On Odroid XU3/4 and other Exynos5422 based boards there is a case, that
different devices on the board are supplied by different regulators
with non-fixed voltages. If one of these devices temporarily requires
higher voltage, there might occur a situation that the spread between
two devices' voltages is so high, that there is a risk of changing
'high' and 'low' states on the interconnection between devices powered
by those regulators.
Uncoupled regulators should be a special case of coupled regulators, so
they should share a common voltage setting path. When enabling,
disabling or setting voltage of a coupled regulator, all coupled
regulators should be locked. Regulator's supplies should be locked, when
setting voltage of a single regulator. Enabling a coupled regulator or
setting its voltage should not be possible if some of its coupled
regulators, has not been registered.
Add function for locking coupled regulators and supplies. Extract
a new function regulator_set_voltage_rdev() from
regulator_set_voltage_unlocked(), which is called when setting
voltage of a single regulator.
Signed-off-by: Maciej Purski <m.purski@samsung.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
On Odroid XU3/4 and other Exynos5422 based boards there is a case, that
different devices on the board are supplied by different regulators
with non-fixed voltages. If one of these devices temporarily requires
higher voltage, there might occur a situation that the spread between
two devices' voltages is so high, that there is a risk of changing
'high' and 'low' states on the interconnection between devices powered
by those regulators.
Introduce new function regulator_balance_voltage(), which
keeps max_spread constraint fulfilled between a group of coupled
regulators. It should be called if a regulator changes its
voltage or after disabling or enabling. Disabled regulators should
follow changes of the enabled ones, but their consumers' demands
shouldn't be taken into account while calculating voltage of other
coupled regulators.
Find voltages, which are closest to suiting all the consumers' demands,
while fulfilling max_spread constraint, keeping the following rules:
- if one regulator is about to rise its voltage, rise others
voltages in order to keep the max_spread
- if a regulator, which has caused rising other regulators, is
lowered, lower other regulators if possible
- if one regulator is about to lower its voltage, but it hasn't caused
rising other regulators, don't change its voltage if it breaks the
max_spread
Change regulators' voltages step by step, keeping max_spread constraint
fulfilled all the time. Function regulator_get_optimal_voltage()
should find the best possible change for the regulator, which doesn't
break max_spread constraint. In function regulator_balance_voltage()
optimize number of steps by finding highest voltage difference on
each iteration.
If a regulator, which is about to change its voltage, is not coupled,
method regulator_get_optimal_voltage() should simply return the lowest
voltage fulfilling consumers' demands.
Coupling should be checked only if the system is in PM_SUSPEND_ON state.
Signed-off-by: Maciej Purski <m.purski@samsung.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
On Odroid XU3/4 and other Exynos5422 based boards there is a case, that
different devices on the board are supplied by different regulators
with non-fixed voltages. If one of these devices temporarily requires
higher voltage, there might occur a situation that the spread between
two devices' voltages is so high, that there is a risk of changing
'high' and 'low' states on the interconnection between devices powered
by those regulators.
Fill coupling descriptor with data obtained from DTS using previously
defined of_functions. Fail to register a regulator, if some data
inconsistency occurs. If some coupled regulators are not yet registered,
don't fail to register, but try to resolve them in late init call.
Signed-off-by: Maciej Purski <m.purski@samsung.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
On Odroid XU3/4 and other Exynos5422 based boards there is a case, that
different devices on the board are supplied by different regulators
with non-fixed voltages. If one of these devices temporarily requires
higher voltage, there might occur a situation that the spread between
devices' voltages is so high, that there is a risk of changing
'high' and 'low' states on the interconnection between devices powered
by those regulators.
Add new structure "coupling_desc" to regulator_dev, which contains
pointers to all coupled regulators including the owner of the structure,
number of coupled regulators and counter of currently resolved
regulators.
Add of_functions to parse all data needed in regulator coupling.
Provide method to check DTS data consistency. Check if each coupled
regulator's max_spread is equal and if their lists of regulators match.
Signed-off-by: Maciej Purski <m.purski@samsung.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Some regulators require keeping their voltage spread below defined
max_spread.
Add properties to provide information on regulators' coupling.
Signed-off-by: Maciej Purski <m.purski@samsung.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
Setting voltage, enabling/disabling regulators requires operations on
all regulators related with the regulator being changed. Therefore,
all of them should be locked for the whole operation. With the current
locking implementation, adding additional dependency (regulators
coupling) causes deadlocks in some cases.
Introduce a possibility to attempt to lock a mutex multiple times
by the same task without waiting on a mutex. This should handle all
reasonable coupling-supplying combinations, especially when two coupled
regulators share common supplies. The only situation that should be
forbidden is simultaneous coupling and supplying between a pair of
regulators.
The idea is based on clk core.
Signed-off-by: Maciej Purski <m.purski@samsung.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Move the Multiple-epoch facility handling into it and rename it to
kvm_s390_get_tod_clock().
This leaves us with:
- kvm_s390_set_tod_clock()
- kvm_s390_get_tod_clock()
- kvm_s390_get_tod_clock_fast()
So all Multiple-epoch facility is hidden in these functions.
Signed-off-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Collin Walling <walling@linux.ibm.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Janosch Frank <frankja@linux.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
KVM is allocated with kzalloc(), so these members are already 0.
Signed-off-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Janosch Frank <frankja@linux.ibm.com>
Reviewed-by: Collin Walling <walling@linux.ibm.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Janosch Frank <frankja@linux.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
In KVM code we use masks to test/set control registers.
Let's define the ones we use in arch/s390/include/asm/ctl_reg.h and
replace all occurrences in KVM code.
As we will be needing the define for Clock-comparator sign control soon,
let's also add it.
Suggested-by: Collin L. Walling <walling@linux.ibm.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Collin Walling <walling@linux.ibm.com>
Acked-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Acked-by: Janosch Frank <frankja@linux.ibm.com>
Signed-off-by: Janosch Frank <frankja@linux.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Introduces a new function to reset the crypto attributes for all
vcpus whether they are running or not. Each vcpu in KVM will
be removed from SIE prior to resetting the crypto attributes in its
SIE state description. After all vcpus have had their crypto attributes
reset the vcpus will be restored to SIE.
This function is incorporated into the kvm_s390_vm_set_crypto(kvm)
function to fix a reported issue whereby the crypto key wrapping
attributes could potentially get out of synch for running vcpus.
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Reported-by: Halil Pasic <pasic@linux.vnet.ibm.com>
Signed-off-by: Tony Krowiak <akrowiak@linux.vnet.ibm.com>
Signed-off-by: Janosch Frank <frankja@linux.vnet.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Up to now we always expected to have the storage key facility
available for our (non-VSIE) KVM guests. For huge page support, we
need to be able to disable it, so let's introduce that now.
We add the use_skf variable to manage KVM storage key facility
usage. Also we rename use_skey in the mm context struct to uses_skeys
to make it more clear that it is an indication that the vm actively
uses storage keys.
Signed-off-by: Janosch Frank <frankja@linux.vnet.ibm.com>
Reviewed-by: Farhan Ali <alifm@linux.vnet.ibm.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
This patch adds support to core apr service, which is used to query
status of other static and dynamic services on the dsp.
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Reviewed-and-tested-by: Rohit kumar <rohitkr@codeaurora.org>
Acked-by: Banajit Goswami <bgoswami@codeaurora.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
These duplicate includes have been found with scripts/checkincludes.pl but
they have been removed manually to avoid removing false positives.
Signed-off-by: Pravin Shedge <pravin.shedge4linux@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
As a side effect of the following commit, the active TX
serializer may get disabled which may result in distorted
audio output.
ASoC: davinci-mcasp: Add support for multichannel playback
(2952b27e2e)
For example, if a 4 channel I2S playback with two TX serializers
is activated. Later on, if a recording of 2 channels, with only 1 RX
serializer is started, which will also disable one of the TX
serializer because max_active_serializers is only calculated for
RX (recording) stream. This patch fixes this issue.
Signed-off-by: Vishal Thanki <vishalthanki@gmail.com>
Acked-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Many X86 devices using a BYT SoC + RT5640 codec are cheap devices with
generic DMI strings, causing snd_soc_set_dmi_name() to fail to set a
long_name, making it impossible for userspace to have a correct UCM
profile which only uses inputs / outputs which are actually hooked up
on the device.
Our quirks already specify which input the internal mic is connected to
and if a single (mono) speaker is used or if the device has stereo
speakers.
This commit sets a long_name based on the quirks so that userspace can
have UCM profiles doing the right thing based on the long_name.
Note that if we ever encounter the need for a special UCM profile for
some device we can add a quirk to set a specific long_name for the
device,
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Even with our recently tweaked defaults, quite a few bytcr_rt5640 devices
still need quirks to be fully functional. This commits adds quirks where
necessary for the 16 bytcr_rt5640 devices I have access to.
The quirks are added for the following reasons:
1) Devices with only one speaker need the mono quirk to avoid driving an
unused and potentially short-circuited output. 8 of my sample of 16 devs
are mono, 4 of these would work with the defaults if it were not for their
mono speaker.
2) Devices using a different input for the internal mic then the default,
this is the case for 6 of my sample of 16 devices.
3) BYTCR devices without an ACPI channel map, which do not work with the
default of SSP0-AIF2, this is the case for 2 of my sample of 16 devices.
4) Devices which need non-default jack-detect settings, this is the case
for 6 of my sample of 16 devices.
This commit add quirks for the following devices:
Acer Iconia Tab 8 W1-810
Chuwi Vi8
HP Pavilion X2 10-n000nd
HP Stream 7
I.T. Works TW891
Lamina I8270
MSI S100
Pipo W4
PoV-mobii-800w (v2.0)
PoV-mobii-800w (v2.1)
Toshiba Click Mini L9W-B
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Use dmi_first_match() instead of dmi_check_system() + callbacks, this
avoid the need to initialize dmi_system_id.callback for each
byt_rt5640_quirk_table entry.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
As we add more quirks it is useful to have some sort of order in the
quirk list, sort it alphabetically.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Out of the 11 BYTCR devices which I have access to for testing, 6 use
JD1IN4P for jack-detect, 2 use JD1IN4P non-inverted and the other 3 use
JD2IN4N, the ones not using JD1IN4P are all also special in other ways and
need a DMI quirk regardless.
All 5 BYT (non CR) devices which I have access to use JD2IN4N.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Currently we've 2 places with BYTCR defaults: 1. The generic catch-all
DMI_SYS_VENDOR=="Insyde" DMI quirk which selects SSP0-AIF1 for generic
Insyde BYTCR tablets without the ACPI channel package; and 2. the
defaults in the if (is_bytcr) {} code block.
Currently these are not identical, both select IN3 as the internal mic
output, but the "Insyde" DMI quirk leaves out the DIFF_MIC quirk. The
DIFF_MIC quirk should be enabled by default, because enabling diff. input
helps a lot for devices with a differential mic, where as it is a nop on
devices with a normal mic.
This commit adds the DIFF_MIC quirk to the "Insyde" DMI quirk path, by
adding a new BYTCR_INPUT_DEFAULTS define and using that in both code paths
which set BYTCR defaults.
Having a single place where the BYTCR input defaults are defined also
allows defining jack-detect defaults in a single place in a follow-up
commit.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Out of the 11 BYTCR devices which I have access to for testing,
7 use IN3 for the internal mic and only 1 uses IN1 for the internal mic,
the other 3 use DMIC1.
So IN3 clearly is a better default, using IN3 as default avoids the need
to add DMI quirks for some of these devices.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Add code to support setting jack-detect parameters through quirks and
extend the existing DMI quirk table entries for the Asus T100TA and the
Dell Venue 8 Pro 5830 to enable jack detection.
Tested-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
This fixes the following 3 issues:
1) The sys_vendor match should be for "Dell Inc." not "DellInc.",
without this fixed the quirk never gets applied
2) DMIC1 is used not DMIC2, this was not a problem sofar because for
regular BYT boards (rather then BYTCR) we default to DMIC1 and because
of 1. the quirk was not being applied
3) The Dell Venue 8 5830 Pro only has a single speaker
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Use device-properties for setting up the dmic, based on the
BYT_RT5640_MAP() value, instead of using the codec specific
rt5640_dmic_enable() function for this. This also removes the need
for the BYT_RT5640_DMIC_EN quirk, which was always set together with
a MAP() quirk of DMIC1_MAP or DMIC2_MAP.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Instead of passing a global GPIO number for the enable GPIO, pass
a descriptor looked up with the standard devm_gpiod_get_optional()
call.
This regulator supports passing platform data, but enable/sleep
regulators are looked up from the device tree exclusively, so
we can need not touch other files.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Krzysztof Kozlowski <krzk@kernel.org>
Tested-by: Krzysztof Kozlowski <krzk@kernel.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
Instead of passing a global GPIO number, pass a descriptor looked
up from the device tree configuration node.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Krzysztof Kozlowski <krzk@kernel.org>
Tested-by: Krzysztof Kozlowski <krzk@kernel.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
Instead of passing a global GPIO number, pass a descriptor looked
up with the standard devm_gpiod_get_optional() call.
We have augmented the GPIO core to look up the regulator special
GPIO "wlf,ldoena" in commit 6a537d4846
"gpio: of: Support regulator nonstandard GPIO properties".
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Acked-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Use new return type vm_fault_t for fault handler
in struct vm_operations_struct. For now, this is
just documenting that the function returns a
VM_FAULT value rather than an errno. Once all
instances are converted, vm_fault_t will become
a distinct type.
commit 1c8f422059 ("mm: change return type to
vm_fault_t")
Signed-off-by: Souptick Joarder <jrdr.linux@gmail.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Although it does not seem possible to break the host by passing bad
parameters when creating a TCE table in KVM, it is still better to get
an early clear indication of that than debugging weird effect this might
bring.
This adds some sanity checks that the page size is 4KB..16GB as this is
what the actual LoPAPR supports and that the window actually fits 64bit
space.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Acked-by: Balbir Singh <bsingharora@gmail.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
At the moment we only support in the host the IOMMU page sizes which
the guest is aware of, which is 4KB/64KB/16MB. However P9 does not support
16MB IOMMU pages, 2MB and 1GB pages are supported instead. We can still
emulate bigger guest pages (for example 16MB) with smaller host pages
(4KB/64KB/2MB).
This allows the physical IOMMU pages to use a page size smaller or equal
than the guest visible IOMMU page size.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
The other TCE handlers use page shift from the guest visible TCE table
(described by kvmppc_spapr_tce_iommu_table) so let's make H_STUFF_TCE
handlers do the same thing.
This should cause no behavioral change now but soon we will allow
the iommu_table::it_page_shift being different from from the emulated
table page size so this will play a role.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Acked-by: Balbir Singh <bsingharora@gmail.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
There's no need to store the xenstore page or event channel in
xen_start_info if they are locally initialized.
This also fixes PVH local xenstore initialization due to the lack of
xen_start_info in that case.
Signed-off-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
Reviewed-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Juergen Gross <jgross@suse.com>
We now have interrupts hard-disabled when coming back from
kvmppc_hv_entry_trampoline, so this changes the comment to reflect
that.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Although Linux doesn't use PURR and SPURR ((Scaled) Processor
Utilization of Resources Register), other OSes depend on them.
On POWER8 they count at a rate depending on whether the VCPU is
idle or running, the activity of the VCPU, and the value in the
RWMR (Region-Weighting Mode Register). Hardware expects the
hypervisor to update the RWMR when a core is dispatched to reflect
the number of online VCPUs in the vcore.
This adds code to maintain a count in the vcore struct indicating
how many VCPUs are online. In kvmppc_run_core we use that count
to set the RWMR register on POWER8. If the core is split because
of a static or dynamic micro-threading mode, we use the value for
8 threads. The RWMR value is not relevant when the host is
executing because Linux does not use the PURR or SPURR register,
so we don't bother saving and restoring the host value.
For the sake of old userspace which does not set the KVM_REG_PPC_ONLINE
register, we set online to 1 if it was 0 at the time of a KVM_RUN
ioctl.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
This adds a new KVM_REG_PPC_ONLINE register which userspace can set
to 0 or 1 via the GET/SET_ONE_REG interface to indicate whether it
considers the VCPU to be offline (0), that is, not currently running,
or online (1). This will be used in a later patch to configure the
register which controls PURR and SPURR accumulation.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
The commit 289ca025ee ("ALSA: Use priority list for managing device
list") changed the way to register/disconnect/free devices via a
single priority list. This helped to make behavior consistent, but it
also changed a slight behavior change: namely, the control device is
registered earlier than others, while it was supposed to be the very
last one.
I've put SNDRV_DEV_CONTROL in the current position as the release of
ctl elements often conflict with the private ctl elements some PCM or
other components may create, which often leads to a double-free.
But, the order of register and disconnect should be indeed fixed as
expected in the early days: the control device gets registered at
last, and disconnected at first.
This patch changes the priority list order to move SNDRV_DEV_CONTROL
as the last guy to assure the register / disconnect order. Meanwhile,
for keeping the messy resource release order, manually treat the
control and lowlevel devices as last freed one.
Additional note:
The lowlevel device is the device where a card driver creates at
probe. And, we still keep the release order control -> lowlevel, as
there might be link from a control element back to a lowlevel object.
Fixes: 289ca025ee ("ALSA: Use priority list for managing device list")
Reported-by: Tzung-Bi Shih <tzungbi@google.com>
Tested-by: Tzung-Bi Shih <tzungbi@google.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The Bananapi M2 Ultra has a Realtek RTL8211E RGMII PHY tied to the GMAC.
The PMIC's DC1SW output provides power for the PHY, while the ALDO2
output provides I/O voltages on both sides.
Acked-by: Maxime Ripard <maxime.ripard@bootlin.com>
Signed-off-by: Chen-Yu Tsai <wens@csie.org>
The R40 SoC has a GMAC (gigabit capable Ethernet controller). Add a
device node for it. The only publicly available board for this SoC
uses an RGMII PHY. Add a pinmux node for it as well.
Since this SoC also has an old 10/100 Mbps EMAC, which also has an
MDIO bus controller, the MDIO bus for the GMAC is labeled "gmac_mdio".
Acked-by: Maxime Ripard <maxime.ripard@bootlin.com>
Signed-off-by: Chen-Yu Tsai <wens@csie.org>
The device nodes dereference (&foo) usages should be sorted by the label
names, barring any parsing order issues such as the #include statement
for the PMIC's .dtsi file that must come after the PMIC.
Move the mmc and ohci blocks in front of the PMIC's regulator blocks.
Acked-by: Maxime Ripard <maxime.ripard@bootlin.com>
Signed-off-by: Chen-Yu Tsai <wens@csie.org>
There's a GMAC configuration register, which exists on A64/A83T/H3/H5 in
the syscon part, in the CCU of R40 SoC.
Export a regmap of the CCU.
Read access is not restricted to all registers, but only the GMAC
register is allowed to be written.
Signed-off-by: Icenowy Zheng <icenowy@aosc.io>
Acked-by: Maxime Ripard <maxime.ripard@bootlin.com>
Signed-off-by: Chen-Yu Tsai <wens@csie.org>
As we need to register a regmap on the R40 CCU, there needs to be a
device structure bound to the CCU device node.
Rewrite the R40 CCU driver initial code to make it a proper platform
driver, thus we will have a platform device bound to it.
Signed-off-by: Icenowy Zheng <icenowy@aosc.io>
Acked-by: Maxime Ripard <maxime.ripard@bootlin.com>
Signed-off-by: Chen-Yu Tsai <wens@csie.org>
A radix guest can execute tlbie instructions to invalidate TLB entries.
After a tlbie or a group of tlbies, it must then do the architected
sequence eieio; tlbsync; ptesync to ensure that the TLB invalidation
has been processed by all CPUs in the system before it can rely on
no CPU using any translation that it just invalidated.
In fact it is the ptesync which does the actual synchronization in
this sequence, and hardware has a requirement that the ptesync must
be executed on the same CPU thread as the tlbies which it is expected
to order. Thus, if a vCPU gets moved from one physical CPU to
another after it has done some tlbies but before it can get to do the
ptesync, the ptesync will not have the desired effect when it is
executed on the second physical CPU.
To fix this, we do a ptesync in the exit path for radix guests. If
there are any pending tlbies, this will wait for them to complete.
If there aren't, then ptesync will just do the same as sync.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
When a vcpu priority (CPPR) is set to a lower value (masking more
interrupts), we stop processing interrupts already in the queue
for the priorities that have now been masked.
If those interrupts were previously re-routed to a different
CPU, they might still be stuck until the older one that has
them in its queue processes them. In the case of guest CPU
unplug, that can be never.
To address that without creating additional overhead for
the normal interrupt processing path, this changes H_CPPR
handling so that when such a priority change occurs, we
scan the interrupt queue for that vCPU, and for any
interrupt in there that has been re-routed, we replace it
with a dummy and force a re-trigger.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Tested-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
The current partition table unmap code clears the _PAGE_PRESENT bit
out of the pte, which leaves pud_huge/pmd_huge true and does not
clear pud_present/pmd_present. This can confuse subsequent page
faults and possibly lead to the guest looping doing continual
hypervisor page faults.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
The standard eieio ; tlbsync ; ptesync must follow tlbie to ensure it
is ordered with respect to subsequent operations.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Currently, the HV KVM guest entry/exit code adds the timebase offset
from the vcore struct to the timebase on guest entry, and subtracts
it on guest exit. Which is fine, except that it is possible for
userspace to change the offset using the SET_ONE_REG interface while
the vcore is running, as there is only one timebase offset per vcore
but potentially multiple VCPUs in the vcore. If that were to happen,
KVM would subtract a different offset on guest exit from that which
it had added on guest entry, leading to the timebase being out of sync
between cores in the host, which then leads to bad things happening
such as hangs and spurious watchdog timeouts.
To fix this, we add a new field 'tb_offset_applied' to the vcore struct
which stores the offset that is currently applied to the timebase.
This value is set from the vcore tb_offset field on guest entry, and
is what is subtracted from the timebase on guest exit. Since it is
zero when the timebase offset is not applied, we can simplify the
logic in kvmhv_start_timing and kvmhv_accumulate_time.
In addition, we had secondary threads reading the timebase while
running concurrently with code on the primary thread which would
eventually add or subtract the timebase offset from the timebase.
This occurred while saving or restoring the DEC register value on
the secondary threads. Although no specific incorrect behaviour has
been observed, this is a race which should be fixed. To fix it, we
move the DEC saving code to just before we call kvmhv_commence_exit,
and the DEC restoring code to after the point where we have waited
for the primary thread to switch the MMU context and add the timebase
offset. That way we are sure that the timebase contains the guest
timebase value in both cases.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Since the max8998 MFD driver supports instantiation by DT, platform data
retrieval is handled in MFD probe and cell drivers should get use
the pdata field of max8998_dev struct to obtain them.
Fixes: ee999fb3f1 ("mfd: max8998: Add support for Device Tree")
Signed-off-by: Paweł Chmiel <pawel.mikolaj.chmiel@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Add support for configuring the machine constraints
valid_modes_mask element based on a list of allowed modes
specified via a device tree property.
Signed-off-by: David Collins <collinsd@codeaurora.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
Add a common device tree property for regulator nodes to support
the specification of allowed operating modes.
Signed-off-by: David Collins <collinsd@codeaurora.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
Regulators attached via RPMh on Qualcomm sdm845 apparently are
write-only. Specifically you can send a request for a certain voltage
but you can't read back to see what voltage you've requested. What
this means is that at bootup we have absolutely no idea what voltage
we could be at.
As discussed in the patches to try to support the RPMh regulators [1],
the fact that regulators are write-only means that its driver's
get_voltage_sel() should return an error code if it's called before
any calls to set_voltage_sel(). This causes problems in
machine_constraints_voltage() when trying to apply the constraints.
A proposed fix was to come up with an error code that could be
returned by get_voltage_sel() which would cause the regulator
framework to simply try setting the voltage with the current
constraints.
In this patch I propose the error code -ENOTRECOVERABLE. In errno.h
this error is described as "State not recoverable". Though the error
code was originally intended "for robust mutexes", the description of
the error code seems to apply here because we can't read the state of
the regulator. Also note that the only existing user of this error
code in the regulator framework is tps65090-regulator.c which returns
this error code from the enable() call (not get_voltage() or
get_voltage_sel()), so there should be no existing regulators that
might accidentally get the new behavior. (Side note is that tps65090
seems to interpret this error code to mean an error that you can't
recover from rather than some data that can't be recovered).
[1] https://patchwork.kernel.org/patch/10340897/
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Mark Brown <broonie@kernel.org>