Checkpoint/Restore in Userspace (CRIU) requires to reconstruct posix timers
with the same timer ID on restore. It uses sys_timer_create() and relies on
the monotonic increasing timer ID provided by this syscall. It creates and
deletes timers until the desired ID is reached. This is can loop for a long
time, when the checkpointed process had a very sparse timer ID range.
It has been debated to implement a new syscall to allow the creation of
timers with a given timer ID, but that's tideous due to the 32/64bit compat
issues of sigevent_t and of dubious value.
The restore mechanism of CRIU creates the timers in a state where all
threads of the restored process are held on a barrier and cannot issue
syscalls. That means the restorer task has exclusive control.
This allows to address this issue with a prctl() so that the restorer
thread can do:
if (prctl(PR_TIMER_CREATE_RESTORE_IDS, PR_TIMER_CREATE_RESTORE_IDS_ON))
goto linear_mode;
create_timers_with_explicit_ids();
prctl(PR_TIMER_CREATE_RESTORE_IDS, PR_TIMER_CREATE_RESTORE_IDS_OFF);
This is backwards compatible because the prctl() fails on older kernels and
CRIU can fall back to the linear timer ID mechanism. CRIU versions which do
not know about the prctl() just work as before.
Implement the prctl() and modify timer_create() so that it copies the
requested timer ID from userspace by utilizing the existing timer_t
pointer, which is used to copy out the allocated timer ID on success.
If the prctl() is disabled, which it is by default, timer_create() works as
before and does not try to read from the userspace pointer.
There is no problem when a broken or rogue user space application enables
the prctl(). If the user space pointer does not contain a valid ID, then
timer_create() fails. If the data is not initialized, but constains a
random valid ID, timer_create() will create that random timer ID or fail if
the ID is already given out.
As CRIU must use the raw syscall to avoid manipulating the internal state
of the restored process, this has no library dependencies and can be
adopted by CRIU right away.
Recreating two timers with IDs 1000000 and 2000000 takes 1.5 seconds with
the create/delete method. With the prctl() it takes 3 microseconds.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Cyrill Gorcunov <gorcunov@gmail.com>
Tested-by: Cyrill Gorcunov <gorcunov@gmail.com>
Link: https://lore.kernel.org/all/87jz8vz0en.ffs@tglx
The readout of /proc/$PID/timers holds sighand::siglock with interrupts
disabled. That is required to protect against concurrent modifications of
the task::signal::posix_timers list because the list is not RCU safe.
With the conversion of the timer storage to a RCU protected hlist, this is
not longer required.
The only requirement is to protect the returned entry against a concurrent
free, which is trivial as the timers are RCU protected.
Removing the trylock of sighand::siglock is benign because the life time of
task_struct::signal is bound to the life time of the task_struct itself.
There are two scenarios where this matters:
1) The process is life and not about to be checkpointed
2) The process is stopped via ptrace for checkpointing
#1 is a racy snapshot of the armed timers and nothing can rely on it. It's
not more than debug information and it has been that way before because
sighand lock is dropped when the buffer is full and the restart of
the iteration might find a completely different set of timers.
The task and therefore task::signal cannot be freed as timers_start()
acquired a reference count via get_pid_task().
#2 the process is stopped for checkpointing so nothing can delete or create
timers at this point. Neither can the process exit during the traversal.
If CRIU fails to observe an exit in progress prior to the dissimination
of the timers, then there are more severe problems to solve in the CRIU
mechanics as they can't rely on posix timers being enabled in the first
place.
Therefore replace the lock acquisition with rcu_read_lock() and switch the
timer storage traversal over to seq_hlist_*_rcu().
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/all/20250308155624.465175807@linutronix.de
struct k_itimer has the hlist_node, which is used for lookup in the hash
bucket, and the timer lock in the same cache line.
That's obviously bad, if one CPU fiddles with a timer and the other is
walking the hash bucket on which that timer is queued.
Avoid this by restructuring struct k_itimer, so that the read mostly (only
modified during setup and teardown) fields are in the first cache line and
the lock and the rest of the fields which get written to are in cacheline
2-N.
Reduces cacheline contention in a test case of 64 processes creating and
accessing 20000 timers each by almost 30% according to perf.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/all/20250308155624.341108067@linutronix.de
The hash distribution of hash_32() is suboptimal. jhash32() provides a way
better distribution, which evens out the length of the hash bucket lists,
which in turn avoids large outliers in list walk times.
Due to the sparse ID space (thanks CRIU) there is no guarantee that the
timers will be fully evenly distributed over the hash buckets, but the
behaviour is way better than with hash_32() even for randomly sparse ID
spaces.
For a pathological test case with 64 processes creating and accessing
20000 timers each, this results in a runtime reduction of ~10% and a
significantly reduced runtime variation.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/all/20250308155624.279080328@linutronix.de
Eric and Ben reported a significant performance bottleneck on the global
hash, which is used to store posix timers for lookup.
Eric tried to do a lockless validation of a new timer ID before trying to
insert the timer, but that does not solve the problem.
For the non-contended case this is a pointless exercise and for the
contended case this extra lookup just creates enough interleaving that all
tasks can make progress.
There are actually two real solutions to the problem:
1) Provide a per process (signal struct) xarray storage
2) Implement a smarter hash like the one in the futex code
#1 works perfectly fine for most cases, but the fact that CRIU enforced a
linear increasing timer ID to restore timers makes this problematic.
It's easy enough to create a sparse timer ID space, which amounts very
fast to a large junk of memory consumed for the xarray. 2048 timers with
a ID offset of 512 consume more than one megabyte of memory for the
xarray storage.
#2 The main advantage of the futex hash is that it uses per hash bucket
locks instead of a global hash lock. Aside of that it is scaled
according to the number of CPUs at boot time.
Experiments with artifical benchmarks have shown that a scaled hash with
per bucket locks comes pretty close to the xarray performance and in some
scenarios it performes better.
Test 1:
A single process creates 20000 timers and afterwards invokes
timer_getoverrun(2) on each of them:
mainline Eric newhash xarray
create 23 ms 23 ms 9 ms 8 ms
getoverrun 14 ms 14 ms 5 ms 4 ms
Test 2:
A single process creates 50000 timers and afterwards invokes
timer_getoverrun(2) on each of them:
mainline Eric newhash xarray
create 98 ms 219 ms 20 ms 18 ms
getoverrun 62 ms 62 ms 10 ms 9 ms
Test 3:
A single process creates 100000 timers and afterwards invokes
timer_getoverrun(2) on each of them:
mainline Eric newhash xarray
create 313 ms 750 ms 48 ms 33 ms
getoverrun 261 ms 260 ms 20 ms 14 ms
Erics changes create quite some overhead in the create() path due to the
double list walk, as the main issue according to perf is the list walk
itself. With 100k timers each hash bucket contains ~200 timers, which in
the worst case need to be all inspected. The same problem applies for
getoverrun() where the lookup has to walk through the hash buckets to find
the timer it is looking for.
The scaled hash obviously reduces hash collisions and lock contention
significantly. This becomes more prominent with concurrency.
Test 4:
A process creates 63 threads and all threads wait on a barrier before
each instance creates 20000 timers and afterwards invokes
timer_getoverrun(2) on each of them. The threads are pinned on
seperate CPUs to achive maximum concurrency. The numbers are the
average times per thread:
mainline Eric newhash xarray
create 180239 ms 38599 ms 579 ms 813 ms
getoverrun 2645 ms 2642 ms 32 ms 7 ms
Test 5:
A process forks 63 times and all forks wait on a barrier before each
instance creates 20000 timers and afterwards invokes
timer_getoverrun(2) on each of them. The processes are pinned on
seperate CPUs to achive maximum concurrency. The numbers are the
average times per process:
mainline eric newhash xarray
create 157253 ms 40008 ms 83 ms 60 ms
getoverrun 2611 ms 2614 ms 40 ms 4 ms
So clearly the reduction of lock contention with Eric's changes makes a
significant difference for the create() loop, but it does not mitigate the
problem of long list walks, which is clearly visible on the getoverrun()
side because that is purely dominated by the lookup itself. Once the timer
is found, the syscall just reads from the timer structure with no other
locks or code paths involved and returns.
The reason for the difference between the thread and the fork case for the
new hash and the xarray is that both suffer from contention on
sighand::siglock and the xarray suffers additionally from contention on the
xarray lock on insertion.
The only case where the reworked hash slighly outperforms the xarray is a
tight loop which creates and deletes timers.
Test 4:
A process creates 63 threads and all threads wait on a barrier before
each instance runs a loop which creates and deletes a timer 100000
times in a row. The threads are pinned on seperate CPUs to achive
maximum concurrency. The numbers are the average times per thread:
mainline Eric newhash xarray
loop 5917 ms 5897 ms 5473 ms 7846 ms
Test 5:
A process forks 63 times and all forks wait on a barrier before each
each instance runs a loop which creates and deletes a timer 100000
times in a row. The processes are pinned on seperate CPUs to achive
maximum concurrency. The numbers are the average times per process:
mainline Eric newhash xarray
loop 5137 ms 7828 ms 891 ms 872 ms
In both test there is not much contention on the hash, but the ucount
accounting for the signal and in the thread case the sighand::siglock
contention (plus the xarray locking) contribute dominantly to the overhead.
As the memory consumption of the xarray in the sparse ID case is
significant, the scaled hash with per bucket locks seems to be the better
overall option. While the xarray has faster lookup times for a large number
of timers, the actual syscall usage, which requires the lookup is not an
extreme hotpath. Most applications utilize signal delivery and all syscalls
except timer_getoverrun(2) are all but cheap.
So implement a scaled hash with per bucket locks, which offers the best
tradeoff between performance and memory consumption.
Reported-by: Eric Dumazet <edumazet@google.com>
Reported-by: Benjamin Segall <bsegall@google.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/all/20250308155624.216091571@linutronix.de
The global hash_lock protecting the posix timer hash table can be heavily
contended especially when there is an extensive linear search for a timer
ID.
Timer IDs are handed out by monotonically increasing next_posix_timer_id
and then validating that there is no timer with the same ID in the hash
table. Both operations happen with the global hash lock held.
To reduce the hash lock contention the hash will be reworked to a scaled
hash with per bucket locks, which requires to handle the ID counter
lockless.
Prepare for this by making next_posix_timer_id an atomic_t, which can be
used lockless with atomic_inc_return().
[ tglx: Adopted from Eric's series, massaged change log and simplified it ]
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Acked-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/all/20250219125522.2535263-2-edumazet@google.com
Link: https://lore.kernel.org/all/20250308155624.151545978@linutronix.de
The lookup and locking of posix timers requires the same repeating pattern
at all usage sites:
tmr = lock_timer(tiner_id);
if (!tmr)
return -EINVAL;
....
unlock_timer(tmr);
Solve this with a guard implementation, which works in most places out of
the box except for those, which need to unlock the timer inside the guard
scope.
Though the only places where this matters are timer_delete() and
timer_settime(). In both cases the timer pointer needs to be preserved
across the end of the scope, which is solved by storing the pointer in a
variable outside of the scope.
timer_settime() also has to protect the timer with RCU before unlocking,
which obviously can't use guard(rcu) before leaving the guard scope as that
guard is cleaned up before the unlock. Solve this by providing the RCU
protection open coded.
[ tglx: Made it work and added change log ]
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/all/20250224162103.GD11590@noisy.programming.kicks-ass.net
Link: https://lore.kernel.org/all/20250308155624.087465658@linutronix.de
sys_timer_delete() and the do_exit() cleanup function itimer_delete() are
doing the same thing, but have needlessly different implementations instead
of sharing the code.
The other oddity of timer deletion is the fact that the timer is not
invalidated before the actual deletion happens, which allows concurrent
lookups to succeed.
That's wrong because a timer which is in the process of being deleted
should not be visible and any actions like signal queueing, delivery and
rearming should not happen once the task, which invoked timer_delete(), has
the timer locked.
Rework the code so that:
1) The signal queueing and delivery code ignore timers which are marked
invalid
2) The deletion implementation between sys_timer_delete() and
itimer_delete() is shared
3) The timer is invalidated and removed from the linked lists before
the deletion callback of the relevant clock is invoked.
That requires to rework timer_wait_running() as it does a lookup of
the timer when relocking it at the end. In case of deletion this
lookup would fail due to the preceding invalidation and the wait loop
would terminate prematurely.
But due to the preceding invalidation the timer cannot be accessed by
other tasks anymore, so there is no way that the timer has been freed
after the timer lock has been dropped.
Move the re-validation out of timer_wait_running() and handle it at
the only other usage site, timer_settime().
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/all/87zfht1exf.ffs@tglx
A timer is only valid in the hashtable when both timer::it_signal and
timer::it_id are set to their final values, but timers are added without
those values being set.
The timer ID is allocated when the timer is added to the hash in invalid
state. The ID is taken from a monotonically increasing per process counter
which wraps around after reaching INT_MAX. The hash insertion validates
that there is no timer with the allocated ID in the hash table which
belongs to the same process. That opens a mostly theoretical race condition:
If other threads of the same process manage to create/delete timers in
rapid succession before the newly created timer is fully initialized and
wrap around to the timer ID which was handed out, then a duplicate timer ID
will be inserted into the hash table.
Prevent this by:
1) Setting timer::it_id before inserting the timer into the hashtable.
2) Storing the signal pointer in timer::it_signal with bit 0 set before
inserting it into the hashtable.
Bit 0 acts as a invalid bit, which means that the regular lookup for
sys_timer_*() will fail the comparison with the signal pointer.
But the lookup on insertion masks out bit 0 and can therefore detect a
timer which is not yet valid, but allocated in the hash table. Bit 0
in the pointer is cleared once the initialization of the timer
completed.
[ tglx: Fold ID and signal iniitializaion into one patch and massage change
log and comments. ]
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/all/20250219125522.2535263-3-edumazet@google.com
Link: https://lore.kernel.org/all/20250308155623.572035178@linutronix.de
Frederic pointed out that the memory operations to initialize the timer are
not guaranteed to be visible, when __lock_timer() observes timer::it_signal
valid under timer::it_lock:
T0 T1
--------- -----------
do_timer_create()
// A
new_timer->.... = ....
spin_lock(current->sighand)
// B
WRITE_ONCE(new_timer->it_signal, current->signal)
spin_unlock(current->sighand)
sys_timer_*()
t = __lock_timer()
spin_lock(&timr->it_lock)
// observes B
if (timr->it_signal == current->signal)
return timr;
if (!t)
return;
// Is not guaranteed to observe A
Protect the write of timer::it_signal, which makes the timer valid, with
timer::it_lock as well. This guarantees that T1 must observe the
initialization A completely, when it observes the valid signal pointer
under timer::it_lock. sighand::siglock must still be taken to protect the
signal::posix_timers list.
Reported-by: Frederic Weisbecker <frederic@kernel.org>
Suggested-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/all/20250308155623.507944489@linutronix.de
The size argument of strscpy() is only required when the destination
pointer is not a fixed sized array or when the copy needs to be smaller
than the size of the fixed sized destination array.
For fixed sized destination arrays and full copies, strscpy() automatically
determines the length of the destination buffer if the size argument is
omitted.
This makes the explicit sizeof() unnecessary. Remove it.
[ tglx: Massaged change log ]
Signed-off-by: Thorsten Blum <thorsten.blum@linux.dev>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/all/20250311110624.495718-2-thorsten.blum@linux.dev
This reverts commit f590308536 ("timer debug: Hide kernel addresses via
%pK in /proc/timer_list")
The timer list helper SEQ_printf() uses either the real seq_printf() for
procfs output or vprintk() to print to the kernel log, when invoked from
SysRq-q. It uses %pK for printing pointers.
In the past %pK was prefered over %p as it would not leak raw pointer
values into the kernel log. Since commit ad67b74d24 ("printk: hash
addresses printed with %p") the regular %p has been improved to avoid this
issue.
Furthermore, restricted pointers ("%pK") were never meant to be used
through printk(). They can still unintentionally leak raw pointers or
acquire sleeping looks in atomic contexts.
Switch to the regular pointer formatting which is safer, easier to reason
about and sufficient here.
Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/lkml/20250113171731-dc10e3c1-da64-4af0-b767-7c7070468023@linutronix.de/
Link: https://lore.kernel.org/all/20250311-restricted-pointers-timer-v1-1-6626b91e54ab@linutronix.de
exit_itimers() loops through every timer in the process to delete it. This
requires taking the system-wide hash_lock for each of these timers, and
contends with other processes trying to create or delete timers.
When a process creates hundreds of thousands of timers, and then exits
while other processes contend with it, this can trigger softlockups on
CONFIG_PREEMPT=n.
Add a cond_resched() invocation into the loop to allow the system to make
progress.
Signed-off-by: Ben Segall <bsegall@google.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/all/xm2634gg2n23.fsf@google.com
Clang and GCC complain about overlapped initialisers in the
hrtimer_clock_to_base_table definition. With `make W=1` and CONFIG_WERROR=y
(which is default nowadays) this breaks the build:
CC kernel/time/hrtimer.o
kernel/time/hrtimer.c:124:21: error: initializer overrides prior initialization of this subobject [-Werror,-Winitializer-overrides]
124 | [CLOCK_REALTIME] = HRTIMER_BASE_REALTIME,
kernel/time/hrtimer.c:122:27: note: previous initialization is here
122 | [0 ... MAX_CLOCKS - 1] = HRTIMER_MAX_CLOCK_BASES,
(and similar for CLOCK_MONOTONIC, CLOCK_BOOTTIME, and CLOCK_TAI).
hrtimer_clockid_to_base(), which uses the table, is only used in
__hrtimer_init(), which is not a hotpath.
Therefore replace the table lookup with a switch case in
hrtimer_clockid_to_base() to avoid this warning.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/all/20250214134424.3367619-1-andriy.shevchenko@linux.intel.com
The sanity checks in hrtimer_update_function() are expensive for high
frequency usage like in the io/uring code due to locking.
Hide the sanity checks behind CONFIG_PROVE_LOCKING, which has a decent
chance to be enabled on a regular basis for testing.
Fixes: 8f02e3563b ("hrtimers: Introduce hrtimer_update_function()")
Reported-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/all/87ikpllali.ffs@tglx
Pull Kbuild fixes from Masahiro Yamada:
- Suppress false-positive -Wformat-{overflow,truncation}-non-kprintf
warnings regardless of the W= option
- Avoid CONFIG_TRIM_UNUSED_KSYMS dropping symbols passed to symbol_get()
- Fix a build regression of the Debian linux-headers package
* tag 'kbuild-fixes-v6.14' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
kbuild: install-extmod-build: add missing quotation marks for CC variable
kbuild: fix misspelling in scripts/Makefile.lib
kbuild: keep symbols for symbol_get() even with CONFIG_TRIM_UNUSED_KSYMS
scripts/Makefile.extrawarn: Do not show clang's non-kprintf warnings at W=1
Pull power management fix from Rafael Wysocki:
"Fix a recently introduced kernel crash due to a NULL pointer
dereference during system-wide suspend (Rafael Wysocki)"
* tag 'pm-6.14-rc2-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
PM: sleep: core: Restrict power.set_active propagation
Pull kvm fixes from Paolo Bonzini:
"ARM:
- Correctly clean the BSS to the PoC before allowing EL2 to access it
on nVHE/hVHE/protected configurations
- Propagate ownership of debug registers in protected mode after the
rework that landed in 6.14-rc1
- Stop pretending that we can run the protected mode without a GICv3
being present on the host
- Fix a use-after-free situation that can occur if a vcpu fails to
initialise the NV shadow S2 MMU contexts
- Always evaluate the need to arm a background timer for fully
emulated guest timers
- Fix the emulation of EL1 timers in the absence of FEAT_ECV
- Correctly handle the EL2 virtual timer, specially when HCR_EL2.E2H==0
s390:
- move some of the guest page table (gmap) logic into KVM itself,
inching towards the final goal of completely removing gmap from the
non-kvm memory management code.
As an initial set of cleanups, move some code from mm/gmap into kvm
and start using __kvm_faultin_pfn() to fault-in pages as needed;
but especially stop abusing page->index and page->lru to aid in the
pgdesc conversion.
x86:
- Add missing check in the fix to defer starting the huge page
recovery vhost_task
- SRSO_USER_KERNEL_NO does not need SYNTHESIZED_F"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (31 commits)
KVM: x86/mmu: Ensure NX huge page recovery thread is alive before waking
KVM: remove kvm_arch_post_init_vm
KVM: selftests: Fix spelling mistake "initally" -> "initially"
kvm: x86: SRSO_USER_KERNEL_NO is not synthesized
KVM: arm64: timer: Don't adjust the EL2 virtual timer offset
KVM: arm64: timer: Correctly handle EL1 timer emulation when !FEAT_ECV
KVM: arm64: timer: Always evaluate the need for a soft timer
KVM: arm64: Fix nested S2 MMU structures reallocation
KVM: arm64: Fail protected mode init if no vgic hardware is present
KVM: arm64: Flush/sync debug state in protected mode
KVM: s390: selftests: Streamline uc_skey test to issue iske after sske
KVM: s390: remove the last user of page->index
KVM: s390: move PGSTE softbits
KVM: s390: remove useless page->index usage
KVM: s390: move gmap_shadow_pgt_lookup() into kvm
KVM: s390: stop using lists to keep track of used dat tables
KVM: s390: stop using page->index for non-shadow gmaps
KVM: s390: move some gmap shadowing functions away from mm/gmap.c
KVM: s390: get rid of gmap_translate()
KVM: s390: get rid of gmap_fault()
...
Commit 3775fc538f ("PM: sleep: core: Synchronize runtime PM status of
parents and children") exposed an issue related to simple_pm_bus_pm_ops
that uses pm_runtime_force_suspend() and pm_runtime_force_resume() as
bus type PM callbacks for the noirq phases of system-wide suspend and
resume.
The problem is that pm_runtime_force_suspend() does not distinguish
runtime-suspended devices from devices for which runtime PM has never
been enabled, so if it sees a device with runtime PM status set to
RPM_ACTIVE, it will assume that runtime PM is enabled for that device
and so it will attempt to suspend it with the help of its runtime PM
callbacks which may not be ready for that. As it turns out, this
causes simple_pm_bus_runtime_suspend() to crash due to a NULL pointer
dereference.
Another problem related to the above commit and simple_pm_bus_pm_ops is
that setting runtime PM status of a device handled by the latter to
RPM_ACTIVE will actually prevent it from being resumed because
pm_runtime_force_resume() only resumes devices with runtime PM status
set to RPM_SUSPENDED.
To mitigate these issues, do not allow power.set_active to propagate
beyond the parent of the device with DPM_FLAG_SMART_SUSPEND set that
will need to be resumed, which should be a sufficient stop-gap for the
time being, but they will need to be properly addressed in the future
because in general during system-wide resume it is necessary to resume
all devices in a dependency chain in which at least one device is going
to be resumed.
Fixes: 3775fc538f ("PM: sleep: core: Synchronize runtime PM status of parents and children")
Closes: https://lore.kernel.org/linux-pm/1c2433d4-7e0f-4395-b841-b8eac7c25651@nvidia.com/
Reported-by: Jon Hunter <jonathanh@nvidia.com>
Tested-by: Johan Hovold <johan+linaro@kernel.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Link: https://patch.msgid.link/6137505.lOV4Wx5bFT@rjwysocki.net
Pull hardening fixes from Kees Cook:
"Address a KUnit stack initialization regression that got tickled on
m68k, and solve a Clang(v14 and earlier) bug found by 0day:
- Fix stackinit KUnit regression on m68k
- Use ARRAY_SIZE() for memtostr*()/strtomem*()"
* tag 'hardening-v6.14-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux:
string.h: Use ARRAY_SIZE() for memtostr*()/strtomem*()
compiler.h: Introduce __must_be_byte_array()
compiler.h: Move C string helpers into C-only kernel section
stackinit: Fix comment for test_small_end
stackinit: Keep selftest union size small on m68k
Pull seccomp fix from Kees Cook:
"This is really a work-around for x86_64 having grown a syscall to
implement uretprobe, which has caused problems since v6.11.
This may change in the future, but for now, this fixes the unintended
seccomp filtering when uretprobe switched away from traps, and does so
with something that should be easy to backport.
- Allow uretprobe on x86_64 to avoid behavioral complications (Eyal
Birger)"
* tag 'seccomp-v6.14-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux:
selftests/seccomp: validate uretprobe syscall passes through seccomp
seccomp: passthrough uretprobe systemcall without filtering
Pull execve fix from Kees Cook:
"This is an alpha-specific fix, but since it touched ELF I was asked to
carry it.
- alpha/elf: Fix misc/setarch test of util-linux by removing 32bit
support (Eric W. Biederman)"
* tag 'execve-v6.14-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux:
alpha/elf: Fix misc/setarch test of util-linux by removing 32bit support
Pull SCSI fixes from James Bottomley:
"A number of fairly small fixes, mostly in drivers but two in the core
to change a retry for depopulation (a trendy new hdd thing that
reorganizes blocks away from failing elements) and one to fix a GFP_
annotation to avoid a lock dependency (the third core patch is all in
testing)"
* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
scsi: qla1280: Fix kernel oops when debug level > 2
scsi: ufs: core: Fix error return with query response
scsi: storvsc: Set correct data length for sending SCSI command without payload
scsi: ufs: core: Fix use-after free in init error and remove paths
scsi: core: Do not retry I/Os during depopulation
scsi: core: Use GFP_NOIO to avoid circular locking dependency
scsi: ufs: Fix toggling of clk_gating.state when clock gating is not allowed
scsi: ufs: core: Ensure clk_gating.lock is used only after initialization
scsi: ufs: core: Simplify temperature exception event handling
scsi: target: core: Add line break to status show
scsi: ufs: core: Fix the HIGH/LOW_TEMP Bit Definitions
scsi: core: Add passthrough tests for success and no failure definitions
Pull i2c reverts from Wolfram Sang:
"It turned out the new mechanism for handling created devices does not
handle all muxing cases.
Revert the changes to give a proper solution more time"
* tag 'i2c-for-6.14-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux:
Revert "i2c: Replace list-based mechanism for handling auto-detected clients"
Revert "i2c: Replace list-based mechanism for handling userspace-created clients"
Pull rust fixes from Miguel Ojeda:
- Do not export KASAN ODR symbols to avoid gendwarfksyms warnings
- Fix future Rust 1.86.0 (to be released 2025-04-03) x86_64 builds
- Clean future Rust 1.86.0 (to be released 2025-04-03) warning
- Fix future GCC 15 (to be released in a few months) builds
- Fix `rusttest` target in macOS
* tag 'rust-fixes-6.14' of https://github.com/Rust-for-Linux/linux:
x86: rust: set rustc-abi=x86-softfloat on rustc>=1.86.0
rust: kbuild: do not export generated KASAN ODR symbols
rust: kbuild: add -fzero-init-padding-bits to bindgen_skip_cflags
rust: init: use explicit ABI to clean warning in future compilers
rust: kbuild: use host dylib naming in rusttestlib-kernel
Pull ftrace fix from Steven Rostedt:
"Function graph fix of notrace functions.
When the function graph tracer was restructured to use the global
section of the meta data in the shadow stack, the bit logic was
changed. There's a TRACE_GRAPH_NOTRACE_BIT that is the bit number in
the mask that tells if the function graph tracer is currently in the
"notrace" mode. The TRACE_GRAPH_NOTRACE is the mask with that bit set.
But when the code we restructured, the TRACE_GRAPH_NOTRACE_BIT was
used when it should have been the TRACE_GRAPH_NOTRACE mask. This made
notrace not work properly"
* tag 'ftrace-v6.14-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:
fgraph: Fix set_graph_notrace with setting TRACE_GRAPH_NOTRACE_BIT
Pull x86 fix from Ingo Molnar:
"Fix a build regression on GCC 15 builds, caused by GCC changing the
default C version that is overriden in the main Makefile but not in
the x86 boot code Makefile"
* tag 'x86-urgent-2025-02-08' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/boot: Use '-std=gnu11' to fix build with GCC 15
Pull timer fixes from Ingo Molnar:
"Fix a PREEMPT_RT bug in the clocksource verification code that caused
false positive warnings.
Also fix a timer migration setup bug when new CPUs are added"
* tag 'timers-urgent-2025-02-08' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
timers/migration: Fix off-by-one root mis-connection
clocksource: Use migrate_disable() to avoid calling get_random_u32() in atomic context
Pull scheduler fixes from Ingo Molnar:
"Fix a cfs_rq->h_nr_runnable accounting bug that trips up a defensive
SCHED_WARN_ON() on certain workloads. The bug is believed to be
(accidentally) self-correcting, hence no behavioral side effects are
expected.
Also print se.slice in debug output, since this value can now be set
via the syscall ABI and can be useful to track"
* tag 'sched-urgent-2025-02-08' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched/debug: Provide slice length for fair tasks
sched/fair: Fix inaccurate h_nr_runnable accounting with delayed dequeue
Pull irq fix from Ingo Molnar:
"Another followup fix for the procps genirq output formatting
regression caused by an optimization"
* tag 'irq-urgent-2025-02-08' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
genirq: Remove leading space from irq_chip::irq_print_chip() callbacks
Pull locking fix from Ingo Molnar:
"Fix a dangling pointer bug in the futex code used by the uring code.
It isn't causing problems at the moment due to uring ABI limitations
leaving it essentially unused in current usages, but is a good idea to
fix nevertheless"
* tag 'locking-urgent-2025-02-08' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
futex: Pass in task to futex_queue()
The code was restructured where the function graph notrace code, that
would not trace a function and all its children is done by setting a
NOTRACE flag when the function that is not to be traced is hit.
There's a TRACE_GRAPH_NOTRACE_BIT which defines the bit in the flags and a
TRACE_GRAPH_NOTRACE which is the mask with that bit set. But the
restructuring used TRACE_GRAPH_NOTRACE_BIT when it should have used
TRACE_GRAPH_NOTRACE.
For example:
# cd /sys/kernel/tracing
# echo set_track_prepare stack_trace_save > set_graph_notrace
# echo function_graph > current_tracer
# cat trace
[..]
0) | __slab_free() {
0) | free_to_partial_list() {
0) | arch_stack_walk() {
0) | __unwind_start() {
0) 0.501 us | get_stack_info();
Where a non filter trace looks like:
# echo > set_graph_notrace
# cat trace
0) | free_to_partial_list() {
0) | set_track_prepare() {
0) | stack_trace_save() {
0) | arch_stack_walk() {
0) | __unwind_start() {
Where the filter should look like:
# cat trace
0) | free_to_partial_list() {
0) | _raw_spin_lock_irqsave() {
0) 0.350 us | preempt_count_add();
0) 0.351 us | do_raw_spin_lock();
0) 2.440 us | }
Cc: stable@vger.kernel.org
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Link: https://lore.kernel.org/20250208001511.535be150@batman.local.home
Fixes: b84214890a ("function_graph: Move graph notrace bit to shadow stack global var")
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
-Wenum-enum-conversion was strengthened in clang-19 to warn for C, which
caused the kernel to move it to W=1 in commit 75b5ab134b ("kbuild:
Move -Wenum-{compare-conditional,enum-conversion} into W=1") because
there were numerous instances that would break builds with -Werror.
Unfortunately, this is not a full solution, as more and more developers,
subsystems, and distributors are building with W=1 as well, so they
continue to see the numerous instances of this warning.
Since the move to W=1, there have not been many new instances that have
appeared through various build reports and the ones that have appeared
seem to be following similar existing patterns, suggesting that most
instances of this warning will not be real issues. The only alternatives
for silencing this warning are adding casts (which is generally seen as
an ugly practice) or refactoring the enums to macro defines or a unified
enum (which may be undesirable because of type safety in other parts of
the code).
Move the warning to W=2, where warnings that occur frequently but may be
relevant should reside.
Cc: stable@vger.kernel.org
Fixes: 75b5ab134b ("kbuild: Move -Wenum-{compare-conditional,enum-conversion} into W=1")
Link: https://lore.kernel.org/ZwRA9SOcOjjLJcpi@google.com/
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull smb client fixes from Steve French:
- Three DFS fixes: DFS mount fix, fix for noisy log msg and one to
remove some unused code
- SMB3 Lease fix
* tag 'v6.14rc1-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6:
smb: client: change lease epoch type from unsigned int to __u16
smb: client: get rid of kstrdup() in get_ses_refpath()
smb: client: fix noisy when tree connecting to DFS interlink targets
smb: client: don't trust DFSREF_STORAGE_SERVER bit
Pull drm fixes from Dave Airlie:
"Just regular drm fixes, amdgpu, xe and i915 mostly, but a few
scattered fixes. I think one of the i915 fixes fixes some build combos
that Guenter was seeing.
amdgpu:
- Add new tiling flag for DCC write compress disable
- Add BO metadata flag for DCC
- Fix potential out of bounds access in display
- Seamless boot fix
- CONFIG_FRAME_WARN fix
- PSR1 fix
xe:
- OA uAPI related fixes
- Fix SRIOV migration initialization
- Restore devcoredump to a sane state
i915:
- Fix the build error with clamp after WARN_ON on gcc 13.x+
- HDCP related fixes
- PMU fix zero delta busyness issue
- Fix page cleanup on DMA remap failure
- Drop 64bpp YUV formats from ICL+ SDR planes
- GuC log related fix
- DisplayPort related fixes
ivpu:
- Fix error handling
komeda:
- add return check
zynqmp:
- fix locking in DP code
ast:
- fix AST DP timeout
cec:
- fix broken CEC adapter check"
* tag 'drm-fixes-2025-02-08' of https://gitlab.freedesktop.org/drm/kernel: (29 commits)
drm/i915/dp: Fix potential infinite loop in 128b/132b SST
Revert "drm/amd/display: Use HW lock mgr for PSR1"
drm/amd/display: Respect user's CONFIG_FRAME_WARN more for dml files
accel/amdxdna: Add MODULE_FIRMWARE() declarations
drm/i915/dp: Iterate DSC BPP from high to low on all platforms
drm/xe: Fix and re-enable xe_print_blob_ascii85()
drm/xe/devcoredump: Move exec queue snapshot to Contexts section
drm/xe/oa: Set stream->pollin in xe_oa_buffer_check_unlocked
drm/xe/pf: Fix migration initialization
drm/xe/oa: Preserve oa_ctrl unused bits
drm/amd/display: Fix seamless boot sequence
drm/amd/display: Fix out-of-bound accesses
drm/amdgpu: add a BO metadata flag to disable write compression for Vulkan
drm/i915/backlight: Return immediately when scale() finds invalid parameters
drm/i915/dp: Return min bpc supported by source instead of 0
drm/i915/dp: fix the Adaptive sync Operation mode for SDP
drm/i915/guc: Debug print LRC state entries only if the context is pinned
drm/i915: Drop 64bpp YUV formats from ICL+ SDR planes
drm/i915: Fix page cleanup on DMA remap failure
drm/i915/pmu: Fix zero delta busyness issue
...
Pull ibft fixes from Konrad Rzeszutek Wilk:
"Two tiny fixes to IBFT code: one for Kconfig and another for IPv6"
* tag 'stable/for-linus-6.14-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/ibft:
iscsi_ibft: Fix UBSAN shift-out-of-bounds warning in ibft_attr_show_nic()
firmware: iscsi_ibft: fix ISCSI_IBFT Kconfig entry
Pull block fixes from Jens Axboe:
- MD pull request via Song:
- fix an error handling path for md-linear
- NVMe pull request via Keith:
- Connection fixes for fibre channel transport (Daniel)
- Endian fixes (Keith, Christoph)
- Cleanup fix for host memory buffer (Francis)
- Platform specific power quirks (Georg)
- Target memory leak (Sagi)
- Use appropriate controller state accessor (Daniel)
- Fixup for a regression introduced last week, where sunvdc wasn't
updated for an API change, causing compilation failures on sparc64.
* tag 'block-6.14-20250207' of git://git.kernel.dk/linux:
drivers/block/sunvdc.c: update the correct AIP call
md: Fix linear_set_limits()
nvme-fc: use ctrl state getter
nvme: make nvme_tls_attrs_group static
nvmet: add a missing endianess conversion in nvmet_execute_admin_connect
nvmet: the result field in nvmet_alloc_ctrl_args is little endian
nvmet: fix a memory leak in controller identify
nvme-fc: do not ignore connectivity loss during connecting
nvme: handle connectivity loss in nvme_set_queue_count
nvme-fc: go straight to connecting state when initializing
nvme-pci: Add TUXEDO IBP Gen9 to Samsung sleep quirk
nvme-pci: Add TUXEDO InfinityFlex to Samsung sleep quirk
nvme-pci: remove redundant dma frees in hmb
nvmet: fix rw control endian access
While attempting to build a Debian packages with CC="ccache gcc", I
saw the following error as builddeb builds linux-headers-$KERNELVERSION:
make HOSTCC=ccache gcc VPATH= srcroot=. -f ./scripts/Makefile.build obj=debian/linux-headers-6.14.0-rc1/usr/src/linux-headers-6.14.0-rc1/scripts
make[6]: *** No rule to make target 'gcc'. Stop.
Upon investigation, it seems that one instance of $(CC) variable reference
in ./scripts/package/install-extmod-build was missing quotation marks,
causing the above error.
Add the missing quotation marks around $(CC) to fix build.
Fixes: 5f73e7d038 ("kbuild: refactor cross-compiling linux-headers package")
Co-developed-by: Mingcong Bai <jeffbai@aosc.io>
Signed-off-by: Mingcong Bai <jeffbai@aosc.io>
Tested-by: WangYuli <wangyuli@uniontech.com>
Signed-off-by: WangYuli <wangyuli@uniontech.com>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Pull power management fixes from Rafael Wysocki:
"These fix a handful of issues in the amd-pstate driver, the airoha
cpufreq driver build, a (recently added) possible NULL pointer
dereference in the cpufreq code and a possible memory leak in the
power capping subsystem:
- Fix cpufreq_policy reference counting and prevent max_perf from
going above the current limit in amd-pstate, and drop a redundant
goto label from it (Dhananjay Ugwekar)
- Prevent the per-policy boost_enabled flag in amd-pstate from
getting out of sync with the actual state after boot failures
(Lifeng Zheng)
- Fix a recently added possible NULL pointer dereference in the
cpufreq core (Aboorva Devarajan)
- Fix a build issue related to CONFIG_OF and COMPILE_TEST
dependencies in the airoha cpufreq driver (Arnd Bergmann)
- Fix a possible memory leak in the power capping subsystem (Joe
Hattori)"
* tag 'pm-6.14-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
cpufreq/amd-pstate: Fix cpufreq_policy ref counting
cpufreq: prevent NULL dereference in cpufreq_online()
cpufreq: airoha: modify CONFIG_OF dependency
cpufreq/amd-pstate: Fix max_perf updation with schedutil
cpufreq/amd-pstate: Remove the goto label in amd_pstate_update_limits
cpufreq/amd-pstate: Fix per-policy boost flag incorrect when fail
powercap: call put_device() on an error path in powercap_register_control_type()
Pull ACPI fixes from Rafael Wysocki:
"These fix three assorted issues, including one recent regression:
- Add an ACPI IRQ override quirk for Eluktronics MECH-17 to make the
internal keyboard work (Gannon Kolding)
- Make acpi_data_prop_read() reflect the OF counterpart behavior in
error cases (Andy Shevchenko)
- Remove recently added strict ACPI PRM handler address checks that
prevented PRM from working on some platforms in the field (Aubrey
Li)"
* tag 'acpi-6.14-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
ACPI: PRM: Remove unnecessary strict handler address checks
ACPI: resource: IRQ override for Eluktronics MECH-17
ACPI: property: Fix return value for nval == 0 in acpi_data_prop_read()
Pull gpio fixes from Bartosz Golaszewski:
- fix interrupt support in gpio-pca953x
- fix configfs attribute locking in gpio-sim
- limit the visibility of the GPIO_GRGPIO Kconfig symbol to OF systems
only
- update MAINTAINERS
* tag 'gpio-fixes-for-v6.14-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux:
MAINTAINERS: Use my kernel.org address for ACPI GPIO work
gpio: GPIO_GRGPIO should depend on OF
gpio: sim: lock hog configfs items if present
gpio: pca953x: Improve interrupt support
