Problem and motivation: Once a breakpoint perf event (PERF_TYPE_BREAKPOINT)
is created, there is no flexibility to change the breakpoint type
(bp_type), breakpoint address (bp_addr), or breakpoint length (bp_len). The
only option is to close the perf event and configure a new breakpoint
event. This inflexibility has a significant performance overhead. For
example, sampling-based, lightweight performance profilers (and also
concurrency bug detection tools), monitor different addresses for a short
duration using PERF_TYPE_BREAKPOINT and change the address (bp_addr) to
another address or change the kind of breakpoint (bp_type) from "write" to
a "read" or vice-versa or change the length (bp_len) of the address being
monitored. The cost of these modifications is prohibitive since it involves
unmapping the circular buffer associated with the perf event, closing the
perf event, opening another perf event and mmaping another circular buffer.
Solution: The new ioctl flag for perf events,
PERF_EVENT_IOC_MODIFY_ATTRIBUTES, introduced in this patch takes a pointer
to a struct perf_event_attr as an argument to update an old breakpoint
event with new address, type, and size. This facility allows retaining a
previous mmaped perf events ring buffer and avoids having to close and
reopen another perf event.
This patch supports only changing PERF_TYPE_BREAKPOINT event type; future
implementations can extend this feature. The patch replicates some of its
functionality of modify_user_hw_breakpoint() in
kernel/events/hw_breakpoint.c. modify_user_hw_breakpoint cannot be called
directly since perf_event_ctx_lock() is already held in _perf_ioctl().
Evidence: Experiments show that the baseline (not able to modify an already
created breakpoint) costs an order of magnitude (~10x) more than the
suggested optimization (having the ability to dynamically modifying a
configured breakpoint via ioctl). When the breakpoints typically do not
trap, the speedup due to the suggested optimization is ~10x; even when the
breakpoints always trap, the speedup is ~4x due to the suggested
optimization.
Testing: tests posted at
https://github.com/linux-contrib/perf_event_modify_bp demonstrate the
performance significance of this patch. Tests also check the functional
correctness of the patch.
Signed-off-by: Milind Chabbi <chabbi.milind@gmail.com>
[ Using modify_user_hw_breakpoint_check function. ]
[ Reformated PERF_EVENT_IOC_*, so the values are all in one column. ]
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Hari Bathini <hbathini@linux.vnet.ibm.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Oleg Nesterov <onestero@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will.deacon@arm.com>
Link: http://lkml.kernel.org/r/20180312134548.31532-8-jolsa@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Mark reported his arm64 perf fuzzer runs sometimes splat like:
armv8pmu_read_counter+0x1e8/0x2d8
armpmu_event_update+0x8c/0x188
armpmu_read+0xc/0x18
perf_output_read+0x550/0x11e8
perf_event_read_event+0x1d0/0x248
perf_event_exit_task+0x468/0xbb8
do_exit+0x690/0x1310
do_group_exit+0xd0/0x2b0
get_signal+0x2e8/0x17a8
do_signal+0x144/0x4f8
do_notify_resume+0x148/0x1e8
work_pending+0x8/0x14
which asserts that we only call pmu::read() on ACTIVE events.
The above callchain does:
perf_event_exit_task()
perf_event_exit_task_context()
task_ctx_sched_out() // INACTIVE
perf_event_exit_event()
perf_event_set_state(EXIT) // EXIT
sync_child_event()
perf_event_read_event()
perf_output_read()
perf_output_read_group()
leader->pmu->read()
Which results in doing a pmu::read() on an !ACTIVE event.
I _think_ this is 'new' since we added attr.inherit_stat, which added
the perf_event_read_event() to the exit path, without that
perf_event_read_output() would only trigger from samples and for
@event to trigger a sample, it's leader _must_ be ACTIVE too.
Still, adding this check makes it consistent with the @sub case for
the siblings.
Reported-and-Tested-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Peter 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: 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: Ingo Molnar <mingo@kernel.org>
Pull timer fixes from Thomas Gleixner:
"A small set of fixes from the timer departement:
- Add a missing timer wheel clock forward when migrating timers off a
unplugged CPU to prevent operating on a stale clock base and
missing timer deadlines.
- Use the proper shift count to extract data from a register value to
prevent evaluating unrelated bits
- Make the error return check in the FSL timer driver work correctly.
Checking an unsigned variable for less than zero does not really
work well.
- Clarify the confusing comments in the ARC timer code"
* 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
timers: Forward timer base before migrating timers
clocksource/drivers/arc_timer: Update some comments
clocksource/drivers/mips-gic-timer: Use correct shift count to extract data
clocksource/drivers/fsl_ftm_timer: Fix error return checking
Pull libnvdimm fixes from Dan Williams:
"A 4.16 regression fix, three fixes for -stable, and a cleanup fix:
- During the merge window support for the new ACPI NVDIMM Platform
Capabilities structure disabled support for "deep flush", a
force-unit- access like mechanism for persistent memory. Restore
that mechanism.
- VFIO like RDMA is yet one more memory registration / pinning
interface that is incompatible with Filesystem-DAX. Disable long
term pins of Filesystem-DAX mappings via VFIO.
- The Filesystem-DAX detection to prevent long terms pins mistakenly
also disabled Device-DAX pins which are not subject to the same
block- map collision concerns.
- Similar to the setup path, softlockup warnings can trigger in the
shutdown path for large persistent memory namespaces. Teach
for_each_device_pfn() to perform cond_resched() in all cases.
- Boaz noticed that the might_sleep() in dax_direct_access() is stale
as of the v4.15 kernel.
These have received a build success notification from the 0day robot,
and the longterm pin fixes have appeared in -next. However, I recently
rebased the tree to remove some other fixes that need to be reworked
after review feedback.
* 'libnvdimm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm:
memremap: fix softlockup reports at teardown
libnvdimm: re-enable deep flush for pmem devices via fsync()
vfio: disable filesystem-dax page pinning
dax: fix vma_is_fsdax() helper
dax: ->direct_access does not sleep anymore
Pull printk fix from Petr Mladek:
"Make sure that we wake up userspace loggers. This fixes a race
introduced by the console waiter logic during this merge window"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/pmladek/printk:
printk: Wake klogd when passing console_lock owner
On CPU hotunplug the enqueued timers of the unplugged CPU are migrated to a
live CPU. This happens from the control thread which initiated the unplug.
If the CPU on which the control thread runs came out from a longer idle
period then the base clock of that CPU might be stale because the control
thread runs prior to any event which forwards the clock.
In such a case the timers from the unplugged CPU are queued on the live CPU
based on the stale clock which can cause large delays due to increased
granularity of the outer timer wheels which are far away from base:;clock.
But there is a worse problem than that. The following sequence of events
illustrates it:
- CPU0 timer1 is queued expires = 59969 and base->clk = 59131.
The timer is queued at wheel level 2, with resulting expiry time = 60032
(due to level granularity).
- CPU1 enters idle @60007, with next timer expiry @60020.
- CPU0 is hotplugged at @60009
- CPU1 exits idle and runs the control thread which migrates the
timers from CPU0
timer1 is now queued in level 0 for immediate handling in the next
softirq because the requested expiry time 59969 is before CPU1 base->clk
60007
- CPU1 runs code which forwards the base clock which succeeds because the
next expiring timer. which was collected at idle entry time is still set
to 60020.
So it forwards beyond 60007 and therefore misses to expire the migrated
timer1. That timer gets expired when the wheel wraps around again, which
takes between 63 and 630ms depending on the HZ setting.
Address both problems by invoking forward_timer_base() for the control CPUs
timer base. All other places, which might run into a similar problem
(mod_timer()/add_timer_on()) already invoke forward_timer_base() to avoid
that.
[ tglx: Massaged comment and changelog ]
Fixes: a683f390b9 ("timers: Forward the wheel clock whenever possible")
Co-developed-by: Neeraj Upadhyay <neeraju@codeaurora.org>
Signed-off-by: Neeraj Upadhyay <neeraju@codeaurora.org>
Signed-off-by: Lingutla Chandrasekhar <clingutla@codeaurora.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Anna-Maria Gleixner <anna-maria@linutronix.de>
Cc: linux-arm-msm@vger.kernel.org
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20180118115022.6368-1-clingutla@codeaurora.org
