Linus Torvalds [Sun, 20 Dec 2020 21:08:19 +0000 (13:08 -0800)]
Merge tag '5.11-rc-smb3-part2' of git://git.samba.org/sfrench/cifs-2.6
Pull cifs fixes from Steve French:
"Four small CIFS/SMB3 fixes (witness protocol and reconnect related),
and two that add ability to get and set auditing information in the
security descriptor (SACL), which can be helpful not just for backup
scenarios ("smbinfo secdesc" etc.) but also for improving security"
* tag '5.11-rc-smb3-part2' of git://git.samba.org/sfrench/cifs-2.6:
Add SMB 2 support for getting and setting SACLs
SMB3: Add support for getting and setting SACLs
cifs: Avoid error pointer dereference
cifs: Re-indent cifs_swn_reconnect()
cifs: Unlock on errors in cifs_swn_reconnect()
cifs: Delete a stray unlock in cifs_swn_reconnect()
Linus Torvalds [Sun, 20 Dec 2020 19:21:06 +0000 (11:21 -0800)]
Merge tag 'perf-tools-2020-12-19' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux
Pull perf tools updates from Arnaldo Carvalho de Melo:
"perf record:
- Fix memory leak when using '--user-regs=?' to list registers
aarch64 support:
- Add aarch64 registers to 'perf record's' --user-regs command line
option
aarch64 hw tracing support:
- Decode memory tagging properties
- Improve ARM's auxtrace support
- Add support for ARMv8.3-SPE
perf kvm:
- Add kvm-stat for arm64
perf stat:
- Add --quiet option
Cleanups:
- Fixup function names wrt what is in libperf and what is in
tools/perf
Build:
- Allow building without libbpf in older systems
New kernel features:
- Initial support for data/code page size sample type, more to come
perf annotate:
- Support MIPS instruction extended support
perf stack unwinding:
- Fix separate debug info files when using elfutils' libdw's unwinder
perf vendor events:
- Update Intel's Skylake client events to v50
- Add JSON metrics for ARM's imx8mm DDR Perf
- Support printing metric groups for system PMUs
perf build id:
- Prep work for supporting having the build id provided by the kernel
in PERF_RECORD_MMAP2 metadata events
perf stat:
- Support regex pattern in --for-each-cgroup
pipe mode:
- Allow to use stdio functions for pipe mode
- Support 'perf report's' --header-only for pipe mode
- Support pipe mode display in 'perf evlist'
Documentation:
- Update information about CAP_PERFMON"
* tag 'perf-tools-2020-12-19' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux: (134 commits)
perf mem: Factor out a function to generate sort order
perf sort: Add sort option for data page size
perf script: Support data page size
tools headers UAPI: Update asm-generic/unistd.h
tools headers cpufeatures: Sync with the kernel sources
tools headers UAPI: Sync linux/prctl.h with the kernel sources
tools headers UAPI: Sync linux/fscrypt.h with the kernel sources
tools headers UAPI: Sync linux/const.h with the kernel headers
tools arch x86: Sync the msr-index.h copy with the kernel sources
perf trace beauty: Update copy of linux/socket.h with the kernel sources
tools headers: Update linux/ctype.h with the kernel sources
tools headers: Add conditional __has_builtin()
tools headers: Get tools's linux/compiler.h closer to the kernel's
tools headers UAPI: Sync linux/stat.h with the kernel sources
tools headers: Syncronize linux/build_bug.h with the kernel sources
perf tools: Reformat record's control fd man text
perf config: Fix example command in manpage to conform to syntax specified in the SYNOPSIS section.
perf test: Make sample-parsing test aware of PERF_SAMPLE_{CODE,DATA}_PAGE_SIZE
perf tools: Add support to read build id from compressed elf
perf debug: Add debug_set_file function
...
Document the device tree bindings of the Canaan Kendryte K210 SoC clock
driver in Documentation/devicetree/bindings/clock/canaan,k210-clk.yaml.
The header file include/dt-bindings/clock/k210-clk.h is modified to
include the complete list of IDs for all clocks of the SoC.
Damien Le Moal [Sun, 20 Dec 2020 08:57:24 +0000 (17:57 +0900)]
dt-bindings: Add Canaan vendor prefix
Update Documentation/devicetree/bindings/vendor-prefixes.yaml to
include "canaan" as a vendor prefix for "Canaan Inc.". Canaan is the
vendor of the Kendryte K210 RISC-V SoC.
Linus Torvalds [Sun, 20 Dec 2020 18:44:05 +0000 (10:44 -0800)]
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Pull KVM updates from Paolo Bonzini:
"Much x86 work was pushed out to 5.12, but ARM more than made up for it.
ARM:
- PSCI relay at EL2 when "protected KVM" is enabled
- New exception injection code
- Simplification of AArch32 system register handling
- Fix PMU accesses when no PMU is enabled
- Expose CSV3 on non-Meltdown hosts
- Cache hierarchy discovery fixes
- PV steal-time cleanups
- Allow function pointers at EL2
- Various host EL2 entry cleanups
- Simplification of the EL2 vector allocation
s390:
- memcg accouting for s390 specific parts of kvm and gmap
- selftest for diag318
- new kvm_stat for when async_pf falls back to sync
x86:
- Tracepoints for the new pagetable code from 5.10
- Catch VFIO and KVM irqfd events before userspace
- Reporting dirty pages to userspace with a ring buffer
- SEV-ES host support
- Nested VMX support for wait-for-SIPI activity state
- New feature flag (AVX512 FP16)
- New system ioctl to report Hyper-V-compatible paravirtualization features
Generic:
- Selftest improvements"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (171 commits)
KVM: SVM: fix 32-bit compilation
KVM: SVM: Add AP_JUMP_TABLE support in prep for AP booting
KVM: SVM: Provide support to launch and run an SEV-ES guest
KVM: SVM: Provide an updated VMRUN invocation for SEV-ES guests
KVM: SVM: Provide support for SEV-ES vCPU loading
KVM: SVM: Provide support for SEV-ES vCPU creation/loading
KVM: SVM: Update ASID allocation to support SEV-ES guests
KVM: SVM: Set the encryption mask for the SVM host save area
KVM: SVM: Add NMI support for an SEV-ES guest
KVM: SVM: Guest FPU state save/restore not needed for SEV-ES guest
KVM: SVM: Do not report support for SMM for an SEV-ES guest
KVM: x86: Update __get_sregs() / __set_sregs() to support SEV-ES
KVM: SVM: Add support for CR8 write traps for an SEV-ES guest
KVM: SVM: Add support for CR4 write traps for an SEV-ES guest
KVM: SVM: Add support for CR0 write traps for an SEV-ES guest
KVM: SVM: Add support for EFER write traps for an SEV-ES guest
KVM: SVM: Support string IO operations for an SEV-ES guest
KVM: SVM: Support MMIO for an SEV-ES guest
KVM: SVM: Create trace events for VMGEXIT MSR protocol processing
KVM: SVM: Create trace events for VMGEXIT processing
...
Linus Torvalds [Sun, 20 Dec 2020 18:12:06 +0000 (10:12 -0800)]
Merge tag 'rtc-5.11' of git://git.kernel.org/pub/scm/linux/kernel/git/abelloni/linux
Pull RTC updates from Alexandre Belloni:
"Subsystem:
- Remove nvram ABI. There was no complaints about the deprecation for
the last 3 years.
- Improve RTC device allocation and registration
- Now available for ARCH=um
Drivers:
- at91rm9200: correction and sam9x60 support
- ds1307: improve ACPI support
- mxc: now DT only
- pcf2127: watchdog support now needs the reset-source property
- pcf8523: set range
- rx6110: i2c support"
* tag 'rtc-5.11' of git://git.kernel.org/pub/scm/linux/kernel/git/abelloni/linux: (43 commits)
rtc: pcf2127: only use watchdog when explicitly available
dt-bindings: rtc: add reset-source property
rtc: fix RTC removal
rtc: s3c: Remove dead code related to periodic tick handling
rtc: s3c: Disable all enable (RTC, tick) bits in the probe
rtc: ep93xx: Fix NULL pointer dereference in ep93xx_rtc_read_time
rtc: test: remove debug message
rtc: mxc{,_v2}: enable COMPILE_TEST
rtc: enable RTC framework on ARCH=um
rtc: pcf8523: use BIT
rtc: pcf8523: set range
rtc: pcf8523: switch to devm_rtc_allocate_device
rtc: destroy mutex when releasing the device
rtc: shrink devm_rtc_allocate_device()
rtc: rework rtc_register_device() resource management
rtc: nvmem: emit an error message when nvmem registration fails
rtc: add devm_ prefix to rtc_nvmem_register()
rtc: nvmem: remove nvram ABI
Documentation: list RTC devres helpers in devres.rst
rtc: omap: use devm_pinctrl_register()
...
Pavel Begunkov [Sun, 20 Dec 2020 13:21:44 +0000 (13:21 +0000)]
io_uring: end waiting before task cancel attempts
Get rid of TASK_UNINTERRUPTIBLE and waiting with finish_wait before
going for next iteration in __io_uring_task_cancel(), because
__io_uring_files_cancel() doesn't expect that sheduling is disallowed.
Linus Torvalds [Sun, 20 Dec 2020 18:04:49 +0000 (10:04 -0800)]
Merge tag 'gfs2-for-5.11' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2
Pull gfs2 updates from Andreas Gruenbacher:
- Don't wait for unfreeze of the wrong filesystems
- Remove an obsolete delete_work_func hack and an incorrect
sb_start_write
- Minor documentation updates and cosmetic care
* tag 'gfs2-for-5.11' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2:
gfs2: in signal_our_withdraw wait for unfreeze of _this_ fs only
gfs2: Remove sb_start_write from gfs2_statfs_sync
gfs2: remove trailing semicolons from macro definitions
Revert "GFS2: Prevent delete work from occurring on glocks used for create"
gfs2: Make inode operations static
MAINTAINERS: Add gfs2 bug tracker link
Documentation: Update filesystems/gfs2.rst
Jens Axboe [Sun, 20 Dec 2020 17:45:02 +0000 (10:45 -0700)]
io_uring: make ctx cancel on exit targeted to actual ctx
Before IORING_SETUP_ATTACH_WQ, we could just cancel everything on the
io-wq when exiting. But that's not the case if they are shared, so
cancel for the specific ctx instead.
clk: vc5: Use "idt,voltage-microvolt" instead of "idt,voltage-microvolts"
Commit 45c940184b501fc6 ("dt-bindings: clk: versaclock5: convert to
yaml") accidentally changed "idt,voltage-microvolts" to
"idt,voltage-microvolt" in the DT bindings, while the driver still used
the former.
Update the driver to match the bindings, as
Documentation/devicetree/bindings/property-units.txt actually recommends
using "microvolt".
Paul Cercueil [Sat, 12 Dec 2020 13:57:33 +0000 (13:57 +0000)]
clk: ingenic: Fix divider calculation with div tables
The previous code assumed that a higher hardware value always resulted
in a bigger divider, which is correct for the regular clocks, but is
an invalid assumption when a divider table is provided for the clock.
Perfect example of this is the PLL0_HALF clock, which applies a /2
divider with the hardware value 0, and a /1 divider otherwise.
Jernej Skrabec [Wed, 2 Dec 2020 20:38:17 +0000 (21:38 +0100)]
clk: sunxi-ng: Make sure divider tables have sentinel
Two clock divider tables are missing sentinel at the end. Effect of that
is that clock framework reads past the last entry. Fix that with adding
sentinel at the end.
Terry Zhou [Fri, 6 Nov 2020 10:00:39 +0000 (11:00 +0100)]
clk: mvebu: a3700: fix the XTAL MODE pin to MPP1_9
There is an error in the current code that the XTAL MODE
pin was set to NB MPP1_31 which should be NB MPP1_9.
The latch register of NB MPP1_9 has different offset of 0x8.
Sascha Hauer [Mon, 30 Nov 2020 09:10:33 +0000 (10:10 +0100)]
clk: si5351: Wait for bit clear after PLL reset
Documentation states that SI5351_PLL_RESET_B and SI5351_PLL_RESET_A bits
are self clearing bits, so wait until they are cleared before
continuing.
This fixes a case when the clock doesn't come up properly after a PLL
reset. It worked properly when the frequency was below 900MHz, but with
900MHz it only works when we are waiting for the bit to clear.
Linus Torvalds [Sat, 19 Dec 2020 21:03:12 +0000 (13:03 -0800)]
Merge tag 'close-range-cloexec-unshare-v5.11' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux
Pull close_range fix from Christian Brauner:
"syzbot reported a bug when asking close_range() to unshare the file
descriptor table and making all fds close-on-exec.
If CLOSE_RANGE_UNSHARE the caller will receive a private file
descriptor table in case their file descriptor table is currently
shared before operating on the requested file descriptor range.
For the case where the caller has requested all file descriptors to be
actually closed via e.g. close_range(3, ~0U, CLOSE_RANGE_UNSHARE) the
kernel knows that the caller does not need any of the file descriptors
anymore and will optimize the close operation by only copying all
files in the range from 0 to 3 and no others.
However, if the caller requested CLOSE_RANGE_CLOEXEC together with
CLOSE_RANGE_UNSHARE the caller wants to still make use of the file
descriptors so the kernel needs to copy all of them and can't
optimize.
The original patch didn't account for this and thus could cause oopses
as evidenced by the syzbot report because it assumed that all fds had
been copied. Fix this by handling the CLOSE_RANGE_CLOEXEC case and
copying all fds if the two flags are specified together.
This should've been caught in the selftests but the original patch
didn't cover this case and I didn't catch it during review. So in
addition to the bugfix I'm also adding selftests. They will reliably
reproduce the bug on a non-fixed kernel and allows us to catch
regressions and verify correct behavior.
Note, the kernel selftest tree contained a bunch of changes that made
the original selftest fail to compile so there are small fixups in
here make them compile without warnings"
* tag 'close-range-cloexec-unshare-v5.11' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux:
selftests/core: add regression test for CLOSE_RANGE_UNSHARE | CLOSE_RANGE_CLOEXEC
selftests/core: add test for CLOSE_RANGE_UNSHARE | CLOSE_RANGE_CLOEXEC
selftests/core: handle missing syscall number for close_range
selftests/core: fix close_range_test build after XFAIL removal
close_range: unshare all fds for CLOSE_RANGE_UNSHARE | CLOSE_RANGE_CLOEXEC
Kan Liang [Wed, 16 Dec 2020 18:57:59 +0000 (10:57 -0800)]
perf mem: Factor out a function to generate sort order
Now, "--phys-data" is the only option which impacts the sort order. A
simple "if else" is enough to handle the option. But there will be more
options added, e.g. "--data-page-size", which also impact the sort
order. The code will become too complex to be maintained.
Divide the sort order string into several small pieces. The first piece
is always the default sort string for LOAD/STORE. Appends the specific
sort string if related option is applied.
Linus Torvalds [Sat, 19 Dec 2020 20:50:10 +0000 (12:50 -0800)]
Merge branch 'pcmcia-next' of git://git.kernel.org/pub/scm/linux/kernel/git/brodo/linux
Pull pcmcia updates from Dominik Brodowski:
"Besides a few PCMCIA odd fixes, the NEC VRC4173 CARDU driver is
removed, as it has not compiled in ages"
* 'pcmcia-next' of git://git.kernel.org/pub/scm/linux/kernel/git/brodo/linux:
pcmcia: omap: Fix error return code in omap_cf_probe()
pcmcia: Remove NEC VRC4173 CARDU
pcmcia: db1xxx_ss: remove unneeded semicolon
pcmcia/electra_cf: Fix some return values in 'electra_cf_probe()' in case of error
Linus Torvalds [Sat, 19 Dec 2020 19:58:46 +0000 (11:58 -0800)]
Merge tag 'for-v5.11' of git://git.kernel.org/pub/scm/linux/kernel/git/sre/linux-power-supply
Pull power supply and reset updates from Sebastian Reichel:
"Battery/charger driver changes:
- collie_battery, generic-adc-battery, s3c-adc-battery: convert to
GPIO descriptors (incl ARM board files)
- misc cleanup and fixes
Reset drivers:
- new poweroff driver for force disabling a regulator
- use printk format symbol resolver
- ocelot: add support for Luton and Jaguar2"
* tag 'for-v5.11' of git://git.kernel.org/pub/scm/linux/kernel/git/sre/linux-power-supply: (31 commits)
power: supply: Fix a typo in warning message
Documentation: DT: binding documentation for regulator-poweroff
power: reset: new driver regulator-poweroff
power: supply: ab8500: Use dev_err_probe() for IIO channels
power: supply: ab8500_fg: Request all IRQs as threaded
power: supply: ab8500_charger: Oneshot threaded IRQs
power: supply: ab8500: Convert to dev_pm_ops
power: supply: ab8500: Use local helper
power: supply: wm831x_power: remove unneeded break
power: supply: bq24735: Drop unused include
power: supply: bq24190_charger: Drop unused include
power: supply: generic-adc-battery: Use GPIO descriptors
power: supply: collie_battery: Convert to GPIO descriptors
power: supply: bq24190_charger: fix reference leak
power: supply: s3c-adc-battery: Convert to GPIO descriptors
power: reset: Use printk format symbol resolver
power: supply: axp20x_usb_power: Use power efficient workqueue for debounce
power: supply: axp20x_usb_power: fix typo
power: supply: max8997-charger: Improve getting charger status
power: supply: max8997-charger: Fix platform data retrieval
...
Linus Torvalds [Sat, 19 Dec 2020 19:53:47 +0000 (11:53 -0800)]
Merge tag 'hsi-for-5.11' of git://git.kernel.org/pub/scm/linux/kernel/git/sre/linux-hsi
Pull HSI updates from Sebastian Reichel:
"Misc cleanups"
* tag 'hsi-for-5.11' of git://git.kernel.org/pub/scm/linux/kernel/git/sre/linux-hsi:
HSI: core: fix a kernel-doc markup
HSI: omap_ssi: Don't jump to free ID in ssi_add_controller()
Linus Torvalds [Sat, 19 Dec 2020 19:51:32 +0000 (11:51 -0800)]
Merge tag 'pwm/for-5.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/thierry.reding/linux-pwm
Pull pwm updates from Thierry Reding:
"This is a fairly big release cycle from the PWM framework's point of
view.
There's a large patcheset here which converts drivers to use the new
devm_platform_ioremap_resource() helper and a bunch of minor fixes to
existing drivers. Some of the existing drivers also add support for
more hardware, such as Atmel SAMA 5D2 and Mediatek MT8183.
Finally there's a couple of new drivers for Intel Keem Bay and LGM
SoCs as well as the DesignWare PWM controller"
* tag 'pwm/for-5.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/thierry.reding/linux-pwm: (66 commits)
pwm: sun4i: Remove erroneous else branch
pwm: sl28cpld: Set driver data before registering the PWM chip
pwm: Remove unused function pwmchip_add_inversed()
pwm: imx27: Fix overflow for bigger periods
pwm: bcm2835: Support apply function for atomic configuration
pwm: keembay: Fix build failure with -Os
pwm: core: Use octal permission
pwm: lpss: Make compilable with COMPILE_TEST
pwm: Fix dependencies on HAS_IOMEM
pwm: Use -EINVAL for unsupported polarity
pwm: sti: Remove unnecessary blank line
pwm: sti: Avoid conditional gotos
pwm: Add PWM fan controller driver for LGM SoC
Add DT bindings YAML schema for PWM fan controller of LGM SoC
pwm: Add DesignWare PWM Controller Driver
dt-bindings: pwm: mtk-disp: add MT8167 SoC binding
pwm: mediatek: Add MT8183 SoC support
pwm: mediatek: Always use bus clock
dt-bindings: pwm: pwm-mediatek: Add documentation for MT8183 SoC
pwm: Add PWM driver for Intel Keem Bay
...
Claudiu Beznea [Thu, 19 Nov 2020 15:43:17 +0000 (17:43 +0200)]
clk: at91: sama7g5: register cpu clock
Register CPU clock as being the master clock prescaler. This would
be used by DVFS. The block schema of SAMA7G5's PMC contains also a divider
between master clock prescaler and CPU (PMC_CPU_RATIO.RATIO) but the
frequencies supported by SAMA7G5 could be directly received from
CPUPLL + master clock prescaler and the extra divider would do no work in
case it would be enabled.
Claudiu Beznea [Thu, 19 Nov 2020 15:43:16 +0000 (17:43 +0200)]
clk: at91: clk-master: re-factor master clock
Re-factor master clock driver by splitting it into 2 clocks: prescaller
and divider clocks. Based on registered clock flags the prescaler's rate
could be changed at runtime. This is necessary for platforms supporting
DVFS (e.g. SAMA7G5) where master clock could be changed at run-time.
Claudiu Beznea [Thu, 19 Nov 2020 15:43:15 +0000 (17:43 +0200)]
clk: at91: sama7g5: do not allow cpu pll to go higher than 1GHz
Since CPU PLL feeds both CPU clock and MCK0, MCK0 cannot go higher
than 200MHz and MCK0 maximum prescaller is 5 limit the CPU PLL at
1GHz to avoid MCK0 overclocking while CPU PLL is changed by DVFS.
Claudiu Beznea [Thu, 19 Nov 2020 15:43:14 +0000 (17:43 +0200)]
clk: at91: sama7g5: decrease lower limit for MCK0 rate
On SAMA7G5 CPU clock is changed at run-time by DVFS. Since MCK0 and
CPU clock shares the same parent clock (CPUPLL clock) the MCK0 is
also changed by DVFS to avoid over/under clocking of MCK0 consumers.
The lower limit is changed to be able to set MCK0 accordingly by
DVFS.
Claudiu Beznea [Thu, 19 Nov 2020 15:43:13 +0000 (17:43 +0200)]
clk: at91: sama7g5: remove mck0 from parent list of other clocks
MCK0 is changed at runtime by DVFS. Due to this, since not all IPs
are glitch free aware at MCK0 changes, remove MCK0 from parent list
of other clocks (e.g. generic clock, programmable/system clock, MCKX).
Claudiu Beznea [Thu, 19 Nov 2020 15:43:07 +0000 (17:43 +0200)]
clk: at91: sama7g5: fix compilation error
pmc_data_allocate() has been changed. pmc_data_free() was removed.
Adapt the code taking this into consideration. With this the programmable
clocks were also saved in sama7g5_pmc so that they could be later
referenced.
Uwe Kleine-König [Fri, 20 Nov 2020 13:21:21 +0000 (14:21 +0100)]
clk: bcm: dvp: drop a variable that is assigned to only
The third parameter to devm_platform_get_and_ioremap_resource() is used
only to provide the used resource. As this variable isn't used
afterwards, switch to the function devm_platform_ioremap_resource()
which doesn't provide this output parameter.
Linus Torvalds [Sat, 19 Dec 2020 19:39:50 +0000 (11:39 -0800)]
Merge branch 'akpm' (patches from Andrew)
Merge still more updates from Andrew Morton:
"18 patches.
Subsystems affected by this patch series: mm (memcg and cleanups) and
epoll"
* emailed patches from Andrew Morton <[email protected]>:
mm/Kconfig: fix spelling mistake "whats" -> "what's"
selftests/filesystems: expand epoll with epoll_pwait2
epoll: wire up syscall epoll_pwait2
epoll: add syscall epoll_pwait2
epoll: convert internal api to timespec64
epoll: eliminate unnecessary lock for zero timeout
epoll: replace gotos with a proper loop
epoll: pull all code between fetch_events and send_event into the loop
epoll: simplify and optimize busy loop logic
epoll: move eavail next to the list_empty_careful check
epoll: pull fatal signal checks into ep_send_events()
epoll: simplify signal handling
epoll: check for events when removing a timed out thread from the wait queue
mm/memcontrol:rewrite mem_cgroup_page_lruvec()
mm, kvm: account kvm_vcpu_mmap to kmemcg
mm/memcg: remove unused definitions
mm/memcg: warning on !memcg after readahead page charged
mm/memcg: bail early from swap accounting if memcg disabled
Willem de Bruijn [Fri, 18 Dec 2020 22:05:35 +0000 (14:05 -0800)]
epoll: convert internal api to timespec64
Patch series "add epoll_pwait2 syscall", v4.
Enable nanosecond timeouts for epoll.
Analogous to pselect and ppoll, introduce an epoll_wait syscall
variant that takes a struct timespec instead of int timeout.
This patch (of 4):
Make epoll more consistent with select/poll: pass along the timeout as
timespec64 pointer.
In anticipation of additional changes affecting all three polling
mechanisms:
- add epoll_pwait2 syscall with timespec semantics,
and share poll_select_set_timeout implementation.
- compute slack before conversion to absolute time,
to save one ktime_get_ts64 call.
epoll: eliminate unnecessary lock for zero timeout
We call ep_events_available() under lock when timeout is 0, and then call
it without locks in the loop for the other cases.
Instead, call ep_events_available() without lock for all cases. For
non-zero timeouts, we will recheck after adding the thread to the wait
queue. For zero timeout cases, by definition, user is opportunistically
polling and will have to call epoll_wait again in the future.
Note that this lock was kept in c5a282e9635e9 because the whole loop was
historically under lock.
This patch results in a 1% CPU/RPC reduction in RPC benchmarks.
ep_events_available() is called multiple times around the busy loop logic,
even though the logic is generally not used. ep_reset_busy_poll_napi_id()
is similarly always called, even when busy loop is not used.
Eliminate ep_reset_busy_poll_napi_id() and inline it inside
ep_busy_loop(). Make ep_busy_loop() return whether there are any events
available after the busy loop. This will eliminate unnecessary loads and
branches, and simplifies the loop.
epoll: pull fatal signal checks into ep_send_events()
To simplify the code, pull in checking the fatal signals into
ep_send_events(). ep_send_events() is called only from ep_poll().
Note that, previously, we were always checking fatal events, but it is
checked only if eavail is true. This should be fine because the goal of
that check is to quickly return from epoll_wait() when there is a pending
fatal signal.
epoll: check for events when removing a timed out thread from the wait queue
Patch series "simplify ep_poll".
This patch series is a followup based on the suggestions and feedback by
Linus:
https://lkml.kernel.org/r/CAHk-=wizk=OxUyQPbO8MS41w2Pag1kniUV5WdD5qWL-gq1kjDA@mail.gmail.com
The first patch in the series is a fix for the epoll race in presence of
timeouts, so that it can be cleanly backported to all affected stable
kernels.
The rest of the patch series simplify the ep_poll() implementation. Some
of these simplifications result in minor performance enhancements as well.
We have kept these changes under self tests and internal benchmarks for a
few days, and there are minor (1-2%) performance enhancements as a result.
This patch (of 8):
After abc610e01c66 ("fs/epoll: avoid barrier after an epoll_wait(2)
timeout"), we break out of the ep_poll loop upon timeout, without checking
whether there is any new events available. Prior to that patch-series we
always called ep_events_available() after exiting the loop.
This can cause races and missed wakeups. For example, consider the
following scenario reported by Guantao Liu:
Suppose we have an eventfd added using EPOLLET to an epollfd.
Thread 1: Sleeps for just below 5ms and then writes to an eventfd.
Thread 2: Calls epoll_wait with a timeout of 5 ms. If it sees an
event of the eventfd, it will write back on that fd.
Thread 3: Calls epoll_wait with a negative timeout.
Prior to abc610e01c66, it is guaranteed that Thread 3 will wake up either
by Thread 1 or Thread 2. After abc610e01c66, Thread 3 can be blocked
indefinitely if Thread 2 sees a timeout right before the write to the
eventfd by Thread 1. Thread 2 will be woken up from
schedule_hrtimeout_range and, with evail 0, it will not call
ep_send_events().
To fix this issue:
1) Simplify the timed_out case as suggested by Linus.
2) while holding the lock, recheck whether the thread was woken up
after its time out has reached.
Note that (2) is different from Linus' original suggestion: It do not set
"eavail = ep_events_available(ep)" to avoid unnecessary contention (when
there are too many timed-out threads and a small number of events), as
well as races mentioned in the discussion thread.
This is the first patch in the series so that the backport to stable
releases is straightforward.
Hui Su [Fri, 18 Dec 2020 22:01:41 +0000 (14:01 -0800)]
mm/memcontrol:rewrite mem_cgroup_page_lruvec()
mem_cgroup_page_lruvec() in memcontrol.c and mem_cgroup_lruvec() in
memcontrol.h is very similar except for the param(page and memcg) which
also can be convert to each other.
So rewrite mem_cgroup_page_lruvec() with mem_cgroup_lruvec().
Shakeel Butt [Fri, 18 Dec 2020 22:01:38 +0000 (14:01 -0800)]
mm, kvm: account kvm_vcpu_mmap to kmemcg
A VCPU of a VM can allocate couple of pages which can be mmap'ed by the
user space application. At the moment this memory is not charged to the
memcg of the VMM. On a large machine running large number of VMs or
small number of VMs having large number of VCPUs, this unaccounted
memory can be very significant. So, charge this memory to the memcg of
the VMM. Please note that lifetime of these allocations corresponds to
the lifetime of the VMM.
Alex Shi [Fri, 18 Dec 2020 22:01:31 +0000 (14:01 -0800)]
mm/memcg: warning on !memcg after readahead page charged
Add VM_WARN_ON_ONCE_PAGE() macro.
Since readahead page is charged on memcg too, in theory we don't have to
check this exception now. Before safely remove them all, add a warning
for the unexpected !memcg.
Alex Shi [Fri, 18 Dec 2020 22:01:28 +0000 (14:01 -0800)]
mm/memcg: bail early from swap accounting if memcg disabled
Patch series "bail out early for memcg disable".
These 2 patches are indepenedent from per memcg lru lock, and may
encounter unexpected warning, so let's move out them from per memcg
lru locking patchset.
This patch (of 2):
We could bail out early when memcg wasn't enabled.
selftests/core: add regression test for CLOSE_RANGE_UNSHARE | CLOSE_RANGE_CLOEXEC
This test is a minimalized version of the reproducer given by syzbot
(cf. [1]).
After introducing CLOSE_RANGE_CLOEXEC syzbot reported a crash when
CLOSE_RANGE_CLOEXEC is specified in conjunction with
CLOSE_RANGE_UNSHARE. When CLOSE_RANGE_UNSHARE is specified the caller
will receive a private file descriptor table in case their file
descriptor table is currently shared.
For the case where the caller has requested all file descriptors to be
actually closed via e.g. close_range(3, ~0U, 0) the kernel knows that
the caller does not need any of the file descriptors anymore and will
optimize the close operation by only copying all files in the range from
0 to 3 and no others.
However, if the caller requested CLOSE_RANGE_CLOEXEC together with
CLOSE_RANGE_UNSHARE the caller wants to still make use of the file
descriptors so the kernel needs to copy all of them and can't optimize.
The original patch didn't account for this and thus could cause oopses
as evidenced by the syzbot report. Add tests for this regression.
We first create a huge gap in the fd table. When we now call
CLOSE_RANGE_UNSHARE with a shared fd table and and with ~0U as upper
bound the kernel will only copy up to fd1 file descriptors into the new
fd table. If the kernel is buggy and doesn't handle CLOSE_RANGE_CLOEXEC
correctly it will not have copied all file descriptors and we will oops!
This test passes on a fixed kernel and will trigger an oops on a buggy
kernel.
Tobias Klauser [Fri, 18 Dec 2020 14:54:12 +0000 (15:54 +0100)]
selftests/core: fix close_range_test build after XFAIL removal
XFAIL was removed in commit 9847d24af95c ("selftests/harness: Refactor
XFAIL into SKIP") and its use in close_range_test was already replaced
by commit 1d44d0dd61b6 ("selftests: core: use SKIP instead of XFAIL in
close_range_test.c"). However, commit 23afeaeff3d9 ("selftests: core:
add tests for CLOSE_RANGE_CLOEXEC") introduced usage of XFAIL in
TEST(close_range_cloexec). Use SKIP there as well.
close_range: unshare all fds for CLOSE_RANGE_UNSHARE | CLOSE_RANGE_CLOEXEC
After introducing CLOSE_RANGE_CLOEXEC syzbot reported a crash when
CLOSE_RANGE_CLOEXEC is specified in conjunction with CLOSE_RANGE_UNSHARE.
When CLOSE_RANGE_UNSHARE is specified the caller will receive a private
file descriptor table in case their file descriptor table is currently
shared.
For the case where the caller has requested all file descriptors to be
actually closed via e.g. close_range(3, ~0U, 0) the kernel knows that
the caller does not need any of the file descriptors anymore and will
optimize the close operation by only copying all files in the range from
0 to 3 and no others.
However, if the caller requested CLOSE_RANGE_CLOEXEC together with
CLOSE_RANGE_UNSHARE the caller wants to still make use of the file
descriptors so the kernel needs to copy all of them and can't optimize.
The original patch didn't account for this and thus could cause oopses
as evidenced by the syzbot report because it assumed that all fds had
been copied. Fix this by handling the CLOSE_RANGE_CLOEXEC case.
syzbot reported
==================================================================
BUG: KASAN: null-ptr-deref in instrument_atomic_read include/linux/instrumented.h:71 [inline]
BUG: KASAN: null-ptr-deref in atomic64_read include/asm-generic/atomic-instrumented.h:837 [inline]
BUG: KASAN: null-ptr-deref in atomic_long_read include/asm-generic/atomic-long.h:29 [inline]
BUG: KASAN: null-ptr-deref in filp_close+0x22/0x170 fs/open.c:1274
Read of size 8 at addr 0000000000000077 by task syz-executor511/8522
Jason Andryuk [Wed, 16 Dec 2020 14:08:38 +0000 (09:08 -0500)]
xen: Kconfig: remove X86_64 depends from XEN_512GB
commit bfda93aee0ec ("xen: Kconfig: nest Xen guest options")
accidentally re-added X86_64 as a dependency to XEN_512GB. It was
originally removed in commit a13f2ef168cb ("x86/xen: remove 32-bit Xen
PV guest support"). Remove it again.
Boris Protopopov [Thu, 17 Dec 2020 20:58:08 +0000 (20:58 +0000)]
Add SMB 2 support for getting and setting SACLs
Fix passing of the additional security info via version
operations. Force new open when getting SACL and avoid
reuse of files that were previously open without
sufficient privileges to access SACLs.
Uwe Kleine-König [Fri, 18 Dec 2020 10:10:54 +0000 (11:10 +0100)]
rtc: pcf2127: only use watchdog when explicitly available
Most boards using the pcf2127 chip (in my bubble) don't make use of the
watchdog functionality and the respective output is not connected. The
effect on such a board is that there is a watchdog device provided that
doesn't work.
So only register the watchdog if the device tree has a "reset-source"
property.
Rasmus Villemoes [Fri, 18 Dec 2020 10:10:53 +0000 (11:10 +0100)]
dt-bindings: rtc: add reset-source property
Some RTCs, e.g. the pcf2127, can be used as a hardware watchdog. But
if the reset pin is not actually wired up, the driver exposes a
watchdog device that doesn't actually work.
Provide a standard binding that can be used to indicate that a given
RTC can perform a reset of the machine, similar to wakeup-source.
Kent Overstreet [Fri, 18 Dec 2020 09:07:11 +0000 (04:07 -0500)]
mm/filemap: fix infinite loop in generic_file_buffered_read()
If iter->count is 0 and iocb->ki_pos is page aligned, this causes
nr_pages to be 0.
Then in generic_file_buffered_read_get_pages() find_get_pages_contig()
returns 0 - because we asked for 0 pages, so we call
generic_file_buffered_read_no_cached_page() which attempts to add a page
to the page cache, which fails with -EEXIST, and then we loop. Oops...
Paul Cercueil [Thu, 17 Dec 2020 00:59:45 +0000 (00:59 +0000)]
dt-bindings/display: abt,y030xx067a: Fix binding
The binding should use "unevaluatedProperties" instead of
"additionalProperties", since it is a SPI device and may have
SPI-related Device Tree properties, for instance the "spi-max-frequency"
property that is present in the example.
Eliminate the following yamllint warnings:
./Documentation/devicetree/bindings/clock/imx8qxp-lpcg.yaml
:32:13:[warning] wrong indentation: expected 14 but found 12 (indentation)
:35:9: [warning] wrong indentation: expected 10 but found 8 (indentation)
Zhen Lei [Mon, 7 Dec 2020 04:48:30 +0000 (12:48 +0800)]
dt-bindings: display: eliminate yamllint warnings
Eliminate the following yamllint warnings:
./Documentation/devicetree/bindings/display/bridge/analogix,anx7625.yaml
:52:9: [warning] wrong indentation: expected 6 but found 8 (indentation)
./Documentation/devicetree/bindings/display/bridge/intel,keembay-dsi.yaml
:42:8: [warning] wrong indentation: expected 8 but found 7 (indentation)
:45:8: [warning] wrong indentation: expected 8 but found 7 (indentation)
./Documentation/devicetree/bindings/display/intel,keembay-msscam.yaml
:21:6: [warning] wrong indentation: expected 6 but found 5 (indentation)
./Documentation/devicetree/bindings/display/panel/novatek,nt36672a.yaml
:25:10: [warning] wrong indentation: expected 10 but found 9 (indentation)
Zhen Lei [Fri, 4 Dec 2020 09:38:12 +0000 (17:38 +0800)]
dt-bindings: devapc: add the required property 'additionalProperties'
When I do dt_binding_check for any YAML file, below wanring is always
reported:
xxx/soc/mediatek/devapc.yaml: 'additionalProperties' is a required property
xxx/soc/mediatek/devapc.yaml: ignoring, error in schema:
warning: no schema found in file: xxx/soc/mediatek/devapc.yaml
Zhen Lei [Fri, 4 Dec 2020 09:38:11 +0000 (17:38 +0800)]
dt-bindings: soc: add the required property 'additionalProperties'
When I do dt_binding_check for any YAML file, below wanring is always
reported:
xxx/soc/litex/litex,soc-controller.yaml: 'additionalProperties' is a required property
xxx/soc/litex/litex,soc-controller.yaml: ignoring, error in schema:
warning: no schema found in file: xxx/soc/litex/litex,soc-controller.yaml
Zhen Lei [Fri, 4 Dec 2020 09:38:10 +0000 (17:38 +0800)]
dt-bindings: serial: add the required property 'additionalProperties'
When I do dt_binding_check for any YAML file, below wanring is always
reported:
xxx/serial/litex,liteuart.yaml: 'additionalProperties' is a required property
xxx/serial/litex,liteuart.yaml: ignoring, error in schema:
warning: no schema found in file: xxx/serial/litex,liteuart.yaml
'additionalProperties' is now required by the meta-schema. Add it for
coda. As a result, 'interrupts', 'interrupt-names' and 'power-domains'
need to be reworked to be defined at the top level.
Rob Herring [Thu, 17 Dec 2020 22:34:29 +0000 (16:34 -0600)]
dt-bindings: Fix JSON pointers
The correct syntax for JSON pointers begins with a '/' after the '#'.
Without a '/', the string should be interpreted as a subschema
identifier. The jsonschema module currently doesn't handle subschema
identifiers and incorrectly allows JSON pointers to begin without a '/'.
Let's fix this before it becomes a problem when jsonschema module is
fixed.
uapi: virtio_ids: add missing device type IDs from OASIS spec
The OASIS virtio spec (1.1) defines several IDs that aren't reflected
in the header yet. Fixing this by adding the missing IDs, even though
they're not yet used by the kernel yet.
Dan Carpenter [Fri, 4 Dec 2020 14:23:36 +0000 (17:23 +0300)]
virtio_ring: Fix two use after free bugs
The "vq" struct is added to the "vdev->vqs" list prematurely. If we
encounter an error later in the function then the "vq" is freed, but
since it is still on the list that could lead to a use after free bug.
Dan Carpenter [Fri, 4 Dec 2020 14:23:00 +0000 (17:23 +0300)]
virtio_ring: Cut and paste bugs in vring_create_virtqueue_packed()
There is a copy and paste bug in the error handling of this code and
it uses "ring_dma_addr" three times instead of "device_event_dma_addr"
and "driver_event_dma_addr".
vdpa_sim: split vdpasim_virtqueue's iov field in out_iov and in_iov
vringh_getdesc_iotlb() manages 2 iovs for writable and readable
descriptors. This is very useful for the block device, where for
each request we have both types of descriptor.
Let's split the vdpasim_virtqueue's iov field in out_iov and
in_iov to use them with vringh_getdesc_iotlb().
We are using VIRTIO terminology for "out" (readable by the device)
and "in" (writable by the device) descriptors.
projects / linux.git / log