There is problem with installing an event in a task that is 'stuck' on
an offline CPU.

Blocked tasks are not dis-assosciated from offlined CPUs, after all, a
blocked task doesn't run and doesn't require a CPU etc.. Only on
wakeup do we ammend the situation and place the task on a available
CPU.

If we hit such a task with perf_install_in_context() we'll loop until
either that task wakes up or the CPU comes back online, if the task
waking depends on the event being installed, we're stuck.

While looking into this issue, I also spotted another problem, if we
hit a task with perf_install_in_context() that is in the middle of
being migrated, that is we observe the old CPU before sending the IPI,
but run the IPI (on the old CPU) while the task is already running on
the new CPU, things also go sideways.

Rework things to rely on task_curr() -- outside of rq->lock -- which
is rather tricky. Imagine the following scenario where we're trying to
install the first event into our task 't':

CPU0            CPU1            CPU2

                (current == t)

t->perf_event_ctxp[] = ctx;
smp_mb();
cpu = task_cpu(t);

                switch(t, n);
                                migrate(t, 2);
                                switch(p, t);

                                ctx = t->perf_event_ctxp[]; // must not be NULL

smp_function_call(cpu, ..);

                generic_exec_single()
                  func();
                    spin_lock(ctx->lock);
                    if (task_curr(t)) // false

                    add_event_to_ctx();
                    spin_unlock(ctx->lock);

                                perf_event_context_sched_in();
                                  spin_lock(ctx->lock);
                                  // sees event

So its CPU0's store of t->perf_event_ctxp[] that must not go 'missing'.
Because if CPU2's load of that variable were to observe NULL, it would
not try to schedule the ctx and we'd have a task running without its
counter, which would be 'bad'.

As long as we observe !NULL, we'll acquire ctx->lock. If we acquire it
first and not see the event yet, then CPU0 must observe task_curr()
and retry. If the install happens first, then we must see the event on
sched-in and all is well.

I think we can translate the first part (until the 'must not be NULL')
of the scenario to a litmus test like:

  C C-peterz

  {
  }

  P0(int *x, int *y)
  {
          int r1;

          WRITE_ONCE(*x, 1);
          smp_mb();
          r1 = READ_ONCE(*y);
  }

  P1(int *y, int *z)
  {
          WRITE_ONCE(*y, 1);
          smp_store_release(z, 1);
  }

  P2(int *x, int *z)
  {
          int r1;
          int r2;

          r1 = smp_load_acquire(z);
	  smp_mb();
          r2 = READ_ONCE(*x);
  }

  exists
  (0:r1=0 /\ 2:r1=1 /\ 2:r2=0)

Where:
  x is perf_event_ctxp[],
  y is our tasks's CPU, and
  z is our task being placed on the rq of CPU2.

The P0 smp_mb() is the one added by this patch, ordering the store to
perf_event_ctxp[] from find_get_context() and the load of task_cpu()
in task_function_call().

The smp_store_release/smp_load_acquire model the RCpc locking of the
rq->lock and the smp_mb() of P2 is the context switch switching from
whatever CPU2 was running to our task 't'.

This litmus test evaluates into:

  Test C-peterz Allowed
  States 7
  0:r1=0; 2:r1=0; 2:r2=0;
  0:r1=0; 2:r1=0; 2:r2=1;
  0:r1=0; 2:r1=1; 2:r2=1;
  0:r1=1; 2:r1=0; 2:r2=0;
  0:r1=1; 2:r1=0; 2:r2=1;
  0:r1=1; 2:r1=1; 2:r2=0;
  0:r1=1; 2:r1=1; 2:r2=1;
  No
  Witnesses
  Positive: 0 Negative: 7
  Condition exists (0:r1=0 /\ 2:r1=1 /\ 2:r2=0)
  Observation C-peterz Never 0 7
  Hash=e427f41d9146b2a5445101d3e2fcaa34

And the strong and weak model agree.
Reported-by: NMark Rutland <mark.rutland@arm.com>
Tested-by: NMark Rutland <mark.rutland@arm.com>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: Will Deacon <will.deacon@arm.com>
Cc: jeremy.linton@arm.com
Link: http://lkml.kernel.org/r/20161209135900.GU3174@twins.programming.kicks-ass.netSigned-off-by: NIngo Molnar <mingo@kernel.org>

The warning introduced in commit:

  864c2357 ("perf/core: Do not set cpuctx->cgrp for unscheduled cgroups")

assumed that a cgroup switch always precedes list_del_event. This is
not the case. Remove warning.

Make sure that cpuctx->cgrp is NULL until a cgroup event is sched in
or ctx->nr_cgroups == 0.
Signed-off-by: NDavid Carrillo-Cisneros <davidcc@google.com>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Nilay Vaish <nilayvaish@gmail.com>
Cc: Paul Turner <pjt@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ravi V Shankar <ravi.v.shankar@intel.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vegard Nossum <vegard.nossum@gmail.com>
Cc: Vikas Shivappa <vikas.shivappa@linux.intel.com>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/1480841177-27299-1-git-send-email-davidcc@google.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Since long already bpf_func is not only about struct sk_buff * as
input anymore. Make it generic as void *, so that callers don't
need to cast for it each time they call BPF_PROG_RUN().
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Acked-by: NAlexei Starovoitov <ast@kernel.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The token table passed into match_token() must be null-terminated, which
it currently is not in the perf's address filter string parser, as caught
by Vince's perf_fuzzer and KASAN.

It doesn't blow up otherwise because of the alignment padding of the table
to the next element in the .rodata, which is luck.

Fixing by adding a null-terminator to the token table.
Reported-by: NVince Weaver <vincent.weaver@maine.edu>
Tested-by: NVince Weaver <vincent.weaver@maine.edu>
Signed-off-by: NAlexander Shishkin <alexander.shishkin@linux.intel.com>
Acked-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: dvyukov@google.com
Cc: stable@vger.kernel.org # v4.7+
Fixes: 375637bc ("perf/core: Introduce address range filtering")
Link: http://lkml.kernel.org/r/877f81f264.fsf@ashishki-desk.ger.corp.intel.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

Commit:

  db4a8356 ("perf/core: Set cgroup in CPU contexts for new cgroup events")

failed to verify that event->cgrp is actually the scheduled cgroup
in a CPU before setting cpuctx->cgrp. This patch fixes that.

Now that there is a different path for scheduled and unscheduled
cgroup, add a warning to catch when cpuctx->cgrp is still set after
the last cgroup event has been unsheduled.

To verify the bug:

  # Create 2 cgroups.
  mkdir /dev/cgroups/devices/g1
  mkdir /dev/cgroups/devices/g2

  # launch a task, bind it to a cpu and move it to g1
  CPU=2
  while :; do : ; done &
  P=$!

  taskset -pc $CPU $P
  echo $P > /dev/cgroups/devices/g1/tasks

  # monitor g2 (it runs no tasks) and observe output
  perf stat -e cycles -I 1000 -C $CPU -G g2

  #           time             counts unit events
     1.000091408          7,579,527      cycles                    g2
     2.000350111      <not counted>      cycles                    g2
     3.000589181      <not counted>      cycles                    g2
     4.000771428      <not counted>      cycles                    g2

  # note first line that displays that a task run in g2, despite
  # g2 having no tasks. This is because cpuctx->cgrp was wrongly
  # set when context of new event was installed.
  # After applying the fix we obtain the right output:

  perf stat -e cycles -I 1000 -C $CPU -G g2
  #           time             counts unit events
     1.000119615      <not counted>      cycles                    g2
     2.000389430      <not counted>      cycles                    g2
     3.000590962      <not counted>      cycles                    g2
Signed-off-by: NDavid Carrillo-Cisneros <davidcc@google.com>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Nilay Vaish <nilayvaish@gmail.com>
Cc: Paul Turner <pjt@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vegard Nossum <vegard.nossum@gmail.com>
Link: http://lkml.kernel.org/r/1478026378-86083-1-git-send-email-davidcc@google.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

The trinity syscall fuzzer triggered following WARN() on powerpc:

  WARNING: CPU: 9 PID: 2998 at arch/powerpc/kernel/hw_breakpoint.c:278
  ...
  NIP [c00000000093aedc] .hw_breakpoint_handler+0x28c/0x2b0
  LR [c00000000093aed8] .hw_breakpoint_handler+0x288/0x2b0
  Call Trace:
  [c0000002f7933580] [c00000000093aed8] .hw_breakpoint_handler+0x288/0x2b0 (unreliable)
  [c0000002f7933630] [c0000000000f671c] .notifier_call_chain+0x7c/0xf0
  [c0000002f79336d0] [c0000000000f6abc] .__atomic_notifier_call_chain+0xbc/0x1c0
  [c0000002f7933780] [c0000000000f6c40] .notify_die+0x70/0xd0
  [c0000002f7933820] [c00000000001a74c] .do_break+0x4c/0x100
  [c0000002f7933920] [c0000000000089fc] handle_dabr_fault+0x14/0x48

Followed by a lockdep warning:

  ===============================
  [ INFO: suspicious RCU usage. ]
  4.8.0-rc5+ #7 Tainted: G        W
  -------------------------------
  ./include/linux/rcupdate.h:556 Illegal context switch in RCU read-side critical section!

  other info that might help us debug this:

  rcu_scheduler_active = 1, debug_locks = 0
  2 locks held by ls/2998:
   #0:  (rcu_read_lock){......}, at: [<c0000000000f6a00>] .__atomic_notifier_call_chain+0x0/0x1c0
   #1:  (rcu_read_lock){......}, at: [<c00000000093ac50>] .hw_breakpoint_handler+0x0/0x2b0

  stack backtrace:
  CPU: 9 PID: 2998 Comm: ls Tainted: G        W       4.8.0-rc5+ #7
  Call Trace:
  [c0000002f7933150] [c00000000094b1f8] .dump_stack+0xe0/0x14c (unreliable)
  [c0000002f79331e0] [c00000000013c468] .lockdep_rcu_suspicious+0x138/0x180
  [c0000002f7933270] [c0000000001005d8] .___might_sleep+0x278/0x2e0
  [c0000002f7933300] [c000000000935584] .mutex_lock_nested+0x64/0x5a0
  [c0000002f7933410] [c00000000023084c] .perf_event_ctx_lock_nested+0x16c/0x380
  [c0000002f7933500] [c000000000230a80] .perf_event_disable+0x20/0x60
  [c0000002f7933580] [c00000000093aeec] .hw_breakpoint_handler+0x29c/0x2b0
  [c0000002f7933630] [c0000000000f671c] .notifier_call_chain+0x7c/0xf0
  [c0000002f79336d0] [c0000000000f6abc] .__atomic_notifier_call_chain+0xbc/0x1c0
  [c0000002f7933780] [c0000000000f6c40] .notify_die+0x70/0xd0
  [c0000002f7933820] [c00000000001a74c] .do_break+0x4c/0x100
  [c0000002f7933920] [c0000000000089fc] handle_dabr_fault+0x14/0x48

While it looks like the first WARN() is probably valid, the other one is
triggered by disabling event via perf_event_disable() from atomic context.

The event is disabled here in case we were not able to emulate
the instruction that hit the breakpoint. By disabling the event
we unschedule the event and make sure it's not scheduled back.

But we can't call perf_event_disable() from atomic context, instead
we need to use the event's pending_disable irq_work method to disable it.
Reported-by: NJan Stancek <jstancek@redhat.com>
Signed-off-by: NJiri Olsa <jolsa@kernel.org>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Huang Ying <ying.huang@intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Michael Neuling <mikey@neuling.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20161026094824.GA21397@kravaSigned-off-by: NIngo Molnar <mingo@kernel.org>

perf/core: Protect PMU device removal with a 'pmu_bus_running' check, to fix CONFIG_DEBUG_TEST_DRIVER_REMOVE=y kernel panic

CAI Qian reported a crash in the PMU uncore device removal code,
enabled by the CONFIG_DEBUG_TEST_DRIVER_REMOVE=y option:

  https://marc.info/?l=linux-kernel&m=147688837328451

The reason for the crash is that perf_pmu_unregister() tries to remove
a PMU device which is not added at this point. We add PMU devices
only after pmu_bus is registered, which happens in the
perf_event_sysfs_init() call and sets the 'pmu_bus_running' flag.

The fix is to get the 'pmu_bus_running' flag state at the point
the PMU is taken out of the PMU list and remove the device
later only if it's set.
Reported-by: NCAI Qian <caiqian@redhat.com>
Tested-by: NCAI Qian <caiqian@redhat.com>
Signed-off-by: NJiri Olsa <jolsa@kernel.org>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rob Herring <robh@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20161020111011.GA13361@kravaSigned-off-by: NIngo Molnar <mingo@kernel.org>

An "exclusive" PMU is the one that can only have one event scheduled in
at any given time. There may be more than one of such PMUs in a system,
though, like Intel PT and BTS. It should be allowed to have one event
for either of those inside the same context (there may be other constraints
that may prevent this, but those would be hardware-specific). However,
the exclusivity code is written so that only one event from any of the
"exclusive" PMUs is allowed in a context.

Fix this by making the exclusive event filter explicitly match two events'
PMUs.
Signed-off-by: NAlexander Shishkin <alexander.shishkin@linux.intel.com>
Acked-by: NPeter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Arnaldo Carvalho de Melo <acme@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: vince@deater.net
Link: http://lkml.kernel.org/r/20160920154811.3255-3-alexander.shishkin@linux.intel.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

In the mmap_close() path we need to stop all the AUX events that are
writing data to the AUX area that we are unmapping, before we can
safely free the pages. To determine if an event needs to be stopped,
we're comparing its ->rb against the one that's getting unmapped.
However, a SET_OUTPUT ioctl may turn up inside an AUX transaction
and swizzle event::rb to some other ring buffer, but the transaction
will keep writing data to the old ring buffer until the event gets
scheduled out. At this point, mmap_close() will skip over such an
event and will proceed to free the AUX area, while it's still being
used by this event, which will set off a warning in the mmap_close()
path and cause a memory corruption.

To avoid this, always stop an AUX event before its ->rb is updated;
this will release the (potentially) last reference on the AUX area
of the buffer. If the event gets restarted, its new ring buffer will
be used. If another SET_OUTPUT comes and switches it back to the
old ring buffer that's getting unmapped, it's also fine: this
ring buffer's aux_mmap_count will be zero and AUX transactions won't
start any more.
Reported-by: NVince Weaver <vincent.weaver@maine.edu>
Signed-off-by: NAlexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: vince@deater.net
Link: http://lkml.kernel.org/r/20160906132353.19887-2-alexander.shishkin@linux.intel.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

The newly added bpf_overflow_handler function is only built of both
CONFIG_EVENT_TRACING and CONFIG_BPF_SYSCALL are enabled, but the caller
only checks the latter:

kernel/events/core.c: In function 'perf_event_alloc':
kernel/events/core.c:9106:27: error: 'bpf_overflow_handler' undeclared (first use in this function)

This changes the caller so we also skip this call if CONFIG_EVENT_TRACING
is disabled entirely.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Fixes: aa6a5f3c ("perf, bpf: add perf events core support for BPF_PROG_TYPE_PERF_EVENT programs")
Acked-by: NAlexei Starovoitov <ast@kernel.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

PERF_EF_START is a flag to indicate to the PMU ->add() callback that, as
well as claiming the PMU resources required by the event being added,
it should also start the PMU.

Passing this flag to the ->start() callback doesn't make sense, because
->start() always tries to start the PMU. Remove it.
Signed-off-by: NWill Deacon <will.deacon@arm.com>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: mark.rutland@arm.com
Link: http://lkml.kernel.org/r/1471257765-29662-1-git-send-email-will.deacon@arm.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

This effectively reverts commit:

  71e7bc2b ("perf/core: Check return value of the perf_event_read() IPI")

... and puts in a comment explaining why we ignore the return value.
Reported-by: NVegard Nossum <vegard.nossum@gmail.com>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: David Carrillo-Cisneros <davidcc@google.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Fixes: 71e7bc2b ("perf/core: Check return value of the perf_event_read() IPI")
Signed-off-by: NIngo Molnar <mingo@kernel.org>

Allow attaching BPF_PROG_TYPE_PERF_EVENT programs to sw and hw perf events
via overflow_handler mechanism.
When program is attached the overflow_handlers become stacked.
The program acts as a filter.
Returning zero from the program means that the normal perf_event_output handler
will not be called and sampling event won't be stored in the ring buffer.

The overflow_handler_context==NULL is an additional safety check
to make sure programs are not attached to hw breakpoints and watchdog
in case other checks (that prevent that now anyway) get accidentally
relaxed in the future.

The program refcnt is incremented in case perf_events are inhereted
when target task is forked.
Similar to kprobe and tracepoint programs there is no ioctl to
detach the program or swap already attached program. The user space
expected to close(perf_event_fd) like it does right now for kprobe+bpf.
That restriction simplifies the code quite a bit.

The invocation of overflow_handler in __perf_event_overflow() is now
done via READ_ONCE, since that pointer can be replaced when the program
is attached while perf_event itself could have been active already.
There is no need to do similar treatment for event->prog, since it's
assigned only once before it's accessed.
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When tearing down an AUX buf for an event via perf_mmap_close(),
__perf_event_output_stop() is called on the event's CPU to ensure that
trace generation is halted before the process of unmapping and
freeing the buffer pages begins.

The callback is performed via cpu_function_call(), which ensures that it
runs with interrupts disabled and is therefore not preemptible.
Unfortunately, the current code grabs the per-cpu context pointer using
get_cpu_ptr(), which unnecessarily disables preemption and doesn't pair
the call with put_cpu_ptr(), leading to a preempt_count() imbalance and
a BUG when freeing the AUX buffer later on:

  WARNING: CPU: 1 PID: 2249 at kernel/events/ring_buffer.c:539 __rb_free_aux+0x10c/0x120
  Modules linked in:
  [...]
  Call Trace:
   [<ffffffff813379dd>] dump_stack+0x4f/0x72
   [<ffffffff81059ff6>] __warn+0xc6/0xe0
   [<ffffffff8105a0c8>] warn_slowpath_null+0x18/0x20
   [<ffffffff8112761c>] __rb_free_aux+0x10c/0x120
   [<ffffffff81128163>] rb_free_aux+0x13/0x20
   [<ffffffff8112515e>] perf_mmap_close+0x29e/0x2f0
   [<ffffffff8111da30>] ? perf_iterate_ctx+0xe0/0xe0
   [<ffffffff8115f685>] remove_vma+0x25/0x60
   [<ffffffff81161796>] exit_mmap+0x106/0x140
   [<ffffffff8105725c>] mmput+0x1c/0xd0
   [<ffffffff8105cac3>] do_exit+0x253/0xbf0
   [<ffffffff8105e32e>] do_group_exit+0x3e/0xb0
   [<ffffffff81068d49>] get_signal+0x249/0x640
   [<ffffffff8101c273>] do_signal+0x23/0x640
   [<ffffffff81905f42>] ? _raw_write_unlock_irq+0x12/0x30
   [<ffffffff81905f69>] ? _raw_spin_unlock_irq+0x9/0x10
   [<ffffffff81901896>] ? __schedule+0x2c6/0x710
   [<ffffffff810022a4>] exit_to_usermode_loop+0x74/0x90
   [<ffffffff81002a56>] prepare_exit_to_usermode+0x26/0x30
   [<ffffffff81906d1b>] retint_user+0x8/0x10

This patch uses this_cpu_ptr() instead of get_cpu_ptr(), since preemption is
already disabled by the caller.
Signed-off-by: NWill Deacon <will.deacon@arm.com>
Reviewed-by: NAlexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Fixes: 95ff4ca2 ("perf/core: Free AUX pages in unmap path")
Link: http://lkml.kernel.org/r/20160824091905.GA16944@arm.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

Introduce the flag PMU_EV_CAP_READ_ACTIVE_PKG, useful for uncore events,
that allows a PMU to signal the generic perf code that an event is readable
in the current CPU if the event is active in a CPU in the same package as
the current CPU.

This is an optimization that avoids a unnecessary IPI for the common case
where uncore events are run and read in the same package but in
different CPUs.

As an example, the IPI removal speeds up perf_read() in my Haswell system
as follows:

  - For event UNC_C_LLC_LOOKUP: From 260 us to 31 us.
  - For event RAPL's power/energy-cores/: From to 255 us to 27 us.

For the optimization to work, all events in the group must have it
(similarly to PERF_EV_CAP_SOFTWARE).
Signed-off-by: NDavid Carrillo-Cisneros <davidcc@google.com>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: David Carrillo-Cisneros <davidcc@google.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Paul Turner <pjt@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vegard Nossum <vegard.nossum@gmail.com>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/1471467307-61171-4-git-send-email-davidcc@google.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

Currently, PERF_GROUP_SOFTWARE is used in the group_flags field of a
group's leader to indicate that is_software_event(event) is true for all
events in a group. This is the only usage of event->group_flags.

This pattern of setting a group level flags when all events in the group
share a property is useful for the flag introduced in the next patch and
for future CQM/CMT flags. So this patches generalizes group_flags to work
as an aggregate of event level flags.

PERF_GROUP_SOFTWARE denotes an inmutable event's property. All other flags
that I intend to add are also determinable at event initialization.
To better convey the above, this patch renames event's group_flags to
group_caps and PERF_GROUP_SOFTWARE to PERF_EV_CAP_SOFTWARE.

Individual event flags are stored in the new event->event_caps. Since the
cap flags do not change after event initialization, there is no need to
serialize event_caps. This new field is used when events are added to a
context, similarly to how PERF_GROUP_SOFTWARE and is_software_event()
worked.

Lastly, for consistency, updates is_software_event() to rely in event_cap
instead of the context index.
Signed-off-by: NDavid Carrillo-Cisneros <davidcc@google.com>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Paul Turner <pjt@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vegard Nossum <vegard.nossum@gmail.com>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/1471467307-61171-3-git-send-email-davidcc@google.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

When decoding the perf_regs mask in perf_output_sample_regs(),
we loop through the mask using find_first_bit and find_next_bit functions.

While the exisiting code works fine in most of the case, the logic
is broken for big-endian 32-bit kernels.

When reading a u64 mask using (u32 *)(&val)[0], find_*_bit() assumes
that it gets the lower 32 bits of u64, but instead it gets the upper
32 bits - which is wrong.

The fix is to swap the words of the u64 to handle this case.
This is _not_ a regular endianness swap.
Suggested-by: NYury Norov <ynorov@caviumnetworks.com>
Signed-off-by: NMadhavan Srinivasan <maddy@linux.vnet.ibm.com>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: NYury Norov <ynorov@caviumnetworks.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: linuxppc-dev@lists.ozlabs.org
Link: http://lkml.kernel.org/r/1471426568-31051-2-git-send-email-maddy@linux.vnet.ibm.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

The call to smp_call_function_single in perf_event_read() may fail if
an invalid or not online CPU index is passed. Warn user if such bug is
present and return error.
Signed-off-by: NDavid Carrillo-Cisneros <davidcc@google.com>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Paul Turner <pjt@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vegard Nossum <vegard.nossum@gmail.com>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/1471467307-61171-2-git-send-email-davidcc@google.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

At this time the perf_addr_filter_needs_mmap() function will _not_
return true on a user space 'stop' filter.  But stop filters need
exactly the same kind of mapping that range and start filters get.
Signed-off-by: NMathieu Poirier <mathieu.poirier@linaro.org>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: NAlexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/1468860187-318-4-git-send-email-mathieu.poirier@linaro.orgSigned-off-by: NIngo Molnar <mingo@kernel.org>

Function perf_event_mmap() is called by the MM subsystem each time
part of a binary is loaded in memory.  There can be several mapping
for a binary, many times unrelated to the code section.

Each time a section of a binary is mapped address filters are
updated, event when the map doesn't pertain to the code section.
The end result is that filters are configured based on the last map
event that was received rather than the last mapping of the code
segment.

For example if we have an executable 'main' that calls library
'libcstest.so.1.0', and that we want to collect traces on code
that is in that library.  The perf cmd line for this scenario
would be:

  perf record -e cs_etm// --filter 'filter 0x72c/0x40@/opt/lib/libcstest.so.1.0' --per-thread ./main

Resulting in binaries being mapped this way:

  root@linaro-nano:~# cat /proc/1950/maps
  00400000-00401000 r-xp 00000000 08:02 33169     /home/linaro/main
  00410000-00411000 r--p 00000000 08:02 33169     /home/linaro/main
  00411000-00412000 rw-p 00001000 08:02 33169     /home/linaro/main
  7fa2464000-7fa2474000 rw-p 00000000 00:00 0
  7fa2474000-7fa25a4000 r-xp 00000000 08:02 543   /lib/aarch64-linux-gnu/libc-2.21.so
  7fa25a4000-7fa25b3000 ---p 00130000 08:02 543   /lib/aarch64-linux-gnu/libc-2.21.so
  7fa25b3000-7fa25b7000 r--p 0012f000 08:02 543   /lib/aarch64-linux-gnu/libc-2.21.so
  7fa25b7000-7fa25b9000 rw-p 00133000 08:02 543   /lib/aarch64-linux-gnu/libc-2.21.so
  7fa25b9000-7fa25bd000 rw-p 00000000 00:00 0
  7fa25bd000-7fa25be000 r-xp 00000000 08:02 38308 /opt/lib/libcstest.so.1.0
  7fa25be000-7fa25cd000 ---p 00001000 08:02 38308 /opt/lib/libcstest.so.1.0
  7fa25cd000-7fa25ce000 r--p 00000000 08:02 38308 /opt/lib/libcstest.so.1.0
  7fa25ce000-7fa25cf000 rw-p 00001000 08:02 38308 /opt/lib/libcstest.so.1.0
  7fa25cf000-7fa25eb000 r-xp 00000000 08:02 574   /lib/aarch64-linux-gnu/ld-2.21.so
  7fa25ef000-7fa25f2000 rw-p 00000000 00:00 0
  7fa25f7000-7fa25f9000 rw-p 00000000 00:00 0
  7fa25f9000-7fa25fa000 r--p 00000000 00:00 0     [vvar]
  7fa25fa000-7fa25fb000 r-xp 00000000 00:00 0     [vdso]
  7fa25fb000-7fa25fc000 r--p 0001c000 08:02 574   /lib/aarch64-linux-gnu/ld-2.21.so
  7fa25fc000-7fa25fe000 rw-p 0001d000 08:02 574   /lib/aarch64-linux-gnu/ld-2.21.so
  7ff2ea8000-7ff2ec9000 rw-p 00000000 00:00 0     [stack]
  root@linaro-nano:~#

Before 'main()' can execute 'libcstest.so.1.0' has to be loaded in
memory.  Once that has been done perf_event_mmap() has been called
4 times, with the last map starting at address 0x7fa25ce000 and
the address filter configured to start filtering when the
IP has passed over address 0x0x7fa25ce72c (0x7fa25ce000 + 0x72c).

But that is wrong since the code segment for library 'libcstest.so.1.0'
as been mapped at 0x7fa25bd000, resulting in traces not being
collected.

This patch corrects the situation by requesting that address
filters be updated only if the mapped event is for a code
segment.
Signed-off-by: NMathieu Poirier <mathieu.poirier@linaro.org>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: NAlexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/1468860187-318-3-git-send-email-mathieu.poirier@linaro.orgSigned-off-by: NIngo Molnar <mingo@kernel.org>

Binary file names have to be supplied for both range and start/stop
filters but the current code only processes the filename if an
address range filter is specified.  This code adds processing of
the filename for start/stop filters.
Signed-off-by: NMathieu Poirier <mathieu.poirier@linaro.org>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: NAlexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/1468860187-318-2-git-send-email-mathieu.poirier@linaro.orgSigned-off-by: NIngo Molnar <mingo@kernel.org>

Vincent reported triggering the WARN_ON_ONCE() in event_function_local().

While thinking through cases I noticed that by using event_function()
directly, we miss the inactive case usually handled by
event_function_call().

Therefore construct a blend of event_function_call() and
event_function() that handles the cases relevant to
event_function_local().
Reported-by: NVince Weaver <vincent.weaver@maine.edu>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org # 4.5+
Fixes: fae3fde6 ("perf: Collapse and fix event_function_call() users")
Signed-off-by: NIngo Molnar <mingo@kernel.org>

For perf record -b, which requires the pmu::sched_task callback the
current code is rather expensive:

     7.68%  sched-pipe  [kernel.vmlinux]    [k] perf_pmu_sched_task
     5.95%  sched-pipe  [kernel.vmlinux]    [k] __switch_to
     5.20%  sched-pipe  [kernel.vmlinux]    [k] __intel_pmu_disable_all
     3.95%  sched-pipe  perf                [.] worker_thread

The problem is that it will iterate all registered PMUs, most of which
will not have anything to do. Avoid this by keeping an explicit list
of PMUs that have requested the callback.

The perf_sched_cb_{inc,dec}() functions already takes the required pmu
argument, and now that these functions are no longer called from NMI
context we can use them to manage a list.

With this patch applied the function doesn't show up in the top 4
anymore (it dropped to 18th place).

     6.67%  sched-pipe  [kernel.vmlinux]    [k] __switch_to
     6.18%  sched-pipe  [kernel.vmlinux]    [k] __intel_pmu_disable_all
     3.92%  sched-pipe  [kernel.vmlinux]    [k] switch_mm_irqs_off
     3.71%  sched-pipe  perf                [.] worker_thread
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: NIngo Molnar <mingo@kernel.org>

In order to allow optimizing perf_pmu_sched_task() we must ensure
perf_sched_cb_{inc,dec}() are no longer called from NMI context; this
means that pmu::{start,stop}() can no longer use them.

Prepare for this by reworking the whole large PEBS setup code.

The current code relied on the cpuc->pebs_enabled state, however since
that reflects the current active state as per pmu::{start,stop}() we
can no longer rely on this.

Introduce two counters: cpuc->n_pebs and cpuc->n_large_pebs which
count the total number of PEBS events and the number of PEBS events
that have FREERUNNING set, resp.. With this we can tell if the current
setup requires a single record interrupt threshold or can use a larger
buffer.

This also improves the code in that it re-enables the large threshold
once the PEBS event that required single record gets removed.
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: NIngo Molnar <mingo@kernel.org>

Groups of events are supposed to be scheduled atomically, such that it
is possible to derive meaningful ratios between their values.

We take great pains to achieve this when scheduling event groups to a
PMU in group_sched_in(), calling {start,commit}_txn() (which fall back
to perf_pmu_{disable,enable}() if necessary) to provide this guarantee.
However we don't mirror this in group_sched_out(), and in some cases
events will not be scheduled out atomically.

For example, if we disable an event group with PERF_EVENT_IOC_DISABLE,
we'll cross-call __perf_event_disable() for the group leader, and will
call group_sched_out() without having first disabled the relevant PMU.
We will disable/enable the PMU around each pmu->del() call, but between
each call the PMU will be enabled and events may count.

Avoid this by explicitly disabling and enabling the PMU around event
removal in group_sched_out(), mirroring what we do in group_sched_in().
Signed-off-by: NMark Rutland <mark.rutland@arm.com>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/1469553141-28314-1-git-send-email-mark.rutland@arm.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

There's a perf stat bug easy to observer on a machine with only one cgroup:

  $ perf stat -e cycles -I 1000 -C 0 -G /
  #          time             counts unit events
      1.000161699      <not counted>      cycles                    /
      2.000355591      <not counted>      cycles                    /
      3.000565154      <not counted>      cycles                    /
      4.000951350      <not counted>      cycles                    /

We'd expect some output there.

The underlying problem is that there is an optimization in
perf_cgroup_sched_{in,out}() that skips the switch of cgroup events
if the old and new cgroups in a task switch are the same.

This optimization interacts with the current code in two ways
that cause a CPU context's cgroup (cpuctx->cgrp) to be NULL even if a
cgroup event matches the current task. These are:

  1. On creation of the first cgroup event in a CPU: In current code,
  cpuctx->cpu is only set in perf_cgroup_sched_in, but due to the
  aforesaid optimization, perf_cgroup_sched_in will run until the next
  cgroup switches in that CPU. This may happen late or never happen,
  depending on system's number of cgroups, CPU load, etc.

  2. On deletion of the last cgroup event in a cpuctx: In list_del_event,
  cpuctx->cgrp is set NULL. Any new cgroup event will not be sched in
  because cpuctx->cgrp == NULL until a cgroup switch occurs and
  perf_cgroup_sched_in is executed (updating cpuctx->cgrp).

This patch fixes both problems by setting cpuctx->cgrp in list_add_event,
mirroring what list_del_event does when removing a cgroup event from CPU
context, as introduced in:

  commit 68cacd29 ("perf_events: Fix stale ->cgrp pointer in update_cgrp_time_from_cpuctx()")

With this patch, cpuctx->cgrp is always set/clear when installing/removing
the first/last cgroup event in/from the CPU context. With cpuctx->cgrp
correctly set, event_filter_match works as intended when events are
sched in/out.

After the fix, the output is as expected:

  $ perf stat -e cycles -I 1000 -a -G /
  #         time             counts unit events
     1.004699159          627342882      cycles                    /
     2.007397156          615272690      cycles                    /
     3.010019057          616726074      cycles                    /
Signed-off-by: NDavid Carrillo-Cisneros <davidcc@google.com>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Paul Turner <pjt@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vegard Nossum <vegard.nossum@gmail.com>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/1470124092-113192-1-git-send-email-davidcc@google.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

Vegard Nossum reported that perf fuzzing generates a NULL
pointer dereference crash:

> Digging a bit deeper into this, it seems the event itself is getting
> created by perf_event_open() and it gets added to the pmu_event_list
> through:
>
> perf_event_open()
>  - perf_event_alloc()
>     - account_event()
>        - account_pmu_sb_event()
>           - attach_sb_event()
>
> so at this point the event is being attached but its ->ctx is still
> NULL. It seems like ->ctx is set just a bit later in
> perf_event_open(), though.
>
> But before that, __schedule() comes along and creates a stack trace
> similar to the one above:
>
> __schedule()
>  - __perf_event_task_sched_out()
>    - perf_iterate_sb()
>      - perf_iterate_sb_cpu()
>         - event_filter_match()
>           - perf_cgroup_match()
>             - __get_cpu_context()
>               - (dereference ctx which is NULL)
>
> So I guess the question is... should the event be attached (= put on
> the list) before ->ctx gets set? Or should the cgroup code check for a
> NULL ->ctx?

The latter seems like the simplest solution. Moving the list-add later
creates a bit of a mess.
Reported-by: NVegard Nossum <vegard.nossum@gmail.com>
Tested-by: NVegard Nossum <vegard.nossum@gmail.com>
Tested-by: NVince Weaver <vincent.weaver@maine.edu>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: David Carrillo-Cisneros <davidcc@google.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Fixes: f2fb6bef ("perf/core: Optimize side-band event delivery")
Link: http://lkml.kernel.org/r/20160804123724.GN6862@twins.programming.kicks-ass.netSigned-off-by: NIngo Molnar <mingo@kernel.org>

When the perf interrupt handler exceeds a threshold warning messages
are displayed on console:

  [12739.31793] perf interrupt took too long (2504 > 2500), lowering kernel.perf_event_max_sample_rate to 50000
  [71340.165065] perf interrupt took too long (5005 > 5000), lowering kernel.perf_event_max_sample_rate to 25000

Many customers and users are confused by the message wondering if
something is wrong or they need to take action to fix a problem.
Since a user can not do anything to fix the issue, the message is really
more informational than a warning. Adjust the log level accordingly.
Signed-off-by: NDavid Ahern <dsa@cumulusnetworks.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/1470084569-438-1-git-send-email-dsa@cumulusnetworks.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

This patch adds support for non-linear data on raw records. It
extends raw records to have one or multiple fragments that will
be written linearly into the ring slot, where each fragment can
optionally have a custom callback handler to walk and extract
complex, possibly non-linear data.

If a callback handler is provided for a fragment, then the new
__output_custom() will be used instead of __output_copy() for
the perf_output_sample() part. perf_prepare_sample() does all
the size calculation only once, so perf_output_sample() doesn't
need to redo the same work anymore, meaning real_size and padding
will be cached in the raw record. The raw record becomes 32 bytes
in size without holes; to not increase it further and to avoid
doing unnecessary recalculations in fast-path, we can reuse
next pointer of the last fragment, idea here is borrowed from
ZERO_OR_NULL_PTR(), which should keep the perf_output_sample()
path for PERF_SAMPLE_RAW minimal.

This facility is needed for BPF's event output helper as a first
user that will, in a follow-up, add an additional perf_raw_frag
to its perf_raw_record in order to be able to more efficiently
dump skb context after a linear head meta data related to it.
skbs can be non-linear and thus need a custom output function to
dump buffers. Currently, the skb data needs to be copied twice;
with the help of __output_custom() this work only needs to be
done once. Future users could be things like XDP/BPF programs
that work on different context though and would thus also have
a different callback function.

The few users of raw records are adapted to initialize their frag
data from the raw record itself, no change in behavior for them.
The code is based upon a PoC diff provided by Peter Zijlstra [1].

  [1] http://thread.gmane.org/gmane.linux.network/421294Suggested-by: NPeter Zijlstra <peterz@infradead.org>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Acked-by: NAlexei Starovoitov <ast@kernel.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Actually a nice symmetric startup/teardown pair which fits properly into
the state machine concept. In the long run we should be able to invoke
the startup callback for the boot CPU via the state machine and get
rid of the init function which invokes it on the boot CPU.

Note: This comes actually before the perf hardware callbacks. In the notifier
model the hardware callbacks have a higher priority than the core
callback. But that's solely for CPU offline so that hardware migration of
events happens before the core is notified about the outgoing CPU.

With the symetric state array model we have the following ordering:

 UP:     core -> hardware
 DOWN:   hardware -> core
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NAnna-Maria Gleixner <anna-maria@linutronix.de>
Reviewed-by: NSebastian Siewior <bigeasy@linutronix.de>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Cc: rt@linutronix.de
Link: http://lkml.kernel.org/r/20160713153333.587514098@linutronix.deSigned-off-by: NIngo Molnar <mingo@kernel.org>

The following commit:

  66eb579e ("perf: allow for PMU-specific event filtering")

added the pmu::filter_match() callback. This was intended to
avoid HW constraints on events from resulting in extremely
pessimistic scheduling.

However, pmu::filter_match() is only called for the leader of each event
group. When the leader is a SW event, we do not filter the groups, and
may fail at pmu::add() time, and when this happens we'll give up on
scheduling any event groups later in the list until they are rotated
ahead of the failing group.

This can result in extremely sub-optimal event scheduling behaviour,
e.g. if running the following on a big.LITTLE platform:

$ taskset -c 0 ./perf stat \
 -e 'a57{context-switches,armv8_cortex_a57/config=0x11/}' \
 -e 'a53{context-switches,armv8_cortex_a53/config=0x11/}' \
 ls

     <not counted>      context-switches                                              (0.00%)
     <not counted>      armv8_cortex_a57/config=0x11/                                 (0.00%)
                24      context-switches                                              (37.36%)
          57589154      armv8_cortex_a53/config=0x11/                                 (37.36%)

Here the 'a53' event group was always eligible to be scheduled, but
the 'a57' group never eligible to be scheduled, as the task was always
affine to a Cortex-A53 CPU. The SW (group leader) event in the 'a57'
group was eligible, but the HW event failed at pmu::add() time,
resulting in ctx_flexible_sched_in giving up on scheduling further
groups with HW events.

One way of avoiding this is to check pmu::filter_match() on siblings
as well as the group leader. If any of these fail their
pmu::filter_match() call, we must skip the entire group before
attempting to add any events.
Signed-off-by: NMark Rutland <mark.rutland@arm.com>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will.deacon@arm.com>
Fixes: 66eb579e ("perf: allow for PMU-specific event filtering")
Link: http://lkml.kernel.org/r/1465917041-15339-1-git-send-email-mark.rutland@arm.com
[ Small readability edits. ]
Signed-off-by: NIngo Molnar <mingo@kernel.org>

Jann Horn reported following analysis that could potentially result
in a very hard to trigger (if not impossible) UAF race, to quote his
event timeline:

 - Set up a process with threads T1, T2 and T3
 - Let T1 set up a socket filter F1 that invokes another filter F2
   through a BPF map [tail call]
 - Let T1 trigger the socket filter via a unix domain socket write,
   don't wait for completion
 - Let T2 call PERF_EVENT_IOC_SET_BPF with F2, don't wait for completion
 - Now T2 should be behind bpf_prog_get(), but before bpf_prog_put()
 - Let T3 close the file descriptor for F2, dropping the reference
   count of F2 to 2
 - At this point, T1 should have looked up F2 from the map, but not
   finished executing it
 - Let T3 remove F2 from the BPF map, dropping the reference count of
   F2 to 1
 - Now T2 should call bpf_prog_put() (wrong BPF program type), dropping
   the reference count of F2 to 0 and scheduling bpf_prog_free_deferred()
   via schedule_work()
 - At this point, the BPF program could be freed
 - BPF execution is still running in a freed BPF program

While at PERF_EVENT_IOC_SET_BPF time it's only guaranteed that the perf
event fd we're doing the syscall on doesn't disappear from underneath us
for whole syscall time, it may not be the case for the bpf fd used as
an argument only after we did the put. It needs to be a valid fd pointing
to a BPF program at the time of the call to make the bpf_prog_get() and
while T2 gets preempted, F2 must have dropped reference to 1 on the other
CPU. The fput() from the close() in T3 should also add additionally delay
to the reference drop via exit_task_work() when bpf_prog_release() gets
called as well as scheduling bpf_prog_free_deferred().

That said, it makes nevertheless sense to move the BPF prog destruction
generally after RCU grace period to guarantee that such scenario above,
but also others as recently fixed in ceb56070 ("bpf, perf: delay release
of BPF prog after grace period") with regards to tail calls won't happen.
Integrating bpf_prog_free_deferred() directly into the RCU callback is
not allowed since the invocation might happen from either softirq or
process context, so we're not permitted to block. Reviewing all bpf_prog_put()
invocations from eBPF side (note, cBPF -> eBPF progs don't use this for
their destruction) with call_rcu() look good to me.

Since we don't know whether at the time of attaching the program, we're
already part of a tail call map, we need to use RCU variant. However, due
to this, there won't be severely more stress on the RCU callback queue:
situations with above bpf_prog_get() and bpf_prog_put() combo in practice
normally won't lead to releases, but even if they would, enough effort/
cycles have to be put into loading a BPF program into the kernel already.
Reported-by: NJann Horn <jannh@google.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Acked-by: NAlexei Starovoitov <ast@kernel.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit dead9f29 ("perf: Fix race in BPF program unregister") moved
destruction of BPF program from free_event_rcu() callback to __free_event(),
which is problematic if used with tail calls: if prog A is attached as
trace event directly, but at the same time present in a tail call map used
by another trace event program elsewhere, then we need to delay destruction
via RCU grace period since it can still be in use by the program doing the
tail call (the prog first needs to be dropped from the tail call map, then
trace event with prog A attached destroyed, so we get immediate destruction).

Fixes: dead9f29 ("perf: Fix race in BPF program unregister")
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Acked-by: NAlexei Starovoitov <ast@kernel.org>
Cc: Jann Horn <jann@thejh.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

There is no way to end up in _free_event() with event::pmu being NULL.
The latter is initialized in event allocation path and remains set
forever. In case of allocation failure, the error path doesn't use
_free_event().

Having the check, however, suggests that it is possible to have a
event::pmu==NULL situation in _free_event() and confuses the robots.

This patch gets rid of the check.
Reported-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NAlexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: eranian@google.com
Cc: vince@deater.net
Link: http://lkml.kernel.org/r/1465303455-26032-1-git-send-email-alexander.shishkin@linux.intel.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

unaccount_pmu_sb_event() did not check for attributes in event->attr
before calling detach_sb_event(), while account_pmu_event() did.

This caused NULL pointer reference in cgroup events that did not
have any of the attributes checked by account_pmu_event().

To trigger the bug just wait for a cgroup event to terminate, e.g.:

  $ mkdir /dev/cgroup/devices/test
  $ perf stat -e cycles -a -G test sleep 0

... see crash ...
Signed-off-by: NDavid Carrillo-Cisneros <davidcc@google.com>
Reviewed-by: NStephane Eranian <eranian@google.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Zheng <zheng.z.yan@intel.com>
Link: http://lkml.kernel.org/r/1464809585-66072-1-git-send-email-davidcc@google.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

Change the return code for sampling event not supported from -ENOTSUPP
to -EOPNOTSUPP.

This allows userspace to identify this case specifically, instead of
printing the catch-all error message it did previously.

Technically this is an ABI change, but we think we can get away
with it.

Old behavior:
 -------
 | # perf record ls
 | Error:
 | The sys_perf_event_open() syscall returned with 524 (Unknown error 524)
 | for event (cycles:ppp).
 | /bin/dmesg may provide additional information.
 | No CONFIG_PERF_EVENTS=y kernel support configured?

New behavior:
 -------
 | # perf record ls
 | Error:
 | PMU Hardware doesn't support sampling/overflow-interrupts.
Signed-off-by: NVineet Gupta <vgupta@synopsys.com>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: <acme@redhat.com>
Cc: <linux-snps-arc@lists.infradead.org>
Cc: <vincent.weaver@maine.edu>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: Vineet Gupta <Vineet.Gupta1@synopsys.com>
Link: http://lkml.kernel.org/r/1462786660-2900-3-git-send-email-vgupta@synopsys.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

This patch fixes an issue which was introduced by commit:

  91a612ee ("perf/core: Fix dynamic interrupt throttle")

... which commit unconditionally sets the perf_sample_allowed_ns value
to !0. But that could trigger a bug in the following corner case:

The user can disable the dynamic interrupt throttle mechanism by setting
perf_cpu_time_max_percent to 0. Then they change perf_event_max_sample_rate.
For this case, the mechanism will be enabled implicitly, because
perf_sample_allowed_ns becomes !0 - which is not what we want.

This patch only updates perf_sample_allowed_ns when the dynamic
interrupt throttle mechanism is enabled.
Signed-off-by: NKan Liang <kan.liang@intel.com>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: acme@kernel.org
Link: http://lkml.kernel.org/r/1462260366-3160-1-git-send-email-kan.liang@intel.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

perf/core: Rename the perf_event_aux*() APIs to perf_event_sb*(), to separate them from AUX ring-buffer records

There are now two different things called AUX in perf, the
infrastructure to deliver the mmap/comm/task records and the
AUX part in the mmap buffer (with associated AUX_RECORD).

Since the former is internal, rename it to side-band to reduce
the confusion factor.

No change in functionality.
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: NIngo Molnar <mingo@kernel.org>

The perf_event_aux() function iterates all PMUs and all events in
their respective per-CPU contexts to find the events to deliver
side-band records to.

For example, the brk test case in lkp triggers many mmap() operations,
which, if we're also running perf, results in many perf_event_aux()
invocations.

If we enable uncore PMU support (even when uncore events are not used),
dozens of uncore PMUs will be iterated, which can significantly
decrease brk_test's throughput.

For example, the brk throughput:

  without uncore PMUs: 2647573 ops_per_sec
  with    uncore PMUs: 1768444 ops_per_sec

... a 33% reduction.

To get at the per-CPU events that need side-band records, this patch
puts these events on a per-CPU list, this avoids iterating the PMUs
and any events that do not need side-band records.

Per task events are unchanged to avoid extra overhead on the context
switch paths.
Suggested-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Reported-by: NHuang, Ying <ying.huang@linux.intel.com>
Signed-off-by: NKan Liang <kan.liang@intel.com>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/1458757477-3781-1-git-send-email-kan.liang@intel.comSigned-off-by: NIngo Molnar <mingo@kernel.org>