Since Bluetooth 3.0 there's a HCI command available for reading the
encryption key size of an BR/EDR connection. This information is
essential e.g. for generating an LTK using SMP over BR/EDR, so store
it as part of struct hci_conn.
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Add a new function to register a PWM chip with channels that have their
initial polarity as specified by an additional parameter. This benefits
drivers of controllers that by default operate with inversed polarity
by removing the need to modify the polarity during initialization.
Signed-off-by: Tim Kryger <tim.kryger@gmail.com>
Signed-off-by: Jonathan Richardson <jonathar@broadcom.com>
[thierry.reding@gmail.com: export pwmchip_add_with_polarity()]
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
Currently, leapsecond adjustments are done at tick time. As a result,
the leapsecond was applied at the first timer tick *after* the
leapsecond (~1-10ms late depending on HZ), rather then exactly on the
second edge.
This was in part historical from back when we were always tick based,
but correcting this since has been avoided since it adds extra
conditional checks in the gettime fastpath, which has performance
overhead.
However, it was recently pointed out that ABS_TIME CLOCK_REALTIME
timers set for right after the leapsecond could fire a second early,
since some timers may be expired before we trigger the timekeeping
timer, which then applies the leapsecond.
This isn't quite as bad as it sounds, since behaviorally it is similar
to what is possible w/ ntpd made leapsecond adjustments done w/o using
the kernel discipline. Where due to latencies, timers may fire just
prior to the settimeofday call. (Also, one should note that all
applications using CLOCK_REALTIME timers should always be careful,
since they are prone to quirks from settimeofday() disturbances.)
However, the purpose of having the kernel do the leap adjustment is to
avoid such latencies, so I think this is worth fixing.
So in order to properly keep those timers from firing a second early,
this patch modifies the ntp and timekeeping logic so that we keep
enough state so that the update_base_offsets_now accessor, which
provides the hrtimer core the current time, can check and apply the
leapsecond adjustment on the second edge. This prevents the hrtimer
core from expiring timers too early.
This patch does not modify any other time read path, so no additional
overhead is incurred. However, this also means that the leap-second
continues to be applied at tick time for all other read-paths.
Apologies to Richard Cochran, who pushed for similar changes years
ago, which I resisted due to the concerns about the performance
overhead.
While I suspect this isn't extremely critical, folks who care about
strict leap-second correctness will likely want to watch
this. Potentially a -stable candidate eventually.
Originally-suggested-by: Richard Cochran <richardcochran@gmail.com>
Reported-by: Daniel Bristot de Oliveira <bristot@redhat.com>
Reported-by: Prarit Bhargava <prarit@redhat.com>
Signed-off-by: John Stultz <john.stultz@linaro.org>
Cc: Richard Cochran <richardcochran@gmail.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jiri Bohac <jbohac@suse.cz>
Cc: Shuah Khan <shuahkh@osg.samsung.com>
Cc: Ingo Molnar <mingo@kernel.org>
Link: http://lkml.kernel.org/r/1434063297-28657-4-git-send-email-john.stultz@linaro.org
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
In commit cd7d8498c9 ("tcp: change tcp_skb_pcount() location") we stored
gso_segs in a temporary cache hot location.
This patch does the same for gso_size.
This allows to save 2 cache line misses in tcp xmit path for
the last packet that is considered but not sent because of
various conditions (cwnd, tso defer, receiver window, TSQ...)
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
* zte/soc:
ARM: zx: Add basic defconfig support for ZX296702
ARM: dts: zx: add an initial zx296702 dts and doc
clk: zx: add clock support to zx296702
dt-bindings: Add #defines for ZTE ZX296702 clocks
Enable HW cacheline start padding and align RX WQE size to cacheline
while considering HW start padding. Also, fix dma_unmap call to use
the correct SKB data buffer size.
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Previously we configured HW MTU to be netdev->mtu, actually we
need to configure netdev->mtu + (ETH_HLEN + VLAN_HLEN + ETH_FCS_LEN).
Also, query MTU can not fail, hence make the relevant helper a
void functionm, add mlx5e_set_dev_port_mtu, helper function to
handle MTU setting.
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Samsung updates for v4.2
- add failure(exception) handling
: of_iomap(), of_find_device_by_node() and kstrdup()
- add common poweroff to use PS_HOLD based for all of exynos SoCs
- add exnos_get/set_boot_addr() helper
- constify platform_device_id and irq_domain_ops
- get current parent clock for power domain on/off
- use core_initcall to register power domain driver
- make exynos_core_restart() less verbose
- add support coupled CPUidle for exynos3250
- fix exynos_boot_secondary() return value on timeout
- fix clk_enable() in s3c24xx adc
- fix missing of_node_put() for power domains
* tag 'samsung-mach-1' of git://git.kernel.org/pub/scm/linux/kernel/git/kgene/linux-samsung: (301 commits)
ARM: EXYNOS: register power domain driver from core_initcall
ARM: EXYNOS: use PS_HOLD based poweroff for all supported SoCs
ARM: SAMSUNG: Constify platform_device_id
ARM: EXYNOS: Constify irq_domain_ops
ARM: EXYNOS: add coupled cpuidle support for Exynos3250
ARM: EXYNOS: add exynos_get_boot_addr() helper
ARM: EXYNOS: add exynos_set_boot_addr() helper
ARM: EXYNOS: make exynos_core_restart() less verbose
ARM: EXYNOS: fix exynos_boot_secondary() return value on timeout
ARM: EXYNOS: Get current parent clock for power domain on/off
ARM: SAMSUNG: fix clk_enable() WARNing in S3C24XX ADC
ARM: EXYNOS: Add missing of_node_put() when parsing power domains
ARM: EXYNOS: Handle of_find_device_by_node() and kstrdup() failures
ARM: EXYNOS: Handle of of_iomap() failure
Linux 4.1-rc4
....
Declare nfcmrvl platform_data structure and few DT parameters
for nfcmrvl driver.
Signed-off-by: Vincent Cuissard <cuissard@marvell.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Linux 4.1-rc6
Conflicts:
arch/arm/boot/dts/zynq-7000.dtsi
Resolution summary:
Mainline had an earlier version of the commit, resolve in favor of the
newer patch in next/dt branch.
Some NFC controller supports UART as host interface.
As with SPI, a lot of code can be shared between vendor
drivers. This patch add the generic support of UART and
provides some extension API for vendor specific needs.
This code is strongly inspired by the Bluetooth HCI ldisc
implementation. NCI UART vendor drivers will have to register
themselves to this layer via nci_uart_register.
Underlying tty will have to be configured from user land
thanks to an ioctl.
Signed-off-by: Vincent Cuissard <cuissard@marvell.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Samsung DT updates for v4.2
- for exyos3250
: use s3c6410-rtc instead of exynos3250-rtc
: add JPEG codec node and support it on exynos3250-rinato
: use s3c-rtc clock id for exynos3250-rinato and monk boards
- for exynos4
: add JPEG codec node and syscon property to MIPI DPHY
: remove obsolete MIPI DPHY reg property
: enable s3c-rtc on exynos4412-trats2
- for exynos5
: add syscon property to MIPI DPHY for exynos5420
: enable s3c-rtc on exynos5420-arndale-octa
: add missing irq pinctrl for max77686 on exynos5250-smdk5250
: clk: add bindings for 32kHz clocks from s2mps11
: fix pinctrl for s2mps11-irq on exynos5420-arndale-octa
- for exynos5422-odroidxu3
: add mmc detect gpio and LEDs
: add HS400 support, simple-audio-card and rtc_src clock
* tag 'samsung-dt-3' of git://git.kernel.org/pub/scm/linux/kernel/git/kgene/linux-samsung:
ARM: dts: Add syscon property to the MIPI DPHY for exynos4415
ARM: dts: Remove obsolete MIPI DPHY 'reg' property for exynos4
ARM: dts: Use last parent for clocks during power domain on/off
ARM: dts: add support JPEG codec for exynos3250-rinato
ARM: dts: support simple-audio-card for exynos5420 and exynos5422-odroidxu3
ARM: dts: add jpeg-codec node for exynos4 and exynos4x12
ARM: dts: Enable S3C RTC on exynos4412-trats2 and exynos5420-arndale-octa
ARM: dts: Use define for s3c-rtc clock id for exynos3250-monk
ARM: dts: Use define for s3c-rtc clock id for exynos3250-rinato
ARM: dts: Use s3c6410-rtc instead of exynos3250-rtc for exynos3250/4415
ARM: dts: add 'rtc_src' clock to rtc node for exynos5422-odroidxu3
clk: samsung: Add bindings for 32kHz clocks from s2mps11
ARM: dts: fix pinctrl for s2mps11-irq on exynos5420-arndale-octa
ARM: dts: Add syscon property to the MIPI phy in exynos5420
ARM: dts: Add HS400 support for exynos5422-odroidxu3
ARM: dts: Add LEDs for exynos5422-odroidxu3
ARM: dts: add mmc detect gpio for exynos5422-odroidxu3
ARM: dts: add JPEG codec device node for exynos3250
ARM: dts: Add missing irq pinctrl for max77686 on smdk5250
Fixes userspace compilation errors like:
error: unknown type name ‘uint32_t’
Signed-off-by: Mikko Rapeli <mikko.rapeli@iki.fi>
Signed-off-by: Rob Clark <robdclark@gmail.com>
Using fb modifier flag, support NV12MT format in MDP4.
v2:
- rework the modifier's description [Daniel Vetter's comment]
- drop .set_mode_config() callback [Rob Clark's comment]
v3:
- change VENDOR's name and restrict usage to NV12 [pointed by Daniel]
Signed-off-by: Rob Clark <robdclark@gmail.com>
This patch supplies a new framework API; mbox_request_channel_byname().
It works by supplying the usual client pointer as the first argument and
a string as the second. The API will search the client's node for a
'mbox-names' property then request a channel in the normal way using the
requested string's index as the expected second 'index' argument.
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Jassi Brar <jaswinder.singh@linaro.org>
Now we have the bus and controller code added to find and initialize
the extended capabilities. Now we need to use them in stream code to
decouple stream, manage links etc
So this patch adds the stream handling code for extended capabilities
introduced in preceding patches
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The controller needs to support the new capabilities and allow
reading, parsing and initializing of these capabilities, so this patch
does it
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The new HDA controllers from Intel support new capabilities like
multilink, pipe processing, SPIB, GTS etc In order to use them we
create an extended HDA bus which embed the hdac bus and contains the
fields for extended configurations
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Yet another regression by the transition to regmap cache; for better
usability, we had the fake mute control using the zero amp value for
Conexant codecs, and this was forgotten in the new hda core code.
Since the bits 4-7 are unused for the amp registers (as we follow the
syntax of AMP_GET verb), the bit 4 is now used to indicate the fake
mute. For setting this flag, snd_hda_codec_amp_update() becomes a
function from a simple macro. The bonus is that it gained a proper
function description.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This function can be called by an IOMMU driver to request
that a device's default domain is direct mapped.
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Define the pseudo-PHY address (30) which is used by all Broadcom
Ethernet switches in a shared header file.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We utilize inline functions from the PHY library, make sure that we do
include phy.h in brcmphy.h in order for the code including brcmphy.h not
to have to resolve this inclusion dependency.
Fixes: 705314797b ("net: phy: broadcom: move shadow 0x1C register accessors to brcmphy.h")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Johannes Berg says:
====================
For this round we mostly have fixes:
* mesh fixes from Alexis Green and Chun-Yeow Yeoh,
* a documentation fix from Jakub Kicinski,
* a missing channel release (from Michal Kazior),
* a fix for a signal strength reporting bug (from Sara Sharon),
* handle deauth while associating (myself),
* don't report mangled TX SKB back to userspace for status (myself),
* handle aggregation session timeouts properly in fast-xmit (myself)
However, there are also a few cleanups and one big change that
affects all drivers (and that required me to pull in your tree)
to change the mac80211 HW flags to use an unsigned long bitmap
so that we can extend them more easily - we're running out of
flags even with a cleanup to remove the two unused ones.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
SCM_SECURITY was originally only implemented for datagram sockets,
not for stream sockets. However, SCM_CREDENTIALS is supported on
Unix stream sockets. For consistency, implement Unix stream support
for SCM_SECURITY as well. Also clean up the existing code and get
rid of the superfluous UNIXSID macro.
Motivated by https://bugzilla.redhat.com/show_bug.cgi?id=1224211,
where systemd was using SCM_CREDENTIALS and assumed wrongly that
SCM_SECURITY was also supported on Unix stream sockets.
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Acked-by: Paul Moore <paul@paul-moore.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This adds create/remove window ioctls to create and remove DMA windows.
sPAPR defines a Dynamic DMA windows capability which allows
para-virtualized guests to create additional DMA windows on a PCI bus.
The existing linux kernels use this new window to map the entire guest
memory and switch to the direct DMA operations saving time on map/unmap
requests which would normally happen in a big amounts.
This adds 2 ioctl handlers - VFIO_IOMMU_SPAPR_TCE_CREATE and
VFIO_IOMMU_SPAPR_TCE_REMOVE - to create and remove windows.
Up to 2 windows are supported now by the hardware and by this driver.
This changes VFIO_IOMMU_SPAPR_TCE_GET_INFO handler to return additional
information such as a number of supported windows and maximum number
levels of TCE tables.
DDW is added as a capability, not as a SPAPR TCE IOMMU v2 unique feature
as we still want to support v2 on platforms which cannot do DDW for
the sake of TCE acceleration in KVM (coming soon).
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
[aw: for the vfio related changes]
Acked-by: Alex Williamson <alex.williamson@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The existing implementation accounts the whole DMA window in
the locked_vm counter. This is going to be worse with multiple
containers and huge DMA windows. Also, real-time accounting would requite
additional tracking of accounted pages due to the page size difference -
IOMMU uses 4K pages and system uses 4K or 64K pages.
Another issue is that actual pages pinning/unpinning happens on every
DMA map/unmap request. This does not affect the performance much now as
we spend way too much time now on switching context between
guest/userspace/host but this will start to matter when we add in-kernel
DMA map/unmap acceleration.
This introduces a new IOMMU type for SPAPR - VFIO_SPAPR_TCE_v2_IOMMU.
New IOMMU deprecates VFIO_IOMMU_ENABLE/VFIO_IOMMU_DISABLE and introduces
2 new ioctls to register/unregister DMA memory -
VFIO_IOMMU_SPAPR_REGISTER_MEMORY and VFIO_IOMMU_SPAPR_UNREGISTER_MEMORY -
which receive user space address and size of a memory region which
needs to be pinned/unpinned and counted in locked_vm.
New IOMMU splits physical pages pinning and TCE table update
into 2 different operations. It requires:
1) guest pages to be registered first
2) consequent map/unmap requests to work only with pre-registered memory.
For the default single window case this means that the entire guest
(instead of 2GB) needs to be pinned before using VFIO.
When a huge DMA window is added, no additional pinning will be
required, otherwise it would be guest RAM + 2GB.
The new memory registration ioctls are not supported by
VFIO_SPAPR_TCE_IOMMU. Dynamic DMA window and in-kernel acceleration
will require memory to be preregistered in order to work.
The accounting is done per the user process.
This advertises v2 SPAPR TCE IOMMU and restricts what the userspace
can do with v1 or v2 IOMMUs.
In order to support memory pre-registration, we need a way to track
the use of every registered memory region and only allow unregistration
if a region is not in use anymore. So we need a way to tell from what
region the just cleared TCE was from.
This adds a userspace view of the TCE table into iommu_table struct.
It contains userspace address, one per TCE entry. The table is only
allocated when the ownership over an IOMMU group is taken which means
it is only used from outside of the powernv code (such as VFIO).
As v2 IOMMU supports IODA2 and pre-IODA2 IOMMUs (which do not support
DDW API), this creates a default DMA window for IODA2 for consistency.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
[aw: for the vfio related changes]
Acked-by: Alex Williamson <alex.williamson@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The Subnet Administrator (SA) is not a component of the RoCE spec.
Therefore, it should not be a capability of a RoCE port.
Signed-off-by: Moni Shoua <monis@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Jonathan writes:
Third round of new IIO drivers, cleanups and functionality for the 4.2 cycle.
Given Linus announced a 4.8rc coming up, hopefully time for one more
lot of IIO patches this cycle. Some of these are actually
improvements / fixes for patches earlier in the cycle.
New device support
* st_accel driver - support devices with 8 bit channels.
Cleanup
* A general cleanup of the iio tools under /tools/ from Hartmut.
I'm more than a little embarassed by how bad some of these were! Are well,
much more refined and less bug prone now.
These cover lots of stuff like unhandled error returns, memory leaks as
well as general refactoring to tidy the code up.
* iio_simple_dummy - fix memory leaks in the init functions, drop some
pointless error returns from functions that never generate errors and
make the module parameter explicitly unsigned.
* More buffer handling reworks from Lars-Peter, this time targetting hardware
buffers (a little used corner that looks likely to get more use in the near
future). Specifically:
- Always compute the masklength as inkernel buffer users may need it.
- Add a means of labeling which buffer modes a given buffer implementation
supports.
- In the case of hardware buffers, require strict scan matching rather than
matching to a superset. Currently the demux is bypassed by these drivers
(this may well not change for efficiency reasons) so allowing a superset
of channels to be selected would otherwise lead to more data than requested
confusing userspace.
Driver funcationality improvments
* mmc35240 - adds a compensation to the raw values as borrowed form Memsic's
own input driver.
* mma8452
- event support
- event debouncing
- high pass filter configuration
- triggers
* vf610 - allow conversion mode to be adjusted
Fixlets
* mmc35240
- Off by one error that by coincidence had no real effect.
- i2c_device_name should be lowercase.
- Lack of null terminator at end of attributes array.
- Avoid computing the fractional part of the magnetic field by moving
the scaling into userspace where floating point is available to simplify
the maths.
- Use a smaller sleep before assuming the measurement is done. This is
safe and improves the possible polling rate.
- Fix sensitivity on z-axis - datasheet disagrees with Memsic's releasedd
code and the value used in the code seems to be correct.
* stk3310 - make a local variable signed to ensure error handling works.
* twl4030
- fix calculation of the temperature sense current - bug unlikely
to have ever been noticed as the difference is small.
- Fix errors in descriptions.
The i.MX device tree changes for 4.2:
- Add device tree for i.MX7D SoC and imx7d-sdb board
- New i.MX6 board support: Armadeus Systems APF6, Gateworks GW5510,
and aristainetos2 boards
- Change LVDS to use simple-panel for nitrogen6x and sabrelite boards
- Add Wifi/Bluetooth devices support for cubox-i board
- Remove unused regulators and correct OTG roles setting for
imx6sl-warp board
- Add I2C support for imx23-olinuxino board
- Move imx6qdl HDMI device to a better place
- Add power-domain for imx6qdl CODA device
* tag 'imx-dt-4.2' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux: (24 commits)
ARM: dts: imx6dl: add imx6dl gpt specific compatible string
ARM: dts: imx6: add DT for aristainetos2 board
ARM: dts: cubox-i/hummingboard: Fix the license text
ARM: dts: sabrelite: use simple-panel instead of display-timings for LVDS0
ARM: dts: nitrogen6x: use simple-panel instead of display-timings for LVDS0
ARM: dts: add imx7d-sdb support
ARM: dts: add imx7d soc dtsi file
ARM: dts: Armadeus Systems APF6 family support (i.MX6)
ARM: dts: vf610: Nomenclature fixup for PTC12 pin used in RMII mode.
ARM: dts: cubox-i: add support for Broadcom Wifi/Bluetooth devices
Document: dt: binding: imx: update document for imx7d support
ARM: dts: imx6qdl: Add power-domain phandle to CODA device node
ARM: dts: Gateworks GW5510 support (i.MX6)
ARM: dts: imx6sl-warp: Fix OTG roles
ARM: dts: imx6sl-warp: Remove USB regulators
ARM: dts: imx6sl-warp: Remove unused regulator
ARM: dts: add pinfunc include file to support imx7d
ARM: mxs: fix in tree users of ssd1306
ARM: dts: imx6qdl-hummingboard: Add PCIe support
ARM: dts: imx23-olinuxino: Add i2c support
...
The i.MX SoC updates for 4.2:
- Add new SoC i.MX7D support, which integrates two Cortex-A7 and one
Cortex-M4 cores.
- Support suspend from IRAM on i.MX53, so that DDR pins can be set to
high impedance for more power saving during suspend.
- Move i.MX clock drivers from arch/arm/mach-imx to drivers/clk/imx.
- Move i.MX GPT timer driver from arch/arm/mach-imx into
drivers/clocksource.
- A couple of clock driver update for VF610 and i.MX6Q.
- A few random code correction and improvement.
* tag 'imx-soc-4.2' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux: (44 commits)
ARM: imx: imx7d requires anatop
clocksource: timer-imx-gpt: remove include of <asm/mach/time.h>
ARM: imx: move timer driver into drivers/clocksource
ARM: imx: remove platform headers from timer driver
ARM: imx: provide gpt device specific irq functions
ARM: imx: get rid of variable timer_base
ARM: imx: define gpt register offset per device type
ARM: imx: move clock event variables into imx_timer
ARM: imx: set up .set_next_event hook via imx_gpt_data
ARM: imx: setup tctl register in device specific function
ARM: imx: initialize gpt device type for DT boot
ARM: imx: define an enum for gpt timer device type
ARM: imx: move timer resources into a structure
ARM: imx: use relaxed IO accessor in timer driver
ARM: imx: make imx51/3 suspend optional
ARM: clk-imx6q: refine sata's parent
ARM: imx: clk-v610: Add clock for I2C2 and I2C3
ARM: mach-imx: iomux-imx31: Use DECLARE_BITMAP
ARM: imx: add imx7d clk tree support
ARM: clk: imx: update pllv3 to support imx7
...
Conflicts:
arch/arm/mach-imx/Kconfig
It has been exceptionally useful to exercise the logic that handles
local immediate errors and RDMA connection loss. To enable
developers to test this regularly and repeatably, add logic to
simulate connection loss every so often.
Fault injection is disabled by default. It is enabled with
$ sudo echo xxx > /sys/kernel/debug/sunrpc/inject_fault/disconnect
where "xxx" is a large positive number of transport method calls
before a disconnect. A value of several thousand is usually a good
number that allows reasonable forward progress while still causing a
lot of connection drops.
These hooks are disabled when SUNRPC_DEBUG is turned off.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
This was only ever set to nfs_writeback_release_common(), a function
which is completely empty. Let's just drop this function pointer and
simplify the code a bit.
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
RDMA xprts don't have a sock_xprt, but an rdma_xprt, so the
xs_swapper_enable/disable functions will likely oops when fed an RDMA
xprt. Turn these functions into rpc_xprt_ops so that that doesn't
occur. For now the RDMA versions are no-ops that just return -EINVAL
on an attempt to swapon.
Cc: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Jeff Layton <jeff.layton@primarydata.com>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Split xs_swapper into enable/disable functions and eliminate the
"enable" flag.
Currently, it's racy if you have multiple swapon/swapoff operations
running in parallel over the same xprt. Also fix it so that we only
set it to a memalloc socket on a 0->1 transition and only clear it
on a 1->0 transition.
Cc: Mel Gorman <mgorman@suse.de>
Signed-off-by: Jeff Layton <jeff.layton@primarydata.com>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Jerome reported seeing a warning pop when working with a swapfile on
NFS. The nfs_swap_activate can end up calling sk_set_memalloc while
holding the rcu_read_lock and that function can sleep.
To fix that, we need to take a reference to the xprt while holding the
rcu_read_lock, set the socket up for swapping and then drop that
reference. But, xprt_put is not exported and having NFS deal with the
underlying xprt is a bit of layering violation anyway.
Fix this by adding a set of activate/deactivate functions that take a
rpc_clnt pointer instead of an rpc_xprt, and have nfs_swap_activate and
nfs_swap_deactivate call those.
Also, add a per-rpc_clnt atomic counter to keep track of the number of
active swapfiles associated with it. When the counter does a 0->1
transition, we enable swapping on the xprt, when we do a 1->0 transition
we disable swapping on it.
This also allows us to be a bit more selective with the RPC_TASK_SWAPPER
flag. If non-swapper and swapper clnts are sharing a xprt, then we only
need to flag the tasks from the swapper clnt with that flag.
Acked-by: Mel Gorman <mgorman@suse.de>
Reported-by: Jerome Marchand <jmarchan@redhat.com>
Signed-off-by: Jeff Layton <jeff.layton@primarydata.com>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Commit 73f7d1ca32 "ACPI / init: Run acpi_early_init() before
timekeeping_init()" moved the ACPI subsystem initialization,
including the ACPI mode enabling, to an earlier point in the
initialization sequence, to allow the timekeeping subsystem
use ACPI early. Unfortunately, that resulted in boot regressions
on some systems and the early ACPI initialization was moved toward
its original position in the kernel initialization code by commit
c4e1acbb35 "ACPI / init: Invoke early ACPI initialization later".
However, that turns out to be insufficient, as boot is still broken
on the Tyan S8812 mainboard.
To fix that issue, split the ACPI early initialization code into
two pieces so the majority of it still located in acpi_early_init()
and the part switching over the platform into the ACPI mode goes into
a new function, acpi_subsystem_init(), executed at the original early
ACPI initialization spot.
That fixes the Tyan S8812 boot problem, but still allows ACPI
tables to be loaded earlier which is useful to the EFI code in
efi_enter_virtual_mode().
Link: https://bugzilla.kernel.org/show_bug.cgi?id=97141
Fixes: 73f7d1ca32 "ACPI / init: Run acpi_early_init() before timekeeping_init()"
Reported-and-tested-by: Marius Tolzmann <tolzmann@molgen.mpg.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Toshi Kani <toshi.kani@hp.com>
Reviewed-by: Hanjun Guo <hanjun.guo@linaro.org>
Reviewed-by: Lee, Chun-Yi <jlee@suse.com>
Currently it is not made explicit why clk_fixed_set_rate() can ignore
its arguments and unconditionally return success. Add a comment
to explain this.
We also mark the clk_ops table const since it should never be
modified at runtime.
Suggested-by: Stephen Boyd <sboyd@codeaurora.org>
Signed-off-by: Daniel Thompson <daniel.thompson@linaro.org>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
