1. 03 4月, 2010 12 次提交
    • A
      perf tools: sort_dimension__add shouldn't die · 2aefa4f7
      Arnaldo Carvalho de Melo 提交于
      Propagate error instead.
      
      LKML-Reference: <new-submission>
      Cc: Frédéric Weisbecker <fweisbec@gmail.com>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Cc: Paul Mackerras <paulus@samba.org>
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      2aefa4f7
    • A
      perf session: Remove one more exit() call from library code · ad5b217b
      Arnaldo Carvalho de Melo 提交于
      Return NULL instead and make the caller propagate the error.
      
      LKML-Reference: <new-submission>
      Cc: Frédéric Weisbecker <fweisbec@gmail.com>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Cc: Paul Mackerras <paulus@samba.org>
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      ad5b217b
    • A
      perf hist: Only allocate callchain_node if processing callchains · b9fb9304
      Arnaldo Carvalho de Melo 提交于
      The struct callchain_node size is 120 bytes, that are never used when
      there are no callchains or '-g none' is specified, so conditionally
      allocate it, reducing sizeof(struct hist_entry) from 210 bytes to only
      96, greatly speeding the non-callchain processing.
      
      LKML-Reference: <new-submission>
      Cc: Frédéric Weisbecker <fweisbec@gmail.com>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Cc: Paul Mackerras <paulus@samba.org>
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      b9fb9304
    • A
      perf kmem: Fixup the symbol address before using it · 71cf8b8f
      Arnaldo Carvalho de Melo 提交于
      We get absolute addresses in the events, but relative ones from the
      symbol subsystem, so calculate the absolute address by asking for the
      map where the symbol was found, that has the place where the DSO was
      actually loaded.
      
      For the core kernel this poses no problems if the kernel is not
      relocated by things like kexec, or if we use /proc/kallsyms, but for
      modules we were getting really large, negative offsets.
      
      LKML-Reference: <new-submission>
      Cc: Frédéric Weisbecker <fweisbec@gmail.com>
      Cc: Li Zefan <lizf@cn.fujitsu.com>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Cc: Paul Mackerras <paulus@samba.org>
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      71cf8b8f
    • A
      perf kmem: Resolve kernel symbols again · e727ca73
      Arnaldo Carvalho de Melo 提交于
      Due to the assumption in perf_session__new that the kernel maps would be
      created using the fake PERF_RECORD_MMAP event in a perf.data file 'perf
      kmem --stat caller', that doesn't have such event, ends up not being
      able to resolve the kernel addresses.
      
      Fix it by calling perf_session__create_kernel_maps() in __cmd_kmem().
      
      LKML-Reference: <new-submission>
      Cc: Frédéric Weisbecker <fweisbec@gmail.com>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Cc: Paul Mackerras <paulus@samba.org>
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      e727ca73
    • A
      perf hist: Replace ->print() routines by ->snprintf() equivalents · a4e3b956
      Arnaldo Carvalho de Melo 提交于
      Then hist_entry__fprintf will just us the newly introduced
      hist_entry__snprintf, add the newline and fprintf it to the supplied
      FILE descriptor.
      
      This allows us to remove the use_browser checking in the color_printf
      routines, that now got color_snprintf variants too.
      
      The newt TUI browser (and other GUIs that may come in the future) don't
      have to worry about stdio specific stuff in the strings they get from
      the se->snprintf routines and instead use whatever means to do the
      equivalent.
      
      Also the newt TUI browser don't have to use the fmemopen() hack, instead
      it can use the se->snprintf routines directly. For now tho use the
      hist_entry__snprintf routine to reduce the patch size.
      
      Cc: Frédéric Weisbecker <fweisbec@gmail.com>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Cc: Paul Mackerras <paulus@samba.org>
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      a4e3b956
    • A
      perf record: Add a fallback to the reference relocation symbol · 70162138
      Arnaldo Carvalho de Melo 提交于
      Usually "_text" is enough, but I received reports that its not always
      available, so fallback to "_stext" for the symbol we use to check if we
      need to apply any relocation to all the symbols in the kernel symtab,
      for when, for instance, kexec is being used.
      Reported-by: NDarren Hart <dvhltc@us.ibm.com>
      Cc: Darren Hart <dvhltc@us.ibm.com>
      Cc: Frédéric Weisbecker <fweisbec@gmail.com>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Cc: Paul Mackerras <paulus@samba.org>
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      70162138
    • A
      perf tools: Allow specifying O= to build files in a separate directory · c29ede61
      Arnaldo Carvalho de Melo 提交于
      Avoiding polluting the source tree with build files.
      Reported-by: NSteven Rostedt <rostedt@goodmis.org>
      Cc: Steven Rostedt <rostedt@goodmis.org>
      Cc: Frédéric Weisbecker <fweisbec@gmail.com>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Cc: Paul Mackerras <paulus@samba.org>
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      c29ede61
    • A
      perf tools: Use -o $(BITBUCKET) in one more case · 8b2c551f
      Arnaldo Carvalho de Melo 提交于
      As described in 1703f2c3 some gcc versions has issues using /dev/null, so
      use the mechanism used elsewhere.
      
      Cc: Frédéric Weisbecker <fweisbec@gmail.com>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Cc: Paul Mackerras <paulus@samba.org>
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      8b2c551f
    • A
      perf report: Add progress bars · 5f4d3f88
      Arnaldo Carvalho de Melo 提交于
      For when we are processing the events and inserting the entries in the
      browser.
      
      Experimentation here: naming "ui_something" we may be treading into
      creating a TUI/GUI set of routines that can then be implemented in terms
      of multiple backends.
      
      Also the time it takes for adding things to the "browser" takes, visually
      (I guess I should do some profiling here ;-) ), more time than for
      processing the events...
      
      That means we probably need to create a custom hist_entry browser, so
      that we reuse the structures we have in place instead of duplicating
      them in newt.
      
      But progress was made and at least we can see something while long files
      are being loaded, that must be one of UI 101 bullet points :-)
      
      Cc: Frédéric Weisbecker <fweisbec@gmail.com>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Cc: Paul Mackerras <paulus@samba.org>
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      5f4d3f88
    • A
      perf symbols: map_groups__find_symbol must return the map too · 7e5e1b14
      Arnaldo Carvalho de Melo 提交于
      Tools need to know from which map in the map_group a symbol was resolved
      to, so that, for isntance, we can annotate kernel modules symbols by
      getting its precise name, etc.
      
      Also add the _by_name variants for completeness.
      
      Cc: Frédéric Weisbecker <fweisbec@gmail.com>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Cc: Paul Mackerras <paulus@samba.org>
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      7e5e1b14
    • A
      perf symbols: Move more map_groups methods to map.c · c6e718ff
      Arnaldo Carvalho de Melo 提交于
      While writing a standalone test app that uses the symbol system to
      find kernel space symbols I noticed these also need to be moved.
      
      Cc: Frédéric Weisbecker <fweisbec@gmail.com>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Cc: Paul Mackerras <paulus@samba.org>
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      c6e718ff
  2. 26 3月, 2010 16 次提交
  3. 24 3月, 2010 1 次提交
  4. 23 3月, 2010 8 次提交
    • A
      perf report: Implement Newt callgraphs · 4ded2b25
      Arnaldo Carvalho de Melo 提交于
      Starts collapsed, allows annotating by pressing 'A' or 'a' on
      the symbol, be it the top level one or any of the symbols in the
      chains.
      
      It (ab)uses the only tree widget in newt, that is actually a
      checkbox tree that we use with just one option ('.'), end result
      is usable but we really need to create a custom widget tree so
      that we can use the data structures we have (hist_entry rb_tree
      + callchain rb_tree + lists), so that we reduce the memory
      footprint by not creating a mirror set of data structures in the
      newtCheckboxTree widget.
      
      Thanks to Frédéric Weisbacker for fixing the orphanage problem
      in 301fde27, without that we were tripping a newt bug (fix
      already sent to newt's maintainer).
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      Cc: Avi Kivity <avi@redhat.com>
      Cc: Frédéric Weisbecker <fweisbec@gmail.com>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Cc: Paul Mackerras <paulus@samba.org>
      LKML-Reference: <1269291169-29820-1-git-send-email-acme@infradead.org>
      Signed-off-by: NIngo Molnar <mingo@elte.hu>
      4ded2b25
    • A
      perf tools: Exit browser before printing usage when unkown option passed · 478b0973
      Arnaldo Carvalho de Melo 提交于
      If not the screen will get garbled when using newt.
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      Cc: Frédéric Weisbecker <fweisbec@gmail.com>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Cc: Paul Mackerras <paulus@samba.org>
      LKML-Reference: <1269274229-20442-5-git-send-email-acme@infradead.org>
      Signed-off-by: NIngo Molnar <mingo@elte.hu>
      478b0973
    • A
      perf stat: Better report failure to collect system wide stats · 084ab9f8
      Arnaldo Carvalho de Melo 提交于
      Before:
      
      [acme@doppio linux-2.6-tip]$ perf stat -a sleep 1s
      
       Performance counter stats for 'sleep 1s':
      
        <not counted>  task-clock-msecs
        <not counted>  context-switches
        <not counted>  CPU-migrations
        <not counted>  page-faults
        <not counted>  cycles
        <not counted>  instructions
        <not counted>  branches
        <not counted>  branch-misses
        <not counted>  cache-references
        <not counted>  cache-misses
      
          1.016998463  seconds time elapsed
      
      [acme@doppio linux-2.6-tip]$
      
      Now:
      
      [acme@doppio linux-2.6-tip]$ perf stat -a sleep 1s
      No permission to collect system-wide stats.
      Consider tweaking /proc/sys/kernel/perf_event_paranoid.
      [acme@doppio linux-2.6-tip]$
      Reported-by: NIngo Molnar <mingo@elte.hu>
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      Cc: Frédéric Weisbecker <fweisbec@gmail.com>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Cc: Paul Mackerras <paulus@samba.org>
      LKML-Reference: <1269274229-20442-4-git-send-email-acme@infradead.org>
      Signed-off-by: NIngo Molnar <mingo@elte.hu>
      084ab9f8
    • M
      perf probe: Add NO_DWARF make option · ca721e45
      Masami Hiramatsu 提交于
      Add NO_DWARF make option for testing build without libdw.
      Signed-off-by: NMasami Hiramatsu <mhiramat@redhat.com>
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      Cc: Frederic Weisbecker <fweisbec@gmail.com>
      Cc: Paul Mackerras <paulus@samba.org>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      LKML-Reference: <1269274229-20442-3-git-send-email-acme@infradead.org>
      Signed-off-by: NIngo Molnar <mingo@elte.hu>
      ca721e45
    • M
      perf probe: Cleanup debuginfo related code · 4b4da7f7
      Masami Hiramatsu 提交于
      Cleanup debuginfo related code to eliminate fragile code which
      pointed by Ingo (Thanks!).
      1) Invert logic of NO_DWARF_SUPPORT to DWARF_SUPPORT.
      2) For removing assymetric/local variable ifdefs, introduce
        more helper functions.
      3) Change options order to reduce the number of ifdefs.
      Reported-by: NIngo Molnar <mingo@elte.hu>
      Signed-off-by: NMasami Hiramatsu <mhiramat@redhat.com>
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      Cc: Frederic Weisbecker <fweisbec@gmail.com>
      Cc: Paul Mackerras <paulus@samba.org>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      LKML-Reference: <1269274229-20442-2-git-send-email-acme@infradead.org>
      Signed-off-by: NIngo Molnar <mingo@elte.hu>
      4b4da7f7
    • A
      perf newt: Properly restore the screen when error exiting · f3a1f0ea
      Arnaldo Carvalho de Melo 提交于
      Show an OK message box with the last message sent via pr_err,
      etc.
      Reported-by: NIngo Molnar <mingo@elte.hu>
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      Cc: Frédéric Weisbecker <fweisbec@gmail.com>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Cc: Paul Mackerras <paulus@samba.org>
      LKML-Reference: <1269274229-20442-1-git-send-email-acme@infradead.org>
      Signed-off-by: NIngo Molnar <mingo@elte.hu>
      f3a1f0ea
    • F
      perf: Fix orphan callchain branches · 301fde27
      Frederic Weisbecker 提交于
      Callchains have markers inside their capture to tell we
      enter a context (kernel, user, ...).
      
      Those are not displayed in the callchains but they are
      incidentally an active part of the radix tree where
      callchains are stored, just like any other address.
      
      If we have the two following callchains:
      
      addr1 -> addr2 -> user context -> addr3
      addr1 -> addr2 -> user context -> addr4
      addr1 -> addr2 -> addr 5
      
      This is pretty common if addr1 and addr2 are part of an
      interrupt path, addr3 and addr4 are user addresses and
      addr5 is a kernel non interrupt path.
      
      This will be stored as follows in the tree:
      
                         addr1
                         addr2
                         /   \
                        /     addr5
                  user context
                     /    \
                   addr3  addr4
      
      But we ignore the context markers in the report, hence
      the addr3 and addr4 will appear as orphan branches:
      
          |--28.30%-- hrtimer_interrupt
          |          smp_apic_timer_interrupt
          |          apic_timer_interrupt
          |          |           <------------- here, no parent!
          |          |          |
          |          |          |--11.11%-- 0x7fae7bccb875
          |          |          |
          |          |          |--11.11%-- 0xffffffffff60013b
          |          |          |
          |          |          |--11.11%-- __pthread_mutex_lock_internal
          |          |          |
          |          |          |--11.11%-- __errno_location
      
      Fix this by removing the context markers when we process the
      callchains to the tree.
      Reported-by: NArnaldo Carvalho de Melo <acme@ghostprotocols.net>
      Signed-off-by: NFrederic Weisbecker <fweisbec@gmail.com>
      Cc: Paul Mackerras <paulus@samba.org>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      LKML-Reference: <1269274173-20328-1-git-send-email-acme@infradead.org>
      Signed-off-by: NIngo Molnar <mingo@elte.hu>
      301fde27
    • I
      Merge commit 'v2.6.34-rc2' into perf/core · d2f1e15b
      Ingo Molnar 提交于
      Merge reason: Pick up latest perf fixes from upstream.
      Signed-off-by: NIngo Molnar <mingo@elte.hu>
      d2f1e15b
  5. 20 3月, 2010 3 次提交