- 14 Feb, 2019 21 commits
-
-
Arnaldo Carvalho de Melo authored
When the libcrypto feature test was added we forgot to add its FEATURE_CHECK_LDFLAGS pointing to the library needed to link with the test-all.bin feature test fast path binary, so even when it was introduced we got this: $ cat /tmp/build/perf/feature/test-all.make.output /usr/bin/ld: /tmp/ccjKeJJU.o: in function `main_test_libcrypto': /home/acme/git/perf/tools/build/feature/test-libcrypto.c:10: undefined reference to `MD5_Init' /usr/bin/ld: /home/acme/git/perf/tools/build/feature/test-libcrypto.c:11: undefined reference to `MD5_Update' /usr/bin/ld: /home/acme/git/perf/tools/build/feature/test-libcrypto.c:12: undefined reference to `MD5_Final' /usr/bin/ld: /home/acme/git/perf/tools/build/feature/test-libcrypto.c:14: undefined reference to `SHA1' collect2: error: ld returned 1 exit status $ cat /tmp/build/perf/feature/test-libcrypto. test-libcrypto.bin test-libcrypto.d test-libcrypto.make.output $ cat /tmp/build/perf/feature/test-libcrypto.make.output $ Fix it, so that we keep the fast path, which, at this point, will fail with the unwind-ARCH feature tests, that will be fixed in a followup patch: $ make -C tools/perf O=/tmp/build/perf ... libcrypto: [ on ] <SNIP> $ cat /tmp/build/perf/feature/test-all.make.output $ ldd /tmp/build/perf/feature/test-all.bin | grep libcrypto libcrypto.so.1.1 => /lib64/libcrypto.so.1.1 (0x00007f9892805000) $ $ grep libcrypto /tmp/build/perf/FEATURE-DUMP feature-libcrypto=1 $ With the unwind-ARCH tests fixed, we now finally manage to get test-all.bin built and linked with the features it tests, among them the ones fixed in this patchkit: $ ldd /tmp/build/perf/feature/test-all.bin | egrep 'unwind|crypto' libcrypto.so.1.1 => /lib64/libcrypto.so.1.1 (0x00007f95cf2b8000) libunwind-x86_64.so.8 => /lib64/libunwind-x86_64.so.8 (0x00007f95cf294000) libunwind.so.8 => /lib64/libunwind.so.8 (0x00007f95cf278000) $ Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Carl Love <cel@us.ibm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: John McCutchan <johnmccutchan@google.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Sonny Rao <sonnyrao@chromium.org> Cc: Stephane Eranian <eranian@google.com> Cc: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> Fixes: 8ee46460 ("perf build: Add libcrypto feature detection") Link: https://lkml.kernel.org/n/tip-rexc248jorf5b4l3qjn888cz@git.kernel.orgSigned-off-by:
Arnaldo Carvalho de Melo <acme@redhat.com>
-
Arnaldo Carvalho de Melo authored
When a test is in the FEATURE_TESTS_BASIC list in tools/build/Makefile.feature must be added to tools/build/feature/test-all.c, because the successfull compilation and linking of that test-all.bin file means that all the features listed in FEATURE_TESTS_BASIC are present in the system, so we don't have to go on feature by feature test building them. Since reallocarray() is expected to be present in modern systems, it has a place in FEATURE_TESTS_BASIC, so that we speed up the build process building just that file. For older systems, such as ubuntu:16.04 (build failure reported by Jin Yao) debian:8, and for the current flagship RHEL distro, RHEL7, the build will fail as test-all.bin (without test-reallocarray.c included) passes but reallocarray() isn't present, making the build fail with: CC /tmp/build/perf/libbpf.o MKDIR /tmp/build/perf/fs/ CC /tmp/build/perf/fs/tracing_path.o LD /tmp/build/perf/fd/libapi-in.o CC /tmp/build/perf/bpf.o libbpf.c: In function 'bpf_object__add_program': libbpf.c:367:10: error: implicit declaration of function 'reallocarray' [-Werror=implicit-function-declaration] progs = reallocarray(progs, nr_progs + 1, sizeof(progs[0])); ^ libbpf.c:367:2: error: nested extern declaration of 'reallocarray' [-Werror=nested-externs] progs = reallocarray(progs, nr_progs + 1, sizeof(progs[0])); ^ libbpf.c:367:8: error: assignment makes pointer from integer without a cast [-Werror=int-conversion] progs = reallocarray(progs, nr_progs + 1, sizeof(progs[0])); ^ libbpf.c: In function 'bpf_object__elf_collect': libbpf.c:887:10: error: assignment makes pointer from integer without a cast [-Werror=int-conversion] reloc = reallocarray(reloc, nr_reloc, ^ libbpf.c: In function 'bpf_program__reloc_text': libbpf.c:1394:12: error: assignment makes pointer from integer without a cast [-Werror=int-conversion] new_insn = reallocarray(prog->insns, new_cnt, sizeof(*insn)); ^ CC /tmp/build/perf/nlattr.o Even with: $ grep reallocarray /tmp/build/perf/FEATURE-DUMP feature-reallocarray=1 $ Which ubuntu:16.04.5 LTS doesn't have: perfbuilder@38a153a1bba8:/$ head -2 /etc/os-release NAME="Ubuntu" VERSION="16.04.5 LTS (Xenial Xerus)" perfbuilder@38a153a1bba8:/$ find /usr/include/ -name "*.h" | xargs grep -w reallocarray perfbuilder@38a153a1bba8:/$ Fix it by including it to test-all.c, which ends up forcing the individual tests to be triggered and for the build process to notice that indeed reallocarray() is not there: perfbuilder@38a153a1bba8:/$ cat /tmp/build/perf/feature/test-all.make.output In file included from test-all.c:178:0: test-reallocarray.c: In function 'main_test_reallocarray': test-reallocarray.c:7:11: error: implicit declaration of function 'reallocarray' [-Werror=implicit-function-declaration] return !!reallocarray(NULL, 1, 1); ^ cc1: all warnings being treated as errors perfbuilder@38a153a1bba8:/$ That is the only test that is failing on Ubuntu 16.03.5 LTS, so all tests are forced: perfbuilder@38a153a1bba8:/tmp/build/perf/feature$ ls -lSr *.make.output <SNIP successful tests> -rw-r--r--. 1 perfbuilder perfbuilder 0 Feb 14 15:00 test-dwarf.make.output -rw-r--r--. 1 perfbuilder perfbuilder 0 Feb 14 14:16 test-cplus-demangle.make.output -rw-r--r--. 1 perfbuilder perfbuilder 0 Feb 14 15:00 test-bpf.make.output -rw-r--r--. 1 perfbuilder perfbuilder 0 Feb 14 15:00 test-backtrace.make.output -rw-r--r--. 1 perfbuilder perfbuilder 104 Feb 14 15:00 test-bionic.make.output -rw-r--r--. 1 perfbuilder perfbuilder 107 Feb 14 15:00 test-libunwind-x86.make.output -rw-r--r--. 1 perfbuilder perfbuilder 115 Feb 14 15:00 test-libunwind-aarch64.make.output -rw-r--r--. 1 perfbuilder perfbuilder 122 Feb 14 15:00 test-libbabeltrace.make.output -rw-r--r--. 1 perfbuilder perfbuilder 254 Feb 14 15:00 test-reallocarray.make.output -rw-r--r--. 1 perfbuilder perfbuilder 312 Feb 14 15:00 test-all.make.output perfbuilder@38a153a1bba8:/tmp/build/perf/feature$ And that reallocarray() one shows: perfbuilder@38a153a1bba8:/tmp/build/perf/feature$ cat test-reallocarray.make.output test-reallocarray.c: In function 'main': test-reallocarray.c:7:11: error: implicit declaration of function 'reallocarray' [-Werror=implicit-function-declaration] return !!reallocarray(NULL, 1, 1); ^ cc1: all warnings being treated as errors perfbuilder@38a153a1bba8:/tmp/build/perf/feature$ Which now generates the expected result: perfbuilder@38a153a1bba8:~$ grep reallocarray /tmp/build/perf/FEATURE-DUMP feature-reallocarray=0 perfbuilder@38a153a1bba8:~$ The fallback mechanism kicks in and libbpf and perf are again buildable in systems without reallocarray(): $ cat tools/include/tools/libc_compat.h // SPDX-License-Identifier: (LGPL-2.0+ OR BSD-2-Clause) /* Copyright (C) 2018 Netronome Systems, Inc. */ #ifndef __TOOLS_LIBC_COMPAT_H #define __TOOLS_LIBC_COMPAT_H #include <stdlib.h> #include <linux/overflow.h> #ifdef COMPAT_NEED_REALLOCARRAY static inline void *reallocarray(void *ptr, size_t nmemb, size_t size) { size_t bytes; if (unlikely(check_mul_overflow(nmemb, size, &bytes))) return NULL; return realloc(ptr, bytes); } #endif #endif $ Reported-by:Jin Yao <yao.jin@linux.intel.com> Acked-by:
Jiri Olsa <jolsa@kernel.org> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexei Starovoitov <ast@fb.com> Cc: Andrii Nakryiko <andrii.nakryiko@gmail.com> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Jakub Kicinski <jakub.kicinski@netronome.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Song Liu <songliubraving@fb.com> Cc: Yonghong Song <yhs@fb.com> Fixes: 531b014e ("tools: bpf: make use of reallocarray") Link: https://lkml.kernel.org/n/tip-aonqku8axii8rxki5g11w40b@git.kernel.orgSigned-off-by:
-
Arnaldo Carvalho de Melo authored
As it is not normally available on x86_64 not being tested on test-all.c but being in FEATURE_TESTS_BASIC ends up implying that those features are present, which leads to trying to link with those libraries and a build failure now that test-all.c is finally again building successfully: /usr/bin/ld: cannot find -lunwind-x86 /usr/bin/ld: cannot find -lunwind-aarch64 collect2: error: ld returned 1 exit status make[3]: *** [Makefile:199: /tmp/build/perf/plugin_jbd2.so] Error 1 make[3]: *** Waiting for unfinished jobs.... /usr/bin/ld: cannot find -lunwind-x86 /usr/bin/ld: cannot find -lunwind-aarch64 So remove those features from there and explicitely test them. And then move this patch to just before the last one that allows this to be exposed, so that we keep the tree bisectable. With all this in place we get, at this point: $ ldd /tmp/build/perf/feature/test-libunwind.bin linux-vdso.so.1 (0x00007fffa09c6000) libunwind-x86_64.so.8 => /lib64/libunwind-x86_64.so.8 (0x00007fbcf4451000) libunwind.so.8 => /lib64/libunwind.so.8 (0x00007fbcf4435000) liblzma.so.5 => /lib64/liblzma.so.5 (0x00007fbcf440c000) libelf.so.1 => /lib64/libelf.so.1 (0x00007fbcf43f2000) libc.so.6 => /lib64/libc.so.6 (0x00007fbcf422c000) libgcc_s.so.1 => /lib64/libgcc_s.so.1 (0x00007fbcf4211000) /lib64/ld-linux-x86-64.so.2 (0x00007fbcf4491000) libpthread.so.0 => /lib64/libpthread.so.0 (0x00007fbcf41ed000) libz.so.1 => /lib64/libz.so.1 (0x00007fbcf41d3000) $ cat /tmp/build/perf/feature/test-libunwind-x86.make.output test-libunwind-x86.c:2:10: fatal error: libunwind-x86.h: No such file or directory #include <libunwind-x86.h> ^~~~~~~~~~~~~~~~~ compilation terminated. $ cat /tmp/build/perf/feature/test-libunwind-aarch64.make.output test-libunwind-aarch64.c:2:10: fatal error: libunwind-aarch64.h: No such file or directory #include <libunwind-aarch64.h> ^~~~~~~~~~~~~~~~~~~~~ compilation terminated. $ $ ldd ~/bin/perf | grep unwind libunwind-x86_64.so.8 => /lib64/libunwind-x86_64.so.8 (0x00007f5ceb24b000) libunwind.so.8 => /lib64/libunwind.so.8 (0x00007f5ceb22f000) $ Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: He Kuang <hekuang@huawei.com> Cc: Jean Pihet <jean.pihet@linaro.org> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Cc: Will Deacon <will.deacon@arm.com> Link: https://lkml.kernel.org/n/tip-vs6kwqsvwk7oxhs6z9mq87pp@git.kernel.orgSigned-off-by: -
Arnaldo Carvalho de Melo authored
Since it is not yet that generally available, avoid testing for the presence of libcoresight in the fast path test-all.bin feature test. # dnf search opencsd No matches found. # dnf search OpenCSD No matches found. # cat /etc/fedora-release Fedora release 29 (Twenty Nine) # I.e. right now, in my system test-all.bin is failing all the time since Fedora29 doesn't have libopencsd available: $ cat /tmp/build/perf/feature/test-all.make.output In file included from test-all.c:174: test-libopencsd.c:2:10: fatal error: opencsd/c_api/opencsd_c_api.h: No such file or directory #include <opencsd/c_api/opencsd_c_api.h> ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ compilation terminated. See: 6ab2b762 ("perf build: Disable libbabeltrace check by default") For the rationale, as soon as libopencsd becomes more generally packaged and available, we do the same thing we did with babeltrace, enabling it by default, as done in: 24787afb ("perf tools: Enable LIBBABELTRACE by default") For now, to explicitely ask for opencsd, make sure you have it installed and use: make -C tools/perf CORESIGHT=1 The feature test output will be there as an empty file: $ ls -la /tmp/build/perf/feature/test-libopencsd.make.output Because the binary used for the feature check was successfully built: $ ls -la /tmp/build/perf/feature/test-libopencsd.bin -rwxrwxr-x. 1 acme acme 18336 Feb 12 14:49 /tmp/build/perf/feature/test-libopencsd.bin $ ldd /tmp/build/perf/feature/test-libopencsd.bin linux-vdso.so.1 (0x00007fffe18cc000) libopencsd_c_api.so.0 => /lib64/libopencsd_c_api.so.0 (0x00007fb8e67f6000) libopencsd.so.0 => /lib64/libopencsd.so.0 (0x00007fb8e676f000) libc.so.6 => /lib64/libc.so.6 (0x00007fb8e65a9000) libstdc++.so.6 => /lib64/libstdc++.so.6 (0x00007fb8e6411000) libm.so.6 => /lib64/libm.so.6 (0x00007fb8e628d000) libgcc_s.so.1 => /lib64/libgcc_s.so.1 (0x00007fb8e6272000) /lib64/ld-linux-x86-64.so.2 (0x00007fb8e6828000) $ And the resulting perf binary will be linked with it: -rw-rw-r--. 1 acme acme 0 Feb 12 14:49 /tmp/build/perf/feature/test-libopencsd.make.output $ ldd ~/bin/perf | grep opencsd libopencsd_c_api.so.0 => /lib64/libopencsd_c_api.so.0 (0x00007fd43097f000) libopencsd.so.0 => /lib64/libopencsd.so.0 (0x00007fd4308f8000) $ To make sure this gets built before pushing things upstream I have a ubuntu:19.04-x-arm64 container that has: [root@quaco x-arm64]# grep CORESIGHT Dockerfile ENV EXTRA_MAKE_ARGS=CORESIGHT=1 [root@quaco x-arm64]# So that I always build with libopencsd before pushing things upstream. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: linux-arm-kernel@lists.infradead.org Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Cc: Mike Leach <mike.leach@arm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Suzuki Poulouse <suzuki.poulose@arm.com> Link: https://lkml.kernel.org/n/tip-20vyy39jw9jgrijesi30fgox@git.kernel.orgSigned-off-by: -
Arnaldo Carvalho de Melo authored
Just like it does with 'sshd', to reduce the feedback loop when doing system wide tracing on on a gnome GUI. Need to figure out how to auto-filter the calls to other UI components tho. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-rjopq5y92itgokppdhe8sc6z@git.kernel.orgSigned-off-by:
-
Arnaldo Carvalho de Melo authored
Since we need it to resolve the AIO symbols, otherwise we fail with: $ cat /tmp/build/perf/feature/test-all.make.output /usr/bin/ld: /tmp/ccEqrj36.o: undefined reference to symbol 'aio_return64@@GLIBC_2.2.5' /usr/bin/ld: //usr/lib64/librt.so.1: error adding symbols: DSO missing from command line collect2: error: ld returned 1 exit status $ When we added the aio support in 'perf record' only the test-libaio.bin target got the -lrt, i.e. the feature detection slow path. Fix it. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Alexey Budankov <alexey.budankov@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Fixes: 2a07d814 ("tools build feature: Check if libaio is available") Signed-off-by:
-
Arnaldo Carvalho de Melo authored
When introducing the possibility for selecting if the common prefix to options such as the waitid ones, i.e. all 'waitid' options start with 'W', so, to make it make it more compact if configured to suppress it, 'perf trace' will do so, other examples include mmap's PROT_ prefix for its 'prot' argument, etc, which, when showing the syscall argument name ends up producing duplicated info that clutters the screen, i.e.: # perf trace -e mmap --max-events 2 sleep 1 0.000 ( 0.014 ms): sleep/20886 mmap(len: 112595, prot: PROT_READ, flags: MAP_PRIVATE, fd: 3) = 0x7f3e986d2000 0.041 ( 0.005 ms): sleep/20886 mmap(len: 8192, prot: PROT_READ|PROT_WRITE, flags: MAP_PRIVATE|MAP_ANONYMOUS) = 0x7f3e986d0000 # So it is possible to suppress that and make it more compact by having this in your ~/.perfconfig: # cat ~/.perfconfig [trace] show_prefix = no # # perf trace -e mmap --max-events 2 sleep 1 0.000 ( 0.014 ms): sleep/8009 mmap(len: 112595, prot: READ, flags: PRIVATE, fd: 3) = 0x7ff2373de000 0.040 ( 0.005 ms): sleep/8009 mmap(len: 8192, prot: READ|WRITE, flags: PRIVATE|ANONYMOUS) = 0x7ff2373dc000 # To have it look more like strace's output, we instead want to suppress the arg name and show the prefix, so use: # cat ~/.perfconfig [trace] show_prefix = yes show_arg_names = no # # perf trace -e mmap --max-events 2 sleep 1 0.000 ( 0.006 ms): sleep/15513 mmap(NULL, 112595, PROT_READ, MAP_PRIVATE, 3, 0) = 0x7f7a9b6d3000 0.020 ( 0.002 ms): sleep/15513 mmap(NULL, 8192, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS) = 0x7f7a9b6d1000 # When this logic was introduced a bug came with it when processing the waitid 'option' arg that ended up expecting 3 strings when just two were being provided, fix it. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Fixes: c65c83ff ("perf trace: Allow asking for not suppressing common string prefixes") Signed-off-by: -
Arnaldo Carvalho de Melo authored
We were crashing when processing a negative fd: Program received signal SIGSEGV, Segmentation fault. 0x0000000000609bbf in syscall_arg__scnprintf_ioctl_cmd (bf=0x1172eca "", size=2038, arg=0x7fffffff8360) at trace/beauty/ioctl.c:182 182 if (file->dev_maj == USB_DEVICE_MAJOR) Missing separate debuginfos, use: dnf debuginfo-install bzip2-libs-1.0.6-28.fc29.x86_64 elfutils-libelf-0.174-5.fc29.x86_64 elfutils-libs-0.174-5.fc29.x86_64 glib2-2.58.3-1.fc29.x86_64 libbabeltrace-1.5.6-1.fc29.x86_64 libunwind-1.2.1-6.fc29.x86_64 libuuid-2.32.1-1.fc29.x86_64 libxcrypt-4.4.3-2.fc29.x86_64 numactl-libs-2.0.12-1.fc29.x86_64 openssl-libs-1.1.1a-1.fc29.x86_64 pcre-8.42-6.fc29.x86_64 perl-libs-5.28.1-427.fc29.x86_64 popt-1.16-15.fc29.x86_64 python2-libs-2.7.15-11.fc29.x86_64 slang-2.3.2-4.fc29.x86_64 xz-libs-5.2.4-3.fc29.x86_64 (gdb) bt #0 0x0000000000609bbf in syscall_arg__scnprintf_ioctl_cmd (bf=0x1172eca "", size=2038, arg=0x7fffffff8360) at trace/beauty/ioctl.c:182 #1 0x000000000048e295 in syscall__scnprintf_val (sc=0x123b500, bf=0x1172eca "", size=2038, arg=0x7fffffff8360, val=21519) at builtin-trace.c:1594 #2 0x000000000048e60d in syscall__scnprintf_args (sc=0x123b500, bf=0x1172ec6 "-1, ", size=2042, args=0x7ffff6a7c034 "\377\377\377\377", augmented_args=0x7ffff6a7c064, augmented_args_size=4, trace=0x7fffffffa8d0, thread=0x1175cd0) at builtin-trace.c:1661 #3 0x000000000048f04e in trace__sys_enter (trace=0x7fffffffa8d0, evsel=0xb260b0, event=0x7ffff6a7bfe8, sample=0x7fffffff84f0) at builtin-trace.c:1880 #4 0x00000000004915a4 in trace__handle_event (trace=0x7fffffffa8d0, event=0x7ffff6a7bfe8, sample=0x7fffffff84f0) at builtin-trace.c:2590 #5 0x0000000000491eed in __trace__deliver_event (trace=0x7fffffffa8d0, event=0x7ffff6a7bfe8) at builtin-trace.c:2818 #6 0x0000000000492030 in trace__deliver_event (trace=0x7fffffffa8d0, event=0x7ffff6a7bfe8) at builtin-trace.c:2845 #7 0x0000000000492896 in trace__run (trace=0x7fffffffa8d0, argc=0, argv=0x7fffffffdb58) at builtin-trace.c:3040 #8 0x000000000049603a in cmd_trace (argc=0, argv=0x7fffffffdb58) at builtin-trace.c:3952 #9 0x00000000004d5103 in main (argc=1, argv=0x7fffffffdb58) at perf.c:474 (gdb) p fd $1 = -1 (gdb) p file $7 = (struct file *) 0xfffffffffffffff0 (gdb) p ((struct thread_trace *)arg->thread)->files.table + fd $8 = (struct file *) 0xfffffffffffffff0 (gdb) Check for that and return NULL instead. This problem was introduced recently, the other codepaths leading to thread_trace__files_entry() check for negative fds, like thread__fd_path(), but we need to do it at thread_trace__files_entry() as more users are now calling it directly. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Fixes: 2d473389 ("perf trace beauty: Export function to get the files for a thread") Link: https://lkml.kernel.org/n/tip-oq7bvaaf07gsd4yqty3107u2@git.kernel.orgSigned-off-by: -
Arnaldo Carvalho de Melo authored
It is possible to pass a negative number as the fd and that has to be handled, so stop using 'unsigned int fd' in the ioctl syscall 'cmd' beautifier. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-b7qwa0l19dswa09h3s41akfu@git.kernel.orgSigned-off-by:
-
Arnaldo Carvalho de Melo authored
Since we get all the tests in a single .c file for a first test, tools/build/feature/test-all.c, if individual tests set that define and fail to undef it at its end, then it the test-all.c build will fail due to defining _GNU_SOURCE multiple times, getting us to the slow path, so undef it at the end in tests that define it. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-w6s00jfo1xabgphzczadl59b@git.kernel.orgSigned-off-by:
-
Song Liu authored
Synthesizing BPF events is only supported for root. Silent warning msg when non-root user runs perf-record. Reported-by:
David Carrillo-Cisneros <davidca@fb.com> Signed-off-by:
Song Liu <songliubraving@fb.com> Tested-by:
-
Paul Clarke authored
Descriptions of metrics for POWER9 processors can be found in the "POWER9 Performance Monitor Unit User’s Guide", which is currently available on the "IBM Portal for OpenPOWER" (https://www-355.ibm.com/systems/power/openpower/welcome.xhtml) at https://www-355.ibm.com/systems/power/openpower/posting.xhtml?postingId=4948CDE1963C9BCA852582F800718190 This patch is for metric groups: - general and other metrics not in a metric group. Signed-off-by:
Paul Clarke <pc@us.ibm.com> Cc: Ananth N Mavinakayanahalli <ananth@linux.vnet.ibm.com> Cc: Carl Love <cel@us.ibm.com> Cc: Madhavan Srinivasan <maddy@linux.vnet.ibm.com> Cc: Michael Ellerman <mpe@ellerman.id.au> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> Cc: linuxppc-dev@ozlabs.org Link: http://lkml.kernel.org/r/20190209181429.23950-5-pc@us.ibm.comSigned-off-by:
-
Paul Clarke authored
perf vendor events power9: Branch_prediction, instruction_stats, latency, lsu_rejects, memory, prefetch & translation metrics Descriptions of metrics for POWER9 processors can be found in the "POWER9 Performance Monitor Unit User’s Guide", which is currently available on the "IBM Portal for OpenPOWER" (https://www-355.ibm.com/systems/power/openpower/welcome.xhtml) at https://www-355.ibm.com/systems/power/openpower/posting.xhtml?postingId=4948CDE1963C9BCA852582F800718190 This patch is for metric groups: - branch_prediction - instruction_stats_percent_per_ref - latency - lsu_rejects - memory - prefetch - translation Plus, some whitespace changes. Signed-off-by:
-
Paul Clarke authored
Descriptions of metrics for POWER9 processors can be found in the "POWER9 Performance Monitor Unit User’s Guide", which is currently available on the "IBM Portal for OpenPOWER" (https://www-355.ibm.com/systems/power/openpower/welcome.xhtml) at https://www-355.ibm.com/systems/power/openpower/posting.xhtml?postingId=4948CDE1963C9BCA852582F800718190 This patch is for metric groups: - dl1_reloads_percent_per_inst - dl1_reloads_percent_per_ref - instruction_misses_percent_per_inst - l2_stats - l3_stats - pteg_reloads_percent_per_inst - pteg_reloads_percent_per_ref Signed-off-by:
-
Paul Clarke authored
Descriptions of metrics for POWER9 processors can be found in the "POWER9 Performance Monitor Unit User’s Guide", which is currently available on the "IBM Portal for OpenPOWER" (https://www-355.ibm.com/systems/power/openpower/welcome.xhtml) at https://www-355.ibm.com/systems/power/openpower/posting.xhtml?postingId=4948CDE1963C9BCA852582F800718190 This patch is for metric groups: - cpi_breakdown - estimated_dcache_miss_cpi Signed-off-by:
-
Paul Clarke authored
POWER8 metrics are not well publicized. Some are here: https://www.ibm.com/support/knowledgecenter/en/SSFK5S_2.2.0/com.ibm.cluster.pedev.v2r2.pedev100.doc/bl7ug_derivedmetricspower8.htm This patch is for metric groups: - translation - general and other metrics not in a metric group. Signed-off-by:
-
Paul Clarke authored
perf vendor events power8: Branch_prediction, latency, bus_stats, instruction_mix & instruction_stats metrics POWER8 metrics are not well publicized. Some are here: https://www.ibm.com/support/knowledgecenter/en/SSFK5S_2.2.0/com.ibm.cluster.pedev.v2r2.pedev100.doc/bl7ug_derivedmetricspower8.htm This patch is for metric groups: - branch_prediction - latency - bus_stats - instruction_mix - instruction_stats_percent_per_ref Signed-off-by:
-
Paul Clarke authored
perf vendor events power8: Dl1_reload, instruction_misses, l2_stats, lsu_rejects, memory & pteg_reloads metrics POWER8 metrics are not well publicized. Some are here: https://www.ibm.com/support/knowledgecenter/en/SSFK5S_2.2.0/com.ibm.cluster.pedev.v2r2.pedev100.doc/bl7ug_derivedmetricspower8.htm This patch is for metric groups: - dl1_reloads_percent_per_inst - dl1_reloads_percent_per_ref - instruction_misses_percent_per_inst - l2_stats - lsu_rejects - memory - pteg_reloads_percent_per_inst - pteg_reloads_percent_per_ref Signed-off-by:
-
Paul Clarke authored
POWER8 metrics are not well publicized. Some are here: https://www.ibm.com/support/knowledgecenter/en/SSFK5S_2.2.0/com.ibm.cluster.pedev.v2r2.pedev100.doc/bl7ug_derivedmetricspower8.htm This patch is for metric groups: - cpi_breakdown - estimated_dcache_miss_cpi Signed-off-by:
-
Thomas Richter authored
On IBM z13 machine types 2964 and 2965 the descriptor sizes for sampling and diagnostic sampling entries might be missing in the trailer entry and are set to zero. This leads to a perf report failure when processing diagnostic sampling entries. This patch adds missing descriptor sizes when the trailer entry contains zero for these fields. Output before: [root@s38lp82 perf]# ./perf report --stdio | fgrep Samples 0xabbf0 [0x8]: failed to process type: 68 Error: failed to process sample [root@s38lp82 perf]# Output after: [root@s38lp82 perf]# ./perf report --stdio | fgrep Samples # Total Lost Samples: 0 # Samples: 3K of event 'SF_CYCLES_BASIC_DIAG' # Samples: 162 of event 'CF_DIAG' [root@s38lp82 perf]# Fixes: 2b1444f2 ("perf report: Add raw report support for s390 auxiliary trace") Signed-off-by:
Thomas Richter <tmricht@linux.ibm.com> Reviewed-by:
Hendrik Brueckner <brueckner@linux.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Link: http://lkml.kernel.org/r/20190211100627.85714-1-tmricht@linux.ibm.comSigned-off-by:
-
Mathieu Poirier authored
After 'commit e22c1c75 ("perf thread: Don't include symbol.h, symbol_conf.h is enough")' Compilation of the perf tools is broken when using the functionality provided by the openCSD library: [...] ... timerfd: [ on ] ... sched_getcpu: [ on ] ... sdt: [ OFF ] ... setns: [ on ] ... libopencsd: [ on ] [...] CC util/arm-spe.o CC util/arm-spe-pkt-decoder.o CC util/s390-cpumsf.o CC util/cs-etm.o CC util/parse-branch-options.o util/cs-etm.c: In function ‘cs_etm__mem_access’: util/cs-etm.c:297:24: error: storage size of ‘al’ isn’t known struct addr_location al; And rightly so since file cs-etm.c doesn't include symbol.h, something that is rectified in this patch. Signed-off-by:
Mathieu Poirier <mathieu.poirier@linaro.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Suzuki K Poulouse <suzuki.poulose@arm.com> Cc: linux-arm-kernel@lists.infradead.org Link: http://lkml.kernel.org/r/20190208223543.31836-1-mathieu.poirier@linaro.orgSigned-off-by:
-
- 11 Feb, 2019 7 commits
-
-
Alexey Budankov authored
Implement --affinity=node|cpu option for the record mode defaulting to system affinity mask bouncing. Signed-off-by:
Alexey Budankov <alexey.budankov@linux.intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/083f5422-ece9-10dd-8305-bf59c860f10f@linux.intel.comSigned-off-by:
-
Kan Liang authored
A microcode patch is also needed for Goldmont while counter freezing feature is enabled. Otherwise, there will be some issues, e.g. PMI lost. Signed-off-by:
Kan Liang <kan.liang@linux.intel.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: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vince Weaver <vincent.weaver@maine.edu> Cc: bp@alien8.de Link: https://lkml.kernel.org/r/1549319013-4522-5-git-send-email-kan.liang@linux.intel.comSigned-off-by:
Ingo Molnar <mingo@kernel.org>
-
Kan Liang authored
Clean up counter freezing quirk to use the new facility to check for min microcode revisions. Rename the counter freezing quirk related functions. Because other platforms, e.g. Goldmont, also needs to call the quirk. Only check the boot CPU, assuming models and features are consistent over all CPUs. Signed-off-by:
-
Kan Liang authored
Clean up SNB PEBS quirk to use the new facility to check for min microcode revisions. Only check the boot CPU, assuming models and features are consistent over all CPUs. Signed-off-by:
-
Andi Kleen authored
KVM added a workaround for PEBS events leaking into guests with commit: 26a4f3c0 ("perf/x86: disable PEBS on a guest entry.") This uses the VT entry/exit list to add an extra disable of the PEBS_ENABLE MSR. Intel also added a fix for this issue to microcode updates on Haswell/Broadwell/Skylake. It turns out using the MSR entry/exit list makes VM exits significantly slower. The list is only needed for disabling PEBS, because the GLOBAL_CTRL change gets optimized by KVM into changing the VMCS. Check for the microcode updates that have the microcode fix for leaking PEBS, and disable the extra entry/exit list entry for PEBS_ENABLE. In addition we always clear the GLOBAL_CTRL for the PEBS counter while running in the guest, which is enough to make them never fire at the wrong side of the host/guest transition. The overhead for VM exits with the filtering active with the patch is reduced from 8% to 4%. The microcode patch has already been merged into future platforms. This patch is one-off thing. The quirks is used here. For other old platforms which doesn't have microcode patch and quirks, extra disable of the PEBS_ENABLE MSR is still required. Signed-off-by:
Andi Kleen <ak@linux.intel.com> Signed-off-by:
-
Ingo Molnar authored
Signed-off-by: -
Kan Liang authored
For bug workarounds or checks, it is useful to check for specific microcode revisions. Add a new generic function to match the CPU with stepping. Add the other function to check the min microcode revisions for the matched CPU. A new table format is introduced to facilitate the quirk to fill the related information. This does not change the existing x86_cpu_id because it's an ABI shared with modules, and also has quite different requirements, as in no wildcards, but everything has to be matched exactly. Originally-by:
Thomas Gleixner <tglx@linutronix.de> Signed-off-by:
Borislav Petkov <bp@alien8.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: eranian@google.com Link: https://lkml.kernel.org/r/1549319013-4522-1-git-send-email-kan.liang@linux.intel.comSigned-off-by:
-
- 09 Feb, 2019 3 commits
-
-
Ingo Molnar authored
Merge tag 'perf-core-for-mingo-5.1-20190206' 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: Hardware tracing: Adrian Hunter: - Handle calls optimized into jumps to a different symbol in the thread stack routines used to process hardware traces (Adrian Hunter) Intel PT: Adrian Hunter: - Fix overlap calculation for padding. - Fix CYC timestamp calculation after OVF. - Packet splitting can only happen in 32-bit. - Add timestamp to auxtrace errors. ARM CoreSight: Leo Yan: - Add last instruction information in packet - Set sample flags for instruction range, exception and return packets and for a trace discontinuity. - Add exception number in exception packet - Change tuple from traceID-CPU# to traceID-metadata - Add traceID in packet Mathieu Poirier: - Add "sinks" group to PMU directory - Use event attributes to send sink information to kernel - Remove set_drv_config() API, no longer used. perf annotate: Jiri Olsa: - Delay symbol annotation to the resort phase, speeding up 'perf report' startup. perf record: Alexey Budankov: - Allow binding userspace buffers to NUMA nodes. Symbols: Adrian Hunter: - Fix calculating of symbol sizes when splitting kallsyms into maps for kcore processing. Vendor events: William Cohen: - Intel: Fix Load_Miss_Real_Latency on CLX Misc: Arnaldo Carvalho de Melo: - Streamline headers, removing includes when all that is needed are just forward declarations, fixup the fallout for cases where headers should have been explicitely included but were instead obtained indirectly, by sheer luck. - Add fallback versions for CPU_{OR,EQUAL}(), so that code using it continue to build on older systems where those were not yet introduced or in systems using some other libc than the GNU one where those helpers aren't present. Documentation: Changbin Du: - Add documentation for BPF event selection. Signed-off-by:
-
Ingo Molnar authored
Signed-off-by: -
Ingo Molnar authored
Merge tag 'perf-urgent-for-mingo-5.0-20190205' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/urgent Pull perf/urgent fixes from Arnaldo Carvalho de Melo: perf trace: Arnaldo Carvalho de Melo: Fix handling of probe:vfs_getname when the probed routine is inlined in multiple places, fixing the collection of the 'filename' parameter in open syscalls. perf test: Gustavo A. R. Silva: Fix bitwise operator usage in evsel-tp-sched test, which made tat test always detect fields as signed. Jiri Olsa: Filter out hidden symbols from labels, added in systems where the annobin plugin is used, such as RHEL8, which, if left in place make the DWARF unwind 'perf test' to fail on PPC. Tony Jones: Fix 'perf_event_attr' tests when building with python3. perf mem/c2c: Ravi Bangoria: Fix perf_mem_events on PowerPC. tools headers UAPI: Arnaldo Carvalho de Melo: Sync linux/in.h copy from the kernel sources, silencing a perf build warning. Signed-off-by:
-
- 06 Feb, 2019 9 commits
-
-
Adrian Hunter authored
The timestamp can use useful to find part of a trace that has an error without outputting all of the trace e.g. using the itrace 's' option to skip initial number of events. Signed-off-by:
Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@redhat.com> Link: http://lkml.kernel.org/r/20190206103947.15750-6-adrian.hunter@intel.comSigned-off-by:
-
Adrian Hunter authored
Data is copied when the trace is stopped, so packets are never split between buffers except when processing if the buffer cannot fit in the address space which can only happen on 32-bit systems. Change the logic to reflect that. Signed-off-by:
-
Adrian Hunter authored
CYC packet timestamp calculation depends upon CBR which was being cleared upon overflow (OVF). That can cause errors due to failing to synchronize with sideband events. Even if a CBR change has been lost, the old CBR is still a better estimate than zero. So remove the clearing of CBR. Signed-off-by:
-
Adrian Hunter authored
Auxtrace records might have up to 7 bytes of padding appended. Adjust the overlap accordingly. Signed-off-by:
-
Adrian Hunter authored
Define auxtrace record alignment so that it can be referenced elsewhere. Note this is preparation for patch "perf intel-pt: Fix overlap calculation for padding" Signed-off-by:
-
Adrian Hunter authored
The compiler might optimize a call/ret combination by making it a jmp. However the thread-stack does not presently cater for that, so that such control flow is not visible in the call graph. Make it visible by recording on the stack a branch to the start of a different symbol. Note, that means when a ret pops the stack, all jmps must be popped off first. Example: $ cat jmp-to-fn.c __attribute__((noinline)) int bar(void) { return -1; } __attribute__((noinline)) int foo(void) { return bar() + 1; } int main() { return foo(); } $ gcc -ggdb3 -Wall -Wextra -O2 -o jmp-to-fn jmp-to-fn.c $ objdump -d jmp-to-fn <SNIP> 0000000000001040 <main>: 1040: 31 c0 xor %eax,%eax 1042: e9 09 01 00 00 jmpq 1150 <foo> <SNIP> 0000000000001140 <bar>: 1140: b8 ff ff ff ff mov $0xffffffff,%eax 1145: c3 retq <SNIP> 0000000000001150 <foo>: 1150: 31 c0 xor %eax,%eax 1152: e8 e9 ff ff ff callq 1140 <bar> 1157: 83 c0 01 add $0x1,%eax 115a: c3 retq <SNIP> $ perf record -o jmp-to-fn.perf.data -e intel_pt/cyc/u ./jmp-to-fn [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0,017 MB jmp-to-fn.perf.data ] $ perf script -i jmp-to-fn.perf.data --itrace=be -s ~/libexec/perf-core/scripts/python/export-to-sqlite.py jmp-to-fn.db branches calls 2019-01-08 13:24:58.783069 Creating database... 2019-01-08 13:24:58.794650 Writing records... 2019-01-08 13:24:59.008050 Adding indexes 2019-01-08 13:24:59.015802 Done $ ~/libexec/perf-core/scripts/python/exported-sql-viewer.py jmp-to-fn.db Before: main -> bar After: main -> foo -> bar Committer testing: Install the python2-pyside package, then select these menu options on the GUI: "Reports" "Context sensitive callgraphs" Then go on expanding the symbols, to get, full picture when doing this on a fedora:29 with gcc version 8.2.1 20181215 (Red Hat 8.2.1-6) (GCC): jmp-to-fn PID:TID _start (ld-2.28.so) __libc_start_main main foo bar To verify that indeed, this fixes the problem. Signed-off-by: -
Adrian Hunter authored
Make thread_stack__no_call_return() more readable by adding more local variables. Signed-off-by:
-
Adrian Hunter authored
If 'cp' is checked in thread_stack__push_cp() a number of error checks can be removed, reducing code size and improving readability. Signed-off-by:
-
Adrian Hunter authored
Kallsyms symbols do not have a size, so the size becomes the distance to the next symbol. Consequently the recently added trampoline symbols end up with large sizes because the trampolines are some distance from one another and the main kernel map. However, symbols that end outside their map can disrupt the symbol tree because, after mapping, it can appear incorrectly that they overlap other symbols. Add logic to truncate symbol size to the end of the corresponding map. Signed-off-by:
-
