IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an
email to Administrator. User accounts are meant only to access repo
and report issues and/or generate pull requests.
This is a purpose-specific Git hosting for
BaseALT
projects. Thank you for your understanding!
Только зарегистрированные пользователи имеют доступ к сервису!
Для получения аккаунта, обратитесь к администратору.
Commit b1c36aae51c9 ("regulator: Convert SY8106A binding to a schema")
converts sy8106a-regulator.txt to silergy,sy8106a.yaml, but missed to
adjust its reference in MAINTAINERS.
Hence, ./scripts/get_maintainer.pl --self-test=patterns complains about
a broken reference.
Repair this file reference in SY8106A REGULATOR DRIVER.
Signed-off-by: Lukas Bulwahn <lukas.bulwahn@gmail.com>
Link: https://lore.kernel.org/r/20211005075451.29691-11-lukas.bulwahn@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>
This reverts commit 0da6736ecd10b45e535b100acd58df2db4c099d8.
The MODULE_DEVICE_TABLE already creates proper alias. Having another
MODULE_ALIAS causes the alias to be duplicated:
$ modinfo max14577-regulator.ko
alias: platform:max77836-regulator
alias: platform:max14577-regulator
description: Maxim 14577/77836 regulator driver
alias: platform:max77836-regulator
alias: platform:max14577-regulator
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Fixes: 0da6736ecd10 ("regulator: max14577: Add proper module aliases strings")
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Link: https://lore.kernel.org/r/20210916144102.120980-1-krzysztof.kozlowski@canonical.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Fix a typo in the pm8009 LDO7 declaration, it uses resource name ldo%s6
instead of ldo%s7.
Fixes: 951384cabc5d ("regulator: qcom-rpmh-regulator: add pm8009-1 chip revision")
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Link: https://lore.kernel.org/r/20210901114350.1106073-1-dmitry.baryshkov@linaro.org
Signed-off-by: Mark Brown <broonie@kernel.org>
vctrl_enable() and vctrl_disable() call regulator_enable() and
regulator_disable(), respectively. However, vctrl_* are regulator ops
and should not be calling the locked regulator APIs. Doing so results in
a lockdep warning.
Instead of exporting more internal regulator ops, model the ctrl supply
as an actual supply to vctrl-regulator. At probe time this driver still
needs to use the consumer API to fetch its constraints, but otherwise
lets the regulator core handle the upstream supply for it.
The enable/disable/is_enabled ops are not removed, but now only track
state internally. This preserves the original behavior with the ops
being available, but one could argue that the original behavior was
already incorrect: the internal state would not match the upstream
supply if that supply had another consumer that enabled the supply,
while vctrl-regulator was not enabled.
The lockdep warning is as follows:
WARNING: possible circular locking dependency detected
5.14.0-rc6 #2 Not tainted
------------------------------------------------------
swapper/0/1 is trying to acquire lock:
ffffffc011306d00 (regulator_list_mutex){+.+.}-{3:3}, at:
regulator_lock_dependent (arch/arm64/include/asm/current.h:19
include/linux/ww_mutex.h:111
drivers/regulator/core.c:329)
but task is already holding lock:
ffffff8004a77160 (regulator_ww_class_mutex){+.+.}-{3:3}, at:
regulator_lock_recursive (drivers/regulator/core.c:156
drivers/regulator/core.c:263)
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #2 (regulator_ww_class_mutex){+.+.}-{3:3}:
__mutex_lock_common (include/asm-generic/atomic-instrumented.h:606
include/asm-generic/atomic-long.h:29
kernel/locking/mutex.c:103
kernel/locking/mutex.c:144
kernel/locking/mutex.c:963)
ww_mutex_lock (kernel/locking/mutex.c:1199)
regulator_lock_recursive (drivers/regulator/core.c:156
drivers/regulator/core.c:263)
regulator_lock_dependent (drivers/regulator/core.c:343)
regulator_enable (drivers/regulator/core.c:2808)
set_machine_constraints (drivers/regulator/core.c:1536)
regulator_register (drivers/regulator/core.c:5486)
devm_regulator_register (drivers/regulator/devres.c:196)
reg_fixed_voltage_probe (drivers/regulator/fixed.c:289)
platform_probe (drivers/base/platform.c:1427)
[...]
-> #1 (regulator_ww_class_acquire){+.+.}-{0:0}:
regulator_lock_dependent (include/linux/ww_mutex.h:129
drivers/regulator/core.c:329)
regulator_enable (drivers/regulator/core.c:2808)
set_machine_constraints (drivers/regulator/core.c:1536)
regulator_register (drivers/regulator/core.c:5486)
devm_regulator_register (drivers/regulator/devres.c:196)
reg_fixed_voltage_probe (drivers/regulator/fixed.c:289)
[...]
-> #0 (regulator_list_mutex){+.+.}-{3:3}:
__lock_acquire (kernel/locking/lockdep.c:3052 (discriminator 4)
kernel/locking/lockdep.c:3174 (discriminator 4)
kernel/locking/lockdep.c:3789 (discriminator 4)
kernel/locking/lockdep.c:5015 (discriminator 4))
lock_acquire (arch/arm64/include/asm/percpu.h:39
kernel/locking/lockdep.c:438
kernel/locking/lockdep.c:5627)
__mutex_lock_common (include/asm-generic/atomic-instrumented.h:606
include/asm-generic/atomic-long.h:29
kernel/locking/mutex.c:103
kernel/locking/mutex.c:144
kernel/locking/mutex.c:963)
mutex_lock_nested (kernel/locking/mutex.c:1125)
regulator_lock_dependent (arch/arm64/include/asm/current.h:19
include/linux/ww_mutex.h:111
drivers/regulator/core.c:329)
regulator_enable (drivers/regulator/core.c:2808)
vctrl_enable (drivers/regulator/vctrl-regulator.c:400)
_regulator_do_enable (drivers/regulator/core.c:2617)
_regulator_enable (drivers/regulator/core.c:2764)
regulator_enable (drivers/regulator/core.c:308
drivers/regulator/core.c:2809)
_set_opp (drivers/opp/core.c:819 drivers/opp/core.c:1072)
dev_pm_opp_set_rate (drivers/opp/core.c:1164)
set_target (drivers/cpufreq/cpufreq-dt.c:62)
__cpufreq_driver_target (drivers/cpufreq/cpufreq.c:2216
drivers/cpufreq/cpufreq.c:2271)
cpufreq_online (drivers/cpufreq/cpufreq.c:1488 (discriminator 2))
cpufreq_add_dev (drivers/cpufreq/cpufreq.c:1563)
subsys_interface_register (drivers/base/bus.c:?)
cpufreq_register_driver (drivers/cpufreq/cpufreq.c:2819)
dt_cpufreq_probe (drivers/cpufreq/cpufreq-dt.c:344)
[...]
other info that might help us debug this:
Chain exists of:
regulator_list_mutex --> regulator_ww_class_acquire --> regulator_ww_class_mutex
Possible unsafe locking scenario:
CPU0 CPU1
---- ----
lock(regulator_ww_class_mutex);
lock(regulator_ww_class_acquire);
lock(regulator_ww_class_mutex);
lock(regulator_list_mutex);
*** DEADLOCK ***
6 locks held by swapper/0/1:
#0: ffffff8002d32188 (&dev->mutex){....}-{3:3}, at:
__device_driver_lock (drivers/base/dd.c:1030)
#1: ffffffc0111a0520 (cpu_hotplug_lock){++++}-{0:0}, at:
cpufreq_register_driver (drivers/cpufreq/cpufreq.c:2792 (discriminator 2))
#2: ffffff8002a8d918 (subsys mutex#9){+.+.}-{3:3}, at:
subsys_interface_register (drivers/base/bus.c:1033)
#3: ffffff800341bb90 (&policy->rwsem){+.+.}-{3:3}, at:
cpufreq_online (include/linux/bitmap.h:285
include/linux/cpumask.h:405
drivers/cpufreq/cpufreq.c:1399)
#4: ffffffc011f0b7b8 (regulator_ww_class_acquire){+.+.}-{0:0}, at:
regulator_enable (drivers/regulator/core.c:2808)
#5: ffffff8004a77160 (regulator_ww_class_mutex){+.+.}-{3:3}, at:
regulator_lock_recursive (drivers/regulator/core.c:156
drivers/regulator/core.c:263)
stack backtrace:
CPU: 1 PID: 1 Comm: swapper/0 Not tainted 5.14.0-rc6 #2 7c8f8996d021ed0f65271e6aeebf7999de74a9fa
Hardware name: Google Scarlet (DT)
Call trace:
dump_backtrace (arch/arm64/kernel/stacktrace.c:161)
show_stack (arch/arm64/kernel/stacktrace.c:218)
dump_stack_lvl (lib/dump_stack.c:106 (discriminator 2))
dump_stack (lib/dump_stack.c:113)
print_circular_bug (kernel/locking/lockdep.c:?)
check_noncircular (kernel/locking/lockdep.c:?)
__lock_acquire (kernel/locking/lockdep.c:3052 (discriminator 4)
kernel/locking/lockdep.c:3174 (discriminator 4)
kernel/locking/lockdep.c:3789 (discriminator 4)
kernel/locking/lockdep.c:5015 (discriminator 4))
lock_acquire (arch/arm64/include/asm/percpu.h:39
kernel/locking/lockdep.c:438
kernel/locking/lockdep.c:5627)
__mutex_lock_common (include/asm-generic/atomic-instrumented.h:606
include/asm-generic/atomic-long.h:29
kernel/locking/mutex.c:103
kernel/locking/mutex.c:144
kernel/locking/mutex.c:963)
mutex_lock_nested (kernel/locking/mutex.c:1125)
regulator_lock_dependent (arch/arm64/include/asm/current.h:19
include/linux/ww_mutex.h:111
drivers/regulator/core.c:329)
regulator_enable (drivers/regulator/core.c:2808)
vctrl_enable (drivers/regulator/vctrl-regulator.c:400)
_regulator_do_enable (drivers/regulator/core.c:2617)
_regulator_enable (drivers/regulator/core.c:2764)
regulator_enable (drivers/regulator/core.c:308
drivers/regulator/core.c:2809)
_set_opp (drivers/opp/core.c:819 drivers/opp/core.c:1072)
dev_pm_opp_set_rate (drivers/opp/core.c:1164)
set_target (drivers/cpufreq/cpufreq-dt.c:62)
__cpufreq_driver_target (drivers/cpufreq/cpufreq.c:2216
drivers/cpufreq/cpufreq.c:2271)
cpufreq_online (drivers/cpufreq/cpufreq.c:1488 (discriminator 2))
cpufreq_add_dev (drivers/cpufreq/cpufreq.c:1563)
subsys_interface_register (drivers/base/bus.c:?)
cpufreq_register_driver (drivers/cpufreq/cpufreq.c:2819)
dt_cpufreq_probe (drivers/cpufreq/cpufreq-dt.c:344)
[...]
Reported-by: Brian Norris <briannorris@chromium.org>
Fixes: f8702f9e4aa7 ("regulator: core: Use ww_mutex for regulators locking")
Fixes: e9153311491d ("regulator: vctrl-regulator: Avoid deadlock getting and setting the voltage")
Signed-off-by: Chen-Yu Tsai <wenst@chromium.org>
Link: https://lore.kernel.org/r/20210825033704.3307263-3-wenst@chromium.org
Signed-off-by: Mark Brown <broonie@kernel.org>
In commit e9153311491d ("regulator: vctrl-regulator: Avoid deadlock getting
and setting the voltage"), all calls to get/set the voltage of the
control regulator were switched to unlocked versions to avoid deadlocks.
However, the call in the probe path is done without regulator locks
held. In this case the locked version should be used.
Switch back to the locked regulator_get_voltage() in the probe path to
avoid any mishaps.
Fixes: e9153311491d ("regulator: vctrl-regulator: Avoid deadlock getting and setting the voltage")
Signed-off-by: Chen-Yu Tsai <wenst@chromium.org>
Link: https://lore.kernel.org/r/20210825033704.3307263-2-wenst@chromium.org
Signed-off-by: Mark Brown <broonie@kernel.org>
The helper to send IRQ notification for regulator errors had still
old description mentioning calling BUG() as a last resort when
error status reading has kept failing for more times than a given
threshold.
The impementation calling BUG() did never end-up in-tree but was
replaced by hopefully more sophisticated handler trying to power-off
the system.
Fix the documentation to reflect actual behaviour.
Signed-off-by: Matti Vaittinen <matti.vaittinen@fi.rohmeurope.com>
Link: https://lore.kernel.org/r/20210823075651.GA3717293@localhost.localdomain
Signed-off-by: Mark Brown <broonie@kernel.org>
From testing on hardware the poll_enable_time isn't required and
sometimes causes the driver probe to fail so let's remove it.
Signed-off-by: Alistair Francis <alistair@alistair23.me>
Link: https://lore.kernel.org/r/20210803084456.198-5-alistair@alistair23.me
Signed-off-by: Mark Brown <broonie@kernel.org>
Enable regmap cache to reduce i2c transactions and corresponding
interrupts if regulator is accessed frequently. Since the register map
is small, we use a FLAT regmap cache.
Signed-off-by: Jisheng Zhang <Jisheng.Zhang@synaptics.com>
Link: https://lore.kernel.org/r/20210803165211.3b00db29@xhacker.debian
Signed-off-by: Mark Brown <broonie@kernel.org>
Enable regmap cache to reduce i2c transactions and corresponding
interrupts if regulator is accessed frequently. Since the register map
is small -- there's only one register in sy8824c and sy8824e, there
are only two registers in sy20276 and sy20278, so we use a FLAT regmap
cache.
Signed-off-by: Jisheng Zhang <Jisheng.Zhang@synaptics.com>
Link: https://lore.kernel.org/r/20210803165043.042ec24d@xhacker.debian
Signed-off-by: Mark Brown <broonie@kernel.org>
The arrays containing the regulator's voltage ranges are
currently named after the first ldo which uses such range.
However, it sounds a lot clearer if those are named with
the voltage range instead.
No functional changes.
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Link: https://lore.kernel.org/r/1bdff1d1f23753b69c8044160decfad1e8553d08.1627121912.git.mchehab+huawei@kernel.org
Signed-off-by: Mark Brown <broonie@kernel.org>
Instead of returning error directly, use dev_err_probe. This avoids
messages in the dmesg log for devices which will be probed again later.
Signed-off-by: Chris Morgan <macromorgan@hotmail.com>
Link: https://lore.kernel.org/r/20210721165716.19915-1-macroalpha82@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Reg will be reset in below two conditions.
1. 'Enable' pin from H to L.
2. Both PAVDD and NAVDD are all disabled.
And 'Enable' pin also control i2c communication capability.
This patch is to Seperate the if condition in enable/disable callback for
reg cache manipulation.
Signed-off-by: ChiYuan Huang <cy_huang@richtek.com>
Link: https://lore.kernel.org/r/1626407746-23156-1-git-send-email-u0084500@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>
In addition to the ability of merging some power outputs, this chip has
an overdrive mode.
BCORE1, BCORE2 and BPRO have this ability, in which case the legal
current draw is increased from 2 amps to 2.5 amps (at the expense of
a quiescent current increase), and the configurable current limits
are doubled.
If a current higher than maximum half-current mode is requested, enable
overdrive, and scale the current limit down.
Symmetrically, scale the current limit up when querying a overdrive-enabled
regulator.
Signed-off-by: Vincent Pelletier <plr.vincent@gmail.com>
Reviewed-by: Adam Thomson <Adam.Thomson.Opensource@diasemi.com>
Link: https://lore.kernel.org/r/824518e6391b783a12eba9ff0527f06607a34bfb.1626160826.git.plr.vincent@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Don't populate the const array func_base on the stack but instead it
static. Makes the object code smaller by 55 bytes:
Before:
text data bss dec hex filename
6422 3216 64 9702 25e6 drivers/regulator/rt6245-regulator.o
After:
text data bss dec hex filename
6303 3280 64 9647 25af drivers/regulator/rt6245-regulator.o
Reduction of 55 bytes
(gcc version 10.3.0)
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Link: https://lore.kernel.org/r/20210715141531.27672-1-colin.king@canonical.com
Signed-off-by: Mark Brown <broonie@kernel.org>
These APIs aren't used anywhere and most-likely exist because of the
general principle of C APIs, where if an API function does an
allocation/registration, it must also have an equivalent
deallocation/deregistration counterpart.
For devm_ functions this isn't all that true (for all cases), as the idea
of these function is to provide an auto-cleanup logic on drivers/system
de-init.
Removing these discourages any weird logic that could be created with
such an API functions.
Alexandru Ardelean (4):
regulator: devres: remove devm_regulator_unregister_notifier()
function
regulator: devres: remove devm_regulator_unregister() function
regulator: devres: remove
devm_regulator_bulk_unregister_supply_alias()
regulator: devres: unexport devm_regulator_unregister_supply_alias()
drivers/regulator/devres.c | 105 +----------------------------
include/linux/regulator/consumer.h | 23 -------
include/linux/regulator/driver.h | 1 -
3 files changed, 2 insertions(+), 127 deletions(-)
--
2.31.1
The ROHM BD71837/47/50/78 do support enabling/disabling the under/over
voltage protection. Add support for enabling/disabling the protection
according to the device-tree information.
Signed-off-by: Matti Vaittinen <matti.vaittinen@fi.rohmeurope.com>
Link: https://lore.kernel.org/r/20210705105416.GA1189560@localhost.localdomain
Signed-off-by: Mark Brown <broonie@kernel.org>
Fix warning caused by a blank/empty line:
../include/linux/regulator/machine.h:115: warning: bad line:
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Link: https://lore.kernel.org/r/20210628015422.8845-1-rdunlap@infradead.org
Signed-off-by: Mark Brown <broonie@kernel.org>
Instead of linear mapping, Use linear range to map all voltage selection.
Signed-off-by: ChiYuan Huang <cy_huang@richtek.com>
Reviewed-by: Axel Lin <axel.lin@ingics.com>
Acked-by: Lee Jones <lee.jones@linaro.org>
Link: https://lore.kernel.org/r/1625553939-9109-1-git-send-email-u0084500@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The TPS65910 regulator now gets a deferred probe until supply regulator is
registered. Silence noisy error message about the deferred probe.
Reported-by: Matt Merhar <mattmerhar@protonmail.com> # Ouya T30
Tested-by: Matt Merhar <mattmerhar@protonmail.com> # Ouya T30
Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
Link: https://lore.kernel.org/r/20210705201211.16082-1-digetx@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The shift setting can be calculated via the corresponding mask field,
so remove these shift fields.
The usage of da_vsel_mask is different from other mask defines because
current code does shift regval before mask with the da_vsel_mask.
Do proper shit to da_vsel_mask setting so we can calculate the shift.
Signed-off-by: Axel Lin <axel.lin@ingics.com>
Link: https://lore.kernel.org/r/20210629130503.2183574-1-axel.lin@ingics.com
Signed-off-by: Mark Brown <broonie@kernel.org>
This API hook isn't used anywhere outside of the regulator devres code.
This function is needed for the devm_regulator_bulk_register_supply_alias()
function on the error path, to cleanup any previously registered supply
aliases.
This change makes the devm_regulator_unregister_supply_alias() local to the
regulator core framework, to avoid it being used in any weird logic.
It's also removing the doc-string for
devm_regulator_unregister_supply_alias(), since it doesn't need to be
documented anymore, as no other external consumer should use it.
Signed-off-by: Alexandru Ardelean <aardelean@deviqon.com>
Link: https://lore.kernel.org/r/20210625122324.327585-5-aardelean@deviqon.com
Signed-off-by: Mark Brown <broonie@kernel.org>
This API hook isn't used anywhere and most-likely exists because of the
general principle of C APIs, where if an API function does an
allocation/registration, it must also have an equivalent
deallocation/deregistration counterpart.
For devm_ functions this isn't all that true (for all cases), as the idea
of these function is to provide an auto-cleanup logic on drivers/system
de-init.
Removing this also discourages any weird logic that could be created with
such an API function.
Signed-off-by: Alexandru Ardelean <aardelean@deviqon.com>
Link: https://lore.kernel.org/r/20210625122324.327585-4-aardelean@deviqon.com
Signed-off-by: Mark Brown <broonie@kernel.org>
This API hook isn't used anywhere and most-likely exists because of the
general principle of C APIs, where if an API function does an
allocation/registration, it must also have an equivalent
deallocation/deregistration counterpart.
For devm_ functions this isn't all that true (for all cases), as the idea
of these function is to provide an auto-cleanup logic on drivers/system
de-init.
Removing this also discourages any weird logic that could be created with
such an API function.
Signed-off-by: Alexandru Ardelean <aardelean@deviqon.com>
Link: https://lore.kernel.org/r/20210625122324.327585-3-aardelean@deviqon.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Since config.dev = pdev->dev.parent in current code, so
dev_get_drvdata(rdev->dev.parent) call in hi6421_regulator_enable
returns the drvdata of the mfd device rather than the regulator. Fix it.
This was broken while converting to use simplified DT parsing because the
config.dev changed from pdev->dev to pdev->dev.parent for parsing the
parent's of_node.
Fixes: 29dc269a85ef ("regulator: hi6421: Convert to use simplified DT parsing")
Signed-off-by: Axel Lin <axel.lin@ingics.com>
Link: https://lore.kernel.org/r/20210630095959.2411543-1-axel.lin@ingics.com
Signed-off-by: Mark Brown <broonie@kernel.org>
I know nothing about zone_device pages and !device_private pages; but if
try_to_migrate_one() will do nothing for them, then it's better that
try_to_migrate() filter them first, than trawl through all their vmas.
Signed-off-by: Hugh Dickins <hughd@google.com>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Reviewed-by: Alistair Popple <apopple@nvidia.com>
Link: https://lore.kernel.org/lkml/1241d356-8ec9-f47b-a5ec-9b2bf66d242@google.com/
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Yang Shi <shy828301@gmail.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
In the unlikely race case that page_mlock_one() finds VM_LOCKED has been
cleared by the time it got page table lock, page_vma_mapped_walk_done()
must be called before returning, either explicitly, or by a final call
to page_vma_mapped_walk() - otherwise the page table remains locked.
Fixes: cd62734ca60d ("mm/rmap: split try_to_munlock from try_to_unmap")
Signed-off-by: Hugh Dickins <hughd@google.com>
Reviewed-by: Alistair Popple <apopple@nvidia.com>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Reported-by: kernel test robot <oliver.sang@intel.com>
Link: https://lore.kernel.org/lkml/20210711151446.GB4070@xsang-OptiPlex-9020/
Link: https://lore.kernel.org/lkml/f71f8523-cba7-3342-40a7-114abc5d1f51@google.com/
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Yang Shi <shy828301@gmail.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The kernel recovers in due course from missing Mlocked pages: but there
was no point in calling page_mlock() (formerly known as
try_to_munlock()) on a THP, because nothing got done even when it was
found to be mapped in another VM_LOCKED vma.
It's true that we need to be careful: Mlocked accounting of pte-mapped
THPs is too difficult (so consistently avoided); but Mlocked accounting
of only-pmd-mapped THPs is supposed to work, even when multiple mappings
are mlocked and munlocked or munmapped. Refine the tests.
There is already a VM_BUG_ON_PAGE(PageDoubleMap) in page_mlock(), so
page_mlock_one() does not even have to worry about that complication.
(I said the kernel recovers: but would page reclaim be likely to split
THP before rediscovering that it's VM_LOCKED? I've not followed that up)
Fixes: 9a73f61bdb8a ("thp, mlock: do not mlock PTE-mapped file huge pages")
Signed-off-by: Hugh Dickins <hughd@google.com>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Link: https://lore.kernel.org/lkml/cfa154c-d595-406-eb7d-eb9df730f944@google.com/
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Yang Shi <shy828301@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Parallel developments in mm/rmap.c have left behind some out-of-date
comments: try_to_migrate_one() also accepts TTU_SYNC (already commented
in try_to_migrate() itself), and try_to_migrate() returns nothing at
all.
TTU_SPLIT_FREEZE has just been deleted, so reword the comment about it
in mm/huge_memory.c; and TTU_IGNORE_ACCESS was removed in 5.11, so
delete the "recently referenced" comment from try_to_unmap_one() (once
upon a time the comment was near the removed codeblock, but they drifted
apart).
Signed-off-by: Hugh Dickins <hughd@google.com>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Reviewed-by: Alistair Popple <apopple@nvidia.com>
Link: https://lore.kernel.org/lkml/563ce5b2-7a44-5b4d-1dfd-59a0e65932a9@google.com/
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Yang Shi <shy828301@gmail.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
