perf/core improvements and fixes
Subsystems (7)
Arnaldo Carvalho de Melo
maintainerHi Ingo,
Please consider pulling, I pulled tip/perf/urgent into
tip/pref/core, IIRC was just a fast forward at that point, yeap, just did it again and it still is:
$ git checkout -b t tip/perf/core Branch 't' set up to track remote branch 'perf/core' from 'tip'. Switched to a new branch 't' $ git merge tip/perf/urgent Updating d15d356887e7..c7a286577d75 Fast-forward <SNIP>
IIRC Jiri needs this for a pile of patches he submitted and
that I'll process next,
Best regards,
- Arnaldo
Test results at the end of this message, as usual.
The following changes since commit 6b89d4c1ae8596a8c9240f169ef108704de373f2:
perf/x86/intel: Fix INTEL_FLAGS_EVENT_CONSTRAINT* masking (2019-05-10 08:04:17 +0200)
are available in the Git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux.git tags/perf-core-for-mingo-5.2-20190517
for you to fetch changes up to 4fc4d8dfa056dfd48afe73b9ea3b7570ceb80b9c:
perf stat: Support 'percore' event qualifier (2019-05-16 14:17:24 -0300)
----------------------------------------------------------------
perf/core improvements and fixes:
perf.data:
Alexey Budankov:
- Streaming compression of perf ring buffer into PERF_RECORD_COMPRESSED user space records, resulting in ~3-5x perf.data file size reduction on variety of tested workloads what saves storage space on larger server systems where perf.data size can easily reach several tens or even hundreds of GiBs, especially when profiling with DWARF-based stacks and tracing of context switches.
perf record:
Arnaldo Carvalho de Melo
- Improve -user-regs/intr-regs suggestions to overcome errors.
perf annotate:
Jin Yao:
- Remove hist__account_cycles() from callback, speeding up branch processing (perf record -b).
perf stat:
- Add a 'percore' event qualifier, e.g.: -e cpu/event=0,umask=0x3,percore=1/, that sums up the event counts for both hardware threads in a core.
We can already do this with --per-core, but it's often useful to do this together with other metrics that are collected per hardware thread.
I.e. now its possible to do this per-event, and have it mixed with other events not aggregated by core.
core libraries:
Donald Yandt:
- Check for errors when doing fgets(/proc/version).
Jiri Olsa:
- Speed up report for perf compiled with linbunwind.
tools headers:
Arnaldo Carvalho de Melo
- Update memcpy_64.S, x86's kvm.h and pt_regs.h.
arm64:
Florian Fainelli:
- Map Brahma-B53 CPUID to cortex-a53 events.
- Add Cortex-A57 and Cortex-A72 events.
csky:
Mao Han:
- Add DWARF register mappings for libdw, allowing --call-graph=dwarf to work on the C-SKY arch.
x86:
Andi Kleen/Kan Liang:
- Add support for recording and printing XMM registers, available, for instance, on Icelake.
Kan Liang:
- Add uncore_upi (Intel's "Ultra Path Interconnect" events) JSON support. UPI replaced the Intel QuickPath Interconnect (QPI) in Xeon Skylake-SP.
Intel PT:
Adrian Hunter
. Fix instructions sampling rate.
. Timestamp fixes.
. Improve exported-sql-viewer GUI, allowing, for instance, to copy'n'paste
the trees, useful for e-mailing.
Documentation:
Thomas Richter:
- Add description for 'perf --debug stderr=1', which redirects stderr to stdout.
libtraceevent:
Tzvetomir Stoyanov:
- Add man pages for the various APIs.
----------------------------------------------------------------
Adrian Hunter (9):
perf scripts python: exported-sql-viewer.py: Move view creation
perf scripts python: exported-sql-viewer.py: Fix error when shrinking / enlarging font perf scripts python: exported-sql-viewer.py: Add tree level
perf scripts python: exported-sql-viewer.py: Add copy to clipboard
perf scripts python: exported-sql-viewer.py: Add context menu
perf scripts python: exported-sql-viewer.py: Add 'About' dialog box perf intel-pt: Fix instructions sampling rate
perf intel-pt: Fix improved sample timestamp
perf intel-pt: Fix sample timestamp wrt non-taken branches
Alexey Budankov (11):
perf session: Define 'bytes_transferred' and 'bytes_compressed' metrics perf record: Implement COMPRESSED event record and its attributes perf mmap: Implement dedicated memory buffer for data compression perf tools: Introduce Zstd streaming based compression API perf record: Implement compression for serial trace streaming perf record: Implement compression for AIO trace streaming perf report: Add stub processing of compressed events for -D perf record: Implement -z,--compression_level[=<n>] option perf report: Implement perf.data record decompression
perf inject: Enable COMPRESSED record decompression
perf tests: Implement Zstd comp/decomp integration test
Andi Kleen (1):
perf tools x86: Add support for recording and printing XMM registers
Arnaldo Carvalho de Melo (8):
tools arch: Update arch/x86/lib/memcpy_64.S copy used in 'perf bench mem memcpy' tools arch uapi: Sync the x86 kvm.h copy
tools x86 uapi asm: Sync the pt_regs.h copy with the kernel sources
tools pci: Do not delete pcitest.sh in 'make clean'
perf record: Fix suggestion to get list of registers usable with --user-regs and --intr-regs perf parse-regs: Improve error output when faced with unknown register name perf build tests: Add NO_LIBZSTD=1 to make_minimal
perf test zstd: Fixup verbose mode output
Colin Ian King (1):
perf test: Fix spelling mistake "leadking" -> "leaking"
Donald Yandt (1):
perf machine: Null-terminate version char array upon fgets(/proc/version) error
Florian Fainelli (3):
perf vendor events arm64: Remove [[:xdigit:]] wildcard perf vendor events arm64: Map Brahma-B53 CPUID to cortex-a53 events perf vendor events arm64: Add Cortex-A57 and Cortex-A72 events
Jin Yao (4):
perf annotate: Remove hist__account_cycles() from callback perf tools: Add a 'percore' event qualifier
perf stat: Factor out aggregate counts printing perf stat: Support 'percore' event qualifier
Jiri Olsa (1):
perf tools: Speed up report for perf compiled with linwunwind
Kan Liang (4):
perf vendor events intel: Add uncore_upi JSON support perf parse-regs: Split parse_regs
perf parse-regs: Add generic support for arch__intr/user_reg_mask() perf regs x86: Add X86 specific arch__intr_reg_mask()
Mao Han (1):
csky: Add support for libdw
Thomas Richter (1):
perf docs: Add description for stderr
Tzvetomir Stoyanov (27):
tools lib traceevent: Remove hard coded install paths from pkg-config file tools lib traceevent: Introduce man pages
tools lib traceevent: Add support for man pages with multiple names tools lib traceevent: Man pages for tep_handler related APIs
tools lib traceevent: Man page for header_page APIs
tools lib traceevent: Man page for get/set cpus APIs
tools lib traceevent: Man page for file endian APIs
tools lib traceevent: Man page for host endian APIs
tools lib traceevent: Man page for page size APIs
tools lib traceevent: Man page for tep_strerror()
tools lib traceevent: Man pages for event handler APIs
tools lib traceevent: Man pages for function related libtraceevent APIs tools lib traceevent: Man pages for registering print function
tools lib traceevent: Man page for tep_read_number()
tools lib traceevent: Man pages for event find APIs
tools lib traceevent: Man page for list events APIs
tools lib traceevent: Man pages for libtraceevent event get APIs tools lib traceevent: Man pages for find field APIs
tools lib traceevent: Man pages for get field value APIs
tools lib traceevent: Man pages for print field APIs
tools lib traceevent: Man page for tep_read_number_field()
tools lib traceevent: Man pages for event fields APIs
tools lib traceevent: Man pages for event filter APIs
tools lib traceevent: Man pages for parse event APIs
tools lib traceevent: Man page for tep_parse_header_page()
tools lib traceevent: Man pages for APIs used to extract common fields from a record tools lib traceevent: Man pages for trace sequences APIs
Zenghui Yu (1):
perf jevents: Remove unused variable
tools/arch/csky/include/uapi/asm/perf_regs.h | 51 ++++ tools/arch/x86/include/uapi/asm/kvm.h | 1 + tools/arch/x86/include/uapi/asm/perf_regs.h | 23 +- tools/arch/x86/lib/memcpy_64.S | 3 +- tools/lib/traceevent/Documentation/Makefile | 207 +++++++++++++ tools/lib/traceevent/Documentation/asciidoc.conf | 120 ++++++++ .../Documentation/libtraceevent-commands.txt | 153 ++++++++++ .../Documentation/libtraceevent-cpus.txt | 77 +++++ .../Documentation/libtraceevent-endian_read.txt | 78 +++++ .../Documentation/libtraceevent-event_find.txt | 103 +++++++ .../Documentation/libtraceevent-event_get.txt | 99 ++++++ .../Documentation/libtraceevent-event_list.txt | 122 ++++++++ .../Documentation/libtraceevent-field_find.txt | 118 +++++++ .../Documentation/libtraceevent-field_get_val.txt | 122 ++++++++ .../Documentation/libtraceevent-field_print.txt | 126 ++++++++ .../Documentation/libtraceevent-field_read.txt | 81 +++++ .../Documentation/libtraceevent-fields.txt | 105 +++++++ .../Documentation/libtraceevent-file_endian.txt | 91 ++++++ .../Documentation/libtraceevent-filter.txt | 209 +++++++++++++ .../Documentation/libtraceevent-func_apis.txt | 183 +++++++++++ .../Documentation/libtraceevent-func_find.txt | 88 ++++++ .../Documentation/libtraceevent-handle.txt | 101 ++++++ .../Documentation/libtraceevent-header_page.txt | 102 +++++++ .../Documentation/libtraceevent-host_endian.txt | 104 +++++++ .../Documentation/libtraceevent-long_size.txt | 78 +++++ .../Documentation/libtraceevent-page_size.txt | 82 +++++ .../Documentation/libtraceevent-parse_event.txt | 90 ++++++ .../Documentation/libtraceevent-parse_head.txt | 82 +++++ .../Documentation/libtraceevent-record_parse.txt | 137 +++++++++ .../libtraceevent-reg_event_handler.txt | 156 ++++++++++ .../Documentation/libtraceevent-reg_print_func.txt | 155 ++++++++++ .../Documentation/libtraceevent-set_flag.txt | 104 +++++++ .../Documentation/libtraceevent-strerror.txt | 85 ++++++ .../Documentation/libtraceevent-tseq.txt | 158 ++++++++++ .../lib/traceevent/Documentation/libtraceevent.txt | 203 ++++++++++++ .../lib/traceevent/Documentation/manpage-1.72.xsl | 14 + .../lib/traceevent/Documentation/manpage-base.xsl | 35 +++ .../Documentation/manpage-bold-literal.xsl | 17 ++ .../traceevent/Documentation/manpage-normal.xsl | 13 + .../Documentation/manpage-suppress-sp.xsl | 21 ++ tools/lib/traceevent/Makefile | 46 ++- tools/lib/traceevent/libtraceevent.pc.template | 4 +- tools/pci/Makefile | 4 +- tools/perf/Documentation/perf-list.txt | 12 + tools/perf/Documentation/perf-record.txt | 8 +- tools/perf/Documentation/perf-stat.txt | 4 + tools/perf/Documentation/perf.data-file-format.txt | 24 ++ tools/perf/Documentation/perf.txt | 2 + tools/perf/Makefile.config | 6 +- tools/perf/arch/csky/Build | 1 + tools/perf/arch/csky/Makefile | 3 + tools/perf/arch/csky/include/perf_regs.h | 100 ++++++ tools/perf/arch/csky/util/Build | 2 + tools/perf/arch/csky/util/dwarf-regs.c | 49 +++ tools/perf/arch/csky/util/unwind-libdw.c | 77 +++++ tools/perf/arch/x86/include/perf_regs.h | 26 +- tools/perf/arch/x86/util/perf_regs.c | 44 +++ tools/perf/builtin-annotate.c | 4 +- tools/perf/builtin-inject.c | 4 + tools/perf/builtin-record.c | 229 ++++++++++++-- tools/perf/builtin-report.c | 16 +- tools/perf/builtin-stat.c | 21 ++ tools/perf/perf.h | 1 + .../arm64/arm/cortex-a57-a72/core-imp-def.json | 179 +++++++++++ tools/perf/pmu-events/arch/arm64/mapfile.csv | 5 +- tools/perf/pmu-events/jevents.c | 2 +- tools/perf/scripts/python/exported-sql-viewer.py | 340 ++++++++++++++++++++- tools/perf/tests/dso-data.c | 4 +- tools/perf/tests/make | 2 +- tools/perf/tests/shell/record+zstd_comp_decomp.sh | 34 +++ tools/perf/util/Build | 2 + tools/perf/util/annotate.c | 2 +- tools/perf/util/compress.h | 53 ++++ tools/perf/util/env.h | 11 + tools/perf/util/event.c | 1 + tools/perf/util/event.h | 7 + tools/perf/util/evlist.c | 8 +- tools/perf/util/evlist.h | 2 +- tools/perf/util/evsel.c | 2 + tools/perf/util/evsel.h | 3 + tools/perf/util/header.c | 53 ++++ tools/perf/util/header.h | 1 + .../perf/util/intel-pt-decoder/intel-pt-decoder.c | 31 +- tools/perf/util/machine.c | 3 +- tools/perf/util/mmap.c | 102 ++----- tools/perf/util/mmap.h | 16 +- tools/perf/util/parse-events.c | 27 ++ tools/perf/util/parse-events.h | 1 + tools/perf/util/parse-events.l | 1 + tools/perf/util/parse-regs-options.c | 33 +- tools/perf/util/parse-regs-options.h | 3 +- tools/perf/util/perf_regs.c | 10 + tools/perf/util/perf_regs.h | 3 + tools/perf/util/session.c | 133 +++++++- tools/perf/util/session.h | 14 + tools/perf/util/stat-display.c | 107 +++++-- tools/perf/util/stat.c | 8 +- tools/perf/util/thread.c | 3 +- tools/perf/util/tool.h | 2 + tools/perf/util/unwind-libunwind-local.c | 6 - tools/perf/util/unwind-libunwind.c | 10 + tools/perf/util/zstd.c | 111 +++++++ 102 files changed, 5703 insertions(+), 216 deletions(-) create mode 100644 tools/arch/csky/include/uapi/asm/perf_regs.h create mode 100644 tools/lib/traceevent/Documentation/Makefile create mode 100644 tools/lib/traceevent/Documentation/asciidoc.conf create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-commands.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-cpus.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-endian_read.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-event_find.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-event_get.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-event_list.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-field_find.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-field_get_val.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-field_print.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-field_read.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-fields.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-file_endian.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-filter.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-func_apis.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-func_find.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-handle.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-header_page.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-host_endian.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-long_size.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-page_size.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-parse_event.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-parse_head.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-record_parse.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-reg_event_handler.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-reg_print_func.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-set_flag.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-strerror.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-tseq.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent.txt create mode 100644 tools/lib/traceevent/Documentation/manpage-1.72.xsl create mode 100644 tools/lib/traceevent/Documentation/manpage-base.xsl create mode 100644 tools/lib/traceevent/Documentation/manpage-bold-literal.xsl create mode 100644 tools/lib/traceevent/Documentation/manpage-normal.xsl create mode 100644 tools/lib/traceevent/Documentation/manpage-suppress-sp.xsl create mode 100644 tools/perf/arch/csky/Build create mode 100644 tools/perf/arch/csky/Makefile create mode 100644 tools/perf/arch/csky/include/perf_regs.h create mode 100644 tools/perf/arch/csky/util/Build create mode 100644 tools/perf/arch/csky/util/dwarf-regs.c create mode 100644 tools/perf/arch/csky/util/unwind-libdw.c create mode 100644 tools/perf/pmu-events/arch/arm64/arm/cortex-a57-a72/core-imp-def.json create mode 100755 tools/perf/tests/shell/record+zstd_comp_decomp.sh create mode 100644 tools/perf/util/zstd.c
Test results:
The first ones are container based builds of tools/perf with and without libelf support. Where clang is available, it is also used to build perf with/without libelf, and building with LIBCLANGLLVM=1 (built-in clang) with gcc and clang when clang and its devel libraries are installed.
The objtool and samples/bpf/ builds are disabled now that I'm switching from using the sources in a local volume to fetching them from a http server to build it inside the container, to make it easier to build in a container cluster. Those will come back later.
Several are cross builds, the ones with -x-ARCH and the android one, and those may not have all the features built, due to lack of multi-arch devel packages, available and being used so far on just a few, like debian:experimental-x-{arm64,mipsel}.
The 'perf test' one will perform a variety of tests exercising tools/perf/util/, tools/lib/{bpf,traceevent,etc}, as well as run perf commands with a variety of command line event specifications to then intercept the sys_perf_event syscall to check that the perf_event_attr fields are set up as expected, among a variety of other unit tests.
Then there is the 'make -C tools/perf build-test' ones, that build tools/perf/ with a variety of feature sets, exercising the build with an incomplete set of features as well as with a complete one. It is planned to have it run on each of the containers mentioned above, using some container orchestration infrastructure. Get in contact if interested in helping having this in place.
$ export PERF_TARBALL=http://192.168.124.1/perf/perf-5.1.0.tar.xz $ dm 1 alpine:3.4 : Ok gcc (Alpine 5.3.0) 5.3.0 2 alpine:3.5 : Ok gcc (Alpine 6.2.1) 6.2.1 20160822 3 alpine:3.6 : Ok gcc (Alpine 6.3.0) 6.3.0 4 alpine:3.7 : Ok gcc (Alpine 6.4.0) 6.4.0 5 alpine:3.8 : Ok gcc (Alpine 6.4.0) 6.4.0 6 alpine:3.9 : Ok gcc (Alpine 8.3.0) 8.3.0 7 alpine:edge : Ok gcc (Alpine 8.3.0) 8.3.0 8 amazonlinux:1 : Ok gcc (GCC) 4.8.5 20150623 (Red Hat 4.8.5-28) 9 amazonlinux:2 : Ok gcc (GCC) 7.3.1 20180303 (Red Hat 7.3.1-5) 10 android-ndk:r12b-arm : Ok arm-linux-androideabi-gcc (GCC) 4.9.x 20150123 (prerelease) 11 android-ndk:r15c-arm : Ok arm-linux-androideabi-gcc (GCC) 4.9.x 20150123 (prerelease) 12 centos:5 : Ok gcc (GCC) 4.1.2 20080704 (Red Hat 4.1.2-55) 13 centos:6 : Ok gcc (GCC) 4.4.7 20120313 (Red Hat 4.4.7-23) 14 centos:7 : Ok gcc (GCC) 4.8.5 20150623 (Red Hat 4.8.5-36) 15 clearlinux:latest : Ok gcc (Clear Linux OS for Intel Architecture) 9.0.1 20190501 (prerelease) gcc-8-branch@270761 16 debian:8 : Ok gcc (Debian 4.9.2-10+deb8u2) 4.9.2 17 debian:9 : Ok gcc (Debian 6.3.0-18+deb9u1) 6.3.0 20170516 18 debian:experimental : Ok gcc (Debian 8.3.0-7) 8.3.0 19 debian:experimental-x-arm64 : Ok aarch64-linux-gnu-gcc (Debian 8.3.0-7) 8.3.0 20 debian:experimental-x-mips : Ok mips-linux-gnu-gcc (Debian 8.3.0-7) 8.3.0 21 debian:experimental-x-mips64 : Ok mips64-linux-gnuabi64-gcc (Debian 8.3.0-7) 8.3.0 22 debian:experimental-x-mipsel : Ok mipsel-linux-gnu-gcc (Debian 8.3.0-7) 8.3.0 23 fedora:20 : Ok gcc (GCC) 4.8.3 20140911 (Red Hat 4.8.3-7) 24 fedora:22 : Ok gcc (GCC) 5.3.1 20160406 (Red Hat 5.3.1-6) 25 fedora:23 : Ok gcc (GCC) 5.3.1 20160406 (Red Hat 5.3.1-6) 26 fedora:24 : Ok gcc (GCC) 6.3.1 20161221 (Red Hat 6.3.1-1) 27 fedora:24-x-ARC-uClibc : Ok arc-linux-gcc (ARCompact ISA Linux uClibc toolchain 2017.09-rc2) 7.1.1 20170710 28 fedora:25 : Ok gcc (GCC) 6.4.1 20170727 (Red Hat 6.4.1-1) 29 fedora:26 : Ok gcc (GCC) 7.3.1 20180130 (Red Hat 7.3.1-2) 30 fedora:27 : Ok gcc (GCC) 7.3.1 20180712 (Red Hat 7.3.1-6) 31 fedora:28 : Ok gcc (GCC) 8.3.1 20190223 (Red Hat 8.3.1-2) 32 fedora:29 : Ok gcc (GCC) 8.3.1 20190223 (Red Hat 8.3.1-2) 33 fedora:30 : Ok gcc (GCC) 9.1.1 20190503 (Red Hat 9.1.1-1) 34 fedora:30-x-ARC-glibc : Ok arc-linux-gcc (ARC HS GNU/Linux glibc toolchain 2019.03-rc1) 8.3.1 20190225 35 fedora:30-x-ARC-uClibc : Ok arc-linux-gcc (ARCv2 ISA Linux uClibc toolchain 2019.03-rc1) 8.3.1 20190225 36 fedora:rawhide : Ok gcc (GCC) 9.0.1 20190418 (Red Hat 9.0.1-0.14) 37 mageia:5 : Ok gcc (GCC) 4.9.2 38 mageia:6 : Ok gcc (Mageia 5.5.0-1.mga6) 5.5.0 39 opensuse:15.0 : Ok gcc (SUSE Linux) 7.4.0 40 opensuse:15.1 : Ok gcc (SUSE Linux) 7.4.0 41 opensuse:42.3 : Ok gcc (SUSE Linux) 4.8.5 42 opensuse:tumbleweed : Ok gcc (SUSE Linux) 8.3.1 20190226 [gcc-8-branch revision 269204] 43 oraclelinux:6 : Ok gcc (GCC) 4.4.7 20120313 (Red Hat 4.4.7-23.0.1) 44 oraclelinux:7 : Ok gcc (GCC) 4.8.5 20150623 (Red Hat 4.8.5-36.0.1) 45 ubuntu:12.04.5 : Ok gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3 46 ubuntu:14.04.4 : Ok gcc (Ubuntu 4.8.4-2ubuntu1~14.04.4) 4.8.4 47 ubuntu:16.04 : Ok gcc (Ubuntu 5.4.0-6ubuntu1~16.04.11) 5.4.0 20160609 48 ubuntu:16.04-x-arm : Ok arm-linux-gnueabihf-gcc (Ubuntu/Linaro 5.4.0-6ubuntu1~16.04.9) 5.4.0 20160609 49 ubuntu:16.04-x-arm64 : Ok aarch64-linux-gnu-gcc (Ubuntu/Linaro 5.4.0-6ubuntu1~16.04.9) 5.4.0 20160609 50 ubuntu:16.04-x-powerpc : Ok powerpc-linux-gnu-gcc (Ubuntu 5.4.0-6ubuntu1~16.04.9) 5.4.0 20160609 51 ubuntu:16.04-x-powerpc64 : Ok powerpc64-linux-gnu-gcc (Ubuntu/IBM 5.4.0-6ubuntu1~16.04.9) 5.4.0 20160609 52 ubuntu:16.04-x-powerpc64el : Ok powerpc64le-linux-gnu-gcc (Ubuntu/IBM 5.4.0-6ubuntu1~16.04.9) 5.4.0 20160609 53 ubuntu:16.04-x-s390 : Ok s390x-linux-gnu-gcc (Ubuntu 5.4.0-6ubuntu1~16.04.9) 5.4.0 20160609 54 ubuntu:17.10 : Ok gcc (Ubuntu 7.2.0-8ubuntu3.2) 7.2.0 55 ubuntu:18.04 : Ok gcc (Ubuntu 7.4.0-1ubuntu1~18.04) 7.4.0 56 ubuntu:18.04-x-arm : Ok arm-linux-gnueabihf-gcc (Ubuntu/Linaro 7.4.0-1ubuntu1~18.04) 7.4.0 57 ubuntu:18.04-x-arm64 : Ok aarch64-linux-gnu-gcc (Ubuntu/Linaro 7.4.0-1ubuntu1~18.04) 7.4.0 58 ubuntu:18.04-x-m68k : Ok m68k-linux-gnu-gcc (Ubuntu 7.4.0-1ubuntu1~18.04) 7.4.0 59 ubuntu:18.04-x-powerpc : Ok powerpc-linux-gnu-gcc (Ubuntu 7.4.0-1ubuntu1~18.04) 7.4.0 60 ubuntu:18.04-x-powerpc64 : Ok powerpc64-linux-gnu-gcc (Ubuntu 7.4.0-1ubuntu1~18.04) 7.4.0 61 ubuntu:18.04-x-powerpc64el : Ok powerpc64le-linux-gnu-gcc (Ubuntu 7.4.0-1ubuntu1~18.04) 7.4.0 62 ubuntu:18.04-x-riscv64 : Ok riscv64-linux-gnu-gcc (Ubuntu 7.4.0-1ubuntu1~18.04) 7.4.0 63 ubuntu:18.04-x-s390 : Ok s390x-linux-gnu-gcc (Ubuntu 7.4.0-1ubuntu1~18.04) 7.4.0 64 ubuntu:18.04-x-sh4 : Ok sh4-linux-gnu-gcc (Ubuntu 7.4.0-1ubuntu1~18.04) 7.4.0 65 ubuntu:18.04-x-sparc64 : Ok sparc64-linux-gnu-gcc (Ubuntu 7.4.0-1ubuntu1~18.04) 7.4.0 66 ubuntu:18.10 : Ok gcc (Ubuntu 8.2.0-7ubuntu1) 8.2.0 67 ubuntu:19.04 : Ok gcc (Ubuntu 8.3.0-6ubuntu1) 8.3.0 68 ubuntu:19.04-x-alpha : Ok alpha-linux-gnu-gcc (Ubuntu 8.3.0-6ubuntu1) 8.3.0 69 ubuntu:19.04-x-arm64 : Ok aarch64-linux-gnu-gcc (Ubuntu/Linaro 8.3.0-6ubuntu1) 8.3.0 70 ubuntu:19.04-x-hppa : Ok hppa-linux-gnu-gcc (Ubuntu 8.3.0-6ubuntu1) 8.3.0
The getname_flags related tests failing at the end (tests 65, 66 and 67) are being investigated, getname_flags() seems to have become just a tail call from getname(), something in this are changed and we're not anymore being able to add a probe at a suitable place to collect the just copied from userspace pathname.
# uname -a
Linux quaco 5.1.0-rc7+ #1 SMP Thu May 2 09:47:59 EDT 2019 x86_64 x86_64 x86_64 GNU/Linux
# git log --oneline -1
4fc4d8dfa056 perf stat: Support 'percore' event qualifier
# perf version --build-options
perf version 5.1.g4fc4d8
dwarf: [ on ] # HAVE_DWARF_SUPPORT
dwarf_getlocations: [ on ] # HAVE_DWARF_GETLOCATIONS_SUPPORT
glibc: [ on ] # HAVE_GLIBC_SUPPORT
gtk2: [ on ] # HAVE_GTK2_SUPPORT
syscall_table: [ on ] # HAVE_SYSCALL_TABLE_SUPPORT
libbfd: [ on ] # HAVE_LIBBFD_SUPPORT
libelf: [ on ] # HAVE_LIBELF_SUPPORT
libnuma: [ on ] # HAVE_LIBNUMA_SUPPORT
numa_num_possible_cpus: [ on ] # HAVE_LIBNUMA_SUPPORT
libperl: [ on ] # HAVE_LIBPERL_SUPPORT
libpython: [ on ] # HAVE_LIBPYTHON_SUPPORT
libslang: [ on ] # HAVE_SLANG_SUPPORT
libcrypto: [ on ] # HAVE_LIBCRYPTO_SUPPORT
libunwind: [ on ] # HAVE_LIBUNWIND_SUPPORT
libdw-dwarf-unwind: [ on ] # HAVE_DWARF_SUPPORT
zlib: [ on ] # HAVE_ZLIB_SUPPORT
lzma: [ on ] # HAVE_LZMA_SUPPORT
get_cpuid: [ on ] # HAVE_AUXTRACE_SUPPORT
bpf: [ on ] # HAVE_LIBBPF_SUPPORT
aio: [ on ] # HAVE_AIO_SUPPORT
zstd: [ on ] # HAVE_ZSTD_SUPPORT
# perf test
1: vmlinux symtab matches kallsyms : Ok
2: Detect openat syscall event : Ok
3: Detect openat syscall event on all cpus : Ok
4: Read samples using the mmap interface : Ok
5: Test data source output : Ok
6: Parse event definition strings : Ok
7: Simple expression parser : Ok
8: PERF_RECORD_* events & perf_sample fields : Ok
9: Parse perf pmu format : Ok
10: DSO data read : Ok
11: DSO data cache : Ok
12: DSO data reopen : Ok
13: Roundtrip evsel->name : Ok
14: Parse sched tracepoints fields : Ok
15: syscalls:sys_enter_openat event fields : Ok
16: Setup struct perf_event_attr : Ok
17: Match and link multiple hists : Ok
18: 'import perf' in python : Ok
19: Breakpoint overflow signal handler : Ok
20: Breakpoint overflow sampling : Ok
21: Breakpoint accounting : Ok
22: Watchpoint :
22.1: Read Only Watchpoint : Skip
22.2: Write Only Watchpoint : Ok
22.3: Read / Write Watchpoint : Ok
22.4: Modify Watchpoint : Ok
23: Number of exit events of a simple workload : Ok
24: Software clock events period values : Ok
25: Object code reading : Ok
26: Sample parsing : Ok
27: Use a dummy software event to keep tracking : Ok
28: Parse with no sample_id_all bit set : Ok
29: Filter hist entries : Ok
30: Lookup mmap thread : Ok
31: Share thread mg : Ok
32: Sort output of hist entries : Ok
33: Cumulate child hist entries : Ok
34: Track with sched_switch : Ok
35: Filter fds with revents mask in a fdarray : Ok
36: Add fd to a fdarray, making it autogrow : Ok
37: kmod_path__parse : Ok
38: Thread map : Ok
39: LLVM search and compile :
39.1: Basic BPF llvm compile : Ok
39.2: kbuild searching : Ok
39.3: Compile source for BPF prologue generation : Ok
39.4: Compile source for BPF relocation : Ok
40: Session topology : Ok
41: BPF filter :
41.1: Basic BPF filtering : Ok
41.2: BPF pinning : Ok
41.3: BPF prologue generation : Ok
41.4: BPF relocation checker : Ok
42: Synthesize thread map : Ok
43: Remove thread map : Ok
44: Synthesize cpu map : Ok
45: Synthesize stat config : Ok
46: Synthesize stat : Ok
47: Synthesize stat round : Ok
48: Synthesize attr update : Ok
49: Event times : Ok
50: Read backward ring buffer : Ok
51: Print cpu map : Ok
52: Probe SDT events : Ok
53: is_printable_array : Ok
54: Print bitmap : Ok
55: perf hooks : Ok
56: builtin clang support : Skip (not compiled in)
57: unit_number__scnprintf : Ok
58: mem2node : Ok
59: x86 rdpmc : Ok
60: Convert perf time to TSC : Ok
61: DWARF unwind : Ok
62: x86 instruction decoder - new instructions : Ok
63: x86 bp modify : Ok
64: probe libc's inet_pton & backtrace it with ping : Ok
65: Use vfs_getname probe to get syscall args filenames : FAILED!
66: Add vfs_getname probe to get syscall args filenames : FAILED!
67: Check open filename arg using perf trace + vfs_getname: FAILED!
68: Zstd perf.data compression/decompression : Ok $ time make -C tools/perf build-test
make: Entering directory '/home/acme/git/perf/tools/perf'
- tarpkg: ./tests/perf-targz-src-pkg .
make_doc_O: make doc
make_cscope_O: make cscope
make_no_newt_O: make NO_NEWT=1
make_with_clangllvm_O: make LIBCLANGLLVM=1
make_no_demangle_O: make NO_DEMANGLE=1
make_debug_O: make DEBUG=1
make_no_libelf_O: make NO_LIBELF=1
make_no_gtk2_O: make NO_GTK2=1
make_no_libunwind_O: make NO_LIBUNWIND=1
make_no_libpython_O: make NO_LIBPYTHON=1
make_perf_o_O: make perf.o
make_install_O: make install
make_pure_O: make
make_util_map_o_O: make util/map.o
make_no_scripts_O: make NO_LIBPYTHON=1 NO_LIBPERL=1
make_no_backtrace_O: make NO_BACKTRACE=1
make_static_O: make LDFLAGS=-static
make_install_prefix_slash_O: make install prefix=/tmp/krava/
make_no_libaudit_O: make NO_LIBAUDIT=1
make_no_libnuma_O: make NO_LIBNUMA=1
make_no_libdw_dwarf_unwind_O: make NO_LIBDW_DWARF_UNWIND=1
make_no_libbionic_O: make NO_LIBBIONIC=1
make_no_libperl_O: make NO_LIBPERL=1
make_clean_all_O: make clean all
make_util_pmu_bison_o_O: make util/pmu-bison.o
make_with_babeltrace_O: make LIBBABELTRACE=1
make_no_slang_O: make NO_SLANG=1
make_install_prefix_O: make install prefix=/tmp/krava
make_minimal_O: make NO_LIBPERL=1 NO_LIBPYTHON=1 NO_NEWT=1 NO_GTK2=1 NO_DEMANGLE=1 NO_LIBELF=1 NO_LIBUNWIND=1 NO_BACKTRACE=1 NO_LIBNUMA=1 NO_LIBAUDIT=1 NO_LIBBIONIC=1 NO_LIBDW_DWARF_UNWIND=1 NO_AUXTRACE=1 NO_LIBBPF=1 NO_LIBCRYPTO=1 NO_SDT=1 NO_JVMTI=1 NO_LIBZSTD=1
make_help_O: make help
make_no_libbpf_O: make NO_LIBBPF=1
make_install_bin_O: make install-bin
make_no_ui_O: make NO_NEWT=1 NO_SLANG=1 NO_GTK2=1
make_no_auxtrace_O: make NO_AUXTRACE=1
make_tags_O: make tags
OK
make: Leaving directory '/home/acme/git/perf/tools/perf'
$ Arnaldo Carvalho de Melo
maintainerPATCH 1/73 perf annotate: Remove hist__account_cycles() from callback
From: Jin Yao <[email protected]>
The hist__account_cycles() function is executed when the hist_iter__branch_callback() is called.
But it looks it's not necessary. In hist__account_cycles, it already walks on all branch entries.
This patch moves the hist__account_cycles out of callback, now the data processing is much faster than before.
Previous code has an issue that the ch[offset].num++ (in __symbol__account_cycles) is executed repeatedly since hist__account_cycles is called in each hist_iter__branch_callback, so the counting of ch[offset].num is not correct (too big).
With this patch, the issue is fixed. And we don't need the code of "ch->reset >= ch->num / 2" to check if there are too many overlaps (in annotation__count_and_fill), otherwise some data would be hidden.
Now, we can try, for example:
perf record -b ...
perf annotate or perf report -s symbol
The before/after output should be no change.
v3:
---
Fix the crash in stdio mode.
Like previous code, it needs the checking of ui__has_annotation() before hist__account_cycles()
v2:
---
1. Cover the similar perf report
2. Remove the checking code "ch->reset >= ch->num / 2"
--- tools/perf/builtin-annotate.c | 4 ++-- tools/perf/builtin-report.c | 11 +++++------ tools/perf/util/annotate.c | 2 +- 3 files changed, 8 insertions(+), 9 deletions(-)
tools/perf/builtin-annotate.c +2 -2
@@ -159,8 +159,6 @@ static int hist_iter__branch_callback(struct hist_entry_iter *iter,struct perf_evsel *evsel = iter->evsel;int err; -hist__account_cycles (sample->branch_stack, al, sample, false); - bi = he->branch_info; err =addr_map_symbol__inc_samples (&bi->from, sample, evsel);
@@ -199,6 +197,8 @@ static int process_branch_callback(struct perf_evsel *evsel,if (a.map !=NULL ) a.map->dso->hit =1 ; +hist__account_cycles (sample->branch_stack, al, sample, false); + ret =hist_entry_iter__add (&iter, &a,PERF_MAX_STACK_DEPTH , ann);return ret; }
tools/perf/builtin-report.c +5 -6
@@ -136,9 +136,6 @@ static int hist_iter__report_callback(struct hist_entry_iter *iter,if (!ui__has_annotation () && !rep->symbol_ipc)return 0 ; -hist__account_cycles (sample->branch_stack, al, sample, - rep->nonany_branch_mode); -if (sort__mode ==SORT_MODE__BRANCH ) { bi = he->branch_info; err =addr_map_symbol__inc_samples (&bi->from, sample, evsel);
@@ -181,9 +178,6 @@ static int hist_iter__branch_callback(struct hist_entry_iter *iter,if (!ui__has_annotation () && !rep->symbol_ipc)return 0 ; -hist__account_cycles (sample->branch_stack, al, sample, - rep->nonany_branch_mode); - bi = he->branch_info; err =addr_map_symbol__inc_samples (&bi->from, sample, evsel);if (err)
@@ -282,6 +276,11 @@ static int process_sample_event(struct perf_tool *tool,if (al.map !=NULL ) al.map->dso->hit =1 ; +if (ui__has_annotation () || rep->symbol_ipc) { +hist__account_cycles (sample->branch_stack, &al, sample, + rep->nonany_branch_mode); + } + ret =hist_entry_iter__add (&iter, &al, rep->max_stack, rep);if (ret <0 )pr_debug ("problem adding hist entry, skipping event\n" );
tools/perf/util/annotate.c +1 -1
@@ -1021,7 +1021,7 @@ static void annotation__count_and_fill(struct annotation *notes, u64 start, u64float ipc = n_insn / ((double )ch->cycles / (double )ch->num);/* Hide data when there are too many overlaps. */ -if (ch->reset >=0x7fff || ch->reset >= ch->num /2 ) +if (ch->reset >=0x7fff )return ;for (offset = start; offset <= end; offset++) {
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 2/73 perf test: Fix spelling mistake "leadking" -> "leaking"
In reply to Arnaldo Carvalho de MeloFrom: Colin Ian King <[email protected]>
There are a couple of spelling mistakes in test assert messages. Fix them.
--- tools/perf/tests/dso-data.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
tools/perf/tests/dso-data.c +2 -2
@@ -304,7 +304,7 @@ int test__dso_data_cache(struct test *test __maybe_unused, int subtest __maybe_u/* Make sure we did not leak any file descriptor. */ nr_end =open_files_cnt ();pr_debug ("nr start %ld, nr stop %ld\n" , nr, nr_end); -TEST_ASSERT_VAL ("failed leadking files" , nr = nr_end); +TEST_ASSERT_VAL ("failed leaking files" , nr = nr_end);return 0 ; }
@@ -380,6 +380,6 @@ int test__dso_data_reopen(struct test *test __maybe_unused, int subtest __maybe_/* Make sure we did not leak any file descriptor. */ nr_end =open_files_cnt ();pr_debug ("nr start %ld, nr stop %ld\n" , nr, nr_end); -TEST_ASSERT_VAL ("failed leadking files" , nr = nr_end); +TEST_ASSERT_VAL ("failed leaking files" , nr = nr_end);return 0 ; }
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 3/73 csky: Add support for libdw
In reply to Arnaldo Carvalho de MeloFrom: Mao Han <[email protected]>
This patch add support for DWARF register mappings and libdw registers initialization, which is used by perf callchain analyzing when --call-graph=dwarf is given.
Here is the elfutils csky backend patch set:
https://sourceware.org/ml/elfutils-devel/2019-q2/msg00007.html
--- tools/arch/csky/include/uapi/asm/perf_regs.h | 51 ++++++++++ tools/perf/Makefile.config | 6 +- tools/perf/arch/csky/Build | 1 + tools/perf/arch/csky/Makefile | 3 + tools/perf/arch/csky/include/perf_regs.h | 100 +++++++++++++++++++ tools/perf/arch/csky/util/Build | 2 + tools/perf/arch/csky/util/dwarf-regs.c | 49 +++++++++ tools/perf/arch/csky/util/unwind-libdw.c | 77 ++++++++++++++ 8 files changed, 288 insertions(+), 1 deletion(-) create mode 100644 tools/arch/csky/include/uapi/asm/perf_regs.h create mode 100644 tools/perf/arch/csky/Build create mode 100644 tools/perf/arch/csky/Makefile create mode 100644 tools/perf/arch/csky/include/perf_regs.h create mode 100644 tools/perf/arch/csky/util/Build create mode 100644 tools/perf/arch/csky/util/dwarf-regs.c create mode 100644 tools/perf/arch/csky/util/unwind-libdw.c
tools/arch/csky/include/uapi/asm/perf_regs.h new file +51 -0
@@ -0,0 +1,51 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +// Copyright (C) 2019 Hangzhou C-SKY Microsystems co.,ltd. + +#ifndef _ASM_CSKY_PERF_REGS_H +#define _ASM_CSKY_PERF_REGS_H + +/* Index of struct pt_regs */ +enum perf_event_csky_regs { +PERF_REG_CSKY_TLS , +PERF_REG_CSKY_LR , +PERF_REG_CSKY_PC , +PERF_REG_CSKY_SR , +PERF_REG_CSKY_SP , +PERF_REG_CSKY_ORIG_A0 , +PERF_REG_CSKY_A0 , +PERF_REG_CSKY_A1 , +PERF_REG_CSKY_A2 , +PERF_REG_CSKY_A3 , +PERF_REG_CSKY_REGS0 , +PERF_REG_CSKY_REGS1 , +PERF_REG_CSKY_REGS2 , +PERF_REG_CSKY_REGS3 , +PERF_REG_CSKY_REGS4 , +PERF_REG_CSKY_REGS5 , +PERF_REG_CSKY_REGS6 , +PERF_REG_CSKY_REGS7 , +PERF_REG_CSKY_REGS8 , +PERF_REG_CSKY_REGS9 , +#if defined(__CSKYABIV2__) +PERF_REG_CSKY_EXREGS0 , +PERF_REG_CSKY_EXREGS1 , +PERF_REG_CSKY_EXREGS2 , +PERF_REG_CSKY_EXREGS3 , +PERF_REG_CSKY_EXREGS4 , +PERF_REG_CSKY_EXREGS5 , +PERF_REG_CSKY_EXREGS6 , +PERF_REG_CSKY_EXREGS7 , +PERF_REG_CSKY_EXREGS8 , +PERF_REG_CSKY_EXREGS9 , +PERF_REG_CSKY_EXREGS10 , +PERF_REG_CSKY_EXREGS11 , +PERF_REG_CSKY_EXREGS12 , +PERF_REG_CSKY_EXREGS13 , +PERF_REG_CSKY_EXREGS14 , +PERF_REG_CSKY_HI , +PERF_REG_CSKY_LO , +PERF_REG_CSKY_DCSR , +#endif +PERF_REG_CSKY_MAX , +}; +#endif /* _ASM_CSKY_PERF_REGS_H */
tools/perf/Makefile.config +5 -1
@@ -59,6 +59,10 @@ ifeq ($(SRCARCH),arm64) LIBUNWIND_LIBS = -lunwind -lunwind-aarch64 endif +ifeq ($(SRCARCH),csky) + NO_PERF_REGS := 0 +endif + ifeq ($(ARCH),s390) NO_PERF_REGS := 0 NO_SYSCALL_TABLE := 0
@@ -77,7 +81,7 @@ endif # Disable it on all other architectures in case libdw unwind # support is detected in system. Add supported architectures # to the check. -ifneq ($(SRCARCH),$(filter $(SRCARCH),x86 arm arm64 powerpc s390)) +ifneq ($(SRCARCH),$(filter $(SRCARCH),x86 arm arm64 powerpc s390 csky)) NO_LIBDW_DWARF_UNWIND := 1 endif
tools/perf/arch/csky/Build new file +1 -0
@@ -0,0 +1 @@ +perf-y += util/
tools/perf/arch/csky/Makefile new file +3 -0
@@ -0,0 +1,3 @@ +ifndef NO_DWARF +PERF_HAVE_DWARF_REGS := 1 +endif
tools/perf/arch/csky/include/perf_regs.h new file +100 -0
100 lines not on this page · raw message
tools/perf/arch/csky/util/Build new file +2 -0
@@ -0,0 +1,2 @@ +perf-$(CONFIG_DWARF) += dwarf-regs.o +perf-$(CONFIG_LIBDW_DWARF_UNWIND) += unwind-libdw.o
tools/perf/arch/csky/util/dwarf-regs.c new file +49 -0
49 lines not on this page · raw message
tools/perf/arch/csky/util/unwind-libdw.c new file +77 -0
77 lines not on this page · raw message
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 4/73 tools lib traceevent: Remove hard coded install paths from pkg-config file
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Install directories of header and library files are hard coded in pkg-config template file.
They must be configurable, the Makefile should set them on the compilation / install stage.
--- tools/lib/traceevent/Makefile | 13 +++++++++---- tools/lib/traceevent/libtraceevent.pc.template | 4 ++-- 2 files changed, 11 insertions(+), 6 deletions(-)
tools/lib/traceevent/Makefile +9 -4
@@ -50,6 +50,9 @@ man_dir = $(prefix)/share/man man_dir_SQ = '$(subst ','\'',$(man_dir))' pkgconfig_dir ?= $(word 1,$(shell $(PKG_CONFIG) \ --variable pc_path pkg-config | tr ":" " ")) +includedir_relative = traceevent +includedir = $(prefix)/include/$(includedir_relative) +includedir_SQ = '$(subst ','\'',$(includedir))' export man_dir man_dir_SQ INSTALL export DESTDIR DESTDIR_SQ
@@ -279,6 +282,8 @@ define do_install_pkgconfig_file cp -f ${PKG_CONFIG_FILE}.template ${PKG_CONFIG_FILE}; \ sed -i "s|INSTALL_PREFIX|${1}|g" ${PKG_CONFIG_FILE}; \ sed -i "s|LIB_VERSION|${EVENT_PARSE_VERSION}|g" ${PKG_CONFIG_FILE}; \ + sed -i "s|LIB_DIR|${libdir}|g" ${PKG_CONFIG_FILE}; \ + sed -i "s|HEADER_DIR|$(includedir)|g" ${PKG_CONFIG_FILE}; \ $(call do_install,$(PKG_CONFIG_FILE),$(pkgconfig_dir),644); \ else \ (echo Failed to locate pkg-config directory) 1>&2; \
@@ -300,10 +305,10 @@ install_pkgconfig: install_headers: $(call QUIET_INSTALL, headers) \ - $(call do_install,event-parse.h,$(prefix)/include/traceevent,644); \ - $(call do_install,event-utils.h,$(prefix)/include/traceevent,644); \ - $(call do_install,trace-seq.h,$(prefix)/include/traceevent,644); \ - $(call do_install,kbuffer.h,$(prefix)/include/traceevent,644) + $(call do_install,event-parse.h,$(DESTDIR)$(includedir_SQ),644); \ + $(call do_install,event-utils.h,$(DESTDIR)$(includedir_SQ),644); \ + $(call do_install,trace-seq.h,$(DESTDIR)$(includedir_SQ),644); \ + $(call do_install,kbuffer.h,$(DESTDIR)$(includedir_SQ),644) install: install_lib
tools/lib/traceevent/libtraceevent.pc.template +2 -2
@@ -1,6 +1,6 @@ prefix=INSTALL_PREFIX -libdir=${prefix}/lib64 -includedir=${prefix}/include/traceevent +libdir=LIB_DIR +includedir=HEADER_DIR Name: libtraceevent URL: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 5/73 perf tools: Speed up report for perf compiled with linwunwind
In reply to Arnaldo Carvalho de MeloFrom: Jiri Olsa <[email protected]>
When compiled with libunwind, perf does some preparatory work when processing side-band events. This is not needed when report actually don't unwind dwarf callchains, so it's disabled with dwarf_callchain_users bool.
However we could move that check to higher level and shield more unwanted code for normal report processing, giving us following speed up on kernel build profile:
Before:
$ perf record make -j40 ... $ ll ../../perf.data -rw-------. 1 jolsa jolsa 461783932 Apr 26 09:11 perf.data $ perf stat -e cycles:u,instructions:u perf report -i perf.data > out
Performance counter stats for 'perf report -i perf.data':
78,669,920,155 cycles:u
99,076,431,951 instructions:u # 1.26 insn per cycle
55.382823668 seconds time elapsed
27.512341000 seconds user
27.712871000 seconds sysAfter:
$ perf stat -e cycles:u,instructions:u perf report -i perf.data > out
Performance counter stats for 'perf report -i perf.data':
59,626,798,904 cycles:u
88,583,575,849 instructions:u # 1.49 insn per cycle
21.296935559 seconds time elapsed
20.010191000 seconds user
1.202935000 seconds sysThe speed is higher with profile having many side-band events, because these trigger libunwind preparatory code.
This does not apply for perf compiled with libdw for dwarf unwind, only for build with libunwind.
--- tools/perf/util/thread.c | 3 ++- tools/perf/util/unwind-libunwind-local.c | 6 ------ tools/perf/util/unwind-libunwind.c | 10 ++++++++++ 3 files changed, 12 insertions(+), 7 deletions(-)
tools/perf/util/thread.c +2 -1
@@ -15,6 +15,7 @@#include "map.h" #include "symbol.h" #include "unwind.h" +#include "callchain.h" #include <api/fs/fs.h>
@@ -327,7 +328,7 @@ static int thread__prepare_access(struct thread *thread) {int err =0 ; -if (symbol_conf.use_callchain) +if (dwarf_callchain_users) err =__thread__prepare_access (thread);return err;
tools/perf/util/unwind-libunwind-local.c +0 -6
@@ -617,8 +617,6 @@ static unw_accessors_t accessors = {static int _unwind__prepare_access (struct thread *thread) { -if (!dwarf_callchain_users) -return 0 ; thread->addr_space =unw_create_addr_space (&accessors,0 );if (!thread->addr_space) {pr_err ("unwind: Can't create unwind address space.\n" );
@@ -631,15 +629,11 @@ static int _unwind__prepare_access(struct thread *thread)static void _unwind__flush_access (struct thread *thread) { -if (!dwarf_callchain_users) -return ;unw_flush_cache (thread->addr_space,0 ,0 ); }static void _unwind__finish_access (struct thread *thread) { -if (!dwarf_callchain_users) -return ;unw_destroy_addr_space (thread->addr_space); }
tools/perf/util/unwind-libunwind.c +10 -0
@@ -5,6 +5,7 @@#include "session.h" #include "debug.h" #include "env.h" +#include "callchain.h" struct unwind_libunwind_ops __weak *local_unwind_libunwind_ops;struct unwind_libunwind_ops __weak *x86_32_unwind_libunwind_ops;
@@ -24,6 +25,9 @@ int unwind__prepare_access(struct thread *thread, struct map *map,struct unwind_libunwind_ops *ops = local_unwind_libunwind_ops;int err; +if (!dwarf_callchain_users) +return 0 ; +if (thread->addr_space) {pr_debug ("unwind: thread map already set, dso=%s\n" , map->dso->name);
@@ -65,12 +69,18 @@ int unwind__prepare_access(struct thread *thread, struct map *map,void unwind__flush_access (struct thread *thread) { +if (!dwarf_callchain_users) +return ; +if (thread->unwind_libunwind_ops) thread->unwind_libunwind_ops->flush_access (thread); }void unwind__finish_access (struct thread *thread) { +if (!dwarf_callchain_users) +return ; +if (thread->unwind_libunwind_ops) thread->unwind_libunwind_ops->finish_access (thread); }
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 6/73 tools arch: Update arch/x86/lib/memcpy_64.S copy used in 'perf bench mem memcpy'
In reply to Arnaldo Carvalho de MeloFrom: Arnaldo Carvalho de Melo <[email protected]>
To bring in the change made in this cset:
b69656fa7ea2 ("x86/uaccess: Fix up the fixup")
Silencing this perf build warning:
Warning: Kernel ABI header at 'tools/arch/x86/lib/memcpy_64.S' differs from latest version at 'arch/x86/lib/memcpy_64.S' diff -u tools/arch/x86/lib/memcpy_64.S arch/x86/lib/memcpy_64.S
No changes in the tooling using this, that was just to ease some objtool return checking.
--- tools/arch/x86/lib/memcpy_64.S | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
tools/arch/x86/lib/memcpy_64.S +2 -1
@@ -257,6 +257,7 @@ ENTRY(__memcpy_mcsafe) /* Copy successful. Return zero */ .L_done_memcpy_trap: xorl %eax, %eax +.L_done: ret ENDPROC(__memcpy_mcsafe) EXPORT_SYMBOL_GPL(__memcpy_mcsafe)
@@ -273,7 +274,7 @@ EXPORT_SYMBOL_GPL(__memcpy_mcsafe) addl %edx, %ecx .E_trailing_bytes: mov %ecx, %eax - ret + jmp .L_done /* * For write fault handling, given the destination is unaligned,
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 7/73 tools arch uapi: Sync the x86 kvm.h copy
In reply to Arnaldo Carvalho de MeloFrom: Arnaldo Carvalho de Melo <[email protected]>
To get the changes in:
59073aaf6de0 ("kvm: x86: Add exception payload fields to kvm_vcpu_events")
This silences the following perf build warning:
Warning: Kernel ABI header at 'tools/arch/x86/include/uapi/asm/kvm.h' differs from latest version at 'arch/x86/include/uapi/asm/kvm.h' diff -u tools/arch/x86/include/uapi/asm/kvm.h arch/x86/include/uapi/asm/kvm.h
The changes in this file are in something not used at this time in any tools/perf/ tool.
--- tools/arch/x86/include/uapi/asm/kvm.h | 1 + 1 file changed, 1 insertion(+)
tools/arch/x86/include/uapi/asm/kvm.h +1 -0
@@ -381,6 +381,7 @@ struct kvm_sync_regs {#define KVM_X86_QUIRK_LINT0_REENABLED (1 << 0)#define KVM_X86_QUIRK_CD_NW_CLEARED (1 << 1)#define KVM_X86_QUIRK_LAPIC_MMIO_HOLE (1 << 2) +#define KVM_X86_QUIRK_OUT_7E_INC_RIP (1 << 3)#define KVM_STATE_NESTED_GUEST_MODE 0x00000001#define KVM_STATE_NESTED_RUN_PENDING 0x00000002
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 8/73 tools x86 uapi asm: Sync the pt_regs.h copy with the kernel sources
In reply to Arnaldo Carvalho de MeloFrom: Arnaldo Carvalho de Melo <[email protected]>
To get the changes in:
878068ea270e ("perf/x86: Support outputting XMM registers")
That will be used in a followup patch to allow users to ask for some or all of those registers to be collected in certain contatexts.
This silences the following perf build warning:
Warning: Kernel ABI header at 'tools/arch/x86/include/uapi/asm/perf_regs.h' differs from latest version at 'arch/x86/include/uapi/asm/perf_regs.h' diff -u tools/arch/x86/include/uapi/asm/perf_regs.h arch/x86/include/uapi/asm/perf_regs.h
--- tools/arch/x86/include/uapi/asm/perf_regs.h | 23 ++++++++++++++++++++- 1 file changed, 22 insertions(+), 1 deletion(-)
tools/arch/x86/include/uapi/asm/perf_regs.h +22 -1
@@ -27,8 +27,29 @@ enum perf_event_x86_regs {PERF_REG_X86_R13 ,PERF_REG_X86_R14 ,PERF_REG_X86_R15 , - +/* These are the limits for the GPRs. */ PERF_REG_X86_32_MAX =PERF_REG_X86_GS +1 ,PERF_REG_X86_64_MAX =PERF_REG_X86_R15 +1 , + +/* These all need two bits set because they are 128bit */ +PERF_REG_X86_XMM0 =32 , +PERF_REG_X86_XMM1 =34 , +PERF_REG_X86_XMM2 =36 , +PERF_REG_X86_XMM3 =38 , +PERF_REG_X86_XMM4 =40 , +PERF_REG_X86_XMM5 =42 , +PERF_REG_X86_XMM6 =44 , +PERF_REG_X86_XMM7 =46 , +PERF_REG_X86_XMM8 =48 , +PERF_REG_X86_XMM9 =50 , +PERF_REG_X86_XMM10 =52 , +PERF_REG_X86_XMM11 =54 , +PERF_REG_X86_XMM12 =56 , +PERF_REG_X86_XMM13 =58 , +PERF_REG_X86_XMM14 =60 , +PERF_REG_X86_XMM15 =62 , + +/* These include both GPRs and XMMX registers */ + PERF_REG_X86_XMM_MAX = PERF_REG_X86_XMM15 + 2, };#endif /* _ASM_X86_PERF_REGS_H */
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 9/73 tools pci: Do not delete pcitest.sh in 'make clean'
In reply to Arnaldo Carvalho de MeloFrom: Arnaldo Carvalho de Melo <[email protected]>
When running 'make -C tools clean' I noticed that a revision controlled file was being deleted:
$ git diff diff --git a/tools/pci/pcitest.sh b/tools/pci/pcitest.sh deleted file mode 100644 index 75ed48ff2990..000000000000 --- a/tools/pci/pcitest.sh +++ /dev/null @@ -1,72 +0,0 @@ -#!/bin/sh -# SPDX-License-Identifier: GPL-2.0 - -echo "BAR tests" -echo <SNIP>
So I changed the make variables to fix that, testing it should produce the same intended result while not deleting revision controlled files.
$ make O=/tmp/build/pci -C tools/pci install make: Entering directory '/home/acme/git/perf/tools/pci' make -f /home/acme/git/perf/tools/build/Makefile.build dir=. obj=pcitest install -d -m 755 /usr/bin; \ for program in /tmp/build/pci/pcitest pcitest.sh; do \ install $program /usr/bin; \ done install: cannot change permissions of ‘/usr/bin’: Operation not permitted install: cannot create regular file '/usr/bin/pcitest': Permission denied install: cannot create regular file '/usr/bin/pcitest.sh': Permission denied make: *** [Makefile:46: install] Error 1 make: Leaving directory '/home/acme/git/perf/tools/pci' $ ls -la /tmp/build/pci/pcitest -rwxrwxr-x. 1 acme acme 27152 May 13 13:52 /tmp/build/pci/pcitest $ /tmp/build/pci/pcitest can't open PCI Endpoint Test device: No such file or directory $
--- tools/pci/Makefile | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
tools/pci/Makefile +2 -2
@@ -14,7 +14,7 @@ MAKEFLAGS += -r CFLAGS += -O2 -Wall -g -D_GNU_SOURCE -I$(OUTPUT)include -ALL_TARGETS := pcitest pcitest.sh +ALL_TARGETS := pcitest ALL_PROGRAMS := $(patsubst %,$(OUTPUT)%,$(ALL_TARGETS)) all: $(ALL_PROGRAMS)
@@ -44,7 +44,7 @@ clean: install: $(ALL_PROGRAMS) install -d -m 755 $(DESTDIR)$(bindir); \ - for program in $(ALL_PROGRAMS); do \ + for program in $(ALL_PROGRAMS) pcitest.sh; do \ install $$program $(DESTDIR)$(bindir); \ done
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 10/73 perf record: Fix suggestion to get list of registers usable with --user-regs and --intr-regs
In reply to Arnaldo Carvalho de MeloFrom: Arnaldo Carvalho de Melo <[email protected]>
$ perf record -h -I
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-I, --intr-regs[=<any register>]
sample selected machine registers on interrupt, use -I ? to list register names
$ m $ perf record -I ? Workload failed: No such file or directory $
After:
$ perf record -h -I
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-I, --intr-regs[=<any register>]
sample selected machine registers on interrupt, use '-I?' to list register names
$
$ perf record -I? available registers: AX BX CX DX SI DI BP SP IP FLAGS CS SS R8 R9 R10 R11 R12 R13 R14 R15
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-I, --intr-regs[=<any register>]
sample selected machine registers on interrupt, use '-I?' to list register names
$
--- tools/perf/builtin-record.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
tools/perf/builtin-record.c +2 -2
@@ -2017,10 +2017,10 @@ static struct option __record_options[] = { "use per-thread mmaps"),OPT_CALLBACK_OPTARG ('I' ,"intr-regs" , &record.opts.sample_intr_regs,NULL ,"any register" ,"sample selected machine registers on interrupt," -" use -I ? to list register names" , parse_regs), +" use '-I?' to list register names" , parse_regs),OPT_CALLBACK_OPTARG (0 ,"user-regs" , &record.opts.sample_user_regs,NULL ,"any register" ,"sample selected machine registers on interrupt," -" use -I ? to list register names" , parse_regs), +" use '-I?' to list register names" , parse_regs),OPT_BOOLEAN (0 ,"running-time" , &record.opts.running_time,"Record running/enabled time of read (:S) events" ), OPT_CALLBACK('k', "clockid", &record.opts,
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 11/73 perf parse-regs: Improve error output when faced with unknown register name
In reply to Arnaldo Carvalho de MeloFrom: Arnaldo Carvalho de Melo <[email protected]>
Add quotes around the register name and suggest using 'perf record -I?' to get the list of available registers.
Before:
# perf record -Idi,xmm20,xmm1 Warning: unknown register xmm20, check man page
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-I, --intr-regs[=<any register>]
sample selected machine registers on interrupt, use -I ? to list register names
#
# perf record -Idi,xmm20,xmm1 Warning: unknown register "xmm20", check man page or run "perf record -I?"
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-I, --intr-regs[=<any register>]
sample selected machine registers on interrupt, use -I ? to list register names
#
--- tools/perf/util/parse-regs-options.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)
tools/perf/util/parse-regs-options.c +1 -2
@@ -48,8 +48,7 @@ parse_regs(const struct option *opt, const char *str, int unset)break ; }if (!r->name) { -ui__warning ("unknown register %s," -" check man page\n" , s); +ui__warning ("Unknown register \"%s\", check man page or run \"perf record -I?\"\n" , s); goto error; }
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 12/73 perf tools x86: Add support for recording and printing XMM registers
In reply to Arnaldo Carvalho de MeloFrom: Andi Kleen <[email protected]>
Icelake and later platforms support collecting XMM registers with PEBS event.
Add support for 'perf script' to dump them, and support for the register parser in 'perf record -I=' ... to configure them.
For now they are just printed in hex, we could potentially later add other formats too.
Committer testing:
Before:
# perf record -IXMM0 Warning: unknown register XMM0, check man page or run 'perf record -I?'
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
#
# perf record -I? available registers: AX BX CX DX SI DI BP SP IP FLAGS CS SS R8 R9 R10 R11 R12 R13 R14 R15
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
#
After:
# perf record -IXMM0 Error: The sys_perf_event_open() syscall returned with 22 (Invalid argument) for event (cycles). /bin/dmesg | grep -i perf may provide additional information.
#
# perf record -I? available registers: AX BX CX DX SI DI BP SP IP FLAGS CS SS R8 R9 R10 R11 R12 R13 R14 R15 XMM0 XMM1 XMM2 XMM3 XMM4 XMM5 XMM6 XMM7 XMM8 XMM9 XMM10 XMM11 XMM12 XMM13 XMM14 XMM15
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-I, --intr-regs[=<any register>]
sample selected machine registers on interrupt, use -I ? to list register names
#
More work is needed to, when faced with such error, warn the user that that register is not available on the running platform.
--- tools/perf/arch/x86/include/perf_regs.h | 25 +++++++++++++++++++++++-- tools/perf/arch/x86/util/perf_regs.c | 16 ++++++++++++++++ tools/perf/util/perf_regs.h | 1 + 3 files changed, 40 insertions(+), 2 deletions(-)
tools/perf/arch/x86/include/perf_regs.h +23 -2
@@ -8,9 +8,9 @@void perf_regs_load (u64 *regs); +#define PERF_REGS_MAX PERF_REG_X86_XMM_MAX#ifndef HAVE_ARCH_X86_64_SUPPORT #define PERF_REGS_MASK ((1ULL << PERF_REG_X86_32_MAX) - 1) -#define PERF_REGS_MAX PERF_REG_X86_32_MAX#define PERF_SAMPLE_REGS_ABI PERF_SAMPLE_REGS_ABI_32#else #define REG_NOSUPPORT ((1ULL << PERF_REG_X86_DS) | \
@@ -18,7 +18,6 @@ void perf_regs_load(u64 *regs); (1ULL << PERF_REG_X86_FS) | \ (1ULL << PERF_REG_X86_GS))#define PERF_REGS_MASK (((1ULL << PERF_REG_X86_64_MAX) - 1) & ~REG_NOSUPPORT) -#define PERF_REGS_MAX PERF_REG_X86_64_MAX#define PERF_SAMPLE_REGS_ABI PERF_SAMPLE_REGS_ABI_64#endif #define PERF_REG_IP PERF_REG_X86_IP
@@ -77,6 +76,28 @@ static inline const char *perf_reg_name(int id)case PERF_REG_X86_R15 :return "R15" ;#endif /* HAVE_ARCH_X86_64_SUPPORT */ + +#define XMM (x) \ + case PERF_REG_X86_XMM ## x: \ + case PERF_REG_X86_XMM ## x + 1: \ + return "XMM" #x; + XMM(0) + XMM(1) + XMM(2) + XMM(3) + XMM(4) + XMM(5) + XMM(6) + XMM(7) + XMM(8) + XMM(9) + XMM(10) + XMM(11) + XMM(12) + XMM(13) + XMM(14) + XMM(15) +#undef XMMdefault :return NULL ; }
tools/perf/arch/x86/util/perf_regs.c +16 -0
@@ -31,6 +31,22 @@ const struct sample_reg sample_reg_masks[] = {SMPL_REG (R14 ,PERF_REG_X86_R14 ),SMPL_REG (R15 ,PERF_REG_X86_R15 ), #endif +SMPL_REG2 (XMM0 ,PERF_REG_X86_XMM0 ), +SMPL_REG2 (XMM1 ,PERF_REG_X86_XMM1 ), +SMPL_REG2 (XMM2 ,PERF_REG_X86_XMM2 ), +SMPL_REG2 (XMM3 ,PERF_REG_X86_XMM3 ), +SMPL_REG2 (XMM4 ,PERF_REG_X86_XMM4 ), +SMPL_REG2 (XMM5 ,PERF_REG_X86_XMM5 ), +SMPL_REG2 (XMM6 ,PERF_REG_X86_XMM6 ), +SMPL_REG2 (XMM7 ,PERF_REG_X86_XMM7 ), +SMPL_REG2 (XMM8 ,PERF_REG_X86_XMM8 ), +SMPL_REG2 (XMM9 ,PERF_REG_X86_XMM9 ), +SMPL_REG2 (XMM10 ,PERF_REG_X86_XMM10 ), +SMPL_REG2 (XMM11 ,PERF_REG_X86_XMM11 ), +SMPL_REG2 (XMM12 ,PERF_REG_X86_XMM12 ), +SMPL_REG2 (XMM13 ,PERF_REG_X86_XMM13 ), +SMPL_REG2 (XMM14 ,PERF_REG_X86_XMM14 ), +SMPL_REG2 (XMM15 ,PERF_REG_X86_XMM15 ),SMPL_REG_END };
tools/perf/util/perf_regs.h +1 -0
@@ -12,6 +12,7 @@ struct sample_reg {uint64_t mask; };#define SMPL_REG (n, b) { .name = #n, .mask = 1ULL << (b) } +#define SMPL_REG2 (n, b) { .name = #n, .mask = 3ULL << (b) }#define SMPL_REG_END { .name = NULL }extern const struct sample_reg sample_reg_masks[];
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 13/73 perf scripts python: exported-sql-viewer.py: Move view creation
In reply to Arnaldo Carvalho de MeloFrom: Adrian Hunter <[email protected]>
As preparation for adding support for copying to clipboard, create view in TreeWindowBase instead of derived classes.
Committer testing:
Tested using an old .db used to test some older patches:
$ python ~acme/libexec/perf-core/scripts/python/exported-sql-viewer.py ~/c/adrian.hunter/simple-retpoline.db
Nothing breaks.
--- tools/perf/scripts/python/exported-sql-viewer.py | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-)
tools/perf/scripts/python/exported-sql-viewer.py +2 -3
@@ -877,9 +877,10 @@ class TreeWindowBase(QMdiSubWindow): super(TreeWindowBase, self).__init__(parent) self.model = None - self.view = None self.find_bar = None + self.view = QTreeView() + def DisplayFound(self, ids): if not len(ids): return False
@@ -921,7 +922,6 @@ class CallGraphWindow(TreeWindowBase): self.model = LookupCreateModel("Context-Sensitive Call Graph", lambda x=glb: CallGraphModel(x)) - self.view = QTreeView() self.view.setModel(self.model) for c, w in ((0, 250), (1, 100), (2, 60), (3, 70), (4, 70), (5, 100)):
@@ -944,7 +944,6 @@ class CallTreeWindow(TreeWindowBase): self.model = LookupCreateModel("Call Tree", lambda x=glb: CallTreeModel(x)) - self.view = QTreeView() self.view.setModel(self.model) for c, w in ((0, 230), (1, 100), (2, 100), (3, 70), (4, 70), (5, 100)):
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 14/73 perf scripts python: exported-sql-viewer.py: Fix error when shrinking / enlarging font
In reply to Arnaldo Carvalho de MeloFrom: Adrian Hunter <[email protected]>
Fix the following error if shrink / enlarge font is used with the help window.
Traceback (most recent call last):
File "tools/perf/scripts/python/exported-sql-viewer.py", line 2791, in ShrinkFont
ShrinkFont(win.view)
AttributeError: 'HelpWindow' object has no attribute 'view'
Committer testing:
Before, matches above output:
$ python ~acme/libexec/perf-core/scripts/python/exported-sql-viewer.py ~/c/adrian.hunter/simple-retpoline.db
Traceback (most recent call last):
File "/home/acme/libexec/perf-core/scripts/python/exported-sql-viewer.py", line 2780, in EnlargeFont
EnlargeFont(win.view)
AttributeError: 'HelpWindow' object has no attribute 'view'
$After:
No more tracebacks, but the fonts don't get enlarged, which is kinda frustrating...
--- tools/perf/scripts/python/exported-sql-viewer.py | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-)
tools/perf/scripts/python/exported-sql-viewer.py +10 -4
@@ -2755,6 +2755,14 @@ class MainWindow(QMainWindow): help_menu = menu.addMenu("&Help") help_menu.addAction(CreateAction("&Exported SQL Viewer Help", "Helpful information", self.Help, self, QKeySequence.HelpContents)) + def Try(self, fn): + win = self.mdi_area.activeSubWindow() + if win: + try: + fn(win.view) + except: + pass + def Find(self): win = self.mdi_area.activeSubWindow() if win:
@@ -2772,12 +2780,10 @@ class MainWindow(QMainWindow): pass def ShrinkFont(self): - win = self.mdi_area.activeSubWindow() - ShrinkFont(win.view) + self.Try(ShrinkFont) def EnlargeFont(self): - win = self.mdi_area.activeSubWindow() - EnlargeFont(win.view) + self.Try(EnlargeFont) def EventMenu(self, events, reports_menu): branches_events = 0
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 15/73 perf scripts python: exported-sql-viewer.py: Add tree level
In reply to Arnaldo Carvalho de MeloFrom: Adrian Hunter <[email protected]>
As preparation for adding support for copying to clipboard, keep track of what level each item is in tree items.
--- tools/perf/scripts/python/exported-sql-viewer.py | 4 ++++ 1 file changed, 4 insertions(+)
tools/perf/scripts/python/exported-sql-viewer.py +4 -0
@@ -456,6 +456,10 @@ class CallGraphLevelItemBase(object): self.query_done = False; self.child_count = 0 self.child_items = [] + if parent_item: + self.level = parent_item.level + 1 + else: + self.level = 0 def getChildItem(self, row): return self.child_items[row]
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 16/73 perf scripts python: exported-sql-viewer.py: Add copy to clipboard
In reply to Arnaldo Carvalho de MeloFrom: Adrian Hunter <[email protected]>
Add support for copying to clipboard. Two menu options are added to copy the selected rows / columns with normal spacing, or as comma-separated-values. In the case of trees, only entire rows can be copied.
Comitter testing:
$ python ~acme/libexec/perf-core/scripts/python/exported-sql-viewer.py ~/c/adrian.hunter/simple-retpoline.db
Select the lines, press control+C and on the same terminal, press control+shift+V and voilà:
Call Path Object Count Time (ns) Time (%) Branch Count Branch Count (%)
▼ 14503:14503
▼ _start ld-2.28.so 1 156267 100.0 10602 100.0
unknown unknown 1 2276 1.5 1 0.0
▼ _dl_start ld-2.28.so 1 137047 87.7 10088 95.2 ▶ unknown unknown 4 4127 3.0 4 0.0
_dl_setup_hash ld-2.28.so 1 0 0.0 1 0.0
▶ _dl_sysdep_start ld-2.28.so 1 131342 95.8 9981 98.9
▼ _dl_init ld-2.28.so 1 9142 5.9 326 3.1
▼ call_init.part.0 ld-2.28.so 3 9133 99.9 319 97.9
▶ _init libc-2.28.so 1 6877 75.3 110 34.5
▶ check_stdfiles_vtables libc-2.28.so 1 76 0.8 2 0.6
▶ init_cacheinfo libc-2.28.so 1 1991 21.8 197 61.8
▶ _start simple-retpoline 1 7457 4.8 182 1.7
--- .../scripts/python/exported-sql-viewer.py | 217 ++++++++++++++++++ 1 file changed, 217 insertions(+)
tools/perf/scripts/python/exported-sql-viewer.py +217 -0
@@ -884,6 +884,8 @@ class TreeWindowBase(QMdiSubWindow): self.find_bar = None self.view = QTreeView() + self.view.setSelectionMode(QAbstractItemView.ContiguousSelection) + self.view.CopyCellsToClipboard = CopyTreeCellsToClipboard def DisplayFound(self, ids): if not len(ids):
@@ -1652,6 +1654,8 @@ class BranchWindow(QMdiSubWindow): self.view = QTreeView() self.view.setUniformRowHeights(True) + self.view.setSelectionMode(QAbstractItemView.ContiguousSelection) + self.view.CopyCellsToClipboard = CopyTreeCellsToClipboard self.view.setModel(self.model) self.ResizeColumnsToContents()
@@ -2264,6 +2268,207 @@ class ResizeColumnsToContentsBase(QObject): self.data_model.rowsInserted.disconnect(self.UpdateColumnWidths) self.ResizeColumnsToContents() +# Convert value to CSV + +def ToCSValue(val): + if '"' in val: + val = val.replace('"', '""') + if "," in val or '"' in val: + val = '"' + val + '"' + return val + +# Key to sort table model indexes by row / column, assuming fewer than 1000 columns + +glb_max_cols = 1000 + +def RowColumnKey(a): + return a.row() * glb_max_cols + a.column() + +# Copy selected table cells to clipboard + +def CopyTableCellsToClipboard(view, as_csv=False, with_hdr=False): + indexes = sorted(view.selectedIndexes(), key=RowColumnKey) + idx_cnt = len(indexes) + if not idx_cnt: + return + if idx_cnt == 1: + with_hdr=False + min_row = indexes[0].row() + max_row = indexes[0].row() + min_col = indexes[0].column() + max_col = indexes[0].column() + for i in indexes: + min_row = min(min_row, i.row()) + max_row = max(max_row, i.row()) + min_col = min(min_col, i.column()) + max_col = max(max_col, i.column()) + if max_col > glb_max_cols: + raise RuntimeError("glb_max_cols is too low") + max_width = [0] * (1 + max_col - min_col) + for i in indexes: + c = i.column() - min_col + max_width[c] = max(max_width[c], len(str(i.data()))) + text = "" + pad = "" + sep = "" + if with_hdr: + model = indexes[0].model() + for col in range(min_col, max_col + 1): + val = model.headerData(col, Qt.Horizontal) + if as_csv: + text += sep + ToCSValue(val) + sep = "," + else: + c = col - min_col + max_width[c] = max(max_width[c], len(val)) + width = max_width[c] + align = model.headerData(col, Qt.Horizontal, Qt.TextAlignmentRole) + if align & Qt.AlignRight: + val = val.rjust(width) + text += pad + sep + val + pad = " " * (width - len(val)) + sep = " " + text += "\n" + pad = "" + sep = "" + last_row = min_row + for i in indexes: + if i.row() > last_row: + last_row = i.row() + text += "\n" + pad = "" + sep = "" + if as_csv: + text += sep + ToCSValue(str(i.data())) + sep = "," + else: + width = max_width[i.column() - min_col] + if i.data(Qt.TextAlignmentRole) & Qt.AlignRight: + val = str(i.data()).rjust(width) + else: + val = str(i.data()) + text += pad + sep + val + pad = " " * (width - len(val)) + sep = " " + QApplication.clipboard().setText(text) + +def CopyTreeCellsToClipboard(view, as_csv=False, with_hdr=False): + indexes = view.selectedIndexes() + if not len(indexes): + return + + selection = view.selectionModel() + + first = None + for i in indexes: + above = view.indexAbove(i) + if not selection.isSelected(above): + first = i + break + + if first is None: + raise RuntimeError("CopyTreeCellsToClipboard internal error") + + model = first.model() + row_cnt = 0 + col_cnt = model.columnCount(first) + max_width = [0] * col_cnt + + indent_sz = 2 + indent_str = " " * indent_sz + + expanded_mark_sz = 2 + if sys.version_info[0] == 3: + expanded_mark = "\u25BC " + not_expanded_mark = "\u25B6 " + else: + expanded_mark = unicode(chr(0xE2) + chr(0x96) + chr(0xBC) + " ", "utf-8") + not_expanded_mark = unicode(chr(0xE2) + chr(0x96) + chr(0xB6) + " ", "utf-8") + leaf_mark = " " + + if not as_csv: + pos = first + while True: + row_cnt += 1 + row = pos.row() + for c in range(col_cnt): + i = pos.sibling(row, c) + if c: + n = len(str(i.data())) + else: + n = len(str(i.data()).strip()) + n += (i.internalPointer().level - 1) * indent_sz + n += expanded_mark_sz + max_width[c] = max(max_width[c], n) + pos = view.indexBelow(pos) + if not selection.isSelected(pos): + break + + text = "" + pad = "" + sep = "" + if with_hdr: + for c in range(col_cnt): + val = model.headerData(c, Qt.Horizontal, Qt.DisplayRole).strip() + if as_csv: + text += sep + ToCSValue(val) + sep = "," + else: + max_width[c] = max(max_width[c], len(val)) + width = max_width[c] + align = model.headerData(c, Qt.Horizontal, Qt.TextAlignmentRole) + if align & Qt.AlignRight: + val = val.rjust(width) + text += pad + sep + val + pad = " " * (width - len(val)) + sep = " " + text += "\n" + pad = "" + sep = "" + + pos = first + while True: + row = pos.row() + for c in range(col_cnt): + i = pos.sibling(row, c) + val = str(i.data()) + if not c: + if model.hasChildren(i): + if view.isExpanded(i): + mark = expanded_mark + else: + mark = not_expanded_mark + else: + mark = leaf_mark + val = indent_str * (i.internalPointer().level - 1) + mark + val.strip() + if as_csv: + text += sep + ToCSValue(val) + sep = "," + else: + width = max_width[c] + if c and i.data(Qt.TextAlignmentRole) & Qt.AlignRight: + val = val.rjust(width) + text += pad + sep + val + pad = " " * (width - len(val)) + sep = " " + pos = view.indexBelow(pos) + if not selection.isSelected(pos): + break + text = text.rstrip() + "\n" + pad = "" + sep = "" + + QApplication.clipboard().setText(text) + +def CopyCellsToClipboard(view, as_csv=False, with_hdr=False): + view.CopyCellsToClipboard(view, as_csv, with_hdr) + +def CopyCellsToClipboardHdr(view): + CopyCellsToClipboard(view, False, True) + +def CopyCellsToClipboardCSV(view): + CopyCellsToClipboard(view, True, True) + # Table window class TableWindow(QMdiSubWindow, ResizeColumnsToContentsBase):
@@ -2282,6 +2487,8 @@ class TableWindow(QMdiSubWindow, ResizeColumnsToContentsBase): self.view.verticalHeader().setVisible(False) self.view.sortByColumn(-1, Qt.AscendingOrder) self.view.setSortingEnabled(True) + self.view.setSelectionMode(QAbstractItemView.ContiguousSelection) + self.view.CopyCellsToClipboard = CopyTableCellsToClipboard self.ResizeColumnsToContents()
@@ -2398,6 +2605,8 @@ class TopCallsWindow(QMdiSubWindow, ResizeColumnsToContentsBase): self.view.setModel(self.model) self.view.setEditTriggers(QAbstractItemView.NoEditTriggers) self.view.verticalHeader().setVisible(False) + self.view.setSelectionMode(QAbstractItemView.ContiguousSelection) + self.view.CopyCellsToClipboard = CopyTableCellsToClipboard self.ResizeColumnsToContents()
@@ -2735,6 +2944,8 @@ class MainWindow(QMainWindow): file_menu.addAction(CreateExitAction(glb.app, self)) edit_menu = menu.addMenu("&Edit") + edit_menu.addAction(CreateAction("&Copy", "Copy to clipboard", self.CopyToClipboard, self, QKeySequence.Copy)) + edit_menu.addAction(CreateAction("Copy as CS&V", "Copy to clipboard as CSV", self.CopyToClipboardCSV, self)) edit_menu.addAction(CreateAction("&Find...", "Find items", self.Find, self, QKeySequence.Find)) edit_menu.addAction(CreateAction("Fetch &more records...", "Fetch more records", self.FetchMoreRecords, self, [QKeySequence(Qt.Key_F8)])) edit_menu.addAction(CreateAction("&Shrink Font", "Make text smaller", self.ShrinkFont, self, [QKeySequence("Ctrl+-")]))
@@ -2767,6 +2978,12 @@ class MainWindow(QMainWindow): except: pass + def CopyToClipboard(self): + self.Try(CopyCellsToClipboardHdr) + + def CopyToClipboardCSV(self): + self.Try(CopyCellsToClipboardCSV) + def Find(self): win = self.mdi_area.activeSubWindow() if win:
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 17/73 perf scripts python: exported-sql-viewer.py: Add context menu
In reply to Arnaldo Carvalho de MeloFrom: Adrian Hunter <[email protected]>
Add a context menu (right-click) that provides options for copying to clipboard, including, for trees, the ability to copy only the cell under the mouse pointer.
Committer testing:
$ python ~acme/libexec/perf-core/scripts/python/exported-sql-viewer.py ~/c/adrian.hunter/simple-retpoline.db
Simply right click and pick "Copy selection", that at this point has just the first line, not expanded, then see what was copied by pressing shift+control+v on a terminal:
Call Path,Object,Count,Time (ns),Time (%),Branch Count,Branch Count (%) ▶ simple-retpolin,,,,,,
Ditto after expanding, i.e. the selection continues to be just one line:
Call Path Object Count Time (ns) Time (%) Branch Count Branch Count (%)
▼ simple-retpolin
Now select all the lines with the mouse and control+shift+v again:
Call Path Object Count Time (ns) Time (%) Branch Count Branch Count (%)
▼ 14503:14503
▼ _start ld-2.28.so 1 156267 100.0 10602 100.0
▶ unknown unknown 1 2276 1.5 1 0.0
▶ _dl_start ld-2.28.so 1 137047 87.7 10088 95.2
▶ _dl_init ld-2.28.so 1 9142 5.9 326 3.1
▼ _start simple-retpoline 1 7457 4.8 182 1.7
▶ unknown unknown 1 805 10.8 1 0.5
▶ __libc_start_main libc-2.28.so 1 6347 85.1 179 98.4--- .../scripts/python/exported-sql-viewer.py | 41 +++++++++++++++++++ 1 file changed, 41 insertions(+)
tools/perf/scripts/python/exported-sql-viewer.py +41 -0
@@ -887,6 +887,8 @@ class TreeWindowBase(QMdiSubWindow): self.view.setSelectionMode(QAbstractItemView.ContiguousSelection) self.view.CopyCellsToClipboard = CopyTreeCellsToClipboard + self.context_menu = TreeContextMenu(self.view) + def DisplayFound(self, ids): if not len(ids): return False
@@ -1660,6 +1662,8 @@ class BranchWindow(QMdiSubWindow): self.ResizeColumnsToContents() + self.context_menu = TreeContextMenu(self.view) + self.find_bar = FindBar(self, self, True) self.finder = ChildDataItemFinder(self.model.root)
@@ -2469,6 +2473,39 @@ def CopyCellsToClipboardHdr(view): def CopyCellsToClipboardCSV(view): CopyCellsToClipboard(view, True, True) +# Context menu + +class ContextMenu(object): + + def __init__(self, view): + self.view = view + self.view.setContextMenuPolicy(Qt.CustomContextMenu) + self.view.customContextMenuRequested.connect(self.ShowContextMenu) + + def ShowContextMenu(self, pos): + menu = QMenu(self.view) + self.AddActions(menu) + menu.exec_(self.view.mapToGlobal(pos)) + + def AddCopy(self, menu): + menu.addAction(CreateAction("&Copy selection", "Copy to clipboard", lambda: CopyCellsToClipboardHdr(self.view), self.view)) + menu.addAction(CreateAction("Copy selection as CS&V", "Copy to clipboard as CSV", lambda: CopyCellsToClipboardCSV(self.view), self.view)) + + def AddActions(self, menu): + self.AddCopy(menu) + +class TreeContextMenu(ContextMenu): + + def __init__(self, view): + super(TreeContextMenu, self).__init__(view) + + def AddActions(self, menu): + i = self.view.currentIndex() + text = str(i.data()).strip() + if len(text): + menu.addAction(CreateAction('Copy "' + text + '"', "Copy to clipboard", lambda: QApplication.clipboard().setText(text), self.view)) + self.AddCopy(menu) + # Table window class TableWindow(QMdiSubWindow, ResizeColumnsToContentsBase):
@@ -2492,6 +2529,8 @@ class TableWindow(QMdiSubWindow, ResizeColumnsToContentsBase): self.ResizeColumnsToContents() + self.context_menu = ContextMenu(self.view) + self.find_bar = FindBar(self, self, True) self.finder = ChildDataItemFinder(self.data_model)
@@ -2608,6 +2647,8 @@ class TopCallsWindow(QMdiSubWindow, ResizeColumnsToContentsBase): self.view.setSelectionMode(QAbstractItemView.ContiguousSelection) self.view.CopyCellsToClipboard = CopyTableCellsToClipboard + self.context_menu = ContextMenu(self.view) + self.ResizeColumnsToContents() self.find_bar = FindBar(self, self, True)
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 18/73 perf scripts python: exported-sql-viewer.py: Add 'About' dialog box
In reply to Arnaldo Carvalho de MeloFrom: Adrian Hunter <[email protected]>
With support for Python 2 or 3 and PySide 1 or 2 (Qt 4 or 5), it is useful to see what versions are in use. Add an 'About' dialog box that displays Python, PySide, Qt and database server (SQLite or PostgreSQL) version numbers.
Committer testing:
$ python ~acme/libexec/perf-core/scripts/python/exported-sql-viewer.py ~/c/adrian.hunter/simple-retpoline.db
Then go to 'Help', then 'About', select all the lines with the mouse press 'Control+C', then, on the same terminal press control+shift+V which shows my current environment:
Python version: 2.7.16 PySide version: 1 Qt version: 4.8.7 SQLite version: 3.26.0
--- .../scripts/python/exported-sql-viewer.py | 59 +++++++++++++++++++ 1 file changed, 59 insertions(+)
tools/perf/scripts/python/exported-sql-viewer.py +59 -0
@@ -2913,6 +2913,60 @@ class HelpOnlyWindow(QMainWindow): self.setCentralWidget(self.text) +# PostqreSQL server version + +def PostqreSQLServerVersion(db): + query = QSqlQuery(db) + QueryExec(query, "SELECT VERSION()") + if query.next(): + v_str = query.value(0) + v_list = v_str.strip().split(" ") + if v_list[0] == "PostgreSQL" and v_list[2] == "on": + return v_list[1] + return v_str + return "Unknown" + +# SQLite version + +def SQLiteVersion(db): + query = QSqlQuery(db) + QueryExec(query, "SELECT sqlite_version()") + if query.next(): + return query.value(0) + return "Unknown" + +# About dialog + +class AboutDialog(QDialog): + + def __init__(self, glb, parent=None): + super(AboutDialog, self).__init__(parent) + + self.setWindowTitle("About Exported SQL Viewer") + self.setMinimumWidth(300) + + pyside_version = "1" if pyside_version_1 else "2" + + text = "<pre>" + text += "Python version: " + sys.version.split(" ")[0] + "\n" + text += "PySide version: " + pyside_version + "\n" + text += "Qt version: " + qVersion() + "\n" + if glb.dbref.is_sqlite3: + text += "SQLite version: " + SQLiteVersion(glb.db) + "\n" + else: + text += "PostqreSQL version: " + PostqreSQLServerVersion(glb.db) + "\n" + text += "</pre>" + + self.text = QTextBrowser() + self.text.setHtml(text) + self.text.setReadOnly(True) + self.text.setOpenExternalLinks(True) + + self.vbox = QVBoxLayout() + self.vbox.addWidget(self.text) + + self.setLayout(self.vbox); + # Font resize def ResizeFont(widget, diff):
@@ -3010,6 +3064,7 @@ class MainWindow(QMainWindow): help_menu = menu.addMenu("&Help") help_menu.addAction(CreateAction("&Exported SQL Viewer Help", "Helpful information", self.Help, self, QKeySequence.HelpContents)) + help_menu.addAction(CreateAction("&About Exported SQL Viewer", "About this application", self.About, self)) def Try(self, fn): win = self.mdi_area.activeSubWindow()
@@ -3095,6 +3150,10 @@ class MainWindow(QMainWindow): def Help(self): HelpWindow(self.glb, self) + def About(self): + dialog = AboutDialog(self.glb, self) + dialog.exec_() + # XED Disassembler class xed_state_t(Structure):
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 19/73 perf vendor events intel: Add uncore_upi JSON support
In reply to Arnaldo Carvalho de MeloFrom: Kan Liang <[email protected]>
Perf cannot parse UPI (Intel's "Ultra Path Interconnect" [1]) events.
# perf stat -e UPI_DATA_BANDWIDTH_TX
event syntax error: 'UPI_DATA_BANDWIDTH_TX'
\___ parser error
Run 'perf list' for a list of valid eventsThe JSON lists call the box UPI LL, while perf calls it upi. Add conversion support to JSON to convert the unit properly.
Committer notes:
[1] https://en.wikipedia.org/wiki/Intel_Ultra_Path_Interconnect
"The Intel Ultra Path Interconnect (UPI) is a point-to-point processor interconnect developed by Intel which replaced the Intel QuickPath Interconnect (QPI) in Xeon Skylake-SP platforms starting in 2017.
UPI is a low-latency coherent interconnect for scalable multiprocessor systems with a shared address space. It uses a directory-based home snoop coherency protocol with a transfer speed of up to 10.4 GT/s. Supporting processors typically have two or three UPI links."
--- tools/perf/pmu-events/jevents.c | 1 + 1 file changed, 1 insertion(+)
tools/perf/pmu-events/jevents.c +1 -0
@@ -235,6 +235,7 @@ static struct map { { "iMPH-U", "uncore_arb" }, { "CPU-M-CF", "cpum_cf" }, { "CPU-M-SF", "cpum_sf" }, + { "UPI LL", "uncore_upi" }, {} };
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 20/73 perf machine: Null-terminate version char array upon fgets(/proc/version) error
In reply to Arnaldo Carvalho de MeloFrom: Donald Yandt <[email protected]>
If fgets() fails due to any other error besides end-of-file, the version char array may not even be null-terminated.
--- tools/perf/util/machine.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
tools/perf/util/machine.c +2 -1
@@ -1234,8 +1234,9 @@ static char *get_kernel_version(const char *root_dir)if (!file)return NULL ; - version[0 ] ='\0' ; tmp =fgets (version,sizeof (version), file); +if (!tmp) + *version ='\0' ;fclose (file); name =strstr (version, prefix);
Signature · 1 line
2.20.1
Re: [PATCH 20/73] perf machine: Null-terminate version char array upon fgets(/proc/version) error
Thank you Arnaldo for signing my patch.
I think we should use version 4 of my patch and return NULL instead of null-terminating for efficiency.
Thanks,
Donald
Quoted 38 lines · Arnaldo Carvalho de Melo
From: Donald Yandt <[email protected]>
If
fgets()fails due to any other error besides end-of-file, the version char array may not even be null-terminated.Signed-off-by Donald Yandt <[email protected]>Cc Alexander Shishkin <[email protected]>Cc Avi Kivity <[email protected]>Cc Jiri Olsa <[email protected]>Cc Peter Zijlstra <[email protected]>Cc Yanmin Zhang <[email protected]>Fixes a1645ce12adb ("perf: 'perf kvm' tool for monitoring guest performance from host")Signed-off-by Arnaldo Carvalho de Melo <[email protected]>---
tools/perf/util/machine.c| 3 ++-1 file changed, 2 insertions(+), 1 deletion(-)tools/perf/util/machine.c +2
-1@@ -1234,8 +1234,9 @@ static char *get_kernel_version(const char *root_dir)if (!file)return NULL ; - version[0 ] ='\0' ; tmp =fgets (version,sizeof (version), file); +if (!tmp) + *version ='\0' ;fclose (file); name =strstr (version, prefix);Signature · 1 line
2.20.1
Re: [PATCH 20/73] perf machine: Null-terminate version char array upon fgets(/proc/version) error
Em Fri, May 17, 2019 at 08:05:39PM -0400, Donald Yandt escreveu:
Quoted 3 lines
Thank you Arnaldo for signing my patch.
I think we should use version 4 of my patch and return NULL instead of null-terminating for efficiency.
Please resubmit then, v3 is already upstream.
- Arnaldo
Quoted 42 lines
Thanks,
Donald
Quoted 38 lines · Arnaldo Carvalho de Melo
From: Donald Yandt <[email protected]>
If
fgets()fails due to any other error besides end-of-file, the version char array may not even be null-terminated.Signed-off-by Donald Yandt <[email protected]>Cc Alexander Shishkin <[email protected]>Cc Avi Kivity <[email protected]>Cc Jiri Olsa <[email protected]>Cc Peter Zijlstra <[email protected]>Cc Yanmin Zhang <[email protected]>Fixes a1645ce12adb ("perf: 'perf kvm' tool for monitoring guest performance from host")Signed-off-by Arnaldo Carvalho de Melo <[email protected]>---
tools/perf/util/machine.c| 3 ++-1 file changed, 2 insertions(+), 1 deletion(-)tools/perf/util/machine.c +2
-1@@ -1234,8 +1234,9 @@ static char *get_kernel_version(const char *root_dir)if (!file)return NULL ; - version[0 ] ='\0' ; tmp =fgets (version,sizeof (version), file); +if (!tmp) + *version ='\0' ;fclose (file); name =strstr (version, prefix);Signature · 1 line
2.20.1
Signature · 1 line
- Arnaldo
Arnaldo Carvalho de Melo
maintainerPATCH 21/73 tools lib traceevent: Introduce man pages
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Initial support for libtraceevent man pages - Documentation directory, templates, configurations, Makefiles.
The first man page is also part of the patch - summary of the library and all its APIs.
Building of the documentation is integrated into the libtraceevent build process, new targets are added to its Makefile:
make help
make doc
make doc-clean
make doc-install
make doc-uninstall
--- tools/lib/traceevent/Documentation/Makefile | 207 ++++++++++++++++++ .../traceevent/Documentation/asciidoc.conf | 91 ++++++++ .../Documentation/libtraceevent.txt | 203 +++++++++++++++++ .../traceevent/Documentation/manpage-1.72.xsl | 14 ++ .../traceevent/Documentation/manpage-base.xsl | 35 +++ .../Documentation/manpage-bold-literal.xsl | 17 ++ .../Documentation/manpage-normal.xsl | 13 ++ .../Documentation/manpage-suppress-sp.xsl | 21 ++ tools/lib/traceevent/Makefile | 33 +++ 9 files changed, 634 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/Makefile create mode 100644 tools/lib/traceevent/Documentation/asciidoc.conf create mode 100644 tools/lib/traceevent/Documentation/libtraceevent.txt create mode 100644 tools/lib/traceevent/Documentation/manpage-1.72.xsl create mode 100644 tools/lib/traceevent/Documentation/manpage-base.xsl create mode 100644 tools/lib/traceevent/Documentation/manpage-bold-literal.xsl create mode 100644 tools/lib/traceevent/Documentation/manpage-normal.xsl create mode 100644 tools/lib/traceevent/Documentation/manpage-suppress-sp.xsl
tools/lib/traceevent/Documentation/Makefile new file +207 -0
@@ -0,0 +1,207 @@ +include ../../../scripts/Makefile.include +include ../../../scripts/utilities.mak + +# This Makefile and manpage XSL files were taken from tools/perf/Documentation +# and modified for libtraceevent. + +MAN3_TXT= \ + $(wildcard libtraceevent-*.txt) \ + libtraceevent.txt + +MAN_TXT = $(MAN3_TXT) +_MAN_XML=$(patsubst %.txt,%.xml,$(MAN_TXT)) +_MAN_HTML=$(patsubst %.txt,%.html,$(MAN_TXT)) +_DOC_MAN3=$(patsubst %.txt,%.3,$(MAN3_TXT)) + +MAN_XML=$(addprefix $(OUTPUT),$(_MAN_XML)) +MAN_HTML=$(addprefix $(OUTPUT),$(_MAN_HTML)) +DOC_MAN3=$(addprefix $(OUTPUT),$(_DOC_MAN3)) + +# Make the path relative to DESTDIR, not prefix +ifndef DESTDIR +prefix?=$(HOME) +endif +bindir?=$(prefix)/bin +htmldir?=$(prefix)/share/doc/libtraceevent-doc +pdfdir?=$(prefix)/share/doc/libtraceevent-doc +mandir?=$(prefix)/share/man +man3dir=$(mandir)/man3 + +ASCIIDOC=asciidoc +ASCIIDOC_EXTRA = --unsafe -f asciidoc.conf +ASCIIDOC_HTML = xhtml11 +MANPAGE_XSL = manpage-normal.xsl +XMLTO_EXTRA = +INSTALL?=install +RM ?= rm -f + +ifdef USE_ASCIIDOCTOR +ASCIIDOC = asciidoctor +ASCIIDOC_EXTRA = -a compat-mode +ASCIIDOC_EXTRA += -I. -rasciidoctor-extensions +ASCIIDOC_EXTRA += -a mansource="libtraceevent" -a manmanual="libtraceevent Manual" +ASCIIDOC_HTML = xhtml5 +endif + +XMLTO=xmlto + +_tmp_tool_path := $(call get-executable,$(ASCIIDOC)) +ifeq ($(_tmp_tool_path),) + missing_tools = $(ASCIIDOC) +endif + +ifndef USE_ASCIIDOCTOR +_tmp_tool_path := $(call get-executable,$(XMLTO)) +ifeq ($(_tmp_tool_path),) + missing_tools += $(XMLTO) +endif +endif + +# +# For asciidoc ... +# -7.1.2, no extra settings are needed. +# 8.0-, set ASCIIDOC8. +# + +# +# For docbook-xsl ... +# -1.68.1, set ASCIIDOC_NO_ROFF? (based on changelog from 1.73.0) +# 1.69.0, no extra settings are needed? +# 1.69.1-1.71.0, set DOCBOOK_SUPPRESS_SP? +# 1.71.1, no extra settings are needed? +# 1.72.0, set DOCBOOK_XSL_172. +# 1.73.0-, set ASCIIDOC_NO_ROFF +# + +# +# If you had been using DOCBOOK_XSL_172 in an attempt to get rid +# of 'the ".ft C" problem' in your generated manpages, and you +# instead ended up with weird characters around callouts, try +# using ASCIIDOC_NO_ROFF instead (it works fine with ASCIIDOC8). +# + +ifdef ASCIIDOC8 +ASCIIDOC_EXTRA += -a asciidoc7compatible +endif +ifdef DOCBOOK_XSL_172 +ASCIIDOC_EXTRA += -a libtraceevent-asciidoc-no-roff +MANPAGE_XSL = manpage-1.72.xsl +else + ifdef ASCIIDOC_NO_ROFF + # docbook-xsl after 1.72 needs the regular XSL, but will not + # pass-thru raw roff codes from asciidoc.conf, so turn them off. + ASCIIDOC_EXTRA += -a libtraceevent-asciidoc-no-roff + endif +endif +ifdef MAN_BOLD_LITERAL +XMLTO_EXTRA += -m manpage-bold-literal.xsl +endif +ifdef DOCBOOK_SUPPRESS_SP +XMLTO_EXTRA += -m manpage-suppress-sp.xsl +endif + +SHELL_PATH ?= $(SHELL) +# Shell quote; +SHELL_PATH_SQ = $(subst ','\'',$(SHELL_PATH)) + +DESTDIR ?= +DESTDIR_SQ = '$(subst ','\'',$(DESTDIR))' + +export DESTDIR DESTDIR_SQ + +# +# Please note that there is a minor bug in asciidoc. +# The version after 6.0.3 _will_ include the patch found here: +# http://marc.theaimsgroup.com/?l=libtraceevent&m=111558757202243&w=2 +# +# Until that version is released you may have to apply the patch +# yourself - yes, all 6 characters of it! +# +QUIET_SUBDIR0 = +$(MAKE) -C # space to separate -C and subdir +QUIET_SUBDIR1 = + +ifneq ($(findstring $(MAKEFLAGS),w),w) +PRINT_DIR = --no-print-directory +else # "make -w" +NO_SUBDIR = : +endif + +ifneq ($(findstring $(MAKEFLAGS),s),s) +ifneq ($(V),1) + QUIET_ASCIIDOC = @echo ' ASCIIDOC '$@; + QUIET_XMLTO = @echo ' XMLTO '$@; + QUIET_SUBDIR0 = +@subdir= + QUIET_SUBDIR1 = ;$(NO_SUBDIR) \ + echo ' SUBDIR ' $$subdir; \ + $(MAKE) $(PRINT_DIR) -C $$subdir + export V +endif +endif + +all: html man + +man: man3 +man3: $(DOC_MAN3) + +html: $(MAN_HTML) + +$(MAN_HTML) $(DOC_MAN3): asciidoc.conf + +install: install-man + +check-man-tools: +ifdef missing_tools + $(error "You need to install $(missing_tools) for man pages") +endif + +do-install-man: man + $(call QUIET_INSTALL, Documentation-man) \ + $(INSTALL) -d -m 755 $(DESTDIR)$(man3dir); \ + $(INSTALL) -m 644 $(DOC_MAN3) $(DESTDIR)$(man3dir); + +install-man: check-man-tools man do-install-man + +uninstall: uninstall-man + +uninstall-man: + $(call QUIET_UNINST, Documentation-man) \ + $(Q)$(RM) $(addprefix $(DESTDIR)$(man3dir)/,$(DOC_MAN3)) + + +ifdef missing_tools + DO_INSTALL_MAN = $(warning Please install $(missing_tools) to have the man pages installed) +else + DO_INSTALL_MAN = do-install-man +endif + +CLEAN_FILES = \ + $(MAN_XML) $(addsuffix +,$(MAN_XML)) \ + $(MAN_HTML) $(addsuffix +,$(MAN_HTML)) \ + $(DOC_MAN3) *.3 + +clean: + $(call QUIET_CLEAN, Documentation) $(RM) $(CLEAN_FILES) + +ifdef USE_ASCIIDOCTOR +$(OUTPUT)%.3 : $(OUTPUT)%.txt + $(QUIET_ASCIIDOC)$(RM) $@+ $@ && \ + $(ASCIIDOC) -b manpage -d manpage \ + $(ASCIIDOC_EXTRA) -alibtraceevent_version=$(EVENT_PARSE_VERSION) -o $@+ $< && \ + mv $@+ $@ +endif + +$(OUTPUT)%.3 : $(OUTPUT)%.xml + $(QUIET_XMLTO)$(RM) $@ && \ + $(XMLTO) -o $(OUTPUT). -m $(MANPAGE_XSL) $(XMLTO_EXTRA) man $< + +$(OUTPUT)%.xml : %.txt + $(QUIET_ASCIIDOC)$(RM) $@+ $@ && \ + $(ASCIIDOC) -b docbook -d manpage \ + $(ASCIIDOC_EXTRA) -alibtraceevent_version=$(EVENT_PARSE_VERSION) -o $@+ $< && \ + mv $@+ $@ + +$(MAN_HTML): $(OUTPUT)%.html : %.txt + $(QUIET_ASCIIDOC)$(RM) $@+ $@ && \ + $(ASCIIDOC) -b $(ASCIIDOC_HTML) -d manpage \ + $(ASCIIDOC_EXTRA) -aperf_version=$(EVENT_PARSE_VERSION) -o $@+ $< && \ + mv $@+ $@
tools/lib/traceevent/Documentation/asciidoc.conf new file +91 -0
91 lines not on this page · raw message
tools/lib/traceevent/Documentation/libtraceevent.txt new file +203 -0
203 lines not on this page · raw message
tools/lib/traceevent/Documentation/manpage-1.72.xsl new file +14 -0
@@ -0,0 +1,14 @@ +<!-- manpage-1.72.xsl: + special settings for manpages rendered from asciidoc+docbook + handles peculiarities in docbook-xsl 1.72.0 --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<xsl:import href="manpage-base.xsl"/> + +<!-- these are the special values for the roff control characters + needed for docbook-xsl 1.72.0 --> +<xsl:param name="git.docbook.backslash">▓</xsl:param> +<xsl:param name="git.docbook.dot" >⌂</xsl:param> + +</xsl:stylesheet>
tools/lib/traceevent/Documentation/manpage-base.xsl new file +35 -0
@@ -0,0 +1,35 @@ +<!-- manpage-base.xsl: + special formatting for manpages rendered from asciidoc+docbook --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<!-- these params silence some output from xmlto --> +<xsl:param name="man.output.quietly" select="1"/> +<xsl:param name="refentry.meta.get.quietly" select="1"/> + +<!-- convert asciidoc callouts to man page format; + git.docbook.backslash and git.docbook.dot params + must be supplied by another XSL file or other means --> +<xsl:template match="co"> + <xsl:value-of select="concat( + $git.docbook.backslash,'fB(', + substring-after(@id,'-'),')', + $git.docbook.backslash,'fR')"/> +</xsl:template> +<xsl:template match="calloutlist"> + <xsl:value-of select="$git.docbook.dot"/> + <xsl:text>sp </xsl:text> + <xsl:apply-templates/> + <xsl:text> </xsl:text> +</xsl:template> +<xsl:template match="callout"> + <xsl:value-of select="concat( + $git.docbook.backslash,'fB', + substring-after(@arearefs,'-'), + '. ',$git.docbook.backslash,'fR')"/> + <xsl:apply-templates/> + <xsl:value-of select="$git.docbook.dot"/> + <xsl:text>br </xsl:text> +</xsl:template> + +</xsl:stylesheet>
tools/lib/traceevent/Documentation/manpage-bold-literal.xsl new file +17 -0
@@ -0,0 +1,17 @@ +<!-- manpage-bold-literal.xsl: + special formatting for manpages rendered from asciidoc+docbook --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<!-- render literal text as bold (instead of plain or monospace); + this makes literal text easier to distinguish in manpages + viewed on a tty --> +<xsl:template match="literal"> + <xsl:value-of select="$git.docbook.backslash"/> + <xsl:text>fB</xsl:text> + <xsl:apply-templates/> + <xsl:value-of select="$git.docbook.backslash"/> + <xsl:text>fR</xsl:text> +</xsl:template> + +</xsl:stylesheet>
tools/lib/traceevent/Documentation/manpage-normal.xsl new file +13 -0
@@ -0,0 +1,13 @@ +<!-- manpage-normal.xsl: + special settings for manpages rendered from asciidoc+docbook + handles anything we want to keep away from docbook-xsl 1.72.0 --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<xsl:import href="manpage-base.xsl"/> + +<!-- these are the normal values for the roff control characters --> +<xsl:param name="git.docbook.backslash">\</xsl:param> +<xsl:param name="git.docbook.dot" >.</xsl:param> + +</xsl:stylesheet>
tools/lib/traceevent/Documentation/manpage-suppress-sp.xsl new file +21 -0
@@ -0,0 +1,21 @@ +<!-- manpage-suppress-sp.xsl: + special settings for manpages rendered from asciidoc+docbook + handles erroneous, inline .sp in manpage output of some + versions of docbook-xsl --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<!-- attempt to work around spurious .sp at the tail of the line + that some versions of docbook stylesheets seem to add --> +<xsl:template match="simpara"> + <xsl:variable name="content"> + <xsl:apply-templates/> + </xsl:variable> + <xsl:value-of select="normalize-space($content)"/> + <xsl:if test="not(ancestor::authorblurb) and + not(ancestor::personblurb)"> + <xsl:text> </xsl:text> + </xsl:if> +</xsl:template> + +</xsl:stylesheet>
tools/lib/traceevent/Makefile +33 -0
@@ -56,6 +56,7 @@ includedir_SQ = '$(subst ','\'',$(includedir))' export man_dir man_dir_SQ INSTALL export DESTDIR DESTDIR_SQ +export EVENT_PARSE_VERSION set_plugin_dir := 1
@@ -318,6 +319,38 @@ clean: $(RM) TRACEEVENT-CFLAGS tags TAGS; \ $(RM) $(PKG_CONFIG_FILE) +PHONY += doc +doc: + $(call descend,Documentation) + +PHONY += doc-clean +doc-clean: + $(call descend,Documentation,clean) + +PHONY += doc-install +doc-install: + $(call descend,Documentation,install) + +PHONY += doc-uninstall +doc-uninstall: + $(call descend,Documentation,uninstall) + +PHONY += help +help: + @echo 'Possible targets:' + @echo'' + @echo ' all - default, compile the library and the'\ + 'plugins' + @echo ' plugins - compile the plugins' + @echo ' install - install the library, the plugins,'\ + 'the header and pkgconfig files' + @echo ' clean - clean the library and the plugins object files' + @echo ' doc - compile the documentation files - man'\ + 'and html pages, in the Documentation directory' + @echo ' doc-clean - clean the documentation files' + @echo ' doc-install - install the man pages' + @echo ' doc-uninstall - uninstall the man pages' + @echo'' PHONY += force plugins force:
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 22/73 tools lib traceevent: Add support for man pages with multiple names
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Added support for man pages with multiple names, used to combine the description of several APIs into one page.
--- .../traceevent/Documentation/asciidoc.conf | 35 +++++++++++++++++-- 1 file changed, 32 insertions(+), 3 deletions(-)
tools/lib/traceevent/Documentation/asciidoc.conf +32 -3
@@ -74,12 +74,41 @@ template::[header-declarations] <refmeta> <refentrytitle>{mantitle}</refentrytitle> <manvolnum>{manvolnum}</manvolnum> -<refmiscinfo class="source">trace-cmd</refmiscinfo> +<refmiscinfo class="source">libtraceevent</refmiscinfo> <refmiscinfo class="version">{libtraceevent_version}</refmiscinfo> -<refmiscinfo class="manual">trace-cmd Manual</refmiscinfo> +<refmiscinfo class="manual">libtraceevent Manual</refmiscinfo> </refmeta> <refnamediv> - <refname>{manname}</refname> + <refname>{manname1}</refname> + <refname>{manname2}</refname> + <refname>{manname3}</refname> + <refname>{manname4}</refname> + <refname>{manname5}</refname> + <refname>{manname6}</refname> + <refname>{manname7}</refname> + <refname>{manname8}</refname> + <refname>{manname9}</refname> + <refname>{manname10}</refname> + <refname>{manname11}</refname> + <refname>{manname12}</refname> + <refname>{manname13}</refname> + <refname>{manname14}</refname> + <refname>{manname15}</refname> + <refname>{manname16}</refname> + <refname>{manname17}</refname> + <refname>{manname18}</refname> + <refname>{manname19}</refname> + <refname>{manname20}</refname> + <refname>{manname21}</refname> + <refname>{manname22}</refname> + <refname>{manname23}</refname> + <refname>{manname24}</refname> + <refname>{manname25}</refname> + <refname>{manname26}</refname> + <refname>{manname27}</refname> + <refname>{manname28}</refname> + <refname>{manname29}</refname> + <refname>{manname30}</refname> <refpurpose>{manpurpose}</refpurpose> </refnamediv> endif::backend-docbook[]
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 23/73 tools lib traceevent: Man pages for tep_handler related APIs
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Added 4 new man pages, describing libtraceevent APIs:
tep_register_comm(),
tep_override_comm(),
tep_is_pid_registered(),
tep_data_comm_from_pid(),
tep_data_pid_from_comm(),
tep_cmdline_pid(),
tep_alloc(),
tep_free(),
tep_get_long_size(),
tep_set_long_size(),
tep_set_flag(),
tep_clear_flag(),
tep_test_flag()
--- .../Documentation/libtraceevent-commands.txt | 153 ++++++++++++++++++ .../Documentation/libtraceevent-handle.txt | 101 ++++++++++++ .../Documentation/libtraceevent-long_size.txt | 78 +++++++++ .../Documentation/libtraceevent-set_flag.txt | 104 ++++++++++++ 4 files changed, 436 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-commands.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-handle.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-long_size.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-set_flag.txt
tools/lib/traceevent/Documentation/libtraceevent-commands.txt new file +153 -0
@@ -0,0 +1,153 @@ +libtraceevent(3) +================ + +NAME +---- +tep_register_comm, tep_override_comm, tep_pid_is_registered, +tep_data_comm_from_pid, tep_data_pid_from_comm, tep_cmdline_pid - +Manage pid to process name mappings. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +int *tep_register_comm*(struct tep_handle pass:[*]_tep_, const char pass:[*]_comm_, int _pid_); +int *tep_override_comm*(struct tep_handle pass:[*]_tep_, const char pass:[*]_comm_, int _pid_); +bool *tep_is_pid_registered*(struct tep_handle pass:[*]_tep_, int _pid_); +const char pass:[*]*tep_data_comm_from_pid*(struct tep_handle pass:[*]_pevent_, int _pid_); +struct cmdline pass:[*]*tep_data_pid_from_comm*(struct tep_handle pass:[*]_pevent_, const char pass:[*]_comm_, struct cmdline pass:[*]_next_); +int *tep_cmdline_pid*(struct tep_handle pass:[*]_pevent_, struct cmdline pass:[*]_cmdline_); +-- + +DESCRIPTION +----------- +These functions can be used to handle the mapping between pid and process name. +The library builds a cache of these mappings, which is used to display the name +of the process, instead of its pid. This information can be retrieved from +tracefs/saved_cmdlines file. + +The _tep_register_comm()_ function registers a _pid_ / process name mapping. +If a command with the same _pid_ is already registered, an error is returned. +The _pid_ argument is the process ID, the _comm_ argument is the process name, +_tep_ is the event context. The _comm_ is duplicated internally. + +The _tep_override_comm()_ function registers a _pid_ / process name mapping. +If a process with the same pid is already registered, the process name string is +udapted with the new one. The _pid_ argument is the process ID, the _comm_ +argument is the process name, _tep_ is the event context. The _comm_ is +duplicated internally. + +The _tep_is_pid_registered()_ function checks if a pid has a process name +mapping registered. The _pid_ argument is the process ID, _tep_ is the event +context. + +The _tep_data_comm_from_pid()_ function returns the process name for a given +pid. The _pid_ argument is the process ID, _tep_ is the event context. +The returned string should not be freed, but will be freed when the _tep_ +handler is closed. + +The _tep_data_pid_from_comm()_ function returns a pid for a given process name. +The _comm_ argument is the process name, _tep_ is the event context. +The argument _next_ is the cmdline structure to search for the next pid. +As there may be more than one pid for a given process, the result of this call +can be passed back into a recurring call in the _next_ parameter, to search for +the next pid. If _next_ is NULL, it will return the first pid associated with +the _comm_. The function performs a linear search, so it may be slow. + +The _tep_cmdline_pid()_ function returns the pid associated with a given +_cmdline_. The _tep_ argument is the event context. + +RETURN VALUE +------------ +_tep_register_comm()_ function returns 0 on success. In case of an error -1 is +returned and errno is set to indicate the cause of the problem: ENOMEM, if there +is not enough memory to duplicate the _comm_ or EEXIST if a mapping for this +_pid_ is already registered. + +_tep_override_comm()_ function returns 0 on success. In case of an error -1 is +returned and errno is set to indicate the cause of the problem: ENOMEM, if there +is not enough memory to duplicate the _comm_. + +_tep_is_pid_registered()_ function returns true if the _pid_ has a process name +mapped to it, false otherwise. + +_tep_data_comm_from_pid()_ function returns the process name as string, or the +string "<...>" if there is no mapping for the given pid. + +_tep_data_pid_from_comm()_ function returns a pointer to a struct cmdline, that +holds a pid for a given process, or NULL if none is found. This result can be +passed back into a recurring call as the _next_ parameter of the function. + +_tep_cmdline_pid()_ functions returns the pid for the give cmdline. If _cmdline_ + is NULL, then -1 is returned. + +EXAMPLE +------- +The following example registers pid for command "ls", in context of event _tep_ +and performs various searches for pid / process name mappings: +[source,c] +-- +#include <event-parse.h> +... +int ret; +int ls_pid = 1021; +struct tep_handle *tep = tep_alloc(); +... + ret = tep_register_comm(tep, "ls", ls_pid); + if (ret != 0 && errno == EEXIST) + ret = tep_override_comm(tep, "ls", ls_pid); + if (ret != 0) { + /* Failed to register pid / command mapping */ + } +... + if (tep_is_pid_registered(tep, ls_pid) == 0) { + /* Command mapping for ls_pid is not registered */ + } +... + const char *comm = tep_data_comm_from_pid(tep, ls_pid); + if (comm) { + /* Found process name for ls_pid */ + } +... + int pid; + struct cmdline *cmd = tep_data_pid_from_comm(tep, "ls", NULL); + while (cmd) { + pid = tep_cmdline_pid(tep, cmd); + /* Found pid for process "ls" */ + cmd = tep_data_pid_from_comm(tep, "ls", cmd); + } +-- +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
tools/lib/traceevent/Documentation/libtraceevent-handle.txt new file +101 -0
101 lines not on this page · raw message
tools/lib/traceevent/Documentation/libtraceevent-long_size.txt new file +78 -0
78 lines not on this page · raw message
tools/lib/traceevent/Documentation/libtraceevent-set_flag.txt new file +104 -0
104 lines not on this page · raw message
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 24/73 tools lib traceevent: Man page for header_page APIs
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create a man page for libtraceevent APIs:
tep_get_header_page_size(),
tep_get_header_timestamp_size(),
tep_is_old_format()
--- .../libtraceevent-header_page.txt | 102 ++++++++++++++++++ 1 file changed, 102 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-header_page.txt
tools/lib/traceevent/Documentation/libtraceevent-header_page.txt new file +102 -0
@@ -0,0 +1,102 @@ +libtraceevent(3) +================ + +NAME +---- +tep_get_header_page_size, tep_get_header_timestamp_size, tep_is_old_format - +Get the data stored in the header page, in kernel context. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +int *tep_get_header_page_size*(struct tep_handle pass:[*]_tep_); +int *tep_get_header_timestamp_size*(struct tep_handle pass:[*]_tep_); +bool *tep_is_old_format*(struct tep_handle pass:[*]_tep_); +-- +DESCRIPTION +----------- +These functions retrieve information from kernel context, stored in tracefs +events/header_page. Old kernels do not have header page info, so default values +from user space context are used. + +The _tep_get_header_page_size()_ function returns the size of a long integer, +in kernel context. The _tep_ argument is trace event parser context. +This information is retrieved from tracefs events/header_page, "commit" field. + +The _tep_get_header_timestamp_size()_ function returns the size of timestamps, +in kernel context. The _tep_ argument is trace event parser context. This +information is retrieved from tracefs events/header_page, "timestamp" field. + +The _tep_is_old_format()_ function returns true if the kernel predates +the addition of events/header_page, otherwise it returns false. + +RETURN VALUE +------------ +The _tep_get_header_page_size()_ function returns the size of a long integer, +in bytes. + +The _tep_get_header_timestamp_size()_ function returns the size of timestamps, +in bytes. + +The _tep_is_old_format()_ function returns true, if an old kernel is used to +generate the tracing data, which has no event/header_page. If the kernel is new, +or _tep_ is NULL, false is returned. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +... +struct tep_handle *tep = tep_alloc(); +... + int longsize; + int timesize; + bool old; + + longsize = tep_get_header_page_size(tep); + timesize = tep_get_header_timestamp_size(tep); + old = tep_is_old_format(tep); + + printf ("%s kernel is used to generate the tracing data.\n", + old?"Old":"New"); + printf("The size of a long integer is %d bytes.\n", longsize); + printf("The timestamps size is %d bytes.\n", timesize); +... +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 25/73 tools lib traceevent: Man page for get/set cpus APIs
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man pages for libtraceevent APIs:
tep_get_cpus(),
tep_set_cpus()
--- .../Documentation/libtraceevent-cpus.txt | 77 +++++++++++++++++++ 1 file changed, 77 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-cpus.txt
tools/lib/traceevent/Documentation/libtraceevent-cpus.txt new file +77 -0
@@ -0,0 +1,77 @@ +libtraceevent(3) +================ + +NAME +---- +tep_get_cpus, tep_set_cpus - Get / set the number of CPUs, which have a tracing +buffer representing it. Note, the buffer may be empty. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +int *tep_get_cpus*(struct tep_handle pass:[*]_tep_); +void *tep_set_cpus*(struct tep_handle pass:[*]_tep_, int _cpus_); +-- + +DESCRIPTION +----------- +The _tep_get_cpus()_ function gets the number of CPUs, which have a tracing +buffer representing it. The _tep_ argument is trace event parser context. + +The _tep_set_cpus()_ function sets the number of CPUs, which have a tracing +buffer representing it. The _tep_ argument is trace event parser context. +The _cpu_ argument is the number of CPUs with tracing data. + +RETURN VALUE +------------ +The _tep_get_cpus()_ functions returns the number of CPUs, which have tracing +data recorded. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +... +struct tep_handle *tep = tep_alloc(); +... + tep_set_cpus(tep, 5); +... + printf("We have tracing data for %d CPUs", tep_get_cpus(tep)); +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 26/73 tools lib traceevent: Man page for file endian APIs
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man pages for libtraceevent APIs:
tep_is_file_bigendian(),
tep_set_file_bigendian()
--- .../libtraceevent-file_endian.txt | 91 +++++++++++++++++++ 1 file changed, 91 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-file_endian.txt
tools/lib/traceevent/Documentation/libtraceevent-file_endian.txt new file +91 -0
@@ -0,0 +1,91 @@ +libtraceevent(3) +================ + +NAME +---- +tep_is_file_bigendian, tep_set_file_bigendian - Get / set the endianness of the +raw data being accessed by the tep handler. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +enum *tep_endian* { + TEP_LITTLE_ENDIAN = 0, + TEP_BIG_ENDIAN +}; + +bool *tep_is_file_bigendian*(struct tep_handle pass:[*]_tep_); +void *tep_set_file_bigendian*(struct tep_handle pass:[*]_tep_, enum tep_endian _endian_); + +-- +DESCRIPTION +----------- +The _tep_is_file_bigendian()_ function gets the endianness of the raw data, +being accessed by the tep handler. The _tep_ argument is trace event parser +context. + +The _tep_set_file_bigendian()_ function sets the endianness of raw data being +accessed by the tep handler. The _tep_ argument is trace event parser context. +[verse] +-- +The _endian_ argument is the endianness: + _TEP_LITTLE_ENDIAN_ - the raw data is in little endian format, + _TEP_BIG_ENDIAN_ - the raw data is in big endian format. +-- +RETURN VALUE +------------ +The _tep_is_file_bigendian()_ function returns true if the data is in bigendian +format, false otherwise. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +... +struct tep_handle *tep = tep_alloc(); +... + tep_set_file_bigendian(tep, TEP_LITTLE_ENDIAN); +... + if (tep_is_file_bigendian(tep)) { + /* The raw data is in big endian */ + } else { + /* The raw data is in little endian */ + } +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 27/73 tools lib traceevent: Man page for host endian APIs
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man pages for libtraceevent APIs:
tep_is_bigendian(),
tep_is_local_bigendian(),
tep_set_local_bigendian()
--- .../libtraceevent-host_endian.txt | 104 ++++++++++++++++++ 1 file changed, 104 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-host_endian.txt
tools/lib/traceevent/Documentation/libtraceevent-host_endian.txt new file +104 -0
@@ -0,0 +1,104 @@ +libtraceevent(3) +================ + +NAME +---- +tep_is_bigendian, tep_is_local_bigendian, tep_set_local_bigendian - Get / set +the endianness of the local machine. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +enum *tep_endian* { + TEP_LITTLE_ENDIAN = 0, + TEP_BIG_ENDIAN +}; + +int *tep_is_bigendian*(void); +bool *tep_is_local_bigendian*(struct tep_handle pass:[*]_tep_); +void *tep_set_local_bigendian*(struct tep_handle pass:[*]_tep_, enum tep_endian _endian_); +-- + +DESCRIPTION +----------- + +The _tep_is_bigendian()_ gets the endianness of the machine, executing +the function. + +The _tep_is_local_bigendian()_ function gets the endianness of the local +machine, saved in the _tep_ handler. The _tep_ argument is the trace event +parser context. This API is a bit faster than _tep_is_bigendian()_, as it +returns cached endianness of the local machine instead of checking it each time. + +The _tep_set_local_bigendian()_ function sets the endianness of the local +machine in the _tep_ handler. The _tep_ argument is trace event parser context. +The _endian_ argument is the endianness: +[verse] +-- + _TEP_LITTLE_ENDIAN_ - the machine is little endian, + _TEP_BIG_ENDIAN_ - the machine is big endian. +-- + +RETURN VALUE +------------ +The _tep_is_bigendian()_ function returns non zero if the endianness of the +machine, executing the code, is big endian and zero otherwise. + +The _tep_is_local_bigendian()_ function returns true, if the endianness of the +local machine, saved in the _tep_ handler, is big endian, or false otherwise. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +... +struct tep_handle *tep = tep_alloc(); +... + if (tep_is_bigendian()) + tep_set_local_bigendian(tep, TEP_BIG_ENDIAN); + else + tep_set_local_bigendian(tep, TEP_LITTLE_ENDIAN); +... + if (tep_is_local_bigendian(tep)) + printf("This machine you are running on is bigendian\n"); + else + printf("This machine you are running on is little endian\n"); + +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 28/73 tools lib traceevent: Man page for page size APIs
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man pages for libtraceevent APIs:
tep_get_page_size()
tep_set_page_size()
--- .../Documentation/libtraceevent-page_size.txt | 82 +++++++++++++++++++ 1 file changed, 82 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-page_size.txt
tools/lib/traceevent/Documentation/libtraceevent-page_size.txt new file +82 -0
@@ -0,0 +1,82 @@ +libtraceevent(3) +================ + +NAME +---- +tep_get_page_size, tep_set_page_size - Get / set the size of a memory page on +the machine, where the trace is generated + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +int *tep_get_page_size*(struct tep_handle pass:[*]_tep_); +void *tep_set_page_size*(struct tep_handle pass:[*]_tep_, int _page_size_); +-- + +DESCRIPTION +----------- +The _tep_get_page_size()_ function returns the size of a memory page on +the machine, where the trace is generated. The _tep_ argument is trace +event parser context. + +The _tep_set_page_size()_ function stores in the _tep_ context the size of a +memory page on the machine, where the trace is generated. +The _tep_ argument is trace event parser context. +The _page_size_ argument is the size of a memory page, in bytes. + +RETURN VALUE +------------ +The _tep_get_page_size()_ function returns size of the memory page, in bytes. + +EXAMPLE +------- +[source,c] +-- +#include <unistd.h> +#include <event-parse.h> +... +struct tep_handle *tep = tep_alloc(); +... + int page_size = getpagesize(); + + tep_set_page_size(tep, page_size); + + printf("The page size for this machine is %d\n", tep_get_page_size(tep)); + +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 29/73 tools lib traceevent: Man page for tep_strerror()
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man page for tep_strerror() libtraceevent API.
--- .../Documentation/libtraceevent-strerror.txt | 85 +++++++++++++++++++ 1 file changed, 85 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-strerror.txt
tools/lib/traceevent/Documentation/libtraceevent-strerror.txt new file +85 -0
@@ -0,0 +1,85 @@ +libtraceevent(3) +================ + +NAME +---- +tep_strerror - Returns a string describing regular errno and tep error number. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +int *tep_strerror*(struct tep_handle pass:[*]_tep_, enum tep_errno _errnum_, char pass:[*]_buf_, size_t _buflen_); + +-- +DESCRIPTION +----------- +The _tep_strerror()_ function converts tep error number into a human +readable string. +The _tep_ argument is trace event parser context. The _errnum_ is a regular +errno, defined in errno.h, or a tep error number. The string, describing this +error number is copied in the _buf_ argument. The _buflen_ argument is +the size of the _buf_. + +It as a thread safe wrapper around strerror_r(). The library function has two +different behaviors - POSIX and GNU specific. The _tep_strerror()_ API always +behaves as the POSIX version - the error string is copied in the user supplied +buffer. + +RETURN VALUE +------------ +The _tep_strerror()_ function returns 0, if a valid _errnum_ is passed and the +string is copied into _buf_. If _errnum_ is not a valid error number, +-1 is returned and _buf_ is not modified. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +... +struct tep_handle *tep = tep_alloc(); +... +char buf[32]; +char *pool = calloc(1, 128); +if (tep == NULL) { + tep_strerror(tep, TEP_ERRNO__MEM_ALLOC_FAILED, buf, 32); + printf ("The pool is not initialized, %s", buf); +} +... +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 30/73 tools lib traceevent: Man pages for event handler APIs
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man pages for libtraceevent APIs:
tep_register_event_handler()
tep_unregister_event_handler()
--- .../libtraceevent-reg_event_handler.txt | 156 ++++++++++++++++++ 1 file changed, 156 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-reg_event_handler.txt
tools/lib/traceevent/Documentation/libtraceevent-reg_event_handler.txt new file +156 -0
@@ -0,0 +1,156 @@ +libtraceevent(3) +================ + +NAME +---- +tep_register_event_handler, tep_unregister_event_handler - Register / +unregisters a callback function to parse an event information. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +enum *tep_reg_handler* { + _TEP_REGISTER_SUCCESS_, + _TEP_REGISTER_SUCCESS_OVERWRITE_, +}; + +int *tep_register_event_handler*(struct tep_handle pass:[*]_tep_, int _id_, const char pass:[*]_sys_name_, const char pass:[*]_event_name_, tep_event_handler_func _func_, void pass:[*]_context_); +int *tep_unregister_event_handler*(struct tep_handle pass:[*]tep, int id, const char pass:[*]sys_name, const char pass:[*]event_name, tep_event_handler_func func, void pass:[*]_context_); + +typedef int (*pass:[*]tep_event_handler_func*)(struct trace_seq pass:[*]s, struct tep_record pass:[*]record, struct tep_event pass:[*]event, void pass:[*]context); +-- + +DESCRIPTION +----------- +The _tep_register_event_handler()_ function registers a handler function, +which is going to be called to parse the information for a given event. +The _tep_ argument is the trace event parser context. The _id_ argument is +the id of the event. The _sys_name_ argument is the name of the system, +the event belongs to. The _event_name_ argument is the name of the event. +If _id_ is >= 0, it is used to find the event, otherwise _sys_name_ and +_event_name_ are used. The _func_ is a pointer to the function, which is going +to be called to parse the event information. The _context_ argument is a pointer +to the context data, which will be passed to the _func_. If a handler function +for the same event is already registered, it will be overridden with the new +one. This mechanism allows a developer to override the parsing of a given event. +If for some reason the default print format is not sufficient, the developer +can register a function for an event to be used to parse the data instead. + +The _tep_unregister_event_handler()_ function unregisters the handler function, +previously registered with _tep_register_event_handler()_. The _tep_ argument +is the trace event parser context. The _id_, _sys_name_, _event_name_, _func_, +and _context_ are the same arguments, as when the callback function _func_ was +registered. + +The _tep_event_handler_func_ is the type of the custom event handler +function. The _s_ argument is the trace sequence, it can be used to create a +custom string, describing the event. A _record_ to get the event from is passed +as input parameter and also the _event_ - the handle to the record's event. The +_context_ is custom context, set when the custom event handler is registered. + +RETURN VALUE +------------ +The _tep_register_event_handler()_ function returns _TEP_REGISTER_SUCCESS_ +if the new handler is registered successfully or +_TEP_REGISTER_SUCCESS_OVERWRITE_ if an existing handler is overwritten. +If there is not enough memory to complete the registration, +TEP_ERRNO__MEM_ALLOC_FAILED is returned. + +The _tep_unregister_event_handler()_ function returns 0 if _func_ was removed +successful or, -1 if the event was not found. + +The _tep_event_handler_func_ should return -1 in case of an error, +or 0 otherwise. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +#include <trace-seq.h> +... +struct tep_handle *tep = tep_alloc(); +... +int timer_expire_handler(struct trace_seq *s, struct tep_record *record, + struct tep_event *event, void *context) +{ + trace_seq_printf(s, "hrtimer="); + + if (tep_print_num_field(s, "0x%llx", event, "timer", record, 0) == -1) + tep_print_num_field(s, "0x%llx", event, "hrtimer", record, 1); + + trace_seq_printf(s, " now="); + + tep_print_num_field(s, "%llu", event, "now", record, 1); + + tep_print_func_field(s, " function=%s", event, "function", record, 0); + + return 0; +} +... + int ret; + + ret = tep_register_event_handler(tep, -1, "timer", "hrtimer_expire_entry", + timer_expire_handler, NULL); + if (ret < 0) { + char buf[32]; + + tep_strerror(tep, ret, buf, 32) + printf("Failed to register handler for hrtimer_expire_entry: %s\n", buf); + } else { + switch (ret) { + case TEP_REGISTER_SUCCESS: + printf ("Registered handler for hrtimer_expire_entry\n"); + break; + case TEP_REGISTER_SUCCESS_OVERWRITE: + printf ("Overwrote handler for hrtimer_expire_entry\n"); + break; + } + } +... + ret = tep_unregister_event_handler(tep, -1, "timer", "hrtimer_expire_entry", + timer_expire_handler, NULL); + if ( ret ) + printf ("Failed to unregister handler for hrtimer_expire_entry\n"); + +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*trace-seq.h* + Header file to include in order to have access to trace sequences + related APIs. Trace sequences are used to allow a function to call + several other functions to create a string of data to use. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 31/73 tools lib traceevent: Man pages for function related libtraceevent APIs
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Added new man pages, describing function related libtraceevent APIs:
tep_register_function(),
tep_register_print_string(),
tep_find_function(),
tep_find_function_address(),
tep_set_function_resolver(),
tep_reset_function_resolver()
--- .../Documentation/libtraceevent-func_apis.txt | 183 ++++++++++++++++++ .../Documentation/libtraceevent-func_find.txt | 88 +++++++++ 2 files changed, 271 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-func_apis.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-func_find.txt
tools/lib/traceevent/Documentation/libtraceevent-func_apis.txt new file +183 -0
@@ -0,0 +1,183 @@ +libtraceevent(3) +================ + +NAME +---- +tep_find_function, tep_find_function_address, tep_set_function_resolver, +tep_reset_function_resolver, tep_register_function, tep_register_print_string - +function related tep APIs + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +typedef char pass:[*](*tep_func_resolver_t*)(void pass:[*]_priv_, unsigned long long pass:[*]_addrp_, char pass:[**]_modp_); +int *tep_set_function_resolver*(struct tep_handle pass:[*]_tep_, tep_func_resolver_t pass:[*]_func_, void pass:[*]_priv_); +void *tep_reset_function_resolver*(struct tep_handle pass:[*]_tep_); +const char pass:[*]*tep_find_function*(struct tep_handle pass:[*]_tep_, unsigned long long _addr_); +unsigned long long *tep_find_function_address*(struct tep_handle pass:[*]_tep_, unsigned long long _addr_); +int *tep_register_function*(struct tep_handle pass:[*]_tep_, char pass:[*]_name_, unsigned long long _addr_, char pass:[*]_mod_); +int *tep_register_print_string*(struct tep_handle pass:[*]_tep_, const char pass:[*]_fmt_, unsigned long long _addr_); +-- + +DESCRIPTION +----------- +Some tools may have already a way to resolve the kernel functions. These APIs +allow them to keep using it instead of duplicating all the entries inside. + +The _tep_func_resolver_t_ type is the prototype of the alternative kernel +functions resolver. This function receives a pointer to its custom context +(set with the _tep_set_function_resolver()_ call ) and the address of a kernel +function, which has to be resolved. In case of success, it should return +the name of the function and its module (if any) in _modp_. + +The _tep_set_function_resolver()_ function registers _func_ as an alternative +kernel functions resolver. The _tep_ argument is trace event parser context. +The _priv_ argument is a custom context of the _func_ function. The function +resolver is used by the APIs _tep_find_function()_, +_tep_find_function_address()_, and _tep_print_func_field()_ to resolve +a function address to a function name. + +The _tep_reset_function_resolver()_ function resets the kernel functions +resolver to the default function. The _tep_ argument is trace event parser +context. + + +These APIs can be used to find function name and start address, by given +address. The given address does not have to be exact, it will select +the function that would contain it. + +The _tep_find_function()_ function returns the function name, which contains the +given address _addr_. The _tep_ argument is the trace event parser context. + +The _tep_find_function_address()_ function returns the function start address, +by given address _addr_. The _addr_ does not have to be exact, it will select +the function that would contain it. The _tep_ argument is the trace event +parser context. + +The _tep_register_function()_ function registers a function name mapped to an +address and (optional) module. This mapping is used in case the function tracer +or events have "%pF" or "%pS" parameter in its format string. It is common to +pass in the kallsyms function names with their corresponding addresses with this +function. The _tep_ argument is the trace event parser context. The _name_ is +the name of the function, the string is copied internally. The _addr_ is +the start address of the function. The _mod_ is the kernel module +the function may be in (NULL for none). + +The _tep_register_print_string()_ function registers a string by the address +it was stored in the kernel. Some strings internal to the kernel with static +address are passed to certain events. The "%s" in the event's format field +which has an address needs to know what string would be at that address. The +tep_register_print_string() supplies the parsing with the mapping between kernel +addresses and those strings. The _tep_ argument is the trace event parser +context. The _fmt_ is the string to register, it is copied internally. +The _addr_ is the address the string was located at. + + +RETURN VALUE +------------ +The _tep_set_function_resolver()_ function returns 0 in case of success, or -1 +in case of an error. + +The _tep_find_function()_ function returns the function name, or NULL in case +it cannot be found. + +The _tep_find_function_address()_ function returns the function start address, +or 0 in case it cannot be found. + +The _tep_register_function()_ function returns 0 in case of success. In case of +an error -1 is returned, and errno is set to the appropriate error number. + +The _tep_register_print_string()_ function returns 0 in case of success. In case +of an error -1 is returned, and errno is set to the appropriate error number. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +... +struct tep_handle *tep = tep_alloc(); +... +char *my_resolve_kernel_addr(void *context, + unsigned long long *addrp, char **modp) +{ + struct db *function_database = context; + struct symbol *sym = sql_lookup(function_database, *addrp); + + if (!sym) + return NULL; + + *modp = sym->module_name; + return sym->name; +} + +void show_function( unsigned long long addr) +{ + unsigned long long fstart; + const char *fname; + + if (tep_set_function_resolver(tep, my_resolve_kernel_addr, + function_database) != 0) { + /* failed to register my_resolve_kernel_addr */ + } + + /* These APIs use my_resolve_kernel_addr() to resolve the addr */ + fname = tep_find_function(tep, addr); + fstart = tep_find_function_address(tep, addr); + + /* + addr is in function named fname, starting at fstart address, + at offset (addr - fstart) + */ + + tep_reset_function_resolver(tep); + +} +... + if (tep_register_function(tep, "kvm_exit", + (unsigned long long) 0x12345678, "kvm") != 0) { + /* Failed to register kvm_exit address mapping */ + } +... + if (tep_register_print_string(tep, "print string", + (unsigned long long) 0x87654321, NULL) != 0) { + /* Failed to register "print string" address mapping */ + } +... +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
tools/lib/traceevent/Documentation/libtraceevent-func_find.txt new file +88 -0
88 lines not on this page · raw message
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 32/73 tools lib traceevent: Man pages for registering print function
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man pages for libtraceevent APIs:
tep_register_print_function()
tep_unregister_print_function()
--- .../libtraceevent-reg_print_func.txt | 155 ++++++++++++++++++ 1 file changed, 155 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-reg_print_func.txt
tools/lib/traceevent/Documentation/libtraceevent-reg_print_func.txt new file +155 -0
@@ -0,0 +1,155 @@ +libtraceevent(3) +================ + +NAME +---- +tep_register_print_function,tep_unregister_print_function - +Registers / Unregisters a helper function. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +enum *tep_func_arg_type* { + TEP_FUNC_ARG_VOID, + TEP_FUNC_ARG_INT, + TEP_FUNC_ARG_LONG, + TEP_FUNC_ARG_STRING, + TEP_FUNC_ARG_PTR, + TEP_FUNC_ARG_MAX_TYPES +}; + +typedef unsigned long long (*pass:[*]tep_func_handler*)(struct trace_seq pass:[*]s, unsigned long long pass:[*]args); + +int *tep_register_print_function*(struct tep_handle pass:[*]_tep_, tep_func_handler _func_, enum tep_func_arg_type _ret_type_, char pass:[*]_name_, _..._); +int *tep_unregister_print_function*(struct tep_handle pass:[*]_tep_, tep_func_handler _func_, char pass:[*]_name_); +-- + +DESCRIPTION +----------- +Some events may have helper functions in the print format arguments. +This allows a plugin to dynamically create a way to process one of +these functions. + +The _tep_register_print_function()_ registers such helper function. The _tep_ +argument is the trace event parser context. The _func_ argument is a pointer +to the helper function. The _ret_type_ argument is the return type of the +helper function, value from the _tep_func_arg_type_ enum. The _name_ is the name +of the helper function, as seen in the print format arguments. The _..._ is a +variable list of _tep_func_arg_type_ enums, the _func_ function arguments. +This list must end with _TEP_FUNC_ARG_VOID_. See 'EXAMPLE' section. + +The _tep_unregister_print_function()_ unregisters a helper function, previously +registered with _tep_register_print_function()_. The _tep_ argument is the +trace event parser context. The _func_ and _name_ arguments are the same, used +when the helper function was registered. + +The _tep_func_handler_ is the type of the helper function. The _s_ argument is +the trace sequence, it can be used to create a custom string. +The _args_ is a list of arguments, defined when the helper function was +registered. + +RETURN VALUE +------------ +The _tep_register_print_function()_ function returns 0 in case of success. +In case of an error, TEP_ERRNO_... code is returned. + +The _tep_unregister_print_function()_ returns 0 in case of success, or -1 in +case of an error. + +EXAMPLE +------- +Some events have internal functions calls, that appear in the print format +output. For example "tracefs/events/i915/g4x_wm/format" has: +[source,c] +-- +print fmt: "pipe %c, frame=%u, scanline=%u, wm %d/%d/%d, sr %s/%d/%d/%d, hpll %s/%d/%d/%d, fbc %s", + ((REC->pipe) + 'A'), REC->frame, REC->scanline, REC->primary, + REC->sprite, REC->cursor, yesno(REC->cxsr), REC->sr_plane, + REC->sr_cursor, REC->sr_fbc, yesno(REC->hpll), REC->hpll_plane, + REC->hpll_cursor, REC->hpll_fbc, yesno(REC->fbc) +-- +Notice the call to function _yesno()_ in the print arguments. In the kernel +context, this function has the following implementation: +[source,c] +-- +static const char *yesno(int x) +{ + static const char *yes = "yes"; + static const char *no = "no"; + + return x ? yes : no; +} +-- +The user space event parser has no idea how to handle this _yesno()_ function. +The _tep_register_print_function()_ API can be used to register a user space +helper function, mapped to the kernel's _yesno()_: +[source,c] +-- +#include <event-parse.h> +#include <trace-seq.h> +... +struct tep_handle *tep = tep_alloc(); +... +static const char *yes_no_helper(int x) +{ + return x ? "yes" : "no"; +} +... + if ( tep_register_print_function(tep, + yes_no_helper, + TEP_FUNC_ARG_STRING, + "yesno", + TEP_FUNC_ARG_INT, + TEP_FUNC_ARG_VOID) != 0) { + /* Failed to register yes_no_helper function */ + } + +/* + Now, when the event parser encounters this yesno() function, it will know + how to handle it. +*/ +... + if (tep_unregister_print_function(tep, yes_no_helper, "yesno") != 0) { + /* Failed to unregister yes_no_helper function */ + } +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*trace-seq.h* + Header file to include in order to have access to trace sequences + related APIs. Trace sequences are used to allow a function to call + several other functions to create a string of data to use. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 33/73 tools lib traceevent: Man page for tep_read_number()
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man page for tep_read_number() libtraceevent API.
--- .../libtraceevent-endian_read.txt | 78 +++++++++++++++++++ 1 file changed, 78 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-endian_read.txt
tools/lib/traceevent/Documentation/libtraceevent-endian_read.txt new file +78 -0
@@ -0,0 +1,78 @@ +libtraceevent(3) +================ + +NAME +---- +tep_read_number - Reads a number from raw data. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +unsigned long long *tep_read_number*(struct tep_handle pass:[*]_tep_, const void pass:[*]_ptr_, int _size_); +-- + +DESCRIPTION +----------- +The _tep_read_number()_ function reads an integer from raw data, taking into +account the endianness of the raw data and the current host. The _tep_ argument +is the trace event parser context. The _ptr_ is a pointer to the raw data, where +the integer is, and the _size_ is the size of the integer. + +RETURN VALUE +------------ +The _tep_read_number()_ function returns the integer in the byte order of +the current host. In case of an error, 0 is returned. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +... +struct tep_handle *tep = tep_alloc(); +... +void process_record(struct tep_record *record) +{ + int offset = 24; + int data = tep_read_number(tep, record->data + offset, 4); + + /* Read the 4 bytes at the offset 24 of data as an integer */ +} +... +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 34/73 tools lib traceevent: Man pages for event find APIs
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man pages for libtraceevent APIs:
tep_find_event()
tep_find_event_by_name()
tep_find_event_by_record()
--- .../libtraceevent-event_find.txt | 103 ++++++++++++++++++ 1 file changed, 103 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-event_find.txt
tools/lib/traceevent/Documentation/libtraceevent-event_find.txt new file +103 -0
@@ -0,0 +1,103 @@ +libtraceevent(3) +================ + +NAME +---- +tep_find_event,tep_find_event_by_name,tep_find_event_by_record - +Find events by given key. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +struct tep_event pass:[*]*tep_find_event*(struct tep_handle pass:[*]_tep_, int _id_); +struct tep_event pass:[*]*tep_find_event_by_name*(struct tep_handle pass:[*]_tep_, const char pass:[*]_sys_, const char pass:[*]_name_); +struct tep_event pass:[*]*tep_find_event_by_record*(struct tep_handle pass:[*]_tep_, struct tep_record pass:[*]_record_); +-- + +DESCRIPTION +----------- +This set of functions can be used to search for an event, based on a given +criteria. All functions require a pointer to a _tep_, trace event parser +context. + +The _tep_find_event()_ function searches for an event by given event _id_. The +event ID is assigned dynamically and can be viewed in event's format file, +"ID" field. + +The tep_find_event_by_name()_ function searches for an event by given +event _name_, under the system _sys_. If the _sys_ is NULL (not specified), +the first event with _name_ is returned. + +The tep_find_event_by_record()_ function searches for an event from a given +_record_. + +RETURN VALUE +------------ +All these functions return a pointer to the found event, or NULL if there is no +such event. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +... +struct tep_handle *tep = tep_alloc(); +... +struct tep_event *event; + +event = tep_find_event(tep, 1857); +if (event == NULL) { + /* There is no event with ID 1857 */ +} + +event = tep_find_event_by_name(tep, "kvm", "kvm_exit"); +if (event == NULL) { + /* There is no kvm_exit event, from kvm system */ +} + +void event_from_record(struct tep_record *record) +{ + struct tep_event *event = tep_find_event_by_record(tep, record); + if (event == NULL) { + /* There is no event from given record */ + } +} +... +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 35/73 tools lib traceevent: Man page for list events APIs
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man page for libtraceevent APIs:
tep_list_events()
tep_list_events_copy()
--- .../libtraceevent-event_list.txt | 122 ++++++++++++++++++ 1 file changed, 122 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-event_list.txt
tools/lib/traceevent/Documentation/libtraceevent-event_list.txt new file +122 -0
@@ -0,0 +1,122 @@ +libtraceevent(3) +================ + +NAME +---- +tep_list_events, tep_list_events_copy - +Get list of events, sorted by given criteria. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +enum *tep_event_sort_type* { + _TEP_EVENT_SORT_ID_, + _TEP_EVENT_SORT_NAME_, + _TEP_EVENT_SORT_SYSTEM_, +}; + +struct tep_event pass:[*]pass:[*]*tep_list_events*(struct tep_handle pass:[*]_tep_, enum tep_event_sort_type _sort_type_); +struct tep_event pass:[*]pass:[*]*tep_list_events_copy*(struct tep_handle pass:[*]_tep_, enum tep_event_sort_type _sort_type_); +-- + +DESCRIPTION +----------- +The _tep_list_events()_ function returns an array of pointers to the events, +sorted by the _sort_type_ criteria. The last element of the array is NULL. +The returned memory must not be freed, it is managed by the library. +The function is not thread safe. The _tep_ argument is trace event parser +context. The _sort_type_ argument is the required sort criteria: +[verse] +-- + _TEP_EVENT_SORT_ID_ - sort by the event ID. + _TEP_EVENT_SORT_NAME_ - sort by the event (name, system, id) triplet. + _TEP_EVENT_SORT_SYSTEM_ - sort by the event (system, name, id) triplet. +-- + +The _tep_list_events_copy()_ is a thread safe version of _tep_list_events()_. +It has the same behavior, but the returned array is allocated internally and +must be freed by the caller. Note that the content of the array must not be +freed (see the EXAMPLE below). + +RETURN VALUE +------------ +The _tep_list_events()_ function returns an array of pointers to events. +In case of an error, NULL is returned. The returned array must not be freed, +it is managed by the library. + +The _tep_list_events_copy()_ function returns an array of pointers to events. +In case of an error, NULL is returned. The returned array must be freed by +the caller. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +... +struct tep_handle *tep = tep_alloc(); +... +int i; +struct tep_event_format **events; + +i=0; +events = tep_list_events(tep, TEP_EVENT_SORT_ID); +if (events == NULL) { + /* Failed to get the events, sorted by ID */ +} else { + while(events[i]) { + /* walk through the list of the events, sorted by ID */ + i++; + } +} + +i=0; +events = tep_list_events_copy(tep, TEP_EVENT_SORT_NAME); +if (events == NULL) { + /* Failed to get the events, sorted by name */ +} else { + while(events[i]) { + /* walk through the list of the events, sorted by name */ + i++; + } + free(events); +} + +... +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 36/73 tools lib traceevent: Man pages for libtraceevent event get APIs
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man pages for libtraceevent APIs:
tep_get_event(),
tep_get_first_event(),
tep_get_events_count()
--- .../Documentation/libtraceevent-event_get.txt | 99 +++++++++++++++++++ 1 file changed, 99 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-event_get.txt
tools/lib/traceevent/Documentation/libtraceevent-event_get.txt new file +99 -0
@@ -0,0 +1,99 @@ +libtraceevent(3) +================ + +NAME +---- +tep_get_event, tep_get_first_event, tep_get_events_count - Access events. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +struct tep_event pass:[*]*tep_get_event*(struct tep_handle pass:[*]_tep_, int _index_); +struct tep_event pass:[*]*tep_get_first_event*(struct tep_handle pass:[*]_tep_); +int *tep_get_events_count*(struct tep_handle pass:[*]_tep_); +-- + +DESCRIPTION +----------- +The _tep_get_event()_ function returns a pointer to event at the given _index_. +The _tep_ argument is trace event parser context, the _index_ is the index of +the requested event. + +The _tep_get_first_event()_ function returns a pointer to the first event. +As events are stored in an array, this function returns the pointer to the +beginning of the array. The _tep_ argument is trace event parser context. + +The _tep_get_events_count()_ function returns the number of the events +in the array. The _tep_ argument is trace event parser context. + +RETURN VALUE +------------ +The _tep_get_event()_ returns a pointer to the event located at _index_. +NULL is returned in case of error, in case there are no events or _index_ is +out of range. + +The _tep_get_first_event()_ returns a pointer to the first event. NULL is +returned in case of error, or in case there are no events. + +The _tep_get_events_count()_ returns the number of the events. 0 is +returned in case of error, or in case there are no events. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +... +struct tep_handle *tep = tep_alloc(); +... +int i,count = tep_get_events_count(tep); +struct tep_event *event, *events = tep_get_first_event(tep); + +if (events == NULL) { + /* There are no events */ +} else { + for (i = 0; i < count; i++) { + event = (events+i); + /* process events[i] */ + } + + /* Get the last event */ + event = tep_get_event(tep, count-1); +} +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 37/73 tools lib traceevent: Man pages for find field APIs
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man pages for libtraceevent APIs:
tep_find_common_field(),
tep_find_field()
tep_find_any_field()
--- .../libtraceevent-field_find.txt | 118 ++++++++++++++++++ 1 file changed, 118 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-field_find.txt
tools/lib/traceevent/Documentation/libtraceevent-field_find.txt new file +118 -0
@@ -0,0 +1,118 @@ +libtraceevent(3) +================ + +NAME +---- +tep_find_common_field, tep_find_field, tep_find_any_field - +Search for a field in an event. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +struct tep_format_field pass:[*]*tep_find_common_field*(struct tep_event pass:[*]_event_, const char pass:[*]_name_); +struct tep_format_field pass:[*]*tep_find_field*(struct tep_event_ormat pass:[*]_event_, const char pass:[*]_name_); +struct tep_format_field pass:[*]*tep_find_any_field*(struct tep_event pass:[*]_event_, const char pass:[*]_name_); +-- + +DESCRIPTION +----------- +These functions search for a field with given name in an event. The field +returned can be used to find the field content from within a data record. + +The _tep_find_common_field()_ function searches for a common field with _name_ +in the _event_. + +The _tep_find_field()_ function searches for an event specific field with +_name_ in the _event_. + +The _tep_find_any_field()_ function searches for any field with _name_ in the +_event_. + +RETURN VALUE +------------ +The _tep_find_common_field(), _tep_find_field()_ and _tep_find_any_field()_ +functions return a pointer to the found field, or NULL in case there is no field +with the requested name. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +... +void get_htimer_info(struct tep_handle *tep, struct tep_record *record) +{ + struct tep_format_field *field; + struct tep_event *event; + long long softexpires; + int mode; + int pid; + + event = tep_find_event_by_name(tep, "timer", "hrtimer_start"); + + field = tep_find_common_field(event, "common_pid"); + if (field == NULL) { + /* Cannot find "common_pid" field in the event */ + } else { + /* Get pid from the data record */ + pid = tep_read_number(tep, record->data + field->offset, + field->size); + } + + field = tep_find_field(event, "softexpires"); + if (field == NULL) { + /* Cannot find "softexpires" event specific field in the event */ + } else { + /* Get softexpires parameter from the data record */ + softexpires = tep_read_number(tep, record->data + field->offset, + field->size); + } + + field = tep_find_any_field(event, "mode"); + if (field == NULL) { + /* Cannot find "mode" field in the event */ + } else + { + /* Get mode parameter from the data record */ + mode = tep_read_number(tep, record->data + field->offset, + field->size); + } +} +... +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 38/73 tools lib traceevent: Man pages for get field value APIs
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man pages for libtraceevent APIs:
tep_get_any_field_val(),
tep_get_common_field_val(),
tep_get_field_val(),
tep_get_field_raw()
--- .../libtraceevent-field_get_val.txt | 122 ++++++++++++++++++ 1 file changed, 122 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-field_get_val.txt
tools/lib/traceevent/Documentation/libtraceevent-field_get_val.txt new file +122 -0
@@ -0,0 +1,122 @@ +libtraceevent(3) +================ + +NAME +---- +tep_get_any_field_val, tep_get_common_field_val, tep_get_field_val, +tep_get_field_raw - Get value of a field. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* +*#include <trace-seq.h>* + +int *tep_get_any_field_val*(struct trace_seq pass:[*]_s_, struct tep_event pass:[*]_event_, const char pass:[*]_name_, struct tep_record pass:[*]_record_, unsigned long long pass:[*]_val_, int _err_); +int *tep_get_common_field_val*(struct trace_seq pass:[*]_s_, struct tep_event pass:[*]_event_, const char pass:[*]_name_, struct tep_record pass:[*]_record_, unsigned long long pass:[*]_val_, int _err_); +int *tep_get_field_val*(struct trace_seq pass:[*]_s_, struct tep_event pass:[*]_event_, const char pass:[*]_name_, struct tep_record pass:[*]_record_, unsigned long long pass:[*]_val_, int _err_); +void pass:[*]*tep_get_field_raw*(struct trace_seq pass:[*]_s_, struct tep_event pass:[*]_event_, const char pass:[*]_name_, struct tep_record pass:[*]_record_, int pass:[*]_len_, int _err_); +-- + +DESCRIPTION +----------- +These functions can be used to find a field and retrieve its value. + +The _tep_get_any_field_val()_ function searches in the _record_ for a field +with _name_, part of the _event_. If the field is found, its value is stored in +_val_. If there is an error and _err_ is not zero, then an error string is +written into _s_. + +The _tep_get_common_field_val()_ function does the same as +_tep_get_any_field_val()_, but searches only in the common fields. This works +for any event as all events include the common fields. + +The _tep_get_field_val()_ function does the same as _tep_get_any_field_val()_, +but searches only in the event specific fields. + +The _tep_get_field_raw()_ function searches in the _record_ for a field with +_name_, part of the _event_. If the field is found, a pointer to where the field +exists in the record's raw data is returned. The size of the data is stored in +_len_. If there is an error and _err_ is not zero, then an error string is +written into _s_. + +RETURN VALUE +------------ +The _tep_get_any_field_val()_, _tep_get_common_field_val()_ and +_tep_get_field_val()_ functions return 0 on success, or -1 in case of an error. + +The _tep_get_field_raw()_ function returns a pointer to field's raw data, and +places the length of this data in _len_. In case of an error NULL is returned. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +#include <trace-seq.h> +... +struct tep_handle *tep = tep_alloc(); +... +struct tep_event *event = tep_find_event_by_name(tep, "kvm", "kvm_exit"); +... +void process_record(struct tep_record *record) +{ + int len; + char *comm; + struct tep_event_format *event; + unsigned long long val; + + event = tep_find_event_by_record(pevent, record); + if (event != NULL) { + if (tep_get_common_field_val(NULL, event, "common_type", + record, &val, 0) == 0) { + /* Got the value of common type field */ + } + if (tep_get_field_val(NULL, event, "pid", record, &val, 0) == 0) { + /* Got the value of pid specific field */ + } + comm = tep_get_field_raw(NULL, event, "comm", record, &len, 0); + if (comm != NULL) { + /* Got a pointer to the comm event specific field */ + } + } +} +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*trace-seq.h* + Header file to include in order to have access to trace sequences + related APIs. Trace sequences are used to allow a function to call + several other functions to create a string of data to use. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 39/73 tools lib traceevent: Man pages for print field APIs
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man pages for libtraceevent APIs:
tep_print_field(),
tep_print_fields(),
tep_print_num_field(),
tep_print_func_field()
--- .../libtraceevent-field_print.txt | 126 ++++++++++++++++++ 1 file changed, 126 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-field_print.txt
tools/lib/traceevent/Documentation/libtraceevent-field_print.txt new file +126 -0
@@ -0,0 +1,126 @@ +libtraceevent(3) +================ + +NAME +---- +tep_print_field, tep_print_fields, tep_print_num_field, tep_print_func_field - +Print the field content. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* +*#include <trace-seq.h>* + +void *tep_print_field*(struct trace_seq pass:[*]_s_, void pass:[*]_data_, struct tep_format_field pass:[*]_field_); +void *tep_print_fields*(struct trace_seq pass:[*]_s_, void pass:[*]_data_, int _size_, struct tep_event pass:[*]_event_); +int *tep_print_num_field*(struct trace_seq pass:[*]_s_, const char pass:[*]_fmt_, struct tep_event pass:[*]_event_, const char pass:[*]_name_, struct tep_record pass:[*]_record_, int _err_); +int *tep_print_func_field*(struct trace_seq pass:[*]_s_, const char pass:[*]_fmt_, struct tep_event pass:[*]_event_, const char pass:[*]_name_, struct tep_record pass:[*]_record_, int _err_); +-- + +DESCRIPTION +----------- +These functions print recorded field's data, according to the field's type. + +The _tep_print_field()_ function extracts from the recorded raw _data_ value of +the _field_ and prints it into _s_, according to the field type. + +The _tep_print_fields()_ prints each field name followed by the record's field +value according to the field's type: +[verse] +-- +"field1_name=field1_value field2_name=field2_value ..." +-- +It iterates all fields of the _event_, and calls _tep_print_field()_ for each of +them. + +The _tep_print_num_field()_ function prints a numeric field with given format +string. A search is performed in the _event_ for a field with _name_. If such +field is found, its value is extracted from the _record_ and is printed in the +_s_, according to the given format string _fmt_. If the argument _err_ is +non-zero, and an error occures - it is printed in the _s_. + +The _tep_print_func_field()_ function prints a function field with given format +string. A search is performed in the _event_ for a field with _name_. If such +field is found, its value is extracted from the _record_. The value is assumed +to be a function address, and a search is perform to find the name of this +function. The function name (if found) and its address are printed in the _s_, +according to the given format string _fmt_. If the argument _err_ is non-zero, +and an error occures - it is printed in _s_. + +RETURN VALUE +------------ +The _tep_print_num_field()_ and _tep_print_func_field()_ functions return 1 +on success, -1 in case of an error or 0 if the print buffer _s_ is full. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +#include <trace-seq.h> +... +struct tep_handle *tep = tep_alloc(); +... +struct trace_seq seq; +trace_seq_init(&seq); +struct tep_event *event = tep_find_event_by_name(tep, "timer", "hrtimer_start"); +... +void process_record(struct tep_record *record) +{ + struct tep_format_field *field_pid = tep_find_common_field(event, "common_pid"); + + trace_seq_reset(&seq); + + /* Print the value of "common_pid" */ + tep_print_field(&seq, record->data, field_pid); + + /* Print all fields of the "hrtimer_start" event */ + tep_print_fields(&seq, record->data, record->size, event); + + /* Print the value of "expires" field with custom format string */ + tep_print_num_field(&seq, " timer expires in %llu ", event, "expires", record, 0); + + /* Print the address and the name of "function" field with custom format string */ + tep_print_func_field(&seq, " timer function is %s ", event, "function", record, 0); + } + ... +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*trace-seq.h* + Header file to include in order to have access to trace sequences related APIs. + Trace sequences are used to allow a function to call several other functions + to create a string of data to use. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 40/73 tools lib traceevent: Man page for tep_read_number_field()
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man page for libtraceevent API tep_read_number_field().
--- .../libtraceevent-field_read.txt | 81 +++++++++++++++++++ 1 file changed, 81 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-field_read.txt
tools/lib/traceevent/Documentation/libtraceevent-field_read.txt new file +81 -0
@@ -0,0 +1,81 @@ +libtraceevent(3) +================ + +NAME +---- +tep_read_number_field - Reads a number from raw data. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +int *tep_read_number_field*(struct tep_format_field pass:[*]_field_, const void pass:[*]_data_, unsigned long long pass:[*]_value_); +-- + +DESCRIPTION +----------- +The _tep_read_number_field()_ function reads the value of the _field_ from the +raw _data_ and stores it in the _value_. The function sets the _value_ according +to the endianness of the raw data and the current machine and stores it in +_value_. + +RETURN VALUE +------------ +The _tep_read_number_field()_ function retunrs 0 in case of success, or -1 in +case of an error. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +... +struct tep_handle *tep = tep_alloc(); +... +struct tep_event *event = tep_find_event_by_name(tep, "timer", "hrtimer_start"); +... +void process_record(struct tep_record *record) +{ + unsigned long long pid; + struct tep_format_field *field_pid = tep_find_common_field(event, "common_pid"); + + if (tep_read_number_field(field_pid, record->data, &pid) != 0) { + /* Failed to get "common_pid" value */ + } +} +... +-- +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 41/73 tools lib traceevent: Man pages for event fields APIs
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man pages for libtraceevent APIs:
tep_event_common_fields(),
tep_event_fields()
--- .../Documentation/libtraceevent-fields.txt | 105 ++++++++++++++++++ 1 file changed, 105 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-fields.txt
tools/lib/traceevent/Documentation/libtraceevent-fields.txt new file +105 -0
@@ -0,0 +1,105 @@ +libtraceevent(3) +================ + +NAME +---- +tep_event_common_fields, tep_event_fields - Get a list of fields for an event. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +struct tep_format_field pass:[*]pass:[*]*tep_event_common_fields*(struct tep_event pass:[*]_event_); +struct tep_format_field pass:[*]pass:[*]*tep_event_fields*(struct tep_event pass:[*]_event_); +-- + +DESCRIPTION +----------- +The _tep_event_common_fields()_ function returns an array of pointers to common +fields for the _event_. The array is allocated in the function and must be freed +by free(). The last element of the array is NULL. + +The _tep_event_fields()_ function returns an array of pointers to event specific +fields for the _event_. The array is allocated in the function and must be freed +by free(). The last element of the array is NULL. + +RETURN VALUE +------------ +Both _tep_event_common_fields()_ and _tep_event_fields()_ functions return +an array of pointers to tep_format_field structures in case of success, or +NULL in case of an error. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +... +struct tep_handle *tep = tep_alloc(); +... +int i; +struct tep_format_field **fields; +struct tep_event *event = tep_find_event_by_name(tep, "kvm", "kvm_exit"); +if (event != NULL) { + fields = tep_event_common_fields(event); + if (fields != NULL) { + i = 0; + while (fields[i]) { + /* + walk through the list of the common fields + of the kvm_exit event + */ + i++; + } + free(fields); + } + fields = tep_event_fields(event); + if (fields != NULL) { + i = 0; + while (fields[i]) { + /* + walk through the list of the event specific + fields of the kvm_exit event + */ + i++; + } + free(fields); + } +} +... +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 42/73 tools lib traceevent: Man pages for event filter APIs
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Added new man pages, describing libtraceevent event filter APIs:
tep_filter_alloc()
tep_filter_free()
tep_filter_reset()
tep_filter_add_filter_str()
tep_filter_strerror()
tep_event_filtered()
tep_filter_remove_event()
tep_filter_match()
tep_filter_copy()
tep_filter_compare()
tep_filter_make_string()
--- .../Documentation/libtraceevent-filter.txt | 209 ++++++++++++++++++ 1 file changed, 209 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-filter.txt
tools/lib/traceevent/Documentation/libtraceevent-filter.txt new file +209 -0
@@ -0,0 +1,209 @@ +libtraceevent(3) +================ + +NAME +---- +tep_filter_alloc, tep_filter_free, tep_filter_reset, tep_filter_make_string, +tep_filter_copy, tep_filter_compare, tep_filter_match, tep_event_filtered, +tep_filter_remove_event, tep_filter_strerror, tep_filter_add_filter_str - +Event filter related APIs. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +struct tep_event_filter pass:[*]*tep_filter_alloc*(struct tep_handle pass:[*]_tep_); +void *tep_filter_free*(struct tep_event_filter pass:[*]_filter_); +void *tep_filter_reset*(struct tep_event_filter pass:[*]_filter_); +enum tep_errno *tep_filter_add_filter_str*(struct tep_event_filter pass:[*]_filter_, const char pass:[*]_filter_str_); +int *tep_event_filtered*(struct tep_event_filter pass:[*]_filter_, int _event_id_); +int *tep_filter_remove_event*(struct tep_event_filter pass:[*]_filter_, int _event_id_); +enum tep_errno *tep_filter_match*(struct tep_event_filter pass:[*]_filter_, struct tep_record pass:[*]_record_); +int *tep_filter_copy*(struct tep_event_filter pass:[*]_dest_, struct tep_event_filter pass:[*]_source_); +int *tep_filter_compare*(struct tep_event_filter pass:[*]_filter1_, struct tep_event_filter pass:[*]_filter2_); +char pass:[*]*tep_filter_make_string*(struct tep_event_filter pass:[*]_filter_, int _event_id_); +int *tep_filter_strerror*(struct tep_event_filter pass:[*]_filter_, enum tep_errno _err_, char pass:[*]buf, size_t _buflen_); +-- + +DESCRIPTION +----------- +Filters can be attached to traced events. They can be used to filter out various +events when outputting them. Each event can be filtered based on its parameters, +described in the event's format file. This set of functions can be used to +create, delete, modify and attach event filters. + +The _tep_filter_alloc()_ function creates a new event filter. The _tep_ argument +is the trace event parser context. + +The _tep_filter_free()_ function frees an event filter and all resources that it +had used. + +The _tep_filter_reset()_ function removes all rules from an event filter and +resets it. + +The _tep_filter_add_filter_str()_ function adds a new rule to the _filter_. The +_filter_str_ argument is the filter string, that contains the rule. + +The _tep_event_filtered()_ function checks if the event with _event_id_ has +_filter_. + +The _tep_filter_remove_event()_ function removes a _filter_ for an event with +_event_id_. + +The _tep_filter_match()_ function tests if a _record_ matches given _filter_. + +The _tep_filter_copy()_ function copies a _source_ filter into a _dest_ filter. + +The _tep_filter_compare()_ function compares two filers - _filter1_ and _filter2_. + +The _tep_filter_make_string()_ function constructs a string, displaying +the _filter_ contents for given _event_id_. + +The _tep_filter_strerror()_ function copies the _filter_ error buffer into the +given _buf_ with the size _buflen_. If the error buffer is empty, in the _buf_ +is copied a string, describing the error _err_. + +RETURN VALUE +------------ +The _tep_filter_alloc()_ function returns a pointer to the newly created event +filter, or NULL in case of an error. + +The _tep_filter_add_filter_str()_ function returns 0 if the rule was +successfully added or a negative error code. Use _tep_filter_strerror()_ to see +actual error message in case of an error. + +The _tep_event_filtered()_ function returns 1 if the filter is found for given +event, or 0 otherwise. + +The _tep_filter_remove_event()_ function returns 1 if the vent was removed, or +0 if the event was not found. + +The _tep_filter_match()_ function returns _tep_errno_, according to the result: +[verse] +-- +_pass:[TEP_ERRNO__FILTER_MATCH]_ - filter found for event, the record matches. +_pass:[TEP_ERRNO__FILTER_MISS]_ - filter found for event, the record does not match. +_pass:[TEP_ERRNO__FILTER_NOT_FOUND]_ - no filter found for record's event. +_pass:[TEP_ERRNO__NO_FILTER]_ - no rules in the filter. +-- +or any other _tep_errno_, if an error occurred during the test. + +The _tep_filter_copy()_ function returns 0 on success or -1 if not all rules + were copied. + +The _tep_filter_compare()_ function returns 1 if the two filters hold the same +content, or 0 if they do not. + +The _tep_filter_make_string()_ function returns a string, which must be freed +with free(), or NULL in case of an error. + +The _tep_filter_strerror()_ function returns 0 if message was filled +successfully, or -1 in case of an error. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +... +struct tep_handle *tep = tep_alloc(); +... +char errstr[200]; +int ret; + +struct tep_event_filter *filter = tep_filter_alloc(tep); +struct tep_event_filter *filter1 = tep_filter_alloc(tep); +ret = tep_filter_add_filter_str(filter, "sched/sched_wakeup:target_cpu==1"); +if(ret < 0) { + tep_filter_strerror(filter, ret, errstr, sizeof(errstr)); + /* Failed to add a new rule to the filter, the error string is in errstr */ +} +if (tep_filter_copy(filter1, filter) != 0) { + /* Failed to copy filter in filter1 */ +} +... +if (tep_filter_compare(filter, filter1) != 1) { + /* Both filters are different */ +} +... +void process_record(struct tep_handle *tep, struct tep_record *record) +{ + struct tep_event *event; + char *fstring; + + event = tep_find_event_by_record(tep, record); + + if (tep_event_filtered(filter, event->id) == 1) { + /* The event has filter */ + fstring = tep_filter_make_string(filter, event->id); + if (fstring != NULL) { + /* The filter for the event is in fstring */ + free(fstring); + } + } + + switch (tep_filter_match(filter, record)) { + case TEP_ERRNO__FILTER_MATCH: + /* The filter matches the record */ + break; + case TEP_ERRNO__FILTER_MISS: + /* The filter does not match the record */ + break; + case TEP_ERRNO__FILTER_NOT_FOUND: + /* No filter found for record's event */ + break; + case TEP_ERRNO__NO_FILTER: + /* There are no rules in the filter */ + break + default: + /* An error occurred during the test */ + break; + } + + if (tep_filter_remove_event(filter, event->id) == 1) { + /* The event was removed from the filter */ + } +} + +... +tep_filter_reset(filter); +... +tep_filter_free(filter); +tep_filter_free(filter1); +... +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 43/73 tools lib traceevent: Man pages for parse event APIs
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man pages for libtraceevent APIs:
tep_parse_event(),
tep_parse_format()
--- .../libtraceevent-parse_event.txt | 90 +++++++++++++++++++ 1 file changed, 90 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-parse_event.txt
tools/lib/traceevent/Documentation/libtraceevent-parse_event.txt new file +90 -0
@@ -0,0 +1,90 @@ +libtraceevent(3) +================ + +NAME +---- +tep_parse_event, tep_parse_format - Parse the event format information + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +enum tep_errno *tep_parse_event*(struct tep_handle pass:[*]_tep_, const char pass:[*]_buf_, unsigned long _size_, const char pass:[*]_sys_); +enum tep_errno *tep_parse_format*(struct tep_handle pass:[*]_tep_, struct tep_event pass:[*]pass:[*]_eventp_, const char pass:[*]_buf_, unsigned long _size_, const char pass:[*]_sys_); +-- + +DESCRIPTION +----------- +The _tep_parse_event()_ function parses the event format and creates an event +structure to quickly parse raw data for a given event. The _tep_ argument is +the trace event parser context. The created event structure is stored in the +_tep_ context. The _buf_ argument is a buffer with _size_, where the event +format data is. The event format data can be taken from +tracefs/events/.../.../format files. The _sys_ argument is the system of +the event. + +The _tep_parse_format()_ function does the same as _tep_parse_event()_. The only +difference is in the extra _eventp_ argument, where the newly created event +structure is returned. + +RETURN VALUE +------------ +Both _tep_parse_event()_ and _tep_parse_format()_ functions return 0 on success, +or TEP_ERRNO__... in case of an error. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +... +struct tep_handle *tep = tep_alloc(); +... +char *buf; +int size; +struct tep_event *event = NULL; +buf = read_file("/sys/kernel/tracing/events/ftrace/print/format", &size); +if (tep_parse_event(tep, buf, size, "ftrace") != 0) { + /* Failed to parse the ftrace print format */ +} + +if (tep_parse_format(tep, &event, buf, size, "ftrace") != 0) { + /* Failed to parse the ftrace print format */ +} +... +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 44/73 tools lib traceevent: Man page for tep_parse_header_page()
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man page for tep_parse_header_page() libtraceevent API.
--- .../libtraceevent-parse_head.txt | 82 +++++++++++++++++++ 1 file changed, 82 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-parse_head.txt
tools/lib/traceevent/Documentation/libtraceevent-parse_head.txt new file +82 -0
@@ -0,0 +1,82 @@ +libtraceevent(3) +================ + +NAME +---- +tep_parse_header_page - Parses the data stored in the header page. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +int *tep_parse_header_page*(struct tep_handle pass:[*]_tep_, char pass:[*]_buf_, unsigned long _size_, int _long_size_); +-- + +DESCRIPTION +----------- +The _tep_parse_header_page()_ function parses the header page data from _buf_, +and initializes the _tep_, trace event parser context, with it. The buffer +_buf_ is with _size_, and is supposed to be copied from +tracefs/events/header_page. + +Some old kernels do not have header page info, in this case the +_tep_parse_header_page()_ function can be called with _size_ equal to 0. The +_tep_ context is initialized with default values. The _long_size_ can be used in +this use case, to set the size of a long integer to be used. + +RETURN VALUE +------------ +The _tep_parse_header_page()_ function returns 0 in case of success, or -1 +in case of an error. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +... +struct tep_handle *tep = tep_alloc(); +... +char *buf; +int size; +buf = read_file("/sys/kernel/tracing/events/header_page", &size); +if (tep_parse_header_page(tep, buf, size, sizeof(unsigned long)) != 0) { + /* Failed to parse the header page */ +} +... +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 45/73 tools lib traceevent: Man pages for APIs used to extract common fields from a record
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man pages for libtraceevent APIs:
tep_data_type(),
tep_data_pid(),
tep_data_preempt_count(),
tep_data_flags()
--- .../libtraceevent-record_parse.txt | 137 ++++++++++++++++++ 1 file changed, 137 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-record_parse.txt
tools/lib/traceevent/Documentation/libtraceevent-record_parse.txt new file +137 -0
@@ -0,0 +1,137 @@ +libtraceevent(3) +================ + +NAME +---- +tep_data_type, tep_data_pid,tep_data_preempt_count, tep_data_flags - +Extract common fields from a record. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* + +enum *trace_flag_type* { + _TRACE_FLAG_IRQS_OFF_, + _TRACE_FLAG_IRQS_NOSUPPORT_, + _TRACE_FLAG_NEED_RESCHED_, + _TRACE_FLAG_HARDIRQ_, + _TRACE_FLAG_SOFTIRQ_, +}; + +int *tep_data_type*(struct tep_handle pass:[*]_tep_, struct tep_record pass:[*]_rec_); +int *tep_data_pid*(struct tep_handle pass:[*]_tep_, struct tep_record pass:[*]_rec_); +int *tep_data_preempt_count*(struct tep_handle pass:[*]_tep_, struct tep_record pass:[*]_rec_); +int *tep_data_flags*(struct tep_handle pass:[*]_tep_, struct tep_record pass:[*]_rec_); +-- + +DESCRIPTION +----------- +This set of functions can be used to extract common fields from a record. + +The _tep_data_type()_ function gets the event id from the record _rec_. +It reads the "common_type" field. The _tep_ argument is the trace event parser +context. + +The _tep_data_pid()_ function gets the process id from the record _rec_. +It reads the "common_pid" field. The _tep_ argument is the trace event parser +context. + +The _tep_data_preempt_count()_ function gets the preemption count from the +record _rec_. It reads the "common_preempt_count" field. The _tep_ argument is +the trace event parser context. + +The _tep_data_flags()_ function gets the latency flags from the record _rec_. +It reads the "common_flags" field. The _tep_ argument is the trace event parser +context. Supported latency flags are: +[verse] +-- + _TRACE_FLAG_IRQS_OFF_, Interrupts are disabled. + _TRACE_FLAG_IRQS_NOSUPPORT_, Reading IRQ flag is not supported by the architecture. + _TRACE_FLAG_NEED_RESCHED_, Task needs rescheduling. + _TRACE_FLAG_HARDIRQ_, Hard IRQ is running. + _TRACE_FLAG_SOFTIRQ_, Soft IRQ is running. +-- + +RETURN VALUE +------------ +The _tep_data_type()_ function returns an integer, representing the event id. + +The _tep_data_pid()_ function returns an integer, representing the process id + +The _tep_data_preempt_count()_ function returns an integer, representing the +preemption count. + +The _tep_data_flags()_ function returns an integer, representing the latency +flags. Look at the _trace_flag_type_ enum for supported flags. + +All these functions in case of an error return a negative integer. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +... +struct tep_handle *tep = tep_alloc(); +... +void process_record(struct tep_record *record) +{ + int data; + + data = tep_data_type(tep, record); + if (data >= 0) { + /* Got the ID of the event */ + } + + data = tep_data_pid(tep, record); + if (data >= 0) { + /* Got the process ID */ + } + + data = tep_data_preempt_count(tep, record); + if (data >= 0) { + /* Got the preemption count */ + } + + data = tep_data_flags(tep, record); + if (data >= 0) { + /* Got the latency flags */ + } +} +... +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 46/73 tools lib traceevent: Man pages for trace sequences APIs
In reply to Arnaldo Carvalho de MeloFrom: Tzvetomir Stoyanov <[email protected]>
Create man pages for trace sequences libtraceevent APIs:
trace_seq_init(),
trace_seq_destroy(),
trace_seq_reset(),
trace_seq_terminate(),
trace_seq_putc(),
trace_seq_puts(),
trace_seq_printf(),
trace_seq_vprintf(),
trace_seq_do_fprintf(),
trace_seq_do_printf()
--- .../Documentation/libtraceevent-tseq.txt | 158 ++++++++++++++++++ 1 file changed, 158 insertions(+) create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-tseq.txt
tools/lib/traceevent/Documentation/libtraceevent-tseq.txt new file +158 -0
@@ -0,0 +1,158 @@ +libtraceevent(3) +================ + +NAME +---- +trace_seq_init, trace_seq_destroy, trace_seq_reset, trace_seq_terminate, +trace_seq_putc, trace_seq_puts, trace_seq_printf, trace_seq_vprintf, +trace_seq_do_fprintf, trace_seq_do_printf - +Initialize / destroy a trace sequence. + +SYNOPSIS +-------- +[verse] +-- +*#include <event-parse.h>* +*#include <trace-seq.h>* + +void *trace_seq_init*(struct trace_seq pass:[*]_s_); +void *trace_seq_destroy*(struct trace_seq pass:[*]_s_); +void *trace_seq_reset*(struct trace_seq pass:[*]_s_); +void *trace_seq_terminate*(struct trace_seq pass:[*]_s_); +int *trace_seq_putc*(struct trace_seq pass:[*]_s_, unsigned char _c_); +int *trace_seq_puts*(struct trace_seq pass:[*]_s_, const char pass:[*]_str_); +int *trace_seq_printf*(struct trace_seq pass:[*]_s_, const char pass:[*]_fmt_, _..._); +int *trace_seq_vprintf*(struct trace_seq pass:[*]_s_, const char pass:[*]_fmt_, va_list _args_); +int *trace_seq_do_printf*(struct trace_seq pass:[*]_s_); +int *trace_seq_do_fprintf*(struct trace_seq pass:[*]_s_, FILE pass:[*]_fp_); +-- + +DESCRIPTION +----------- +Trace sequences are used to allow a function to call several other functions +to create a string of data to use. + +The _trace_seq_init()_ function initializes the trace sequence _s_. + +The _trace_seq_destroy()_ function destroys the trace sequence _s_ and frees +all its resources that it had used. + +The _trace_seq_reset()_ function re-initializes the trace sequence _s_. All +characters already written in _s_ will be deleted. + +The _trace_seq_terminate()_ function terminates the trace sequence _s_. It puts +the null character pass:['\0'] at the end of the buffer. + +The _trace_seq_putc()_ function puts a single character _c_ in the trace +sequence _s_. + +The _trace_seq_puts()_ function puts a NULL terminated string _str_ in the +trace sequence _s_. + +The _trace_seq_printf()_ function puts a formated string _fmt _with +variable arguments _..._ in the trace sequence _s_. + +The _trace_seq_vprintf()_ function puts a formated string _fmt _with +list of arguments _args_ in the trace sequence _s_. + +The _trace_seq_do_printf()_ function prints the buffer of trace sequence _s_ to +the standard output stdout. + +The _trace_seq_do_fprintf()_ function prints the buffer of trace sequence _s_ +to the given file _fp_. + +RETURN VALUE +------------ +Both _trace_seq_putc()_ and _trace_seq_puts()_ functions return the number of +characters put in the trace sequence, or 0 in case of an error + +Both _trace_seq_printf()_ and _trace_seq_vprintf()_ functions return 0 if the +trace oversizes the buffer's free space, the number of characters printed, or +a negative value in case of an error. + +Both _trace_seq_do_printf()_ and _trace_seq_do_fprintf()_ functions return the +number of printed characters, or -1 in case of an error. + +EXAMPLE +------- +[source,c] +-- +#include <event-parse.h> +#include <trace-seq.h> +... +struct trace_seq seq; +trace_seq_init(&seq); +... +void foo_seq_print(struct trace_seq *tseq, char *format, ...) +{ + va_list ap; + va_start(ap, format); + if (trace_seq_vprintf(tseq, format, ap) <= 0) { + /* Failed to print in the trace sequence */ + } + va_end(ap); +} + +trace_seq_reset(&seq); + +char *str = " MAN page example"; +if (trace_seq_puts(&seq, str) != strlen(str)) { + /* Failed to put str in the trace sequence */ +} +if (trace_seq_putc(&seq, ':') != 1) { + /* Failed to put ':' in the trace sequence */ +} +if (trace_seq_printf(&seq, " trace sequence: %d", 1) <= 0) { + /* Failed to print in the trace sequence */ +} +foo_seq_print( &seq, " %d\n", 2); + +trace_seq_terminate(&seq); +... + +if (trace_seq_do_printf(&seq) < 0 ) { + /* Failed to print the sequence buffer to the standard output */ +} +FILE *fp = fopen("trace.txt", "w"); +if (trace_seq_do_fprintf(&seq, fp) < 0 ) [ + /* Failed to print the sequence buffer to the trace.txt file */ +} + +trace_seq_destroy(&seq); +... +-- + +FILES +----- +[verse] +-- +*event-parse.h* + Header file to include in order to have access to the library APIs. +*trace-seq.h* + Header file to include in order to have access to trace sequences related APIs. +*-ltraceevent* + Linker switch to add when building a program that uses the library. +-- + +SEE ALSO +-------- +_libtraceevent(3)_, _trace-cmd(1)_ + +AUTHOR +------ +[verse] +-- +*Steven Rostedt* <[email protected]>, author of *libtraceevent*. +*Tzvetomir Stoyanov* <[email protected]>, author of this man page. +-- +REPORTING BUGS +-------------- +Report bugs to <[email protected]> + +LICENSE +------- +libtraceevent is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 47/73 perf build tests: Add NO_LIBZSTD=1 to make_minimal
In reply to Arnaldo Carvalho de MeloFrom: Arnaldo Carvalho de Melo <[email protected]>
So that we can test the ifdef parts for this feature.
--- tools/perf/tests/make | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/tests/make +1 -1
@@ -107,7 +107,7 @@ make_minimal := NO_LIBPERL=1 NO_LIBPYTHON=1 NO_NEWT=1 NO_GTK2=1 make_minimal += NO_DEMANGLE=1 NO_LIBELF=1 NO_LIBUNWIND=1 NO_BACKTRACE=1 make_minimal += NO_LIBNUMA=1 NO_LIBAUDIT=1 NO_LIBBIONIC=1 make_minimal += NO_LIBDW_DWARF_UNWIND=1 NO_AUXTRACE=1 NO_LIBBPF=1 -make_minimal += NO_LIBCRYPTO=1 NO_SDT=1 NO_JVMTI=1 +make_minimal += NO_LIBCRYPTO=1 NO_SDT=1 NO_JVMTI=1 NO_LIBZSTD=1 # $(run) contains all available tests run := make_pure
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 48/73 perf session: Define 'bytes_transferred' and 'bytes_compressed' metrics
In reply to Arnaldo Carvalho de MeloFrom: Alexey Budankov <[email protected]>
Define 'bytes_transferred' and 'bytes_compressed' metrics to calculate ratio in the end of the data collection:
compression ratio = bytes_transferred / bytes_compressed
The 'bytes_transferred' metric accumulates the amount of bytes that was extracted from the mmaped kernel buffers for compression, while 'bytes_compressed' accumulates the amount of bytes that was received after applying compression.
--- tools/perf/builtin-record.c | 14 +++++++++++++- tools/perf/util/env.h | 1 + tools/perf/util/session.h | 2 ++ 3 files changed, 16 insertions(+), 1 deletion(-)
tools/perf/builtin-record.c +13 -1
@@ -1186,6 +1186,7 @@ static int __cmd_record(struct record *rec, int argc, const char **argv)bool disabled = false, draining = false;struct perf_evlist *sb_evlist =NULL ;int fd; +float ratio =0 ;atexit (record__sig_exit);signal (SIGCHLD , sig_handler);
@@ -1491,6 +1492,11 @@ static int __cmd_record(struct record *rec, int argc, const char **argv)record__mmap_read_all (rec, true);record__aio_mmap_read_sync (rec); +if (rec->session->bytes_transferred && rec->session->bytes_compressed) { + ratio = (float )rec->session->bytes_transferred/(float )rec->session->bytes_compressed; + session->header.env.comp_ratio = ratio +0.5 ; + } +if (forks) {int exit_status;
@@ -1537,9 +1543,15 @@ static int __cmd_record(struct record *rec, int argc, const char **argv)else samples[0 ] ='\0' ; -fprintf (stderr,"[ perf record: Captured and wrote %.3f MB %s%s%s ]\n" , +fprintf (stderr,"[ perf record: Captured and wrote %.3f MB %s%s%s" ,perf_data__size (data) /1024.0 /1024.0 , data->path, postfix, samples); +if (ratio) { +fprintf (stderr,", compressed (original %.3f MB, ratio is %.3f)" , + rec->session->bytes_transferred /1024.0 /1024.0 , + ratio); + } +fprintf (stderr," ]\n" ); } out_delete_session:
tools/perf/util/env.h +1 -0
@@ -62,6 +62,7 @@ struct perf_env {struct cpu_topology_map *cpu;struct cpu_cache_level *caches;int caches_cnt; +u32 comp_ratio;struct numa_node *numa_nodes;struct memory_node *memory_nodes;unsigned long long memory_bsize;
tools/perf/util/session.h +2 -0
@@ -35,6 +35,8 @@ struct perf_session {struct ordered_events ordered_events;struct perf_data *data;struct perf_tool *tool; +u64 bytes_transferred; +u64 bytes_compressed; };struct perf_tool ;
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 49/73 perf record: Implement COMPRESSED event record and its attributes
In reply to Arnaldo Carvalho de MeloFrom: Alexey Budankov <[email protected]>
Implemented PERF_RECORD_COMPRESSED event, related data types, header feature and functions to write, read and print feature attributes from the trace header section.
comp_mmap_len preserves the size of mmaped kernel buffer that was used during collection. comp_mmap_len size is used on loading stage as the size of decomp buffer for decompression of COMPRESSED events content.
Committer notes:
Fixed up conflict with BPF_PROG_INFO and BTF_BTF header features.
--- .../Documentation/perf.data-file-format.txt | 24 +++++++++ tools/perf/builtin-record.c | 8 +++ tools/perf/perf.h | 1 + tools/perf/util/env.h | 10 ++++ tools/perf/util/event.c | 1 + tools/perf/util/event.h | 7 +++ tools/perf/util/header.c | 53 +++++++++++++++++++ tools/perf/util/header.h | 1 + 8 files changed, 105 insertions(+)
tools/perf/Documentation/perf.data-file-format.txt +24 -0
@@ -272,6 +272,19 @@ struct { Two uint64_t for the time of first sample and the time of last sample. + HEADER_COMPRESSED = 27, + +struct { + u32 version; + u32 type; + u32 level; + u32 ratio; + u32 mmap_len; +}; + +Indicates that trace contains records of PERF_RECORD_COMPRESSED type +that have perf_events records in compressed form. + other bits are reserved and should ignored for now HEADER_FEAT_BITS = 256,
@@ -437,6 +450,17 @@ struct auxtrace_error_event { Describes a header feature. These are records used in pipe-mode that contain information that otherwise would be in perf.data file's header. + PERF_RECORD_COMPRESSED = 81, + +struct compressed_event { + struct perf_event_header header; + char data[]; +}; + +The header is followed by compressed data frame that can be decompressed +into array of perf trace records. The size of the entire compressed event +record including the header is limited by the max value of header.size. + Event types Define the event attributes with their IDs.
tools/perf/builtin-record.c +8 -0
@@ -372,6 +372,11 @@ static int record__mmap_flush_parse(const struct option *opt,return 0 ; } +static int record__comp_enabled (struct record *rec) +{ +return rec->opts.comp_level >0 ; +} + static int process_synthesized_event(struct perf_tool *tool, union perf_event *event, struct perf_sample *sample __maybe_unused,
@@ -888,6 +893,8 @@ static void record__init_features(struct record *rec)perf_header__clear_feat (&session->header,HEADER_CLOCKID );perf_header__clear_feat (&session->header,HEADER_DIR_FORMAT ); +if (!record__comp_enabled (rec)) +perf_header__clear_feat (&session->header,HEADER_COMPRESSED );perf_header__clear_feat (&session->header,HEADER_STAT ); }
@@ -1245,6 +1252,7 @@ static int __cmd_record(struct record *rec, int argc, const char **argv) err =-1 ; goto out_child; } + session->header.env.comp_mmap_len = session->evlist->mmap_len; err =bpf__apply_obj_config ();if (err) {
tools/perf/perf.h +1 -0
@@ -86,6 +86,7 @@ struct record_opts {int nr_cblocks;int affinity;int mmap_flush; +unsigned int comp_level; };enum perf_affinity {
tools/perf/util/env.h +10 -0
@@ -63,6 +63,10 @@ struct perf_env {struct cpu_cache_level *caches;int caches_cnt;u32 comp_ratio; +u32 comp_ver; +u32 comp_type; +u32 comp_level; +u32 comp_mmap_len;struct numa_node *numa_nodes;struct memory_node *memory_nodes;unsigned long long memory_bsize;
@@ -81,6 +85,12 @@ struct perf_env { } bpf_progs; }; +enum perf_compress_type { +PERF_COMP_NONE =0 , +PERF_COMP_ZSTD , +PERF_COMP_MAX +}; +struct bpf_prog_info_node ;struct btf_node ;
tools/perf/util/event.c +1 -0
@@ -68,6 +68,7 @@ static const char *perf_event__names[] = { [PERF_RECORD_EVENT_UPDATE] = "EVENT_UPDATE", [PERF_RECORD_TIME_CONV] = "TIME_CONV", [PERF_RECORD_HEADER_FEATURE] = "FEATURE", + [PERF_RECORD_COMPRESSED] = "COMPRESSED", };static const char *perf_ns__names[] = {
tools/perf/util/event.h +7 -0
@@ -255,6 +255,7 @@ enum perf_user_event_type { /* above any possible kernel type */PERF_RECORD_EVENT_UPDATE =78 ,PERF_RECORD_TIME_CONV =79 ,PERF_RECORD_HEADER_FEATURE =80 , +PERF_RECORD_COMPRESSED =81 , PERF_RECORD_HEADER_MAX };
@@ -627,6 +628,11 @@ struct feature_event {char data[]; }; +struct compressed_event { +struct perf_event_header header; +char data[]; +}; +union perf_event {struct perf_event_header header;struct mmap_event mmap;
@@ -660,6 +666,7 @@ union perf_event {struct feature_event feat;struct ksymbol_event ksymbol_event;struct bpf_event bpf_event; +struct compressed_event pack; };void perf_event__print_totals (void );
tools/perf/util/header.c +53 -0
77 lines not on this page · raw message
tools/perf/util/header.h +1 -0
@@ -42,6 +42,7 @@ enum {HEADER_DIR_FORMAT ,HEADER_BPF_PROG_INFO ,HEADER_BPF_BTF , +HEADER_COMPRESSED ,HEADER_LAST_FEATURE , HEADER_FEAT_BITS = 256, };
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 50/73 perf mmap: Implement dedicated memory buffer for data compression
In reply to Arnaldo Carvalho de MeloFrom: Alexey Budankov <[email protected]>
Implemented mmap data buffer that is used as the memory to operate on when compressing data in case of serial trace streaming.
--- tools/perf/builtin-record.c | 8 +++++++- tools/perf/util/evlist.c | 8 +++++--- tools/perf/util/evlist.h | 2 +- tools/perf/util/mmap.c | 30 ++++++++++++++++++++++++++++-- tools/perf/util/mmap.h | 4 +++- 5 files changed, 44 insertions(+), 8 deletions(-)
tools/perf/builtin-record.c +7 -1
@@ -372,6 +372,8 @@ static int record__mmap_flush_parse(const struct option *opt,return 0 ; } +static unsigned int comp_level_max =22 ; +static int record__comp_enabled (struct record *rec) {return rec->opts.comp_level >0 ;
@@ -587,7 +589,7 @@ static int record__mmap_evlist(struct record *rec, opts->auxtrace_mmap_pages, opts->auxtrace_snapshot_mode, opts->nr_cblocks, opts->affinity, - opts->mmap_flush) < 0) { + opts->mmap_flush, opts->comp_level) < 0) {if (errno ==EPERM ) {pr_err ("Permission error mapping pages.\n" "Consider increasing "
@@ -2298,6 +2300,10 @@ int cmd_record(int argc, const char **argv)pr_debug ("affinity: %s\n" , affinity_tags[rec->opts.affinity]);pr_debug ("mmap flush: %d\n" , rec->opts.mmap_flush); +if (rec->opts.comp_level > comp_level_max) + rec->opts.comp_level = comp_level_max; +pr_debug ("comp level: %d\n" , rec->opts.comp_level); + err =__cmd_record (&record, argc, argv); out:perf_evlist__delete (rec->evlist);
tools/perf/util/evlist.c +5 -3
@@ -1009,7 +1009,8 @@ int perf_evlist__parse_mmap_pages(const struct option *opt, const char *str,*/ int perf_evlist__mmap_ex (struct perf_evlist *evlist,unsigned int pages,unsigned int auxtrace_pages, -bool auxtrace_overwrite,int nr_cblocks,int affinity,int flush) +bool auxtrace_overwrite,int nr_cblocks,int affinity,int flush, +int comp_level) {struct perf_evsel *evsel;const struct cpu_map *cpus = evlist->cpus;
@@ -1019,7 +1020,8 @@ int perf_evlist__mmap_ex(struct perf_evlist *evlist, unsigned int pages,* Its value is decided by evsel's write_backward. * So &mp should not be passed through const pointer. */ -struct mmap_params mp = { .nr_cblocks = nr_cblocks, .affinity = affinity, .flush = flush }; +struct mmap_params mp = { .nr_cblocks = nr_cblocks, .affinity = affinity, .flush = flush, + .comp_level = comp_level };if (!evlist->mmap) evlist->mmap =perf_evlist__alloc_mmap (evlist, false);
@@ -1051,7 +1053,7 @@ int perf_evlist__mmap_ex(struct perf_evlist *evlist, unsigned int pages,int perf_evlist__mmap (struct perf_evlist *evlist,unsigned int pages) { -return perf_evlist__mmap_ex (evlist, pages,0 , false,0 ,PERF_AFFINITY_SYS ,1 ); +return perf_evlist__mmap_ex (evlist, pages,0 , false,0 ,PERF_AFFINITY_SYS ,1 ,0 ); }int perf_evlist__create_maps (struct perf_evlist *evlist,struct target *target)
tools/perf/util/evlist.h +1 -1
@@ -178,7 +178,7 @@ unsigned long perf_event_mlock_kb_in_pages(void);int perf_evlist__mmap_ex (struct perf_evlist *evlist,unsigned int pages,unsigned int auxtrace_pages,bool auxtrace_overwrite,int nr_cblocks, -int affinity,int flush); +int affinity,int flush,int comp_level);int perf_evlist__mmap (struct perf_evlist *evlist,unsigned int pages);void perf_evlist__munmap (struct perf_evlist *evlist);
tools/perf/util/mmap.c +28 -2
@@ -157,6 +157,10 @@ void __weak auxtrace_mmap_params__set_idx(struct auxtrace_mmap_params *mp __mayb }#ifdef HAVE_AIO_SUPPORT +static int perf_mmap__aio_enabled (struct perf_mmap *map) +{ +return map->aio.nr_cblocks >0 ; +}#ifdef HAVE_LIBNUMA_SUPPORT static int perf_mmap__aio_alloc (struct perf_mmap *map,int idx)
@@ -198,7 +202,7 @@ static int perf_mmap__aio_bind(struct perf_mmap *map, int idx, int cpu, int affi return 0; } -#else +#else /* !HAVE_LIBNUMA_SUPPORT */ static int perf_mmap__aio_alloc (struct perf_mmap *map,int idx) { map->aio.data[idx] =malloc (perf_mmap__mmap_len (map));
@@ -359,7 +363,12 @@ int perf_mmap__aio_push(struct perf_mmap *md, void *to, int idx, return rc; } -#else +#else /* !HAVE_AIO_SUPPORT */ +static int perf_mmap__aio_enabled (struct perf_mmap *map__maybe_unused ) +{ +return 0 ; +} +static int perf_mmap__aio_mmap (struct perf_mmap *map__maybe_unused ,struct mmap_params *mp__maybe_unused ) {
@@ -374,6 +383,10 @@ static void perf_mmap__aio_munmap(struct perf_mmap *map __maybe_unused)void perf_mmap__munmap (struct perf_mmap *map) {perf_mmap__aio_munmap (map); +if (map->data !=NULL ) { +munmap (map->data,perf_mmap__mmap_len (map)); + map->data =NULL ; + }if (map->base !=NULL ) {munmap (map->base,perf_mmap__mmap_len (map)); map->base =NULL ;
@@ -442,6 +455,19 @@ int perf_mmap__mmap(struct perf_mmap *map, struct mmap_params *mp, int fd, int c map->flush = mp->flush; + map->comp_level = mp->comp_level; + +if (map->comp_level && !perf_mmap__aio_enabled (map)) { + map->data =mmap (NULL ,perf_mmap__mmap_len (map),PROT_READ |PROT_WRITE , +MAP_PRIVATE |MAP_ANONYMOUS ,0 ,0 ); +if (map->data ==MAP_FAILED ) { +pr_debug2 ("failed to mmap data buffer, error %d\n" , + errno); + map->data =NULL ; +return -1 ; + } + } +if (auxtrace_mmap__mmap (&map->auxtrace_mmap, &mp->auxtrace_mp, map->base, fd))return -1 ;
tools/perf/util/mmap.h +3 -1
@@ -40,6 +40,8 @@ struct perf_mmap {#endif cpu_set_t affinity_mask;u64 flush; +void *data; +int comp_level; };/*
@@ -71,7 +73,7 @@ enum bkw_mmap_state { };struct mmap_params { -int prot, mask, nr_cblocks, affinity, flush; +int prot, mask, nr_cblocks, affinity, flush, comp_level;struct auxtrace_mmap_params auxtrace_mp; };
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 51/73 perf tools: Introduce Zstd streaming based compression API
In reply to Arnaldo Carvalho de MeloFrom: Alexey Budankov <[email protected]>
Implemented functions are based on Zstd streaming compression API.
The functions are used in runtime to compress data that come from mmaped kernel buffer. zstd_init(), zstd_fini() are used for initialization and finalization to allocate and deallocate internal zstd objects. zstd_compress_stream_to_records() is used to convert parts of mmaped kernel buffer into an array of PERF_RECORD_COMPRESSED records.
--- tools/perf/util/Build | 2 ++ tools/perf/util/compress.h | 42 +++++++++++++++++++++++ tools/perf/util/zstd.c | 70 ++++++++++++++++++++++++++++++++++++++ 3 files changed, 114 insertions(+) create mode 100644 tools/perf/util/zstd.c
tools/perf/util/Build +2 -0
@@ -145,6 +145,8 @@ perf-y += scripting-engines/ perf-$(CONFIG_ZLIB) += zlib.o perf-$(CONFIG_LZMA) += lzma.o +perf-$(CONFIG_ZSTD) += zstd.o + perf-y += demangle-java.o perf-y += demangle-rust.o
tools/perf/util/compress.h +42 -0
52 lines not on this page · raw message
tools/perf/util/zstd.c new file +70 -0
70 lines not on this page · raw message
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 52/73 perf record: Implement compression for serial trace streaming
In reply to Arnaldo Carvalho de MeloFrom: Alexey Budankov <[email protected]>
Compression is implemented using the functions from zstd.c. As the memory to operate on the compression uses mmap->data buffer.
If Zstd streaming compression API fails for some reason the data to be compressed are just copied into the memory buffers using plain memcpy().
Compressed trace frame consists of an array of PERF_RECORD_COMPRESSED records. Each element of the array is not longer that PERF_SAMPLE_MAX_SIZE and consists of perf_event_header followed by the compressed chunk that is decompressed on the loading stage.
Comitter notes:
Undo some unnecessary line breaks, remove some unnecessary () around zstd_data to then just get its address, and fix conflicts with BPF_PROG_INFO/BPF_BTF patchkits.
--- tools/perf/builtin-record.c | 51 +++++++++++++++++++++++++++++++++++-- tools/perf/util/session.h | 2 ++ 2 files changed, 51 insertions(+), 2 deletions(-)
tools/perf/builtin-record.c +49 -2
@@ -133,6 +133,9 @@ static int record__write(struct record *rec, struct perf_mmap *map __maybe_unusereturn 0 ; } +static size_t zstd_compress (struct perf_session *session,void *dst,size_t dst_size, +void *src,size_t src_size); +#ifdef HAVE_AIO_SUPPORT static int record__aio_write (struct aiocb *cblock,int trace_fd,void *buf,size_t size,off_t off)
@@ -392,6 +395,11 @@ static int record__pushfn(struct perf_mmap *map, void *to, void *bf, size_t size {struct record *rec = to; +if (record__comp_enabled (rec)) { + size =zstd_compress (rec->session, map->data,perf_mmap__mmap_len (map), bf, size); + bf = map->data; + } + rec->samples++;return record__write (rec, map, bf, size); }
@@ -778,6 +786,37 @@ static void record__adjust_affinity(struct record *rec, struct perf_mmap *map) } } +static size_t process_comp_header (void *record,size_t increment) +{ +struct compressed_event *event = record; +size_t size =sizeof (*event); + +if (increment) { + event->header.size += increment; +return increment; + } + + event->header.type =PERF_RECORD_COMPRESSED ; + event->header.size = size; + +return size; +} + +static size_t zstd_compress (struct perf_session *session,void *dst,size_t dst_size, +void *src,size_t src_size) +{ +size_t compressed; +size_t max_record_size =PERF_SAMPLE_MAX_SIZE -sizeof (struct compressed_event ) -1 ; + + compressed =zstd_compress_stream_to_records (&session->zstd_data, dst, dst_size, src, src_size, + max_record_size, process_comp_header); + + session->bytes_transferred += src_size; + session->bytes_compressed += compressed; + +return compressed; +} +static int record__mmap_read_evlist (struct record *rec,struct perf_evlist *evlist,bool overwrite,bool synch) {
@@ -1225,6 +1264,14 @@ static int __cmd_record(struct record *rec, int argc, const char **argv) fd =perf_data__fd (data); rec->session = session; +if (zstd_init (&session->zstd_data, rec->opts.comp_level) <0 ) { +pr_err ("Compression initialization failed.\n" ); +return -1 ; + } + + session->header.env.comp_type =PERF_COMP_ZSTD ; + session->header.env.comp_level = rec->opts.comp_level; +record__init_features (rec);if (rec->opts.use_clockid && rec->opts.clockid_res_ns)
@@ -1565,6 +1612,7 @@ static int __cmd_record(struct record *rec, int argc, const char **argv) } out_delete_session: +zstd_fini (&session->zstd_data);perf_session__delete (session);if (!opts->no_bpf_event)
@@ -2294,8 +2342,7 @@ int cmd_record(int argc, const char **argv)if (rec->opts.nr_cblocks > nr_cblocks_max) rec->opts.nr_cblocks = nr_cblocks_max; -if (verbose >0 ) -pr_info ("nr_cblocks: %d\n" , rec->opts.nr_cblocks); +pr_debug ("nr_cblocks: %d\n" , rec->opts.nr_cblocks);pr_debug ("affinity: %s\n" , affinity_tags[rec->opts.affinity]);pr_debug ("mmap flush: %d\n" , rec->opts.mmap_flush);
tools/perf/util/session.h +2 -0
@@ -8,6 +8,7 @@#include "machine.h" #include "data.h" #include "ordered-events.h" +#include "util/compress.h" #include <linux/kernel.h> #include <linux/rbtree.h> #include <linux/perf_event.h>
@@ -37,6 +38,7 @@ struct perf_session {struct perf_tool *tool;u64 bytes_transferred;u64 bytes_compressed; +struct zstd_data zstd_data; };struct perf_tool ;
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 53/73 perf record: Implement compression for AIO trace streaming
In reply to Arnaldo Carvalho de MeloFrom: Alexey Budankov <[email protected]>
Compression is implemented using the functions from zstd.c. As the memory to operate on the compression uses mmap->aio.data[] buffers. If Zstd streaming compression API fails for some reason the data to be compressed are just copied into the memory buffers using plain memcpy().
Compressed trace frame consists of an array of PERF_RECORD_COMPRESSED records. Each element of the array is not longer that PERF_SAMPLE_MAX_SIZE and consists of perf_event_header followed by the compressed chunk that is decompressed on the loading stage.
perf_mmap__aio_push() is replaced by perf_mmap__push() which is now used in the both serial and AIO streaming cases. perf_mmap__push() is extended with positive return values to signify absence of data ready for processing.
--- tools/perf/builtin-record.c | 114 ++++++++++++++++++++++++++++-------- tools/perf/util/mmap.c | 76 +----------------------- tools/perf/util/mmap.h | 12 ---- 3 files changed, 89 insertions(+), 113 deletions(-)
tools/perf/builtin-record.c +88 -26
@@ -133,6 +133,8 @@ static int record__write(struct record *rec, struct perf_mmap *map __maybe_unusereturn 0 ; } +static int record__aio_enabled (struct record *rec); +static int record__comp_enabled (struct record *rec);static size_t zstd_compress (struct perf_session *session,void *dst,size_t dst_size,void *src,size_t src_size);
@@ -186,9 +188,9 @@ static int record__aio_complete(struct perf_mmap *md, struct aiocb *cblock)if (rem_size ==0 ) { cblock->aio_fildes =-1 ;/* -* md->refcount is incremented in perf_mmap__push() for -* every enqueued aio write request so decrement it because -* the request is now complete. +* md->refcount is incremented in record__aio_pushfn() for +* every aio write request started in record__aio_push() so +* decrement it because the request is now complete. */ perf_mmap__put (md); rc =1 ;
@@ -243,18 +245,89 @@ static int record__aio_sync(struct perf_mmap *md, bool sync_all) }while (1 ); } -static int record__aio_pushfn (void *to,struct aiocb *cblock,void *bf,size_t size,off_t off) +struct record_aio { +struct record *rec; +void *data; +size_t size; +}; + +static int record__aio_pushfn (struct perf_mmap *map,void *to,void *buf,size_t size) { -struct record *rec = to; -int ret, trace_fd = rec->session->data->file.fd; +struct record_aio *aio = to; - rec->samples++; +/* +* map->base data pointed by buf is copied into free map->aio.data[] buffer +* to release space in the kernel buffer as fast as possible, calling +* perf_mmap__consume() from perf_mmap__push() function. +* +* That lets the kernel to proceed with storing more profiling data into +* the kernel buffer earlier than other per-cpu kernel buffers are handled. +* +* Coping can be done in two steps in case the chunk of profiling data +* crosses the upper bound of the kernel buffer. In this case we first move +* part of data from map->start till the upper bound and then the reminder +* from the beginning of the kernel buffer till the end of the data chunk. +*/ - ret =record__aio_write (cblock, trace_fd, bf, size, off); +if (record__comp_enabled (aio->rec)) { + size =zstd_compress (aio->rec->session, aio->data + aio->size, +perf_mmap__mmap_len (map) - aio->size, + buf, size); + }else { +memcpy (aio->data + aio->size, buf, size); + } + +if (!aio->size) { +/* +* Increment map->refcount to guard map->aio.data[] buffer +* from premature deallocation because map object can be +* released earlier than aio write request started on +* map->aio.data[] buffer is complete. +* +* perf_mmap__put() is done at record__aio_complete() +* after started aio request completion or at record__aio_push() +* if the request failed to start. +*/ +perf_mmap__get (map); + } + + aio->size += size; + +return size; +} + +static int record__aio_push (struct record *rec,struct perf_mmap *map,off_t *off) +{ +int ret, idx; +int trace_fd = rec->session->data->file.fd; +struct record_aio aio = { .rec = rec, .size =0 }; + +/* +* Call record__aio_sync() to wait till map->aio.data[] buffer +* becomes available after previous aio write operation. +*/ + + idx =record__aio_sync (map, false); + aio.data = map->aio.data[idx]; + ret =perf_mmap__push (map, &aio, record__aio_pushfn); +if (ret !=0 )/* ret > 0 - no data, ret < 0 - error */ +return ret; + + rec->samples++; + ret =record__aio_write (&(map->aio.cblocks[idx]), trace_fd, aio.data, aio.size, *off);if (!ret) { - rec->bytes_written += size; + *off += aio.size; + rec->bytes_written += aio.size;if (switch_output_size (rec))trigger_hit (&switch_output_trigger); + }else { +/* +* Decrement map->refcount incremented in record__aio_pushfn() +* back if record__aio_write() operation failed to start, otherwise +* map->refcount is decremented in record__aio_complete() after +* aio write operation finishes successfully. +*/ +perf_mmap__put (map); }return ret;
@@ -276,7 +349,7 @@ static void record__aio_mmap_read_sync(struct record *rec)struct perf_evlist *evlist = rec->evlist;struct perf_mmap *maps = evlist->mmap; -if (!rec->opts.nr_cblocks) +if (!record__aio_enabled (rec))return ;for (i =0 ; i < evlist->nr_mmaps; i++) {
@@ -310,13 +383,8 @@ static int record__aio_parse(const struct option *opt,#else /* HAVE_AIO_SUPPORT */ static int nr_cblocks_max =0 ; -static int record__aio_sync (struct perf_mmap *md__maybe_unused ,bool sync_all__maybe_unused ) -{ -return -1 ; -} - -static int record__aio_pushfn (void *to__maybe_unused ,struct aiocb *cblock__maybe_unused , -void *bf__maybe_unused ,size_t size__maybe_unused ,off_t off__maybe_unused ) +static int record__aio_push (struct record *rec__maybe_unused ,struct perf_mmap *map__maybe_unused , +off_t *off__maybe_unused ) {return -1 ; }
@@ -825,7 +893,7 @@ static int record__mmap_read_evlist(struct record *rec, struct perf_evlist *evliint rc =0 ;struct perf_mmap *maps;int trace_fd = rec->data.file.fd; -off_t off; +off_t off =0 ;if (!evlist)return 0 ;
@@ -851,20 +919,14 @@ static int record__mmap_read_evlist(struct record *rec, struct perf_evlist *evli map->flush =1 ; }if (!record__aio_enabled (rec)) { -if (perf_mmap__push (map, rec, record__pushfn) !=0 ) { +if (perf_mmap__push (map, rec, record__pushfn) <0 ) {if (synch) map->flush = flush; rc =-1 ; goto out; } }else { -int idx; -/* -* Call record__aio_sync() to wait till map->data buffer -* becomes available after previous aio write request. -*/ - idx =record__aio_sync (map, false); -if (perf_mmap__aio_push (map, rec, idx, record__aio_pushfn, &off) !=0 ) { +if (record__aio_push (rec, map, &off) <0 ) {record__aio_set_pos (trace_fd, off);if (synch) map->flush = flush;
tools/perf/util/mmap.c +1 -75
88 lines not on this page · raw message
tools/perf/util/mmap.h +0 -12
@@ -101,18 +101,6 @@ union perf_event *perf_mmap__read_event(struct perf_mmap *map);int perf_mmap__push (struct perf_mmap *md,void *to,int push (struct perf_mmap *map,void *to,void *buf,size_t size)); -#ifdef HAVE_AIO_SUPPORT -int perf_mmap__aio_push (struct perf_mmap *md,void *to,int idx, -int push (void *to,struct aiocb *cblock,void *buf,size_t size,off_t off), -off_t *off); -#else -static inline int perf_mmap__aio_push (struct perf_mmap *md__maybe_unused ,void *to__maybe_unused ,int idx__maybe_unused , -int push (void *to,struct aiocb *cblock,void *buf,size_t size,off_t off)__maybe_unused , -off_t *off__maybe_unused ) -{ -return 0 ; -} -#endif size_t perf_mmap__mmap_len (struct perf_mmap *map);
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 54/73 perf report: Add stub processing of compressed events for -D
In reply to Arnaldo Carvalho de MeloFrom: Alexey Budankov <[email protected]>
Committer note:
Split from a larger patch, this only dumps PERF_RECORD_COMPRESSED as unhandled, so that when we introduce the record part in the next patch, we don't see unhandled events when using 'perf record -D'.
Changed it so that we dump the event if the handler is just a stub, i.e. for the case where we don't have ZSTD linked but we're processing a perf.data file generated by a tool with that linked.
Also when failing to decompress we can't just dump the uncompressed event and return 0, we have to propagate the error.
--- tools/perf/util/session.c | 19 ++++++++++++++++++- tools/perf/util/tool.h | 2 ++ 2 files changed, 20 insertions(+), 1 deletion(-)
tools/perf/util/session.c +18 -1
@@ -358,6 +358,14 @@ static int process_stat_round_stub(struct perf_session *perf_session __maybe_unureturn 0 ; } +static int perf_session__process_compressed_event_stub (struct perf_session *session__maybe_unused , +union perf_event *event__maybe_unused , +u64 file_offset__maybe_unused ) +{ +dump_printf (": unhandled!\n" ); +return 0 ; +} +void perf_tool__fill_defaults (struct perf_tool *tool) {if (tool->sample ==NULL )
@@ -430,6 +438,8 @@ void perf_tool__fill_defaults(struct perf_tool *tool) tool->time_conv = process_event_op2_stub;if (tool->feature ==NULL ) tool->feature = process_event_op2_stub; +if (tool->compressed ==NULL ) + tool->compressed = perf_session__process_compressed_event_stub; }static void swap_sample_id_all (union perf_event *event,void *data)
@@ -1373,7 +1383,9 @@ static s64 perf_session__process_user_event(struct perf_session *session,int fd =perf_data__fd (session->data);int err; -dump_event (session->evlist, event, file_offset, &sample); +if (event->header.type !=PERF_RECORD_COMPRESSED || + tool->compressed == perf_session__process_compressed_event_stub) +dump_event (session->evlist, event, file_offset, &sample);/* These events are processed right away */ switch (event->header.type) {
@@ -1426,6 +1438,11 @@ static s64 perf_session__process_user_event(struct perf_session *session,return tool->time_conv (session, event);case PERF_RECORD_HEADER_FEATURE :return tool->feature (session, event); +case PERF_RECORD_COMPRESSED : + err = tool->compressed (session, event, file_offset); +if (err) +dump_event (session->evlist, event, file_offset, &sample); +return err;default :return -EINVAL ; }
tools/perf/util/tool.h +2 -0
@@ -28,6 +28,7 @@ typedef int (*event_attr_op)(struct perf_tool *tool,typedef int (*event_op2 )(struct perf_session *session,union perf_event *event);typedef s64 (*event_op3 )(struct perf_session *session,union perf_event *event); +typedef int (*event_op4 )(struct perf_session *session,union perf_event *event,u64 data);typedef int (*event_oe )(struct perf_tool *tool,union perf_event *event,struct ordered_events *oe);
@@ -72,6 +73,7 @@ struct perf_tool { stat, stat_round, feature; +event_op4 compressed;event_op3 auxtrace;bool ordered_events;bool ordering_requires_timestamps;
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 55/73 perf record: Implement -z,--compression_level[=<n>] option
In reply to Arnaldo Carvalho de MeloFrom: Alexey Budankov <[email protected]>
Implemented -z,--compression_level[=<n>] option that enables compression of mmaped kernel data buffers content in runtime during perf record mode collection. Default option value is 1 (fastest compression).
Compression overhead has been measured for serial and AIO streaming when profiling matrix multiplication workload:
-------------------------------------------------------------
| SERIAL | AIO-1 |
----------------------------------------------------------------| |-z | OVH(x) | ratio(x) size(MiB) | OVH(x) | ratio(x) size(MiB) | |---------------------------------------------------------------| | 0 | 1,00 | 1,000 179,424 | 1,00 | 1,000 187,527 | | 1 | 1,04 | 8,427 181,148 | 1,01 | 8,474 188,562 | | 2 | 1,07 | 8,055 186,953 | 1,03 | 7,912 191,773 | | 3 | 1,04 | 8,283 181,908 | 1,03 | 8,220 191,078 | | 5 | 1,09 | 8,101 187,705 | 1,05 | 7,780 190,065 | | 8 | 1,05 | 9,217 179,191 | 1,12 | 6,111 193,024 |
-----------------------------------------------------------------
OVH = (Execution time with -z N) / (Execution time with -z 0)
ratio - compression ratio
size - number of bytes that was compressed
size ~= trace size x ratio
Committer notes:
Testing it I noticed that it failed to disable build id processing when compression is enabled, and as we'd have to uncompress everything to look for the PERF_RECORD_{MMAP,SAMPLE,etc} to figure out which build ids to read from DSOs, we better disable build id processing when compression is enabled, logging with pr_debug() when doing so:
Original patch:
# perf record -z2 ^C[ perf record: Woken up 1 times to write data ] 0x1746e0 [0x76]: failed to process type: 81 [Invalid argument] [ perf record: Captured and wrote 1.568 MB perf.data, compressed (original 0.452 MB, ratio is 3.995) ] #
After auto-disabling build id processing when compression is enabled:
$ perf record -z2 sleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.001 MB perf.data, compressed (original 0.001 MB, ratio is 2.292) ] $ perf record -v -z2 sleep 1 Compression enabled, disabling build id collection at the end of the session. <SNIP extra -v pr_debug() messages> [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.001 MB perf.data, compressed (original 0.001 MB, ratio is 2.305) ] $
Also, with parts of the patch originally after this one moved to just before this one we get:
$ perf record -z2 sleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.001 MB perf.data, compressed (original 0.001 MB, ratio is 2.371) ]
$ perf report -D | grep COMPRESS
0 0x1b8 [0x155]: PERF_RECORD_COMPRESSED: unhandled!
0 0x30d [0x80]: PERF_RECORD_COMPRESSED: unhandled!
COMPRESSED events: 2
COMPRESSED events: 0
$I.e. when faced with PERF_RECORD_COMPRESSED that we still have no code to process, we just show it as not being handled, skip them and continue, while before we had:
$ perf report -D | grep COMPRESS 0x1b8 [0x169]: failed to process type: 81 [Invalid argument] Error: failed to process sample 0 0x1b8 [0x169]: PERF_RECORD_COMPRESSED $
--- tools/perf/Documentation/perf-record.txt | 5 ++++ tools/perf/builtin-record.c | 30 ++++++++++++++++++++++++ 2 files changed, 35 insertions(+)
tools/perf/Documentation/perf-record.txt +5 -0
@@ -478,6 +478,11 @@ Also at some cases executing less output write syscalls with bigger data size can take less time than executing more output write syscalls with smaller data size thus lowering runtime profiling overhead. +-z:: +--compression-level[=n]:: +Produce compressed trace using specified level n (default: 1 - fastest compression, +22 - smallest trace) + --all-kernel:: Configure all used events to run in kernel space.
tools/perf/builtin-record.c +30 -0
@@ -443,6 +443,25 @@ static int record__mmap_flush_parse(const struct option *opt,return 0 ; } +#ifdef HAVE_ZSTD_SUPPORT +static unsigned int comp_level_default =1 ; + +static int record__parse_comp_level (const struct option *opt,const char *str,int unset) +{ +struct record_opts *opts = opt->value; + +if (unset) { + opts->comp_level =0 ; + }else { +if (str) + opts->comp_level =strtol (str,NULL ,0 ); +if (!opts->comp_level) + opts->comp_level = comp_level_default; + } + +return 0 ; +} +#endif static unsigned int comp_level_max =22 ;static int record__comp_enabled (struct record *rec)
@@ -2200,6 +2219,11 @@ static struct option __record_options[] = { OPT_CALLBACK(0, "affinity", &record.opts, "node|cpu", "Set affinity mask of trace reading thread to NUMA node cpu mask or cpu of processed mmap buffer", record__parse_affinity), +#ifdef HAVE_ZSTD_SUPPORT +OPT_CALLBACK_OPTARG ('z' ,"compression-level" , &record.opts, &comp_level_default, +"n" ,"Compressed records using specified level (default: 1 - fastest compression, 22 - greatest compression)" , + record__parse_comp_level), +#endifOPT_END () };
@@ -2259,6 +2283,12 @@ int cmd_record(int argc, const char **argv) "cgroup monitoring only available in system-wide mode"); } + +if (rec->opts.comp_level !=0 ) { +pr_debug ("Compression enabled, disabling build id collection at the end of the session.\n" ); + rec->no_buildid = true; + } +if (rec->opts.record_switch_events && !perf_can_record_switch_events ()) {ui__error ("kernel does not support recording context switch events\n" );
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 56/73 perf report: Implement perf.data record decompression
In reply to Arnaldo Carvalho de MeloFrom: Alexey Budankov <[email protected]>
zstd_init(, comp_level = 0) initializes decompression part of API only hat now consists of zstd_decompress_stream() function.
The perf.data PERF_RECORD_COMPRESSED records are decompressed using zstd_decompress_stream() function into a linked list of mmaped memory regions of mmap_comp_len size (struct decomp).
After decompression of one COMPRESSED record its content is iterated and fetched for usual processing. The mmaped memory regions with decompressed events are kept in the linked list till the tool process termination.
When dumping raw records (e.g., perf report -D --header) file offsets of events from compressed records are printed as zero.
Committer notes:
Since now we have support for processing PERF_RECORD_COMPRESSED, we see none, in raw form, like we saw in the previous patch commiter notes, they were decompressed into the usual PERF_RECORD_{FORK,MMAP,COMM,etc} records, we only see the stats for those PERF_RECORD_COMPRESSED events, and since I used the file generated in the commiter notes for the previous patch, there they are, 2 compressed records:
$ perf report --header-only | grep cmdline
# cmdline : /home/acme/bin/perf record -z2 sleep 1
$ perf report -D | grep COMPRESS
COMPRESSED events: 2
COMPRESSED events: 0
$ perf report --stdio
# To display the perf.data header info, please use --header/--header-only options.
#
#
# Total Lost Samples: 0
#
# Samples: 15 of event 'cycles:u'
# Event count (approx.): 962227
#
# Overhead Command Shared Object Symbol
# ........ ....... ................ ...........................
#
46.99% sleep libc-2.28.so [.] _dl_addr
29.24% sleep [unknown] [k] 0xffffffffaea00a67
16.45% sleep libc-2.28.so [.] __GI__IO_un_link.part.1
5.92% sleep ld-2.28.so [.] _dl_setup_hash
1.40% sleep libc-2.28.so [.] __nanosleep
0.00% sleep [unknown] [k] 0xffffffffaea00163 #
# (Tip: To see callchains in a more compact form: perf report -g folded) #
$
--- tools/perf/builtin-report.c | 5 +- tools/perf/util/compress.h | 11 ++++ tools/perf/util/session.c | 116 +++++++++++++++++++++++++++++++++++- tools/perf/util/session.h | 10 ++++ tools/perf/util/zstd.c | 41 +++++++++++++ 5 files changed, 181 insertions(+), 2 deletions(-)
tools/perf/builtin-report.c +4 -1
@@ -1258,6 +1258,9 @@ int cmd_report(int argc, const char **argv)if (session ==NULL )return -1 ; +if (zstd_init (&(session->zstd_data),0 ) <0 ) +pr_warning ("Decompression initialization failed. Reported data may be incomplete.\n" ); +if (report.queue_size) {ordered_events__set_alloc_size (&session->ordered_events, report.queue_size);
@@ -1448,7 +1451,7 @@ int cmd_report(int argc, const char **argv) error:if (report.ptime_range)zfree (&report.ptime_range); - +zstd_fini (&(session->zstd_data));perf_session__delete (session);return ret; }
tools/perf/util/compress.h +11 -0
@@ -20,6 +20,7 @@ bool lzma_is_compressed(const char *input);struct zstd_data {#ifdef HAVE_ZSTD_SUPPORT ZSTD_CStream *cstream; +ZSTD_DStream *dstream;#endif };
@@ -31,6 +32,9 @@ int zstd_fini(struct zstd_data *data);size_t zstd_compress_stream_to_records (struct zstd_data *data,void *dst,size_t dst_size,void *src,size_t src_size,size_t max_record_size,size_t process_header (void *record,size_t increment)); + +size_t zstd_decompress_stream (struct zstd_data *data,void *src,size_t src_size, +void *dst,size_t dst_size);#else /* !HAVE_ZSTD_SUPPORT */ static inline int zstd_init (struct zstd_data *data__maybe_unused ,int level__maybe_unused )
@@ -52,6 +56,13 @@ size_t zstd_compress_stream_to_records(struct zstd_data *data __maybe_unused, {return 0 ; } + +static inline size_t zstd_decompress_stream (struct zstd_data *data__maybe_unused ,void *src__maybe_unused , +size_t src_size__maybe_unused ,void *dst__maybe_unused , +size_t dst_size__maybe_unused ) +{ +return 0 ; +}#endif #endif /* PERF_COMPRESS_H */
tools/perf/util/session.c +115 -1
164 lines not on this page · raw message
tools/perf/util/session.h +10 -0
@@ -39,6 +39,16 @@ struct perf_session {u64 bytes_transferred;u64 bytes_compressed;struct zstd_data zstd_data; +struct decomp *decomp; +struct decomp *decomp_last; +}; + +struct decomp { +struct decomp *next; +u64 file_pos; +u64 head; +size_t size; +char data[]; };struct perf_tool ;
tools/perf/util/zstd.c +41 -0
56 lines not on this page · raw message
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 57/73 perf inject: Enable COMPRESSED record decompression
In reply to Arnaldo Carvalho de MeloFrom: Alexey Budankov <[email protected]>
Initialized decompression part of Zstd based API so COMPRESSED records would be decompressed into the resulting output data file.
--- tools/perf/builtin-inject.c | 4 ++++ 1 file changed, 4 insertions(+)
tools/perf/builtin-inject.c +4 -0
@@ -837,6 +837,9 @@ int cmd_inject(int argc, const char **argv)if (inject.session ==NULL )return -1 ; +if (zstd_init (&(inject.session->zstd_data),0 ) <0 ) +pr_warning ("Decompression initialization failed.\n" ); +if (inject.build_ids) {/* * to make sure the mmap records are ordered correctly
@@ -867,6 +870,7 @@ int cmd_inject(int argc, const char **argv) ret =__cmd_inject (&inject); out_delete: +zstd_fini (&(inject.session->zstd_data));perf_session__delete (inject.session);return ret; }
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 58/73 perf tests: Implement Zstd comp/decomp integration test
In reply to Arnaldo Carvalho de MeloFrom: Alexey Budankov <[email protected]>
Introduce a basic integration test for Zstd based record compression/decompression using 'perf record' and 'perf report'.
Committer notes:
Reduce a bit the freq (from 25 kHz to 5 kHz) and the number of /dev/null records read (from 1000 to 500), reducing the time it takes to something more in line with the time existing 'perf test' entries take to run.
With that in place:
$ time perf test zstd 68: Zstd perf.data compression/decompression : Ok
real 0m10.376s
user 0m0.105s
sys 0m0.440s
$ grep "model name" /proc/cpuinfo | head -1 model name : Intel(R) Core(TM) i7-8650U CPU @ 1.90GHz $
--- .../tests/shell/record+zstd_comp_decomp.sh | 35 +++++++++++++++++++ 1 file changed, 35 insertions(+) create mode 100755 tools/perf/tests/shell/record+zstd_comp_decomp.sh
tools/perf/tests/shell/record+zstd_comp_decomp.sh new file +35 -0
@@ -0,0 +1,35 @@ +#!/bin/sh +# Zstd perf.data compression/decompression + +trace_file=$(mktemp /tmp/perf.data.XXX) +perf_tool=perf +output=/dev/null + +skip_if_no_z_record() { + $perf_tool record -h 2>&1 | grep '\-z, \-\-compression\-level' +} + +collect_z_record() { + echo "Collecting compressed record file:" + $perf_tool record -o $trace_file -g -z -F 5000 -- \ + dd count=500 if=/dev/random of=/dev/null > $output 2>&1 +} + +check_compressed_stats() { + echo "Checking compressed events stats:" + $perf_tool report -i $trace_file --header --stats | \ + grep -E "(# compressed : Zstd,)|(COMPRESSED events:)" > $output 2>&1 +} + +check_compressed_output() { + $perf_tool inject -i $trace_file -o $trace_file.decomp && + $perf_tool report -i $trace_file --stdio | head -n -3 > $trace_file.comp.output && + $perf_tool report -i $trace_file.decomp --stdio | head -n -3 > $trace_file.decomp.output && + diff $trace_file.comp.output $trace_file.decomp.output > $output 2>&1 +} + +skip_if_no_z_record || exit 2 +collect_z_record && check_compressed_stats && check_compressed_output +err=$? +rm -f $trace_file* +exit $err
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 59/73 perf test zstd: Fixup verbose mode output
In reply to Arnaldo Carvalho de MeloFrom: Arnaldo Carvalho de Melo <[email protected]>
The shell tests should not redirect useful output to /dev/null, as that is done automatically by 'perf test' in non verbose mode, so remove that from the zstd comp/decomp test, fixing up verbose mode.
Before:
$ perf test zstd
68: Zstd perf.data compression/decompression : Ok
$ perf test -v zstd
68: Zstd perf.data compression/decompression :
--- start ---
test child forked, pid 11956
-z, --compression-level[=<n>]
Collecting compressed record file:
Checking compressed events stats:
test child finished with 0
---- end ----
Zstd perf.data compression/decompression: Ok
$Now:
$ perf test zstd
68: Zstd perf.data compression/decompression : Ok
$ perf test -v zstd
68: Zstd perf.data compression/decompression :
--- start ---
test child forked, pid 12695
Collecting compressed record file:
0+500 records in
72+1 records out
37361 bytes (37 kB, 36 KiB) copied, 9.83796 s, 3.8 kB/s
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.001 MB /tmp/perf.data.rzq, compressed (original 0.004 MB, ratio is 3.679) ]
Checking compressed events stats:
# compressed : Zstd, level = 1, ratio = 4
COMPRESSED events: 3
test child finished with 0
---- end ----
Zstd perf.data compression/decompression: Ok
$--- tools/perf/tests/shell/record+zstd_comp_decomp.sh | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-)
tools/perf/tests/shell/record+zstd_comp_decomp.sh +4 -5
@@ -3,29 +3,28 @@ trace_file=$(mktemp /tmp/perf.data.XXX) perf_tool=perf -output=/dev/null skip_if_no_z_record() { - $perf_tool record -h 2>&1 | grep '\-z, \-\-compression\-level' + $perf_tool record -h 2>&1 | grep -q '\-z, \-\-compression\-level' } collect_z_record() { echo "Collecting compressed record file:" $perf_tool record -o $trace_file -g -z -F 5000 -- \ - dd count=500 if=/dev/random of=/dev/null > $output 2>&1 + dd count=500 if=/dev/random of=/dev/null } check_compressed_stats() { echo "Checking compressed events stats:" $perf_tool report -i $trace_file --header --stats | \ - grep -E "(# compressed : Zstd,)|(COMPRESSED events:)" > $output 2>&1 + grep -E "(# compressed : Zstd,)|(COMPRESSED events:)" } check_compressed_output() { $perf_tool inject -i $trace_file -o $trace_file.decomp && $perf_tool report -i $trace_file --stdio | head -n -3 > $trace_file.comp.output && $perf_tool report -i $trace_file.decomp --stdio | head -n -3 > $trace_file.decomp.output && - diff $trace_file.comp.output $trace_file.decomp.output > $output 2>&1 + diff $trace_file.comp.output $trace_file.decomp.output } skip_if_no_z_record || exit 2
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 60/73 perf jevents: Remove unused variable
In reply to Arnaldo Carvalho de MeloFrom: Zenghui Yu <[email protected]>
Address gcc warning:
pmu-events/jevents.c: In function ‘save_arch_std_events’:
pmu-events/jevents.c:417:15: warning: unused variable ‘sb’ [-Wunused-variable]
struct stat *sb = data;
^~--- tools/perf/pmu-events/jevents.c | 1 - 1 file changed, 1 deletion(-)
tools/perf/pmu-events/jevents.c +0 -1
@@ -415,7 +415,6 @@ static int save_arch_std_events(void *data, char *name, char *event, char *metric_name, char *metric_group) { struct event_struct *es; -struct stat *sb = data; es =malloc (sizeof (*es));if (!es)
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 61/73 perf vendor events arm64: Remove [[:xdigit:]] wildcard
In reply to Arnaldo Carvalho de MeloFrom: Florian Fainelli <[email protected]>
ARM64's implementation of get_cpuidr_str() masks out the revision bits [3:0] while reading the CPU identifier, there is no need for the [[:xdigit:]] wildcard.
--- tools/perf/pmu-events/arch/arm64/mapfile.csv | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/pmu-events/arch/arm64/mapfile.csv +1 -1
@@ -12,7 +12,7 @@ # # #Family-model,Version,Filename,EventType -0x00000000410fd03[[:xdigit:]],v1,arm/cortex-a53,core +0x00000000410fd030,v1,arm/cortex-a53,core 0x00000000420f5160,v1,cavium/thunderx2,core 0x00000000430f0af0,v1,cavium/thunderx2,core 0x00000000480fd010,v1,hisilicon/hip08,core
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 62/73 perf vendor events arm64: Map Brahma-B53 CPUID to cortex-a53 events
In reply to Arnaldo Carvalho de MeloFrom: Florian Fainelli <[email protected]>
Broadcom's Brahma-B53 CPUs support the same type of events that the Cortex-A53 supports, recognize its CPUID and map it to the cortex-a53 events.
--- tools/perf/pmu-events/arch/arm64/mapfile.csv | 1 + 1 file changed, 1 insertion(+)
tools/perf/pmu-events/arch/arm64/mapfile.csv +1 -0
@@ -13,6 +13,7 @@ # #Family-model,Version,Filename,EventType 0x00000000410fd030,v1,arm/cortex-a53,core +0x00000000420f1000,v1,arm/cortex-a53,core 0x00000000420f5160,v1,cavium/thunderx2,core 0x00000000430f0af0,v1,cavium/thunderx2,core 0x00000000480fd010,v1,hisilicon/hip08,core
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 63/73 perf vendor events arm64: Add Cortex-A57 and Cortex-A72 events
In reply to Arnaldo Carvalho de MeloFrom: Florian Fainelli <[email protected]>
The Cortex-A57 and Cortex-A72 both support all ARMv8 recommended events up to the RC_ST_SPEC (0x91) event with the exception of:
- L1D_CACHE_REFILL_INNER (0x44)
- L1D_CACHE_REFILL_OUTER (0x45)
- L1D_TLB_RD (0x4E)
- L1D_TLB_WR (0x4F)
- L2D_TLB_REFILL_RD (0x5C)
- L2D_TLB_REFILL_WR (0x5D)
- L2D_TLB_RD (0x5E)
- L2D_TLB_WR (0x5F)
- STREX_SPEC (0x6F)
Create an appropriate JSON file for mapping those events and update the mapfile.csv for matching the Cortex-A57 and Cortex-A72 MIDR to that file.
--- .../arm/cortex-a57-a72/core-imp-def.json | 179 ++++++++++++++++++ tools/perf/pmu-events/arch/arm64/mapfile.csv | 2 + 2 files changed, 181 insertions(+) create mode 100644 tools/perf/pmu-events/arch/arm64/arm/cortex-a57-a72/core-imp-def.json
tools/perf/pmu-events/arch/arm64/arm/cortex-a57-a72/core-imp-def.json new file +179 -0
@@ -0,0 +1,179 @@ +[ + { + "ArchStdEvent": "L1D_CACHE_RD", + }, + { + "ArchStdEvent": "L1D_CACHE_WR", + }, + { + "ArchStdEvent": "L1D_CACHE_REFILL_RD", + }, + { + "ArchStdEvent": "L1D_CACHE_REFILL_WR", + }, + { + "ArchStdEvent": "L1D_CACHE_WB_VICTIM", + }, + { + "ArchStdEvent": "L1D_CACHE_WB_CLEAN", + }, + { + "ArchStdEvent": "L1D_CACHE_INVAL", + }, + { + "ArchStdEvent": "L1D_TLB_REFILL_RD", + }, + { + "ArchStdEvent": "L1D_TLB_REFILL_WR", + }, + { + "ArchStdEvent": "L2D_CACHE_RD", + }, + { + "ArchStdEvent": "L2D_CACHE_WR", + }, + { + "ArchStdEvent": "L2D_CACHE_REFILL_RD", + }, + { + "ArchStdEvent": "L2D_CACHE_REFILL_WR", + }, + { + "ArchStdEvent": "L2D_CACHE_WB_VICTIM", + }, + { + "ArchStdEvent": "L2D_CACHE_WB_CLEAN", + }, + { + "ArchStdEvent": "L2D_CACHE_INVAL", + }, + { + "ArchStdEvent": "BUS_ACCESS_RD", + }, + { + "ArchStdEvent": "BUS_ACCESS_WR", + }, + { + "ArchStdEvent": "BUS_ACCESS_SHARED", + }, + { + "ArchStdEvent": "BUS_ACCESS_NOT_SHARED", + }, + { + "ArchStdEvent": "BUS_ACCESS_NORMAL", + }, + { + "ArchStdEvent": "BUS_ACCESS_PERIPH", + }, + { + "ArchStdEvent": "MEM_ACCESS_RD", + }, + { + "ArchStdEvent": "MEM_ACCESS_WR", + }, + { + "ArchStdEvent": "UNALIGNED_LD_SPEC", + }, + { + "ArchStdEvent": "UNALIGNED_ST_SPEC", + }, + { + "ArchStdEvent": "UNALIGNED_LDST_SPEC", + }, + { + "ArchStdEvent": "LDREX_SPEC", + }, + { + "ArchStdEvent": "STREX_PASS_SPEC", + }, + { + "ArchStdEvent": "STREX_FAIL_SPEC", + }, + { + "ArchStdEvent": "LD_SPEC", + }, + { + "ArchStdEvent": "ST_SPEC", + }, + { + "ArchStdEvent": "LDST_SPEC", + }, + { + "ArchStdEvent": "DP_SPEC", + }, + { + "ArchStdEvent": "ASE_SPEC", + }, + { + "ArchStdEvent": "VFP_SPEC", + }, + { + "ArchStdEvent": "PC_WRITE_SPEC", + }, + { + "ArchStdEvent": "CRYPTO_SPEC", + }, + { + "ArchStdEvent": "BR_IMMED_SPEC", + }, + { + "ArchStdEvent": "BR_RETURN_SPEC", + }, + { + "ArchStdEvent": "BR_INDIRECT_SPEC", + }, + { + "ArchStdEvent": "ISB_SPEC", + }, + { + "ArchStdEvent": "DSB_SPEC", + }, + { + "ArchStdEvent": "DMB_SPEC", + }, + { + "ArchStdEvent": "EXC_UNDEF", + }, + { + "ArchStdEvent": "EXC_SVC", + }, + { + "ArchStdEvent": "EXC_PABORT", + }, + { + "ArchStdEvent": "EXC_DABORT", + }, + { + "ArchStdEvent": "EXC_IRQ", + }, + { + "ArchStdEvent": "EXC_FIQ", + }, + { + "ArchStdEvent": "EXC_SMC", + }, + { + "ArchStdEvent": "EXC_HVC", + }, + { + "ArchStdEvent": "EXC_TRAP_PABORT", + }, + { + "ArchStdEvent": "EXC_TRAP_DABORT", + }, + { + "ArchStdEvent": "EXC_TRAP_OTHER", + }, + { + "ArchStdEvent": "EXC_TRAP_IRQ", + }, + { + "ArchStdEvent": "EXC_TRAP_FIQ", + }, + { + "ArchStdEvent": "RC_LD_SPEC", + }, + { + "ArchStdEvent": "RC_ST_SPEC", + }, +]
tools/perf/pmu-events/arch/arm64/mapfile.csv +2 -0
@@ -14,6 +14,8 @@ #Family-model,Version,Filename,EventType 0x00000000410fd030,v1,arm/cortex-a53,core 0x00000000420f1000,v1,arm/cortex-a53,core +0x00000000410fd070,v1,arm/cortex-a57-a72,core +0x00000000410fd080,v1,arm/cortex-a57-a72,core 0x00000000420f5160,v1,cavium/thunderx2,core 0x00000000430f0af0,v1,cavium/thunderx2,core 0x00000000480fd010,v1,hisilicon/hip08,core
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 64/73 perf parse-regs: Split parse_regs
In reply to Arnaldo Carvalho de MeloFrom: Kan Liang <[email protected]>
The available registers for --int-regs and --user-regs may be different, e.g. XMM registers.
Split parse_regs into two dedicated functions for --int-regs and --user-regs respectively.
Modify the warning message. "--user-regs=?" should be applied to show the available registers for --user-regs.
--- tools/perf/Documentation/perf-record.txt | 3 ++- tools/perf/builtin-record.c | 4 ++-- tools/perf/util/parse-regs-options.c | 19 ++++++++++++++++--- tools/perf/util/parse-regs-options.h | 3 ++- 4 files changed, 22 insertions(+), 7 deletions(-)
tools/perf/Documentation/perf-record.txt +2 -1
@@ -406,7 +406,8 @@ symbolic names, e.g. on x86, ax, si. To list the available registers use --intr-regs=ax,bx. The list of register is architecture dependent. --user-regs:: -Capture user registers at sample time. Same arguments as -I. +Similar to -I, but capture user registers at sample time. To list the available +user registers use --user-regs=\?. --running-time:: Record running and enabled time for read events (:S)
tools/perf/builtin-record.c +2 -2
@@ -2168,10 +2168,10 @@ static struct option __record_options[] = { "use per-thread mmaps"),OPT_CALLBACK_OPTARG ('I' ,"intr-regs" , &record.opts.sample_intr_regs,NULL ,"any register" ,"sample selected machine registers on interrupt," -" use '-I?' to list register names" , parse_regs), +" use '-I?' to list register names" , parse_intr_regs),OPT_CALLBACK_OPTARG (0 ,"user-regs" , &record.opts.sample_user_regs,NULL ,"any register" ,"sample selected machine registers on interrupt," -" use '-I?' to list register names" , parse_regs), +" use '--user-regs=?' to list register names" , parse_user_regs),OPT_BOOLEAN (0 ,"running-time" , &record.opts.running_time,"Record running/enabled time of read (:S) events" ), OPT_CALLBACK('k', "clockid", &record.opts,
tools/perf/util/parse-regs-options.c +16 -3
@@ -5,8 +5,8 @@#include <subcmd/parse-options.h> #include "util/parse-regs-options.h" -int -parse_regs (const struct option *opt,const char *str,int unset) +static int +__parse_regs (const struct option *opt,const char *str,int unset,bool intr) {uint64_t *mode = (uint64_t *)opt->value;const struct sample_reg *r;
@@ -48,7 +48,8 @@ parse_regs(const struct option *opt, const char *str, int unset)break ; }if (!r->name) { -ui__warning ("Unknown register \"%s\", check man page or run \"perf record -I?\"\n" , s); +ui__warning ("Unknown register \"%s\", check man page or run \"perf record %s?\"\n" , + s, intr ?"-I" :"--user-regs=" ); goto error; }
@@ -69,3 +70,15 @@ parse_regs(const struct option *opt, const char *str, int unset)free (os);return ret; } + +int +parse_user_regs (const struct option *opt,const char *str,int unset) +{ +return __parse_regs (opt, str, unset, false); +} + +int +parse_intr_regs (const struct option *opt,const char *str,int unset) +{ +return __parse_regs (opt, str, unset, true); +}
tools/perf/util/parse-regs-options.h +2 -1
@@ -2,5 +2,6 @@#ifndef _PERF_PARSE_REGS_OPTIONS_H#define _PERF_PARSE_REGS_OPTIONS_H 1struct option ; -int parse_regs (const struct option *opt,const char *str,int unset); +int parse_user_regs (const struct option *opt,const char *str,int unset); +int parse_intr_regs (const struct option *opt,const char *str,int unset);#endif /* _PERF_PARSE_REGS_OPTIONS_H */
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 65/73 perf parse-regs: Add generic support for arch__intr/user_reg_mask()
In reply to Arnaldo Carvalho de MeloFrom: Kan Liang <[email protected]>
There may be different register mask for use with intr or user on some platforms, e.g. Icelake.
Add weak functions arch__intr_reg_mask() and arch__user_reg_mask() to return intr and user register mask respectively.
Check mask before printing or comparing the register name.
Generic code always return PERF_REGS_MASK. No functional change.
--- tools/perf/util/parse-regs-options.c | 13 ++++++++++--- tools/perf/util/perf_regs.c | 10 ++++++++++ tools/perf/util/perf_regs.h | 2 ++ 3 files changed, 22 insertions(+), 3 deletions(-)
tools/perf/util/parse-regs-options.c +10 -3
@@ -12,6 +12,7 @@ __parse_regs(const struct option *opt, const char *str, int unset, bool intr)const struct sample_reg *r;char *s, *os =NULL , *p;int ret =-1 ; +uint64_t mask;if (unset)return 0 ;
@@ -22,6 +23,11 @@ __parse_regs(const struct option *opt, const char *str, int unset, bool intr)if (*mode)return -1 ; +if (intr) + mask =arch__intr_reg_mask (); +else + mask =arch__user_reg_mask (); +/* str may be NULL in case no arg is passed to -I */ if (str) {/* because str is read-only */
@@ -37,14 +43,15 @@ __parse_regs(const struct option *opt, const char *str, int unset, bool intr)if (!strcmp (s,"?" )) {fprintf (stderr,"available registers: " );for (r = sample_reg_masks; r->name; r++) { -fprintf (stderr,"%s " , r->name); +if (r->mask & mask) +fprintf (stderr,"%s " , r->name); }fputc ('\n' , stderr);/* just printing available regs */ return -1 ; }for (r = sample_reg_masks; r->name; r++) { -if (!strcasecmp (s, r->name)) +if ((r->mask & mask) && !strcasecmp (s, r->name))break ; }if (!r->name) {
@@ -65,7 +72,7 @@ __parse_regs(const struct option *opt, const char *str, int unset, bool intr)/* default to all possible regs */ if (*mode ==0 ) - *mode =PERF_REGS_MASK ; + *mode = mask; error:free (os);return ret;
tools/perf/util/perf_regs.c +10 -0
@@ -13,6 +13,16 @@ int __weak arch_sdt_arg_parse_op(char *old_op __maybe_unused,return SDT_ARG_SKIP ; } +uint64_t __weak arch__intr_reg_mask (void ) +{ +return PERF_REGS_MASK ; +} + +uint64_t __weak arch__user_reg_mask (void ) +{ +return PERF_REGS_MASK ; +} +#ifdef HAVE_PERF_REGS_SUPPORT int perf_reg_value (u64 *valp,struct regs_dump *regs,int id) {
tools/perf/util/perf_regs.h +2 -0
@@ -23,6 +23,8 @@ enum { };int arch_sdt_arg_parse_op (char *old_op,char **new_op); +uint64_t arch__intr_reg_mask (void ); +uint64_t arch__user_reg_mask (void );#ifdef HAVE_PERF_REGS_SUPPORT #include <perf_regs.h>
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 66/73 perf regs x86: Add X86 specific arch__intr_reg_mask()
In reply to Arnaldo Carvalho de MeloFrom: Kan Liang <[email protected]>
XMM registers can be collected on Icelake and later platforms.
Add specific arch__intr_reg_mask(), which creating an event to check if the kernel and hardware can collect XMM registers.
Test on Skylake which doesn't support XMM registers collection. There is nothing changed.
#perf record -I?
available registers: AX BX CX DX SI DI BP SP IP FLAGS CS SS R8 R9 R10 R11 R12 R13 R14 R15
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-I, --intr-regs[=<any register>]
sample selected machine registers on
interrupt, use '-I?' to list register names
#perf record -I
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.905 MB perf.data (2520 samples) ]
#perf evlist -v
cycles: size: 112, { sample_period, sample_freq }: 4000, sample_type:
IP|TID|TIME|CPU|PERIOD|REGS_INTR, read_format: ID, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, precise_ip: 3,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1, ksymbol:
1, bpf_event: 1, sample_regs_intr: 0xff0fff
Test on Icelake which support XMM registers collection.
#perf record -I?
available registers: AX BX CX DX SI DI BP SP IP FLAGS CS SS R8 R9 R10 R11 R12 R13 R14 R15 XMM0 XMM1 XMM2 XMM3 XMM4 XMM5 XMM6 XMM7 XMM8 XMM9 XMM10 XMM11 XMM12 XMM13 XMM14 XMM15
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-I, --intr-regs[=<any register>]
sample selected machine registers on
interrupt, use '-I?' to list register names
#perf record -I
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.800 MB perf.data (318 samples) ]
#perf evlist -v
cycles: size: 112, { sample_period, sample_freq }: 4000, sample_type:
IP|TID|TIME|CPU|PERIOD|REGS_INTR, read_format: ID, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, precise_ip: 3,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1, ksymbol:
1, bpf_event: 1, sample_regs_intr: 0xffffffff00ff0fff
Committer notes:
Don't set attr.sample_period as a named struct init, as it is part of an unnamed union in 'struct perf_event_attr', and doing so breaks the build on older gcc versions, such as:
gcc version 4.1.2 20080704 (Red Hat 4.1.2-55) gcc version 4.4.7 20120313 (Red Hat 4.4.7-23) (GCC)
arch/x86/util/perf_regs.c: In function 'arch__intr_reg_mask':
arch/x86/util/perf_regs.c:279: error: unknown field 'sample_period' specified in initializer
cc1: warnings being treated as errors
arch/x86/util/perf_regs.c:279: warning: missing braces around initializer arch/x86/util/perf_regs.c:279: warning: (near initialization for 'attr.<anonymous>')
--- tools/perf/arch/x86/include/perf_regs.h | 1 + tools/perf/arch/x86/util/perf_regs.c | 28 +++++++++++++++++++++++++ 2 files changed, 29 insertions(+)
tools/perf/arch/x86/include/perf_regs.h +1 -0
@@ -9,6 +9,7 @@void perf_regs_load (u64 *regs);#define PERF_REGS_MAX PERF_REG_X86_XMM_MAX +#define PERF_XMM_REGS_MASK (~((1ULL << PERF_REG_X86_XMM0) - 1))#ifndef HAVE_ARCH_X86_64_SUPPORT #define PERF_REGS_MASK ((1ULL << PERF_REG_X86_32_MAX) - 1)#define PERF_SAMPLE_REGS_ABI PERF_SAMPLE_REGS_ABI_32
tools/perf/arch/x86/util/perf_regs.c +28 -0
@@ -270,3 +270,31 @@ int arch_sdt_arg_parse_op(char *old_op, char **new_op)return SDT_ARG_VALID ; } + +uint64_t arch__intr_reg_mask (void ) +{ +struct perf_event_attr attr = { + .type =PERF_TYPE_HARDWARE , + .config =PERF_COUNT_HW_CPU_CYCLES , + .sample_type =PERF_SAMPLE_REGS_INTR , + .sample_regs_intr =PERF_XMM_REGS_MASK , + .precise_ip =1 , + .disabled =1 , + .exclude_kernel =1 , + }; +int fd; +/* +* In an unnamed union, init it here to build on older gcc versions +*/ + attr.sample_period =1 ; + +event_attr_init (&attr); + + fd =sys_perf_event_open (&attr,0 ,-1 ,-1 ,0 ); +if (fd !=-1 ) { +close (fd); +return (PERF_XMM_REGS_MASK |PERF_REGS_MASK ); + } + +return PERF_REGS_MASK ; +}
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 67/73 perf intel-pt: Fix instructions sampling rate
In reply to Arnaldo Carvalho de MeloFrom: Adrian Hunter <[email protected]>
The timestamp used to determine if an instruction sample is made, is an estimate based on the number of instructions since the last known timestamp. A consequence is that it might go backwards, which results in extra samples. Change it so that a sample is only made when the timestamp goes forwards.
Note this does not affect a sampling period of 0 or sampling periods specified as a count of instructions.
Example:
Before:
$ perf script --itrace=i10us ls 13812 [003] 2167315.222583: 3270 instructions:u: 7fac71e2e494 __GI___tunables_init+0xf4 (/lib/x86_64-linux-gnu/ld-2.28.so) ls 13812 [003] 2167315.222667: 30902 instructions:u: 7fac71e2da0f _dl_cache_libcmp+0x2f (/lib/x86_64-linux-gnu/ld-2.28.so) ls 13812 [003] 2167315.222667: 10 instructions:u: 7fac71e2d9ff _dl_cache_libcmp+0x1f (/lib/x86_64-linux-gnu/ld-2.28.so) ls 13812 [003] 2167315.222667: 8 instructions:u: 7fac71e2d9ea _dl_cache_libcmp+0xa (/lib/x86_64-linux-gnu/ld-2.28.so) ls 13812 [003] 2167315.222667: 14 instructions:u: 7fac71e2d9ea _dl_cache_libcmp+0xa (/lib/x86_64-linux-gnu/ld-2.28.so) ls 13812 [003] 2167315.222667: 6 instructions:u: 7fac71e2d9ff _dl_cache_libcmp+0x1f (/lib/x86_64-linux-gnu/ld-2.28.so) ls 13812 [003] 2167315.222667: 14 instructions:u: 7fac71e2d9ff _dl_cache_libcmp+0x1f (/lib/x86_64-linux-gnu/ld-2.28.so) ls 13812 [003] 2167315.222667: 4 instructions:u: 7fac71e2dab2 _dl_cache_libcmp+0xd2 (/lib/x86_64-linux-gnu/ld-2.28.so) ls 13812 [003] 2167315.222728: 16423 instructions:u: 7fac71e2477a _dl_map_object_deps+0x1ba (/lib/x86_64-linux-gnu/ld-2.28.so) ls 13812 [003] 2167315.222734: 12731 instructions:u: 7fac71e27938 _dl_name_match_p+0x68 (/lib/x86_64-linux-gnu/ld-2.28.so) ...
After:
$ perf script --itrace=i10us ls 13812 [003] 2167315.222583: 3270 instructions:u: 7fac71e2e494 __GI___tunables_init+0xf4 (/lib/x86_64-linux-gnu/ld-2.28.so) ls 13812 [003] 2167315.222667: 30902 instructions:u: 7fac71e2da0f _dl_cache_libcmp+0x2f (/lib/x86_64-linux-gnu/ld-2.28.so) ls 13812 [003] 2167315.222728: 16479 instructions:u: 7fac71e2477a _dl_map_object_deps+0x1ba (/lib/x86_64-linux-gnu/ld-2.28.so) ...
--- tools/perf/util/intel-pt-decoder/intel-pt-decoder.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-)
tools/perf/util/intel-pt-decoder/intel-pt-decoder.c +10 -3
@@ -888,16 +888,20 @@ static uint64_t intel_pt_next_period(struct intel_pt_decoder *decoder) timestamp = decoder->timestamp + decoder->timestamp_insn_cnt; masked_timestamp = timestamp & decoder->period_mask;if (decoder->continuous_period) { -if (masked_timestamp != decoder->last_masked_timestamp) +if (masked_timestamp > decoder->last_masked_timestamp)return 1 ; }else { timestamp +=1 ; masked_timestamp = timestamp & decoder->period_mask; -if (masked_timestamp != decoder->last_masked_timestamp) { +if (masked_timestamp > decoder->last_masked_timestamp) { decoder->last_masked_timestamp = masked_timestamp; decoder->continuous_period = true; } } + +if (masked_timestamp < decoder->last_masked_timestamp) +return decoder->period_ticks; +return decoder->period_ticks - (timestamp - masked_timestamp); }
@@ -926,7 +930,10 @@ static void intel_pt_sample_insn(struct intel_pt_decoder *decoder)case INTEL_PT_PERIOD_TICKS : timestamp = decoder->timestamp + decoder->timestamp_insn_cnt; masked_timestamp = timestamp & decoder->period_mask; - decoder->last_masked_timestamp = masked_timestamp; +if (masked_timestamp > decoder->last_masked_timestamp) + decoder->last_masked_timestamp = masked_timestamp; +else + decoder->last_masked_timestamp += decoder->period_ticks;break ;case INTEL_PT_PERIOD_NONE :case INTEL_PT_PERIOD_MTC :
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 68/73 perf intel-pt: Fix improved sample timestamp
In reply to Arnaldo Carvalho de MeloFrom: Adrian Hunter <[email protected]>
The decoder uses its current timestamp in samples. Usually that is a timestamp that has already passed, but in some cases it is a timestamp for a branch that the decoder is walking towards, and consequently hasn't reached.
The intel_pt_sample_time() function decides which is which, but was not handling TNT packets exactly correctly.
In the case of TNT, the timestamp applies to the first branch, so the decoder must first walk to that branch.
That means intel_pt_sample_time() should return true for TNT, and this patch makes that change. However, if the first branch is a non-taken branch (i.e. a 'N'), then intel_pt_sample_time() needs to return false for subsequent taken branches in the same TNT packet.
To handle that, introduce a new state INTEL_PT_STATE_TNT_CONT to distinguish the cases.
Note that commit 3f04d98e972b5 ("perf intel-pt: Improve sample timestamp") was also a stable fix and appears, for example, in v4.4 stable tree as commit a4ebb58fd124 ("perf intel-pt: Improve sample timestamp").
--- tools/perf/util/intel-pt-decoder/intel-pt-decoder.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-)
tools/perf/util/intel-pt-decoder/intel-pt-decoder.c +10 -3
@@ -58,6 +58,7 @@ enum intel_pt_pkt_state {INTEL_PT_STATE_NO_IP ,INTEL_PT_STATE_ERR_RESYNC ,INTEL_PT_STATE_IN_SYNC , +INTEL_PT_STATE_TNT_CONT ,INTEL_PT_STATE_TNT ,INTEL_PT_STATE_TIP , INTEL_PT_STATE_TIP_PGD,
@@ -72,8 +73,9 @@ static inline bool intel_pt_sample_time(enum intel_pt_pkt_state pkt_state)case INTEL_PT_STATE_NO_IP :case INTEL_PT_STATE_ERR_RESYNC :case INTEL_PT_STATE_IN_SYNC : -case INTEL_PT_STATE_TNT : +case INTEL_PT_STATE_TNT_CONT :return true; +case INTEL_PT_STATE_TNT :case INTEL_PT_STATE_TIP :case INTEL_PT_STATE_TIP_PGD :case INTEL_PT_STATE_FUP :
@@ -1261,7 +1263,9 @@ static int intel_pt_walk_tnt(struct intel_pt_decoder *decoder)return -ENOENT ; } decoder->tnt.count -=1 ; -if (!decoder->tnt.count) +if (decoder->tnt.count) + decoder->pkt_state =INTEL_PT_STATE_TNT_CONT ; +else decoder->pkt_state =INTEL_PT_STATE_IN_SYNC ; decoder->tnt.payload <<=1 ; decoder->state.from_ip = decoder->ip;
@@ -1292,7 +1296,9 @@ static int intel_pt_walk_tnt(struct intel_pt_decoder *decoder)if (intel_pt_insn.branch ==INTEL_PT_BR_CONDITIONAL ) { decoder->tnt.count -=1 ; -if (!decoder->tnt.count) +if (decoder->tnt.count) + decoder->pkt_state =INTEL_PT_STATE_TNT_CONT ; +else decoder->pkt_state =INTEL_PT_STATE_IN_SYNC ;if (decoder->tnt.payload &BIT63 ) { decoder->tnt.payload <<=1 ;
@@ -2372,6 +2378,7 @@ const struct intel_pt_state *intel_pt_decode(struct intel_pt_decoder *decoder) err =intel_pt_walk_trace (decoder);break ;case INTEL_PT_STATE_TNT : +case INTEL_PT_STATE_TNT_CONT : err =intel_pt_walk_tnt (decoder);if (err == -EAGAIN ) err =intel_pt_walk_trace (decoder);
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 69/73 perf intel-pt: Fix sample timestamp wrt non-taken branches
In reply to Arnaldo Carvalho de MeloFrom: Adrian Hunter <[email protected]>
The sample timestamp is updated to ensure that the timestamp represents the time of the sample and not a branch that the decoder is still walking towards. The sample timestamp is updated when the decoder returns, but the decoder does not return for non-taken branches. Update the sample timestamp then also.
Note that commit 3f04d98e972b5 ("perf intel-pt: Improve sample timestamp") was also a stable fix and appears, for example, in v4.4 stable tree as commit a4ebb58fd124 ("perf intel-pt: Improve sample timestamp").
--- tools/perf/util/intel-pt-decoder/intel-pt-decoder.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
tools/perf/util/intel-pt-decoder/intel-pt-decoder.c +4 -1
@@ -1318,8 +1318,11 @@ static int intel_pt_walk_tnt(struct intel_pt_decoder *decoder)return 0 ; } decoder->ip += intel_pt_insn.length; -if (!decoder->tnt.count) +if (!decoder->tnt.count) { + decoder->sample_timestamp = decoder->timestamp; + decoder->sample_insn_cnt = decoder->timestamp_insn_cnt;return -EAGAIN ; + } decoder->tnt.payload <<=1 ;continue ; }
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 70/73 perf docs: Add description for stderr
In reply to Arnaldo Carvalho de MeloFrom: Thomas Richter <[email protected]>
'perf report' displays recorded data on the screen and emits warnings and debug messages in the status line (last one on screen).
perf also supports the possibility to write all debug messages to stderr (instead of writing them to the status line).
This is achieved with the following command:
# ./perf --debug stderr=1 report -vvvvv -i ~/fast.data 2>/tmp/2 # ll /tmp/2 -rw-rw-r-- 1 tmricht tmricht 5420835 May 7 13:46 /tmp/2 #
The usage of variable stderr=1 is not documented, so add it to the perf man page.
--- tools/perf/Documentation/perf.txt | 2 ++ 1 file changed, 2 insertions(+)
tools/perf/Documentation/perf.txt +2 -0
@@ -22,6 +22,8 @@ OPTIONS verbose - general debug messages ordered-events - ordered events object debug messages data-convert - data convert command debug messages + stderr - write debug output (option -v) to stderr + in browser mode --buildid-dir:: Setup buildid cache directory. It has higher priority than
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 71/73 perf tools: Add a 'percore' event qualifier
In reply to Arnaldo Carvalho de MeloFrom: Jin Yao <[email protected]>
Add a 'percore' event qualifier, like cpu/event=0,umask=0x3,percore=1/, that sums up the event counts for both hardware threads in a core.
We can already do this with --per-core, but it's often useful to do this together with other metrics that are collected per hardware thread. So we need to support this per-core counting on a event level.
This can be implemented in only the user tool, no kernel support needed.
v4:
---
1. Add Arnaldo's patch which updates the documentation for this new qualifier.
2. Rebase to latest perf/core branch
v3:
---
Simplify the code according to Jiri's comments.
Before:
"return term->val.percore ? true : false;"
Now:
"return term->val.percore;"
v2:
---
Change the qualifier name from 'coresum' to 'percore' according to comments from Jiri and Andi.
--- tools/perf/Documentation/perf-list.txt | 12 ++++++++++++ tools/perf/util/evsel.c | 2 ++ tools/perf/util/evsel.h | 3 +++ tools/perf/util/parse-events.c | 27 ++++++++++++++++++++++++++ tools/perf/util/parse-events.h | 1 + tools/perf/util/parse-events.l | 1 + 6 files changed, 46 insertions(+)
tools/perf/Documentation/perf-list.txt +12 -0
@@ -199,6 +199,18 @@ also be supplied. For example: perf stat -C 0 -e 'hv_gpci/dtbp_ptitc,phys_processor_idx=0x2/' ... +EVENT QUALIFIERS: + +It is also possible to add extra qualifiers to an event: + +percore: + +Sums up the event counts for all hardware threads in a core, e.g.: + + + perf stat -e cpu/event=0,umask=0x3,percore=1/ + + EVENT GROUPS ------------
tools/perf/util/evsel.c +2 -0
@@ -813,6 +813,8 @@ static void apply_config_terms(struct perf_evsel *evsel,break ;case PERF_EVSEL__CONFIG_TERM_DRV_CFG :break ; +case PERF_EVSEL__CONFIG_TERM_PERCORE : +break ;default :break ; }
tools/perf/util/evsel.h +3 -0
@@ -50,6 +50,7 @@ enum term_type {PERF_EVSEL__CONFIG_TERM_OVERWRITE ,PERF_EVSEL__CONFIG_TERM_DRV_CFG ,PERF_EVSEL__CONFIG_TERM_BRANCH , + PERF_EVSEL__CONFIG_TERM_PERCORE, }; struct perf_evsel_config_term {
@@ -67,6 +68,7 @@ struct perf_evsel_config_term {bool overwrite;char *branch;unsigned long max_events; +bool percore; } val;bool weak; };
@@ -158,6 +160,7 @@ struct perf_evsel {struct perf_evsel **metric_events;bool collect_stat;bool weak_group; +bool percore;const char *pmu_name;struct {perf_evsel__sb_cb_t *cb;
tools/perf/util/parse-events.c +27 -0
@@ -950,6 +950,7 @@ static const char *config_term_names[__PARSE_EVENTS__TERM_TYPE_NR] = { [PARSE_EVENTS__TERM_TYPE_OVERWRITE] = "overwrite", [PARSE_EVENTS__TERM_TYPE_NOOVERWRITE] = "no-overwrite", [PARSE_EVENTS__TERM_TYPE_DRV_CFG] = "driver-config", + [PARSE_EVENTS__TERM_TYPE_PERCORE] = "percore", };static bool config_term_shrinked;
@@ -970,6 +971,7 @@ config_term_avail(int term_type, struct parse_events_error *err)case PARSE_EVENTS__TERM_TYPE_CONFIG2 :case PARSE_EVENTS__TERM_TYPE_NAME :case PARSE_EVENTS__TERM_TYPE_SAMPLE_PERIOD : +case PARSE_EVENTS__TERM_TYPE_PERCORE :return true;default :if (!err)
@@ -1061,6 +1063,14 @@ do { \case PARSE_EVENTS__TERM_TYPE_MAX_EVENTS :CHECK_TYPE_VAL (NUM );break ; +case PARSE_EVENTS__TERM_TYPE_PERCORE : +CHECK_TYPE_VAL (NUM ); +if ((unsigned int )term->val.num >1 ) { + err->str =strdup ("expected 0 or 1" ); + err->idx = term->err_val; +return -EINVAL ; + } +break ;default : err->str =strdup ("unknown term" ); err->idx = term->err_term;
@@ -1199,6 +1209,10 @@ do { \case PARSE_EVENTS__TERM_TYPE_DRV_CFG :ADD_CONFIG_TERM (DRV_CFG , drv_cfg, term->val.str);break ; +case PARSE_EVENTS__TERM_TYPE_PERCORE : +ADD_CONFIG_TERM (PERCORE , percore, + term->val.num ? true : false); +break ;default :break ; }
@@ -1260,6 +1274,18 @@ int parse_events_add_tool(struct parse_events_state *parse_state,return add_event_tool (list, &parse_state->idx, tool_event); } +static bool config_term_percore (struct list_head *config_terms) +{ +struct perf_evsel_config_term *term; + +list_for_each_entry (term, config_terms, list) { +if (term->type ==PERF_EVSEL__CONFIG_TERM_PERCORE ) +return term->val.percore; + } + +return false; +} + int parse_events_add_pmu(struct parse_events_state *parse_state, struct list_head *list, char *name, struct list_head *head_config,
@@ -1333,6 +1359,7 @@ int parse_events_add_pmu(struct parse_events_state *parse_state, evsel->metric_name = info.metric_name; evsel->pmu_name = name; evsel->use_uncore_alias = use_uncore_alias; + evsel->percore =config_term_percore (&evsel->config_terms); }return evsel ?0 : -ENOMEM ;
tools/perf/util/parse-events.h +1 -0
@@ -75,6 +75,7 @@ enum {PARSE_EVENTS__TERM_TYPE_NOOVERWRITE ,PARSE_EVENTS__TERM_TYPE_OVERWRITE ,PARSE_EVENTS__TERM_TYPE_DRV_CFG , +PARSE_EVENTS__TERM_TYPE_PERCORE , __PARSE_EVENTS__TERM_TYPE_NR, };
tools/perf/util/parse-events.l +1 -0
@@ -283,6 +283,7 @@ inherit { return term(yyscanner, PARSE_EVENTS__TERM_TYPE_INHERIT); } no-inherit { return term(yyscanner, PARSE_EVENTS__TERM_TYPE_NOINHERIT); } overwrite { return term(yyscanner, PARSE_EVENTS__TERM_TYPE_OVERWRITE); } no-overwrite { return term(yyscanner, PARSE_EVENTS__TERM_TYPE_NOOVERWRITE); } +percore { return term(yyscanner, PARSE_EVENTS__TERM_TYPE_PERCORE); } , { return ','; } "/" { BEGIN(INITIAL); return '/'; } {name_minus} { return str(yyscanner, PE_NAME); }
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 72/73 perf stat: Factor out aggregate counts printing
In reply to Arnaldo Carvalho de MeloFrom: Jin Yao <[email protected]>
Move the aggregate counts printing to a new function print_counter_aggrdata, which will be used in following patches.
--- tools/perf/util/stat-display.c | 64 +++++++++++++++++++++------------- 1 file changed, 39 insertions(+), 25 deletions(-)
tools/perf/util/stat-display.c +39 -25
@@ -594,6 +594,41 @@ static void aggr_cb(struct perf_stat_config *config, } } +static void print_counter_aggrdata (struct perf_stat_config *config, +struct perf_evsel *counter,int s, +char *prefix,bool metric_only, +bool *first) +{ +struct aggr_data ad; +FILE *output = config->output; +u64 ena, run, val; +int id, nr; +double uval; + + ad.id = id = config->aggr_map->map[s]; + ad.val = ad.ena = ad.run =0 ; + ad.nr =0 ; +if (!collect_data (config, counter, aggr_cb, &ad)) +return ; + + nr = ad.nr; + ena = ad.ena; + run = ad.run; + val = ad.val; +if (*first && metric_only) { + *first = false; +aggr_printout (config, counter, id, nr); + } +if (prefix && !metric_only) +fprintf (output,"%s" , prefix); + + uval = val * counter->scale; +printout (config, id, nr, counter, uval, prefix, + run, ena,1.0 , &rt_stat); +if (!metric_only) +fputc ('\n' , output); +} +static void print_aggr (struct perf_stat_config *config,struct perf_evlist *evlist,char *prefix)
@@ -601,9 +636,7 @@ static void print_aggr(struct perf_stat_config *config,bool metric_only = config->metric_only;FILE *output = config->output;struct perf_evsel *counter; -int s, id, nr; -double uval; -u64 ena, run, val; +int s;bool first;if (!(config->aggr_map || config->aggr_get_id))
@@ -616,33 +649,14 @@ static void print_aggr(struct perf_stat_config *config,* Without each counter has its own line. */ for (s =0 ; s < config->aggr_map->nr; s++) { -struct aggr_data ad;if (prefix && metric_only)fprintf (output,"%s" , prefix); - ad.id = id = config->aggr_map->map[s]; first = true;evlist__for_each_entry (evlist, counter) { - ad.val = ad.ena = ad.run =0 ; - ad.nr =0 ; -if (!collect_data (config, counter, aggr_cb, &ad)) -continue ; - nr = ad.nr; - ena = ad.ena; - run = ad.run; - val = ad.val; -if (first && metric_only) { - first = false; -aggr_printout (config, counter, id, nr); - } -if (prefix && !metric_only) -fprintf (output,"%s" , prefix); - - uval = val * counter->scale; -printout (config, id, nr, counter, uval, prefix, - run, ena,1.0 , &rt_stat); -if (!metric_only) -fputc ('\n' , output); +print_counter_aggrdata (config, counter, s, + prefix, metric_only, + &first); }if (metric_only)fputc ('\n' , output);
Signature · 1 line
2.20.1
Arnaldo Carvalho de Melo
maintainerPATCH 73/73 perf stat: Support 'percore' event qualifier
In reply to Arnaldo Carvalho de MeloFrom: Jin Yao <[email protected]>
With this patch, we can use the 'percore' event qualifier in perf-stat.
root@skl:/tmp# perf stat -e cpu/event=0,umask=0x3,percore=1/,cpu/event=0,umask=0x3/ -a -A -I1000
1.000773050 S0-C0 98,352,832 cpu/event=0,umask=0x3,percore=1/ (50.01%)
1.000773050 S0-C1 103,763,057 cpu/event=0,umask=0x3,percore=1/ (50.02%)
1.000773050 S0-C2 196,776,995 cpu/event=0,umask=0x3,percore=1/ (50.02%)
1.000773050 S0-C3 176,493,779 cpu/event=0,umask=0x3,percore=1/ (50.02%)
1.000773050 CPU0 47,699,641 cpu/event=0,umask=0x3/ (50.02%)
1.000773050 CPU1 49,052,451 cpu/event=0,umask=0x3/ (49.98%)
1.000773050 CPU2 102,771,422 cpu/event=0,umask=0x3/ (49.98%)
1.000773050 CPU3 100,784,662 cpu/event=0,umask=0x3/ (49.98%)
1.000773050 CPU4 43,171,342 cpu/event=0,umask=0x3/ (49.98%)
1.000773050 CPU5 54,152,158 cpu/event=0,umask=0x3/ (49.98%)
1.000773050 CPU6 93,618,410 cpu/event=0,umask=0x3/ (49.98%)
1.000773050 CPU7 74,477,589 cpu/event=0,umask=0x3/ (49.99%)In this example, we count the event 'ref-cycles' per-core and per-CPU in one perf stat command-line. From the output, we can see:
S0-C0 = CPU0 + CPU4
S0-C1 = CPU1 + CPU5
S0-C2 = CPU2 + CPU6
S0-C3 = CPU3 + CPU7
So the result is expected (tiny difference is ignored).
Note that, the 'percore' event qualifier needs to use with option '-A'.
--- tools/perf/Documentation/perf-stat.txt | 4 +++ tools/perf/builtin-stat.c | 21 +++++++++++++ tools/perf/util/stat-display.c | 43 +++++++++++++++++++++++--- tools/perf/util/stat.c | 8 +++-- 4 files changed, 69 insertions(+), 7 deletions(-)
tools/perf/Documentation/perf-stat.txt +4 -0
@@ -43,6 +43,10 @@ report:: param1 and param2 are defined as formats for the PMU in /sys/bus/event_source/devices/<pmu>/format/* + 'percore' is a event qualifier that sums up the event counts for both + hardware threads in a core. For example: + perf stat -A -a -e cpu/event,percore=1/,otherevent ... + - a symbolically formed event like 'pmu/config=M,config1=N,config2=K/' where M, N, K are numbers (in decimal, hex, octal format). Acceptable values for each of 'config', 'config1' and 'config2'
tools/perf/builtin-stat.c +21 -0
@@ -847,6 +847,18 @@ static int perf_stat__get_core_cached(struct perf_stat_config *config,return perf_stat__get_aggr (config, perf_stat__get_core, map, idx); } +static bool term_percore_set (void ) +{ +struct perf_evsel *counter; + +evlist__for_each_entry (evsel_list, counter) { +if (counter->percore) +return true; + } + +return false; +} +static int perf_stat_init_aggr_mode (void ) {int nr;
@@ -867,6 +879,15 @@ static int perf_stat_init_aggr_mode(void) stat_config.aggr_get_id = perf_stat__get_core_cached;break ;case AGGR_NONE : +if (term_percore_set ()) { +if (cpu_map__build_core_map (evsel_list->cpus, + &stat_config.aggr_map)) { +perror ("cannot build core map" ); +return -1 ; + } + stat_config.aggr_get_id = perf_stat__get_core_cached; + } +break ;case AGGR_GLOBAL :case AGGR_THREAD :case AGGR_UNSET :
tools/perf/util/stat-display.c +39 -4
61 lines not on this page · raw message
tools/perf/util/stat.c +5 -3
@@ -277,9 +277,11 @@ process_counter_values(struct perf_stat_config *config, struct perf_evsel *evselif (!evsel->snapshot)perf_evsel__compute_deltas (evsel, cpu, thread, count);perf_counts_values__scale (count, config->scale,NULL ); -if (config->aggr_mode ==AGGR_NONE ) -perf_stat__update_shadow_stats (evsel, count->val, cpu, - &rt_stat); +if ((config->aggr_mode ==AGGR_NONE ) && (!evsel->percore)) { +perf_stat__update_shadow_stats (evsel, count->val, + cpu, &rt_stat); + } +if (config->aggr_mode ==AGGR_THREAD ) {if (config->stats) perf_stat__update_shadow_stats(evsel,
Signature · 1 line
2.20.1
Quoted 384 lines · Arnaldo Carvalho de Melo
Hi Ingo,
Please consider pulling, I pulled tip/perf/urgent into
tip/pref/core, IIRC was just a fast forward at that point, yeap, just did it again and it still is:$ git checkout -b t tip/perf/core Branch 't' set up to track remote branch 'perf/core' from 'tip'. Switched to a new branch 't' $ git merge tip/perf/urgent Updating d15d356887e7..c7a286577d75 Fast-forward <SNIP>IIRC Jiri needs this for a pile of patches he submitted and
that I'll process next,Best regards,
- Arnaldo
Test results at the end of this message, as usual.
The following changes since commit
6b89d4c1ae8596a8c9240f169ef108704de373f2:perf/x86/intel: Fix INTEL_FLAGS_EVENT_CONSTRAINT* masking (2019-05-10 08:04:17 +0200)
are available in the Git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux.gittags/perf-core-for-mingo-5.2-20190517for you to fetch changes up to
4fc4d8dfa056dfd48afe73b9ea3b7570ceb80b9c:perf stat: Support 'percore' event qualifier (2019-05-16 14:17:24 -0300)
----------------------------------------------------------------
perf/core improvements and fixes:perf.data:
Alexey Budankov:
- Streaming compression of perf ring buffer into
PERF_RECORD_COMPRESSEDuser space records, resulting in ~3-5x perf.data file size reduction on variety of tested workloads what saves storage space on larger server systems where perf.data size can easily reach several tens or even hundreds of GiBs, especially when profiling with DWARF-based stacks and tracing of context switches.perf record:
Arnaldo Carvalho de Melo
- Improve -user-regs/intr-regs suggestions to overcome errors.
perf annotate:
Jin Yao:
- Remove
hist__account_cycles()from callback, speeding up branch processing (perf record -b).perf stat:
- Add a 'percore' event qualifier, e.g.: -e cpu/event=0,umask=0x3,percore=1/, that sums up the event counts for both hardware threads in a core.
We can already do this with --per-core, but it's often useful to do this together with other metrics that are collected per hardware thread.
I.e. now its possible to do this per-event, and have it mixed with other events not aggregated by core.
core libraries:
Donald Yandt:
- Check for errors when doing fgets(/proc/version).
Jiri Olsa:
- Speed up report for perf compiled with linbunwind.
tools headers:
Arnaldo Carvalho de Melo
- Update
memcpy_64.S, x86'skvm.handpt_regs.h.arm64:
Florian Fainelli:
- Map Brahma-B53 CPUID to cortex-a53 events.
- Add Cortex-A57 and Cortex-A72 events.
csky:
Mao Han:
- Add DWARF register mappings for libdw, allowing --call-graph=dwarf to work on the C-SKY arch.
x86:
Andi Kleen/Kan Liang:
- Add support for recording and printing XMM registers, available, for instance, on Icelake.
Kan Liang:
- Add
uncore_upi(Intel's "Ultra Path Interconnect" events) JSON support. UPI replaced the Intel QuickPath Interconnect (QPI) in Xeon Skylake-SP.Intel PT:
Adrian Hunter
. Fix instructions sampling rate.
. Timestamp fixes.
. Improve exported-sql-viewer GUI, allowing, for instance, to copy'n'paste
the trees, useful for e-mailing.Documentation:
Thomas Richter:
- Add description for 'perf --debug stderr=1', which redirects stderr to stdout.
libtraceevent:
Tzvetomir Stoyanov:
- Add man pages for the various APIs.
Signed-off-by Arnaldo Carvalho de Melo <[email protected]>----------------------------------------------------------------
Adrian Hunter (9):
perf scripts python:exported-sql-viewer.py: Move view creation
perf scripts python:exported-sql-viewer.py: Fix error when shrinking / enlarging font perf scripts python:exported-sql-viewer.py: Add tree level
perf scripts python:exported-sql-viewer.py: Add copy to clipboard
perf scripts python:exported-sql-viewer.py: Add context menu
perf scripts python:exported-sql-viewer.py: Add 'About' dialog box perf intel-pt: Fix instructions sampling rate
perf intel-pt: Fix improved sample timestamp
perf intel-pt: Fix sample timestamp wrt non-taken branchesAlexey Budankov (11):
perf session: Define 'bytes_transferred' and 'bytes_compressed' metrics perf record: Implement COMPRESSED event record and its attributes perf mmap: Implement dedicated memory buffer for data compression perf tools: Introduce Zstd streaming based compression API perf record: Implement compression for serial trace streaming perf record: Implement compression for AIO trace streaming perf report: Add stub processing of compressed events for -D perf record: Implement -z,--compression_level[=<n>] option perf report: Implement perf.data record decompression
perf inject: Enable COMPRESSED record decompression
perf tests: Implement Zstd comp/decomp integration testAndi Kleen (1):
perf tools x86: Add support for recording and printing XMM registersArnaldo Carvalho de Melo (8):
tools arch: Updatearch/x86/lib/memcpy_64.Scopy used in 'perf bench mem memcpy' tools arch uapi: Sync the x86kvm.hcopy
tools x86 uapi asm: Sync thept_regs.hcopy with the kernel sources
tools pci: Do not deletepcitest.shin 'make clean'
perf record: Fix suggestion to get list of registers usable with --user-regs and --intr-regs perf parse-regs: Improve error output when faced with unknown register name perf build tests: Add NO_LIBZSTD=1 tomake_minimal
perf test zstd: Fixup verbose mode outputColin Ian King (1):
perf test: Fix spelling mistake "leadking" -> "leaking"Donald Yandt (1):
perf machine: Null-terminate version char array upon fgets(/proc/version) errorFlorian Fainelli (3):
perf vendor events arm64: Remove [[:xdigit:]] wildcard perf vendor events arm64: Map Brahma-B53 CPUID to cortex-a53 events perf vendor events arm64: Add Cortex-A57 and Cortex-A72 eventsJin Yao (4):
perf annotate: Removehist__account_cycles()from callback perf tools: Add a 'percore' event qualifier
perf stat: Factor out aggregate counts printing perf stat: Support 'percore' event qualifierJiri Olsa (1):
perf tools: Speed up report for perf compiled with linwunwindKan Liang (4):
perf vendor events intel: Adduncore_upiJSON support perf parse-regs: Splitparse_regs
perf parse-regs: Add generic support for arch__intr/user_reg_mask() perf regs x86: Add X86 specificarch__intr_reg_mask()Mao Han (1):
csky: Add support for libdwThomas Richter (1):
perf docs: Add description for stderrTzvetomir Stoyanov (27):
tools lib traceevent: Remove hard coded install paths from pkg-config file tools lib traceevent: Introduce man pages
tools lib traceevent: Add support for man pages with multiple names tools lib traceevent: Man pages fortep_handlerrelated APIs
tools lib traceevent: Man page forheader_pageAPIs
tools lib traceevent: Man page for get/set cpus APIs
tools lib traceevent: Man page for file endian APIs
tools lib traceevent: Man page for host endian APIs
tools lib traceevent: Man page for page size APIs
tools lib traceevent: Man page fortep_strerror()
tools lib traceevent: Man pages for event handler APIs
tools lib traceevent: Man pages for function related libtraceevent APIs tools lib traceevent: Man pages for registering print function
tools lib traceevent: Man page fortep_read_number()
tools lib traceevent: Man pages for event find APIs
tools lib traceevent: Man page for list events APIs
tools lib traceevent: Man pages for libtraceevent event get APIs tools lib traceevent: Man pages for find field APIs
tools lib traceevent: Man pages for get field value APIs
tools lib traceevent: Man pages for print field APIs
tools lib traceevent: Man page fortep_read_number_field()
tools lib traceevent: Man pages for event fields APIs
tools lib traceevent: Man pages for event filter APIs
tools lib traceevent: Man pages for parse event APIs
tools lib traceevent: Man page fortep_parse_header_page()
tools lib traceevent: Man pages for APIs used to extract common fields from a record tools lib traceevent: Man pages for trace sequences APIsZenghui Yu (1):
perf jevents: Remove unused variabletools/arch/csky/include/uapi/asm/perf_regs.h | 51 ++++ tools/arch/x86/include/uapi/asm/kvm.h | 1 + tools/arch/x86/include/uapi/asm/perf_regs.h | 23 +- tools/arch/x86/lib/memcpy_64.S | 3 +- tools/lib/traceevent/Documentation/Makefile | 207 +++++++++++++ tools/lib/traceevent/Documentation/asciidoc.conf | 120 ++++++++ .../Documentation/libtraceevent-commands.txt | 153 ++++++++++ .../Documentation/libtraceevent-cpus.txt | 77 +++++ .../Documentation/libtraceevent-endian_read.txt | 78 +++++ .../Documentation/libtraceevent-event_find.txt | 103 +++++++ .../Documentation/libtraceevent-event_get.txt | 99 ++++++ .../Documentation/libtraceevent-event_list.txt | 122 ++++++++ .../Documentation/libtraceevent-field_find.txt | 118 +++++++ .../Documentation/libtraceevent-field_get_val.txt | 122 ++++++++ .../Documentation/libtraceevent-field_print.txt | 126 ++++++++ .../Documentation/libtraceevent-field_read.txt | 81 +++++ .../Documentation/libtraceevent-fields.txt | 105 +++++++ .../Documentation/libtraceevent-file_endian.txt | 91 ++++++ .../Documentation/libtraceevent-filter.txt | 209 +++++++++++++ .../Documentation/libtraceevent-func_apis.txt | 183 +++++++++++ .../Documentation/libtraceevent-func_find.txt | 88 ++++++ .../Documentation/libtraceevent-handle.txt | 101 ++++++ .../Documentation/libtraceevent-header_page.txt | 102 +++++++ .../Documentation/libtraceevent-host_endian.txt | 104 +++++++ .../Documentation/libtraceevent-long_size.txt | 78 +++++ .../Documentation/libtraceevent-page_size.txt | 82 +++++ .../Documentation/libtraceevent-parse_event.txt | 90 ++++++ .../Documentation/libtraceevent-parse_head.txt | 82 +++++ .../Documentation/libtraceevent-record_parse.txt | 137 +++++++++ .../libtraceevent-reg_event_handler.txt | 156 ++++++++++ .../Documentation/libtraceevent-reg_print_func.txt | 155 ++++++++++ .../Documentation/libtraceevent-set_flag.txt | 104 +++++++ .../Documentation/libtraceevent-strerror.txt | 85 ++++++ .../Documentation/libtraceevent-tseq.txt | 158 ++++++++++ .../lib/traceevent/Documentation/libtraceevent.txt | 203 ++++++++++++ .../lib/traceevent/Documentation/manpage-1.72.xsl | 14 + .../lib/traceevent/Documentation/manpage-base.xsl | 35 +++ .../Documentation/manpage-bold-literal.xsl | 17 ++ .../traceevent/Documentation/manpage-normal.xsl | 13 + .../Documentation/manpage-suppress-sp.xsl | 21 ++ tools/lib/traceevent/Makefile | 46 ++- tools/lib/traceevent/libtraceevent.pc.template | 4 +- tools/pci/Makefile | 4 +- tools/perf/Documentation/perf-list.txt | 12 + tools/perf/Documentation/perf-record.txt | 8 +- tools/perf/Documentation/perf-stat.txt | 4 + tools/perf/Documentation/perf.data-file-format.txt | 24 ++ tools/perf/Documentation/perf.txt | 2 + tools/perf/Makefile.config | 6 +- tools/perf/arch/csky/Build | 1 + tools/perf/arch/csky/Makefile | 3 + tools/perf/arch/csky/include/perf_regs.h | 100 ++++++ tools/perf/arch/csky/util/Build | 2 + tools/perf/arch/csky/util/dwarf-regs.c | 49 +++ tools/perf/arch/csky/util/unwind-libdw.c | 77 +++++ tools/perf/arch/x86/include/perf_regs.h | 26 +- tools/perf/arch/x86/util/perf_regs.c | 44 +++ tools/perf/builtin-annotate.c | 4 +- tools/perf/builtin-inject.c | 4 + tools/perf/builtin-record.c | 229 ++++++++++++-- tools/perf/builtin-report.c | 16 +- tools/perf/builtin-stat.c | 21 ++ tools/perf/perf.h | 1 + .../arm64/arm/cortex-a57-a72/core-imp-def.json | 179 +++++++++++ tools/perf/pmu-events/arch/arm64/mapfile.csv | 5 +- tools/perf/pmu-events/jevents.c | 2 +- tools/perf/scripts/python/exported-sql-viewer.py | 340 ++++++++++++++++++++- tools/perf/tests/dso-data.c | 4 +- tools/perf/tests/make | 2 +- tools/perf/tests/shell/record+zstd_comp_decomp.sh | 34 +++ tools/perf/util/Build | 2 + tools/perf/util/annotate.c | 2 +- tools/perf/util/compress.h | 53 ++++ tools/perf/util/env.h | 11 + tools/perf/util/event.c | 1 + tools/perf/util/event.h | 7 + tools/perf/util/evlist.c | 8 +- tools/perf/util/evlist.h | 2 +- tools/perf/util/evsel.c | 2 + tools/perf/util/evsel.h | 3 + tools/perf/util/header.c | 53 ++++ tools/perf/util/header.h | 1 + .../perf/util/intel-pt-decoder/intel-pt-decoder.c | 31 +- tools/perf/util/machine.c | 3 +- tools/perf/util/mmap.c | 102 ++----- tools/perf/util/mmap.h | 16 +- tools/perf/util/parse-events.c | 27 ++ tools/perf/util/parse-events.h | 1 + tools/perf/util/parse-events.l | 1 + tools/perf/util/parse-regs-options.c | 33 +- tools/perf/util/parse-regs-options.h | 3 +- tools/perf/util/perf_regs.c | 10 + tools/perf/util/perf_regs.h | 3 + tools/perf/util/session.c | 133 +++++++- tools/perf/util/session.h | 14 + tools/perf/util/stat-display.c | 107 +++++-- tools/perf/util/stat.c | 8 +- tools/perf/util/thread.c | 3 +- tools/perf/util/tool.h | 2 + tools/perf/util/unwind-libunwind-local.c | 6 - tools/perf/util/unwind-libunwind.c | 10 + tools/perf/util/zstd.c | 111 +++++++ 102 files changed, 5703 insertions(+), 216 deletions(-) create mode 100644 tools/arch/csky/include/uapi/asm/perf_regs.h create mode 100644 tools/lib/traceevent/Documentation/Makefile create mode 100644 tools/lib/traceevent/Documentation/asciidoc.conf create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-commands.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-cpus.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-endian_read.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-event_find.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-event_get.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-event_list.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-field_find.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-field_get_val.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-field_print.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-field_read.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-fields.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-file_endian.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-filter.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-func_apis.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-func_find.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-handle.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-header_page.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-host_endian.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-long_size.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-page_size.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-parse_event.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-parse_head.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-record_parse.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-reg_event_handler.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-reg_print_func.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-set_flag.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-strerror.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent-tseq.txt create mode 100644 tools/lib/traceevent/Documentation/libtraceevent.txt create mode 100644 tools/lib/traceevent/Documentation/manpage-1.72.xsl create mode 100644 tools/lib/traceevent/Documentation/manpage-base.xsl create mode 100644 tools/lib/traceevent/Documentation/manpage-bold-literal.xsl create mode 100644 tools/lib/traceevent/Documentation/manpage-normal.xsl create mode 100644 tools/lib/traceevent/Documentation/manpage-suppress-sp.xsl create mode 100644 tools/perf/arch/csky/Build
Quoted 8 lines
create mode 100644 tools/perf/arch/csky/Makefile create mode 100644 tools/perf/arch/csky/include/perf_regs.h create mode 100644 tools/perf/arch/csky/util/Build create mode 100644 tools/perf/arch/csky/util/dwarf-regs.c create mode 100644 tools/perf/arch/csky/util/unwind-libdw.c create mode 100644 tools/perf/pmu-events/arch/arm64/arm/cortex-a57-a72/core-imp-def.json create mode 100755 tools/perf/tests/shell/record+zstd_comp_decomp.sh create mode 100644 tools/perf/util/zstd.c
Pulled, thanks a lot Arnaldo!
Ingo
PATCH 1/1 tools/headers: Synchronize kernel ABI headers
In reply to Arnaldo Carvalho de MeloPick up the latest v5.2-to-be kernel ABI headers and synchronize them with tooling:
- arch/x86/entry/syscalls/syscall_64.tbl => tools/perf/arch/x86/entry/syscalls/syscall_64.tbl # new syscalls - arch/x86/include/asm/cpufeatures.h => tools/arch/x86/include/asm/cpufeatures.h # new CPUID flags - include/uapi/drm/drm.h => tools/include/uapi/drm/drm.h # new 'syncobj' DRM ABI - include/uapi/drm/i915_drm.h => tools/include/uapi/drm/i915_drm.h # new extensible DRM ABI - include/uapi/linux/fcntl.h => tools/include/uapi/linux/fcntl.h # new AT_RECURSIVE - include/uapi/linux/fs.h => tools/include/uapi/linux/fs.h # new SYNC_FILE_RANGE_WRITE_AND_WAIT - include/uapi/linux/mount.h => tools/include/uapi/linux/mount.h # new VFS system calls: fspick, fsmount, fsconfig, fsopen, move_mount, open_tree - include/uapi/linux/sched.h => tools/include/uapi/linux/sched.h # new CLONE_PIDFD
All of these are new ABI additions with no impact on existing tooling, so we copy the kernel headers with no other changes necessary.
--- tools/arch/x86/include/asm/cpufeatures.h | 3 + tools/include/uapi/drm/drm.h | 37 ++++ tools/include/uapi/drm/i915_drm.h | 254 +++++++++++++++------- tools/include/uapi/linux/fcntl.h | 2 + tools/include/uapi/linux/fs.h | 3 + tools/include/uapi/linux/mount.h | 62 ++++++ tools/include/uapi/linux/sched.h | 1 + tools/perf/arch/x86/entry/syscalls/syscall_64.tbl | 6 + 8 files changed, 295 insertions(+), 73 deletions(-)
tools/arch/x86/include/asm/cpufeatures.h +3 -0
@@ -344,6 +344,7 @@/* Intel-defined CPU features, CPUID level 0x00000007:0 (EDX), word 18 */ #define X86_FEATURE_AVX512_4VNNIW (18*32+ 2)/* AVX-512 Neural Network Instructions */ #define X86_FEATURE_AVX512_4FMAPS (18*32+ 3)/* AVX-512 Multiply Accumulation Single precision */ +#define X86_FEATURE_MD_CLEAR (18*32+10)/* VERW clears CPU buffers */ #define X86_FEATURE_TSX_FORCE_ABORT (18*32+13)/* "" TSX_FORCE_ABORT */ #define X86_FEATURE_PCONFIG (18*32+18)/* Intel PCONFIG */ #define X86_FEATURE_SPEC_CTRL (18*32+26)/* "" Speculation Control (IBRS + IBPB) */
@@ -382,5 +383,7 @@#define X86_BUG_SPECTRE_V2 X86_BUG(16)/* CPU is affected by Spectre variant 2 attack with indirect branches */ #define X86_BUG_SPEC_STORE_BYPASS X86_BUG(17)/* CPU is affected by speculative store bypass attack */ #define X86_BUG_L1TF X86_BUG(18)/* CPU is affected by L1 Terminal Fault */ +#define X86_BUG_MDS X86_BUG(19)/* CPU is affected by Microarchitectural data sampling */ +#define X86_BUG_MSBDS_ONLY X86_BUG(20)/* CPU is only affected by the MSDBS variant of BUG_MDS */ #endif /* _ASM_X86_CPUFEATURES_H */
tools/include/uapi/drm/drm.h +37 -0
@@ -649,6 +649,7 @@ struct drm_gem_open {#define DRM_CAP_PAGE_FLIP_TARGET 0x11#define DRM_CAP_CRTC_IN_VBLANK_EVENT 0x12#define DRM_CAP_SYNCOBJ 0x13 +#define DRM_CAP_SYNCOBJ_TIMELINE 0x14/** DRM_IOCTL_GET_CAP ioctl argument type */ struct drm_get_cap {
@@ -735,8 +736,18 @@ struct drm_syncobj_handle {__u32 pad; }; +struct drm_syncobj_transfer { +__u32 src_handle; +__u32 dst_handle; +__u64 src_point; +__u64 dst_point; +__u32 flags; +__u32 pad; +}; +#define DRM_SYNCOBJ_WAIT_FLAGS_WAIT_ALL (1 << 0)#define DRM_SYNCOBJ_WAIT_FLAGS_WAIT_FOR_SUBMIT (1 << 1) +#define DRM_SYNCOBJ_WAIT_FLAGS_WAIT_AVAILABLE (1 << 2)/* wait for time point to become available */ struct drm_syncobj_wait {__u64 handles;/* absolute timeout */
@@ -747,12 +758,33 @@ struct drm_syncobj_wait {__u32 pad; }; +struct drm_syncobj_timeline_wait { +__u64 handles; +/* wait on specific timeline point for every handles*/ +__u64 points; +/* absolute timeout */ +__s64 timeout_nsec; +__u32 count_handles; +__u32 flags; +__u32 first_signaled;/* only valid when not waiting all */ +__u32 pad; +}; + +struct drm_syncobj_array {__u64 handles;__u32 count_handles;__u32 pad; }; +struct drm_syncobj_timeline_array { +__u64 handles; +__u64 points; +__u32 count_handles; +__u32 pad; +}; + +/* Query current scanout sequence number */ struct drm_crtc_get_sequence {__u32 crtc_id;/* requested crtc_id */
@@ -909,6 +941,11 @@ extern "C" {#define DRM_IOCTL_MODE_GET_LEASE DRM_IOWR(0xC8, struct drm_mode_get_lease)#define DRM_IOCTL_MODE_REVOKE_LEASE DRM_IOWR(0xC9, struct drm_mode_revoke_lease) +#define DRM_IOCTL_SYNCOBJ_TIMELINE_WAIT DRM_IOWR(0xCA, struct drm_syncobj_timeline_wait) +#define DRM_IOCTL_SYNCOBJ_QUERY DRM_IOWR(0xCB, struct drm_syncobj_timeline_array) +#define DRM_IOCTL_SYNCOBJ_TRANSFER DRM_IOWR(0xCC, struct drm_syncobj_transfer) +#define DRM_IOCTL_SYNCOBJ_TIMELINE_SIGNAL DRM_IOWR(0xCD, struct drm_syncobj_timeline_array) +/** * Device specific ioctls should only be in their respective headers * The device specific ioctl range is from 0x40 to 0x9f.
tools/include/uapi/drm/i915_drm.h +181 -73
356 lines not on this page · raw message
tools/include/uapi/linux/fcntl.h +2 -0
@@ -91,5 +91,7 @@#define AT_STATX_FORCE_SYNC 0x2000/* - Force the attributes to be sync'd with the server */ #define AT_STATX_DONT_SYNC 0x4000/* - Don't sync attributes with the server */ +#define AT_RECURSIVE 0x8000/* Apply to the entire subtree */ +#endif /* _UAPI_LINUX_FCNTL_H */
tools/include/uapi/linux/fs.h +3 -0
@@ -320,6 +320,9 @@ struct fscrypt_key {#define SYNC_FILE_RANGE_WAIT_BEFORE 1#define SYNC_FILE_RANGE_WRITE 2#define SYNC_FILE_RANGE_WAIT_AFTER 4 +#define SYNC_FILE_RANGE_WRITE_AND_WAIT (SYNC_FILE_RANGE_WRITE | \ + SYNC_FILE_RANGE_WAIT_BEFORE | \ + SYNC_FILE_RANGE_WAIT_AFTER)/* * Flags for preadv2/pwritev2:
tools/include/uapi/linux/mount.h +62 -0
66 lines not on this page · raw message
tools/include/uapi/linux/sched.h +1 -0
@@ -10,6 +10,7 @@#define CLONE_FS 0x00000200/* set if fs info shared between processes */ #define CLONE_FILES 0x00000400/* set if open files shared between processes */ #define CLONE_SIGHAND 0x00000800/* set if signal handlers and blocked signals shared */ +#define CLONE_PIDFD 0x00001000/* set if a pidfd should be placed in parent */ #define CLONE_PTRACE 0x00002000/* set if we want to let tracing continue on the child too */ #define CLONE_VFORK 0x00004000/* set if the parent wants the child to wake it up on mm_release */ #define CLONE_PARENT 0x00008000/* set if we want to have the same parent as the cloner */
tools/perf/arch/x86/entry/syscalls/syscall_64.tbl +6 -0
@@ -343,6 +343,12 @@ 332 common statx __x64_sys_statx 333 common io_pgetevents __x64_sys_io_pgetevents 334 common rseq __x64_sys_rseq +335 common open_tree __x64_sys_open_tree +336 common move_mount __x64_sys_move_mount +337 common fsopen __x64_sys_fsopen +338 common fsconfig __x64_sys_fsconfig +339 common fsmount __x64_sys_fsmount +340 common fspick __x64_sys_fspick # don't use numbers 387 through 423, add new calls after the last # 'common' entry 424 common pidfd_send_signal __x64_sys_pidfd_send_signal
Re: [PATCH] tools/headers: Synchronize kernel ABI headers
Em Sat, May 18, 2019 at 10:42:23AM +0200, Ingo Molnar escreveu:
Quoted 11 lines
Pick up the latest v5.2-to-be kernel ABI headers and synchronize them with tooling:
- arch/x86/entry/syscalls/syscall_64.tbl => tools/perf/arch/x86/entry/syscalls/syscall_64.tbl # new syscalls - arch/x86/include/asm/cpufeatures.h => tools/arch/x86/include/asm/cpufeatures.h # new CPUID flags - include/uapi/drm/drm.h => tools/include/uapi/drm/drm.h # new 'syncobj' DRM ABI - include/uapi/drm/i915_drm.h => tools/include/uapi/drm/i915_drm.h # new extensible DRM ABI - include/uapi/linux/fcntl.h => tools/include/uapi/linux/fcntl.h # new AT_RECURSIVE - include/uapi/linux/fs.h => tools/include/uapi/linux/fs.h # new SYNC_FILE_RANGE_WRITE_AND_WAIT - include/uapi/linux/mount.h => tools/include/uapi/linux/mount.h # new VFS system calls: fspick, fsmount, fsconfig, fsopen, move_mount, open_tree - include/uapi/linux/sched.h => tools/include/uapi/linux/sched.h # new CLONE_PIDFD
Quoted 1 line
All of these are new ABI additions with no impact on existing tooling,
There is some impact in a number of them, for instance:
[acme@quaco perf]$ diff -u tools/perf/arch/x86/entry/syscalls/syscall_64.tbl arch/x86/entry/syscalls/syscall_64.tbl
x86/entry/syscalls/syscall_64.tbl renamed from perf/arch/x86/entry/syscalls/syscall_64.tbl +6 -0
@@ -343,6 +343,12 @@ 332 common statx __x64_sys_statx 333 common io_pgetevents __x64_sys_io_pgetevents 334 common rseq __x64_sys_rseq +335 common open_tree __x64_sys_open_tree +336 common move_mount __x64_sys_move_mount +337 common fsopen __x64_sys_fsopen +338 common fsconfig __x64_sys_fsconfig +339 common fsmount __x64_sys_fsmount +340 common fspick __x64_sys_fspick # don't use numbers 387 through 423, add new calls after the last # 'common' entry 424 common pidfd_send_signal __x64_sys_pidfd_send_signal
[acme@quaco perf]$
These will enable 'perf trace' to know about these syscalls, i.e. we will be able to say:
perf trace -e fs*
And have it trace just those fsopen, fsconfig, fsmount and fspick syscalls (or any other that starts with fs, that is). Looking at this makes one think if we should have a new syscall group like we have here:
[acme@quaco perf]$ ls -la tools/perf/trace/strace/groups/ total 16
drwxrwxr-x. 2 acme acme 4096 May 13 14:46 . drwxrwxr-x. 3 acme acme 4096 Apr 10 10:20 .. -rw-rw-r--. 1 acme acme 136 May 13 14:46 file -rw-rw-r--. 1 acme acme 584 May 13 14:46 string [acme@quaco perf]$
One for file oriented syscalls, the other for syscalls that operate on strings (pathnames, etc).
That i915_drm.h looks related to:
[acme@quaco perf]$ tools/perf/trace/beauty/drm_ioctl.sh | head #ifndef DRM_COMMAND_BASE
#define DRM_COMMAND_BASE 0x40
#endif
static const char *drm_ioctl_cmds[] = {
[0x00] = "VERSION",
[0x01] = "GET_UNIQUE",
[0x02] = "GET_MAGIC",
[0x03] = "IRQ_BUSID",
[0x04] = "GET_MAP",
[0x05] = "GET_CLIENT",
[acme@quaco perf]$ tools/perf/trace/beauty/drm_ioctl.sh | tail
[DRM_COMMAND_BASE + 0x31] = "I915_REG_READ",
[DRM_COMMAND_BASE + 0x32] = "I915_GET_RESET_STATS",
[DRM_COMMAND_BASE + 0x33] = "I915_GEM_USERPTR",
[DRM_COMMAND_BASE + 0x34] = "I915_GEM_CONTEXT_GETPARAM",
[DRM_COMMAND_BASE + 0x35] = "I915_GEM_CONTEXT_SETPARAM",
[DRM_COMMAND_BASE + 0x36] = "I915_PERF_OPEN",
[DRM_COMMAND_BASE + 0x37] = "I915_PERF_ADD_CONFIG",
[DRM_COMMAND_BASE + 0x38] = "I915_PERF_REMOVE_CONFIG",
[DRM_COMMAND_BASE + 0x39] = "I915_QUERY",
};[acme@quaco perf]$
So I'll check that tools/perf/trace/beauty/drm_ioctl.sh is sane wrt any new additions to that header, ditto for the other headers, I'll go thru them soon.
- Arnaldo
Quoted 611 lines
so we copy the kernel headers with no other changes necessary.
Signed-off-by Ingo Molnar <[email protected]>--- tools/arch/x86/include/asm/cpufeatures.h | 3 + tools/include/uapi/drm/drm.h | 37 ++++ tools/include/uapi/drm/i915_drm.h | 254 +++++++++++++++------- tools/include/uapi/linux/fcntl.h | 2 + tools/include/uapi/linux/fs.h | 3 + tools/include/uapi/linux/mount.h | 62 ++++++ tools/include/uapi/linux/sched.h | 1 + tools/perf/arch/x86/entry/syscalls/syscall_64.tbl | 6 + 8 files changed, 295 insertions(+), 73 deletions(-)tools/arch/x86/include/asm/cpufeatures.h +3
-0@@ -344,6 +344,7 @@/* Intel-defined CPU features, CPUID level 0x00000007:0 (EDX), word 18 */ #define X86_FEATURE_AVX512_4VNNIW (18*32+ 2)/* AVX-512 Neural Network Instructions */ #define X86_FEATURE_AVX512_4FMAPS (18*32+ 3)/* AVX-512 Multiply Accumulation Single precision */ +#define X86_FEATURE_MD_CLEAR (18*32+10)/* VERW clears CPU buffers */ #define X86_FEATURE_TSX_FORCE_ABORT (18*32+13)/* "" TSX_FORCE_ABORT */ #define X86_FEATURE_PCONFIG (18*32+18)/* Intel PCONFIG */ #define X86_FEATURE_SPEC_CTRL (18*32+26)/* "" Speculation Control (IBRS + IBPB) */ @@ -382,5 +383,7 @@#define X86_BUG_SPECTRE_V2 X86_BUG(16)/* CPU is affected by Spectre variant 2 attack with indirect branches */ #define X86_BUG_SPEC_STORE_BYPASS X86_BUG(17)/* CPU is affected by speculative store bypass attack */ #define X86_BUG_L1TF X86_BUG(18)/* CPU is affected by L1 Terminal Fault */ +#define X86_BUG_MDS X86_BUG(19)/* CPU is affected by Microarchitectural data sampling */ +#define X86_BUG_MSBDS_ONLY X86_BUG(20)/* CPU is only affected by the MSDBS variant of BUG_MDS */ #endif /* _ASM_X86_CPUFEATURES_H */ tools/include/uapi/drm/drm.h +37
-0@@ -649,6 +649,7 @@ struct drm_gem_open {#define DRM_CAP_PAGE_FLIP_TARGET 0x11#define DRM_CAP_CRTC_IN_VBLANK_EVENT 0x12#define DRM_CAP_SYNCOBJ 0x13 +#define DRM_CAP_SYNCOBJ_TIMELINE 0x14/** DRM_IOCTL_GET_CAP ioctl argument type */ struct drm_get_cap {@@ -735,8 +736,18 @@ struct drm_syncobj_handle {__u32 pad; }; +struct drm_syncobj_transfer { +__u32 src_handle; +__u32 dst_handle; +__u64 src_point; +__u64 dst_point; +__u32 flags; +__u32 pad; +}; +#define DRM_SYNCOBJ_WAIT_FLAGS_WAIT_ALL (1 << 0)#define DRM_SYNCOBJ_WAIT_FLAGS_WAIT_FOR_SUBMIT (1 << 1) +#define DRM_SYNCOBJ_WAIT_FLAGS_WAIT_AVAILABLE (1 << 2)/* wait for time point to become available */ struct drm_syncobj_wait {__u64 handles;/* absolute timeout */ @@ -747,12 +758,33 @@ struct drm_syncobj_wait {__u32 pad; }; +struct drm_syncobj_timeline_wait { +__u64 handles; +/* wait on specific timeline point for every handles*/ +__u64 points; +/* absolute timeout */ +__s64 timeout_nsec; +__u32 count_handles; +__u32 flags; +__u32 first_signaled;/* only valid when not waiting all */ +__u32 pad; +}; + +struct drm_syncobj_array {__u64 handles;__u32 count_handles;__u32 pad; }; +struct drm_syncobj_timeline_array { +__u64 handles; +__u64 points; +__u32 count_handles; +__u32 pad; +}; + +/* Query current scanout sequence number */ struct drm_crtc_get_sequence {__u32 crtc_id;/* requested crtc_id */ @@ -909,6 +941,11 @@ extern "C" {#define DRM_IOCTL_MODE_GET_LEASE DRM_IOWR(0xC8, struct drm_mode_get_lease)#define DRM_IOCTL_MODE_REVOKE_LEASE DRM_IOWR(0xC9, struct drm_mode_revoke_lease) +#define DRM_IOCTL_SYNCOBJ_TIMELINE_WAIT DRM_IOWR(0xCA, struct drm_syncobj_timeline_wait) +#define DRM_IOCTL_SYNCOBJ_QUERY DRM_IOWR(0xCB, struct drm_syncobj_timeline_array) +#define DRM_IOCTL_SYNCOBJ_TRANSFER DRM_IOWR(0xCC, struct drm_syncobj_transfer) +#define DRM_IOCTL_SYNCOBJ_TIMELINE_SIGNAL DRM_IOWR(0xCD, struct drm_syncobj_timeline_array) +/** * Device specific ioctls should only be in their respective headers * The device specific ioctl range is from 0x40 to 0x9f. tools/include/uapi/drm/i915_drm.h +181
-73@@ -62,6 +62,28 @@ extern "C" {#define I915_ERROR_UEVENT "ERROR"#define I915_RESET_UEVENT "RESET" +/* +* i915_user_extension: Base class for defining a chain of extensions +* +* Many interfaces need to grow over time. In most cases we can simply +* extend the struct and have userspace pass in more data. Another option, +* as demonstrated by Vulkan's approach to providing extensions for forward +* and backward compatibility, is to use a list of optional structs to +* provide those extra details. +* +* The key advantage to using an extension chain is that it allows us to +* redefine the interface more easily than an ever growing struct of +* increasing complexity, and for large parts of that interface to be +* entirely optional. The downside is more pointer chasing; chasing across +* the __user boundary with pointers encapsulated inside u64. +*/ +struct i915_user_extension { +__u64 next_extension; +__u32 name; +__u32 flags;/* All undefined bits must be zero. */ +__u32 rsvd[4 ];/* Reserved for future use; must be zero. */ +}; +/* * MOCS indexes used for GPU surfaces, defining the cacheability of the * surface data and the coherency for this data wrt. CPU vs. GPU accesses. @@ -99,9 +121,23 @@ enum drm_i915_gem_engine_class {I915_ENGINE_CLASS_VIDEO =2 ,I915_ENGINE_CLASS_VIDEO_ENHANCE =3 , +/* should be kept compact */ + I915_ENGINE_CLASS_INVALID = -1 }; +/* +* There may be more than one engine fulfilling any role within the system. +* Each engine of a class is given a unique instance number and therefore +* any engine can be specified by its class:instance tuplet. APIs that allow +* access to any engine in the system will use struct i915_engine_class_instance +* for this identification. +*/ +struct i915_engine_class_instance { +__u16 engine_class;/* see enum drm_i915_gem_engine_class */ +__u16 engine_instance; +}; +/** * DOC: perf_events exposed by i915 through /sys/bus/event_sources/drivers/i915 * @@ -319,6 +355,7 @@ typedef struct _drm_i915_sarea {#define DRM_I915_PERF_ADD_CONFIG 0x37#define DRM_I915_PERF_REMOVE_CONFIG 0x38#define DRM_I915_QUERY 0x39 +/* Must be kept compact -- no holes */ #define DRM_IOCTL_I915_INIT DRM_IOW( DRM_COMMAND_BASE + DRM_I915_INIT, drm_i915_init_t)#define DRM_IOCTL_I915_FLUSH DRM_IO ( DRM_COMMAND_BASE + DRM_I915_FLUSH)@@ -367,6 +404,7 @@ typedef struct _drm_i915_sarea {#define DRM_IOCTL_I915_GET_SPRITE_COLORKEY DRM_IOWR(DRM_COMMAND_BASE + DRM_I915_GET_SPRITE_COLORKEY, struct drm_intel_sprite_colorkey)#define DRM_IOCTL_I915_GEM_WAIT DRM_IOWR(DRM_COMMAND_BASE + DRM_I915_GEM_WAIT, struct drm_i915_gem_wait)#define DRM_IOCTL_I915_GEM_CONTEXT_CREATE DRM_IOWR (DRM_COMMAND_BASE + DRM_I915_GEM_CONTEXT_CREATE, struct drm_i915_gem_context_create) +#define DRM_IOCTL_I915_GEM_CONTEXT_CREATE_EXT DRM_IOWR (DRM_COMMAND_BASE + DRM_I915_GEM_CONTEXT_CREATE, struct drm_i915_gem_context_create_ext)#define DRM_IOCTL_I915_GEM_CONTEXT_DESTROY DRM_IOW (DRM_COMMAND_BASE + DRM_I915_GEM_CONTEXT_DESTROY, struct drm_i915_gem_context_destroy)#define DRM_IOCTL_I915_REG_READ DRM_IOWR (DRM_COMMAND_BASE + DRM_I915_REG_READ, struct drm_i915_reg_read)#define DRM_IOCTL_I915_GET_RESET_STATS DRM_IOWR (DRM_COMMAND_BASE + DRM_I915_GET_RESET_STATS, struct drm_i915_reset_stats)@@ -476,6 +514,7 @@ typedef struct drm_i915_irq_wait {#define I915_SCHEDULER_CAP_ENABLED (1ul << 0)#define I915_SCHEDULER_CAP_PRIORITY (1ul << 1)#define I915_SCHEDULER_CAP_PREEMPTION (1ul << 2) +#define I915_SCHEDULER_CAP_SEMAPHORES (1ul << 3)#define I915_PARAM_HUC_STATUS 42@@ -559,6 +598,8 @@ typedef struct drm_i915_irq_wait {*/ #define I915_PARAM_MMAP_GTT_COHERENT 52 +/* Must be kept compact -- no holes and well documented */ + typedef struct drm_i915_getparam { __s32 param; /*@@ -574,6 +615,7 @@ typedef struct drm_i915_getparam {#define I915_SETPARAM_TEX_LRU_LOG_GRANULARITY 2#define I915_SETPARAM_ALLOW_BATCHBUFFER 3#define I915_SETPARAM_NUM_USED_FENCES 4 +/* Must be kept compact -- no holes */ typedef struct drm_i915_setparam { int param;@@ -972,7 +1014,7 @@ struct drm_i915_gem_execbuffer2 {* struct drm_i915_gem_exec_fence *fences. */ __u64 cliprects_ptr; -#define I915_EXEC_RING_MASK (7<<0) +#define I915_EXEC_RING_MASK (0x3f)#define I915_EXEC_DEFAULT (0<<0)#define I915_EXEC_RENDER (1<<0)#define I915_EXEC_BSD (2<<0)@@ -1120,32 +1162,34 @@ struct drm_i915_gem_busy {* as busy may become idle before the ioctl is completed. * * Furthermore, if the object is busy, which engine is busy is only -* provided as a guide. There are race conditions which prevent the -* report of which engines are busy from being always accurate. -* However, the converse is not true. If the object is idle, the -* result of the ioctl, that all engines are idle, is accurate. +* provided as a guide and only indirectly by reporting its class +* (there may be more than one engine in each class). There are race +* conditions which prevent the report of which engines are busy from +* being always accurate. However, the converse is not true. If the +* object is idle, the result of the ioctl, that all engines are idle, +* is accurate. * * The returned dword is split into two fields to indicate both -* the engines on which the object is being read, and the -* engine on which it is currently being written (if any). +* the engine classess on which the object is being read, and the +* engine class on which it is currently being written (if any). * * The low word (bits 0:15) indicate if the object is being written * to by any engine (there can only be one, as the GEM implicit * synchronisation rules force writes to be serialised). Only the -* engine for the last write is reported. +* engine class (offset by 1, I915_ENGINE_CLASS_RENDER is reported as +* 1 not 0 etc) for the last write is reported. * -* The high word (bits 16:31) are a bitmask of which engines are -* currently reading from the object. Multiple engines may be +* The high word (bits 16:31) are a bitmask of which engines classes +* are currently reading from the object. Multiple engines may be * reading from the object simultaneously. * -* The value of each engine is the same as specified in the -* EXECBUFFER2 ioctl, i.e. I915_EXEC_RENDER, I915_EXEC_BSD etc. -* Note I915_EXEC_DEFAULT is a symbolic value and is mapped to -* the I915_EXEC_RENDER engine for execution, and so it is never +* The value of each engine class is the same as specified in the +* I915_CONTEXT_SET_ENGINES parameter and via perf, i.e. +* I915_ENGINE_CLASS_RENDER, I915_ENGINE_CLASS_COPY, etc. * reported as active itself. Some hardware may have parallel * execution engines, e.g. multiple media engines, which are -* mapped to the same identifier in the EXECBUFFER2 ioctl and -* so are not separately reported for busyness. +* mapped to the same class identifier and so are not separately +* reported for busyness. * * Caveat emptor: * Only the boolean result of this query is reliable; that is whether @@ -1412,65 +1456,17 @@ struct drm_i915_gem_wait { };struct drm_i915_gem_context_create { -/* output: id of new context*/ -__u32 ctx_id; -__u32 pad; -}; - -struct drm_i915_gem_context_destroy { -__u32 ctx_id; +__u32 ctx_id;/* output: id of new context*/ __u32 pad; }; -struct drm_i915_reg_read { -/* -* Register offset. -* For 64bit wide registers where the upper 32bits don't immediately -* follow the lower 32bits, the offset of the lower 32bits must -* be specified -*/ -__u64 offset; -#define I915_REG_READ_8B_WA (1ul << 0) - -__u64 val;/* Return value */ -}; -/* Known registers: -* -* Render engine timestamp - 0x2358 + 64bit - gen7+ -* - Note this register returns an invalid value if using the default -* single instruction 8byte read, in order to workaround that pass -* flag I915_REG_READ_8B_WA in offset field. -* -*/ - -struct drm_i915_reset_stats { -__u32 ctx_id; -__u32 flags; - -/* All resets since boot/module reload, for all contexts */ -__u32 reset_count; - -/* Number of batches lost when active in GPU, for this context */ -__u32 batch_active; - -/* Number of batches lost pending for execution, for this context */ -__u32 batch_pending; - -__u32 pad; -}; - -struct drm_i915_gem_userptr { -__u64 user_ptr; -__u64 user_size; +struct drm_i915_gem_context_create_ext { +__u32 ctx_id;/* output: id of new context*/ __u32 flags; -#define I915_USERPTR_READ_ONLY 0x1 -#define I915_USERPTR_UNSYNCHRONIZED 0x80000000 -/** -* Returned handle for the object. -* -* Object handles are nonzero. -*/ -__u32 handle; +#define I915_CONTEXT_CREATE_FLAGS_USE_EXTENSIONS (1u << 0) +#define I915_CONTEXT_CREATE_FLAGS_UNKNOWN \ + (-(I915_CONTEXT_CREATE_FLAGS_USE_EXTENSIONS << 1)) +__u64 extensions; };struct drm_i915_gem_context_param {@@ -1491,6 +1487,28 @@ struct drm_i915_gem_context_param {* drm_i915_gem_context_param_sseu. */ #define I915_CONTEXT_PARAM_SSEU 0x7 + +/* +* Not all clients may want to attempt automatic recover of a context after +* a hang (for example, some clients may only submit very small incremental +* batches relying on known logical state of previous batches which will never +* recover correctly and each attempt will hang), and so would prefer that +* the context is forever banned instead. +* +* If set to false (0), after a reset, subsequent (and in flight) rendering +* from this context is discarded, and the client will need to create a new +* context to use instead. +* +* If set to true (1), the kernel will automatically attempt to recover the +* context by skipping the hanging batch and executing the next batch starting +* from the default context state (discarding the incomplete logical context +* state lost due to the reset). +* +* On creation, all new contexts are marked as recoverable. +*/ +#define I915_CONTEXT_PARAM_RECOVERABLE 0x8 +/* Must be kept compact -- no holes and well documented */ +__u64 value; };@@ -1519,8 +1537,7 @@ struct drm_i915_gem_context_param_sseu {/* * Engine class & instance to be configured or queried. */ -__u16 engine_class; -__u16 engine_instance; +struct i915_engine_class_instance engine;/* * Unused for now. Must be cleared to zero. @@ -1553,6 +1570,96 @@ struct drm_i915_gem_context_param_sseu {__u32 rsvd; }; +struct drm_i915_gem_context_create_ext_setparam { +#define I915_CONTEXT_CREATE_EXT_SETPARAM 0 +struct i915_user_extension base; +struct drm_i915_gem_context_param param; +}; + +struct drm_i915_gem_context_destroy { +__u32 ctx_id; +__u32 pad; +}; + +/* +* DRM_I915_GEM_VM_CREATE - +* +* Create a new virtual memory address space (ppGTT) for use within a context +* on the same file. Extensions can be provided to configure exactly how the +* address space is setup upon creation. +* +* The id of new VM (bound to the fd) for use with I915_CONTEXT_PARAM_VM is +* returned in the outparam @id. +* +* No flags are defined, with all bits reserved and must be zero. +* +* An extension chain maybe provided, starting with @extensions, and terminated +* by the @next_extension being 0. Currently, no extensions are defined. +* +* DRM_I915_GEM_VM_DESTROY - +* +* Destroys a previously created VM id, specified in @id. +* +* No extensions or flags are allowed currently, and so must be zero. +*/ +struct drm_i915_gem_vm_control { +__u64 extensions; +__u32 flags; +__u32 vm_id; +}; + +struct drm_i915_reg_read { +/* +* Register offset. +* For 64bit wide registers where the upper 32bits don't immediately +* follow the lower 32bits, the offset of the lower 32bits must +* be specified +*/ +__u64 offset; +#define I915_REG_READ_8B_WA (1ul << 0) + +__u64 val;/* Return value */ +}; + +/* Known registers: +* +* Render engine timestamp - 0x2358 + 64bit - gen7+ +* - Note this register returns an invalid value if using the default +* single instruction 8byte read, in order to workaround that pass +* flag I915_REG_READ_8B_WA in offset field. +* +*/ + +struct drm_i915_reset_stats { +__u32 ctx_id; +__u32 flags; + +/* All resets since boot/module reload, for all contexts */ +__u32 reset_count; + +/* Number of batches lost when active in GPU, for this context */ +__u32 batch_active; + +/* Number of batches lost pending for execution, for this context */ +__u32 batch_pending; + +__u32 pad; +}; + +struct drm_i915_gem_userptr { +__u64 user_ptr; +__u64 user_size; +__u32 flags; +#define I915_USERPTR_READ_ONLY 0x1 +#define I915_USERPTR_UNSYNCHRONIZED 0x80000000 +/** +* Returned handle for the object. +* +* Object handles are nonzero. +*/ +__u32 handle; +}; +enum drm_i915_oa_format {I915_OA_FORMAT_A13 =1 ,/* HSW only */ I915_OA_FORMAT_A29 ,/* HSW only */ @@ -1714,6 +1821,7 @@ struct drm_i915_perf_oa_config {struct drm_i915_query_item {__u64 query_id;#define DRM_I915_QUERY_TOPOLOGY_INFO 1 +/* Must be kept compact -- no holes and well documented */ /* * When set to zero by userspace, this is filled with the size of the tools/include/uapi/linux/fcntl.h +2
-0@@ -91,5 +91,7 @@#define AT_STATX_FORCE_SYNC 0x2000/* - Force the attributes to be sync'd with the server */ #define AT_STATX_DONT_SYNC 0x4000/* - Don't sync attributes with the server */ +#define AT_RECURSIVE 0x8000/* Apply to the entire subtree */ +#endif /* _UAPI_LINUX_FCNTL_H */ tools/include/uapi/linux/fs.h +3
-0@@ -320,6 +320,9 @@ struct fscrypt_key {#define SYNC_FILE_RANGE_WAIT_BEFORE 1#define SYNC_FILE_RANGE_WRITE 2#define SYNC_FILE_RANGE_WAIT_AFTER 4 +#define SYNC_FILE_RANGE_WRITE_AND_WAIT (SYNC_FILE_RANGE_WRITE | \ + SYNC_FILE_RANGE_WAIT_BEFORE | \ + SYNC_FILE_RANGE_WAIT_AFTER)/* * Flags for preadv2/pwritev2: tools/include/uapi/linux/mount.h +62
-0@@ -55,4 +55,66 @@#define MS_MGC_VAL 0xC0ED0000#define MS_MGC_MSK 0xffff0000 +/* +* open_tree() flags. +*/ +#define OPEN_TREE_CLONE 1/* Clone the target tree and attach the clone */ +#define OPEN_TREE_CLOEXEC O_CLOEXEC/* Close the file on execve() */ + +/* +* move_mount() flags. +*/ +#define MOVE_MOUNT_F_SYMLINKS 0x00000001/* Follow symlinks on from path */ +#define MOVE_MOUNT_F_AUTOMOUNTS 0x00000002/* Follow automounts on from path */ +#define MOVE_MOUNT_F_EMPTY_PATH 0x00000004/* Empty from path permitted */ +#define MOVE_MOUNT_T_SYMLINKS 0x00000010/* Follow symlinks on to path */ +#define MOVE_MOUNT_T_AUTOMOUNTS 0x00000020/* Follow automounts on to path */ +#define MOVE_MOUNT_T_EMPTY_PATH 0x00000040/* Empty to path permitted */ +#define MOVE_MOUNT__MASK 0x00000077 + +/* +* fsopen() flags. +*/ +#define FSOPEN_CLOEXEC 0x00000001 + +/* +* fspick() flags. +*/ +#define FSPICK_CLOEXEC 0x00000001 +#define FSPICK_SYMLINK_NOFOLLOW 0x00000002 +#define FSPICK_NO_AUTOMOUNT 0x00000004 +#define FSPICK_EMPTY_PATH 0x00000008 + +/* +* The type of fsconfig() call made. +*/ +enum fsconfig_command { +FSCONFIG_SET_FLAG =0 ,/* Set parameter, supplying no value */ +FSCONFIG_SET_STRING =1 ,/* Set parameter, supplying a string value */ +FSCONFIG_SET_BINARY =2 ,/* Set parameter, supplying a binary blob value */ +FSCONFIG_SET_PATH =3 ,/* Set parameter, supplying an object by path */ +FSCONFIG_SET_PATH_EMPTY =4 ,/* Set parameter, supplying an object by (empty) path */ +FSCONFIG_SET_FD =5 ,/* Set parameter, supplying an object by fd */ +FSCONFIG_CMD_CREATE =6 ,/* Invoke superblock creation */ +FSCONFIG_CMD_RECONFIGURE =7 ,/* Invoke superblock reconfiguration */ +}; + +/* +* fsmount() flags. +*/ +#define FSMOUNT_CLOEXEC 0x00000001 + +/* +* Mount attributes. +*/ +#define MOUNT_ATTR_RDONLY 0x00000001/* Mount read-only */ +#define MOUNT_ATTR_NOSUID 0x00000002/* Ignore suid and sgid bits */ +#define MOUNT_ATTR_NODEV 0x00000004/* Disallow access to device special files */ +#define MOUNT_ATTR_NOEXEC 0x00000008/* Disallow program execution */ +#define MOUNT_ATTR__ATIME 0x00000070/* Setting on how atime should be updated */ +#define MOUNT_ATTR_RELATIME 0x00000000/* - Update atime relative to mtime/ctime. */ +#define MOUNT_ATTR_NOATIME 0x00000010/* - Do not update access times. */ +#define MOUNT_ATTR_STRICTATIME 0x00000020/* - Always perform atime updates */ +#define MOUNT_ATTR_NODIRATIME 0x00000080/* Do not update directory access times */ +#endif /* _UAPI_LINUX_MOUNT_H */ tools/include/uapi/linux/sched.h +1
-0@@ -10,6 +10,7 @@#define CLONE_FS 0x00000200/* set if fs info shared between processes */ #define CLONE_FILES 0x00000400/* set if open files shared between processes */ #define CLONE_SIGHAND 0x00000800/* set if signal handlers and blocked signals shared */ +#define CLONE_PIDFD 0x00001000/* set if a pidfd should be placed in parent */ #define CLONE_PTRACE 0x00002000/* set if we want to let tracing continue on the child too */ #define CLONE_VFORK 0x00004000/* set if the parent wants the child to wake it up on mm_release */ #define CLONE_PARENT 0x00008000/* set if we want to have the same parent as the cloner */ tools/perf/arch/x86/entry/syscalls/syscall_64.tbl +6
-0@@ -343,6 +343,12 @@ 332 common statx __x64_sys_statx 333 common io_pgetevents __x64_sys_io_pgetevents 334 common rseq __x64_sys_rseq +335 common open_tree __x64_sys_open_tree +336 common move_mount __x64_sys_move_mount +337 common fsopen __x64_sys_fsopen +338 common fsconfig __x64_sys_fsconfig +339 common fsmount __x64_sys_fsmount +340 common fspick __x64_sys_fspick # don't use numbers 387 through 423, add new calls after the last # 'common' entry 424 common pidfd_send_signal __x64_sys_pidfd_send_signal
Signature · 1 line
- Arnaldo
Re: [PATCH] tools/headers: Synchronize kernel ABI headers
Quoted 51 lines · Arnaldo Carvalho de Melo
Em Sat, May 18, 2019 at 10:42:23AM +0200, Ingo Molnar escreveu:
Quoted 11 lines
Pick up the latest v5.2-to-be kernel ABI headers and synchronize them with tooling:
- arch/x86/entry/syscalls/syscall_64.tbl => tools/perf/arch/x86/entry/syscalls/syscall_64.tbl # new syscalls - arch/x86/include/asm/cpufeatures.h => tools/arch/x86/include/asm/cpufeatures.h # new CPUID flags - include/uapi/drm/drm.h => tools/include/uapi/drm/drm.h # new 'syncobj' DRM ABI - include/uapi/drm/i915_drm.h => tools/include/uapi/drm/i915_drm.h # new extensible DRM ABI - include/uapi/linux/fcntl.h => tools/include/uapi/linux/fcntl.h # new AT_RECURSIVE - include/uapi/linux/fs.h => tools/include/uapi/linux/fs.h # new SYNC_FILE_RANGE_WRITE_AND_WAIT - include/uapi/linux/mount.h => tools/include/uapi/linux/mount.h # new VFS system calls: fspick, fsmount, fsconfig, fsopen, move_mount, open_tree - include/uapi/linux/sched.h => tools/include/uapi/linux/sched.h # new CLONE_PIDFDQuoted 1 line
All of these are new ABI additions with no impact on existing tooling,
There is some impact in a number of them, for instance:
[acme@quaco perf]$ diff -u
tools/perf/arch/x86/entry/syscalls/syscall_64.tblarch/x86/entry/syscalls/syscall_64.tblx86/entry/syscalls/syscall_64.tbl renamed from perf/arch/x86/entry/syscalls/syscall_64.tbl +6
-0@@ -343,6 +343,12 @@ 332 common statx __x64_sys_statx 333 common io_pgetevents __x64_sys_io_pgetevents 334 common rseq __x64_sys_rseq +335 common open_tree __x64_sys_open_tree +336 common move_mount __x64_sys_move_mount +337 common fsopen __x64_sys_fsopen +338 common fsconfig __x64_sys_fsconfig +339 common fsmount __x64_sys_fsmount +340 common fspick __x64_sys_fspick # don't use numbers 387 through 423, add new calls after the last # 'common' entry 424 common pidfd_send_signal __x64_sys_pidfd_send_signal[acme@quaco perf]$
These will enable 'perf trace' to know about these syscalls, i.e. we will be able to say:
perf trace -e fs*
And have it trace just those fsopen, fsconfig, fsmount and fspick syscalls (or any other that starts with fs, that is). Looking at this makes one think if we should have a new syscall group like we have here:
[acme@quaco perf]$ ls -la
tools/perf/trace/strace/groups/total 16
drwxrwxr-x. 2 acme acme 4096 May 13 14:46 . drwxrwxr-x. 3 acme acme 4096 Apr 10 10:20 ..x86/entry/syscalls/syscall_64.tbl renamed from perf/arch/x86/entry/syscalls/syscall_64.tbl +0
-2-rw-rw-r--. 1 acme acme 136 May 13 14:46 file -rw-rw-r--. 1 acme acme 584 May 13 14:46 string[acme@quaco perf]$
Quoted 2 lines
One for file oriented syscalls, the other for syscalls that operate on strings (pathnames, etc).
Good point, I missed that!
Quoted 3 lines
So I'll check that
tools/perf/trace/beauty/drm_ioctl.shis sane wrt any new additions to that header, ditto for the other headers, I'll go thru them soon.
Thanks!
Ingo