Boot and suspend/resume should not be slowed down in kernels built with
CONFIG_RCU_LAZY=y. In particular, suspend can sometimes fail in such
kernels.
This commit therefore adds rcu_async_hurry(), rcu_async_relax(), and
rcu_async_should_hurry() functions that track whether or not either
a boot or a suspend/resume operation is in progress. This will
enable a later commit to refrain from laziness during those times.
Export rcu_async_should_hurry(), rcu_async_hurry(), and rcu_async_relax()
for later use by rcutorture.
[ paulmck: Apply feedback from Steve Rostedt. ]
Fixes: 3cb278e73b ("rcu: Make call_rcu() lazy to save power")
Signed-off-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Daniel Borkmann says:
====================
bpf 2023-01-16
We've added 6 non-merge commits during the last 8 day(s) which contain
a total of 6 files changed, 22 insertions(+), 24 deletions(-).
The main changes are:
1) Mitigate a Spectre v4 leak in unprivileged BPF from speculative
pointer-as-scalar type confusion, from Luis Gerhorst.
2) Fix a splat when pid 1 attaches a BPF program that attempts to
send killing signal to itself, from Hao Sun.
3) Fix BPF program ID information in BPF_AUDIT_UNLOAD as well as
PERF_BPF_EVENT_PROG_UNLOAD events, from Paul Moore.
4) Fix BPF verifier warning triggered from invalid kfunc call in
backtrack_insn, also from Hao Sun.
5) Fix potential deadlock in htab_lock_bucket from same bucket index
but different map_locked index, from Tonghao Zhang.
* tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:
bpf: Fix pointer-leak due to insufficient speculative store bypass mitigation
bpf: hash map, avoid deadlock with suitable hash mask
bpf: remove the do_idr_lock parameter from bpf_prog_free_id()
bpf: restore the ebpf program ID for BPF_AUDIT_UNLOAD and PERF_BPF_EVENT_PROG_UNLOAD
bpf: Skip task with pid=1 in send_signal_common()
bpf: Skip invalid kfunc call in backtrack_insn
====================
Link: https://lore.kernel.org/r/20230116230745.21742-1-daniel@iogearbox.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Currently we allow to load any tracing program as sleepable,
but BPF_TRACE_RAW_TP can't sleep. Making the check explicit
for tracing programs attach types, so sleepable BPF_TRACE_RAW_TP
will fail to load.
Updating the verifier error to mention iter programs as well.
Acked-by: Song Liu <song@kernel.org>
Acked-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/r/20230117223705.440975-1-jolsa@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
msi_create_device_irq_domain() creates a firmware node for the new domain,
which is never freed. kmemleak reports:
unreferenced object 0xffff888120ba9a00 (size 96):
comm "systemd-modules", pid 221, jiffies 4294893411 (age 635.732s)
hex dump (first 32 bytes):
00 00 00 00 00 00 00 00 e0 19 8b 83 ff ff ff ff ................
00 00 00 00 00 00 00 00 18 9a ba 20 81 88 ff ff ........... ....
backtrace:
[<000000008cdbc98d>] __irq_domain_alloc_fwnode+0x51/0x2b0
[<00000000c57acf9d>] msi_create_device_irq_domain+0x283/0x670
[<000000009b567982>] __pci_enable_msix_range+0x49e/0xdb0
[<0000000077cc1445>] pci_alloc_irq_vectors_affinity+0x11f/0x1c0
[<00000000532e9ef5>] mlx5_irq_table_create+0x24c/0x940 [mlx5_core]
[<00000000fabd2b80>] mlx5_load+0x1fa/0x680 [mlx5_core]
[<000000006bb22ae4>] mlx5_init_one+0x485/0x670 [mlx5_core]
[<00000000eaa5e1ad>] probe_one+0x4c2/0x720 [mlx5_core]
[<00000000df8efb43>] local_pci_probe+0xd6/0x170
[<0000000085cb9924>] pci_device_probe+0x231/0x6e0
Use the proper free operation for the firmware wnode so the name is freed
during error unwind of msi_create_device_irq_domain() and also free the
node in msi_remove_device_irq_domain() if it was automatically allocated.
To avoid extra NULL pointer checks make irq_domain_free_fwnode() tolerant
of NULL.
Fixes: 27a6dea3eb ("genirq/msi: Provide msi_create/free_device_irq_domain()")
Reported-by: Omri Barazi <obarazi@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Kalle Valo <kvalo@kernel.org>
Tested-by: Leon Romanovsky <leonro@nvidia.com>
Link: https://lore.kernel.org/r/0-v2-24af6665e2da+c9-msi_leak_jgg@nvidia.com
Map irq vector into group, which allows to abstract the algorithm for
a generic use case outside of the interrupt core.
Rename irq_build_affinity_masks as group_cpus_evenly, so the API can be
reused for blk-mq to make default queue mapping even though irq vectors
aren't involved.
No functional change, just rename vector as group.
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Link: https://lore.kernel.org/r/20221227022905.352674-5-ming.lei@redhat.com
Prepare for abstracting irq_build_affinity_masks() into a public function
for assigning all CPUs evenly into several groups.
Don't pass irq_affinity_desc array to irq_build_affinity_masks, instead
return a cpumask array by storing each assigned group into one element of
the array.
This allows to provide a generic interface for grouping all CPUs evenly
from a NUMA and CPU locality viewpoint, and the cost is one extra allocation
in irq_build_affinity_masks(), which should be fine since it is done via
GFP_KERNEL and irq_build_affinity_masks() is a slow path anyway.
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: John Garry <john.g.garry@oracle.com>
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Link: https://lore.kernel.org/r/20221227022905.352674-4-ming.lei@redhat.com
Several mutexes are taken while setting up console serial ports. In
particular, the tty_port->mutex and @console_mutex are taken:
serial_pnp_probe
serial8250_register_8250_port
uart_add_one_port (locks tty_port->mutex)
uart_configure_port
register_console (locks @console_mutex)
In order to synchronize kgdb's tty_find_polling_driver() with
register_console(), commit 6193bc9084 ("tty: serial: kgdboc:
synchronize tty_find_polling_driver() and register_console()") takes
the @console_mutex. However, this leads to the following call chain
(with locking):
platform_probe
kgdboc_probe
configure_kgdboc (locks @console_mutex)
tty_find_polling_driver
uart_poll_init (locks tty_port->mutex)
uart_set_options
This is clearly deadlock potential due to the reverse lock ordering.
Since uart_set_options() requires holding @console_mutex in order to
serialize early initialization of the serial-console lock, take the
@console_mutex in uart_poll_init() instead of configure_kgdboc().
Since configure_kgdboc() was using @console_mutex for safe traversal
of the console list, change it to use the SRCU iterator instead.
Add comments to uart_set_options() kerneldoc mentioning that it
requires holding @console_mutex (aka the console_list_lock).
Fixes: 6193bc9084 ("tty: serial: kgdboc: synchronize tty_find_polling_driver() and register_console()")
Signed-off-by: John Ogness <john.ogness@linutronix.de>
Reviewed-by: Sergey Senozhatsky <senozhatsky@chromium.org>
Reviewed-by: Petr Mladek <pmladek@suse.com>
[pmladek@suse.com: Export console_srcu_read_lock_is_held() to fix build kgdboc as a module.]
Signed-off-by: Petr Mladek <pmladek@suse.com>
Link: https://lore.kernel.org/r/20230112161213.1434854-1-john.ogness@linutronix.de
The kernel commit 9a5418bc48 ("sched/core: Use kfree_rcu() in
do_set_cpus_allowed()") introduces a bug for kernels built with non-SMP
configs. Calling sched_setaffinity() on such a uniprocessor kernel will
cause cpumask_copy() to be called with a NULL pointer leading to general
protection fault. This is not really a problem in real use cases as
there aren't that many uniprocessor kernel configs in use and calling
sched_setaffinity() on such a uniprocessor system doesn't make sense.
Fix this problem by making sure cpumask_copy() will not be called in
such a case.
Fixes: 9a5418bc48 ("sched/core: Use kfree_rcu() in do_set_cpus_allowed()")
Reported-by: kernel test robot <yujie.liu@intel.com>
Signed-off-by: Waiman Long <longman@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20230115193122.563036-1-longman@redhat.com
In presence of a lot of small weight tasks like sched_idle tasks, normal
or high weight tasks can see their ideal runtime (sched_slice) to increase
to hundreds ms whereas it normally stays below sysctl_sched_latency.
2 normal tasks running on a CPU will have a max sched_slice of 12ms
(half of the sched_period). This means that they will make progress
every sysctl_sched_latency period.
If we now add 1000 idle tasks on the CPU, the sched_period becomes
3006 ms and the ideal runtime of the normal tasks becomes 609 ms.
It will even become 1500ms if the idle tasks belongs to an idle cgroup.
This means that the scheduler will look for picking another waiting task
after 609ms running time (1500ms respectively). The idle tasks change
significantly the way the 2 normal tasks interleave their running time
slot whereas they should have a small impact.
Such long sched_slice can delay significantly the release of resources
as the tasks can wait hundreds of ms before the next running slot just
because of idle tasks queued on the rq.
Cap the ideal_runtime to sysctl_sched_latency to make sure that tasks will
regularly make progress and will not be significantly impacted by
idle/background tasks queued on the rq.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Tested-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Link: https://lore.kernel.org/r/20230113133613.257342-1-vincent.guittot@linaro.org
Pull module fix from Luis Chamberlain:
"Just one fix for modules by Nick"
* tag 'modules-6.2-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/mcgrof/linux:
kallsyms: Fix scheduling with interrupts disabled in self-test
To mitigate Spectre v4, 2039f26f3a ("bpf: Fix leakage due to
insufficient speculative store bypass mitigation") inserts lfence
instructions after 1) initializing a stack slot and 2) spilling a
pointer to the stack.
However, this does not cover cases where a stack slot is first
initialized with a pointer (subject to sanitization) but then
overwritten with a scalar (not subject to sanitization because
the slot was already initialized). In this case, the second write
may be subject to speculative store bypass (SSB) creating a
speculative pointer-as-scalar type confusion. This allows the
program to subsequently leak the numerical pointer value using,
for example, a branch-based cache side channel.
To fix this, also sanitize scalars if they write a stack slot
that previously contained a pointer. Assuming that pointer-spills
are only generated by LLVM on register-pressure, the performance
impact on most real-world BPF programs should be small.
The following unprivileged BPF bytecode drafts a minimal exploit
and the mitigation:
[...]
// r6 = 0 or 1 (skalar, unknown user input)
// r7 = accessible ptr for side channel
// r10 = frame pointer (fp), to be leaked
//
r9 = r10 # fp alias to encourage ssb
*(u64 *)(r9 - 8) = r10 // fp[-8] = ptr, to be leaked
// lfence added here because of pointer spill to stack.
//
// Ommitted: Dummy bpf_ringbuf_output() here to train alias predictor
// for no r9-r10 dependency.
//
*(u64 *)(r10 - 8) = r6 // fp[-8] = scalar, overwrites ptr
// 2039f26f3a: no lfence added because stack slot was not STACK_INVALID,
// store may be subject to SSB
//
// fix: also add an lfence when the slot contained a ptr
//
r8 = *(u64 *)(r9 - 8)
// r8 = architecturally a scalar, speculatively a ptr
//
// leak ptr using branch-based cache side channel:
r8 &= 1 // choose bit to leak
if r8 == 0 goto SLOW // no mispredict
// architecturally dead code if input r6 is 0,
// only executes speculatively iff ptr bit is 1
r8 = *(u64 *)(r7 + 0) # encode bit in cache (0: slow, 1: fast)
SLOW:
[...]
After running this, the program can time the access to *(r7 + 0) to
determine whether the chosen pointer bit was 0 or 1. Repeat this 64
times to recover the whole address on amd64.
In summary, sanitization can only be skipped if one scalar is
overwritten with another scalar. Scalar-confusion due to speculative
store bypass can not lead to invalid accesses because the pointer
bounds deducted during verification are enforced using branchless
logic. See 979d63d50c ("bpf: prevent out of bounds speculation on
pointer arithmetic") for details.
Do not make the mitigation depend on !env->allow_{uninit_stack,ptr_leaks}
because speculative leaks are likely unexpected if these were enabled.
For example, leaking the address to a protected log file may be acceptable
while disabling the mitigation might unintentionally leak the address
into the cached-state of a map that is accessible to unprivileged
processes.
Fixes: 2039f26f3a ("bpf: Fix leakage due to insufficient speculative store bypass mitigation")
Signed-off-by: Luis Gerhorst <gerhorst@cs.fau.de>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Henriette Hofmeier <henriette.hofmeier@rub.de>
Link: https://lore.kernel.org/bpf/edc95bad-aada-9cfc-ffe2-fa9bb206583c@cs.fau.de
Link: https://lore.kernel.org/bpf/20230109150544.41465-1-gerhorst@cs.fau.de
Low level noinstr context-tracking code is calling out to instrumented
code on KASAN:
vmlinux.o: warning: objtool: __ct_user_enter+0x72: call to __kasan_check_write() leaves .noinstr.text section
vmlinux.o: warning: objtool: __ct_user_exit+0x47: call to __kasan_check_write() leaves .noinstr.text section
Use even lower level atomic methods to avoid the instrumentation.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20230112195542.458034262@infradead.org
Robot reported that trace_hardirqs_{on,off}() tickle the forbidden
_rcuidle() tracepoint through local_irq_{en,dis}able().
For 'sane' configs, these calls will only happen with RCU enabled and
as such can use the regular tracepoint. This also means it's possible
to trace them from NMI context again.
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20230112195541.477416709@infradead.org
Current arch_cpu_idle() is called with IRQs disabled, but will return
with IRQs enabled.
However, the very first thing the generic code does after calling
arch_cpu_idle() is raw_local_irq_disable(). This means that
architectures that can idle with IRQs disabled end up doing a
pointless 'enable-disable' dance.
Therefore, push this IRQ disabling into the idle function, meaning
that those architectures can avoid the pointless IRQ state flipping.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Tested-by: Tony Lindgren <tony@atomide.com>
Tested-by: Ulf Hansson <ulf.hansson@linaro.org>
Reviewed-by: Gautham R. Shenoy <gautham.shenoy@amd.com>
Acked-by: Mark Rutland <mark.rutland@arm.com> [arm64]
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Guo Ren <guoren@kernel.org>
Acked-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20230112195540.618076436@infradead.org
Pull timer doc fixes from Ingo Molnar:
- Fix various DocBook formatting errors in kernel/time/ that generated
(justified) warnings during a kernel-doc build.
* tag 'timers-urgent-2023-01-12' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
time: Fix various kernel-doc problems
Pull scheduler fixes from Ingo Molnar:
- Fix scheduler frequency invariance bug related to overly long
tickless periods triggering an integer overflow and disabling the
feature.
- Fix use-after-free bug in dup_user_cpus_ptr().
- Fix do_set_cpus_allowed() deadlock scenarios related to calling
kfree() with the pi_lock held. NOTE: the rcu_free() is the 'lazy'
solution here - we looked at patches to free the structure after the
pi_lock got dropped, but that looked quite a bit messier - and none
of this is truly performance critical. We can revisit this if it's
too lazy of a solution ...
* tag 'sched-urgent-2023-01-12' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched/core: Use kfree_rcu() in do_set_cpus_allowed()
sched/core: Fix use-after-free bug in dup_user_cpus_ptr()
sched/core: Fix arch_scale_freq_tick() on tickless systems
The rcu_boost_kthread_setaffinity() function is invoked at
rcutree_online_cpu() and rcutree_offline_cpu() time, early in the online
timeline and late in the offline timeline, respectively. It is also
invoked from rcutree_dead_cpu(), however, in the absence of userspace
manipulations (for which userspace must take responsibility), this call
is redundant with that from rcutree_offline_cpu(). This redundancy can
be demonstrated by printing out the relevant cpumasks
This commit therefore removes the call to rcu_boost_kthread_setaffinity()
from rcutree_dead_cpu().
Signed-off-by: Zqiang <qiang1.zhang@intel.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
It has been reported that isolated CPUs can suffer from interference due to
per-CPU kworkers waking up just to die.
A surge of workqueue activity during initial setup of a latency-sensitive
application (refresh_vm_stats() being one of the culprits) can cause extra
per-CPU kworkers to be spawned. Then, said latency-sensitive task can be
running merrily on an isolated CPU only to be interrupted sometime later by
a kworker marked for death (cf. IDLE_WORKER_TIMEOUT, 5 minutes after last
kworker activity).
Prevent this by affining kworkers to the wq_unbound_cpumask (which doesn't
contain isolated CPUs, cf. HK_TYPE_WQ) before waking them up after marking
them with WORKER_DIE.
Changing the affinity does require a sleepable context, leverage the newly
introduced pool->idle_cull_work to get that.
Remove dying workers from pool->workers and keep track of them in a
separate list. This intentionally prevents for_each_loop_worker() from
iterating over workers that are marked for death.
Rename destroy_worker() to set_working_dying() to better reflect its
effects and relationship with wake_dying_workers().
Signed-off-by: Valentin Schneider <vschneid@redhat.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
put_unbound_pool() currently passes wq_manager_inactive() as exit condition
to rcuwait_wait_event(), which grabs pool->lock to check for
pool->flags & POOL_MANAGER_ACTIVE
A later patch will require destroy_worker() to be invoked with
wq_pool_attach_mutex held, which needs to be acquired before
pool->lock. A mutex cannot be acquired within rcuwait_wait_event(), as
it could clobber the task state set by rcuwait_wait_event()
Instead, restructure the waiting logic to acquire any necessary lock
outside of rcuwait_wait_event().
Since further work cannot be inserted into unbound pwqs that have reached
->refcnt==0, this is bound to make forward progress as eventually the
worklist will be drained and need_more_worker(pool) will remain false,
preventing any worker from stealing the manager position from us.
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Valentin Schneider <vschneid@redhat.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
A later patch will require a sleepable context in the idle worker timeout
function. Converting worker_pool.idle_timer to a delayed_work gives us just
that, however this would imply turning all idle_timer expiries into
scheduler events (waking up a worker to handle the dwork).
Instead, implement a "custom dwork" where the timer callback does some
extra checks before queuing the associated work.
No change in functionality intended.
Signed-off-by: Valentin Schneider <vschneid@redhat.com>
Reviewed-by: Lai Jiangshan <jiangshanlai@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Later patches will reuse this code, move it into reusable functions.
Signed-off-by: Valentin Schneider <vschneid@redhat.com>
Reviewed-by: Lai Jiangshan <jiangshanlai@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
When unbind_workers() reads wq_unbound_cpumask to set the affinity of
freshly-unbound kworkers, it only holds wq_pool_attach_mutex. This isn't
sufficient as wq_unbound_cpumask is only protected by wq_pool_mutex.
Make wq_unbound_cpumask protected with wq_pool_attach_mutex and also
remove the need of temporary saved_cpumask.
Fixes: 10a5a651e3 ("workqueue: Restrict kworker in the offline CPU pool running on housekeeping CPUs")
Reported-by: Valentin Schneider <vschneid@redhat.com>
Signed-off-by: Lai Jiangshan <jiangshan.ljs@antgroup.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
s390 doesn't use irq_domains, so it has no place to set
IRQ_DOMAIN_FLAG_ISOLATED_MSI. Instead of continuing to abuse the iommu
subsystem to convey this information add a simple define which s390 can
make statically true. The define will cause msi_device_has_isolated() to
return true.
Remove IOMMU_CAP_INTR_REMAP from the s390 iommu driver.
Link: https://lore.kernel.org/r/8-v3-3313bb5dd3a3+10f11-secure_msi_jgg@nvidia.com
Reviewed-by: Matthew Rosato <mjrosato@linux.ibm.com>
Tested-by: Matthew Rosato <mjrosato@linux.ibm.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
This will replace irq_domain_check_msi_remap() in following patches.
The new API makes it more clear what "msi_remap" actually means from a
functional perspective instead of identifying an implementation specific
HW feature.
Isolated MSI means that HW modeled by an irq_domain on the path from the
initiating device to the CPU will validate that the MSI message specifies
an interrupt number that the device is authorized to trigger. This must
block devices from triggering interrupts they are not authorized to
trigger. Currently authorization means the MSI vector is one assigned to
the device.
This is interesting for securing VFIO use cases where a rouge MSI (eg
created by abusing a normal PCI MemWr DMA) must not allow the VFIO
userspace to impact outside its security domain, eg userspace triggering
interrupts on kernel drivers, a VM triggering interrupts on the
hypervisor, or a VM triggering interrupts on another VM.
As this is actually modeled as a per-irq_domain property, not a global
platform property, correct the interface to accept the device parameter
and scan through only the part of the irq_domains hierarchy originating
from the source device.
Locate the new code in msi.c as it naturally only works with
CONFIG_GENERIC_MSI_IRQ, which also requires CONFIG_IRQ_DOMAIN and
IRQ_DOMAIN_HIERARCHY.
Link: https://lore.kernel.org/r/1-v3-3313bb5dd3a3+10f11-secure_msi_jgg@nvidia.com
Tested-by: Matthew Rosato <mjrosato@linux.ibm.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
With the introduction of threaded interrupt handlers, it is virtually
never safe to call disable_irq() from non-premptible context.
Thus: Update the documentation, add an explicit might_sleep() to catch any
offenders. This is more obvious and straight forward than the implicit
might_sleep() check deeper down in the disable_irq() call chain.
Fixes: 3aa551c9b4 ("genirq: add threaded interrupt handler support")
Signed-off-by: Manfred Spraul <manfred@colorfullife.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20221216150441.200533-3-manfred@colorfullife.com
The nanosleep syscalls use the restart_block mechanism, with a quirk:
The `type` and `rmtp`/`compat_rmtp` fields are set up unconditionally on
syscall entry, while the rest of the restart_block is only set up in the
unlikely case that the syscall is actually interrupted by a signal (or
pseudo-signal) that doesn't have a signal handler.
If the restart_block was set up by a previous syscall (futex(...,
FUTEX_WAIT, ...) or poll()) and hasn't been invalidated somehow since then,
this will clobber some of the union fields used by futex_wait_restart() and
do_restart_poll().
If userspace afterwards wrongly calls the restart_syscall syscall,
futex_wait_restart()/do_restart_poll() will read struct fields that have
been clobbered.
This doesn't actually lead to anything particularly interesting because
none of the union fields contain trusted kernel data, and
futex(..., FUTEX_WAIT, ...) and poll() aren't syscalls where it makes much
sense to apply seccomp filters to their arguments.
So the current consequences are just of the "if userspace does bad stuff,
it can damage itself, and that's not a problem" flavor.
But still, it seems like a hazard for future developers, so invalidate the
restart_block when partly setting it up in the nanosleep syscalls.
Signed-off-by: Jann Horn <jannh@google.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20230105134403.754986-1-jannh@google.com
The various internal size limit macros have names and/or values that
do not fit well to their current usage.
Rename the macros so that their purpose is clear and, if needed,
provide a more appropriate value. In general, the new macros and
values will lead to less memory usage. The new macros are...
PRINTK_MESSAGE_MAX:
This is the maximum size for a formatted message on a console,
devkmsg, or syslog. It does not matter which format the message has
(normal or extended). It replaces the use of CONSOLE_EXT_LOG_MAX for
console and devkmsg. It replaces the use of CONSOLE_LOG_MAX for
syslog.
Historically, normal messages have been allowed to print up to 1kB,
whereas extended messages have been allowed to print up to 8kB.
However, the difference in lengths of these message types is not
significant and in multi-line records, normal messages are probably
larger. Also, because 1kB is only slightly above the allowed record
size, multi-line normal messages could be easily truncated during
formatting.
This new macro should be significantly larger than the allowed
record size to allow sufficient space for extended or multi-line
prefix text. A value of 2kB should be plenty of space. For normal
messages this represents a doubling of the historically allowed
amount. For extended messages it reduces the excessive 8kB size,
thus reducing memory usage needed for message formatting.
PRINTK_PREFIX_MAX:
This is the maximum size allowed for a record prefix (used by
console and syslog). It replaces PREFIX_MAX. The value is left
unchanged.
PRINTKRB_RECORD_MAX:
This is the maximum size allowed to be reserved for a record in the
ringbuffer. It is used by all readers and writers with the printk
ringbuffer. It replaces LOG_LINE_MAX.
Previously this was set to "1kB - PREFIX_MAX", which makes some
sense if 1kB is the limit for normal message output and prefixes are
enabled. However, with the allowance of larger output and the
existence of multi-line records, the value is rather bizarre.
Round the value up to 1kB.
Signed-off-by: John Ogness <john.ogness@linutronix.de>
Reviewed-by: Petr Mladek <pmladek@suse.com>
Signed-off-by: Petr Mladek <pmladek@suse.com>
Link: https://lore.kernel.org/r/20230109100800.1085541-9-john.ogness@linutronix.de
Replace the buffers in struct devkmsg_user with a struct
printk_buffers. This reduces the number of buffers to keep track of.
As a side-effect, @text_buf was 8kB large, even though it only
needed to be the max size of a ringbuffer record. By switching to
struct printk_buffers, ~7kB less memory is allocated when opening
/dev/kmsg.
And since struct printk_buffers will be used now, reduce duplicate
code by calling printk_get_next_message() to handle the record
reading and formatting.
Note that since /dev/kmsg never suppresses records based on
loglevel, printk_get_next_message() is extended with an extra
bool argument to specify if suppression is allowed.
Signed-off-by: John Ogness <john.ogness@linutronix.de>
Reviewed-by: Petr Mladek <pmladek@suse.com>
Signed-off-by: Petr Mladek <pmladek@suse.com>
Link: https://lore.kernel.org/r/20230109100800.1085541-8-john.ogness@linutronix.de
Currently "dropped messages" are separately printed immediately
before printing the printk message. Since normal consoles are
now using an output buffer that is much larger than previously,
the "dropped message" could be prepended to the printk message
and then output everything in a single write() call.
Introduce a helper function console_prepend_dropped() to prepend
an existing message with a "dropped message". This simplifies
the code by allowing all message formatting to be handled
together and then only requires a single write() call to output
the full message. And since this helper does not require any
locking, it can be used in the future for other console printing
contexts as well.
Note that console_prepend_dropped() is defined as a NOP for
!CONFIG_PRINTK. Although the function will never be called for
!CONFIG_PRINTK, compiling the function can lead to warnings of
"always true" conditionals due to the size macro values used
in !CONFIG_PRINTK.
Signed-off-by: John Ogness <john.ogness@linutronix.de>
Reviewed-by: Petr Mladek <pmladek@suse.com>
Signed-off-by: Petr Mladek <pmladek@suse.com>
Link: https://lore.kernel.org/r/20230109100800.1085541-7-john.ogness@linutronix.de
Code for performing the console output is intermixed with code that
is formatting the output for that console. Introduce a new helper
function printk_get_next_message() to handle the reading and
formatting of the printk text. The helper does not require any
locking so that in the future it can be used for other printing
contexts as well.
This also introduces a new struct printk_message to wrap the struct
printk_buffers, adding metadata about its contents. This allows
users of printk_get_next_message() to receive all relevant
information about the message that was read and formatted.
Why is struct printk_message a wrapper struct?
It is intentional that a wrapper struct is introduced instead of
adding the metadata directly to struct printk_buffers. The upcoming
atomic consoles support multiple printing contexts per CPU. This
means that while a CPU is formatting a message, it can be
interrupted and the interrupting context may also format a (possibly
different) message. Since the printk buffers are rather large,
there will only be one struct printk_buffers per CPU and it must be
shared by the possible contexts of that CPU.
If the metadata was part of struct printk_buffers, interrupting
contexts would clobber the metadata being prepared by the
interrupted context. This could be handled by robustifying the
message formatting functions to cope with metadata unexpectedly
changing. However, this would require significant amounts of extra
data copying, also adding significant complexity to the code.
Instead, the metadata can live on the stack of the formatting
context and the message formatting functions do not need to be
concerned about the metadata changing underneath them.
Note that the message formatting functions can handle unexpected
text buffer changes. So it is perfectly OK if a shared text buffer
is clobbered by an interrupting context. The atomic console
implementation will recognize the interruption and avoid printing
the (probably garbage) text buffer.
Signed-off-by: John Ogness <john.ogness@linutronix.de>
Reviewed-by: Petr Mladek <pmladek@suse.com>
Signed-off-by: Petr Mladek <pmladek@suse.com>
Link: https://lore.kernel.org/r/20230109100800.1085541-6-john.ogness@linutronix.de
Introduce a new struct printk_buffers to contain all the buffers
needed to read and format a printk message for output. Putting the
buffers inside a struct reduces the number of buffer pointers that
need to be tracked. Also, it allows usage of the sizeof() macro for
the buffer sizes, rather than expecting certain sized buffers being
passed in.
Note that since the output buffer for normal consoles is now
CONSOLE_EXT_LOG_MAX instead of CONSOLE_LOG_MAX, multi-line
messages that may have been previously truncated will now be
printed in full. This should be considered a feature and not a bug
since the CONSOLE_LOG_MAX restriction was about limiting static
buffer usage rather than limiting printed text.
Signed-off-by: John Ogness <john.ogness@linutronix.de>
Reviewed-by: Petr Mladek <pmladek@suse.com>
Signed-off-by: Petr Mladek <pmladek@suse.com>
Link: https://lore.kernel.org/r/20230109100800.1085541-5-john.ogness@linutronix.de
