- 13 Apr, 2013 3 commits
-
-
Anton Arapov authored
Hijack the return address and replace it with a trampoline address. Signed-off-by:
Anton Arapov <anton@redhat.com> Acked-by:
Srikar Dronamraju <srikar@linux.vnet.ibm.com> Signed-off-by:
Oleg Nesterov <oleg@redhat.com>
-
Anton Arapov authored
Allocate trampoline page, as the very first one in uprobed task xol area, and fill it with breakpoint opcode. Also introduce get_trampoline_vaddr() helper, to wrap the trampoline address extraction from area->vaddr. That removes confusion and eases the debug experience in case ->vaddr notion will be changed. Signed-off-by:
-
Anton Arapov authored
Enclose return probes implementation, introduce ->ret_handler() and update existing code to rely on ->handler() *and* ->ret_handler() for uprobe and uretprobe respectively. Signed-off-by:
-
- 04 Apr, 2013 9 commits
-
-
Oleg Nesterov authored
Change write_opcode() to use copy_highpage() + copy_to_page() and simplify the code. Signed-off-by:
-
Oleg Nesterov authored
Extract the kmap_atomic/memcpy/kunmap_atomic code from xol_get_insn_slot() into the new simple helper, copy_to_page(). It will have more users soon. Signed-off-by:
-
Oleg Nesterov authored
__copy_insn(filp) can only be called after valid_vma() returns T, vma->vm_file passed as "filp" can not be NULL. Signed-off-by:
-
Oleg Nesterov authored
Change __copy_insn() to use copy_from_page() and simplify the code. Signed-off-by:
-
Oleg Nesterov authored
No functional changes. Rename copy_opcode() into copy_from_page() and add the new "int len" argument to make it more more generic for the new users. Signed-off-by:
-
Ananth N Mavinakayanahalli authored
prepare_uprobe() already checks if the underlying unstruction (on file) is a trap variant. We don't need to check this again. Signed-off-by:
Ananth N Mavinakayanahalli <ananth@in.ibm.com> Acked-by:
-
Ananth N Mavinakayanahalli authored
Powerpc has many trap variants that could be used by entities like gdb. Currently, running gdb on a program being traced by uprobes causes an endless loop since uprobes doesn't understand that the trap was inserted by some other entity and a SIGTRAP needs to be delivered. Teach uprobes to ignore breakpoints that do not belong to it. Signed-off-by:
-
Ananth N Mavinakayanahalli authored
Some architectures like powerpc have multiple variants of the trap instruction. Introduce an additional helper is_trap_insn() for run-time handling of non-uprobe traps on such architectures. While there, change is_swbp_at_addr() to is_trap_at_addr() for reading clarity. With this change, the uprobe registration path will supercede any trap instruction inserted at the requested location, while taking care of delivering the SIGTRAP for cases where the trap notification came in for an address without a uprobe. See [1] for a more detailed explanation. [1] https://lists.ozlabs.org/pipermail/linuxppc-dev/2013-March/104771.html This change was suggested by Oleg Nesterov. Signed-off-by:
-
Oleg Nesterov authored
Cleanup. Now that we have f_inode/file_inode() we can use it instead of vm_file->f_mapping->host. This should not make any difference for uprobes, but in theory this change is more correct. We use this inode as a key, to compare it with uprobe->inode set by uprobe_register(inode), and the caller uses d_inode. Signed-off-by:
-
- 01 Apr, 2013 21 commits
-
-
Ingo Molnar authored
Merge tag 'perf-core-for-mingo' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo: * Revert "perf sched: Handle PERF_RECORD_EXIT events" to get 'perf sched lat' back working. * We don't use Newt anymore, just plain libslang. * Kill a bunch of die() calls, from Namhyung Kim. * Add --no-demangle to report/top, from Namhyung Kim. * Fix dependency of the python binding wrt libtraceevent, from Naohiro Aota. * Introduce per core aggregation in 'perf stat', from Stephane Eranian. * Add memory profiling via PEBS, from Stephane Eranian. Signed-off-by:
Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by:
Ingo Molnar <mingo@kernel.org>
-
Arnaldo Carvalho de Melo authored
Initial motivation was to avoid the confusing exit when when '/' is pressed in non verbose mode, as specified in the help line searches are only available in verbose mode. Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/n/tip-20xezxim2y4agmkx7f3sucll@git.kernel.orgSigned-off-by:
-
Arnaldo Carvalho de Melo authored
Now that the map browser shares the input routine with the hists browser, there is no need for using any libnewt routine, so remove all traces except for honouring NO_NEWT=1 on the makefile command line as an indication that TUI support is not needed, in fact it just sets NO_SLANG=1. Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/n/tip-wae5o7xca9m52bj1re28jc5j@git.kernel.orgSigned-off-by:
-
Arnaldo Carvalho de Melo authored
Instead of an ad-hoc, libnewt based equivalent. Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/n/tip-elrijp95pijt66y6mmij4xm1@git.kernel.orgSigned-off-by:
-
Naohiro Aota authored
The python/perf.so binding lacks dependency for libtraceevent.a so that it cause the following error building python/perf.so. This patch introduce the dependency for it. $ make python/perf.so CHK -fstack-protector-all CHK -Wstack-protector CHK -Wvolatile-register-var CHK -D_FORTIFY_SOURCE=2 CHK bionic CHK libelf CHK libdw CHK libunwind CHK -DLIBELF_MMAP CHK libaudit CHK libnewt CHK gtk2 CHK -DHAVE_GTK_INFO_BAR CHK perl CHK python CHK python version CHK libbfd CHK -DHAVE_STRLCPY CHK -DHAVE_ON_EXIT CHK -DBACKTRACE_SUPPORT CHK libnuma GEN python/perf.so x86_64-pc-linux-gnu-gcc: error: ../lib/traceevent/libtraceevent.a: No such file or directory error: command 'x86_64-pc-linux-gnu-gcc' failed with exit status 1 cp: cannot stat 'python_ext_build/lib/perf.so': No such file or directory make: *** [python/perf.so] Error 1 Signed-off-by:Naohiro Aota <naota@elisp.net> Cc: Ingo Molnar <mingo@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/87wqswzznx.fsf@locke.i-did-not-set--mail-host-address--so-tickle-meSigned-off-by:
-
Arnaldo Carvalho de Melo authored
Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/n/tip-k85ajz97xbrd8fkt2a8pp7q1@git.kernel.orgSigned-off-by:
-
Arnaldo Carvalho de Melo authored
This reverts commit 0439539f. This caused this segfault: [root@sandy linux]# perf sched rec ^C[ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 1.306 MB perf.data (~57062 samples) ] perf [root@sandy linux]# perf sched lat perf: builtin-sched.c:781: thread_atoms_search: Assertion `!(thread != atoms->thread)' failed. Aborted (core dumped) [root@sandy linux]# Further investigation is needed to check that even with machine__remove_thread() not really deleting the thread referenced in the PERF_RECORD_EXIT (it goes to machine->dead_threads, because references may still exist to them in things like hist, etc) some event later comes for this dead thread and then machine__findnew_thread() will create a new thead instance that will not be the same as the one referenced by work_atoms->thread in thread_atoms_search(). For now just revert this patch to get the 'perf sched lat' back working. Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> echo Link: http://lkml.kernel.org/n/tip-`ranpwd -l 24`@git.kernel.org Link: http://lkml.kernel.org/n/tip-hg4s6e5txiwqe00h8rdg1sin@git.kernel.orgSigned-off-by:
-
Namhyung Kim authored
The symbol addresses in a dso have relative offsets from the start of a mapping. So in order to ouput correct offset value from @ip, one of them should be converted. Signed-off-by:
Namhyung Kim <namhyung@kernel.org> Cc: Andi Kleen <ak@linux.intel.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/r/1359040242-8269-19-git-send-email-eranian@google.comSigned-off-by:
-
Stephane Eranian authored
Leverages the PERF_RECORD_MISC_MMAP_DATA bit in the RECORD_MMAP record header. When the bit is set then the mapping type is set to MAP__VARIABLE. Signed-off-by:
Stephane Eranian <eranian@google.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung.kim@lge.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1359040242-8269-17-git-send-email-eranian@google.comSigned-off-by:
-
Stephane Eranian authored
This new command is a wrapper on top of perf record and perf report to make it easier to configure for memory access profiling. To record loads: $ perf mem -t load rec ..... To record stores: $ perf mem -t store rec ..... To get the report: $ perf mem -t load rep Signed-off-by:
-
Stephane Eranian authored
This patch adds the --mem-mode option to perf report. This mode requires a perf.data file created with memory access samples. Signed-off-by:
-
Stephane Eranian authored
We use the -W option to obtain the cost of the memory accesses. Data address sampling is obtained via the -d option. Signed-off-by:
-
Stephane Eranian authored
This patch adds the sorting and histogram support functions to enable profiling of memory accesses. The following sorting orders are added: - symbol_daddr: data address symbol (or raw address) - dso_daddr: data address shared object - locked: access uses locked transaction - tlb : TLB access - mem : memory level of the access (L1, L2, L3, RAM, ...) - snoop: access snoop mode Signed-off-by:
-
Andi Kleen authored
perf record has a new option -W that enables weightened sampling. Add sorting support in top/report for the average weight per sample and the total weight sum. This allows to both compare relative cost per event and the total cost over the measurement period. Add the necessary glue to perf report, record and the library. v2: Merge with new hist refactoring. v3: Fix manpage. Remove value check. Rename global_weight to weight and weight to local_weight. v4: Readd sort keys to manpage v5: Move weight to end v6: Move weight to template v7: Rename weight key. Original patch from Andi modified by Stephane Eranian <eranian@google.com> to include ONLY the weight supporting code and apply to pristine 3.8.0-rc4. Signed-off-by:
Andi Kleen <ak@linux.intel.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung.kim@lge.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1359040242-8269-6-git-send-email-eranian@google.com [ committer note: changed to cope with fc5871ed and the hists_link perf test entry ] Signed-off-by:
-
Stephane Eranian authored
Type of mapping was lost and made it hard for a tool to distinguish code vs. data mmaps. Perf has the ability to distinguish the two. Use a bit in the header->misc bitmask to keep track of the mmap type. If PERF_RECORD_MISC_MMAP_DATA is set then the mapping is not executable (!VM_EXEC). If not set, then the mapping is executable. Signed-off-by:
-
Stephane Eranian authored
This patch adds support for PEBS Precise Store which is available on Intel Sandy Bridge and Ivy Bridge processors. To use Precise store, the proper PEBS event must be used: mem_trans_retired:precise_stores. For the perf tool, the generic mem-stores event exported via sysfs can be used directly. Signed-off-by:
-
Stephane Eranian authored
Make the PEBS Load Latency threshold register layout and encoding visible to user level tools. Signed-off-by:
-
Stephane Eranian authored
This patch adds support for memory profiling using the PEBS Load Latency facility. Load accesses are sampled by HW and the instruction address, data address, load latency, data source, tlb, locked information can be saved in the sampling buffer if using the PERF_SAMPLE_COST (for latency), PERF_SAMPLE_ADDR, PERF_SAMPLE_DATA_SRC types. To enable PEBS Load Latency, users have to use the model specific event: - on NHM/WSM: MEM_INST_RETIRED:LATENCY_ABOVE_THRESHOLD - on SNB/IVB: MEM_TRANS_RETIRED:LATENCY_ABOVE_THRESHOLD To make things easier, this patch also exports a generic alias via sysfs: mem-loads. It export the right event encoding based on the host CPU and can be used directly by the perf tool. Loosely based on Intel's Lin Ming patch posted on LKML in July 2011. Signed-off-by:
-
Stephane Eranian authored
This patch adds PERF_SAMPLE_DATA_SRC. PERF_SAMPLE_DATA_SRC collects the data source, i.e., where did the data associated with the sampled instruction come from. Information is stored in a perf_mem_data_src structure. It contains opcode, mem level, tlb, snoop, lock information, subject to availability in hardware. Signed-off-by:
-
Andi Kleen authored
For some events it's useful to weight sample with a hardware provided number. This expresses how expensive the action the sample represent was. This allows the profiler to scale the samples to be more informative to the programmer. There is already the period which is used similarly, but it means something different, so I chose to not overload it. Instead a new sample type for WEIGHT is added. Can be used for multiple things. Initially it is used for TSX abort costs and profiling by memory latencies (so to make expensive load appear higher up in the histograms). The concept is quite generic and can be extended to many other kinds of events or architectures, as long as the hardware provides suitable auxillary values. In principle it could be also used for software tracepoints. This adds the generic glue. A new optional sample format for a 64-bit weight value. Signed-off-by:
-
Stephane Eranian authored
This patch adds a flags field to each event constraint. It can be used to store event specific features which can then later be used by scheduling code or low-level x86 code. The flags are propagated into event->hw.flags during the get_event_constraint() call. They are cleared during the put_event_constraint() call. This mechanism is going to be used by the PEBS-LL patches. It avoids defining yet another table to hold event specific information. Signed-off-by:
-
- 26 Mar, 2013 3 commits
-
-
Stephane Eranian authored
This patch extends Jiri's changes to make generic events mapping visible via sysfs. The patch extends the mechanism to non-generic events by allowing the mappings to be hardcoded in strings. This mechanism will be used by the PEBS-LL patch later on. Signed-off-by:
-
Andi Kleen authored
Add a way for the CPU initialization code to register additional events, and merge them into the events attribute directory. Used in the next patch. Signed-off-by:
-
Namhyung Kim authored
It's sometimes useful to see undemangled raw symbol name for example other tools using the perf output to do manipulation of binaries. Signed-off-by:
William Cohen <wcohen@redhat.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: William Cohen <wcohen@redhat.com> BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=55571 Link: http://lkml.kernel.org/r/1364203098-17741-1-git-send-email-namhyung@kernel.orgSigned-off-by:
-
- 25 Mar, 2013 3 commits
-
-
Stephane Eranian authored
This patch adds the --per-core option to perf stat. This option is used to aggregate system-wide counts on a per physical core basis. On processors with hyperthreading, this means counts of all HT threads running on a physical core are aggregated. This mode is useful to find imblance between physical cores running an uniform workload. Cores are identified by socket: S0-C1, means physical core 1 on socket 0. Note that cores are identified using their physical core id, thus their numbering may not be continuous. Per core aggregation can be combined with interval printing: # perf stat -a --per-core -I 1000 -e cycles sleep 1000 # time core cpus counts events 1.000090030 S0-C0 1 4,765,747 cycles 1.000090030 S0-C1 1 5,580,647 cycles 1.000090030 S0-C2 1 221,181 cycles 1.000090030 S0-C3 1 266,092 cycles Signed-off-by: -
Stephane Eranian authored
To make it more obvious what this option does as suggested by Andi on LKML. Signed-off-by:
-
Stephane Eranian authored
Refactor aggregation code by introducing a single aggr_mode variable and an enum for aggregation. Also refactor cpumap code having to do with cpu to socket mappings. All in preparation for extended modes, such as cpu -> core. Also fix socket aggregation and ensure that sockets are printed in increasing order. Signed-off-by:
-
- 21 Mar, 2013 1 commit
-
-
Namhyung Kim authored
It's for calculating whole trace data size during reading. However relation functions are called only in this file, no need to conditionalize it with tricky +1 offset and rename the variable to more meaningful name like trace_data_size. Signed-off-by:
-
