The reason is that sched_ext would like to have a 'final' call that
knows the next task. By placing put_prev_task() right next to
set_next_task() (as it already is for sched_core) this becomes
trivial.
the idea being that since the task has switched classes, we need to drop
the sched_delayed logic and have that task be deactivated per its previous
dequeue_task(..., DEQUEUE_SLEEP).
Unfortunately, this leaves the task on_rq. This is missing the tail end of
dequeue_entities() that issues __block_task(), which __sched_setscheduler()
won't have done due to not using DEQUEUE_DELAYED - not that it should, as
it is pretty much a fair_sched_class specific thing.
Make switched_from_fair() properly deactivate sched_delayed tasks upon
class changes via __block_task(), as if a
dequeue_task(..., DEQUEUE_DELAYED)
had been issued.
Fixes: 2e0199df252a ("sched/fair: Prepare exit/cleanup paths for delayed_dequeue") Reported-by: "Paul E. McKenney" <paulmck@kernel.org> Reported-by: Chen Yu <yu.c.chen@intel.com> Signed-off-by: Valentin Schneider <vschneid@redhat.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/20240829135353.1524260-1-vschneid@redhat.com
Peter Zijlstra [Mon, 22 May 2023 11:46:30 +0000 (13:46 +0200)]
sched/eevdf: Use sched_attr::sched_runtime to set request/slice suggestion
Allow applications to directly set a suggested request/slice length using
sched_attr::sched_runtime.
The implementation clamps the value to: 0.1[ms] <= slice <= 100[ms]
which is 1/10 the size of HZ=1000 and 10 times the size of HZ=100.
Applications should strive to use their periodic runtime at a high
confidence interval (95%+) as the target slice. Using a smaller slice
will introduce undue preemptions, while using a larger value will
increase latency.
For all the following examples assume a scheduling quantum of 8, and for
consistency all examples have W=4:
{A,B,C,D}(w=1,r=8):
ABCD...
+---+---+---+---
t=0, V=1.5 t=1, V=3.5
A |------< A |------<
B |------< B |------<
C |------< C |------<
D |------< D |------<
---+*------+-------+--- ---+--*----+-------+---
t=2, V=5.5 t=3, V=7.5
A |------< A |------<
B |------< B |------<
C |------< C |------<
D |------< D |------<
---+----*--+-------+--- ---+------*+-------+---
t=0, V=1.25 t=2, V=5.25
A |--------------< A |--------------<
B |--------------< B |--------------<
C |------< C |------<
---+*------+-------+--- ---+----*--+-------+---
t=4, V=8.25 t=6, V=12.25
A |--------------< A |--------------<
B |--------------< B |--------------<
C |------< C |------<
---+-------*-------+--- ---+-------+---*---+---
Note: 1 heavy task -- because q=8, double r such that the deadline of the w=2
task doesn't go below q.
Note: observe the full schedule becomes: W*max(r_i/w_i) = 4*2q = 8q in length.
Note: the period of the heavy task is half the full period at:
W*(r_i/w_i) = 4*(2q/2) = 4q
t=0, V=1.5 t=1, V=3.5
A |--------------< A |---------------<
B |------< B |------<
C |--------------< C |--------------<
D |--------------< D |--------------<
---+*------+-------+--- ---+--*----+-------+---
t=3, V=7.5 t=5, V=11.5
A |---------------< A |---------------<
B |------< B |------<
C |--------------< C |--------------<
D |--------------< D |--------------<
---+------*+-------+--- ---+-------+--*----+---
t=6, V=13.5
A |---------------<
B |------<
C |--------------<
D |--------------<
---+-------+----*--+---
Note: 1 short task -- again double r so that the deadline of the short task
won't be below q. Made B short because its not the leftmost task, but is
eligible with the 0,1,2,3 spread.
Note: like with the heavy task, the period of the short task observes:
W*(r_i/w_i) = 4*(1q/1) = 4q
t=0, V=1.25 t=1, V=3.25
A |--------------< A |--------------<
B |------< B |------<
C |------< C |------<
---+*------+-------+--- ---+--*----+-------+---
t=3, V=7.25 t=5, V=11.25
A |--------------< A |--------------<
B |------< B |------<
C |------< C |------<
---+------*+-------+--- ---+-------+--*----+---
t=6, V=13.25
A |--------------<
B |------<
C |------<
---+-------+----*--+---
Note: 1 heavy and 1 short task -- combine them all.
Note: both the short and heavy task end up with a period of 4q
t=0, V=1 t=2, V=5
A |--------------< A |--------------<
B |------< B |------<
C |------< C |------<
---+*------+-------+--- ---+----*--+-------+---
t=3, V=7 t=5, V=11
A |--------------< A |--------------<
B |------< B |------<
C |------< C |------<
---+------*+-------+--- ---+-------+--*----+---
t=7, V=15
A |--------------<
B |------<
C |------<
---+-------+------*+---
Note: as before but permuted
~~~
From all this it can be deduced that, for the steady state:
- the total period (P) of a schedule is: W*max(r_i/w_i)
- the average period of a task is: W*(r_i/w_i)
- each task obtains the fair share: w_i/W of each full period P
1 massive_intr:(5) | 846018.956 ms | 779188 | avg: 0.273 ms | max: 58.337 ms | sum:212545.245 ms |
2 massive_intr:(5) | 853450.693 ms | 792269 | avg: 0.275 ms | max: 71.193 ms | sum:218263.588 ms |
3 massive_intr:(5) | 843888.920 ms | 771456 | avg: 0.277 ms | max: 92.405 ms | sum:213353.221 ms |
1 chromium-browse:(8) | 53015.889 ms | 131766 | avg: 0.463 ms | max: 36.341 ms | sum:60959.230 ms |
2 chromium-browse:(8) | 53864.088 ms | 136962 | avg: 0.480 ms | max: 27.091 ms | sum:65687.681 ms |
3 chromium-browse:(9) | 53637.904 ms | 132637 | avg: 0.481 ms | max: 24.756 ms | sum:63781.673 ms |
1 cyclictest:(5) | 12615.604 ms | 639689 | avg: 0.471 ms | max: 32.272 ms | sum:301351.094 ms |
2 cyclictest:(5) | 12511.583 ms | 642578 | avg: 0.448 ms | max: 44.243 ms | sum:287632.830 ms |
3 cyclictest:(5) | 12545.867 ms | 635953 | avg: 0.475 ms | max: 25.530 ms | sum:302374.658 ms |
100ms massive_intr 500us cyclictest PREEMPT_SHORT
1 massive_intr:(5) | 839843.919 ms | 837384 | avg: 0.264 ms | max: 74.366 ms | sum:221476.885 ms |
2 massive_intr:(5) | 852449.913 ms | 845086 | avg: 0.252 ms | max: 68.162 ms | sum:212595.968 ms |
3 massive_intr:(5) | 839180.725 ms | 836883 | avg: 0.266 ms | max: 69.742 ms | sum:222812.038 ms |
1 chromium-browse:(11) | 54591.481 ms | 138388 | avg: 0.458 ms | max: 35.427 ms | sum:63401.508 ms |
2 chromium-browse:(8) | 52034.541 ms | 132276 | avg: 0.436 ms | max: 31.826 ms | sum:57732.958 ms |
3 chromium-browse:(8) | 55231.771 ms | 141892 | avg: 0.469 ms | max: 27.607 ms | sum:66538.697 ms |
1 cyclictest:(5) | 13156.391 ms | 667412 | avg: 0.373 ms | max: 38.247 ms | sum:249174.502 ms |
2 cyclictest:(5) | 12688.939 ms | 665144 | avg: 0.374 ms | max: 33.548 ms | sum:248509.392 ms |
3 cyclictest:(5) | 13475.623 ms | 669110 | avg: 0.370 ms | max: 37.819 ms | sum:247673.390 ms |
As per the numbers the, this makes cyclictest (short slice) it's
max-delay more consistent and consistency drops the sum-delay. The
trade-off is that the massive_intr (long slice) gets more context
switches and a slight increase in sum-delay.
Chunxin contributed did_preempt_short() where a task that lost slice
protection from PREEMPT_SHORT gets rescheduled once it becomes
in-eligible.
Peter Zijlstra [Fri, 31 May 2024 13:49:40 +0000 (15:49 +0200)]
sched/fair: Avoid re-setting virtual deadline on 'migrations'
During OSPM24 Youssef noted that migrations are re-setting the virtual
deadline. Notably everything that does a dequeue-enqueue, like setting
nice, changing preferred numa-node, and a myriad of other random crap,
will cause this to happen.
This shouldn't be. Preserve the relative virtual deadline across such
dequeue/enqueue cycles.
Peter Zijlstra [Fri, 26 Apr 2024 11:00:50 +0000 (13:00 +0200)]
sched/eevdf: Fixup PELT vs DELAYED_DEQUEUE
Note that tasks that are kept on the runqueue to burn off negative
lag, are not in fact runnable anymore, they'll get dequeued the moment
they get picked.
As such, don't count this time towards runnable.
Thanks to Valentin for spotting I had this backwards initially.
Peter Zijlstra [Thu, 23 May 2024 10:25:32 +0000 (12:25 +0200)]
sched/fair: Implement delayed dequeue
Extend / fix 86bfbb7ce4f6 ("sched/fair: Add lag based placement") by
noting that lag is fundamentally a temporal measure. It should not be
carried around indefinitely.
OTOH it should also not be instantly discarded, doing so will allow a
task to game the system by purposefully (micro) sleeping at the end of
its time quantum.
Since lag is intimately tied to the virtual time base, a wall-time
based decay is also insufficient, notably competition is required for
any of this to make sense.
Instead, delay the dequeue and keep the 'tasks' on the runqueue,
competing until they are eligible.
Strictly speaking, we only care about keeping them until the 0-lag
point, but that is a difficult proposition, instead carry them around
until they get picked again, and dequeue them at that point.
Peter Zijlstra [Mon, 1 Jul 2024 19:38:11 +0000 (21:38 +0200)]
sched: Teach dequeue_task() about special task states
Since special task states must not suffer spurious wakeups, and the
proposed delayed dequeue can cause exactly these (under some boundary
conditions), propagate this knowledge into dequeue_task() such that it
can do the right thing.
Peter Zijlstra [Thu, 23 May 2024 09:57:43 +0000 (11:57 +0200)]
sched/fair: Implement ENQUEUE_DELAYED
Doing a wakeup on a delayed dequeue task is about as simple as it
sounds -- remove the delayed mark and enjoy the fact it was actually
still on the runqueue.
Peter Zijlstra [Thu, 23 May 2024 09:26:25 +0000 (11:26 +0200)]
sched/fair: Prepare pick_next_task() for delayed dequeue
Delayed dequeue's natural end is when it gets picked again. Ensure
pick_next_task() knows what to do with delayed tasks.
Note, this relies on the earlier patch that made pick_next_task()
state invariant -- it will restart the pick on dequeue, because
obviously the just dequeued task is no longer eligible.
Peter Zijlstra [Wed, 5 Jun 2024 10:09:11 +0000 (12:09 +0200)]
sched/uclamg: Handle delayed dequeue
Delayed dequeue has tasks sit around on the runqueue that are not
actually runnable -- specifically, they will be dequeued the moment
they get picked.
One side-effect is that such a task can get migrated, which leads to a
'nested' dequeue_task() scenario that messes up uclamp if we don't
take care.
Notably, dequeue_task(DEQUEUE_SLEEP) can 'fail' and keep the task on
the runqueue. This however will have removed the task from uclamp --
per uclamp_rq_dec() in dequeue_task(). So far so good.
However, if at that point the task gets migrated -- or nice adjusted
or any of a myriad of operations that does a dequeue-enqueue cycle --
we'll pass through dequeue_task()/enqueue_task() again. Without
modification this will lead to a double decrement for uclamp, which is
wrong.
Peter Zijlstra [Thu, 23 May 2024 08:55:59 +0000 (10:55 +0200)]
sched: Prepare generic code for delayed dequeue
While most of the delayed dequeue code can be done inside the
sched_class itself, there is one location where we do not have an
appropriate hook, namely ttwu_runnable().
Add an ENQUEUE_DELAYED call to the on_rq path to deal with waking
delayed dequeue tasks.
Peter Zijlstra [Thu, 23 May 2024 08:48:09 +0000 (10:48 +0200)]
sched: Split DEQUEUE_SLEEP from deactivate_task()
As a preparation for dequeue_task() failing, and a second code-path
needing to take care of the 'success' path, split out the DEQEUE_SLEEP
path from deactivate_task().
Much thanks to Libo for spotting and fixing a TASK_ON_RQ_MIGRATING
ordering fail.
Peter Zijlstra [Wed, 3 Apr 2024 07:50:41 +0000 (09:50 +0200)]
sched/fair: Re-organize dequeue_task_fair()
Working towards delaying dequeue, notably also inside the hierachy,
rework dequeue_task_fair() such that it can 'resume' an interrupted
hierarchy walk.
Peter Zijlstra [Wed, 3 Apr 2024 07:50:07 +0000 (09:50 +0200)]
sched/fair: Cleanup pick_task_fair() vs throttle
Per 54d27365cae8 ("sched/fair: Prevent throttling in early
pick_next_task_fair()") the reason check_cfs_rq_runtime() is under the
'if (curr)' check is to ensure the (downward) traversal does not
result in an empty cfs_rq.
But then the pick_task_fair() 'copy' of all this made it restart the
traversal anyway, so that seems to solve the issue too.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Ben Segall <bsegall@google.com> Reviewed-by: Valentin Schneider <vschneid@redhat.com> Tested-by: Valentin Schneider <vschneid@redhat.com> Link: https://lkml.kernel.org/r/20240727105028.501679876@infradead.org
Qais Yousef [Mon, 10 Jun 2024 19:20:16 +0000 (20:20 +0100)]
sched/rt: Clean up usage of rt_task()
rt_task() checks if a task has RT priority. But depends on your
dictionary, this could mean it belongs to RT class, or is a 'realtime'
task, which includes RT and DL classes.
Since this has caused some confusion already on discussion [1], it
seemed a clean up is due.
I define the usage of rt_task() to be tasks that belong to RT class.
Make sure that it returns true only for RT class and audit the users and
replace the ones required the old behavior with the new realtime_task()
which returns true for RT and DL classes. Introduce similar
realtime_prio() to create similar distinction to rt_prio() and update
the users that required the old behavior to use the new function.
Move MAX_DL_PRIO to prio.h so it can be used in the new definitions.
Document the functions to make it more obvious what is the difference
between them. PI-boosted tasks is a factor that must be taken into
account when choosing which function to use.
Rename task_is_realtime() to realtime_task_policy() as the old name is
confusing against the new realtime_task().
Dan Carpenter [Thu, 1 Aug 2024 15:44:03 +0000 (10:44 -0500)]
sched/debug: Fix fair_server_period_max value
This code has an integer overflow or sign extension bug which was caught
by gcc-13:
kernel/sched/debug.c:341:57: error: integer overflow in expression of
type 'long int' results in '-100663296' [-Werror=overflow]
341 | static unsigned long fair_server_period_max = (1 << 22) * NSEC_PER_USEC; /* ~4 seconds */
The result is that "fair_server_period_max" is set to 0xfffffffffa000000
(585 years) instead of instead of 0xfa000000 (4 seconds) that was
intended.
Fix this by changing the type to shift from (1 << 22) to (1UL << 22).
Closes: https://lore.kernel.org/all/CA+G9fYtE2GAbeqU+AOCffgo2oH0RTJUxU+=Pi3cFn4di_KgBAQ@mail.gmail.com/ Fixes: d741f297bcea ("sched/fair: Fair server interface") Reported-by: Linux Kernel Functional Testing <lkft@linaro.org> Reported-by: Arnd Bergmann <arnd@kernel.org> Reported-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/a936b991-e464-4bdf-94ab-08e25d364986@stanley.mountain
Tejun Heo [Mon, 5 Aug 2024 22:39:10 +0000 (12:39 -1000)]
sched/fair: Make balance_fair() test sched_fair_runnable() instead of rq->nr_running
balance_fair() skips newidle balancing if rq->nr_running - there are already
tasks on the rq, so no need to try to pull tasks. This tests the total
number of queued tasks on the CPU instead of only the fair class, but is
still correct as the rq can currently only have fair class tasks while
balance_fair() is running.
However, with the addition of sched_ext below the fair class, this will not
hold anymore and make put_prev_task_balance() skip sched_ext's balance()
incorrectly as, when a CPU has only lower priority class tasks,
rq->nr_running would still be positive and balance_fair() would return 1
even when fair doesn't have any tasks to run.
Update balance_fair() to use sched_fair_runnable() which tests
rq->cfs.nr_running which is updated by bandwidth throttling. Note that
pick_next_task_fair() already uses sched_fair_runnable() in its optimized
path for the same purpose.
Reported-by: Peter Zijlstra <peterz@infradead.org> Signed-off-by: Tejun Heo <tj@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Chengming Zhou <chengming.zhou@linux.dev> Reviewed-by: K Prateek Nayak <kprateek.nayak@amd.com> Link: https://lore.kernel.org/r/ZrFUjlCf7x3TNXB8@slm.duckdns.org
sched/core: Fix picking of tasks for core scheduling with DL server
* Use simple CFS pick_task for DL pick_task
DL server's pick_task calls CFS's pick_next_task_fair(), this is wrong
because core scheduling's pick_task only calls CFS's pick_task() for
evaluation / checking of the CFS task (comparing across CPUs), not for
actually affirmatively picking the next task. This causes RB tree
corruption issues in CFS that were found by syzbot.
* Make pick_task_fair clear DL server
A DL task pick might set ->dl_server, but it is possible the task will
never run (say the other HT has a stop task). If the CFS task is picked
in the future directly (say without DL server), ->dl_server will be
set. So clear it in pick_task_fair().
This fixes the KASAN issue reported by syzbot in set_next_entity().
(DL refactoring suggestions by Vineeth Pillai).
Reported-by: Suleiman Souhlal <suleiman@google.com> Signed-off-by: "Joel Fernandes (Google)" <joel@joelfernandes.org> Signed-off-by: Daniel Bristot de Oliveira <bristot@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Vineeth Pillai <vineeth@bitbyteword.org> Tested-by: Juri Lelli <juri.lelli@redhat.com> Link: https://lore.kernel.org/r/b10489ab1f03d23e08e6097acea47442e7d6466f.1716811044.git.bristot@kernel.org
sched/core: Fix priority checking for DL server picks
In core scheduling, a DL server pick (which is CFS task) should be
given higher priority than tasks in other classes.
Not doing so causes CFS starvation. A kselftest is added later to
demonstrate this. A CFS task that is competing with RT tasks can
be completely starved without this and the DL server's boosting
completely ignored.
Fix these problems.
Reported-by: Suleiman Souhlal <suleiman@google.com> Signed-off-by: "Joel Fernandes (Google)" <joel@joelfernandes.org> Signed-off-by: Daniel Bristot de Oliveira <bristot@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Vineeth Pillai <vineeth@bitbyteword.org> Tested-by: Juri Lelli <juri.lelli@redhat.com> Link: https://lore.kernel.org/r/48b78521d86f3b33c24994d843c1aad6b987dda9.1716811044.git.bristot@kernel.org
Among the motivations for the DL servers is the real-time throttling
mechanism. This mechanism works by throttling the rt_rq after
running for a long period without leaving space for fair tasks.
The base dl server avoids this problem by boosting fair tasks instead
of throttling the rt_rq. The point is that it boosts without waiting
for potential starvation, causing some non-intuitive cases.
For example, an IRQ dispatches two tasks on an idle system, a fair
and an RT. The DL server will be activated, running the fair task
before the RT one. This problem can be avoided by deferring the
dl server activation.
By setting the defer option, the dl_server will dispatch an
SCHED_DEADLINE reservation with replenished runtime, but throttled.
The dl_timer will be set for the defer time at (period - runtime) ns
from start time. Thus boosting the fair rq at defer time.
If the fair scheduler has the opportunity to run while waiting
for defer time, the dl server runtime will be consumed. If
the runtime is completely consumed before the defer time, the
server will be replenished while still in a throttled state. Then,
the dl_timer will be reset to the new defer time
If the fair server reaches the defer time without consuming
its runtime, the server will start running, following CBS rules
(thus without breaking SCHED_DEADLINE). Then the server will
continue the running state (without deferring) until it fair
tasks are able to execute as regular fair scheduler (end of
the starvation).
Peter Zijlstra [Mon, 27 May 2024 12:06:50 +0000 (14:06 +0200)]
sched/fair: Add trivial fair server
Use deadline servers to service fair tasks.
This patch adds a fair_server deadline entity which acts as a container
for fair entities and can be used to fix starvation when higher priority
(wrt fair) tasks are monopolizing CPU(s).
According to the cgroup hierarchy, A should preempt B. But current
check_preempt_wakeup_fair() treats cgroup se and task separately, so B
will preempt A unexpectedly.
Unify the wakeup logic by {c,p}se_is_idle only. This makes SCHED_IDLE of
a task a relative policy that is effective only within its own cgroup,
similar to the behavior of NICE.
Also fix se_is_idle() definition when !CONFIG_FAIR_GROUP_SCHED.
Fixes: 304000390f88 ("sched: Cgroup SCHED_IDLE support") Signed-off-by: Tianchen Ding <dtcccc@linux.alibaba.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Josh Don <joshdon@google.com> Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org> Link: https://lkml.kernel.org/r/20240626023505.1332596-1-dtcccc@linux.alibaba.com
Phil Auld [Wed, 15 May 2024 13:37:05 +0000 (09:37 -0400)]
sched: remove HZ_BW feature hedge
As a hedge against unexpected user issues commit 88c56cfeaec4
("sched/fair: Block nohz tick_stop when cfs bandwidth in use")
included a scheduler feature to disable the new functionality.
It's been a few releases (v6.6) and no screams, so remove it.
sched/fair: Remove cfs_rq::nr_spread_over and cfs_rq::exec_clock
nr_spread_over tracks the number of instances where the difference
between a scheduling entity's virtual runtime and the minimum virtual
runtime in the runqueue exceeds three times the scheduler latency,
indicating significant disparity in task scheduling.
Commit that removed its usage: 5e963f2bd: sched/fair: Commit to EEVDF
cfs_rq->exec_clock was used to account for time spent executing tasks.
Commit that removed its usage: 5d69eca542ee1 sched: Unify runtime
accounting across classes
cfs_rq::nr_spread_over and cfs_rq::exec_clock are not used anymore in
eevdf. Remove them from struct cfs_rq.
Peilin He [Tue, 16 Jul 2024 02:42:44 +0000 (10:42 +0800)]
sched/core: Add WARN_ON_ONCE() to check overflow for migrate_disable()
Background
==========
When repeated migrate_disable() calls are made with missing the
corresponding migrate_enable() calls, there is a risk of
'migration_disabled' going upper overflow because
'migration_disabled' is a type of unsigned short whose max value is
65535.
In PREEMPT_RT kernel, if 'migration_disabled' goes upper overflow, it may
make the migrate_disable() ineffective within local_lock_irqsave(). This
is because, during the scheduling procedure, the value of
'migration_disabled' will be checked, which can trigger CPU migration.
Consequently, the count of 'rcu_read_lock_nesting' may leak due to
local_lock_irqsave() and local_unlock_irqrestore() occurring on different
CPUs.
Usecase
========
For example, When I developed a driver, I encountered a warning like
"WARNING: CPU: 4 PID: 260 at kernel/rcu/tree_plugin.h:315
rcu_note_context_switch+0xa8/0x4e8" warning. It took me half a month
to locate this issue. Ultimately, I discovered that the lack of upper
overflow detection mechanism in migrate_disable() was the root cause,
leading to a significant amount of time spent on problem localization.
If the upper overflow detection mechanism was added to migrate_disable(),
the root cause could be very quickly and easily identified.
Effect
======
Using WARN_ON_ONCE() to check if 'migration_disabled' is upper overflow
can help developers identify the issue quickly.
Suggested-by: Peter Zijlstra <peterz@infradead.org> Signed-off-by: Peilin He<he.peilin@zte.com.cn> Signed-off-by: xu xin <xu.xin16@zte.com.cn> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Yunkai Zhang <zhang.yunkai@zte.com.cn> Reviewed-by: Qiang Tu <tu.qiang35@zte.com.cn> Reviewed-by: Kun Jiang <jiang.kun2@zte.com.cn> Reviewed-by: Fan Yu <fan.yu9@zte.com.cn> Link: https://lkml.kernel.org/r/20240716104244764N2jD8gnBpnsLjCDnQGQ8c@zte.com.cn
Zhang Qiao [Thu, 27 Jun 2024 13:33:59 +0000 (21:33 +0800)]
sched: Initialize the vruntime of a new task when it is first enqueued
When creating a new task, we initialize vruntime of the newly task at
sched_cgroup_fork(). However, the timing of executing this action is too
early and may not be accurate.
Because it uses current CPU to init the vruntime, but the new task
actually runs on the cpu which be assigned at wake_up_new_task().
To optimize this case, we pass ENQUEUE_INITIAL flag to activate_task()
in wake_up_new_task(), in this way, when place_entity is called in
enqueue_entity(), the vruntime of the new task will be initialized.
In addition, place_entity() in task_fork_fair() was introduced for two
reasons:
1. Previously, the __enqueue_entity() was in task_new_fair(),
in order to provide vruntime for enqueueing the newly task, the
vruntime assignment equation "se->vruntime = cfs_rq->min_vruntime" was
introduced by commit e9acbff6484d ("sched: introduce se->vruntime").
This is the initial state of place_entity().
2. commit 4d78e7b656aa ("sched: new task placement for vruntime") added
child_runs_first task placement feature which based on vruntime, this
also requires the new task's vruntime value.
After removing the child_runs_first and enqueue_entity() from
task_fork_fair(), this place_entity() no longer makes sense, so remove
it also.
Because when cpuset_cpu_inactive() fails in sched_cpu_deactivate(),
the cpu offline failed, but sched_smt_present is decremented before
calling sched_cpu_deactivate(), it leads to unbalanced dec/inc, so
fix it by incrementing sched_smt_present in the error path.
Fixes: c5511d03ec09 ("sched/smt: Make sched_smt_present track topology") Cc: stable@kernel.org Signed-off-by: Yang Yingliang <yangyingliang@huawei.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Chen Yu <yu.c.chen@intel.com> Reviewed-by: Tim Chen <tim.c.chen@linux.intel.com> Link: https://lore.kernel.org/r/20240703031610.587047-3-yangyingliang@huaweicloud.com
sched/cputime: Fix mul_u64_u64_div_u64() precision for cputime
In extreme test scenarios:
the 14th field utime in /proc/xx/stat is greater than sum_exec_runtime,
utime = 18446744073709518790 ns, rtime = 135989749728000 ns
In cputime_adjust() process, stime is greater than rtime due to
mul_u64_u64_div_u64() precision problem.
before call mul_u64_u64_div_u64(),
stime = 175136586720000, rtime = 135989749728000, utime = 1416780000.
after call mul_u64_u64_div_u64(),
stime = 135989949653530
Trigger condition:
1). User task run in kernel mode most of time
2). ARM64 architecture
3). TICK_CPU_ACCOUNTING=y
CONFIG_VIRT_CPU_ACCOUNTING_NATIVE is not set
Fix mul_u64_u64_div_u64() conversion precision by reset stime to rtime
Merge tag 'kbuild-fixes-v6.11' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild
Pull Kbuild fixes from Masahiro Yamada:
- Fix RPM package build error caused by an incorrect locale setup
- Mark modules.weakdep as ghost in RPM package
- Fix the odd combination of -S and -c in stack protector scripts,
which is an error with the latest Clang
* tag 'kbuild-fixes-v6.11' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
kbuild: Fix '-S -c' in x86 stack protector scripts
kbuild: rpm-pkg: ghost modules.weakdep file
kbuild: rpm-pkg: Fix C locale setup
minmax: simplify and clarify min_t()/max_t() implementation
This simplifies the min_t() and max_t() macros by no longer making them
work in the context of a C constant expression.
That means that you can no longer use them for static initializers or
for array sizes in type definitions, but there were only a couple of
such uses, and all of them were converted (famous last words) to use
MIN_T/MAX_T instead.
Cc: David Laight <David.Laight@aculab.com> Cc: Lorenzo Stoakes <lorenzo.stoakes@oracle.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Commit 3a7e02c040b1 ("minmax: avoid overly complicated constant
expressions in VM code") added the simpler MIN_T/MAX_T macros in order
to avoid some excessive expansion from the rather complicated regular
min/max macros.
The complexity of those macros stems from two issues:
(a) trying to use them in situations that require a C constant
expression (in static initializers and for array sizes)
(b) the type sanity checking
and MIN_T/MAX_T avoids both of these issues.
Now, in the whole (long) discussion about all this, it was pointed out
that the whole type sanity checking is entirely unnecessary for
min_t/max_t which get a fixed type that the comparison is done in.
But that still leaves min_t/max_t unnecessarily complicated due to
worries about the C constant expression case.
However, it turns out that there really aren't very many cases that use
min_t/max_t for this, and we can just force-convert those.
This does exactly that.
Which in turn will then allow for much simpler implementations of
min_t()/max_t(). All the usual "macros in all upper case will evaluate
the arguments multiple times" rules apply.
We should do all the same things for the regular min/max() vs MIN/MAX()
cases, but that has the added complexity of various drivers defining
their own local versions of MIN/MAX, so that needs another level of
fixes first.
kbuild: Fix '-S -c' in x86 stack protector scripts
After a recent change in clang to stop consuming all instances of '-S'
and '-c' [1], the stack protector scripts break due to the kernel's use
of -Werror=unused-command-line-argument to catch cases where flags are
not being properly consumed by the compiler driver:
$ echo | clang -o - -x c - -S -c -Werror=unused-command-line-argument
clang: error: argument unused during compilation: '-c' [-Werror,-Wunused-command-line-argument]
This results in CONFIG_STACKPROTECTOR getting disabled because
CONFIG_CC_HAS_SANE_STACKPROTECTOR is no longer set.
'-c' and '-S' both instruct the compiler to stop at different stages of
the pipeline ('-S' after compiling, '-c' after assembling), so having
them present together in the same command makes little sense. In this
case, the test wants to stop before assembling because it is looking at
the textual assembly output of the compiler for either '%fs' or '%gs',
so remove '-c' from the list of arguments to resolve the error.
All versions of GCC continue to work after this change, along with
versions of clang that do or do not contain the change mentioned above.
Cc: stable@vger.kernel.org Fixes: 4f7fd4d7a791 ("[PATCH] Add the -fstack-protector option to the CFLAGS") Fixes: 60a5317ff0f4 ("x86: implement x86_32 stack protector") Link: https://github.com/llvm/llvm-project/commit/6461e537815f7fa68cef06842505353cf5600e9c Signed-off-by: Nathan Chancellor <nathan@kernel.org> Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Merge tag 'v6.11-merge' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux
Pull turbostat updates from Len Brown:
- Enable turbostat extensions to add both perf and PMT (Intel
Platform Monitoring Technology) counters via the cmdline
- Demonstrate PMT access with built-in support for Meteor Lake's
Die C6 counter
* tag 'v6.11-merge' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux:
tools/power turbostat: version 2024.07.26
tools/power turbostat: Include umask=%x in perf counter's config
tools/power turbostat: Document PMT in turbostat.8
tools/power turbostat: Add MTL's PMT DC6 builtin counter
tools/power turbostat: Add early support for PMT counters
tools/power turbostat: Add selftests for added perf counters
tools/power turbostat: Add selftests for SMI, APERF and MPERF counters
tools/power turbostat: Move verbose counter messages to level 2
tools/power turbostat: Move debug prints from stdout to stderr
tools/power turbostat: Fix typo in turbostat.8
tools/power turbostat: Add perf added counter example to turbostat.8
tools/power turbostat: Fix formatting in turbostat.8
tools/power turbostat: Extend --add option with perf counters
tools/power turbostat: Group SMI counter with APERF and MPERF
tools/power turbostat: Add ZERO_ARRAY for zero initializing builtin array
tools/power turbostat: Replace enum rapl_source and cstate_source with counter_source
tools/power turbostat: Remove anonymous union from rapl_counter_info_t
tools/power/turbostat: Switch to new Intel CPU model defines
Merge tag 'cxl-for-6.11' of git://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl
Pull CXL updates from Dave Jiang:
"Core:
- A CXL maturity map has been added to the documentation to detail
the current state of CXL enabling.
It provides the status of the current state of various CXL features
to inform current and future contributors of where things are and
which areas need contribution.
- A notifier handler has been added in order for a newly created CXL
memory region to trigger the abstract distance metrics calculation.
This should bring parity for CXL memory to the same level vs
hotplugged DRAM for NUMA abstract distance calculation. The
abstract distance reflects relative performance used for memory
tiering handling.
- An addition for XOR math has been added to address the CXL DPA to
SPA translation.
CXL address translation did not support address interleave math
with XOR prior to this change.
Fixes:
- Fix to address race condition in the CXL memory hotplug notifier
- Add missing MODULE_DESCRIPTION() for CXL modules
- Fix incorrect vendor debug UUID define
Misc:
- A warning has been added to inform users of an unsupported
configuration when mixing CXL VH and RCH/RCD hierarchies
- The ENXIO error code has been replaced with EBUSY for inject poison
limit reached via debugfs and cxl-test support
- Moving the PCI config read in cxl_dvsec_rr_decode() to avoid
unnecessary PCI config reads
- A refactor to a common struct for DRAM and general media CXL
events"
* tag 'cxl-for-6.11' of git://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl:
cxl/core/pci: Move reading of control register to immediately before usage
cxl: Remove defunct code calculating host bridge target positions
cxl/region: Verify target positions using the ordered target list
cxl: Restore XOR'd position bits during address translation
cxl/core: Fold cxl_trace_hpa() into cxl_dpa_to_hpa()
cxl/test: Replace ENXIO with EBUSY for inject poison limit reached
cxl/memdev: Replace ENXIO with EBUSY for inject poison limit reached
cxl/acpi: Warn on mixed CXL VH and RCH/RCD Hierarchy
cxl/core: Fix incorrect vendor debug UUID define
Documentation: CXL Maturity Map
cxl/region: Simplify cxl_region_nid()
cxl/region: Support to calculate memory tier abstract distance
cxl/region: Fix a race condition in memory hotplug notifier
cxl: add missing MODULE_DESCRIPTION() macros
cxl/events: Use a common struct for DRAM and General Media events
Merge tag 'unicode-next-6.11' of git://git.kernel.org/pub/scm/linux/kernel/git/krisman/unicode
Pull unicode update from Gabriel Krisman Bertazi:
"Two small fixes to silence the compiler and static analyzers tools
from Ben Dooks and Jeff Johnson"
* tag 'unicode-next-6.11' of git://git.kernel.org/pub/scm/linux/kernel/git/krisman/unicode:
unicode: add MODULE_DESCRIPTION() macros
unicode: make utf8 test count static
In the same way as for other similar files, mark as ghost the new file
generated by depmod for configured weak dependencies for modules,
modules.weakdep, so that although it is not included in the package,
claim the ownership on it.
Signed-off-by: Jose Ignacio Tornos Martinez <jtornosm@redhat.com> Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Merge tag '6.11-rc-smb-client-fixes-part2' of git://git.samba.org/sfrench/cifs-2.6
Pull more smb client updates from Steve French:
- fix for potential null pointer use in init cifs
- additional dynamic trace points to improve debugging of some common
scenarios
- two SMB1 fixes (one addressing reconnect with POSIX extensions, one a
mount parsing error)
* tag '6.11-rc-smb-client-fixes-part2' of git://git.samba.org/sfrench/cifs-2.6:
smb3: add dynamic trace point for session setup key expired failures
smb3: add four dynamic tracepoints for copy_file_range and reflink
smb3: add dynamic tracepoint for reflink errors
cifs: mount with "unix" mount option for SMB1 incorrectly handled
cifs: fix reconnect with SMB1 UNIX Extensions
cifs: fix potential null pointer use in destroy_workqueue in init_cifs error path
Merge tag 'block-6.11-20240726' of git://git.kernel.dk/linux
Pull block fixes from Jens Axboe:
- NVMe pull request via Keith:
- Fix request without payloads cleanup (Leon)
- Use new protection information format (Francis)
- Improved debug message for lost pci link (Bart)
- Another apst quirk (Wang)
- Use appropriate sysfs api for printing chars (Markus)
- ublk async device deletion fix (Ming)
- drbd kerneldoc fixups (Simon)
- Fix deadlock between sd removal and release (Yang)
* tag 'block-6.11-20240726' of git://git.kernel.dk/linux:
nvme-pci: add missing condition check for existence of mapped data
ublk: fix UBLK_CMD_DEL_DEV_ASYNC handling
block: fix deadlock between sd_remove & sd_release
drbd: Add peer_device to Kernel doc
nvme-core: choose PIF from QPIF if QPIFS supports and PIF is QTYPE
nvme-pci: Fix the instructions for disabling power management
nvme: remove redundant bdev local variable
nvme-fabrics: Use seq_putc() in __nvmf_concat_opt_tokens()
nvme/pci: Add APST quirk for Lenovo N60z laptop
Merge tag 'vfs-6.11-rc1.fixes.3' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Pull vfs fixes from Christian Brauner:
"This contains two fixes for this merge window:
VFS:
- I noticed that it is possible for a privileged user to mount most
filesystems with a non-initial user namespace in sb->s_user_ns.
When fsopen() is called in a non-init namespace the caller's
namespace is recorded in fs_context->user_ns. If the returned file
descriptor is then passed to a process privileged in init_user_ns,
that process can call fsconfig(fd_fs, FSCONFIG_CMD_CREATE*),
creating a new superblock with sb->s_user_ns set to the namespace
of the process which called fsopen().
This is problematic as only filesystems that raise FS_USERNS_MOUNT
are known to be able to support a non-initial s_user_ns. Others may
suffer security issues, on-disk corruption or outright crash the
kernel. Prevent that by restricting such delegation to filesystems
that allow FS_USERNS_MOUNT.
Note, that this delegation requires a privileged process to
actually create the superblock so either the privileged process is
cooperaing or someone must have tricked a privileged process into
operating on a fscontext file descriptor whose origin it doesn't
know (a stupid idea).
The bug dates back to about 5 years afaict.
Misc:
- Fix hostfs parsing when the mount request comes in via the legacy
mount api.
In the legacy mount api hostfs allows to specify the host directory
mount without any key.
Restore that behavior"
* tag 'vfs-6.11-rc1.fixes.3' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
hostfs: fix the host directory parse when mounting.
fs: don't allow non-init s_user_ns for filesystems without FS_USERNS_MOUNT
Merge tag 'rust-6.11' of https://github.com/Rust-for-Linux/linux
Pull Rust updates from Miguel Ojeda:
"The highlight is the establishment of a minimum version for the Rust
toolchain, including 'rustc' (and bundled tools) and 'bindgen'.
The initial minimum will be the pinned version we currently have, i.e.
we are just widening the allowed versions. That covers three stable
Rust releases: 1.78.0, 1.79.0, 1.80.0 (getting released tomorrow),
plus beta, plus nightly.
This should already be enough for kernel developers in distributions
that provide recent Rust compiler versions routinely, such as Arch
Linux, Debian Unstable (outside the freeze period), Fedora Linux,
Gentoo Linux (especially the testing channel), Nix (unstable) and
openSUSE Slowroll and Tumbleweed.
In addition, the kernel is now being built-tested by Rust's pre-merge
CI. That is, every change that is attempting to land into the Rust
compiler is tested against the kernel, and it is merged only if it
passes. Similarly, the bindgen tool has agreed to build the kernel in
their CI too.
Thus, with the pre-merge CI in place, both projects hope to avoid
unintentional changes to Rust that break the kernel. This means that,
in general, apart from intentional changes on their side (that we will
need to workaround conditionally on our side), the upcoming Rust
compiler versions should generally work.
In addition, the Rust project has proposed getting the kernel into
stable Rust (at least solving the main blockers) as one of its three
flagship goals for 2024H2 [1].
I would like to thank Niko, Sid, Emilio et al. for their help
promoting the collaboration between Rust and the kernel.
Toolchain and infrastructure:
- Support several Rust toolchain versions.
- Support several bindgen versions.
- Remove 'cargo' requirement and simplify 'rusttest', thanks to
'alloc' having been dropped last cycle.
- Provide proper error reporting for the 'rust-analyzer' target.
'kernel' crate:
- Add 'uaccess' module with a safe userspace pointers abstraction.
- Add 'page' module with a 'struct page' abstraction.
- Support more complex generics in workqueue's 'impl_has_work!'
macro.
'macros' crate:
- Add 'firmware' field support to the 'module!' macro.
- Improve 'module!' macro documentation.
Documentation:
- Provide instructions on what packages should be installed to build
the kernel in some popular Linux distributions.
- Introduce the new kernel.org LLVM+Rust toolchains.
- Explain '#[no_std]'.
And a few other small bits"
Link: https://rust-lang.github.io/rust-project-goals/2024h2/index.html#flagship-goals
* tag 'rust-6.11' of https://github.com/Rust-for-Linux/linux: (26 commits)
docs: rust: quick-start: add section on Linux distributions
rust: warn about `bindgen` versions 0.66.0 and 0.66.1
rust: start supporting several `bindgen` versions
rust: work around `bindgen` 0.69.0 issue
rust: avoid assuming a particular `bindgen` build
rust: start supporting several compiler versions
rust: simplify Clippy warning flags set
rust: relax most deny-level lints to warnings
rust: allow `dead_code` for never constructed bindings
rust: init: simplify from `map_err` to `inspect_err`
rust: macros: indent list item in `paste!`'s docs
rust: add abstraction for `struct page`
rust: uaccess: add typed accessors for userspace pointers
uaccess: always export _copy_[from|to]_user with CONFIG_RUST
rust: uaccess: add userspace pointers
kbuild: rust-analyzer: improve comment documentation
kbuild: rust-analyzer: better error handling
docs: rust: no_std is used
rust: alloc: add __GFP_HIGHMEM flag
rust: alloc: fix typo in docs for GFP_NOWAIT
...
Merge tag 'apparmor-pr-2024-07-25' of git://git.kernel.org/pub/scm/linux/kernel/git/jj/linux-apparmor
Pull apparmor updates from John Johansen:
"Cleanups
- optimization: try to avoid refing the label in apparmor_file_open
- remove useless static inline function is_deleted
- use kvfree_sensitive to free data->data
- fix typo in kernel doc
Bug fixes:
- unpack transition table if dfa is not present
- test: add MODULE_DESCRIPTION()
- take nosymfollow flag into account
- fix possible NULL pointer dereference
- fix null pointer deref when receiving skb during sock creation"
* tag 'apparmor-pr-2024-07-25' of git://git.kernel.org/pub/scm/linux/kernel/git/jj/linux-apparmor:
apparmor: unpack transition table if dfa is not present
apparmor: try to avoid refing the label in apparmor_file_open
apparmor: test: add MODULE_DESCRIPTION()
apparmor: take nosymfollow flag into account
apparmor: fix possible NULL pointer dereference
apparmor: fix typo in kernel doc
apparmor: remove useless static inline function is_deleted
apparmor: use kvfree_sensitive to free data->data
apparmor: Fix null pointer deref when receiving skb during sock creation
Merge tag 'landlock-6.11-rc1-houdini-fix' of git://git.kernel.org/pub/scm/linux/kernel/git/mic/linux
Pull landlock fix from Mickaël Salaün:
"Jann Horn reported a sandbox bypass for Landlock. This includes the
fix and new tests. This should be backported"
* tag 'landlock-6.11-rc1-houdini-fix' of git://git.kernel.org/pub/scm/linux/kernel/git/mic/linux:
selftests/landlock: Add cred_transfer test
landlock: Don't lose track of restrictions on cred_transfer
Merge tag 'devicetree-fixes-for-6.11-1' of git://git.kernel.org/pub/scm/linux/kernel/git/robh/linux
Pull more devicetree updates from Rob Herring:
"Most of this is a treewide change to of_property_for_each_u32() which
was small enough to do in one go before rc1 and avoids the need to
create of_property_for_each_u32_some_new_name().
- Treewide conversion of of_property_for_each_u32() to drop internal
arguments making struct property opaque
- Add binding for Amlogic A4 SoC watchdog
- Fix constraints for AD7192 'single-channel' property"
* tag 'devicetree-fixes-for-6.11-1' of git://git.kernel.org/pub/scm/linux/kernel/git/robh/linux:
dt-bindings: iio: adc: ad7192: Fix 'single-channel' constraints
of: remove internal arguments from of_property_for_each_u32()
dt-bindings: watchdog: add support for Amlogic A4 SoCs
Merge tag 'iommu-fixes-v6.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/iommu/linux
Pull iommu fixes from Will Deacon:
"We're still resolving a regression with the handling of unexpected
page faults on SMMUv3, but we're not quite there with a fix yet.
- Fix NULL dereference when freeing domain in Unisoc SPRD driver
- Separate assignment statements with semicolons in AMD page-table
code
- Fix Tegra erratum workaround when the CPU is using 16KiB pages"
* tag 'iommu-fixes-v6.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/iommu/linux:
iommu: arm-smmu: Fix Tegra workaround for PAGE_SIZE mappings
iommu/amd: Convert comma to semicolon
iommu: sprd: Avoid NULL deref in sprd_iommu_hw_en
Merge tag 'firewire-fixes-6.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394
Pull firewire fixes from Takashi Sakamoto:
"The recent integration of compiler collections introduced the
technology to check flexible array length at runtime by providing
proper annotations. In v6.10 kernel, a patch was merged into firewire
subsystem to utilize it, however the annotation was inadequate.
There is also the related change for the flexible array in sound
subsystem, but it causes a regression where the data in the payload of
isochronous packet is incorrect for some devices. These bugs are now
fixed"
* tag 'firewire-fixes-6.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394:
ALSA: firewire-lib: fix wrong value as length of header for CIP_NO_HEADER case
Revert "firewire: Annotate struct fw_iso_packet with __counted_by()"
Merge tag 'spi-fix-v6.11-merge-window' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi
Pull spi fixes from Mark Brown:
"The bulk of this is a series of fixes for the microchip-core driver
mostly originating from one of their customers, I also applied an
additional patch adding support for controlling the word size which
came along with it since it's still the merge window and clearly had a
bunch of fairly thorough testing.
We also have a fix for the compatible used to bind spidev to the
BH2228FV"
* tag 'spi-fix-v6.11-merge-window' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi:
spi: spidev: add correct compatible for Rohm BH2228FV
dt-bindings: trivial-devices: fix Rohm BH2228FV compatible string
spi: microchip-core: add support for word sizes of 1 to 32 bits
spi: microchip-core: ensure TX and RX FIFOs are empty at start of a transfer
spi: microchip-core: fix init function not setting the master and motorola modes
spi: microchip-core: only disable SPI controller when register value change requires it
spi: microchip-core: defer asserting chip select until just before write to TX FIFO
spi: microchip-core: fix the issues in the isr
Merge tag 'regulator-fix-v6.11-merge-window' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator
Pull regulator fixes from Mark Brown:
"These two commits clean up the excessively loose dependencies for the
RZG2L USB VBCTRL regulator driver, ensuring it shouldn't prompt for
people who can't use it"
* tag 'regulator-fix-v6.11-merge-window' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator:
regulator: Further restrict RZG2L USB VBCTRL regulator dependencies
regulator: renesas-usb-vbus-regulator: Update the default
Merge tag 'regmap-fix-v6.11-merge-window' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regmap
Pull regmap fix from Mark Brown:
"Arnd sent a workaround for a false positive warning which was showing
up with GCC 14.1"
* tag 'regmap-fix-v6.11-merge-window' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regmap:
regmap: maple: work around gcc-14.1 false-positive warning
Merge tag 'clk-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux
Pull clk fixes from Stephen Boyd:
"A few clk driver fixes for the merge window to fix the build and boot
on some SoCs.
- Initialize struct clk_init_data in the TI da8xx-cfgchip driver so
that stack contents aren't used for things like clk flags leading
to unexpected behavior
- Don't leak stack contents in a debug print in the new Sophgo clk
driver
- Disable the new T-Head clk driver on 32-bit targets to fix the
build due to a division
- Fix Samsung Exynos4 fin_pll wreckage from the clkdev rework done
last cycle by using a struct clk_hw directly instead of a struct
clk consumer"
* tag 'clk-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux:
clk: samsung: fix getting Exynos4 fin_pll rate from external clocks
clk: T-Head: Disable on 32-bit Targets
clk: sophgo: clk-sg2042-pll: Fix uninitialized variable in debug output
clk: davinci: da8xx-cfgchip: Initialize clk_init_data before use
Merge tag 'i3c/for-6.11' of git://git.kernel.org/pub/scm/linux/kernel/git/i3c/linux
Pull i3c updates from Alexandre Belloni:
"This cycle, there are new features for the Designware controller and
fixes for the other IPs:
- dw: optional apb clock and power management support, IBI handling
fixes
- mipi-i3c-hci: IBI handling fixes
- svc: a few fixes"
* tag 'i3c/for-6.11' of git://git.kernel.org/pub/scm/linux/kernel/git/i3c/linux:
dt-bindings: i3c: add header for generic I3C flags
i3c: master: svc: Fix error code in svc_i3c_master_do_daa_locked()
i3c: master: Enhance i3c_bus_type visibility for device searching & event monitoring
i3c: dw: Add power management support
i3c: dw: Add some functions for reusability
i3c: dw: Save timing registers and other values
i3c: master: svc: Improve DAA STOP handle code logic
i3c: dw: Add optional apb clock
i3c: dw: Use new *_enabled clk API
dt-bindings: i3c: dw: Add apb clock binding
i3c: master: svc: Convert comma to semicolon
i3c: mipi-i3c-hci: Round IBI data chunk size to HW supported value
i3c: mipi-i3c-hci: Error out instead on BUG_ON() in IBI DMA setup
i3c: mipi-i3c-hci: Set IBI Status and Data Ring base addresses
i3c: mipi-i3c-hci: Switch to lower_32_bits()/upper_32_bits() helpers
i3c: dw: Remove ibi_capable property
i3c: dw: Fix IBI intr programming
i3c: dw: Fix clearing queue thld
i3c: mipi-i3c-hci: Fix number of DAT/DCT entries for HCI versions < 1.1
i3c: master: svc: resend target address when get NACK
Merge tag 'thermal-6.11-rc1-3' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull thermal control fix from Rafael Wysocki:
"Prevent the thermal core from flooding the kernel log with useless
messages if thermal zone temperature can never be determined (or its
sensor has failed permanently) and make it finally give up and disable
defective thermal zones (Rafael Wysocki)"
* tag 'thermal-6.11-rc1-3' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
thermal: core: Back off when polling thermal zones on errors
thermal: trip: Split thermal_zone_device_set_mode()
Merge tag 'mm-hotfixes-stable-2024-07-26-14-33' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
Pull misc hotfixes from Andrew Morton:
"11 hotfixes, 7 of which are cc:stable. 7 are MM, 4 are other"
* tag 'mm-hotfixes-stable-2024-07-26-14-33' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm:
nilfs2: handle inconsistent state in nilfs_btnode_create_block()
selftests/mm: skip test for non-LPA2 and non-LVA systems
mm/page_alloc: fix pcp->count race between drain_pages_zone() vs __rmqueue_pcplist()
mm: memcg: add cacheline padding after lruvec in mem_cgroup_per_node
alloc_tag: outline and export free_reserved_page()
decompress_bunzip2: fix rare decompression failure
mm/huge_memory: avoid PMD-size page cache if needed
mm: huge_memory: use !CONFIG_64BIT to relax huge page alignment on 32 bit machines
mm: fix old/young bit handling in the faulting path
dt-bindings: arm: update James Clark's email address
MAINTAINERS: mailmap: update James Clark's email address
Merge tag 'timers-urgent-2024-07-26' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull timer migration updates from Thomas Gleixner:
"Fixes and minor updates for the timer migration code:
- Stop testing the group->parent pointer as it is not guaranteed to
be stable over a chain of operations by design.
This includes a warning which would be nice to have but it produces
false positives due to the racy nature of the check.
- Plug a race between CPUs going in and out of idle and a CPU hotplug
operation. The latter can create and connect a new hierarchy level
which is missed in the concurrent updates of CPUs which go into
idle. As a result the events of such a CPU might not be processed
and timers go stale.
Cure it by splitting the hotplug operation into a prepare and
online callback. The prepare callback is guaranteed to run on an
online and therefore active CPU. This CPU updates the hierarchy and
being online ensures that there is always at least one migrator
active which handles the modified hierarchy correctly when going
idle. The online callback which runs on the incoming CPU then just
marks the CPU active and brings it into operation.
- Improve tracing and polish the code further so it is more obvious
what's going on"
* tag 'timers-urgent-2024-07-26' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
timers/migration: Fix grammar in comment
timers/migration: Spare write when nothing changed
timers/migration: Rename childmask by groupmask to make naming more obvious
timers/migration: Read childmask and parent pointer in a single place
timers/migration: Use a single struct for hierarchy walk data
timers/migration: Improve tracing
timers/migration: Move hierarchy setup into cpuhotplug prepare callback
timers/migration: Do not rely always on group->parent
Merge tag 'riscv-for-linus-6.11-mw2' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux
Pull more RISC-V updates from Palmer Dabbelt:
- Support for NUMA (via SRAT and SLIT), console output (via SPCR), and
cache info (via PPTT) on ACPI-based systems.
- The trap entry/exit code no longer breaks the return address stack
predictor on many systems, which results in an improvement to trap
latency.
- Support for HAVE_ARCH_STACKLEAK.
- The sv39 linear map has been extended to support 128GiB mappings.
- The frequency of the mtime CSR is now visible via hwprobe.
* tag 'riscv-for-linus-6.11-mw2' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux: (21 commits)
RISC-V: Provide the frequency of time CSR via hwprobe
riscv: Extend sv39 linear mapping max size to 128G
riscv: enable HAVE_ARCH_STACKLEAK
riscv: signal: Remove unlikely() from WARN_ON() condition
riscv: Improve exception and system call latency
RISC-V: Select ACPI PPTT drivers
riscv: cacheinfo: initialize cacheinfo's level and type from ACPI PPTT
riscv: cacheinfo: remove the useless input parameter (node) of ci_leaf_init()
RISC-V: ACPI: Enable SPCR table for console output on RISC-V
riscv: boot: remove duplicated targets line
trace: riscv: Remove deprecated kprobe on ftrace support
riscv: cpufeature: Extract common elements from extension checking
riscv: Introduce vendor variants of extension helpers
riscv: Add vendor extensions to /proc/cpuinfo
riscv: Extend cpufeature.c to detect vendor extensions
RISC-V: run savedefconfig for defconfig
RISC-V: hwprobe: sort EXT_KEY()s in hwprobe_isa_ext0() alphabetically
ACPI: NUMA: replace pr_info with pr_debug in arch_acpi_numa_init
ACPI: NUMA: change the ACPI_NUMA to a hidden option
ACPI: NUMA: Add handler for SRAT RINTC affinity structure
...
Merge tag 'for-linus-6.11-rc1a-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip
Pull xen fixes from Juergen Gross:
"Two fixes for issues introduced in this merge window:
- fix enhanced debugging in the Xen multicall handling
- two patches fixing a boot failure when running as dom0 in PVH mode"
* tag 'for-linus-6.11-rc1a-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
x86/xen: fix memblock_reserve() usage on PVH
x86/xen: move xen_reserve_extra_memory()
xen: fix multicall debug data referencing
Hongbo Li [Thu, 25 Jul 2024 06:51:30 +0000 (14:51 +0800)]
hostfs: fix the host directory parse when mounting.
hostfs not keep the host directory when mounting. When the host
directory is none (default), fc->source is used as the host root
directory, and this is wrong. Here we use `parse_monolithic` to
handle the old mount path for parsing the root directory. For new
mount path, The `parse_param` is used for the host directory parse.
fs: don't allow non-init s_user_ns for filesystems without FS_USERNS_MOUNT
Christian noticed that it is possible for a privileged user to mount
most filesystems with a non-initial user namespace in sb->s_user_ns.
When fsopen() is called in a non-init namespace the caller's namespace
is recorded in fs_context->user_ns. If the returned file descriptor is
then passed to a process priviliged in init_user_ns, that process can
call fsconfig(fd_fs, FSCONFIG_CMD_CREATE), creating a new superblock
with sb->s_user_ns set to the namespace of the process which called
fsopen().
This is problematic. We cannot assume that any filesystem which does not
set FS_USERNS_MOUNT has been written with a non-initial s_user_ns in
mind, increasing the risk for bugs and security issues.
Prevent this by returning EPERM from sget_fc() when FS_USERNS_MOUNT is
not set for the filesystem and a non-initial user namespace will be
used. sget() does not need to be updated as it always uses the user
namespace of the current context, or the initial user namespace if
SB_SUBMOUNT is set.
Fixes: cb50b348c71f ("convenience helpers: vfs_get_super() and sget_fc()") Reported-by: Christian Brauner <brauner@kernel.org> Signed-off-by: Seth Forshee (DigitalOcean) <sforshee@kernel.org> Link: https://lore.kernel.org/r/20240724-s_user_ns-fix-v1-1-895d07c94701@kernel.org Reviewed-by: Alexander Mikhalitsyn <aleksandr.mikhalitsyn@canonical.com> Signed-off-by: Christian Brauner <brauner@kernel.org>
ALSA: firewire-lib: fix wrong value as length of header for CIP_NO_HEADER case
In a commit 1d717123bb1a ("ALSA: firewire-lib: Avoid
-Wflex-array-member-not-at-end warning"), DEFINE_FLEX() macro was used to
handle variable length of array for header field in struct fw_iso_packet
structure. The usage of macro has a side effect that the designated
initializer assigns the count of array to the given field. Therefore
CIP_HEADER_QUADLETS (=2) is assigned to struct fw_iso_packet.header,
while the original designated initializer assigns zero to all fields.
With CIP_NO_HEADER flag, the change causes invalid length of header in
isochronous packet for 1394 OHCI IT context. This bug affects all of
devices supported by ALSA fireface driver; RME Fireface 400, 800, UCX, UFX,
and 802.
This commit fixes the bug by replacing it with the alternative version of
macro which corresponds no initializer.
The header_length field is byte unit, thus it can not express the number of
elements in header field. It seems that the argument for counted_by
attribute can have no arithmetic expression, therefore this commit just
reverts the issued commit.
minmax: avoid overly complicated constant expressions in VM code
The minmax infrastructure is overkill for simple constants, and can
cause huge expansions because those simple constants are then used by
other things.
For example, 'pageblock_order' is a core VM constant, but because it was
implemented using 'min_t()' and all the type-checking that involves, it
actually expanded to something like 2.5kB of preprocessor noise.
And when that simple constant was then used inside other expansions:
the end result was that one statement expanding to 253kB in size.
There are probably other cases of this, but this one case certainly
stood out.
I've added 'MIN_T()' and 'MAX_T()' macros for this kind of "core simple
constant with specific type" use. These macros skip the type checking,
and as such need to be very sparingly used only for obvious cases that
have active issues like this.
minmax: avoid overly complex min()/max() macro arguments in xen
We have some very fancy min/max macros that have tons of sanity checking
to warn about mixed signedness etc.
This is all things that a sane compiler should warn about, but there are
no sane compiler interfaces for this, and '-Wsign-compare' is broken [1]
and not useful.
So then we compensate (some would say over-compensate) by doing the
checks manually with some truly horrid macro games.
And no, we can't just use __builtin_types_compatible_p(), because the
whole question of "does it make sense to compare these two values" is a
lot more complicated than that.
For example, it makes a ton of sense to compare unsigned values with
simple constants like "5", even if that is indeed a signed type. So we
have these very strange macros to try to make sensible type checking
decisions on the arguments to 'min()' and 'max()'.
But that can cause enormous code expansion if the min()/max() macros are
used with complicated expressions, and particularly if you nest these
things so that you get the first big expansion then expanded again.
The xen setup.c file ended up ballooning to over 50MB of preprocessed
noise that takes 15s to compile (obviously depending on the build host),
largely due to one single line.
So let's split that one single line to just be simpler. I think it ends
up being more legible to humans too at the same time. Now that single
file compiles in under a second.
nilfs2: handle inconsistent state in nilfs_btnode_create_block()
Syzbot reported that a buffer state inconsistency was detected in
nilfs_btnode_create_block(), triggering a kernel bug.
It is not appropriate to treat this inconsistency as a bug; it can occur
if the argument block address (the buffer index of the newly created
block) is a virtual block number and has been reallocated due to
corruption of the bitmap used to manage its allocation state.
So, modify nilfs_btnode_create_block() and its callers to treat it as a
possible filesystem error, rather than triggering a kernel bug.
Dev Jain [Thu, 18 Jul 2024 05:25:04 +0000 (10:55 +0530)]
selftests/mm: skip test for non-LPA2 and non-LVA systems
Post my improvement of the test in e4a4ba415419 ("selftests/mm:
va_high_addr_switch: dynamically initialize testcases to enable LPA2
testing"):
The test begins to fail on 4k and 16k pages, on non-LPA2 systems. To
reduce noise in the CI systems, let us skip the test when higher address
space is not implemented.
Link: https://lkml.kernel.org/r/20240718052504.356517-1-dev.jain@arm.com Fixes: e4a4ba415419 ("selftests/mm: va_high_addr_switch: dynamically initialize testcases to enable LPA2 testing") Signed-off-by: Dev Jain <dev.jain@arm.com> Reviewed-by: Ryan Roberts <ryan.roberts@arm.com> Cc: Anshuman Khandual <anshuman.khandual@arm.com> Cc: Mark Brown <broonie@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Li Zhijian [Tue, 23 Jul 2024 06:44:28 +0000 (14:44 +0800)]
mm/page_alloc: fix pcp->count race between drain_pages_zone() vs __rmqueue_pcplist()
It's expected that no page should be left in pcp_list after calling
zone_pcp_disable() in offline_pages(). Previously, it's observed that
offline_pages() gets stuck [1] due to some pages remaining in pcp_list.
Cause:
There is a race condition between drain_pages_zone() and __rmqueue_pcplist()
involving the pcp->count variable. See below scenario:
CPU0 CPU1
---------------- ---------------
spin_lock(&pcp->lock);
__rmqueue_pcplist() {
zone_pcp_disable() {
/* list is empty */
if (list_empty(list)) {
/* add pages to pcp_list */
alloced = rmqueue_bulk()
mutex_lock(&pcp_batch_high_lock)
...
__drain_all_pages() {
drain_pages_zone() {
/* read pcp->count, it's 0 here */
count = READ_ONCE(pcp->count)
/* 0 means nothing to drain */
/* update pcp->count */
pcp->count += alloced << order;
...
...
spin_unlock(&pcp->lock);
In this case, after calling zone_pcp_disable() though, there are still some
pages in pcp_list. And these pages in pcp_list are neither movable nor
isolated, offline_pages() gets stuck as a result.
Solution:
Expand the scope of the pcp->lock to also protect pcp->count in
drain_pages_zone(), to ensure no pages are left in the pcp list after
zone_pcp_disable()
Roman Gushchin [Tue, 23 Jul 2024 17:12:44 +0000 (17:12 +0000)]
mm: memcg: add cacheline padding after lruvec in mem_cgroup_per_node
Oliver Sand reported a performance regression caused by commit 98c9daf5ae6b ("mm: memcg: guard memcg1-specific members of struct
mem_cgroup_per_node"), which puts some fields of the mem_cgroup_per_node
structure under the CONFIG_MEMCG_V1 config option. Apparently it causes a
false cache sharing between lruvec and lru_zone_size members of the
structure. Fix it by adding an explicit padding after the lruvec member.
Even though the padding is not required with CONFIG_MEMCG_V1 set, it seems
like the introduced memory overhead is not significant enough to warrant
another divergence in the mem_cgroup_per_node layout, so the padding is
added unconditionally.
Link: https://lkml.kernel.org/r/20240723171244.747521-1-roman.gushchin@linux.dev Fixes: 98c9daf5ae6b ("mm: memcg: guard memcg1-specific members of struct mem_cgroup_per_node") Signed-off-by: Roman Gushchin <roman.gushchin@linux.dev> Reported-by: kernel test robot <oliver.sang@intel.com> Closes: https://lore.kernel.org/oe-lkp/202407121335.31a10cb6-oliver.sang@intel.com Tested-by: Oliver Sang <oliver.sang@intel.com> Acked-by: Shakeel Butt <shakeel.butt@linux.dev> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Michal Hocko <mhocko@kernel.org> Cc: Muchun Song <muchun.song@linux.dev> Cc: Roman Gushchin <roman.gushchin@linux.dev> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
alloc_tag: outline and export free_reserved_page()
Outline and export free_reserved_page() because modules use it and it in
turn uses page_ext_{get|put} which should not be exported. The same
result could be obtained by outlining {get|put}_page_tag_ref() but that
would have higher performance impact as these functions are used in more
performance critical paths.
Link: https://lkml.kernel.org/r/20240717212844.2749975-1-surenb@google.com Fixes: dcfe378c81f7 ("lib: introduce support for page allocation tagging") Signed-off-by: Suren Baghdasaryan <surenb@google.com> Reported-by: kernel test robot <lkp@intel.com> Closes: https://lore.kernel.org/oe-kbuild-all/202407080044.DWMC9N9I-lkp@intel.com/ Suggested-by: Christoph Hellwig <hch@infradead.org> Suggested-by: Vlastimil Babka <vbabka@suse.cz> Acked-by: Vlastimil Babka <vbabka@suse.cz> Cc: Kees Cook <keescook@chromium.org> Cc: Kent Overstreet <kent.overstreet@linux.dev> Cc: Pasha Tatashin <pasha.tatashin@soleen.com> Cc: Sourav Panda <souravpanda@google.com> Cc: <stable@vger.kernel.org> [6.10] Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The decompression code parses a huffman tree and counts the number of
symbols for a given bit length. In rare cases, there may be >= 256
symbols with a given bit length, causing the unsigned char to overflow.
This causes a decompression failure later when the code tries and fails to
find the bit length for a given symbol.
Since the maximum number of symbols is 258, use unsigned short instead.
Link: https://lkml.kernel.org/r/20240717162016.1514077-1-ross.lagerwall@citrix.com Fixes: bc22c17e12c1 ("bzip2/lzma: library support for gzip, bzip2 and lzma decompression") Signed-off-by: Ross Lagerwall <ross.lagerwall@citrix.com> Cc: Alain Knaff <alain@knaff.lu> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
projects / thirdparty / linux.git / log
