| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
This patch converts the resource management in PCI es1968 driver with
devres as a clean up. Each manual resource management is converted
with the corresponding devres helper, and the card object release is
managed now via card->private_free instead of a lowlevel snd_device.
This should give no user-visible functional changes.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
This patch converts the resource management in PCI es1938 driver with
devres as a clean up. Each manual resource management is converted
with the corresponding devres helper, and the card object release is
managed now via card->private_free instead of a lowlevel snd_device.
This should give no user-visible functional changes.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
This patch converts the resource management in PCI esn137x drivers
with devres as a clean up. Each manual resource management is
converted with the corresponding devres helper, the devres helper is
used for the DMA buffer page allocations, and the card object release
is managed now via card->private_free instead of a lowlevel
snd_device.
This should give no user-visible functional changes.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
This patch converts the resource management in PCI cs5530 driver with
devres as a clean up. Each manual resource management is converted
with the corresponding devres helper, and the card object release is
managed now via card->private_free instead of a lowlevel snd_device.
This should give no user-visible functional changes.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
This patch converts the resource management in PCI cs4281 driver with
devres as a clean up. Each manual resource management is converted
with the corresponding devres helper, and the card object release is
managed now via card->private_free instead of a lowlevel snd_device.
This should give no user-visible functional changes.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
This patch converts the resource management in PCI cmipci driver with
devres as a clean up. Each manual resource management is converted
with the corresponding devres helper, and the card object release is
managed now via card->private_free instead of a lowlevel snd_device.
This should give no user-visible functional changes.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
This patch converts the resource management in PCI bt87x driver with
devres as a clean up. Each manual resource management is converted
with the corresponding devres helper, and the card object release is
managed now via card->private_free instead of a lowlevel snd_device.
This should give no user-visible functional changes.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
This patch converts the resource management in PCI azt3328 driver with
devres as a clean up. Each manual resource management is converted
with the corresponding devres helper, and the card object release is
managed now via card->private_free instead of a lowlevel snd_device.
This should give no user-visible functional changes.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
This patch converts the resource management in PCI als4000 driver with
devres as a clean up. Each manual resource management is converted
with the corresponding devres helper, and the card object release is
managed now via card->private_free instead of a lowlevel snd_device.
This should give no user-visible functional changes.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
This patch converts the resource management in PCI als300 driver with
devres as a clean up. Each manual resource management is converted
with the corresponding devres helper, and the card object release is
managed now via card->private_free instead of a lowlevel snd_device.
This should give no user-visible functional changes.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
This patch converts the resource management in PCI ad1889 driver with
devres as a clean up. Each manual resource management is converted
with the corresponding devres helper, and the card object release is
managed now via card->private_free instead of a lowlevel snd_device.
Also, the unnecessary ac97 free callbacks are removed, too.
This should give no user-visible functional changes.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
This patch is an attempt to slightly simplify the resource management
in HD-audio code, by using some device-managed APIs. Only a few
resources like PCI enablement and PCI resources managed via devres,
but most of the rest code dealing with HD-audio core stuff couldn't be
changed so much, hence the changes in this patch are pretty small in
the end. A special caveat is needed for the card object: we can't
move the card object release into devres, because the driver is
involved with the component stuff and its unregistiration doesn't work
well from devres release at all.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
Like the previous patch, this patch converts the resource allocations
with device-managed API calls, so that we can reduce resource-free
calls.
The atiixp drivers are simpler than intel8x0, and even the irq can be
allocated with devres.
The end result is a good amount of code reduction.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
This patch refactors the intel8x0 and intel8x0m driver codes using
devres and gets rid of the driver remove callback.
The conversion is fairly straightforward: each API call is replaced
with the device-managed API function, e.g. pci_enable_device() ->
pcim_enable_device(), and so on. The buffer descriptor list is
allocated with a new API, snd_devm_alloc_pages().
A slight code structure change is that the intel8x0 object is
allocated as a card's private_data instead of the own lowlevel
snd_device object. This simplifies the resource management.
And, the take-down procedure is triggered via card->private_free, and
it's registered at the end of the whole initialization, i.e. after the
all resources get properly managed.
The only not-devres-managed resource is the irq handler. Since we
need to release at suspend and re-acquire at resume (otherwise
something weird happens on some machines), this is still managed
manually. But the rest are all freed automatically.
The end result is a good amount of code reduction.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
Give brief explanations about the device-managed resources and the
newly introduced snd_devm_card_new() helper.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
This patch adds a devres-supported helper for requesting an ISA DMA
channel that will be automatically freed at the device unbinding.
It'll be used by quite a few ISA sound drivers.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
As a second step for preliminary to widen the devres usages among
sound drivers, this patch adds a new ALSA core API function,
snd_devm_card_new(), to create a snd_card object via devres.
When a card object is created by this new function, snd_card_free() is
called automatically and the card object resource gets released at the
device unbinding time.
However, the story isn't that simple. A caveat is that we have to
call snd_card_free() at the very first of the whole resource release
procedure, in order to assure that the all exposed devices on
user-space are deleted and sync with processes accessing those devices
before releasing resources.
For achieving it, snd_card_register() adds a new devres action to
trigger snd_card_free() automatically when the given card object is a
"managed" one. Since usually snd_card_register() is the last step of
the initialization, this should work in most cases.
With all these tricks, some drivers can get rid of the whole driver
remove callback code.
About a bit of implementation details: the patch adds two new flags to
snd_card object: managed and releasing. The former indicates that the
object was created via snd_devm_card_new(), and the latter is used for
avoiding the double-free of snd_card_free() calls. Both flags are
fairly internal and likely uninteresting to normal users.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
This is a preparation for allowing devres usages more widely in
various sound drivers. As a first step, this patch adds a new
allocator function, snd_devm_alloc_pages(), to manage the allocated
pages via devres, so that the pages will be automagically released as
device unbinding.
Unlike the old snd_dma_alloc_pages(), the new function returns
directly the snd_dma_buffer pointer. The caller needs NULL-check for
the allocation error appropriately.
Also, since a real device pointer is mandatory for devres,
SNDRV_DMA_TYPE_CONTINUOUS or SNDRV_DMA_TYPE_VMALLOC type can't be used
for this function.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
There is a spelling mistake in the Kconfig text. Fix it.
Signed-off-by: Colin Ian King <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
During probe the parent_clk_id is set to -1 which should not be used to
array index within hsdiv_rates[].
Signed-off-by: Peter Ujfalusi <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
In case of an error within j721e_audio_startup() the domain->active must
be decremented to avoid unbalanced counter.
Signed-off-by: Peter Ujfalusi <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
This is a Lenovo ThinkStation machine which uses the codec alc623.
There are 2 issues on this machine, the 1st one is the pop noise in
the lineout, the 2nd one is there are 2 Front Mics and pulseaudio
can't handle them, After applying the fixup of
ALC623_FIXUP_LENOVO_THINKSTATION_P340 to this machine, the 2 issues
are fixed.
Cc: <[email protected]>
Signed-off-by: Hui Wang <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
The BIOS on MSI Mortar B550m WiFi (MS-7C94) board with AMDGPU seems
disabling the other pins than HDMI although it has more outputs
including DP.
This patch adds the board to the allow list for enabling all pins.
Reported-by: Damjan Georgievski <[email protected]>
Cc: <[email protected]>
Link: https://lore.kernel.org/r/CAEk1YH4Jd0a8vfZxORVu7qg+Zsc-K+pR187ezNq8QhJBPW4gpw@mail.gmail.com
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
SB16 CSP driver may hit potentially a typical ABBA deadlock in two
code paths:
In snd_sb_csp_stop():
spin_lock_irqsave(&p->chip->mixer_lock, flags);
spin_lock(&p->chip->reg_lock);
In snd_sb_csp_load():
spin_lock_irqsave(&p->chip->reg_lock, flags);
spin_lock(&p->chip->mixer_lock);
Also the similar pattern is seen in snd_sb_csp_start().
Although the practical impact is very small (those states aren't
triggered in the same running state and this happens only on a real
hardware, decades old ISA sound boards -- which must be very difficult
to find nowadays), it's a real scenario and has to be fixed.
This patch addresses those deadlocks by splitting the locks in
snd_sb_csp_start() and snd_sb_csp_stop() for avoiding the nested
locks.
Reported-by: Jia-Ju Bai <[email protected]>
Cc: <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
While using the DMIC recording, the garbled data will be captured by the
DMIC. It is caused by the critical power of PLL closed in the jack detect
function.
Signed-off-by: Oder Chiou <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
For Stoneyridge platform, it is required to invoke DMA driver stop
first rather than invoking DWC I2S controller stop.
Enable dai_link structure stop_dma_fist flag to reverse the stop
sequence.
Signed-off-by: Vijendar Mukunda <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
On stream stop, currently CPU DAI stop sequence invoked first
followed by DMA. For Few platforms, it is required to stop the
DMA first before stopping CPU DAI.
Introduced new flag in dai_link structure for reordering stop sequence.
Based on flag check, ASoC core will re-order the stop sequence.
Fixes: 4378f1fbe92405 ("ASoC: soc-pcm: Use different sequence for start/stop trigger")
Signed-off-by: Vijendar Mukunda <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
SoundWire registers are only accessable after sdw components are succesfully
binded. Setup irqs at that point instead of doing at probe.
Signed-off-by: Srinivas Kandagatla <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
A property regex goes under 'patternProperties', not 'properties'
schema. Otherwise, the regex is interpretted as a fixed string.
Fixes: 17c2d247ddd2 ("ASoC: dt-bindings: renesas: rsnd: tidyup properties")
Cc: Mark Brown <[email protected]>
Cc: Kuninori Morimoto <[email protected]>
Cc: [email protected]
Signed-off-by: Rob Herring <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
Currently the snd_compr.lock mutex isn't initialized in the API
functions although the lock is used many places in other code in
compress offload API. It's because the object was expected to be
initialized via snd_compress_register(), but this was never used by
ASoC, which is the only user. Instead, ASoC initializes the mutex by
itself, and this is error-prone.
This patch moves the mutex initialization into the more appropriate
place, snd_compress_new(), for avoiding the missing init.
Reviewed-by: Mark Brown <[email protected]>
Acked-by: Vinod Koul <[email protected]>
Reviewed-by: Peter Ujfalusi <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
snd_compress_register() and snd_compress_deregister() API functions
have been never used by in-tree drivers.
Let's clean up the dead code.
Acked-by: Vinod Koul <[email protected]>
Reviewed-by: Peter Ujfalusi <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
Recently we've added a new usb_mixer element type, USB_MIXER_BESPOKEN,
but it wasn't added in the table in snd_usb_mixer_dump_cval(). This
is no big problem since each bespoken type should have its own dump
method, but it still isn't disallowed to use the standard one, so we
should cover it as well. Along with it, define the table with the
explicit array initializer for avoiding other pitfalls.
Fixes: 785b6f29a795 ("ALSA: usb-audio: scarlett2: Fix wrong resume call")
Reported-by: Pavel Machek <[email protected]>
Cc: <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
currenlty wcd938x has only soundwire interface and depends on
symbols from wcd938x soundwire module, so make this dependency
explicit in Kconfig
Without this one of the randconfig endup setting
CONFIG_SND_SOC_WCD938X=y
CONFIG_SND_SOC_WCD938X_SDW=m
resulting in some undefined reference to wcd938x_sdw* symbols.
Reported-by: kernel test robot <[email protected]>
Fixes: 045442228868 ("ASoC: codecs: wcd938x: add audio routing and Kconfig")
Signed-off-by: Srinivas Kandagatla <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
The ADL descriptor was missing an ACPI power setting, causing the DSP
to enter D3 even with a D0i1-compatible wake-on-voice/hotwording
capture stream.
Fixes: 4ad03f894b3c ('ASoC: SOF: Intel: Update ADL P to use its own descriptor')
Reviewed-by: Ranjani Sridharan <[email protected]>
Signed-off-by: Sathya Prakash M R <[email protected]>
Signed-off-by: Pierre-Louis Bossart <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
The ALC5631 does not like multi-write accesses, avoid them. This fixes:
rt5631 4-001a: Unable to sync registers 0x3a-0x3c. -121
errors on resume from suspend (and all registers after the registers in
the error not being synced).
Inspired by commit 2d30e9494f1e ("ASoC: rt5651: Fix regcache sync errors
on resume") from Hans de Geode, which fixed the same errors on ALC5651.
Signed-off-by: Maxim Schwalm <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
The intel8x0 driver tries to measure the AC97 bus clock at the probe
time because there have been machines that are driven in different
rate (44.1kHz vs 48kHz). This was the past and currently most of
usages of this driver are on VM, and those are certainly fixed with
48k clock, hence it's useless and waste of time to measure.
This patch is an optimization, setting the fixed 48k rate if it's
detected to be running on a VM.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
On Intel platforms, we always enable all pins no matter what BIOS
sets, but for others, we still take the conservative approach, and the
HD-audio HDMI codec driver enables the streams only for the pins that
are enabled by BIOS, and skips the disabled pins. This seems not good
for some machines that have (as expected) broken BIOS, resulting in
the lack of needed outputs.
There is an allow-list in the driver code to enable all pins, but its
addition needs a recompilation. This patch adds a module option for
user's convenience to enable all pins forcibly on the fly. User can
pass snd_hda_codec_hdmi.enable_all_pins=1 option for testing, and once
confirmed, we may add a static entry in force_connect_list[].
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
The reset procedure in had_do_reset() is exactly for the recently
introduced PCM sync_stop ops. Replace the call with the new ops and
clean up the unnecessary code and flags.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
The ice1724 driver looks for the matching card_info entry twice, once
at reading EEPROM and another for the rest tasks in the probe call.
Reduce the second one by referring to the already assigned card_info
entry.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
The variable status is being initialized with a value that is never
read, the assignment is redundant and can be removed.
Addresses-Coverity: ("Unused value")
Signed-off-by: Colin Ian King <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
If a 32-bit application is being used with a 64-bit kernel and is using
the mmap mechanism to write data, then the SNDRV_PCM_IOCTL_SYNC_PTR
ioctl results in calling snd_pcm_ioctl_sync_ptr_compat(). Make this use
pcm_lib_apply_appl_ptr() so that the substream's ack() method, if
defined, is called.
The snd_pcm_sync_ptr() function, used in the 64-bit ioctl case, already
uses snd_pcm_ioctl_sync_ptr_compat().
Fixes: 9027c4639ef1 ("ALSA: pcm: Call ack() whenever appl_ptr is updated")
Signed-off-by: Alan Young <[email protected]>
Cc: <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
Thierry Reding <[email protected]>
Thierry Reding <[email protected]>:
From: Thierry Reding <[email protected]>
This small series addresses a minor issue with how IOMMU support is
wired up on various Tegra generations. Currently the virtual "card"
device is used to allocate DMA memory for, but since that device does
not actually exist, the path to memory cannot be correctly described.
To address this, this series moves to using the ADMAIF as the DMA device
for audio. This is a real device that can have a proper DMA mask set and
with which a stream ID can be associated with in the SMMU. The memory
accesses technically originate from the ADMA controller (that the ADMAIF
uses), but DMA channel are dynamically allocated at runtime while DMA
memory is allocated at driver load time, drivers won't have access to
the ADMA device yet.
Further patches will be required to correct this issue on Tegra186 and
Tegra210, but I wanted to get feedback on this approach first.
Changes in v2:
- add backwards-compatibility fallback
Thierry
Thierry Reding (2):
ASoC: tegra: Use ADMAIF component for DMA allocations
arm64: tegra: Enable audio IOMMU support on Tegra194
arch/arm64/boot/dts/nvidia/tegra194.dtsi | 4 ++++
sound/soc/tegra/tegra_pcm.c | 30 ++++++++++++++----------
2 files changed, 22 insertions(+), 12 deletions(-)
--
2.32.0
|
|
Currently the tlv320aic31xx driver has regulator support but does not
enable the regulators during probe, deferring this until something causes
ASoC to make the card active. It does put the device into cache only mode
but only when the component level probe is called, however if interrupts
are in use the driver will access the regmap before then which if the
regulators are not powered on would cause I/O problems.
Signed-off-by: Mark Brown <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
None of the TAS2505 outputs are stereo, do not pretend they are by
implementing them using SOC*DOUBLE* macros referencing the same
register twice, use SOC*SINGLE* instead. Fix volume ranges and mute
control for the codec according to datasheet.
Fixes: b4525b6196cd7 ("ASoC: tlv320aic32x4: add support for TAS2505")
Signed-off-by: Marek Vasut <[email protected]>
Cc: Claudius Heine <[email protected]>
Cc: Mark Brown <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
DMA memory is currently allocated for the soundcard device, which is a
virtual device added for the sole purpose of "stitching" together the
audio device. It is not a real device and therefore doesn't have a DMA
mask or a description of the path to and from memory of accesses.
Memory accesses really originate from the ADMA controller that provides
the DMA channels used by the PCM component. However, since the DMA
memory is allocated up-front and the DMA channels aren't known at that
point, there is no way of knowing the DMA channel provider at allocation
time.
The next best physical device in the memory path is the ADMAIF. Use it
as the device to allocate DMA memory to. iommus and interconnects device
tree properties can thus be added to the ADMAIF device tree node to
describe the memory access path for audio.
Signed-off-by: Thierry Reding <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
|
|
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 <[email protected]>
Reviewed-by: Shakeel Butt <[email protected]>
Reviewed-by: Alistair Popple <[email protected]>
Link: https://lore.kernel.org/lkml/[email protected]/
Cc: Andrew Morton <[email protected]>
Cc: Jason Gunthorpe <[email protected]>
Cc: Ralph Campbell <[email protected]>
Cc: Christoph Hellwig <[email protected]>
Cc: Yang Shi <[email protected]>
Cc: Kirill A. Shutemov <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
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 <[email protected]>
Reviewed-by: Alistair Popple <[email protected]>
Reviewed-by: Shakeel Butt <[email protected]>
Reported-by: kernel test robot <[email protected]>
Link: https://lore.kernel.org/lkml/20210711151446.GB4070@xsang-OptiPlex-9020/
Link: https://lore.kernel.org/lkml/[email protected]/
Cc: Andrew Morton <[email protected]>
Cc: Jason Gunthorpe <[email protected]>
Cc: Ralph Campbell <[email protected]>
Cc: Christoph Hellwig <[email protected]>
Cc: Yang Shi <[email protected]>
Cc: Kirill A. Shutemov <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
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 <[email protected]>
Reviewed-by: Shakeel Butt <[email protected]>
Acked-by: Kirill A. Shutemov <[email protected]>
Link: https://lore.kernel.org/lkml/[email protected]/
Cc: Andrew Morton <[email protected]>
Cc: Alistair Popple <[email protected]>
Cc: Jason Gunthorpe <[email protected]>
Cc: Ralph Campbell <[email protected]>
Cc: Christoph Hellwig <[email protected]>
Cc: Yang Shi <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
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 <[email protected]>
Reviewed-by: Shakeel Butt <[email protected]>
Reviewed-by: Alistair Popple <[email protected]>
Link: https://lore.kernel.org/lkml/[email protected]/
Cc: Andrew Morton <[email protected]>
Cc: Jason Gunthorpe <[email protected]>
Cc: Ralph Campbell <[email protected]>
Cc: Christoph Hellwig <[email protected]>
Cc: Yang Shi <[email protected]>
Cc: Kirill A. Shutemov <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|