The clock dividers might be used with clock stop bit enabled or not.
Current tables only support recommended values from the datasheet. This
might result in warnings like below because no valid clock divider is
found. Resulting in a 0 divider.
There are Renesas ARM Trusted Firmware version out there which e.g.
configure 0x201 (shifted logical right by 2: 0x80) and with this match
the added { STPnHCK | 0, 1 }:
https://github.com/renesas-rcar/arm-trusted-firmware/blob/rcar_gen3_v2.3/drivers/renesas/rcar/emmc/emmc_init.c#L108
------------[ cut here ]------------
sd1h: Zero divisor and CLK_DIVIDER_ALLOW_ZERO not set
WARNING: CPU: 1 PID: 1 at drivers/clk/clk-divider.c:141 divider_recalc_rate+0x48/0x70
Modules linked in:
CPU: 1 PID: 1 Comm: swapper/0 Not tainted 6.1.52 #1
Hardware name: Custom board based on r8a7796 (DT)
pstate: 40000005 (nZcv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
pc : divider_recalc_rate+0x48/0x70
...
------------[ cut here ]------------
Fixes: bb6d3fa98a ("clk: renesas: rcar-gen3: Switch to new SD clock handling")
Signed-off-by: Dirk Behme <dirk.behme@de.bosch.com>
[wsa: extended the table to 5 entries, added comments, reword commit message a little]
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Tested-by: Dirk Behme <dirk.behme@de.bosch.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/20230928080317.28224-1-wsa+renesas@sang-engineering.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Since commit 636b21b501 ("PCI: Revoke mappings like devmem"), mmappable
sysfs entries have started to receive their f_mapping from the iomem
pseudo filesystem, so that CONFIG_IO_STRICT_DEVMEM is honored in sysfs
(and procfs) as well as in /dev/[k]mem.
This resulted in a userspace-visible regression:
1. Open a sysfs PCI resource file (eg. /sys/bus/pci/devices/*/resource0)
2. Use lseek(fd, 0, SEEK_END) to determine its size
Expected result: a PCI region size is returned.
Actual result: 0 is returned.
The reason is that PCI resource files residing in sysfs use
generic_file_llseek(), which relies on f_mapping->host inode to get the
file size. As f_mapping is now redefined, f_mapping->host points to an
anonymous zero-sized iomem_inode which has nothing to do with sysfs file
in question.
Implement a custom llseek method for sysfs PCI resources, which is
almost the same as proc_bus_pci_lseek() used for procfs entries.
This makes sysfs and procfs entries consistent with regards to seeking,
but also introduces userspace-visible changes to seeking PCI resources
in sysfs:
- SEEK_DATA and SEEK_HOLE are no longer supported;
- Seeking past the end of the file is prohibited while previously
offsets up to MAX_NON_LFS were accepted (reading from these offsets
was always invalid).
Signed-off-by: Valentine Sinitsyn <valesini@yandex-team.ru>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://lore.kernel.org/r/20230925084013.309399-2-valesini@yandex-team.ru
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
As of now, seeking in sysfs files is handled by generic_file_llseek().
There are situations where one may want to customize seeking logic:
- Many sysfs entries are fixed files while generic_file_llseek() accepts
past-the-end positions. Not only being useless by itself, this
also means a bug in userspace code will trigger not at lseek(), but at
some later point making debugging harder.
- generic_file_llseek() relies on f_mapping->host to get the file size
which might not be correct for all sysfs entries.
See commit 636b21b501 ("PCI: Revoke mappings like devmem") as an example.
Implement llseek method to override this behavior at sysfs attribute
level. The method is optional, and if it is absent,
generic_file_llseek() is called to preserve backwards compatibility.
Signed-off-by: Valentine Sinitsyn <valesini@yandex-team.ru>
Link: https://lore.kernel.org/r/20230925084013.309399-1-valesini@yandex-team.ru
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
While searching for possible refactor of napi_schedule_prep and
__napi_schedule it was notice that the mtk eth driver disable the
interrupt for rx and tx AFTER napi is scheduled.
While this is a very hard to repro case it might happen to have
situation where the interrupt is disabled and never enabled again as the
napi completes and the interrupt is enabled before.
This is caused by the fact that a napi driven by interrupt expect a
logic with:
1. interrupt received. napi prepared -> interrupt disabled -> napi
scheduled
2. napi triggered. ring cleared -> interrupt enabled -> wait for new
interrupt
To prevent this case, disable the interrupt BEFORE the napi is
scheduled.
Fixes: 656e705243 ("net-next: mediatek: add support for MT7623 ethernet")
Cc: stable@vger.kernel.org
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Link: https://lore.kernel.org/r/20231002140805.568-1-ansuelsmth@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
This allows splicing zeroed pages into a pipe, and allows discarding
pages from a pipe by splicing them to /dev/zero. Writing to /dev/zero
should have the same effect as writing to /dev/null, and a
"splice_write" implementation exists only for /dev/null.
(The /dev/zero splice_read implementation could be optimized by
pushing references to the global zero page to the pipe, but that's an
optimization for another day.)
Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Link: https://lore.kernel.org/r/20230919073743.1066313-1-max.kellermann@ionos.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Prepare for the coming implementation by GCC and Clang of the __counted_by
attribute. Flexible array members annotated with __counted_by can have
their accesses bounds-checked at run-time checking via CONFIG_UBSAN_BOUNDS
(for array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family
functions).
As found with Coccinelle[1], add __counted_by for struct hpets.
[1] https://github.com/kees/kernel-tools/blob/trunk/coccinelle/examples/counted_by.cocci
Cc: Clemens Ladisch <clemens@ladisch.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: "Gustavo A. R. Silva" <gustavoars@kernel.org>
Link: https://lore.kernel.org/r/20230922175348.work.056-kees@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The word "wich"
was corrected to "which" for spelling accuracy.
Signed-off-by: Kuan-Wei Chiu <visitorckw@gmail.com>
Acked-by: Carlos Llamas <cmllamas@google.com>
Link: https://lore.kernel.org/r/20230903204250.2697370-1-visitorckw@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Prepare for the coming implementation by GCC and Clang of the __counted_by
attribute. Flexible array members annotated with __counted_by can have
their accesses bounds-checked at run-time checking via CONFIG_UBSAN_BOUNDS
(for array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family
functions).
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Reviewed-by: Kees Cook <keescook@chromium.org>
Reviewed-by: "Gustavo A. R. Silva" <gustavoars@kernel.org>
Link: https://lore.kernel.org/r/5c3b7459b820e22e2ac6ce892d4aadcc119cc919.1696065263.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In a future patch HAS_IOPORT=n will result in inb()/outb() and friends
not being declared. We thus need to add HAS_IOPORT as dependency for
those drivers using them.
This was previously done in commit b5c75b68b7 ("comedi: add HAS_IOPORT
dependencies"), but that has been reverted because it made it impossible
to select configuration options for several comedi drivers. This is a
do-over that avoids that.
Since the original patch, modifications have been made to various comedi
modules so that they can still be built even if the port I/O functions
have not been declared, so the configuration options for building those
modules no longer need to depend on HAS_IOPORT.
Make the COMEDI_ISA_DRIVERS menu option (which allows configuration
options for ISA and PC/104 drivers to be selected) depend on HAS_IOPORT,
and also depend on ISA || ISA_BUS || PC104.
Co-developed-by: Arnd Bergmann <arnd@kernel.org>
Signed-off-by: Arnd Bergmann <arnd@kernel.org>
Co-developed-by: Niklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: Niklas Schnelle <schnelle@linux.ibm.com>
Cc: Arnd Bergmann <arnd@kernel.org>
Cc: Niklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Link: https://lore.kernel.org/r/20230913170712.111719-14-abbotti@mev.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In a future patch, the port I/O functions (`inb()`, `outb()`, and
friends will only be declared in the `HAS_IOPORT` configuration option
is enabled.
The amplc_dio200_common module is used by the amplc_dio200 module (for
ISA cards) and the amplc_dio200_pci module (for PCI and PCI Express
cards). It supports both port I/O and memory-mapped I/O.
Port I/O and memory-mapped I/O is confined to the `dio200___read8()`,
`dio200___read32()`, `dio200___write8()` and `dio200___write32()`
functions. Conditionally compile two versions of those functions. If
the `CONFIG_HAS_IOPORT` macro is defined, call either the port I/O or
memory mapped I/O functions depending on the `mmio` member of the
`struct comedi_device`. If the `CONFIG_HAS_IOPORT` macro is undefined
only call the memory-mapped I/O functions.
Add a run-time check to `amplc_dio200_common_attach()` to return an
error if the device wants to use port I/O when the `CONFIG_HAS_IOPORT`
macro is undefined.
The changes allow the module to be built even if the port I/O functions
have not been declared.
Cc: Arnd Bergmann <arnd@kernel.org>
Cc: Niklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Link: https://lore.kernel.org/r/20230913170712.111719-13-abbotti@mev.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The `dio200_read8()`, `dio200_write8()`, `dio200_read32()` and
`dio200_write32()` functions apply a right-shift to the register offset
for some devices and then perform the actual register access. Factor
the register access part out to new functions `dio200___read8()`,
`dio200___write8()`, `dio200___read32()`, and `dio200___write32()`.
This will reduce duplicated code in a subsequent patch that will
conditionally compile support for port I/O as part of the `HAS_IOPORT`
changes.
Cc: Arnd Bergmann <arnd@kernel.org>
Cc: Niklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Link: https://lore.kernel.org/r/20230913170712.111719-12-abbotti@mev.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In a future patch, the port I/O functions (`inb()`, `outb()`, and
friends will only be declared in the `HAS_IOPORT` configuration option
is enabled.
The amplc_dio200_pci module supports various Amplicon PCI and PCI
Express devices. Some of the supported devices (the PCI ones) use port
I/O, and some of them (the PCIe ones) only use memory-mapped I/O.
Conditionally compile in support for the devices that need port I/O if
and only if the `CONFIG_HAS_IOPORT` macro is defined.
Add a run-time check in `dio200_pci_auto_attach()` to return an error if
the device actually requires port I/O (based on the PCI BAR resource
flags) but the `HAS_IOPORT` configuration option is not enabled.
Cc: Arnd Bergmann <arnd@kernel.org>
Cc: Niklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Link: https://lore.kernel.org/r/20230913170712.111719-11-abbotti@mev.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In a future patch, the port I/O functions (`inb()`, `outb()`, and
friends will only be declared in the `HAS_IOPORT` configuration option
is enabled.
The "ni_mio_common.c" file contains calls to both port I/O functions and
memory-mapped I/O functions. The file is `#include`d by "ni_atmio.c",
"ni_mio_cs.c", and "ni_pcimio.c" for the ni_atmio, ni_mio_cs, and
ni_pcimio modules, respectively. Only "ni_pcimio.c" defines the
`PCIDMA` macro before including "ni_mio_common.c" and various bits of
code in "ni_mio_common.c" is conditionally compiled according to whether
that macro is defined or not. Currently, the port I/O function calls
are compiled in regardless of whether the `PCIDMA` macro is defined or
not. However, the fact is that the ni_atmio and ni_mio_cs modules will
never call the memory-mapped I/O functions, and the ni_pcimio module
will never call the port I/O functions.
Calls to the port I/O and memory-mapped I/O functions is confined to the
`ni_writel()`, `ni_writew()`, `ni_writeb()`, `ni_readl()`, `ni_readw()`,
and `ni_readb()` functions which do a run-time test to decide whether to
call the port I/O functions or the memory-mapped I/O functions.
Conditionally compile two variants of the functions so they only call
the port I/O functions if the `PCIDMA` macro is undefined (for the
ni_atmio and ni_mio_cs modules), and only call the memory-mapped I/O
functions if the `PCIDMA` macro is defined (for the ni_pcimio module).
Add a run-time check in the `ni_E_init()` function to return an error if
the comedi device has been set up to use port I/O if `PCIDMA` is
defined, or has been set up to use memory-mapped I/O if `PCIDMA` is not
defined.
The changes make it possible to build the ni_pcimio module even if the
port I/O functions have not been declared. (The ni_atmio and ni_mio_cs
modules do still require the port I/O functions to be declared.)
Cc: Arnd Bergmann <arnd@kernel.org>
Cc: Niklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Link: https://lore.kernel.org/r/20230913170712.111719-10-abbotti@mev.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In a future patch, the port I/O functions (`inb()`, `outb()`, and
friends will only be declared in the `HAS_IOPORT` configuration option
is enabled.
The ni_labpc_common module is used by the ni_labpc module (for ISA
cards), the ni_labpc_cs module (for PCMCIA cards), and the ni_labpc_pci
module (for PCI cards). The ISA and PCMCIA cards use port I/O and the
PCI cards use memory-mapped I/O.
Conditionally compile the parts of the module that use the port I/O
functions so they are compiled if and only if the `CONFIG_HAS_IOPORT`
macro is defined, so that the module can be built if the port I/O
functions have not been declared.
Add a run-time check in the `labpc_common_attach()` to return an error
if the comedi device wants to use port I/O when the `CONFIG_HAS_IOPORT`
macro is undefined.
The changes allow the module to be built even if the port I/O functions
have not been declared.
Cc: Arnd Bergmann <arnd@kernel.org>
Cc: Niklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Link: https://lore.kernel.org/r/20230913170712.111719-9-abbotti@mev.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In a future patch, the port I/O functions (`inb()`, `outb()`, and
friends will only be declared in the `HAS_IOPORT` configuration option
is enabled.
The comedi_8255 module supports both port I/O and memory-mapped I/O.
Conditionally compile the parts of the module that use port I/O if and
only if the `CONFIG_HAS_IOPORT` macro is defined so that it can still be
built if the port I/O functions have not been declared. If the
`CONFIG_HAS_IOPORT` macro is undefined, replace the GPL-exported
`subdev_8255_io_init()` function with a dummy static inline version that
just returns `-ENXIO`.
Cc: Arnd Bergmann <arnd@kernel.org>
Cc: Niklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Link: https://lore.kernel.org/r/20230913170712.111719-8-abbotti@mev.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Comedi drivers can initialize an 8255 subdevice in I/O space by calling
`subdev_8255_init()`, or in memory-mapped I/O space by calling
`subdev_8255_mm_init()`, or by supplying a call-back function pointer
and context to either of those functions. Change it so that a new
function `subdev_8255_cb_init()` shall be called instead when supplying
a callback function and context, and remove the call-back function
parameter from `subdev_8255_init()` and `subdev_8255_mm_init()`.
Also rename `subdev_8255_init()` to `subdev_8255_io_init()`. The
parameters are changing, so might as well rename it at the same time.
Also rename the `regbase` member of `struct subdev_8255_private` to
`context` since this holds the context for the call-back function call.
Cc: Arnd Bergmann <arnd@kernel.org>
Cc: Niklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Link: https://lore.kernel.org/r/20230913170712.111719-7-abbotti@mev.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In a future patch, the port I/O functions (`inb()`, `outb()`, and
friends will only be declared in the `HAS_IOPORT` configuration option
is enabled.
The 8255_pci module supports PCI digital I/O devices from various
manufacturers that consist of one or more 8255 Programmable Peripheral
Interface chips (or equivalent hardware) to provide their digital I/O
ports. Some of the devices use port I/O and some only use memory-mapped
I/O.
Conditionally compile in support for the devices that need port I/O if
and only if the `CONFIG_HAS_IOPORT` macro is defined. Change
`pci_8255_auto_attach()` to return an error if the device actually
requires port I/O (based on the PCI BAR resource flags) but the
`HAS_IOPORT` configuration is not enabled.
Cc: Arnd Bergmann <arnd@kernel.org>
Cc: Niklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Link: https://lore.kernel.org/r/20230913170712.111719-6-abbotti@mev.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The comedi_8254 module supports both port I/O and memory-mapped I/O.
In a future patch, the port I/O functions (`inb()`, `outb()`, and
friends) will only be declared if the `HAS_IOPORT` configuration option
is enabled.
Conditionally compile the parts of the module that use port I/O so they
are compiled if and only if the `CONFIG_HAS_IOPORT` macro is defined, so
that it can still be built if the port I/O functions have not been
declared. If `CONFIG_HAS_IOPORT` is undefined, replace the GPL-exported
`comedi_8254_io_alloc()` function with a dummy static inline version
that just returns `ERR_PTR(-ENXIO)`.
Cc: Arnd Bergmann <arnd@kernel.org>
Cc: Niklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Link: https://lore.kernel.org/r/20230913170712.111719-5-abbotti@mev.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
`comedi_8254_init()` and `comedi_8254_mm_init()` return `NULL` on
failure, but the failure is not necessarily due to lack of memory.
Change them to return an `ERR_PTR` value on failure and rename the
functions to make it obvious the API has changed. `comedi_8254_init()`
has been replaced with `comedi_8254_io_alloc()`, and
`comedi_8254_mm_init()` has been replaced with `comedi_8254_mm_alloc()`.
Cc: Arnd Bergmann <arnd@kernel.org>
Cc: Niklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Link: https://lore.kernel.org/r/20230913170712.111719-4-abbotti@mev.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Rework the comedi_8254 module to use a call-back function for register
access. This will make it easier to isolate the parts that will depend
on the `CONFIG_HAS_IOPORT` macro being defined and also allows the
possibility of supplying an external callback function during
initialization by a variant of the `comedi_8254_init()` and
`comedi_8254_mm_init()` functions, although that has not been
implemented yet.
The `struct comedi_8254` members have been changed to use a pointer to a
callback function and a context of type `unsigned long`. The
`comedi_8254_init()` and `comedi_8254_mm_init()` functions use an
internal callback function and set the context to the base address of
the registers (for `comedi_8254_mm_init()` that involves converting a
`void __iomem *` to `unsigned long`).
A minor change to `dio200_subdev_8254_offset()` in the
amplc_dio200_common module has been made due to the changes in `struct
comedi_8254`.
Cc: Arnd Bergmann <arnd@kernel.org>
Cc: Niklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Link: https://lore.kernel.org/r/20230913170712.111719-3-abbotti@mev.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The ni_pcidio module does not depend on the comedi_8255 module, so
change the `COMEDI_NI_PCIDIO` configuration option to not select
`COMEDI_8255` and remove the inherited dependency on `HAS_IOPORT`.
Cc: Arnd Bergmann <arnd@kernel.org>
Cc: Niklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Link: https://lore.kernel.org/r/20230913170712.111719-2-abbotti@mev.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Prepare for the coming implementation by GCC and Clang of the __counted_by
attribute. Flexible array members annotated with __counted_by can have
their accesses bounds-checked at run-time checking via CONFIG_UBSAN_BOUNDS
(for array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family
functions).
As found with Coccinelle[1], add __counted_by for struct bcm_vk_wkent.
Additionally, since the element count member must be set before accessing
the annotated flexible array member, move its initialization earlier.
[1] https://github.com/kees/kernel-tools/blob/trunk/coccinelle/examples/counted_by.cocci
Cc: Scott Branden <scott.branden@broadcom.com>
Cc: Broadcom internal kernel review list <bcm-kernel-feedback-list@broadcom.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: "Gustavo A. R. Silva" <gustavoars@kernel.org>
Link: https://lore.kernel.org/r/20230922175057.work.558-kees@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Because thermal_zone_device_exec() has no users any more and there are
no plans to use it anywhere, revert commit 9a99a996d1 ("thermal: core:
Introduce thermal_zone_device_exec()") that introduced it.
No functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Rearrange the code handling notifications from the platform firmware
regarding trip point updates to carry out one loop over trip points
instead of two of them by using thermal_zone_for_each_trip() for that,
which is more straightforward than using a combination of
thermal_zone_device_exec() and for_each_thermal_trip(), each with its
own callback function.
No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Modify int340x_thermal_update_trips() to use thermal_zone_for_each_trip()
for walking trips instead of using the trips[] table passed to the
thermal zone registration function.
For this purpose, store active trip point indices in the priv fieids of
the corresponding thermal_trip structures.
No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>
`Box::from_raw()` is `#[must_use]`, which means the result cannot
go unused.
In Rust 1.71.0, this was not detected because the block expression
swallows the diagnostic [1]:
unsafe { Box::from_raw(self.ptr.as_ptr()) };
It would have been detected, however, if the line had been instead:
unsafe { Box::from_raw(self.ptr.as_ptr()); }
i.e. the semicolon being inside the `unsafe` block, rather than
outside.
In Rust 1.72.0, the compiler started warning about this [2], so
without this patch we will get:
error: unused return value of `alloc::boxed::Box::<T>::from_raw` that must be used
--> rust/kernel/sync/arc.rs:302:22
|
302 | unsafe { Box::from_raw(self.ptr.as_ptr()) };
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
= note: call `drop(Box::from_raw(ptr))` if you intend to drop the `Box`
= note: `-D unused-must-use` implied by `-D warnings`
help: use `let _ = ...` to ignore the resulting value
|
302 | unsafe { let _ = Box::from_raw(self.ptr.as_ptr()); };
| +++++++ +
Thus add an add an explicit `drop()` as the `#[must_use]`'s
annotation suggests (instead of the more general help line).
Link: https://github.com/rust-lang/rust/issues/104253 [1]
Link: https://github.com/rust-lang/rust/pull/112529 [2]
Reviewed-by: Martin Rodriguez Reboredo <yakoyoku@gmail.com>
Reviewed-by: Gary Guo <gary@garyguo.net>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Reviewed-by: Andreas Hindborg <a.hindborg@samsung.com>
Reviewed-by: Björn Roy Baron <bjorn3_gh@protonmail.com>
Link: https://lore.kernel.org/r/20230823160244.188033-2-ojeda@kernel.org
Signed-off-by: Miguel Ojeda <ojeda@kernel.org>
Before Google adopted FQ for its production servers,
we had to ensure AF4 packets would get a higher share
than BE1 ones.
As discussed this week in Netconf 2023 in Paris, it is time
to upstream this for public use.
After this patch FQ can replace pfifo_fast, with the following
differences :
- FQ uses WRR instead of strict prio, to avoid starvation of
low priority packets.
- We make sure each band/prio tracks its own usage against sch->limit.
This was done to make sure flood of low priority packets would not
prevent AF4 packets to be queued. Contributed by Willem.
- priomap can be changed, if needed (default value are the ones
coming from pfifo_fast).
In this patch, we set default band weights so that :
- high prio (band=0) packets get 90% of the bandwidth
if they compete with low prio (band=2) packets.
- high prio packets get 75% of the bandwidth
if they compete with medium prio (band=1) packets.
Following patch in this series adds the possibility to tune
the per-band weights.
As we added many fields in 'struct fq_sched_data', we had
to make sure to have the first cache line read-mostly, and
avoid wasting precious cache lines.
More optimizations are possible but will be sent separately.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Dave Taht <dave.taht@gmail.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
pfifo_fast prio2band[] is renamed to sch_default_prio2band[]
and exported because we want to share it in FQ.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Dave Taht <dave.taht@gmail.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Reviewed-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Now that both enqueue() and dequeue() need to use ktime_get_ns(),
there is no point wasting 8 bytes in struct fq_sched_data.
This makes room for future fields. ;)
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Dave Taht <dave.taht@gmail.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Reviewed-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The Cypress 'touchkey' hardware is available on Exynos4412-based Midas
family boards, so enable the driver for it to increase testing coverage.
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Link: https://lore.kernel.org/r/20231005111718.1096634-1-m.szyprowski@samsung.com
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Combine acpi_thermal_update_passive_trip() and
acpi_thermal_update_active_trip() into one common function called
acpi_thermal_update_trip(), so as to reduce code duplication and
prepare the code in question for subsequent changes.
No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Put the get_active_temp() function next to the analogous
get_passive_temp() one to allow subsequent changes to be easier to
follow.
No functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Add a wrapper around for_each_thermal_trip(), called
thermal_zone_for_each_trip(), that will invoke the former under the
thermal zone lock and pass its return value to the caller.
Two drivers will be modified subsequently to use this new function.
No functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>
There is no reason why the KUnit Tests for the property entry API can
only be built-in. Add support for building these tests as a loadable
module, like is supported by most other tests.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/98388154383df9d4ced73946efd18318aeea50e2.1695820382.git.geert+renesas@glider.be
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The kernel documentation validator is not happy with:
drivers/base/core.c:67: warning: Function parameter or member 'flags' not described in '__fwnode_link_add'
Add missing parameter description.
Fixes: 6a6dfdf8b3 ("driver core: fw_devlink: Allow marking a fwnode link as being part of a cycle")
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20230919195048.3197551-1-andriy.shevchenko@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Most of the struct fwnode_handle members are for exclusive use with
device links framework. Clarify this by adding a respective comment.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Reviewed-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Link: https://lore.kernel.org/r/20230904104046.1682875-1-andriy.shevchenko@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Together with enabling the Function Measurement Block
zpci_fmb_enable_device() also resets the software counters. This allows
to use "echo 0 > /sys/kernel/debug/pci/<dev>/statistics" followed by
echo "1 > /../statistics" to reset all counters. In commit c76c067e48
("s390/pci: Use dma-iommu layer") this use of the now obsolete counters
in struct zpci_device was missed as was their removal. Fix this by
resetting the new counters and removing the old ones.
Fixes: c76c067e48 ("s390/pci: Use dma-iommu layer")
Signed-off-by: Niklas Schnelle <schnelle@linux.ibm.com>
Reviewed-by: Matthew Rosato <mjrosato@linux.ibm.com>
Link: https://lore.kernel.org/r/20231004-dma_iommu_fix-v1-1-129777cd8232@linux.ibm.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
msm_iommu platforms do not select either CONFIG_IOMMU_DMA or
CONFIG_ARM_DMA_USE_IOMMU so they create a IOMMU_DOMAIN_DMA domain by
default and never populate it. This acts like a BLOCKED domain and breaks
the GPU driver on the platform.
Detect this and force use of IDENTITY instead.
Fixes: 98ac73f99b ("iommu: Require a default_domain for all iommu drivers")
Reported-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/linux-iommu/CAA8EJprz7VVmBG68U9zLuqPd0UdSRHYoLDJSP6tCj6H6qanuTQ@mail.gmail.com/
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Jerry Snitselaar <jsnitsel@redhat.com>
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/0-v1-20700abdf239+19c-iommu_no_dma_iommu_jgg@nvidia.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
