link/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
Add check for the return value of clk_prepare() and return the error if
it fails in order to catch the error.
Fixes: 4a2f83b7f780 ("spi: atmel-quadspi: add runtime pm support")
Signed-off-by: Chen Ni <[email protected]>
Link: https://msgid.link/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
Remove the empty spi_imx_cleanup function.
It's ok if a driver does not set the controller->cleanup pointer, the
caller does a NULL check.
Signed-off-by: Martin Kaiser <[email protected]>
Link: https://msgid.link/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
The AXI SPI Engine IP core v1.2 added support for SPI_CS_HIGH. This
provides the driver implementation to make use of this feature when
supported hardware is detected.
Signed-off-by: David Lechner <[email protected]>
Link: https://msgid.link/r/20240508-spi-axi-spi-engine-add-spi_cs_high-support-v1-1-695dd8e45f00@baylibre.com
Signed-off-by: Mark Brown <[email protected]>
|
|
Instead of 'if (!ret)' switch to "check for the error first" rule.
Signed-off-by: Andy Shevchenko <[email protected]>
Link: https://msgid.link/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
Convert the module to be property provider agnostic and allow
it to be used on non-OF platforms.
Include mod_devicetable.h explicitly to replace the dropped of.h
which included mod_devicetable.h indirectly.
Signed-off-by: Andy Shevchenko <[email protected]>
Link: https://msgid.link/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
Simple type conversion with no functional change implied.
While at it, adjust indentation where it makes sense.
Signed-off-by: Andy Shevchenko <[email protected]>
Link: https://msgid.link/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
With a typedef for the txrx_*() callbacks the code looks neater.
Note that typedef for a function is okay to have.
Signed-off-by: Andy Shevchenko <[email protected]>
Link: https://msgid.link/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
The SPI "hard" controller within the PolarFire SoC is capable of
handling eight CS lines, but only one CS line is wired. Therefore, use
GPIO descriptors to configure additional CS lines.
Signed-off-by: Prajna Rajendra Kumar <[email protected]>
Link: https://msgid.link/r/[email protected]
Acked-by: Conor Dooley <[email protected]>
Signed-off-by: Mark Brown <[email protected]>
|
|
The SPI "hard" controller in PolarFire SoC has eight CS lines, but only
one CS line is wired. When the 'num-cs' property is not specified in
the device tree, the driver defaults to the MAX_CS value, which has
been fixed to 1 to match the hardware configuration; however, when the
'num-cs' property is explicitly defined in the device tree, it
overrides the default value.
Fixes: 9ac8d17694b6 ("spi: add support for microchip fpga spi controllers")
Signed-off-by: Prajna Rajendra Kumar <[email protected]>
Reviewed-by: Conor Dooley <[email protected]>
Link: https://msgid.link/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
When the requested dirmap accesses are outside of the window, it is
probably more sensible to return -EINVAL rather than an "unsupported"
error code. If however the operation in itself is not supported, then
-EOPNOTSUP is likely going to be preferred as it is a standard error
code.
>From a caller (and reviewer) point of view, distinguising between the
two may be helpful because somehow one can be "fixed" while the other
will always be refused no matter how hard we try.
As part of a wider work to bring spi-nand continuous reads, it was
useful to easily catch the upper limit direct mapping boundaries for
each controller, with the idea of enlarging this area from a page to an
eraseblock, without risking too many regressions.
Signed-off-by: Miquel Raynal <[email protected]>
Link: https://msgid.link/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
If the request is out of range, returning -EINVAL seems a better pick
than -ENOTSUPP.
>From a caller (and reviewer) point of view, distinguising between the
two may be helpful because somehow one can be "fixed" while the other
will always be refused no matter how hard we try.
As part of a wider work to bring spi-nand continuous reads, it was
useful to easily catch the upper limit direct mapping boundaries for
each controller, with the idea of enlarging this area from a page to an
eraseblock, without risking too many regressions.
In all other cases, as part of a wider work towards using -EOPNOTSUP
rather than -ENOTSUPP (which is not a SUSV4 code), let's change the
error code to be uniform across spi-mem controller drivers.
Finally, reword a little bit the conditions to clarify what is intended
(ie. checking for the presence of a direct mapping, and also ensuring we
create a dirmap only on DATA_IN flows).
Signed-off-by: Miquel Raynal <[email protected]>
Link: https://msgid.link/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
If the request is out of range, returning -EINVAL seems
sensible. However if there is no direct mapping available (which is a
possible case), no direct mapping will ever be allowed, hence -EOPNOTSUP
is probably more relevant in this case.
>From a caller (and reviewer) point of view, distinguising between the
two may be helpful because somehow one can be "fixed" while the other
will always be refused no matter how hard we try.
As part of a wider work to bring spi-nand continuous reads, it was
useful to easily catch the upper limit direct mapping boundaries for
each controller, with the idea of enlarging this area from a page to an
eraseblock, without risking too many regressions.
Signed-off-by: Miquel Raynal <[email protected]>
Link: https://msgid.link/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
The driver does not support dirmap write operations, return -EOPTNOTSUPP
in this case.
Most controllers have a maximum linear mapping area. Requests beyond
this limit can be considered invalid, rather than unsupported.
>From a caller (and reviewer) point of view, distinguising between the
two may be helpful because somehow one can be "fixed" while the other
will always be refused no matter how hard we try.
As part of a wider work to bring spi-nand continuous reads, it was
useful to easily catch the upper limit direct mapping boundaries for
each controller, with the idea of enlarging this area from a page to an
eraseblock, without risking too many regressions.
Signed-off-by: Miquel Raynal <[email protected]>
Link: https://msgid.link/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
Add optional reset control support for spi-cadence to properly bring
the SPI device into an operating condition.
Co-developed-by: Eng Lee Teh <[email protected]>
Signed-off-by: Eng Lee Teh <[email protected]>
Co-developed-by: Ley Foon Tan <[email protected]>
Signed-off-by: Ley Foon Tan <[email protected]>
Signed-off-by: Ji Sheng Teoh <[email protected]>
Link: https://msgid.link/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
to 0
These drivers don't use the driver_data member of struct i2c_device_id,
so don't explicitly initialize this member.
This prepares putting driver_data in an anonymous union which requires
either no initialization or named designators. But it's also a nice
cleanup on its own.
While add it, also remove commas after the sentinel entries.
Signed-off-by: Uwe Kleine-König <[email protected]>
Link: https://msgid.link/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
Add support of the new compatible "st,stm32mp13-pwr-reg" for STM32MP13.
Signed-off-by: Patrick Delaunay <[email protected]>
Link: https://msgid.link/r/20240523200353.v4.2.I04ec53442753147c35efad1307b6ec133f53b471@changeid
Signed-off-by: Mark Brown <[email protected]>
|
|
- Use kunit_kcalloc() to allocate the defaults table so that it will be
freed when the test case ends.
- kfree() the buf and *data buffers on the error paths.
- Use kunit_add_action_or_reset() instead of kunit_add_action() so that
if it fails it will call regmap_exit().
Signed-off-by: Richard Fitzgerald <[email protected]>
Link: https://msgid.link/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
Currently, when an adapter defines a max_write_len quirk,
the data will be chunked into data sizes equal to the
max_write_len quirk value. But the payload will be increased by
the size of the register address before transmission. The
resulting value always ends up larger than the limit set
by the quirk.
Avoid this error by setting regmap's max_write to the quirk's
max_write_len minus the number of bytes for the register and
padding. This allows the chunking to work correctly for this
limited case without impacting other use-cases.
Signed-off-by: Jim Wylder <[email protected]>
Link: https://msgid.link/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
|
|
Assign the correct name to ID 82 and fix the ID of SMB2360.
Fixes: e025171d1ab1 ("soc: qcom: socinfo: Add SMB2360 PMIC")
Signed-off-by: Konrad Dybcio <[email protected]>
Reviewed-by: Neil Armstrong <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Bjorn Andersson <[email protected]>
|
|
As with previous platforms, qseecom and the uefisecapp provides access
to EFI variables. Add X1E CRD and QCP devices to the allowlist.
Signed-off-by: Bjorn Andersson <[email protected]>
Reviewed-by: Konrad Dybcio <[email protected]>
Tested-by: Konrad Dybcio <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Bjorn Andersson <[email protected]>
|
|
Fix the 'make W=1' warnings:
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/firewire/uapi-test.o
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/firewire/packet-serdes-test.o
Signed-off-by: Jeff Johnson <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Sakamoto <[email protected]>
|
|
A potential deadlock was reported with the config file at
https://web.archive.org/web/20240522052129/https://0x0.st/XPN_.txt
In this particular configuration, the deadlock doesn't exist because
the warning triggered at a point before modules were even available.
However, the deadlock can be real because any module loaded would
invoke async_synchronize_full.
The issue is spurious for software crypto algorithms which aren't
themselves involved in async probing. However, it would be hard to
avoid for a PCI crypto driver using async probing.
In this particular call trace, the problem is easily avoided because
the only reason the module is being requested during probing is the
add_early_randomness call in the hwrng core. This feature is
vestigial since there is now a kernel thread dedicated to doing
exactly this.
So remove add_early_randomness as it is no longer needed.
Reported-by: Nícolas F. R. A. Prado <[email protected]>
Reported-by: Eric Biggers <[email protected]>
Fixes: 1b6d7f9eb150 ("tpm: add session encryption protection to tpm2_get_random()")
Link: https://lore.kernel.org/r/119dc5ed-f159-41be-9dda-1a056f29888d@notapiano/
Signed-off-by: Herbert Xu <[email protected]>
Reviewed-by: Jarkko Sakkinen <[email protected]>
Tested-by: Nícolas F. R. A. Prado <[email protected]>
Signed-off-by: Herbert Xu <[email protected]>
|
|
Pull ipmi updates from Corey Minyard:
"Mostly updates for deprecated interfaces, platform.remove and
converting from a tasklet to a BH workqueue.
Also use HAS_IOPORT for disabling inb()/outb()"
* tag 'for-linus-6.10-1' of https://github.com/cminyard/linux-ipmi:
ipmi: kcs_bmc_npcm7xx: Convert to platform remove callback returning void
ipmi: kcs_bmc_aspeed: Convert to platform remove callback returning void
ipmi: ipmi_ssif: Convert to platform remove callback returning void
ipmi: ipmi_si_platform: Convert to platform remove callback returning void
ipmi: ipmi_powernv: Convert to platform remove callback returning void
ipmi: bt-bmc: Convert to platform remove callback returning void
char: ipmi: handle HAS_IOPORT dependencies
ipmi: Convert from tasklet to BH workqueue
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/abelloni/linux
Pull RTC updates from Alexandre Belloni:
"There is one new driver and then most of the changes are the device
tree bindings conversions to yaml.
New driver:
- Epson RX8111
Drivers:
- Many Device Tree bindings conversions to dtschema
- pcf8563: wakeup-source support"
* tag 'rtc-6.10' of git://git.kernel.org/pub/scm/linux/kernel/git/abelloni/linux:
pcf8563: add wakeup-source support
rtc: rx8111: handle VLOW flag
rtc: rx8111: demote warnings to debug level
rtc: rx6110: Constify struct regmap_config
dt-bindings: rtc: convert trivial devices into dtschema
dt-bindings: rtc: stmp3xxx-rtc: convert to dtschema
dt-bindings: rtc: pxa-rtc: convert to dtschema
rtc: Add driver for Epson RX8111
dt-bindings: rtc: Add Epson RX8111
rtc: mcp795: drop unneeded MODULE_ALIAS
rtc: nuvoton: Modify part number value
rtc: test: Split rtc unit test into slow and normal speed test
dt-bindings: rtc: nxp,lpc1788-rtc: convert to dtschema
dt-bindings: rtc: digicolor-rtc: move to trivial-rtc
dt-bindings: rtc: alphascale,asm9260-rtc: convert to dtschema
dt-bindings: rtc: armada-380-rtc: convert to dtschema
rtc: cros-ec: provide ID table for avoiding fallback match
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/i3c/linux
Pull i3c updates from Alexandre Belloni:
"Runtime PM (power management) is improved and hot-join support has
been added to the dw controller driver.
Core:
- Allow device driver to trigger controller runtime PM
Drivers:
- dw: hot-join support
- svc: better IBI handling"
* tag 'i3c/for-6.10' of git://git.kernel.org/pub/scm/linux/kernel/git/i3c/linux:
i3c: dw: Add hot-join support.
i3c: master: Enable runtime PM for master controller
i3c: master: svc: fix invalidate IBI type and miss call client IBI handler
i3c: master: svc: change ENXIO to EAGAIN when IBI occurs during start frame
i3c: Add comment for -EAGAIN in i3c_device_do_priv_xfers()
|
|
When running Cilium connectivity test suite with netkit in L2 mode, we
found that compared to tcx a few tests were failing which pushed traffic
into an L7 proxy sitting in host namespace. The problem in particular is
around the invocation of eth_type_trans() in netkit.
In case of tcx, this is run before the tcx ingress is triggered inside
host namespace and thus if the BPF program uses the bpf_skb_change_type()
helper the newly set type is retained. However, in case of netkit, the
late eth_type_trans() invocation overrides the earlier decision from the
BPF program which eventually leads to the test failure.
Instead of eth_type_trans(), split out the relevant parts, meaning, reset
of mac header and call to eth_skb_pkt_type() before the BPF program is run
in order to have the same behavior as with tcx, and refactor a small helper
called eth_skb_pull_mac() which is run in case it's passed up the stack
where the mac header must be pulled. With this all connectivity tests pass.
Fixes: 35dfaad7188c ("netkit, bpf: Add bpf programmable net device")
Signed-off-by: Daniel Borkmann <[email protected]>
Acked-by: Nikolay Aleksandrov <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Alexei Starovoitov <[email protected]>
|