linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [GIT PULL] uprobes/perf: pre-filtering
@ 2013-02-08 18:03 Oleg Nesterov
  2013-02-11  9:43 ` Ingo Molnar
  0 siblings, 1 reply; 2+ messages in thread
From: Oleg Nesterov @ 2013-02-08 18:03 UTC (permalink / raw)
  To: Ingo Molnar
  Cc: Anton Arapov, Arnaldo Carvalho de Melo, Frank Eigler, Jiri Olsa,
	Josh Stone, Masami Hiramatsu, Srikar Dronamraju, Steven Rostedt,
	Suzuki K. Poulose, linux-kernel

OK, nobody has comments. Ingo, please pull from

  git://git.kernel.org/pub/scm/linux/kernel/git/oleg/misc uprobes/core

Changes:
	- fix the conflict with d24d7dbf which added the (unsafe) check
	  into create_trace_uprobe()

	- drop "Kill the "uprobe != NULL" check in uprobe_unregister()",
	  the same patch from Sasha Levin was already applied as c91368c4

	- drop "perf: Ensure we do not free event->parent before event"
	  http://marc.info/?l=linux-kernel&m=136000475509419

	  Currently it is not necessary and it wasn't reviewed. But perhaps
	  I'll resend it later.

Simple performance test:

	# time perl -e 'syscall -1 for 1..100_000'
	real    0m0.040s
	user    0m0.027s
	sys     0m0.010s
	
	# perf probe -x /lib/libc.so.6 syscall
	# perf record -e probe_libc:syscall sleep 100 &

Before this series:

	# time perl -e 'syscall -1 for 1..100_000'
	real    0m1.714s
	user    0m0.103s
	sys     0m1.607s

After:
	real    0m0.037s
	user    0m0.013s
	sys	0m0.023s

The untraced process is not penalized.

Josh Stone (1):
      uprobes: Add exports for module use

Oleg Nesterov (37):
      uprobes: Move __set_bit(UPROBE_SKIP_SSTEP) into alloc_uprobe()
      uprobes: Kill the pointless inode/uc checks in register/unregister
      uprobes: Kill uprobe_consumer->filter()
      uprobes: Introduce filter_chain()
      uprobes: _unregister() should always do register_for_each_vma(false)
      uprobes: _register() should always do register_for_each_vma(true)
      uprobes: Introduce uprobe->register_rwsem
      uprobes: Change filter_chain() to iterate ->consumers list
      uprobes: Kill UPROBE_RUN_HANDLER flag
      uprobes: Kill uprobe->copy_mutex
      uprobes: Kill uprobe_events, use RB_EMPTY_ROOT() instead
      uprobes: Introduce uprobe_is_active()
      uprobes: Kill uprobes_mutex[], separate alloc_uprobe() and __uprobe_register()
      uprobes: Rationalize the usage of filter_chain()
      uprobes: Reintroduce uprobe_consumer->filter()
      uprobes: Teach handler_chain() to filter out the probed task
      uprobes/x86: Change __skip_sstep() to actually skip the whole insn
      uprobes: Change handle_swbp() to expose bp_vaddr to handler_chain()
      uprobes: Move alloc_page() from xol_add_vma() to xol_alloc_area()
      uprobes: Fold xol_alloc_area() into get_xol_area()
      uprobes: Turn add_utask() into get_utask()
      uprobes: Do not play with utask in xol_get_insn_slot()
      uprobes: Fix utask->xol_vaddr leak in pre_ssout()
      uprobes: Do not allocate current->utask unnecessary
      uprobes: Kill the bogus IS_ERR_VALUE(xol_vaddr) check
      uprobes/tracing: Fix dentry/mount leak in create_trace_uprobe()
      uprobes/tracing: Fully initialize uprobe_trace_consumer before uprobe_register()
      uprobes/tracing: Ensure inode != NULL in create_trace_uprobe()
      uprobes/tracing: Introduce is_trace_uprobe_enabled()
      uprobes/tracing: Kill uprobe_trace_consumer, embed uprobe_consumer into trace_uprobe
      uprobes/perf: Always increment trace_uprobe->nhit
      perf: Introduce hw_perf_event->tp_target and ->tp_list
      uprobes: Introduce uprobe_apply()
      uprobes/perf: Teach trace_uprobe/perf code to track the active perf_event's
      uprobes/perf: Teach trace_uprobe/perf code to pre-filter
      uprobes/perf: Teach trace_uprobe/perf code to use UPROBE_HANDLER_REMOVE
      uprobes/perf: Avoid uprobe_apply() whenever possible

 arch/x86/kernel/uprobes.c   |    4 +-
 include/linux/perf_event.h  |    9 +-
 include/linux/uprobes.h     |   23 ++-
 kernel/events/core.c        |    5 +-
 kernel/events/uprobes.c     |  465 +++++++++++++++++++++++--------------------
 kernel/ptrace.c             |    6 +
 kernel/trace/trace_probe.h  |    1 -
 kernel/trace/trace_uprobe.c |  217 +++++++++++++++-----
 8 files changed, 456 insertions(+), 274 deletions(-)


^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [GIT PULL] uprobes/perf: pre-filtering
  2013-02-08 18:03 [GIT PULL] uprobes/perf: pre-filtering Oleg Nesterov
@ 2013-02-11  9:43 ` Ingo Molnar
  0 siblings, 0 replies; 2+ messages in thread
From: Ingo Molnar @ 2013-02-11  9:43 UTC (permalink / raw)
  To: Oleg Nesterov
  Cc: Ingo Molnar, Anton Arapov, Arnaldo Carvalho de Melo,
	Frank Eigler, Jiri Olsa, Josh Stone, Masami Hiramatsu,
	Srikar Dronamraju, Steven Rostedt, Suzuki K. Poulose,
	linux-kernel


* Oleg Nesterov <oleg@redhat.com> wrote:

> OK, nobody has comments. Ingo, please pull from
> 
>   git://git.kernel.org/pub/scm/linux/kernel/git/oleg/misc uprobes/core
> 
> Changes:
> 	- fix the conflict with d24d7dbf which added the (unsafe) check
> 	  into create_trace_uprobe()
> 
> 	- drop "Kill the "uprobe != NULL" check in uprobe_unregister()",
> 	  the same patch from Sasha Levin was already applied as c91368c4
> 
> 	- drop "perf: Ensure we do not free event->parent before event"
> 	  http://marc.info/?l=linux-kernel&m=136000475509419
> 
> 	  Currently it is not necessary and it wasn't reviewed. But perhaps
> 	  I'll resend it later.
> 
> Simple performance test:
> 
> 	# time perl -e 'syscall -1 for 1..100_000'
> 	real    0m0.040s
> 	user    0m0.027s
> 	sys     0m0.010s
> 	
> 	# perf probe -x /lib/libc.so.6 syscall
> 	# perf record -e probe_libc:syscall sleep 100 &
> 
> Before this series:
> 
> 	# time perl -e 'syscall -1 for 1..100_000'
> 	real    0m1.714s
> 	user    0m0.103s
> 	sys     0m1.607s
> 
> After:
> 	real    0m0.037s
> 	user    0m0.013s
> 	sys	0m0.023s
> 
> The untraced process is not penalized.
> 
> Josh Stone (1):
>       uprobes: Add exports for module use
> 
> Oleg Nesterov (37):
>       uprobes: Move __set_bit(UPROBE_SKIP_SSTEP) into alloc_uprobe()
>       uprobes: Kill the pointless inode/uc checks in register/unregister
>       uprobes: Kill uprobe_consumer->filter()
>       uprobes: Introduce filter_chain()
>       uprobes: _unregister() should always do register_for_each_vma(false)
>       uprobes: _register() should always do register_for_each_vma(true)
>       uprobes: Introduce uprobe->register_rwsem
>       uprobes: Change filter_chain() to iterate ->consumers list
>       uprobes: Kill UPROBE_RUN_HANDLER flag
>       uprobes: Kill uprobe->copy_mutex
>       uprobes: Kill uprobe_events, use RB_EMPTY_ROOT() instead
>       uprobes: Introduce uprobe_is_active()
>       uprobes: Kill uprobes_mutex[], separate alloc_uprobe() and __uprobe_register()
>       uprobes: Rationalize the usage of filter_chain()
>       uprobes: Reintroduce uprobe_consumer->filter()
>       uprobes: Teach handler_chain() to filter out the probed task
>       uprobes/x86: Change __skip_sstep() to actually skip the whole insn
>       uprobes: Change handle_swbp() to expose bp_vaddr to handler_chain()
>       uprobes: Move alloc_page() from xol_add_vma() to xol_alloc_area()
>       uprobes: Fold xol_alloc_area() into get_xol_area()
>       uprobes: Turn add_utask() into get_utask()
>       uprobes: Do not play with utask in xol_get_insn_slot()
>       uprobes: Fix utask->xol_vaddr leak in pre_ssout()
>       uprobes: Do not allocate current->utask unnecessary
>       uprobes: Kill the bogus IS_ERR_VALUE(xol_vaddr) check
>       uprobes/tracing: Fix dentry/mount leak in create_trace_uprobe()
>       uprobes/tracing: Fully initialize uprobe_trace_consumer before uprobe_register()
>       uprobes/tracing: Ensure inode != NULL in create_trace_uprobe()
>       uprobes/tracing: Introduce is_trace_uprobe_enabled()
>       uprobes/tracing: Kill uprobe_trace_consumer, embed uprobe_consumer into trace_uprobe
>       uprobes/perf: Always increment trace_uprobe->nhit
>       perf: Introduce hw_perf_event->tp_target and ->tp_list
>       uprobes: Introduce uprobe_apply()
>       uprobes/perf: Teach trace_uprobe/perf code to track the active perf_event's
>       uprobes/perf: Teach trace_uprobe/perf code to pre-filter
>       uprobes/perf: Teach trace_uprobe/perf code to use UPROBE_HANDLER_REMOVE
>       uprobes/perf: Avoid uprobe_apply() whenever possible
> 
>  arch/x86/kernel/uprobes.c   |    4 +-
>  include/linux/perf_event.h  |    9 +-
>  include/linux/uprobes.h     |   23 ++-
>  kernel/events/core.c        |    5 +-
>  kernel/events/uprobes.c     |  465 +++++++++++++++++++++++--------------------
>  kernel/ptrace.c             |    6 +
>  kernel/trace/trace_probe.h  |    1 -
>  kernel/trace/trace_uprobe.c |  217 +++++++++++++++-----
>  8 files changed, 456 insertions(+), 274 deletions(-)

Pulled, thanks a lot Oleg!

	Ingo

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2013-02-11  9:44 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-02-08 18:03 [GIT PULL] uprobes/perf: pre-filtering Oleg Nesterov
2013-02-11  9:43 ` Ingo Molnar

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).