Currently a number of Makefiles break when used with toolchains that
pass extra flags in CC and other cross-compile related variables (such
as --sysroot).
Thus we get this error when we use a toolchain that puts --sysroot in
the CC var:
~/src/linux/tools$ make iio
[snip]
iio_event_monitor.c:18:10: fatal error: unistd.h: No such file or directory
#include <unistd.h>
^~~~~~~~~~
This occurs because we clobber several env vars related to
cross-compiling with lines like this:
CC = $(CROSS_COMPILE)gcc
Although this will point to a valid cross-compiler, we lose any extra
flags that might exist in the CC variable, which can break toolchains
that rely on them (for example, those that use --sysroot).
This easily shows up using a Yocto SDK:
$ . [snip]/sdk/environment-setup-cortexa8hf-neon-poky-linux-gnueabi
$ echo $CC
arm-poky-linux-gnueabi-gcc -march=armv7-a -mfpu=neon -mfloat-abi=hard
-mcpu=cortex-a8
--sysroot=[snip]/sdk/sysroots/cortexa8hf-neon-poky-linux-gnueabi
$ echo $CROSS_COMPILE
arm-poky-linux-gnueabi-
$ echo ${CROSS_COMPILE}gcc
krm-poky-linux-gnueabi-gcc
Although arm-poky-linux-gnueabi-gcc is a cross-compiler, we've lost the
--sysroot and other flags that enable us to find the right libraries to
link against, so we can't find unistd.h and other libraries and headers.
Normally with the --sysroot flag we would find unistd.h in the sdk
directory in the sysroot:
$ find [snip]/sdk/sysroots -path '*/usr/include/unistd.h'
[snip]/sdk/sysroots/cortexa8hf-neon-poky-linux-gnueabi/usr/include/unistd.h
The perf Makefile adds CC = $(CROSS_COMPILE)gcc if and only if CC is not
already set, and it compiles correctly with the above toolchain.
So, generalize the logic that perf uses in the common Makefile and
remove the manual CC = $(CROSS_COMPILE)gcc lines from each Makefile.
Note that this patch does not fix cross-compile for all the tools (some
have other bugs), but it does fix it for all except usb and acpi, which
still have other unrelated issues.
I tested both with and without the patch on native and cross-build and
there appear to be no regressions.
Link: http://lkml.kernel.org/r/20180107214028.23771-1-martin@martingkelly.com
Signed-off-by: Martin Kelly <martin@martingkelly.com>
Acked-by: Mark Brown <broonie@kernel.org>
Cc: Tejun Heo <tj@kernel.org>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Linus Walleij <linus.walleij@linaro.org>
Cc: "K. Y. Srinivasan" <kys@microsoft.com>
Cc: Haiyang Zhang <haiyangz@microsoft.com>
Cc: Stephen Hemminger <sthemmin@microsoft.com>
Cc: Jonathan Cameron <jic23@kernel.org>
Cc: Pali Rohar <pali.rohar@gmail.com>
Cc: Richard Purdie <rpurdie@rpsys.net>
Cc: Jacek Anaszewski <jacek.anaszewski@gmail.com>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Robert Moore <robert.moore@intel.com>
Cc: Lv Zheng <lv.zheng@intel.com>
Cc: "Rafael J. Wysocki" <rafael.j.wysocki@intel.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Valentina Manea <valentina.manea.m@gmail.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Mario Limonciello <mario.limonciello@dell.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
- add -cgskip option to reduce callgraph output size
- add -cgfilter option to focus on a list of devices
- add -result option for exporting batch test results
- removed all phoronix hooks, use -result to enable batch testing
- change -usbtopo to -devinfo, now prints all devices
- add -gzip option to read/write logs in gz format
- add -bufsize option to manually control ftrace buffer size
- add -sync option to run filesystem sync prior to test
- add -display option to enable/disable the display prior to test
- add -rs option to enable/disable runtime suspend on all devices for test
- add installed config files to search path
- add kernel error/warning links into the timeline
- fix callgraph trace to better handle interrupts
- include command string and kernel params in timeline output header
Signed-off-by: Todd Brandt <todd.e.brandt@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
- add -cgskip option to reduce callgraph output size
- add -cgfilter option to focus on a list of devices
- add -result option for exporting batch test results
- removed all phoronix hooks, use -result to enable batch testing
- changed argument -f to match sleegraph, -f = -callgraph
- use -fstat for function status instead of -f
- add -verbose option to print out timeline stats and kernel options
- include command string and kernel params in timeline output header
Signed-off-by: Todd Brandt <todd.e.brandt@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
- name change: analyze_boot.py to bootgraph.py
- name change: analyze_suspend.py to sleepgraph.py
- added config files for easier sleepgraph usage
- added example.cfg which describes all config options
- added cgskip.txt definition for slimmer callgraphs
Signed-off-by: Todd Brandt <todd.e.brandt@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Daniel Borkmann says:
====================
pull-request: bpf 2018-02-20
The following pull-request contains BPF updates for your *net* tree.
The main changes are:
1) Fix a memory leak in LPM trie's map_free() callback function, where
the trie structure itself was not freed since initial implementation.
Also a synchronize_rcu() was needed in order to wait for outstanding
programs accessing the trie to complete, from Yonghong.
2) Fix sock_map_alloc()'s error path in order to correctly propagate
the -EINVAL error in case of too large allocation requests. This
was just recently introduced when fixing close hooks via ULP layer,
fix from Eric.
3) Do not use GFP_ATOMIC in __cpu_map_entry_alloc(). Reason is that this
will not work with the recent __ptr_ring_init_queue_alloc() conversion
to kvmalloc_array(), where in case of fallback to vmalloc() that GFP
flag is invalid, from Jason.
4) Fix two recent syzkaller warnings: i) fix bpf_prog_array_copy_to_user()
when a prog query with a big number of ids was performed where we'd
otherwise trigger a warning from allocator side, ii) fix a missing
mlock precharge on arraymaps, from Daniel.
5) Two fixes for bpftool in order to avoid breaking JSON output when used
in batch mode, from Quentin.
6) Move a pr_debug() in libbpf in order to avoid having an otherwise
uninitialized variable in bpf_program__reloc_text(), from Jeremy.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
The default values for nreader and nwriter are apparently not all that
user-friendly, resulting in people doing scalability tests that ran all
runs at large scale. This commit therefore makes both the nreaders and
nwriters module default to the number of CPUs, and adds a comment to
rcuperf.c stating that the number of CPUs should be specified using the
nr_cpus kernel boot parameter. This commit also eliminates the redundant
rcuperf scripting specification of default values for these parameters.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
The rcuperf trace-event processing counted every "done" trace event
as a piggyback, which is incorrect because the task that started the
grace period didn't piggyback at all. This commit fixes this problem
by recording the task that started a given grace period and ignoring
that task's "done" record for that grace period.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
The rcuperf event-trace processing assumes that expedited grace periods
start and end on the same task, an assumption that was violated by moving
expedited grace-period processing to workqueues. This commit removes
this now-fallacious assumption from rcuperf's event-trace processing.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
The purpose of jitter is to expose concurrency bugs due to invalid
assumptions about forward progress. There is usually little point
in jitter when measuring performance. This commit therefore defaults
jitter off when running rcuperf. You can override this by specifying
the kvm.sh "--jitter" argument -after- the "--torture rcuperf"
argument. No idea why you would want this, but if you do, that is
how you do it.
One example of a conccurrency bug that this jitter might expose is one
in which the developer assumed that a given short region of code would be
guaranteed to execute within some short time limit. Such assumptions are
invalid in virtualized environments because the hupervisor can preempt
the guest OS at any point, even when the guest OS thinks that it has
disabled interrupts.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
The 512 megabyte memory size has served quite well, but more memory
is required when using large trace buffers on large systems. This
commit therefore adds a --memory argument to the kvm.sh script, which
allows the memory size to be specified on the command line, for example,
"--memory 768", --memory 800M", or "--memory 2G".
Reported-by: Sebastian Siewior <bigeasy@linutronix.de>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
This commit adds support of the qemu command qemu-system-aarch64
to rcutorture.
Signed-off-by: Lihao Liang <lianglihao@huawei.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
The kvm.sh header comment is a bit of a relic, so this commit brings
it up to date.
Reported-by: Lihao Liang <lianglihao@huawei.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Fixes a GCC maybe-uninitialized warning introduced by 48cca7e44f.
"text" is only initialized inside the if statement so only print debug
info there.
Fixes: 48cca7e44f ("libbpf: add support for bpf_call")
Signed-off-by: Jeremy Cline <jeremy@jcline.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Before this change, the '--graph-funcs', '--nograph-funcs' and
'--trace-funcs' options didn't work as expected when the <func> doesn't
exist. Because the kernel side hid possible errors.
$ sudo ./perf ftrace -a --graph-depth 1 --graph-funcs abcdefg
0) 0.140 us | rcu_all_qs();
3) 0.304 us | mutex_unlock();
0) 0.153 us | find_vma();
3) 0.088 us | __fsnotify_parent();
0) 6.145 us | handle_mm_fault();
3) 0.089 us | fsnotify();
3) 0.161 us | __sb_end_write();
3) 0.710 us | SyS_close();
3) 7.848 us | exit_to_usermode_loop();
On the example above, I specified the function filter 'abcdefg' but all
functions are enabled. The expected result is for all functions to be
filtered, since there is no such function ('abcdefg')
The original fix is to make the kernel support '\0' as end of string:
https://lkml.org/lkml/2018/1/16/116
But above fix cannot be compatible with old kernels. Then Namhyung Kim
suggest adding a space after function name.
This patch will append an '\n' when write tracing file. After this fix,
the perf will report correct error state. Also let it print an error if
reset_tracing_files() fails.
Committer testing:
Now it prints:
# perf ftrace -a --graph-depth 1 --graph-funcs abcdefg
failed to set tracing filters
#
And for an existing function:
# perf ftrace -a --graph-depth 1 --graph-funcs SyS_open
3) | SyS_open() {
3) ! 494.899 us | }
0) + 23.910 us | SyS_open();
1) + 17.115 us | SyS_open();
1) + 13.900 us | SyS_open();
------------------------------------------
3) qemu-sy-2817 => pickup-1290
------------------------------------------
3) + 20.021 us | SyS_open();
#
Signed-off-by: Changbin Du <changbin.du@intel.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1519007609-14551-1-git-send-email-changbin.du@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo:
- Fix wrong jump arrow in systems with branch records with cycles,
i.e. Intel's >= Skylake (Jin Yao)
- Fix 'perf record --per-thread' problem introduced when
implementing 'perf stat --per-thread (Jin Yao)
- Use arch__compare_symbol_names() to fix 'perf test vmlinux',
that was using strcmp(symbol names) while the dso routines
doing symbol lookups used the arch overridable one, making
this test fail in architectures that overrided that function
with something other than strcmp() (Jiri Olsa)
- Add 'perf script --show-round-event' to display
PERF_RECORD_FINISHED_ROUND entries (Jiri Olsa)
- Fix dwarf unwind for stripped binaries in 'perf test' (Jiri Olsa)
- Use ordered_events for 'perf report --tasks', otherwise we may get
artifacts when PERF_RECORD_FORK gets processed before PERF_RECORD_COMM
(when they got recorded in different CPUs) (Jiri Olsa)
- Add support to display group output for non group events, i.e.
now when one uses 'perf report --group' on a perf.data file
recorded without explicitly grouping events with {} (e.g.
"perf record -e '{cycles,instructions}'" get the same output
that would produce, i.e. see all those non-grouped events in
multiple columns, at the same time (Jiri Olsa)
- Skip non-address kallsyms entries, e.g. '(null)' for !root (Jiri Olsa)
- Kernel maps fixes wrt perf.data(report) versus live system (top)
(Jiri Olsa)
- Fix memory corruption when using 'perf record -j call -g -a <application>'
followed by 'perf report --branch-history' (Jiri Olsa)
- ARM CoreSight fixes (Mathieu Poirier)
- Add inject capability for CoreSight Traces (Robert Waker)
- Update documentation for use of 'perf' + ARM CoreSight (Robert Walker)
- Man pages fixes (Sangwon Hong, Jaecheol Shin)
- Fix some 'perf test' cases on s/390 and x86_64 (some backtraces
changed with a glibc update) (Thomas Richter)
- Add detailed CPUID info in the 'perf.data' headers for s/390 to
then use it in 'perf annotate' (Thomas Richter)
- Add '--interval-count N' to 'perf stat', to use with -I, i.e.
'perf stat -I 1000 --interval-count 2' will show stats every
1000ms, two times (yuzhoujian)
- Add 'perf stat --timeout Nms', that will run for that many
milliseconds and then stop, printing the counters (yuzhoujian)
- Fix description for 'perf report --mem-modex (Andi Kleen)
- Use a wildcard to remove the vfs_getname probe in the
'perf test' shell based test cases (Arnaldo Carvalho de Melo)
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Send a cookie with sendmsg() on PF_RDS sockets, and process the
returned batched cookies in do_recv_completion()
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In some situations the vfs_getname is being added both as requested and
with a _1 suffix (inlines?):
probe:vfs_getname_1 (on getname_flags:63@acme/git/linux/fs/namei.c with pathname)
This ends up making the cleanup to miss that one, as it removes just
'probe:vfs_getname', which makes the second test to use this probe point
to fail, since it finds that leftover from the first test, use a
wildcard to remove both.
Before:
# perf test 60 61 62 63
60: Use vfs_getname probe to get syscall args filenames : FAILED!
61: probe libc's inet_pton & backtrace it with ping : Ok
62: Check open filename arg using perf trace + vfs_getname: FAILED!
63: Add vfs_getname probe to get syscall args filenames : Ok
After:
# perf test 60 61 62 63
60: Use vfs_getname probe to get syscall args filenames : Ok
61: probe libc's inet_pton & backtrace it with ping : Ok
62: Check open filename arg using perf trace + vfs_getname: Ok
63: Add vfs_getname probe to get syscall args filenames : Ok
#
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Masami Hiramatsu <mhiramat@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-2k5kutwr4ds36adiakyb4yvy@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Using Fedora 27 and latest Linux kernel the test case
trace+probe_libc_inet_pton.sh fails again on s390. This time is the
inlining of functions which does not match. After an update of the
glibc (from 2.26-16 to 2.26-24) the output is different
The expected output is:
__inet_pton (/usr/lib64/libc-2.26.so)
gaih_inet (inlined)
....
The actual output is:
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 0.061/0.061/0.061/0.000 ms
0.000 probe_libc:inet_pton:(3ffb2140448))
__inet_pton (inlined)
gaih_inet.constprop.7 (/usr/lib64/libc-2.26.so)
...
Fix this by being less strict on 'inlined' verses library name and
accept both
Signed-off-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Link: http://lkml.kernel.org/r/20180214070303.55757-1-tmricht@linux.vnet.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The function get_cpuid_str() is called by perf_pmu__getcpuid() and on
s390 returns a complete description of the CPU and its capabilities,
which is a comma separated list.
To map the CPU type with the value defined in the
pmu-events/arch/s390/mapfile.csv, introduce an architecture specific
cpuid compare function named strcmp_cpuid_str()
The currently used regex algorithm is defined as the weak default and
will be used if no platform specific one is defined. This matches the
current behavior.
Signed-off-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Reviewed-by: Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Link: http://lkml.kernel.org/r/20180213151419.80737-3-tmricht@linux.vnet.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When perf record ... is setup to record data, the s390 cpu information
was a fixed string "IBM/S390".
Replace this string with one containing more information about the
machine. The information included in the cpuid is a comma separated
list:
manufacturer,type,model-capacity,model[,version,authorization]
with
- manufacturer: up to 16 byte name of the manufacturer (IBM).
- type: a four digit number refering to the machine
generation.
- model-capacitiy: up to 16 characters describing number
of cpus etc.
- model: up to 16 characters describing model.
- version: the CPU-MF counter facility version number,
available on LPARs only, omitted on z/VM guests.
- authorization: the CPU-MF counter facility authorization level,
available on LPARs only, omitted on z/VM guests.
Before:
[root@s8360047 perf]# ./perf record -- sleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.001 MB perf.data (4 samples) ]
[root@s8360047 perf]# ./perf report --header | fgrep cpuid
# cpuid : IBM/S390
[root@s8360047 perf]#
After:
[root@s35lp76 perf]# ./perf report --header|fgrep cpuid
# cpuid : IBM,3906,704,M03,3.5,002f
[root@s35lp76 perf]#
Signed-off-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Reviewed-by: Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Link: http://lkml.kernel.org/r/20180213151419.80737-1-tmricht@linux.vnet.ibm.com
[ Use scnprintf instead of strncat to fix build errors on gcc GNU C99 5.4.0 20160609 -march=zEC12 -m64 -mzarch -ggdb3 -O6 -std=gnu99 -fPIC -fno-omit-frame-pointer -funwind-tables -fstack-protector-all ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Jin Yao reported memory corrupton in perf report with
branch info used for stack trace:
> Following command lines will cause perf crash.
> perf record -j call -g -a <application>
> perf report --branch-history
>
> *** Error in `perf': double free or corruption (!prev): 0x00000000104aa040 ***
> ======= Backtrace: =========
> /lib/x86_64-linux-gnu/libc.so.6(+0x77725)[0x7f6b37254725]
> /lib/x86_64-linux-gnu/libc.so.6(+0x7ff4a)[0x7f6b3725cf4a]
> /lib/x86_64-linux-gnu/libc.so.6(cfree+0x4c)[0x7f6b37260abc]
> perf[0x51b914]
> perf(hist_entry_iter__add+0x1e5)[0x51f305]
> perf[0x43cf01]
> perf[0x4fa3bf]
> perf[0x4fa923]
> perf[0x4fd396]
> perf[0x4f9614]
> perf(perf_session__process_events+0x89e)[0x4fc38e]
> perf(cmd_report+0x15d2)[0x43f202]
> perf[0x4a059f]
> perf(main+0x631)[0x427b71]
> /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf0)[0x7f6b371fd830]
> perf(_start+0x29)[0x427d89]
For the cumulative output, we allocate the he_cache array based on the
--max-stack option value and populate it with data from 'callchain_cursor'.
The --max-stack option value does not ensure now the limit for number of
callchain_cursor nodes, so the cumulative iter code will allocate smaller array
than it's actually needed and cause above corruption.
I think the --max-stack limit does not apply here anyway, because we add
callchain data as normal hist entries, while the --max-stack control the limit
of single entry callchain depth.
Using the callchain_cursor.nr as he_cache array count to fix this. Also
removing struct hist_entry_iter::max_stack, because there's no longer any use
for it.
We need more fixes to ensure that the branch stack code follows properly the
logic of --max-stack, which is not the case at the moment.
Original-patch-by: Jin Yao <yao.jin@linux.intel.com>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Reported-by: Jin Yao <yao.jin@linux.intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20180216123619.GA9945@krava
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When we use perf report interactive annotate view, we can see
the position of jump arrow is not correct. For example,
1. perf record -b ...
2. perf report
3. In interactive mode, select Annotate 'function'
Percent│ IPC Cycle
│ if (flag)
1.37 │0.4┌── 1 ↓ je 82
│ │ x += x / y + y / x;
0.00 │0.4│ 1310 movsd (%rsp),%xmm0
0.00 │0.4│ 565 movsd 0x8(%rsp),%xmm4
│0.4│ movsd 0x8(%rsp),%xmm1
│0.4│ movsd (%rsp),%xmm3
│0.4│ divsd %xmm4,%xmm0
0.00 │0.4│ 579 divsd %xmm3,%xmm1
│0.4│ movsd (%rsp),%xmm2
│0.4│ addsd %xmm1,%xmm0
│0.4│ addsd %xmm2,%xmm0
0.00 │0.4│ movsd %xmm0,(%rsp)
│ │ volatile double x = 1212121212, y = 121212;
│ │
│ │ s_randseed = time(0);
│ │ srand(s_randseed);
│ │
│ │ for (i = 0; i < 2000000000; i++) {
1.37 │0.4└─→ 82: sub $0x1,%ebx
28.21 │0.48 17 ↑ jne 38
The jump arrow in above example is not correct. It should add the
width of IPC and Cycle.
With this patch, the result is:
Percent│ IPC Cycle
│ if (flag)
1.37 │0.48 1 ┌──je 82
│ │ x += x / y + y / x;
0.00 │0.48 1310 │ movsd (%rsp),%xmm0
0.00 │0.48 565 │ movsd 0x8(%rsp),%xmm4
│0.48 │ movsd 0x8(%rsp),%xmm1
│0.48 │ movsd (%rsp),%xmm3
│0.48 │ divsd %xmm4,%xmm0
0.00 │0.48 579 │ divsd %xmm3,%xmm1
│0.48 │ movsd (%rsp),%xmm2
│0.48 │ addsd %xmm1,%xmm0
│0.48 │ addsd %xmm2,%xmm0
0.00 │0.48 │ movsd %xmm0,(%rsp)
│ │ volatile double x = 1212121212, y = 121212;
│ │
│ │ s_randseed = time(0);
│ │ srand(s_randseed);
│ │
│ │ for (i = 0; i < 2000000000; i++) {
1.37 │0.48 82:└─→sub $0x1,%ebx
28.21 │0.48 17 ↑ jne 38
Committer notes:
Please note that only from LBRv5 (according to Jiri) onwards, i.e. >=
Skylake is that we'll have the cycles counts in each branch record
entry, so to see the Cycles and IPC columns, and be able to test this
patch, one need a capable hardware.
While applying this I first tested it on a Broadwell class machine and
couldn't get those columns, will add code to the annotate browser to
warn the user about that, i.e. you have branch records, but no cycles,
use a more recent hardware to get the cycles and IPC columns.
Signed-off-by: Jin Yao <yao.jin@linux.intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jin Yao <yao.jin@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1517223473-14750-1-git-send-email-yao.jin@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
