sched_ext/scx_qmap: Pick idle CPU for direct dispatch on !wakeup enqueues
Because there was no way to directly dispatch to the local DSQ of a remote
CPU from ops.enqueue(), scx_qmap skipped looking for an idle CPU on !wakeup
enqueues. This restriction was removed and sched_ext now allows
SCX_DSQ_LOCAL_ON verdicts for direct dispatches.
Factor out pick_direct_dispatch_cpu() from ops.select_cpu() and use it to
direct dispatch from ops.enqueue() on !wakeup enqueues.
Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: David Vernet <void@manifault.com> Cc: Dan Schatzberg <schatzberg.dan@gmail.com> Cc: Changwoo Min <changwoo@igalia.com> Cc: Andrea Righi <righi.andrea@gmail.com>
sched_ext: Allow SCX_DSQ_LOCAL_ON for direct dispatches
In ops.dispatch(), SCX_DSQ_LOCAL_ON can be used to dispatch the task to the
local DSQ of any CPU. However, during direct dispatch from ops.select_cpu()
and ops.enqueue(), this isn't allowed. This is because dispatching to the
local DSQ of a remote CPU requires locking both the task's current and new
rq's and such double locking can't be done directly from ops.enqueue().
While waking up a task, as ops.select_cpu() can pick any CPU and both
ops.select_cpu() and ops.enqueue() can use SCX_DSQ_LOCAL as the dispatch
target to dispatch to the DSQ of the picked CPU, the BPF scheduler can still
do whatever it wants to do. However, while a task is being enqueued for a
different reason, e.g. after its slice expiration, only ops.enqueue() is
called and there's no way for the BPF scheduler to directly dispatch to the
local DSQ of a remote CPU. This gap in API forces schedulers into
work-arounds which are not straightforward or optimal such as skipping
direct dispatches in such cases.
Implement deferred enqueueing to allow directly dispatching to the local DSQ
of a remote CPU from ops.select_cpu() and ops.enqueue(). Such tasks are
temporarily queued on rq->scx.ddsp_deferred_locals. When the rq lock can be
safely released, the tasks are taken off the list and queued on the target
local DSQs using dispatch_to_local_dsq().
v2: - Add missing return after queue_balance_callback() in
schedule_deferred(). (David).
- dispatch_to_local_dsq() now assumes that @rq is locked but unpinned
and thus no longer takes @rf. Updated accordingly.
- UP build warning fix.
Signed-off-by: Tejun Heo <tj@kernel.org> Tested-by: Andrea Righi <righi.andrea@gmail.com> Acked-by: David Vernet <void@manifault.com> Cc: Dan Schatzberg <schatzberg.dan@gmail.com> Cc: Changwoo Min <changwoo@igalia.com>
sched_ext: s/SCX_RQ_BALANCING/SCX_RQ_IN_BALANCE/ and add SCX_RQ_IN_WAKEUP
SCX_RQ_BALANCING is used to mark that the rq is currently in balance().
Rename it to SCX_RQ_IN_BALANCE and add SCX_RQ_IN_WAKEUP which marks whether
the rq is currently enqueueing for a wakeup. This will be used to implement
direct dispatching to local DSQ of another CPU.
Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: David Vernet <void@manifault.com>
sched_ext: Unpin and repin rq lock from balance_scx()
sched_ext often needs to migrate tasks across CPUs right before execution
and thus uses the balance path to dispatch tasks from the BPF scheduler.
balance_scx() is called with rq locked and pinned but is passed @rf and thus
allowed to unpin and unlock. Currently, @rf is passed down the call stack so
the rq lock is unpinned just when double locking is needed.
This creates unnecessary complications such as having to explicitly
manipulate lock pinning for core scheduling. We also want to use
dispatch_to_local_dsq_lock() from other paths which are called with rq
locked but unpinned.
rq lock handling in the dispatch path is straightforward outside the
migration implementation and extending the pinning protection down the
callstack doesn't add enough meaningful extra protection to justify the
extra complexity.
Unpin and repin rq lock from the outer balance_scx() and drop @rf passing
and lock pinning handling from the inner functions. UP is updated to call
balance_one() instead of balance_scx() to avoid adding NULL @rf handling to
balance_scx(). AS this makes balance_scx() unused in UP, it's put inside a
CONFIG_SMP block.
No functional changes intended outside of lock annotation updates.
Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: David Vernet <void@manifault.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Andrea Righi <righi.andrea@gmail.com>
task_linked_on_dsq() exists as a helper because it used to test both the
rbtree and list nodes. It now only tests the list node and the list node
will soon be used for something else too. The helper doesn't improve
anything materially and the naming will become confusing. Open-code the list
node testing and remove task_linked_on_dsq()
Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: David Vernet <void@manifault.com>
Move struct balance_callback definition upward so that it's visible to
class-specific rq struct definitions. This will be used to embed a struct
balance_callback in struct scx_rq.
No functional changes.
Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: David Vernet <void@manifault.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org>
sched_ext: Make scx_bpf_reenqueue_local() skip tasks that are being migrated
When a running task is migrated to another CPU, the stop_task is used to
preempt the running task and migrate it. This, expectedly, invokes
ops.cpu_release(). If the BPF scheduler then calls
scx_bpf_reenqueue_local(), it re-enqueues all tasks on the local DSQ
including the task which is being migrated.
This creates an unnecessary re-enqueue of a task which is about to be
deactivated and re-activated for migration anyway. It can also cause
confusion for the BPF scheduler as scx_bpf_task_cpu() of the task and its
allowed CPUs may not agree while migration is pending.
scx_bpf_reenqueue_local() is used to re-enqueue tasks on the local DSQ from
ops.cpu_release(). Because the BPF scheduler may dispatch tasks to the same
local DSQ, to avoid processing the same tasks repeatedly, it first takes the
number of queued tasks and processes the task at the head of the queue that
number of times.
This is incorrect as a task can be dispatched to the same local DSQ with
SCX_ENQ_HEAD. Such a task will be processed repeatedly until the count is
exhausted and the succeeding tasks won't be processed at all.
Fix it by first moving all candidate tasks to a private list and then
processing that list. While at it, remove the WARNs. They're rather
superflous as later steps will check them anyway.
DSQs are very opaque in the consumption path. The BPF scheduler has no way
of knowing which tasks are being considered and which is picked. This patch
adds BPF DSQ iterator.
- Allows iterating tasks queued on a DSQ in the dispatch order or reverse
from anywhere using bpf_for_each(scx_dsq) or calling the iterator kfuncs
directly.
- Has ordering guarantee where only tasks which were already queued when the
iteration started are visible and consumable during the iteration.
v5: - Add a comment to the naked list_empty(&dsq->list) test in
consume_dispatch_q() to explain the reasoning behind the lockless test
and by extension why nldsq_next_task() isn't used there.
- scx_qmap changes separated into its own patch.
v4: - bpf_iter_scx_dsq_new() declaration in common.bpf.h was using the wrong
type for the last argument (bool rev instead of u64 flags). Fix it.
v3: - Alexei pointed out that the iterator is too big to allocate on stack.
Added a prep patch to reduce the size of the cursor. Now
bpf_iter_scx_dsq is 48 bytes and bpf_iter_scx_dsq_kern is 40 bytes on
64bit.
- u32_before() comparison factored out.
v2: - scx_bpf_consume_task() is separated out into a separate patch.
- DSQ seq and iter flags don't need to be u64. Use u32.
sched_ext: Take out ->priq and ->flags from scx_dsq_node
struct scx_dsq_node contains two data structure nodes to link the containing
task to a DSQ and a flags field that is protected by the lock of the
associated DSQ. One reason why they are grouped into a struct is to use the
type independently as a cursor node when iterating tasks on a DSQ. However,
when iterating, the cursor only needs to be linked on the FIFO list and the
rb_node part ends up inflating the size of the iterator data structure
unnecessarily making it potentially too expensive to place it on stack.
Take ->priq and ->flags out of scx_dsq_node and put them in sched_ext_entity
as ->dsq_priq and ->dsq_flags, respectively. scx_dsq_node is renamed to
scx_dsq_list_node and the field names are renamed accordingly. This will
help implementing DSQ task iterator that can be allocated on stack.
sched, sched_ext: Move some declarations from kernel/sched/ext.h to sched.h
While sched_ext was out of tree, everything sched_ext specific which can be
put in kernel/sched/ext.h was put there to ease forward porting. However,
kernel/sched/sched.h is the better location for some of them. Relocate.
- struct sched_enq_and_set_ctx, sched_deq_and_put_task() and
sched_enq_and_set_task().
- scx_enabled() and scx_switched_all().
- for_active_class_range() and for_each_active_class(). sched_class
declarations are moved above the class iterators for this.
No functional changes intended.
Signed-off-by: Tejun Heo <tj@kernel.org> Suggested-by: Linus Torvalds <torvalds@linux-foundation.org> Cc: David Vernet <void@manifault.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de>
sched, sched_ext: Open code for_balance_class_range()
For flexibility, sched_ext allows the BPF scheduler to select the CPU to
execute a task on at dispatch time so that e.g. a queue can be shared across
multiple CPUs. To enable this, the dispatch path is executed from balance()
so that a dispatched task can be hot-migrated to its target CPU. This means
that sched_ext needs its balance() method invoked before every
pick_next_task() even when the CPU is waking up from SCHED_IDLE.
for_balance_class_range() defined in kernel/sched/ext.h implements this
selective iteration promotion. However, the indirection obfuscates more than
helps. Open code the iteration promotion in put_prev_task_balance() and
remove for_balance_class_range().
No functional changes intended.
Signed-off-by: Tejun Heo <tj@kernel.org> Suggested-by: Linus Torvalds <torvalds@linux-foundation.org> Acked-by: David Vernet <void@manifault.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de>
Tejun Heo [Thu, 27 Jun 2024 01:27:02 +0000 (15:27 -1000)]
sched_ext: Disallow loading BPF scheduler if isolcpus= domain isolation is in effect
sched_domains regulate the load balancing for sched_classes. A machine can
be partitioned into multiple sections that are not load-balanced across
using either isolcpus= boot param or cpuset partitions. In such cases, tasks
that are in one partition are expected to stay within that partition.
cpuset configured partitions are always reflected in each member task's
cpumask. As SCX always honors the task cpumasks, the BPF scheduler is
automatically in compliance with the configured partitions.
However, for isolcpus= domain isolation, the isolated CPUs are simply
omitted from the top-level sched_domain[s] without further restrictions on
tasks' cpumasks, so, for example, a task currently running in an isolated
CPU may have more CPUs in its allowed cpumask while expected to remain on
the same CPU.
There is no straightforward way to enforce this partitioning preemptively on
BPF schedulers and erroring out after a violation can be surprising.
isolcpus= domain isolation is being replaced with cpuset partitions anyway,
so keep it simple and simply disallow loading a BPF scheduler if isolcpus=
domain isolation is in effect.
sched_ext: Account for idle policy when setting p->scx.weight in scx_ops_enable_task()
When initializing p->scx.weight, scx_ops_enable_task() wasn't considering
whether the task is SCHED_IDLE. Update it to use WEIGHT_IDLEPRIO as the
source weight for SCHED_IDLE tasks. This leaves reweight_task_scx() the sole
user of set_task_scx_weight(). Open code it. @weight is going to be provided
by sched core in the future anyway.
v2: Use the newly available @lw->weight to set @p->scx.weight in
reweight_task_scx().
Signed-off-by: Tejun Heo <tj@kernel.org> Cc: David Vernet <void@manifault.com> Cc: Peter Zijlstra <peterz@infradead.org>
Tejun Heo [Mon, 24 Jun 2024 20:51:49 +0000 (10:51 -1000)]
sched, sched_ext: Simplify dl_prio() case handling in sched_fork()
sched_fork() returns with -EAGAIN if dl_prio(@p). a7a9fc549293 ("sched_ext:
Add boilerplate for extensible scheduler class") added scx_pre_fork() call
before it and then scx_cancel_fork() on the exit path. This is silly as the
dl_prio() block can just be moved above the scx_pre_fork() call.
Move the dl_prio() block above the scx_pre_fork() call and remove the now
unnecessary scx_cancel_fork() invocation.
Signed-off-by: Tejun Heo <tj@kernel.org> Suggested-by: Thomas Gleixner <tglx@linutronix.de> Cc: Ingo Molnar <mingo@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: David Vernet <void@manifault.com>
rq contains many useful fields to implement a custom scheduler. For
example, various clock signals like clock_task and clock_pelt can be
used to track load. It also contains stats in other sched_classes, which
are useful to drive scheduling decisions in ext.
tj: Put the new helper below scx_bpf_task_*() helpers.
Merge branch 'sched/core' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip into for-6.11
d32960528702 ("sched/fair: set_load_weight() must also call reweight_task()
for SCHED_IDLE tasks") applied to sched/core changes how reweight_task() is
called causing conflicts with e83edbf88f18 ("sched: Add
sched_class->reweight_task()"). Resolve the conflicts by taking
set_load_weight() changes from d32960528702 and updating
sched_class->reweight_task() to take pointer to struct load_weight instead
of int prio.
Tejun Heo [Wed, 26 Jun 2024 01:29:58 +0000 (15:29 -1000)]
sched/fair: set_load_weight() must also call reweight_task() for SCHED_IDLE tasks
When a task's weight is being changed, set_load_weight() is called with
@update_load set. As weight changes aren't trivial for the fair class,
set_load_weight() calls fair.c::reweight_task() for fair class tasks.
However, set_load_weight() first tests task_has_idle_policy() on entry and
skips calling reweight_task() for SCHED_IDLE tasks. This is buggy as
SCHED_IDLE tasks are just fair tasks with a very low weight and they would
incorrectly skip load, vlag and position updates.
Fix it by updating reweight_task() to take struct load_weight as idle weight
can't be expressed with prio and making set_load_weight() call
reweight_task() for SCHED_IDLE tasks too when @update_load is set.
Fixes: 9059393e4ec1 ("sched/fair: Use reweight_entity() for set_user_nice()") Suggested-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Tejun Heo <tj@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: stable@vger.kernel.org # v4.15+ Link: http://lkml.kernel.org/r/20240624102331.GI31592@noisy.programming.kicks-ass.net
Tvrtko Ursulin [Tue, 25 Jun 2024 13:50:00 +0000 (14:50 +0100)]
sched/psi: Optimise psi_group_change a bit
The current code loops over the psi_states only to call a helper which
then resolves back to the action needed for each state using a switch
statement. That is effectively creating a double indirection of a kind
which, given how all the states need to be explicitly listed and handled
anyway, we can simply remove. Both the for loop and the switch statement
that is.
The benefit is both in the code size and CPU time spent in this function.
YMMV but on my Steam Deck, while in a game, the patch makes the CPU usage
go from ~2.4% down to ~1.2%. Text size at the same time went from 0x323 to
0x2c1.
Signed-off-by: Tvrtko Ursulin <tursulin@ursulin.net> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Chengming Zhou <chengming.zhou@linux.dev> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Link: https://lkml.kernel.org/r/20240625135000.38652-1-tursulin@igalia.com
sched_ext: Swap argument positions in kcalloc() call to avoid compiler warning
alloc_exit_info() calls kcalloc() but puts in the size of the element as the
first argument which triggers the following gcc warning:
kernel/sched/ext.c:3815:32: warning: ‘kmalloc_array_noprof’ sizes
specified with ‘sizeof’ in the earlier argument and not in the later
argument [-Wcalloc-transposed-args]
Fix it by swapping the positions of the first two arguments. No functional
changes.
Tejun Heo [Tue, 25 Jun 2024 02:00:37 +0000 (16:00 -1000)]
sched_ext: Drop tools_clean target from the top-level Makefile
2a52ca7c9896 ("sched_ext: Add scx_simple and scx_example_qmap example
schedulers") added the tools_clean target which is triggered by mrproper.
The tools_clean target triggers the sched_ext_clean target in tools/. This
unfortunately makes mrproper fail when no BTF enabled kernel image is found:
Makefile:83: *** Cannot find a vmlinux for VMLINUX_BTF at any of " ../../vmlinux /sys/kernel/btf/vmlinux/boot/vmlinux-4.15.0-136-generic". Stop.
Makefile:192: recipe for target 'sched_ext_clean' failed
make[2]: *** [sched_ext_clean] Error 2
Makefile:1361: recipe for target 'sched_ext' failed
make[1]: *** [sched_ext] Error 2
Makefile:240: recipe for target '__sub-make' failed
make: *** [__sub-make] Error 2
Clean targets shouldn't fail like this but also it's really odd for mrproper
to single out and trigger the sched_ext_clean target when no other clean
targets under tools/ are triggered.
Fix builds by dropping the tools_clean target from the top-level Makefile.
The offending Makefile line is shared across BPF targets under tools/. Let's
revisit them later.
Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Jon Hunter <jonathanh@nvidia.com> Link: http://lkml.kernel.org/r/ac065f1f-8754-4626-95db-2c9fcf02567b@nvidia.com Fixes: 2a52ca7c9896 ("sched_ext: Add scx_simple and scx_example_qmap example schedulers") Cc: David Vernet <void@manifault.com>
David Vernet [Sun, 23 Jun 2024 08:10:36 +0000 (03:10 -0500)]
sched_ext: Make scx_bpf_cpuperf_set() @cpu arg signed
The scx_bpf_cpuperf_set() kfunc allows a BPF program to set the relative
performance target of a specified CPU. Commit d86adb4fc065 ("sched_ext: Add
cpuperf support") defined the @cpu argument to be unsigned. Let's update it
to be signed to match the norm for the rest of ext.c and the kernel.
Note that the kfunc declaration of scx_bpf_cpuperf_set() in the
common.bpf.h header in tools/sched_ext already listed the cpu as signed, so
this also fixes the build for tools/sched_ext and the sched_ext selftests
due to kfunc declarations now being emitted in vmlinux.h based on BTF (thus
causing the compiler to error due to observing conflicting types).
Tejun Heo [Fri, 21 Jun 2024 22:37:22 +0000 (12:37 -1000)]
sched_ext: Add cpuperf support
sched_ext currently does not integrate with schedutil. When schedutil is the
governor, frequencies are left unregulated and usually get stuck close to
the highest performance level from running RT tasks.
Add CPU performance monitoring and scaling support by integrating into
schedutil. The following kfuncs are added:
- scx_bpf_cpuperf_cap(): Query the relative performance capacity of
different CPUs in the system.
- scx_bpf_cpuperf_cur(): Query the current performance level of a CPU
relative to its max performance.
- scx_bpf_cpuperf_set(): Set the current target performance level of a CPU.
This gives direct control over CPU performance setting to the BPF scheduler.
The only changes on the schedutil side are accounting for the utilization
factor from sched_ext and disabling frequency holding heuristics as it may
not apply well to sched_ext schedulers which may have a lot weaker
connection between tasks and their current / last CPU.
With cpuperf support added, there is no reason to block uclamp. Enable while
at it.
A toy implementation of cpuperf is added to scx_qmap as a demonstration of
the feature.
v2: Ignore cpu_util_cfs_boost() when scx_switched_all() in sugov_get_util()
to avoid factoring in stale util metric. (Christian)
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com> Cc: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Cc: Viresh Kumar <viresh.kumar@linaro.org> Cc: Christian Loehle <christian.loehle@arm.com>
Tejun Heo [Fri, 21 Jun 2024 22:37:03 +0000 (12:37 -1000)]
cpufreq_schedutil: Refactor sugov_cpu_is_busy()
sugov_cpu_is_busy() is used to avoid decreasing performance level while the
CPU is busy and called by sugov_update_single_freq() and
sugov_update_single_perf(). Both callers repeat the same pattern to first
test for uclamp and then the business. Let's refactor so that the tests
aren't repeated.
The new helper is named sugov_hold_freq() and tests both the uclamp
exception and CPU business. No functional changes. This will make adding
more exception conditions easier.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com> Reviewed-by: Christian Loehle <christian.loehle@arm.com> Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Cc: Viresh Kumar <viresh.kumar@linaro.org>
Tejun Heo [Thu, 20 Jun 2024 22:15:02 +0000 (12:15 -1000)]
sched, sched_ext: Replace scx_next_task_picked() with sched_class->switch_class()
scx_next_task_picked() is used by sched_ext to notify the BPF scheduler when
a CPU is taken away by a task dispatched from a higher priority sched_class
so that the BPF scheduler can, e.g., punt the task[s] which was running or
were waiting for the CPU to other CPUs.
Replace the sched_ext specific hook scx_next_task_picked() with a new
sched_class operation switch_class().
The changes are straightforward and the code looks better afterwards.
However, when !CONFIG_SCHED_CLASS_EXT, this ends up adding an unused hook
which is unlikely to be useful to other sched_classes. For further
discussion on this subject, please refer to the following:
Tejun Heo [Tue, 18 Jun 2024 20:09:21 +0000 (10:09 -1000)]
sched_ext: Add vtime-ordered priority queue to dispatch_q's
Currently, a dsq is always a FIFO. A task which is dispatched earlier gets
consumed or executed earlier. While this is sufficient when dsq's are used
for simple staging areas for tasks which are ready to execute, it'd make
dsq's a lot more useful if they can implement custom ordering.
This patch adds a vtime-ordered priority queue to dsq's. When the BPF
scheduler dispatches a task with the new scx_bpf_dispatch_vtime() helper, it
can specify the vtime tha the task should be inserted at and the task is
inserted into the priority queue in the dsq which is ordered according to
time_before64() comparison of the vtime values.
A DSQ can either be a FIFO or priority queue and automatically switches
between the two depending on whether scx_bpf_dispatch() or
scx_bpf_dispatch_vtime() is used. Using the wrong variant while the DSQ
already has the other type queued is not allowed and triggers an ops error.
Built-in DSQs must always be FIFOs.
This makes it very easy for the BPF schedulers to implement proper vtime
based scheduling within each dsq very easy and efficient at a negligible
cost in terms of code complexity and overhead.
scx_simple and scx_example_flatcg are updated to default to weighted
vtime scheduling (the latter within each cgroup). FIFO scheduling can be
selected with -f option.
v4: - As allowing mixing priority queue and FIFO on the same DSQ sometimes
led to unexpected starvations, DSQs now error out if both modes are
used at the same time and the built-in DSQs are no longer allowed to
be priority queues.
- Explicit type struct scx_dsq_node added to contain fields needed to be
linked on DSQs. This will be used to implement stateful iterator.
- Tasks are now always linked on dsq->list whether the DSQ is in FIFO or
PRIQ mode. This confines PRIQ related complexities to the enqueue and
dequeue paths. Other paths only need to look at dsq->list. This will
also ease implementing BPF iterator.
- Print p->scx.dsq_flags in debug dump.
v3: - SCX_TASK_DSQ_ON_PRIQ flag is moved from p->scx.flags into its own
p->scx.dsq_flags. The flag is protected with the dsq lock unlike other
flags in p->scx.flags. This led to flag corruption in some cases.
- Add comments explaining the interaction between using consumption of
p->scx.slice to determine vtime progress and yielding.
v2: - p->scx.dsq_vtime was not initialized on load or across cgroup
migrations leading to some tasks being stalled for extended period of
time depending on how saturated the machine is. Fixed.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com>
Tejun Heo [Tue, 18 Jun 2024 20:09:20 +0000 (10:09 -1000)]
sched_ext: Implement core-sched support
The core-sched support is composed of the following parts:
- task_struct->scx.core_sched_at is added. This is a timestamp which can be
used to order tasks. Depending on whether the BPF scheduler implements
custom ordering, it tracks either global FIFO ordering of all tasks or
local-DSQ ordering within the dispatched tasks on a CPU.
- prio_less() is updated to call scx_prio_less() when comparing SCX tasks.
scx_prio_less() calls ops.core_sched_before() if available or uses the
core_sched_at timestamp. For global FIFO ordering, the BPF scheduler
doesn't need to do anything. Otherwise, it should implement
ops.core_sched_before() which reflects the ordering.
- When core-sched is enabled, balance_scx() balances all SMT siblings so
that they all have tasks dispatched if necessary before pick_task_scx() is
called. pick_task_scx() picks between the current task and the first
dispatched task on the local DSQ based on availability and the
core_sched_at timestamps. Note that FIFO ordering is expected among the
already dispatched tasks whether running or on the local DSQ, so this path
always compares core_sched_at instead of calling into
ops.core_sched_before().
qmap_core_sched_before() is added to scx_qmap. It scales the
distances from the heads of the queues to compare the tasks across different
priority queues and seems to behave as expected.
v3: Fixed build error when !CONFIG_SCHED_SMT reported by Andrea Righi.
v2: Sched core added the const qualifiers to prio_less task arguments.
Explicitly drop them for ops.core_sched_before() task arguments. BPF
enforces access control through the verifier, so the qualifier isn't
actually operative and only gets in the way when interacting with
various helpers.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com> Reviewed-by: Josh Don <joshdon@google.com> Cc: Andrea Righi <andrea.righi@canonical.com>
Tejun Heo [Tue, 18 Jun 2024 20:09:20 +0000 (10:09 -1000)]
sched_ext: Bypass BPF scheduler while PM events are in progress
PM operations freeze userspace. Some BPF schedulers have active userspace
component and may misbehave as expected across PM events. While the system
is frozen, nothing too interesting is happening in terms of scheduling and
we can get by just fine with the fallback FIFO behavior. Let's make things
easier by always bypassing the BPF scheduler while PM events are in
progress.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com>
Add ops.cpu_online/offline() which are invoked when CPUs come online and
offline respectively. As the enqueue path already automatically bypasses
tasks to the local dsq on a deactivated CPU, BPF schedulers are guaranteed
to see tasks only on CPUs which are between online() and offline().
If the BPF scheduler doesn't implement ops.cpu_online/offline(), the
scheduler is automatically exited with SCX_ECODE_RESTART |
SCX_ECODE_RSN_HOTPLUG. Userspace can implement CPU hotpplug support
trivially by simply reinitializing and reloading the scheduler.
scx_qmap is updated to print out online CPUs on hotplug events. Other
schedulers are updated to restart based on ecode.
v3: - The previous implementation added @reason to
sched_class.rq_on/offline() to distinguish between CPU hotplug events
and topology updates. This was buggy and fragile as the methods are
skipped if the current state equals the target state. Instead, add
scx_rq_[de]activate() which are directly called from
sched_cpu_de/activate(). This also allows ops.cpu_on/offline() to
sleep which can be useful.
- ops.dispatch() could be called on a CPU that the BPF scheduler was
told to be offline. The dispatch patch is updated to bypass in such
cases.
v2: - To accommodate lock ordering change between scx_cgroup_rwsem and
cpus_read_lock(), CPU hotplug operations are put into its own SCX_OPI
block and enabled eariler during scx_ope_enable() so that
cpus_read_lock() can be dropped before acquiring scx_cgroup_rwsem.
- Auto exit with ECODE added.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com> Acked-by: Josh Don <joshdon@google.com> Acked-by: Hao Luo <haoluo@google.com> Acked-by: Barret Rhoden <brho@google.com>
Scheduler classes are strictly ordered and when a higher priority class has
tasks to run, the lower priority ones lose access to the CPU. Being able to
monitor and act on these events are necessary for use cases includling
strict core-scheduling and latency management.
This patch adds two operations ops.cpu_acquire() and .cpu_release(). The
former is invoked when a CPU becomes available to the BPF scheduler and the
opposite for the latter. This patch also implements
scx_bpf_reenqueue_local() which can be called from .cpu_release() to trigger
requeueing of all tasks in the local dsq of the CPU so that the tasks can be
reassigned to other available CPUs.
scx_pair is updated to use .cpu_acquire/release() along with
%SCX_KICK_WAIT to make the pair scheduling guarantee strict even when a CPU
is preempted by a higher priority scheduler class.
scx_qmap is updated to use .cpu_acquire/release() to empty the local
dsq of a preempted CPU. A similar approach can be adopted by BPF schedulers
that want to have a tight control over latency.
v4: Use the new SCX_KICK_IDLE to wake up a CPU after re-enqueueing.
v3: Drop the const qualifier from scx_cpu_release_args.task. BPF enforces
access control through the verifier, so the qualifier isn't actually
operative and only gets in the way when interacting with various
helpers.
v2: Add p->scx.kf_mask annotation to allow calling scx_bpf_reenqueue_local()
from ops.cpu_release() nested inside ops.init() and other sleepable
operations.
Signed-off-by: David Vernet <dvernet@meta.com> Reviewed-by: Tejun Heo <tj@kernel.org> Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Josh Don <joshdon@google.com> Acked-by: Hao Luo <haoluo@google.com> Acked-by: Barret Rhoden <brho@google.com>
David Vernet [Tue, 18 Jun 2024 20:09:20 +0000 (10:09 -1000)]
sched_ext: Implement SCX_KICK_WAIT
If set when calling scx_bpf_kick_cpu(), the invoking CPU will busy wait for
the kicked cpu to enter the scheduler. See the following for example usage:
Tejun Heo [Tue, 18 Jun 2024 20:09:19 +0000 (10:09 -1000)]
sched_ext: Track tasks that are subjects of the in-flight SCX operation
When some SCX operations are in flight, it is known that the subject task's
rq lock is held throughout which makes it safe to access certain fields of
the task - e.g. its current task_group. We want to add SCX kfunc helpers
that can make use of this guarantee - e.g. to help determining the currently
associated CPU cgroup from the task's current task_group.
As it'd be dangerous call such a helper on a task which isn't rq lock
protected, the helper should be able to verify the input task and reject
accordingly. This patch adds sched_ext_entity.kf_tasks[] that track the
tasks which are currently being operated on by a terminal SCX operation. The
new SCX_CALL_OP_[2]TASK[_RET]() can be used when invoking SCX operations
which take tasks as arguments and the scx_kf_allowed_on_arg_tasks() can be
used by kfunc helpers to verify the input task status.
Note that as sched_ext_entity.kf_tasks[] can't handle nesting, the tracking
is currently only limited to terminal SCX operations. If needed in the
future, this restriction can be removed by moving the tracking to the task
side with a couple per-task counters.
v2: Updated to reflect the addition of SCX_KF_SELECT_CPU.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com>
Tejun Heo [Tue, 18 Jun 2024 20:09:19 +0000 (10:09 -1000)]
sched_ext: Implement tickless support
Allow BPF schedulers to indicate tickless operation by setting p->scx.slice
to SCX_SLICE_INF. A CPU whose current task has infinte slice goes into
tickless operation.
scx_central is updated to use tickless operations for all tasks and
instead use a BPF timer to expire slices. This also uses the SCX_ENQ_PREEMPT
and task state tracking added by the previous patches.
Currently, there is no way to pin the timer on the central CPU, so it may
end up on one of the worker CPUs; however, outside of that, the worker CPUs
can go tickless both while running sched_ext tasks and idling.
While scx_central itself is too barebone to be useful as a
production scheduler, a more featureful central scheduler can be built using
the same approach. Google's experience shows that such an approach can have
significant benefits for certain applications such as VM hosting.
v4: Allow operation even if BPF_F_TIMER_CPU_PIN is not available.
v3: Pin the central scheduler's timer on the central_cpu using
BPF_F_TIMER_CPU_PIN.
v2: Convert to BPF inline iterators.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com> Acked-by: Josh Don <joshdon@google.com> Acked-by: Hao Luo <haoluo@google.com> Acked-by: Barret Rhoden <brho@google.com>
Tejun Heo [Tue, 18 Jun 2024 20:09:19 +0000 (10:09 -1000)]
sched_ext: Add task state tracking operations
Being able to track the task runnable and running state transitions are
useful for a variety of purposes including latency tracking and load factor
calculation.
Currently, BPF schedulers don't have a good way of tracking these
transitions. Becoming runnable can be determined from ops.enqueue() but
becoming quiescent can only be inferred from the lack of subsequent enqueue.
Also, as the local dsq can have multiple tasks and some events are handled
in the sched_ext core, it's difficult to determine when a given task starts
and stops executing.
This patch adds sched_ext_ops.runnable(), .running(), .stopping() and
.quiescent() operations to track the task runnable and running state
transitions. They're mostly self explanatory; however, we want to ensure
that running <-> stopping transitions are always contained within runnable
<-> quiescent transitions which is a bit different from how the scheduler
core behaves. This adds a bit of complication. See the comment in
dequeue_task_scx().
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com> Acked-by: Josh Don <joshdon@google.com> Acked-by: Hao Luo <haoluo@google.com> Acked-by: Barret Rhoden <brho@google.com>
Tejun Heo [Tue, 18 Jun 2024 20:09:19 +0000 (10:09 -1000)]
sched_ext: Make watchdog handle ops.dispatch() looping stall
The dispatch path retries if the local DSQ is still empty after
ops.dispatch() either dispatched or consumed a task. This is both out of
necessity and for convenience. It has to retry because the dispatch path
might lose the tasks to dequeue while the rq lock is released while trying
to migrate tasks across CPUs, and the retry mechanism makes ops.dispatch()
implementation easier as it only needs to make some forward progress each
iteration.
However, this makes it possible for ops.dispatch() to stall CPUs by
repeatedly dispatching ineligible tasks. If all CPUs are stalled that way,
the watchdog or sysrq handler can't run and the system can't be saved. Let's
address the issue by breaking out of the dispatch loop after 32 iterations.
It is unlikely but not impossible for ops.dispatch() to legitimately go over
the iteration limit. We want to come back to the dispatch path in such cases
as not doing so risks stalling the CPU by idling with runnable tasks
pending. As the previous task is still current in balance_scx(),
resched_curr() doesn't do anything - it will just get cleared. Let's instead
use scx_kick_bpf() which will trigger reschedule after switching to the next
task which will likely be the idle task.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com>
Tejun Heo [Tue, 18 Jun 2024 20:09:19 +0000 (10:09 -1000)]
sched_ext: Add a central scheduler which makes all scheduling decisions on one CPU
This patch adds a new example scheduler, scx_central, which demonstrates
central scheduling where one CPU is responsible for making all scheduling
decisions in the system using scx_bpf_kick_cpu(). The central CPU makes
scheduling decisions for all CPUs in the system, queues tasks on the
appropriate local dsq's and preempts the worker CPUs. The worker CPUs in
turn preempt the central CPU when it needs tasks to run.
Currently, every CPU depends on its own tick to expire the current task. A
follow-up patch implementing tickless support for sched_ext will allow the
worker CPUs to go full tickless so that they can run completely undisturbed.
v3: - Kumar fixed a bug where the dispatch path could overflow the dispatch
buffer if too many are dispatched to the fallback DSQ.
- Use the new SCX_KICK_IDLE to wake up non-central CPUs.
- Dropped '-p' option.
v2: - Use RESIZABLE_ARRAY() instead of fixed MAX_CPUS and use SCX_BUG[_ON]()
to simplify error handling.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com> Acked-by: Josh Don <joshdon@google.com> Acked-by: Hao Luo <haoluo@google.com> Acked-by: Barret Rhoden <brho@google.com> Cc: Kumar Kartikeya Dwivedi <memxor@gmail.com> Cc: Julia Lawall <julia.lawall@inria.fr>
Tejun Heo [Tue, 18 Jun 2024 20:09:19 +0000 (10:09 -1000)]
sched_ext: Implement scx_bpf_kick_cpu() and task preemption support
It's often useful to wake up and/or trigger reschedule on other CPUs. This
patch adds scx_bpf_kick_cpu() kfunc helper that BPF scheduler can call to
kick the target CPU into the scheduling path.
As a sched_ext task relinquishes its CPU only after its slice is depleted,
this patch also adds SCX_KICK_PREEMPT and SCX_ENQ_PREEMPT which clears the
slice of the target CPU's current task to guarantee that sched_ext's
scheduling path runs on the CPU.
If SCX_KICK_IDLE is specified, the target CPU is kicked iff the CPU is idle
to guarantee that the target CPU will go through at least one full sched_ext
scheduling cycle after the kicking. This can be used to wake up idle CPUs
without incurring unnecessary overhead if it isn't currently idle.
As a demonstration of how backward compatibility can be supported using BPF
CO-RE, tools/sched_ext/include/scx/compat.bpf.h is added. It provides
__COMPAT_scx_bpf_kick_cpu_IDLE() which uses SCX_KICK_IDLE if available or
becomes a regular kicking otherwise. This allows schedulers to use the new
SCX_KICK_IDLE while maintaining support for older kernels. The plan is to
temporarily use compat helpers to ease API updates and drop them after a few
kernel releases.
v5: - SCX_KICK_IDLE added. Note that this also adds a compat mechanism for
schedulers so that they can support kernels without SCX_KICK_IDLE.
This is useful as a demonstration of how new feature flags can be
added in a backward compatible way.
- kick_cpus_irq_workfn() reimplemented so that it touches the pending
cpumasks only as necessary to reduce kicking overhead on machines with
a lot of CPUs.
- tools/sched_ext/include/scx/compat.bpf.h added.
v4: - Move example scheduler to its own patch.
v3: - Make scx_example_central switch all tasks by default.
- Convert to BPF inline iterators.
v2: - Julia Lawall reported that scx_example_central can overflow the
dispatch buffer and malfunction. As scheduling for other CPUs can't be
handled by the automatic retry mechanism, fix by implementing an
explicit overflow and retry handling.
- Updated to use generic BPF cpumask helpers.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com> Acked-by: Josh Don <joshdon@google.com> Acked-by: Hao Luo <haoluo@google.com> Acked-by: Barret Rhoden <brho@google.com>
Tejun Heo [Tue, 18 Jun 2024 20:09:18 +0000 (10:09 -1000)]
tools/sched_ext: Add scx_show_state.py
There are states which are interesting but don't quite fit the interface
exposed under /sys/kernel/sched_ext. Add tools/scx_show_state.py to show
them.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com>
Tejun Heo [Tue, 18 Jun 2024 20:09:18 +0000 (10:09 -1000)]
sched_ext: Print debug dump after an error exit
If a BPF scheduler triggers an error, the scheduler is aborted and the
system is reverted to the built-in scheduler. In the process, a lot of
information which may be useful for figuring out what happened can be lost.
This patch adds debug dump which captures information which may be useful
for debugging including runqueue and runnable thread states at the time of
failure. The following shows a debug dump after triggering the watchdog:
EXIT: runnable task stall (stress[1530] failed to run for 6.841s)
It shows that CPU 3 was running the watchdog when it triggered the error
condition and the scx_qmap thread has been queued on CPU 0 for over 5
seconds but failed to run. It also prints out scx_qmap specific information
- e.g. which tasks are queued on each FIFO and so on using the dump_*() ops.
This dump has proved pretty useful for developing and debugging BPF
schedulers.
Debug dump is generated automatically when the BPF scheduler exits due to an
error. The debug buffer used in such cases is determined by
sched_ext_ops.exit_dump_len and defaults to 32k. If the debug dump overruns
the available buffer, the output is truncated and marked accordingly.
Debug dump output can also be read through the sched_ext_dump tracepoint.
When read through the tracepoint, there is no length limit.
SysRq-D can be used to trigger debug dump at any time while a BPF scheduler
is loaded. This is non-destructive - the scheduler keeps running afterwards.
The output can be read through the sched_ext_dump tracepoint.
v2: - The size of exit debug dump buffer can now be customized using
sched_ext_ops.exit_dump_len.
- sched_ext_ops.dump*() added to enable dumping of BPF scheduler
specific information.
- Tracpoint output and SysRq-D triggering added.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com>
David Vernet [Tue, 18 Jun 2024 20:09:18 +0000 (10:09 -1000)]
sched_ext: Print sched_ext info when dumping stack
It would be useful to see what the sched_ext scheduler state is, and what
scheduler is running, when we're dumping a task's stack. This patch
therefore adds a new print_scx_info() function that's called in the same
context as print_worker_info() and print_stop_info(). An example dump
follows.
v3: - scx_ops_enable_state_str[] definition moved to an earlier patch as
it's now used by core implementation.
- Convert jiffy delta to msecs using jiffies_to_msecs() instead of
multiplying by (HZ / MSEC_PER_SEC). The conversion is implemented in
jiffies_delta_msecs().
v2: - We are now using scx_ops_enable_state_str[] outside
CONFIG_SCHED_DEBUG. Move it outside of CONFIG_SCHED_DEBUG and to the
top. This was reported by Changwoo and Andrea.
Signed-off-by: David Vernet <void@manifault.com> Reported-by: Changwoo Min <changwoo@igalia.com> Reported-by: Andrea Righi <andrea.righi@canonical.com> Signed-off-by: Tejun Heo <tj@kernel.org>
Tejun Heo [Tue, 18 Jun 2024 20:09:18 +0000 (10:09 -1000)]
sched_ext: Allow BPF schedulers to disallow specific tasks from joining SCHED_EXT
BPF schedulers might not want to schedule certain tasks - e.g. kernel
threads. This patch adds p->scx.disallow which can be set by BPF schedulers
in such cases. The field can be changed anytime and setting it in
ops.prep_enable() guarantees that the task can never be scheduled by
sched_ext.
scx_qmap is updated with the -d option to disallow a specific PID:
David Vernet [Tue, 18 Jun 2024 20:09:18 +0000 (10:09 -1000)]
sched_ext: Implement runnable task stall watchdog
The most common and critical way that a BPF scheduler can misbehave is by
failing to run runnable tasks for too long. This patch implements a
watchdog.
* All tasks record when they become runnable.
* A watchdog work periodically scans all runnable tasks. If any task has
stayed runnable for too long, the BPF scheduler is aborted.
* scheduler_tick() monitors whether the watchdog itself is stuck. If so, the
BPF scheduler is aborted.
Because the watchdog only scans the tasks which are currently runnable and
usually very infrequently, the overhead should be negligible.
scx_qmap is updated so that it can be told to stall user and/or
kernel tasks.
A detected task stall looks like the following:
sched_ext: BPF scheduler "qmap" errored, disabling
sched_ext: runnable task stall (dbus-daemon[953] failed to run for 6.478s)
scx_check_timeout_workfn+0x10e/0x1b0
process_one_work+0x287/0x560
worker_thread+0x234/0x420
kthread+0xe9/0x100
ret_from_fork+0x1f/0x30
Note that this patch exposes scx_ops_error[_type]() in kernel/sched/ext.h to
inline scx_notify_sched_tick().
v4: - While disabling, cancel_delayed_work_sync(&scx_watchdog_work) was
being called before forward progress was guaranteed and thus could
lead to system lockup. Relocated.
- While enabling, it was comparing msecs against jiffies without
conversion leading to spurious load failures on lower HZ kernels.
Fixed.
- runnable list management is now used by core bypass logic and moved to
the patch implementing sched_ext core.
v3: - bpf_scx_init_member() was incorrectly comparing ops->timeout_ms
against SCX_WATCHDOG_MAX_TIMEOUT which is in jiffies without
conversion leading to spurious load failures in lower HZ kernels.
Fixed.
v2: - Julia Lawall noticed that the watchdog code was mixing msecs and
jiffies. Fix by using jiffies for everything.
Signed-off-by: David Vernet <dvernet@meta.com> Reviewed-by: Tejun Heo <tj@kernel.org> Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Josh Don <joshdon@google.com> Acked-by: Hao Luo <haoluo@google.com> Acked-by: Barret Rhoden <brho@google.com> Cc: Julia Lawall <julia.lawall@inria.fr>
Tejun Heo [Tue, 18 Jun 2024 20:09:17 +0000 (10:09 -1000)]
sched_ext: Add scx_simple and scx_example_qmap example schedulers
Add two simple example BPF schedulers - simple and qmap.
* simple: In terms of scheduling, it behaves identical to not having any
operation implemented at all. The two operations it implements are only to
improve visibility and exit handling. On certain homogeneous
configurations, this actually can perform pretty well.
* qmap: A fixed five level priority scheduler to demonstrate queueing PIDs
on BPF maps for scheduling. While not very practical, this is useful as a
simple example and will be used to demonstrate different features.
v7: - Compat helpers stripped out in prepartion of upstreaming as the
upstreamed patchset will be the baselinfe. Utility macros that can be
used to implement compat features are kept.
- Explicitly disable map autoattach on struct_ops to avoid trying to
attach twice while maintaining compatbility with older libbpf.
v6: - Common header files reorganized and cleaned up. Compat helpers are
added to demonstrate how schedulers can maintain backward
compatibility with older kernels while making use of newly added
features.
- simple_select_cpu() added to keep track of the number of local
dispatches. This is needed because the default ops.select_cpu()
implementation is updated to dispatch directly and won't call
ops.enqueue().
- Updated to reflect the sched_ext API changes. Switching all tasks is
the default behavior now and scx_qmap supports partial switching when
`-p` is specified.
- tools/sched_ext/Kconfig dropped. This will be included in the doc
instead.
v5: - Improve Makefile. Build artifects are now collected into a separate
dir which change be changed. Install and help targets are added and
clean actually cleans everything.
- MEMBER_VPTR() improved to improve access to structs. ARRAY_ELEM_PTR()
and RESIZEABLE_ARRAY() are added to support resizable arrays in .bss.
- Add scx_common.h which provides common utilities to user code such as
SCX_BUG[_ON]() and RESIZE_ARRAY().
- Use SCX_BUG[_ON]() to simplify error handling.
v4: - Dropped _example prefix from scheduler names.
v3: - Rename scx_example_dummy to scx_example_simple and restructure a bit
to ease later additions. Comment updates.
- Added declarations for BPF inline iterators. In the future, hopefully,
these will be consolidated into a generic BPF header so that they
don't need to be replicated here.
v2: - Updated with the generic BPF cpumask helpers.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com> Acked-by: Josh Don <joshdon@google.com> Acked-by: Hao Luo <haoluo@google.com> Acked-by: Barret Rhoden <brho@google.com>
Tejun Heo [Tue, 18 Jun 2024 20:09:17 +0000 (10:09 -1000)]
sched_ext: Implement BPF extensible scheduler class
Implement a new scheduler class sched_ext (SCX), which allows scheduling
policies to be implemented as BPF programs to achieve the following:
1. Ease of experimentation and exploration: Enabling rapid iteration of new
scheduling policies.
2. Customization: Building application-specific schedulers which implement
policies that are not applicable to general-purpose schedulers.
3. Rapid scheduler deployments: Non-disruptive swap outs of scheduling
policies in production environments.
sched_ext leverages BPF’s struct_ops feature to define a structure which
exports function callbacks and flags to BPF programs that wish to implement
scheduling policies. The struct_ops structure exported by sched_ext is
struct sched_ext_ops, and is conceptually similar to struct sched_class. The
role of sched_ext is to map the complex sched_class callbacks to the more
simple and ergonomic struct sched_ext_ops callbacks.
For more detailed discussion on the motivations and overview, please refer
to the cover letter.
Later patches will also add several example schedulers and documentation.
This patch implements the minimum core framework to enable implementation of
BPF schedulers. Subsequent patches will gradually add functionalities
including safety guarantee mechanisms, nohz and cgroup support.
include/linux/sched/ext.h defines struct sched_ext_ops. With the comment on
top, each operation should be self-explanatory. The followings are worth
noting:
- Both "sched_ext" and its shorthand "scx" are used. If the identifier
already has "sched" in it, "ext" is used; otherwise, "scx".
- In sched_ext_ops, only .name is mandatory. Every operation is optional and
if omitted a simple but functional default behavior is provided.
- A new policy constant SCHED_EXT is added and a task can select sched_ext
by invoking sched_setscheduler(2) with the new policy constant. However,
if the BPF scheduler is not loaded, SCHED_EXT is the same as SCHED_NORMAL
and the task is scheduled by CFS. When the BPF scheduler is loaded, all
tasks which have the SCHED_EXT policy are switched to sched_ext.
- To bridge the workflow imbalance between the scheduler core and
sched_ext_ops callbacks, sched_ext uses simple FIFOs called dispatch
queues (dsq's). By default, there is one global dsq (SCX_DSQ_GLOBAL), and
one local per-CPU dsq (SCX_DSQ_LOCAL). SCX_DSQ_GLOBAL is provided for
convenience and need not be used by a scheduler that doesn't require it.
SCX_DSQ_LOCAL is the per-CPU FIFO that sched_ext pulls from when putting
the next task on the CPU. The BPF scheduler can manage an arbitrary number
of dsq's using scx_bpf_create_dsq() and scx_bpf_destroy_dsq().
- sched_ext guarantees system integrity no matter what the BPF scheduler
does. To enable this, each task's ownership is tracked through
p->scx.ops_state and all tasks are put on scx_tasks list. The disable path
can always recover and revert all tasks back to CFS. See p->scx.ops_state
and scx_tasks.
- A task is not tied to its rq while enqueued. This decouples CPU selection
from queueing and allows sharing a scheduling queue across an arbitrary
subset of CPUs. This adds some complexities as a task may need to be
bounced between rq's right before it starts executing. See
dispatch_to_local_dsq() and move_task_to_local_dsq().
- One complication that arises from the above weak association between task
and rq is that synchronizing with dequeue() gets complicated as dequeue()
may happen anytime while the task is enqueued and the dispatch path might
need to release the rq lock to transfer the task. Solving this requires a
bit of complexity. See the logic around p->scx.sticky_cpu and
p->scx.ops_qseq.
- Both enable and disable paths are a bit complicated. The enable path
switches all tasks without blocking to avoid issues which can arise from
partially switched states (e.g. the switching task itself being starved).
The disable path can't trust the BPF scheduler at all, so it also has to
guarantee forward progress without blocking. See scx_ops_enable() and
scx_ops_disable_workfn().
- When sched_ext is disabled, static_branches are used to shut down the
entry points from hot paths.
v7: - scx_ops_bypass() was incorrectly and unnecessarily trying to grab
scx_ops_enable_mutex which can lead to deadlocks in the disable path.
Fixed.
- Fixed TASK_DEAD handling bug in scx_ops_enable() path which could lead
to use-after-free.
- Consolidated per-cpu variable usages and other cleanups.
v6: - SCX_NR_ONLINE_OPS replaced with SCX_OPI_*_BEGIN/END so that multiple
groups can be expressed. Later CPU hotplug operations are put into
their own group.
- SCX_OPS_DISABLING state is replaced with the new bypass mechanism
which allows temporarily putting the system into simple FIFO
scheduling mode bypassing the BPF scheduler. In addition to the shut
down path, this will also be used to isolate the BPF scheduler across
PM events. Enabling and disabling the bypass mode requires iterating
all runnable tasks. rq->scx.runnable_list addition is moved from the
later watchdog patch.
- ops.prep_enable() is replaced with ops.init_task() and
ops.enable/disable() are now called whenever the task enters and
leaves sched_ext instead of when the task becomes schedulable on
sched_ext and stops being so. A new operation - ops.exit_task() - is
called when the task stops being schedulable on sched_ext.
- scx_bpf_dispatch() can now be called from ops.select_cpu() too. This
removes the need for communicating local dispatch decision made by
ops.select_cpu() to ops.enqueue() via per-task storage.
SCX_KF_SELECT_CPU is added to support the change.
- SCX_TASK_ENQ_LOCAL which told the BPF scheudler that
scx_select_cpu_dfl() wants the task to be dispatched to the local DSQ
was removed. Instead, scx_bpf_select_cpu_dfl() now dispatches directly
if it finds a suitable idle CPU. If such behavior is not desired,
users can use scx_bpf_select_cpu_dfl() which returns the verdict in a
bool out param.
- scx_select_cpu_dfl() was mishandling WAKE_SYNC and could end up
queueing many tasks on a local DSQ which makes tasks to execute in
order while other CPUs stay idle which made some hackbench numbers
really bad. Fixed.
- The current state of sched_ext can now be monitored through files
under /sys/sched_ext instead of /sys/kernel/debug/sched/ext. This is
to enable monitoring on kernels which don't enable debugfs.
- sched_ext wasn't telling BPF that ops.dispatch()'s @prev argument may
be NULL and a BPF scheduler which derefs the pointer without checking
could crash the kernel. Tell BPF. This is currently a bit ugly. A
better way to annotate this is expected in the future.
- scx_exit_info updated to carry pointers to message buffers instead of
embedding them directly. This decouples buffer sizes from API so that
they can be changed without breaking compatibility.
- exit_code added to scx_exit_info. This is used to indicate different
exit conditions on non-error exits and will be used to handle e.g. CPU
hotplugs.
- The patch "sched_ext: Allow BPF schedulers to switch all eligible
tasks into sched_ext" is folded in and the interface is changed so
that partial switching is indicated with a new ops flag
%SCX_OPS_SWITCH_PARTIAL. This makes scx_bpf_switch_all() unnecessasry
and in turn SCX_KF_INIT. ops.init() is now called with
SCX_KF_SLEEPABLE.
- Code reorganized so that only the parts necessary to integrate with
the rest of the kernel are in the header files.
- Changes to reflect the BPF and other kernel changes including the
addition of bpf_sched_ext_ops.cfi_stubs.
v5: - To accommodate 32bit configs, p->scx.ops_state is now atomic_long_t
instead of atomic64_t and scx_dsp_buf_ent.qseq which uses
load_acquire/store_release is now unsigned long instead of u64.
- Fix the bug where bpf_scx_btf_struct_access() was allowing write
access to arbitrary fields.
- Distinguish kfuncs which can be called from any sched_ext ops and from
anywhere. e.g. scx_bpf_pick_idle_cpu() can now be called only from
sched_ext ops.
- Rename "type" to "kind" in scx_exit_info to make it easier to use on
languages in which "type" is a reserved keyword.
- Since cff9b2332ab7 ("kernel/sched: Modify initial boot task idle
setup"), PF_IDLE is not set on idle tasks which haven't been online
yet which made scx_task_iter_next_filtered() include those idle tasks
in iterations leading to oopses. Update scx_task_iter_next_filtered()
to directly test p->sched_class against idle_sched_class instead of
using is_idle_task() which tests PF_IDLE.
- Other updates to match upstream changes such as adding const to
set_cpumask() param and renaming check_preempt_curr() to
wakeup_preempt().
v4: - SCHED_CHANGE_BLOCK replaced with the previous
sched_deq_and_put_task()/sched_enq_and_set_tsak() pair. This is
because upstream is adaopting a different generic cleanup mechanism.
Once that lands, the code will be adapted accordingly.
- task_on_scx() used to test whether a task should be switched into SCX,
which is confusing. Renamed to task_should_scx(). task_on_scx() now
tests whether a task is currently on SCX.
- scx_has_idle_cpus is barely used anymore and replaced with direct
check on the idle cpumask.
- SCX_PICK_IDLE_CORE added and scx_pick_idle_cpu() improved to prefer
fully idle cores.
- ops.enable() now sees up-to-date p->scx.weight value.
- ttwu_queue path is disabled for tasks on SCX to avoid confusing BPF
schedulers expecting ->select_cpu() call.
- Use cpu_smt_mask() instead of topology_sibling_cpumask() like the rest
of the scheduler.
v3: - ops.set_weight() added to allow BPF schedulers to track weight changes
without polling p->scx.weight.
- move_task_to_local_dsq() was losing SCX-specific enq_flags when
enqueueing the task on the target dsq because it goes through
activate_task() which loses the upper 32bit of the flags. Carry the
flags through rq->scx.extra_enq_flags.
- scx_bpf_dispatch(), scx_bpf_pick_idle_cpu(), scx_bpf_task_running()
and scx_bpf_task_cpu() now use the new KF_RCU instead of
KF_TRUSTED_ARGS to make it easier for BPF schedulers to call them.
- The kfunc helper access control mechanism implemented through
sched_ext_entity.kf_mask is improved. Now SCX_CALL_OP*() is always
used when invoking scx_ops operations.
v2: - balance_scx_on_up() is dropped. Instead, on UP, balance_scx() is
called from put_prev_taks_scx() and pick_next_task_scx() as necessary.
To determine whether balance_scx() should be called from
put_prev_task_scx(), SCX_TASK_DEQD_FOR_SLEEP flag is added. See the
comment in put_prev_task_scx() for details.
- sched_deq_and_put_task() / sched_enq_and_set_task() sequences replaced
with SCHED_CHANGE_BLOCK().
- Unused all_dsqs list removed. This was a left-over from previous
iterations.
- p->scx.kf_mask is added to track and enforce which kfunc helpers are
allowed. Also, init/exit sequences are updated to make some kfuncs
always safe to call regardless of the current BPF scheduler state.
Combined, this should make all the kfuncs safe.
- BPF now supports sleepable struct_ops operations. Hacky workaround
removed and operations and kfunc helpers are tagged appropriately.
- BPF now supports bitmask / cpumask helpers. scx_bpf_get_idle_cpumask()
and friends are added so that BPF schedulers can use the idle masks
with the generic helpers. This replaces the hacky kfunc helpers added
by a separate patch in V1.
- CONFIG_SCHED_CLASS_EXT can no longer be enabled if SCHED_CORE is
enabled. This restriction will be removed by a later patch which adds
core-sched support.
- Add MAINTAINERS entries and other misc changes.
Signed-off-by: Tejun Heo <tj@kernel.org> Co-authored-by: David Vernet <dvernet@meta.com> Acked-by: Josh Don <joshdon@google.com> Acked-by: Hao Luo <haoluo@google.com> Acked-by: Barret Rhoden <brho@google.com> Cc: Andrea Righi <andrea.righi@canonical.com>
Tejun Heo [Tue, 18 Jun 2024 20:09:17 +0000 (10:09 -1000)]
sched_ext: Add boilerplate for extensible scheduler class
This adds dummy implementations of sched_ext interfaces which interact with
the scheduler core and hook them in the correct places. As they're all
dummies, this doesn't cause any behavior changes. This is split out to help
reviewing.
v2: balance_scx_on_up() dropped. This will be handled in sched_ext proper.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com> Acked-by: Josh Don <joshdon@google.com> Acked-by: Hao Luo <haoluo@google.com> Acked-by: Barret Rhoden <brho@google.com>
Tejun Heo [Tue, 18 Jun 2024 20:09:17 +0000 (10:09 -1000)]
sched: Add normal_policy()
A new BPF extensible sched_class will need to dynamically change how a task
picks its sched_class. For example, if the loaded BPF scheduler progs fail,
the tasks will be forced back on CFS even if the task's policy is set to the
new sched_class. To support such mapping, add normal_policy() which wraps
testing for %SCHED_NORMAL. This doesn't cause any behavior changes.
v2: Update the description with more details on the expected use.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com> Acked-by: Josh Don <joshdon@google.com> Acked-by: Hao Luo <haoluo@google.com> Acked-by: Barret Rhoden <brho@google.com>
Tejun Heo [Tue, 18 Jun 2024 20:09:17 +0000 (10:09 -1000)]
sched: Factor out update_other_load_avgs() from __update_blocked_others()
RT, DL, thermal and irq load and utilization metrics need to be decayed and
updated periodically and before consumption to keep the numbers reasonable.
This is currently done from __update_blocked_others() as a part of the fair
class load balance path. Let's factor it out to update_other_load_avgs().
Pure refactor. No functional changes.
This will be used by the new BPF extensible scheduling class to ensure that
the above metrics are properly maintained.
v2: Refreshed on top of tip:sched/core.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com>
Tejun Heo [Tue, 18 Jun 2024 20:09:16 +0000 (10:09 -1000)]
sched: Factor out cgroup weight conversion functions
Factor out sched_weight_from/to_cgroup() which convert between scheduler
shares and cgroup weight. No functional change. The factored out functions
will be used by a new BPF extensible sched_class so that the weights can be
exposed to the BPF programs in a way which is consistent cgroup weights and
easier to interpret.
The weight conversions will be used regardless of cgroup usage. It's just
borrowing the cgroup weight range as it's more intuitive.
CGROUP_WEIGHT_MIN/DFL/MAX constants are moved outside CONFIG_CGROUPS so that
the conversion helpers can always be defined.
v2: The helpers are now defined regardless of COFNIG_CGROUPS.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com> Acked-by: Josh Don <joshdon@google.com> Acked-by: Hao Luo <haoluo@google.com> Acked-by: Barret Rhoden <brho@google.com>
Tejun Heo [Tue, 18 Jun 2024 20:09:16 +0000 (10:09 -1000)]
sched: Add sched_class->switching_to() and expose check_class_changing/changed()
When a task switches to a new sched_class, the prev and new classes are
notified through ->switched_from() and ->switched_to(), respectively, after
the switching is done.
A new BPF extensible sched_class will have callbacks that allow the BPF
scheduler to keep track of relevant task states (like priority and cpumask).
Those callbacks aren't called while a task is on a different sched_class.
When a task comes back, we wanna tell the BPF progs the up-to-date state
before the task gets enqueued, so we need a hook which is called before the
switching is committed.
This patch adds ->switching_to() which is called during sched_class switch
through check_class_changing() before the task is restored. Also, this patch
exposes check_class_changing/changed() in kernel/sched/sched.h. They will be
used by the new BPF extensible sched_class to implement implicit sched_class
switching which is used e.g. when falling back to CFS when the BPF scheduler
fails or unloads.
This is a prep patch and doesn't cause any behavior changes. The new
operation and exposed functions aren't used yet.
v3: Refreshed on top of tip:sched/core.
v2: Improve patch description w/ details on planned use.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com> Acked-by: Josh Don <joshdon@google.com> Acked-by: Hao Luo <haoluo@google.com> Acked-by: Barret Rhoden <brho@google.com>
Tejun Heo [Tue, 18 Jun 2024 20:09:16 +0000 (10:09 -1000)]
sched: Add sched_class->reweight_task()
Currently, during a task weight change, sched core directly calls
reweight_task() defined in fair.c if @p is on CFS. Let's make it a proper
sched_class operation instead. CFS's reweight_task() is renamed to
reweight_task_fair() and now called through sched_class.
While it turns a direct call into an indirect one, set_load_weight() isn't
called from a hot path and this change shouldn't cause any noticeable
difference. This will be used to implement reweight_task for a new BPF
extensible sched_class so that it can keep its cached task weight
up-to-date.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com> Acked-by: Josh Don <joshdon@google.com> Acked-by: Hao Luo <haoluo@google.com> Acked-by: Barret Rhoden <brho@google.com>
Tejun Heo [Tue, 18 Jun 2024 20:09:16 +0000 (10:09 -1000)]
sched: Allow sched_cgroup_fork() to fail and introduce sched_cancel_fork()
A new BPF extensible sched_class will need more control over the forking
process. It wants to be able to fail from sched_cgroup_fork() after the new
task's sched_task_group is initialized so that the loaded BPF program can
prepare the task with its cgroup association is established and reject fork
if e.g. allocation fails.
Allow sched_cgroup_fork() to fail by making it return int instead of void
and adding sched_cancel_fork() to undo sched_fork() in the error path.
sched_cgroup_fork() doesn't fail yet and this patch shouldn't cause any
behavior changes.
v2: Patch description updated to detail the expected use.
Signed-off-by: Tejun Heo <tj@kernel.org> Reviewed-by: David Vernet <dvernet@meta.com> Acked-by: Josh Don <joshdon@google.com> Acked-by: Hao Luo <haoluo@google.com> Acked-by: Barret Rhoden <brho@google.com>
Tejun Heo [Tue, 18 Jun 2024 20:09:16 +0000 (10:09 -1000)]
sched: Restructure sched_class order sanity checks in sched_init()
Currently, sched_init() checks that the sched_class'es are in the expected
order by testing each adjacency which is a bit brittle and makes it
cumbersome to add optional sched_class'es. Instead, let's verify whether
they're in the expected order using sched_class_above() which is what
matters.
Signed-off-by: Tejun Heo <tj@kernel.org> Suggested-by: Peter Zijlstra <peterz@infradead.org> Reviewed-by: David Vernet <dvernet@meta.com>
Andrii Nakryiko [Mon, 17 Jun 2024 21:38:32 +0000 (14:38 -0700)]
Merge branch 'bpf-support-resilient-split-btf'
Alan Maguire says:
====================
bpf: support resilient split BTF
Split BPF Type Format (BTF) provides huge advantages in that kernel
modules only have to provide type information for types that they do not
share with the core kernel; for core kernel types, split BTF refers to
core kernel BTF type ids. So for a STRUCT sk_buff, a module that
uses that structure (or a pointer to it) simply needs to refer to the
core kernel type id, saving the need to define the structure and its many
dependents. This cuts down on duplication and makes BTF as compact
as possible.
However, there is a downside. This scheme requires the references from
split BTF to base BTF to be valid not just at encoding time, but at use
time (when the module is loaded). Even a small change in kernel types
can perturb the type ids in core kernel BTF, and - if the new reproducible
BTF option is not used - pahole's parallel processing of compilation units
can lead to different type ids for the same kernel if the BTF is
regenerated.
So we have a robustness problem for split BTF for cases where a module is
not always compiled at the same time as the kernel. This problem is
particularly acute for distros which generally want module builders to be
able to compile a module for the lifetime of a Linux stable-based release,
and have it continue to be valid over the lifetime of that release, even
as changes in data structures (and hence BTF types) accrue. Today it's not
possible to generate BTF for modules that works beyond the initial
kernel it is compiled against - kernel bugfixes etc invalidate the split
BTF references to vmlinux BTF, and BTF is no longer usable for the
module.
The goal of this series is to provide options to provide additional
context for cases like this. That context comes in the form of
distilled base BTF; it stands in for the base BTF, and contains
information about the types referenced from split BTF, but not their
full descriptions. The modified split BTF will refer to type ids in
this .BTF.base section, and when the kernel loads such modules it
will use that .BTF.base to map references from split BTF to the
equivalent current vmlinux base BTF types. Once this relocation
process has succeeded, the module BTF available in /sys/kernel/btf
will look exactly as if it was built with the current vmlinux;
references to base types will be fixed up etc.
A module builder - using this series along with the pahole changes -
can then build a module with distilled base BTF via an out-of-tree
module build, i.e.
make -C . M=path/2/module
The module will have a .BTF section (the split BTF) and a
.BTF.base section. The latter is small in size - distilled base
BTF does not need full struct/union/enum information for named
types for example. For 2667 modules built with distilled base BTF,
the average size observed was 1556 bytes (stddev 1563). The overall
size added to this 2667 modules was 5.3Mb.
Note that for the in-tree modules, this approach is not needed as
split and base BTF in the case of in-tree modules are always built
and re-built together.
The series first focuses on generating split BTF with distilled base
BTF; then relocation support is added to allow split BTF with
an associated distlled base to be relocated with a new base BTF.
Next Eduard's patch allows BTF ELF parsing to work with both
.BTF and .BTF.base sections; this ensures that bpftool will be
able to dump BTF for a module with a .BTF.base section for example,
or indeed dump relocated BTF where a module and a "-B vmlinux"
is supplied.
Then we add support to resolve_btfids to ignore base BTF - i.e.
to avoid relocation - if a .BTF.base section is found. This ensures
the .BTF.ids section is populated with ids relative to the distilled
base (these will be relocated as part of module load).
Finally the series supports storage of .BTF.base data/size in modules
and supports sharing of relocation code with the kernel to allow
relocation of module BTF. For the kernel, this relocation
process happens at module load time, and we relocate split BTF
references to point at types in the current vmlinux BTF. As part of
this, .BTF.ids references need to be mapped also.
So concretely, what happens is
- we generate split BTF in the .BTF section of a module that refers to
types in the .BTF.base section as base types; the latter are not full
type descriptions but provide information about the base type. So
a STRUCT sk_buff would be represented as a FWD struct sk_buff in
distilled base BTF for example.
- when the module is loaded, the split BTF is relocated with vmlinux
BTF; in the case of the FWD struct sk_buff, we find the STRUCT sk_buff
in vmlinux BTF and map all split BTF references to the distilled base
FWD sk_buff, replacing them with references to the vmlinux BTF
STRUCT sk_buff.
A previous approach to this problem [1] utilized standalone BTF for such
cases - where the BTF is not defined relative to base BTF so there is no
relocation required. The problem with that approach is that from
the verifier perspective, some types are special, and having a custom
representation of a core kernel type that did not necessarily match the
current representation is not tenable. So the approach taken here was
to preserve the split BTF model while minimizing the representation of
the context needed to relocate split and current vmlinux BTF.
To generate distilled .BTF.base sections the associated dwarves
patch (to be applied on the "next" branch there) is needed [3]
Without it, things will still work but modules will not be built
with a .BTF.base section.
Changes since v5[4]:
- Update search of distilled types to return the first occurrence
of a string (or a string+size pair); this allows us to iterate
over all matches in distilled base BTF (Andrii, patch 3)
- Update to use BTF field iterators (Andrii, patches 1, 3 and 8)
- Update tests to cover multiple match and associated error cases
(Eduard, patch 4)
- Rename elf_sections_info to btf_elf_secs, remove use of
libbpf_get_error(), reset btf->owns_base when relocation
succeeds (Andrii, patch 5)
Changes since v4[5]:
- Moved embeddedness, duplicate name checks to relocation time
and record struct/union size for all distilled struct/unions
instead of using forwards. This allows us to carry out
type compatibility checks based on the base BTF we want to
relocate with (Eduard, patches 1, 3)
- Moved to using qsort() instead of qsort_r() as support for
qsort_r() appears to be missing in Android libc (Andrii, patch 3)
- Sorting/searching now incorporates size matching depending
on BTF kind and embeddedness of struct/union (Eduard, Andrii,
patch 3)
- Improved naming of various types during relocation to avoid
confusion (Andrii, patch 3)
- Incorporated Eduard's patch (patch 5) which handles .BTF.base
sections internally in btf_parse_elf(). This makes ELF parsing
work with split BTF, split BTF with a distilled base, split
BTF with a distilled base _and_ base BTF (by relocating) etc.
Having this avoids the need for bpftool changes; it will work
as-is with .BTF.base sections (Eduard, patch 4)
- Updated resolve_btfids to _not_ relocate BTF for modules
where a .BTF.base section is present; in that one case we
do not want to relocate BTF as the .BTF.ids section should
reflect ids in .BTF.base which will later be relocated on
module load (Eduard, Andrii, patch 5)
Changes since v3[6]:
- distill now checks for duplicate-named struct/unions and records
them as a sized struct/union to help identify which of the
multiple base BTF structs/unions it refers to (Eduard, patch 1)
- added test support for multiple name handling (Eduard, patch 2)
- simplified the string mapping when updating split BTF to use
base BTF instead of distilled base. Since the only string
references split BTF can make to base BTF are the names of
the base types, create a string map from distilled string
offset -> base BTF string offset and update string offsets
by visiting all strings in split BTF; this saves having to
do costly searches of base BTF (Eduard, patch 7,10)
- fixed bpftool manpage and indentation issues (Quentin, patch 11)
Also explored Eduard's suggestion of doing an implicit fallback
to checking for .BTF.base section in btf__parse() when it is
called to get base BTF. However while it is doable, it turned
out to be difficult operationally. Since fallback is implicit
we do not know the source of the BTF - was it from .BTF or
.BTF.base? In bpftool, we want to try first standalone BTF,
then split, then split with distilled base. Having a way
to explicitly request .BTF.base via btf__parse_opts() fits
that model better.
Changes since v2[7]:
- submitted patch to use --btf_features in Makefile.btf for pahole
v1.26 and later separately (Andrii). That has landed in bpf-next
now.
- distilled base now encodes ENUM64 as fwd ENUM (size 8), eliminating
the need for support for ENUM64 in btf__add_fwd (patch 1, Andrii)
- moved to distilling only named types, augmenting split BTF with
associated reference types; this simplifies greatly the distilled
base BTF and the mapping operation between distilled and base
BTF when relocating (most of the series changes, Andrii)
- relocation now iterates over base BTF, looking for matches based
on name in distilled BTF. Distilled BTF is pre-sorted by name
(Andrii, patch 8)
- removed most redundant compabitiliby checks aside from struct
size for base types/embedded structs and kind compatibility
(since we only match on name) (Andrii, patch 8)
- btf__parse_opts() now replaces btf_parse() internally in libbpf
(Eduard, patch 3)
Changes since RFC [8]:
- updated terminology; we replace clunky "base reference" BTF with
distilling base BTF into a .BTF.base section. Similarly BTF
reconcilation becomes BTF relocation (Andrii, most patches)
- add distilled base BTF by default for out-of-tree modules
(Alexei, patch 8)
- distill algorithm updated to record size of embedded struct/union
by recording it as a 0-vlen STRUCT/UNION with size preserved
(Andrii, patch 2)
- verify size match on relocation for such STRUCT/UNIONs (Andrii,
patch 9)
- with embedded STRUCT/UNION recording size, we can have bpftool
dump a header representation using .BTF.base + .BTF sections
rather than special-casing and refusing to use "format c" for
that case (patch 5)
- match enum with enum64 and vice versa (Andrii, patch 9)
- ensure that resolve_btfids works with BTF without .BTF.base
section (patch 7)
- update tests to cover embedded types, arrays and function
prototypes (patches 3, 12)
Alan Maguire [Thu, 13 Jun 2024 09:50:11 +0000 (10:50 +0100)]
resolve_btfids: Handle presence of .BTF.base section
Now that btf_parse_elf() handles .BTF.base section presence,
we need to ensure that resolve_btfids uses .BTF.base when present
rather than the vmlinux base BTF passed in via the -B option.
Detect .BTF.base section presence and unset the base BTF path
to ensure that BTF ELF parsing will do the right thing.
Eduard Zingerman [Thu, 13 Jun 2024 09:50:10 +0000 (10:50 +0100)]
libbpf: Make btf_parse_elf process .BTF.base transparently
Update btf_parse_elf() to check if .BTF.base section is present.
The logic is as follows:
if .BTF.base section exists:
distilled_base := btf_new(.BTF.base)
if distilled_base:
btf := btf_new(.BTF, .base_btf=distilled_base)
if base_btf:
btf_relocate(btf, base_btf)
else:
btf := btf_new(.BTF)
return btf
In other words:
- if .BTF.base section exists, load BTF from it and use it as a base
for .BTF load;
- if base_btf is specified and .BTF.base section exist, relocate newly
loaded .BTF against base_btf.
Alan Maguire [Thu, 13 Jun 2024 09:50:09 +0000 (10:50 +0100)]
selftests/bpf: Extend distilled BTF tests to cover BTF relocation
Ensure relocated BTF looks as expected; in this case identical to
original split BTF, with a few duplicate anonymous types added to
split BTF by the relocation process. Also add relocation tests
for edge cases like missing type in base BTF and multiple types
of the same name.
Alan Maguire [Thu, 13 Jun 2024 09:50:08 +0000 (10:50 +0100)]
libbpf: Split BTF relocation
Map distilled base BTF type ids referenced in split BTF and their
references to the base BTF passed in, and if the mapping succeeds,
reparent the split BTF to the base BTF.
Relocation is done by first verifying that distilled base BTF
only consists of named INT, FLOAT, ENUM, FWD, STRUCT and
UNION kinds; then we sort these to speed lookups. Once sorted,
the base BTF is iterated, and for each relevant kind we check
for an equivalent in distilled base BTF. When found, the
mapping from distilled -> base BTF id and string offset is recorded.
In establishing mappings, we need to ensure we check STRUCT/UNION
size when the STRUCT/UNION is embedded in a split BTF STRUCT/UNION,
and when duplicate names exist for the same STRUCT/UNION. Otherwise
size is ignored in matching STRUCT/UNIONs.
Once all mappings are established, we can update type ids
and string offsets in split BTF and reparent it to the new base.
Alan Maguire [Thu, 13 Jun 2024 09:50:07 +0000 (10:50 +0100)]
selftests/bpf: Test distilled base, split BTF generation
Test generation of split+distilled base BTF, ensuring that
- named base BTF STRUCTs and UNIONs are represented as 0-vlen sized
STRUCT/UNIONs
- named ENUM[64]s are represented as 0-vlen named ENUM[64]s
- anonymous struct/unions are represented in full in split BTF
- anonymous enums are represented in full in split BTF
- types unreferenced from split BTF are not present in distilled
base BTF
Also test that with vmlinux BTF and split BTF based upon it,
we only represent needed base types referenced from split BTF
in distilled base.
Alan Maguire [Thu, 13 Jun 2024 09:50:06 +0000 (10:50 +0100)]
libbpf: Add btf__distill_base() creating split BTF with distilled base BTF
To support more robust split BTF, adding supplemental context for the
base BTF type ids that split BTF refers to is required. Without such
references, a simple shuffling of base BTF type ids (without any other
significant change) invalidates the split BTF. Here the attempt is made
to store additional context to make split BTF more robust.
This context comes in the form of distilled base BTF providing minimal
information (name and - in some cases - size) for base INTs, FLOATs,
STRUCTs, UNIONs, ENUMs and ENUM64s along with modified split BTF that
points at that base and contains any additional types needed (such as
TYPEDEF, PTR and anonymous STRUCT/UNION declarations). This
information constitutes the minimal BTF representation needed to
disambiguate or remove split BTF references to base BTF. The rules
are as follows:
- INT, FLOAT, FWD are recorded in full.
- if a named base BTF STRUCT or UNION is referred to from split BTF, it
will be encoded as a zero-member sized STRUCT/UNION (preserving
size for later relocation checks). Only base BTF STRUCT/UNIONs
that are either embedded in split BTF STRUCT/UNIONs or that have
multiple STRUCT/UNION instances of the same name will _need_ size
checks at relocation time, but as it is possible a different set of
types will be duplicates in the later to-be-resolved base BTF,
we preserve size information for all named STRUCT/UNIONs.
- if an ENUM[64] is named, a ENUM forward representation (an ENUM
with no values) of the same size is used.
- in all other cases, the type is added to the new split BTF.
Avoiding struct/union/enum/enum64 expansion is important to keep the
distilled base BTF representation to a minimum size.
When successful, new representations of the distilled base BTF and new
split BTF that refers to it are returned. Both need to be freed by the
caller.
So to take a simple example, with split BTF with a type referring
to "struct sk_buff", we will generate distilled base BTF with a
0-member STRUCT sk_buff of the appropriate size, and the split BTF
will refer to it instead.
Tools like pahole can utilize such split BTF to populate the .BTF
section (split BTF) and an additional .BTF.base section. Then
when the split BTF is loaded, the distilled base BTF can be used
to relocate split BTF to reference the current (and possibly changed)
base BTF.
So for example if "struct sk_buff" was id 502 when the split BTF was
originally generated, we can use the distilled base BTF to see that
id 502 refers to a "struct sk_buff" and replace instances of id 502
with the current (relocated) base BTF sk_buff type id.
Distilled base BTF is small; when building a kernel with all modules
using distilled base BTF as a test, overall module size grew by only
5.3Mb total across ~2700 modules.
Compilers can generate the code
r1 = r2
r1 += 0x1
if r2 < 1000 goto ...
use knowledge of r2 range in subsequent r1 operations
So remember constant delta between r2 and r1 and update r1 after 'if' condition.
Unfortunately LLVM still uses this pattern for loops with 'can_loop' construct:
for (i = 0; i < 1000 && can_loop; i++)
The "undo" pass was introduced in LLVM
https://reviews.llvm.org/D121937
to prevent this optimization, but it cannot cover all cases.
Instead of fighting middle end optimizer in BPF backend teach the verifier
about this pattern.
__bpf_skc_lookup() safely handles incorrect values of tuple len,
hence we can allow zero to be passed as tuple len.
This patch alone doesn't make an observable verifier difference.
It's a trivial improvement that might simplify bpf programs.
====================
bpf: make trusted args nullable
Current verifier checks for the arg to be nullable after checking for
certain pointer types. It prevents programs to pass NULL to kfunc args
even if they are marked as nullable. This patchset adjusts verifier and
changes bpf crypto kfuncs to allow null for IV parameter which is
optional for some ciphers. Benchmark shows ~4% improvements when there
is no need to initialise 0-sized dynptr.
v3:
- add special selftest for nullable parameters
v2:
- adjust kdoc accordingly
====================
Vadim Fedorenko [Thu, 13 Jun 2024 21:18:17 +0000 (14:18 -0700)]
selftests: bpf: add testmod kfunc for nullable params
Add special test to be sure that only __nullable BTF params can be
replaced by NULL. This patch adds fake kfuncs in bpf_testmod to
properly test different params.
Vadim Fedorenko [Thu, 13 Jun 2024 21:18:14 +0000 (14:18 -0700)]
bpf: crypto: make state and IV dynptr nullable
Some ciphers do not require state and IV buffer, but with current
implementation 0-sized dynptr is always needed. With adjustment to
verifier we can provide NULL instead of 0-sized dynptr. Make crypto
kfuncs ready for this.
Vadim Fedorenko [Thu, 13 Jun 2024 21:18:13 +0000 (14:18 -0700)]
bpf: verifier: make kfuncs args nullalble
Some arguments to kfuncs might be NULL in some cases. But currently it's
not possible to pass NULL to any BTF structures because the check for
the suffix is located after all type checks. Move it to earlier place
to allow nullable args.
====================
bpf: Support dumping kfunc prototypes from BTF
This patchset enables both detecting as well as dumping compilable
prototypes for kfuncs.
The first commit instructs pahole to DECL_TAG kfuncs when available.
This requires v1.27 which was released on 6/11/24. With it, users will
be able to look at BTF inside vmlinux (or modules) and check if the
kfunc they want is available.
The final commit teaches bpftool how to dump kfunc prototypes. This
is done for developer convenience.
The rest of the commits are fixups to enable selftests to use the
newly dumped kfunc prototypes. With these, selftests will regularly
exercise the newly added codepaths.
Tested with and without the required pahole changes:
=== Changelog ===
From v4:
* Change bpf_session_cookie() return type
* Only fixup used fentry test kfunc prototypes
* Extract out projection detection into shared btf_is_projection_of()
* Fix kernel test robot build warnings about doc comments
From v3:
* Teach selftests to use dumped prototypes
From v2:
* Update Makefile.btf with pahole flag
* More error checking
* Output formatting changes
* Drop already-merged commit
From v1:
* Add __weak annotation
* Use btf_dump for kfunc prototypes
* Update kernel bpf_rdonly_cast() signature
====================
Daniel Xu [Wed, 12 Jun 2024 15:58:36 +0000 (09:58 -0600)]
bpftool: Support dumping kfunc prototypes from BTF
This patch enables dumping kfunc prototypes from bpftool. This is useful
b/c with this patch, end users will no longer have to manually define
kfunc prototypes. For the kernel tree, this also means we can optionally
drop kfunc prototypes from:
Daniel Xu [Wed, 12 Jun 2024 15:58:35 +0000 (09:58 -0600)]
bpf: selftests: xfrm: Opt out of using generated kfunc prototypes
The xfrm_info selftest locally defines an aliased type such that folks
with CONFIG_XFRM_INTERFACE=m/n configs can still build the selftests.
See commit aa67961f3243 ("selftests/bpf: Allow building bpf tests with CONFIG_XFRM_INTERFACE=[m|n]").
Thus, it is simpler if this selftest opts out of using enerated kfunc
prototypes. The preprocessor macro this commit uses will be introduced
in the final commit.
Daniel Xu [Wed, 12 Jun 2024 15:58:34 +0000 (09:58 -0600)]
bpf: selftests: nf: Opt out of using generated kfunc prototypes
The bpf-nf selftests play various games with aliased types such that
folks with CONFIG_NF_CONNTRACK=m/n configs can still build the
selftests. See commits:
1058b6a78db2 ("selftests/bpf: Do not fail build if CONFIG_NF_CONNTRACK=m/n") 92afc5329a5b ("selftests/bpf: Fix build errors if CONFIG_NF_CONNTRACK=m")
Thus, it is simpler if these selftests opt out of using generated kfunc
prototypes. The preprocessor macro this commit uses will be introduced
in the final commit.
Daniel Xu [Wed, 12 Jun 2024 15:58:33 +0000 (09:58 -0600)]
bpf: treewide: Align kfunc signatures to prog point-of-view
Previously, kfunc declarations in bpf_kfuncs.h (and others) used "user
facing" types for kfuncs prototypes while the actual kfunc definitions
used "kernel facing" types. More specifically: bpf_dynptr vs
bpf_dynptr_kern, __sk_buff vs sk_buff, and xdp_md vs xdp_buff.
It wasn't an issue before, as the verifier allows aliased types.
However, since we are now generating kfunc prototypes in vmlinux.h (in
addition to keeping bpf_kfuncs.h around), this conflict creates
compilation errors.
Fix this conflict by using "user facing" types in kfunc definitions.
This results in more casts, but otherwise has no additional runtime
cost.
Note, similar to 5b268d1ebcdc ("bpf: Have bpf_rdonly_cast() take a const
pointer"), we also make kfuncs take const arguments where appropriate in
order to make the kfunc more permissive.
Daniel Xu [Wed, 12 Jun 2024 15:58:32 +0000 (09:58 -0600)]
bpf: verifier: Relax caller requirements for kfunc projection type args
Currently, if a kfunc accepts a projection type as an argument (eg
struct __sk_buff *), the caller must exactly provide exactly the same
type with provable provenance.
However in practice, kfuncs that accept projection types _must_ cast to
the underlying type before use b/c projection type layouts are
completely made up. Thus, it is ok to relax the verifier rules around
implicit conversions.
We will use this functionality in the next commit when we align kfuncs
to user-facing types.
Daniel Xu [Wed, 12 Jun 2024 15:58:30 +0000 (09:58 -0600)]
bpf: Make bpf_session_cookie() kfunc return long *
We will soon be generating kfunc prototypes from BTF. As part of that,
we need to align the manual signatures in bpf_kfuncs.h with the actual
kfunc definitions. There is currently a conflicting signature for
bpf_session_cookie() w.r.t. return type.
The original intent was to return long * and not __u64 *. You can see
evidence of that intent in a3a5113393cc ("selftests/bpf: Add kprobe
session cookie test").
The prototype in progs/map_percpu_stats.c is not in line with how the
actual kfuncs are defined in kernel/bpf/map_iter.c. This causes
compilation errors when kfunc prototypes are generated from BTF.
The prototype in progs/nested_trust_common.h is not in line with how the
actual kfuncs are defined in kernel/bpf/cpumask.c. This causes compilation
errors when kfunc prototypes are generated from BTF.
Daniel Xu [Wed, 12 Jun 2024 15:58:27 +0000 (09:58 -0600)]
bpf: selftests: Fix fentry test kfunc prototypes
Some prototypes in progs/get_func_ip_test.c were not in line with how the
actual kfuncs are defined in net/bpf/test_run.c. This causes compilation
errors when kfunc prototypes are generated from BTF.
bpf_iter_task_vma_new() is defined as taking a u64 as its 3rd argument.
u64 is a unsigned long long. bpf_experimental.h was defining the
prototype as unsigned long.
Daniel Xu [Wed, 12 Jun 2024 15:58:25 +0000 (09:58 -0600)]
kbuild: bpf: Tell pahole to DECL_TAG kfuncs
With [0], pahole can now discover kfuncs and inject DECL_TAG
into BTF. With this commit, we will start shipping said DECL_TAGs
to downstream consumers if pahole supports it.
This is useful for feature probing kfuncs as well as generating
compilable prototypes. This is particularly important as kfuncs
do not have stable ABI.
Kenta Tada [Fri, 7 Jun 2024 11:17:04 +0000 (20:17 +0900)]
bpftool: Query only cgroup-related attach types
When CONFIG_NETKIT=y,
bpftool-cgroup shows error even if the cgroup's path is correct:
$ bpftool cgroup tree /sys/fs/cgroup
CgroupPath
ID AttachType AttachFlags Name
Error: can't query bpf programs attached to /sys/fs/cgroup: No such device or address
>From strace and kernel tracing, I found netkit returned ENXIO and this command failed.
I think this AttachType(BPF_NETKIT_PRIMARY) is not relevant to cgroup.
bpftool-cgroup should query just only cgroup-related attach types.
====================
Intel Wired LAN Driver Updates 2024-06-03
This series includes miscellaneous improvements for the ice as well as a
cleanup to the Makefiles for all Intel net drivers.
Andy fixes all of the Intel net driver Makefiles to use the documented
'*-y' syntax for specifying object files to link into kernel driver
modules, rather than the '*-objs' syntax which works but is documented as
reserved for user-space host programs.
Jacob has a cleanup to refactor rounding logic in the ice driver into a
common roundup_u64 helper function.
Michal Schmidt replaces irq_set_affinity_hint() to use
irq_update_affinity_hint() which behaves better with user-applied affinity
settings.
projects / thirdparty / kernel / linux.git / log
