To support wire to MSI domains via the MSI infrastructure it is required to
use the firmware node of the device which implements this for creating the
MSI domain. Otherwise the existing firmware match mechanisms to find the
correct irqdomain for a wired interrupt which is connected to a wire to MSI
bridge would fail.
This cannot be used for the general case because not all devices provide
firmware nodes and all regular per device MSI domains are directly
associated to the device and have not be searched for.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Anup Patel <apatel@ventanamicro.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240127161753.114685-11-apatel@ventanamicro.com
irq_create_fwspec_mapping() requires translation of the firmware spec to a
hardware interrupt number and the trigger type information.
Wired interrupts which are connected to a wire to MSI bridge, like MBIGEN
are allocated that way. So far MBIGEN provides a regular irqdomain which
then hooks backwards into the MSI infrastructure. That's an unholy mess and
will be replaced with per device MSI domains which are regular MSI domains.
Interrupts on MSI domains are not supported by irq_create_fwspec_mapping(),
but for making the wire to MSI bridges sane it makes sense to provide a
special allocation/free interface in the MSI infrastructure. That avoids
the backdoors into the core MSI allocation code and just shares all the
regular MSI infrastructure.
Provide an optional translation callback in msi_domain_ops which can be
utilized by these wire to MSI bridges. No other MSI domain should provide a
translation callback. The default translation callback of the MSI
irqdomains will warn when it is invoked on a non-prepared MSI domain.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Anup Patel <apatel@ventanamicro.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240127161753.114685-8-apatel@ventanamicro.com
RISC-V uses xRET instructions on return from interrupt and to go back
to user-space; the xRET instruction is not core serializing.
Use FENCE.I for providing core serialization as follows:
- by calling sync_core_before_usermode() on return from interrupt (cf.
ipi_sync_core()),
- via switch_mm() and sync_core_before_usermode() (respectively, for
uthread->uthread and kthread->uthread transitions) before returning
to user-space.
On RISC-V, the serialization in switch_mm() is activated by resetting
the icache_stale_mask of the mm at prepare_sync_core_cmd().
Suggested-by: Palmer Dabbelt <palmer@dabbelt.com>
Signed-off-by: Andrea Parri <parri.andrea@gmail.com>
Reviewed-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Link: https://lore.kernel.org/r/20240131144936.29190-5-parri.andrea@gmail.com
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
The membarrier system call requires a full memory barrier after storing
to rq->curr, before going back to user-space. The barrier is only
needed when switching between processes: the barrier is implied by
mmdrop() when switching from kernel to userspace, and it's not needed
when switching from userspace to kernel.
Rely on the feature/mechanism ARCH_HAS_MEMBARRIER_CALLBACKS and on the
primitive membarrier_arch_switch_mm(), already adopted by the PowerPC
architecture, to insert the required barrier.
Fixes: fab957c11e ("RISC-V: Atomic and Locking Code")
Signed-off-by: Andrea Parri <parri.andrea@gmail.com>
Reviewed-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Link: https://lore.kernel.org/r/20240131144936.29190-2-parri.andrea@gmail.com
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
Real-world BPF applications keep growing in size. Medium-sized production
application can easily have 50K+ verified instructions, and its line
info section in .BTF.ext has more than 3K entries.
When verifier emits log with log_level>=1, it annotates assembly code
with matched original C source code. Currently it uses linear search
over line info records to find a match. As complexity of BPF
applications grows, this O(K * N) approach scales poorly.
So, let's instead of linear O(N) search for line info record use faster
equivalent O(log(N)) binary search algorithm. It's not a plain binary
search, as we don't look for exact match. It's an upper bound search
variant, looking for rightmost line info record that starts at or before
given insn_off.
Some unscientific measurements were done before and after this change.
They were done in VM and fluctuate a bit, but overall the speed up is
undeniable.
BASELINE
========
File Program Duration (us) Insns
-------------------------------- ---------------- ------------- ------
katran.bpf.o balancer_ingress 2497130 343552
pyperf600.bpf.linked3.o on_event 12389611 627288
strobelight_pyperf_libbpf.o on_py_event 387399 52445
-------------------------------- ---------------- ------------- ------
BINARY SEARCH
=============
File Program Duration (us) Insns
-------------------------------- ---------------- ------------- ------
katran.bpf.o balancer_ingress 2339312 343552
pyperf600.bpf.linked3.o on_event 5602203 627288
strobelight_pyperf_libbpf.o on_py_event 294761 52445
-------------------------------- ---------------- ------------- ------
While Katran's speed up is pretty modest (about 105ms, or 6%), for
production pyperf BPF program (on_py_event) it's much greater already,
going from 387ms down to 295ms (23% improvement).
Looking at BPF selftests's biggest pyperf example, we can see even more
dramatic improvement, shaving more than 50% of time, going from 12.3s
down to 5.6s.
Different amount of improvement is the function of overall amount of BPF
assembly instructions in .bpf.o files (which contributes to how much
line info records there will be and thus, on average, how much time linear
search will take), among other things:
$ llvm-objdump -d katran.bpf.o | wc -l
3863
$ llvm-objdump -d strobelight_pyperf_libbpf.o | wc -l
6997
$ llvm-objdump -d pyperf600.bpf.linked3.o | wc -l
87854
Granted, this only applies to debugging cases (e.g., using veristat, or
failing verification in production), but seems worth doing to improve
overall developer experience anyways.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/bpf/20240214002311.2197116-1-andrii@kernel.org
When queue_work_on() is used to queue a BH work item on a remote CPU, the
work item is queued on that CPU but kick_pool() raises softirq on the local
CPU. This leads to stalls as the work item won't be executed until something
else on the remote CPU schedules a BH work item or tasklet locally.
Fix it by bouncing raising softirq to the target CPU using per-cpu irq_work.
Signed-off-by: Tejun Heo <tj@kernel.org>
Fixes: 4cb1ef6460 ("workqueue: Implement BH workqueues to eventually replace tasklets")
Currently, if rcu_torture_writer() sees fewer than ten grace periods
having elapsed during a call to stutter_wait() that actually waited,
the rtort_pipe_count warning is emitted. This has worked well for
a long time. Except that the rcutorture TREE07 scenario now does a
short-term 14-second RCU CPU stall, which can most definitely case
false-positive rtort_pipe_count warnings.
This commit therefore changes rcu_torture_writer() to compute the
full expected holdoff and stall duration, and to refuse to report any
rtort_pipe_count warnings until after all stalls have completed.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Cc: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
The comments added in commit 1ef990c4b36b ("srcu: No need to
advance/accelerate if no callback enqueued") are a bit confusing.
The comments are describing a scenario for code that was moved and is
no longer the way it was (snapshot after advancing). Improve the code
comments to reflect this and also document why acceleration can never
fail.
Cc: Frederic Weisbecker <frederic@kernel.org>
Cc: Neeraj Upadhyay <neeraj.iitr10@gmail.com>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
To allow more flexible arrangements while still provide a single kernel
for distros, provide a boot time parameter to enable/disable lazy RCU.
Specify:
rcutree.enable_rcu_lazy=[y|1|n|0]
Which also requires
rcu_nocbs=all
at boot time to enable/disable lazy RCU.
To disable it by default at build time when CONFIG_RCU_LAZY=y, the new
CONFIG_RCU_LAZY_DEFAULT_OFF can be used.
Signed-off-by: Qais Yousef (Google) <qyousef@layalina.io>
Tested-by: Andrea Righi <andrea.righi@canonical.com>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
The variable name jiffies_till_flush is too generic and therefore:
* It may shadow a global variable
* It doesn't tell on what it operates
Make the name more precise, along with the related APIs.
Reviewed-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
TREE04 running on short iterations can produce writer stalls of the
following kind:
??? Writer stall state RTWS_EXP_SYNC(4) g3968 f0x0 ->state 0x2 cpu 0
task:rcu_torture_wri state:D stack:14568 pid:83 ppid:2 flags:0x00004000
Call Trace:
<TASK>
__schedule+0x2de/0x850
? trace_event_raw_event_rcu_exp_funnel_lock+0x6d/0xb0
schedule+0x4f/0x90
synchronize_rcu_expedited+0x430/0x670
? __pfx_autoremove_wake_function+0x10/0x10
? __pfx_synchronize_rcu_expedited+0x10/0x10
do_rtws_sync.constprop.0+0xde/0x230
rcu_torture_writer+0x4b4/0xcd0
? __pfx_rcu_torture_writer+0x10/0x10
kthread+0xc7/0xf0
? __pfx_kthread+0x10/0x10
ret_from_fork+0x2f/0x50
? __pfx_kthread+0x10/0x10
ret_from_fork_asm+0x1b/0x30
</TASK>
Waiting for an expedited grace period and polling for an expedited
grace period both are operations that internally rely on the same
workqueue performing necessary asynchronous work.
However, a dependency chain is involved between those two operations,
as depicted below:
====== CPU 0 ======= ====== CPU 1 =======
synchronize_rcu_expedited()
exp_funnel_lock()
mutex_lock(&rcu_state.exp_mutex);
start_poll_synchronize_rcu_expedited
queue_work(rcu_gp_wq, &rnp->exp_poll_wq);
synchronize_rcu_expedited_queue_work()
queue_work(rcu_gp_wq, &rew->rew_work);
wait_event() // A, wait for &rew->rew_work completion
mutex_unlock() // B
//======> switch to kworker
sync_rcu_do_polled_gp() {
synchronize_rcu_expedited()
exp_funnel_lock()
mutex_lock(&rcu_state.exp_mutex); // C, wait B
....
} // D
Since workqueues are usually implemented on top of several kworkers
handling the queue concurrently, the above situation wouldn't deadlock
most of the time because A then doesn't depend on D. But in case of
memory stress, a single kworker may end up handling alone all the works
in a serialized way. In that case the above layout becomes a problem
because A then waits for D, closing a circular dependency:
A -> D -> C -> B -> A
This however only happens when CONFIG_RCU_EXP_KTHREAD=n. Indeed
synchronize_rcu_expedited() is otherwise implemented on top of a kthread
worker while polling still relies on rcu_gp_wq workqueue, breaking the
above circular dependency chain.
Fix this with making expedited grace period to always rely on kthread
worker. The workqueue based implementation is essentially a duplicate
anyway now that the per-node initialization is performed by per-node
kthread workers.
Meanwhile the CONFIG_RCU_EXP_KTHREAD switch is still kept around to
manage the scheduler policy of these kthread workers.
Reported-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Reported-by: Thomas Gleixner <tglx@linutronix.de>
Suggested-by: Joel Fernandes <joel@joelfernandes.org>
Suggested-by: Paul E. McKenney <paulmck@kernel.org>
Suggested-by: Neeraj upadhyay <Neeraj.Upadhyay@amd.com>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
Affine the parallel expedited gp kworkers to their respective RCU node
in order to make them close to the cache their are playing with.
This reuses the boost kthreads machinery that probe into CPU hotplug
operations such that the kthreads become/stay affine to their respective
node as soon/long as they contain online CPUs. Otherwise and if the
current CPU going down was the last online on the leaf node, the related
kthread is affine to the housekeeping CPUs.
In the long run, this affinity VS CPU hotplug operation game should
probably be implemented at the generic kthread level.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
[boqun: s/* rcu_boost_task/*rcu_boost_task as reported by checkpatch]
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
When CONFIG_RCU_EXP_KTHREAD=n, the expedited grace period per node
initialization is performed in parallel via workqueues (one work per
node).
However in CONFIG_RCU_EXP_KTHREAD=y, this per node initialization is
performed by a single kworker serializing each node initialization (one
work for all nodes).
The second part is certainly less scalable and efficient beyond a single
leaf node.
To improve this, expand this single kworker into per-node kworkers. This
new layout is eventually intended to remove the workqueues based
implementation since it will essentially now become duplicate code.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
The expedited kthread worker performing the per node initialization is
going to be split into per node kthreads. As such, the future per node
kthread creation will need to be called from CPU hotplug callbacks
instead of an initcall, right beside the per node boost kthread
creation.
To prepare for that, move the kthread worker creation above
rcutree_prepare_cpu() as a first step to make the review smoother for
the upcoming modifications.
No intended functional change.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
This mutex is currently protecting per node boost kthreads creation and
affinity setting across CPU hotplug operations.
Since the expedited kworkers will soon be split per node as well, they
will be subject to the same concurrency constraints against hotplug.
Therefore their creation and affinity tuning operations will be grouped
with those of boost kthreads and then rely on the same mutex.
To prepare for that, generalize its name.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
Just like is done for the kworker performing nodes initialization,
gracefully handle the possible allocation failure of the RCU expedited
grace period main kworker.
While at it perform a rename of the related checking functions to better
reflect the expedited specifics.
Reviewed-by: Kalesh Singh <kaleshsingh@google.com>
Fixes: 9621fbee44 ("rcu: Move expedited grace period (GP) work to RT kthread_worker")
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
Under CONFIG_RCU_EXP_KTHREAD=y, the nodes initialization for expedited
grace periods is queued to a kworker. However if the allocation of that
kworker failed, the nodes initialization is performed synchronously by
the caller instead.
Now the check for kworker initialization failure relies on the kworker
pointer to be NULL while its value might actually encapsulate an
allocation failure error.
Make sure to handle this case.
Reviewed-by: Kalesh Singh <kaleshsingh@google.com>
Fixes: 9621fbee44 ("rcu: Move expedited grace period (GP) work to RT kthread_worker")
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
When an expedited grace period is ending, care must be taken so that all
the quiescent states propagated up to the root are correctly ordered
against the wake up of the main expedited grace period workqueue.
This ordering is already carried through the root rnp locking augmented
by an smp_mb__after_unlock_lock() barrier.
Therefore the explicit smp_mb() placed before the wake up is not needed
and can be removed.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
Currently, only rdp_gp->nocb_timer is used, for nocb_timer of
no-rdp_gp structure, the timer_pending() is always return false,
this commit therefore need to check rdp_gp->nocb_timer in
__call_rcu_nocb_wake().
Signed-off-by: Zqiang <qiang.zhang1211@gmail.com>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
For the kernels built with CONFIG_RCU_NOCB_CPU_DEFAULT_ALL=y and
CONFIG_RCU_LAZY=y, the following scenarios will trigger WARN_ON_ONCE()
in the rcu_nocb_bypass_lock() and rcu_nocb_wait_contended() functions:
CPU2 CPU11
kthread
rcu_nocb_cb_kthread ksys_write
rcu_do_batch vfs_write
rcu_torture_timer_cb proc_sys_write
__kmem_cache_free proc_sys_call_handler
kmemleak_free drop_caches_sysctl_handler
delete_object_full drop_slab
__delete_object shrink_slab
put_object lazy_rcu_shrink_scan
call_rcu rcu_nocb_flush_bypass
__call_rcu_commn rcu_nocb_bypass_lock
raw_spin_trylock(&rdp->nocb_bypass_lock) fail
atomic_inc(&rdp->nocb_lock_contended);
rcu_nocb_wait_contended WARN_ON_ONCE(smp_processor_id() != rdp->cpu);
WARN_ON_ONCE(atomic_read(&rdp->nocb_lock_contended)) |
|_ _ _ _ _ _ _ _ _ _same rdp and rdp->cpu != 11_ _ _ _ _ _ _ _ _ __|
Reproduce this bug with "echo 3 > /proc/sys/vm/drop_caches".
This commit therefore uses rcu_nocb_try_flush_bypass() instead of
rcu_nocb_flush_bypass() in lazy_rcu_shrink_scan(). If the nocb_bypass
queue is being flushed, then rcu_nocb_try_flush_bypass will return
directly.
Signed-off-by: Zqiang <qiang.zhang1211@gmail.com>
Reviewed-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
Currently the call_rcu() function interleaves NOCB and !NOCB enqueue
code in a complicated way such that:
* The bypass enqueue code may or may not have enqueued and may or may
not have locked the ->nocb_lock. Everything that follows is in a
Schrödinger locking state for the unwary reviewer's eyes.
* The was_alldone is always set but only used in NOCB related code.
* The NOCB wake up is distantly related to the locking hopefully
performed by the bypass enqueue code that did not enqueue on the
bypass list.
Unconfuse the whole and gather NOCB and !NOCB specific enqueue code to
their own functions.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
Currently IRQs are disabled on call_rcu() and then depending on the
context:
* If the CPU is in nocb mode:
- If the callback is enqueued in the bypass list, IRQs are re-enabled
implictly by rcu_nocb_try_bypass()
- If the callback is enqueued in the normal list, IRQs are re-enabled
implicitly by __call_rcu_nocb_wake()
* If the CPU is NOT in nocb mode, IRQs are reenabled explicitly from call_rcu()
This makes the code a bit hard to follow, especially as it interleaves
with nocb locking.
To make the IRQ flags coverage clearer and also in order to prepare for
moving all the nocb enqueue code to its own function, always re-enable
the IRQ flags explicitly from call_rcu().
Reviewed-by: Neeraj Upadhyay (AMD) <neeraj.iitr10@gmail.com>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
A full barrier is issued from nocb_gp_wait() upon callbacks advancing
to order grace period completion with callbacks execution.
However these two events are already ordered by the
smp_mb__after_unlock_lock() barrier within the call to
raw_spin_lock_rcu_node() that is necessary for callbacks advancing to
happen.
The following litmus test shows the kind of guarantee that this barrier
provides:
C smp_mb__after_unlock_lock
{}
// rcu_gp_cleanup()
P0(spinlock_t *rnp_lock, int *gpnum)
{
// Grace period cleanup increase gp sequence number
spin_lock(rnp_lock);
WRITE_ONCE(*gpnum, 1);
spin_unlock(rnp_lock);
}
// nocb_gp_wait()
P1(spinlock_t *rnp_lock, spinlock_t *nocb_lock, int *gpnum, int *cb_ready)
{
int r1;
// Call rcu_advance_cbs() from nocb_gp_wait()
spin_lock(nocb_lock);
spin_lock(rnp_lock);
smp_mb__after_unlock_lock();
r1 = READ_ONCE(*gpnum);
WRITE_ONCE(*cb_ready, 1);
spin_unlock(rnp_lock);
spin_unlock(nocb_lock);
}
// nocb_cb_wait()
P2(spinlock_t *nocb_lock, int *cb_ready, int *cb_executed)
{
int r2;
// rcu_do_batch() -> rcu_segcblist_extract_done_cbs()
spin_lock(nocb_lock);
r2 = READ_ONCE(*cb_ready);
spin_unlock(nocb_lock);
// Actual callback execution
WRITE_ONCE(*cb_executed, 1);
}
P3(int *cb_executed, int *gpnum)
{
int r3;
WRITE_ONCE(*cb_executed, 2);
smp_mb();
r3 = READ_ONCE(*gpnum);
}
exists (1:r1=1 /\ 2:r2=1 /\ cb_executed=2 /\ 3:r3=0) (* Bad outcome. *)
Here the bad outcome only occurs if the smp_mb__after_unlock_lock() is
removed. This barrier orders the grace period completion against
callbacks advancing and even later callbacks invocation, thanks to the
opportunistic propagation via the ->nocb_lock to nocb_cb_wait().
Therefore the smp_mb() placed after callbacks advancing can be safely
removed.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
The LOAD-ACQUIRE access performed on rdp->nocb_cb_sleep advertizes
ordering callback execution against grace period completion. However
this is contradicted by the following:
* This LOAD-ACQUIRE doesn't pair with anything. The only counterpart
barrier that can be found is the smp_mb() placed after callbacks
advancing in nocb_gp_wait(). However the barrier is placed _after_
->nocb_cb_sleep write.
* Callbacks can be concurrently advanced between the LOAD-ACQUIRE on
->nocb_cb_sleep and the call to rcu_segcblist_extract_done_cbs() in
rcu_do_batch(), making any ordering based on ->nocb_cb_sleep broken.
* Both rcu_segcblist_extract_done_cbs() and rcu_advance_cbs() are called
under the nocb_lock, the latter hereby providing already the desired
ACQUIRE semantics.
Therefore it is safe to access ->nocb_cb_sleep with a simple compiler
barrier.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
As BPF applications grow in size and complexity and are separated into
multiple .bpf.c files that are statically linked together, it becomes
harder and harder to match verifier's BPF assembly level output to
original C code. While often annotated C source code is unique enough to
be able to identify the file it belongs to, quite often this is actually
problematic as parts of source code can be quite generic.
Long story short, it is very useful to see source code file name and
line number information along with the original C code. Verifier already
knows this information, we just need to output it.
This patch extends verifier log with file name and line number
information, emitted next to original (presumably C) source code,
annotating BPF assembly output, like so:
; <original C code> @ <filename>.bpf.c:<line>
If file name has directory names in it, they are stripped away. This
should be fine in practice as file names tend to be pretty unique with
C code anyways, and keeping log size smaller is always good.
In practice this might look something like below, where some code is
coming from application files, while others are from libbpf's usdt.bpf.h
header file:
; if (STROBEMETA_READ( @ strobemeta_probe.bpf.c:534
5592: (79) r1 = *(u64 *)(r10 -56) ; R1_w=mem_or_null(id=1589,sz=7680) R10=fp0
5593: (7b) *(u64 *)(r10 -56) = r1 ; R1_w=mem_or_null(id=1589,sz=7680) R10=fp0
5594: (79) r3 = *(u64 *)(r10 -8) ; R3_w=scalar() R10=fp0 fp-8=mmmmmmmm
...
170: (71) r1 = *(u8 *)(r8 +15) ; frame1: R1_w=scalar(...) R8_w=map_value(map=__bpf_usdt_spec,ks=4,vs=208)
171: (67) r1 <<= 56 ; frame1: R1_w=scalar(...)
172: (c7) r1 s>>= 56 ; frame1: R1_w=scalar(smin=smin32=-128,smax=smax32=127)
; val <<= arg_spec->arg_bitshift; @ usdt.bpf.h:183
173: (67) r1 <<= 32 ; frame1: R1_w=scalar(...)
174: (77) r1 >>= 32 ; frame1: R1_w=scalar(smin=0,smax=umax=0xffffffff,var_off=(0x0; 0xffffffff))
175: (79) r2 = *(u64 *)(r10 -8) ; frame1: R2_w=scalar() R10=fp0 fp-8=mmmmmmmm
176: (6f) r2 <<= r1 ; frame1: R1_w=scalar(smin=0,smax=umax=0xffffffff,var_off=(0x0; 0xffffffff)) R2_w=scalar()
177: (7b) *(u64 *)(r10 -8) = r2 ; frame1: R2_w=scalar(id=61) R10=fp0 fp-8_w=scalar(id=61)
; if (arg_spec->arg_signed) @ usdt.bpf.h:184
178: (bf) r3 = r2 ; frame1: R2_w=scalar(id=61) R3_w=scalar(id=61)
179: (7f) r3 >>= r1 ; frame1: R1_w=scalar(smin=0,smax=umax=0xffffffff,var_off=(0x0; 0xffffffff)) R3_w=scalar()
; if (arg_spec->arg_signed) @ usdt.bpf.h:184
180: (71) r4 = *(u8 *)(r8 +14)
181: safe
log_fixup tests needed a minor adjustment as verifier log output
increased a bit and that test is quite sensitive to such changes.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/r/20240212235944.2816107-1-andrii@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
For program types that don't have named context type name (e.g., BPF
iterator programs or tracepoint programs), ctx_tname will be a non-NULL
empty string. For such programs it shouldn't be possible to have
PTR_TO_CTX argument for global subprogs based on type name alone.
arg:ctx tag is the only way to have PTR_TO_CTX passed into global
subprog for such program types.
Fix this loophole, which currently would assume PTR_TO_CTX whenever
user uses a pointer to anonymous struct as an argument to their global
subprogs. This happens in practice with the following (quite common, in
practice) approach:
typedef struct { /* anonymous */
int x;
} my_type_t;
int my_subprog(my_type_t *arg) { ... }
User's intent is to have PTR_TO_MEM argument for `arg`, but verifier
will complain about expecting PTR_TO_CTX.
This fix also closes unintended s390x-specific KPROBE handling of
PTR_TO_CTX case. Selftest change is necessary to accommodate this.
Fixes: 91cc1a9974 ("bpf: Annotate context types")
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/r/20240212233221.2575350-4-andrii@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Expected canonical argument type for global function arguments
representing PTR_TO_CTX is `bpf_user_pt_regs_t *ctx`. This currently
works on s390x by accident because kernel resolves such typedef to
underlying struct (which is anonymous on s390x), and erroneously
accepting it as expected context type. We are fixing this problem next,
which would break s390x arch, so we need to handle `bpf_user_pt_regs_t`
case explicitly for KPROBE programs.
Fixes: 91cc1a9974 ("bpf: Annotate context types")
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/r/20240212233221.2575350-3-andrii@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Return result of btf_get_prog_ctx_type() is never used and callers only
check NULL vs non-NULL case to determine if given type matches expected
PTR_TO_CTX type. So rename function to `btf_is_prog_ctx_type()` and
return a simple true/false. We'll use this simpler interface to handle
kprobe program type's special typedef case in the next patch.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/r/20240212233221.2575350-2-andrii@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Originally, this patch removed a redundant check in
BPF_CGROUP_RUN_PROG_INET_EGRESS, as the check was already being done in
the function it called, __cgroup_bpf_run_filter_skb. For v2, it was
reccomended that I remove the check from __cgroup_bpf_run_filter_skb,
and add the checks to the other macro that calls that function,
BPF_CGROUP_RUN_PROG_INET_INGRESS.
To sum it up, checking that the socket exists and that it is a full
socket is now part of both macros BPF_CGROUP_RUN_PROG_INET_EGRESS and
BPF_CGROUP_RUN_PROG_INET_INGRESS, and it is no longer part of the
function they call, __cgroup_bpf_run_filter_skb.
v3->v4: Fixed weird merge conflict.
v2->v3: Sent to bpf-next instead of generic patch
v1->v2: Addressed feedback about where check should be removed.
Signed-off-by: Oliver Crumrine <ozlinuxc@gmail.com>
Acked-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/r/7lv62yiyvmj5a7eozv2iznglpkydkdfancgmbhiptrgvgan5sy@3fl3onchgdz3
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Collect argument information from the type information of stub functions to
mark arguments of BPF struct_ops programs with PTR_MAYBE_NULL if they are
nullable. A nullable argument is annotated by suffixing "__nullable" at
the argument name of stub function.
For nullable arguments, this patch sets a struct bpf_ctx_arg_aux to label
their reg_type with PTR_TO_BTF_ID | PTR_TRUSTED | PTR_MAYBE_NULL. This
makes the verifier to check programs and ensure that they properly check
the pointer. The programs should check if the pointer is null before
accessing the pointed memory.
The implementer of a struct_ops type should annotate the arguments that can
be null. The implementer should define a stub function (empty) as a
placeholder for each defined operator. The name of a stub function should
be in the pattern "<st_op_type>__<operator name>". For example, for
test_maybe_null of struct bpf_testmod_ops, it's stub function name should
be "bpf_testmod_ops__test_maybe_null". You mark an argument nullable by
suffixing the argument name with "__nullable" at the stub function.
Since we already has stub functions for kCFI, we just reuse these stub
functions with the naming convention mentioned earlier. These stub
functions with the naming convention is only required if there are nullable
arguments to annotate. For functions having not nullable arguments, stub
functions are not necessary for the purpose of this patch.
This patch will prepare a list of struct bpf_ctx_arg_aux, aka arg_info, for
each member field of a struct_ops type. "arg_info" will be assigned to
"prog->aux->ctx_arg_info" of BPF struct_ops programs in
check_struct_ops_btf_id() so that it can be used by btf_ctx_access() later
to set reg_type properly for the verifier.
Signed-off-by: Kui-Feng Lee <thinker.li@gmail.com>
Link: https://lore.kernel.org/r/20240209023750.1153905-4-thinker.li@gmail.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Currently tracing is supposed not to allow for bpf_spin_{lock,unlock}()
helper calls. This is to prevent deadlock for the following cases:
- there is a prog (prog-A) calling bpf_spin_{lock,unlock}().
- there is a tracing program (prog-B), e.g., fentry, attached
to bpf_spin_lock() and/or bpf_spin_unlock().
- prog-B calls bpf_spin_{lock,unlock}().
For such a case, when prog-A calls bpf_spin_{lock,unlock}(),
a deadlock will happen.
The related source codes are below in kernel/bpf/helpers.c:
notrace BPF_CALL_1(bpf_spin_lock, struct bpf_spin_lock *, lock)
notrace BPF_CALL_1(bpf_spin_unlock, struct bpf_spin_lock *, lock)
notrace is supposed to prevent fentry prog from attaching to
bpf_spin_{lock,unlock}().
But actually this is not the case and fentry prog can successfully
attached to bpf_spin_lock(). Siddharth Chintamaneni reported
the issue in [1]. The following is the macro definition for
above BPF_CALL_1:
#define BPF_CALL_x(x, name, ...) \
static __always_inline \
u64 ____##name(__BPF_MAP(x, __BPF_DECL_ARGS, __BPF_V, __VA_ARGS__)); \
typedef u64 (*btf_##name)(__BPF_MAP(x, __BPF_DECL_ARGS, __BPF_V, __VA_ARGS__)); \
u64 name(__BPF_REG(x, __BPF_DECL_REGS, __BPF_N, __VA_ARGS__)); \
u64 name(__BPF_REG(x, __BPF_DECL_REGS, __BPF_N, __VA_ARGS__)) \
{ \
return ((btf_##name)____##name)(__BPF_MAP(x,__BPF_CAST,__BPF_N,__VA_ARGS__));\
} \
static __always_inline \
u64 ____##name(__BPF_MAP(x, __BPF_DECL_ARGS, __BPF_V, __VA_ARGS__))
#define BPF_CALL_1(name, ...) BPF_CALL_x(1, name, __VA_ARGS__)
The notrace attribute is actually applied to the static always_inline function
____bpf_spin_{lock,unlock}(). The actual callback function
bpf_spin_{lock,unlock}() is not marked with notrace, hence
allowing fentry prog to attach to two helpers, and this
may cause the above mentioned deadlock. Siddharth Chintamaneni
actually has a reproducer in [2].
To fix the issue, a new macro NOTRACE_BPF_CALL_1 is introduced which
will add notrace attribute to the original function instead of
the hidden always_inline function and this fixed the problem.
[1] https://lore.kernel.org/bpf/CAE5sdEigPnoGrzN8WU7Tx-h-iFuMZgW06qp0KHWtpvoXxf1OAQ@mail.gmail.com/
[2] https://lore.kernel.org/bpf/CAE5sdEg6yUc_Jz50AnUXEEUh6O73yQ1Z6NV2srJnef0ZrQkZew@mail.gmail.com/
Fixes: d83525ca62 ("bpf: introduce bpf_spin_lock")
Signed-off-by: Yonghong Song <yonghong.song@linux.dev>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/bpf/20240207070102.335167-1-yonghong.song@linux.dev
Since 20d59ee551 ("libbpf: add bpf_core_cast() macro"), libbpf is now
exporting a const arg version of bpf_rdonly_cast(). This causes the
following conflicting type error when generating kfunc prototypes from
BTF:
In file included from skeleton/pid_iter.bpf.c:5:
/home/dxu/dev/linux/tools/bpf/bpftool/bootstrap/libbpf/include/bpf/bpf_core_read.h:297:14: error: conflicting types for 'bpf_rdonly_cast'
extern void *bpf_rdonly_cast(const void *obj__ign, __u32 btf_id__k) __ksym __weak;
^
./vmlinux.h:135625:14: note: previous declaration is here
extern void *bpf_rdonly_cast(void *obj__ign, u32 btf_id__k) __weak __ksym;
This is b/c the kernel defines bpf_rdonly_cast() with non-const arg.
Since const arg is more permissive and thus backwards compatible, we
change the kernel definition as well to avoid conflicting type errors.
Signed-off-by: Daniel Xu <dxu@dxuuu.xyz>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Yonghong Song <yonghong.song@linux.dev>
Link: https://lore.kernel.org/bpf/dfd3823f11ffd2d4c838e961d61ec9ae8a646773.1707080349.git.dxu@dxuuu.xyz
Commit a8b9cf62ad ("ftrace: Fix DIRECT_CALLS to use SAVE_REGS by
default") attempted to fix an issue with direct trampolines on x86, see
its description for details. However, it wrongly referenced the
HAVE_DYNAMIC_FTRACE_WITH_REGS config option and the problem is still
present.
Add the missing "CONFIG_" prefix for the logic to work as intended.
Link: https://lore.kernel.org/linux-trace-kernel/20240213132434.22537-1-petr.pavlu@suse.com
Fixes: a8b9cf62ad ("ftrace: Fix DIRECT_CALLS to use SAVE_REGS by default")
Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Pull timer fix from Borislav Petkov:
- Make sure a warning is issued when a hrtimer gets queued after the
timers have been migrated on the CPU down path and thus said timer
will get ignored
* tag 'timers_urgent_for_v6.8_rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
hrtimer: Report offline hrtimer enqueue
Pull misc fixes from Andrew Morton:
"21 hotfixes. 12 are cc:stable and the remainder pertain to post-6.7
issues or aren't considered to be needed in earlier kernel versions"
* tag 'mm-hotfixes-stable-2024-02-10-11-16' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (21 commits)
nilfs2: fix potential bug in end_buffer_async_write
mm/damon/sysfs-schemes: fix wrong DAMOS tried regions update timeout setup
nilfs2: fix hang in nilfs_lookup_dirty_data_buffers()
MAINTAINERS: Leo Yan has moved
mm/zswap: don't return LRU_SKIP if we have dropped lru lock
fs,hugetlb: fix NULL pointer dereference in hugetlbs_fill_super
mailmap: switch email address for John Moon
mm: zswap: fix objcg use-after-free in entry destruction
mm/madvise: don't forget to leave lazy MMU mode in madvise_cold_or_pageout_pte_range()
arch/arm/mm: fix major fault accounting when retrying under per-VMA lock
selftests: core: include linux/close_range.h for CLOSE_RANGE_* macros
mm/memory-failure: fix crash in split_huge_page_to_list from soft_offline_page
mm: memcg: optimize parent iteration in memcg_rstat_updated()
nilfs2: fix data corruption in dsync block recovery for small block sizes
mm/userfaultfd: UFFDIO_MOVE implementation should use ptep_get()
exit: wait_task_zombie: kill the no longer necessary spin_lock_irq(siglock)
fs/proc: do_task_stat: use sig->stats_lock to gather the threads/children stats
fs/proc: do_task_stat: move thread_group_cputime_adjusted() outside of lock_task_sighand()
getrusage: use sig->stats_lock rather than lock_task_sighand()
getrusage: move thread_group_cputime_adjusted() outside of lock_task_sighand()
...
Turn kill_pid_info() into kill_pid_info_type(), this allows to pass any
pid_type to group_send_sig_info(), despite its name it should work fine
even if type = PIDTYPE_PID.
Change pidfd_send_signal() to use PIDTYPE_PID or PIDTYPE_TGID depending
on PIDFD_THREAD.
While at it kill another TODO comment in pidfd_show_fdinfo(). As Christian
expains fdinfo reports f_flags, userspace can already detect PIDFD_THREAD.
Reviewed-by: Tycho Andersen <tandersen@netflix.com>
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Link: https://lore.kernel.org/r/20240209130650.GA8048@redhat.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
