- 25 Jan, 2018 9 commits
-
-
Andi Kleen authored
Signed-off-by:
Andi Kleen <ak@linux.intel.com> Link: https://lkml.kernel.org/r/20180118234518.GA27753@tassilo.jf.intel.comSigned-off-by:
Arnaldo Carvalho de Melo <acme@redhat.com>
-
Andi Kleen authored
Signed-off-by:
-
Andi Kleen authored
Signed-off-by:
-
Andi Kleen authored
Signed-off-by:
-
Andi Kleen authored
Signed-off-by:
-
Andi Kleen authored
Signed-off-by:
-
Andi Kleen authored
Signed-off-by:
-
Andi Kleen authored
Signed-off-by:
-
Andi Kleen authored
Signed-off-by:
-
- 23 Jan, 2018 9 commits
-
-
Hendrik Brueckner authored
Change the Makefile and build process to no longer require audit-libs interfaces when the architecture provides system call tables. Committer notes: Its not enough to hook into the NO_LIBAUDIT makefile block, we need to define a CONFIG_TRACE that gets selected by both architectures generating the syscall tables from the kernel headers and from detecting the availability of libaudit. With that in place we will not link against libaudit even if the necessary files are available for that, in fact we will not even try to detect its availability, speeding up a bit the feature detection phase. Signed-off-by:
Hendrik Brueckner <brueckner@linux.vnet.ibm.com> Reviewed-by:
Thomas Richter <tmricht@linux.vnet.ibm.com> Tested-by:
-
Hendrik Brueckner authored
Replace the errno_to_name() from the audit-libs with the newly introduced arch_syscalls__strerrno() function. With this change: 1. With replacing errno_to_name() from audit-libs, perf trace does no longer require audit-lib interfaces. 2. In addition to 1, the audit-libs dependency can be removed for architectures that support syscall tables in perf. This is achieved in a follow-up commit. 3. With the architecture specific errno number/name mapping, perf trace reports can work across architectures. Signed-off-by: -
Hendrik Brueckner authored
Introduce a script that generates a mapping of errno numbers to their names for each architecture that is supported by perf (i.e. has a subdirectory in tools/perf/arch/). The errno mapping is generated as part of the trace beautifiers and can be used by including the trace/beauty/arch_errno_names.c file. Then, use arch_syscalls__strerrno() to look up an errno value to obtain the errno name (e.g. ENOENT) for a particular architecture. Signed-off-by:
-
Hendrik Brueckner authored
This is a pre-req to generate an architecture specific mapping of errno numbers to their names. This errno mapping can be used by perf trace to support cross-architecture trace reports and to get rid of the audit-libs dependency. Signed-off-by:
-
Hendrik Brueckner authored
For each arch in tools/perf/arch, grab a copy of errno.h. This is a pre-req to generate an architecture specific mapping of errno numbers to their names. This errno mapping can be used by perf trace to support cross-architecture trace reports and to get rid of the audit-libs dependency. Signed-off-by:
-
Jiri Olsa authored
Display the state of the rest of the features (FEATURE_TESTS_EXTRA) on a 'make VF=1' build. These features are detected manually by perf's Makefile.config so they can't be displayed with the main list, but only after we're done in Makefile.config. $ make VF=1 BUILD: Doing 'make -j4' parallel build Auto-detecting system features: ... dwarf: [ on ] ... dwarf_getlocations: [ on ] ... glibc: [ on ] ... gtk2: [ on ] SNIP ... timerfd: [ on ] ... sched_getcpu: [ on ] ... sdt: [ on ] ... setns: [ on ] extra features: ... bionic: [ OFF ] ... compile-32: [ on ] ... compile-x32: [ OFF ] ... cplus-demangle: [ on ] ... hello: [ OFF ] ... libbabeltrace: [ on ] ... liberty: [ on ] ... liberty-z: [ on ] ... libunwind-debug-frame: [ OFF ] ... libunwind-debug-frame-arm: [ OFF ] ... libunwind-debug-frame-aarch64: [ OFF ] SNIP Signed-off-by:Jiri Olsa <jolsa@kernel.org> Tested-by:
-
Mathieu Poirier authored
Commit (93d10af2 perf tools: Optimize sample parsing for ordered events) breaks intelPT trace decoding by invariably returning an error if the event type isn't a PERF_SAMPLE_TIME. With this patch the timestamp is initialised and processing is allowed to continue if the error returned by function perf_evlist__parse_sample_timestamp() is not a fault. Signed-off-by:
Mathieu Poirier <mathieu.poirier@linaro.org> Acked-by:
Adrian Hunter <adrian.hunter@intel.com> Acked-by:
Jiri Olsa <jolsa@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Fixes: 93d10af2 ("perf tools: Optimize sample parsing for ordered events") Link: http://lkml.kernel.org/r/1515616312-27645-1-git-send-email-mathieu.poirier@linaro.orgSigned-off-by:
-
Wang YanQing authored
I've meet a strange behavior with these commands on my gentoo box: 1: perf kmem record 2: CTRL-C to stop 1 3: perf report 4: "Enter", "Enter", "Run scripts for all samples", "event_analyzing_sample". Then 'perf report' says: " No kallsyms or vmlinux with build-id xxxx was found /lib/modules/4.10.0+/build/vmlinux with build id xxxx not found, continuing without symbols ". It is strange because I am sure /lib/modules/4.10.0+/build/vmlinux is right for perf.data. After digging, I found out the reason is that "perf report" generates many open fds, then "script_browse" uses popen to run "perf script" which run out of open files. The gentoo box has a small default value for "max open files", 1024. Yes, "ulimit -n " with a bigger number could fix it, but I think that using O_CLOEXEC in do_open is a better way. Signed-off-by:
Wang YanQing <udknight@gmail.com> Acked-by:
-
Arnaldo Carvalho de Melo authored
To be more generally available and get the build on centos:5 to work after we use O_CLOEXEC in the next patch, in the util/dso.c file. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Cc: Wang YanQing <udknight@gmail.com> Link: https://lkml.kernel.org/n/tip-vsjbiydh15pfqomxw1kx64ex@git.kernel.orgSigned-off-by:
-
- 18 Jan, 2018 5 commits
-
-
Arnaldo Carvalho de Melo authored
When clang is not linked with 'perf' we should just add a debug message about that before doing the fallback to calling the external compiler. I.e. just the "-95" warning below gets turned into a debug message: # cat sys_enter_open.c #include "bpf.h" SEC("syscalls:sys_enter_open") int func(void *ctx) { struct { char *ptr; char path[256]; } filename = { .ptr = *((char **)(ctx + 16)), }; int len = bpf_probe_read_str(filename.path, sizeof(filename.path), filename.ptr); if (len > 0) { if (len == 1) perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, &filename, len + sizeof(filename.ptr)); else if (len < 256) perf_event_output(ctx, &__bpf_stdout__, BPF_F_CURRENT_CPU, &filename, len + sizeof(filename.ptr)); } return 0; } # trace -e open,sys_enter_open.c bpf: builtin compilation failed: -95, try external compiler 0.000 ( ): __bpf_stdout__:@......./proc/self/task/11160/comm..) 0.014 ( 0.116 ms): qemu-system-x8/6721 open(filename: /proc/self/task/11160/comm, flags: RDWR) = 91 2335.411 ( ): __bpf_stdout__:FB..~.../etc/resolv.conf....) 2335.421 ( 0.030 ms): chronyd/883 open(filename: /etc/resolv.conf, flags: CLOEXEC) = 5 ^C# Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-z5aak9oay448ffj37giz94yr@git.kernel.orgSigned-off-by: -
Arnaldo Carvalho de Melo authored
So that we can get it working for TUI, where using just pr_err() would end up making the message emitted to stderr to be erased by the TUI exit routine restoring the terminal to its previous state. Now we can see that trying to use a tracepoint field as one of the --field entries isn't working: # perf top --stdio --no-children -e syscalls:sys_enter_write --fields pid,sym,count Error: Unknown --fields key: `count' Usage: perf top [<options>] --fields <key[,keys...]> output field(s): overhead, period, sample plus all of sort keys # Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-usy9hhy7umdd4bbblkn63t8w@git.kernel.orgSigned-off-by: -
Adrian Hunter authored
There is never a need to synthesize a 'swapped' sample, so all callers to perf_event__synthesize_sample() pass 'false' as the value to 'swapped'. So get rid of the unused 'swapped' parameter. Signed-off-by:
-
Adrian Hunter authored
PERF_SAMPLE_CPU contains the cpu number in the first 4 bytes and the second 4 bytes are reserved. Ensure the reserved bytes are zero in perf_event__synthesize_sample(). Signed-off-by:
-
Adrian Hunter authored
Both 'perf inject' and internal tools consume cpu endian samples, so there is never a need to do any swapping when synthesizing samples. Signed-off-by:
-
- 17 Jan, 2018 17 commits
-
-
Ingo Molnar authored
Merge tag 'perf-core-for-mingo-4.16-20180117' 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: - Fix various per event 'max-stack' and 'call-graph=dwarf' issues, mostly in 'perf trace', allowing to use 'perf trace --call-graph' with 'dwarf' and 'fp' to setup the callgraph details for the syscall events and make that apply to other events, whilhe allowing to override that on a per-event basis, using '-e sched:*switch/call-graph=dwarf/' for instance (Arnaldo Carvalho de Melo) - Improve the --time percent support in record/report/script (Jin Yao) - Fix copyfile_offset update of output offset (Jiri Olsa) - Add python script to profile and resolve physical mem type (Kan Liang) - Add ARM Statistical Profiling Extensions (SPE) support (Kim Phillips) - Remove trailing semicolon in the evlist code (Luis de Bethencourt) - Fix incorrect handling of type _TERM_DRV_CFG (Mathieu Poirier) - Use asprintf when possible in libtraceevent (Federico Vaga) - Fix bad force_token escape sequence in libtraceevent (Michael Sartain) - Add UL suffix to MISSING_EVENTS in libtraceevent (Michael Sartain) - value of unknown symbolic fields in libtraceevent (Jan Kiszka) - libtraceevent updates: (Steven Rostedt) o Show value of flags that have not been parsed o Simplify pointer print logic and fix %pF o Handle new pointer processing of bprint strings o Show contents (in hex) of data of unrecognized type records o Fix get_field_str() for dynamic strings - Add missing break in FALSE case of pevent_filter_clear_trivial() (Taeung Song) - Fix failed memory allocation for get_cpuid_str (Thomas Richter) Signed-off-by:
Ingo Molnar <mingo@kernel.org>
-
Ingo Molnar authored
Signed-off-by: -
Thomas Richter authored
In x86 architecture dependend part function get_cpuid_str() mallocs a 128 byte buffer, but does not check if the memory allocation succeeded or not. When the memory allocation fails, function __get_cpuid() is called with first parameter being a NULL pointer. However this function references its first parameter and operates on a NULL pointer which might cause core dumps. Signed-off-by:
-
Jin Yao authored
Previously it was only allowed to use at most 10 time slices in 'perf script --time'. This patch removes this limitation. For example, following command line is OK (12 time slices) perf script --time 1%/1,1%/2,1%/3,1%/4,1%/5,1%/6,1%/7,1%/8,1%/9,1%/10,1%/11,1%/12 Signed-off-by:
Jin Yao <yao.jin@linux.intel.com> Suggested-by:
-
Jin Yao authored
Previously it was only allowed to use at most 10 time slices in 'perf report --time'. This patch removes this limitation. For example, following command line is OK (12 time slices) perf report --stdio --time 1%/1,1%/2,1%/3,1%/4,1%/5,1%/6,1%/7,1%/8,1%/9,1%/10,1%/11,1%/12 Signed-off-by:
-
Jin Yao authored
Previously we use a magic number 10 to limit the number of time slices. It's not very good. This patch creates a new function perf_time__range_alloc() to allocate time slices buffer. The number of buffer entries is determined by the number of comma in string but at least it will allocate one entry even if no comma is found. Signed-off-by:
-
Jin Yao authored
Add a time slices indication to the perf report header. For example, # perf report --stdio --time 10% # Total Lost Samples: 0 # # Samples: 9K of event 'cycles:ppp' (time slices: 10%) # Event count (approx.): 8951288803 Signed-off-by:
-
Jin Yao authored
Previously, the time percent slice needs an index to specify which one the user wants. It may be easier to use if the index can be omitted. So with this patch, for example, perf report --stdio --time 10%/1 should be equivalent to perf report --stdio --time 10% Signed-off-by:
-
Jin Yao authored
The command line like 'perf report --stdio --time 1abc%/1' could be accepted by perf. It looks not very good. This patch uses strtod() to replace original atof() and check the entire string. Now for the same command line, it would return error message "Invalid time string". root@skl:/tmp# perf report --stdio --time 1abc%/1 Invalid time string Signed-off-by:
-
Jin Yao authored
The following message will be returned to user when executing 'perf script --time' if perf data file doesn't contain the first/last sample time. "HINT: no first/last sample time found in perf data. Please use latest perf binary to execute 'perf record' (if '--buildid-all' is enabled, needs to set '--timestamp-boundary')." Signed-off-by:
-
Jin Yao authored
The following message will be returned to user when executing 'perf report --time' if perf data file doesn't contain the first/last sample time. "HINT: no first/last sample time found in perf data. Please use latest perf binary to execute 'perf record' (if '--buildid-all' is enabled, needs to set '--timestamp-boundary')." Signed-off-by:
-
Arnaldo Carvalho de Melo authored
When we use a global DWARF setting as in: perf record --call-graph dwarf According to 5c0cf224 ("perf record: Store data mmaps for dwarf unwind") we need to set up some extra perf_event_attr bits. But when we instead do a per event dwarf setting: perf record -e cycles/call-graph=dwarf/ This was not being done, make them equivalent. This didn't produce any output changes in my tests while fixing up loose ends in the per-event settings, I found it just by comparing the perf_event_attr fields trying to find an explanation for those problems. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Hendrick Brueckner <brueckner@linux.vnet.ibm.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Noel Grandin <noelgrandin@gmail.com> Cc: Thomas Richter <tmricht@linux.vnet.ibm.com> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-6476r53h2o38skbs9qa4ust4@git.kernel.orgSigned-off-by:
-
Arnaldo Carvalho de Melo authored
The per-event max-stack setting wasn't overriding the global --max-stack setting: # perf trace --no-syscalls --max-stack 4 -e probe_libc:inet_pton/call-graph=dwarf,max-stack=2/ ping -6 -c 1 ::1 PING ::1(::1) 56 data bytes 64 bytes from ::1: icmp_seq=1 ttl=64 time=0.072 ms --- ::1 ping statistics --- 1 packets transmitted, 1 received, 0% packet loss, time 0ms rtt min/avg/max/mdev = 0.072/0.072/0.072/0.000 ms 0.000 probe_libc:inet_pton:(7feb7a998350)) __inet_pton (inlined) gaih_inet.constprop.7 (/usr/lib64/libc-2.26.so) __GI_getaddrinfo (inlined) [0xffffaa39b6108f3f] (/usr/bin/ping) # Fix it: # perf trace --no-syscalls --max-stack 4 -e probe_libc:inet_pton/call-graph=dwarf,max-stack=2/ ping -6 -c 1 ::1 PING ::1(::1) 56 data bytes 64 bytes from ::1: icmp_seq=1 ttl=64 time=0.073 ms --- ::1 ping statistics --- 1 packets transmitted, 1 received, 0% packet loss, time 0ms rtt min/avg/max/mdev = 0.073/0.073/0.073/0.000 ms 0.000 probe_libc:inet_pton:(7f1083221350)) __inet_pton (inlined) gaih_inet.constprop.7 (/usr/lib64/libc-2.26.so) # Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Hendrick Brueckner <brueckner@linux.vnet.ibm.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Thomas Richter <tmricht@linux.vnet.ibm.com> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-ic3g837xg8ob3kcpkspxwz0g@git.kernel.orgSigned-off-by: -
Arnaldo Carvalho de Melo authored
If we use: perf trace --max-stack=4 then the syscall events will use DWARF callchains, when available (libunwind enabled in the build) and the printing will stop at 4 levels. When we introduced support for tracepoint events this ended up not applying for them, fix it. Before: # perf trace --call-graph=dwarf --no-syscalls -e probe_libc:inet_pton ping -6 -c 1 ::1 PING ::1(::1) 56 data bytes 64 bytes from ::1: icmp_seq=1 ttl=64 time=0.058 ms --- ::1 ping statistics --- 1 packets transmitted, 1 received, 0% packet loss, time 0ms rtt min/avg/max/mdev = 0.058/0.058/0.058/0.000 ms 0.000 probe_libc:inet_pton:(7fc6c2a16350)) # After: # perf trace --call-graph=dwarf --no-syscalls -e probe_libc:inet_pton ping -6 -c 1 ::1 PING ::1(::1) 56 data bytes 64 bytes from ::1: icmp_seq=1 ttl=64 time=0.087 ms --- ::1 ping statistics --- 1 packets transmitted, 1 received, 0% packet loss, time 0ms rtt min/avg/max/mdev = 0.087/0.087/0.087/0.000 ms 0.000 probe_libc:inet_pton:(7fbf9a041350)) __inet_pton (inlined) gaih_inet.constprop.7 (/usr/lib64/libc-2.26.so) __GI_getaddrinfo (inlined) [0xffffaa947cb67f3f] (/usr/bin/ping) __libc_start_main (/usr/lib64/libc-2.26.so) [0xffffaa947cb68379] (/usr/bin/ping) # Reported-by: -
Arnaldo Carvalho de Melo authored
When setting up DWARF callchains on specific events, without using 'record' or 'trace' --call-graph, but instead doing it like: perf trace -e cycles/call-graph=dwarf/ The unwind__prepare_access() call in thread__insert_map() when we process PERF_RECORD_MMAP(2) metadata events were not being performed, precluding us from using per-event DWARF callchains, handling them just when we asked for all events to be DWARF, using "--call-graph dwarf". We do it in the PERF_RECORD_MMAP because we have to look at one of the executable maps to figure out the executable type (64-bit, 32-bit) of the DSO laid out in that mmap. Also to look at the architecture where the perf.data file was recorded. All this probably should be deferred to when we process a sample for some thread that has callchains, so that we do this processing only for the threads with samples, not for all of them. For now, fix using DWARF on specific events. Before: # perf trace --no-syscalls -e probe_libc:inet_pton/call-graph=dwarf/ ping -6 -c 1 ::1 PING ::1(::1) 56 data bytes 64 bytes from ::1: icmp_seq=1 ttl=64 time=0.048 ms --- ::1 ping statistics --- 1 packets transmitted, 1 received, 0% packet loss, time 0ms rtt min/avg/max/mdev = 0.048/0.048/0.048/0.000 ms 0.000 probe_libc:inet_pton:(7fe9597bb350)) Problem processing probe_libc:inet_pton callchain, skipping... # After: # perf trace --no-syscalls -e probe_libc:inet_pton/call-graph=dwarf/ ping -6 -c 1 ::1 PING ::1(::1) 56 data bytes 64 bytes from ::1: icmp_seq=1 ttl=64 time=0.060 ms --- ::1 ping statistics --- 1 packets transmitted, 1 received, 0% packet loss, time 0ms rtt min/avg/max/mdev = 0.060/0.060/0.060/0.000 ms 0.000 probe_libc:inet_pton:(7fd4aa930350)) __inet_pton (inlined) gaih_inet.constprop.7 (/usr/lib64/libc-2.26.so) __GI_getaddrinfo (inlined) [0xffffaa804e51af3f] (/usr/bin/ping) __libc_start_main (/usr/lib64/libc-2.26.so) [0xffffaa804e51b379] (/usr/bin/ping) # # perf trace --call-graph=dwarf --no-syscalls -e probe_libc:inet_pton/call-graph=dwarf/ ping -6 -c 1 ::1 PING ::1(::1) 56 data bytes 64 bytes from ::1: icmp_seq=1 ttl=64 time=0.057 ms --- ::1 ping statistics --- 1 packets transmitted, 1 received, 0% packet loss, time 0ms rtt min/avg/max/mdev = 0.057/0.057/0.057/0.000 ms 0.000 probe_libc:inet_pton:(7f9363b9e350)) __inet_pton (inlined) gaih_inet.constprop.7 (/usr/lib64/libc-2.26.so) __GI_getaddrinfo (inlined) [0xffffa9e8a14e0f3f] (/usr/bin/ping) __libc_start_main (/usr/lib64/libc-2.26.so) [0xffffa9e8a14e1379] (/usr/bin/ping) # # perf trace --call-graph=fp --no-syscalls -e probe_libc:inet_pton/call-graph=dwarf/ ping -6 -c 1 ::1 PING ::1(::1) 56 data bytes 64 bytes from ::1: icmp_seq=1 ttl=64 time=0.077 ms --- ::1 ping statistics --- 1 packets transmitted, 1 received, 0% packet loss, time 0ms rtt min/avg/max/mdev = 0.077/0.077/0.077/0.000 ms 0.000 probe_libc:inet_pton:(7f4947e1c350)) __inet_pton (inlined) gaih_inet.constprop.7 (/usr/lib64/libc-2.26.so) __GI_getaddrinfo (inlined) [0xffffaa716d88ef3f] (/usr/bin/ping) __libc_start_main (/usr/lib64/libc-2.26.so) [0xffffaa716d88f379] (/usr/bin/ping) # # perf trace --no-syscalls -e probe_libc:inet_pton/call-graph=fp/ ping -6 -c 1 ::1 PING ::1(::1) 56 data bytes 64 bytes from ::1: icmp_seq=1 ttl=64 time=0.078 ms --- ::1 ping statistics --- 1 packets transmitted, 1 received, 0% packet loss, time 0ms rtt min/avg/max/mdev = 0.078/0.078/0.078/0.000 ms 0.000 probe_libc:inet_pton:(7fa157696350)) __GI___inet_pton (/usr/lib64/libc-2.26.so) getaddrinfo (/usr/lib64/libc-2.26.so) [0xffffa9ba39c74f40] (/usr/bin/ping) # Acked-by:Namhyung Kim <namhyung@kernel.org> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Hendrick Brueckner <brueckner@linux.vnet.ibm.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Thomas Richter <tmricht@linux.vnet.ibm.com> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/r/20180116182650.GE16107@kernel.orgSigned-off-by:
-
Arnaldo Carvalho de Melo authored
When setting the "dwarf" unwinder for a specific event and not specifying the max-stack, the attr.sample_max_stack ended up using an uninitialized callchain_param.max_stack, fix it by using designated initializers for that callchain_param variable, zeroing all non explicitely initialized struct members. Here is what happened: # perf trace -vv --no-syscalls --max-stack 4 -e probe_libc:inet_pton/call-graph=dwarf/ ping -6 -c 1 ::1 callchain: type DWARF callchain: stack dump size 8192 perf_event_attr: type 2 size 112 config 0x730 { sample_period, sample_freq } 1 sample_type IP|TID|TIME|ADDR|CALLCHAIN|CPU|PERIOD|RAW|REGS_USER|STACK_USER|DATA_SRC exclude_callchain_user 1 { wakeup_events, wakeup_watermark } 1 sample_regs_user 0xff0fff sample_stack_user 8192 sample_max_stack 50656 sys_perf_event_open failed, error -75 Value too large for defined data type # perf trace -vv --no-syscalls --max-stack 4 -e probe_libc:inet_pton/call-graph=dwarf/ ping -6 -c 1 ::1 callchain: type DWARF callchain: stack dump size 8192 perf_event_attr: type 2 size 112 config 0x730 sample_type IP|TID|TIME|ADDR|CALLCHAIN|CPU|PERIOD|RAW|REGS_USER|STACK_USER|DATA_SRC exclude_callchain_user 1 sample_regs_user 0xff0fff sample_stack_user 8192 sample_max_stack 30448 sys_perf_event_open failed, error -75 Value too large for defined data type # Now the attr.sample_max_stack is set to zero and the above works as expected: # perf trace --no-syscalls --max-stack 4 -e probe_libc:inet_pton/call-graph=dwarf/ ping -6 -c 1 ::1 PING ::1(::1) 56 data bytes 64 bytes from ::1: icmp_seq=1 ttl=64 time=0.072 ms --- ::1 ping statistics --- 1 packets transmitted, 1 received, 0% packet loss, time 0ms rtt min/avg/max/mdev = 0.072/0.072/0.072/0.000 ms 0.000 probe_libc:inet_pton:(7feb7a998350)) __inet_pton (inlined) gaih_inet.constprop.7 (/usr/lib64/libc-2.26.so) __GI_getaddrinfo (inlined) [0xffffaa39b6108f3f] (/usr/bin/ping) # Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Hendrick Brueckner <brueckner@linux.vnet.ibm.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Thomas Richter <tmricht@linux.vnet.ibm.com> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-is9tramondqa9jlxxsgcm9iz@git.kernel.orgSigned-off-by: -
Kim Phillips authored
'perf record' and 'perf report --dump-raw-trace' supported in this release. Example usage: # perf record -e arm_spe/ts_enable=1,pa_enable=1/ dd if=/dev/zero of=/dev/null count=10000 # perf report --dump-raw-trace Note that the perf.data file is portable, so the report can be run on another architecture host if necessary. Output will contain raw SPE data and its textual representation, such as: 0x5c8 [0x30]: PERF_RECORD_AUXTRACE size: 0x200000 offset: 0 ref: 0x1891ad0e idx: 1 tid: 2227 cpu: 1 . . ... ARM SPE data: size 2097152 bytes . 00000000: 49 00 LD . 00000002: b2 c0 3b 29 0f 00 00 ff ff VA 0xffff00000f293bc0 . 0000000b: b3 c0 eb 24 fb 00 00 00 80 PA 0xfb24ebc0 ns=1 . 00000014: 9a 00 00 LAT 0 XLAT . 00000017: 42 16 EV RETIRED L1D-ACCESS TLB-ACCESS . 00000019: b0 00 c4 15 08 00 00 ff ff PC 0xff00000815c400 el3 ns=1 . 00000022: 98 00 00 LAT 0 TOT . 00000025: 71 36 6c 21 2c 09 00 00 00 TS 39395093558 . 0000002e: 49 00 LD . 00000030: b2 80 3c 29 0f 00 00 ff ff VA 0xffff00000f293c80 . 00000039: b3 80 ec 24 fb 00 00 00 80 PA 0xfb24ec80 ns=1 . 00000042: 9a 00 00 LAT 0 XLAT . 00000045: 42 16 EV RETIRED L1D-ACCESS TLB-ACCESS . 00000047: b0 f4 11 16 08 00 00 ff ff PC 0xff0000081611f4 el3 ns=1 . 00000050: 98 00 00 LAT 0 TOT . 00000053: 71 36 6c 21 2c 09 00 00 00 TS 39395093558 . 0000005c: 48 00 INSN-OTHER . 0000005e: 42 02 EV RETIRED . 00000060: b0 2c ef 7f 08 00 00 ff ff PC 0xff0000087fef2c el3 ns=1 . 00000069: 98 00 00 LAT 0 TOT . 0000006c: 71 d1 6f 21 2c 09 00 00 00 TS 39395094481 ... Other release notes: - applies to acme's perf/{core,urgent} branches, likely elsewhere - Report is self-contained within the tool. Record requires enabling the kernel SPE driver by setting CONFIG_ARM_SPE_PMU. - The intel-bts implementation was used as a starting point; its min/default/max buffer sizes and power of 2 pages granularity need to be revisited for ARM SPE - Recording across multiple SPE clusters/domains not supported - Snapshot support (record -S), and conversion to native perf events (e.g., via 'perf inject --itrace'), are also not supported - Technically both cs-etm and spe can be used simultaneously, however disabled for simplicity in this release Signed-off-by:Kim Phillips <kim.phillips@arm.com> Reviewed-by:
Dongjiu Geng <gengdongjiu@huawei.com> Acked-by:
-
