Use the define RATE_11M present in the file "device.h" instead of the
magic number 3. So the code is more clear.
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Oscar Carter <oscar.carter@gmx.com>
Link: https://lore.kernel.org/r/20200414153849.5785-4-oscar.carter@gmx.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use the BIT() macro instead of the bit left shift operator. So the code
is more clear.
It's safe to remove the casting to u16 type because the value obtained
never exceeds 16 bits. So the casting is unnecessary.
Signed-off-by: Oscar Carter <oscar.carter@gmx.com>
Link: https://lore.kernel.org/r/20200414153849.5785-3-oscar.carter@gmx.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The kernel requires at least GCC 4.8 in order to build, and so there is
no need to cater for the pre-4.7 gcov format.
Remove the obsolete code.
Acked-by: Peter Oberparleiter <oberpar@linux.ibm.com>
Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: Will Deacon <will@kernel.org>
Passing volatile-qualified pointers to the arm64 implementations of the
load-acquire/store-release macros results in a re-load from the stack
and a bunch of associated stack-protector churn due to the temporary
result variable inheriting the volatile semantics thanks to the use of
'typeof()'.
Define these temporary variables using 'unqual_scalar_typeof' to drop
the volatile qualifier in the case that they are scalar types.
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Acked-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Will Deacon <will@kernel.org>
Passing volatile-qualified pointers to the asm-generic implementations of
the load-acquire macros results in a re-load from the stack due to the
temporary result variable inheriting the volatile semantics thanks to the
use of 'typeof()'.
Define these temporary variables using 'unqual_scalar_typeof' to drop
the volatile qualifier in the case that they are scalar types.
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Will Deacon <will@kernel.org>
Passing a volatile-qualified pointer to READ_ONCE() is an absolute
trainwreck for code generation: the use of 'typeof()' to define a
temporary variable inside the macro means that the final evaluation in
macro scope ends up forcing a read back from the stack. When stack
protector is enabled (the default for arm64, at least), this causes
the compiler to vomit up all sorts of junk.
Unfortunately, dropping pointer qualifiers inside the macro poses quite
a challenge, especially since the pointed-to type is permitted to be an
aggregate, and this is relied upon by mm/ code accessing things like
'pmd_t'. Based on numerous hacks and discussions on the mailing list,
this is the best I've managed to come up with.
Introduce '__unqual_scalar_typeof()' which takes an expression and, if
the expression is an optionally qualified 8, 16, 32 or 64-bit scalar
type, evaluates to the unqualified type. Other input types, including
aggregates, remain unchanged. Hopefully READ_ONCE() on volatile aggregate
pointers isn't something we do on a fast-path.
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Reported-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Will Deacon <will@kernel.org>
{READ,WRITE}_ONCE() cannot guarantee atomicity for arbitrary data sizes.
This can be surprising to callers that might incorrectly be expecting
atomicity for accesses to aggregate structures, although there are other
callers where tearing is actually permissable (e.g. if they are using
something akin to sequence locking to protect the access).
Linus sayeth:
| We could also look at being stricter for the normal READ/WRITE_ONCE(),
| and require that they are
|
| (a) regular integer types
|
| (b) fit in an atomic word
|
| We actually did (b) for a while, until we noticed that we do it on
| loff_t's etc and relaxed the rules. But maybe we could have a
| "non-atomic" version of READ/WRITE_ONCE() that is used for the
| questionable cases?
The slight snag is that we also have to support 64-bit accesses on 32-bit
architectures, as these appear to be widespread and tend to work out ok
if either the architecture supports atomic 64-bit accesses (x86, armv7)
or if the variable being accesses represents a virtual address and
therefore only requires 32-bit atomicity in practice.
Take a step in that direction by introducing a variant of
'compiletime_assert_atomic_type()' and use it to check the pointer
argument to {READ,WRITE}_ONCE(). Expose __{READ,WRITE}_ONCE() variants
which are allowed to tear and convert the one broken caller over to the
new macros.
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Will Deacon <will@kernel.org>
The implementations of {READ,WRITE}_ONCE() suffer from a significant
amount of indirection and complexity due to a historic GCC bug:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58145
which was originally worked around by 230fa253df ("kernel: Provide
READ_ONCE and ASSIGN_ONCE").
Since GCC 4.8 is fairly vintage at this point and we emit a warning if
we detect it during the build, return {READ,WRITE}_ONCE() to their former
glory with an implementation that is easier to understand and, crucially,
more amenable to optimisation. A side effect of this simplification is
that WRITE_ONCE() no longer returns a value, but nobody seems to be
relying on that and the new behaviour is aligned with smp_store_release().
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Will Deacon <will@kernel.org>
Fix a typo in the readl/writel accessor conversion where val is used
instead of pol changing the behavior of the original code.
Cc: stable@vger.kernel.org
Fixes: 6c73698904 pinctrl: qcom: Introduce readl/writel accessors
Signed-off-by: Ansuel Smith <ansuelsmth@gmail.com>
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Link: https://lore.kernel.org/r/20200414003726.25347-1-ansuelsmth@gmail.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
After commit 6f87359e8b ("pinctrl: actions: Fix functions groups names
for S700 SoC") following error has been observed while booting Linux on
Cubieboard7-lite(based on S700 SoC).
[ 1.206245] pinctrl-s700 e01b0000.pinctrl: invalid group "i2c0_mfp" for function "i2c0"
This commit fixes it by using correct name for i2c0_group.
Fixes: 6f87359e8b ("pinctrl: actions: Fix functions groups names for S700 SoC")
Signed-off-by: Amit Singh Tomar <amittomer25@gmail.com>
Link: https://lore.kernel.org/r/1586757913-5438-1-git-send-email-amittomer25@gmail.com
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
As of commit 153969fd95 ("ARM: versatile: Drop CLCD platform data"),
the IB2 module is not accessed in the platform code, so let's remove
mapping it.
Cc: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/20200409222014.31828-1-robh@kernel.org
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Now that there's a DT based sched_clock driver in
drivers/clocksource/timer-versatile.c and all the Arm reference
platforms are DT only, the non-DT versatile sched_clock code can be
removed.
Cc: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/20200409221952.31287-1-robh@kernel.org
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Support for ARM Ltd. RealView systems depends on ARCH_MULTIPLATFORM,
which selects USE_OF.
Support for ARMv6 and ARMv7 variants depends on ARCH_MULTI_V6 or
ARCH_MULTI_V7, which both select ARCH_MULTI_V6_V7 and thus
MIGHT_HAVE_CACHE_L2X0.
Support for ARMv7 variants depends on ARCH_MULTI_V7, which selects
HAVE_SMP.
Hence there is no need for the affected RealView-specific symbols to
select any of them.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Cc: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Support for ARM Ltd. Integrator systems depends on ARCH_MULTIPLATFORM.
As the latter selects SPARSE_IRQ, there is no need for ARCH_INTEGRATOR
to select SPARSE_IRQ.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Cc: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Use the error code returned by the vnt_control_out function as the
returned value of the vnt_rf_write_embedded function instead of a
boolean value.
Then, fix all vnt_rf_write_embedded calls removing the "and" operations
and replace with a direct assignment to the ret variable and add a check
condition after every call.
Also replace the boolean values true or false in the vnt_rf_set_txpower
function to 0 or error code EINVAL to follow the coding style guide.
The vnt_rf_set_txpower function is called only in the vnt_rf_setpower
function that already returns error codes. The calls to this function
(vnt_rf_set_txpower) not use the returned values, so they not need to be
fixed.
Signed-off-by: Oscar Carter <oscar.carter@gmx.com>
Link: https://lore.kernel.org/r/20200414153849.5785-2-oscar.carter@gmx.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Make function arguments alignment match open parenthesis.
Found using checkpatch.pl.
Signed-off-by: R Veera Kumar <vkor@vkten.in>
Reviewed-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Link: https://lore.kernel.org/r/20200413165905.6081-1-vkor@vkten.in
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Make function arguments alignment match open parenthesis.
Found using checkpatch.pl.
Signed-off-by: R Veera Kumar <vkor@vkten.in>
Link: https://lore.kernel.org/r/20200413160753.3787-1-vkor@vkten.in
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Variable 'n' in 'mt7621_pcie_init_virtual_bridges' function
can be directly initialized when it is declared instead of
doing it before using it first.
Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Link: https://lore.kernel.org/r/20200413155018.31921-3-sergio.paracuellos@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add a comment explaining a bit what is going on in this function.
It would be helfulp for other people for a better understanding
of the code.
Suggested-by: NeilBrown <neil@brown.name>
Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Link: https://lore.kernel.org/r/20200413155018.31921-2-sergio.paracuellos@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Definition 'PCIE_P2P_MAX' is '3'. The value here is not a MAXimum.
It is a count or a number. It is how many masks there are.
The masks are numbered 0, 1, 2 so the maximum is 2. Hence rename
variable into 'PCIE_P2P_CNT' which is a more accurate name.
Suggested-by: NeilBrown <neil@brown.name>
Signed-off-by: Sergio Paracuellos <sergio.paracuellos@gmail.com>
Link: https://lore.kernel.org/r/20200413155018.31921-1-sergio.paracuellos@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Update the TODO list associated to the wfx driver with the last
progresses and last remarks made by reviewers.
The items about support for P2P and mesh have also been removed. Indeed,
it seems that the device does not fully support them.
Signed-off-by: Jérôme Pouiller <jerome.pouiller@silabs.com>
Link: https://lore.kernel.org/r/20200415161147.69738-21-Jerome.Pouiller@silabs.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently, wfx_update_filtering() apply the value of wvif->filter_mcast
to the hardware. But an attentive reader will note that
wfx_update_filtering() is always called after set wvif->filter_mcast.
Thus, it not necessary to store filter_mcast in the struct wfx_vif. We
can just pass it as parameter.
Also rename wfx_update_filtering() in wfx_filter_mcast() to reflect this
change.
Signed-off-by: Jérôme Pouiller <jerome.pouiller@silabs.com>
Link: https://lore.kernel.org/r/20200415161147.69738-20-Jerome.Pouiller@silabs.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently, wfx_update_filtering() is sometime called only to apply the
value of wvif->filter_beacon to the hardware. It is nicer to have a
specific function for beacon filtering.
In add, an attentive reader would note that wfx_update_filtering() is
always called after change of wvif->filter_beacon. Thus, it not
necessary to store filter_beacon in the struct wfx_vif. We can just pass
it as parameter.
Signed-off-by: Jérôme Pouiller <jerome.pouiller@silabs.com>
Link: https://lore.kernel.org/r/20200415161147.69738-18-Jerome.Pouiller@silabs.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
It is now useless to store filter_prbreq and filter_bssid in struct
wfx_vif. We can just pass them as parameters to hif_set_rx_filter().
Signed-off-by: Jérôme Pouiller <jerome.pouiller@silabs.com>
Link: https://lore.kernel.org/r/20200415161147.69738-17-Jerome.Pouiller@silabs.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
hif_set_rx_filter() apply changes on filter_prbreq and filter_bssid to
the hardware. Each time filter_prbreq and filter_bssid are changed,
hif_set_rx_filter() is called.
Currently, on extra call to hif_set_rx_filter() is made from
wfx_update_filtering(). This call is useless. Drop it.
Signed-off-by: Jérôme Pouiller <jerome.pouiller@silabs.com>
Link: https://lore.kernel.org/r/20200415161147.69738-16-Jerome.Pouiller@silabs.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The struct wfx_grp_addr_table is only instantiated on one place. There
is no justifiation for this struct. Moreover, it is not consistent with
other fields related to filtering (filter_bssid, etc...).
Signed-off-by: Jérôme Pouiller <jerome.pouiller@silabs.com>
Link: https://lore.kernel.org/r/20200415161147.69738-15-Jerome.Pouiller@silabs.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Filters provided by HIF API are sometime inclusive, sometime exclusive.
This patch align the behavior and name of the probe request filter with
the other filters.
Signed-off-by: Jérôme Pouiller <jerome.pouiller@silabs.com>
Link: https://lore.kernel.org/r/20200415161147.69738-14-Jerome.Pouiller@silabs.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Filters provided by HIF API are sometime inclusive, sometime exclusive.
This patch align the behavior and name of the beacon filter with the
other filters. Also avoid double negation: "disable filter"
Signed-off-by: Jérôme Pouiller <jerome.pouiller@silabs.com>
Link: https://lore.kernel.org/r/20200415161147.69738-13-Jerome.Pouiller@silabs.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
wfx_fwd_probe_req() is a function of two lines called from only one
place. In order to uniformize all filtering functions, drop
wfx_fwd_probe_req().
Signed-off-by: Jérôme Pouiller <jerome.pouiller@silabs.com>
Link: https://lore.kernel.org/r/20200415161147.69738-12-Jerome.Pouiller@silabs.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Device is able to automatically reply to probe request (therefore, the
probe requests are not forwarded to host). This feature is enabled by
wfx_start_ap(). However, further calls to wfx_configure_filter() can
disable it.
So, relocate the handling of probe request filtering in
wfx_configure_filter().
Note that wfx_configure_filter() is always called by mac80211 when an AP
start.
Signed-off-by: Jérôme Pouiller <jerome.pouiller@silabs.com>
Link: https://lore.kernel.org/r/20200415161147.69738-11-Jerome.Pouiller@silabs.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
It is not very clear why driver ask to firmware to stop filtering probe
request during scan. It is particularly weird because scan is not
supported in AP mode. In add, it does not restore previous filtering
state after scanning.
It seems to be a mistake and should be dropped.
Signed-off-by: Jérôme Pouiller <jerome.pouiller@silabs.com>
Link: https://lore.kernel.org/r/20200415161147.69738-10-Jerome.Pouiller@silabs.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Firmware is able to detect detect changes about ERP protection in
beacons and automatically enable/disable ERP protection. However, it not
bring performance improvements and we are more confident in the ERP
handling of mac80211. So, the patch disable this feature.
Signed-off-by: Jérôme Pouiller <jerome.pouiller@silabs.com>
Link: https://lore.kernel.org/r/20200415161147.69738-9-Jerome.Pouiller@silabs.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
After hif_reset(), device stop to receive any RF data. So, it is
useless to disable beacon filtering. In add, if necessary, mac80211 will
call wfx_configure_filter().
Signed-off-by: Jérôme Pouiller <jerome.pouiller@silabs.com>
Link: https://lore.kernel.org/r/20200415161147.69738-8-Jerome.Pouiller@silabs.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The only reason we ask to mac80211 to wait for a dtim before to
associate is to get an accurate DTIM value. However, we can temporary
set the dtim period to 1. Then we can waiting for
BSS_CHANGED_BEACON_INFO for the real value. Thus, the association
process is faster.
Signed-off-by: Jérôme Pouiller <jerome.pouiller@silabs.com>
Link: https://lore.kernel.org/r/20200415161147.69738-7-Jerome.Pouiller@silabs.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When the device joins an AP, beacon are forwarded to the host. Once has
retrieved all necessary parameters, we can start to filter the beacon
(and only beacon with changed/new data will be forwarded).
Currently, the driver detect beacons in data Rx process. It is far more
easier to just wait for the BSS_CHANGED_BEACON_INFO event.
Signed-off-by: Jérôme Pouiller <jerome.pouiller@silabs.com>
Link: https://lore.kernel.org/r/20200415161147.69738-5-Jerome.Pouiller@silabs.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
wfx_configure_filter() lacks of coherency. In add, some corner cases
seems to not been handled properly. Rework the whole function().
Signed-off-by: Jérôme Pouiller <jerome.pouiller@silabs.com>
Link: https://lore.kernel.org/r/20200415161147.69738-4-Jerome.Pouiller@silabs.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently, filtering are not update if the device is not connected.
There is no obvious rationale for that. Updating filtering while it is
not necessary is harmless. On the other hand, it is difficult to ensure
that the filtering rule has been correctly sent to the firmware.
Just drop this useless condition.
Signed-off-by: Jérôme Pouiller <jerome.pouiller@silabs.com>
Link: https://lore.kernel.org/r/20200415161147.69738-2-Jerome.Pouiller@silabs.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The "mali_0" or "mali_1" clock trees should not be updated while the
clock is running. Enforce this by setting CLK_SET_RATE_GATE on the
"mali_0" and "mali_1" gates. This makes the CCF switch to the "mali_1"
tree when "mali_0" is currently active and vice versa, which is exactly
what the vendor driver does when updating the frequency of the mali
clock.
Also propagate set_rate requests from the gate to the divider and from
the divider to the the mux so the GPU clock frequency can be updated at
runtime (which will be required for GPU DVFS). Don't propagate rate
changes to the mux parents because we don't want to change the MPLL
clocks (these are reserved for audio).
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
Acked-by: Neil Armstrong <narmstrong@baylibre.com>
Link: https://lore.kernel.org/r/20200414195031.224021-3-martin.blumenstingl@googlemail.com
The "mali_0" or "mali_1" clock trees should not be updated while the
clock is running. Enforce this by setting CLK_SET_RATE_GATE on the
"mali_0" and "mali_1" gates. This makes the CCF switch to the "mali_1"
tree when "mali_0" is currently active and vice versa, which is exactly
what the vendor driver does when updating the frequency of the mali
clock.
Also propagate set_rate requests from the gate to the divider and from
the divider to the the mux so the GPU clock frequency can be updated at
runtime (which will be required for GPU DVFS). Don't propagate rate
changes to the mux parents because we don't want to change the MPLL
clocks (these are reserved for audio).
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
Acked-by: Neil Armstrong <narmstrong@baylibre.com>
Link: https://lore.kernel.org/r/20200414195031.224021-2-martin.blumenstingl@googlemail.com
Michael Kerrisk suggested to replace numeric clock IDs with symbolic names.
Now the content of these files looks like this:
$ cat /proc/774/timens_offsets
monotonic 864000 0
boottime 1728000 0
For setting offsets, both representations of clocks (numeric and symbolic)
can be used.
As for compatibility, it is acceptable to change things as long as
userspace doesn't care. The format of timens_offsets files is very new and
there are no userspace tools yet which rely on this format.
But three projects crun, util-linux and criu rely on the interface of
setting time offsets and this is why it's required to continue supporting
the numeric clock IDs on write.
Fixes: 04a8682a71 ("fs/proc: Introduce /proc/pid/timens_offsets")
Suggested-by: Michael Kerrisk <mtk.manpages@gmail.com>
Signed-off-by: Andrei Vagin <avagin@gmail.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Michael Kerrisk <mtk.manpages@gmail.com>
Acked-by: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Dmitry Safonov <0x7f454c46@gmail.com>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20200411154031.642557-1-avagin@gmail.com
The check was always succeeding even in case of a mismatch due to the
HDCP_STATUS_ENC bit being set. Make sure both bits are actually set.
Signed-off-by: Oliver Barta <oliver.barta@aptiv.com>
Fixes: 2320175feb ("drm/i915: Implement HDCP for HDMI")
Reviewed-by: Ramalingam C <ramalingam.c@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200328104100.12162-1-oliver.barta@aptiv.com
The Enhanced Connection Complete event is use in conjunction with LL
Privacy and not Extended Advertising.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Although the vSGIs are not directly visible to the host, they still
get moved around by the CPU hotplug, for example. This results in
the kernel moaning on the console, such as:
genirq: irq_chip GICv4.1-sgi did not update eff. affinity mask of irq 38
Updating the effective affinity on set_affinity() fixes it.
Reviewed-by: Zenghui Yu <yuzenghui@huawei.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
When a vPE is made resident, the GIC starts parsing the virtual pending
table to deliver pending interrupts. This takes place asynchronously,
and can at times take a long while. Long enough that the vcpu enters
the guest and hits WFI before any interrupt has been signaled yet.
The vcpu then exits, blocks, and now gets a doorbell. Rince, repeat.
In order to avoid the above, a (optional on GICv4, mandatory on v4.1)
feature allows the GIC to feedback to the hypervisor whether it is
done parsing the VPT by clearing the GICR_VPENDBASER.Dirty bit.
The hypervisor can then wait until the GIC is ready before actually
running the vPE.
Plug the detection code as well as polling on vPE schedule. While
at it, tidy-up the kernel message that displays the GICv4 optional
features.
Reviewed-by: Zenghui Yu <yuzenghui@huawei.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
PLANE_CTL_FORMAT_AYUV is already supported, according to hardware
specification.
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
Signed-off-by: Bob Paauwe <bob.j.paauwe@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200407215546.5445-2-bob.j.paauwe@intel.com
