Horatiu Vultur says:
====================
net: bridge: mrp: Add support for Media Redundancy Protocol(MRP)
Media Redundancy Protocol is a data network protocol standardized by
International Electrotechnical Commission as IEC 62439-2. It allows rings of
Ethernet switches to overcome any single failure with recovery time faster than
STP. It is primarily used in Industrial Ethernet applications.
Based on the previous RFC[1][2][3][4][5], and patches[6][7][8], the MRP state
machine and all the timers were moved to userspace, except for the timers used
to generate MRP Test frames. In this way the userspace doesn't know and should
not know if the HW or the kernel will generate the MRP Test frames. The
following changes were added to the bridge to support the MRP:
- the existing netlink interface was extended with MRP support,
- allow to detect when a MRP frame was received on a MRP ring port
- allow MRP instance to forward/terminate MRP frames
- generate MRP Test frames in case the HW doesn't have support for this
To be able to offload MRP support to HW, the switchdev API was extend.
With these changes the userspace doesn't do the following because already the
kernel/HW will do:
- doesn't need to forward/terminate MRP frames
- doesn't need to generate MRP Test frames
- doesn't need to detect when the ring is open/closed.
The userspace application that is using the new netlink can be found here[9].
The current implementation both in kernel and userspace supports only 2 roles:
MRM - this one is responsible to send MRP_Test and MRP_Topo frames on both
ring ports. It needs to process MRP_Test to know if the ring is open or
closed. This operation is desired to be offloaded to the HW because it
requires to generate and process up to 4000 frames per second. Whenever it
detects that the ring is open it sends MRP_Topo frames to notify all MRC about
changes in the topology. MRM needs also to process MRP_LinkChange frames,
these frames are generated by the MRC. When the ring is open then the state
of both ports is to forward frames and when the ring is closed then the
secondary port is blocked.
MRC - this one is responsible to forward MRP frames between the ring ports.
In case one of the ring ports gets a link down or up, then MRC will generate
a MRP_LinkChange frames. This node should also process MRP_Topo frames and to
clear its FDB when it receives this frame.
Userspace
Deamon +----------+ Client
+
|
+--------------|-----------------------------------------+
Kernel |
+ Netlink
| + Interrupt
| |
+--------------|------------------------------|----------+
HW | Switchdev |
+ |
The user interacts using the client (called 'mrp'), the client talks to the
deamon (called 'mrp_server'), which talks with the kernel using netlink. The
kernel will try to offload the requests to the HW via switchdev API.
If this will be accepted then in the future the netlink interface can be
expended with multiple attributes which are required by different roles of the
MRP. Like Media Redundancy Automanager(MRA), Media Interconnect Manager(MIM) and
Media Interconnect Client(MIC).
[1] https://www.spinics.net/lists/netdev/msg623647.html
[2] https://www.spinics.net/lists/netdev/msg624378.html
[3] https://www.spinics.net/lists/netdev/msg627500.html
[4] https://www.spinics.net/lists/netdev/msg641005.html
[5] https://www.spinics.net/lists/netdev/msg643991.html
[6] https://www.spinics.net/lists/netdev/msg645378.html
[7] https://www.spinics.net/lists/kernel/msg3484685.html
[8] https://www.spinics.net/lists/netdev/msg646202.html
[9] https://github.com/microchip-ung/mrp/tree/patch-v9
-v4:
- fix comments in br_mrp.c
- use skb_header_pointer to get br_mrp_tlv_hdr
- fix line over 80 characters
-v3:
- fix unused variables
-v2:
- drop patch 4
- add port flag BR_MRP_LOST_CONT;
- another fix for bisectability
-v1:
- fix bisectability issues
- in case of errors use extack
-RFC v5:
- use nla_parse_nested
- rework the usage of the rcu in br_mrp
- reorder patches
- few other small issues raised by Nikolay
-RFC v4:
- extend existing netlink interface to add mrp support
- use rcu locks
-RFC v3:
- move MRP state machine in userspace
- create generic netlink interface for configuring the HW using switchdev API
-RFC v2:
- extend switchdev API to offload to HW
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
It is not possible to have the MRP and STP running at the same time on the
bridge, therefore add check when enabling the STP to check if MRP is already
enabled. In that case return error.
Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
To integrate MRP into the bridge, the bridge needs to do the following:
- detect if the MRP frame was received on MRP ring port in that case it would be
processed otherwise just forward it as usual.
- enable parsing of MRP
- before whenever the bridge was set up, it would set all the ports in
forwarding state. Add an extra check to not set ports in forwarding state if
the port is an MRP ring port. The reason of this change is that if the MRP
instance initially sets the port in blocked state by setting the bridge up it
would overwrite this setting.
Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Implement netlink interface to configure MRP. The implementation
will do sanity checks over the attributes and then eventually call the MRP
interface.
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Implement the MRP API.
In case the HW can't generate MRP Test frames then the SW will try to generate
the frames. In case that also the SW will fail in generating the frames then a
error is return to the userspace. The userspace is responsible to generate all
the other MRP frames regardless if the test frames are generated by HW or SW.
The forwarding/termination of MRP frames is happening in the kernel and is done
by the MRP instance. The userspace application doesn't do the forwarding.
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Implement the MRP api for switchdev.
These functions will just eventually call the switchdev functions:
switchdev_port_obj_add/del and switchdev_port_attr_set.
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Extend switchdev API to add support for MRP. The HW is notified in
following cases:
SWITCHDEV_OBJ_ID_MRP: This is used when a MRP instance is added/removed
from the MRP ring.
SWITCHDEV_OBJ_ID_RING_ROLE_MRP: This is used when the role of the node
changes. The current supported roles are MRM and MRC.
SWITCHDEV_OBJ_ID_RING_TEST_MRP: This is used when to start/stop sending
MRP_Test frames on the mrp ring ports. This is called only on nodes that have
the role MRM. In case this fails then the SW will generate the frames.
SWITCHDEV_OBJ_ID_RING_STATE_STATE: This is used when the ring changes it states
to open or closed. This is required to notify HW because the MRP_Test frame
contains the field MRP_InState which contains this information.
SWITCHDEV_ATTR_ID_MRP_PORT_STATE: This is used when the port's state is
changed. It can be in blocking/forwarding mode.
SWITCHDEV_ATTR_ID_MRP_PORT_ROLE: This is used when port's role changes. The
roles of the port can be primary/secondary. This is required to notify HW
because the MRP_Test frame contains the field MRP_PortRole that contains this
information.
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Define the MRP interface.
This interface is used by the netlink to update the MRP instances and by the MRP
to make the calls to switchdev to offload it to HW.
It defines an MRP instance 'struct br_mrp' which is a list of MRP instances.
Which will be part of the 'struct net_bridge'. Each instance has 2 ring ports,
a bridge and an ID.
In case the HW can't generate MRP Test frames then the SW will generate those.
br_mrp_add - adds a new MRP instance.
br_mrp_del - deletes an existing MRP instance. Each instance has an ID(ring_id).
br_mrp_set_port_state - changes the port state. The port can be in forwarding
state, which means that the frames can pass through or in blocked state which
means that the frames can't pass through except MRP frames. This will
eventually call the switchdev API to notify the HW. This information is used
also by the SW bridge to know how to forward frames in case the HW doesn't
have this capability.
br_mrp_set_port_role - a port role can be primary or secondary. This
information is required to be pushed to HW in case the HW can generate
MRP_Test frames. Because the MRP_Test frames contains a file with this
information. Otherwise the HW will not be able to generate the frames
correctly.
br_mrp_set_ring_state - a ring can be in state open or closed. State open means
that the mrp port stopped receiving MRP_Test frames, while closed means that
the mrp port received MRP_Test frames. Similar with br_mrp_port_role, this
information is pushed in HW because the MRP_Test frames contain this
information.
br_mrp_set_ring_role - a ring can have the following roles MRM or MRC. For the
role MRM it is expected that the HW can terminate the MRP frames, notify the
SW that it stopped receiving MRP_Test frames and trapp all the other MRP
frames. While for MRC mode it is expected that the HW can forward the MRP
frames only between the MRP ports and copy MRP_Topology frames to CPU. In
case the HW doesn't support a role it needs to return an error code different
than -EOPNOTSUPP.
br_mrp_start_test - this starts/stops the generation of MRP_Test frames. To stop
the generation of frames the interval needs to have a value of 0. In this case
the userspace needs to know if the HW supports this or not. Not to have
duplicate frames(generated by HW and SW). Because if the HW supports this then
the SW will not generate anymore frames and will expect that the HW will
notify when it stopped receiving MRP frames using the function
br_mrp_port_open.
br_mrp_port_open - this function is used by drivers to notify the userspace via
a netlink callback that one of the ports stopped receiving MRP_Test frames.
This function is called only when the node has the role MRM. It is not
supposed to be called from userspace.
br_mrp_port_switchdev_add - this corresponds to the function br_mrp_add,
and will notify the HW that a MRP instance is added. The function gets
as parameter the MRP instance.
br_mrp_port_switchdev_del - this corresponds to the function br_mrp_del,
and will notify the HW that a MRP instance is removed. The function
gets as parameter the ID of the MRP instance that is removed.
br_mrp_port_switchdev_set_state - this corresponds to the function
br_mrp_set_port_state. It would notify the HW if it should block or not
non-MRP frames.
br_mrp_port_switchdev_set_port - this corresponds to the function
br_mrp_set_port_role. It would set the port role, primary or secondary.
br_mrp_switchdev_set_role - this corresponds to the function
br_mrp_set_ring_role and would set one of the role MRM or MRC.
br_mrp_switchdev_set_ring_state - this corresponds to the function
br_mrp_set_ring_state and would set the ring to be open or closed.
br_mrp_switchdev_send_ring_test - this corresponds to the function
br_mrp_start_test. This will notify the HW to start or stop generating
MRP_Test frames. Value 0 for the interval parameter means to stop generating
the frames.
br_mrp_port_open - this function is used to notify the userspace that the port
lost the continuity of MRP Test frames.
Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds a new port attribute, IFLA_BRPORT_MRP_RING_OPEN, which allows
to notify the userspace when the port lost the continuite of MRP frames.
This attribute is set by kernel whenever the SW or HW detects that the ring is
being open or closed.
Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
To integrate MRP into the bridge, first the bridge needs to be aware of ports
that are part of an MRP ring and which rings are on the bridge.
Therefore extend bridge interface with the following:
- add new flag(BR_MPP_AWARE) to the net bridge ports, this bit will be
set when the port is added to an MRP instance. In this way it knows if
the frame was received on MRP ring port
- add new flag(BR_MRP_LOST_CONT) to the net bridge ports, this bit will be set
when the port lost the continuity of MRP Test frames.
- add a list of MRP instances
Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add the option BRIDGE_MRP to allow to build in or not MRP support.
The default value is N.
Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add new nested netlink attribute to configure the MRP. These attributes are used
by the userspace to add/delete/configure MRP instances and by the kernel to
notify the userspace when the MRP ring gets open/closed. MRP nested attribute
has the following attributes:
IFLA_BRIDGE_MRP_INSTANCE - the parameter type is br_mrp_instance which contains
the instance id, and the ifindex of the two ports. The ports can't be part of
multiple instances. This is used to create/delete MRP instances.
IFLA_BRIDGE_MRP_PORT_STATE - the parameter type is u32. Which can be forwarding,
blocking or disabled.
IFLA_BRIDGE_MRP_PORT_ROLE - the parameter type is br_mrp_port_role which
contains the instance id and the role. The role can be primary or secondary.
IFLA_BRIDGE_MRP_RING_STATE - the parameter type is br_mrp_ring_state which
contains the instance id and the state. The state can be open or closed.
IFLA_BRIDGE_MRP_RING_ROLE - the parameter type is br_mrp_ring_role which
contains the instance id and the ring role. The role can be MRM or MRC.
IFLA_BRIDGE_MRP_START_TEST - the parameter type is br_mrp_start_test which
contains the instance id, the interval at which to send the MRP_Test frames,
how many test frames can be missed before declaring the ring open and the
period which represent for how long to send the test frames.
Also add the file include/uapi/linux/mrp_bridge.h which defines all the types
used by MRP that are also needed by the userpace.
Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Read the temperature sensor register from the correct location for the
88E2110 PHY. There is no enable/disable bit on 2110, so make
mv3310_hwmon_config() run on 88X3310 only.
Fixes: 62d0153547 ("net: phy: marvell10g: add support for the 88x2110 PHY")
Cc: Maxime Chevallier <maxime.chevallier@bootlin.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Reviewed-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
The IRQ postinstall handling had open-coded pipe fault mask selection
that never got updated for gen11. Switch it to use
gen8_de_pipe_fault_mask() to ensure we don't miss updates for new
platforms.
Cc: José Roberto de Souza <jose.souza@intel.com>
Fixes: d506a65d56 ("drm/i915: Catch GTT fault errors for gen11+ planes")
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200424231423.4065231-1-matthew.d.roper@intel.com
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
My intent was to not let users set a zero drop_batch_size,
it seems I once again messed with min()/max().
Fixes: 9d18562a22 ("fq_codel: add batch ability to fq_codel_drop()")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
tls_data_ready() invokes sk_psock_get(), which returns a reference of
the specified sk_psock object to "psock" with increased refcnt.
When tls_data_ready() returns, local variable "psock" becomes invalid,
so the refcount should be decreased to keep refcount balanced.
The reference counting issue happens in one exception handling path of
tls_data_ready(). When "psock->ingress_msg" is empty but "psock" is not
NULL, the function forgets to decrease the refcnt increased by
sk_psock_get(), causing a refcnt leak.
Fix this issue by calling sk_psock_put() on all paths when "psock" is
not NULL.
Signed-off-by: Xiyu Yang <xiyuyang19@fudan.edu.cn>
Signed-off-by: Xin Tan <tanxin.ctf@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
x25_connect() invokes x25_get_neigh(), which returns a reference of the
specified x25_neigh object to "x25->neighbour" with increased refcnt.
When x25 connect success and returns, the reference still be hold by
"x25->neighbour", so the refcount should be decreased in
x25_disconnect() to keep refcount balanced.
The reference counting issue happens in x25_disconnect(), which forgets
to decrease the refcnt increased by x25_get_neigh() in x25_connect(),
causing a refcnt leak.
Fix this issue by calling x25_neigh_put() before x25_disconnect()
returns.
Signed-off-by: Xiyu Yang <xiyuyang19@fudan.edu.cn>
Signed-off-by: Xin Tan <tanxin.ctf@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
bpf_exec_tx_verdict() invokes sk_psock_get(), which returns a reference
of the specified sk_psock object to "psock" with increased refcnt.
When bpf_exec_tx_verdict() returns, local variable "psock" becomes
invalid, so the refcount should be decreased to keep refcount balanced.
The reference counting issue happens in one exception handling path of
bpf_exec_tx_verdict(). When "policy" equals to NULL but "psock" is not
NULL, the function forgets to decrease the refcnt increased by
sk_psock_get(), causing a refcnt leak.
Fix this issue by calling sk_psock_put() on this error path before
bpf_exec_tx_verdict() returns.
Signed-off-by: Xiyu Yang <xiyuyang19@fudan.edu.cn>
Signed-off-by: Xin Tan <tanxin.ctf@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The variable err is being initializeed with a value that is never read
and it is being updated later with a new value. The initialization
is redundant and can be removed.
Addresses-Coverity: ("Unused value")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The Aquantia AQC100 controller enables a SFP+ port, so the driver should
configure the media type as '_TYPE_FIBRE' instead of '_TYPE_TP'.
Signed-off-by: Richard Clark <richard.xnu.clark@gmail.com>
Cc: Igor Russkikh <irusskikh@marvell.com>
Cc: "David S. Miller" <davem@davemloft.net>
Acked-by: Igor Russkikh <irusskikh@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
A spin lock is held in insert_report_filterlist(), so the krealloc()
should use GFP_ATOMIC. This commit therefore makes this change.
Reviewed-by: Marco Elver <elver@google.com>
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit simplifies and clarifies the highest level KCSAN Kconfig
help text.
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
When using -Wextra, gcc complains about torture_preempt_schedule()
when its definition is empty (i.e., when CONFIG_PREEMPTION is not
set/enabled). Fix these warnings by adding an empty do-while block
for that macro when CONFIG_PREEMPTION is not set.
Fixes these build warnings:
../kernel/locking/locktorture.c:119:29: warning: suggest braces around empty body in an ‘if’ statement [-Wempty-body]
../kernel/locking/locktorture.c:166:29: warning: suggest braces around empty body in an ‘if’ statement [-Wempty-body]
../kernel/locking/locktorture.c:337:29: warning: suggest braces around empty body in an ‘if’ statement [-Wempty-body]
../kernel/locking/locktorture.c:490:29: warning: suggest braces around empty body in an ‘if’ statement [-Wempty-body]
../kernel/locking/locktorture.c:528:29: warning: suggest braces around empty body in an ‘if’ statement [-Wempty-body]
../kernel/locking/locktorture.c:553:29: warning: suggest braces around empty body in an ‘if’ statement [-Wempty-body]
I have verified that there is no object code change (with gcc 7.5.0).
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Josh Triplett <josh@joshtriplett.org>
Signed-off-by: "Paul E. McKenney" <paulmck@kernel.org>
The n_barrier_successes, n_barrier_attempts, and
n_rcu_torture_barrier_error variables are updated (without access
markings) by the main rcu_barrier() test kthread, and accessed (also
without access markings) by the rcu_torture_stats() kthread. This of
course can result in KCSAN complaints.
Because the accesses are in diagnostic prints, this commit uses
data_race() to excuse the diagnostic prints from the data race. If this
were to ever cause bogus statistics prints (for example, due to store
tearing), any misleading information would be disambiguated by the
presence or absence of an rcutorture splat.
This data race was reported by KCSAN. Not appropriate for backporting
due to failure being unlikely and due to the mild consequences of the
failure, namely a confusing rcutorture console message.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Joel Fernandes (Google) <joel@joelfernandes.org>
System hangs or killed rcutorture guest OSes can result in truncated
"Reader Pipe:" lines, which can in turn result in false-positive
reader-batch near-miss warnings. This commit therefore adjusts the
reader-batch checks to account for possible line truncation.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit adds stubs for KCSAN's data_race(), ASSERT_EXCLUSIVE_WRITER(),
and ASSERT_EXCLUSIVE_ACCESS() macros to allow code using these macros to
move ahead.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
When all quiescent states have been seen, it is normally the grace-period
kthread that is in trouble. Although the existing stack trace from
the current CPU might possibly provide useful information, experience
indicates that there is too much noise for this to be worthwhile.
This commit therefore removes this stack trace from the output.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
If the grace-period kthread is starved, idle threads' extended quiescent
states are not reported. These idle threads thus wrongly appear to
be blocking the current grace period. This commit therefore tags such
idle threads as probable false positives when the grace-period kthread
is being starved.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Although the accesses used to determine whether or not an expedited
stall should be printed are an integral part of the concurrency algorithm
governing use of the corresponding variables, the values that are simply
printed are ancillary. As such, it is best to use data_race() for these
accesses in order to provide the greatest latitude in the use of KCSAN
for the other accesses that are an integral part of the algorithm. This
commit therefore changes the relevant uses of READ_ONCE() to data_race().
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit replaces the schedule_on_each_cpu(ftrace_sync) instances
with synchronize_rcu_tasks_rude().
Suggested-by: Steven Rostedt <rostedt@goodmis.org>
Cc: Ingo Molnar <mingo@redhat.com>
[ paulmck: Make Kconfig adjustments noted by kbuild test robot. ]
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit allows TASKS_TRACE_RCU to be used independently of TASKS_RCU
and vice versa.
[ paulmck: Fix conditional compilation per kbuild test robot feedback. ]
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit adds a failure-return count for smp_call_function_single(),
and adds this to the console messages for rcutorture writer stalls and at
the end of rcutorture testing.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit adds a TRACE02 scenario which enables preemption and RCU
Tasks Trace IPIs, more specifically, disabling heavyweight readers.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit adds a counter for the number of times the quiescent state
was an idle task associated with an offline CPU, and prints this count
at the end of rcutorture runs and at stall time.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit adds counts of the number of calls and number of successful
calls to rcu_dynticks_zero_in_eqs(), which are printed at the end
of rcutorture runs and at stall time. This allows evaluation of the
effectiveness of rcu_dynticks_zero_in_eqs().
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit scans the CPUs, adding each CPU's idle task to the list of
tasks that need quiescent states.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The idle task corresponding to an offline CPU can appear to be running
while that CPU is offline. This commit therefore adds checks for this
situation, treating it as a quiescent state. Because the tasklist scan
and the holdout-list scan now exclude CPU-hotplug operations, readers
on the CPU-hotplug paths are still waited for.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit disables CPU hotplug across RCU tasks trace scans, which
is a first step towards correctly recognizing idle tasks "running" on
offline CPUs.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The rcu_read_unlock_trace() can invoke rcu_read_unlock_trace_special(),
which in turn can call wake_up(). Therefore, if any scheduler lock is
held across a call to rcu_read_unlock_trace(), self-deadlock can occur.
This commit therefore uses the irq_work facility to defer the wake_up()
to a clean environment where no scheduler locks will be held.
Reported-by: Steven Rostedt <rostedt@goodmis.org>
[ paulmck: Update #includes for m68k per kbuild test robot. ]
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Systems running CPU-bound real-time task do not want IPIs sent to CPUs
executing nohz_full userspace tasks. Battery-powered systems don't
want IPIs sent to idle CPUs in low-power mode. Unfortunately, RCU tasks
trace can and will send such IPIs in some cases.
Both of these situations occur only when the target CPU is in RCU
dyntick-idle mode, in other words, when RCU is not watching the
target CPU. This suggests that CPUs in dyntick-idle mode should use
memory barriers in outermost invocations of rcu_read_lock_trace()
and rcu_read_unlock_trace(), which would allow the RCU tasks trace
grace period to directly read out the target CPU's read-side state.
One challenge is that RCU tasks trace is not targeting a specific
CPU, but rather a task. And that task could switch from one CPU to
another at any time.
This commit therefore uses try_invoke_on_locked_down_task()
and checks for task_curr() in trc_inspect_reader_notrunning().
When this condition holds, the target task is running and cannot move.
If CONFIG_TASKS_TRACE_RCU_READ_MB=y, the new rcu_dynticks_zero_in_eqs()
function can be used to check if the specified integer (in this case,
t->trc_reader_nesting) is zero while the target CPU remains in that same
dyntick-idle sojourn. If so, the target task is in a quiescent state.
If not, trc_read_check_handler() must indicate failure so that the
grace-period kthread can take appropriate action or retry after an
appropriate delay, as the case may be.
With this change, given CONFIG_TASKS_TRACE_RCU_READ_MB=y, if a given
CPU remains idle or a given task continues executing in nohz_full mode,
the RCU tasks trace grace-period kthread will detect this without the
need to send an IPI.
Suggested-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit provides a new TASKS_TRACE_RCU_READ_MB Kconfig option that
enables use of read-side memory barriers by both rcu_read_lock_trace()
and rcu_read_unlock_trace() when the are executed with the
current->trc_reader_special.b.need_mb flag set. This flag is currently
never set. Doing that is the subject of a later commit.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit adds a grace-period count and a count of IPIs sent since
boot, which is printed in response to rcutorture writer stalls and at
the end of rcutorture testing. These counts will be used to evaluate
various schemes to reduce the number of IPIs sent.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit splits ->trc_reader_need_end by using the rcu_special union.
This change permits readers to check to see if a memory barrier is
required without any added overhead in the common case where no such
barrier is required. This commit also adds the read-side checking.
Later commits will add the machinery to properly set the new
->trc_reader_special.b.need_mb field.
This commit also makes rcu_read_unlock_trace_special() tolerate nested
read-side critical sections within interrupt and NMI handlers.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit provides a rcupdate.rcu_task_ipi_delay kernel boot parameter
that specifies how old the RCU tasks trace grace period must be before
the grace-period kthread starts sending IPIs. This delay allows more
tasks to pass through rcu_tasks_qs() quiescent states, thus reducing
(or even eliminating) the number of IPIs that must be sent.
On a short rcutorture test setting this kernel boot parameter to HZ/2
resulted in zero IPIs for all 877 RCU-tasks trace grace periods that
elapsed during that test.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit adds a place to record the grace-period start in jiffies.
This will be used by later commits for debugging purposes and to throttle
IPIs early in the grace period.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit makes the calls to rcu_tasks_qs() detect and report
quiescent states for RCU tasks trace. If the task is in a quiescent
state and if ->trc_reader_checked is not yet set, the task sets its own
->trc_reader_checked. This will cause the grace-period kthread to
remove it from the holdout list if it still remains there.
[ paulmck: Fix conditional compilation per kbuild test robot feedback. ]
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit adds state for each RCU-tasks flavor to the rcutorture
writer stall output. The initial state is minimal, but you have to
start somewhere.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
[ paulmck: Fixes based on feedback from kbuild test robot. ]
This commit pushes the #ifdef CONFIG_TASKS_RCU_GENERIC from
kernel/rcu/update.c to kernel/rcu/tasks.h in order to improve
readability as more APIs are added.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
