This RTC driver starts counting from 2000 to avoid Y2K problem. Also it
only supports 100 years range for all RTCs. Provide that info to RTC
framework. Also remove check for 100 years range in s3c_rtc_settime(),
as RTC core won't pass any invalid values to the driver, now that
correct range is set.
Here is the rationale on 100 years range limitation. Info on different
Samsung RTCs (credit goes to Krzysztof Kozlowski):
- All S3C chips have only 8-bit wide year register (can store 100
years range in BCD format)
- S5Pv210 and Exynos chips have 12-bit year register (can store 1000
years range in BCD format)
But in reality we usually can't make use of those 12 bits either:
- RTCs might think that both 2000 and 2100 years are leap years. So
when the YEAR register is 0, RTC goes from 28 Feb to 29 Feb, and
when the YEAR register is 100, RTC also goes from 28 Feb to 29 Feb.
This is of course incorrect: RTC breaks leap year criteria, which
breaks the time contiguity, which leads to inability to use the RTC
after year of 2099. It was found for example on Exynos850 SoC.
- Despite having 12 bits for holding the year value, RTC might
overflow the year value internally much earlier. For example, on
Exynos850 the RTC overflows when YEAR=159, making the next YEAR=0.
This way RTC actually has range of 160 years, not 1000 as one may
think.
All that said, there is no sense in trying to increase the time range
for more than 100 years on RTCs that seem capable of that. It also
doesn't have too much practical value -- current hardware will be
probably obsolete by 2100.
Tested manually on Exynos850 RTC:
$ date -s "1999-12-31 23:59:50"
$ hwclock -w -f /dev/rtc0
$ date -s "2100-01-01 00:00:00"
$ hwclock -w -f /dev/rtc0
$ date -s "2000-01-01 00:00:00"
$ hwclock -w -f /dev/rtc0
$ hwclock -r -f /dev/rtc0
$ date -s "2099-12-31 23:59:50"
$ hwclock -w -f /dev/rtc0
$ hwclock -r -f /dev/rtc0
Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Link: https://lore.kernel.org/r/20211021202256.28517-4-semen.protsenko@linaro.org
Create dedicated functions for I/O operations and BCD conversion. It can
be useful to separate those from representation conversion and other
stuff found in RTC callbacks.
This patch does not introduce any functional changes, it's merely
refactoring change.
Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Link: https://lore.kernel.org/r/20211021202256.28517-3-semen.protsenko@linaro.org
devm_rtc_device_register() is deprecated. Use devm_rtc_allocate_device()
and devm_rtc_register_device() API instead. This change doesn't change
the behavior, but allows for further improvements.
Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Link: https://lore.kernel.org/r/20211021202256.28517-2-semen.protsenko@linaro.org
Some SoCs have an RTC supported by this RTC driver, but do not have an
early clock provider declared here. Currently, this prevents the RTC
driver from probing, because it expects a global struct to already be
allocated. Fix probing the driver by copying the missing pieces from the
clock provider setup function, replacing them with the devm variants.
Signed-off-by: Samuel Holland <samuel@sholland.org>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Link: https://lore.kernel.org/r/20210928080335.36706-7-samuel@sholland.org
Update mailing list, website, and tree. Drop Michal Marek as a
maintainer, who has not participated in a long time.
Signed-off-by: Julia Lawall <Julia.Lawall@inria.fr>
Reviewed-by: Kees Cook <keescook@chromium.org>
Set regmap raw read/write from spi max_transfer_size
so regmap_raw_read/write can split the access into chunks
Signed-off-by: Lucas Tanure <tanureal@opensource.cirrus.com>
Reviewed-by: Charles Keepax <ckeepax@opensource.cirrus.com>
[André: fix build warning]
Signed-off-by: André Almeida <andrealmeid@collabora.com>
Link: https://lore.kernel.org/r/20211021132721.13669-1-andrealmeid@collabora.com
Signed-off-by: Mark Brown <broonie@kernel.org>
TPS80031 driver was upstreamed back in 2013 and never got a user. I noticed
that driver is abandoned while was about to write a patch to switch it to a
new power-off API (that I'm working on). Driver requires platform data that
nobody provides. Instead of changing the dead code, let's remove it.
Dmitry Osipenko (3):
rtc: tps80031: Remove driver
regulator: tps80031: Remove driver
mfd: tps80031: Remove driver
drivers/mfd/Kconfig | 14 -
drivers/mfd/Makefile | 1 -
drivers/mfd/tps80031.c | 526 -----------------
drivers/regulator/Kconfig | 9 -
drivers/regulator/Makefile | 1 -
drivers/regulator/tps80031-regulator.c | 753 -------------------------
drivers/rtc/Kconfig | 8 -
drivers/rtc/Makefile | 1 -
drivers/rtc/rtc-tps80031.c | 324 -----------
include/linux/mfd/tps80031.h | 637 ---------------------
10 files changed, 2274 deletions(-)
delete mode 100644 drivers/mfd/tps80031.c
delete mode 100644 drivers/regulator/tps80031-regulator.c
delete mode 100644 drivers/rtc/rtc-tps80031.c
delete mode 100644 include/linux/mfd/tps80031.h
--
2.32.0
We know the upper bound on # of mixers (ie. two), so lets just allocate
this on the stack.
Fixes:
BUG: sleeping function called from invalid context at include/linux/sched/mm.h:201
in_atomic(): 1, irqs_disabled(): 128, non_block: 0, pid: 0, name: swapper/0
INFO: lockdep is turned off.
irq event stamp: 43642
hardirqs last enabled at (43641): [<ffffffe24dd276bc>] cpuidle_enter_state+0x158/0x25c
hardirqs last disabled at (43642): [<ffffffe24dfff450>] enter_el1_irq_or_nmi+0x10/0x1c
softirqs last enabled at (43620): [<ffffffe24d4103fc>] __do_softirq+0x1e4/0x464
softirqs last disabled at (43615): [<ffffffe24d48bd90>] __irq_exit_rcu+0x104/0x150
CPU: 0 PID: 0 Comm: swapper/0 Tainted: G W 5.15.0-rc3-debug+ #105
Hardware name: Google Lazor (rev1 - 2) with LTE (DT)
Call trace:
dump_backtrace+0x0/0x18c
show_stack+0x24/0x30
dump_stack_lvl+0xa0/0xd4
dump_stack+0x18/0x34
___might_sleep+0x1e0/0x1f0
__might_sleep+0x78/0x8c
slab_pre_alloc_hook.constprop.0+0x48/0x6c
__kmalloc+0xc8/0x21c
dpu_crtc_vblank_callback+0x158/0x1f8
dpu_encoder_vblank_callback+0x70/0xc4
dpu_encoder_phys_vid_vblank_irq+0x50/0x12c
dpu_core_irq+0x1bc/0x1d0
dpu_irq+0x1c/0x28
msm_irq+0x34/0x40
__handle_irq_event_percpu+0x15c/0x308
handle_irq_event_percpu+0x3c/0x90
handle_irq_event+0x54/0x98
handle_level_irq+0xa0/0xd0
handle_irq_desc+0x2c/0x44
generic_handle_domain_irq+0x28/0x34
dpu_mdss_irq+0x90/0xe8
handle_irq_desc+0x2c/0x44
handle_domain_irq+0x54/0x80
gic_handle_irq+0xd4/0x148
call_on_irq_stack+0x2c/0x54
do_interrupt_handler+0x4c/0x64
el1_interrupt+0x30/0xd0
el1h_64_irq_handler+0x18/0x24
el1h_64_irq+0x78/0x7c
arch_local_irq_enable+0xc/0x14
cpuidle_enter+0x44/0x5c
do_idle+0x248/0x268
cpu_startup_entry+0x30/0x48
rest_init+0x188/0x19c
arch_call_rest_init+0x1c/0x28
start_kernel+0x704/0x744
__primary_switched+0xc0/0xc8
Fixes: 78d9b458cc ("drm/msm/dpu: Add CRC support for DPU")
Signed-off-by: Rob Clark <robdclark@chromium.org>
Link: https://lore.kernel.org/r/20211023160016.3322052-2-robdclark@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>
For the upcoming AMX support it's necessary to do a proper integration with
KVM. Currently KVM allocates two FPU structs which are used for saving the user
state of the vCPU thread and restoring the guest state when entering
vcpu_run() and doing the reverse operation before leaving vcpu_run().
With the new fpstate mechanism this can be reduced to one extra buffer by
swapping the fpstate pointer in current:🧵:fpu. This makes the
upcoming support for AMX and XFD simpler because then fpstate information
(features, sizes, xfd) are always consistent and it does not require any
nasty workarounds.
Convert the KVM FPU code over to this new scheme.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20211022185313.019454292@linutronix.de
On the Amlogic AXG series, the TODDR FIFO may get out of sync with the TDM
decoder if the decoder is started before the FIFO. The channel appears
shifted in memory in an unpredictable way.
To fix this, the trick is to start the FIFO before the TDM decoder. This
way the FIFO is already waiting when the 1st channel is produced and it is
correctly placed in memory.
Jerome Brunet (2):
ASoC: meson: axg-card: make links nonatomic
ASoC: meson: axg-tdm-interface: manage formatters in trigger
sound/soc/meson/axg-card.c | 1 +
sound/soc/meson/axg-tdm-interface.c | 26 +++++++++++++++++++++-----
2 files changed, 22 insertions(+), 5 deletions(-)
--
2.33.0
Fixed regulator can't change voltage and regulator_sync_voltage()
returns -EINVAL in this case. Make regulator_sync_voltage() to succeed
for regulators that are incapable to change voltage.
On NVIDIA Tegra power management driver needs to sync voltage and we have
one device (Trimslice) that uses fixed regulator which is getting synced.
The syncing error isn't treated as fatal, but produces a noisy error
message. This patch silences that error.
Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
Link: https://lore.kernel.org/r/20211021183308.27786-1-digetx@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The MFD for SY7636A still isn't merged so we need a dependency even if
it ends up not having a real driver due to the header file.
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Mark Brown <broonie@kernel.org>
For the upcoming AMX support it's necessary to do a proper integration with
KVM. Currently KVM allocates two FPU structs which are used for saving the user
state of the vCPU thread and restoring the guest state when entering
vcpu_run() and doing the reverse operation before leaving vcpu_run().
With the new fpstate mechanism this can be reduced to one extra buffer by
swapping the fpstate pointer in current:🧵:fpu. This makes the
upcoming support for AMX and XFD simpler because then fpstate information
(features, sizes, xfd) are always consistent and it does not require any
nasty workarounds.
Provide:
- An allocator which initializes the state properly
- A replacement for the existing FPU swap mechanim
Aside of the reduced memory footprint, this also makes state switching
more efficient when TIF_FPU_NEED_LOAD is set. It does not require a
memcpy as the state is already correct in the to be swapped out fpstate.
The existing interfaces will be removed once KVM is converted over.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20211022185312.954684740@linutronix.de
Repalce kthread_create/wake_up_process() with kthread_run()
to simplify the code.
Signed-off-by: Cai Huoqing <caihuoqing@baidu.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
When the EEPROM data is stored on a MTD partition print an error message
when reading this MTD partition failed. This is currently happening
often in OpenWrt because the initial data was written with using a flash
driver which ignores the error detection data and now OpenWrt uses a
driver which checks it.
With this patch a better error message is shown:
[ 8.986988] mt7915e 0000:01:00.0: WA Firmware Version: DEV_000000, Build Time: 20201105222323
[ 9.100508] mt7915e 0000:01:00.0: reading EEPROM from mtd factory failed: -117
[ 9.144289] mt7915e: probe of 0000:01:00.0 failed with error -22
mt7915 does not work without an EEPROM, MT7922 still works.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Add debugfs knobs to read MCU utilization, which helps user know
firmware status more easily to narrow down CPU bottleneck issues.
Co-developed-by: Ryder Lee <ryder.lee@mediatek.com>
Signed-off-by: Ryder Lee <ryder.lee@mediatek.com>
Signed-off-by: Shayne Chen <shayne.chen@mediatek.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
With this patch, driver can support single rate, (HE)GI and HE_LTF
configuration through .set_bitrate_mask().
Tested-by: MeiChia Chiu <meichia.chiu@mediatek.com>
Signed-off-by: Ryder Lee <ryder.lee@mediatek.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Make mt7915_mcu_sta_he_tlv() as a part of mt7915_mcu_add_rate_ctrl() as
firmware rate control should get HE rate information from sta_rec_he,
and reduce a global function accordingly.
Signed-off-by: Ryder Lee <ryder.lee@mediatek.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Band0 and band1 share the same hardware, so band0 will stop Tx/Rx when
band1 performs Rx calibration. cal_cache is introduced to solve such
corner cases by moving necessary datas from channel_switch to bootup.
Co-developed-by: Peter Chiu <chui-hao.chiu@mediatek.com>
Signed-off-by: Peter Chiu <chui-hao.chiu@mediatek.com>
Signed-off-by: Ryder Lee <ryder.lee@mediatek.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
For the upcoming AMX support it's necessary to do a proper integration with
KVM. To avoid more nasty hackery in KVM which violate encapsulation extend
struct fpu and fpstate so the fpstate switching can be consolidated and
simplified.
Currently KVM allocates two FPU structs which are used for saving the user
state of the vCPU thread and restoring the guest state when entering
vcpu_run() and doing the reverse operation before leaving vcpu_run().
With the new fpstate mechanism this can be reduced to one extra buffer by
swapping the fpstate pointer in current:🧵:fpu. This makes the
upcoming support for AMX and XFD simpler because then fpstate information
(features, sizes, xfd) are always consistent and it does not require any
nasty workarounds.
Add fpu::__task_fpstate to save the regular fpstate pointer while the task
is inside vcpu_run(). Add some state fields to fpstate to indicate the
nature of the state.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20211022185312.896403942@linutronix.de
commit 7f4b792031 ("mt76: mt7615: add ibss support") introduced IBSS
and commit f4ec7fdf7f ("mt76: mt7615: enable support for mesh")
meshpoint support.
Both used in the "get_omac_idx"-function:
if (~mask & BIT(HW_BSSID_0))
return HW_BSSID_0;
With commit d8d59f66d1 ("mt76: mt7615: support 16 interfaces") the
ibss and meshpoint mode should "prefer hw bssid slot 1-3". However,
with that change the ibss or meshpoint mode will not send any beacon on
the mt7622 wifi anymore. Devices were still able to exchange data but
only if a bssid already existed. Two mt7622 devices will never be able
to communicate.
This commits reverts the preferation of slot 1-3 for ibss and
meshpoint. Only NL80211_IFTYPE_STATION will still prefer slot 1-3.
Tested on Banana Pi R64.
Fixes: d8d59f66d1 ("mt76: mt7615: support 16 interfaces")
Signed-off-by: Nick Hainke <vincent@systemli.org>
Acked-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/20211007225725.2615-1-vincent@systemli.org
ARRAY_SIZE() is of type size_t, so the format specfier should
be %zu instead of %lu.
Fixes this build warning:
../drivers/net/wireless/mediatek/mt76/mt7921/main.c: In function ‘mt7921_get_et_stats’:
../drivers/net/wireless/mediatek/mt76/mt7921/main.c:1024:26: warning: format ‘%lu’ expects argument of type ‘long unsigned int’, but argument 4 has type ‘unsigned int’ [-Wformat=]
dev_err(dev->mt76.dev, "ei: %d SSTATS_LEN: %lu",
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Sean Wang <sean.wang@mediatek.com>
Cc: linux-wireless@vger.kernel.org
Cc: Felix Fietkau <nbd@nbd.name>
Cc: Lorenzo Bianconi <lorenzo.bianconi83@gmail.com>
Cc: Ryder Lee <ryder.lee@mediatek.com>
Cc: Kalle Valo <kvalo@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/20211022233251.29987-1-rdunlap@infradead.org
-----BEGIN PGP SIGNATURE-----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=uhGw
-----END PGP SIGNATURE-----
Merge tag 'block-5.15-2021-10-22' of git://git.kernel.dk/linux-block
Pull block fixes from Jens Axboe:
"Fix for the cgroup code not ussing irq safe stats updates, and one fix
for an error handling condition in add_partition()"
* tag 'block-5.15-2021-10-22' of git://git.kernel.dk/linux-block:
block: fix incorrect references to disk objects
blk-cgroup: blk_cgroup_bio_start() should use irq-safe operations on blkg->iostat_cpu
-----BEGIN PGP SIGNATURE-----
iQJEBAABCAAuFiEEwPw5LcreJtl1+l5K99NY+ylx4KYFAmFzfyQQHGF4Ym9lQGtl
cm5lbC5kawAKCRD301j7KXHgpiO/D/9cqYHpjGDwyftzQFJFfEy9ny6nlLm6lJef
hsrZjC0S649FnXc0YHVLDH3/nos0XsQUYvVJnAMW9EHB6x/95JRUyxzouVz1Fewp
w8Z+lOKymIf3X1LQoB6KQXH5ayohNtUo6HA0Ye/v+iEG+bq/lo9tCMSshpJs3afq
UWW8RxGhrMHfqfgn/8Kkz8fEqZjXz7tssZ+1AFftTxKbk97ZWPahwjvO+xLFWl/m
NbMkHf3xeAvDL747ccrVBOerRZUPySXZElgkPzdjQ4y5HHZrpxt/ZR9Xu7XRzgkJ
7SEmsJ80vla19u3eW/oAn3T4EEGS3qWlei8T47kKIoT1W52S3rqjwsV/30re16GW
sGMWdFiH/GW3VnOxs0/a4/q70je3E9DicSTs4SALTwnvjQ+vrunWgG6ojtxLcieT
Br+km8nmDPug1wxoH2gQLN/EhGcH5hQvi4ZMiMH8MWalYpEkIADOOvAwp0GDwVoE
6DxWeYs57rdSQnSLxDah+mAqBokqswJ/ZmuBOO/iSqXCImehLs0VL1Y+TsThVbRy
epnBdqLk5PbDpODcYTl7on3MD3hpoHjbpnAPah0py57sroiY73sNE/ms1AUsqYPs
fAe5tjFwhGhVWRiZMGOAG6kgTtSdxG134c0Lyvy6xACTR8rJfgcnWMwFJDWK2GDn
ReGYJcgEOA==
=ywLV
-----END PGP SIGNATURE-----
Merge tag 'io_uring-5.15-2021-10-22' of git://git.kernel.dk/linux-block
Pull io_uring fixes from Jens Axboe:
"Two fixes for the max workers limit API that was introduced this
series: one fix for an issue with that code, and one fixing a linked
timeout regression in this series"
* tag 'io_uring-5.15-2021-10-22' of git://git.kernel.dk/linux-block:
io_uring: apply worker limits to previous users
io_uring: fix ltimeout unprep
io_uring: apply max_workers limit to all future users
io-wq: max_worker fixes
If CONFIG_BLOCK isn't set, then it's an empty struct anyway. Just make
it generally available, so we don't break the compile:
kernel/sched/core.c: In function ‘sched_submit_work’:
kernel/sched/core.c:6346:35: error: ‘struct task_struct’ has no member named ‘plug’
6346 | blk_flush_plug(tsk->plug, true);
| ^~
kernel/sched/core.c: In function ‘io_schedule_prepare’:
kernel/sched/core.c:8357:20: error: ‘struct task_struct’ has no member named ‘plug’
8357 | if (current->plug)
| ^~
kernel/sched/core.c:8358:39: error: ‘struct task_struct’ has no member named ‘plug’
8358 | blk_flush_plug(current->plug, true);
| ^~
Reported-by: Nathan Chancellor <nathan@kernel.org>
Fixes: 008f75a20e ("block: cleanup the flush plug helpers")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The current logic of requests with IOSQE_ASYNC is first queueing it to
io-worker, then execute it in a synchronous way. For unbound works like
pollable requests(e.g. read/write a socketfd), the io-worker may stuck
there waiting for events for a long time. And thus other works wait in
the list for a long time too.
Let's introduce a new way for unbound works (currently pollable
requests), with this a request will first be queued to io-worker, then
executed in a nonblock try rather than a synchronous way. Failure of
that leads it to arm poll stuff and then the worker can begin to handle
other works.
The detail process of this kind of requests is:
step1: original context:
queue it to io-worker
step2: io-worker context:
nonblock try(the old logic is a synchronous try here)
|
|--fail--> arm poll
|
|--(fail/ready)-->synchronous issue
|
|--(succeed)-->worker finish it's job, tw
take over the req
This works much better than the old IOSQE_ASYNC logic in cases where
unbound max_worker is relatively small. In this case, number of
io-worker eazily increments to max_worker, new worker cannot be created
and running workers stuck there handling old works in IOSQE_ASYNC mode.
In my 64-core machine, set unbound max_worker to 20, run echo-server,
turns out:
(arguments: register_file, connetion number is 1000, message size is 12
Byte)
original IOSQE_ASYNC: 76664.151 tps
after this patch: 166934.985 tps
Suggested-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Hao Xu <haoxu@linux.alibaba.com>
Link: https://lore.kernel.org/r/20211018133445.103438-1-haoxu@linux.alibaba.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Original code assumed fixed and correct BTF header length. That's not
always the case, though, so fix this bug with a proper additional check.
And use actual header length instead of sizeof(struct btf_header) in
sanity checks.
Fixes: 8a138aed4a ("bpf: btf: Add BTF support to libbpf")
Reported-by: Evgeny Vereshchagin <evvers@ya.ru>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20211023003157.726961-2-andrii@kernel.org
btf_header's str_off+str_len or type_off+type_len can overflow as they
are u32s. This will lead to bypassing the sanity checks during BTF
parsing, resulting in crashes afterwards. Fix by using 64-bit signed
integers for comparison.
Fixes: d812362450 ("libbpf: Fix BTF data layout checks and allow empty BTF")
Reported-by: Evgeny Vereshchagin <evvers@ya.ru>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20211023003157.726961-1-andrii@kernel.org
When enabling CONFIG_CGROUP_BPF, kmemleak can be observed by running
the command as below:
$mount -t cgroup -o none,name=foo cgroup cgroup/
$umount cgroup/
unreferenced object 0xc3585c40 (size 64):
comm "mount", pid 425, jiffies 4294959825 (age 31.990s)
hex dump (first 32 bytes):
01 00 00 80 84 8c 28 c0 00 00 00 00 00 00 00 00 ......(.........
00 00 00 00 00 00 00 00 6c 43 a0 c3 00 00 00 00 ........lC......
backtrace:
[<e95a2f9e>] cgroup_bpf_inherit+0x44/0x24c
[<1f03679c>] cgroup_setup_root+0x174/0x37c
[<ed4b0ac5>] cgroup1_get_tree+0x2c0/0x4a0
[<f85b12fd>] vfs_get_tree+0x24/0x108
[<f55aec5c>] path_mount+0x384/0x988
[<e2d5e9cd>] do_mount+0x64/0x9c
[<208c9cfe>] sys_mount+0xfc/0x1f4
[<06dd06e0>] ret_fast_syscall+0x0/0x48
[<a8308cb3>] 0xbeb4daa8
This is because that since the commit 2b0d3d3e4f ("percpu_ref: reduce
memory footprint of percpu_ref in fast path") root_cgrp->bpf.refcnt.data
is allocated by the function percpu_ref_init in cgroup_bpf_inherit which
is called by cgroup_setup_root when mounting, but not freed along with
root_cgrp when umounting. Adding cgroup_bpf_offline which calls
percpu_ref_kill to cgroup_kill_sb can free root_cgrp->bpf.refcnt.data in
umount path.
This patch also fixes the commit 4bfc0bb2c6 ("bpf: decouple the lifetime
of cgroup_bpf from cgroup itself"). A cgroup_bpf_offline is needed to do a
cleanup that frees the resources which are allocated by cgroup_bpf_inherit
in cgroup_setup_root.
And inside cgroup_bpf_offline, cgroup_get() is at the beginning and
cgroup_put is at the end of cgroup_bpf_release which is called by
cgroup_bpf_offline. So cgroup_bpf_offline can keep the balance of
cgroup's refcount.
Fixes: 2b0d3d3e4f ("percpu_ref: reduce memory footprint of percpu_ref in fast path")
Fixes: 4bfc0bb2c6 ("bpf: decouple the lifetime of cgroup_bpf from cgroup itself")
Signed-off-by: Quanyang Wang <quanyang.wang@windriver.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Roman Gushchin <guro@fb.com>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Link: https://lore.kernel.org/bpf/20211018075623.26884-1-quanyang.wang@windriver.com
1. The ufd in generic_map_update_batch() should be read from batch.map_fd;
2. A call to fdget() should be followed by a symmetric call to fdput().
Fixes: aa2e93b8e5 ("bpf: Add generic support for update and delete batch ops")
Signed-off-by: Xu Kuohai <xukuohai@huawei.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20211019032934.1210517-1-xukuohai@huawei.com
Lorenz Bauer says:
====================
Fix some inconsistencies of bpf_jit_limit on non-x86 platforms.
I've dropped exposing bpf_jit_current since we couldn't agree on
file modes, correct names, etc.
====================
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Expose the maximum amount of useable memory from the riscv JIT.
Signed-off-by: Lorenz Bauer <lmb@cloudflare.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Luke Nelson <luke.r.nels@gmail.com>
Acked-by: Björn Töpel <bjorn@kernel.org>
Link: https://lore.kernel.org/bpf/20211014142554.53120-2-lmb@cloudflare.com
