Starting with the renameat and renameat2 syscall, that both receive as
second and fourth parameters a pathname:
# perf trace -e rename* mv one ANOTHER
LLVM: dumping /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.o
mv: cannot stat 'one': No such file or directory
renameat2(AT_FDCWD, "one", AT_FDCWD, "ANOTHER", RENAME_NOREPLACE) = -1 ENOENT (No such file or directory)
#
Since the per CPU scratch buffer map has space for two maximum sized
pathnames, the verifier is satisfied that there will be no overrun.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-x2uboyg5kx2wqeru288209b6@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
I.e. for a syscall that has its second argument being a string, its
difficult these days to find 'open' being used in the wild :-)
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-yf3kbzirqrukd3fb2sp5qx4p@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
So, we use a PERF_COUNT_SW_BPF_OUTPUT to output the augmented sys_enter
payload, i.e. to output more than just the raw syscall args, and if
something goes wrong when handling an unfiltered syscall, we bail out
and just return 1 in the bpf program associated with
raw_syscalls:sys_enter, meaning, don't filter that tracepoint, in which
case what will appear in the perf ring buffer isn't the BPF_OUTPUT
event, but the original raw_syscalls:sys_enter event with its normal
payload.
Now that we're switching to using a bpf_tail_call +
BPF_MAP_TYPE_PROG_ARRAY we're going to use this in the common case, so a
bug where raw_syscalls:sys_enter wasn't being handled by
trace__sys_enter() surfaced and for that case, instead of using the
strace-like augmenter (trace__sys_enter()), we continued to use the
normal generic tracepoint handler:
(gdb) p evsel
$2 = (struct perf_evsel *) 0xc03e40
(gdb) p evsel->name
$3 = 0xbc56c0 "raw_syscalls:sys_enter"
(gdb) p ((struct perf_evsel *) 0xc03e40)->name
$4 = 0xbc56c0 "raw_syscalls:sys_enter"
(gdb) p ((struct perf_evsel *) 0xc03e40)->handler
$5 = (void *) 0x495eb3 <trace__event_handler>
This resulted in this:
0.027 raw_syscalls:sys_enter:NR 12 (0, 7fcfcac64c9b, 4d, 7fcfcac64c9b, 7fcfcac6ce00, 19)
... [continued]: brk()) = 0x563b88677000
I.e. only the sys_exit tracepoint was being properly handled, but since
the sys_enter went to the generic trace__event_handler() we printed it
using libtraceevent's formatter instead of 'perf trace's strace-like
one.
Fix it by setting trace__sys_enter() as the handler for
raw_syscalls:sys_enter and setup the tp_field tracepoint field
accessors.
Now, to test it we just make raw_syscalls:sys_enter return 1 right after
checking if the pid is filtered, making it not use
bpf_perf_output_event() but rather ask for the tracepoint not to be
filtered and the result is the expected one:
brk(NULL) = 0x556f42d6e000
I.e. raw_syscalls:sys_enter returns 1, gets handled by
trace__sys_enter() and gets it combined with the raw_syscalls:sys_exit
in a strace-like way.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-0mkocgk31nmy0odknegcby4z@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
I.e. look for "syscalls_sys_enter" and "syscalls_sys_exit" BPF maps of
type PROG_ARRAY and populate it with the handlers as specified per
syscall, for now only 'open' is wiring it to something, in time all
syscalls that need to copy arguments entering a syscall or returning
from one will set these to the right handlers, reusing when possible
pre-existing ones.
Next step is to use bpf_tail_call() into that.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-t0p4u43i9vbpzs1xtowna3gb@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This is a step in the direction of being able to use a
BPF_MAP_TYPE_PROG_ARRAY to handle syscalls that need to copy pointer
payloads in addition to the raw tracepoint syscall args.
There is a first example in
tools/perf/examples/bpf/augmented_raw_syscalls.c for the 'open' syscall.
Next step is to introduce the prog array map and use this 'open'
augmenter, then use that augmenter in other syscalls that also only copy
the first arg as a string, and then show how to use with a syscall that
reads more than one filename, like 'rename', etc.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-pys4v57x5qqrybb4cery2mc8@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Will be used to assign to syscalls that don't need augmentation, i.e.
those with just integer args.
All syscalls will be in a BPF_MAP_TYPE_PROG_ARRAY, and the
bpf_tail_call() keyed by the syscall id will either find nothing in
place, which means the syscall is being filtered, or a function that
will either add things like filenames to the ring buffer, right after
the raw syscall args, or be this unaugmented handler that will just
return 1, meaning don't filter the original
raw_syscalls:sys_{enter,exit} tracepoint.
For now it is not really being used, this is just leg work to break the
patch into smaller pieces.
It introduces a trace__find_bpf_program_by_title() helper that in turn
uses libbpf's bpf_object__find_program_by_title() on the BPF object with
the __augmented_syscalls__ map. "title" is how libbpf calls the SEC()
argument for functions, i.e. the ELF section that follows a convention
to specify what BPF program (a function with this SEC() marking) should
be connected to which tracepoint, kprobes, etc.
In perf anything that is of the form SEC("sys:event_name") will be
connected to that tracepoint by perf's BPF loader.
In this case its something that will be bpf_tail_call()ed from either
the "raw_syscalls:sys_enter" or "raw_syscall:sys_exit" tracepoints, so
its named "!raw_syscalls:unaugmented" to convey that idea, i.e. its not
going to be directly attached to a tracepoint, thus it starts with a
"!".
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-meucpjx2u0slpkayx56lxqq6@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The ev_qualifier is an array with the syscall ids passed via -e on the
command line, sort it as we'll search it when setting up the
BPF_MAP_TYPE_PROG_ARRAY.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-c8hprylp3ai6e0z9burn2r3s@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
We can conceivably have multiple BPF object files for other purposes, so
better look just on the BPF object containing the __augmented_syscalls__
map for all things augmented_syscalls related.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-3jt8knkuae9lt705r1lns202@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
So that we can use it when looking for other components of that object
file, such as other programs to add to the BPF_MAP_TYPE_PROG_ARRAY and
use with bpf_tail_call().
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-1ibmz7ouv6llqxajy7m8igtd@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Circa v5.2 this started to fail:
# perf trace -e /wb/augmented_raw_syscalls.o
event syntax error: '/wb/augmented_raw_syscalls.o'
\___ Operation not permitted
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf trace [<options>] [<command>]
or: perf trace [<options>] -- <command> [<options>]
or: perf trace record [<options>] [<command>]
or: perf trace record [<options>] -- <command> [<options>]
-e, --event <event> event/syscall selector. use 'perf list' to list available events
#
In verbose mode we some -EPERM when creating a BPF map:
# perf trace -v -e /wb/augmented_raw_syscalls.o
<SNIP>
libbpf: failed to create map (name: '__augmented_syscalls__'): Operation not permitted
libbpf: failed to load object '/wb/augmented_raw_syscalls.o'
bpf: load objects failed: err=-1: (Operation not permitted)
event syntax error: '/wb/augmented_raw_syscalls.o'
\___ Operation not permitted
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf trace [<options>] [<command>]
or: perf trace [<options>] -- <command> [<options>]
or: perf trace record [<options>] [<command>]
or: perf trace record [<options>] -- <command> [<options>]
-e, --event <event> event/syscall selector. use 'perf list' to list available events
#
If we bumped 'ulimit -l 128' to get it from the 64k default to double that, it
worked, so use the recently added rlimit__bump_memlock() helper:
# perf trace -e /wb/augmented_raw_syscalls.o -e open*,*sleep sleep 1
0.000 ( 0.007 ms): sleep/28042 openat(dfd: CWD, filename: "/etc/ld.so.cache", flags: RDONLY|CLOEXEC) = 3
0.022 ( 0.004 ms): sleep/28042 openat(dfd: CWD, filename: "/lib64/libc.so.6", flags: RDONLY|CLOEXEC) = 3
0.201 ( 0.007 ms): sleep/28042 openat(dfd: CWD, filename: "", flags: RDONLY|CLOEXEC) = 3
0.241 (1000.421 ms): sleep/28042 nanosleep(rqtp: 0x7ffd6c3e6ed0) = 0
#
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-j6f2ioa6hj9dinzpjvlhcjoc@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Eroding a bit more the tools/perf/util/util.h hodpodge header.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-natazosyn9rwjka25tvcnyi0@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Based on the following report from Smatch, fix the potential NULL
pointer dereference check.
tools/perf/builtin-trace.c:1044
thread_trace__new() error: we previously assumed 'ttrace' could be
null (see line 1041).
tools/perf/builtin-trace.c
1037 static struct thread_trace *thread_trace__new(void)
1038 {
1039 struct thread_trace *ttrace = zalloc(sizeof(struct thread_trace));
1040
1041 if (ttrace)
1042 ttrace->files.max = -1;
1043
1044 ttrace->syscall_stats = intlist__new(NULL);
^^^^^^^^
1045
1046 return ttrace;
1047 }
Signed-off-by: Leo Yan <leo.yan@linaro.org>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alexey Budankov <alexey.budankov@linux.intel.com>
Cc: Alexios Zavras <alexios.zavras@intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Changbin Du <changbin.du@intel.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Eric Saint-Etienne <eric.saint.etienne@oracle.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: Konstantin Khlebnikov <khlebnikov@yandex-team.ru>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Cc: Song Liu <songliubraving@fb.com>
Cc: Suzuki Poulouse <suzuki.poulose@arm.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Thomas Richter <tmricht@linux.ibm.com>
Cc: linux-arm-kernel@lists.infradead.org
Link: http://lkml.kernel.org/r/20190702103420.27540-6-leo.yan@linaro.org
[ Just made it look like other tools/perf constructors, same end result ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
We got the sane_ctype.h headers from git and kept using it so far, but
since that code originally came from the kernel sources to the git
sources, perhaps its better to just use the one in the kernel, so that
we can leverage tools/perf/check_headers.sh to be notified when our copy
gets out of sync, i.e. when fixes or goodies are added to the code we've
copied.
This will help with things like tools/lib/string.c where we want to have
more things in common with the kernel, such as strim(), skip_spaces(),
etc so as to go on removing the things that we have in tools/perf/util/
and instead using the code in the kernel, indirectly and removing things
like EXPORT_SYMBOL(), etc, getting notified when fixes and improvements
are made to the original code.
Hopefully this also should help with reducing the difference of code
hosted in tools/ to the one in the kernel proper.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-7k9868l713wqtgo01xxygn12@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
We can't just add the consumed bytes to the arg->augmented.args member,
as it is not void *, so it will access (consumed * sizeof(struct augmented_arg))
in the next augmented arg, totally wrong, cast the member to void pointe
before adding the number of bytes consumed, duh.
With this and hardcoding handling the 'renameat' and 'renameat2'
syscalls in the tools/perf/examples/bpf/augmented_raw_syscalls.c eBPF
proggie, we get:
mv/24388 renameat2(AT_FDCWD, "/tmp/build/perf/util/.bpf-event.o.tmp", AT_FDCWD, "/tmp/build/perf/util/.bpf-event.o.cmd", RENAME_NOREPLACE) = 0
mv/24394 renameat2(AT_FDCWD, "/tmp/build/perf/util/.perf-hooks.o.tmp", AT_FDCWD, "/tmp/build/perf/util/.perf-hooks.o.cmd", RENAME_NOREPLACE) = 0
mv/24398 renameat2(AT_FDCWD, "/tmp/build/perf/util/.pmu-bison.o.tmp", AT_FDCWD, "/tmp/build/perf/util/.pmu-bison.o.cmd", RENAME_NOREPLACE) = 0
mv/24401 renameat2(AT_FDCWD, "/tmp/build/perf/util/.expr-bison.o.tmp", AT_FDCWD, "/tmp/build/perf/util/.expr-bison.o.cmd", RENAME_NOREPLACE) = 0
mv/24406 renameat2(AT_FDCWD, "/tmp/build/perf/util/.pmu.o.tmp", AT_FDCWD, "/tmp/build/perf/util/.pmu.o.cmd", RENAME_NOREPLACE) = 0
mv/24407 renameat2(AT_FDCWD, "/tmp/build/perf/util/.pmu-flex.o.tmp", AT_FDCWD, "/tmp/build/perf/util/.pmu-flex.o.cmd", RENAME_NOREPLACE) = 0
mv/24416 renameat2(AT_FDCWD, "/tmp/build/perf/util/.parse-events-flex.o.tmp", AT_FDCWD, "/tmp/build/perf/util/.parse-events-flex.o.cmd", RENAME_NOREPLACE) = 0
I.e. it works with two string args in the same syscall.
Now back to taming the verifier...
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Fixes: 8195168e87 ("perf trace: Consume the augmented_raw_syscalls payload")
Link: https://lkml.kernel.org/n/tip-n1w59lpxks6m1le7fpo6rmyw@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Rename the 'i' variable to 'nr_used' and use set 'nr_allocated' since
the start of this function, leaving the final assignment of the longer
named trace->ev_qualifier_ids.nr state to 'nr_used' at the end of the
function.
No change in behaviour intended.
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Mike Leach <mike.leach@linaro.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Suzuki K Poulose <suzuki.poulose@arm.com>
Link: https://lkml.kernel.org/n/tip-kpgyn8xjdjgt0timrrnniquv@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
We were just skipping the syscalls not available in a particular
architecture without reflecting this in the number of entries in the
ev_qualifier_ids.nr variable, fix it.
This was done with the most minimalistic way, reusing the index variable
'i', a followup patch will further clean this by making 'i' renamed to
'nr_used' and using 'nr_allocated' in a few more places.
Reported-by: Leo Yan <leo.yan@linaro.org>
Tested-by: Leo Yan <leo.yan@linaro.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Mike Leach <mike.leach@linaro.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Suzuki K Poulose <suzuki.poulose@arm.com>
Fixes: 04c41bcb86 ("perf trace: Skip unknown syscalls when expanding strace like syscall groups")
Link: https://lkml.kernel.org/r/20190613181514.GC1402@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
perf record:
Alexey Budankov:
- Allow mixing --user-regs with --call-graph=dwarf, making sure that
the minimal set of registers for DWARF unwinding is present in the
set of user registers requested to be present in each sample, while
warning the user that this may make callchains unreliable if more
that the minimal set of registers is needed to unwind.
yuzhoujian:
- Add support to collect callchains from kernel or user space only,
IOW allow setting the perf_event_attr.exclude_callchain_{kernel,user}
bits from the command line.
perf trace:
Arnaldo Carvalho de Melo:
- Remove x86_64 specific syscall numbers from the augmented_raw_syscalls
BPF in-kernel collector of augmented raw_syscalls:sys_{enter,exit}
payloads, use instead the syscall numbers obtainer either by the
arch specific syscalltbl generators or from audit-libs.
- Allow 'perf trace' to ask for the number of bytes to collect for
string arguments, for now ask for PATH_MAX, i.e. the whole
pathnames, which ends up being just a way to speficy which syscall
args are pathnames and thus should be read using bpf_probe_read_str().
- Skip unknown syscalls when expanding strace like syscall groups.
This helps using the 'string' group of syscalls to work in arm64,
where some of the syscalls present in x86_64 that deal with
strings, for instance 'access', are deprecated and this should not
be asked for tracing.
Leo Yan:
- Exit when failing to build eBPF program.
perf config:
Arnaldo Carvalho de Melo:
- Bail out when a handler returns failure for a key-value pair. This
helps with cases where processing a key-value pair is not just a
matter of setting some tool specific knob, involving, for instance
building a BPF program to then attach to the list of events 'perf
trace' will use, e.g. augmented_raw_syscalls.c.
perf.data:
Kan Liang:
- Read and store die ID information available in new Intel processors
in CPUID.1F in the CPU topology written in the perf.data header.
perf stat:
Kan Liang:
- Support per-die aggregation.
Documentation:
Arnaldo Carvalho de Melo:
- Update perf.data documentation about the CPU_TOPOLOGY, MEM_TOPOLOGY,
CLOCKID and DIR_FORMAT headers.
Song Liu:
- Add description of headers HEADER_BPF_PROG_INFO and HEADER_BPF_BTF.
Leo Yan:
- Update default value for llvm.clang-bpf-cmd-template in 'man perf-config'.
JVMTI:
Jiri Olsa:
- Address gcc string overflow warning for strncpy()
core:
- Remove superfluous nthreads system_wide setup in perf_evsel__alloc_fd().
Intel PT:
Adrian Hunter:
- Add support for samples to contain IPC ratio, collecting cycles
information from CYC packets, showing the IPC info periodically, because
Intel PT does not update the cycle count on every branch or instruction,
the incremental values will often be zero. When there are values, they
will be the number of instructions and number of cycles since the last
update, and thus represent the average IPC since the last IPC value.
E.g.:
# perf record --cpu 1 -m200000 -a -e intel_pt/cyc/u sleep 0.0001
rounding mmap pages size to 1024M (262144 pages)
[ perf record: Woken up 0 times to write data ]
[ perf record: Captured and wrote 2.208 MB perf.data ]
# perf script --insn-trace --xed -F+ipc,-dso,-cpu,-tid
#
<SNIP + add line numbering to make sense of IPC counts e.g.: (18/3)>
1 cc1 63501.650479626: 7f5219ac27bf _int_free+0x3f jnz 0x7f5219ac2af0 IPC: 0.81 (36/44)
2 cc1 63501.650479626: 7f5219ac27c5 _int_free+0x45 cmp $0x1f, %rbp
3 cc1 63501.650479626: 7f5219ac27c9 _int_free+0x49 jbe 0x7f5219ac2b00
4 cc1 63501.650479626: 7f5219ac27cf _int_free+0x4f test $0x8, %al
5 cc1 63501.650479626: 7f5219ac27d1 _int_free+0x51 jnz 0x7f5219ac2b00
6 cc1 63501.650479626: 7f5219ac27d7 _int_free+0x57 movq 0x13c58a(%rip), %rcx
7 cc1 63501.650479626: 7f5219ac27de _int_free+0x5e mov %rdi, %r12
8 cc1 63501.650479626: 7f5219ac27e1 _int_free+0x61 movq %fs:(%rcx), %rax
9 cc1 63501.650479626: 7f5219ac27e5 _int_free+0x65 test %rax, %rax
10 cc1 63501.650479626: 7f5219ac27e8 _int_free+0x68 jz 0x7f5219ac2821
11 cc1 63501.650479626: 7f5219ac27ea _int_free+0x6a leaq -0x11(%rbp), %rdi
12 cc1 63501.650479626: 7f5219ac27ee _int_free+0x6e mov %rdi, %rsi
13 cc1 63501.650479626: 7f5219ac27f1 _int_free+0x71 shr $0x4, %rsi
14 cc1 63501.650479626: 7f5219ac27f5 _int_free+0x75 cmpq %rsi, 0x13caf4(%rip)
15 cc1 63501.650479626: 7f5219ac27fc _int_free+0x7c jbe 0x7f5219ac2821
16 cc1 63501.650479626: 7f5219ac2821 _int_free+0xa1 cmpq 0x13f138(%rip), %rbp
17 cc1 63501.650479626: 7f5219ac2828 _int_free+0xa8 jnbe 0x7f5219ac28d8
18 cc1 63501.650479626: 7f5219ac28d8 _int_free+0x158 testb $0x2, 0x8(%rbx)
19 cc1 63501.650479628: 7f5219ac28dc _int_free+0x15c jnz 0x7f5219ac2ab0 IPC: 6.00 (18/3)
<SNIP>
- Allow using time ranges with Intel PT, i.e. these features, already
present but not optimially usable with Intel PT, should be now:
Select the second 10% time slice:
$ perf script --time 10%/2
Select from 0% to 10% time slice:
$ perf script --time 0%-10%
Select the first and second 10% time slices:
$ perf script --time 10%/1,10%/2
Select from 0% to 10% and 30% to 40% slices:
$ perf script --time 0%-10%,30%-40%
cs-etm (ARM):
Mathieu Poirier:
- Add support for CPU-wide trace scenarios.
s390:
Thomas Richter:
- Fix missing kvm module load for s390.
- Fix OOM error in TUI mode on s390
- Support s390 diag event display when doing analysis on !s390
architectures.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
-----BEGIN PGP SIGNATURE-----
iHUEABYIAB0WIQR2GiIUctdOfX2qHhGyPKLppCJ+JwUCXP/1xQAKCRCyPKLppCJ+
J9xcAQCwOITAshE7op7HbKUPtkqiMNu+hpNa3skhxEpGHvKO0AEArpBXtuvEP8EU
PZsp+8vcVrlZ+dZutttgvkRz25mScg8=
=kfFb
-----END PGP SIGNATURE-----
Merge tag 'perf-core-for-mingo-5.3-20190611' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core
Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo:
perf record:
Alexey Budankov:
- Allow mixing --user-regs with --call-graph=dwarf, making sure that
the minimal set of registers for DWARF unwinding is present in the
set of user registers requested to be present in each sample, while
warning the user that this may make callchains unreliable if more
that the minimal set of registers is needed to unwind.
yuzhoujian:
- Add support to collect callchains from kernel or user space only,
IOW allow setting the perf_event_attr.exclude_callchain_{kernel,user}
bits from the command line.
perf trace:
Arnaldo Carvalho de Melo:
- Remove x86_64 specific syscall numbers from the augmented_raw_syscalls
BPF in-kernel collector of augmented raw_syscalls:sys_{enter,exit}
payloads, use instead the syscall numbers obtainer either by the
arch specific syscalltbl generators or from audit-libs.
- Allow 'perf trace' to ask for the number of bytes to collect for
string arguments, for now ask for PATH_MAX, i.e. the whole
pathnames, which ends up being just a way to speficy which syscall
args are pathnames and thus should be read using bpf_probe_read_str().
- Skip unknown syscalls when expanding strace like syscall groups.
This helps using the 'string' group of syscalls to work in arm64,
where some of the syscalls present in x86_64 that deal with
strings, for instance 'access', are deprecated and this should not
be asked for tracing.
Leo Yan:
- Exit when failing to build eBPF program.
perf config:
Arnaldo Carvalho de Melo:
- Bail out when a handler returns failure for a key-value pair. This
helps with cases where processing a key-value pair is not just a
matter of setting some tool specific knob, involving, for instance
building a BPF program to then attach to the list of events 'perf
trace' will use, e.g. augmented_raw_syscalls.c.
perf.data:
Kan Liang:
- Read and store die ID information available in new Intel processors
in CPUID.1F in the CPU topology written in the perf.data header.
perf stat:
Kan Liang:
- Support per-die aggregation.
Documentation:
Arnaldo Carvalho de Melo:
- Update perf.data documentation about the CPU_TOPOLOGY, MEM_TOPOLOGY,
CLOCKID and DIR_FORMAT headers.
Song Liu:
- Add description of headers HEADER_BPF_PROG_INFO and HEADER_BPF_BTF.
Leo Yan:
- Update default value for llvm.clang-bpf-cmd-template in 'man perf-config'.
JVMTI:
Jiri Olsa:
- Address gcc string overflow warning for strncpy()
core:
- Remove superfluous nthreads system_wide setup in perf_evsel__alloc_fd().
Intel PT:
Adrian Hunter:
- Add support for samples to contain IPC ratio, collecting cycles
information from CYC packets, showing the IPC info periodically, because
Intel PT does not update the cycle count on every branch or instruction,
the incremental values will often be zero. When there are values, they
will be the number of instructions and number of cycles since the last
update, and thus represent the average IPC since the last IPC value.
E.g.:
# perf record --cpu 1 -m200000 -a -e intel_pt/cyc/u sleep 0.0001
rounding mmap pages size to 1024M (262144 pages)
[ perf record: Woken up 0 times to write data ]
[ perf record: Captured and wrote 2.208 MB perf.data ]
# perf script --insn-trace --xed -F+ipc,-dso,-cpu,-tid
#
<SNIP + add line numbering to make sense of IPC counts e.g.: (18/3)>
1 cc1 63501.650479626: 7f5219ac27bf _int_free+0x3f jnz 0x7f5219ac2af0 IPC: 0.81 (36/44)
2 cc1 63501.650479626: 7f5219ac27c5 _int_free+0x45 cmp $0x1f, %rbp
3 cc1 63501.650479626: 7f5219ac27c9 _int_free+0x49 jbe 0x7f5219ac2b00
4 cc1 63501.650479626: 7f5219ac27cf _int_free+0x4f test $0x8, %al
5 cc1 63501.650479626: 7f5219ac27d1 _int_free+0x51 jnz 0x7f5219ac2b00
6 cc1 63501.650479626: 7f5219ac27d7 _int_free+0x57 movq 0x13c58a(%rip), %rcx
7 cc1 63501.650479626: 7f5219ac27de _int_free+0x5e mov %rdi, %r12
8 cc1 63501.650479626: 7f5219ac27e1 _int_free+0x61 movq %fs:(%rcx), %rax
9 cc1 63501.650479626: 7f5219ac27e5 _int_free+0x65 test %rax, %rax
10 cc1 63501.650479626: 7f5219ac27e8 _int_free+0x68 jz 0x7f5219ac2821
11 cc1 63501.650479626: 7f5219ac27ea _int_free+0x6a leaq -0x11(%rbp), %rdi
12 cc1 63501.650479626: 7f5219ac27ee _int_free+0x6e mov %rdi, %rsi
13 cc1 63501.650479626: 7f5219ac27f1 _int_free+0x71 shr $0x4, %rsi
14 cc1 63501.650479626: 7f5219ac27f5 _int_free+0x75 cmpq %rsi, 0x13caf4(%rip)
15 cc1 63501.650479626: 7f5219ac27fc _int_free+0x7c jbe 0x7f5219ac2821
16 cc1 63501.650479626: 7f5219ac2821 _int_free+0xa1 cmpq 0x13f138(%rip), %rbp
17 cc1 63501.650479626: 7f5219ac2828 _int_free+0xa8 jnbe 0x7f5219ac28d8
18 cc1 63501.650479626: 7f5219ac28d8 _int_free+0x158 testb $0x2, 0x8(%rbx)
19 cc1 63501.650479628: 7f5219ac28dc _int_free+0x15c jnz 0x7f5219ac2ab0 IPC: 6.00 (18/3)
<SNIP>
- Allow using time ranges with Intel PT, i.e. these features, already
present but not optimially usable with Intel PT, should be now:
Select the second 10% time slice:
$ perf script --time 10%/2
Select from 0% to 10% time slice:
$ perf script --time 0%-10%
Select the first and second 10% time slices:
$ perf script --time 10%/1,10%/2
Select from 0% to 10% and 30% to 40% slices:
$ perf script --time 0%-10%,30%-40%
cs-etm (ARM):
Mathieu Poirier:
- Add support for CPU-wide trace scenarios.
s390:
Thomas Richter:
- Fix missing kvm module load for s390.
- Fix OOM error in TUI mode on s390
- Support s390 diag event display when doing analysis on !s390
architectures.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
We have $INSTALL_DIR/share/perf-core/strace/groups/string files with
syscalls that should be selected when 'string' is used, meaning, in this
case, syscalls that receive as one of its arguments a string, like a
pathname.
But those were first selected and tested on x86_64, and end up failing
in architectures where some of those syscalls are not available, like
the 'access' syscall on arm64, which makes using 'perf trace -e string'
in such archs to fail.
Since this the routine doing the validation is used only when reading
such files, do not fail when some syscall is not found in the
syscalltbl, instead just use pr_debug() to register that in case people
are suspicious of problems.
Now using 'perf trace -e string' should work on arm64, selecting only
the syscalls that have a string and are available on that architecture.
Reported-by: Leo Yan <leo.yan@linaro.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Martin KaFai Lau <kafai@fb.com>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Mike Leach <mike.leach@linaro.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: Suzuki K Poulose <suzuki.poulose@arm.com>
Cc: Yonghong Song <yhs@fb.com>
Link: https://lkml.kernel.org/r/20190610184754.GU21245@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
On my Juno board with ARM64 CPUs, perf trace command reports the eBPF
program building failure but the command will not exit and continue to
run. If we define an eBPF event in config file, the event will be
parsed with below flow:
perf_config()
`> trace__config()
`> parse_events_option()
`> parse_events__scanner()
`-> parse_events_parse()
`> parse_events_load_bpf()
`> llvm__compile_bpf()
Though the low level functions return back error values when detect eBPF
building failure, but parse_events_option() returns 1 for this case and
trace__config() passes 1 to perf_config(); perf_config() doesn't treat
the returned value 1 as failure and it continues to parse other
configurations. Thus the perf command continues to run even without
enabling eBPF event successfully.
This patch changes error handling in trace__config(), when it detects
failure it will return -1 rather than directly pass error value (1);
finally, perf_config() will directly bail out and perf will exit for
this case.
Committer notes:
Simplified the patch to just check directly the return of
parse_events_option() and it it is non-zero, change err from its initial
zero value to -1.
Signed-off-by: Leo Yan <leo.yan@linaro.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Martin KaFai Lau <kafai@fb.com>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Mike Leach <mike.leach@linaro.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: Suzuki Poulouse <suzuki.poulose@arm.com>
Cc: Yonghong Song <yhs@fb.com>
Fixes: ac96287cae ("perf trace: Allow specifying a set of events to add in perfconfig")
Link: https://lkml.kernel.org/n/tip-x4i63f5kscykfok0hqim3zma@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Based on 1 normalized pattern(s):
released under the gpl v2 and only v2 not any later version
extracted by the scancode license scanner the SPDX license identifier
GPL-2.0-only
has been chosen to replace the boilerplate/reference in 12 file(s).
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Steve Winslow <swinslow@gmail.com>
Reviewed-by: Alexios Zavras <alexios.zavras@intel.com>
Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org>
Reviewed-by: Allison Randal <allison@lohutok.net>
Cc: linux-spdx@vger.kernel.org
Link: https://lkml.kernel.org/r/20190529141332.526460839@linutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
For instance, the rename* family uses "oldname", "newname", so check if
"name" is at the end and treat it as a filename.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-wjy7j4bk06g7atzwoz1mid24@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
To support the SCA_FILENAME beautifier in more than one syscall arg, as
needed for syscalls such as the rename* family, we need to, after
processing one such arg, bump the augmented pointers so that the next
augmented arg don't reuse data for the previous augmented arguments.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-4e4cmzyjxb3wkonfo1x9a27y@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Since we know what args are strings from reading the syscall
descriptions in tracefs and also already mark such args to be beautified
using the syscall_arg__scnprintf_filename() helper, all we need is to
fill in this info in the 'syscalls' BPF map we were using to state which
syscalls the user is interested in, i.e. the syscall filter.
Right now just set that with PATH_MAX and unroll the syscall arg in the
BPF program, as the verifier isn't liking something clang generates when
unrolling the loop.
This also makes the augmented_raw_syscalls.c program support all arches,
since we removed that set of defines with the hard coded syscall
numbers, all should be automatically set for all arches, with the
syscall id mapping done correcly.
Doing baby steps here, i.e. just the first string arg for a syscall is
printed, syscalls with more than one, say, the various rename* syscalls,
need further work, but lets get first something that the BPF verifier
accepts before increasing the complexity
To test it, something like:
# perf trace -e string -e /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c
With:
# cat ~/.perfconfig
[llvm]
dump-obj = true
clang-opt = -g
[trace]
#add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c
show_zeros = yes
show_duration = no
no_inherit = yes
show_timestamp = no
show_arg_names = no
args_alignment = 40
show_prefix = yes
#
That commented add_events line is needed for developing this
augmented_raw_syscalls.c BPF program, as if we add it via the
'add_events' mechanism so as to shorten the 'perf trace' command lines,
then we end up not setting up the -v option which precludes us having
access to the bpf verifier log :-\
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@fb.com>
Cc: Andrii Nakryiko <andriin@fb.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Jesper Dangaard Brouer <brouer@redhat.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: Yonghong Song <yhs@fb.com>
Link: https://lkml.kernel.org/n/tip-dn863ya0cbsqycxuy0olvbt1@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
So that one can just define a strarray and process it as a set of flags,
similar to syscall_arg__scnprintf_strarray() with plain arrays.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-nnt25wkpkow2w0yefhi6sb7q@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Use existing beautifiers for the first arg, fd, assigned using the
heuristic that looks for syscall arg names and associates SCA_FD with
'fd' named argumes, and wire up the recently introduced fsconfig cmd
table generator.
Now it should be possible to just use:
perf trace -e fsconfig
As root and see all fsconfig syscalls with its args beautified, more
work needed to look at the command and according to it handle the 'key',
'value' and 'aux' args, using the 'fcntl' and 'futex' beautifiers as a
starting point to see how to suppress sets of these last three args that
may not be used by the 'cmd' arg, etc.
# cat sys_fsconfig.c
#define _GNU_SOURCE /* See feature_test_macros(7) */
#include <unistd.h>
#include <sys/syscall.h> /* For SYS_xxx definitions */
#include <fcntl.h>
#define __NR_fsconfig 431
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 */
};
static inline int sys_fsconfig(int fd, int cmd, const char *key, const void *value, int aux)
{
syscall(__NR_fsconfig, fd, cmd, key, value, aux);
}
int main(int argc, char *argv[])
{
int fd = 0, aux = 0;
open("/foo", 0);
sys_fsconfig(fd++, FSCONFIG_SET_FLAG, "/foo1", "/bar1", aux++);
sys_fsconfig(fd++, FSCONFIG_SET_STRING, "/foo2", "/bar2", aux++);
sys_fsconfig(fd++, FSCONFIG_SET_BINARY, "/foo3", "/bar3", aux++);
sys_fsconfig(fd++, FSCONFIG_SET_PATH, "/foo4", "/bar4", aux++);
sys_fsconfig(fd++, FSCONFIG_SET_PATH_EMPTY, "/foo5", "/bar5", aux++);
sys_fsconfig(fd++, FSCONFIG_SET_FD, "/foo6", "/bar6", aux++);
sys_fsconfig(fd++, FSCONFIG_CMD_CREATE, "/foo7", "/bar7", aux++);
sys_fsconfig(fd++, FSCONFIG_CMD_RECONFIGURE, "/foo8", "/bar8", aux++);
return 0;
}
# trace -e fsconfig ./sys_fsconfig
fsconfig(0, FSCONFIG_SET_FLAG, 0x40201b, 0x402015, 0) = -1 EINVAL (Invalid argument)
fsconfig(1, FSCONFIG_SET_STRING, 0x402027, 0x402021, 1) = -1 EINVAL (Invalid argument)
fsconfig(2, FSCONFIG_SET_BINARY, 0x402033, 0x40202d, 2) = -1 EINVAL (Invalid argument)
fsconfig(3, FSCONFIG_SET_PATH, 0x40203f, 0x402039, 3) = -1 EBADF (Bad file descriptor)
fsconfig(4, FSCONFIG_SET_PATH_EMPTY, 0x40204b, 0x402045, 4) = -1 EBADF (Bad file descriptor)
fsconfig(5, FSCONFIG_SET_FD, 0x402057, 0x402051, 5) = -1 EINVAL (Invalid argument)
fsconfig(6, FSCONFIG_CMD_CREATE, 0x402063, 0x40205d, 6) = -1 EINVAL (Invalid argument)
fsconfig(7, FSCONFIG_CMD_RECONFIGURE, 0x40206f, 0x402069, 7) = -1 EINVAL (Invalid argument)
#
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-fb04b76cm59zfuv1wzu40uxy@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Use existing beautifiers for the first 2 args (dfd, path) and wire up
the recently introduced fspick flags table generator.
Now it should be possible to just use:
perf trace -e fspick
As root and see all move_mount syscalls with its args beautified, either
using the vfs_getname perf probe method or using the
augmented_raw_syscalls.c eBPF helper to get the pathnames, the other
args should work in all cases, i.e. all that is needed can be obtained
directly from the raw_syscalls:sys_enter tracepoint args.
# cat sys_fspick.c
#define _GNU_SOURCE /* See feature_test_macros(7) */
#include <unistd.h>
#include <sys/syscall.h> /* For SYS_xxx definitions */
#include <fcntl.h>
#define __NR_fspick 433
#define FSPICK_CLOEXEC 0x00000001
#define FSPICK_SYMLINK_NOFOLLOW 0x00000002
#define FSPICK_NO_AUTOMOUNT 0x00000004
#define FSPICK_EMPTY_PATH 0x00000008
static inline int sys_fspick(int fd, const char *path, int flags)
{
syscall(__NR_fspick, fd, path, flags);
}
int main(int argc, char *argv[])
{
int flags = 0, fd = 0;
open("/foo", 0);
sys_fspick(fd++, "/foo1", flags);
flags |= FSPICK_CLOEXEC;
sys_fspick(fd++, "/foo2", flags);
flags |= FSPICK_SYMLINK_NOFOLLOW;
sys_fspick(fd++, "/foo3", flags);
flags |= FSPICK_NO_AUTOMOUNT;
sys_fspick(fd++, "/foo4", flags);
flags |= FSPICK_EMPTY_PATH;
return sys_fspick(fd++, "/foo5", flags);
}
# perf trace -e fspick ./sys_fspick
LLVM: dumping /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.o
fspick(0, "/foo1", 0) = -1 ENOENT (No such file or directory)
fspick(1, "/foo2", FSPICK_CLOEXEC) = -1 ENOENT (No such file or directory)
fspick(2, "/foo3", FSPICK_CLOEXEC|FSPICK_SYMLINK_NOFOLLOW) = -1 ENOENT (No such file or directory)
fspick(3, "/foo4", FSPICK_CLOEXEC|FSPICK_SYMLINK_NOFOLLOW|FSPICK_NO_AUTOMOUNT) = -1 ENOENT (No such file or directory)
fspick(4, "/foo5", FSPICK_CLOEXEC|FSPICK_SYMLINK_NOFOLLOW|FSPICK_NO_AUTOMOUNT|FSPICK_EMPTY_PATH) = -1 ENOENT (No such file or directory)
#
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-erau5xjtt8wvgnhvdbchstuk@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Use existing beautifiers for the first 4 args (to/from fds, pathnames)
and wire up the recently introduced move_mount flags table generator.
Now it should be possible to just use:
perf trace -e move_mount
As root and see all move_mount syscalls with its args beautified, except
for the filenames, that need work in the augmented_raw_syscalls.c eBPF
helper to pass more than one, see comment in the
augmented_raw_syscalls.c source code, the other args should work in all
cases, i.e. all that is needed can be obtained directly from the
raw_syscalls:sys_enter tracepoint args.
Running without the strace "skin" (.perfconfig setting output formatting
switches to look like strace output + BPF to collect strings, as we
still need to support collecting multiple string args for the same
syscall, like with move_mount):
# cat sys_move_mount.c
#define _GNU_SOURCE /* See feature_test_macros(7) */
#include <unistd.h>
#include <sys/syscall.h> /* For SYS_xxx definitions */
#define __NR_move_mount 429
#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 */
static inline int sys_move_mount(int from_fd, const char *from_pathname,
int to_fd, const char *to_pathname,
int flags)
{
syscall(__NR_move_mount, from_fd, from_pathname, to_fd, to_pathname, flags);
}
int main(int argc, char *argv[])
{
int flags = 0, from_fd = 0, to_fd = 100;
sys_move_mount(from_fd++, "/foo", to_fd++, "bar", flags);
flags |= MOVE_MOUNT_F_SYMLINKS;
sys_move_mount(from_fd++, "/foo1", to_fd++, "bar1", flags);
flags |= MOVE_MOUNT_F_AUTOMOUNTS;
sys_move_mount(from_fd++, "/foo2", to_fd++, "bar2", flags);
flags |= MOVE_MOUNT_F_EMPTY_PATH;
sys_move_mount(from_fd++, "/foo3", to_fd++, "bar3", flags);
flags |= MOVE_MOUNT_T_SYMLINKS;
sys_move_mount(from_fd++, "/foo4", to_fd++, "bar4", flags);
flags |= MOVE_MOUNT_T_AUTOMOUNTS;
sys_move_mount(from_fd++, "/foo5", to_fd++, "bar5", flags);
flags |= MOVE_MOUNT_T_EMPTY_PATH;
return sys_move_mount(from_fd++, "/foo6", to_fd++, "bar6", flags);
}
# mv ~/.perfconfig ~/.perfconfig.OFF
# perf trace -e move_mount ./sys_move_mount
0.000 ( 0.009 ms): sys_move_mount/28971 move_mount(from_pathname: 0x402010, to_dfd: 100, to_pathname: 0x402015) = -1 ENOENT (No such file or directory)
0.011 ( 0.003 ms): sys_move_mount/28971 move_mount(from_dfd: 1, from_pathname: 0x40201e, to_dfd: 101, to_pathname: 0x402019, flags: F_SYMLINKS) = -1 ENOENT (No such file or directory)
0.016 ( 0.002 ms): sys_move_mount/28971 move_mount(from_dfd: 2, from_pathname: 0x402029, to_dfd: 102, to_pathname: 0x402024, flags: F_SYMLINKS|F_AUTOMOUNTS) = -1 ENOENT (No such file or directory)
0.020 ( 0.002 ms): sys_move_mount/28971 move_mount(from_dfd: 3, from_pathname: 0x402034, to_dfd: 103, to_pathname: 0x40202f, flags: F_SYMLINKS|F_AUTOMOUNTS|F_EMPTY_PATH) = -1 ENOENT (No such file or directory)
0.023 ( 0.002 ms): sys_move_mount/28971 move_mount(from_dfd: 4, from_pathname: 0x40203f, to_dfd: 104, to_pathname: 0x40203a, flags: F_SYMLINKS|F_AUTOMOUNTS|F_EMPTY_PATH|T_SYMLINKS) = -1 ENOENT (No such file or directory)
0.027 ( 0.002 ms): sys_move_mount/28971 move_mount(from_dfd: 5, from_pathname: 0x40204a, to_dfd: 105, to_pathname: 0x402045, flags: F_SYMLINKS|F_AUTOMOUNTS|F_EMPTY_PATH|T_SYMLINKS|T_AUTOMOUNTS) = -1 ENOENT (No such file or directory)
0.031 ( 0.017 ms): sys_move_mount/28971 move_mount(from_dfd: 6, from_pathname: 0x402055, to_dfd: 106, to_pathname: 0x402050, flags: F_SYMLINKS|F_AUTOMOUNTS|F_EMPTY_PATH|T_SYMLINKS|T_AUTOMOUNTS|T_EMPTY_PATH) = -1 ENOENT (No such file or directory)
#
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-83rim8g4k0s4gieieh5nnlck@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add a 'path' member to 'struct perf_data'. It will keep the configured
path for the data (const char *). The path in struct perf_data_file is
now dynamically allocated (duped) from it.
This scheme is useful/used in following patches where struct
perf_data::path holds the 'configure' directory path and struct
perf_data_file::path holds the allocated path for specific files.
Also it actually makes the code little simpler.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alexey Budankov <alexey.budankov@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/20190221094145.9151-3-jolsa@kernel.org
[ Fixup data-convert-bt.c missing conversion ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Just like it does with 'sshd', to reduce the feedback loop when doing
system wide tracing on on a gnome GUI.
Need to figure out how to auto-filter the calls to other UI components
tho.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/n/tip-rjopq5y92itgokppdhe8sc6z@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
We were crashing when processing a negative fd:
Program received signal SIGSEGV, Segmentation fault.
0x0000000000609bbf in syscall_arg__scnprintf_ioctl_cmd (bf=0x1172eca "", size=2038, arg=0x7fffffff8360) at trace/beauty/ioctl.c:182
182 if (file->dev_maj == USB_DEVICE_MAJOR)
Missing separate debuginfos, use: dnf debuginfo-install bzip2-libs-1.0.6-28.fc29.x86_64 elfutils-libelf-0.174-5.fc29.x86_64 elfutils-libs-0.174-5.fc29.x86_64 glib2-2.58.3-1.fc29.x86_64 libbabeltrace-1.5.6-1.fc29.x86_64 libunwind-1.2.1-6.fc29.x86_64 libuuid-2.32.1-1.fc29.x86_64 libxcrypt-4.4.3-2.fc29.x86_64 numactl-libs-2.0.12-1.fc29.x86_64 openssl-libs-1.1.1a-1.fc29.x86_64 pcre-8.42-6.fc29.x86_64 perl-libs-5.28.1-427.fc29.x86_64 popt-1.16-15.fc29.x86_64 python2-libs-2.7.15-11.fc29.x86_64 slang-2.3.2-4.fc29.x86_64 xz-libs-5.2.4-3.fc29.x86_64
(gdb) bt
#0 0x0000000000609bbf in syscall_arg__scnprintf_ioctl_cmd (bf=0x1172eca "", size=2038, arg=0x7fffffff8360) at trace/beauty/ioctl.c:182
#1 0x000000000048e295 in syscall__scnprintf_val (sc=0x123b500, bf=0x1172eca "", size=2038, arg=0x7fffffff8360, val=21519)
at builtin-trace.c:1594
#2 0x000000000048e60d in syscall__scnprintf_args (sc=0x123b500, bf=0x1172ec6 "-1, ", size=2042, args=0x7ffff6a7c034 "\377\377\377\377",
augmented_args=0x7ffff6a7c064, augmented_args_size=4, trace=0x7fffffffa8d0, thread=0x1175cd0) at builtin-trace.c:1661
#3 0x000000000048f04e in trace__sys_enter (trace=0x7fffffffa8d0, evsel=0xb260b0, event=0x7ffff6a7bfe8, sample=0x7fffffff84f0)
at builtin-trace.c:1880
#4 0x00000000004915a4 in trace__handle_event (trace=0x7fffffffa8d0, event=0x7ffff6a7bfe8, sample=0x7fffffff84f0) at builtin-trace.c:2590
#5 0x0000000000491eed in __trace__deliver_event (trace=0x7fffffffa8d0, event=0x7ffff6a7bfe8) at builtin-trace.c:2818
#6 0x0000000000492030 in trace__deliver_event (trace=0x7fffffffa8d0, event=0x7ffff6a7bfe8) at builtin-trace.c:2845
#7 0x0000000000492896 in trace__run (trace=0x7fffffffa8d0, argc=0, argv=0x7fffffffdb58) at builtin-trace.c:3040
#8 0x000000000049603a in cmd_trace (argc=0, argv=0x7fffffffdb58) at builtin-trace.c:3952
#9 0x00000000004d5103 in main (argc=1, argv=0x7fffffffdb58) at perf.c:474
(gdb) p fd
$1 = -1
(gdb) p file
$7 = (struct file *) 0xfffffffffffffff0
(gdb) p ((struct thread_trace *)arg->thread)->files.table + fd
$8 = (struct file *) 0xfffffffffffffff0
(gdb)
Check for that and return NULL instead.
This problem was introduced recently, the other codepaths leading to
thread_trace__files_entry() check for negative fds, like thread__fd_path(),
but we need to do it at thread_trace__files_entry() as more users are now
calling it directly.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Fixes: 2d473389f8 ("perf trace beauty: Export function to get the files for a thread")
Link: https://lkml.kernel.org/n/tip-oq7bvaaf07gsd4yqty3107u2@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Hardware tracing:
Adrian Hunter:
- Handle calls optimized into jumps to a different symbol
in the thread stack routines used to process hardware traces (Adrian Hunter)
Intel PT:
Adrian Hunter:
- Fix overlap calculation for padding.
- Fix CYC timestamp calculation after OVF.
- Packet splitting can only happen in 32-bit.
- Add timestamp to auxtrace errors.
ARM CoreSight:
Leo Yan:
- Add last instruction information in packet
- Set sample flags for instruction range, exception and
return packets and for a trace discontinuity.
- Add exception number in exception packet
- Change tuple from traceID-CPU# to traceID-metadata
- Add traceID in packet
Mathieu Poirier:
- Add "sinks" group to PMU directory
- Use event attributes to send sink information to kernel
- Remove set_drv_config() API, no longer used.
perf annotate:
Jiri Olsa:
- Delay symbol annotation to the resort phase, speeding up 'perf report'
startup.
perf record:
Alexey Budankov:
- Allow binding userspace buffers to NUMA nodes.
Symbols:
Adrian Hunter:
- Fix calculating of symbol sizes when splitting kallsyms into
maps for kcore processing.
Vendor events:
William Cohen:
- Intel: Fix Load_Miss_Real_Latency on CLX
Misc:
Arnaldo Carvalho de Melo:
- Streamline headers, removing includes when all that is needed are
just forward declarations, fixup the fallout for cases where headers
should have been explicitely included but were instead obtained
indirectly, by sheer luck.
- Add fallback versions for CPU_{OR,EQUAL}(), so that code using it
continue to build on older systems where those were not yet introduced
or in systems using some other libc than the GNU one where those
helpers aren't present.
Documentation:
Changbin Du:
- Add documentation for BPF event selection.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
-----BEGIN PGP SIGNATURE-----
iHUEABYIAB0WIQR2GiIUctdOfX2qHhGyPKLppCJ+JwUCXFsqugAKCRCyPKLppCJ+
JzpwAQDEh1mNZoxfdGZEi9d+8p2hnRlOs3GOUG4iGnqAYfae4QEAkMJ0V1wrmkdw
NXgV+PgWfDcgbD4Cn90eWA8M6KEcbgA=
=ogOF
-----END PGP SIGNATURE-----
Merge tag 'perf-core-for-mingo-5.1-20190206' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core
Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo:
Hardware tracing:
Adrian Hunter:
- Handle calls optimized into jumps to a different symbol
in the thread stack routines used to process hardware traces (Adrian Hunter)
Intel PT:
Adrian Hunter:
- Fix overlap calculation for padding.
- Fix CYC timestamp calculation after OVF.
- Packet splitting can only happen in 32-bit.
- Add timestamp to auxtrace errors.
ARM CoreSight:
Leo Yan:
- Add last instruction information in packet
- Set sample flags for instruction range, exception and
return packets and for a trace discontinuity.
- Add exception number in exception packet
- Change tuple from traceID-CPU# to traceID-metadata
- Add traceID in packet
Mathieu Poirier:
- Add "sinks" group to PMU directory
- Use event attributes to send sink information to kernel
- Remove set_drv_config() API, no longer used.
perf annotate:
Jiri Olsa:
- Delay symbol annotation to the resort phase, speeding up 'perf report'
startup.
perf record:
Alexey Budankov:
- Allow binding userspace buffers to NUMA nodes.
Symbols:
Adrian Hunter:
- Fix calculating of symbol sizes when splitting kallsyms into
maps for kcore processing.
Vendor events:
William Cohen:
- Intel: Fix Load_Miss_Real_Latency on CLX
Misc:
Arnaldo Carvalho de Melo:
- Streamline headers, removing includes when all that is needed are
just forward declarations, fixup the fallout for cases where headers
should have been explicitely included but were instead obtained
indirectly, by sheer luck.
- Add fallback versions for CPU_{OR,EQUAL}(), so that code using it
continue to build on older systems where those were not yet introduced
or in systems using some other libc than the GNU one where those
helpers aren't present.
Documentation:
Changbin Du:
- Add documentation for BPF event selection.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Several places were using definitions found in symbols.h but not
including it, getting it by sheer luck from some other headers that now
are in the process of removing that include because they don't need it
or because simply having struct forward declarations is enough, fix it.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-xbcvvx296d70kpg9wb0qmeq9@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Lots of places get the map.h file indirectly, and since we're going to
remove it from machine.h, then those need to include it directly, do it
now, before we remove that dep.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-ob8jehdjda8h5jsrv9dqj9tf@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
With a suitably defined "probe:vfs_getname" probe, 'perf trace' can
"beautify" its output, so syscalls like open() or openat() can print the
"filename" argument instead of just its hex address, like:
$ perf trace -e open -- touch /dev/null
[...]
0.590 ( 0.014 ms): touch/18063 open(filename: /dev/null, flags: CREAT|NOCTTY|NONBLOCK|WRONLY, mode: IRUGO|IWUGO) = 3
[...]
The output without such beautifier looks like:
0.529 ( 0.011 ms): touch/18075 open(filename: 0xc78cf288, flags: CREAT|NOCTTY|NONBLOCK|WRONLY, mode: IRUGO|IWUGO) = 3
However, when the vfs_getname probe expands to multiple probes and it is
not the first one that is hit, the beautifier fails, as following:
0.326 ( 0.010 ms): touch/18072 open(filename: , flags: CREAT|NOCTTY|NONBLOCK|WRONLY, mode: IRUGO|IWUGO) = 3
Fix it by hooking into all the expanded probes (inlines), now, for instance:
[root@quaco ~]# perf probe -l
probe:vfs_getname (on getname_flags:73@fs/namei.c with pathname)
probe:vfs_getname_1 (on getname_flags:73@fs/namei.c with pathname)
[root@quaco ~]# perf trace -e open* sleep 1
0.010 ( 0.005 ms): sleep/5588 openat(dfd: CWD, filename: /etc/ld.so.cache, flags: RDONLY|CLOEXEC) = 3
0.029 ( 0.006 ms): sleep/5588 openat(dfd: CWD, filename: /lib64/libc.so.6, flags: RDONLY|CLOEXEC) = 3
0.194 ( 0.008 ms): sleep/5588 openat(dfd: CWD, filename: /usr/lib/locale/locale-archive, flags: RDONLY|CLOEXEC) = 3
[root@quaco ~]#
Works, further verified with:
[root@quaco ~]# perf test vfs
65: Use vfs_getname probe to get syscall args filenames : Ok
66: Add vfs_getname probe to get syscall args filenames : Ok
67: Check open filename arg using perf trace + vfs_getname: Ok
[root@quaco ~]#
Reported-by: Michael Petlan <mpetlan@redhat.com>
Tested-by: Michael Petlan <mpetlan@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-mv8kolk17xla1smvmp3qabv1@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Where we don't have "raw_syscalls:sys_enter", so we need to look for a
"*syscalls:sys_enter*" to initialize the offsets for the
__augmented_syscalls__ evsel, which is the case with etcsnoop, that was
segfaulting, fixed:
# trace -e /home/acme/git/perf/tools/perf/examples/bpf/etcsnoop.c
0.000 ( ): gnome-shell/2105 openat(dfd: CWD, filename: "/etc/localtime") ...
631.834 ( ): cat/6521 openat(dfd: CWD, filename: "/etc/ld.so.cache", flags: RDONLY|CLOEXEC) ...
632.637 ( ): bash/6521 openat(dfd: CWD, filename: "/etc/passwd") ...
^C#
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Fixes: b9b6a2ea2b ("perf trace: Do not hardcode the size of the tracepoint common_ fields")
Link: https://lkml.kernel.org/n/tip-0tjwcit8qitsmh4nyvf2b0jo@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
We were not taking into account the "... [continued]" printed
characters, fix it.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/n/tip-qt20y0acmf8k0bzisce8kw95@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When we get the sys_enter for a syscall we check if the last one is
still waiting for its matching sys_exit, if so we print this:
468.753 ( ): firefox/32382 poll(ufds: 0x7f3988d3dd00, nfds: 7, timeout_msecs: 4294967295) ...
449.575 ( 0.004 ms): Softwar~cThrea/32434 futex(uaddr: 0x7f39a18a9b70, op: WAKE|PRIVATE_FLAG, val: 1) = 0
At some point we'll get that poll sys_exit event and will print a "[continued]" line.
While making the sizing of the alignment after the syscall arg list and
its result configurable, so that we can mimic strace, which uses a
smaller alingment by default, a bug was introduced where the closing
parens appeared before the syscall name and its arg list, fix it.
Fixes: 4b8a240ed5 ("perf trace: Add alignment spaces after the closing parens")
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/n/tip-oi45i54s59h1w1kmgpzrfuum@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
So that beautifiers can access things like dev_maj.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/n/tip-wm5o51f206c5pi063dsaeraq@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
We keep a table for the fds to map them back to pathnames when showing
'fd' based APIs such as write(), store as well the major number for the
device the path is in, to use in things like choosing the right ioctl
'cmd' beautifier.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/n/tip-qjkds7bnk7v7fk2xhqsb0a4v@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
So that we can have that table expanded when setting other attributes.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/n/tip-hzvpe3qwafe6sqcq3bhtbxds@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
So that we can add more per file attributes besides the pathname, such
as which ioctl beautifier to use, for cases such as the sound and
usbdeffs ioctls, that both use the 'U' command, so we have to
differentiate at the major number for the device file.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/n/tip-1895cmhrdz2dkl5prf2cj2yj@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
We shouldn't hardcode the size of the tracepoint common_ fields, use the
offset of the 'id'/'__syscallnr' field in the sys_enter event instead.
This caused the augmented syscalls code to fail on a particular build of a
PREEMPT_RT_FULL kernel where these extra 'common_migrate_disable' and
'common_padding' fields were before the syscall id one:
# cat /sys/kernel/debug/tracing/events/raw_syscalls/sys_enter/format
name: sys_enter
ID: 22
format:
field:unsigned short common_type; offset:0; size:2; signed:0;
field:unsigned char common_flags; offset:2; size:1; signed:0;
field:unsigned char common_preempt_count; offset:3; size:1; signed:0;
field:int common_pid; offset:4; size:4; signed:1;
field:unsigned short common_migrate_disable; offset:8; size:2; signed:0;
field:unsigned short common_padding; offset:10; size:2; signed:0;
field:long id; offset:16; size:8; signed:1;
field:unsigned long args[6]; offset:24; size:48; signed:0;
print fmt: "NR %ld (%lx, %lx, %lx, %lx, %lx, %lx)", REC->id, REC->args[0], REC->args[1], REC->args[2], REC->args[3], REC->args[4], REC->args[5]
#
All those 'common_' prefixed fields are zeroed when they hit a BPF tracepoint
hook, we better just discard those, i.e. somehow pass an offset to the
BPF program from the start of the ctx and make adjustments in the 'perf trace'
handlers to adjust the offset of the syscall arg offsets obtained from tracefs.
Till then, fix it the quick way and add this to the augmented_raw_syscalls.c to
bet it to work in such kernels:
diff --git a/tools/perf/examples/bpf/augmented_raw_syscalls.c b/tools/perf/examples/bpf/augmented_raw_syscalls.c
index 53c233370fae..1f746f931e13 100644
--- a/tools/perf/examples/bpf/augmented_raw_syscalls.c
+++ b/tools/perf/examples/bpf/augmented_raw_syscalls.c
@@ -38,12 +38,14 @@ struct bpf_map SEC("maps") syscalls = {
struct syscall_enter_args {
unsigned long long common_tp_fields;
+ long rt_common_tp_fields;
long syscall_nr;
unsigned long args[6];
};
struct syscall_exit_args {
unsigned long long common_tp_fields;
+ long rt_common_tp_fields;
long syscall_nr;
long ret;
};
Just to check that this was the case. Fix it properly later, for now remove the
hardcoding of the offset in the 'perf trace' side and document the situation
with this patch.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/n/tip-2pqavrktqkliu5b9nzouio21@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
While updating 'perf trace' on an machine with an old precompiled
augmented_raw_syscalls.o that didn't setup the syscall map the new 'perf
trace' codebase notices the augmented_raw_syscalls.o eBPF event, decides
to use it instead of the old raw_syscalls:sys_{enter,exit} method, but
then because we don't have the syscall map tries to set the tracepoint
filter on the sys_{enter,exit} evsels, that are NULL, segfaulting.
Make the code more robust by checking it those tracepoints have
their respective evsels in place before trying to set the tp filter.
With this we still get everything to work, just not setting up the
syscall filters, which is better than a segfault. Now to update the
precompiled augmented_raw_syscalls.o and continue development :-)
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/n/tip-3ft5rjdl05wgz2pwpx2z8btu@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Also to make it match 'strace' output, for regression testing.
Both now produce this option, when 'perf trace' uses a .perfconfig
asking for the strace like output:
mmap(0x7faf66e6a000, 1363968, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x22000) = 0x7faf66e6a000
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/n/tip-27qhouo1kaac2iyl85nfnsf5@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This actually so far, AFAIK is available only in x86, so the code was
put in place with x86 prefixes, in arches where it is not available it
will just not be called, so no further mechanisms are needed at this
time.
Later, when other arches wire this up, we'll just look at the uname
(live sessions) or perf_env data in the perf.data header to auto-wire
the right beautifier.
With this the output is the same as produced by 'strace' when used with
the following ~/.perfconfig:
# cat ~/.perfconfig
[llvm]
dump-obj = true
[trace]
add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.o
show_zeros = yes
show_duration = no
no_inherit = yes
show_timestamp = no
show_arg_names = no
args_alignment = -40
show_prefix = yes
#
And, on fedora 29, since the string tables are generated from the kernel
sources, we don't know about 0x3001, just like strace:
--- /tmp/strace 2018-12-17 11:22:08.707586721 -0300
+++ /tmp/trace 2018-12-18 11:11:32.037512729 -0300
@@ -1,49 +1,49 @@
-arch_prctl(0x3001 /* ARCH_??? */, 0x7ffc8a92dc80) = -1 EINVAL (Invalid argument)
+arch_prctl(0x3001 /* ARCH_??? */, 0x7ffe4eb93ae0) = -1 EINVAL (Invalid argument)
-arch_prctl(ARCH_SET_FS, 0x7faf6700f540) = 0
+arch_prctl(ARCH_SET_FS, 0x7fb507364540) = 0
And that seems to be related to the CET/Shadow Stack feature, that
userland in Fedora 29 (glibc 2.28) are querying the kernel about, that
0x3001 seems to be ARCH_CET_STATUS, I'll check the situation and test
with a fedora 29 kernel to see if the other codes are used.
A diff that ignores the different pointers for different runs needs to
be put in place in the upcoming regression tests comparing 'perf trace's
output to strace's.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/n/tip-73a9prs8ktkrt97trtdmdjs8@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>