linux/tools/perf
Arnaldo Carvalho de Melo 59f3bd7802 perf augmented_raw_syscalls: Use a PERCPU_ARRAY map to copy more string bytes
The previous method, copying to the BPF stack limited us in how many
bytes we could copy from strings, use a PERCPU_ARRAY map like devised by
the sysdig guys[1] to copy more bytes:

Before:

  # trace --no-inherit -e openat touch `python -c "print "$s" 'a' * 2000"`
  touch: cannot touch 'aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa': File name too long
  openat(AT_FDCWD, "/etc/ld.so.cache", O_RDONLY|O_CLOEXEC) = 3
  openat(AT_FDCWD, "/lib64/libc.so.6", O_RDONLY|O_CLOEXEC) = 3
  openat(AT_FDCWD, "/usr/lib/locale/locale-archive", O_RDONLY|O_CLOEXEC) = 3
  openat(AT_FDCWD, "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa", O_CREAT|O_NOCTTY|O_NONBLOCK|O_WRONLY, S_IRUGO|S_IWUGO) = -1 ENAMETOOLONG (File name too long)
  openat(AT_FDCWD, "/usr/share/locale/locale.alias", O_RDONLY|O_CLOEXEC) = 3
  openat(AT_FDCWD, "/usr/share/locale/en_US.UTF-8/LC_MESSAGES/coreutils.mo", O_RDONLY) = -1 ENOENT (No such file or directory)
  openat(AT_FDCWD, "/usr/share/locale/en_US.utf8/LC_MESSAGES/coreutils.mo", O_RDONLY) = -1 ENOENT (No such file or directory)
  <SNIP some openat calls>
  #

After:

  [root@quaco acme]# trace --no-inherit -e openat touch `python -c "print "$s" 'a' * 2000"`
  <STRIP what is the same as in the 'before' part>
  openat(AT_FDCWD, "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa", O_CREAT|O_NOCTTY|O_NONBLOC) = -1 ENAMETOOLONG (File name too long)
  <STRIP what is the same as in the 'before' part>

If we leave something like 'perf trace -e string' to trace all syscalls
with a string, and then do some 'perf top', to get some annotation for
the augmented_raw_syscalls.o BPF program we get:

       │     → callq  *ffffffffc45576d1                                                                                                          ▒
       │                augmented_args->filename.size = probe_read_str(&augmented_args->filename.value,                                          ▒
  0.05 │       mov    %eax,0x40(%r13)

Looking with pahole, expanding types, asking for hex offsets and sizes,
and use of BTF type information to see what is at that 0x40 offset from
%r13:

  # pahole -F btf -C augmented_args_filename --expand_types --hex /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.o
  struct augmented_args_filename {
	struct syscall_enter_args {
		long long unsigned int common_tp_fields;                                 /*     0   0x8 */
		long int           syscall_nr;                                           /*   0x8   0x8 */
		long unsigned int  args[6];                                              /*  0x10  0x30 */
	} args; /*     0  0x40 */
	/* --- cacheline 1 boundary (64 bytes) --- */
	struct augmented_filename {
		unsigned int       size;                                                 /*  0x40   0x4 */
		int                reserved;                                             /*  0x44   0x4 */
		char               value[4096];                                          /*  0x48 0x1000 */
	} filename; /*  0x40 0x1008 */

	/* size: 4168, cachelines: 66, members: 2 */
	/* last cacheline: 8 bytes */
  };
  #

Then looking if PATH_MAX leaves some signature in the tests:

       │                if (augmented_args->filename.size < sizeof(augmented_args->filename.value)) {                                            ▒
       │       cmp    $0xfff,%rdi

0xfff == 4095
sizeof(augmented_args->filename.value) == PATH_MAX == 4096

[1] https://sysdig.com/blog/the-art-of-writing-ebpf-programs-a-primer/

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Andrii Nakryiko <andriin@fb.com>
Cc: Daniel Borkmann <borkmann@iogearbox.net>
Cc: Gianluca Borello <g.borello@gmail.com>
Cc: Jesper Dangaard Brouer <brouer@redhat.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
cc: Martin Lau <kafai@fb.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Cc: Yonghong Song <yhs@fb.com>
Link: https://lkml.kernel.org/n/tip-76gce2d2ghzq537ubwhjkone@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-04-01 14:49:24 -03:00
..
arch tools headers: Update x86's syscall_64.tbl and uapi/asm-generic/unistd 2019-03-28 14:41:11 -03:00
bench perf/core improvements and fixes: 2019-03-22 22:51:21 +01:00
Documentation perf config: Fix an error in the config template documentation 2019-03-19 16:52:04 -03:00
examples/bpf perf augmented_raw_syscalls: Use a PERCPU_ARRAY map to copy more string bytes 2019-04-01 14:49:24 -03:00
include/bpf perf bpf: Automatically add BTF ELF markers 2019-03-06 09:45:37 -03:00
jvmti perf jvmti: Separate jvmti cmlr check 2018-11-21 22:39:58 -03:00
pmu-events perf vendor events: Remove P8 HW events which are not supported 2019-03-19 16:52:03 -03:00
python perf python: Make twatch.py work with both python2 and python3 2018-02-19 12:28:08 -03:00
scripts perf scripts python: exported-sql-viewer.py: Fix python3 support 2019-03-28 15:53:16 -03:00
tests perf tools: Save bpf_prog_info and BTF of new BPF programs 2019-03-21 11:27:04 -03:00
trace perf trace: Add 'string' event alias to select syscalls with string args 2019-04-01 14:49:24 -03:00
ui perf tools report: Add custom scripts to script menu 2019-03-11 16:33:20 -03:00
util perf pmu: Fix parser error for uncore event alias 2019-03-28 15:53:27 -03:00
.gitignore perf tools: Add trace/beauty/generated/ into .gitignore 2018-02-05 13:58:02 -03:00
Build perf tools: Rename build libperf to perf 2019-02-14 15:18:08 -03:00
builtin-annotate.c perf data: Add global path holder 2019-02-22 16:52:07 -03:00
builtin-bench.c perf bench: Add epoll_ctl(2) benchmark 2018-11-21 22:39:55 -03:00
builtin-buildid-cache.c perf data: Add global path holder 2019-02-22 16:52:07 -03:00
builtin-buildid-list.c perf data: Add global path holder 2019-02-22 16:52:07 -03:00
builtin-c2c.c perf c2c: Fix c2c report for empty numa node 2019-03-06 18:15:24 -03:00
builtin-config.c perf config: Show the configuration when no arguments are provided 2018-12-18 12:24:00 -03:00
builtin-data.c License cleanup: add SPDX GPL-2.0 license identifier to files with no license 2017-11-02 11:10:55 +01:00
builtin-diff.c perf diff: Support --pid/--tid filter options 2019-03-06 18:06:16 -03:00
builtin-evlist.c perf data: Add global path holder 2019-02-22 16:52:07 -03:00
builtin-ftrace.c perf ftrace: Append an EOL when write tracing files 2018-02-19 09:49:12 -03:00
builtin-help.c perf help: Remove needless use of strncpy() 2018-12-17 14:59:18 -03:00
builtin-inject.c perf data: Add global path holder 2019-02-22 16:52:07 -03:00
builtin-kallsyms.c pref tools: Add missing map.h includes 2019-02-06 10:00:38 -03:00
builtin-kmem.c perf data: Add global path holder 2019-02-22 16:52:07 -03:00
builtin-kvm.c perf data: Add global path holder 2019-02-22 16:52:07 -03:00
builtin-list.c perf list: Filter metrics too 2019-03-19 11:56:19 -03:00
builtin-lock.c perf data: Add global path holder 2019-02-22 16:52:07 -03:00
builtin-mem.c perf data: Add global path holder 2019-02-22 16:52:07 -03:00
builtin-probe.c perf namespaces: Remove namespaces.h from .h headers 2019-01-25 15:12:09 +01:00
builtin-record.c perf tools: Save bpf_prog_info and BTF of new BPF programs 2019-03-21 11:27:04 -03:00
builtin-report.c perf report: Show all sort keys in help output 2019-03-19 16:15:42 -03:00
builtin-sched.c perf data: Add global path holder 2019-02-22 16:52:07 -03:00
builtin-script.c perf script: Support relative time 2019-03-19 16:52:03 -03:00
builtin-stat.c perf stat: Fix --no-scale 2019-03-19 16:52:03 -03:00
builtin-timechart.c perf data: Add global path holder 2019-02-22 16:52:07 -03:00
builtin-top.c perf tools: Save bpf_prog_info and BTF of new BPF programs 2019-03-21 11:27:04 -03:00
builtin-trace.c perf data: Add global path holder 2019-02-22 16:52:07 -03:00
builtin-version.c perf version: Print status for syscall_table 2018-04-12 10:33:34 -03:00
builtin.h perf script: Add array bound checking to list_scripts 2019-03-11 16:33:19 -03:00
check-headers.sh tools headers uapi: Sync asm-generic/mman-common.h and linux/mman.h 2019-03-28 14:31:56 -03:00
command-list.txt perf help: Add missing subcommand version 2018-09-19 14:53:36 -03:00
CREDITS
design.txt perf/doc: Update design.txt for exclude_{host|guest} flags 2019-01-21 11:01:18 +01:00
Makefile perf tools: Disable parallelism for 'make clean' 2018-08-20 08:54:58 -03:00
Makefile.config perf build: Check what binutils's 'disassembler()' signature to use 2019-03-20 16:42:10 -03:00
Makefile.perf tools headers uapi: Sync asm-generic/mman-common.h and linux/mman.h 2019-03-28 14:31:56 -03:00
MANIFEST perf tools: Get all of tools/{arch,include}/ in the MANIFEST 2017-09-25 10:39:43 -03:00
perf-archive.sh License cleanup: add SPDX GPL-2.0 license identifier to files with no license 2017-11-02 11:10:55 +01:00
perf-completion.sh perf tools: Auto-complete for events with ':' 2017-12-27 12:16:00 -03:00
perf-read-vdso.c perf tools: Make find_vdso_map() more modular 2019-01-08 13:28:13 -03:00
perf-sys.h Drop a bunch of metag references 2018-02-23 14:29:59 +00:00
perf-with-kcore.sh
perf.c perf bpf: Save bpf_prog_info in a rbtree in perf_env 2019-03-19 16:52:06 -03:00
perf.h perf record: Replace option --bpf-event with --no-bpf-event 2019-03-19 16:52:06 -03:00