Or Gerlitz says:
====================
mlx5e-pedit 2017-03-28
This series adds support for offloading modifications of packet headers using
ConnectX-5 HW header re-write as an action applied during packet steering.
The offloaded SW mechanism is TC's pedit action. The offloading is
supported for E-Switch steering of VF traffic in the SRIOV
switchdev mode and for NIC (non eswitch) RX.
One use-case for this offload on virtual networks, is when the hypervisor
implements flow based router such as Open-Stack's DVR, where L2 headers
of guest packets re-written with routers' MAC addresses and the IP TTL
is decremented.
Another use case (which can be applied in parallel with routing) is
stateless NAT where guest L3/L4 headers are re-written.
The series is built as follows: the 1st six patches are preperations which
don't yet add new functionality, patches 7-8 add the FW APIs (data-structures
and commands) for header re-write, and patch nine allows offloading driver
to access pedit keys.
The 10th patch is somehow the core of the series, where we translate from
the pedit way to represent set of header modification elements to the FW
API for that same matter.
Once a set of HW modification is established, we register it with the FW
and get a modify header ID. When this ID is used with an action during
packet steering, the HW applies the header modification on the packet.
Patches 11 and 12 implement the above logic as an offload for pedit action
for the NIC and E-Switch use-cases.
I'd like to thanks Elijah Shakkour <elijahs@mellanox.com> for implementing
and helping me testing this functionality on HW simulator, before it could
be done with FW.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
I inadvertently applied the v5 version of this patch, whereas
the agreed upon version was v5. Revert this one so we can apply
the right one.
This reverts commit 7fc6b87a9f.
Now that the remaining drivers have been converted to one request queue
per gendisk, let's warn if a request queue gets registered more than
once. This will catch future drivers which might do it inadvertently or
any old drivers that I may have missed.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
As the .q_usage_counter is used by both legacy and
mq path, we need to block new I/O if queue becomes
dead in blk_queue_enter().
So rename it and we can use this function in both
paths.
Reviewed-by: Bart Van Assche <bart.vanassche@sandisk.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Jens Axboe <axboe@fb.com>
Commit 26988efe11 ("regulator: core: Allow to get voltage count and
list from parent") introduces the propagation of the parent voltage
count and list for regulators that don't provide this information
themselves. The goal is to support simple switch regulators, however as
a side effect normal continuous regulators can leak details of their
supplies and provide consumers with inconsistent information.
Limit the propagation of the voltage count and list to switch
regulators.
Fixes: 26988efe11 ("regulator: core: Allow to get voltage count and
list from parent")
Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
Reviewed-by: Javier Martinez Canillas <javier@osg.samsung.com>
Tested-by: Javier Martinez Canillas <javier@osg.samsung.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Invoke the proper function while initializing
a dynamically allocated spi_message to avoid
NULL pointer dereference during resources deallocation.
Signed-off-by: Emiliano Ingrassia <ingrassia@epigenesys.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The gpio-matrix-keypad driver normally sets inactive columns as inputs
while scanning. This does not work for all hardware, which may require
the inactive columns to be actively driven in order to overcome any
pull-ups/downs on the columns.
Signed-off-by: David Rivshin <drivshin@allworx.com>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
There is an include loop between netdevice.h, dsa.h, devlink.h because
of NETDEV_ALIGN, making it impossible to use devlink structures in
dsa.h.
Break this loop by taking dsa.h out of netdevice.h, add a forward
declaration of dsa_switch_tree and netdev_set_default_ethtool_ops()
function, which is what netdevice.h requires.
No longer having dsa.h in netdevice.h means the includes in dsa.h no
longer get included. This breaks a few other files which depend on
these includes. Add these directly in the affected file.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Refactor inet_netconf_notify_devconf to take the event as an input arg.
Signed-off-by: David Ahern <dsa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Remove the standalone SMD implementation as we have transitioned the
client drivers to use the RPMSG based one.
Also remove all dependencies on QCOM_SMD from Kconfig files, in order to
keep them selectable in the absence of the removed symbol.
Acked-by: Andy Gross <andy.gross@linaro.org>
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
By moving these client drivers to use RPMSG instead of the direct SMD
API we can reuse them ontop of the newly added GLINK wire-protocol
support found in the 820 and 835 Qualcomm platforms.
As the new (RPMSG-based) and old SMD implementations are mutually
exclusive we have to change all client drivers in one commit, to make
sure we have a working system before and after this transition.
Acked-by: Andy Gross <andy.gross@linaro.org>
Acked-by: Kalle Valo <kvalo@codeaurora.org>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
The current way to implement an always on PM domain consists of returning
-EBUSY from the ->power_off() callback. This is a bit different compared to
using the always on genpd governor, which prevents the PM domain from being
powered off via runtime suspend, but not via system suspend.
The approach to return -EBUSY from the ->power_off() callback to support
always on PM domains in genpd is suboptimal. That is because it requires
genpd to follow the regular execution path of the power off sequence, which
ends by invoking the ->power_off() callback.
To enable genpd to early abort the power off sequence for always on PM
domains, it needs static information about these configurations. Therefore
let's add a new genpd configuration flag, GENPD_FLAG_ALWAYS_ON.
Users of the new GENPD_FLAG_ALWAYS_ON flag, are by genpd required to make
sure the PM domain is powered on before calling pm_genpd_init(). Moreover,
users don't need to implement the ->power_off() callback, as genpd doesn't
ever invoke it.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Similar to OF endpoints, endpoint type nodes can be also supported on
ACPI. In order to make it possible for drivers to ignore the matter,
add a type for fwnode_endpoint and a function to parse them.
On ACPI, find the child node index instead of relying on the "endpoint"
property.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The function to obtain a fwnode related to a struct device is useful for
drivers that use the fwnode property API: it allows not being aware of the
underlying firmware implementation.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
of_fwnode_handle() returns a struct fwnode_handle of the struct
device_node. This may be used on the fwnode property API.
Use a macro instead of a function in order to support const and non-const
arguments.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
fwnode_handle_get() is used to obtain a reference to a fwnode_handle
container. In this case this is OF specific struct device_node.
This complements fwnode_handle_put() which is already implemented.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This follows DT implementation of of_graph_* APIs but we call them
fwnode_graph_* instead. For DT nodes the existing of_graph_* implementation
will be used. For ACPI we use the new ACPI graph implementation instead.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
DT has had concept of remote endpoints for some time already. It makes
possible to reference another firmware node through a property called
remote-endpoint. This is already used by some subsystems like v4l2 for
parsing hardware properties related to camera.
This patch adds ACPI support for remote endpoints utilizing _DSD
hierarchical data extensions.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Since now we have means to enumerate all children of any fwnode even in
ACPI we can implement fwnode_get_named_child_node(). This is similar than
device_get_named_child_node() with the exception that it can be called to
any fwnode handle. Make device_get_named_child_node() call directly this
new function.
This is useful in cases where we need to be able to find child nodes which
are not direct descendants of the parent device.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The ACPI _DSD hierarchical data extension makes it possible to have
hierarchies deeper than one level in similar way than DT allows. These
"subsubnodes" have not been accessible because device property
implementation only provides device_get_next_child_node() that is limited
to direct descendants of a device.
We need this ability in order support things like remote endpoints
currently supported in DT with of_graph_* APIs.
Modify acpi_get_next_subnode() to accept fwnode handle instead and update
callers accordingly. Also add a new function fwnode_get_next_child_node()
that works directly with fwnodes and modify device_get_next_child_node() to
call it directly. While there add a macro fwnode_for_each_child_node()
analogous to the current device_for_each_child_node() but it works with
fwnodes instead of devices.
Link: http://www.uefi.org/sites/default/files/resources/_DSD-hierarchical-data-extension-UUID-v1.pdf
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Now that ACPI has support for returning parent firmware node for both types
of nodes we can expose this to others as well. This adds a new function
fwnode_get_parent() that can be used for DT and ACPI nodes to retrieve the
parent firmware node.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Sometimes it is useful to be able to navigate firmware node hierarchy
upwards toward parent nodes. ACPI device nodes are pretty much already
supported because ACPICA provides acpi_get_parent(). ACPI data nodes,
however, are all below the same parent ACPI device. Their hierarchy is
created by "linking" each other using references in the value field.
Add parent pointer to the parent data node while we create them so it is
easy to navigate the hierarchy backwards. We use this parent pointer in a
new function acpi_node_get_parent() that is able to extract parent of both
ACPI firmware node types.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
blkg_conf_prep() currently calls blkg_lookup_create() while holding
request queue spinlock. This means allocating memory for struct
blkcg_gq has to be made non-blocking. This causes occasional -ENOMEM
failures in call paths like below:
pcpu_alloc+0x68f/0x710
__alloc_percpu_gfp+0xd/0x10
__percpu_counter_init+0x55/0xc0
cfq_pd_alloc+0x3b2/0x4e0
blkg_alloc+0x187/0x230
blkg_create+0x489/0x670
blkg_lookup_create+0x9a/0x230
blkg_conf_prep+0x1fb/0x240
__cfqg_set_weight_device.isra.105+0x5c/0x180
cfq_set_weight_on_dfl+0x69/0xc0
cgroup_file_write+0x39/0x1c0
kernfs_fop_write+0x13f/0x1d0
__vfs_write+0x23/0x120
vfs_write+0xc2/0x1f0
SyS_write+0x44/0xb0
entry_SYSCALL_64_fastpath+0x18/0xad
In the code path above, percpu allocator cannot call vmalloc() due to
queue spinlock.
A failure in this call path gives grief to tools which are trying to
configure io weights. We see occasional failures happen shortly after
reboots even when system is not under any memory pressure. Machines
with a lot of cpus are more vulnerable to this condition.
Update blkg_create() function to temporarily drop the rcu and queue
locks when it is allowed by gfp mask.
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Tahsin Erdogan <tahsin@google.com>
Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@fb.com>
Some of the Broadcom iProc SoCs have FlexRM ring manager
which provides a ring-based programming interface to various
offload engines (e.g. RAID, Crypto, etc).
This patch adds a common mailbox driver for Broadcom FlexRM
ring manager which can be shared by various offload engine
drivers (implemented as mailbox clients).
Reviewed-by: Ray Jui <ray.jui@broadcom.com>
Reviewed-by: Scott Branden <scott.branden@broadcom.com>
Reviewed-by: Pramod KUMAR <pramod.kumar@broadcom.com>
Signed-off-by: Anup Patel <anup.patel@broadcom.com>
Signed-off-by: Jassi Brar <jaswinder.singh@linaro.org>
This driver is now used only on platforms which support device tree, so
it is safe to remove legacy platform data based initialization code.
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Acked-by: Arnd Bergmann <arnd@arndb.de>
For plat-samsung:
Acked-by: Krzysztof Kozlowski <krzk@kernel.org>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
User configures latency target, but the latency threshold for each
request size isn't fixed. For a SSD, the IO latency highly depends on
request size. To calculate latency threshold, we sample some data, eg,
average latency for request size 4k, 8k, 16k, 32k .. 1M. The latency
threshold of each request size will be the sample latency (I'll call it
base latency) plus latency target. For example, the base latency for
request size 4k is 80us and user configures latency target 60us. The 4k
latency threshold will be 80 + 60 = 140us.
To sample data, we calculate the order base 2 of rounded up IO sectors.
If the IO size is bigger than 1M, it will be accounted as 1M. Since the
calculation does round up, the base latency will be slightly smaller
than actual value. Also if there isn't any IO dispatched for a specific
IO size, we will use the base latency of smaller IO size for this IO
size.
But we shouldn't sample data at any time. The base latency is supposed
to be latency where disk isn't congested, because we use latency
threshold to schedule IOs between cgroups. If disk is congested, the
latency is higher, using it for scheduling is meaningless. Hence we only
do the sampling when block throttling is in the LOW limit, with
assumption disk isn't congested in such state. If the assumption isn't
true, eg, low limit is too high, calculated latency threshold will be
higher.
Hard disk is completely different. Latency depends on spindle seek
instead of request size. Currently this feature is SSD only, we probably
can use a fixed threshold like 4ms for hard disk though.
Signed-off-by: Shaohua Li <shli@fb.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Currently there is no way to know the request size when the request is
finished. Next patch will need this info. We could add extra field to
record the size, but blk_issue_stat has enough space to record it, so
this patch just overloads blk_issue_stat. With this, we will have 49bits
to track time, which still is very long time.
Signed-off-by: Shaohua Li <shli@fb.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
A cgroup gets assigned a low limit, but the cgroup could never dispatch
enough IO to cross the low limit. In such case, the queue state machine
will remain in LIMIT_LOW state and all other cgroups will be throttled
according to low limit. This is unfair for other cgroups. We should
treat the cgroup idle and upgrade the state machine to lower state.
We also have a downgrade logic. If the state machine upgrades because of
cgroup idle (real idle), the state machine will downgrade soon as the
cgroup is below its low limit. This isn't what we want. A more
complicated case is cgroup isn't idle when queue is in LIMIT_LOW. But
when queue gets upgraded to lower state, other cgroups could dispatch
more IO and this cgroup can't dispatch enough IO, so the cgroup is below
its low limit and looks like idle (fake idle). In this case, the queue
should downgrade soon. The key to determine if we should do downgrade is
to detect if cgroup is truely idle.
Unfortunately it's very hard to determine if a cgroup is real idle. This
patch uses the 'think time check' idea from CFQ for the purpose. Please
note, the idea doesn't work for all workloads. For example, a workload
with io depth 8 has disk utilization 100%, hence think time is 0, eg,
not idle. But the workload can run higher bandwidth with io depth 16.
Compared to io depth 16, the io depth 8 workload is idle. We use the
idea to roughly determine if a cgroup is idle.
We treat a cgroup idle if its think time is above a threshold (by
default 1ms for SSD and 100ms for HD). The idea is think time above the
threshold will start to harm performance. HD is much slower so a longer
think time is ok.
The patch (and the latter patches) uses 'unsigned long' to track time.
We convert 'ns' to 'us' with 'ns >> 10'. This is fast but loses
precision, should not a big deal.
Signed-off-by: Shaohua Li <shli@fb.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Add the definitions related to creation/deletion of a modify header
context and the modify header steering action which are used for HW
packet header modify (re-write) as part of steering. Add as well the
modify header id into two intermediate structs and set it to the FTE.
Note that as the push/pop vlan steering actions are emulated by the
ewitch management code, we're not breaking any compatibility while
changing their values to make room for the modify header action which
is not emulated and whose value is part of the FW API. The new bit
values for the emulated actions are at the end of the possible range.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Hadar Hen Zion <hadarh@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
There are bunch of places in the code where the intermediate struct
that keeps the elements related to flow actions is initialized with
the same default values. Put that into a small DECLARE type helper.
This patch doesn't change any functionality.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Hadar Hen Zion <hadarh@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Add a calc_num_ports callback to the generic driver and verify that the
device has the required endpoints there instead of in core.
Note that the generic driver num_ports field was never used.
Signed-off-by: Johan Hovold <johan@kernel.org>
Allow subdrivers to modify the port-endpoint mapping by passing the
endpoint descriptors to calc_num_ports.
The callback can now also be used to verify that the required endpoints
exists and abort probing otherwise.
This will allow us to get rid of a few hacks in subdrivers that are
already modifying the port-endpoint mapping (or aborting probe due to
missing endpoints), but only after the port structures have been setup.
Signed-off-by: Johan Hovold <johan@kernel.org>
This fixes compilation for files, that try to include the
cpcap header in alphabetically sorted #include lists.
Acked-by: Pavel Machek <pavel@ucw.cz>
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Sebastian Reichel <sre@kernel.org>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
The patch fix syntax errors introduced by commit 0c8893c9095d
("clockevents: Add a clkevt-of mechanism like clksrc-of").
Fixes: 0c8893c9095d ("clockevents: Add a clkevt-of mechanism like clksrc-of")
Signed-off-by: Alexander Kochetkov <al.kochet@gmail.com>
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Introduce a simple data structure for collecting correctable errors
along with accessors. More detailed description in the code itself.
The error decoding is done with the decoding chain now and
mce_first_notifier() gets to see the error first and the CEC decides
whether to log it and then the rest of the chain doesn't hear about it -
basically the main reason for the CE collector - or to continue running
the notifiers.
When the CEC hits the action threshold, it will try to soft-offine the
page containing the ECC and then the whole decoding chain gets to see
the error.
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-edac <linux-edac@vger.kernel.org>
Link: http://lkml.kernel.org/r/20170327093304.10683-5-bp@alien8.de
Signed-off-by: Ingo Molnar <mingo@kernel.org>
We switched from "struct task_struct"->security to "struct cred"->security
in Linux 2.6.29. But not all LSM modules were happy with that change.
TOMOYO LSM module is an example which want to use per "struct task_struct"
security blob, for TOMOYO's security context is defined based on "struct
task_struct" rather than "struct cred". AppArmor LSM module is another
example which want to use it, for AppArmor is currently abusing the cred
a little bit to store the change_hat and setexeccon info. Although
security_task_free() hook was revived in Linux 3.4 because Yama LSM module
wanted to release per "struct task_struct" security blob,
security_task_alloc() hook and "struct task_struct"->security field were
not revived. Nowadays, we are getting proposals of lightweight LSM modules
which want to use per "struct task_struct" security blob.
We are already allowing multiple concurrent LSM modules (up to one fully
armored module which uses "struct cred"->security field or exclusive hooks
like security_xfrm_state_pol_flow_match(), plus unlimited number of
lightweight modules which do not use "struct cred"->security nor exclusive
hooks) as long as they are built into the kernel. But this patch does not
implement variable length "struct task_struct"->security field which will
become needed when multiple LSM modules want to use "struct task_struct"->
security field. Although it won't be difficult to implement variable length
"struct task_struct"->security field, let's think about it after we merged
this patch.
Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Acked-by: John Johansen <john.johansen@canonical.com>
Acked-by: Serge Hallyn <serge@hallyn.com>
Acked-by: Casey Schaufler <casey@schaufler-ca.com>
Tested-by: Djalal Harouni <tixxdz@gmail.com>
Acked-by: José Bollo <jobol@nonadev.net>
Cc: Paul Moore <paul@paul-moore.com>
Cc: Stephen Smalley <sds@tycho.nsa.gov>
Cc: Eric Paris <eparis@parisplace.org>
Cc: Kees Cook <keescook@chromium.org>
Cc: James Morris <james.l.morris@oracle.com>
Cc: José Bollo <jobol@nonadev.net>
Signed-off-by: James Morris <james.l.morris@oracle.com>
This resolves a merge issue in the gadget code, and we want the USB
fixes in here as well.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
