RALINK_PCI_CONFIG_DATA_VIRTUAL_REG is a very long name. Make it a bit
shorter renaming it to RALINK_PCI_CONFIG_DATA.
Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Tested-by: NeilBrown <neil@brown.name>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Some preprocessor definitions are using a custom implementation of
BIT macro. Just use linux kernel BIT macro instead.
Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Tested-by: NeilBrown <neil@brown.name>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
RALINK_PCI_BASE has no sense and this driver has base address readed
and mapped from device tree. Remove remaining uses of it and
change code to use pcie_read and pcie_write functions in places
where this was being used.
Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Tested-by: NeilBrown <neil@brown.name>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
RALINK_PCI_PCICFG_ADDR and RALINK_PCI_PCIMSK_ADDR are defined to be directly
referenced for read and write. Use pcie_read and pcie_write instead changing
its definition to a simple relative offset to pcie base address.
Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Tested-by: NeilBrown <neil@brown.name>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There are some includes that are being used that are not really
needed to correct driver compilation. Remove them and reorder the
rest alphabetically.
Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Tested-by: NeilBrown <neil@brown.name>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use pcie_[read|write] fucntions to read and write controller registers.
Define those only by offset and pass controller offset + register offset
relative to base address to functions.
Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Tested-by: NeilBrown <neil@brown.name>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
write_config function is always called with bus and func
being 0. Avoid those params and just use 0 inside the
function. Review parameter types changing for more proper
ones.
Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Tested-by: NeilBrown <neil@brown.name>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
read_config function is always called with bus and func
being 0. Avoid those params and just use 0 inside the
function. Return readed value instead pass a reference
parameter.
Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Tested-by: NeilBrown <neil@brown.name>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Driver is using now pci subsystem generics reads and writes and requesting
bus resources without using legacy code functions. Because of this there is
a lot of dead code that can be removed.
Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Tested-by: NeilBrown <neil@brown.name>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
map_bus callback is called before every .read/.write operation.
Implement it and change custom read write operations for the
pci subsystem generics. Make the probe function to don't use
legacy stuff and request bus resources directly. Get pci register
base and ranges from device tree.
The driver is not using PCI_LEGACY code anymore and shall use the
PCI_DRIVERS_GENERIC option to correct compile it. Add also new
Kconfig file for this controller setting there its correct dependencies.
Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Tested-by: NeilBrown <neil@brown.name>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Return statements in functions returning bool should use true or false
instead of an integer value. This code was detected with the help of
Coccinelle.
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
When the user supplies a ctrl_loss_tmo < 0, we warn them that this will
cause the fabrics layer to attempt reconnection forever. However, in
reality the fabrics layer never attempts to reconnect because the
condition to test whether we should reconnect is backwards in this case.
Signed-off-by: Tal Shorer <tal.shorer@gmail.com>
Reviewed-by: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
This patch implements the Namespace Write Protect feature described in
"NVMe TP 4005a Namespace Write Protect". In this version, we implement
No Write Protect and Write Protect states for target ns which can be
toggled by set-features commands from the host side.
For write-protect state transition, we need to flush the ns specified
as a part of command so we also add helpers for carrying out synchronous
flush operations.
Signed-off-by: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
[hch: fixed an incorrect endianess conversion, minor cleanups]
Signed-off-by: Christoph Hellwig <hch@lst.de>
NVMe 1.3 TP 4005 introduces new filed (NSATTR). This field indicates
whether given namespace is write protected or not. This patch sets the
gendisk associated with the namespace to read only based on the identify
namespace nsattr field.
Signed-off-by: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
ANA Phase 3 draft had the 'reserved' field in the group descriptor
format set to '23:17' (so that the first namespace identifier started
at byte 24), but that got move with the approved TP to '31:17'
(so that the first namespace identifier started at byte 32).
Signed-off-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
The offset of the DSP core needs to be taken into account for the DSP
preloader control get and put. Currently the dsp->preloaded variable
will only ever be read/updated on the first DSP, whilst this doesn't
affect the operation of the control the readback will be incorrect.
Signed-off-by: Ajit Pandey <ajit.pandey@cirrus.com>
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Cc: stable@vger.kernel.org
Replace GPL v2.0+ license statements with SPDX license
identifiers.
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
Replace GPL v2.0 and v2.0+ license statements with SPDX license
identifiers.
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
The driver expects to find the device id in rt5677_of_match.data, however
it is currently assigned to rt5677_of_match.type. Fix this.
The problem was found with the help of clang:
sound/soc/codecs/rt5677.c:5010:36: warning: expression which evaluates to
zero treated as a null pointer constant of type 'const void *'
[-Wnon-literal-null-conversion]
{ .compatible = "realtek,rt5677", RT5677 },
^~~~~~
Fixes: ddc9e69b9d ("ASoC: rt5677: Hide platform data in the module sources")
Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
Reviewed-by: Guenter Roeck <groeck@chromium.org>
Acked-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
A copy-and-pasted comment from another code sequence is removed from
gasket core init sequence.
Signed-off-by: Todd Poynor <toddpoynor@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The apex device is left out of reset mode at the end of device
probe/initialize processing. Add a call to enter reset at the end of
the sequence, triggering power gating and other low power features.
Signed-off-by: Todd Poynor <toddpoynor@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Keep mutex held across the unregistration operation, until the
driver_desc field of the global table is removed, to prevent a
concurrent accessor from looking up the driver_desc while
gasket_unregister_device() is in the processing of removing it.
Reported-by: Guenter Roeck <groeck@chromium.org>
Signed-off-by: Todd Poynor <toddpoynor@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The gasket framework no longer provides callbacks to the device driver
for sysfs setup and teardown. Move the sysfs setup code to the device
probe function. Apex does not implement sysfs cleanup code.
Signed-off-by: Todd Poynor <toddpoynor@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Gasket device drivers now call into the gasket framework to initialize
and de-initialize, rather than the other way around. The calling code
can perform sysfs setup and cleanup actions without callbacks from the
framework. Remove the sysfs setup and cleanup callbacks.
Signed-off-by: Todd Poynor <toddpoynor@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Gasket device drivers are now in charge of the device add and remove
sequences; the framework callbacks for these are deleted. Move the
apex device add callback code to the probe function. Apex did not
implement the removal callback.
Signed-off-by: Todd Poynor <toddpoynor@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Gasket device drivers are now in charge of orchestrating the device add
and removal sequences, so the callbacks from the framework to the device
drivers for these events are no longer needed.
Signed-off-by: Todd Poynor <toddpoynor@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Gasket framework now places device drivers in charge of calling APIs to
enable and disable gasket device operations. Make the appropriate calls
from the apex driver.
Signed-off-by: Todd Poynor <toddpoynor@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Move gasket device enable/disable functions from internal calls to
external calls from the gasket device drivers. The device driver will
call these functions at appropriate times in its processing, placing
the device driver in control of this sequence and reducing the need for
callbacks from framework back to the device drivers during the
enable/disable sequences.
Signed-off-by: Todd Poynor <toddpoynor@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
These are not implemented for apex, and are now being removed from the
gasket framework.
Signed-off-by: Todd Poynor <toddpoynor@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Device enable/disable operations are moving from being initiated through
the gasket framework to being initiated by the gasket device driver.
The driver can perform any processing needed for these operations before
or after the calls into the framework. Neither of these callbacks are
implemented for the only gasket driver upstream today, apex.
Signed-off-by: Todd Poynor <toddpoynor@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remaining info-level logs in gasket core converted to debug-level; the
information is not needed during normal system operation.
Signed-off-by: Todd Poynor <toddpoynor@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Apex driver moves PCI core calls like probe, enable, and remove from
gasket to apex. Call new functions in gasket to register apex as a PCI
device to the gasket framework.
Signed-off-by: Todd Poynor <toddpoynor@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove gasket wrapping of PCI probe, enable, disable, and remove
functions. Replace with calls to add and remove PCI gasket devices,
to be called by the gasket device drivers.
Signed-off-by: Todd Poynor <toddpoynor@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If kcalloc() returns NULL in put_mapping(), continue to clean up state,
including dropping the reference on the struct device and free attribute
memory.
Signed-off-by: Todd Poynor <toddpoynor@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The base address used for DMA operations on the second-level table
did incorrectly include the offset for the table entry. The offset
was then added again which lead to incorrect behavior.
Operations on the L1 table are not affected.
The calculation of the base address is changed to point to the
beginning of the L2 table.
Fixes: bfee0cf0ee ("iommu/omap: Use DMA-API for performing cache flushes")
Acked-by: Suman Anna <s-anna@ti.com>
Signed-off-by: Ralf Goebel <ralf.goebel@imago-technologies.com>
Signed-off-by: Joerg Roedel <jroedel@suse.de>
All iommu drivers use the default_iommu_map_sg implementation, and there
is no good reason to ever override it. Just expose it as iommu_map_sg
directly and remove the indirection, specially in our post-spectre world
where indirect calls are horribly expensive.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Joerg Roedel <jroedel@suse.de>
There are some powerpc selftests, as tm/tm-unavailable, that run for a long
period (>120 seconds), and if it is interrupted, as pressing CRTL-C
(SIGINT), the foreground process (harness) dies but the child process and
threads continue to execute (with PPID = 1 now) in background.
In this case, you'd think the whole test exited, but there are remaining
threads and processes being executed in background. Sometimes these
zombies processes are doing annoying things, as consuming the whole CPU or
dumping things to STDOUT.
This patch fixes this problem by attaching an empty signal handler to
SIGINT in the harness process. This handler will interrupt (EINTR) the
parent process waitpid() call, letting the code to follow through the
normal flow, which will kill all the processes in the child process group.
This patch also fixes a typo.
Signed-off-by: Breno Leitao <leitao@debian.org>
Signed-off-by: Gustavo Romero <gromero@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
set_memory_np() is used to mark kernel mappings not present, but it has
it's own open coded mechanism which does not have the L1TF protection of
inverting the address bits.
Replace the open coded PTE manipulation with the L1TF protecting low level
PTE routines.
Passes the CPA self test.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Some cases in THP like:
- MADV_FREE
- mprotect
- split
mark the PMD non present for temporarily to prevent races. The window for
an L1TF attack in these contexts is very small, but it wants to be fixed
for correctness sake.
Use the proper low level functions for pmd/pud_mknotpresent() to address
this.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
