linux/tools/perf/util/intel-pt-decoder/intel-pt-insn-decoder.c

313 lines
7.6 KiB

perf script: Add 'brstackinsn' for branch stacks Implement printing instruction sequences as hex dump for branch stacks. This relies on the x86 instruction decoder used by the PT decoder to find the lengths of instructions to dump them individually. This is good enough for pattern matching. This allows to study hot paths for individual samples, together with branch misprediction and cycle count / IPC information if available (on Skylake systems). % perf record -b ... % perf script -F brstackinsn ... read_hpet+67: ffffffff9905b843 insn: 74 ea # PRED ffffffff9905b82f insn: 85 c9 ffffffff9905b831 insn: 74 12 ffffffff9905b833 insn: f3 90 ffffffff9905b835 insn: 48 8b 0f ffffffff9905b838 insn: 48 89 ca ffffffff9905b83b insn: 48 c1 ea 20 ffffffff9905b83f insn: 39 f2 ffffffff9905b841 insn: 89 d0 ffffffff9905b843 insn: 74 ea # PRED Only works when no special branch filters are specified. Occasionally the path does not reach up to the sample IP, as the LBRs may be frozen before executing a final jump. In this case we print a special message. The instruction dumper piggy backs on the existing infrastructure from the IP PT decoder. An earlier iteration of this patch relied on a disassembler, but this version only uses the existing instruction decoder. Committer note: Added hint about how to get suitable perf.data files for use with '-F brstackinsm': $ perf record usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.018 MB perf.data (8 samples) ] $ $ perf script -F brstackinsn Display of branch stack assembler requested, but non all-branch filter set Hint: run 'perf record -b ...' $ Signed-off-by: Andi Kleen <ak@linux.intel.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Michael Ellerman <mpe@ellerman.id.au> Link: http://lkml.kernel.org/r/20170223234634.583-1-andi@firstfloor.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-02-23 15:46:34 -08:00
perf tools: Fix build on powerpc broken by pt/bts It is theoretically possible to process perf.data files created on x86 and that contain Intel PT or Intel BTS data, on any other architecture, which is why it is possible for there to be build errors on powerpc caused by pt/bts. The errors were: util/intel-pt-decoder/intel-pt-insn-decoder.c: In function ‘intel_pt_insn_decoder’: util/intel-pt-decoder/intel-pt-insn-decoder.c:138:3: error: switch missing default case [-Werror=switch-default] switch (insn->immediate.nbytes) { ^ cc1: all warnings being treated as errors linux-acme.git/tools/perf/perf-obj/libperf.a(libperf-in.o): In function `intel_pt_synth_branch_sample': sources/linux-acme.git/tools/perf/util/intel-pt.c:871: undefined reference to `tsc_to_perf_time' linux-acme.git/tools/perf/perf-obj/libperf.a(libperf-in.o): In function `intel_pt_sample': sources/linux-acme.git/tools/perf/util/intel-pt.c:915: undefined reference to `tsc_to_perf_time' sources/linux-acme.git/tools/perf/util/intel-pt.c:962: undefined reference to `tsc_to_perf_time' linux-acme.git/tools/perf/perf-obj/libperf.a(libperf-in.o): In function `intel_pt_process_event': sources/linux-acme.git/tools/perf/util/intel-pt.c:1454: undefined reference to `perf_time_to_tsc' Signed-off-by: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> Cc: Wang Nan <wangnan0@huawei.com> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1441046384-28663-1-git-send-email-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-08-31 21:39:44 +03:00
perf script: Minimize "not reaching sample" for '-F +brstackinsn' In some situations 'perf script -F +brstackinsn' sees a lot of "not reaching sample" messages. This happens when the last LBR block before the sample contains a branch that is not in the LBR, and the instruction dumping stops. $ perf record -b emacs -Q --batch '()' [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.396 MB perf.data (443 samples) ] $ perf script -F +brstackinsn ... 00007f0ab2d171a4 insn: 41 0f 94 c0 00007f0ab2d171a8 insn: 83 fa 01 00007f0ab2d171ab insn: 74 d3 # PRED 6 cycles [313] 1.00 IPC 00007f0ab2d17180 insn: 45 84 c0 00007f0ab2d17183 insn: 74 28 ... not reaching sample ... $ perf script -F +brstackinsn | grep -c reach 136 $ This is a problem for further analysis that wants to see the full code upto the sample. There are two common cases where the message is bogus: - The LBR only logs taken branches, but the branch might be a conditional branch that is not taken (that is the most common case actually) - The LBR sampling uses a filter ignoring some branches, but the perf script check checks for all branches. This patch fixes these two conditions, by only checking for conditional branches, as well as checking the perf_event_attr's branch filter attributes. For the test case above it fixes all the messages: $ ./perf script -F +brstackinsn | grep -c reach 0 Note that there are still conditions when the message is hit -- sometimes there can be a unconditional branch that misses the LBR update before the sample -- but they are much more rare now. Signed-off-by: Andi Kleen <ak@linux.intel.com> Reviewed-by: Adrian Hunter <adrian.hunter@intel.com> Link: https://lore.kernel.org/r/20240229161828.386397-1-ak@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2024-02-29 08:18:28 -08:00
perf script: Fix LBR skid dump problems in brstackinsn This is a fix for another instance of the skid problem Milian recently found [1] The LBRs don't freeze at the exact same time as the PMI is triggered. The perf script brstackinsn code that dumps LBR assembler assumes that the last branch in the LBR leads to the sample point. But with skid it's possible that the CPU executes one or more branches before the sample, but which do not appear in the LBR. What happens then is either that the sample point is before the last LBR branch. In this case the dumper sees a negative length and ignores it. Or it the sample point is long after the last branch. Then the dumper sees a very long block and dumps it upto its block limit (16k bytes), which is noise in the output. On typical sample session this can happen regularly. This patch tries to detect and handle the situation. On the last block that is dumped by the LBR dumper we always stop on the first branch. If the block length is negative just scan forward to the first branch. Otherwise scan until a branch is found. The PT decoder already has a function that uses the instruction decoder to detect branches, so we can just reuse it here. Then when a terminating branch is found print an indication and stop dumping. This might miss a few instructions, but at least shows no runaway blocks. Signed-off-by: Andi Kleen <ak@linux.intel.com> Acked-by: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Milian Wolff <milian.wolff@kdab.com> Link: http://lkml.kernel.org/r/20181120050617.4119-1-andi@firstfloor.org [ Resolved conflict with dd2e18e9ac20 ("perf tools: Support 'srccode' output") ] Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-11-19 21:06:17 -08:00
perf script: Minimize "not reaching sample" for '-F +brstackinsn' In some situations 'perf script -F +brstackinsn' sees a lot of "not reaching sample" messages. This happens when the last LBR block before the sample contains a branch that is not in the LBR, and the instruction dumping stops. $ perf record -b emacs -Q --batch '()' [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.396 MB perf.data (443 samples) ] $ perf script -F +brstackinsn ... 00007f0ab2d171a4 insn: 41 0f 94 c0 00007f0ab2d171a8 insn: 83 fa 01 00007f0ab2d171ab insn: 74 d3 # PRED 6 cycles [313] 1.00 IPC 00007f0ab2d17180 insn: 45 84 c0 00007f0ab2d17183 insn: 74 28 ... not reaching sample ... $ perf script -F +brstackinsn | grep -c reach 136 $ This is a problem for further analysis that wants to see the full code upto the sample. There are two common cases where the message is bogus: - The LBR only logs taken branches, but the branch might be a conditional branch that is not taken (that is the most common case actually) - The LBR sampling uses a filter ignoring some branches, but the perf script check checks for all branches. This patch fixes these two conditions, by only checking for conditional branches, as well as checking the perf_event_attr's branch filter attributes. For the test case above it fixes all the messages: $ ./perf script -F +brstackinsn | grep -c reach 0 Note that there are still conditions when the message is hit -- sometimes there can be a unconditional branch that misses the LBR update before the sample -- but they are much more rare now. Signed-off-by: Andi Kleen <ak@linux.intel.com> Reviewed-by: Adrian Hunter <adrian.hunter@intel.com> Link: https://lore.kernel.org/r/20240229161828.386397-1-ak@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2024-02-29 08:18:28 -08:00
perf script: Fix LBR skid dump problems in brstackinsn This is a fix for another instance of the skid problem Milian recently found [1] The LBRs don't freeze at the exact same time as the PMI is triggered. The perf script brstackinsn code that dumps LBR assembler assumes that the last branch in the LBR leads to the sample point. But with skid it's possible that the CPU executes one or more branches before the sample, but which do not appear in the LBR. What happens then is either that the sample point is before the last LBR branch. In this case the dumper sees a negative length and ignores it. Or it the sample point is long after the last branch. Then the dumper sees a very long block and dumps it upto its block limit (16k bytes), which is noise in the output. On typical sample session this can happen regularly. This patch tries to detect and handle the situation. On the last block that is dumped by the LBR dumper we always stop on the first branch. If the block length is negative just scan forward to the first branch. Otherwise scan until a branch is found. The PT decoder already has a function that uses the instruction decoder to detect branches, so we can just reuse it here. Then when a terminating branch is found print an indication and stop dumping. This might miss a few instructions, but at least shows no runaway blocks. Signed-off-by: Andi Kleen <ak@linux.intel.com> Acked-by: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Milian Wolff <milian.wolff@kdab.com> Link: http://lkml.kernel.org/r/20181120050617.4119-1-andi@firstfloor.org [ Resolved conflict with dd2e18e9ac20 ("perf tools: Support 'srccode' output") ] Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-11-19 21:06:17 -08:00
perf script: Add 'brstackinsn' for branch stacks Implement printing instruction sequences as hex dump for branch stacks. This relies on the x86 instruction decoder used by the PT decoder to find the lengths of instructions to dump them individually. This is good enough for pattern matching. This allows to study hot paths for individual samples, together with branch misprediction and cycle count / IPC information if available (on Skylake systems). % perf record -b ... % perf script -F brstackinsn ... read_hpet+67: ffffffff9905b843 insn: 74 ea # PRED ffffffff9905b82f insn: 85 c9 ffffffff9905b831 insn: 74 12 ffffffff9905b833 insn: f3 90 ffffffff9905b835 insn: 48 8b 0f ffffffff9905b838 insn: 48 89 ca ffffffff9905b83b insn: 48 c1 ea 20 ffffffff9905b83f insn: 39 f2 ffffffff9905b841 insn: 89 d0 ffffffff9905b843 insn: 74 ea # PRED Only works when no special branch filters are specified. Occasionally the path does not reach up to the sample IP, as the LBRs may be frozen before executing a final jump. In this case we print a special message. The instruction dumper piggy backs on the existing infrastructure from the IP PT decoder. An earlier iteration of this patch relied on a disassembler, but this version only uses the existing instruction decoder. Committer note: Added hint about how to get suitable perf.data files for use with '-F brstackinsm': $ perf record usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.018 MB perf.data (8 samples) ] $ $ perf script -F brstackinsn Display of branch stack assembler requested, but non all-branch filter set Hint: run 'perf record -b ...' $ Signed-off-by: Andi Kleen <ak@linux.intel.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Michael Ellerman <mpe@ellerman.id.au> Link: http://lkml.kernel.org/r/20170223234634.583-1-andi@firstfloor.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-02-23 15:46:34 -08:00
perf script: Add 'brstackinsn' for branch stacks Implement printing instruction sequences as hex dump for branch stacks. This relies on the x86 instruction decoder used by the PT decoder to find the lengths of instructions to dump them individually. This is good enough for pattern matching. This allows to study hot paths for individual samples, together with branch misprediction and cycle count / IPC information if available (on Skylake systems). % perf record -b ... % perf script -F brstackinsn ... read_hpet+67: ffffffff9905b843 insn: 74 ea # PRED ffffffff9905b82f insn: 85 c9 ffffffff9905b831 insn: 74 12 ffffffff9905b833 insn: f3 90 ffffffff9905b835 insn: 48 8b 0f ffffffff9905b838 insn: 48 89 ca ffffffff9905b83b insn: 48 c1 ea 20 ffffffff9905b83f insn: 39 f2 ffffffff9905b841 insn: 89 d0 ffffffff9905b843 insn: 74 ea # PRED Only works when no special branch filters are specified. Occasionally the path does not reach up to the sample IP, as the LBRs may be frozen before executing a final jump. In this case we print a special message. The instruction dumper piggy backs on the existing infrastructure from the IP PT decoder. An earlier iteration of this patch relied on a disassembler, but this version only uses the existing instruction decoder. Committer note: Added hint about how to get suitable perf.data files for use with '-F brstackinsm': $ perf record usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.018 MB perf.data (8 samples) ] $ $ perf script -F brstackinsn Display of branch stack assembler requested, but non all-branch filter set Hint: run 'perf record -b ...' $ Signed-off-by: Andi Kleen <ak@linux.intel.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Michael Ellerman <mpe@ellerman.id.au> Link: http://lkml.kernel.org/r/20170223234634.583-1-andi@firstfloor.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-02-23 15:46:34 -08:00
perf script: Add 'brstackinsn' for branch stacks Implement printing instruction sequences as hex dump for branch stacks. This relies on the x86 instruction decoder used by the PT decoder to find the lengths of instructions to dump them individually. This is good enough for pattern matching. This allows to study hot paths for individual samples, together with branch misprediction and cycle count / IPC information if available (on Skylake systems). % perf record -b ... % perf script -F brstackinsn ... read_hpet+67: ffffffff9905b843 insn: 74 ea # PRED ffffffff9905b82f insn: 85 c9 ffffffff9905b831 insn: 74 12 ffffffff9905b833 insn: f3 90 ffffffff9905b835 insn: 48 8b 0f ffffffff9905b838 insn: 48 89 ca ffffffff9905b83b insn: 48 c1 ea 20 ffffffff9905b83f insn: 39 f2 ffffffff9905b841 insn: 89 d0 ffffffff9905b843 insn: 74 ea # PRED Only works when no special branch filters are specified. Occasionally the path does not reach up to the sample IP, as the LBRs may be frozen before executing a final jump. In this case we print a special message. The instruction dumper piggy backs on the existing infrastructure from the IP PT decoder. An earlier iteration of this patch relied on a disassembler, but this version only uses the existing instruction decoder. Committer note: Added hint about how to get suitable perf.data files for use with '-F brstackinsm': $ perf record usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.018 MB perf.data (8 samples) ] $ $ perf script -F brstackinsn Display of branch stack assembler requested, but non all-branch filter set Hint: run 'perf record -b ...' $ Signed-off-by: Andi Kleen <ak@linux.intel.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Michael Ellerman <mpe@ellerman.id.au> Link: http://lkml.kernel.org/r/20170223234634.583-1-andi@firstfloor.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-02-23 15:46:34 -08:00
perf script: Add 'brstackinsn' for branch stacks Implement printing instruction sequences as hex dump for branch stacks. This relies on the x86 instruction decoder used by the PT decoder to find the lengths of instructions to dump them individually. This is good enough for pattern matching. This allows to study hot paths for individual samples, together with branch misprediction and cycle count / IPC information if available (on Skylake systems). % perf record -b ... % perf script -F brstackinsn ... read_hpet+67: ffffffff9905b843 insn: 74 ea # PRED ffffffff9905b82f insn: 85 c9 ffffffff9905b831 insn: 74 12 ffffffff9905b833 insn: f3 90 ffffffff9905b835 insn: 48 8b 0f ffffffff9905b838 insn: 48 89 ca ffffffff9905b83b insn: 48 c1 ea 20 ffffffff9905b83f insn: 39 f2 ffffffff9905b841 insn: 89 d0 ffffffff9905b843 insn: 74 ea # PRED Only works when no special branch filters are specified. Occasionally the path does not reach up to the sample IP, as the LBRs may be frozen before executing a final jump. In this case we print a special message. The instruction dumper piggy backs on the existing infrastructure from the IP PT decoder. An earlier iteration of this patch relied on a disassembler, but this version only uses the existing instruction decoder. Committer note: Added hint about how to get suitable perf.data files for use with '-F brstackinsm': $ perf record usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.018 MB perf.data (8 samples) ] $ $ perf script -F brstackinsn Display of branch stack assembler requested, but non all-branch filter set Hint: run 'perf record -b ...' $ Signed-off-by: Andi Kleen <ak@linux.intel.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Michael Ellerman <mpe@ellerman.id.au> Link: http://lkml.kernel.org/r/20170223234634.583-1-andi@firstfloor.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-02-23 15:46:34 -08:00