Subsequent commits will allow update_curr() to throttle the hierarchy
when the runtime accounting exceeds allocated quota. Call update_curr()
before the unthrottle event, and in tg_unthrottle_up() to catch up on
any remaining runtime and stabilize the "runtime_remaining" and
"throttle_count" for that cfs_rq.
Doing an update_curr() early ensures the cfs_rq is not throttled right
back up again when the unthrottle is in progress.
Since all callers of unthrottle_cfs_rq(), except two, already update the
rq_clock and call rq_clock_start_loop_update(), move the
update_rq_clock() from unthrottle_cfs_rq() to the callers that don't
update the rq_clock.
Signed-off-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Benjamin Segall <bsegall@google.com>
Tested-by: Aaron Lu <ziqianlu@bytedance.com>
Link: https://patch.msgid.link/20260602052531.11450-1-kprateek.nayak@amd.com
When distribute_cfs_runtime() encounters a local cfs_rq, it adds it to a
local list and unthrottles it at the end, when it is done unthrottling
other cfs_rq(s) on cfs_b->throttled_cfs_rq until the bandwidth runs out.
Instead of using a local list, reuse the local CPU's
rq->throttled_csd_list and the __cfsb_csd_unthrottle() path for
unthrottle.
If this is the first cfs_rq to be queued on the "throttled_csd_list", it
prevents the need for a remote CPUs to interrupt this local CPU if they
themselves are performing async unthrottle.
If this is not the first cfs_rq on the list, there is an async unthrottle
operation pending on this local CPU and the unthrottle can be batched
together.
No functional changes intended.
Signed-off-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Benjamin Segall <bsegall@google.com>
Tested-by: Aaron Lu <ziqianlu@bytedance.com>
Link: https://patch.msgid.link/20260602050005.11160-3-kprateek.nayak@amd.com
Routine conversion of rcu_read_lock(), spin_lock*, and rq_lock usage
within the cfs bandwidth controller to use class guards.
Only notable changes are:
- Checking for "cfs_rq->runtime_remaining <= 0" instead of the inverse
to spot a throttle and break early. This also saves the need
for extra indentation in the unthrottle case.
- Reordering of list_del_rcu() against throttled_clock indicator update
in unthrottle_cfs_rq(). Both are done with "cfs_b->lock" held after
the "cfs_rq->throttled" is cleared which make the reordering safe
against concurrent list modifications.
No functional changes intended.
Signed-off-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Ben Segall <bsegall@google.com>
Tested-by: Aaron Lu <ziqianlu@bytedance.com>
Link: https://patch.msgid.link/20260602050005.11160-2-kprateek.nayak@amd.com
To remove the cfs_rq pointer array in task_group, allocate the combined
cfs_rq and sched_entity using the per-cpu allocator.
This patch implements the following:
- Changes task_group->cfs_rq from 'struct cfs_rq **' to
'struct cfs_rq __percpu *'.
- Updates memory allocation in alloc_fair_sched_group() and
free_fair_sched_group() to use alloc_percpu() and free_percpu()
respectively.
- Uses the inline accessor tg_cfs_rq(tg, cpu) with per_cpu_ptr() to retrieve
the pointer to cfs_rq for the given task group and CPU.
- Replaces direct accesses tg->cfs_rq[cpu] with calls to the new tg_cfs_rq(tg,
cpu) helper.
- Handles the root_task_group: since struct rq is already a per-cpu variable
(runqueues), its embedded cfs_rq (rq->cfs) is also per-cpu. Therefore, we
assign root_task_group.cfs_rq = &runqueues.cfs.
- Cleanup the code in initializing the root task group.
This change places each CPU's cfs_rq and sched_entity in its local per-cpu
memory area to remove the per-task_group pointer arrays.
Signed-off-by: Zecheng Li <zecheng@google.com>
Signed-off-by: Zecheng Li <zli94@ncsu.edu>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: K Prateek Nayak <kprateek.nayak@amd.com>
Reviewed-by: Josh Don <joshdon@google.com>
Link: https://patch.msgid.link/20260522141623.600235-4-zli94@ncsu.edu
Now that struct sched_entity is co-located with struct cfs_rq for non-root task
groups, the task_group->se pointer array is redundant. The associated
sched_entity can be loaded directly from the cfs_rq.
This patch performs the access conversion with the helpers:
- is_root_task_group(tg): checks if a task group is the root task group. It
compares the task group's address with the global root_task_group variable.
- tg_se(tg, cpu): retrieves the cfs_rq and returns the address of the
co-located se. This function checks if tg is the root task group to ensure
behaving the same of previous tg->se[cpu]. Replaces all accesses that use
the tg->se[cpu] pointer array with calls to the new tg_se(tg, cpu) accessor.
- cfs_rq_se(cfs_rq): simplifies access paths like cfs_rq->tg->se[...] to use
the co-located sched_entity. This function also checks if tg is the root
task group to ensure same behavior.
Since tg_se is not in very hot code paths, and the branch is a register
comparison with an immediate value (`&root_task_group`), the performance impact
is expected to be negligible.
Signed-off-by: Zecheng Li <zecheng@google.com>
Signed-off-by: Zecheng Li <zli94@ncsu.edu>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: K Prateek Nayak <kprateek.nayak@amd.com>
Reviewed-by: Josh Don <joshdon@google.com>
Link: https://patch.msgid.link/20260522141623.600235-3-zli94@ncsu.edu
Improve data locality and reduce pointer chasing by allocating struct
cfs_rq and struct sched_entity together for non-root task groups. This
is achieved by introducing a new combined struct cfs_tg_state that
holds both objects in a single allocation.
This patch:
- Introduces struct cfs_tg_state that embeds cfs_rq, sched_entity, and
sched_statistics together in a single structure.
- Updates __schedstats_from_se() in stats.h to use cfs_tg_state for accessing
sched_statistics from a group sched_entity.
- Modifies alloc_fair_sched_group() and free_fair_sched_group() to allocate
and free the new struct as a single unit.
- Modifies the per-CPU pointers in task_group->se and task_group->cfs_rq to
point to the members in the new combined structure.
Signed-off-by: Zecheng Li <zecheng@google.com>
Signed-off-by: Zecheng Li <zli94@ncsu.edu>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: K Prateek Nayak <kprateek.nayak@amd.com>
Reviewed-by: Josh Don <joshdon@google.com>
Link: https://patch.msgid.link/20260522141623.600235-2-zli94@ncsu.edu
Commit ed4fb6d7ef ("hrtimer: Use and report correct timerslack values
for realtime tasks") sets timer_slack_ns to 0 for RT tasks in
__setscheduler_params(). However, when an RT task with SCHED_RESET_ON_FORK
creates child threads, the children inherit timer_slack_ns=0 from the
parent. sched_fork() resets the child's policy to SCHED_NORMAL but does
not restore timer_slack_ns, leaving the child permanently running with
zero slack.
Fix this by restoring timer_slack_ns from default_timer_slack_ns in
sched_fork() when resetting from RT/DL to NORMAL policy, matching the
existing behavior in __setscheduler_params().
Note: this fix alone requires a correct default_timer_slack_ns to be
effective. See the following patch for that fix.
Fixes: ed4fb6d7ef ("hrtimer: Use and report correct timerslack values for realtime tasks")
Reported-by: Qiaoting.Lin <linqiaoting@xiaomi.com>
Signed-off-by: Guanyou.Chen <chenguanyou@xiaomi.com>
Signed-off-by: Chunhui.Li <chunhui.li@mediatek.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260522131000.1664983-2-chenguanyou@xiaomi.com
Now that the proxy path uses ->is_blocked, use the '->is_blocked &&
!->blocked_on' state instead of PROXY_WAKING. Notably, this is where a
blocked_on relation is broken but the donor task might still need a return
migration.
Signed-off-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260526113322.596522894%40infradead.org
Rather than gate the proxy paths with p->blocked_on, use p->is_blocked.
This opens up the state: '->is_blocked && !->blocked_on' for future use.
Notably, only proxy and delayed tasks can be ->on_rq && ->is_blocked, and it is
guaranteed that sched_class::pick_task() will never return a delayed task.
Therefore any task returned from pick_next_task() that has ->is_blocked set,
must be a proxy task.
Suggested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260526113322.477954312%40infradead.org
Current code will 'unconditionally' return migrate on PROXY_WAKING, even if the
task is (still) on the original CPU.
Check task_cpu(p) against p->waking_cpu, which per proxy_set_task_cpu()
preserves the original CPU the task was on. If they do not mis-match, there is
no need to go through the more expensive wakeup path.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260527082916.GP3126523%40noisy.programming.kicks-ass.net
Upon entry to try_to_block_task(), p->is_blocked should be false. After all,
the prior wakeup would have made it so per ttwu_do_wakeup().
Ensure this is the case, rather than clearing it in the path that doesn't set
it.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: John Stultz <jstultz@google.com>
Link: https://patch.msgid.link/20260526113322.364017314%40infradead.org
The reason for the clause in try_to_wake_up() is, per its comment, that
find_proxy_task()'s proxy_deactivate() is not always called with a cleared
p->blocked_on.
However, that seems silly and easily cured. Make sure to always call
proxy_deactivate() with a cleared p->blocked_on such that we might remove this
clause from the common wake-up path.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: John Stultz <jstultz@google.com>
Link: https://patch.msgid.link/20260526113322.244729903%40infradead.org
Add link to the task this task is proxying for, and use it so
the mutex owner can do an intelligent hand-off of the mutex to
the task that the owner is running on behalf.
[jstultz: This patch was split out from larger proxy patch]
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Juri Lelli <juri.lelli@redhat.com>
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Connor O'Brien <connoro@google.com>
Signed-off-by: John Stultz <jstultz@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260512025635.2840817-8-jstultz@google.com
Add a new is_blocked flag to the task struct. This flag is set
by try_to_block_task() and cleared by ttwu_do_wakeup() and
tracks if the task is blocked.
Traditionally this would mirror !p->on_rq, however due things
like DELAY_DEQUEUE and PROXY_EXEC, this can diverge, so its
useful to manage separately.
Additionally with this, we might be able to get rid of the
p->se.sched_delayed (ab)use in the core code (eventually).
Taken whole cloth from Peter's email:
https://lore.kernel.org/lkml/20260501132143.GC1026330@noisy.programming.kicks-ass.net/
With a few additional p->is_blocked = 0 in a few cases where
we return current if blocked_on gets zeroed or there is
no owner. This may hint that these current special cases
might be dropped eventually.
This change also helps resolve wait-queue stalls seen with
proxy-execution. See previous patch attempts for details:
https://lore.kernel.org/lkml/20260430215103.2978955-2-jstultz@google.com/
Reported-by: Vineeth Pillai <vineethrp@google.com>
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: John Stultz <jstultz@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260512025635.2840817-7-jstultz@google.com
This patch adds logic so try_to_wake_up() will notice if we are
waking a task where blocked_on == PROXY_WAKING, and if necessary
dequeue the task so the wakeup will naturally return-migrate the
donor task back to a cpu it can run on.
This helps performance as we do the dequeue and wakeup under the
locks normally taken in the try_to_wake_up() and avoids having
to do proxy_force_return() from __schedule(), which has to
re-take similar locks and then force a pick again loop.
This was split out from the larger proxy patch, and
significantly reworked.
Credits for the original patch go to:
Peter Zijlstra (Intel) <peterz@infradead.org>
Juri Lelli <juri.lelli@redhat.com>
Valentin Schneider <valentin.schneider@arm.com>
Connor O'Brien <connoro@google.com>
Signed-off-by: John Stultz <jstultz@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260512025635.2840817-6-jstultz@google.com
Pull most of the logic out of try_to_block_task() and put it
into block_task() directly, so that we can call block_task() and
not have to worry about the failing cases in try_to_block_task()
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: John Stultz <jstultz@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260512025635.2840817-5-jstultz@google.com
The DL scheduler keeps the current task in the rbtree, since
the deadline value isn't usually chagned while the task is
runnable.
This results in set_next_task() and put_prev_task() being
simpler, but unfortunately this causes complexity elsewhere.
Specifically when update_curr_dl() updates the deadline, it has
to dequeue and then enqueue the task.
From put_prev_task_dl(), we first call update_curr_dl(), and
then call enqueue_pushable_dl_task().
However, with Proxy Exec this goes awry. Since when a mutex is
released, we might wake the waiting rq->donor. This will cause
put_prev_task() to be called on the donor to take it off the
cpu for return migration.
At that point, from put_prev_task_dl() the update_curr_dl()
logic will dequeue & enqueue the task, and the enqueue function
will call enqueue_pushable_dl_task() (since the task_current()
check won't prevent it). Then back up the callstack in
put_prev_task_dl() we'll end up calling
enqueue_pushable_dl_task() again, tripping the
!RB_EMPTY_NODE(&p->pushable_dl_tasks) warning.
So to avoid this, use Peter's suggested[1] approach, and add a
dl_rq->curr pointer that is set/cleared from set_next_task()/
put_prev_task(), which effectively tracks the rq->donor. We can
then use this to avoid adding the active donor to the pushable
list from enqueue_task_dl().
[1]: https://lore.kernel.org/lkml/20260304095123.GP606826@noisy.programming.kicks-ass.net/
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: John Stultz <jstultz@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260512025635.2840817-4-jstultz@google.com
As part of an improvement to handling pushable deadline tasks,
Peter suggested this cleanup[1], to use helper values for
dl_entity and dl_rq in the enqueue_task_dl() and
put_prev_task_dl() functions. There should be no functional
change from this patch.
To make sure this cleanup change doesn't obscure later logic
changes, I've split it into its own patch.
[1]: https://lore.kernel.org/lkml/20260304095123.GP606826@noisy.programming.kicks-ass.net/
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: John Stultz <jstultz@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260512025635.2840817-3-jstultz@google.com
Historically, the prev value from __schedule() was the rq->curr.
This prev value is passed down through numerous functions, and
used in the class scheduler implementations. The fact that
prev was on_cpu until the end of __schedule(), meant it was
stable across the rq lock drops that the class->balance()
implementations often do.
However, with proxy-exec, the prev passed to functions called
by __schedule() is rq->donor, which may not be the same as
rq->curr and may not be on_cpu, this makes the prev value
potentially unstable across rq lock drops.
A recently found issue with proxy-exec, is when we begin doing
return migration from try_to_wake_up(), its possible we may be
waking up the rq->donor. When we do this, we proxy_resched_idle()
to put_prev_set_next() setting the rq->donor to rq->idle, allowing
the rq->donor to be return migrated and allowed to run.
This however runs into trouble, as on another cpu we might be in
the middle of calling __schedule(). Conceptually the rq lock is
held for the majority of the time, but in calling prev_balance()
its possible the class->balance() handler call may briefly drop the rq lock.
This opens a window for try_to_wake_up() to wake and return migrate the
rq->donor before the class logic reacquires the rq lock.
Unfortunately prev_balance() pass in a prev argument, to which we pass
rq->donor. However this prev value can now become stale and incorrect across a
rq lock drop.
So, to correct this, rework the prev_balance() call so that it does not take a
"prev" argument.
Signed-off-by: John Stultz <jstultz@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260512025635.2840817-2-jstultz@google.com
BPF arena (kernel/bpf/arena.c) is compiled only on MMU && 64BIT, while
SCHED_CLASS_EXT depends on BPF_SYSCALL && BPF_JIT && DEBUG_INFO_BTF with no
64BIT requirement. On a 32-bit arch with a BPF JIT, SCX builds while the
arena helpers are absent, so the cid-form code's unconditional calls to
bpf_prog_arena() and bpf_arena_map_kern_vm_start() fail to link:
build_policy.o: undefined reference to `bpf_prog_arena'
build_policy.o: undefined reference to `bpf_arena_map_kern_vm_start'
Guard the three call sites with the same MMU && 64BIT condition that gates
arena.o. A cid-form scheduler needs a BPF arena, which isn't available on
such builds, so it can't run there regardless. cpu-form schedulers don't
touch the arena and are unaffected.
This is a quick workaround to get past the build errors. A fuller fix may
make the whole cid-form path conditional on the same condition, or drop
32-bit support outright.
Fixes: 0e2819cba9 ("sched_ext: Require an arena for cid-form schedulers")
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202605310454.U9iByL2n-lkp@intel.com/
Closes: https://lore.kernel.org/oe-kbuild-all/202605310926.APXMc0RJ-lkp@intel.com/
Signed-off-by: Tejun Heo <tj@kernel.org>
update_tg_load_avg() is called once per leaf cfs_rq from the
__update_blocked_fair() walk that runs inside the NOHZ idle-balance
softirq, and again from update_load_avg() with UPDATE_TG. Its first
operation after the trivial early-outs is unconditionally:
now = sched_clock_cpu(cpu_of(rq_of(cfs_rq)));
if (now - cfs_rq->last_update_tg_load_avg < NSEC_PER_MSEC)
return;
Jakub ran into a system where nohz_idle_balance() was taking 75%
of a CPU (which is handling network traffic and doing many irq_exit_cpu
calls), with 35% of that CPU spent in update_load_avg, and 17% of the
CPU in sched_clock_cpu(), reading the TSC.
In a quick synthetic test, it looks like this patch reduces the
CPU use of sched_balance_update_blocked_averages by about 20%.
Switch the rate-limit to read rq_clock(rq_of(cfs_rq)) instead.
This eliminates the rdtsc, and uses a fairly fresh timestamp,
because all callers of update_tg_load_avg() and clear_tg_load_avg()
hold rq->lock and have called update_rq_clock(rq) within microseconds:
caller pre-state
__update_blocked_fair encloser did update_rq_clock(rq)
update_load_avg's three UPDATE_TG sites under rq->lock after enqueue/dequeue/update_curr
attach_/detach_entity_cfs_rq preceded by update_load_avg(...)
clear_tg_load_avg via offline path rq_clock_start_loop_update(rq) upfront
so rq->clock is fresh at every call. Since cfs_rqs are per-CPU
per-task_group, cfs_rq->last_update_tg_load_avg is always compared
against the same rq's clock; no cross-rq drift.
Signed-off-by: Rik van Riel <riel@surriel.com>
Assisted-by: Claude (Anthropic)
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://patch.msgid.link/20260527110250.6a91718d@fangorn
Commit cd959a3562 ("sched_ext: Add a DL server for sched_ext tasks")
introduced an ext_server deadline server to protect sched_ext tasks from
fair/RT starvation, mirroring the existing fair_server.
Currently, both servers reserve their 50ms/1000ms bandwidth at boot,
regardless of whether a BPF scheduler is loaded. Unused bandwidth is
still reclaimed at runtime by other classes, but the static reservation
prevents the RT class from implicitly using that headroom when one of
the two classes is guaranteed to be empty.
A sysadmin can work around this by writing
/sys/kernel/debug/sched/{fair,ext}_server/cpu*/runtime, but that
requires manual action and not all systems expose debugfs.
A better approach is to make server bandwidth reservations dynamic: only
the scheduling policy that is currently active should register its
reservation, while the inactive one should not artificially hold
capacity (keeping both reservations only when the BPF scheduler is
running in partial mode):
+---------------------------------------------+-------------+------------+
| BPF scheduler state | fair server | ext server |
+---------------------------------------------+-------------+------------+
| not loaded (default boot) | reserved | none |
| loaded full mode (!SCX_OPS_SWITCH_PARTIAL) | none | reserved |
| loaded partial mode (SCX_OPS_SWITCH_PARTIAL)| reserved | reserved |
+---------------------------------------------+-------------+------------+
To achieve this, introduce an "attached/detached" state for each
deadline server, so the kernel can decide whether a server's bandwidth
should be accounted in global bandwidth tracking.
At boot, the system starts with only the fair server contributing to
bandwidth accounting. When a BPF scheduler is enabled, the ext server is
attached and may replace or complement the fair server depending on
whether full or partial mode is used. When sched_ext is disabled, the
system restores the previous deadline bandwidth values and behavior.
The transition logic ensures that switching between scheduling modes is
consistent and reversible, without losing runtime configuration or
requiring manual intervention.
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Juri Lelli <juri.lelli@redhat.com>
Link: https://patch.msgid.link/20260526164420.638711-2-arighi@nvidia.com
Writing runtime or period via the per-CPU dl_server debugfs files
(/sys/kernel/debug/sched/{fair,ext}_server/cpu*/{runtime,period}) on an
offline CPU can trigger two distinct kernel issues:
1) Divide-by-zero in dl_server_apply_params():
Oops: divide error: 0000 [#1] SMP NOPTI
RIP: 0010:dl_server_apply_params+0x239/0x3a0
Call Trace:
sched_server_write_common.isra.0+0x21a/0x3c0
full_proxy_write+0x78/0xd0
vfs_write+0xe7/0x6e0
Both __dl_sub() and __dl_add() divide by cpus internally, which can be
0 once the CPU has been removed from any active root-domain span (this
has been latent since the debugfs interface was introduced).
2) WARN_ON_ONCE in dl_server_start():
WARNING: kernel/sched/deadline.c:1805 at dl_server_start+0x232/0x270
Commit ee6e44dfe6 ("sched/deadline: Stop dl_server before CPU goes
offline") added this check to catch enqueueing the server on an
offline rq.
There's no meaningful semantics for re-configuring the per-CPU dl_server
bandwidth while the CPU is offline, so simply reject the write with
-EBUSY so userspace gets a clear error.
Closes: https://lore.kernel.org/all/20260526092228.3B6891F00A3A@smtp.kernel.org/
Fixes: d741f297bc ("sched/fair: Fair server interface")
Reported-by: Sashiko <sashiko-bot@kernel.org>
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Juri Lelli <juri.lelli@redhat.com>
Tested-by: abaci-kreproducer <abaci@linux.alibaba.com>
Link: https://patch.msgid.link/20260526100502.575774-1-arighi@nvidia.com
Venkat Reported a boot kernel panic next-20260522. Git bisect pointed to
b5ea300a17 ("sched/cache: Make LLC id continuous")
Stacktrace points to llc_mask being null.
NIP [c000000000e58504] _find_first_bit+0x44/0x130
LR [c000000000e58500] _find_first_bit+0x40/0x130
Call Trace:
build_sched_domains+0xad8/0xe50
sched_init_smp+0xa8/0x164
kernel_init_freeable+0x250/0x370
ret_from_kernel_user_thread+0x14/0x1c
On powerpc, cpu_coregroup_mask is available only when the underlying
hardware support coregroup. In shared LPAR, QEMU guest or power9 etc
coregroup isn't supported. In such cases llc_mask was being referenced
when it was null leading to panic.
On powerpc, LLC is at SMT core level. So assumption that coregroup(MC)
domain point to LLC is wrong. Provide a way for archs to say where its
LLC is if it not at MC domain.
Fixes: b5ea300a17 ("sched/cache: Make LLC id continuous")
Closes: https://lore.kernel.org/all/51154de7-3700-4cb4-82f2-1b3a8fa427f7@linux.ibm.com/
Reported-by: Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Co-developed-by: Chen, Yu C <yu.c.chen@intel.com>
Signed-off-by: Shrikanth Hegde <sshegde@linux.ibm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Chen Yu <yu.c.chen@intel.com>
Tested-by: Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Tested-by: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
Link: https://patch.msgid.link/20260529075712.1181039-1-sshegde@linux.ibm.com
|| is a boolean operator, any nonzero (error) return short-circuits
to 1 rather than the actual errno. The caller in scx_init() logs and
propagates this value, so the wrong code reaches upper layers.
Signed-off-by: Cheng-Yang Chou <yphbchou0911@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
The reason for pick_next_task_fair() is the put/set optimization that
avoids touching the common ancestors. However, it is possible to
implement this in the put_prev_task() and set_next_task() calls as
used in put_prev_set_next_task().
Notably, put_prev_set_next_task() is the only site that:
- calls put_prev_task() with a .next argument;
- calls set_next_task() with .first = true.
This means that put_prev_task() can determine the common hierarchy and
stop there, and then set_next_task() can terminate where put_prev_task
stopped.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://patch.msgid.link/20260511120628.057634261@infradead.org
With commit 50653216e4 ("sched: Add support to pick functions to
take rf") removing the balance callback, the pick_task() callback is
in charge of newidle balancing.
This means pick_task_fair() should do so too. This hasn't been a
problem in practise because pick_next_task_fair() is used. However,
since we'll be removing that one shortly, make sure pick_next_task()
is up to scratch.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://patch.msgid.link/20260511120627.944705718@infradead.org
Robots figured out you can read and write this concurrently and got
'upset'. Gemini even noted sched_dynamic_show() can generate
'confusing' output if it observed different values during the
printing.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260511120627.176946327@infradead.org
Commit c9d93a73ce ("sched/fair: Drop redundant RCU read lock in NOHZ
kick path") removed the rcu_read_lock()/unlock() pair from
set_cpu_sd_state_busy() and set_cpu_sd_state_idle() on the assumption
that all callers run in a safe context for rcu_dereference_all(): IRQs
disabled or cpus_write_lock() held.
That assumption is wrong for the CPU hotplug teardown path. When CPUs
are taken offline, set_cpu_sd_state_busy() is invoked via:
cpuhp/N kthread
cpuhp_thread_fun()
cpuhp_invoke_callback()
sched_cpu_deactivate()
nohz_balance_exit_idle()
set_cpu_sd_state_busy()
rcu_dereference_all(per_cpu(sd_llc, cpu))
The cpuhp kthread holds cpu_hotplug_lock (percpu-rwsem) but runs with
preemption and IRQs enabled. As a result, lockdep correctly reports a
suspicious RCU usage on CPU offline, e.g.:
# echo 0 > /sys/devices/system/cpu/cpu1/online
=============================
WARNING: suspicious RCU usage
-----------------------------
kernel/sched/fair.c:12793 suspicious rcu_dereference_check() usage!
...
2 locks held by cpuhp/1/20:
#0: (cpu_hotplug_lock){++++}-{0:0}, at: cpuhp_thread_fun+0x42/0x1ae
#1: (cpuhp_state-down){+.+.}-{0:0}, at: cpuhp_thread_fun+0x72/0x1ae
Call Trace:
lockdep_rcu_suspicious
nohz_balance_exit_idle
sched_cpu_deactivate
cpuhp_invoke_callback
cpuhp_thread_fun
smpboot_thread_fn
Fix this by adding RCU read lock coverage to the one caller that lacks
it: nohz_balance_exit_idle() in the CPU hotplug teardown.
The other callers (nohz_balancer_kick() and nohz_balance_enter_idle())
genuinely run with IRQs disabled, so they remain unchanged.
Fixes: c9d93a73ce ("sched/fair: Drop redundant RCU read lock in NOHZ kick path")
Closes: https://lore.kernel.org/all/38fe0a1d-1a48-435a-910a-c278024d9ac9@samsung.com/
Reported-by: Marek Szyprowski <m.szyprowski@samsung.com>
Suggested-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260522092523.2046095-1-arighi@nvidia.com
ops_cid.set_cmask() expects a cmask. The kernel couldn't write into the
arena, so it translated cpumask -> cmask in kernel memory and passed the
result as a trusted pointer. The BPF cmask helpers all operate on arena
cmasks though, so the BPF side had to word-by-word probe-read the kernel
cmask into an arena cmask via cmask_copy_from_kernel() before any helper
could touch it. It works, but is clumsy.
With direct kernel-side arena access now in place, build the cmask in the
arena. The kernel writes to it through the kern_va side of the dual mapping.
BPF directly dereferences it via an __arena pointer like any other arena
struct.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Build a per-scheduler sub-allocator on top of pages claimed from the BPF
arena registered in the previous patch. Subsequent kernel-managed
arena-resident structures (e.g. per-CPU set_cmask cmask) carve their storage
from this pool.
scx_arena_pool_init() creates a gen_pool. scx_arena_alloc() returns the
kernel VA. On exhaustion, the pool grows by claiming more pages via
bpf_arena_alloc_pages_sleepable(). Chunks are added at the kernel-side
mapping address. Callers translate to the BPF-arena form themselves if
needed.
Allocations sleep (GFP_KERNEL) - they may grow the pool through vzalloc and
arena page allocation. All current consumers run from the enable path (after
ops.init() and the kernel-side arena auto-discovery, before validate_ops()),
where sleeping is fine.
scx_arena_pool_destroy() walks each chunk, returns outstanding ranges to the
gen_pool with gen_pool_free() and then calls gen_pool_destroy(). The
underlying arena pages are released when the arena map itself is torn down,
so the pool destroy doesn't free them explicitly.
v2: Switch scx_arena_alloc() to a loop. (Andrea)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Upcoming patches will let the kernel place arena-resident scratch shared
with the BPF program (e.g. per-CPU set_cmask cmask) so the BPF side can
dereference it directly via __arena pointers, replacing the current
cmask_copy_from_kernel() probe-read loop. That requires each cid-form
scheduler to expose its arena to the kernel. Kernel- side accesses are
recovered by the per-arena scratch-page mechanism.
bpf_scx_reg_cid() walks the struct_ops member progs via
bpf_struct_ops_for_each_prog() and reads each prog's arena via
bpf_prog_arena(). The verifier enforces one arena per program, so each
member prog contributes at most one arena. All non-NULL contributions must
match and at least one member prog must use an arena. The map ref is held on
scx_sched and dropped on sched destroy. cpu-form schedulers (bpf_scx_reg)
are unchanged - no arena requirement.
Signed-off-by: Tejun Heo <tj@kernel.org>
- Spurious WARN in ops_dequeue() racing with concurrent dispatch.
- Self-deadlock between scheduler disable and a concurrent sub-sched
enable.
-----BEGIN PGP SIGNATURE-----
iIQEABYKACwWIQTfIjM1kS57o3GsC/uxYfJx3gVYGQUCahCHGQ4cdGpAa2VybmVs
Lm9yZwAKCRCxYfJx3gVYGdKOAP9C6xYbZSXuPJSugQg7Ogq7GTcMf0EtK7CGVb9x
0pVJigEA1E17Vqf1WTWTp2DOsoPV1adS51wcoTGvklJc0eFRrw0=
=i40E
-----END PGP SIGNATURE-----
Merge tag 'sched_ext-for-7.1-rc4-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext
Pull sched_ext fixes from Tejun Heo:
- Spurious WARN in ops_dequeue() racing with concurrent dispatch
- Self-deadlock between scheduler disable and a concurrent sub-sched
enable
* tag 'sched_ext-for-7.1-rc4-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext:
sched_ext: Fix spurious WARN on stale ops_state in ops_dequeue()
sched_ext: Fix deadlock between scx_root_disable() and concurrent forks
ops_dequeue() can race with finish_dispatch() and spuriously trigger the
"queued task must be in BPF scheduler's custody" warning.
ops_dequeue() snapshots p->scx.ops_state via atomic_long_read_acquire()
and then, in the SCX_OPSS_QUEUED arm, asserts that SCX_TASK_IN_CUSTODY
is set. The two reads are not atomic w.r.t. a concurrent
finish_dispatch() running on another CPU:
CPU 1 CPU 2
===== =====
dequeue_task_scx()
ops_dequeue()
opss = read_acquire(ops_state)
= SCX_OPSS_QUEUED
finish_dispatch()
cmpxchg ops_state:
SCX_OPSS_QUEUED -> SCX_OPSS_DISPATCHING [succeeds]
dispatch_enqueue(SCX_DSQ_GLOBAL,
SCX_ENQ_CLEAR_OPSS)
call_task_dequeue()
p->scx.flags &= ~SCX_TASK_IN_CUSTODY
WARN_ON_ONCE(!(p->scx.flags &
SCX_TASK_IN_CUSTODY))
/* opss is stale: QUEUED,
* but task already claimed */
set_release(ops_state, SCX_OPSS_NONE)
The race has been observed via two distinct call chains: the most common
goes through sched_setaffinity(), a rarer variant through
sched_change_begin().
For SCX_DSQ_GLOBAL / SCX_DSQ_BYPASS, dispatch_enqueue() clears
SCX_TASK_IN_CUSTODY before clearing ops_state to SCX_OPSS_NONE
(intentional, to avoid concurrent non-atomic RMW of p->scx.flags against
ops_dequeue()). The window between those two writes is exactly what
ops_dequeue() observes as "QUEUED without custody".
The observed state is not actually inconsistent, it just means CPU 1 has
already claimed the task and the QUEUED value held by CPU 2 is stale.
Re-read ops_state in that case; the next read is guaranteed to return
SCX_OPSS_DISPATCHING or SCX_OPSS_NONE, both of which exit the switch
cleanly. The retry is bounded: once IN_CUSTODY is cleared, ops_state has
already advanced past QUEUED for this dispatch cycle, and a fresh QUEUED
would require re-enqueue under p's rq lock, which CPU 2 holds.
Changes in v2:
- Use READ_ONCE() for p->scx.flags to ensure fresh reads and prevent
compiler reordering in the lockless path
- Add cpu_relax() to reduce power consumption and improve performance
during the spin-wait
- Use unlikely() to optimize branch prediction for the common case
- Expand the in-code comment to document the race condition and
bounded retry guarantee
Fixes: ebf1ccff79 ("sched_ext: Fix ops.dequeue() semantics")
Suggested-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Samuele Mariotti <smariotti@disroot.org>
Signed-off-by: Paolo Valente <paolo.valente@unimore.it>
Signed-off-by: Tejun Heo <tj@kernel.org>
Sub-sched cap code and other upcoming consumers need bulk cmask ops, both
mutating (and/or/copy/andnot) and predicate (subset/intersects/empty).
cmask_walk_op2() walks the intersection of two ranges word by word;
cmask_walk_op1() walks one range. Both are __always_inline and dispatched on
a compile-time-constant op enum, so each public entry collapses to a
specialized loop with the inner switch reduced to one arm.
Two-cmask ops only touch bits in the intersection of the two ranges; bits
outside are left unchanged. scx_cmask_or_racy() and scx_cmask_copy_racy()
mirror the locking forms but read @src word-by-word through data_race();
callers handle ordering with concurrent writers themselves.
v2: Add scx_cmask_empty().
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
scx_cmask carries @base and @nr_cids but not the bits[] allocation size, so
helpers reshaping the active range have no way to check it fits and later
kfuncs taking caller-provided storage can't validate it.
Add @alloc_words (u64 word count) annotated with __counted_by, and split the
bit-range API into three helpers:
- SCX_CMASK_DEFINE() / __SCX_CMASK_DEFINE() define an on-stack cmask, the
latter taking an explicit capacity for oversized storage.
SCX_CMASK_DEFINE_SHARD() is a thin wrapper that always reserves
SCX_CID_SHARD_MAX_CPUS bits of storage.
- scx_cmask_init() / __scx_cmask_init() initialize a cmask, with the same
tight-vs-explicit split.
- scx_cmask_reframe() reshapes the active range without resizing storage.
The BPF mirror (cmask_init / __cmask_init / cmask_reframe) gets the same
shape.
Add scx_cmask_clear() and scx_cmask_fill() to zero and set the
active-range bits respectively. scx_cpumask_to_cmask() uses
scx_cmask_clear(); scx_cmask_init() would otherwise re-write @alloc_words
on every call.
A later patch uses @alloc_words in scx_cmask_ref_shard() to refuse output
storage that can't hold the requested shard.
v2: Init per-CPU scx_set_cmask_scratch (was zero-init, emitted empty
cmasks). Add nr_cids/alloc_cids check in BPF __cmask_init().
(sashiko AI)
Widen SCX_CMASK_NR_WORDS()/CMASK_NR_WORDS() to compute in u64 so that
@nr_cids near U32_MAX no longer wraps to a small value and bypasses
the bounds check in cmask_reframe(). (Andrea)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
struct scx_cmask is a base-windowed bitmap over cid space. Each bit
represents one cid, so the count of active bits is the count of cids. The
sibling struct scx_cid_shard already uses nr_cids. Rename as a prep so the
following patches that grow the cmask API can use the consistent name.
v2: Also rename src->nr_bits / dst->nr_bits in
cmask_copy_from_kernel(). (sashiko AI)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Recent optimizations of sd->shared assignment moved to allocating a
single instance of per-CPU sched_domain_shared objects per s_data.
Recent optimizations to select_idle_capacity() moved the sd->shared
assignments to "sd_asym" domain when ASYM_CPUCAPACITY is detected but
cache-aware scheduling mandates the presence of "sd_llc_shared" to
compute and cache per-LLC statistics.
Use an "alloc_flags" union in sched_domain_shared to claim a
sched_domain_shared object per sched_domain. Allocation starts searching
for an available / matching sched_domain_shared instance from the first
CPU of sched_domain_span(sd) (sd can be sd_llc, or sd_asym). If the
shared object is claimed by another domain, the instance corresponding
to next CPU in the domain span is explored until a matching / available
instance is found.
In case of a single CPU in sched_domain_span(), the domain will be
degenerated and a temporary overlap of ->shared objects across different
domains is acceptable.
"alloc_flags" forms a union with "nr_idle_scan" and the stale flags are
left as is when the sd->shared is published. The expectation is for the
first load balancing instance to correct the value just like the current
behavior, except the initial value is no longer 0.
Originally-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Andrea Righi <arighi@nvidia.com>
RT migration is done aggressively. When a CPU schedules out a high
priority RT task for a lower priority task, it will look to see if there's
any RT tasks that are waiting to run on another CPU that is of higher
priority than the task this CPU is about to run. If it finds one, it will
pull that task over to the CPU and allow it to run there instead.
Normally, this pulling is done by looking at the RT overloaded mask (rto)
which contains all the CPUs in the scheduler domain with RT tasks that are
waiting to run due to a higher priority RT task currently running on their
CPU. The CPU that is about to schedule a lower priority task will grab the
rq lock of the overloaded CPU and move the RT task from that CPU's runqueue
to the local one and schedule the higher priority RT task.
This caused issues when a lot of CPUs would schedule a lower priority task
at the same time. They would all try to grab the same runqueue lock of
the CPU with the overloaded RT tasks. Only the first CPU that got in will
get that task. All the others would wait until they got the runqueue lock
and see there's nothing to pull and do nothing. On systems with lots of
CPUs, this caused a large latency (up to 500us) which is beyond what
PREEMPT_RT is to allow.
The solution to that was to create an RT_PUSH_IPI logic. When any CPU
wanted to pull a task, instead of grabbing the runqueue lock of the
overloaded CPU, it would start by sending an IPI to the overloaded CPU,
and that IPI handler would have the CPU with the waiting RT task do a push
instead. Then that handler would send an IPI to the next CPU with
overloaded RT tasks, and so on. Note, after the first CPU starts this
process, if another CPU wanted to do a pull, it would see that the process
has already begun and would only increment a counter to have the IPIs
continue again.
The RT_PUSH_IPI solved the latency problem with PREEMPT_RT but could cause
a new issue with non PREEMPT_RT. Namely, softirqs run in a threaded
context on PREEMPT_RT but they can run in an interrupt context in non-RT.
If an IPI lands on a CPU that has just woken up multiple RT tasks and the
current CPU is running a non RT or a low priority RT task, instead of
doing a push, it would simply do a schedule on that CPU. But if a softirq
was also executing on this CPU, the schedule would need to wait until the
softirq finished. Until then, the CPU would still be considered overloaded
as there are RT tasks still waiting to run on it.
A live lock occurred on a workload that was doing heavy networking traffic
on a large machine where the softirqs would run 500us out of 750us. And it
would also be waking up RT tasks, causing the RT pull logic to be
constantly executed.
When a softirq triggered on a CPU with RT tasks queued but not running
yet, and the other CPUs would see this CPU as being overloaded, they would
send an IPI over to it. The CPU would notice that the waiting RT tasks are
of higher priority than the currently running task and simply schedule
that CPU instead. But because the softirq was executing, before it could
schedule, it would receive another IPI to do the same. The amount of IPIs
would slow down the currently running softirq so much that before it could
return back to task context, it would execute another softirq never
allowing the CPU to schedule. This live locked that CPU.
As RT_PUSH_IPI was created to help PREEMPT_RT, make it default off if
PREEMPT_RT is not enabled.
Fixes: b6366f048e ("sched/rt: Use IPI to trigger RT task push migration instead of pulling")
Closes: https://lore.kernel.org/all/20260506235716.2530720-1-tj@kernel.org/
Reported-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260515103740.25ccbed8@gandalf.local.home
K Prateek noticed we weren't setting the rq->next_class in
proxy_resched_idle(), when I was debugging an issue seen with
CONFIG_SCHED_PROXY_EXEC and some of Peter's new patches, and
suggested this fix.
So set rq->next_class when we temporarily switch the donor to
idle, so we don't accidentally call wakeup_preempt_fair()
with idle as the donor.
Suggested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: John Stultz <jstultz@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260514234732.3170197-1-jstultz@google.com
Add to select_idle_capacity() the same SIS_UTIL-controlled idle-scan
mechanism, already used by select_idle_cpu(): when sched_feat(SIS_UTIL)
is enabled and the LLC domain has sched_domain_shared data, derive the
per-attempt scan limit from sd->shared->nr_idle_scan.
That bounds the walk on large LLCs: once nr_idle_scan is exhausted,
return the best CPU seen so far. The early exit is gated on
!has_idle_core so an active idle-core search (SMT with idle cores
reported by test_idle_cores()) isn't cut short before it gets a chance
to find one.
Co-developed-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://patch.msgid.link/20260509180955.1840064-6-arighi@nvidia.com
When SD_ASYM_CPUCAPACITY load balancing considers pulling a misfit task,
capacity_of(dst_cpu) can overstate available compute if the SMT sibling is
busy: the core does not deliver its full nominal capacity.
If SMT is active and dst_cpu is not on a fully idle core, skip this
destination so we do not migrate a misfit expecting a capacity upgrade we
cannot actually provide.
Reported-by: Felix Abecassis <fabecassis@nvidia.com>
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://patch.msgid.link/20260509180955.1840064-5-arighi@nvidia.com
On systems with asymmetric CPU capacity (e.g., ACPI/CPPC reporting
different per-core frequencies), the wakeup path uses
select_idle_capacity() and prioritizes idle CPUs with higher capacity
for better task placement. However, when those CPUs belong to SMT cores,
their effective capacity can be much lower than the nominal capacity
when the sibling thread is busy: SMT siblings compete for shared
resources, so a "high capacity" CPU that is idle but whose sibling is
busy does not deliver its full capacity. This effective capacity
reduction cannot be modeled by the static capacity value alone.
Introduce SMT awareness in the asym-capacity idle selection policy: when
SMT is active, always prefer fully-idle SMT cores over partially-idle
ones.
Prioritizing fully-idle SMT cores yields better task placement because
the effective capacity of partially-idle SMT cores is reduced; always
preferring them when available leads to more accurate capacity usage on
task wakeup.
On an SMT system with asymmetric CPU capacities (NVIDIA Vera Rubin),
SMT-aware idle selection has been shown to improve throughput by around
15-18% over NO_ASYM mainline and by around 60% over ASYM mainline, for
CPU-bound workloads (NVBLAS) running an amount of tasks equal to the
amount of SMT cores.
Reported-by: Felix Abecassis <fabecassis@nvidia.com>
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Reviewed-by: K Prateek Nayak <kprateek.nayak@amd.com>
Link: https://patch.msgid.link/20260511142502.3873984-1-arighi@nvidia.com
On asymmetric CPU capacity systems, the wakeup path uses
select_idle_capacity(), which scans the span of sd_asym_cpucapacity
rather than sd_llc.
The has_idle_cores hint however lives on sd_llc->shared, so the
wakeup-time read of has_idle_cores operates on an LLC-scoped blob while
the actual scan/decision spans the asym domain; nr_busy_cpus also lives
in the same shared sched_domain data, but it's never used in the asym
CPU capacity scenario.
Therefore, move the sched_domain_shared object to sd_asym_cpucapacity
whenever the CPU has a SD_ASYM_CPUCAPACITY_FULL ancestor and that
ancestor is non-overlapping (i.e., not built from SD_NUMA). In that case
the scope of has_idle_cores matches the scope of the wakeup scan.
Fall back to attaching the shared object to sd_llc in three cases:
1) plain symmetric systems (no SD_ASYM_CPUCAPACITY_FULL anywhere);
2) CPUs in an exclusive cpuset that carves out a symmetric capacity
island: has_asym is system-wide but those CPUs have no
SD_ASYM_CPUCAPACITY_FULL ancestor in their hierarchy and follow
the symmetric LLC path in select_idle_sibling();
3) exotic topologies where SD_ASYM_CPUCAPACITY_FULL lands on an
SD_NUMA-built domain. init_sched_domain_shared() keys the shared
blob off cpumask_first(span), which on overlapping NUMA domains
would alias unrelated spans onto the same blob. Keep the shared
object on the LLC there; select_idle_capacity() gracefully skips
the has_idle_cores preference when sd->shared is NULL.
While at it, also rename the per-CPU sd_llc_shared to sd_balance_shared,
as it is no longer strictly tied to the LLC.
Co-developed-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Shrikanth Hegde <sshegde@linux.ibm.com>
Acked-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://patch.msgid.link/20260516055850.1345932-1-arighi@nvidia.com
nohz_balancer_kick() is reached from sched_balance_trigger(), which is
called from sched_tick(). sched_tick() runs with IRQs disabled, so the
additional rcu_read_lock/unlock() used around sched_domain accesses in
this path is redundant. Rely on the existing IRQ-disabled context (and
the rcu_dereference_all() checking) instead.
The same applies to set_cpu_sd_state_idle(), called from the idle entry
path with IRQs disabled, and to set_cpu_sd_state_busy(), reachable via
nohz_balance_exit_idle() from two contexts: nohz_balancer_kick() (IRQs
disabled, as above) and sched_cpu_deactivate() (the CPUHP_AP_ACTIVE
teardown, which runs under cpus_write_lock(), so it cannot race with
sched-domain rebuilds). In both cases the rcu_dereference_all()
validation is sufficient.
No functional change intended.
Suggested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: K Prateek Nayak <kprateek.nayak@amd.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://patch.msgid.link/20260509180955.1840064-2-arighi@nvidia.com
There is a use of sched_smt_active() and explicit use of sched_smt_present.
Remove the explicit usage for better code maintenance and readability.
Note that this differs slightly for update_idle_core. It used to call
static_branch_unlikely earlier and now it will call static_branch_likely.
Signed-off-by: Shrikanth Hegde <sshegde@linux.ibm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <vschneid@redhat.com>
Link: https://patch.msgid.link/20260515172456.542799-5-sshegde@linux.ibm.com
For fastpaths such as wakeup and load balance even minimal code additions
can add up. is_core_idle is accessed during load balance.
Other callsites of is_core_idle make sched_smt_active() check first.
Make the same check in should_we_balance.
Rest of access to cpu_smt_mask isn't in fastpath.
Note: Remove the stale comment above is_core_idle. Enqueue methods
of fair aren't close to it anymore.
Suggested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: Shrikanth Hegde <sshegde@linux.ibm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <vschneid@redhat.com>
Link: https://patch.msgid.link/20260515172456.542799-4-sshegde@linux.ibm.com
Now, that cpu_smt_mask is defined as cpumask_of(cpu) for
CONFIG_SCHED_SMT=n, it is possible to get rid of the ifdeffery.
Effectively,
- This makes sched_smt_present is defined always
- cpumask_weight(cpumask_of(cpu)) == 1. So sched_smt_present_inc/dec
will never enable the sched_smt_present. Which is expected.
- Paths that were compile-time eliminated become runtime guarded
using static keys.
- Defines set_idle_cores, test_idle_cores, etc which could likely benefit
the CONFIG_SCHED_SMT=n systems to use the same optimizations within the
LLC at wakeups.
- This will expose sched_smt_present symbol for CONFIG_SCHED_SMT=n.
Likely not a concern.
- There is a bloat of code CONFIG_SCHED_SMT=n. (NR_CPUS=2048)
add/remove: 24/18 grow/shrink: 26/28 up/down: 6396/-3188 (3208)
Total: Before=30629880, After=30633088, chg +0.01%
- No code bloat for CONFIG_SCHED_SMT=y, which is expected.
- Add comments around stop_core_cpuslocked on why ifdefs are not
removed.
- This leaves the remaining uses of CONFIG_SCHED_SMT mainly for
topology building bits which has a policy based decision.
Signed-off-by: Shrikanth Hegde <sshegde@linux.ibm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Phil Auld <pauld@redhat.com>
Reviewed-by: Valentin Schneider <vschneid@redhat.com>
Acked-by: Tejun Heo <tj@kernel.org>
Tested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Link: https://patch.msgid.link/20260515172456.542799-3-sshegde@linux.ibm.com
util_est_update() must be called after updating util_avg during the dequeue
of a task and only when the task is not delayed dequeue.
Move util_est_update() in update_load_avg().
Fixes: b55945c500 ("sched: Fix pick_next_task_fair() vs try_to_wake_up() race")
Closes: https://lore.kernel.org/all/20260512124653.305275-1-qyousef@layalina.io/
Reported-by: Qais Yousef <qyousef@layalina.io>
Reviewed-and-tested-by: Qais Yousef <qyousef@layalina.io>
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260518102345.268452-1-vincent.guittot@linaro.org
Commit 77baa5bafc ("sched/cputime: Fix mul_u64_u64_div_u64() precision
for cputime") added a clamp in cputime_adjust():
if (unlikely(stime > rtime))
stime = rtime;
The justification was that mul_u64_u64_div_u64() could over-approximate
on some architectures (notably arm64 and the old 32-bit fallback), so
the mathematically impossible stime > rtime was nevertheless reachable
and would underflow utime = rtime - stime.
That premise no longer holds. Commit b29a62d87c ("mul_u64_u64_div_u64:
make it precise always") replaced the fallback implementation with an
exact 128-bit long division, and the x86_64 inline asm already produced
exact results. The helper now returns the mathematically correct
floor(a*b/d) on every architecture, so stime <= rtime is guaranteed by
stime <= stime + utime and the clamp is dead code.
Remove it along with its stale comment.
Signed-off-by: Nicolas Pitre <nico@fluxnic.net>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260514202629.673539-1-nico@fluxnic.net
Enqueue and replenish non-deferred deadline servers when their runtime is
exhausted and the replenishment timer could not be started because it is
too close to the wake-up instant.
Signed-off-by: Yuri Andriaccio <yurand2000@gmail.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260430213835.62217-2-yurand2000@gmail.com
Set the default total bandwidth for SCHED_DEADLINE tasks and servers
to ONE. FIFO/RR tasks are already throttled by fair-servers and
ext-servers, and the sysctl_sched_rt_runtime parameter now only
defines the total bw that is allowed to deadline entities.
Signed-off-by: Yuri Andriaccio <yurand2000@gmail.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260430213835.62217-22-yurand2000@gmail.com
On fork without CLONE_VM, the child gets a new mm,
the parent's preferred_llc value is stale for the
child.
Fix this by resetting the task's preferred_llc to -1.
This bug was reported by sashiko.
Fixes: 47d8696b95 ("sched/cache: Assign preferred LLC ID to processes")
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Co-developed-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/0ec7309d0e24ede97656754d1505b7490403d966.1778703694.git.tim.c.chen@linux.intel.com
sched_cache_present is a global static key, but build_sched_domains()
is called per partition from the "Build new domains" loop in
partition_sched_domains_locked(). Each call unconditionally sets the
key based solely on the has_multi_llcs local variable for that partition.
The call to the last partition set the value even when there
are previous partitions with multiple LLCs.
If partition A (multi-LLC) is built first, the key is enabled. Then
when partition B (single-LLC) is built, the key is disabled. The
multi-LLC partition A is still active but the key is now off.
Fix it by doing a similar thing as sched_energy_present: check the
multi-LLCs during the iteration over all the partitions rather than
checking it on a single partition.
This bug was reported by sashiko.
Fixes: d59f4fd1d3 ("sched/cache: Enable cache aware scheduling for multi LLCs NUMA node")
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Co-developed-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/c541af2547d54509fbfd3b3a1e8072e2e5c7ff68.1778703694.git.tim.c.chen@linux.intel.com
If there are multiple LLCs in the system, cache aware scheduling
should be enabled. However, there is a corner case where, if there
is a single NUMA node and a single LLC per node, cache aware
scheduling will be turned on in the current implementation -
because at this moment, the parent domain has not yet been
degenerated, and it is possible that the current domain has the
same cpu span as its parent. There is no need to turn cache aware
scheduling on in this scenario.
Fix it by iterating the parent domains to find a domain that is
a superset of the current sd_llc, so that later, after the duplicated
parent domains have been degenerated, cache aware scheduling will
take effect.
For example, the expected behavior would be:
2 sockets, 1 LLC per socket: MC span=0-3, PKG span=0-7, has_multi_llcs=true
1 socket, 2 LLCs per socket: MC span=0-3, PKG span=0-7, has_multi_llcs=true
2 sockets, 2 LLCs per socket: MC span=0-3, PKG span=0-7, has_multi_llcs=true
1 socket, 1 LLC per socket: MC span=0-3, PKG span=0-3, has_multi_llcs=false
This bug was reported by sashiko.
Fixes: d59f4fd1d3 ("sched/cache: Enable cache aware scheduling for multi LLCs NUMA node")
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Co-developed-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/6328a8a7f40925cec2a712d81ee58128a4c4444a.1778703694.git.tim.c.chen@linux.intel.com
sched_cache_active_set_unlocked() checks hardware support without
locks:
static void sched_cache_active_set(bool locked)
{
/* hardware does not support */
if (!static_branch_likely(&sched_cache_present)) {
_sched_cache_active_set(false, locked);
return;
}
...
If build_sched_domains() runs concurrently during CPU hotplug,
it can disable sched_cache_present under sched_domains_mutex
and the CPU hotplug lock. If a debugfs write thread evaluates
sched_cache_present as true right before that, and then blocks
or gets preempted, it might proceed to enable sched_cache_active
after the hardware support has been marked as absent. Make it
safer by acquiring cpus_read_lock() and sched_domains_mutex_lock()
when the user changes sched_cache_active via debugfs.
This bug was reported by sashiko.
Fixes: 067a313581 ("sched/cache: Allow the user space to turn on and off cache aware scheduling")
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Co-developed-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/9afddf439687f04bb56b46625bd9f153eb8abad5.1778703694.git.tim.c.chen@linux.intel.com
The currently running task cur may not be a CFS task, such as
an RT or Deadline task. For non-CFS tasks, the task_util(cur)
utilization average is not maintained, so this might pass a
stale or meaningless value to can_migrate_llc().
Check if the task is CFS before getting its task_util().
This bug was reported by sashiko.
Fixes: 714059f79f ("sched/cache: Handle moving single tasks to/from their preferred LLC")
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Co-developed-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/f9161133cf040d286dca11344a112c5ef2a5253d.1778703694.git.tim.c.chen@linux.intel.com
There is a race condition that, after a task is enqueued
on a runqueue, task_llc(p) may change due to CPU hotplug,
because the llc_id is dynamically allocated and adjusted
at runtime.
Therefore, checking task_llc(p) to determine whether the
task is being dequeued from its preferred LLC is unreliable
and can cause inconsistent values.
To fix this problem, record whether p is enqueued on its
preferred LLC, in order to pair with account_llc_dequeue()
to maintain a consistent nr_pref_llc_running per runqueue.
This bug was reported by sashiko, and the solution was once
suggested by Prateek.
Fixes: 46afe3af7e ("sched/cache: Track LLC-preferred tasks per runqueue")
Suggested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Co-developed-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/0c8c6a1571d66792a4d2ff0103ba3cc13e059046.1778703694.git.tim.c.chen@linux.intel.com
mm->sc_stat.cpu is written by task_cache_work() and could be read
locklessly by several functions on other CPUs. Use READ_ONCE and
WRITE_ONCE on mm->sc_stat.cpu access and write to prevent inconsistent
values from compiler optimizations when there are multiple accesses.
For example in get_pref_llc(), if the writer updated the field between
two compiler-generated loads, the validation (e.g., cpu != -1) and
subsequent use (e.g., llc_id(cpu)) could operate on different values,
allowing a negative CPU ID to be used as an index.
Leave plain write in mm_init_sched(), where the mm is not
yet visible to other CPUs.
This bug was reported by sashiko.
Fixes: 47d8696b95 ("sched/cache: Assign preferred LLC ID to processes")
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Co-developed-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/63ea494f12efcf265d7134400a06cd75d7f2c310.1778703694.git.tim.c.chen@linux.intel.com
A concurrent task exit might cause a NULL pointer dereference
in account_mm_sched(). Use the locally cached mm pointer instead,
since the active_mm reference guarantees the structure remains
allocated. Meanwhile, skip the kernel thread because it has
nothing to do with cache aware scheduling.
This bug was reported by sashiko and Vern.
Fixes: df0d984759 ("sched/cache: Introduce infrastructure for cache-aware load balancing")
Reported-by: Vern Hao <haoxing990@gmail.com>
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Co-developed-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/all/09cf7ee3-6e27-4505-9692-4b4a4707c8b2@gmail.com/
Link: https://patch.msgid.link/066d8cfa45d4822bf4367e788c50377c66bbcc82.1778703694.git.tim.c.chen@linux.intel.com
rcu_dereference_all() should be used to access the
sd_llc domain under RCU protection.
This bug was reported by sashiko.
Fixes: df0d984759 ("sched/cache: Introduce infrastructure for cache-aware load balancing")
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Co-developed-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/2dc49455e861215d8059a1c877953f0b95990038.1778703694.git.tim.c.chen@linux.intel.com
Introduce a set of debugfs knobs to control how aggressively the
cache aware scheduling does the task aggregation.
(1) aggr_tolerance
With sched_cache enabled, the scheduler uses a process's footprint
as a proxy for its LLC footprint to determine if aggregating tasks
on the preferred LLC could cause cache contention. If the footprint
exceeds the LLC size, aggregation is skipped. Since the kernel
cannot efficiently track per-task cache usage (resctrl is
user-space only), userspace can provide a more accurate hint.
Introduce /sys/kernel/debug/sched/llc_balancing/aggr_tolerance to
let users control how strictly footprint limits aggregation. Values
range from 0 to 100:
- 0: Cache-aware scheduling is disabled.
- 1: Strict; tasks with footprint larger than LLC size are skipped.
- >=100: Aggressive; tasks are aggregated regardless of footprint.
For example, with a 32MB L3 cache:
- aggr_tolerance=1 -> tasks with footprint > 32MB are skipped.
- aggr_tolerance=99 -> tasks with footprint > 784GB are skipped
(784GB = (1 + (99 - 1) * 256) * 32MB).
Similarly, /sys/kernel/debug/sched/llc_balancing/aggr_tolerance also
controls how strictly the number of active threads is considered when
doing cache aware load balance. The number of SMTs is also considered.
High SMT counts reduce the aggregation capacity, preventing excessive
task aggregation on SMT-heavy systems like Power10/Power11.
Yangyu suggested introducing separate aggregation controls for the
number of active threads and memory footprint checks. Since there are
plans to add per-process/task group controls, fine-grained tunables are
deferred to that implementation.
(2) epoch_period, epoch_affinity_timeout,
imb_pct, overaggr_pct are also turned into tunables.
Suggested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Suggested-by: Madadi Vineeth Reddy <vineethr@linux.ibm.com>
Suggested-by: Shrikanth Hegde <sshegde@linux.ibm.com>
Suggested-by: Tingyin Duan <tingyin.duan@gmail.com>
Suggested-by: Jianyong Wu <jianyong.wu@outlook.com>
Suggested-by: Yangyu Chen <cyy@cyyself.name>
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Co-developed-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Tingyin Duan <tingyin.duan@gmail.com>
Link: https://patch.msgid.link/1c62cc060ba2b33d7b1f0ed98b3390128edbae93.1778703694.git.tim.c.chen@linux.intel.com
Prateek and Tingyin reported that memory-intensive workloads (such as
stream) can saturate memory bandwidth and caches on the preferred LLC
when sched_cache aggregates too many threads.
To mitigate this, estimate a process's memory footprint by comparing
its NUMA balancing fault statistics to the size of the LLC. If the
footprint exceeds the LLC size, skip cache-aware scheduling.
Note that footprint is only an approximation of the memory footprint,
since the kernel lacks suitable metrics to estimate the real working
set. If a user-provided hint is available in the future, it would be
more accurate. A later patch will allow users to provide a hint to
adjust this threshold.
Suggested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Suggested-by: Vern Hao <vernhao@tencent.com>
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Co-developed-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Tingyin Duan <tingyin.duan@gmail.com>
Link: https://patch.msgid.link/95cf64a385bcc12f18dcebe9d59e8d3ba8bb318f.1778703694.git.tim.c.chen@linux.intel.com
Cache aware scheduling needs to know the LLC size that a process
can use, so as to avoid memory-intensive tasks from being
over-aggregated on a single LLC.
Introduce a preparation patch to add get_effective_llc_bytes() to
get the LLC size that a CPU can use. The function can be further
enhanced by subtracting the LLC cache ways reserved by resctrl
(CAT in Intel RDT, etc).
Suggested-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Co-developed-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Tingyin Duan <tingyin.duan@gmail.com>
Link: https://patch.msgid.link/37afee09ff608034da0ce149e72d33b6f4698edf.1778703694.git.tim.c.chen@linux.intel.com
For a single thread, the current wakeup path tends to place it
on the same LLC where it was previously running with cache-hot
data. There is no need to enable cache-aware scheduling for
single-threaded processes for the following reasons:
1. Cache-aware scheduling primarily benefits multi-threaded
processes where threads share data. Single-threaded processes
typically have no inter-thread data sharing and thus gain little.
2. Enabling it incurs the additional overhead of tracking the
thread's residency in the LLCs.
3. Bypassing single-threaded processes avoids excessive
concentration of such tasks on a single LLC.
Nevertheless, this check can be omitted if users explicitly
provide hints for such single-threaded workloads where different
processes have shared memory, e.g., via prctl() or other interfaces
to be added in the future.
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Co-developed-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Tingyin Duan <tingyin.duan@gmail.com>
Link: https://patch.msgid.link/8a59a13aa58fdb48e410ecb2aabd97fe3ea5d256.1778703694.git.tim.c.chen@linux.intel.com
A performance regression was observed by Prateek when running hackbench
with many threads per process (high fd count). To avoid this, processes
with a large number of active threads are excluded from cache-aware
scheduling.
With sched_cache enabled, record the number of active threads in each
process during the periodic task_cache_work(). While iterating over
CPUs, if the currently running task belongs to the same process as the
task that launched task_cache_work(), increment the active thread count.
If the number of active threads within the process exceeds the number
of Cores (divided by the SMT number) in the LLC, do not enable
cache-aware scheduling. However, on systems with a smaller number of
CPUs within 1 LLC, like Power10/Power11 with SMT4 and an LLC size of 4,
this check effectively disables cache-aware scheduling for any process.
One possible solution suggested by Peter is to use an LLC-mask instead
of a single LLC value for preference. Once there are a 'few' LLCs as
preference, this constraint becomes a little easier. It could be an
enhancement in the future.
For users who wish to perform task aggregation regardless, a debugfs knob
is provided for tuning in a subsequent change.
Suggested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Suggested-by: Aaron Lu <ziqianlu@bytedance.com>
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Co-developed-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Tingyin Duan <tingyin.duan@gmail.com>
Link: https://patch.msgid.link/d076cd21a8e6c6341d1e2d927e118db770ebb650.1778703694.git.tim.c.chen@linux.intel.com
Scanning online CPUs to calculate the occupancy might be
time-consuming. Only allow 1 thread of the process to scan
the CPUs at the same time, which is similar to what
NUMA balance does in task_numa_work().
Signed-off-by: Jianyong Wu <wujianyong@hygon.cn>
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/5672b52e588b855b01e5a1a17822f7c6c7237a3d.1778703694.git.tim.c.chen@linux.intel.com
Pull to receive:
39e25a2100 ("sched_ext: Drop NONE early return in scx_disable_and_exit_task()")
b273b75b8d ("sched_ext: INIT_LIST_HEAD() &sch->all in scx_alloc_and_add_sched()")
cceb874eee ("sched_ext: Defer sub_kset base put to scx_sched_free_rcu_work")
6ae315d379 ("sched_ext: Use HK_TYPE_DOMAIN_BOOT to detect isolcpus= domain isolation")
515e3996a4 ("sched_ext: Fix deadlock between scx_root_disable() and concurrent forks")
to prepare for-7.2 for further sub-sched changes.
Signed-off-by: Tejun Heo <tj@kernel.org>
scx_root_disable() enters SCX_DISABLING before it grabs scx_enable_mutex to
clear __scx_switched_all and scx_switching_all. task_should_scx() short-circuits on DISABLING,
so forks in that window land on fair while next_active_class() still skips
fair - the new tasks stall.
This can deadlock the disable path itself: scx_alloc_and_add_sched() runs
under scx_enable_mutex and creates a helper kthread; if that new kthread is
one of the stalled fair tasks, the mutex holder waits forever and
scx_root_disable() can never make progress. Only sub-sched support exposes
this, since sub-sched enables are the only path where
scx_alloc_and_add_sched() can race the root's disable.
Move the DISABLING check after @scx_switching_all. @scx_switching_all
serves as a proxy for __scx_switched_all, so while it's set, forks keep
going to scx. Once cleared, DISABLING applies normally.
v2: Reword in-source comment and description. (Andrea)
Fixes: 337ec00b1d ("sched_ext: Implement cgroup sub-sched enabling and disabling")
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Bulk is hardening of the new sub-scheduler infrastructure.
- UAFs and lifecycle bugs on the sub-sched attach/detach paths: parent
sub_kset freed under a racing child, list_del_rcu on an uninitialized
list head, ops->priv stomped by concurrent attach/detach, and a UAF in
the init-failure error path.
- Task state-machine reorg closing concurrent enable-vs-dead races: a
task exiting during the unlocked init window could trip NULL ops
derefs or skip exit_task() cleanup.
- A scx_link_sched() self-deadlock on scx_sched_lock.
- isolcpus: stop dereferencing the now-RCU-protected HK_TYPE_DOMAIN
cpumask without RCU, and stop rejecting BPF schedulers when only
cpuset isolated partitions are active.
- PREEMPT_RT: disable irq_work runs in hardirq context so dumps show the
failing task rather than the irq_work kthread.
- Assorted !CONFIG_EXT_SUB_SCHED, randconfig, and selftest build fixes.
-----BEGIN PGP SIGNATURE-----
iIQEABYKACwWIQTfIjM1kS57o3GsC/uxYfJx3gVYGQUCagTk1g4cdGpAa2VybmVs
Lm9yZwAKCRCxYfJx3gVYGT6TAP0ZbRHz9ViligecZXIHjEvZQjEV4sn1NLpGi4og
V0Ol2AD/RzqHQZo5+HpMz4hPrcZdkAWcr74cLrNTJ2WQjOk4RgE=
=6Mbx
-----END PGP SIGNATURE-----
Merge tag 'sched_ext-for-7.1-rc3-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext
Pull sched_ext fixes from Tejun Heo:
"The bulk of this is hardening of the new sub-scheduler infrastructure.
- UAFs and lifecycle bugs on the sub-sched attach/detach paths:
parent sub_kset freed under a racing child, list_del_rcu on an
uninitialized list head, ops->priv stomped by concurrent
attach/detach, and a UAF in the init-failure error path
- Task state-machine reorg closing concurrent enable-vs-dead races: a
task exiting during the unlocked init window could trip NULL ops
derefs or skip exit_task() cleanup
- A scx_link_sched() self-deadlock on scx_sched_lock
- isolcpus: stop dereferencing the now-RCU-protected HK_TYPE_DOMAIN
cpumask without RCU, and stop rejecting BPF schedulers when only
cpuset isolated partitions are active
- PREEMPT_RT: disable irq_work runs in hardirq context so dumps show
the failing task rather than the irq_work kthread
- Assorted !CONFIG_EXT_SUB_SCHED, randconfig, and selftest build
fixes"
* tag 'sched_ext-for-7.1-rc3-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext:
sched_ext: Use HK_TYPE_DOMAIN_BOOT to detect isolcpus= domain isolation
sched_ext: Defer sub_kset base put to scx_sched_free_rcu_work
sched_ext: INIT_LIST_HEAD() &sch->all in scx_alloc_and_add_sched()
sched_ext: Drop NONE early return in scx_disable_and_exit_task()
sched_ext: Avoid UAF in scx_root_enable_workfn() init failure path
sched_ext: Clear ops->priv on scx_alloc_and_add_sched() error paths
sched_ext: Fix ops->priv clobber on concurrent attach/detach
selftests/sched_ext: Fix build error in dequeue selftest
sched_ext: Handle SCX_TASK_NONE in disable/switched_from paths
sched_ext: Close sub-sched init race with post-init DEAD recheck
sched_ext: Close root-enable vs sched_ext_dead() race with SCX_TASK_INIT_BEGIN
sched_ext: Replace SCX_TASK_OFF_TASKS flag with SCX_TASK_DEAD state
sched_ext: Inline scx_init_task() and move RESET_RUNNABLE_AT into scx_set_task_state()
sched_ext: Cleanups in preparation for the SCX_TASK_INIT_BEGIN/DEAD work
sched_ext: Use IRQ_WORK_INIT_HARD() to initialize sch->disable_irq_work
sched_ext: Fix !CONFIG_EXT_SUB_SCHED build warnings
sched_ext: Drop unused scx_find_sub_sched() stub
sched_ext: Move scx_error() out of scx_link_sched()'s lock region
- cpuset fixes:
- Partition invalidation could return CPUs still in use by sibling
partitions, producing overlapping effective_cpus.
- cpuset_can_attach() over-reserved DL bandwidth on moves that stayed
within the same root domain.
- Pending DL migration state leaked into later attaches when a later
can_attach() check failed.
- Reorder PF_EXITING and __GFP_HARDWALL checks so dying tasks can
allocate from any node and exit quickly.
- dmem: propagate -ENOMEM instead of spinning forever when the fallback
pool allocation also fails.
- selftests/cgroup: percpu test error-path leak, bogus numeric
comparison of cpuset strings, and a zero-length read() that silently
passed OOM-kill tests.
-----BEGIN PGP SIGNATURE-----
iIQEABYKACwWIQTfIjM1kS57o3GsC/uxYfJx3gVYGQUCagTkzw4cdGpAa2VybmVs
Lm9yZwAKCRCxYfJx3gVYGR+AAQCcYEGJ+yNAzzrTcY8xy7333rorMckSmZt18jzv
1KSqEQD+KjindGNcWP/meQBPnEjcBjix6i961mgnQ99e/UD2HQ4=
=4pT3
-----END PGP SIGNATURE-----
Merge tag 'cgroup-for-7.1-rc3-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup
Pull cgroup fixes from Tejun Heo:
- cpuset fixes:
- Partition invalidation could return CPUs still in use by sibling
partitions, producing overlapping effective_cpus
- cpuset_can_attach() over-reserved DL bandwidth on moves that
stayed within the same root domain
- Pending DL migration state leaked into later attaches when a
later can_attach() check failed
- Reorder PF_EXITING and __GFP_HARDWALL checks so dying tasks can
allocate from any node and exit quickly
- dmem: propagate -ENOMEM instead of spinning forever when the fallback
pool allocation also fails
- selftests/cgroup: percpu test error-path leak, bogus numeric
comparison of cpuset strings, and a zero-length read() that silently
passed OOM-kill tests
* tag 'cgroup-for-7.1-rc3-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup:
cgroup/cpuset: Return only actually allocated CPUs during partition invalidation
selftests/cgroup: Fix error path leaks in test_percpu_basic
cgroup/cpuset: Reserve DL bandwidth only for root-domain moves
cgroup/cpuset: Reset DL migration state on can_attach() failure
selftests/cgroup: Fix string comparison in write_test
selftests/cgroup: Fix cg_read_strcmp() empty string comparison
cgroup/dmem: Return -ENOMEM on failed pool preallocation
cgroup/cpuset: move PF_EXITING check before __GFP_HARDWALL in cpuset_current_node_allowed()
scx_enable() refuses to attach a BPF scheduler when isolcpus=domain is
in effect by comparing housekeeping_cpumask(HK_TYPE_DOMAIN) against
cpu_possible_mask.
Since commit 27c3a5967f ("sched/isolation: Convert housekeeping
cpumasks to rcu pointers"), HK_TYPE_DOMAIN's cpumask is RCU protected
and dereferencing it requires either RCU read lock, the cpu_hotplug
write lock, or the cpuset lock; scx_enable() holds none of these, so
booting with isolcpus=domain and attaching any BPF scheduler triggers
the following lockdep splat:
=============================
WARNING: suspicious RCU usage
-----------------------------
kernel/sched/isolation.c:60 suspicious rcu_dereference_check() usage!
1 lock held by scx_flash/281:
#0: ffffffff8379fce0 (update_mutex){+.+.}-{4:4}, at:
bpf_struct_ops_link_create+0x134/0x1c0
Call Trace:
dump_stack_lvl+0x6f/0xb0
lockdep_rcu_suspicious.cold+0x37/0x70
housekeeping_cpumask+0xcd/0xe0
scx_enable.isra.0+0x17/0x120
bpf_scx_reg+0x5e/0x80
bpf_struct_ops_link_create+0x151/0x1c0
__sys_bpf+0x1e4b/0x33c0
__x64_sys_bpf+0x21/0x30
do_syscall_64+0x117/0xf80
entry_SYSCALL_64_after_hwframe+0x77/0x7f
In addition, commit 03ff735101 ("cpuset: Update HK_TYPE_DOMAIN cpumask
from cpuset") made HK_TYPE_DOMAIN include cpuset isolated partitions as
well, which means the current check also rejects BPF schedulers when a
cpuset partition is active. That contradicts the original intent of
commit 9f391f94a1 ("sched_ext: Disallow loading BPF scheduler if
isolcpus= domain isolation is in effect"), which explicitly noted that
cpuset partitions are honored through per-task cpumasks and should not
be rejected.
Switch to housekeeping_enabled(HK_TYPE_DOMAIN_BOOT), which reads only
the housekeeping flag bit (no RCU dereference) and reflects exactly the
boot-time isolcpus= configuration that the error message refers to.
Fixes: 27c3a5967f ("sched/isolation: Convert housekeeping cpumasks to rcu pointers")
Cc: stable@vger.kernel.org # v7.0+
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Frederic Weisbecker <frederic@kernel.org>
scx_sub_enable_workfn() pins parent->kobj before dropping scx_sched_lock,
but that does not pin parent->sub_kset. Concurrent disable can
kset_unregister and free sub_kset before scx_alloc_and_add_sched()
dereferences it.
Split sub_kset teardown: kobject_del() at disable keeps sysfs removal; defer
kobject_put() to scx_sched_free_rcu_work so the memory survives. A racing
child sees state_in_sysfs=0 with valid memory, sysfs_create_dir() fails, and
the existing exit_kind gate in scx_link_sched() turns it away with -ENOENT.
Fixes: 411d3ef1a7 ("sched_ext: Unregister sub_kset on scheduler disable")
Signed-off-by: Tejun Heo <tj@kernel.org>
On scx_link_sched() error paths (parent disabled, hash insert failure),
&sch->all is never added to scx_sched_all. The cleanup path runs
scx_unlink_sched() unconditionally, which calls list_del_rcu(&sch->all) on a
list_head that was never initialized triggering a corruption warning.
Initialize &sch->all.
Fixes: 54be8de423 ("sched_ext: Factor out scx_link_sched() and scx_unlink_sched()")
Signed-off-by: Tejun Heo <tj@kernel.org>
d3e73a0808 ("sched_ext: Handle SCX_TASK_NONE in disable/switched_from
paths") skipped the trailing scx_set_task_sched(p, NULL) on NONE tasks.
After scx_fail_parent() parks a task at NONE/sched=parent and the parent
is later freed via queue_rcu_work() during root_disable, the preserved
p->scx.sched dangles - print_scx_info() from sched_show_task() reads
sch->ops.name from freed memory.
Drop the early return. __scx_disable_and_exit_task() already short-
circuits on NONE and the SUB_INIT block was cleared by
scx_fail_parent()'s earlier call, so clearing p->scx.sched is the only
work left - and the one thing the path actually needs.
v2: Extend the SUB_INIT block comment to note that the flag is only
set on the sub-enable path, so it's always clear on the NONE
re-entry (Andrea).
Fixes: d3e73a0808 ("sched_ext: Handle SCX_TASK_NONE in disable/switched_from paths")
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Pull to receive:
9a415cc537 ("sched_ext: Avoid UAF in scx_root_enable_workfn() init failure path")
Conflicts with for-7.2's scx_task_iter_relock() rework. The fix moves
put_task_struct(p) past scx_error(); for-7.2 still has it at the old
position. Resolved by dropping the old one.
Signed-off-by: Tejun Heo <tj@kernel.org>
In scx_root_enable_workfn(), put_task_struct(p) is called before scx_error()
dereferences p->comm and p->pid. If the iterator's reference is the last
drop, the task is freed synchronously and the deref becomes a UAF.
Move put_task_struct() past scx_error().
Reported-by: Sashiko <sashiko-bot@kernel.org>
Closes: https://lore.kernel.org/all/20260511214031.AF5E9C2BCB0@smtp.kernel.org/
Fixes: f0e1a0643a ("sched_ext: Implement BPF extensible scheduler class")
Cc: stable@vger.kernel.org # v6.12+
Signed-off-by: Tejun Heo <tj@kernel.org>
Mark !CONFIG_EXT_SUB_SCHED dummy stubs static inline to avoid
-Wunused-function in configs without callers. No functional change.
Signed-off-by: Tejun Heo <tj@kernel.org>
cpuset_can_attach() currently adds the bandwidth of all migrating
SCHED_DEADLINE tasks to sum_migrate_dl_bw. If the source and destination
cpuset effective CPU masks do not overlap, the whole sum is then
reserved in the destination root domain.
set_cpus_allowed_dl(), however, subtracts bandwidth from the source
root domain only when the affinity change really moves the task between
root domains. A DL task can move between cpusets that are still in the
same root domain, so including that task in sum_migrate_dl_bw can reserve
destination bandwidth without a matching source-side subtraction.
Share the root-domain move test with set_cpus_allowed_dl(). Keep
nr_migrate_dl_tasks counting all migrating deadline tasks for cpuset DL
task accounting, but add to sum_migrate_dl_bw only for tasks that need a
root-domain bandwidth move. Keep using the destination cpuset effective
CPU mask and leave the broader can_attach()/attach() transaction model
unchanged.
Fixes: 2ef269ef1a ("cgroup/cpuset: Free DL BW in case can_attach() fails")
Cc: stable@vger.kernel.org # v6.10+
Signed-off-by: Guopeng Zhang <zhangguopeng@kylinos.cn>
Reviewed-by: Waiman Long <longman@redhat.com>
Acked-by: Juri Lelli <juri.lelli@redhat.com>
Tested-by: Juri Lelli <juri.lelli@redhat.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
The tryget_task_struct() calls in scx_sub_disable(),
scx_root_enable_workfn() and scx_sub_enable_workfn() can never fail at
the points they're invoked:
- scx_root_enable_workfn() iterates over scx_tasks under scx_tasks_lock
and rq lock. sched_ext_dead() removes tasks from scx_tasks under the
same scx_tasks_lock before put_task_struct_rcu_user() runs in
finish_task_switch(). So any task observed in scx_tasks must have
usage > 0; put_task_struct_rcu_user() hasn't been called and the
delayed_put_task_struct() callback that decrements usage cannot have
been queued.
- scx_sub_disable() and scx_sub_enable_workfn() iterate via
css_task_iter, which takes a reference on each task in
css_task_iter_next() and holds it until the next iter_next() call, so
usage > 0 is guaranteed by the iter itself.
The actual filter for dead tasks is the SCX_TASK_DEAD check inside
scx_task_iter_next_locked(), not tryget; tryget only fails on zero
usage, a state that can't be reached for tasks visible to these iters.
Commit b7d4b28db7 ("sched_ext: Use SCX_TASK_READY test instead of
tryget_task_struct() during class switch") removed an analogous tryget
in the class-switch loop. Convert the remaining tryget calls to plain
get_task_struct() and update the comment in scx_root_enable_workfn()
that suggested tasks could be observed with zero @usage waiting for an
RCU grace period.
Link: https://lore.kernel.org/all/agCLBxHEUqWIepx8@google.com
Suggested-by: Alice Ryhl <aliceryhl@google.com>
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
scx_alloc_and_add_sched() can fail after @sch has been assigned to
ops->priv. In those cases @sch is torn down (either via kfree() through
the err_free_* chain or via kobject_put() -> scx_kobj_release() -> RCU
work), but @ops->priv is left pointing at the about-to-be-freed pointer.
With the recent -EBUSY gate in scx_root_enable_workfn() and
scx_sub_enable_workfn() that rejects an attach when @ops->priv is still
non-NULL, see commit bbf30b383c ("sched_ext: Fix ops->priv clobber on
concurrent attach/detach"), a dangling @ops->priv permanently locks the
kdata out: every future attach attempt sees a stale binding and returns
-EBUSY even though no scheduler is actually attached.
Clear @ops->priv on the post-assign failure paths so that the kdata
returns to its pre-attach state when the function returns ERR_PTR().
Fixes: bbf30b383c ("sched_ext: Fix ops->priv clobber on concurrent attach/detach")
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Under heavy concurrent attach/detach operations, scx_claim_exit() can
trigger a NULL pointer dereference. This can be reproduced running the
reload_loop kselftests inside a virtme-ng session:
$ vng -v -- ./tools/testing/selftests/sched_ext/runner -t reload_loop
...
BUG: kernel NULL pointer dereference, address: 0000000000000400
RIP: 0010:scx_claim_exit+0x3b/0x120
Call Trace:
<TASK>
bpf_scx_unreg+0x45/0xb0
bpf_struct_ops_map_link_dealloc+0x39/0x50
bpf_link_release+0x18/0x20
__fput+0x10b/0x2e0
__x64_sys_close+0x47/0xa0
The underlying race (diagnosed by Tejun Heo) is a stomp of @ops->priv,
not a missing NULL check:
T2 unreg(K) T1 reg(K)
----------- ---------
sch = ops->priv = sch_b800
scx_disable; flush_disable_work
[scx_root_disable: scx_root=NULL,
mutex_unlock, state=DISABLED]
mutex_lock; state ok
scx_alloc_and_add_sched:
ops->priv = sch_a800
scx_root = sch_a800; init=0
state=ENABLED; mutex_unlock
[flush returns]
RCU_INIT_POINTER(ops->priv, NULL) <-- clobbers sch_a800
kobject_put(sch_b800)
T1 acquires scx_enable_mutex inside scx_root_disable()'s mutex_unlock
window and starts a fresh attach on the same kdata, assigning sch_a800
to @ops->priv. T2 then continues out of scx_disable()/flush_disable_work
and clobbers @ops->priv to NULL, leaking sch_a800; the bpf_link is gone
but state stays SCX_ENABLED, so all future attaches fail with -EBUSY
permanently. The next bpf_scx_unreg() on that kdata then reads NULL
@ops->priv and dereferences it in scx_claim_exit().
Make @ops->priv the lifecycle binding: in scx_root_enable_workfn() and
scx_sub_enable_workfn(), after the existing state check and still under
scx_enable_mutex, refuse with -EBUSY if @ops->priv is non-NULL. This
rejects an attempt to reuse a kdata that is still bound to a previous
scheduler instance, closing the race without changing the unreg side.
Fixes: 105dcd005b ("sched_ext: Introduce scx_prog_sched()")
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Conflict between:
[1] 41e3312861 ("sched_ext: add p->scx.tid and SCX_OPS_TID_TO_TASK lookup")
[2] c941d7391f ("sched_ext: Close root-enable vs sched_ext_dead() race with SCX_TASK_INIT_BEGIN")
in scx_root_enable_workfn()'s post-init block. [1] added a tid hash
insertion under a scoped_guard() for scx_tasks_lock; [2] wraps the same
region in task_rq_lock() for a DEAD recheck. A naive merge would invert the
iter's outer/inner order.
[3] f25ad1e3cb ("sched_ext: Add scx_task_iter_relock() and use it in scx_root_enable_workfn()")
was added to for-7.2 for a clean resolution: scx_task_iter_relock(iter, p)
takes both scx_tasks_lock and @p's rq lock in iter order.
Resolved by routing both sides through [3]'s dual-lock helper: the post-init
region runs under a single scx_task_iter_relock() acquisition, with [2]'s
state machine and [1]'s hash insert in sequence inside it.
Signed-off-by: Tejun Heo <tj@kernel.org>
scx_root_enable_workfn()'s post-init block re-acquires scx_tasks_lock
briefly via a scoped_guard() for the tid hash insertion. c941d7391f
("sched_ext: Close root-enable vs sched_ext_dead() race with
SCX_TASK_INIT_BEGIN") on for-7.1-fixes adds a post-init DEAD recheck that
holds the task's rq lock across the state-machine updates in the same
region. A naive merge would acquire scx_tasks_lock while the rq lock is
held, inverting the iter's outer/inner order (scx_tasks_lock then rq lock).
Add scx_task_iter_relock(iter, p), the counterpart to
scx_task_iter_unlock(), that re-acquires scx_tasks_lock and, if @p is
non-NULL, @p's rq lock. The locks are tracked in @iter so subsequent
iteration releases them.
Use it in scx_root_enable_workfn()'s post-init block and drop the
now-redundant scoped_guard on the hash insertion. The post-init region now
runs with both scx_tasks_lock and the task's rq lock held across the init
failure check, the state-machine updates and the hash insert.
v2: Move scx_task_iter_relock() earlier to ease the for-7.1-fixes merge.
Signed-off-by: Tejun Heo <tj@kernel.org>
scx_fail_parent() leaves cgroup tasks at (state=NONE, sched=parent,
sched_class=ext) until the parent itself is torn down by the scx_error() it
raised. When the later root_disable iterates them, two paths trip on NONE.
scx_disable_and_exit_task() re-enters the wrapper at NONE: the inner switch
returns early but the trailing scx_set_task_sched(p, NULL) clobbers the
parent sched left by scx_fail_parent(), and scx_set_task_state(p, NONE)
wastes a write on an already-NONE task. switched_from_scx() then calls
scx_disable_task(), which WARNs on non-ENABLED state and writes state=READY,
producing a NONE -> READY transition the validation matrix rejects.
Treat NONE as "nothing to do" in both paths. Add a NONE early-return at the
top of scx_disable_and_exit_task() and a parallel NONE check in
switched_from_scx() next to task_dead_and_done().
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
scx_sub_enable_workfn()'s init pass and scx_sub_disable() migration both
drop the rq lock to call __scx_init_task() against the other sched. A
TASK_DEAD @p can fall through sched_ext_dead() in that window.
sched_ext_dead() runs ops.exit_task() on the sched @p was attached to, not
on the sched whose init just completed, so the new allocation leaks.
Reuse the DEAD signal set by sched_ext_dead(). After __scx_init_task()
returns, take task_rq_lock(p) and check for DEAD; on hit, call
scx_sub_init_cancel_task() against the sub sched the init ran for and drop
@p; on miss, proceed as before.
Reported-by: zhidao su <suzhidao@xiaomi.com>
Link: https://lore.kernel.org/all/20260429133155.3825247-1-suzhidao@xiaomi.com/
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
scx_root_enable_workfn() drops the iter rq lock for ops.init_task() and a
TASK_DEAD @p can fall through sched_ext_dead() in that window. The race hits
when sched_ext_dead() observes SCX_TASK_INIT (the intermediate state before
@p->scx.sched is published) and dereferences NULL via SCX_HAS_OP(NULL,
exit_task), or observes SCX_TASK_NONE during the unlocked init window and
skips cleanup so exit_task() never runs.
Add SCX_TASK_INIT_BEGIN. The enable path writes NONE -> INIT_BEGIN under the
iter rq lock, then takes the rq lock again after init to walk INIT_BEGIN ->
INIT -> READY. sched_ext_dead() that wins the rq-lock race observes
INIT_BEGIN and sets DEAD without calling into ops; the post-init recheck
unwinds via scx_sub_init_cancel_task().
scx_fork() runs single-threaded against sched_ext_dead() (the task is not on
scx_tasks until scx_post_fork() adds it) so its INIT_BEGIN -> INIT walk
needs no rq-lock pairing; it rolls back to NONE on ops.init_task() failure.
The validation matrix grows the INIT_BEGIN row and the INIT_BEGIN -> DEAD
edge; INIT now requires INIT_BEGIN as the predecessor. scx_sub_disable()'s
migration writes INIT_BEGIN as a synthetic predecessor to satisfy the
tightened verification.
The sub-sched paths still race with sched_ext_dead() during the unlocked
init window. This will be fixed by the next patch.
Reported-by: zhidao su <suzhidao@xiaomi.com>
Link: https://lore.kernel.org/all/20260429133155.3825247-1-suzhidao@xiaomi.com/
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
SCX_TASK_OFF_TASKS marked tasks already through sched_ext_dead() so cgroup
task iteration would skip them. This can be expressed better with a task
state. Replace the flag with SCX_TASK_DEAD.
scx_disable_and_exit_task() resets state to NONE on its way out, so
sched_ext_dead() now sets DEAD after the wrapper returns. The validation
matrix grows NONE -> DEAD, warns on DEAD -> NONE, and tightens READY's
predecessor to INIT or ENABLED so the new DEAD value cannot silently
transition to READY.
Prepares for the following enable vs dead race fix.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Prepare for the SCX_TASK_INIT_BEGIN/DEAD work that follows by collapsing the
scx_init_task() helper. Move the SCX_TASK_RESET_RUNNABLE_AT setting into
scx_set_task_state() on the INIT transition (it was set unconditionally at
every INIT site through the scx_init_task() helper), inline scx_init_task()
into scx_fork() and scx_root_enable_workfn(), and drop the helper.
As a side effect, scx_sub_disable() migration sequence now also sets
RESET_RUNNABLE_AT (it previously wrote INIT directly without going through
scx_init_task()). The flag triggers a runnable_at reset on the next
set_task_runnable(), which is harmless on a task that has just been moved
between scheds.
On root-enable, p->scx.flags is written without the task's rq lock. The task
isn't visible to scx yet, and a follow-up patch restores the lock-held
write.
v2: Note p->scx.flags rq-lock relaxation on root-enable path. (Andrea)
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Cleanups in preparation for the state-machine work that follows:
- Convert three sub-sched call sites that open-code
rcu_assign_pointer(p->scx.sched, ...) to scx_set_task_sched().
- Move scx_get_task_state()/scx_set_task_state() above the SCX task iter
section so scx_task_iter_next_locked() can use them without a forward
declaration.
No functional change.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
ca1d48a86f ("sched_ext: Use offsetofend on both sides of the ops_cid
layout assert") replaced sizeof() with offsetofend() to dodge 32-bit
PPC trailing padding, but the resulting check is tautological: with
CID_OFFSET_MATCH(priv, priv) already enforcing offsetof(priv) equality
and @priv being the same type in both structs, the two offsetofends
are equal by construction. The original protection - catching a stray
field added past @priv in sched_ext_ops_cid - is gone.
Anchor on a zero-size __end[] marker appended after @priv. Its offset
sits flush after @priv regardless of trailing struct padding; if a
field is inserted past @priv, __end shifts and the assert fires.
Closes: https://lore.kernel.org/all/20260508215211.0C03AC2BCB0@smtp.kernel.org/
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
- Fix spurious failures in rseq self-tests (Mark Brown)
- Fix rseq rseq::cpu_id_start ABI regression due to
TCMalloc's creative use of the supposedly read-only field.
The fix is to introduce a new ABI variant based on
a new (larger) rseq area registration size, to keep
the TCMalloc use of rseq backwards compatible on new kernels.
(Thomas Gleixner)
- Fix wakeup_preempt_fair() for not waking up task (Vincent Guittot)
- Fix s64 mult overflow in vruntime_eligible() (Zhan Xusheng)
Signed-off-by: Ingo Molnar <mingo@kernel.org>
-----BEGIN PGP SIGNATURE-----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=2QJp
-----END PGP SIGNATURE-----
Merge tag 'sched-urgent-2026-05-09' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull scheduler fixes from Ingo Molnar:
- Fix spurious failures in rseq self-tests (Mark Brown)
- Fix rseq rseq::cpu_id_start ABI regression due to TCMalloc's creative
use of the supposedly read-only field
The fix is to introduce a new ABI variant based on a new (larger)
rseq area registration size, to keep the TCMalloc use of rseq
backwards compatible on new kernels (Thomas Gleixner)
- Fix wakeup_preempt_fair() for not waking up task (Vincent Guittot)
- Fix s64 mult overflow in vruntime_eligible() (Zhan Xusheng)
* tag 'sched-urgent-2026-05-09' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched/fair: Fix wakeup_preempt_fair() for not waking up task
sched/fair: Fix overflow in vruntime_eligible()
selftests/rseq: Expand for optimized RSEQ ABI v2
rseq: Reenable performance optimizations conditionally
rseq: Implement read only ABI enforcement for optimized RSEQ V2 mode
selftests/rseq: Validate legacy behavior
selftests/rseq: Make registration flexible for legacy and optimized mode
selftests/rseq: Skip tests if time slice extensions are not available
rseq: Revert to historical performance killing behaviour
rseq: Don't advertise time slice extensions if disabled
rseq: Protect rseq_reset() against interrupts
rseq: Set rseq::cpu_id_start to 0 on unregistration
selftests/rseq: Don't run tests with runner scripts outside of the scripts
sizeof() includes trailing struct pad, offsetofend() doesn't. On
32-bit PPC, sched_ext_ops_cid tail-pads 4 bytes past @priv and the
assert trips. Use offsetofend() on both sides.
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202605081637.DbH4SZ1E-lkp@intel.com/
Fixes: 7e655ed7b9 ("sched_ext: Add bpf_sched_ext_ops_cid struct_ops type")
Signed-off-by: Tejun Heo <tj@kernel.org>
