Set i_generation for kernfs inode. This is required to implement
exportfs operations. The generation is 32-bit, so it's possible the
generation wraps up and we find stale files. To reduce the posssibility,
we don't reuse inode numer immediately. When the inode number allocation
wraps, we increase generation number. In this way generation/inode
number consist of a 64-bit number which is unlikely duplicated. This
does make the idr tree more sparse and waste some memory. Since idr
manages 32-bit keys, idr uses a 6-level radix tree, each level covers 6
bits of the key. In a 100k inode kernfs, the worst case will have around
300k radix tree node. Each node is 576bytes, so the tree will use about
~150M memory. Sounds not too bad, if this really is a problem, we should
find better data structure.
Acked-by: Tejun Heo <tj@kernel.org>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Shaohua Li <shli@fb.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
kernfs uses ida to manage inode number. The problem is we can't get
kernfs_node from inode number with ida. Switching to use idr, next patch
will add an API to get kernfs_node from inode number.
Acked-by: Tejun Heo <tj@kernel.org>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Shaohua Li <shli@fb.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Up until recently sync_file were create to export a single dma-fence to
userspace, and so we could canabalise a bit insie dma-fence to mark
whether or not we had enable polling for the sync_file itself. However,
with the advent of syncobj, we do allow userspace to create multiple
sync_files for a single dma-fence. (Similarly, that the sw-sync
validation framework also started returning multiple sync-files wrapping
a single dma-fence for a syncpt also triggering the problem.)
This patch reverts my suggestion in commit e241655373
("dma-buf/sync_file: only enable fence signalling on poll()") to use a
single bit in the shared dma-fence and restores the sync_file->flags for
tracking the bits individually.
Reported-by: Gustavo Padovan <gustavo.padovan@collabora.com>
Fixes: f1e8c67123 ("dma-buf/sw-sync: Use an rbtree to sort fences in the timeline")
Fixes: e9083420bb ("drm: introduce sync objects (v4)")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Sumit Semwal <sumit.semwal@linaro.org>
Cc: Sean Paul <seanpaul@chromium.org>
Cc: Gustavo Padovan <gustavo@padovan.org>
Cc: dri-devel@lists.freedesktop.org
Cc: <drm-intel-fixes@lists.freedesktop.org> # v4.13-rc1+
Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170728212951.7818-1-chris@chris-wilson.co.uk
This has proven immensely useful for debugging memory leaks and
overallocation (which is a rather serious concern on the platform,
given that we typically run at about 256MB of CMA out of up to 1GB
total memory, with framebuffers that are about 8MB ecah).
The state of the art without this is to dump debug logs from every GL
application, guess as to kernel allocations based on bo_stats, and try
to merge that all together into a global picture of memory allocation
state. With this, you can add a couple of calls to the debug build of
the 3D driver and get a pretty detailed view of GPU memory usage from
/debug/dri/0/bo_stats (or when we debug print to dmesg on allocation
failure).
The Mesa side currently labels at the gallium resource level (so you
see that a 1920x20 pixmap has been created, presumably for the window
system panel), but we could extend that to be even more useful with
glObjectLabel() names being sent all the way down to the kernel.
(partial) example of sorted debugfs output with Mesa labeling all
resources:
kernel BO cache: 16392kb BOs (3)
tiling shadow 1920x1080: 8160kb BOs (1)
resource 1920x1080@32/0: 8160kb BOs (1)
scanout resource 1920x1080@32/0: 8100kb BOs (1)
kernel: 8100kb BOs (1)
v2: Use strndup_user(), use lockdep assertion instead of just a
comment, fix an array[-1] reference, extend comment about name
freeing.
Signed-off-by: Eric Anholt <eric@anholt.net>
Link: https://patchwork.freedesktop.org/patch/msgid/20170725182718.31468-2-eric@anholt.net
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Pull KVM fixes from Paolo Bonzini:
"s390:
- SRCU fix
PPC:
- host crash fixes
x86:
- bugfixes, including making nested posted interrupts really work
Generic:
- tweaks to kvm_stat and to uevents"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
KVM: LAPIC: Fix reentrancy issues with preempt notifiers
tools/kvm_stat: add '-f help' to get the available event list
tools/kvm_stat: use variables instead of hard paths in help output
KVM: nVMX: Fix loss of L2's NMI blocking state
KVM: nVMX: Fix posted intr delivery when vcpu is in guest mode
x86: irq: Define a global vector for nested posted interrupts
KVM: x86: do mask out upper bits of PAE CR3
KVM: make pid available for uevents without debugfs
KVM: s390: take srcu lock when getting/setting storage keys
KVM: VMX: remove unused field
KVM: PPC: Book3S HV: Fix host crash on changing HPT size
KVM: PPC: Book3S HV: Enable TM before accessing TM registers
Pull arm64 fixes from Will Deacon:
"I'd been collecting these whilst we debugged a CPU hotplug failure,
but we ended up diagnosing that one to tglx, who has taken a fix via
the -tip tree separately.
We're seeing some NFS issues that we haven't gotten to the bottom of
yet, and we've uncovered some issues with our backtracing too so there
might be another fixes pull before we're done.
Summary:
- Ensure we have a guard page after the kernel image in vmalloc
- Fix incorrect prefetch stride in copy_page
- Ensure irqs are disabled in die()
- Fix for event group validation in QCOM L2 PMU driver
- Fix requesting of PMU IRQs on AMD Seattle
- Minor cleanups and fixes"
* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
arm64: mmu: Place guard page after mapping of kernel image
drivers/perf: arm_pmu: Request PMU SPIs with IRQF_PER_CPU
arm64: sysreg: Fix unprotected macro argmuent in write_sysreg
perf: qcom_l2: fix column exclusion check
arm64/lib: copy_page: use consistent prefetch stride
arm64/numa: Drop duplicate message
perf: Convert to using %pOF instead of full_name
arm64: Convert to using %pOF instead of full_name
arm64: traps: disable irq in die()
arm64: atomics: Remove '&' from '+&' asm constraint in lse atomics
arm64: uaccess: Remove redundant __force from addr cast in __range_ok
Pull device mapper fixes from Mike Snitzer:
- a few DM integrity fixes that improve performance. One that address
inefficiencies in the on-disk journal device layout. Another that
makes use of the block layer's on-stack plugging when writing the
journal.
- a dm-bufio fix for the blk_status_t conversion that went in during
the merge window.
- a few DM raid fixes that address correctness when suspending the
device and a validation fix for validation that occurs during device
activation.
- a couple DM zoned target fixes. Important one being the fix to not
use GFP_KERNEL in the IO path due to concerns about deadlock in
low-memory conditions (e.g. swap over a DM zoned device, etc).
- a DM DAX device fix to make sure dm_dax_flush() is called if the
underlying DAX device is operating as a write cache.
* tag 'for-4.13/dm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm:
dm, dax: Make sure dm_dax_flush() is called if device supports it
dm verity fec: fix GFP flags used with mempool_alloc()
dm zoned: use GFP_NOIO in I/O path
dm zoned: remove test for impossible REQ_OP_FLUSH conditions
dm raid: bump target version
dm raid: avoid mddev->suspended access
dm raid: fix activation check in validate_raid_redundancy()
dm raid: remove WARN_ON() in raid10_md_layout_to_format()
dm bufio: fix error code in dm_bufio_write_dirty_buffers()
dm integrity: test for corrupted disk format during table load
dm integrity: WARN_ON if variables representing journal usage get out of sync
dm integrity: use plugging when writing the journal
dm integrity: fix inefficient allocation of journal space
Pull block fixes from Jens Axboe:
"A small collection of fixes that should go into this series. This
contains:
- NVMe pull request from Christoph, with various fixes for nvme
proper and nvme-fc.
- disable runtime PM for blk-mq for now.
With scsi now defaulting to using blk-mq, this reared its head as
an issue. Longer term we'll fix up runtime PM for blk-mq, for now
just disable it to prevent a hang on laptop resume for some folks.
- blk-mq CPU <-> hw queue map fix from Christoph.
- xen/blkfront pull request from Konrad, with two small fixes for the
blkfront driver.
- a few fixups for nbd from Joseph.
- a stable fix for pblk from Javier"
* 'for-linus' of git://git.kernel.dk/linux-block:
lightnvm: pblk: advance bio according to lba index
nvme: validate admin queue before unquiesce
nbd: clear disconnected on reconnect
nvme-pci: fix HMB size calculation
nvme-fc: revise TRADDR parsing
nvme-fc: address target disconnect race conditions in fcp io submit
nvme: fabrics commands should use the fctype field for data direction
nvme: also provide a UUID in the WWID sysfs attribute
xen/blkfront: always allocate grants first from per-queue persistent grants
xen-blkfront: fix mq start/stop race
blk-mq: map queues to all present CPUs
block: disable runtime-pm for blk-mq
xen-blkfront: Fix handling of non-supported operations
nbd: only set sndtimeo if we have a timeout set
nbd: take tx_lock before disconnecting
nbd: allow multiple disconnects to be sent
Pull MMC fixes from Ulf Hansson:
"Here are a couple of mmc fixes intended for v4.13-rc1.
I have also included a couple of cleanup patches in this pull request
for OMAP2+, related to the omap_hsmmc driver. The reason is because of
the changes are also depending on OMAP SoC specific code, so this
simplifies how to deal with this.
Summary:
MMC host:
- sunxi: Correct time phase settings
- omap_hsmmc: Clean up some dead code
- dw_mmc: Fix message printed for deprecated num-slots DT binding
- dw_mmc: Fix DT documentation"
* tag 'mmc-v4.13-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc:
Documentation: dw-mshc: deprecate num-slots
mmc: dw_mmc: fix the wrong condition check of getting num-slots from DT
mmc: host: omap_hsmmc: remove unused platform callbacks
ARM: OMAP2+: hsmmc.c: Remove dead code
mmc: sunxi: Keep default timing phase settings for new timing mode
Arnd reported some false positive warnings with GCC 7:
drivers/hid/wacom_wac.o: warning: objtool: wacom_bpt3_touch()+0x2a5: stack state mismatch: cfa1=7+8 cfa2=6+16
drivers/iio/adc/vf610_adc.o: warning: objtool: vf610_adc_calculate_rates() falls through to next function vf610_adc_sample_set()
drivers/pwm/pwm-hibvt.o: warning: objtool: hibvt_pwm_get_state() falls through to next function hibvt_pwm_remove()
drivers/pwm/pwm-mediatek.o: warning: objtool: mtk_pwm_config() falls through to next function mtk_pwm_enable()
drivers/spi/spi-bcm2835.o: warning: objtool: .text: unexpected end of section
drivers/spi/spi-bcm2835aux.o: warning: objtool: .text: unexpected end of section
drivers/watchdog/digicolor_wdt.o: warning: objtool: dc_wdt_get_timeleft() falls through to next function dc_wdt_restart()
When GCC 7 detects a potential divide-by-zero condition, it sometimes
inserts a UD2 instruction for the case where the divisor is zero,
instead of letting the hardware trap on the divide instruction.
Objtool doesn't consider UD2 to be fatal unless it's annotated with
unreachable(). So it considers the GCC-generated UD2 to be non-fatal,
and it tries to follow the control flow past the UD2 and gets
confused.
Previously, objtool *did* assume UD2 was always a dead end. That
changed with the following commit:
d1091c7fa3 ("objtool: Improve detection of BUG() and other dead ends")
The motivation behind that change was that Peter was planning on using
UD2 for __WARN(), which is *not* a dead end. However, it turns out
that some emulators rely on UD2 being fatal, so he ended up using
'ud0' instead:
9a93848fe7 ("x86/debug: Implement __WARN() using UD0")
For GCC 4.5+, it should be safe to go back to the previous assumption
that UD2 is fatal, even when it's not annotated with unreachable().
But for pre-4.5 versions of GCC, the unreachable() macro isn't
supported, so such cases of UD2 need to be explicitly annotated as
reachable.
Reported-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Fixes: d1091c7fa3 ("objtool: Improve detection of BUG() and other dead ends")
Link: http://lkml.kernel.org/r/e57fa9dfede25f79487da8126ee9cdf7b856db65.1501188854.git.jpoimboe@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Jonathan writes:
First round of IIO new device support, features and cleanups for the 4.14 cycle.
4 completely new drivers in this set and plenty of other stuff.
One ABI change due to a silly mistake a long time back. Hopefully no
one will notice. It effects the numerical order of consumer device
channels which was the reverse of the obvious. It's going the slow
way to allow us some margin to spot if we have broken userspace or
not (seems unlikely)
New Device Support
* ccs811
- new driver for the Volatile Organic Compounds (VOC) sensor.
* dln2 adc
- new driver for the ADC on this flexible usb board.
* EP93xx
- new driver for this Cirrus logic SoC ADC.
* ltc2471
- new ADC driver support the ltc2471 and ltc2473
* st_accel
- add trivial table entries to support H3LIS331DL, LIS331DL, LIS3LV02DL.
* st_gyro
- add L3GD20H support (again) having fixed the various things that were
broken in the first try. Includes devicetree binding.
* stm32 dac
- add support for the DACs in the STM32F4 series
Features
* Documentation
- add missing power attribute documentation to the ABI docs.
* at91-sama5d2
- add hardware trigger and buffered capture support with bindings.
- suspend and resume functionality.
* bmc150
- support for the BOSC0200 ACPI device id seen on some tablets.
* hdc100x
- devicetree bindings
- document supported devices
- match table and device ids.
* hts221
- support active low interrupts (with bindings)
- open drain mode with bindings.
* htu21
- OF match table and bindings.
* lsm6dsx
- open drain mode with bindings
* ltc2497
- add support for board file based consumer mapping.
* ms5367
- OF match table and bindings.
* mt7622
- binding document and OF match table.
- suspend and resume support.
* rpr0521
- triggered buffer support.
* tsys01
- OF match table and bindings.
Cleanups and minor fixes
* core
- fix ordering of IIO channels to entry numbers when using
iio_map_array_register rather than reversing them.
- use the new %pOF format specifier rather than full name for the
device tree nodes.
* ad7280a
- fix potential issue with macro argument reuse.
* ad7766
- drop a pointless NULL value check as it's done in the gpiod code.
* adis16400
- unsigned -> unsigned int.
* at91 adc
- make some init data static to reduce code size.
* at91-sama5d2 ADC
- make some init data static to reduce code size.
* da311
- make some init data static to reduce code size.
* hid-sensor-rotation
- drop an unnecessary static.
* hts221
- refactor the write_with_mask code.
- move the BDU configuration to probe time as there is no reason for it
to change.
- avoid overwriting reserved data during power-down. This is a fix, but
the infrastructure need was too invasive to send it to mainline except
in a merge window. It's not a regression as it was always wrong.
- avoid reconfigure the sampling frequency multiple times by just
doing it in the write_raw function directly.
- refactor the power_on/off calls into a set_enable.
- move the dry-enable logic into trig_set_state as that is the only
place it was used.
* ina219
- fix polling of ina226 conversion ready flag.
* imx7d
- add vendor name in kconfig for consistency with similar parts.
* mcp3422
- Change initial channel to 0 as it feels more logical.
- Check for some errors in probe.
* meson-saradc
- add a check of of_match_device return value.
* mpu3050
- allow open drain for any interrupt type.
* rockchip adc
- add check on of_match_device return value.
* sca3000
- drop a trailing whitespace.
* stm32 adc
- make array stm32h7_adc_ckmodes_spec static.
* stm32 dac
- fix an error message.
* stm32 timers
- fix clock name in docs to match reality after changes.
* st_accel
- explicit OF table (spi).
- add missing entries to OF table (i2c).
- rename of_device_id table to drop the part name.
- adding missing lis3l02dq entry to bindings.
- rename H3LIS331DL_DRIVER_NAME to line up with similar entries in driver.
* st_gyro
- explicit OF table (spi).
* st_magn
- explicit OF table (spi).
- enable multiread for lis3mdl.
* st_pressure
- explicit OF table (spi).
* st_sensors common.
- move st_sensors_of_i2c_probe and rename to make it available for spi
drivers.
* tsc3472
- don't write an extra byte when writing the ATIME register.
- add a link to the datasheet.
* tsl2x7x - continued staging cleanups
- add of_match_table.
- drop redundant power_state sysfs attribute.
- drop wrapper tsl2x7x_i2c_read.
- clean up i2c calls made in tsl2x7x_als_calibrate.
- refactor the read and write _event_value callbacks to handle additional
elements.
- use usleep_range instead of mdelay.
- check return value from tsl2x7x_invoke_change.
* zpa2326
- add some newline to the end of logging macros.
Pull ACPI fixes from Rafael Wysocki:
"These are two fixups for the suspend-to-idle handling in the ACPI
subsystem after recent changes in that area and two simple fixes of
the ACPI NUMA code.
Specifics:
- Add an ACPI module parameter to allow users to override the new
default behavior on some systems where the EC GPE is not disabled
during suspend-to-idle in case the EC on their systems generates
excessive wakeup events and they want to sacrifice some
functionality (like power button wakeups) for extra battery life
while suspended (Rafael Wysocki).
- Fix flushing of the outstanding EC work in the ACPI core
suspend-to-idle code (Rafael Wysocki).
- Add a missing include and fix a messed-up comment in the ACPI NUMA
code (Ross Zwisler)"
* tag 'acpi-4.13-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
ACPI: NUMA: Fix typo in the full name of SRAT
ACPI: NUMA: add missing include in acpi_numa.h
ACPI / PM / EC: Flush all EC work in acpi_freeze_sync()
ACPI / EC: Add parameter to force disable the GPE on suspend
In order to mark relevant fields while setting the MTPPS register
add field select. Otherwise it can cause a misconfiguration in
firmware.
Fixes: ee7f12205a ('net/mlx5e: Implement 1PPS support')
Signed-off-by: Eugenia Emantayev <eugenia@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
That commit was part of the changes moving x86 to the generic CPU hotplug
interrupt migration code. The force flag was required on x86 before the
hierarchical irqdomain rework, but invoking set_affinity() with force=true
stayed and had no side effects.
At some point in the past, the force flag got repurposed to support the
exynos timer interrupt affinity setting to a not yet online CPU, so the
interrupt controller callback does not verify the supplied affinity mask
against cpu_online_mask.
Setting the flag in the CPU hotplug code causes the cpu online masking to
be blocked on these irq controllers and results in potentially affining an
interrupt to the CPU which is unplugged, i.e. instead of moving it away,
it's just reassigned to it.
As the force flags is not longer needed on x86, it's safe to revert that
patch so the ARM irqchips which use the force flag work again.
Add comments to that effect, so this won't happen again.
Note: The online mask handling should be done in the generic code and the
force flag and the masking in the irq chips removed all together, but
that's not a change possible for 4.13.
Fixes: 77f85e66aa ("genirq/cpuhotplug: Set force affinity flag on hotplug migration")
Reported-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Will Deacon <will.deacon@arm.com>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: LAK <linux-arm-kernel@lists.infradead.org>
Link: http://lkml.kernel.org/r/alpine.DEB.2.20.1707271217590.3109@nanos
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Although HID itself is transport-agnostic, occasionally a driver may
want to interact with the low-level transport that a device is connected
through. To do this, we need to know what kind of bus is in use. The
first guess may be to look at the 'bus' field of the 'struct hid_device',
but this field may be emulated in some cases (e.g. uhid).
More ideally, we can check which ll_driver a device is using. This
function introduces a 'hid_is_using_ll_driver' function and makes the
'struct hid_ll_driver' of the four most common transports accessible
through hid.h.
Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com>
Acked-By: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
The number of supported WIDs, if they are supported at all, can be
limited due to resources. Notifying the user space application the
number of available WIDs allows it to utilize them correctly.
Signed-off-by: Ram Amrani <Ram.Amrani@cavium.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Direct Packet Mode support may be disabled, e.g, due to limited
resources. Notifying the user application prevents wasting cycles
on attempting to send these kind of packets.
Signed-off-by: Ram Amrani <Ram.Amrani@cavium.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Since the PMU register interface is banked per CPU, CPU PMU interrrupts
cannot be handled by a CPU other than the one with the PMU asserting the
interrupt. This means that migrating PMU SPIs, as we do during a CPU
hotplug operation doesn't make any sense and can lead to the IRQ being
disabled entirely if we route a spurious IRQ to the new affinity target.
This has been observed in practice on AMD Seattle, where CPUs on the
non-boot cluster appear to take a spurious PMU IRQ when coming online,
which is routed to CPU0 where it cannot be handled.
This patch passes IRQF_PERCPU for PMU SPIs and forcefully sets their
affinity prior to requesting them, ensuring that they cannot
be migrated during hotplug events. This interacts badly with the DB8500
erratum workaround that ping-pongs the interrupt affinity from the handler,
so we avoid passing IRQF_PERCPU in that case by allowing the IRQ flags
to be overridden in the platdata.
Fixes: 3cf7ee98b8 ("drivers/perf: arm_pmu: move irq request/free into probe")
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Retrieve the actual coalesce value from hardware for every Rx/Tx
queue, instead of Rx/Tx coalesce value cached during set coalesce.
Signed-off-by: Rahul Verma <Rahul.Verma@cavium.com>
Signed-off-by: Yuval Mintz <yuval.mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch add the ethtool support to set RX/Tx coalesce
value to the VF associated Rx/Tx queues.
Signed-off-by: Rahul Verma <Rahul.Verma@cavium.com>
Signed-off-by: Yuval Mintz <yuval.mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The patch adds required driver support for reading/configuring the
Energy Efficient Ethernet (EEE) parameters.
Signed-off-by: Sudarsana Reddy Kalluru <sudarsana.kalluru@cavium.com>
Signed-off-by: Yuval Mintz <yuval.mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit b1f5bfc27a ("sctp: don't dereference ptr before leaving
_sctp_walk_{params, errors}()") tried to fix the issue that it
may overstep the chunk end for _sctp_walk_{params, errors} with
'chunk_end > offset(length) + sizeof(length)'.
But it introduced a side effect: When processing INIT, it verifies
the chunks with 'param.v == chunk_end' after iterating all params
by sctp_walk_params(). With the check 'chunk_end > offset(length)
+ sizeof(length)', it would return when the last param is not yet
accessed. Because the last param usually is fwdtsn supported param
whose size is 4 and 'chunk_end == offset(length) + sizeof(length)'
This is a badly issue even causing sctp couldn't process 4-shakes.
Client would always get abort when connecting to server, due to
the failure of INIT chunk verification on server.
The patch is to use 'chunk_end <= offset(length) + sizeof(length)'
instead of 'chunk_end < offset(length) + sizeof(length)' for both
_sctp_walk_params and _sctp_walk_errors.
Fixes: b1f5bfc27a ("sctp: don't dereference ptr before leaving _sctp_walk_{params, errors}()")
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The CPU hotplug related code of this driver can be simplified by:
1) Consolidating the callbacks into a single state. The CPU thread can be
torn down on the CPU which goes offline. There is no point in delaying
that to the CPU dead state
2) Let the core code invoke the online/offline callbacks and remove the
extra for_each_online_cpu() loops.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Chad Dupuis <chad.dupuis@cavium.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
The CPU hotplug related code of this driver can be simplified by:
1) Consolidating the callbacks into a single state. The CPU thread can be
torn down on the CPU which goes offline. There is no point in delaying
that to the CPU dead state
2) Let the core code invoke the online/offline callbacks and remove the
extra for_each_online_cpu() loops.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Backmerge drm-next with -rc2 in it to pull in a couple stm patches that
were previously incorrectly applied to -misc-next. By picking them up in
the correct manner, git will hopefully fix any errant trees that are out
in the wild.
Signed-off-by: Sean Paul <seanpaul@chromium.org>
The bitmap allocator must keep metadata consistent. The easiest way is
to scan after every allocation for each affected block and the entire
chunk. This is rather expensive.
The free path can take advantage of current contig hints to prevent
scanning within the start and end block. If a scan is needed, it can
be done by scanning backwards from the start and forwards from the end
to identify the entire free area this can be combined with. The blocks
can then be updated by some basic checks rather than complete block
scans.
A chunk scan happens when the freed area makes a page free, a block
free, or spans across blocks. This is necessary as the contig hint at
this point could span across blocks. The check uses the minimum of page
size and the block size to allow for variable sized blocks. There is a
tradeoff here with not updating after every free. It is possible a
contig hint in one block can be merged with the contig hint in the next
block. This means the contig hint can be off by up to a page. However,
if the chunk's contig hint is contained in one block, the contig hint
will be accurate.
Signed-off-by: Dennis Zhou <dennisszhou@gmail.com>
Reviewed-by: Josef Bacik <jbacik@fb.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
This patch introduces the bitmap metadata blocks and adds the skeleton
of the code that will be used to maintain these blocks. Each chunk's
bitmap is made up of full metadata blocks. These blocks maintain basic
metadata to help prevent scanning unnecssarily to update hints. Full
scanning methods are used for the skeleton and will be replaced in the
coming patches. A number of helper functions are added as well to do
conversion of pages to blocks and manage offsets. Comments will be
updated as the final version of each function is added.
There exists a relationship between PAGE_SIZE, PCPU_BITMAP_BLOCK_SIZE,
the region size, and unit_size. Every chunk's region (including offsets)
is page aligned at the beginning to preserve alignment. The end is
aligned to LCM(PAGE_SIZE, PCPU_BITMAP_BLOCK_SIZE) to ensure that the end
can fit with the populated page map which is by page and every metadata
block is fully accounted for. The unit_size is already page aligned, but
must also be aligned with PCPU_BITMAP_BLOCK_SIZE to ensure full metadata
blocks.
Signed-off-by: Dennis Zhou <dennisszhou@gmail.com>
Reviewed-by: Josef Bacik <jbacik@fb.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
The percpu memory allocator is experiencing scalability issues when
allocating and freeing large numbers of counters as in BPF.
Additionally, there is a corner case where iteration is triggered over
all chunks if the contig_hint is the right size, but wrong alignment.
This patch replaces the area map allocator with a basic bitmap allocator
implementation. Each subsequent patch will introduce new features and
replace full scanning functions with faster non-scanning options when
possible.
Implementation:
This patchset removes the area map allocator in favor of a bitmap
allocator backed by metadata blocks. The primary goal is to provide
consistency in performance and memory footprint with a focus on small
allocations (< 64 bytes). The bitmap removes the heavy memmove from the
freeing critical path and provides a consistent memory footprint. The
metadata blocks provide a bound on the amount of scanning required by
maintaining a set of hints.
In an effort to make freeing fast, the metadata is updated on the free
path if the new free area makes a page free, a block free, or spans
across blocks. This causes the chunk's contig hint to potentially be
smaller than what it could allocate by up to the smaller of a page or a
block. If the chunk's contig hint is contained within a block, a check
occurs and the hint is kept accurate. Metadata is always kept accurate
on allocation, so there will not be a situation where a chunk has a
later contig hint than available.
Evaluation:
I have primarily done testing against a simple workload of allocation of
1 million objects (2^20) of varying size. Deallocation was done by in
order, alternating, and in reverse. These numbers were collected after
rebasing ontop of a80099a152. I present the worst-case numbers here:
Area Map Allocator:
Object Size | Alloc Time (ms) | Free Time (ms)
----------------------------------------------
4B | 310 | 4770
16B | 557 | 1325
64B | 436 | 273
256B | 776 | 131
1024B | 3280 | 122
Bitmap Allocator:
Object Size | Alloc Time (ms) | Free Time (ms)
----------------------------------------------
4B | 490 | 70
16B | 515 | 75
64B | 610 | 80
256B | 950 | 100
1024B | 3520 | 200
This data demonstrates the inability for the area map allocator to
handle less than ideal situations. In the best case of reverse
deallocation, the area map allocator was able to perform within range
of the bitmap allocator. In the worst case situation, freeing took
nearly 5 seconds for 1 million 4-byte objects. The bitmap allocator
dramatically improves the consistency of the free path. The small
allocations performed nearly identical regardless of the freeing
pattern.
While it does add to the allocation latency, the allocation scenario
here is optimal for the area map allocator. The area map allocator runs
into trouble when it is allocating in chunks where the latter half is
full. It is difficult to replicate this, so I present a variant where
the pages are second half filled. Freeing was done sequentially. Below
are the numbers for this scenario:
Area Map Allocator:
Object Size | Alloc Time (ms) | Free Time (ms)
----------------------------------------------
4B | 4118 | 4892
16B | 1651 | 1163
64B | 598 | 285
256B | 771 | 158
1024B | 3034 | 160
Bitmap Allocator:
Object Size | Alloc Time (ms) | Free Time (ms)
----------------------------------------------
4B | 481 | 67
16B | 506 | 69
64B | 636 | 75
256B | 892 | 90
1024B | 3262 | 147
The data shows a parabolic curve of performance for the area map
allocator. This is due to the memmove operation being the dominant cost
with the lower object sizes as more objects are packed in a chunk and at
higher object sizes, the traversal of the chunk slots is the dominating
cost. The bitmap allocator suffers this problem as well. The above data
shows the inability to scale for the allocation path with the area map
allocator and that the bitmap allocator demonstrates consistent
performance in general.
The second problem of additional scanning can result in the area map
allocator completing in 52 minutes when trying to allocate 1 million
4-byte objects with 8-byte alignment. The same workload takes
approximately 16 seconds to complete for the bitmap allocator.
V2:
Fixed a bug in pcpu_alloc_first_chunk end_offset was setting the bitmap
using bytes instead of bits.
Added a comment to pcpu_cnt_pop_pages to explain bitmap_weight.
Signed-off-by: Dennis Zhou <dennisszhou@gmail.com>
Reviewed-by: Josef Bacik <jbacik@fb.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Currently dm_dax_flush() is not being called, even if underlying dax
device supports write cache, because DAXDEV_WRITE_CACHE is not being
propagated up to the DM dax device.
If the underlying dax device supports write cache, set
DAXDEV_WRITE_CACHE on the DM dax device. This will cause dm_dax_flush()
to be called.
Fixes: abebfbe2f7 ("dm: add ->flush() dax operation support")
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Acked-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
When CONFIG_EEH=y and CONFIG_VFIO_SPAPR_EEH=n, build fails with the
following:
drivers/vfio/pci/vfio_pci.o: In function `.vfio_pci_release':
vfio_pci.c:(.text+0xa98): undefined reference to `.vfio_spapr_pci_eeh_release'
drivers/vfio/pci/vfio_pci.o: In function `.vfio_pci_open':
vfio_pci.c:(.text+0x1420): undefined reference to `.vfio_spapr_pci_eeh_open'
In this case, vfio_pci.c should use the empty definitions of
vfio_spapr_pci_eeh_open and vfio_spapr_pci_eeh_release functions.
This patch fixes it by guarding these function definitions with
CONFIG_VFIO_SPAPR_EEH, the symbol that controls whether vfio_spapr_eeh.c is
built, which is where the non-empty versions of these functions are. We need to
make use of IS_ENABLED() macro because CONFIG_VFIO_SPAPR_EEH is a tristate
option.
This issue was found during a randconfig build. Logs are here:
http://kisskb.ellerman.id.au/kisskb/buildresult/12982362/
Signed-off-by: Murilo Opsfelder Araujo <mopsfelder@gmail.com>
Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
v4l2_subdev_call is a macro returning whatever the callback return
type is, usually 'int'. With gcc-7 and ccache, this can lead to
many wanings like:
media/platform/pxa_camera.c: In function 'pxa_mbus_build_fmts_xlate':
media/platform/pxa_camera.c:766:27: error: ?: using integer constants in boolean context [-Werror=int-in-bool-context]
while (!v4l2_subdev_call(subdev, pad, enum_mbus_code, NULL, &code)) {
media/atomisp/pci/atomisp2/atomisp_cmd.c: In function 'atomisp_s_ae_window':
media/atomisp/pci/atomisp2/atomisp_cmd.c:6414:52: error: ?: using integer constants in boolean context [-Werror=int-in-bool-context]
if (v4l2_subdev_call(isp->inputs[asd->input_curr].camera,
The problem here is that after preprocessing, we the compiler
sees a variation of
if (a ? 0 : 2)
that it thinks is suspicious.
This replaces the ?: operator with an different expression that
does the same thing in a more easily readable way that cannot
tigger the warning
Link: https://lkml.org/lkml/2017/7/14/156
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Simplify and improve the code so that the PID is always available in
the uevent even when debugfs is not available.
This adds a userspace_pid field to struct kvm, as per Radim's
suggestion, so that the PID can be retrieved on destruction too.
Acked-by: Janosch Frank <frankja@linux.vnet.ibm.com>
Fixes: 286de8f6ac ("KVM: trigger uevents when creating or destroying a VM")
Signed-off-by: Claudio Imbrenda <imbrenda@linux.vnet.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Nothing calls this wrapper anymore, so just remove it and rename the
old function to get rid of the double underscore prefix.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
This patch increases the minimum allocation size of percpu memory to
4-bytes. This change will help minimize the metadata overhead
associated with the bitmap allocator. The assumption is that most
allocations will be of objects or structs greater than 2 bytes with
integers or longs being used rather than shorts.
The first chunk regions are now aligned with the minimum allocation
size. The reserved region is expected to be set as a multiple of the
minimum allocation size. The static region is aligned up and the delta
is removed from the dynamic size. This works because the dynamic size is
increased to be page aligned. If the static size is not minimum
allocation size aligned, then there must be a gap that is added to the
dynamic size. The dynamic size will never be smaller than the set value.
Signed-off-by: Dennis Zhou <dennisszhou@gmail.com>
Reviewed-by: Josef Bacik <jbacik@fb.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
The V4L2 flash class initialisation expects struct led_classdev_flash that
describes an indicator but only uses struct led_classdev which is a field
iled_cdev in the struct. Use struct iled_cdev only.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Jacek Anaszewski <jacek.anaszewski@gmail.com>
Reviewed-by: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
FB helper code falls back to a 1024x768 mode if no outputs are connected
or don't report back any modes upon initialization. This can be annoying
because outputs that are added to FB helper later on can't be used with
FB helper if they don't support a matching mode.
The fallback is in place because VGA connectors can happen to report an
unknown connection status even when they are in fact connected.
Some drivers have custom solutions in place to defer FB helper setup
until at least one output is connected. But the logic behind these
solutions is always the same and there is nothing driver-specific about
it, so a better alterative is to fix the FB helper core and add support
for all drivers automatically.
This patch adds support for deferred FB helper setup. It checks all the
connectors for their connection status, and if all of them report to be
disconnected marks the FB helper as needing deferred setup. Whet setup
is deferred, the FB helper core will automatically retry setup after a
hotplug event, and it will keep trying until it succeeds.
v2: Rebase onto my entirely reworked fbdev helper locking. One big
difference is that this version again drops&reacquires the fbdev lock
(which is now fb_helper->lock, but before this patch series it was
mode_config->mutex), because register_framebuffer must be able to
recurse back into fbdev helper code for the initial screen setup.
v3: __drm_fb_helper_initial_config must hold fb_helper->lock upon
return, I've fumbled that in the deferred setup case (Liviu).
v4: I was blind, redo this all. __drm_fb_helper_initial_config
shouldn't need to reacquire fb_helper->lock, that just confuses
callers. I myself got confused by kernel_fb_helper_lock and somehow
thought it's the same as fb_helper->lock. Tsk.
Also simplify the logic a bit (we don't need two functions to probe
connectors), we can stick much closer to the existing code. And update
some comments I've spotted that are outdated.
v5: Don't pass -EAGAIN to drivers, it's just an internal error code
(Liviu).
v6: Add _and_unlock suffix to clarify locking (Maarten)
Cc: Liviu Dudau <liviu@dudau.co.uk>
Cc: John Stultz <john.stultz@linaro.org>
Cc: Thierry Reding <treding@nvidia.com>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Tested-by: John Stultz <john.stultz@linaro.org>
Signed-off-by: Thierry Reding <treding@nvidia.com> (v1)
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@intel.com>
Reviewed-by: Liviu Dudau <liviu@dudau.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20170706130023.28417-3-daniel.vetter@ffwll.ch
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
I need this to be able to apply the deferred fbdev setup patches, I
need the relevant prep work that landed through the drm-intel tree.
Also squash in conflict fixup from Laurent Pinchart.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
