diff options
Diffstat (limited to 'tools/perf')
109 files changed, 8635 insertions, 2931 deletions
diff --git a/tools/perf/.gitignore b/tools/perf/.gitignore index fe08660ce0bd..e1d60d780784 100644 --- a/tools/perf/.gitignore +++ b/tools/perf/.gitignore @@ -13,6 +13,8 @@ perf*.xml perf*.html common-cmds.h perf.data +perf.data.old +perf-archive tags TAGS cscope* diff --git a/tools/perf/Documentation/perf-archive.txt b/tools/perf/Documentation/perf-archive.txt new file mode 100644 index 000000000000..fae174dc7d01 --- /dev/null +++ b/tools/perf/Documentation/perf-archive.txt @@ -0,0 +1,22 @@ +perf-archive(1) +=============== + +NAME +---- +perf-archive - Create archive with object files with build-ids found in perf.data file + +SYNOPSIS +-------- +[verse] +'perf archive' [file] + +DESCRIPTION +----------- +This command runs runs perf-buildid-list --with-hits, and collects the files +with the buildids found so that analisys of perf.data contents can be possible +on another machine. + + +SEE ALSO +-------- +linkperf:perf-record[1], linkperf:perf-buildid-list[1], linkperf:perf-report[1] diff --git a/tools/perf/Documentation/perf-buildid-cache.txt b/tools/perf/Documentation/perf-buildid-cache.txt new file mode 100644 index 000000000000..88bc3b519746 --- /dev/null +++ b/tools/perf/Documentation/perf-buildid-cache.txt @@ -0,0 +1,33 @@ +perf-buildid-cache(1) +===================== + +NAME +---- +perf-buildid-cache - Manage build-id cache. + +SYNOPSIS +-------- +[verse] +'perf buildid-list <options>' + +DESCRIPTION +----------- +This command manages the build-id cache. It can add and remove files to the +cache. In the future it should as well purge older entries, set upper limits +for the space used by the cache, etc. + +OPTIONS +------- +-a:: +--add=:: + Add specified file to the cache. +-r:: +--remove=:: + Remove specified file to the cache. +-v:: +--verbose:: + Be more verbose. + +SEE ALSO +-------- +linkperf:perf-record[1], linkperf:perf-report[1] diff --git a/tools/perf/Documentation/perf-diff.txt b/tools/perf/Documentation/perf-diff.txt new file mode 100644 index 000000000000..8974e208cba6 --- /dev/null +++ b/tools/perf/Documentation/perf-diff.txt @@ -0,0 +1,55 @@ +perf-diff(1) +============== + +NAME +---- +perf-diff - Read two perf.data files and display the differential profile + +SYNOPSIS +-------- +[verse] +'perf diff' [oldfile] [newfile] + +DESCRIPTION +----------- +This command displays the performance difference amongst two perf.data files +captured via perf record. + +If no parameters are passed it will assume perf.data.old and perf.data. + +OPTIONS +------- +-d:: +--dsos=:: + Only consider symbols in these dsos. CSV that understands + file://filename entries. + +-C:: +--comms=:: + Only consider symbols in these comms. CSV that understands + file://filename entries. + +-S:: +--symbols=:: + Only consider these symbols. CSV that understands + file://filename entries. + +-s:: +--sort=:: + Sort by key(s): pid, comm, dso, symbol. + +-t:: +--field-separator=:: + + Use a special separator character and don't pad with spaces, replacing + all occurances of this separator in symbol names (and other output) + with a '.' character, that thus it's the only non valid separator. + +-v:: +--verbose:: + Be verbose, for instance, show the raw counts in addition to the + diff. + +SEE ALSO +-------- +linkperf:perf-record[1] diff --git a/tools/perf/Documentation/perf-kmem.txt b/tools/perf/Documentation/perf-kmem.txt index 44b0ce35c28a..eac4d852e7cd 100644 --- a/tools/perf/Documentation/perf-kmem.txt +++ b/tools/perf/Documentation/perf-kmem.txt @@ -8,16 +8,16 @@ perf-kmem - Tool to trace/measure kernel memory(slab) properties SYNOPSIS -------- [verse] -'perf kmem' {record} [<options>] +'perf kmem' {record|stat} [<options>] DESCRIPTION ----------- -There's two variants of perf kmem: +There are two variants of perf kmem: 'perf kmem record <command>' to record the kmem events of an arbitrary workload. - 'perf kmem' to report kernel memory statistics. + 'perf kmem stat' to report kernel memory statistics. OPTIONS ------- @@ -25,8 +25,11 @@ OPTIONS --input=<file>:: Select the input file (default: perf.data) ---stat=<caller|alloc>:: - Select per callsite or per allocation statistics +--caller:: + Show per-callsite statistics + +--alloc:: + Show per-allocation statistics -s <key[,key2...]>:: --sort=<key[,key2...]>:: diff --git a/tools/perf/Documentation/perf-probe.txt b/tools/perf/Documentation/perf-probe.txt index 9270594e6dfd..2de34075f6a4 100644 --- a/tools/perf/Documentation/perf-probe.txt +++ b/tools/perf/Documentation/perf-probe.txt @@ -8,10 +8,15 @@ perf-probe - Define new dynamic tracepoints SYNOPSIS -------- [verse] -'perf probe' [options] --add 'PROBE' [--add 'PROBE' ...] +'perf probe' [options] --add='PROBE' [...] or -'perf probe' [options] 'PROBE' ['PROBE' ...] - +'perf probe' [options] PROBE +or +'perf probe' [options] --del='[GROUP:]EVENT' [...] +or +'perf probe' --list +or +'perf probe' --line='FUNC[:RLN[+NUM|:RLN2]]|SRC:ALN[+NUM|:ALN2]' DESCRIPTION ----------- @@ -31,19 +36,46 @@ OPTIONS Be more verbose (show parsed arguments, etc). -a:: ---add:: - Define a probe point (see PROBE SYNTAX for detail) +--add=:: + Define a probe event (see PROBE SYNTAX for detail). + +-d:: +--del=:: + Delete a probe event. + +-l:: +--list:: + List up current probe events. + +-L:: +--line=:: + Show source code lines which can be probed. This needs an argument + which specifies a range of the source code. PROBE SYNTAX ------------ Probe points are defined by following syntax. - "FUNC[+OFFS|:RLN|%return][@SRC]|SRC:ALN [ARG ...]" + "[EVENT=]FUNC[+OFFS|:RLN|%return][@SRC]|SRC:ALN [ARG ...]" +'EVENT' specifies the name of new event, if omitted, it will be set the name of the probed function. Currently, event group name is set as 'probe'. 'FUNC' specifies a probed function name, and it may have one of the following options; '+OFFS' is the offset from function entry address in bytes, 'RLN' is the relative-line number from function entry line, and '%return' means that it probes function return. In addition, 'SRC' specifies a source file which has that function. It is also possible to specify a probe point by the source line number by using 'SRC:ALN' syntax, where 'SRC' is the source file path and 'ALN' is the line number. 'ARG' specifies the arguments of this probe point. You can use the name of local variable, or kprobe-tracer argument format (e.g. $retval, %ax, etc). +LINE SYNTAX +----------- +Line range is descripted by following syntax. + + "FUNC[:RLN[+NUM|:RLN2]]|SRC:ALN[+NUM|:ALN2]" + +FUNC specifies the function name of showing lines. 'RLN' is the start line +number from function entry line, and 'RLN2' is the end line number. As same as +probe syntax, 'SRC' means the source file path, 'ALN' is start line number, +and 'ALN2' is end line number in the file. It is also possible to specify how +many lines to show by using 'NUM'. +So, "source.c:100-120" shows lines between 100th to l20th in source.c file. And "func:10+20" shows 20 lines from 10th line of func function. + SEE ALSO -------- linkperf:perf-trace[1], linkperf:perf-record[1] diff --git a/tools/perf/Documentation/perf-report.txt b/tools/perf/Documentation/perf-report.txt index 9dccb180b7af..abfabe9147a4 100644 --- a/tools/perf/Documentation/perf-report.txt +++ b/tools/perf/Documentation/perf-report.txt @@ -39,6 +39,10 @@ OPTIONS Only consider these symbols. CSV that understands file://filename entries. +-s:: +--sort=:: + Sort by key(s): pid, comm, dso, symbol, parent. + -w:: --field-width=:: Force each column width to the provided list, for large terminal diff --git a/tools/perf/Documentation/perf-top.txt b/tools/perf/Documentation/perf-top.txt index 4a7d558dc309..785b9fc32a46 100644 --- a/tools/perf/Documentation/perf-top.txt +++ b/tools/perf/Documentation/perf-top.txt @@ -74,7 +74,7 @@ OPTIONS -s <symbol>:: --sym-annotate=<symbol>:: - Annotate this symbol. Requires -k option. + Annotate this symbol. -v:: --verbose:: diff --git a/tools/perf/Documentation/perf-trace-perl.txt b/tools/perf/Documentation/perf-trace-perl.txt index c5f55f439091..d729cee8d987 100644 --- a/tools/perf/Documentation/perf-trace-perl.txt +++ b/tools/perf/Documentation/perf-trace-perl.txt @@ -8,7 +8,7 @@ perf-trace-perl - Process trace data with a Perl script SYNOPSIS -------- [verse] -'perf trace' [-s [lang]:script[.ext] ] +'perf trace' [-s [Perl]:script[.pl] ] DESCRIPTION ----------- diff --git a/tools/perf/Documentation/perf-trace-python.txt b/tools/perf/Documentation/perf-trace-python.txt new file mode 100644 index 000000000000..a241aca77184 --- /dev/null +++ b/tools/perf/Documentation/perf-trace-python.txt @@ -0,0 +1,625 @@ +perf-trace-python(1) +================== + +NAME +---- +perf-trace-python - Process trace data with a Python script + +SYNOPSIS +-------- +[verse] +'perf trace' [-s [Python]:script[.py] ] + +DESCRIPTION +----------- + +This perf trace option is used to process perf trace data using perf's +built-in Python interpreter. It reads and processes the input file and +displays the results of the trace analysis implemented in the given +Python script, if any. + +A QUICK EXAMPLE +--------------- + +This section shows the process, start to finish, of creating a working +Python script that aggregates and extracts useful information from a +raw perf trace stream. You can avoid reading the rest of this +document if an example is enough for you; the rest of the document +provides more details on each step and lists the library functions +available to script writers. + +This example actually details the steps that were used to create the +'syscall-counts' script you see when you list the available perf trace +scripts via 'perf trace -l'. As such, this script also shows how to +integrate your script into the list of general-purpose 'perf trace' +scripts listed by that command. + +The syscall-counts script is a simple script, but demonstrates all the +basic ideas necessary to create a useful script. Here's an example +of its output (syscall names are not yet supported, they will appear +as numbers): + +---- +syscall events: + +event count +---------------------------------------- ----------- +sys_write 455067 +sys_getdents 4072 +sys_close 3037 +sys_swapoff 1769 +sys_read 923 +sys_sched_setparam 826 +sys_open 331 +sys_newfstat 326 +sys_mmap 217 +sys_munmap 216 +sys_futex 141 +sys_select 102 +sys_poll 84 +sys_setitimer 12 +sys_writev 8 +15 8 +sys_lseek 7 +sys_rt_sigprocmask 6 +sys_wait4 3 +sys_ioctl 3 +sys_set_robust_list 1 +sys_exit 1 +56 1 +sys_access 1 +---- + +Basically our task is to keep a per-syscall tally that gets updated +every time a system call occurs in the system. Our script will do +that, but first we need to record the data that will be processed by +that script. Theoretically, there are a couple of ways we could do +that: + +- we could enable every event under the tracing/events/syscalls + directory, but this is over 600 syscalls, well beyond the number + allowable by perf. These individual syscall events will however be + useful if we want to later use the guidance we get from the + general-purpose scripts to drill down and get more detail about + individual syscalls of interest. + +- we can enable the sys_enter and/or sys_exit syscalls found under + tracing/events/raw_syscalls. These are called for all syscalls; the + 'id' field can be used to distinguish between individual syscall + numbers. + +For this script, we only need to know that a syscall was entered; we +don't care how it exited, so we'll use 'perf record' to record only +the sys_enter events: + +---- +# perf record -c 1 -f -a -M -R -e raw_syscalls:sys_enter + +^C[ perf record: Woken up 1 times to write data ] +[ perf record: Captured and wrote 56.545 MB perf.data (~2470503 samples) ] +---- + +The options basically say to collect data for every syscall event +system-wide and multiplex the per-cpu output into a single stream. +That single stream will be recorded in a file in the current directory +called perf.data. + +Once we have a perf.data file containing our data, we can use the -g +'perf trace' option to generate a Python script that will contain a +callback handler for each event type found in the perf.data trace +stream (for more details, see the STARTER SCRIPTS section). + +---- +# perf trace -g python +generated Python script: perf-trace.py + +The output file created also in the current directory is named +perf-trace.py. Here's the file in its entirety: + +# perf trace event handlers, generated by perf trace -g python +# Licensed under the terms of the GNU GPL License version 2 + +# The common_* event handler fields are the most useful fields common to +# all events. They don't necessarily correspond to the 'common_*' fields +# in the format files. Those fields not available as handler params can +# be retrieved using Python functions of the form common_*(context). +# See the perf-trace-python Documentation for the list of available functions. + +import os +import sys + +sys.path.append(os.environ['PERF_EXEC_PATH'] + \ + '/scripts/python/Perf-Trace-Util/lib/Perf/Trace') + +from perf_trace_context import * +from Core import * + +def trace_begin(): + print "in trace_begin" + +def trace_end(): + print "in trace_end" + +def raw_syscalls__sys_enter(event_name, context, common_cpu, + common_secs, common_nsecs, common_pid, common_comm, + id, args): + print_header(event_name, common_cpu, common_secs, common_nsecs, + common_pid, common_comm) + + print "id=%d, args=%s\n" % \ + (id, args), + +def trace_unhandled(event_name, context, common_cpu, common_secs, common_nsecs, + common_pid, common_comm): + print_header(event_name, common_cpu, common_secs, common_nsecs, + common_pid, common_comm) + +def print_header(event_name, cpu, secs, nsecs, pid, comm): + print "%-20s %5u %05u.%09u %8u %-20s " % \ + (event_name, cpu, secs, nsecs, pid, comm), +---- + +At the top is a comment block followed by some import statements and a +path append which every perf trace script should include. + +Following that are a couple generated functions, trace_begin() and +trace_end(), which are called at the beginning and the end of the +script respectively (for more details, see the SCRIPT_LAYOUT section +below). + +Following those are the 'event handler' functions generated one for +every event in the 'perf record' output. The handler functions take +the form subsystem__event_name, and contain named parameters, one for +each field in the event; in this case, there's only one event, +raw_syscalls__sys_enter(). (see the EVENT HANDLERS section below for +more info on event handlers). + +The final couple of functions are, like the begin and end functions, +generated for every script. The first, trace_unhandled(), is called +every time the script finds an event in the perf.data file that +doesn't correspond to any event handler in the script. This could +mean either that the record step recorded event types that it wasn't +really interested in, or the script was run against a trace file that +doesn't correspond to the script. + +The script generated by -g option option simply prints a line for each +event found in the trace stream i.e. it basically just dumps the event +and its parameter values to stdout. The print_header() function is +simply a utility function used for that purpose. Let's rename the +script and run it to see the default output: + +---- +# mv perf-trace.py syscall-counts.py +# perf trace -s syscall-counts.py + +raw_syscalls__sys_enter 1 00840.847582083 7506 perf id=1, args= +raw_syscalls__sys_enter 1 00840.847595764 7506 perf id=1, args= +raw_syscalls__sys_enter 1 00840.847620860 7506 perf id=1, args= +raw_syscalls__sys_enter 1 00840.847710478 6533 npviewer.bin id=78, args= +raw_syscalls__sys_enter 1 00840.847719204 6533 npviewer.bin id=142, args= +raw_syscalls__sys_enter 1 00840.847755445 6533 npviewer.bin id=3, args= +raw_syscalls__sys_enter 1 00840.847775601 6533 npviewer.bin id=3, args= +raw_syscalls__sys_enter 1 00840.847781820 6533 npviewer.bin id=3, args= +. +. +. +---- + +Of course, for this script, we're not interested in printing every +trace event, but rather aggregating it in a useful way. So we'll get +rid of everything to do with printing as well as the trace_begin() and +trace_unhandled() functions, which we won't be using. That leaves us +with this minimalistic skeleton: + +---- +import os +import sys + +sys.path.append(os.environ['PERF_EXEC_PATH'] + \ + '/scripts/python/Perf-Trace-Util/lib/Perf/Trace') + +from perf_trace_context import * +from Core import * + +def trace_end(): + print "in trace_end" + +def raw_syscalls__sys_enter(event_name, context, common_cpu, + common_secs, common_nsecs, common_pid, common_comm, + id, args): +---- + +In trace_end(), we'll simply print the results, but first we need to +generate some results to print. To do that we need to have our +sys_enter() handler do the necessary tallying until all events have +been counted. A hash table indexed by syscall id is a good way to +store that information; every time the sys_enter() handler is called, +we simply increment a count associated with that hash entry indexed by +that syscall id: + +---- + syscalls = autodict() + + try: + syscalls[id] += 1 + except TypeError: + syscalls[id] = 1 +---- + +The syscalls 'autodict' object is a special kind of Python dictionary +(implemented in Core.py) that implements Perl's 'autovivifying' hashes +in Python i.e. with autovivifying hashes, you can assign nested hash +values without having to go to the trouble of creating intermediate +levels if they don't exist e.g syscalls[comm][pid][id] = 1 will create +the intermediate hash levels and finally assign the value 1 to the +hash entry for 'id' (because the value being assigned isn't a hash +object itself, the initial value is assigned in the TypeError +exception. Well, there may be a better way to do this in Python but +that's what works for now). + +Putting that code into the raw_syscalls__sys_enter() handler, we +effectively end up with a single-level dictionary keyed on syscall id +and having the counts we've tallied as values. + +The print_syscall_totals() function iterates over the entries in the +dictionary and displays a line for each entry containing the syscall +name (the dictonary keys contain the syscall ids, which are passed to +the Util function syscall_name(), which translates the raw syscall +numbers to the corresponding syscall name strings). The output is +displayed after all the events in the trace have been processed, by +calling the print_syscall_totals() function from the trace_end() +handler called at the end of script processing. + +The final script producing the output shown above is shown in its +entirety below (syscall_name() helper is not yet available, you can +only deal with id's for now): + +---- +import os +import sys + +sys.path.append(os.environ['PERF_EXEC_PATH'] + \ + '/scripts/python/Perf-Trace-Util/lib/Perf/Trace') + +from perf_trace_context import * +from Core import * +from Util import * + +syscalls = autodict() + +def trace_end(): + print_syscall_totals() + +def raw_syscalls__sys_enter(event_name, context, common_cpu, + common_secs, common_nsecs, common_pid, common_comm, + id, args): + try: + syscalls[id] += 1 + except TypeError: + syscalls[id] = 1 + +def print_syscall_totals(): + if for_comm is not None: + print "\nsyscall events for %s:\n\n" % (for_comm), + else: + print "\nsyscall events:\n\n", + + print "%-40s %10s\n" % ("event", "count"), + print "%-40s %10s\n" % ("----------------------------------------", \ + "-----------"), + + for id, val in sorted(syscalls.iteritems(), key = lambda(k, v): (v, k), \ + reverse = True): + print "%-40s %10d\n" % (syscall_name(id), val), +---- + +The script can be run just as before: + + # perf trace -s syscall-counts.py + +So those are the essential steps in writing and running a script. The +process can be generalized to any tracepoint or set of tracepoints +you're interested in - basically find the tracepoint(s) you're +interested in by looking at the list of available events shown by +'perf list' and/or look in /sys/kernel/debug/tracing events for +detailed event and field info, record the corresponding trace data +using 'perf record', passing it the list of interesting events, +generate a skeleton script using 'perf trace -g python' and modify the +code to aggregate and display it for your particular needs. + +After you've done that you may end up with a general-purpose script +that you want to keep around and have available for future use. By +writing a couple of very simple shell scripts and putting them in the +right place, you can have your script listed alongside the other +scripts listed by the 'perf trace -l' command e.g.: + +---- +root@tropicana:~# perf trace -l +List of available trace scripts: + workqueue-stats workqueue stats (ins/exe/create/destroy) + wakeup-latency system-wide min/max/avg wakeup latency + rw-by-file <comm> r/w activity for a program, by file + rw-by-pid system-wide r/w activity +---- + +A nice side effect of doing this is that you also then capture the +probably lengthy 'perf record' command needed to record the events for +the script. + +To have the script appear as a 'built-in' script, you write two simple +scripts, one for recording and one for 'reporting'. + +The 'record' script is a shell script with the same base name as your +script, but with -record appended. The shell script should be put +into the perf/scripts/python/bin directory in the kernel source tree. +In that script, you write the 'perf record' command-line needed for +your script: + +---- +# cat kernel-source/tools/perf/scripts/python/bin/syscall-counts-record + +#!/bin/bash +perf record -c 1 -f -a -M -R -e raw_syscalls:sys_enter +---- + +The 'report' script is also a shell script with the same base name as +your script, but with -report appended. It should also be located in +the perf/scripts/python/bin directory. In that script, you write the +'perf trace -s' command-line needed for running your script: + +---- +# cat kernel-source/tools/perf/scripts/python/bin/syscall-counts-report + +#!/bin/bash +# description: system-wide syscall counts +perf trace -s ~/libexec/perf-core/scripts/python/syscall-counts.py +---- + +Note that the location of the Python script given in the shell script +is in the libexec/perf-core/scripts/python directory - this is where +the script will be copied by 'make install' when you install perf. +For the installation to install your script there, your script needs +to be located in the perf/scripts/python directory in the kernel +source tree: + +---- +# ls -al kernel-source/tools/perf/scripts/python + +root@tropicana:/home/trz/src/tip# ls -al tools/perf/scripts/python +total 32 +drwxr-xr-x 4 trz trz 4096 2010-01-26 22:30 . +drwxr-xr-x 4 trz trz 4096 2010-01-26 22:29 .. +drwxr-xr-x 2 trz trz 4096 2010-01-26 22:29 bin +-rw-r--r-- 1 trz trz 2548 2010-01-26 22:29 check-perf-trace.py +drwxr-xr-x 3 trz trz 4096 2010-01-26 22:49 Perf-Trace-Util +-rw-r--r-- 1 trz trz 1462 2010-01-26 22:30 syscall-counts.py +---- + +Once you've done that (don't forget to do a new 'make install', +otherwise your script won't show up at run-time), 'perf trace -l' +should show a new entry for your script: + +---- +root@tropicana:~# perf trace -l +List of available trace scripts: + workqueue-stats workqueue stats (ins/exe/create/destroy) + wakeup-latency system-wide min/max/avg wakeup latency + rw-by-file <comm> r/w activity for a program, by file + rw-by-pid system-wide r/w activity + syscall-counts system-wide syscall counts +---- + +You can now perform the record step via 'perf trace record': + + # perf trace record syscall-counts + +and display the output using 'perf trace report': + + # perf trace report syscall-counts + +STARTER SCRIPTS +--------------- + +You can quickly get started writing a script for a particular set of +trace data by generating a skeleton script using 'perf trace -g +python' in the same directory as an existing perf.data trace file. +That will generate a starter script containing a handler for each of +the event types in the trace file; it simply prints every available +field for each event in the trace file. + +You can also look at the existing scripts in +~/libexec/perf-core/scripts/python for typical examples showing how to +do basic things like aggregate event data, print results, etc. Also, +the check-perf-trace.py script, while not interesting for its results, +attempts to exercise all of the main scripting features. + +EVENT HANDLERS +-------------- + +When perf trace is invoked using a trace script, a user-defined +'handler function' is called for each event in the trace. If there's +no handler function defined for a given event type, the event is +ignored (or passed to a 'trace_handled' function, see below) and the +next event is processed. + +Most of the event's field values are passed as arguments to the +handler function; some of the less common ones aren't - those are +available as calls back into the perf executable (see below). + +As an example, the following perf record command can be used to record +all sched_wakeup events in the system: + + # perf record -c 1 -f -a -M -R -e sched:sched_wakeup + +Traces meant to be processed using a script should be recorded with +the above options: -c 1 says to sample every event, -a to enable +system-wide collection, -M to multiplex the output, and -R to collect +raw samples. + +The format file for the sched_wakep event defines the following fields +(see /sys/kernel/debug/tracing/events/sched/sched_wakeup/format): + +---- + format: + field:unsigned short common_type; + field:unsigned char common_flags; + field:unsigned char common_preempt_count; + field:int common_pid; + field:int common_lock_depth; + + field:char comm[TASK_COMM_LEN]; + field:pid_t pid; + field:int prio; + field:int success; + field:int target_cpu; +---- + +The handler function for this event would be defined as: + +---- +def sched__sched_wakeup(event_name, context, common_cpu, common_secs, + common_nsecs, common_pid, common_comm, + comm, pid, prio, success, target_cpu): + pass +---- + +The handler function takes the form subsystem__event_name. + +The common_* arguments in the handler's argument list are the set of +arguments passed to all event handlers; some of the fields correspond +to the common_* fields in the format file, but some are synthesized, +and some of the common_* fields aren't common enough to to be passed +to every event as arguments but are available as library functions. + +Here's a brief description of each of the invariant event args: + + event_name the name of the event as text + context an opaque 'cookie' used in calls back into perf + common_cpu the cpu the event occurred on + common_secs the secs portion of the event timestamp + common_nsecs the nsecs portion of the event timestamp + common_pid the pid of the current task + common_comm the name of the current process + +All of the remaining fields in the event's format file have +counterparts as handler function arguments of the same name, as can be +seen in the example above. + +The above provides the basics needed to directly access every field of +every event in a trace, which covers 90% of what you need to know to +write a useful trace script. The sections below cover the rest. + +SCRIPT LAYOUT +------------- + +Every perf trace Python script should start by setting up a Python +module search path and 'import'ing a few support modules (see module +descriptions below): + +---- + import os + import sys + + sys.path.append(os.environ['PERF_EXEC_PATH'] + \ + '/scripts/python/Perf-Trace-Util/lib/Perf/Trace') + + from perf_trace_context import * + from Core import * +---- + +The rest of the script can contain handler functions and support +functions in any order. + +Aside from the event handler functions discussed above, every script +can implement a set of optional functions: + +*trace_begin*, if defined, is called before any event is processed and +gives scripts a chance to do setup tasks: + +---- +def trace_begin: + pass +---- + +*trace_end*, if defined, is called after all events have been + processed and gives scripts a chance to do end-of-script tasks, such + as display results: + +---- +def trace_end: + pass +---- + +*trace_unhandled*, if defined, is called after for any event that + doesn't have a handler explicitly defined for it. The standard set + of common arguments are passed into it: + +---- +def trace_unhandled(event_name, context, common_cpu, common_secs, + common_nsecs, common_pid, common_comm): + pass +---- + +The remaining sections provide descriptions of each of the available +built-in perf trace Python modules and their associated functions. + +AVAILABLE MODULES AND FUNCTIONS +------------------------------- + +The following sections describe the functions and variables available +via the various perf trace Python modules. To use the functions and +variables from the given module, add the corresponding 'from XXXX +import' line to your perf trace script. + +Core.py Module +~~~~~~~~~~~~~~ + +These functions provide some essential functions to user scripts. + +The *flag_str* and *symbol_str* functions provide human-readable +strings for flag and symbolic fields. These correspond to the strings +and values parsed from the 'print fmt' fields of the event format +files: + + flag_str(event_name, field_name, field_value) - returns the string represention corresponding to field_value for the flag field field_name of event event_name + symbol_str(event_name, field_name, field_value) - returns the string represention corresponding to field_value for the symbolic field field_name of event event_name + +The *autodict* function returns a special special kind of Python +dictionary that implements Perl's 'autovivifying' hashes in Python +i.e. with autovivifying hashes, you can assign nested hash values +without having to go to the trouble of creating intermediate levels if +they don't exist. + + autodict() - returns an autovivifying dictionary instance + + +perf_trace_context Module +~~~~~~~~~~~~~~~~~~~~~~~~~ + +Some of the 'common' fields in the event format file aren't all that +common, but need to be made accessible to user scripts nonetheless. + +perf_trace_context defines a set of functions that can be used to +access this data in the context of the current event. Each of these +functions expects a context variable, which is the same as the +context variable passed into every event handler as the second +argument. + + common_pc(context) - returns common_preempt count for the current event + common_flags(context) - returns common_flags for the current event + common_lock_depth(context) - returns common_lock_depth for the current event + +Util.py Module +~~~~~~~~~~~~~~ + +Various utility functions for use with perf trace: + + nsecs(secs, nsecs) - returns total nsecs given secs/nsecs pair + nsecs_secs(nsecs) - returns whole secs portion given nsecs + nsecs_nsecs(nsecs) - returns nsecs remainder given nsecs + nsecs_str(nsecs) - returns printable string in the form secs.nsecs + avg(total, n) - returns average given a sum and a total number of values + +SEE ALSO +-------- +linkperf:perf-trace[1] diff --git a/tools/perf/Documentation/perf-trace.txt b/tools/perf/Documentation/perf-trace.txt index 07065efa60e0..8879299cd9df 100644 --- a/tools/perf/Documentation/perf-trace.txt +++ b/tools/perf/Documentation/perf-trace.txt @@ -8,21 +8,56 @@ perf-trace - Read perf.data (created by perf record) and display trace output SYNOPSIS -------- [verse] -'perf trace' [-i <file> | --input=file] symbol_name +'perf trace' {record <script> | report <script> [args] } DESCRIPTION ----------- This command reads the input file and displays the trace recorded. +There are several variants of perf trace: + + 'perf trace' to see a detailed trace of the workload that was + recorded. + + You can also run a set of pre-canned scripts that aggregate and + summarize the raw trace data in various ways (the list of scripts is + available via 'perf trace -l'). The following variants allow you to + record and run those scripts: + + 'perf trace record <script>' to record the events required for 'perf + trace report'. <script> is the name displayed in the output of + 'perf trace --list' i.e. the actual script name minus any language + extension. + + 'perf trace report <script>' to run and display the results of + <script>. <script> is the name displayed in the output of 'perf + trace --list' i.e. the actual script name minus any language + extension. The perf.data output from a previous run of 'perf trace + record <script>' is used and should be present for this command to + succeed. + + See the 'SEE ALSO' section for links to language-specific + information on how to write and run your own trace scripts. + OPTIONS ------- -D:: --dump-raw-trace=:: Display verbose dump of the trace data. --s:: +-L:: +--Latency=:: + Show latency attributes (irqs/preemption disabled, etc). + +-l:: +--list=:: + Display a list of available trace scripts. + +-s ['lang']:: --script=:: Process trace data with the given script ([lang]:script[.ext]). + If the string 'lang' is specified in place of a script name, a + list of supported languages will be displayed instead. -g:: --gen-script=:: @@ -31,4 +66,5 @@ OPTIONS SEE ALSO -------- -linkperf:perf-record[1], linkperf:perf-trace-perl[1] +linkperf:perf-record[1], linkperf:perf-trace-perl[1], +linkperf:perf-trace-python[1] diff --git a/tools/perf/Documentation/perf.txt b/tools/perf/Documentation/perf.txt index 69c832557199..0eeb247dc7d2 100644 --- a/tools/perf/Documentation/perf.txt +++ b/tools/perf/Documentation/perf.txt @@ -12,7 +12,7 @@ SYNOPSIS DESCRIPTION ----------- -Performance counters for Linux are are a new kernel-based subsystem +Performance counters for Linux are a new kernel-based subsystem that provide a framework for all things performance analysis. It covers hardware level (CPU/PMU, Performance Monitoring Unit) features and software features (software counters, tracepoints) as well. diff --git a/tools/perf/Makefile b/tools/perf/Makefile index 23ec66098bdc..54a5b50ff312 100644 --- a/tools/perf/Makefile +++ b/tools/perf/Makefile @@ -237,8 +237,8 @@ lib = lib export prefix bindir sharedir sysconfdir -CC = gcc -AR = ar +CC = $(CROSS_COMPILE)gcc +AR = $(CROSS_COMPILE)ar RM = rm -f TAR = tar FIND = find @@ -250,7 +250,19 @@ PTHREAD_LIBS = -lpthread # explicitly what architecture to check for. Fix this up for yours.. SPARSE_FLAGS = -D__BIG_ENDIAN__ -D__powerpc__ -ifeq ($(shell sh -c "echo 'int foo(void) {char X[2]; return 3;}' | $(CC) -x c -c -Werror -fstack-protector-all - -o /dev/null "$(QUIET_STDERR)" && echo y"), y) +ifeq ($(V), 2) + QUIET_STDERR = ">/dev/null" +else + QUIET_STDERR = ">/dev/null 2>&1" +endif + +BITBUCKET = "/dev/null" + +ifneq ($(shell sh -c "(echo '\#include <stdio.h>'; echo 'int main(void) { return puts(\"hi\"); }') | $(CC) -x c - $(ALL_CFLAGS) -o $(BITBUCKET) "$(QUIET_STDERR)" && echo y"), y) + BITBUCKET = .perf.dev.null +endif + +ifeq ($(shell sh -c "echo 'int foo(void) {char X[2]; return 3;}' | $(CC) -x c -c -Werror -fstack-protector-all - -o $(BITBUCKET) "$(QUIET_STDERR)" && echo y"), y) CFLAGS := $(CFLAGS) -fstack-protector-all endif @@ -274,11 +286,7 @@ SCRIPT_PERL = SCRIPT_SH = TEST_PROGRAMS = -# -# No scripts right now: -# - -# SCRIPT_SH += perf-am.sh +SCRIPT_SH += perf-archive.sh # # No Perl scripts right now: @@ -303,9 +311,6 @@ PROGRAMS += perf # List built-in command $C whose implementation cmd_$C() is not in # builtin-$C.o but is linked in as part of some other command. # -# None right now: -# -# BUILT_INS += perf-init $X # what 'all' will build and 'install' will install, in perfexecdir ALL_PROGRAMS = $(PROGRAMS) $(SCRIPTS) @@ -328,6 +333,7 @@ LIB_FILE=libperf.a LIB_H += ../../include/linux/perf_event.h LIB_H += ../../include/linux/rbtree.h LIB_H += ../../include/linux/list.h +LIB_H += ../../include/linux/hash.h LIB_H += ../../include/linux/stringify.h LIB_H += util/include/linux/bitmap.h LIB_H += util/include/linux/bitops.h @@ -343,23 +349,33 @@ LIB_H += util/include/linux/string.h LIB_H += util/include/linux/types.h LIB_H += util/include/asm/asm-offsets.h LIB_H += util/include/asm/bitops.h +LIB_H += util/include/asm/bug.h LIB_H += util/include/asm/byteorder.h LIB_H += util/include/asm/swab.h LIB_H += util/include/asm/system.h LIB_H += util/include/asm/uaccess.h LIB_H += perf.h +LIB_H += util/cache.h +LIB_H += util/callchain.h +LIB_H += util/build-id.h +LIB_H += util/debug.h LIB_H += util/debugfs.h LIB_H += util/event.h +LIB_H += util/exec_cmd.h LIB_H += util/types.h LIB_H += util/levenshtein.h +LIB_H += util/map.h LIB_H += util/parse-options.h LIB_H += util/parse-events.h LIB_H += util/quote.h LIB_H += util/util.h +LIB_H += util/header.h LIB_H += util/help.h +LIB_H += util/session.h LIB_H += util/strbuf.h LIB_H += util/string.h LIB_H += util/strlist.h +LIB_H += util/svghelper.h LIB_H += util/run-command.h LIB_H += util/sigchain.h LIB_H += util/symbol.h @@ -368,12 +384,13 @@ LIB_H += util/values.h LIB_H += util/sort.h LIB_H += util/hist.h LIB_H += util/thread.h -LIB_H += util/data_map.h +LIB_H += util/trace-event.h LIB_H += util/probe-finder.h LIB_H += util/probe-event.h LIB_OBJS += util/abspath.o LIB_OBJS += util/alias.o +LIB_OBJS += util/build-id.o LIB_OBJS += util/config.o LIB_OBJS += util/ctype.o LIB_OBJS += util/debugfs.o @@ -405,16 +422,17 @@ LIB_OBJS += util/callchain.o LIB_OBJS += util/values.o LIB_OBJS += util/debug.o LIB_OBJS += util/map.o +LIB_OBJS += util/session.o LIB_OBJS += util/thread.o LIB_OBJS += util/trace-event-parse.o LIB_OBJS += util/trace-event-read.o LIB_OBJS += util/trace-event-info.o -LIB_OBJS += util/trace-event-perl.o +LIB_OBJS += util/trace-event-scripting.o LIB_OBJS += util/svghelper.o LIB_OBJS += util/sort.o LIB_OBJS += util/hist.o -LIB_OBJS += util/data_map.o LIB_OBJS += util/probe-event.o +LIB_OBJS += util/util.o BUILTIN_OBJS += builtin-annotate.o @@ -425,9 +443,11 @@ BUILTIN_OBJS += bench/sched-messaging.o BUILTIN_OBJS += bench/sched-pipe.o BUILTIN_OBJS += bench/mem-memcpy.o +BUILTIN_OBJS += builtin-diff.o BUILTIN_OBJS += builtin-help.o BUILTIN_OBJS += builtin-sched.o BUILTIN_OBJS += builtin-buildid-list.o +BUILTIN_OBJS += builtin-buildid-cache.o BUILTIN_OBJS += builtin-list.o BUILTIN_OBJS += builtin-record.o BUILTIN_OBJS += builtin-report.o @@ -437,14 +457,10 @@ BUILTIN_OBJS += builtin-top.o BUILTIN_OBJS += builtin-trace.o BUILTIN_OBJS += builtin-probe.o BUILTIN_OBJS += builtin-kmem.o +BUILTIN_OBJS += builtin-lock.o PERFLIBS = $(LIB_FILE) -ifeq ($(V), 2) - QUIET_STDERR = ">/dev/null" -else - QUIET_STDERR = ">/dev/null 2>&1" -endif # # Platform specific tweaks # @@ -472,53 +488,70 @@ ifeq ($(uname_S),Darwin) PTHREAD_LIBS = endif -ifeq ($(shell sh -c "(echo '\#include <libelf.h>'; echo 'int main(void) { Elf * elf = elf_begin(0, ELF_C_READ, 0); return (long)elf; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o /dev/null $(ALL_LDFLAGS) $(EXTLIBS) "$(QUIET_STDERR)" && echo y"), y) -ifneq ($(shell sh -c "(echo '\#include <gnu/libc-version.h>'; echo 'int main(void) { const char * version = gnu_get_libc_version(); return (long)version; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o /dev/null $(ALL_LDFLAGS) $(EXTLIBS) "$(QUIET_STDERR)" && echo y"), y) +ifeq ($(shell sh -c "(echo '\#include <libelf.h>'; echo 'int main(void) { Elf * elf = elf_begin(0, ELF_C_READ, 0); return (long)elf; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o $(BITBUCKET) $(ALL_LDFLAGS) $(EXTLIBS) "$(QUIET_STDERR)" && echo y"), y) +ifneq ($(shell sh -c "(echo '\#include <gnu/libc-version.h>'; echo 'int main(void) { const char * version = gnu_get_libc_version(); return (long)version; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o $(BITBUCKET) $(ALL_LDFLAGS) $(EXTLIBS) "$(QUIET_STDERR)" && echo y"), y) msg := $(error No gnu/libc-version.h found, please install glibc-dev[el]/glibc-static); endif - ifneq ($(shell sh -c "(echo '\#include <libelf.h>'; echo 'int main(void) { Elf * elf = elf_begin(0, ELF_C_READ_MMAP, 0); return (long)elf; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o /dev/null $(ALL_LDFLAGS) $(EXTLIBS) "$(QUIET_STDERR)" && echo y"), y) + ifneq ($(shell sh -c "(echo '\#include <libelf.h>'; echo 'int main(void) { Elf * elf = elf_begin(0, ELF_C_READ_MMAP, 0); return (long)elf; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o $(BITBUCKET) $(ALL_LDFLAGS) $(EXTLIBS) "$(QUIET_STDERR)" && echo y"), y) BASIC_CFLAGS += -DLIBELF_NO_MMAP endif else msg := $(error No libelf.h/libelf found, please install libelf-dev/elfutils-libelf-devel and glibc-dev[el]); endif -ifneq ($(shell sh -c "(echo '\#include <libdwarf/dwarf.h>'; echo '\#include <libdwarf/libdwarf.h>'; echo 'int main(void) { Dwarf_Debug dbg; Dwarf_Error err; Dwarf_Ranges *rng; dwarf_init(0, DW_DLC_READ, 0, 0, &dbg, &err); dwarf_get_ranges(dbg, 0, &rng, 0, 0, &err); return (long)dbg; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -ldwarf -lelf -o /dev/null $(ALL_LDFLAGS) $(EXTLIBS) "$(QUIET_STDERR)" && echo y"), y) +ifneq ($(shell sh -c "(echo '\#ifndef _MIPS_SZLONG'; echo '\#define _MIPS_SZLONG 0'; echo '\#endif'; echo '\#include <dwarf.h>'; echo '\#include <libdwarf.h>'; echo 'int main(void) { Dwarf_Debug dbg; Dwarf_Error err; Dwarf_Ranges *rng; dwarf_init(0, DW_DLC_READ, 0, 0, &dbg, &err); dwarf_get_ranges(dbg, 0, &rng, 0, 0, &err); return (long)dbg; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -I/usr/include/libdwarf -ldwarf -lelf -o $(BITBUCKET) $(ALL_LDFLAGS) $(EXTLIBS) "$(QUIET_STDERR)" && echo y"), y) msg := $(warning No libdwarf.h found or old libdwarf.h found, disables dwarf support. Please install libdwarf-dev/libdwarf-devel >= 20081231); BASIC_CFLAGS += -DNO_LIBDWARF else + BASIC_CFLAGS += -I/usr/include/libdwarf EXTLIBS += -lelf -ldwarf LIB_OBJS += util/probe-finder.o endif +ifndef NO_LIBPERL PERL_EMBED_LDOPTS = `perl -MExtUtils::Embed -e ldopts 2>/dev/null` PERL_EMBED_CCOPTS = `perl -MExtUtils::Embed -e ccopts 2>/dev/null` +endif -ifneq ($(shell sh -c "(echo '\#include <EXTERN.h>'; echo '\#include <perl.h>'; echo 'int main(void) { perl_alloc(); return 0; }') | $(CC) -x c - $(PERL_EMBED_CCOPTS) -o /dev/null $(PERL_EMBED_LDOPTS) > /dev/null 2>&1 && echo y"), y) +ifneq ($(shell sh -c "(echo '\#include <EXTERN.h>'; echo '\#include <perl.h>'; echo 'int main(void) { perl_alloc(); return 0; }') | $(CC) -x c - $(PERL_EMBED_CCOPTS) -o $(BITBUCKET) $(PERL_EMBED_LDOPTS) > /dev/null 2>&1 && echo y"), y) BASIC_CFLAGS += -DNO_LIBPERL else ALL_LDFLAGS += $(PERL_EMBED_LDOPTS) + LIB_OBJS += util/scripting-engines/trace-event-perl.o LIB_OBJS += scripts/perl/Perf-Trace-Util/Context.o endif +ifndef NO_LIBPYTHON +PYTHON_EMBED_LDOPTS = `python-config --ldflags 2>/dev/null` +PYTHON_EMBED_CCOPTS = `python-config --cflags 2>/dev/null` +endif + +ifneq ($(shell sh -c "(echo '\#include <Python.h>'; echo 'int main(void) { Py_Initialize(); return 0; }') | $(CC) -x c - $(PYTHON_EMBED_CCOPTS) -o /dev/null $(PYTHON_EMBED_LDOPTS) > /dev/null 2>&1 && echo y"), y) + BASIC_CFLAGS += -DNO_LIBPYTHON +else + ALL_LDFLAGS += $(PYTHON_EMBED_LDOPTS) + LIB_OBJS += util/scripting-engines/trace-event-python.o + LIB_OBJS += scripts/python/Perf-Trace-Util/Context.o +endif + ifdef NO_DEMANGLE BASIC_CFLAGS += -DNO_DEMANGLE else - has_bfd := $(shell sh -c "(echo '\#include <bfd.h>'; echo 'int main(void) { bfd_demangle(0, 0, 0); return 0; }') | $(CC) -x c - $(ALL_CFLAGS) -o /dev/null $(ALL_LDFLAGS) $(EXTLIBS) -lbfd "$(QUIET_STDERR)" && echo y") + has_bfd := $(shell sh -c "(echo '\#include <bfd.h>'; echo 'int main(void) { bfd_demangle(0, 0, 0); return 0; }') | $(CC) -x c - $(ALL_CFLAGS) -o $(BITBUCKET) $(ALL_LDFLAGS) $(EXTLIBS) -lbfd "$(QUIET_STDERR)" && echo y") ifeq ($(has_bfd),y) EXTLIBS += -lbfd else - has_bfd_iberty := $(shell sh -c "(echo '\#include <bfd.h>'; echo 'int main(void) { bfd_demangle(0, 0, 0); return 0; }') | $(CC) -x c - $(ALL_CFLAGS) -o /dev/null $(ALL_LDFLAGS) $(EXTLIBS) -lbfd -liberty "$(QUIET_STDERR)" && echo y") + has_bfd_iberty := $(shell sh -c "(echo '\#include <bfd.h>'; echo 'int main(void) { bfd_demangle(0, 0, 0); return 0; }') | $(CC) -x c - $(ALL_CFLAGS) -o $(BITBUCKET) $(ALL_LDFLAGS) $(EXTLIBS) -lbfd -liberty "$(QUIET_STDERR)" && echo y") ifeq ($(has_bfd_iberty),y) EXTLIBS += -lbfd -liberty else - has_bfd_iberty_z := $(shell sh -c "(echo '\#include <bfd.h>'; echo 'int main(void) { bfd_demangle(0, 0, 0); return 0; }') | $(CC) -x c - $(ALL_CFLAGS) -o /dev/null $(ALL_LDFLAGS) $(EXTLIBS) -lbfd -liberty -lz "$(QUIET_STDERR)" && echo y") + has_bfd_iberty_z := $(shell sh -c "(echo '\#include <bfd.h>'; echo 'int main(void) { bfd_demangle(0, 0, 0); return 0; }') | $(CC) -x c - $(ALL_CFLAGS) -o $(BITBUCKET) $(ALL_LDFLAGS) $(EXTLIBS) -lbfd -liberty -lz "$(QUIET_STDERR)" && echo y") ifeq ($(has_bfd_iberty_z),y) EXTLIBS += -lbfd -liberty -lz else - has_cplus_demangle := $(shell sh -c "(echo 'extern char *cplus_demangle(const char *, int);'; echo 'int main(void) { cplus_demangle(0, 0); return 0; }') | $(CC) -x c - $(ALL_CFLAGS) -o /dev/null $(ALL_LDFLAGS) $(EXTLIBS) -liberty "$(QUIET_STDERR)" && echo y") + has_cplus_demangle := $(shell sh -c "(echo 'extern char *cplus_demangle(const char *, int);'; echo 'int main(void) { cplus_demangle(0, 0); return 0; }') | $(CC) -x c - $(ALL_CFLAGS) -o $(BITBUCKET) $(ALL_LDFLAGS) $(EXTLIBS) -liberty "$(QUIET_STDERR)" && echo y") ifeq ($(has_cplus_demangle),y) EXTLIBS += -liberty BASIC_CFLAGS += -DHAVE_CPLUS_DEMANGLE @@ -766,7 +799,7 @@ export TAR INSTALL DESTDIR SHELL_PATH SHELL = $(SHELL_PATH) -all:: shell_compatibility_test $(ALL_PROGRAMS) $(BUILT_INS) $(OTHER_PROGRAMS) PERF-BUILD-OPTIONS +all:: .perf.dev.null shell_compatibility_test $(ALL_PROGRAMS) $(BUILT_INS) $(OTHER_PROGRAMS) PERF-BUILD-OPTIONS ifneq (,$X) $(foreach p,$(patsubst %$X,%,$(filter %$X,$(ALL_PROGRAMS) $(BUILT_INS) perf$X)), test '$p' -ef '$p$X' || $(RM) '$p';) endif @@ -873,12 +906,18 @@ util/hweight.o: ../../lib/hweight.c PERF-CFLAGS util/find_next_bit.o: ../../lib/find_next_bit.c PERF-CFLAGS $(QUIET_CC)$(CC) -o util/find_next_bit.o -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $< -util/trace-event-perl.o: util/trace-event-perl.c PERF-CFLAGS - $(QUIET_CC)$(CC) -o util/trace-event-perl.o -c $(ALL_CFLAGS) $(PERL_EMBED_CCOPTS) -Wno-redundant-decls -Wno-strict-prototypes -Wno-unused-parameter -Wno-shadow $< +util/scripting-engines/trace-event-perl.o: util/scripting-engines/trace-event-perl.c PERF-CFLAGS + $(QUIET_CC)$(CC) -o util/scripting-engines/trace-event-perl.o -c $(ALL_CFLAGS) $(PERL_EMBED_CCOPTS) -Wno-redundant-decls -Wno-strict-prototypes -Wno-unused-parameter -Wno-shadow $< scripts/perl/Perf-Trace-Util/Context.o: scripts/perl/Perf-Trace-Util/Context.c PERF-CFLAGS $(QUIET_CC)$(CC) -o scripts/perl/Perf-Trace-Util/Context.o -c $(ALL_CFLAGS) $(PERL_EMBED_CCOPTS) -Wno-redundant-decls -Wno-strict-prototypes -Wno-unused-parameter -Wno-nested-externs $< +util/scripting-engines/trace-event-python.o: util/scripting-engines/trace-event-python.c PERF-CFLAGS + $(QUIET_CC)$(CC) -o util/scripting-engines/trace-event-python.o -c $(ALL_CFLAGS) $(PYTHON_EMBED_CCOPTS) -Wno-redundant-decls -Wno-strict-prototypes -Wno-unused-parameter -Wno-shadow $< + +scripts/python/Perf-Trace-Util/Context.o: scripts/python/Perf-Trace-Util/Context.c PERF-CFLAGS + $(QUIET_CC)$(CC) -o scripts/python/Perf-Trace-Util/Context.o -c $(ALL_CFLAGS) $(PYTHON_EMBED_CCOPTS) -Wno-redundant-decls -Wno-strict-prototypes -Wno-unused-parameter -Wno-nested-externs $< + perf-%$X: %.o $(PERFLIBS) $(QUIET_LINK)$(CC) $(ALL_CFLAGS) -o $@ $(ALL_LDFLAGS) $(filter %.o,$^) $(LIBS) @@ -988,11 +1027,16 @@ install: all $(INSTALL) perf$X '$(DESTDIR_SQ)$(bindir_SQ)' $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl/Perf-Trace-Util/lib/Perf/Trace' $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl/bin' + $(INSTALL) perf-archive -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)' $(INSTALL) scripts/perl/Perf-Trace-Util/lib/Perf/Trace/* -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl/Perf-Trace-Util/lib/Perf/Trace' $(INSTALL) scripts/perl/*.pl -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl' $(INSTALL) scripts/perl/bin/* -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl/bin' - $(INSTALL) scripts/perl/Perf-Trace-Util/Makefile.PL -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl/Perf-Trace-Util' - $(INSTALL) scripts/perl/Perf-Trace-Util/README -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl/Perf-Trace-Util' + $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/python/Perf-Trace-Util/lib/Perf/Trace' + $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/python/bin' + $(INSTALL) scripts/python/Perf-Trace-Util/lib/Perf/Trace/* -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/python/Perf-Trace-Util/lib/Perf/Trace' + $(INSTALL) scripts/python/*.py -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/python' + $(INSTALL) scripts/python/bin/* -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/python/bin' + ifdef BUILT_INS $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)' $(INSTALL) $(BUILT_INS) '$(DESTDIR_SQ)$(perfexec_instdir_SQ)' @@ -1095,6 +1139,11 @@ clean: .PHONY: .FORCE-PERF-VERSION-FILE TAGS tags cscope .FORCE-PERF-CFLAGS .PHONY: .FORCE-PERF-BUILD-OPTIONS +.perf.dev.null: + touch .perf.dev.null + +.INTERMEDIATE: .perf.dev.null + ### Make sure built-ins do not have dups and listed in perf.c # check-builtins:: diff --git a/tools/perf/bench/sched-messaging.c b/tools/perf/bench/sched-messaging.c index 605a2a959aa8..81cee78181fa 100644 --- a/tools/perf/bench/sched-messaging.c +++ b/tools/perf/bench/sched-messaging.c @@ -1,6 +1,6 @@ /* * - * builtin-bench-messaging.c + * sched-messaging.c * * messaging: Benchmark for scheduler and IPC mechanisms * @@ -320,10 +320,12 @@ int bench_sched_messaging(int argc, const char **argv, num_groups, num_groups * 2 * num_fds, thread_mode ? "threads" : "processes"); printf(" %14s: %lu.%03lu [sec]\n", "Total time", - diff.tv_sec, diff.tv_usec/1000); + diff.tv_sec, + (unsigned long) (diff.tv_usec/1000)); break; case BENCH_FORMAT_SIMPLE: - printf("%lu.%03lu\n", diff.tv_sec, diff.tv_usec/1000); + printf("%lu.%03lu\n", diff.tv_sec, + (unsigned long) (diff.tv_usec/1000)); break; default: /* reaching here is something disaster */ diff --git a/tools/perf/bench/sched-pipe.c b/tools/perf/bench/sched-pipe.c index 238185f97977..4f77c7c27640 100644 --- a/tools/perf/bench/sched-pipe.c +++ b/tools/perf/bench/sched-pipe.c @@ -1,6 +1,6 @@ /* * - * builtin-bench-pipe.c + * sched-pipe.c * * pipe: Benchmark for pipe() * @@ -87,7 +87,8 @@ int bench_sched_pipe(int argc, const char **argv, if (pid) { retpid = waitpid(pid, &wait_stat, 0); assert((retpid == pid) && WIFEXITED(wait_stat)); - return 0; + } else { + exit(0); } switch (bench_format) { @@ -99,7 +100,8 @@ int bench_sched_pipe(int argc, const char **argv, result_usec += diff.tv_usec; printf(" %14s: %lu.%03lu [sec]\n\n", "Total time", - diff.tv_sec, diff.tv_usec/1000); + diff.tv_sec, + (unsigned long) (diff.tv_usec/1000)); printf(" %14lf usecs/op\n", (double)result_usec / (double)loops); @@ -110,7 +112,8 @@ int bench_sched_pipe(int argc, const char **argv, case BENCH_FORMAT_SIMPLE: printf("%lu.%03lu\n", - diff.tv_sec, diff.tv_usec / 1000); + diff.tv_sec, + (unsigned long) (diff.tv_usec / 1000)); break; default: diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c index 0bf2e8f9af57..5ec5de995872 100644 --- a/tools/perf/builtin-annotate.c +++ b/tools/perf/builtin-annotate.c @@ -25,7 +25,7 @@ #include "util/thread.h" #include "util/sort.h" #include "util/hist.h" -#include "util/data_map.h" +#include "util/session.h" static char const *input_name = "perf.data"; @@ -51,39 +51,22 @@ struct sym_priv { struct sym_ext *ext; }; -static struct symbol_conf symbol_conf = { - .priv_size = sizeof(struct sym_priv), - .try_vmlinux_path = true, -}; - static const char *sym_hist_filter; -static int symbol_filter(struct map *map __used, struct symbol *sym) +static int sym__alloc_hist(struct symbol *self) { - if (sym_hist_filter == NULL || - strcmp(sym->name, sym_hist_filter) == 0) { - struct sym_priv *priv = symbol__priv(sym); - const int size = (sizeof(*priv->hist) + - (sym->end - sym->start) * sizeof(u64)); + struct sym_priv *priv = symbol__priv(self); + const int size = (sizeof(*priv->hist) + + (self->end - self->start) * sizeof(u64)); - priv->hist = malloc(size); - if (priv->hist) - memset(priv->hist, 0, size); - return 0; - } - /* - * FIXME: We should really filter it out, as we don't want to go thru symbols - * we're not interested, and if a DSO ends up with no symbols, delete it too, - * but right now the kernel loading routines in symbol.c bail out if no symbols - * are found, fix it later. - */ - return 0; + priv->hist = zalloc(size); + return priv->hist == NULL ? -1 : 0; } /* * collect histogram counts */ -static void hist_hit(struct hist_entry *he, u64 ip) +static int annotate__hist_hit(struct hist_entry *he, u64 ip) { unsigned int sym_size, offset; struct symbol *sym = he->sym; @@ -93,81 +76,127 @@ static void hist_hit(struct hist_entry *he, u64 ip) he->count++; if (!sym || !he->map) - return; + return 0; priv = symbol__priv(sym); - if (!priv->hist) - return; + if (priv->hist == NULL && sym__alloc_hist(sym) < 0) + return -ENOMEM; sym_size = sym->end - sym->start; offset = ip - sym->start; - if (verbose) - fprintf(stderr, "%s: ip=%Lx\n", __func__, - he->map->unmap_ip(he->map, ip)); + pr_debug3("%s: ip=%#Lx\n", __func__, he->map->unmap_ip(he->map, ip)); if (offset >= sym_size) - return; + return 0; h = priv->hist; h->sum++; h->ip[offset]++; - if (verbose >= 3) - printf("%p %s: count++ [ip: %p, %08Lx] => %Ld\n", - (void *)(unsigned long)he->sym->start, - he->sym->name, - (void *)(unsigned long)ip, ip - he->sym->start, - h->ip[offset]); + pr_debug3("%#Lx %s: count++ [ip: %#Lx, %#Lx] => %Ld\n", he->sym->start, + he->sym->name, ip, ip - he->sym->start, h->ip[offset]); + return 0; } -static int hist_entry__add(struct addr_location *al, u64 count) +static int perf_session__add_hist_entry(struct perf_session *self, + struct addr_location *al, u64 count) { bool hit; - struct hist_entry *he = __hist_entry__add(al, NULL, count, &hit); + struct hist_entry *he; + + if (sym_hist_filter != NULL && + (al->sym == NULL || strcmp(sym_hist_filter, al->sym->name) != 0)) { + /* We're only interested in a symbol named sym_hist_filter */ + if (al->sym != NULL) { + rb_erase(&al->sym->rb_node, + &al->map->dso->symbols[al->map->type]); + symbol__delete(al->sym); + } + return 0; + } + + he = __perf_session__add_hist_entry(self, al, NULL, count, &hit); if (he == NULL) return -ENOMEM; - hist_hit(he, al->addr); - return 0; + + return annotate__hist_hit(he, al->addr); } -static int process_sample_event(event_t *event) +static int process_sample_event(event_t *event, struct perf_session *session) { struct addr_location al; - dump_printf("(IP, %d): %d: %p\n", event->header.misc, - event->ip.pid, (void *)(long)event->ip.ip); + dump_printf("(IP, %d): %d: %#Lx\n", event->header.misc, + event->ip.pid, event->ip.ip); - if (event__preprocess_sample(event, &al, symbol_filter) < 0) { - fprintf(stderr, "problem processing %d event, skipping it.\n", - event->header.type); + if (event__preprocess_sample(event, session, &al, NULL) < 0) { + pr_warning("problem processing %d event, skipping it.\n", + event->header.type); return -1; } - if (hist_entry__add(&al, 1)) { - fprintf(stderr, "problem incrementing symbol count, " - "skipping event\n"); + if (!al.filtered && perf_session__add_hist_entry(session, &al, 1)) { + pr_warning("problem incrementing symbol count, " + "skipping event\n"); return -1; } return 0; } -static int parse_line(FILE *file, struct hist_entry *he, u64 len) +struct objdump_line { + struct list_head node; + s64 offset; + char *line; +}; + +static struct objdump_line *objdump_line__new(s64 offset, char *line) +{ + struct objdump_line *self = malloc(sizeof(*self)); + + if (self != NULL) { + self->offset = offset; + self->line = line; + } + + return self; +} + +static void objdump_line__free(struct objdump_line *self) +{ + free(self->line); + free(self); +} + +static void objdump__add_line(struct list_head *head, struct objdump_line *line) +{ + list_add_tail(&line->node, head); +} + +static struct objdump_line *objdump__get_next_ip_line(struct list_head *head, + struct objdump_line *pos) +{ + list_for_each_entry_continue(pos, head, node) + if (pos->offset >= 0) + return pos; + + return NULL; +} + +static int parse_line(FILE *file, struct hist_entry *he, + struct list_head *head) { struct symbol *sym = he->sym; + struct objdump_line *objdump_line; char *line = NULL, *tmp, *tmp2; - static const char *prev_line; - static const char *prev_color; - unsigned int offset; size_t line_len; - u64 start; - s64 line_ip; - int ret; + s64 line_ip, offset = -1; char *c; if (getline(&line, &line_len, file) < 0) return -1; + if (!line) return -1; @@ -176,8 +205,6 @@ static int parse_line(FILE *file, struct hist_entry *he, u64 len) *c = 0; line_ip = -1; - offset = 0; - ret = -2; /* * Strip leading spaces: @@ -198,9 +225,30 @@ static int parse_line(FILE *file, struct hist_entry *he, u64 len) line_ip = -1; } - start = he->map->unmap_ip(he->map, sym->start); - if (line_ip != -1) { + u64 start = map__rip_2objdump(he->map, sym->start); + offset = line_ip - start; + } + + objdump_line = objdump_line__new(offset, line); + if (objdump_line == NULL) { + free(line); + return -1; + } + objdump__add_line(head, objdump_line); + + return 0; +} + +static int objdump_line__print(struct objdump_line *self, + struct list_head *head, + struct hist_entry *he, u64 len) +{ + struct symbol *sym = he->sym; + static const char *prev_line; + static const char *prev_color; + + if (self->offset != -1) { const char *path = NULL; unsigned int hits = 0; double percent = 0.0; @@ -208,15 +256,22 @@ static int parse_line(FILE *file, struct hist_entry *he, u64 len) struct sym_priv *priv = symbol__priv(sym); struct sym_ext *sym_ext = priv->ext; struct sym_hist *h = priv->hist; + s64 offset = self->offset; + struct objdump_line *next = objdump__get_next_ip_line(head, self); + + while (offset < (s64)len && + (next == NULL || offset < next->offset)) { + if (sym_ext) { + if (path == NULL) + path = sym_ext[offset].path; + percent += sym_ext[offset].percent; + } else + hits += h->ip[offset]; + + ++offset; + } - offset = line_ip - start; - if (offset < len) - hits = h->ip[offset]; - - if (offset < len && sym_ext) { - path = sym_ext[offset].path; - percent = sym_ext[offset].percent; - } else if (h->sum) + if (sym_ext == NULL && h->sum) percent = 100.0 * hits / h->sum; color = get_percent_color(percent); @@ -237,12 +292,12 @@ static int parse_line(FILE *file, struct hist_entry *he, u64 len) color_fprintf(stdout, color, " %7.2f", percent); printf(" : "); - color_fprintf(stdout, PERF_COLOR_BLUE, "%s\n", line); + color_fprintf(stdout, PERF_COLOR_BLUE, "%s\n", self->line); } else { - if (!*line) + if (!*self->line) printf(" :\n"); else - printf(" : %s\n", line); + printf(" : %s\n", self->line); } return 0; @@ -368,6 +423,20 @@ static void print_summary(const char *filename) } } +static void hist_entry__print_hits(struct hist_entry *self) +{ + struct symbol *sym = self->sym; + struct sym_priv *priv = symbol__priv(sym); + struct sym_hist *h = priv->hist; + u64 len = sym->end - sym->start, offset; + + for (offset = 0; offset < len; ++offset) + if (h->ip[offset] != 0) + printf("%*Lx: %Lu\n", BITS_PER_LONG / 2, + sym->start + offset, h->ip[offset]); + printf("%*s: %Lu\n", BITS_PER_LONG / 2, "h->sum", h->sum); +} + static void annotate_sym(struct hist_entry *he) { struct map *map = he->map; @@ -377,15 +446,15 @@ static void annotate_sym(struct hist_entry *he) u64 len; char command[PATH_MAX*2]; FILE *file; + LIST_HEAD(head); + struct objdump_line *pos, *n; if (!filename) return; - if (verbose) - fprintf(stderr, "%s: filename=%s, sym=%s, start=%Lx, end=%Lx\n", - __func__, filename, sym->name, - map->unmap_ip(map, sym->start), - map->unmap_ip(map, sym->end)); + pr_debug("%s: filename=%s, sym=%s, start=%#Lx, end=%#Lx\n", __func__, + filename, sym->name, map->unmap_ip(map, sym->start), + map->unmap_ip(map, sym->end)); if (full_paths) d_filename = filename; @@ -408,7 +477,8 @@ static void annotate_sym(struct hist_entry *he) dso, dso->long_name, sym, sym->name); sprintf(command, "objdump --start-address=0x%016Lx --stop-address=0x%016Lx -dS %s|grep -v %s", - map->unmap_ip(map, sym->start), map->unmap_ip(map, sym->end), + map__rip_2objdump(map, sym->start), + map__rip_2objdump(map, sym->end), filename, filename); if (verbose >= 3) @@ -419,20 +489,30 @@ static void annotate_sym(struct hist_entry *he) return; while (!feof(file)) { - if (parse_line(file, he, len) < 0) + if (parse_line(file, he, &head) < 0) break; } pclose(file); + + if (verbose) + hist_entry__print_hits(he); + + list_for_each_entry_safe(pos, n, &head, node) { + objdump_line__print(pos, &head, he, len); + list_del(&pos->node); + objdump_line__free(pos); + } + if (print_line) free_source_line(he, len); } -static void find_annotations(void) +static void perf_session__find_annotations(struct perf_session *self) { struct rb_node *nd; - for (nd = rb_first(&output_hists); nd; nd = rb_next(nd)) { + for (nd = rb_first(&self->hists); nd; nd = rb_next(nd)) { struct hist_entry *he = rb_entry(nd, struct hist_entry, rb_node); struct sym_priv *priv; @@ -453,42 +533,42 @@ static void find_annotations(void) } } -static struct perf_file_handler file_handler = { - .process_sample_event = process_sample_event, - .process_mmap_event = event__process_mmap, - .process_comm_event = event__process_comm, - .process_fork_event = event__process_task, +static struct perf_event_ops event_ops = { + .sample = process_sample_event, + .mmap = event__process_mmap, + .comm = event__process_comm, + .fork = event__process_task, }; static int __cmd_annotate(void) { - struct perf_header *header; - struct thread *idle; int ret; + struct perf_session *session; - idle = register_idle_thread(); - register_perf_file_handler(&file_handler); + session = perf_session__new(input_name, O_RDONLY, force); + if (session == NULL) + return -ENOMEM; - ret = mmap_dispatch_perf_file(&header, input_name, 0, 0, - &event__cwdlen, &event__cwd); + ret = perf_session__process_events(session, &event_ops); if (ret) - return ret; + goto out_delete; if (dump_trace) { event__print_totals(); - return 0; + goto out_delete; } if (verbose > 3) - threads__fprintf(stdout); + perf_session__fprintf(session, stdout); if (verbose > 2) dsos__fprintf(stdout); - collapse__resort(); - output__resort(event__total[0]); - - find_annotations(); + perf_session__collapse_resort(session); + perf_session__output_resort(session, session->event_total[0]); + perf_session__find_annotations(session); +out_delete: + perf_session__delete(session); return ret; } @@ -519,29 +599,17 @@ static const struct option options[] = { OPT_END() }; -static void setup_sorting(void) +int cmd_annotate(int argc, const char **argv, const char *prefix __used) { - char *tmp, *tok, *str = strdup(sort_order); - - for (tok = strtok_r(str, ", ", &tmp); - tok; tok = strtok_r(NULL, ", ", &tmp)) { - if (sort_dimension__add(tok) < 0) { - error("Unknown --sort key: `%s'", tok); - usage_with_options(annotate_usage, options); - } - } + argc = parse_options(argc, argv, options, annotate_usage, 0); - free(str); -} + symbol_conf.priv_size = sizeof(struct sym_priv); + symbol_conf.try_vmlinux_path = true; -int cmd_annotate(int argc, const char **argv, const char *prefix __used) -{ - if (symbol__init(&symbol_conf) < 0) + if (symbol__init() < 0) return -1; - argc = parse_options(argc, argv, options, annotate_usage, 0); - - setup_sorting(); + setup_sorting(annotate_usage, options); if (argc) { /* @@ -557,9 +625,8 @@ int cmd_annotate(int argc, const char **argv, const char *prefix __used) setup_pager(); if (field_sep && *field_sep == '.') { - fputs("'.' is the only non valid --field-separator argument\n", - stderr); - exit(129); + pr_err("'.' is the only non valid --field-separator argument\n"); + return -1; } return __cmd_annotate(); diff --git a/tools/perf/builtin-bench.c b/tools/perf/builtin-bench.c index e043eb83092a..46996774e559 100644 --- a/tools/perf/builtin-bench.c +++ b/tools/perf/builtin-bench.c @@ -31,6 +31,9 @@ struct bench_suite { const char *summary; int (*fn)(int, const char **, const char *); }; + \ +/* sentinel: easy for help */ +#define suite_all { "all", "test all suite (pseudo suite)", NULL } static struct bench_suite sched_suites[] = { { "messaging", @@ -39,6 +42,7 @@ static struct bench_suite sched_suites[] = { { "pipe", "Flood of communication over pipe() between two processes", bench_sched_pipe }, + suite_all, { NULL, NULL, NULL } @@ -48,6 +52,7 @@ static struct bench_suite mem_suites[] = { { "memcpy", "Simple memory copy in various ways", bench_mem_memcpy }, + suite_all, { NULL, NULL, NULL } @@ -66,6 +71,9 @@ static struct bench_subsys subsystems[] = { { "mem", "memory access performance", mem_suites }, + { "all", /* sentinel: easy for help */ + "test all subsystem (pseudo subsystem)", + NULL }, { NULL, NULL, NULL } @@ -75,11 +83,11 @@ static void dump_suites(int subsys_index) { int i; - printf("List of available suites for %s...\n\n", + printf("# List of available suites for %s...\n\n", subsystems[subsys_index].name); for (i = 0; subsystems[subsys_index].suites[i].name; i++) - printf("\t%s: %s\n", + printf("%14s: %s\n", subsystems[subsys_index].suites[i].name, subsystems[subsys_index].suites[i].summary); @@ -110,10 +118,10 @@ static void print_usage(void) printf("\t%s\n", bench_usage[i]); printf("\n"); - printf("List of available subsystems...\n\n"); + printf("# List of available subsystems...\n\n"); for (i = 0; subsystems[i].name; i++) - printf("\t%s: %s\n", + printf("%14s: %s\n", subsystems[i].name, subsystems[i].summary); printf("\n"); } @@ -131,6 +139,37 @@ static int bench_str2int(char *str) return BENCH_FORMAT_UNKNOWN; } +static void all_suite(struct bench_subsys *subsys) /* FROM HERE */ +{ + int i; + const char *argv[2]; + struct bench_suite *suites = subsys->suites; + + argv[1] = NULL; + /* + * TODO: + * preparing preset parameters for + * embedded, ordinary PC, HPC, etc... + * will be helpful + */ + for (i = 0; suites[i].fn; i++) { + printf("# Running %s/%s benchmark...\n", + subsys->name, + suites[i].name); + + argv[1] = suites[i].name; + suites[i].fn(1, argv, NULL); + printf("\n"); + } +} + +static void all_subsystem(void) +{ + int i; + for (i = 0; subsystems[i].suites; i++) + all_suite(&subsystems[i]); +} + int cmd_bench(int argc, const char **argv, const char *prefix __used) { int i, j, status = 0; @@ -155,6 +194,11 @@ int cmd_bench(int argc, const char **argv, const char *prefix __used) goto end; } + if (!strcmp(argv[0], "all")) { + all_subsystem(); + goto end; + } + for (i = 0; subsystems[i].name; i++) { if (strcmp(subsystems[i].name, argv[0])) continue; @@ -165,6 +209,11 @@ int cmd_bench(int argc, const char **argv, const char *prefix __used) goto end; } + if (!strcmp(argv[1], "all")) { + all_suite(&subsystems[i]); + goto end; + } + for (j = 0; subsystems[i].suites[j].name; j++) { if (strcmp(subsystems[i].suites[j].name, argv[1])) continue; diff --git a/tools/perf/builtin-buildid-cache.c b/tools/perf/builtin-buildid-cache.c new file mode 100644 index 000000000000..30a05f552c96 --- /dev/null +++ b/tools/perf/builtin-buildid-cache.c @@ -0,0 +1,133 @@ +/* + * builtin-buildid-cache.c + * + * Builtin buildid-cache command: Manages build-id cache + * + * Copyright (C) 2010, Red Hat Inc. + * Copyright (C) 2010, Arnaldo Carvalho de Melo <acme@redhat.com> + */ +#include "builtin.h" +#include "perf.h" +#include "util/cache.h" +#include "util/debug.h" +#include "util/header.h" +#include "util/parse-options.h" +#include "util/strlist.h" +#include "util/symbol.h" + +static char const *add_name_list_str, *remove_name_list_str; + +static const char * const buildid_cache_usage[] = { + "perf buildid-cache [<options>]", + NULL +}; + +static const struct option buildid_cache_options[] = { + OPT_STRING('a', "add", &add_name_list_str, + "file list", "file(s) to add"), + OPT_STRING('r', "remove", &remove_name_list_str, "file list", + "file(s) to remove"), + OPT_BOOLEAN('v', "verbose", &verbose, "be more verbose"), + OPT_END() +}; + +static int build_id_cache__add_file(const char *filename, const char *debugdir) +{ + char sbuild_id[BUILD_ID_SIZE * 2 + 1]; + u8 build_id[BUILD_ID_SIZE]; + int err; + + if (filename__read_build_id(filename, &build_id, sizeof(build_id)) < 0) { + pr_debug("Couldn't read a build-id in %s\n", filename); + return -1; + } + + build_id__sprintf(build_id, sizeof(build_id), sbuild_id); + err = build_id_cache__add_s(sbuild_id, debugdir, filename, false); + if (verbose) + pr_info("Adding %s %s: %s\n", sbuild_id, filename, + err ? "FAIL" : "Ok"); + return err; +} + +static int build_id_cache__remove_file(const char *filename __used, + const char *debugdir __used) +{ + u8 build_id[BUILD_ID_SIZE]; + char sbuild_id[BUILD_ID_SIZE * 2 + 1]; + + int err; + + if (filename__read_build_id(filename, &build_id, sizeof(build_id)) < 0) { + pr_debug("Couldn't read a build-id in %s\n", filename); + return -1; + } + + build_id__sprintf(build_id, sizeof(build_id), sbuild_id); + err = build_id_cache__remove_s(sbuild_id, debugdir); + if (verbose) + pr_info("Removing %s %s: %s\n", sbuild_id, filename, + err ? "FAIL" : "Ok"); + + return err; +} + +static int __cmd_buildid_cache(void) +{ + struct strlist *list; + struct str_node *pos; + char debugdir[PATH_MAX]; + + snprintf(debugdir, sizeof(debugdir), "%s/%s", getenv("HOME"), + DEBUG_CACHE_DIR); + + if (add_name_list_str) { + list = strlist__new(true, add_name_list_str); + if (list) { + strlist__for_each(pos, list) + if (build_id_cache__add_file(pos->s, debugdir)) { + if (errno == EEXIST) { + pr_debug("%s already in the cache\n", + pos->s); + continue; + } + pr_warning("Couldn't add %s: %s\n", + pos->s, strerror(errno)); + } + + strlist__delete(list); + } + } + + if (remove_name_list_str) { + list = strlist__new(true, remove_name_list_str); + if (list) { + strlist__for_each(pos, list) + if (build_id_cache__remove_file(pos->s, debugdir)) { + if (errno == ENOENT) { + pr_debug("%s wasn't in the cache\n", + pos->s); + continue; + } + pr_warning("Couldn't remove %s: %s\n", + pos->s, strerror(errno)); + } + + strlist__delete(list); + } + } + + return 0; +} + +int cmd_buildid_cache(int argc, const char **argv, const char *prefix __used) +{ + argc = parse_options(argc, argv, buildid_cache_options, + buildid_cache_usage, 0); + + if (symbol__init() < 0) + return -1; + + setup_pager(); + return __cmd_buildid_cache(); +} diff --git a/tools/perf/builtin-buildid-list.c b/tools/perf/builtin-buildid-list.c index 7dee9d19ab7a..d0675c02f81e 100644 --- a/tools/perf/builtin-buildid-list.c +++ b/tools/perf/builtin-buildid-list.c @@ -8,22 +8,24 @@ */ #include "builtin.h" #include "perf.h" +#include "util/build-id.h" #include "util/cache.h" -#include "util/data_map.h" #include "util/debug.h" -#include "util/header.h" #include "util/parse-options.h" +#include "util/session.h" #include "util/symbol.h" static char const *input_name = "perf.data"; static int force; +static bool with_hits; -static const char *const buildid_list_usage[] = { - "perf report [<options>]", +static const char * const buildid_list_usage[] = { + "perf buildid-list [<options>]", NULL }; static const struct option options[] = { + OPT_BOOLEAN('H', "with-hits", &with_hits, "Show only DSOs with hits"), OPT_STRING('i', "input", &input_name, "file", "input file name"), OPT_BOOLEAN('f', "force", &force, "don't complain, do it"), @@ -32,79 +34,21 @@ static const struct option options[] = { OPT_END() }; -static int perf_file_section__process_buildids(struct perf_file_section *self, - int feat, int fd) -{ - if (feat != HEADER_BUILD_ID) - return 0; - - if (lseek(fd, self->offset, SEEK_SET) < 0) { - pr_warning("Failed to lseek to %Ld offset for buildids!\n", - self->offset); - return -1; - } - - if (perf_header__read_build_ids(fd, self->offset, self->size)) { - pr_warning("Failed to read buildids!\n"); - return -1; - } - - return 0; -} - static int __cmd_buildid_list(void) { int err = -1; - struct perf_header *header; - struct perf_file_header f_header; - struct stat input_stat; - int input = open(input_name, O_RDONLY); - - if (input < 0) { - pr_err("failed to open file: %s", input_name); - if (!strcmp(input_name, "perf.data")) - pr_err(" (try 'perf record' first)"); - pr_err("\n"); - goto out; - } - - err = fstat(input, &input_stat); - if (err < 0) { - perror("failed to stat file"); - goto out_close; - } - - if (!force && input_stat.st_uid && (input_stat.st_uid != geteuid())) { - pr_err("file %s not owned by current user or root\n", - input_name); - goto out_close; - } + struct perf_session *session; - if (!input_stat.st_size) { - pr_info("zero-sized file, nothing to do!\n"); - goto out_close; - } - - err = -1; - header = perf_header__new(); - if (header == NULL) - goto out_close; - - if (perf_file_header__read(&f_header, header, input) < 0) { - pr_warning("incompatible file format"); - goto out_close; - } + session = perf_session__new(input_name, O_RDONLY, force); + if (session == NULL) + return -1; - err = perf_header__process_sections(header, input, - perf_file_section__process_buildids); + if (with_hits) + perf_session__process_events(session, &build_id__mark_dso_hit_ops); - if (err < 0) - goto out_close; + dsos__fprintf_buildid(stdout, with_hits); - dsos__fprintf_buildid(stdout); -out_close: - close(input); -out: + perf_session__delete(session); return err; } diff --git a/tools/perf/builtin-diff.c b/tools/perf/builtin-diff.c new file mode 100644 index 000000000000..18b3f505f9db --- /dev/null +++ b/tools/perf/builtin-diff.c @@ -0,0 +1,239 @@ +/* + * builtin-diff.c + * + * Builtin diff command: Analyze two perf.data input files, look up and read + * DSOs and symbol information, sort them and produce a diff. + */ +#include "builtin.h" + +#include "util/debug.h" +#include "util/event.h" +#include "util/hist.h" +#include "util/session.h" +#include "util/sort.h" +#include "util/symbol.h" +#include "util/util.h" + +#include <stdlib.h> + +static char const *input_old = "perf.data.old", + *input_new = "perf.data"; +static char diff__default_sort_order[] = "dso,symbol"; +static int force; +static bool show_displacement; + +static int perf_session__add_hist_entry(struct perf_session *self, + struct addr_location *al, u64 count) +{ + bool hit; + struct hist_entry *he = __perf_session__add_hist_entry(self, al, NULL, + count, &hit); + if (he == NULL) + return -ENOMEM; + + if (hit) + he->count += count; + + return 0; +} + +static int diff__process_sample_event(event_t *event, struct perf_session *session) +{ + struct addr_location al; + struct sample_data data = { .period = 1, }; + + dump_printf("(IP, %d): %d: %#Lx\n", event->header.misc, + event->ip.pid, event->ip.ip); + + if (event__preprocess_sample(event, session, &al, NULL) < 0) { + pr_warning("problem processing %d event, skipping it.\n", + event->header.type); + return -1; + } + + if (al.filtered || al.sym == NULL) + return 0; + + event__parse_sample(event, session->sample_type, &data); + + if (perf_session__add_hist_entry(session, &al, data.period)) { + pr_warning("problem incrementing symbol count, skipping event\n"); + return -1; + } + + session->events_stats.total += data.period; + return 0; +} + +static struct perf_event_ops event_ops = { + .sample = diff__process_sample_event, + .mmap = event__process_mmap, + .comm = event__process_comm, + .exit = event__process_task, + .fork = event__process_task, + .lost = event__process_lost, +}; + +static void perf_session__insert_hist_entry_by_name(struct rb_root *root, + struct hist_entry *he) +{ + struct rb_node **p = &root->rb_node; + struct rb_node *parent = NULL; + struct hist_entry *iter; + + while (*p != NULL) { + parent = *p; + iter = rb_entry(parent, struct hist_entry, rb_node); + if (hist_entry__cmp(he, iter) < 0) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + } + + rb_link_node(&he->rb_node, parent, p); + rb_insert_color(&he->rb_node, root); +} + +static void perf_session__resort_hist_entries(struct perf_session *self) +{ + unsigned long position = 1; + struct rb_root tmp = RB_ROOT; + struct rb_node *next = rb_first(&self->hists); + + while (next != NULL) { + struct hist_entry *n = rb_entry(next, struct hist_entry, rb_node); + + next = rb_next(&n->rb_node); + rb_erase(&n->rb_node, &self->hists); + n->position = position++; + perf_session__insert_hist_entry_by_name(&tmp, n); + } + + self->hists = tmp; +} + +static void perf_session__set_hist_entries_positions(struct perf_session *self) +{ + perf_session__output_resort(self, self->events_stats.total); + perf_session__resort_hist_entries(self); +} + +static struct hist_entry * +perf_session__find_hist_entry(struct perf_session *self, + struct hist_entry *he) +{ + struct rb_node *n = self->hists.rb_node; + + while (n) { + struct hist_entry *iter = rb_entry(n, struct hist_entry, rb_node); + int64_t cmp = hist_entry__cmp(he, iter); + + if (cmp < 0) + n = n->rb_left; + else if (cmp > 0) + n = n->rb_right; + else + return iter; + } + + return NULL; +} + +static void perf_session__match_hists(struct perf_session *old_session, + struct perf_session *new_session) +{ + struct rb_node *nd; + + for (nd = rb_first(&new_session->hists); nd; nd = rb_next(nd)) { + struct hist_entry *pos = rb_entry(nd, struct hist_entry, rb_node); + pos->pair = perf_session__find_hist_entry(old_session, pos); + } +} + +static int __cmd_diff(void) +{ + int ret, i; + struct perf_session *session[2]; + + session[0] = perf_session__new(input_old, O_RDONLY, force); + session[1] = perf_session__new(input_new, O_RDONLY, force); + if (session[0] == NULL || session[1] == NULL) + return -ENOMEM; + + for (i = 0; i < 2; ++i) { + ret = perf_session__process_events(session[i], &event_ops); + if (ret) + goto out_delete; + } + + perf_session__output_resort(session[1], session[1]->events_stats.total); + if (show_displacement) + perf_session__set_hist_entries_positions(session[0]); + + perf_session__match_hists(session[0], session[1]); + perf_session__fprintf_hists(session[1], session[0], + show_displacement, stdout); +out_delete: + for (i = 0; i < 2; ++i) + perf_session__delete(session[i]); + return ret; +} + +static const char * const diff_usage[] = { + "perf diff [<options>] [old_file] [new_file]", + NULL, +}; + +static const struct option options[] = { + OPT_BOOLEAN('v', "verbose", &verbose, + "be more verbose (show symbol address, etc)"), + OPT_BOOLEAN('m', "displacement", &show_displacement, + "Show position displacement relative to baseline"), + OPT_BOOLEAN('D', "dump-raw-trace", &dump_trace, + "dump raw trace in ASCII"), + OPT_BOOLEAN('f', "force", &force, "don't complain, do it"), + OPT_BOOLEAN('m', "modules", &symbol_conf.use_modules, + "load module symbols - WARNING: use only with -k and LIVE kernel"), + OPT_BOOLEAN('P', "full-paths", &symbol_conf.full_paths, + "Don't shorten the pathnames taking into account the cwd"), + OPT_STRING('d', "dsos", &symbol_conf.dso_list_str, "dso[,dso...]", + "only consider symbols in these dsos"), + OPT_STRING('C', "comms", &symbol_conf.comm_list_str, "comm[,comm...]", + "only consider symbols in these comms"), + OPT_STRING('S', "symbols", &symbol_conf.sym_list_str, "symbol[,symbol...]", + "only consider these symbols"), + OPT_STRING('s', "sort", &sort_order, "key[,key2...]", + "sort by key(s): pid, comm, dso, symbol, parent"), + OPT_STRING('t', "field-separator", &symbol_conf.field_sep, "separator", + "separator for columns, no spaces will be added between " + "columns '.' is reserved."), + OPT_END() +}; + +int cmd_diff(int argc, const char **argv, const char *prefix __used) +{ + sort_order = diff__default_sort_order; + argc = parse_options(argc, argv, options, diff_usage, 0); + if (argc) { + if (argc > 2) + usage_with_options(diff_usage, options); + if (argc == 2) { + input_old = argv[0]; + input_new = argv[1]; + } else + input_new = argv[0]; + } + + symbol_conf.exclude_other = false; + if (symbol__init() < 0) + return -1; + + setup_sorting(diff_usage, options); + setup_pager(); + + sort_entry__setup_elide(&sort_dso, symbol_conf.dso_list, "dso", NULL); + sort_entry__setup_elide(&sort_comm, symbol_conf.comm_list, "comm", NULL); + sort_entry__setup_elide(&sort_sym, symbol_conf.sym_list, "symbol", NULL); + + return __cmd_diff(); +} diff --git a/tools/perf/builtin-help.c b/tools/perf/builtin-help.c index 9f810b17c25c..215b584007b1 100644 --- a/tools/perf/builtin-help.c +++ b/tools/perf/builtin-help.c @@ -286,8 +286,7 @@ void list_common_cmds_help(void) puts(" The most commonly used perf commands are:"); for (i = 0; i < ARRAY_SIZE(common_cmds); i++) { - printf(" %s ", common_cmds[i].name); - mput_char(' ', longest - strlen(common_cmds[i].name)); + printf(" %-*s ", longest, common_cmds[i].name); puts(common_cmds[i].help); } } @@ -314,8 +313,6 @@ static const char *cmd_to_page(const char *perf_cmd) return "perf"; else if (!prefixcmp(perf_cmd, "perf")) return perf_cmd; - else if (is_perf_command(perf_cmd)) - return prepend("perf-", perf_cmd); else return prepend("perf-", perf_cmd); } diff --git a/tools/perf/builtin-kmem.c b/tools/perf/builtin-kmem.c index 047fef74bd52..924a9518931a 100644 --- a/tools/perf/builtin-kmem.c +++ b/tools/perf/builtin-kmem.c @@ -6,12 +6,12 @@ #include "util/symbol.h" #include "util/thread.h" #include "util/header.h" +#include "util/session.h" #include "util/parse-options.h" #include "util/trace-event.h" #include "util/debug.h" -#include "util/data_map.h" #include <linux/rbtree.h> @@ -20,9 +20,6 @@ typedef int (*sort_fn_t)(struct alloc_stat *, struct alloc_stat *); static char const *input_name = "perf.data"; -static struct perf_header *header; -static u64 sample_type; - static int alloc_flag; static int caller_flag; @@ -57,11 +54,6 @@ static struct rb_root root_caller_sorted; static unsigned long total_requested, total_allocated; static unsigned long nr_allocs, nr_cross_allocs; -struct raw_event_sample { - u32 size; - char data[0]; -}; - #define PATH_SYS_NODE "/sys/devices/system/node" static void init_cpunode_map(void) @@ -100,23 +92,18 @@ static void setup_cpunode_map(void) if (!dir1) return; - while (true) { - dent1 = readdir(dir1); - if (!dent1) - break; - - if (sscanf(dent1->d_name, "node%u", &mem) < 1) + while ((dent1 = readdir(dir1)) != NULL) { + if (dent1->d_type != DT_DIR || + sscanf(dent1->d_name, "node%u", &mem) < 1) continue; snprintf(buf, PATH_MAX, "%s/%s", PATH_SYS_NODE, dent1->d_name); dir2 = opendir(buf); if (!dir2) continue; - while (true) { - dent2 = readdir(dir2); - if (!dent2) - break; - if (sscanf(dent2->d_name, "cpu%u", &cpu) < 1) + while ((dent2 = readdir(dir2)) != NULL) { + if (dent2->d_type != DT_LNK || + sscanf(dent2->d_name, "cpu%u", &cpu) < 1) continue; cpunode_map[cpu] = mem; } @@ -145,7 +132,7 @@ static void insert_alloc_stat(unsigned long call_site, unsigned long ptr, if (data && data->ptr == ptr) { data->hit++; data->bytes_req += bytes_req; - data->bytes_alloc += bytes_req; + data->bytes_alloc += bytes_alloc; } else { data = malloc(sizeof(*data)); if (!data) @@ -185,7 +172,7 @@ static void insert_caller_stat(unsigned long call_site, if (data && data->call_site == call_site) { data->hit++; data->bytes_req += bytes_req; - data->bytes_alloc += bytes_req; + data->bytes_alloc += bytes_alloc; } else { data = malloc(sizeof(*data)); if (!data) @@ -201,7 +188,7 @@ static void insert_caller_stat(unsigned long call_site, } } -static void process_alloc_event(struct raw_event_sample *raw, +static void process_alloc_event(void *data, struct event *event, int cpu, u64 timestamp __used, @@ -214,10 +201,10 @@ static void process_alloc_event(struct raw_event_sample *raw, int bytes_alloc; int node1, node2; - ptr = raw_field_value(event, "ptr", raw->data); - call_site = raw_field_value(event, "call_site", raw->data); - bytes_req = raw_field_value(event, "bytes_req", raw->data); - bytes_alloc = raw_field_value(event, "bytes_alloc", raw->data); + ptr = raw_field_value(event, "ptr", data); + call_site = raw_field_value(event, "call_site", data); + bytes_req = raw_field_value(event, "bytes_req", data); + bytes_alloc = raw_field_value(event, "bytes_alloc", data); insert_alloc_stat(call_site, ptr, bytes_req, bytes_alloc, cpu); insert_caller_stat(call_site, bytes_req, bytes_alloc); @@ -227,7 +214,7 @@ static void process_alloc_event(struct raw_event_sample *raw, if (node) { node1 = cpunode_map[cpu]; - node2 = raw_field_value(event, "node", raw->data); + node2 = raw_field_value(event, "node", data); if (node1 != node2) nr_cross_allocs++; } @@ -262,7 +249,7 @@ static struct alloc_stat *search_alloc_stat(unsigned long ptr, return NULL; } -static void process_free_event(struct raw_event_sample *raw, +static void process_free_event(void *data, struct event *event, int cpu, u64 timestamp __used, @@ -271,7 +258,7 @@ static void process_free_event(struct raw_event_sample *raw, unsigned long ptr; struct alloc_stat *s_alloc, *s_caller; - ptr = raw_field_value(event, "ptr", raw->data); + ptr = raw_field_value(event, "ptr", data); s_alloc = search_alloc_stat(ptr, 0, &root_alloc_stat, ptr_cmp); if (!s_alloc) @@ -289,66 +276,50 @@ static void process_free_event(struct raw_event_sample *raw, } static void -process_raw_event(event_t *raw_event __used, void *more_data, +process_raw_event(event_t *raw_event __used, void *data, int cpu, u64 timestamp, struct thread *thread) { - struct raw_event_sample *raw = more_data; struct event *event; int type; - type = trace_parse_common_type(raw->data); + type = trace_parse_common_type(data); event = trace_find_event(type); if (!strcmp(event->name, "kmalloc") || !strcmp(event->name, "kmem_cache_alloc")) { - process_alloc_event(raw, event, cpu, timestamp, thread, 0); + process_alloc_event(data, event, cpu, timestamp, thread, 0); return; } if (!strcmp(event->name, "kmalloc_node") || !strcmp(event->name, "kmem_cache_alloc_node")) { - process_alloc_event(raw, event, cpu, timestamp, thread, 1); + process_alloc_event(data, event, cpu, timestamp, thread, 1); return; } if (!strcmp(event->name, "kfree") || !strcmp(event->name, "kmem_cache_free")) { - process_free_event(raw, event, cpu, timestamp, thread); + process_free_event(data, event, cpu, timestamp, thread); return; } } -static int process_sample_event(event_t *event) +static int process_sample_event(event_t *event, struct perf_session *session) { - u64 ip = event->ip.ip; - u64 timestamp = -1; - u32 cpu = -1; - u64 period = 1; - void *more_data = event->ip.__more_data; - struct thread *thread = threads__findnew(event->ip.pid); - - if (sample_type & PERF_SAMPLE_TIME) { - timestamp = *(u64 *)more_data; - more_data += sizeof(u64); - } + struct sample_data data; + struct thread *thread; - if (sample_type & PERF_SAMPLE_CPU) { - cpu = *(u32 *)more_data; - more_data += sizeof(u32); - more_data += sizeof(u32); /* reserved */ - } + memset(&data, 0, sizeof(data)); + data.time = -1; + data.cpu = -1; + data.period = 1; - if (sample_type & PERF_SAMPLE_PERIOD) { - period = *(u64 *)more_data; - more_data += sizeof(u64); - } + event__parse_sample(event, session->sample_type, &data); - dump_printf("(IP, %d): %d/%d: %p period: %Ld\n", - event->header.misc, - event->ip.pid, event->ip.tid, - (void *)(long)ip, - (long long)period); + dump_printf("(IP, %d): %d/%d: %#Lx period: %Ld\n", event->header.misc, + data.pid, data.tid, data.ip, data.period); + thread = perf_session__findnew(session, event->ip.pid); if (thread == NULL) { pr_debug("problem processing %d event, skipping it.\n", event->header.type); @@ -357,40 +328,17 @@ static int process_sample_event(event_t *event) dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid); - process_raw_event(event, more_data, cpu, timestamp, thread); + process_raw_event(event, data.raw_data, data.cpu, + data.time, thread); return 0; } -static int sample_type_check(u64 type) -{ - sample_type = type; - - if (!(sample_type & PERF_SAMPLE_RAW)) { - fprintf(stderr, - "No trace sample to read. Did you call perf record " - "without -R?"); - return -1; - } - - return 0; -} - -static struct perf_file_handler file_handler = { - .process_sample_event = process_sample_event, - .process_comm_event = event__process_comm, - .sample_type_check = sample_type_check, +static struct perf_event_ops event_ops = { + .sample = process_sample_event, + .comm = event__process_comm, }; -static int read_events(void) -{ - register_idle_thread(); - register_perf_file_handler(&file_handler); - - return mmap_dispatch_perf_file(&header, input_name, 0, 0, - &event__cwdlen, &event__cwd); -} - static double fragmentation(unsigned long n_req, unsigned long n_alloc) { if (n_alloc == 0) @@ -399,13 +347,14 @@ static double fragmentation(unsigned long n_req, unsigned long n_alloc) return 100.0 - (100.0 * n_req / n_alloc); } -static void __print_result(struct rb_root *root, int n_lines, int is_caller) +static void __print_result(struct rb_root *root, struct perf_session *session, + int n_lines, int is_caller) { struct rb_node *next; printf("%.102s\n", graph_dotted_line); printf(" %-34s |", is_caller ? "Callsite": "Alloc Ptr"); - printf(" Total_alloc/Per | Total_req/Per | Hit | Ping-pong | Frag\n"); + printf(" Total_alloc/Per | Total_req/Per | Hit | Ping-pong | Frag\n"); printf("%.102s\n", graph_dotted_line); next = rb_first(root); @@ -420,7 +369,7 @@ static void __print_result(struct rb_root *root, int n_lines, int is_caller) if (is_caller) { addr = data->call_site; if (!raw_ip) - sym = thread__find_function(kthread, addr, NULL); + sym = map_groups__find_function(&session->kmaps, addr, NULL); } else addr = data->ptr; @@ -431,7 +380,7 @@ static void __print_result(struct rb_root *root, int n_lines, int is_caller) snprintf(buf, sizeof(buf), "%#Lx", addr); printf(" %-34s |", buf); - printf(" %9llu/%-5lu | %9llu/%-5lu | %6lu | %8lu | %6.3f%%\n", + printf(" %9llu/%-5lu | %9llu/%-5lu | %8lu | %8lu | %6.3f%%\n", (unsigned long long)data->bytes_alloc, (unsigned long)data->bytes_alloc / data->hit, (unsigned long long)data->bytes_req, @@ -461,12 +410,12 @@ static void print_summary(void) printf("Cross CPU allocations: %lu/%lu\n", nr_cross_allocs, nr_allocs); } -static void print_result(void) +static void print_result(struct perf_session *session) { if (caller_flag) - __print_result(&root_caller_sorted, caller_lines, 1); + __print_result(&root_caller_sorted, session, caller_lines, 1); if (alloc_flag) - __print_result(&root_alloc_sorted, alloc_lines, 0); + __print_result(&root_alloc_sorted, session, alloc_lines, 0); print_summary(); } @@ -534,16 +483,27 @@ static void sort_result(void) static int __cmd_kmem(void) { + int err = -EINVAL; + struct perf_session *session = perf_session__new(input_name, O_RDONLY, 0); + if (session == NULL) + return -ENOMEM; + + if (!perf_session__has_traces(session, "kmem record")) + goto out_delete; + setup_pager(); - read_events(); + err = perf_session__process_events(session, &event_ops); + if (err != 0) + goto out_delete; sort_result(); - print_result(); - - return 0; + print_result(session); +out_delete: + perf_session__delete(session); + return err; } static const char * const kmem_usage[] = { - "perf kmem [<options>] {record}", + "perf kmem [<options>] {record|stat}", NULL }; @@ -703,18 +663,17 @@ static int parse_sort_opt(const struct option *opt __used, return 0; } -static int parse_stat_opt(const struct option *opt __used, - const char *arg, int unset __used) +static int parse_caller_opt(const struct option *opt __used, + const char *arg __used, int unset __used) { - if (!arg) - return -1; + caller_flag = (alloc_flag + 1); + return 0; +} - if (strcmp(arg, "alloc") == 0) - alloc_flag = (caller_flag + 1); - else if (strcmp(arg, "caller") == 0) - caller_flag = (alloc_flag + 1); - else - return -1; +static int parse_alloc_opt(const struct option *opt __used, + const char *arg __used, int unset __used) +{ + alloc_flag = (caller_flag + 1); return 0; } @@ -739,14 +698,17 @@ static int parse_line_opt(const struct option *opt __used, static const struct option kmem_options[] = { OPT_STRING('i', "input", &input_name, "file", "input file name"), - OPT_CALLBACK(0, "stat", NULL, "<alloc>|<caller>", - "stat selector, Pass 'alloc' or 'caller'.", - parse_stat_opt), + OPT_CALLBACK_NOOPT(0, "caller", NULL, NULL, + "show per-callsite statistics", + parse_caller_opt), + OPT_CALLBACK_NOOPT(0, "alloc", NULL, NULL, + "show per-allocation statistics", + parse_alloc_opt), OPT_CALLBACK('s', "sort", NULL, "key[,key2...]", "sort by keys: ptr, call_site, bytes, hit, pingpong, frag", parse_sort_opt), OPT_CALLBACK('l', "line", NULL, "num", - "show n lins", + "show n lines", parse_line_opt), OPT_BOOLEAN(0, "raw-ip", &raw_ip, "show raw ip instead of symbol"), OPT_END() @@ -786,22 +748,27 @@ static int __cmd_record(int argc, const char **argv) int cmd_kmem(int argc, const char **argv, const char *prefix __used) { - symbol__init(0); - argc = parse_options(argc, argv, kmem_options, kmem_usage, 0); - if (argc && !strncmp(argv[0], "rec", 3)) - return __cmd_record(argc, argv); - else if (argc) + if (!argc) usage_with_options(kmem_usage, kmem_options); - if (list_empty(&caller_sort)) - setup_sorting(&caller_sort, default_sort_order); - if (list_empty(&alloc_sort)) - setup_sorting(&alloc_sort, default_sort_order); + symbol__init(); - setup_cpunode_map(); + if (!strncmp(argv[0], "rec", 3)) { + return __cmd_record(argc, argv); + } else if (!strcmp(argv[0], "stat")) { + setup_cpunode_map(); + + if (list_empty(&caller_sort)) + setup_sorting(&caller_sort, default_sort_order); + if (list_empty(&alloc_sort)) + setup_sorting(&alloc_sort, default_sort_order); - return __cmd_kmem(); + return __cmd_kmem(); + } else + usage_with_options(kmem_usage, kmem_options); + + return 0; } diff --git a/tools/perf/builtin-lock.c b/tools/perf/builtin-lock.c new file mode 100644 index 000000000000..fb9ab2ad3f92 --- /dev/null +++ b/tools/perf/builtin-lock.c @@ -0,0 +1,678 @@ +#include "builtin.h" +#include "perf.h" + +#include "util/util.h" +#include "util/cache.h" +#include "util/symbol.h" +#include "util/thread.h" +#include "util/header.h" + +#include "util/parse-options.h" +#include "util/trace-event.h" + +#include "util/debug.h" +#include "util/session.h" + +#include <sys/types.h> +#include <sys/prctl.h> +#include <semaphore.h> +#include <pthread.h> +#include <math.h> +#include <limits.h> + +#include <linux/list.h> +#include <linux/hash.h> + +/* based on kernel/lockdep.c */ +#define LOCKHASH_BITS 12 +#define LOCKHASH_SIZE (1UL << LOCKHASH_BITS) + +static struct list_head lockhash_table[LOCKHASH_SIZE]; + +#define __lockhashfn(key) hash_long((unsigned long)key, LOCKHASH_BITS) +#define lockhashentry(key) (lockhash_table + __lockhashfn((key))) + +#define LOCK_STATE_UNLOCKED 0 /* initial state */ +#define LOCK_STATE_LOCKED 1 + +struct lock_stat { + struct list_head hash_entry; + struct rb_node rb; /* used for sorting */ + + /* + * FIXME: raw_field_value() returns unsigned long long, + * so address of lockdep_map should be dealed as 64bit. + * Is there more better solution? + */ + void *addr; /* address of lockdep_map, used as ID */ + char *name; /* for strcpy(), we cannot use const */ + + int state; + u64 prev_event_time; /* timestamp of previous event */ + + unsigned int nr_acquired; + unsigned int nr_acquire; + unsigned int nr_contended; + unsigned int nr_release; + + /* these times are in nano sec. */ + u64 wait_time_total; + u64 wait_time_min; + u64 wait_time_max; +}; + +/* build simple key function one is bigger than two */ +#define SINGLE_KEY(member) \ + static int lock_stat_key_ ## member(struct lock_stat *one, \ + struct lock_stat *two) \ + { \ + return one->member > two->member; \ + } + +SINGLE_KEY(nr_acquired) +SINGLE_KEY(nr_contended) +SINGLE_KEY(wait_time_total) +SINGLE_KEY(wait_time_min) +SINGLE_KEY(wait_time_max) + +struct lock_key { + /* + * name: the value for specify by user + * this should be simpler than raw name of member + * e.g. nr_acquired -> acquired, wait_time_total -> wait_total + */ + const char *name; + int (*key)(struct lock_stat*, struct lock_stat*); +}; + +static const char *sort_key = "acquired"; + +static int (*compare)(struct lock_stat *, struct lock_stat *); + +static struct rb_root result; /* place to store sorted data */ + +#define DEF_KEY_LOCK(name, fn_suffix) \ + { #name, lock_stat_key_ ## fn_suffix } +struct lock_key keys[] = { + DEF_KEY_LOCK(acquired, nr_acquired), + DEF_KEY_LOCK(contended, nr_contended), + DEF_KEY_LOCK(wait_total, wait_time_total), + DEF_KEY_LOCK(wait_min, wait_time_min), + DEF_KEY_LOCK(wait_max, wait_time_max), + + /* extra comparisons much complicated should be here */ + + { NULL, NULL } +}; + +static void select_key(void) +{ + int i; + + for (i = 0; keys[i].name; i++) { + if (!strcmp(keys[i].name, sort_key)) { + compare = keys[i].key; + return; + } + } + + die("Unknown compare key:%s\n", sort_key); +} + +static void insert_to_result(struct lock_stat *st, + int (*bigger)(struct lock_stat *, struct lock_stat *)) +{ + struct rb_node **rb = &result.rb_node; + struct rb_node *parent = NULL; + struct lock_stat *p; + + while (*rb) { + p = container_of(*rb, struct lock_stat, rb); + parent = *rb; + + if (bigger(st, p)) + rb = &(*rb)->rb_left; + else + rb = &(*rb)->rb_right; + } + + rb_link_node(&st->rb, parent, rb); + rb_insert_color(&st->rb, &result); +} + +/* returns left most element of result, and erase it */ +static struct lock_stat *pop_from_result(void) +{ + struct rb_node *node = result.rb_node; + + if (!node) + return NULL; + + while (node->rb_left) + node = node->rb_left; + + rb_erase(node, &result); + return container_of(node, struct lock_stat, rb); +} + +static struct lock_stat *lock_stat_findnew(void *addr, const char *name) +{ + struct list_head *entry = lockhashentry(addr); + struct lock_stat *ret, *new; + + list_for_each_entry(ret, entry, hash_entry) { + if (ret->addr == addr) + return ret; + } + + new = zalloc(sizeof(struct lock_stat)); + if (!new) + goto alloc_failed; + + new->addr = addr; + new->name = zalloc(sizeof(char) * strlen(name) + 1); + if (!new->name) + goto alloc_failed; + strcpy(new->name, name); + + /* LOCK_STATE_UNLOCKED == 0 isn't guaranteed forever */ + new->state = LOCK_STATE_UNLOCKED; + new->wait_time_min = ULLONG_MAX; + + list_add(&new->hash_entry, entry); + return new; + +alloc_failed: + die("memory allocation failed\n"); +} + +static char const *input_name = "perf.data"; + +static int profile_cpu = -1; + +struct raw_event_sample { + u32 size; + char data[0]; +}; + +struct trace_acquire_event { + void *addr; + const char *name; +}; + +struct trace_acquired_event { + void *addr; + const char *name; +}; + +struct trace_contended_event { + void *addr; + const char *name; +}; + +struct trace_release_event { + void *addr; + const char *name; +}; + +struct trace_lock_handler { + void (*acquire_event)(struct trace_acquire_event *, + struct event *, + int cpu, + u64 timestamp, + struct thread *thread); + + void (*acquired_event)(struct trace_acquired_event *, + struct event *, + int cpu, + u64 timestamp, + struct thread *thread); + + void (*contended_event)(struct trace_contended_event *, + struct event *, + int cpu, + u64 timestamp, + struct thread *thread); + + void (*release_event)(struct trace_release_event *, + struct event *, + int cpu, + u64 timestamp, + struct thread *thread); +}; + +static void +report_lock_acquire_event(struct trace_acquire_event *acquire_event, + struct event *__event __used, + int cpu __used, + u64 timestamp, + struct thread *thread __used) +{ + struct lock_stat *st; + + st = lock_stat_findnew(acquire_event->addr, acquire_event->name); + + switch (st->state) { + case LOCK_STATE_UNLOCKED: + break; + case LOCK_STATE_LOCKED: + break; + default: + BUG_ON(1); + break; + } + + st->prev_event_time = timestamp; +} + +static void +report_lock_acquired_event(struct trace_acquired_event *acquired_event, + struct event *__event __used, + int cpu __used, + u64 timestamp, + struct thread *thread __used) +{ + struct lock_stat *st; + + st = lock_stat_findnew(acquired_event->addr, acquired_event->name); + + switch (st->state) { + case LOCK_STATE_UNLOCKED: + st->state = LOCK_STATE_LOCKED; + st->nr_acquired++; + break; + case LOCK_STATE_LOCKED: + break; + default: + BUG_ON(1); + break; + } + + st->prev_event_time = timestamp; +} + +static void +report_lock_contended_event(struct trace_contended_event *contended_event, + struct event *__event __used, + int cpu __used, + u64 timestamp, + struct thread *thread __used) +{ + struct lock_stat *st; + + st = lock_stat_findnew(contended_event->addr, contended_event->name); + + switch (st->state) { + case LOCK_STATE_UNLOCKED: + break; + case LOCK_STATE_LOCKED: + st->nr_contended++; + break; + default: + BUG_ON(1); + break; + } + + st->prev_event_time = timestamp; +} + +static void +report_lock_release_event(struct trace_release_event *release_event, + struct event *__event __used, + int cpu __used, + u64 timestamp, + struct thread *thread __used) +{ + struct lock_stat *st; + u64 hold_time; + + st = lock_stat_findnew(release_event->addr, release_event->name); + + switch (st->state) { + case LOCK_STATE_UNLOCKED: + break; + case LOCK_STATE_LOCKED: + st->state = LOCK_STATE_UNLOCKED; + hold_time = timestamp - st->prev_event_time; + + if (timestamp < st->prev_event_time) { + /* terribly, this can happen... */ + goto end; + } + + if (st->wait_time_min > hold_time) + st->wait_time_min = hold_time; + if (st->wait_time_max < hold_time) + st->wait_time_max = hold_time; + st->wait_time_total += hold_time; + + st->nr_release++; + break; + default: + BUG_ON(1); + break; + } + +end: + st->prev_event_time = timestamp; +} + +/* lock oriented handlers */ +/* TODO: handlers for CPU oriented, thread oriented */ +static struct trace_lock_handler report_lock_ops = { + .acquire_event = report_lock_acquire_event, + .acquired_event = report_lock_acquired_event, + .contended_event = report_lock_contended_event, + .release_event = report_lock_release_event, +}; + +static struct trace_lock_handler *trace_handler; + +static void +process_lock_acquire_event(void *data, + struct event *event __used, + int cpu __used, + u64 timestamp __used, + struct thread *thread __used) +{ + struct trace_acquire_event acquire_event; + u64 tmp; /* this is required for casting... */ + + tmp = raw_field_value(event, "lockdep_addr", data); + memcpy(&acquire_event.addr, &tmp, sizeof(void *)); + acquire_event.name = (char *)raw_field_ptr(event, "name", data); + + if (trace_handler->acquire_event) + trace_handler->acquire_event(&acquire_event, event, cpu, timestamp, thread); +} + +static void +process_lock_acquired_event(void *data, + struct event *event __used, + int cpu __used, + u64 timestamp __used, + struct thread *thread __used) +{ + struct trace_acquired_event acquired_event; + u64 tmp; /* this is required for casting... */ + + tmp = raw_field_value(event, "lockdep_addr", data); + memcpy(&acquired_event.addr, &tmp, sizeof(void *)); + acquired_event.name = (char *)raw_field_ptr(event, "name", data); + + if (trace_handler->acquire_event) + trace_handler->acquired_event(&acquired_event, event, cpu, timestamp, thread); +} + +static void +process_lock_contended_event(void *data, + struct event *event __used, + int cpu __used, + u64 timestamp __used, + struct thread *thread __used) +{ + struct trace_contended_event contended_event; + u64 tmp; /* this is required for casting... */ + + tmp = raw_field_value(event, "lockdep_addr", data); + memcpy(&contended_event.addr, &tmp, sizeof(void *)); + contended_event.name = (char *)raw_field_ptr(event, "name", data); + + if (trace_handler->acquire_event) + trace_handler->contended_event(&contended_event, event, cpu, timestamp, thread); +} + +static void +process_lock_release_event(void *data, + struct event *event __used, + int cpu __used, + u64 timestamp __used, + struct thread *thread __used) +{ + struct trace_release_event release_event; + u64 tmp; /* this is required for casting... */ + + tmp = raw_field_value(event, "lockdep_addr", data); + memcpy(&release_event.addr, &tmp, sizeof(void *)); + release_event.name = (char *)raw_field_ptr(event, "name", data); + + if (trace_handler->acquire_event) + trace_handler->release_event(&release_event, event, cpu, timestamp, thread); +} + +static void +process_raw_event(void *data, int cpu, + u64 timestamp, struct thread *thread) +{ + struct event *event; + int type; + + type = trace_parse_common_type(data); + event = trace_find_event(type); + + if (!strcmp(event->name, "lock_acquire")) + process_lock_acquire_event(data, event, cpu, timestamp, thread); + if (!strcmp(event->name, "lock_acquired")) + process_lock_acquired_event(data, event, cpu, timestamp, thread); + if (!strcmp(event->name, "lock_contended")) + process_lock_contended_event(data, event, cpu, timestamp, thread); + if (!strcmp(event->name, "lock_release")) + process_lock_release_event(data, event, cpu, timestamp, thread); +} + +static int process_sample_event(event_t *event, struct perf_session *session) +{ + struct thread *thread; + struct sample_data data; + + bzero(&data, sizeof(struct sample_data)); + event__parse_sample(event, session->sample_type, &data); + thread = perf_session__findnew(session, data.pid); + + if (thread == NULL) { + pr_debug("problem processing %d event, skipping it.\n", + event->header.type); + return -1; + } + + dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid); + + if (profile_cpu != -1 && profile_cpu != (int) data.cpu) + return 0; + + process_raw_event(data.raw_data, data.cpu, data.time, thread); + + return 0; +} + +/* TODO: various way to print, coloring, nano or milli sec */ +static void print_result(void) +{ + struct lock_stat *st; + char cut_name[20]; + + printf("%18s ", "ID"); + printf("%20s ", "Name"); + printf("%10s ", "acquired"); + printf("%10s ", "contended"); + + printf("%15s ", "total wait (ns)"); + printf("%15s ", "max wait (ns)"); + printf("%15s ", "min wait (ns)"); + + printf("\n\n"); + + while ((st = pop_from_result())) { + bzero(cut_name, 20); + + printf("%p ", st->addr); + + if (strlen(st->name) < 16) { + /* output raw name */ + printf("%20s ", st->name); + } else { + strncpy(cut_name, st->name, 16); + cut_name[16] = '.'; + cut_name[17] = '.'; + cut_name[18] = '.'; + cut_name[19] = '\0'; + /* cut off name for saving output style */ + printf("%20s ", cut_name); + } + + printf("%10u ", st->nr_acquired); + printf("%10u ", st->nr_contended); + + printf("%15llu ", st->wait_time_total); + printf("%15llu ", st->wait_time_max); + printf("%15llu ", st->wait_time_min == ULLONG_MAX ? + 0 : st->wait_time_min); + printf("\n"); + } +} + +static void dump_map(void) +{ + unsigned int i; + struct lock_stat *st; + + for (i = 0; i < LOCKHASH_SIZE; i++) { + list_for_each_entry(st, &lockhash_table[i], hash_entry) { + printf("%p: %s\n", st->addr, st->name); + } + } +} + +static struct perf_event_ops eops = { + .sample = process_sample_event, + .comm = event__process_comm, +}; + +static struct perf_session *session; + +static int read_events(void) +{ + session = perf_session__new(input_name, O_RDONLY, 0); + if (!session) + die("Initializing perf session failed\n"); + + return perf_session__process_events(session, &eops); +} + +static void sort_result(void) +{ + unsigned int i; + struct lock_stat *st; + + for (i = 0; i < LOCKHASH_SIZE; i++) { + list_for_each_entry(st, &lockhash_table[i], hash_entry) { + insert_to_result(st, compare); + } + } +} + +static void __cmd_report(void) +{ + setup_pager(); + select_key(); + read_events(); + sort_result(); + print_result(); +} + +static const char * const report_usage[] = { + "perf lock report [<options>]", + NULL +}; + +static const struct option report_options[] = { + OPT_STRING('k', "key", &sort_key, "acquired", + "key for sorting"), + /* TODO: type */ + OPT_END() +}; + +static const char * const lock_usage[] = { + "perf lock [<options>] {record|trace|report}", + NULL +}; + +static const struct option lock_options[] = { + OPT_STRING('i', "input", &input_name, "file", "input file name"), + OPT_BOOLEAN('v', "verbose", &verbose, "be more verbose (show symbol address, etc)"), + OPT_BOOLEAN('D', "dump-raw-trace", &dump_trace, "dump raw trace in ASCII"), + OPT_END() +}; + +static const char *record_args[] = { + "record", + "-a", + "-R", + "-M", + "-f", + "-m", "1024", + "-c", "1", + "-e", "lock:lock_acquire:r", + "-e", "lock:lock_acquired:r", + "-e", "lock:lock_contended:r", + "-e", "lock:lock_release:r", +}; + +static int __cmd_record(int argc, const char **argv) +{ + unsigned int rec_argc, i, j; + const char **rec_argv; + + rec_argc = ARRAY_SIZE(record_args) + argc - 1; + rec_argv = calloc(rec_argc + 1, sizeof(char *)); + + for (i = 0; i < ARRAY_SIZE(record_args); i++) + rec_argv[i] = strdup(record_args[i]); + + for (j = 1; j < (unsigned int)argc; j++, i++) + rec_argv[i] = argv[j]; + + BUG_ON(i != rec_argc); + + return cmd_record(i, rec_argv, NULL); +} + +int cmd_lock(int argc, const char **argv, const char *prefix __used) +{ + unsigned int i; + + symbol__init(); + for (i = 0; i < LOCKHASH_SIZE; i++) + INIT_LIST_HEAD(lockhash_table + i); + + argc = parse_options(argc, argv, lock_options, lock_usage, + PARSE_OPT_STOP_AT_NON_OPTION); + if (!argc) + usage_with_options(lock_usage, lock_options); + + if (!strncmp(argv[0], "rec", 3)) { + return __cmd_record(argc, argv); + } else if (!strncmp(argv[0], "report", 6)) { + trace_handler = &report_lock_ops; + if (argc) { + argc = parse_options(argc, argv, + report_options, report_usage, 0); + if (argc) + usage_with_options(report_usage, report_options); + } + __cmd_report(); + } else if (!strcmp(argv[0], "trace")) { + /* Aliased to 'perf trace' */ + return cmd_trace(argc, argv, prefix); + } else if (!strcmp(argv[0], "map")) { + /* recycling report_lock_ops */ + trace_handler = &report_lock_ops; + setup_pager(); + read_events(); + dump_map(); + } else { + usage_with_options(lock_usage, lock_options); + } + + return 0; +} diff --git a/tools/perf/builtin-probe.c b/tools/perf/builtin-probe.c index a58e11b7ea80..ad47bd4c50ef 100644 --- a/tools/perf/builtin-probe.c +++ b/tools/perf/builtin-probe.c @@ -35,34 +35,34 @@ #include "perf.h" #include "builtin.h" #include "util/util.h" +#include "util/strlist.h" #include "util/event.h" #include "util/debug.h" +#include "util/debugfs.h" +#include "util/symbol.h" +#include "util/thread.h" #include "util/parse-options.h" #include "util/parse-events.h" /* For debugfs_path */ #include "util/probe-finder.h" #include "util/probe-event.h" -/* Default vmlinux search paths */ -#define NR_SEARCH_PATH 3 -const char *default_search_path[NR_SEARCH_PATH] = { -"/lib/modules/%s/build/vmlinux", /* Custom build kernel */ -"/usr/lib/debug/lib/modules/%s/vmlinux", /* Red Hat debuginfo */ -"/boot/vmlinux-debug-%s", /* Ubuntu */ -}; - #define MAX_PATH_LEN 256 #define MAX_PROBES 128 /* Session management structure */ static struct { - char *vmlinux; - char *release; - int need_dwarf; + bool need_dwarf; + bool list_events; + bool force_add; + bool show_lines; int nr_probe; struct probe_point probes[MAX_PROBES]; + struct strlist *dellist; + struct map_groups kmap_groups; + struct map *kmaps[MAP__NR_TYPES]; + struct line_range line_range; } session; -static bool listing; /* Parse an event definition. Note that any error must die. */ static void parse_probe_event(const char *str) @@ -74,11 +74,30 @@ static void parse_probe_event(const char *str) die("Too many probes (> %d) are specified.", MAX_PROBES); /* Parse perf-probe event into probe_point */ - session.need_dwarf = parse_perf_probe_event(str, pp); + parse_perf_probe_event(str, pp, &session.need_dwarf); pr_debug("%d arguments\n", pp->nr_args); } +static void parse_probe_event_argv(int argc, const char **argv) +{ + int i, len; + char *buf; + + /* Bind up rest arguments */ + len = 0; + for (i = 0; i < argc; i++) + len += strlen(argv[i]) + 1; + buf = zalloc(len + 1); + if (!buf) + die("Failed to allocate memory for binding arguments."); + len = 0; + for (i = 0; i < argc; i++) + len += sprintf(&buf[len], "%s ", argv[i]); + parse_probe_event(buf); + free(buf); +} + static int opt_add_probe_event(const struct option *opt __used, const char *str, int unset __used) { @@ -87,41 +106,57 @@ static int opt_add_probe_event(const struct option *opt __used, return 0; } -#ifndef NO_LIBDWARF -static int open_default_vmlinux(void) +static int opt_del_probe_event(const struct option *opt __used, + const char *str, int unset __used) { - struct utsname uts; - char fname[MAX_PATH_LEN]; - int fd, ret, i; - - ret = uname(&uts); - if (ret) { - pr_debug("uname() failed.\n"); - return -errno; + if (str) { + if (!session.dellist) + session.dellist = strlist__new(true, NULL); + strlist__add(session.dellist, str); } - session.release = uts.release; - for (i = 0; i < NR_SEARCH_PATH; i++) { - ret = snprintf(fname, MAX_PATH_LEN, - default_search_path[i], session.release); - if (ret >= MAX_PATH_LEN || ret < 0) { - pr_debug("Filename(%d,%s) is too long.\n", i, - uts.release); - errno = E2BIG; - return -E2BIG; - } - pr_debug("try to open %s\n", fname); - fd = open(fname, O_RDONLY); - if (fd >= 0) - break; + return 0; +} + +/* Currently just checking function name from symbol map */ +static void evaluate_probe_point(struct probe_point *pp) +{ + struct symbol *sym; + sym = map__find_symbol_by_name(session.kmaps[MAP__FUNCTION], + pp->function, NULL); + if (!sym) + die("Kernel symbol \'%s\' not found - probe not added.", + pp->function); +} + +#ifndef NO_LIBDWARF +static int open_vmlinux(void) +{ + if (map__load(session.kmaps[MAP__FUNCTION], NULL) < 0) { + pr_debug("Failed to load kernel map.\n"); + return -EINVAL; } - return fd; + pr_debug("Try to open %s\n", + session.kmaps[MAP__FUNCTION]->dso->long_name); + return open(session.kmaps[MAP__FUNCTION]->dso->long_name, O_RDONLY); +} + +static int opt_show_lines(const struct option *opt __used, + const char *str, int unset __used) +{ + if (str) + parse_line_range_desc(str, &session.line_range); + INIT_LIST_HEAD(&session.line_range.line_list); + session.show_lines = true; + return 0; } #endif static const char * const probe_usage[] = { "perf probe [<options>] 'PROBEDEF' ['PROBEDEF' ...]", "perf probe [<options>] --add 'PROBEDEF' [--add 'PROBEDEF' ...]", + "perf probe [<options>] --del '[GROUP:]EVENT' ...", "perf probe --list", + "perf probe --line 'LINEDESC'", NULL }; @@ -129,19 +164,22 @@ static const struct option options[] = { OPT_BOOLEAN('v', "verbose", &verbose, "be more verbose (show parsed arguments, etc)"), #ifndef NO_LIBDWARF - OPT_STRING('k', "vmlinux", &session.vmlinux, "file", - "vmlinux/module pathname"), + OPT_STRING('k', "vmlinux", &symbol_conf.vmlinux_name, + "file", "vmlinux pathname"), #endif - OPT_BOOLEAN('l', "list", &listing, "list up current probes"), + OPT_BOOLEAN('l', "list", &session.list_events, + "list up current probe events"), + OPT_CALLBACK('d', "del", NULL, "[GROUP:]EVENT", "delete a probe event.", + opt_del_probe_event), OPT_CALLBACK('a', "add", NULL, #ifdef NO_LIBDWARF - "FUNC[+OFFS|%return] [ARG ...]", + "[EVENT=]FUNC[+OFFS|%return] [ARG ...]", #else - "FUNC[+OFFS|%return|:RLN][@SRC]|SRC:ALN [ARG ...]", + "[EVENT=]FUNC[+OFFS|%return|:RLN][@SRC]|SRC:ALN [ARG ...]", #endif "probe point definition, where\n" - "\t\tGRP:\tGroup name (optional)\n" - "\t\tNAME:\tEvent name\n" + "\t\tGROUP:\tGroup name (optional)\n" + "\t\tEVENT:\tEvent name\n" "\t\tFUNC:\tFunction name\n" "\t\tOFFS:\tOffset from function entry (in byte)\n" "\t\t%return:\tPut the probe at function return\n" @@ -155,12 +193,36 @@ static const struct option options[] = { #endif "\t\t\tkprobe-tracer argument format.)\n", opt_add_probe_event), + OPT_BOOLEAN('f', "force", &session.force_add, "forcibly add events" + " with existing name"), +#ifndef NO_LIBDWARF + OPT_CALLBACK('L', "line", NULL, + "FUNC[:RLN[+NUM|:RLN2]]|SRC:ALN[+NUM|:ALN2]", + "Show source code lines.", opt_show_lines), +#endif OPT_END() }; +/* Initialize symbol maps for vmlinux */ +static void init_vmlinux(void) +{ + symbol_conf.sort_by_name = true; + if (symbol_conf.vmlinux_name == NULL) + symbol_conf.try_vmlinux_path = true; + else + pr_debug("Use vmlinux: %s\n", symbol_conf.vmlinux_name); + if (symbol__init() < 0) + die("Failed to init symbol map."); + + map_groups__init(&session.kmap_groups); + if (map_groups__create_kernel_maps(&session.kmap_groups, + session.kmaps) < 0) + die("Failed to create kernel maps."); +} + int cmd_probe(int argc, const char **argv, const char *prefix __used) { - int i, j, ret; + int i, ret; #ifndef NO_LIBDWARF int fd; #endif @@ -168,54 +230,105 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used) argc = parse_options(argc, argv, options, probe_usage, PARSE_OPT_STOP_AT_NON_OPTION); - for (i = 0; i < argc; i++) - parse_probe_event(argv[i]); + if (argc > 0) { + if (strcmp(argv[0], "-") == 0) { + pr_warning(" Error: '-' is not supported.\n"); + usage_with_options(probe_usage, options); + } + parse_probe_event_argv(argc, argv); + } - if ((session.nr_probe == 0 && !listing) || - (session.nr_probe != 0 && listing)) + if ((!session.nr_probe && !session.dellist && !session.list_events && + !session.show_lines)) usage_with_options(probe_usage, options); - if (listing) { + if (debugfs_valid_mountpoint(debugfs_path) < 0) + die("Failed to find debugfs path."); + + if (session.list_events) { + if (session.nr_probe != 0 || session.dellist) { + pr_warning(" Error: Don't use --list with" + " --add/--del.\n"); + usage_with_options(probe_usage, options); + } + if (session.show_lines) { + pr_warning(" Error: Don't use --list with --line.\n"); + usage_with_options(probe_usage, options); + } show_perf_probe_events(); return 0; } +#ifndef NO_LIBDWARF + if (session.show_lines) { + if (session.nr_probe != 0 || session.dellist) { + pr_warning(" Error: Don't use --line with" + " --add/--del.\n"); + usage_with_options(probe_usage, options); + } + init_vmlinux(); + fd = open_vmlinux(); + if (fd < 0) + die("Could not open debuginfo file."); + ret = find_line_range(fd, &session.line_range); + if (ret <= 0) + die("Source line is not found.\n"); + close(fd); + show_line_range(&session.line_range); + return 0; + } +#endif + + if (session.dellist) { + del_trace_kprobe_events(session.dellist); + strlist__delete(session.dellist); + if (session.nr_probe == 0) + return 0; + } + + /* Add probes */ + init_vmlinux(); + if (session.need_dwarf) #ifdef NO_LIBDWARF die("Debuginfo-analysis is not supported"); #else /* !NO_LIBDWARF */ pr_debug("Some probes require debuginfo.\n"); - if (session.vmlinux) - fd = open(session.vmlinux, O_RDONLY); - else - fd = open_default_vmlinux(); + fd = open_vmlinux(); if (fd < 0) { if (session.need_dwarf) - die("Could not open vmlinux/module file."); + die("Could not open debuginfo file."); - pr_warning("Could not open vmlinux/module file." - " Try to use symbols.\n"); + pr_debug("Could not open vmlinux/module file." + " Try to use symbols.\n"); goto end_dwarf; } /* Searching probe points */ - for (j = 0; j < session.nr_probe; j++) { - pp = &session.probes[j]; + for (i = 0; i < session.nr_probe; i++) { + pp = &session.probes[i]; if (pp->found) continue; lseek(fd, SEEK_SET, 0); ret = find_probepoint(fd, pp); - if (ret < 0) { - if (session.need_dwarf) - die("Could not analyze debuginfo."); - - pr_warning("An error occurred in debuginfo analysis. Try to use symbols.\n"); - break; + if (ret > 0) + continue; + if (ret == 0) { /* No error but failed to find probe point. */ + synthesize_perf_probe_point(pp); + die("Probe point '%s' not found. - probe not added.", + pp->probes[0]); + } + /* Error path */ + if (session.need_dwarf) { + if (ret == -ENOENT) + pr_warning("No dwarf info found in the vmlinux - please rebuild with CONFIG_DEBUG_INFO=y.\n"); + die("Could not analyze debuginfo."); } - if (ret == 0) /* No error but failed to find probe point. */ - die("No probe point found."); + pr_debug("An error occurred in debuginfo analysis." + " Try to use symbols.\n"); + break; } close(fd); @@ -223,11 +336,12 @@ end_dwarf: #endif /* !NO_LIBDWARF */ /* Synthesize probes without dwarf */ - for (j = 0; j < session.nr_probe; j++) { - pp = &session.probes[j]; + for (i = 0; i < session.nr_probe; i++) { + pp = &session.probes[i]; if (pp->found) /* This probe is already found. */ continue; + evaluate_probe_point(pp); ret = synthesize_trace_kprobe_event(pp); if (ret == -E2BIG) die("probe point definition becomes too long."); @@ -236,7 +350,8 @@ end_dwarf: } /* Settng up probe points */ - add_trace_kprobe_events(session.probes, session.nr_probe); + add_trace_kprobe_events(session.probes, session.nr_probe, + session.force_add); return 0; } diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index 0e519c667e3a..771533ced6a8 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -5,10 +5,13 @@ * (or a CPU, or a PID) into the perf.data output file - for * later analysis via perf report. */ +#define _FILE_OFFSET_BITS 64 + #include "builtin.h" #include "perf.h" +#include "util/build-id.h" #include "util/util.h" #include "util/parse-options.h" #include "util/parse-events.h" @@ -17,6 +20,7 @@ #include "util/header.h" #include "util/event.h" #include "util/debug.h" +#include "util/session.h" #include "util/symbol.h" #include <unistd.h> @@ -61,8 +65,9 @@ static int nr_poll = 0; static int nr_cpu = 0; static int file_new = 1; +static off_t post_processing_offset; -struct perf_header *header = NULL; +static struct perf_session *session; struct mmap_data { int counter; @@ -110,21 +115,10 @@ static void write_output(void *buf, size_t size) } } -static void write_event(event_t *buf, size_t size) +static int process_synthesized_event(event_t *event, + struct perf_session *self __used) { - /* - * Add it to the list of DSOs, so that when we finish this - * record session we can pick the available build-ids. - */ - if (buf->header.type == PERF_RECORD_MMAP) - dsos__findnew(buf->mmap.filename); - - write_output(buf, size); -} - -static int process_synthesized_event(event_t *event) -{ - write_event(event, event->header.size); + write_output(event, event->header.size); return 0; } @@ -176,14 +170,14 @@ static void mmap_read(struct mmap_data *md) size = md->mask + 1 - (old & md->mask); old += size; - write_event(buf, size); + write_output(buf, size); } buf = &data[old & md->mask]; size = head - old; old += size; - write_event(buf, size); + write_output(buf, size); md->prev = old; mmap_write_tail(md, old); @@ -216,12 +210,12 @@ static struct perf_header_attr *get_header_attr(struct perf_event_attr *a, int n { struct perf_header_attr *h_attr; - if (nr < header->attrs) { - h_attr = header->attr[nr]; + if (nr < session->header.attrs) { + h_attr = session->header.attr[nr]; } else { h_attr = perf_header_attr__new(a); if (h_attr != NULL) - if (perf_header__add_attr(header, h_attr) < 0) { + if (perf_header__add_attr(&session->header, h_attr) < 0) { perf_header_attr__delete(h_attr); h_attr = NULL; } @@ -276,7 +270,7 @@ static void create_counter(int counter, int cpu, pid_t pid) attr->mmap = track; attr->comm = track; - attr->inherit = (cpu < 0) && inherit; + attr->inherit = inherit; attr->disabled = 1; try_again: @@ -393,11 +387,22 @@ static void open_counters(int cpu, pid_t pid) nr_cpu++; } +static int process_buildids(void) +{ + u64 size = lseek(output, 0, SEEK_CUR); + + session->fd = output; + return __perf_session__process_events(session, post_processing_offset, + size - post_processing_offset, + size, &build_id__mark_dso_hit_ops); +} + static void atexit_header(void) { - header->data_size += bytes_written; + session->header.data_size += bytes_written; - perf_header__write(header, output, true); + process_buildids(); + perf_header__write(&session->header, output, true); } static int __cmd_record(int argc, const char **argv) @@ -408,6 +413,9 @@ static int __cmd_record(int argc, const char **argv) int flags; int err; unsigned long waking = 0; + int child_ready_pipe[2], go_pipe[2]; + const bool forks = target_pid == -1 && argc > 0; + char buf; page_size = sysconf(_SC_PAGE_SIZE); nr_cpus = sysconf(_SC_NPROCESSORS_ONLN); @@ -418,11 +426,25 @@ static int __cmd_record(int argc, const char **argv) signal(SIGCHLD, sig_handler); signal(SIGINT, sig_handler); + if (forks && (pipe(child_ready_pipe) < 0 || pipe(go_pipe) < 0)) { + perror("failed to create pipes"); + exit(-1); + } + if (!stat(output_name, &st) && st.st_size) { - if (!force && !append_file) { - fprintf(stderr, "Error, output file %s exists, use -A to append or -f to overwrite.\n", - output_name); - exit(-1); + if (!force) { + if (!append_file) { + pr_err("Error, output file %s exists, use -A " + "to append or -f to overwrite.\n", + output_name); + exit(-1); + } + } else { + char oldname[PATH_MAX]; + snprintf(oldname, sizeof(oldname), "%s.old", + output_name); + unlink(oldname); + rename(output_name, oldname); } } else { append_file = 0; @@ -440,24 +462,24 @@ static int __cmd_record(int argc, const char **argv) exit(-1); } - header = perf_header__new(); - if (header == NULL) { + session = perf_session__new(output_name, O_WRONLY, force); + if (session == NULL) { pr_err("Not enough memory for reading perf file header\n"); return -1; } if (!file_new) { - err = perf_header__read(header, output); + err = perf_header__read(&session->header, output); if (err < 0) return err; } if (raw_samples) { - perf_header__set_feat(header, HEADER_TRACE_INFO); + perf_header__set_feat(&session->header, HEADER_TRACE_INFO); } else { for (i = 0; i < nr_counters; i++) { if (attrs[i].sample_type & PERF_SAMPLE_RAW) { - perf_header__set_feat(header, HEADER_TRACE_INFO); + perf_header__set_feat(&session->header, HEADER_TRACE_INFO); break; } } @@ -465,56 +487,94 @@ static int __cmd_record(int argc, const char **argv) atexit(atexit_header); - if (!system_wide) { - pid = target_pid; - if (pid == -1) - pid = getpid(); + if (forks) { + pid = fork(); + if (pid < 0) { + perror("failed to fork"); + exit(-1); + } - open_counters(profile_cpu, pid); - } else { - if (profile_cpu != -1) { - open_counters(profile_cpu, target_pid); - } else { - for (i = 0; i < nr_cpus; i++) - open_counters(i, target_pid); + if (!pid) { + close(child_ready_pipe[0]); + close(go_pipe[1]); + fcntl(go_pipe[0], F_SETFD, FD_CLOEXEC); + + /* + * Do a dummy execvp to get the PLT entry resolved, + * so we avoid the resolver overhead on the real + * execvp call. + */ + execvp("", (char **)argv); + + /* + * Tell the parent we're ready to go + */ + close(child_ready_pipe[1]); + + /* + * Wait until the parent tells us to go. + */ + if (read(go_pipe[0], &buf, 1) == -1) + perror("unable to read pipe"); + + execvp(argv[0], (char **)argv); + + perror(argv[0]); + exit(-1); } + + child_pid = pid; + + if (!system_wide) + target_pid = pid; + + close(child_ready_pipe[1]); + close(go_pipe[0]); + /* + * wait for child to settle + */ + if (read(child_ready_pipe[0], &buf, 1) == -1) { + perror("unable to read pipe"); + exit(-1); + } + close(child_ready_pipe[0]); + } + + + if ((!system_wide && !inherit) || profile_cpu != -1) { + open_counters(profile_cpu, target_pid); + } else { + for (i = 0; i < nr_cpus; i++) + open_counters(i, target_pid); } if (file_new) { - err = perf_header__write(header, output, false); + err = perf_header__write(&session->header, output, false); if (err < 0) return err; } - if (!system_wide) - event__synthesize_thread(pid, process_synthesized_event); - else - event__synthesize_threads(process_synthesized_event); - - if (target_pid == -1 && argc) { - pid = fork(); - if (pid < 0) - die("failed to fork"); + post_processing_offset = lseek(output, 0, SEEK_CUR); - if (!pid) { - if (execvp(argv[0], (char **)argv)) { - perror(argv[0]); - exit(-1); - } - } else { - /* - * Wait a bit for the execv'ed child to appear - * and be updated in /proc - * FIXME: Do you know a less heuristical solution? - */ - usleep(1000); - event__synthesize_thread(pid, - process_synthesized_event); - } + err = event__synthesize_kernel_mmap(process_synthesized_event, + session, "_text"); + if (err < 0) { + pr_err("Couldn't record kernel reference relocation symbol.\n"); + return err; + } - child_pid = pid; + err = event__synthesize_modules(process_synthesized_event, session); + if (err < 0) { + pr_err("Couldn't record kernel reference relocation symbol.\n"); + return err; } + if (!system_wide && profile_cpu == -1) + event__synthesize_thread(target_pid, process_synthesized_event, + session); + else + event__synthesize_threads(process_synthesized_event, session); + if (realtime_prio) { struct sched_param param; @@ -525,6 +585,12 @@ static int __cmd_record(int argc, const char **argv) } } + /* + * Let the child rip + */ + if (forks) + close(go_pipe[1]); + for (;;) { int hits = samples; @@ -619,13 +685,13 @@ int cmd_record(int argc, const char **argv, const char *prefix __used) { int counter; - symbol__init(0); - argc = parse_options(argc, argv, options, record_usage, - PARSE_OPT_STOP_AT_NON_OPTION); - if (!argc && target_pid == -1 && !system_wide) + PARSE_OPT_STOP_AT_NON_OPTION); + if (!argc && target_pid == -1 && !system_wide && profile_cpu == -1) usage_with_options(record_usage, options); + symbol__init(); + if (!nr_counters) { nr_counters = 1; attrs[0].type = PERF_TYPE_HARDWARE; diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index 383c4ab4f9af..cfc655d40bb7 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -22,25 +22,20 @@ #include "perf.h" #include "util/debug.h" #include "util/header.h" +#include "util/session.h" #include "util/parse-options.h" #include "util/parse-events.h" -#include "util/data_map.h" #include "util/thread.h" #include "util/sort.h" #include "util/hist.h" static char const *input_name = "perf.data"; -static char *dso_list_str, *comm_list_str, *sym_list_str, - *col_width_list_str; -static struct strlist *dso_list, *comm_list, *sym_list; - static int force; - -static int full_paths; -static int show_nr_samples; +static bool hide_unresolved; +static bool dont_use_callchains; static int show_threads; static struct perf_read_values show_threads_values; @@ -48,445 +43,27 @@ static struct perf_read_values show_threads_values; static char default_pretty_printing_style[] = "normal"; static char *pretty_printing_style = default_pretty_printing_style; -static int exclude_other = 1; - static char callchain_default_opt[] = "fractal,0.5"; -static struct perf_header *header; - -static u64 sample_type; - -struct symbol_conf symbol_conf; - - -static size_t -callchain__fprintf_left_margin(FILE *fp, int left_margin) -{ - int i; - int ret; - - ret = fprintf(fp, " "); - - for (i = 0; i < left_margin; i++) - ret += fprintf(fp, " "); - - return ret; -} - -static size_t ipchain__fprintf_graph_line(FILE *fp, int depth, int depth_mask, - int left_margin) -{ - int i; - size_t ret = 0; - - ret += callchain__fprintf_left_margin(fp, left_margin); - - for (i = 0; i < depth; i++) - if (depth_mask & (1 << i)) - ret += fprintf(fp, "| "); - else - ret += fprintf(fp, " "); - - ret += fprintf(fp, "\n"); - - return ret; -} -static size_t -ipchain__fprintf_graph(FILE *fp, struct callchain_list *chain, int depth, - int depth_mask, int count, u64 total_samples, - int hits, int left_margin) -{ - int i; - size_t ret = 0; - - ret += callchain__fprintf_left_margin(fp, left_margin); - for (i = 0; i < depth; i++) { - if (depth_mask & (1 << i)) - ret += fprintf(fp, "|"); - else - ret += fprintf(fp, " "); - if (!count && i == depth - 1) { - double percent; - - percent = hits * 100.0 / total_samples; - ret += percent_color_fprintf(fp, "--%2.2f%%-- ", percent); - } else - ret += fprintf(fp, "%s", " "); - } - if (chain->sym) - ret += fprintf(fp, "%s\n", chain->sym->name); - else - ret += fprintf(fp, "%p\n", (void *)(long)chain->ip); - - return ret; -} - -static struct symbol *rem_sq_bracket; -static struct callchain_list rem_hits; - -static void init_rem_hits(void) -{ - rem_sq_bracket = malloc(sizeof(*rem_sq_bracket) + 6); - if (!rem_sq_bracket) { - fprintf(stderr, "Not enough memory to display remaining hits\n"); - return; - } - - strcpy(rem_sq_bracket->name, "[...]"); - rem_hits.sym = rem_sq_bracket; -} - -static size_t -__callchain__fprintf_graph(FILE *fp, struct callchain_node *self, - u64 total_samples, int depth, int depth_mask, - int left_margin) -{ - struct rb_node *node, *next; - struct callchain_node *child; - struct callchain_list *chain; - int new_depth_mask = depth_mask; - u64 new_total; - u64 remaining; - size_t ret = 0; - int i; - - if (callchain_param.mode == CHAIN_GRAPH_REL) - new_total = self->children_hit; - else - new_total = total_samples; - - remaining = new_total; - - node = rb_first(&self->rb_root); - while (node) { - u64 cumul; - - child = rb_entry(node, struct callchain_node, rb_node); - cumul = cumul_hits(child); - remaining -= cumul; - - /* - * The depth mask manages the output of pipes that show - * the depth. We don't want to keep the pipes of the current - * level for the last child of this depth. - * Except if we have remaining filtered hits. They will - * supersede the last child - */ - next = rb_next(node); - if (!next && (callchain_param.mode != CHAIN_GRAPH_REL || !remaining)) - new_depth_mask &= ~(1 << (depth - 1)); - - /* - * But we keep the older depth mask for the line seperator - * to keep the level link until we reach the last child - */ - ret += ipchain__fprintf_graph_line(fp, depth, depth_mask, - left_margin); - i = 0; - list_for_each_entry(chain, &child->val, list) { - if (chain->ip >= PERF_CONTEXT_MAX) - continue; - ret += ipchain__fprintf_graph(fp, chain, depth, - new_depth_mask, i++, - new_total, - cumul, - left_margin); - } - ret += __callchain__fprintf_graph(fp, child, new_total, - depth + 1, - new_depth_mask | (1 << depth), - left_margin); - node = next; - } - - if (callchain_param.mode == CHAIN_GRAPH_REL && - remaining && remaining != new_total) { - - if (!rem_sq_bracket) - return ret; - - new_depth_mask &= ~(1 << (depth - 1)); - - ret += ipchain__fprintf_graph(fp, &rem_hits, depth, - new_depth_mask, 0, new_total, - remaining, left_margin); - } - - return ret; -} - - -static size_t -callchain__fprintf_graph(FILE *fp, struct callchain_node *self, - u64 total_samples, int left_margin) -{ - struct callchain_list *chain; - bool printed = false; - int i = 0; - int ret = 0; - - list_for_each_entry(chain, &self->val, list) { - if (chain->ip >= PERF_CONTEXT_MAX) - continue; - - if (!i++ && sort__first_dimension == SORT_SYM) - continue; - - if (!printed) { - ret += callchain__fprintf_left_margin(fp, left_margin); - ret += fprintf(fp, "|\n"); - ret += callchain__fprintf_left_margin(fp, left_margin); - ret += fprintf(fp, "---"); - - left_margin += 3; - printed = true; - } else - ret += callchain__fprintf_left_margin(fp, left_margin); - - if (chain->sym) - ret += fprintf(fp, " %s\n", chain->sym->name); - else - ret += fprintf(fp, " %p\n", (void *)(long)chain->ip); - } - - ret += __callchain__fprintf_graph(fp, self, total_samples, 1, 1, left_margin); - - return ret; -} - -static size_t -callchain__fprintf_flat(FILE *fp, struct callchain_node *self, - u64 total_samples) -{ - struct callchain_list *chain; - size_t ret = 0; - - if (!self) - return 0; - - ret += callchain__fprintf_flat(fp, self->parent, total_samples); - - - list_for_each_entry(chain, &self->val, list) { - if (chain->ip >= PERF_CONTEXT_MAX) - continue; - if (chain->sym) - ret += fprintf(fp, " %s\n", chain->sym->name); - else - ret += fprintf(fp, " %p\n", - (void *)(long)chain->ip); - } - - return ret; -} - -static size_t -hist_entry_callchain__fprintf(FILE *fp, struct hist_entry *self, - u64 total_samples, int left_margin) -{ - struct rb_node *rb_node; - struct callchain_node *chain; - size_t ret = 0; - - rb_node = rb_first(&self->sorted_chain); - while (rb_node) { - double percent; - - chain = rb_entry(rb_node, struct callchain_node, rb_node); - percent = chain->hit * 100.0 / total_samples; - switch (callchain_param.mode) { - case CHAIN_FLAT: - ret += percent_color_fprintf(fp, " %6.2f%%\n", - percent); - ret += callchain__fprintf_flat(fp, chain, total_samples); - break; - case CHAIN_GRAPH_ABS: /* Falldown */ - case CHAIN_GRAPH_REL: - ret += callchain__fprintf_graph(fp, chain, total_samples, - left_margin); - case CHAIN_NONE: - default: - break; - } - ret += fprintf(fp, "\n"); - rb_node = rb_next(rb_node); - } - - return ret; -} - -static size_t -hist_entry__fprintf(FILE *fp, struct hist_entry *self, u64 total_samples) -{ - struct sort_entry *se; - size_t ret; - - if (exclude_other && !self->parent) - return 0; - - if (total_samples) - ret = percent_color_fprintf(fp, - field_sep ? "%.2f" : " %6.2f%%", - (self->count * 100.0) / total_samples); - else - ret = fprintf(fp, field_sep ? "%lld" : "%12lld ", self->count); - - if (show_nr_samples) { - if (field_sep) - fprintf(fp, "%c%lld", *field_sep, self->count); - else - fprintf(fp, "%11lld", self->count); - } - - list_for_each_entry(se, &hist_entry__sort_list, list) { - if (se->elide) - continue; - - fprintf(fp, "%s", field_sep ?: " "); - ret += se->print(fp, self, se->width ? *se->width : 0); - } - - ret += fprintf(fp, "\n"); - - if (callchain) { - int left_margin = 0; - - if (sort__first_dimension == SORT_COMM) { - se = list_first_entry(&hist_entry__sort_list, typeof(*se), - list); - left_margin = se->width ? *se->width : 0; - left_margin -= thread__comm_len(self->thread); - } - - hist_entry_callchain__fprintf(fp, self, total_samples, - left_margin); - } - - return ret; -} - -/* - * - */ - -static void dso__calc_col_width(struct dso *self) -{ - if (!col_width_list_str && !field_sep && - (!dso_list || strlist__has_entry(dso_list, self->name))) { - unsigned int slen = strlen(self->name); - if (slen > dsos__col_width) - dsos__col_width = slen; - } - - self->slen_calculated = 1; -} - -static void thread__comm_adjust(struct thread *self) -{ - char *comm = self->comm; - - if (!col_width_list_str && !field_sep && - (!comm_list || strlist__has_entry(comm_list, comm))) { - unsigned int slen = strlen(comm); - - if (slen > comms__col_width) { - comms__col_width = slen; - threads__col_width = slen + 6; - } - } -} - -static int thread__set_comm_adjust(struct thread *self, const char *comm) -{ - int ret = thread__set_comm(self, comm); - - if (ret) - return ret; - - thread__comm_adjust(self); - - return 0; -} - -static int call__match(struct symbol *sym) -{ - if (sym->name && !regexec(&parent_regex, sym->name, 0, NULL, 0)) - return 1; - - return 0; -} - -static struct symbol **resolve_callchain(struct thread *thread, - struct ip_callchain *chain, - struct symbol **parent) -{ - u8 cpumode = PERF_RECORD_MISC_USER; - struct symbol **syms = NULL; - unsigned int i; - - if (callchain) { - syms = calloc(chain->nr, sizeof(*syms)); - if (!syms) { - fprintf(stderr, "Can't allocate memory for symbols\n"); - exit(-1); - } - } - - for (i = 0; i < chain->nr; i++) { - u64 ip = chain->ips[i]; - struct addr_location al; - - if (ip >= PERF_CONTEXT_MAX) { - switch (ip) { - case PERF_CONTEXT_HV: - cpumode = PERF_RECORD_MISC_HYPERVISOR; break; - case PERF_CONTEXT_KERNEL: - cpumode = PERF_RECORD_MISC_KERNEL; break; - case PERF_CONTEXT_USER: - cpumode = PERF_RECORD_MISC_USER; break; - default: - break; - } - continue; - } - - thread__find_addr_location(thread, cpumode, MAP__FUNCTION, - ip, &al, NULL); - if (al.sym != NULL) { - if (sort__has_parent && !*parent && - call__match(al.sym)) - *parent = al.sym; - if (!callchain) - break; - syms[i] = al.sym; - } - } - - return syms; -} - -/* - * collect histogram counts - */ - -static int hist_entry__add(struct addr_location *al, - struct ip_callchain *chain, u64 count) +static int perf_session__add_hist_entry(struct perf_session *self, + struct addr_location *al, + struct ip_callchain *chain, u64 count) { struct symbol **syms = NULL, *parent = NULL; bool hit; struct hist_entry *he; - if ((sort__has_parent || callchain) && chain) - syms = resolve_callchain(al->thread, chain, &parent); - - he = __hist_entry__add(al, parent, count, &hit); + if ((sort__has_parent || symbol_conf.use_callchain) && chain) + syms = perf_session__resolve_callchain(self, al->thread, + chain, &parent); + he = __perf_session__add_hist_entry(self, al, parent, count, &hit); if (he == NULL) return -ENOMEM; if (hit) he->count += count; - if (callchain) { + if (symbol_conf.use_callchain) { if (!hit) callchain_init(&he->callchain); append_chain(&he->callchain, chain, syms); @@ -496,100 +73,6 @@ static int hist_entry__add(struct addr_location *al, return 0; } -static size_t output__fprintf(FILE *fp, u64 total_samples) -{ - struct hist_entry *pos; - struct sort_entry *se; - struct rb_node *nd; - size_t ret = 0; - unsigned int width; - char *col_width = col_width_list_str; - int raw_printing_style; - - raw_printing_style = !strcmp(pretty_printing_style, "raw"); - - init_rem_hits(); - - fprintf(fp, "# Samples: %Ld\n", (u64)total_samples); - fprintf(fp, "#\n"); - - fprintf(fp, "# Overhead"); - if (show_nr_samples) { - if (field_sep) - fprintf(fp, "%cSamples", *field_sep); - else - fputs(" Samples ", fp); - } - list_for_each_entry(se, &hist_entry__sort_list, list) { - if (se->elide) - continue; - if (field_sep) { - fprintf(fp, "%c%s", *field_sep, se->header); - continue; - } - width = strlen(se->header); - if (se->width) { - if (col_width_list_str) { - if (col_width) { - *se->width = atoi(col_width); - col_width = strchr(col_width, ','); - if (col_width) - ++col_width; - } - } - width = *se->width = max(*se->width, width); - } - fprintf(fp, " %*s", width, se->header); - } - fprintf(fp, "\n"); - - if (field_sep) - goto print_entries; - - fprintf(fp, "# ........"); - if (show_nr_samples) - fprintf(fp, " .........."); - list_for_each_entry(se, &hist_entry__sort_list, list) { - unsigned int i; - - if (se->elide) - continue; - - fprintf(fp, " "); - if (se->width) - width = *se->width; - else - width = strlen(se->header); - for (i = 0; i < width; i++) - fprintf(fp, "."); - } - fprintf(fp, "\n"); - - fprintf(fp, "#\n"); - -print_entries: - for (nd = rb_first(&output_hists); nd; nd = rb_next(nd)) { - pos = rb_entry(nd, struct hist_entry, rb_node); - ret += hist_entry__fprintf(fp, pos, total_samples); - } - - if (sort_order == default_sort_order && - parent_pattern == default_parent_pattern) { - fprintf(fp, "#\n"); - fprintf(fp, "# (For a higher level overview, try: perf report --sort comm,dso)\n"); - fprintf(fp, "#\n"); - } - fprintf(fp, "\n"); - - free(rem_sq_bracket); - - if (show_threads) - perf_read_values_display(fp, &show_threads_values, - raw_printing_style); - - return ret; -} - static int validate_chain(struct ip_callchain *chain, event_t *event) { unsigned int chain_size; @@ -603,108 +86,57 @@ static int validate_chain(struct ip_callchain *chain, event_t *event) return 0; } -static int process_sample_event(event_t *event) +static int process_sample_event(event_t *event, struct perf_session *session) { - u64 ip = event->ip.ip; - u64 period = 1; - void *more_data = event->ip.__more_data; - struct ip_callchain *chain = NULL; - int cpumode; + struct sample_data data = { .period = 1, }; struct addr_location al; - struct thread *thread = threads__findnew(event->ip.pid); - if (sample_type & PERF_SAMPLE_PERIOD) { - period = *(u64 *)more_data; - more_data += sizeof(u64); - } + event__parse_sample(event, session->sample_type, &data); - dump_printf("(IP, %d): %d/%d: %p period: %Ld\n", - event->header.misc, - event->ip.pid, event->ip.tid, - (void *)(long)ip, - (long long)period); + dump_printf("(IP, %d): %d/%d: %#Lx period: %Ld\n", event->header.misc, + data.pid, data.tid, data.ip, data.period); - if (sample_type & PERF_SAMPLE_CALLCHAIN) { + if (session->sample_type & PERF_SAMPLE_CALLCHAIN) { unsigned int i; - chain = (void *)more_data; - - dump_printf("... chain: nr:%Lu\n", chain->nr); + dump_printf("... chain: nr:%Lu\n", data.callchain->nr); - if (validate_chain(chain, event) < 0) { + if (validate_chain(data.callchain, event) < 0) { pr_debug("call-chain problem with event, " "skipping it.\n"); return 0; } if (dump_trace) { - for (i = 0; i < chain->nr; i++) - dump_printf("..... %2d: %016Lx\n", i, chain->ips[i]); + for (i = 0; i < data.callchain->nr; i++) + dump_printf("..... %2d: %016Lx\n", + i, data.callchain->ips[i]); } } - if (thread == NULL) { - pr_debug("problem processing %d event, skipping it.\n", + if (event__preprocess_sample(event, session, &al, NULL) < 0) { + fprintf(stderr, "problem processing %d event, skipping it.\n", event->header.type); return -1; } - dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid); - - if (comm_list && !strlist__has_entry(comm_list, thread->comm)) + if (al.filtered || (hide_unresolved && al.sym == NULL)) return 0; - cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK; - - thread__find_addr_location(thread, cpumode, - MAP__FUNCTION, ip, &al, NULL); - /* - * We have to do this here as we may have a dso with no symbol hit that - * has a name longer than the ones with symbols sampled. - */ - if (al.map && !sort_dso.elide && !al.map->dso->slen_calculated) - dso__calc_col_width(al.map->dso); - - if (dso_list && - (!al.map || !al.map->dso || - !(strlist__has_entry(dso_list, al.map->dso->short_name) || - (al.map->dso->short_name != al.map->dso->long_name && - strlist__has_entry(dso_list, al.map->dso->long_name))))) - return 0; - - if (sym_list && al.sym && !strlist__has_entry(sym_list, al.sym->name)) - return 0; - - if (hist_entry__add(&al, chain, period)) { + if (perf_session__add_hist_entry(session, &al, data.callchain, data.period)) { pr_debug("problem incrementing symbol count, skipping event\n"); return -1; } - event__stats.total += period; - - return 0; -} - -static int process_comm_event(event_t *event) -{ - struct thread *thread = threads__findnew(event->comm.pid); - - dump_printf(": %s:%d\n", event->comm.comm, event->comm.pid); - - if (thread == NULL || - thread__set_comm_adjust(thread, event->comm.comm)) { - dump_printf("problem processing PERF_RECORD_COMM, skipping event.\n"); - return -1; - } - + session->events_stats.total += data.period; return 0; } -static int process_read_event(event_t *event) +static int process_read_event(event_t *event, struct perf_session *session __used) { struct perf_event_attr *attr; - attr = perf_header__find_attr(event->read.id, header); + attr = perf_header__find_attr(event->read.id, &session->header); if (show_threads) { const char *name = attr ? __event_name(attr->type, attr->config) @@ -723,94 +155,110 @@ static int process_read_event(event_t *event) return 0; } -static int sample_type_check(u64 type) +static int perf_session__setup_sample_type(struct perf_session *self) { - sample_type = type; - - if (!(sample_type & PERF_SAMPLE_CALLCHAIN)) { + if (!(self->sample_type & PERF_SAMPLE_CALLCHAIN)) { if (sort__has_parent) { fprintf(stderr, "selected --sort parent, but no" " callchain data. Did you call" " perf record without -g?\n"); - return -1; + return -EINVAL; } - if (callchain) { + if (symbol_conf.use_callchain) { fprintf(stderr, "selected -g but no callchain data." " Did you call perf record without" " -g?\n"); return -1; } - } else if (callchain_param.mode != CHAIN_NONE && !callchain) { - callchain = 1; + } else if (!dont_use_callchains && callchain_param.mode != CHAIN_NONE && + !symbol_conf.use_callchain) { + symbol_conf.use_callchain = true; if (register_callchain_param(&callchain_param) < 0) { fprintf(stderr, "Can't register callchain" " params\n"); - return -1; + return -EINVAL; } } return 0; } -static struct perf_file_handler file_handler = { - .process_sample_event = process_sample_event, - .process_mmap_event = event__process_mmap, - .process_comm_event = process_comm_event, - .process_exit_event = event__process_task, - .process_fork_event = event__process_task, - .process_lost_event = event__process_lost, - .process_read_event = process_read_event, - .sample_type_check = sample_type_check, +static struct perf_event_ops event_ops = { + .sample = process_sample_event, + .mmap = event__process_mmap, + .comm = event__process_comm, + .exit = event__process_task, + .fork = event__process_task, + .lost = event__process_lost, + .read = process_read_event, }; - static int __cmd_report(void) { - struct thread *idle; - int ret; + int ret = -EINVAL; + struct perf_session *session; - idle = register_idle_thread(); - thread__comm_adjust(idle); + session = perf_session__new(input_name, O_RDONLY, force); + if (session == NULL) + return -ENOMEM; if (show_threads) perf_read_values_init(&show_threads_values); - register_perf_file_handler(&file_handler); + ret = perf_session__setup_sample_type(session); + if (ret) + goto out_delete; - ret = mmap_dispatch_perf_file(&header, input_name, force, - full_paths, &event__cwdlen, &event__cwd); + ret = perf_session__process_events(session, &event_ops); if (ret) - return ret; + goto out_delete; if (dump_trace) { event__print_totals(); - return 0; + goto out_delete; } if (verbose > 3) - threads__fprintf(stdout); + perf_session__fprintf(session, stdout); if (verbose > 2) dsos__fprintf(stdout); - collapse__resort(); - output__resort(event__stats.total); - output__fprintf(stdout, event__stats.total); + perf_session__collapse_resort(session); + perf_session__output_resort(session, session->events_stats.total); + fprintf(stdout, "# Samples: %Ld\n#\n", session->events_stats.total); + perf_session__fprintf_hists(session, NULL, false, stdout); + if (sort_order == default_sort_order && + parent_pattern == default_parent_pattern) + fprintf(stdout, "#\n# (For a higher level overview, try: perf report --sort comm,dso)\n#\n"); - if (show_threads) + if (show_threads) { + bool raw_printing_style = !strcmp(pretty_printing_style, "raw"); + perf_read_values_display(stdout, &show_threads_values, + raw_printing_style); perf_read_values_destroy(&show_threads_values); - + } +out_delete: + perf_session__delete(session); return ret; } static int parse_callchain_opt(const struct option *opt __used, const char *arg, - int unset __used) + int unset) { char *tok; char *endptr; - callchain = 1; + /* + * --no-call-graph + */ + if (unset) { + dont_use_callchains = true; + return 0; + } + + symbol_conf.use_callchain = true; if (!arg) return 0; @@ -831,7 +279,7 @@ parse_callchain_opt(const struct option *opt __used, const char *arg, else if (!strncmp(tok, "none", strlen(arg))) { callchain_param.mode = CHAIN_NONE; - callchain = 0; + symbol_conf.use_callchain = false; return 0; } @@ -856,8 +304,7 @@ setup: return 0; } -//static const char * const report_usage[] = { -const char * const report_usage[] = { +static const char * const report_usage[] = { "perf report [<options>] <command>", NULL }; @@ -874,7 +321,7 @@ static const struct option options[] = { OPT_BOOLEAN('f', "force", &force, "don't complain, do it"), OPT_BOOLEAN('m', "modules", &symbol_conf.use_modules, "load module symbols - WARNING: use only with -k and LIVE kernel"), - OPT_BOOLEAN('n', "show-nr-samples", &show_nr_samples, + OPT_BOOLEAN('n', "show-nr-samples", &symbol_conf.show_nr_samples, "Show a column with the number of samples"), OPT_BOOLEAN('T', "threads", &show_threads, "Show per-thread event counters"), @@ -882,78 +329,48 @@ static const struct option options[] = { "pretty printing style key: normal raw"), OPT_STRING('s', "sort", &sort_order, "key[,key2...]", "sort by key(s): pid, comm, dso, symbol, parent"), - OPT_BOOLEAN('P', "full-paths", &full_paths, + OPT_BOOLEAN('P', "full-paths", &symbol_conf.full_paths, "Don't shorten the pathnames taking into account the cwd"), OPT_STRING('p', "parent", &parent_pattern, "regex", "regex filter to identify parent, see: '--sort parent'"), - OPT_BOOLEAN('x', "exclude-other", &exclude_other, + OPT_BOOLEAN('x', "exclude-other", &symbol_conf.exclude_other, "Only display entries with parent-match"), OPT_CALLBACK_DEFAULT('g', "call-graph", NULL, "output_type,min_percent", "Display callchains using output_type and min percent threshold. " "Default: fractal,0.5", &parse_callchain_opt, callchain_default_opt), - OPT_STRING('d', "dsos", &dso_list_str, "dso[,dso...]", + OPT_STRING('d', "dsos", &symbol_conf.dso_list_str, "dso[,dso...]", "only consider symbols in these dsos"), - OPT_STRING('C', "comms", &comm_list_str, "comm[,comm...]", + OPT_STRING('C', "comms", &symbol_conf.comm_list_str, "comm[,comm...]", "only consider symbols in these comms"), - OPT_STRING('S', "symbols", &sym_list_str, "symbol[,symbol...]", + OPT_STRING('S', "symbols", &symbol_conf.sym_list_str, "symbol[,symbol...]", "only consider these symbols"), - OPT_STRING('w', "column-widths", &col_width_list_str, + OPT_STRING('w', "column-widths", &symbol_conf.col_width_list_str, "width[,width...]", "don't try to adjust column width, use these fixed values"), - OPT_STRING('t', "field-separator", &field_sep, "separator", + OPT_STRING('t', "field-separator", &symbol_conf.field_sep, "separator", "separator for columns, no spaces will be added between " "columns '.' is reserved."), + OPT_BOOLEAN('U', "hide-unresolved", &hide_unresolved, + "Only display entries resolved to a symbol"), OPT_END() }; -static void setup_sorting(void) +int cmd_report(int argc, const char **argv, const char *prefix __used) { - char *tmp, *tok, *str = strdup(sort_order); - - for (tok = strtok_r(str, ", ", &tmp); - tok; tok = strtok_r(NULL, ", ", &tmp)) { - if (sort_dimension__add(tok) < 0) { - error("Unknown --sort key: `%s'", tok); - usage_with_options(report_usage, options); - } - } - - free(str); -} + argc = parse_options(argc, argv, options, report_usage, 0); -static void setup_list(struct strlist **list, const char *list_str, - struct sort_entry *se, const char *list_name, - FILE *fp) -{ - if (list_str) { - *list = strlist__new(true, list_str); - if (!*list) { - fprintf(stderr, "problems parsing %s list\n", - list_name); - exit(129); - } - if (strlist__nr_entries(*list) == 1) { - fprintf(fp, "# %s: %s\n", list_name, - strlist__entry(*list, 0)->s); - se->elide = true; - } - } -} + setup_pager(); -int cmd_report(int argc, const char **argv, const char *prefix __used) -{ - if (symbol__init(&symbol_conf) < 0) + if (symbol__init() < 0) return -1; - argc = parse_options(argc, argv, options, report_usage, 0); - - setup_sorting(); + setup_sorting(report_usage, options); if (parent_pattern != default_parent_pattern) { sort_dimension__add("parent"); sort_parent.elide = 1; } else - exclude_other = 0; + symbol_conf.exclude_other = false; /* * Any (unrecognized) arguments left? @@ -961,17 +378,9 @@ int cmd_report(int argc, const char **argv, const char *prefix __used) if (argc) usage_with_options(report_usage, options); - setup_pager(); - - setup_list(&dso_list, dso_list_str, &sort_dso, "dso", stdout); - setup_list(&comm_list, comm_list_str, &sort_comm, "comm", stdout); - setup_list(&sym_list, sym_list_str, &sort_sym, "symbol", stdout); - - if (field_sep && *field_sep == '.') { - fputs("'.' is the only non valid --field-separator argument\n", - stderr); - exit(129); - } + sort_entry__setup_elide(&sort_dso, symbol_conf.dso_list, "dso", stdout); + sort_entry__setup_elide(&sort_comm, symbol_conf.comm_list, "comm", stdout); + sort_entry__setup_elide(&sort_sym, symbol_conf.sym_list, "symbol", stdout); return __cmd_report(); } diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c index 26b782f26ee1..4f5a03e43444 100644 --- a/tools/perf/builtin-sched.c +++ b/tools/perf/builtin-sched.c @@ -6,14 +6,13 @@ #include "util/symbol.h" #include "util/thread.h" #include "util/header.h" +#include "util/session.h" #include "util/parse-options.h" #include "util/trace-event.h" #include "util/debug.h" -#include "util/data_map.h" -#include <sys/types.h> #include <sys/prctl.h> #include <semaphore.h> @@ -22,9 +21,6 @@ static char const *input_name = "perf.data"; -static struct perf_header *header; -static u64 sample_type; - static char default_sort_order[] = "avg, max, switch, runtime"; static char *sort_order = default_sort_order; @@ -141,6 +137,7 @@ struct work_atoms { struct thread *thread; struct rb_node node; u64 max_lat; + u64 max_lat_at; u64 total_lat; u64 nb_atoms; u64 total_runtime; @@ -414,34 +411,33 @@ static u64 get_cpu_usage_nsec_parent(void) return sum; } -static u64 get_cpu_usage_nsec_self(void) +static int self_open_counters(void) { - char filename [] = "/proc/1234567890/sched"; - unsigned long msecs, nsecs; - char *line = NULL; - u64 total = 0; - size_t len = 0; - ssize_t chars; - FILE *file; - int ret; + struct perf_event_attr attr; + int fd; - sprintf(filename, "/proc/%d/sched", getpid()); - file = fopen(filename, "r"); - BUG_ON(!file); + memset(&attr, 0, sizeof(attr)); - while ((chars = getline(&line, &len, file)) != -1) { - ret = sscanf(line, "se.sum_exec_runtime : %ld.%06ld\n", - &msecs, &nsecs); - if (ret == 2) { - total = msecs*1e6 + nsecs; - break; - } - } - if (line) - free(line); - fclose(file); + attr.type = PERF_TYPE_SOFTWARE; + attr.config = PERF_COUNT_SW_TASK_CLOCK; + + fd = sys_perf_event_open(&attr, 0, -1, -1, 0); - return total; + if (fd < 0) + die("Error: sys_perf_event_open() syscall returned" + "with %d (%s)\n", fd, strerror(errno)); + return fd; +} + +static u64 get_cpu_usage_nsec_self(int fd) +{ + u64 runtime; + int ret; + + ret = read(fd, &runtime, sizeof(runtime)); + BUG_ON(ret != sizeof(runtime)); + + return runtime; } static void *thread_func(void *ctx) @@ -450,9 +446,11 @@ static void *thread_func(void *ctx) u64 cpu_usage_0, cpu_usage_1; unsigned long i, ret; char comm2[22]; + int fd; sprintf(comm2, ":%s", this_task->comm); prctl(PR_SET_NAME, comm2); + fd = self_open_counters(); again: ret = sem_post(&this_task->ready_for_work); @@ -462,16 +460,15 @@ again: ret = pthread_mutex_unlock(&start_work_mutex); BUG_ON(ret); - cpu_usage_0 = get_cpu_usage_nsec_self(); + cpu_usage_0 = get_cpu_usage_nsec_self(fd); for (i = 0; i < this_task->nr_events; i++) { this_task->curr_event = i; process_sched_event(this_task, this_task->atoms[i]); } - cpu_usage_1 = get_cpu_usage_nsec_self(); + cpu_usage_1 = get_cpu_usage_nsec_self(fd); this_task->cpu_usage = cpu_usage_1 - cpu_usage_0; - ret = sem_post(&this_task->work_done_sem); BUG_ON(ret); @@ -628,11 +625,6 @@ static void test_calibrations(void) printf("the sleep test took %Ld nsecs\n", T1-T0); } -struct raw_event_sample { - u32 size; - char data[0]; -}; - #define FILL_FIELD(ptr, field, event, data) \ ptr.field = (typeof(ptr.field)) raw_field_value(event, #field, data) @@ -736,18 +728,21 @@ struct trace_migrate_task_event { struct trace_sched_handler { void (*switch_event)(struct trace_switch_event *, + struct perf_session *, struct event *, int cpu, u64 timestamp, struct thread *thread); void (*runtime_event)(struct trace_runtime_event *, + struct perf_session *, struct event *, int cpu, u64 timestamp, struct thread *thread); void (*wakeup_event)(struct trace_wakeup_event *, + struct perf_session *, struct event *, int cpu, u64 timestamp, @@ -760,6 +755,7 @@ struct trace_sched_handler { struct thread *thread); void (*migrate_task_event)(struct trace_migrate_task_event *, + struct perf_session *session, struct event *, int cpu, u64 timestamp, @@ -769,6 +765,7 @@ struct trace_sched_handler { static void replay_wakeup_event(struct trace_wakeup_event *wakeup_event, + struct perf_session *session __used, struct event *event, int cpu __used, u64 timestamp __used, @@ -795,6 +792,7 @@ static u64 cpu_last_switched[MAX_CPUS]; static void replay_switch_event(struct trace_switch_event *switch_event, + struct perf_session *session __used, struct event *event, int cpu, u64 timestamp, @@ -1019,13 +1017,16 @@ add_sched_in_event(struct work_atoms *atoms, u64 timestamp) delta = atom->sched_in_time - atom->wake_up_time; atoms->total_lat += delta; - if (delta > atoms->max_lat) + if (delta > atoms->max_lat) { atoms->max_lat = delta; + atoms->max_lat_at = timestamp; + } atoms->nb_atoms++; } static void latency_switch_event(struct trace_switch_event *switch_event, + struct perf_session *session, struct event *event __used, int cpu, u64 timestamp, @@ -1049,8 +1050,8 @@ latency_switch_event(struct trace_switch_event *switch_event, die("hm, delta: %Ld < 0 ?\n", delta); - sched_out = threads__findnew(switch_event->prev_pid); - sched_in = threads__findnew(switch_event->next_pid); + sched_out = perf_session__findnew(session, switch_event->prev_pid); + sched_in = perf_session__findnew(session, switch_event->next_pid); out_events = thread_atoms_search(&atom_root, sched_out, &cmp_pid); if (!out_events) { @@ -1078,12 +1079,13 @@ latency_switch_event(struct trace_switch_event *switch_event, static void latency_runtime_event(struct trace_runtime_event *runtime_event, + struct perf_session *session, struct event *event __used, int cpu, u64 timestamp, struct thread *this_thread __used) { - struct thread *thread = threads__findnew(runtime_event->pid); + struct thread *thread = perf_session__findnew(session, runtime_event->pid); struct work_atoms *atoms = thread_atoms_search(&atom_root, thread, &cmp_pid); BUG_ON(cpu >= MAX_CPUS || cpu < 0); @@ -1100,6 +1102,7 @@ latency_runtime_event(struct trace_runtime_event *runtime_event, static void latency_wakeup_event(struct trace_wakeup_event *wakeup_event, + struct perf_session *session, struct event *__event __used, int cpu __used, u64 timestamp, @@ -1113,7 +1116,7 @@ latency_wakeup_event(struct trace_wakeup_event *wakeup_event, if (!wakeup_event->success) return; - wakee = threads__findnew(wakeup_event->pid); + wakee = perf_session__findnew(session, wakeup_event->pid); atoms = thread_atoms_search(&atom_root, wakee, &cmp_pid); if (!atoms) { thread_atoms_insert(wakee); @@ -1147,6 +1150,7 @@ latency_wakeup_event(struct trace_wakeup_event *wakeup_event, static void latency_migrate_task_event(struct trace_migrate_task_event *migrate_task_event, + struct perf_session *session, struct event *__event __used, int cpu __used, u64 timestamp, @@ -1162,7 +1166,7 @@ latency_migrate_task_event(struct trace_migrate_task_event *migrate_task_event, if (profile_cpu == -1) return; - migrant = threads__findnew(migrate_task_event->pid); + migrant = perf_session__findnew(session, migrate_task_event->pid); atoms = thread_atoms_search(&atom_root, migrant, &cmp_pid); if (!atoms) { thread_atoms_insert(migrant); @@ -1216,10 +1220,11 @@ static void output_lat_thread(struct work_atoms *work_list) avg = work_list->total_lat / work_list->nb_atoms; - printf("|%11.3f ms |%9llu | avg:%9.3f ms | max:%9.3f ms |\n", + printf("|%11.3f ms |%9llu | avg:%9.3f ms | max:%9.3f ms | max at: %9.6f s\n", (double)work_list->total_runtime / 1e6, work_list->nb_atoms, (double)avg / 1e6, - (double)work_list->max_lat / 1e6); + (double)work_list->max_lat / 1e6, + (double)work_list->max_lat_at / 1e9); } static int pid_cmp(struct work_atoms *l, struct work_atoms *r) @@ -1356,7 +1361,7 @@ static void sort_lat(void) static struct trace_sched_handler *trace_handler; static void -process_sched_wakeup_event(struct raw_event_sample *raw, +process_sched_wakeup_event(void *data, struct perf_session *session, struct event *event, int cpu __used, u64 timestamp __used, @@ -1364,16 +1369,17 @@ process_sched_wakeup_event(struct raw_event_sample *raw, { struct trace_wakeup_event wakeup_event; - FILL_COMMON_FIELDS(wakeup_event, event, raw->data); + FILL_COMMON_FIELDS(wakeup_event, event, data); - FILL_ARRAY(wakeup_event, comm, event, raw->data); - FILL_FIELD(wakeup_event, pid, event, raw->data); - FILL_FIELD(wakeup_event, prio, event, raw->data); - FILL_FIELD(wakeup_event, success, event, raw->data); - FILL_FIELD(wakeup_event, cpu, event, raw->data); + FILL_ARRAY(wakeup_event, comm, event, data); + FILL_FIELD(wakeup_event, pid, event, data); + FILL_FIELD(wakeup_event, prio, event, data); + FILL_FIELD(wakeup_event, success, event, data); + FILL_FIELD(wakeup_event, cpu, event, data); if (trace_handler->wakeup_event) - trace_handler->wakeup_event(&wakeup_event, event, cpu, timestamp, thread); + trace_handler->wakeup_event(&wakeup_event, session, event, + cpu, timestamp, thread); } /* @@ -1391,6 +1397,7 @@ static char next_shortname2 = '0'; static void map_switch_event(struct trace_switch_event *switch_event, + struct perf_session *session, struct event *event __used, int this_cpu, u64 timestamp, @@ -1418,8 +1425,8 @@ map_switch_event(struct trace_switch_event *switch_event, die("hm, delta: %Ld < 0 ?\n", delta); - sched_out = threads__findnew(switch_event->prev_pid); - sched_in = threads__findnew(switch_event->next_pid); + sched_out = perf_session__findnew(session, switch_event->prev_pid); + sched_in = perf_session__findnew(session, switch_event->next_pid); curr_thread[this_cpu] = sched_in; @@ -1469,7 +1476,7 @@ map_switch_event(struct trace_switch_event *switch_event, static void -process_sched_switch_event(struct raw_event_sample *raw, +process_sched_switch_event(void *data, struct perf_session *session, struct event *event, int this_cpu, u64 timestamp __used, @@ -1477,15 +1484,15 @@ process_sched_switch_event(struct raw_event_sample *raw, { struct trace_switch_event switch_event; - FILL_COMMON_FIELDS(switch_event, event, raw->data); + FILL_COMMON_FIELDS(switch_event, event, data); - FILL_ARRAY(switch_event, prev_comm, event, raw->data); - FILL_FIELD(switch_event, prev_pid, event, raw->data); - FILL_FIELD(switch_event, prev_prio, event, raw->data); - FILL_FIELD(switch_event, prev_state, event, raw->data); - FILL_ARRAY(switch_event, next_comm, event, raw->data); - FILL_FIELD(switch_event, next_pid, event, raw->data); - FILL_FIELD(switch_event, next_prio, event, raw->data); + FILL_ARRAY(switch_event, prev_comm, event, data); + FILL_FIELD(switch_event, prev_pid, event, data); + FILL_FIELD(switch_event, prev_prio, event, data); + FILL_FIELD(switch_event, prev_state, event, data); + FILL_ARRAY(switch_event, next_comm, event, data); + FILL_FIELD(switch_event, next_pid, event, data); + FILL_FIELD(switch_event, next_prio, event, data); if (curr_pid[this_cpu] != (u32)-1) { /* @@ -1496,13 +1503,14 @@ process_sched_switch_event(struct raw_event_sample *raw, nr_context_switch_bugs++; } if (trace_handler->switch_event) - trace_handler->switch_event(&switch_event, event, this_cpu, timestamp, thread); + trace_handler->switch_event(&switch_event, session, event, + this_cpu, timestamp, thread); curr_pid[this_cpu] = switch_event.next_pid; } static void -process_sched_runtime_event(struct raw_event_sample *raw, +process_sched_runtime_event(void *data, struct perf_session *session, struct event *event, int cpu __used, u64 timestamp __used, @@ -1510,17 +1518,17 @@ process_sched_runtime_event(struct raw_event_sample *raw, { struct trace_runtime_event runtime_event; - FILL_ARRAY(runtime_event, comm, event, raw->data); - FILL_FIELD(runtime_event, pid, event, raw->data); - FILL_FIELD(runtime_event, runtime, event, raw->data); - FILL_FIELD(runtime_event, vruntime, event, raw->data); + FILL_ARRAY(runtime_event, comm, event, data); + FILL_FIELD(runtime_event, pid, event, data); + FILL_FIELD(runtime_event, runtime, event, data); + FILL_FIELD(runtime_event, vruntime, event, data); if (trace_handler->runtime_event) - trace_handler->runtime_event(&runtime_event, event, cpu, timestamp, thread); + trace_handler->runtime_event(&runtime_event, session, event, cpu, timestamp, thread); } static void -process_sched_fork_event(struct raw_event_sample *raw, +process_sched_fork_event(void *data, struct event *event, int cpu __used, u64 timestamp __used, @@ -1528,15 +1536,16 @@ process_sched_fork_event(struct raw_event_sample *raw, { struct trace_fork_event fork_event; - FILL_COMMON_FIELDS(fork_event, event, raw->data); + FILL_COMMON_FIELDS(fork_event, event, data); - FILL_ARRAY(fork_event, parent_comm, event, raw->data); - FILL_FIELD(fork_event, parent_pid, event, raw->data); - FILL_ARRAY(fork_event, child_comm, event, raw->data); - FILL_FIELD(fork_event, child_pid, event, raw->data); + FILL_ARRAY(fork_event, parent_comm, event, data); + FILL_FIELD(fork_event, parent_pid, event, data); + FILL_ARRAY(fork_event, child_comm, event, data); + FILL_FIELD(fork_event, child_pid, event, data); if (trace_handler->fork_event) - trace_handler->fork_event(&fork_event, event, cpu, timestamp, thread); + trace_handler->fork_event(&fork_event, event, + cpu, timestamp, thread); } static void @@ -1550,7 +1559,7 @@ process_sched_exit_event(struct event *event, } static void -process_sched_migrate_task_event(struct raw_event_sample *raw, +process_sched_migrate_task_event(void *data, struct perf_session *session, struct event *event, int cpu __used, u64 timestamp __used, @@ -1558,80 +1567,64 @@ process_sched_migrate_task_event(struct raw_event_sample *raw, { struct trace_migrate_task_event migrate_task_event; - FILL_COMMON_FIELDS(migrate_task_event, event, raw->data); + FILL_COMMON_FIELDS(migrate_task_event, event, data); - FILL_ARRAY(migrate_task_event, comm, event, raw->data); - FILL_FIELD(migrate_task_event, pid, event, raw->data); - FILL_FIELD(migrate_task_event, prio, event, raw->data); - FILL_FIELD(migrate_task_event, cpu, event, raw->data); + FILL_ARRAY(migrate_task_event, comm, event, data); + FILL_FIELD(migrate_task_event, pid, event, data); + FILL_FIELD(migrate_task_event, prio, event, data); + FILL_FIELD(migrate_task_event, cpu, event, data); if (trace_handler->migrate_task_event) - trace_handler->migrate_task_event(&migrate_task_event, event, cpu, timestamp, thread); + trace_handler->migrate_task_event(&migrate_task_event, session, + event, cpu, timestamp, thread); } static void -process_raw_event(event_t *raw_event __used, void *more_data, - int cpu, u64 timestamp, struct thread *thread) +process_raw_event(event_t *raw_event __used, struct perf_session *session, + void *data, int cpu, u64 timestamp, struct thread *thread) { - struct raw_event_sample *raw = more_data; struct event *event; int type; - type = trace_parse_common_type(raw->data); + + type = trace_parse_common_type(data); event = trace_find_event(type); if (!strcmp(event->name, "sched_switch")) - process_sched_switch_event(raw, event, cpu, timestamp, thread); + process_sched_switch_event(data, session, event, cpu, timestamp, thread); if (!strcmp(event->name, "sched_stat_runtime")) - process_sched_runtime_event(raw, event, cpu, timestamp, thread); + process_sched_runtime_event(data, session, event, cpu, timestamp, thread); if (!strcmp(event->name, "sched_wakeup")) - process_sched_wakeup_event(raw, event, cpu, timestamp, thread); + process_sched_wakeup_event(data, session, event, cpu, timestamp, thread); if (!strcmp(event->name, "sched_wakeup_new")) - process_sched_wakeup_event(raw, event, cpu, timestamp, thread); + process_sched_wakeup_event(data, session, event, cpu, timestamp, thread); if (!strcmp(event->name, "sched_process_fork")) - process_sched_fork_event(raw, event, cpu, timestamp, thread); + process_sched_fork_event(data, event, cpu, timestamp, thread); if (!strcmp(event->name, "sched_process_exit")) process_sched_exit_event(event, cpu, timestamp, thread); if (!strcmp(event->name, "sched_migrate_task")) - process_sched_migrate_task_event(raw, event, cpu, timestamp, thread); + process_sched_migrate_task_event(data, session, event, cpu, timestamp, thread); } -static int process_sample_event(event_t *event) +static int process_sample_event(event_t *event, struct perf_session *session) { + struct sample_data data; struct thread *thread; - u64 ip = event->ip.ip; - u64 timestamp = -1; - u32 cpu = -1; - u64 period = 1; - void *more_data = event->ip.__more_data; - if (!(sample_type & PERF_SAMPLE_RAW)) + if (!(session->sample_type & PERF_SAMPLE_RAW)) return 0; - thread = threads__findnew(event->ip.pid); - - if (sample_type & PERF_SAMPLE_TIME) { - timestamp = *(u64 *)more_data; - more_data += sizeof(u64); - } - - if (sample_type & PERF_SAMPLE_CPU) { - cpu = *(u32 *)more_data; - more_data += sizeof(u32); - more_data += sizeof(u32); /* reserved */ - } + memset(&data, 0, sizeof(data)); + data.time = -1; + data.cpu = -1; + data.period = -1; - if (sample_type & PERF_SAMPLE_PERIOD) { - period = *(u64 *)more_data; - more_data += sizeof(u64); - } + event__parse_sample(event, session->sample_type, &data); - dump_printf("(IP, %d): %d/%d: %p period: %Ld\n", - event->header.misc, - event->ip.pid, event->ip.tid, - (void *)(long)ip, - (long long)period); + dump_printf("(IP, %d): %d/%d: %#Lx period: %Ld\n", event->header.misc, + data.pid, data.tid, data.ip, data.period); + thread = perf_session__findnew(session, data.pid); if (thread == NULL) { pr_debug("problem processing %d event, skipping it.\n", event->header.type); @@ -1640,15 +1633,16 @@ static int process_sample_event(event_t *event) dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid); - if (profile_cpu != -1 && profile_cpu != (int) cpu) + if (profile_cpu != -1 && profile_cpu != (int)data.cpu) return 0; - process_raw_event(event, more_data, cpu, timestamp, thread); + process_raw_event(event, session, data.raw_data, data.cpu, data.time, thread); return 0; } -static int process_lost_event(event_t *event __used) +static int process_lost_event(event_t *event __used, + struct perf_session *session __used) { nr_lost_chunks++; nr_lost_events += event->lost.lost; @@ -1656,34 +1650,24 @@ static int process_lost_event(event_t *event __used) return 0; } -static int sample_type_check(u64 type) -{ - sample_type = type; - - if (!(sample_type & PERF_SAMPLE_RAW)) { - fprintf(stderr, - "No trace sample to read. Did you call perf record " - "without -R?"); - return -1; - } - - return 0; -} - -static struct perf_file_handler file_handler = { - .process_sample_event = process_sample_event, - .process_comm_event = event__process_comm, - .process_lost_event = process_lost_event, - .sample_type_check = sample_type_check, +static struct perf_event_ops event_ops = { + .sample = process_sample_event, + .comm = event__process_comm, + .lost = process_lost_event, }; static int read_events(void) { - register_idle_thread(); - register_perf_file_handler(&file_handler); + int err = -EINVAL; + struct perf_session *session = perf_session__new(input_name, O_RDONLY, 0); + if (session == NULL) + return -ENOMEM; - return mmap_dispatch_perf_file(&header, input_name, 0, 0, - &event__cwdlen, &event__cwd); + if (perf_session__has_traces(session, "record -R")) + err = perf_session__process_events(session, &event_ops); + + perf_session__delete(session); + return err; } static void print_bad_events(void) @@ -1724,9 +1708,9 @@ static void __cmd_lat(void) read_events(); sort_lat(); - printf("\n -----------------------------------------------------------------------------------------\n"); - printf(" Task | Runtime ms | Switches | Average delay ms | Maximum delay ms |\n"); - printf(" -----------------------------------------------------------------------------------------\n"); + printf("\n ---------------------------------------------------------------------------------------------------------------\n"); + printf(" Task | Runtime ms | Switches | Average delay ms | Maximum delay ms | Maximum delay at |\n"); + printf(" ---------------------------------------------------------------------------------------------------------------\n"); next = rb_first(&sorted_atom_root); @@ -1902,13 +1886,18 @@ static int __cmd_record(int argc, const char **argv) int cmd_sched(int argc, const char **argv, const char *prefix __used) { - symbol__init(0); - argc = parse_options(argc, argv, sched_options, sched_usage, PARSE_OPT_STOP_AT_NON_OPTION); if (!argc) usage_with_options(sched_usage, sched_options); + /* + * Aliased to 'perf trace' for now: + */ + if (!strcmp(argv[0], "trace")) + return cmd_trace(argc, argv, prefix); + + symbol__init(); if (!strncmp(argv[0], "rec", 3)) { return __cmd_record(argc, argv); } else if (!strncmp(argv[0], "lat", 3)) { @@ -1932,11 +1921,6 @@ int cmd_sched(int argc, const char **argv, const char *prefix __used) usage_with_options(replay_usage, replay_options); } __cmd_replay(); - } else if (!strcmp(argv[0], "trace")) { - /* - * Aliased to 'perf trace' for now: - */ - return cmd_trace(argc, argv, prefix); } else { usage_with_options(sched_usage, sched_options); } diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c index c70d72003557..e8c85d5aec41 100644 --- a/tools/perf/builtin-stat.c +++ b/tools/perf/builtin-stat.c @@ -44,6 +44,7 @@ #include "util/parse-events.h" #include "util/event.h" #include "util/debug.h" +#include "util/header.h" #include <sys/prctl.h> #include <math.h> @@ -79,6 +80,8 @@ static int fd[MAX_NR_CPUS][MAX_COUNTERS]; static int event_scaled[MAX_COUNTERS]; +static volatile int done = 0; + struct stats { double n, mean, M2; @@ -247,61 +250,64 @@ static int run_perf_stat(int argc __used, const char **argv) unsigned long long t0, t1; int status = 0; int counter; - int pid; + int pid = target_pid; int child_ready_pipe[2], go_pipe[2]; + const bool forks = (target_pid == -1 && argc > 0); char buf; if (!system_wide) nr_cpus = 1; - if (pipe(child_ready_pipe) < 0 || pipe(go_pipe) < 0) { + if (forks && (pipe(child_ready_pipe) < 0 || pipe(go_pipe) < 0)) { perror("failed to create pipes"); exit(1); } - if ((pid = fork()) < 0) - perror("failed to fork"); + if (forks) { + if ((pid = fork()) < 0) + perror("failed to fork"); + + if (!pid) { + close(child_ready_pipe[0]); + close(go_pipe[1]); + fcntl(go_pipe[0], F_SETFD, FD_CLOEXEC); + + /* + * Do a dummy execvp to get the PLT entry resolved, + * so we avoid the resolver overhead on the real + * execvp call. + */ + execvp("", (char **)argv); + + /* + * Tell the parent we're ready to go + */ + close(child_ready_pipe[1]); + + /* + * Wait until the parent tells us to go. + */ + if (read(go_pipe[0], &buf, 1) == -1) + perror("unable to read pipe"); + + execvp(argv[0], (char **)argv); + + perror(argv[0]); + exit(-1); + } - if (!pid) { - close(child_ready_pipe[0]); - close(go_pipe[1]); - fcntl(go_pipe[0], F_SETFD, FD_CLOEXEC); + child_pid = pid; /* - * Do a dummy execvp to get the PLT entry resolved, - * so we avoid the resolver overhead on the real - * execvp call. - */ - execvp("", (char **)argv); - - /* - * Tell the parent we're ready to go + * Wait for the child to be ready to exec. */ close(child_ready_pipe[1]); - - /* - * Wait until the parent tells us to go. - */ - if (read(go_pipe[0], &buf, 1) == -1) + close(go_pipe[0]); + if (read(child_ready_pipe[0], &buf, 1) == -1) perror("unable to read pipe"); - - execvp(argv[0], (char **)argv); - - perror(argv[0]); - exit(-1); + close(child_ready_pipe[0]); } - child_pid = pid; - - /* - * Wait for the child to be ready to exec. - */ - close(child_ready_pipe[1]); - close(go_pipe[0]); - if (read(child_ready_pipe[0], &buf, 1) == -1) - perror("unable to read pipe"); - close(child_ready_pipe[0]); - for (counter = 0; counter < nr_counters; counter++) create_perf_stat_counter(counter, pid); @@ -310,8 +316,12 @@ static int run_perf_stat(int argc __used, const char **argv) */ t0 = rdclock(); - close(go_pipe[1]); - wait(&status); + if (forks) { + close(go_pipe[1]); + wait(&status); + } else { + while(!done); + } t1 = rdclock(); @@ -417,10 +427,13 @@ static void print_stat(int argc, const char **argv) fflush(stdout); fprintf(stderr, "\n"); - fprintf(stderr, " Performance counter stats for \'%s", argv[0]); - - for (i = 1; i < argc; i++) - fprintf(stderr, " %s", argv[i]); + fprintf(stderr, " Performance counter stats for "); + if(target_pid == -1) { + fprintf(stderr, "\'%s", argv[0]); + for (i = 1; i < argc; i++) + fprintf(stderr, " %s", argv[i]); + }else + fprintf(stderr, "task pid \'%d", target_pid); fprintf(stderr, "\'"); if (run_count > 1) @@ -445,6 +458,9 @@ static volatile int signr = -1; static void skip_signal(int signo) { + if(target_pid != -1) + done = 1; + signr = signo; } @@ -461,7 +477,7 @@ static void sig_atexit(void) } static const char * const stat_usage[] = { - "perf stat [<options>] <command>", + "perf stat [<options>] [<command>]", NULL }; @@ -492,7 +508,7 @@ int cmd_stat(int argc, const char **argv, const char *prefix __used) argc = parse_options(argc, argv, options, stat_usage, PARSE_OPT_STOP_AT_NON_OPTION); - if (!argc) + if (!argc && target_pid == -1) usage_with_options(stat_usage, options); if (run_count <= 0) usage_with_options(stat_usage, options); diff --git a/tools/perf/builtin-timechart.c b/tools/perf/builtin-timechart.c index cb58b6605fcc..0d4d8ff7914b 100644 --- a/tools/perf/builtin-timechart.c +++ b/tools/perf/builtin-timechart.c @@ -30,15 +30,12 @@ #include "util/parse-options.h" #include "util/parse-events.h" #include "util/event.h" -#include "util/data_map.h" +#include "util/session.h" #include "util/svghelper.h" static char const *input_name = "perf.data"; static char const *output_name = "output.svg"; - -static u64 sample_type; - static unsigned int numcpus; static u64 min_freq; /* Lowest CPU frequency seen */ static u64 max_freq; /* Highest CPU frequency seen */ @@ -281,33 +278,30 @@ static int cpus_cstate_state[MAX_CPUS]; static u64 cpus_pstate_start_times[MAX_CPUS]; static u64 cpus_pstate_state[MAX_CPUS]; -static int -process_comm_event(event_t *event) +static int process_comm_event(event_t *event, struct perf_session *session __used) { - pid_set_comm(event->comm.pid, event->comm.comm); + pid_set_comm(event->comm.tid, event->comm.comm); return 0; } -static int -process_fork_event(event_t *event) + +static int process_fork_event(event_t *event, struct perf_session *session __used) { pid_fork(event->fork.pid, event->fork.ppid, event->fork.time); return 0; } -static int -process_exit_event(event_t *event) +static int process_exit_event(event_t *event, struct perf_session *session __used) { pid_exit(event->fork.pid, event->fork.time); return 0; } struct trace_entry { - u32 size; unsigned short type; unsigned char flags; unsigned char preempt_count; int pid; - int tgid; + int lock_depth; }; struct power_entry { @@ -481,46 +475,24 @@ static void sched_switch(int cpu, u64 timestamp, struct trace_entry *te) } -static int -process_sample_event(event_t *event) +static int process_sample_event(event_t *event, struct perf_session *session) { - int cursor = 0; - u64 addr = 0; - u64 stamp = 0; - u32 cpu = 0; - u32 pid = 0; + struct sample_data data; struct trace_entry *te; - if (sample_type & PERF_SAMPLE_IP) - cursor++; - - if (sample_type & PERF_SAMPLE_TID) { - pid = event->sample.array[cursor]>>32; - cursor++; - } - if (sample_type & PERF_SAMPLE_TIME) { - stamp = event->sample.array[cursor++]; + memset(&data, 0, sizeof(data)); - if (!first_time || first_time > stamp) - first_time = stamp; - if (last_time < stamp) - last_time = stamp; + event__parse_sample(event, session->sample_type, &data); + if (session->sample_type & PERF_SAMPLE_TIME) { + if (!first_time || first_time > data.time) + first_time = data.time; + if (last_time < data.time) + last_time = data.time; } - if (sample_type & PERF_SAMPLE_ADDR) - addr = event->sample.array[cursor++]; - if (sample_type & PERF_SAMPLE_ID) - cursor++; - if (sample_type & PERF_SAMPLE_STREAM_ID) - cursor++; - if (sample_type & PERF_SAMPLE_CPU) - cpu = event->sample.array[cursor++] & 0xFFFFFFFF; - if (sample_type & PERF_SAMPLE_PERIOD) - cursor++; - te = (void *)&event->sample.array[cursor]; - - if (sample_type & PERF_SAMPLE_RAW && te->size > 0) { + te = (void *)data.raw_data; + if (session->sample_type & PERF_SAMPLE_RAW && data.raw_size > 0) { char *event_str; struct power_entry *pe; @@ -532,19 +504,19 @@ process_sample_event(event_t *event) return 0; if (strcmp(event_str, "power:power_start") == 0) - c_state_start(cpu, stamp, pe->value); + c_state_start(data.cpu, data.time, pe->value); if (strcmp(event_str, "power:power_end") == 0) - c_state_end(cpu, stamp); + c_state_end(data.cpu, data.time); if (strcmp(event_str, "power:power_frequency") == 0) - p_state_change(cpu, stamp, pe->value); + p_state_change(data.cpu, data.time, pe->value); if (strcmp(event_str, "sched:sched_wakeup") == 0) - sched_wakeup(cpu, stamp, pid, te); + sched_wakeup(data.cpu, data.time, data.pid, te); if (strcmp(event_str, "sched:sched_switch") == 0) - sched_switch(cpu, stamp, te); + sched_switch(data.cpu, data.time, te); } return 0; } @@ -597,16 +569,16 @@ static void end_sample_processing(void) } } -static u64 sample_time(event_t *event) +static u64 sample_time(event_t *event, const struct perf_session *session) { int cursor; cursor = 0; - if (sample_type & PERF_SAMPLE_IP) + if (session->sample_type & PERF_SAMPLE_IP) cursor++; - if (sample_type & PERF_SAMPLE_TID) + if (session->sample_type & PERF_SAMPLE_TID) cursor++; - if (sample_type & PERF_SAMPLE_TIME) + if (session->sample_type & PERF_SAMPLE_TIME) return event->sample.array[cursor]; return 0; } @@ -616,8 +588,7 @@ static u64 sample_time(event_t *event) * We first queue all events, sorted backwards by insertion. * The order will get flipped later. */ -static int -queue_sample_event(event_t *event) +static int queue_sample_event(event_t *event, struct perf_session *session) { struct sample_wrapper *copy, *prev; int size; @@ -631,7 +602,7 @@ queue_sample_event(event_t *event) memset(copy, 0, size); copy->next = NULL; - copy->timestamp = sample_time(event); + copy->timestamp = sample_time(event, session); memcpy(©->data, event, event->sample.header.size); @@ -1043,7 +1014,7 @@ static void write_svg_file(const char *filename) svg_close(); } -static void process_samples(void) +static void process_samples(struct perf_session *session) { struct sample_wrapper *cursor; event_t *event; @@ -1054,44 +1025,33 @@ static void process_samples(void) while (cursor) { event = (void *)&cursor->data; cursor = cursor->next; - process_sample_event(event); + process_sample_event(event, session); } } -static int sample_type_check(u64 type) -{ - sample_type = type; - - if (!(sample_type & PERF_SAMPLE_RAW)) { - fprintf(stderr, "No trace samples found in the file.\n" - "Have you used 'perf timechart record' to record it?\n"); - return -1; - } - - return 0; -} - -static struct perf_file_handler file_handler = { - .process_comm_event = process_comm_event, - .process_fork_event = process_fork_event, - .process_exit_event = process_exit_event, - .process_sample_event = queue_sample_event, - .sample_type_check = sample_type_check, +static struct perf_event_ops event_ops = { + .comm = process_comm_event, + .fork = process_fork_event, + .exit = process_exit_event, + .sample = queue_sample_event, }; static int __cmd_timechart(void) { - struct perf_header *header; - int ret; + struct perf_session *session = perf_session__new(input_name, O_RDONLY, 0); + int ret = -EINVAL; + + if (session == NULL) + return -ENOMEM; - register_perf_file_handler(&file_handler); + if (!perf_session__has_traces(session, "timechart record")) + goto out_delete; - ret = mmap_dispatch_perf_file(&header, input_name, 0, 0, - &event__cwdlen, &event__cwd); + ret = perf_session__process_events(session, &event_ops); if (ret) - return EXIT_FAILURE; + goto out_delete; - process_samples(); + process_samples(session); end_sample_processing(); @@ -1101,8 +1061,9 @@ static int __cmd_timechart(void) pr_info("Written %2.1f seconds of trace to %s.\n", (last_time - first_time) / 1000000000.0, output_name); - - return EXIT_SUCCESS; +out_delete: + perf_session__delete(session); + return ret; } static const char * const timechart_usage[] = { @@ -1167,11 +1128,11 @@ static const struct option options[] = { int cmd_timechart(int argc, const char **argv, const char *prefix __used) { - symbol__init(0); - argc = parse_options(argc, argv, options, timechart_usage, PARSE_OPT_STOP_AT_NON_OPTION); + symbol__init(); + if (argc && !strncmp(argv[0], "rec", 3)) return __cmd_record(argc, argv); else if (argc) diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c index e0a374d0e43a..31f2e597800c 100644 --- a/tools/perf/builtin-top.c +++ b/tools/perf/builtin-top.c @@ -20,8 +20,9 @@ #include "perf.h" -#include "util/symbol.h" #include "util/color.h" +#include "util/session.h" +#include "util/symbol.h" #include "util/thread.h" #include "util/util.h" #include <linux/rbtree.h> @@ -79,7 +80,6 @@ static int dump_symtab = 0; static bool hide_kernel_symbols = false; static bool hide_user_symbols = false; static struct winsize winsize; -struct symbol_conf symbol_conf; /* * Source @@ -94,6 +94,7 @@ struct source_line { static char *sym_filter = NULL; struct sym_entry *sym_filter_entry = NULL; +struct sym_entry *sym_filter_entry_sched = NULL; static int sym_pcnt_filter = 5; static int sym_counter = 0; static int display_weighted = -1; @@ -201,10 +202,9 @@ static void parse_source(struct sym_entry *syme) len = sym->end - sym->start; sprintf(command, - "objdump --start-address=0x%016Lx " - "--stop-address=0x%016Lx -dS %s", - map->unmap_ip(map, sym->start), - map->unmap_ip(map, sym->end), path); + "objdump --start-address=%#0*Lx --stop-address=%#0*Lx -dS %s", + BITS_PER_LONG / 4, map__rip_2objdump(map, sym->start), + BITS_PER_LONG / 4, map__rip_2objdump(map, sym->end), path); file = popen(command, "r"); if (!file) @@ -215,7 +215,7 @@ static void parse_source(struct sym_entry *syme) while (!feof(file)) { struct source_line *src; size_t dummy = 0; - char *c; + char *c, *sep; src = malloc(sizeof(struct source_line)); assert(src != NULL); @@ -234,14 +234,11 @@ static void parse_source(struct sym_entry *syme) *source->lines_tail = src; source->lines_tail = &src->next; - if (strlen(src->line)>8 && src->line[8] == ':') { - src->eip = strtoull(src->line, NULL, 16); - src->eip = map->unmap_ip(map, src->eip); - } - if (strlen(src->line)>8 && src->line[16] == ':') { - src->eip = strtoull(src->line, NULL, 16); - src->eip = map->unmap_ip(map, src->eip); - } + src->eip = strtoull(src->line, &sep, 16); + if (*sep == ':') + src->eip = map__objdump_2ip(map, src->eip); + else /* this line has no ip info (e.g. source line) */ + src->eip = 0; } pclose(file); out_assign: @@ -276,6 +273,9 @@ static void record_precise_ip(struct sym_entry *syme, int counter, u64 ip) goto out_unlock; for (line = syme->src->lines; line; line = line->next) { + /* skip lines without IP info */ + if (line->eip == 0) + continue; if (line->eip == ip) { line->count[counter]++; break; @@ -287,17 +287,20 @@ out_unlock: pthread_mutex_unlock(&syme->src->lock); } +#define PATTERN_LEN (BITS_PER_LONG / 4 + 2) + static void lookup_sym_source(struct sym_entry *syme) { struct symbol *symbol = sym_entry__symbol(syme); struct source_line *line; - char pattern[PATH_MAX]; + char pattern[PATTERN_LEN + 1]; - sprintf(pattern, "<%s>:", symbol->name); + sprintf(pattern, "%0*Lx <", BITS_PER_LONG / 4, + map__rip_2objdump(syme->map, symbol->start)); pthread_mutex_lock(&syme->src->lock); for (line = syme->src->lines; line; line = line->next) { - if (strstr(line->line, pattern)) { + if (memcmp(line->line, pattern, PATTERN_LEN) == 0) { syme->src->source = line; break; } @@ -667,7 +670,7 @@ static void prompt_symbol(struct sym_entry **target, const char *msg) } if (!found) { - fprintf(stderr, "Sorry, %s is not active.\n", sym_filter); + fprintf(stderr, "Sorry, %s is not active.\n", buf); sleep(1); return; } else @@ -695,17 +698,15 @@ static void print_mapped_keys(void) fprintf(stdout, "\t[f] profile display filter (count). \t(%d)\n", count_filter); - if (symbol_conf.vmlinux_name) { - fprintf(stdout, "\t[F] annotate display filter (percent). \t(%d%%)\n", sym_pcnt_filter); - fprintf(stdout, "\t[s] annotate symbol. \t(%s)\n", name?: "NULL"); - fprintf(stdout, "\t[S] stop annotation.\n"); - } + fprintf(stdout, "\t[F] annotate display filter (percent). \t(%d%%)\n", sym_pcnt_filter); + fprintf(stdout, "\t[s] annotate symbol. \t(%s)\n", name?: "NULL"); + fprintf(stdout, "\t[S] stop annotation.\n"); if (nr_counters > 1) fprintf(stdout, "\t[w] toggle display weighted/count[E]r. \t(%d)\n", display_weighted ? 1 : 0); fprintf(stdout, - "\t[K] hide kernel_symbols symbols. \t(%s)\n", + "\t[K] hide kernel_symbols symbols. \t(%s)\n", hide_kernel_symbols ? "yes" : "no"); fprintf(stdout, "\t[U] hide user symbols. \t(%s)\n", @@ -725,14 +726,13 @@ static int key_mapped(int c) case 'Q': case 'K': case 'U': + case 'F': + case 's': + case 'S': return 1; case 'E': case 'w': return nr_counters > 1 ? 1 : 0; - case 'F': - case 's': - case 'S': - return symbol_conf.vmlinux_name ? 1 : 0; default: break; } @@ -910,8 +910,12 @@ static int symbol_filter(struct map *map, struct symbol *sym) syme = symbol__priv(sym); syme->map = map; syme->src = NULL; - if (!sym_filter_entry && sym_filter && !strcmp(name, sym_filter)) - sym_filter_entry = syme; + + if (!sym_filter_entry && sym_filter && !strcmp(name, sym_filter)) { + /* schedule initial sym_filter_entry setup */ + sym_filter_entry_sched = syme; + sym_filter = NULL; + } for (i = 0; skip_symbols[i]; i++) { if (!strcmp(skip_symbols[i], name)) { @@ -926,15 +930,19 @@ static int symbol_filter(struct map *map, struct symbol *sym) return 0; } -static void event__process_sample(const event_t *self, int counter) +static void event__process_sample(const event_t *self, + struct perf_session *session, int counter) { u64 ip = self->ip.ip; struct sym_entry *syme; struct addr_location al; u8 origin = self->header.misc & PERF_RECORD_MISC_CPUMODE_MASK; + ++samples; + switch (origin) { case PERF_RECORD_MISC_USER: + ++userspace_samples; if (hide_user_symbols) return; break; @@ -946,9 +954,38 @@ static void event__process_sample(const event_t *self, int counter) return; } - if (event__preprocess_sample(self, &al, symbol_filter) < 0 || - al.sym == NULL) + if (event__preprocess_sample(self, session, &al, symbol_filter) < 0 || + al.filtered) + return; + + if (al.sym == NULL) { + /* + * As we do lazy loading of symtabs we only will know if the + * specified vmlinux file is invalid when we actually have a + * hit in kernel space and then try to load it. So if we get + * here and there are _no_ symbols in the DSO backing the + * kernel map, bail out. + * + * We may never get here, for instance, if we use -K/ + * --hide-kernel-symbols, even if the user specifies an + * invalid --vmlinux ;-) + */ + if (al.map == session->vmlinux_maps[MAP__FUNCTION] && + RB_EMPTY_ROOT(&al.map->dso->symbols[MAP__FUNCTION])) { + pr_err("The %s file can't be used\n", + symbol_conf.vmlinux_name); + exit(1); + } + return; + } + + /* let's see, whether we need to install initial sym_filter_entry */ + if (sym_filter_entry_sched) { + sym_filter_entry = sym_filter_entry_sched; + sym_filter_entry_sched = NULL; + parse_source(sym_filter_entry); + } syme = symbol__priv(al.sym); if (!syme->skip) { @@ -959,20 +996,21 @@ static void event__process_sample(const event_t *self, int counter) if (list_empty(&syme->node) || !syme->node.next) __list_insert_active_sym(syme); pthread_mutex_unlock(&active_symbols_lock); - if (origin == PERF_RECORD_MISC_USER) - ++userspace_samples; - ++samples; } } -static int event__process(event_t *event) +static int event__process(event_t *event, struct perf_session *session) { switch (event->header.type) { case PERF_RECORD_COMM: - event__process_comm(event); + event__process_comm(event, session); break; case PERF_RECORD_MMAP: - event__process_mmap(event); + event__process_mmap(event, session); + break; + case PERF_RECORD_FORK: + case PERF_RECORD_EXIT: + event__process_task(event, session); break; default: break; @@ -999,7 +1037,8 @@ static unsigned int mmap_read_head(struct mmap_data *md) return head; } -static void mmap_read_counter(struct mmap_data *md) +static void perf_session__mmap_read_counter(struct perf_session *self, + struct mmap_data *md) { unsigned int head = mmap_read_head(md); unsigned int old = md->prev; @@ -1052,9 +1091,9 @@ static void mmap_read_counter(struct mmap_data *md) } if (event->header.type == PERF_RECORD_SAMPLE) - event__process_sample(event, md->counter); + event__process_sample(event, self, md->counter); else - event__process(event); + event__process(event, self); old += size; } @@ -1064,13 +1103,13 @@ static void mmap_read_counter(struct mmap_data *md) static struct pollfd event_array[MAX_NR_CPUS * MAX_COUNTERS]; static struct mmap_data mmap_array[MAX_NR_CPUS][MAX_COUNTERS]; -static void mmap_read(void) +static void perf_session__mmap_read(struct perf_session *self) { int i, counter; for (i = 0; i < nr_cpus; i++) { for (counter = 0; counter < nr_counters; counter++) - mmap_read_counter(&mmap_array[i][counter]); + perf_session__mmap_read_counter(self, &mmap_array[i][counter]); } } @@ -1155,11 +1194,18 @@ static int __cmd_top(void) pthread_t thread; int i, counter; int ret; + /* + * FIXME: perf_session__new should allow passing a O_MMAP, so that all this + * mmap reading, etc is encapsulated in it. Use O_WRONLY for now. + */ + struct perf_session *session = perf_session__new(NULL, O_WRONLY, false); + if (session == NULL) + return -ENOMEM; if (target_pid != -1) - event__synthesize_thread(target_pid, event__process); + event__synthesize_thread(target_pid, event__process, session); else - event__synthesize_threads(event__process); + event__synthesize_threads(event__process, session); for (i = 0; i < nr_cpus; i++) { group_fd = -1; @@ -1170,7 +1216,7 @@ static int __cmd_top(void) /* Wait for a minimal set of events before starting the snapshot */ poll(event_array, nr_poll, 100); - mmap_read(); + perf_session__mmap_read(session); if (pthread_create(&thread, NULL, display_thread, NULL)) { printf("Could not create display thread.\n"); @@ -1190,7 +1236,7 @@ static int __cmd_top(void) while (1) { int hits = samples; - mmap_read(); + perf_session__mmap_read(session); if (hits == samples) ret = poll(event_array, nr_poll, 100); @@ -1235,7 +1281,7 @@ static const struct option options[] = { OPT_BOOLEAN('i', "inherit", &inherit, "child tasks inherit counters"), OPT_STRING('s', "sym-annotate", &sym_filter, "symbol name", - "symbol to annotate - requires -k option"), + "symbol to annotate"), OPT_BOOLEAN('z', "zero", &zero, "zero history across updates"), OPT_INTEGER('F', "freq", &freq, @@ -1271,16 +1317,14 @@ int cmd_top(int argc, const char **argv, const char *prefix __used) symbol_conf.priv_size = (sizeof(struct sym_entry) + (nr_counters + 1) * sizeof(unsigned long)); - if (symbol_conf.vmlinux_name == NULL) - symbol_conf.try_vmlinux_path = true; - if (symbol__init(&symbol_conf) < 0) + + symbol_conf.try_vmlinux_path = (symbol_conf.vmlinux_name == NULL); + if (symbol__init() < 0) return -1; if (delay_secs < 1) delay_secs = 1; - parse_source(sym_filter_entry); - /* * User specified count overrides default frequency. */ diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index abb914aa7be6..5db687fc13de 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -7,11 +7,14 @@ #include "util/header.h" #include "util/exec_cmd.h" #include "util/trace-event.h" +#include "util/session.h" static char const *script_name; static char const *generate_script_lang; -static int default_start_script(const char *script __attribute((unused))) +static int default_start_script(const char *script __unused, + int argc __unused, + const char **argv __unused) { return 0; } @@ -21,7 +24,7 @@ static int default_stop_script(void) return 0; } -static int default_generate_script(const char *outfile __attribute ((unused))) +static int default_generate_script(const char *outfile __unused) { return 0; } @@ -41,6 +44,7 @@ static void setup_scripting(void) perf_set_argv_exec_path(perf_exec_path()); setup_perl_scripting(); + setup_python_scripting(); scripting_ops = &default_scripting_ops; } @@ -56,99 +60,55 @@ static int cleanup_scripting(void) #include "util/debug.h" #include "util/trace-event.h" -#include "util/data_map.h" #include "util/exec_cmd.h" static char const *input_name = "perf.data"; -static struct perf_header *header; -static u64 sample_type; - -static int process_sample_event(event_t *event) +static int process_sample_event(event_t *event, struct perf_session *session) { - u64 ip = event->ip.ip; - u64 timestamp = -1; - u32 cpu = -1; - u64 period = 1; - void *more_data = event->ip.__more_data; - struct thread *thread = threads__findnew(event->ip.pid); - - if (sample_type & PERF_SAMPLE_TIME) { - timestamp = *(u64 *)more_data; - more_data += sizeof(u64); - } + struct sample_data data; + struct thread *thread; - if (sample_type & PERF_SAMPLE_CPU) { - cpu = *(u32 *)more_data; - more_data += sizeof(u32); - more_data += sizeof(u32); /* reserved */ - } + memset(&data, 0, sizeof(data)); + data.time = -1; + data.cpu = -1; + data.period = 1; - if (sample_type & PERF_SAMPLE_PERIOD) { - period = *(u64 *)more_data; - more_data += sizeof(u64); - } + event__parse_sample(event, session->sample_type, &data); - dump_printf("(IP, %d): %d/%d: %p period: %Ld\n", - event->header.misc, - event->ip.pid, event->ip.tid, - (void *)(long)ip, - (long long)period); + dump_printf("(IP, %d): %d/%d: %#Lx period: %Ld\n", event->header.misc, + data.pid, data.tid, data.ip, data.period); + thread = perf_session__findnew(session, event->ip.pid); if (thread == NULL) { pr_debug("problem processing %d event, skipping it.\n", event->header.type); return -1; } - dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid); - - if (sample_type & PERF_SAMPLE_RAW) { - struct { - u32 size; - char data[0]; - } *raw = more_data; - + if (session->sample_type & PERF_SAMPLE_RAW) { /* * FIXME: better resolve from pid from the struct trace_entry * field, although it should be the same than this perf * event pid */ - scripting_ops->process_event(cpu, raw->data, raw->size, - timestamp, thread->comm); + scripting_ops->process_event(data.cpu, data.raw_data, + data.raw_size, + data.time, thread->comm); } - event__stats.total += period; + session->events_stats.total += data.period; return 0; } -static int sample_type_check(u64 type) -{ - sample_type = type; - - if (!(sample_type & PERF_SAMPLE_RAW)) { - fprintf(stderr, - "No trace sample to read. Did you call perf record " - "without -R?"); - return -1; - } - - return 0; -} - -static struct perf_file_handler file_handler = { - .process_sample_event = process_sample_event, - .process_comm_event = event__process_comm, - .sample_type_check = sample_type_check, +static struct perf_event_ops event_ops = { + .sample = process_sample_event, + .comm = event__process_comm, }; -static int __cmd_trace(void) +static int __cmd_trace(struct perf_session *session) { - register_idle_thread(); - register_perf_file_handler(&file_handler); - - return mmap_dispatch_perf_file(&header, input_name, - 0, 0, &event__cwdlen, &event__cwd); + return perf_session__process_events(session, &event_ops); } struct script_spec { @@ -260,9 +220,9 @@ static int parse_scriptname(const struct option *opt __used, const char *script, *ext; int len; - if (strcmp(str, "list") == 0) { + if (strcmp(str, "lang") == 0) { list_available_languages(); - return 0; + exit(0); } script = strchr(str, ':'); @@ -299,7 +259,245 @@ static int parse_scriptname(const struct option *opt __used, return 0; } -static const char * const annotate_usage[] = { +#define for_each_lang(scripts_dir, lang_dirent, lang_next) \ + while (!readdir_r(scripts_dir, &lang_dirent, &lang_next) && \ + lang_next) \ + if (lang_dirent.d_type == DT_DIR && \ + (strcmp(lang_dirent.d_name, ".")) && \ + (strcmp(lang_dirent.d_name, ".."))) + +#define for_each_script(lang_dir, script_dirent, script_next) \ + while (!readdir_r(lang_dir, &script_dirent, &script_next) && \ + script_next) \ + if (script_dirent.d_type != DT_DIR) + + +#define RECORD_SUFFIX "-record" +#define REPORT_SUFFIX "-report" + +struct script_desc { + struct list_head node; + char *name; + char *half_liner; + char *args; +}; + +LIST_HEAD(script_descs); + +static struct script_desc *script_desc__new(const char *name) +{ + struct script_desc *s = zalloc(sizeof(*s)); + + if (s != NULL) + s->name = strdup(name); + + return s; +} + +static void script_desc__delete(struct script_desc *s) +{ + free(s->name); + free(s); +} + +static void script_desc__add(struct script_desc *s) +{ + list_add_tail(&s->node, &script_descs); +} + +static struct script_desc *script_desc__find(const char *name) +{ + struct script_desc *s; + + list_for_each_entry(s, &script_descs, node) + if (strcasecmp(s->name, name) == 0) + return s; + return NULL; +} + +static struct script_desc *script_desc__findnew(const char *name) +{ + struct script_desc *s = script_desc__find(name); + + if (s) + return s; + + s = script_desc__new(name); + if (!s) + goto out_delete_desc; + + script_desc__add(s); + + return s; + +out_delete_desc: + script_desc__delete(s); + + return NULL; +} + +static char *ends_with(char *str, const char *suffix) +{ + size_t suffix_len = strlen(suffix); + char *p = str; + + if (strlen(str) > suffix_len) { + p = str + strlen(str) - suffix_len; + if (!strncmp(p, suffix, suffix_len)) + return p; + } + + return NULL; +} + +static char *ltrim(char *str) +{ + int len = strlen(str); + + while (len && isspace(*str)) { + len--; + str++; + } + + return str; +} + +static int read_script_info(struct script_desc *desc, const char *filename) +{ + char line[BUFSIZ], *p; + FILE *fp; + + fp = fopen(filename, "r"); + if (!fp) + return -1; + + while (fgets(line, sizeof(line), fp)) { + p = ltrim(line); + if (strlen(p) == 0) + continue; + if (*p != '#') + continue; + p++; + if (strlen(p) && *p == '!') + continue; + + p = ltrim(p); + if (strlen(p) && p[strlen(p) - 1] == '\n') + p[strlen(p) - 1] = '\0'; + + if (!strncmp(p, "description:", strlen("description:"))) { + p += strlen("description:"); + desc->half_liner = strdup(ltrim(p)); + continue; + } + + if (!strncmp(p, "args:", strlen("args:"))) { + p += strlen("args:"); + desc->args = strdup(ltrim(p)); + continue; + } + } + + fclose(fp); + + return 0; +} + +static int list_available_scripts(const struct option *opt __used, + const char *s __used, int unset __used) +{ + struct dirent *script_next, *lang_next, script_dirent, lang_dirent; + char scripts_path[MAXPATHLEN]; + DIR *scripts_dir, *lang_dir; + char script_path[MAXPATHLEN]; + char lang_path[MAXPATHLEN]; + struct script_desc *desc; + char first_half[BUFSIZ]; + char *script_root; + char *str; + + snprintf(scripts_path, MAXPATHLEN, "%s/scripts", perf_exec_path()); + + scripts_dir = opendir(scripts_path); + if (!scripts_dir) + return -1; + + for_each_lang(scripts_dir, lang_dirent, lang_next) { + snprintf(lang_path, MAXPATHLEN, "%s/%s/bin", scripts_path, + lang_dirent.d_name); + lang_dir = opendir(lang_path); + if (!lang_dir) + continue; + + for_each_script(lang_dir, script_dirent, script_next) { + script_root = strdup(script_dirent.d_name); + str = ends_with(script_root, REPORT_SUFFIX); + if (str) { + *str = '\0'; + desc = script_desc__findnew(script_root); + snprintf(script_path, MAXPATHLEN, "%s/%s", + lang_path, script_dirent.d_name); + read_script_info(desc, script_path); + } + free(script_root); + } + } + + fprintf(stdout, "List of available trace scripts:\n"); + list_for_each_entry(desc, &script_descs, node) { + sprintf(first_half, "%s %s", desc->name, + desc->args ? desc->args : ""); + fprintf(stdout, " %-36s %s\n", first_half, + desc->half_liner ? desc->half_liner : ""); + } + + exit(0); +} + +static char *get_script_path(const char *script_root, const char *suffix) +{ + struct dirent *script_next, *lang_next, script_dirent, lang_dirent; + char scripts_path[MAXPATHLEN]; + char script_path[MAXPATHLEN]; + DIR *scripts_dir, *lang_dir; + char lang_path[MAXPATHLEN]; + char *str, *__script_root; + char *path = NULL; + + snprintf(scripts_path, MAXPATHLEN, "%s/scripts", perf_exec_path()); + + scripts_dir = opendir(scripts_path); + if (!scripts_dir) + return NULL; + + for_each_lang(scripts_dir, lang_dirent, lang_next) { + snprintf(lang_path, MAXPATHLEN, "%s/%s/bin", scripts_path, + lang_dirent.d_name); + lang_dir = opendir(lang_path); + if (!lang_dir) + continue; + + for_each_script(lang_dir, script_dirent, script_next) { + __script_root = strdup(script_dirent.d_name); + str = ends_with(__script_root, suffix); + if (str) { + *str = '\0'; + if (strcmp(__script_root, script_root)) + continue; + snprintf(script_path, MAXPATHLEN, "%s/%s", + lang_path, script_dirent.d_name); + path = strdup(script_path); + free(__script_root); + break; + } + free(__script_root); + } + } + + return path; +} + +static const char * const trace_usage[] = { "perf trace [<options>] <command>", NULL }; @@ -309,37 +507,81 @@ static const struct option options[] = { "dump raw trace in ASCII"), OPT_BOOLEAN('v', "verbose", &verbose, "be more verbose (show symbol address, etc)"), - OPT_BOOLEAN('l', "latency", &latency_format, + OPT_BOOLEAN('L', "Latency", &latency_format, "show latency attributes (irqs/preemption disabled, etc)"), + OPT_CALLBACK_NOOPT('l', "list", NULL, NULL, "list available scripts", + list_available_scripts), OPT_CALLBACK('s', "script", NULL, "name", "script file name (lang:script name, script name, or *)", parse_scriptname), OPT_STRING('g', "gen-script", &generate_script_lang, "lang", "generate perf-trace.xx script in specified language"), + OPT_STRING('i', "input", &input_name, "file", + "input file name"), OPT_END() }; int cmd_trace(int argc, const char **argv, const char *prefix __used) { - int err; + struct perf_session *session; + const char *suffix = NULL; + const char **__argv; + char *script_path; + int i, err; + + if (argc >= 2 && strncmp(argv[1], "rec", strlen("rec")) == 0) { + if (argc < 3) { + fprintf(stderr, + "Please specify a record script\n"); + return -1; + } + suffix = RECORD_SUFFIX; + } + + if (argc >= 2 && strncmp(argv[1], "rep", strlen("rep")) == 0) { + if (argc < 3) { + fprintf(stderr, + "Please specify a report script\n"); + return -1; + } + suffix = REPORT_SUFFIX; + } - symbol__init(0); + if (suffix) { + script_path = get_script_path(argv[2], suffix); + if (!script_path) { + fprintf(stderr, "script not found\n"); + return -1; + } - setup_scripting(); + __argv = malloc((argc + 1) * sizeof(const char *)); + __argv[0] = "/bin/sh"; + __argv[1] = script_path; + for (i = 3; i < argc; i++) + __argv[i - 1] = argv[i]; + __argv[argc - 1] = NULL; - argc = parse_options(argc, argv, options, annotate_usage, 0); - if (argc) { - /* - * Special case: if there's an argument left then assume tha - * it's a symbol filter: - */ - if (argc > 1) - usage_with_options(annotate_usage, options); + execvp("/bin/sh", (char **)__argv); + exit(-1); } + setup_scripting(); + + argc = parse_options(argc, argv, options, trace_usage, + PARSE_OPT_STOP_AT_NON_OPTION); + + if (symbol__init() < 0) + return -1; setup_pager(); + session = perf_session__new(input_name, O_RDONLY, 0); + if (session == NULL) + return -ENOMEM; + + if (!perf_session__has_traces(session, "record -R")) + return -EINVAL; + if (generate_script_lang) { struct stat perf_stat; @@ -366,23 +608,20 @@ int cmd_trace(int argc, const char **argv, const char *prefix __used) return -1; } - header = perf_header__new(); - if (header == NULL) - return -1; - - perf_header__read(header, input); + perf_header__read(&session->header, input); err = scripting_ops->generate_script("perf-trace"); goto out; } if (script_name) { - err = scripting_ops->start_script(script_name); + err = scripting_ops->start_script(script_name, argc, argv); if (err) goto out; } - err = __cmd_trace(); + err = __cmd_trace(session); + perf_session__delete(session); cleanup_scripting(); out: return err; diff --git a/tools/perf/builtin.h b/tools/perf/builtin.h index a3d8bf65f26c..10fe49e7048a 100644 --- a/tools/perf/builtin.h +++ b/tools/perf/builtin.h @@ -16,7 +16,9 @@ extern int check_pager_config(const char *cmd); extern int cmd_annotate(int argc, const char **argv, const char *prefix); extern int cmd_bench(int argc, const char **argv, const char *prefix); +extern int cmd_buildid_cache(int argc, const char **argv, const char *prefix); extern int cmd_buildid_list(int argc, const char **argv, const char *prefix); +extern int cmd_diff(int argc, const char **argv, const char *prefix); extern int cmd_help(int argc, const char **argv, const char *prefix); extern int cmd_sched(int argc, const char **argv, const char *prefix); extern int cmd_list(int argc, const char **argv, const char *prefix); @@ -29,5 +31,6 @@ extern int cmd_trace(int argc, const char **argv, const char *prefix); extern int cmd_version(int argc, const char **argv, const char *prefix); extern int cmd_probe(int argc, const char **argv, const char *prefix); extern int cmd_kmem(int argc, const char **argv, const char *prefix); +extern int cmd_lock(int argc, const char **argv, const char *prefix); #endif diff --git a/tools/perf/command-list.txt b/tools/perf/command-list.txt index 02b09ea17a3e..9afcff2e3ae5 100644 --- a/tools/perf/command-list.txt +++ b/tools/perf/command-list.txt @@ -3,8 +3,11 @@ # command name category [deprecated] [common] # perf-annotate mainporcelain common +perf-archive mainporcelain common perf-bench mainporcelain common +perf-buildid-cache mainporcelain common perf-buildid-list mainporcelain common +perf-diff mainporcelain common perf-list mainporcelain common perf-sched mainporcelain common perf-record mainporcelain common diff --git a/tools/perf/design.txt b/tools/perf/design.txt index f000c30877ac..bd0bb1b1279b 100644 --- a/tools/perf/design.txt +++ b/tools/perf/design.txt @@ -21,7 +21,7 @@ There's one file descriptor per virtual counter used. The special file descriptor is opened via the perf_event_open() system call: - int sys_perf_event_open(struct perf_event_hw_event *hw_event_uptr, + int sys_perf_event_open(struct perf_event_attr *hw_event_uptr, pid_t pid, int cpu, int group_fd, unsigned long flags); @@ -32,9 +32,9 @@ can be used to set the blocking mode, etc. Multiple counters can be kept open at a time, and the counters can be poll()ed. -When creating a new counter fd, 'perf_event_hw_event' is: +When creating a new counter fd, 'perf_event_attr' is: -struct perf_event_hw_event { +struct perf_event_attr { /* * The MSB of the config word signifies if the rest contains cpu * specific (raw) counter configuration data, if unset, the next @@ -101,10 +101,10 @@ enum hw_event_ids { */ PERF_COUNT_HW_CPU_CYCLES = 0, PERF_COUNT_HW_INSTRUCTIONS = 1, - PERF_COUNT_HW_CACHE_REFERENCES = 2, + PERF_COUNT_HW_CACHE_REFERENCES = 2, PERF_COUNT_HW_CACHE_MISSES = 3, PERF_COUNT_HW_BRANCH_INSTRUCTIONS = 4, - PERF_COUNT_HW_BRANCH_MISSES = 5, + PERF_COUNT_HW_BRANCH_MISSES = 5, PERF_COUNT_HW_BUS_CYCLES = 6, }; @@ -131,8 +131,8 @@ software events, selected by 'event_id': */ enum sw_event_ids { PERF_COUNT_SW_CPU_CLOCK = 0, - PERF_COUNT_SW_TASK_CLOCK = 1, - PERF_COUNT_SW_PAGE_FAULTS = 2, + PERF_COUNT_SW_TASK_CLOCK = 1, + PERF_COUNT_SW_PAGE_FAULTS = 2, PERF_COUNT_SW_CONTEXT_SWITCHES = 3, PERF_COUNT_SW_CPU_MIGRATIONS = 4, PERF_COUNT_SW_PAGE_FAULTS_MIN = 5, @@ -399,7 +399,7 @@ Notification of new events is possible through poll()/select()/epoll() and fcntl() managing signals. Normally a notification is generated for every page filled, however one can -additionally set perf_event_hw_event.wakeup_events to generate one every +additionally set perf_event_attr.wakeup_events to generate one every so many counter overflow events. Future work will include a splice() interface to the ring-buffer. diff --git a/tools/perf/perf-archive.sh b/tools/perf/perf-archive.sh new file mode 100644 index 000000000000..45fbe2f07b15 --- /dev/null +++ b/tools/perf/perf-archive.sh @@ -0,0 +1,32 @@ +#!/bin/bash +# perf archive +# Arnaldo Carvalho de Melo <acme@redhat.com> + +PERF_DATA=perf.data +if [ $# -ne 0 ] ; then + PERF_DATA=$1 +fi + +DEBUGDIR=~/.debug/ +BUILDIDS=$(mktemp /tmp/perf-archive-buildids.XXXXXX) + +perf buildid-list -i $PERF_DATA --with-hits > $BUILDIDS +if [ ! -s $BUILDIDS ] ; then + echo "perf archive: no build-ids found" + rm -f $BUILDIDS + exit 1 +fi + +MANIFEST=$(mktemp /tmp/perf-archive-manifest.XXXXXX) + +cut -d ' ' -f 1 $BUILDIDS | \ +while read build_id ; do + linkname=$DEBUGDIR.build-id/${build_id:0:2}/${build_id:2} + filename=$(readlink -f $linkname) + echo ${linkname#$DEBUGDIR} >> $MANIFEST + echo ${filename#$DEBUGDIR} >> $MANIFEST +done + +tar cfj $PERF_DATA.tar.bz2 -C $DEBUGDIR -T $MANIFEST +rm -f $MANIFEST $BUILDIDS +exit 0 diff --git a/tools/perf/perf.c b/tools/perf/perf.c index cf64049bc9bd..57cb107c1f13 100644 --- a/tools/perf/perf.c +++ b/tools/perf/perf.c @@ -48,7 +48,8 @@ int check_pager_config(const char *cmd) return c.val; } -static void commit_pager_choice(void) { +static void commit_pager_choice(void) +{ switch (use_pager) { case 0: setenv("PERF_PAGER", "cat", 1); @@ -70,7 +71,7 @@ static void set_debugfs_path(void) "tracing/events"); } -static int handle_options(const char*** argv, int* argc, int* envchanged) +static int handle_options(const char ***argv, int *argc, int *envchanged) { int handled = 0; @@ -109,7 +110,7 @@ static int handle_options(const char*** argv, int* argc, int* envchanged) *envchanged = 1; } else if (!strcmp(cmd, "--perf-dir")) { if (*argc < 2) { - fprintf(stderr, "No directory given for --perf-dir.\n" ); + fprintf(stderr, "No directory given for --perf-dir.\n"); usage(perf_usage_string); } setenv(PERF_DIR_ENVIRONMENT, (*argv)[1], 1); @@ -124,7 +125,7 @@ static int handle_options(const char*** argv, int* argc, int* envchanged) *envchanged = 1; } else if (!strcmp(cmd, "--work-tree")) { if (*argc < 2) { - fprintf(stderr, "No directory given for --work-tree.\n" ); + fprintf(stderr, "No directory given for --work-tree.\n"); usage(perf_usage_string); } setenv(PERF_WORK_TREE_ENVIRONMENT, (*argv)[1], 1); @@ -168,7 +169,7 @@ static int handle_alias(int *argcp, const char ***argv) { int envchanged = 0, ret = 0, saved_errno = errno; int count, option_count; - const char** new_argv; + const char **new_argv; const char *alias_command; char *alias_string; @@ -210,11 +211,11 @@ static int handle_alias(int *argcp, const char ***argv) if (!strcmp(alias_command, new_argv[0])) die("recursive alias: %s", alias_command); - new_argv = realloc(new_argv, sizeof(char*) * + new_argv = realloc(new_argv, sizeof(char *) * (count + *argcp + 1)); /* insert after command name */ - memcpy(new_argv + count, *argv + 1, sizeof(char*) * *argcp); - new_argv[count+*argcp] = NULL; + memcpy(new_argv + count, *argv + 1, sizeof(char *) * *argcp); + new_argv[count + *argcp] = NULL; *argv = new_argv; *argcp += count - 1; @@ -285,7 +286,9 @@ static void handle_internal_command(int argc, const char **argv) { const char *cmd = argv[0]; static struct cmd_struct commands[] = { + { "buildid-cache", cmd_buildid_cache, 0 }, { "buildid-list", cmd_buildid_list, 0 }, + { "diff", cmd_diff, 0 }, { "help", cmd_help, 0 }, { "list", cmd_list, 0 }, { "record", cmd_record, 0 }, @@ -300,6 +303,7 @@ static void handle_internal_command(int argc, const char **argv) { "sched", cmd_sched, 0 }, { "probe", cmd_probe, 0 }, { "kmem", cmd_kmem, 0 }, + { "lock", cmd_lock, 0 }, }; unsigned int i; static const char ext[] = STRIP_EXTENSION; @@ -387,7 +391,7 @@ static int run_argv(int *argcp, const char ***argv) /* mini /proc/mounts parser: searching for "^blah /mount/point debugfs" */ static void get_debugfs_mntpt(void) { - const char *path = debugfs_find_mountpoint(); + const char *path = debugfs_mount(NULL); if (path) strncpy(debugfs_mntpt, path, sizeof(debugfs_mntpt)); @@ -448,8 +452,8 @@ int main(int argc, const char **argv) setup_path(); while (1) { - static int done_help = 0; - static int was_alias = 0; + static int done_help; + static int was_alias; was_alias = run_argv(&argc, &argv); if (errno != ENOENT) diff --git a/tools/perf/perf.h b/tools/perf/perf.h index 454d5d55f32d..75f941bfba9e 100644 --- a/tools/perf/perf.h +++ b/tools/perf/perf.h @@ -59,6 +59,18 @@ #define cpu_relax() asm volatile ("hint @pause" ::: "memory") #endif +#ifdef __arm__ +#include "../../arch/arm/include/asm/unistd.h" +/* + * Use the __kuser_memory_barrier helper in the CPU helper page. See + * arch/arm/kernel/entry-armv.S in the kernel source for details. + */ +#define rmb() asm volatile("mov r0, #0xffff0fff; mov lr, pc;" \ + "sub pc, r0, #95" ::: "r0", "lr", "cc", \ + "memory") +#define cpu_relax() asm volatile("":::"memory") +#endif + #include <time.h> #include <unistd.h> #include <sys/types.h> diff --git a/tools/perf/scripts/perl/Perf-Trace-Util/Context.c b/tools/perf/scripts/perl/Perf-Trace-Util/Context.c index af78d9a52a7d..01a64ad693f2 100644 --- a/tools/perf/scripts/perl/Perf-Trace-Util/Context.c +++ b/tools/perf/scripts/perl/Perf-Trace-Util/Context.c @@ -31,13 +31,14 @@ #include "EXTERN.h" #include "perl.h" #include "XSUB.h" -#include "../../../util/trace-event-perl.h" +#include "../../../perf.h" +#include "../../../util/trace-event.h" #ifndef PERL_UNUSED_VAR # define PERL_UNUSED_VAR(var) if (0) var = var #endif -#line 41 "Context.c" +#line 42 "Context.c" XS(XS_Perf__Trace__Context_common_pc); /* prototype to pass -Wmissing-prototypes */ XS(XS_Perf__Trace__Context_common_pc) diff --git a/tools/perf/scripts/perl/Perf-Trace-Util/Context.xs b/tools/perf/scripts/perl/Perf-Trace-Util/Context.xs index fb78006c165e..549cf0467d30 100644 --- a/tools/perf/scripts/perl/Perf-Trace-Util/Context.xs +++ b/tools/perf/scripts/perl/Perf-Trace-Util/Context.xs @@ -22,7 +22,8 @@ #include "EXTERN.h" #include "perl.h" #include "XSUB.h" -#include "../../../util/trace-event-perl.h" +#include "../../../perf.h" +#include "../../../util/trace-event.h" MODULE = Perf::Trace::Context PACKAGE = Perf::Trace::Context PROTOTYPES: ENABLE diff --git a/tools/perf/scripts/perl/Perf-Trace-Util/lib/Perf/Trace/Util.pm b/tools/perf/scripts/perl/Perf-Trace-Util/lib/Perf/Trace/Util.pm index 052f132ced24..f869c48dc9b0 100644 --- a/tools/perf/scripts/perl/Perf-Trace-Util/lib/Perf/Trace/Util.pm +++ b/tools/perf/scripts/perl/Perf-Trace-Util/lib/Perf/Trace/Util.pm @@ -44,7 +44,7 @@ sub nsecs_secs { sub nsecs_nsecs { my ($nsecs) = @_; - return $nsecs - nsecs_secs($nsecs); + return $nsecs % $NSECS_PER_SEC; } sub nsecs_str { diff --git a/tools/perf/scripts/perl/bin/check-perf-trace-record b/tools/perf/scripts/perl/bin/check-perf-trace-record index c7ec5de2f535..e6cb1474f8e8 100644 --- a/tools/perf/scripts/perl/bin/check-perf-trace-record +++ b/tools/perf/scripts/perl/bin/check-perf-trace-record @@ -1,7 +1,2 @@ #!/bin/bash -perf record -c 1 -f -a -M -R -e kmem:kmalloc -e irq:softirq_entry - - - - - +perf record -c 1 -f -a -M -R -e kmem:kmalloc -e irq:softirq_entry -e kmem:kfree diff --git a/tools/perf/scripts/perl/bin/check-perf-trace-report b/tools/perf/scripts/perl/bin/check-perf-trace-report deleted file mode 100644 index 89948b015020..000000000000 --- a/tools/perf/scripts/perl/bin/check-perf-trace-report +++ /dev/null @@ -1,5 +0,0 @@ -#!/bin/bash -perf trace -s ~/libexec/perf-core/scripts/perl/check-perf-trace.pl - - - diff --git a/tools/perf/scripts/perl/bin/failed-syscalls-record b/tools/perf/scripts/perl/bin/failed-syscalls-record new file mode 100644 index 000000000000..f8885d389e6f --- /dev/null +++ b/tools/perf/scripts/perl/bin/failed-syscalls-record @@ -0,0 +1,2 @@ +#!/bin/bash +perf record -c 1 -f -a -M -R -e raw_syscalls:sys_exit diff --git a/tools/perf/scripts/perl/bin/failed-syscalls-report b/tools/perf/scripts/perl/bin/failed-syscalls-report new file mode 100644 index 000000000000..8bfc660e5056 --- /dev/null +++ b/tools/perf/scripts/perl/bin/failed-syscalls-report @@ -0,0 +1,4 @@ +#!/bin/bash +# description: system-wide failed syscalls +# args: [comm] +perf trace -s ~/libexec/perf-core/scripts/perl/failed-syscalls.pl $1 diff --git a/tools/perf/scripts/perl/bin/rw-by-file-report b/tools/perf/scripts/perl/bin/rw-by-file-report index f5dcf9cb5bd2..eddb9ccce6a5 100644 --- a/tools/perf/scripts/perl/bin/rw-by-file-report +++ b/tools/perf/scripts/perl/bin/rw-by-file-report @@ -1,5 +1,7 @@ #!/bin/bash -perf trace -s ~/libexec/perf-core/scripts/perl/rw-by-file.pl +# description: r/w activity for a program, by file +# args: <comm> +perf trace -s ~/libexec/perf-core/scripts/perl/rw-by-file.pl $1 diff --git a/tools/perf/scripts/perl/bin/rw-by-pid-report b/tools/perf/scripts/perl/bin/rw-by-pid-report index cea16f78a3a2..7f44c25cc857 100644 --- a/tools/perf/scripts/perl/bin/rw-by-pid-report +++ b/tools/perf/scripts/perl/bin/rw-by-pid-report @@ -1,4 +1,5 @@ #!/bin/bash +# description: system-wide r/w activity perf trace -s ~/libexec/perf-core/scripts/perl/rw-by-pid.pl diff --git a/tools/perf/scripts/perl/bin/wakeup-latency-report b/tools/perf/scripts/perl/bin/wakeup-latency-report index 85769dc456eb..fce3adcb3249 100644 --- a/tools/perf/scripts/perl/bin/wakeup-latency-report +++ b/tools/perf/scripts/perl/bin/wakeup-latency-report @@ -1,4 +1,5 @@ #!/bin/bash +# description: system-wide min/max/avg wakeup latency perf trace -s ~/libexec/perf-core/scripts/perl/wakeup-latency.pl diff --git a/tools/perf/scripts/perl/bin/workqueue-stats-report b/tools/perf/scripts/perl/bin/workqueue-stats-report index aa68435be926..71cfbd182fb9 100644 --- a/tools/perf/scripts/perl/bin/workqueue-stats-report +++ b/tools/perf/scripts/perl/bin/workqueue-stats-report @@ -1,4 +1,5 @@ #!/bin/bash +# description: workqueue stats (ins/exe/create/destroy) perf trace -s ~/libexec/perf-core/scripts/perl/workqueue-stats.pl diff --git a/tools/perf/scripts/perl/failed-syscalls.pl b/tools/perf/scripts/perl/failed-syscalls.pl new file mode 100644 index 000000000000..c18e7e27a84b --- /dev/null +++ b/tools/perf/scripts/perl/failed-syscalls.pl @@ -0,0 +1,38 @@ +# failed system call counts +# (c) 2010, Tom Zanussi <tzanussi@gmail.com> +# Licensed under the terms of the GNU GPL License version 2 +# +# Displays system-wide failed system call totals +# If a [comm] arg is specified, only syscalls called by [comm] are displayed. + +use lib "$ENV{'PERF_EXEC_PATH'}/scripts/perl/Perf-Trace-Util/lib"; +use lib "./Perf-Trace-Util/lib"; +use Perf::Trace::Core; +use Perf::Trace::Context; +use Perf::Trace::Util; + +my %failed_syscalls; + +sub raw_syscalls::sys_exit +{ + my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs, + $common_pid, $common_comm, + $id, $ret) = @_; + + if ($ret < 0) { + $failed_syscalls{$common_comm}++; + } +} + +sub trace_end +{ + printf("\nfailed syscalls by comm:\n\n"); + + printf("%-20s %10s\n", "comm", "# errors"); + printf("%-20s %6s %10s\n", "--------------------", "----------"); + + foreach my $comm (sort {$failed_syscalls{$b} <=> $failed_syscalls{$a}} + keys %failed_syscalls) { + printf("%-20s %10s\n", $comm, $failed_syscalls{$comm}); + } +} diff --git a/tools/perf/scripts/perl/rw-by-file.pl b/tools/perf/scripts/perl/rw-by-file.pl index 61f91561d848..2a39097687b9 100644 --- a/tools/perf/scripts/perl/rw-by-file.pl +++ b/tools/perf/scripts/perl/rw-by-file.pl @@ -18,8 +18,9 @@ use lib "./Perf-Trace-Util/lib"; use Perf::Trace::Core; use Perf::Trace::Util; -# change this to the comm of the program you're interested in -my $for_comm = "perf"; +my $usage = "perf trace -s rw-by-file.pl <comm>\n"; + +my $for_comm = shift or die $usage; my %reads; my %writes; diff --git a/tools/perf/scripts/python/Perf-Trace-Util/Context.c b/tools/perf/scripts/python/Perf-Trace-Util/Context.c new file mode 100644 index 000000000000..957085dd5d8d --- /dev/null +++ b/tools/perf/scripts/python/Perf-Trace-Util/Context.c @@ -0,0 +1,88 @@ +/* + * Context.c. Python interfaces for perf trace. + * + * Copyright (C) 2010 Tom Zanussi <tzanussi@gmail.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + */ + +#include <Python.h> +#include "../../../perf.h" +#include "../../../util/trace-event.h" + +PyMODINIT_FUNC initperf_trace_context(void); + +static PyObject *perf_trace_context_common_pc(PyObject *self, PyObject *args) +{ + static struct scripting_context *scripting_context; + PyObject *context; + int retval; + + if (!PyArg_ParseTuple(args, "O", &context)) + return NULL; + + scripting_context = PyCObject_AsVoidPtr(context); + retval = common_pc(scripting_context); + + return Py_BuildValue("i", retval); +} + +static PyObject *perf_trace_context_common_flags(PyObject *self, + PyObject *args) +{ + static struct scripting_context *scripting_context; + PyObject *context; + int retval; + + if (!PyArg_ParseTuple(args, "O", &context)) + return NULL; + + scripting_context = PyCObject_AsVoidPtr(context); + retval = common_flags(scripting_context); + + return Py_BuildValue("i", retval); +} + +static PyObject *perf_trace_context_common_lock_depth(PyObject *self, + PyObject *args) +{ + static struct scripting_context *scripting_context; + PyObject *context; + int retval; + + if (!PyArg_ParseTuple(args, "O", &context)) + return NULL; + + scripting_context = PyCObject_AsVoidPtr(context); + retval = common_lock_depth(scripting_context); + + return Py_BuildValue("i", retval); +} + +static PyMethodDef ContextMethods[] = { + { "common_pc", perf_trace_context_common_pc, METH_VARARGS, + "Get the common preempt count event field value."}, + { "common_flags", perf_trace_context_common_flags, METH_VARARGS, + "Get the common flags event field value."}, + { "common_lock_depth", perf_trace_context_common_lock_depth, + METH_VARARGS, "Get the common lock depth event field value."}, + { NULL, NULL, 0, NULL} +}; + +PyMODINIT_FUNC initperf_trace_context(void) +{ + (void) Py_InitModule("perf_trace_context", ContextMethods); +} diff --git a/tools/perf/scripts/python/Perf-Trace-Util/lib/Perf/Trace/Core.py b/tools/perf/scripts/python/Perf-Trace-Util/lib/Perf/Trace/Core.py new file mode 100644 index 000000000000..1dc464ee2ca8 --- /dev/null +++ b/tools/perf/scripts/python/Perf-Trace-Util/lib/Perf/Trace/Core.py @@ -0,0 +1,91 @@ +# Core.py - Python extension for perf trace, core functions +# +# Copyright (C) 2010 by Tom Zanussi <tzanussi@gmail.com> +# +# This software may be distributed under the terms of the GNU General +# Public License ("GPL") version 2 as published by the Free Software +# Foundation. + +from collections import defaultdict + +def autodict(): + return defaultdict(autodict) + +flag_fields = autodict() +symbolic_fields = autodict() + +def define_flag_field(event_name, field_name, delim): + flag_fields[event_name][field_name]['delim'] = delim + +def define_flag_value(event_name, field_name, value, field_str): + flag_fields[event_name][field_name]['values'][value] = field_str + +def define_symbolic_field(event_name, field_name): + # nothing to do, really + pass + +def define_symbolic_value(event_name, field_name, value, field_str): + symbolic_fields[event_name][field_name]['values'][value] = field_str + +def flag_str(event_name, field_name, value): + string = "" + + if flag_fields[event_name][field_name]: + print_delim = 0 + keys = flag_fields[event_name][field_name]['values'].keys() + keys.sort() + for idx in keys: + if not value and not idx: + string += flag_fields[event_name][field_name]['values'][idx] + break + if idx and (value & idx) == idx: + if print_delim and flag_fields[event_name][field_name]['delim']: + string += " " + flag_fields[event_name][field_name]['delim'] + " " + string += flag_fields[event_name][field_name]['values'][idx] + print_delim = 1 + value &= ~idx + + return string + +def symbol_str(event_name, field_name, value): + string = "" + + if symbolic_fields[event_name][field_name]: + keys = symbolic_fields[event_name][field_name]['values'].keys() + keys.sort() + for idx in keys: + if not value and not idx: + string = symbolic_fields[event_name][field_name]['values'][idx] + break + if (value == idx): + string = symbolic_fields[event_name][field_name]['values'][idx] + break + + return string + +trace_flags = { 0x00: "NONE", \ + 0x01: "IRQS_OFF", \ + 0x02: "IRQS_NOSUPPORT", \ + 0x04: "NEED_RESCHED", \ + 0x08: "HARDIRQ", \ + 0x10: "SOFTIRQ" } + +def trace_flag_str(value): + string = "" + print_delim = 0 + + keys = trace_flags.keys() + + for idx in keys: + if not value and not idx: + string += "NONE" + break + + if idx and (value & idx) == idx: + if print_delim: + string += " | "; + string += trace_flags[idx] + print_delim = 1 + value &= ~idx + + return string diff --git a/tools/perf/scripts/python/Perf-Trace-Util/lib/Perf/Trace/Util.py b/tools/perf/scripts/python/Perf-Trace-Util/lib/Perf/Trace/Util.py new file mode 100644 index 000000000000..83e91435ed09 --- /dev/null +++ b/tools/perf/scripts/python/Perf-Trace-Util/lib/Perf/Trace/Util.py @@ -0,0 +1,25 @@ +# Util.py - Python extension for perf trace, miscellaneous utility code +# +# Copyright (C) 2010 by Tom Zanussi <tzanussi@gmail.com> +# +# This software may be distributed under the terms of the GNU General +# Public License ("GPL") version 2 as published by the Free Software +# Foundation. + +NSECS_PER_SEC = 1000000000 + +def avg(total, n): + return total / n + +def nsecs(secs, nsecs): + return secs * NSECS_PER_SEC + nsecs + +def nsecs_secs(nsecs): + return nsecs / NSECS_PER_SEC + +def nsecs_nsecs(nsecs): + return nsecs % NSECS_PER_SEC + +def nsecs_str(nsecs): + str = "%5u.%09u" % (nsecs_secs(nsecs), nsecs_nsecs(nsecs)), + return str diff --git a/tools/perf/scripts/python/bin/failed-syscalls-by-pid-record b/tools/perf/scripts/python/bin/failed-syscalls-by-pid-record new file mode 100644 index 000000000000..f8885d389e6f --- /dev/null +++ b/tools/perf/scripts/python/bin/failed-syscalls-by-pid-record @@ -0,0 +1,2 @@ +#!/bin/bash +perf record -c 1 -f -a -M -R -e raw_syscalls:sys_exit diff --git a/tools/perf/scripts/python/bin/failed-syscalls-by-pid-report b/tools/perf/scripts/python/bin/failed-syscalls-by-pid-report new file mode 100644 index 000000000000..1e0c0a860c87 --- /dev/null +++ b/tools/perf/scripts/python/bin/failed-syscalls-by-pid-report @@ -0,0 +1,4 @@ +#!/bin/bash +# description: system-wide failed syscalls, by pid +# args: [comm] +perf trace -s ~/libexec/perf-core/scripts/python/failed-syscalls-by-pid.py $1 diff --git a/tools/perf/scripts/python/bin/syscall-counts-by-pid-record b/tools/perf/scripts/python/bin/syscall-counts-by-pid-record new file mode 100644 index 000000000000..45a8c50359da --- /dev/null +++ b/tools/perf/scripts/python/bin/syscall-counts-by-pid-record @@ -0,0 +1,2 @@ +#!/bin/bash +perf record -c 1 -f -a -M -R -e raw_syscalls:sys_enter diff --git a/tools/perf/scripts/python/bin/syscall-counts-by-pid-report b/tools/perf/scripts/python/bin/syscall-counts-by-pid-report new file mode 100644 index 000000000000..f8044d192271 --- /dev/null +++ b/tools/perf/scripts/python/bin/syscall-counts-by-pid-report @@ -0,0 +1,4 @@ +#!/bin/bash +# description: system-wide syscall counts, by pid +# args: [comm] +perf trace -s ~/libexec/perf-core/scripts/python/syscall-counts-by-pid.py $1 diff --git a/tools/perf/scripts/python/bin/syscall-counts-record b/tools/perf/scripts/python/bin/syscall-counts-record new file mode 100644 index 000000000000..45a8c50359da --- /dev/null +++ b/tools/perf/scripts/python/bin/syscall-counts-record @@ -0,0 +1,2 @@ +#!/bin/bash +perf record -c 1 -f -a -M -R -e raw_syscalls:sys_enter diff --git a/tools/perf/scripts/python/bin/syscall-counts-report b/tools/perf/scripts/python/bin/syscall-counts-report new file mode 100644 index 000000000000..a366aa61612f --- /dev/null +++ b/tools/perf/scripts/python/bin/syscall-counts-report @@ -0,0 +1,4 @@ +#!/bin/bash +# description: system-wide syscall counts +# args: [comm] +perf trace -s ~/libexec/perf-core/scripts/python/syscall-counts.py $1 diff --git a/tools/perf/scripts/python/check-perf-trace.py b/tools/perf/scripts/python/check-perf-trace.py new file mode 100644 index 000000000000..964d934395ff --- /dev/null +++ b/tools/perf/scripts/python/check-perf-trace.py @@ -0,0 +1,83 @@ +# perf trace event handlers, generated by perf trace -g python +# (c) 2010, Tom Zanussi <tzanussi@gmail.com> +# Licensed under the terms of the GNU GPL License version 2 +# +# This script tests basic functionality such as flag and symbol +# strings, common_xxx() calls back into perf, begin, end, unhandled +# events, etc. Basically, if this script runs successfully and +# displays expected results, Python scripting support should be ok. + +import os +import sys + +sys.path.append(os.environ['PERF_EXEC_PATH'] + \ + '/scripts/python/Perf-Trace-Util/lib/Perf/Trace') + +from Core import * +from perf_trace_context import * + +unhandled = autodict() + +def trace_begin(): + print "trace_begin" + pass + +def trace_end(): + print_unhandled() + +def irq__softirq_entry(event_name, context, common_cpu, + common_secs, common_nsecs, common_pid, common_comm, + vec): + print_header(event_name, common_cpu, common_secs, common_nsecs, + common_pid, common_comm) + + print_uncommon(context) + + print "vec=%s\n" % \ + (symbol_str("irq__softirq_entry", "vec", vec)), + +def kmem__kmalloc(event_name, context, common_cpu, + common_secs, common_nsecs, common_pid, common_comm, + call_site, ptr, bytes_req, bytes_alloc, + gfp_flags): + print_header(event_name, common_cpu, common_secs, common_nsecs, + common_pid, common_comm) + + print_uncommon(context) + + print "call_site=%u, ptr=%u, bytes_req=%u, " \ + "bytes_alloc=%u, gfp_flags=%s\n" % \ + (call_site, ptr, bytes_req, bytes_alloc, + + flag_str("kmem__kmalloc", "gfp_flags", gfp_flags)), + +def trace_unhandled(event_name, context, common_cpu, common_secs, common_nsecs, + common_pid, common_comm): + try: + unhandled[event_name] += 1 + except TypeError: + unhandled[event_name] = 1 + +def print_header(event_name, cpu, secs, nsecs, pid, comm): + print "%-20s %5u %05u.%09u %8u %-20s " % \ + (event_name, cpu, secs, nsecs, pid, comm), + +# print trace fields not included in handler args +def print_uncommon(context): + print "common_preempt_count=%d, common_flags=%s, common_lock_depth=%d, " \ + % (common_pc(context), trace_flag_str(common_flags(context)), \ + common_lock_depth(context)) + +def print_unhandled(): + keys = unhandled.keys() + if not keys: + return + + print "\nunhandled events:\n\n", + + print "%-40s %10s\n" % ("event", "count"), + print "%-40s %10s\n" % ("----------------------------------------", \ + "-----------"), + + for event_name in keys: + print "%-40s %10d\n" % (event_name, unhandled[event_name]) diff --git a/tools/perf/scripts/python/failed-syscalls-by-pid.py b/tools/perf/scripts/python/failed-syscalls-by-pid.py new file mode 100644 index 000000000000..0ca02278fe69 --- /dev/null +++ b/tools/perf/scripts/python/failed-syscalls-by-pid.py @@ -0,0 +1,68 @@ +# failed system call counts, by pid +# (c) 2010, Tom Zanussi <tzanussi@gmail.com> +# Licensed under the terms of the GNU GPL License version 2 +# +# Displays system-wide failed system call totals, broken down by pid. +# If a [comm] arg is specified, only syscalls called by [comm] are displayed. + +import os +import sys + +sys.path.append(os.environ['PERF_EXEC_PATH'] + \ + '/scripts/python/Perf-Trace-Util/lib/Perf/Trace') + +from perf_trace_context import * +from Core import * + +usage = "perf trace -s syscall-counts-by-pid.py [comm]\n"; + +for_comm = None + +if len(sys.argv) > 2: + sys.exit(usage) + +if len(sys.argv) > 1: + for_comm = sys.argv[1] + +syscalls = autodict() + +def trace_begin(): + pass + +def trace_end(): + print_error_totals() + +def raw_syscalls__sys_exit(event_name, context, common_cpu, + common_secs, common_nsecs, common_pid, common_comm, + id, ret): + if for_comm is not None: + if common_comm != for_comm: + return + + if ret < 0: + try: + syscalls[common_comm][common_pid][id][ret] += 1 + except TypeError: + syscalls[common_comm][common_pid][id][ret] = 1 + +def print_error_totals(): + if for_comm is not None: + print "\nsyscall errors for %s:\n\n" % (for_comm), + else: + print "\nsyscall errors:\n\n", + + print "%-30s %10s\n" % ("comm [pid]", "count"), + print "%-30s %10s\n" % ("------------------------------", \ + "----------"), + + comm_keys = syscalls.keys() + for comm in comm_keys: + pid_keys = syscalls[comm].keys() + for pid in pid_keys: + print "\n%s [%d]\n" % (comm, pid), + id_keys = syscalls[comm][pid].keys() + for id in id_keys: + print " syscall: %-16d\n" % (id), + ret_keys = syscalls[comm][pid][id].keys() + for ret, val in sorted(syscalls[comm][pid][id].iteritems(), key = lambda(k, v): (v, k), reverse = True): + print " err = %-20d %10d\n" % (ret, val), diff --git a/tools/perf/scripts/python/syscall-counts-by-pid.py b/tools/perf/scripts/python/syscall-counts-by-pid.py new file mode 100644 index 000000000000..af722d6a4b3f --- /dev/null +++ b/tools/perf/scripts/python/syscall-counts-by-pid.py @@ -0,0 +1,64 @@ +# system call counts, by pid +# (c) 2010, Tom Zanussi <tzanussi@gmail.com> +# Licensed under the terms of the GNU GPL License version 2 +# +# Displays system-wide system call totals, broken down by syscall. +# If a [comm] arg is specified, only syscalls called by [comm] are displayed. + +import os +import sys + +sys.path.append(os.environ['PERF_EXEC_PATH'] + \ + '/scripts/python/Perf-Trace-Util/lib/Perf/Trace') + +from perf_trace_context import * +from Core import * + +usage = "perf trace -s syscall-counts-by-pid.py [comm]\n"; + +for_comm = None + +if len(sys.argv) > 2: + sys.exit(usage) + +if len(sys.argv) > 1: + for_comm = sys.argv[1] + +syscalls = autodict() + +def trace_begin(): + pass + +def trace_end(): + print_syscall_totals() + +def raw_syscalls__sys_enter(event_name, context, common_cpu, + common_secs, common_nsecs, common_pid, common_comm, + id, args): + if for_comm is not None: + if common_comm != for_comm: + return + try: + syscalls[common_comm][common_pid][id] += 1 + except TypeError: + syscalls[common_comm][common_pid][id] = 1 + +def print_syscall_totals(): + if for_comm is not None: + print "\nsyscall events for %s:\n\n" % (for_comm), + else: + print "\nsyscall events by comm/pid:\n\n", + + print "%-40s %10s\n" % ("comm [pid]/syscalls", "count"), + print "%-40s %10s\n" % ("----------------------------------------", \ + "----------"), + + comm_keys = syscalls.keys() + for comm in comm_keys: + pid_keys = syscalls[comm].keys() + for pid in pid_keys: + print "\n%s [%d]\n" % (comm, pid), + id_keys = syscalls[comm][pid].keys() + for id, val in sorted(syscalls[comm][pid].iteritems(), \ + key = lambda(k, v): (v, k), reverse = True): + print " %-38d %10d\n" % (id, val), diff --git a/tools/perf/scripts/python/syscall-counts.py b/tools/perf/scripts/python/syscall-counts.py new file mode 100644 index 000000000000..f977e85ff049 --- /dev/null +++ b/tools/perf/scripts/python/syscall-counts.py @@ -0,0 +1,58 @@ +# system call counts +# (c) 2010, Tom Zanussi <tzanussi@gmail.com> +# Licensed under the terms of the GNU GPL License version 2 +# +# Displays system-wide system call totals, broken down by syscall. +# If a [comm] arg is specified, only syscalls called by [comm] are displayed. + +import os +import sys + +sys.path.append(os.environ['PERF_EXEC_PATH'] + \ + '/scripts/python/Perf-Trace-Util/lib/Perf/Trace') + +from perf_trace_context import * +from Core import * + +usage = "perf trace -s syscall-counts.py [comm]\n"; + +for_comm = None + +if len(sys.argv) > 2: + sys.exit(usage) + +if len(sys.argv) > 1: + for_comm = sys.argv[1] + +syscalls = autodict() + +def trace_begin(): + pass + +def trace_end(): + print_syscall_totals() + +def raw_syscalls__sys_enter(event_name, context, common_cpu, + common_secs, common_nsecs, common_pid, common_comm, + id, args): + if for_comm is not None: + if common_comm != for_comm: + return + try: + syscalls[id] += 1 + except TypeError: + syscalls[id] = 1 + +def print_syscall_totals(): + if for_comm is not None: + print "\nsyscall events for %s:\n\n" % (for_comm), + else: + print "\nsyscall events:\n\n", + + print "%-40s %10s\n" % ("event", "count"), + print "%-40s %10s\n" % ("----------------------------------------", \ + "-----------"), + + for id, val in sorted(syscalls.iteritems(), key = lambda(k, v): (v, k), \ + reverse = True): + print "%-40d %10d\n" % (id, val), diff --git a/tools/perf/util/build-id.c b/tools/perf/util/build-id.c new file mode 100644 index 000000000000..04904b35ba81 --- /dev/null +++ b/tools/perf/util/build-id.c @@ -0,0 +1,39 @@ +/* + * build-id.c + * + * build-id support + * + * Copyright (C) 2009, 2010 Red Hat Inc. + * Copyright (C) 2009, 2010 Arnaldo Carvalho de Melo <acme@redhat.com> + */ +#include "build-id.h" +#include "event.h" +#include "symbol.h" +#include <linux/kernel.h> + +static int build_id__mark_dso_hit(event_t *event, struct perf_session *session) +{ + struct addr_location al; + u8 cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK; + struct thread *thread = perf_session__findnew(session, event->ip.pid); + + if (thread == NULL) { + pr_err("problem processing %d event, skipping it.\n", + event->header.type); + return -1; + } + + thread__find_addr_map(thread, session, cpumode, MAP__FUNCTION, + event->ip.ip, &al); + + if (al.map != NULL) + al.map->dso->hit = 1; + + return 0; +} + +struct perf_event_ops build_id__mark_dso_hit_ops = { + .sample = build_id__mark_dso_hit, + .mmap = event__process_mmap, + .fork = event__process_task, +}; diff --git a/tools/perf/util/build-id.h b/tools/perf/util/build-id.h new file mode 100644 index 000000000000..1d981d63cf9a --- /dev/null +++ b/tools/perf/util/build-id.h @@ -0,0 +1,8 @@ +#ifndef PERF_BUILD_ID_H_ +#define PERF_BUILD_ID_H_ 1 + +#include "session.h" + +extern struct perf_event_ops build_id__mark_dso_hit_ops; + +#endif diff --git a/tools/perf/util/data_map.c b/tools/perf/util/data_map.c deleted file mode 100644 index ca0bedf637c2..000000000000 --- a/tools/perf/util/data_map.c +++ /dev/null @@ -1,291 +0,0 @@ -#include "data_map.h" -#include "symbol.h" -#include "util.h" -#include "debug.h" - - -static struct perf_file_handler *curr_handler; -static unsigned long mmap_window = 32; -static char __cwd[PATH_MAX]; - -static int process_event_stub(event_t *event __used) -{ - dump_printf(": unhandled!\n"); - return 0; -} - -void register_perf_file_handler(struct perf_file_handler *handler) -{ - if (!handler->process_sample_event) - handler->process_sample_event = process_event_stub; - if (!handler->process_mmap_event) - handler->process_mmap_event = process_event_stub; - if (!handler->process_comm_event) - handler->process_comm_event = process_event_stub; - if (!handler->process_fork_event) - handler->process_fork_event = process_event_stub; - if (!handler->process_exit_event) - handler->process_exit_event = process_event_stub; - if (!handler->process_lost_event) - handler->process_lost_event = process_event_stub; - if (!handler->process_read_event) - handler->process_read_event = process_event_stub; - if (!handler->process_throttle_event) - handler->process_throttle_event = process_event_stub; - if (!handler->process_unthrottle_event) - handler->process_unthrottle_event = process_event_stub; - - curr_handler = handler; -} - -static const char *event__name[] = { - [0] = "TOTAL", - [PERF_RECORD_MMAP] = "MMAP", - [PERF_RECORD_LOST] = "LOST", - [PERF_RECORD_COMM] = "COMM", - [PERF_RECORD_EXIT] = "EXIT", - [PERF_RECORD_THROTTLE] = "THROTTLE", - [PERF_RECORD_UNTHROTTLE] = "UNTHROTTLE", - [PERF_RECORD_FORK] = "FORK", - [PERF_RECORD_READ] = "READ", - [PERF_RECORD_SAMPLE] = "SAMPLE", -}; - -unsigned long event__total[PERF_RECORD_MAX]; - -void event__print_totals(void) -{ - int i; - for (i = 0; i < PERF_RECORD_MAX; ++i) - pr_info("%10s events: %10ld\n", - event__name[i], event__total[i]); -} - -static int -process_event(event_t *event, unsigned long offset, unsigned long head) -{ - trace_event(event); - - if (event->header.type < PERF_RECORD_MAX) { - dump_printf("%p [%p]: PERF_RECORD_%s", - (void *)(offset + head), - (void *)(long)(event->header.size), - event__name[event->header.type]); - ++event__total[0]; - ++event__total[event->header.type]; - } - - switch (event->header.type) { - case PERF_RECORD_SAMPLE: - return curr_handler->process_sample_event(event); - case PERF_RECORD_MMAP: - return curr_handler->process_mmap_event(event); - case PERF_RECORD_COMM: - return curr_handler->process_comm_event(event); - case PERF_RECORD_FORK: - return curr_handler->process_fork_event(event); - case PERF_RECORD_EXIT: - return curr_handler->process_exit_event(event); - case PERF_RECORD_LOST: - return curr_handler->process_lost_event(event); - case PERF_RECORD_READ: - return curr_handler->process_read_event(event); - case PERF_RECORD_THROTTLE: - return curr_handler->process_throttle_event(event); - case PERF_RECORD_UNTHROTTLE: - return curr_handler->process_unthrottle_event(event); - default: - curr_handler->total_unknown++; - return -1; - } -} - -int perf_header__read_build_ids(int input, off_t offset, off_t size) -{ - struct build_id_event bev; - char filename[PATH_MAX]; - off_t limit = offset + size; - int err = -1; - - while (offset < limit) { - struct dso *dso; - ssize_t len; - - if (read(input, &bev, sizeof(bev)) != sizeof(bev)) - goto out; - - len = bev.header.size - sizeof(bev); - if (read(input, filename, len) != len) - goto out; - - dso = dsos__findnew(filename); - if (dso != NULL) - dso__set_build_id(dso, &bev.build_id); - - offset += bev.header.size; - } - err = 0; -out: - return err; -} - -int mmap_dispatch_perf_file(struct perf_header **pheader, - const char *input_name, - int force, - int full_paths, - int *cwdlen, - char **cwd) -{ - int err; - struct perf_header *header; - unsigned long head, shift; - unsigned long offset = 0; - struct stat input_stat; - size_t page_size; - u64 sample_type; - event_t *event; - uint32_t size; - int input; - char *buf; - - if (curr_handler == NULL) { - pr_debug("Forgot to register perf file handler\n"); - return -EINVAL; - } - - page_size = getpagesize(); - - input = open(input_name, O_RDONLY); - if (input < 0) { - pr_err("Failed to open file: %s", input_name); - if (!strcmp(input_name, "perf.data")) - pr_err(" (try 'perf record' first)"); - pr_err("\n"); - return -errno; - } - - if (fstat(input, &input_stat) < 0) { - pr_err("failed to stat file"); - err = -errno; - goto out_close; - } - - err = -EACCES; - if (!force && input_stat.st_uid && (input_stat.st_uid != geteuid())) { - pr_err("file: %s not owned by current user or root\n", - input_name); - goto out_close; - } - - if (input_stat.st_size == 0) { - pr_info("zero-sized file, nothing to do!\n"); - goto done; - } - - err = -ENOMEM; - header = perf_header__new(); - if (header == NULL) - goto out_close; - - err = perf_header__read(header, input); - if (err < 0) - goto out_delete; - *pheader = header; - head = header->data_offset; - - sample_type = perf_header__sample_type(header); - - err = -EINVAL; - if (curr_handler->sample_type_check && - curr_handler->sample_type_check(sample_type) < 0) - goto out_delete; - - if (!full_paths) { - if (getcwd(__cwd, sizeof(__cwd)) == NULL) { - pr_err("failed to get the current directory\n"); - err = -errno; - goto out_delete; - } - *cwd = __cwd; - *cwdlen = strlen(*cwd); - } else { - *cwd = NULL; - *cwdlen = 0; - } - - shift = page_size * (head / page_size); - offset += shift; - head -= shift; - -remap: - buf = mmap(NULL, page_size * mmap_window, PROT_READ, - MAP_SHARED, input, offset); - if (buf == MAP_FAILED) { - pr_err("failed to mmap file\n"); - err = -errno; - goto out_delete; - } - -more: - event = (event_t *)(buf + head); - - size = event->header.size; - if (!size) - size = 8; - - if (head + event->header.size >= page_size * mmap_window) { - int munmap_ret; - - shift = page_size * (head / page_size); - - munmap_ret = munmap(buf, page_size * mmap_window); - assert(munmap_ret == 0); - - offset += shift; - head -= shift; - goto remap; - } - - size = event->header.size; - - dump_printf("\n%p [%p]: event: %d\n", - (void *)(offset + head), - (void *)(long)event->header.size, - event->header.type); - - if (!size || process_event(event, offset, head) < 0) { - - dump_printf("%p [%p]: skipping unknown header type: %d\n", - (void *)(offset + head), - (void *)(long)(event->header.size), - event->header.type); - - /* - * assume we lost track of the stream, check alignment, and - * increment a single u64 in the hope to catch on again 'soon'. - */ - - if (unlikely(head & 7)) - head &= ~7ULL; - - size = 8; - } - - head += size; - - if (offset + head >= header->data_offset + header->data_size) - goto done; - - if (offset + head < (unsigned long)input_stat.st_size) - goto more; - -done: - err = 0; -out_close: - close(input); - - return err; -out_delete: - perf_header__delete(header); - goto out_close; -} diff --git a/tools/perf/util/data_map.h b/tools/perf/util/data_map.h deleted file mode 100644 index 3180ff7e3633..000000000000 --- a/tools/perf/util/data_map.h +++ /dev/null @@ -1,32 +0,0 @@ -#ifndef __PERF_DATAMAP_H -#define __PERF_DATAMAP_H - -#include "event.h" -#include "header.h" - -typedef int (*event_type_handler_t)(event_t *); - -struct perf_file_handler { - event_type_handler_t process_sample_event; - event_type_handler_t process_mmap_event; - event_type_handler_t process_comm_event; - event_type_handler_t process_fork_event; - event_type_handler_t process_exit_event; - event_type_handler_t process_lost_event; - event_type_handler_t process_read_event; - event_type_handler_t process_throttle_event; - event_type_handler_t process_unthrottle_event; - int (*sample_type_check)(u64 sample_type); - unsigned long total_unknown; -}; - -void register_perf_file_handler(struct perf_file_handler *handler); -int mmap_dispatch_perf_file(struct perf_header **pheader, - const char *input_name, - int force, - int full_paths, - int *cwdlen, - char **cwd); -int perf_header__read_build_ids(int input, off_t offset, off_t file_size); - -#endif diff --git a/tools/perf/util/debug.c b/tools/perf/util/debug.c index 28d520d5a1fb..0905600c3851 100644 --- a/tools/perf/util/debug.c +++ b/tools/perf/util/debug.c @@ -9,6 +9,7 @@ #include "color.h" #include "event.h" #include "debug.h" +#include "util.h" int verbose = 0; int dump_trace = 0; diff --git a/tools/perf/util/debugfs.c b/tools/perf/util/debugfs.c index 06b73ee02c49..a88fefc0cc0a 100644 --- a/tools/perf/util/debugfs.c +++ b/tools/perf/util/debugfs.c @@ -106,16 +106,14 @@ int debugfs_valid_entry(const char *path) return 0; } -/* mount the debugfs somewhere */ +/* mount the debugfs somewhere if it's not mounted */ -int debugfs_mount(const char *mountpoint) +char *debugfs_mount(const char *mountpoint) { - char mountcmd[128]; - /* see if it's already mounted */ if (debugfs_find_mountpoint()) { debugfs_premounted = 1; - return 0; + return debugfs_mountpoint; } /* if not mounted and no argument */ @@ -127,13 +125,14 @@ int debugfs_mount(const char *mountpoint) mountpoint = "/sys/kernel/debug"; } + if (mount(NULL, mountpoint, "debugfs", 0, NULL) < 0) + return NULL; + /* save the mountpoint */ strncpy(debugfs_mountpoint, mountpoint, sizeof(debugfs_mountpoint)); + debugfs_found = 1; - /* mount it */ - snprintf(mountcmd, sizeof(mountcmd), - "/bin/mount -t debugfs debugfs %s", mountpoint); - return system(mountcmd); + return debugfs_mountpoint; } /* umount the debugfs */ diff --git a/tools/perf/util/debugfs.h b/tools/perf/util/debugfs.h index 3cd14f9ae784..83a02879745f 100644 --- a/tools/perf/util/debugfs.h +++ b/tools/perf/util/debugfs.h @@ -15,7 +15,7 @@ extern const char *debugfs_find_mountpoint(void); extern int debugfs_valid_mountpoint(const char *debugfs); extern int debugfs_valid_entry(const char *path); -extern int debugfs_mount(const char *mountpoint); +extern char *debugfs_mount(const char *mountpoint); extern int debugfs_umount(void); extern int debugfs_write(const char *entry, const char *value); extern int debugfs_read(const char *entry, char *buffer, size_t size); diff --git a/tools/perf/util/event.c b/tools/perf/util/event.c index 414b89d1bde9..705ec63548b4 100644 --- a/tools/perf/util/event.c +++ b/tools/perf/util/event.c @@ -1,11 +1,15 @@ #include <linux/types.h> #include "event.h" #include "debug.h" +#include "session.h" +#include "sort.h" #include "string.h" +#include "strlist.h" #include "thread.h" static pid_t event__synthesize_comm(pid_t pid, int full, - int (*process)(event_t *event)) + event__handler_t process, + struct perf_session *session) { event_t ev; char filename[PATH_MAX]; @@ -54,7 +58,7 @@ out_race: if (!full) { ev.comm.tid = pid; - process(&ev); + process(&ev, session); goto out_fclose; } @@ -72,7 +76,7 @@ out_race: ev.comm.tid = pid; - process(&ev); + process(&ev, session); } closedir(tasks); @@ -86,7 +90,8 @@ out_failure: } static int event__synthesize_mmap_events(pid_t pid, pid_t tgid, - int (*process)(event_t *event)) + event__handler_t process, + struct perf_session *session) { char filename[PATH_MAX]; FILE *fp; @@ -105,7 +110,10 @@ static int event__synthesize_mmap_events(pid_t pid, pid_t tgid, while (1) { char bf[BUFSIZ], *pbf = bf; event_t ev = { - .header = { .type = PERF_RECORD_MMAP }, + .header = { + .type = PERF_RECORD_MMAP, + .misc = 0, /* Just like the kernel, see kernel/perf_event.c __perf_event_mmap */ + }, }; int n; size_t size; @@ -141,7 +149,7 @@ static int event__synthesize_mmap_events(pid_t pid, pid_t tgid, ev.mmap.pid = tgid; ev.mmap.tid = pid; - process(&ev); + process(&ev, session); } } @@ -149,15 +157,48 @@ static int event__synthesize_mmap_events(pid_t pid, pid_t tgid, return 0; } -int event__synthesize_thread(pid_t pid, int (*process)(event_t *event)) +int event__synthesize_modules(event__handler_t process, + struct perf_session *session) { - pid_t tgid = event__synthesize_comm(pid, 1, process); + struct rb_node *nd; + + for (nd = rb_first(&session->kmaps.maps[MAP__FUNCTION]); + nd; nd = rb_next(nd)) { + event_t ev; + size_t size; + struct map *pos = rb_entry(nd, struct map, rb_node); + + if (pos->dso->kernel) + continue; + + size = ALIGN(pos->dso->long_name_len + 1, sizeof(u64)); + memset(&ev, 0, sizeof(ev)); + ev.mmap.header.misc = 1; /* kernel uses 0 for user space maps, see kernel/perf_event.c __perf_event_mmap */ + ev.mmap.header.type = PERF_RECORD_MMAP; + ev.mmap.header.size = (sizeof(ev.mmap) - + (sizeof(ev.mmap.filename) - size)); + ev.mmap.start = pos->start; + ev.mmap.len = pos->end - pos->start; + + memcpy(ev.mmap.filename, pos->dso->long_name, + pos->dso->long_name_len + 1); + process(&ev, session); + } + + return 0; +} + +int event__synthesize_thread(pid_t pid, event__handler_t process, + struct perf_session *session) +{ + pid_t tgid = event__synthesize_comm(pid, 1, process, session); if (tgid == -1) return -1; - return event__synthesize_mmap_events(pid, tgid, process); + return event__synthesize_mmap_events(pid, tgid, process, session); } -void event__synthesize_threads(int (*process)(event_t *event)) +void event__synthesize_threads(event__handler_t process, + struct perf_session *session) { DIR *proc; struct dirent dirent, *next; @@ -171,24 +212,100 @@ void event__synthesize_threads(int (*process)(event_t *event)) if (*end) /* only interested in proper numerical dirents */ continue; - event__synthesize_thread(pid, process); + event__synthesize_thread(pid, process, session); } closedir(proc); } -char *event__cwd; -int event__cwdlen; +struct process_symbol_args { + const char *name; + u64 start; +}; + +static int find_symbol_cb(void *arg, const char *name, char type, u64 start) +{ + struct process_symbol_args *args = arg; + + /* + * Must be a function or at least an alias, as in PARISC64, where "_text" is + * an 'A' to the same address as "_stext". + */ + if (!(symbol_type__is_a(type, MAP__FUNCTION) || + type == 'A') || strcmp(name, args->name)) + return 0; + + args->start = start; + return 1; +} + +int event__synthesize_kernel_mmap(event__handler_t process, + struct perf_session *session, + const char *symbol_name) +{ + size_t size; + event_t ev = { + .header = { + .type = PERF_RECORD_MMAP, + .misc = 1, /* kernel uses 0 for user space maps, see kernel/perf_event.c __perf_event_mmap */ + }, + }; + /* + * We should get this from /sys/kernel/sections/.text, but till that is + * available use this, and after it is use this as a fallback for older + * kernels. + */ + struct process_symbol_args args = { .name = symbol_name, }; + + if (kallsyms__parse("/proc/kallsyms", &args, find_symbol_cb) <= 0) + return -ENOENT; + + size = snprintf(ev.mmap.filename, sizeof(ev.mmap.filename), + "[kernel.kallsyms.%s]", symbol_name) + 1; + size = ALIGN(size, sizeof(u64)); + ev.mmap.header.size = (sizeof(ev.mmap) - (sizeof(ev.mmap.filename) - size)); + ev.mmap.pgoff = args.start; + ev.mmap.start = session->vmlinux_maps[MAP__FUNCTION]->start; + ev.mmap.len = session->vmlinux_maps[MAP__FUNCTION]->end - ev.mmap.start ; + + return process(&ev, session); +} + +static void thread__comm_adjust(struct thread *self) +{ + char *comm = self->comm; -struct events_stats event__stats; + if (!symbol_conf.col_width_list_str && !symbol_conf.field_sep && + (!symbol_conf.comm_list || + strlist__has_entry(symbol_conf.comm_list, comm))) { + unsigned int slen = strlen(comm); -int event__process_comm(event_t *self) + if (slen > comms__col_width) { + comms__col_width = slen; + threads__col_width = slen + 6; + } + } +} + +static int thread__set_comm_adjust(struct thread *self, const char *comm) { - struct thread *thread = threads__findnew(self->comm.pid); + int ret = thread__set_comm(self, comm); + + if (ret) + return ret; + + thread__comm_adjust(self); + + return 0; +} + +int event__process_comm(event_t *self, struct perf_session *session) +{ + struct thread *thread = perf_session__findnew(session, self->comm.pid); dump_printf(": %s:%d\n", self->comm.comm, self->comm.pid); - if (thread == NULL || thread__set_comm(thread, self->comm.comm)) { + if (thread == NULL || thread__set_comm_adjust(thread, self->comm.comm)) { dump_printf("problem processing PERF_RECORD_COMM, skipping event.\n"); return -1; } @@ -196,38 +313,104 @@ int event__process_comm(event_t *self) return 0; } -int event__process_lost(event_t *self) +int event__process_lost(event_t *self, struct perf_session *session) { dump_printf(": id:%Ld: lost:%Ld\n", self->lost.id, self->lost.lost); - event__stats.lost += self->lost.lost; + session->events_stats.lost += self->lost.lost; return 0; } -int event__process_mmap(event_t *self) +int event__process_mmap(event_t *self, struct perf_session *session) { - struct thread *thread = threads__findnew(self->mmap.pid); - struct map *map = map__new(&self->mmap, MAP__FUNCTION, - event__cwd, event__cwdlen); + struct thread *thread; + struct map *map; + + dump_printf(" %d/%d: [%#Lx(%#Lx) @ %#Lx]: %s\n", + self->mmap.pid, self->mmap.tid, self->mmap.start, + self->mmap.len, self->mmap.pgoff, self->mmap.filename); + + if (self->mmap.pid == 0) { + static const char kmmap_prefix[] = "[kernel.kallsyms."; + + if (self->mmap.filename[0] == '/') { + char short_module_name[1024]; + char *name = strrchr(self->mmap.filename, '/'), *dot; + + if (name == NULL) + goto out_problem; + + ++name; /* skip / */ + dot = strrchr(name, '.'); + if (dot == NULL) + goto out_problem; + + snprintf(short_module_name, sizeof(short_module_name), + "[%.*s]", (int)(dot - name), name); + strxfrchar(short_module_name, '-', '_'); + + map = perf_session__new_module_map(session, + self->mmap.start, + self->mmap.filename); + if (map == NULL) + goto out_problem; + + name = strdup(short_module_name); + if (name == NULL) + goto out_problem; + + map->dso->short_name = name; + map->end = map->start + self->mmap.len; + } else if (memcmp(self->mmap.filename, kmmap_prefix, + sizeof(kmmap_prefix) - 1) == 0) { + const char *symbol_name = (self->mmap.filename + + sizeof(kmmap_prefix) - 1); + /* + * Should be there already, from the build-id table in + * the header. + */ + struct dso *kernel = __dsos__findnew(&dsos__kernel, + "[kernel.kallsyms]"); + if (kernel == NULL) + goto out_problem; + + kernel->kernel = 1; + if (__perf_session__create_kernel_maps(session, kernel) < 0) + goto out_problem; + + session->vmlinux_maps[MAP__FUNCTION]->start = self->mmap.start; + session->vmlinux_maps[MAP__FUNCTION]->end = self->mmap.start + self->mmap.len; + /* + * Be a bit paranoid here, some perf.data file came with + * a zero sized synthesized MMAP event for the kernel. + */ + if (session->vmlinux_maps[MAP__FUNCTION]->end == 0) + session->vmlinux_maps[MAP__FUNCTION]->end = ~0UL; + + perf_session__set_kallsyms_ref_reloc_sym(session, symbol_name, + self->mmap.pgoff); + } + return 0; + } - dump_printf(" %d/%d: [%p(%p) @ %p]: %s\n", - self->mmap.pid, self->mmap.tid, - (void *)(long)self->mmap.start, - (void *)(long)self->mmap.len, - (void *)(long)self->mmap.pgoff, - self->mmap.filename); + thread = perf_session__findnew(session, self->mmap.pid); + map = map__new(&self->mmap, MAP__FUNCTION, + session->cwd, session->cwdlen); if (thread == NULL || map == NULL) - dump_printf("problem processing PERF_RECORD_MMAP, skipping event.\n"); - else - thread__insert_map(thread, map); + goto out_problem; + + thread__insert_map(thread, map); + return 0; +out_problem: + dump_printf("problem processing PERF_RECORD_MMAP, skipping event.\n"); return 0; } -int event__process_task(event_t *self) +int event__process_task(event_t *self, struct perf_session *session) { - struct thread *thread = threads__findnew(self->fork.pid); - struct thread *parent = threads__findnew(self->fork.ppid); + struct thread *thread = perf_session__findnew(session, self->fork.pid); + struct thread *parent = perf_session__findnew(session, self->fork.ppid); dump_printf("(%d:%d):(%d:%d)\n", self->fork.pid, self->fork.tid, self->fork.ppid, self->fork.ptid); @@ -249,28 +432,28 @@ int event__process_task(event_t *self) return 0; } -void thread__find_addr_location(struct thread *self, u8 cpumode, - enum map_type type, u64 addr, - struct addr_location *al, - symbol_filter_t filter) +void thread__find_addr_map(struct thread *self, + struct perf_session *session, u8 cpumode, + enum map_type type, u64 addr, + struct addr_location *al) { - struct thread *thread = al->thread = self; + struct map_groups *mg = &self->mg; + al->thread = self; al->addr = addr; - if (cpumode & PERF_RECORD_MISC_KERNEL) { + if (cpumode == PERF_RECORD_MISC_KERNEL) { al->level = 'k'; - thread = kthread; - } else if (cpumode & PERF_RECORD_MISC_USER) + mg = &session->kmaps; + } else if (cpumode == PERF_RECORD_MISC_USER) al->level = '.'; else { al->level = 'H'; al->map = NULL; - al->sym = NULL; return; } try_again: - al->map = thread__find_map(thread, type, al->addr); + al->map = map_groups__find(mg, type, al->addr); if (al->map == NULL) { /* * If this is outside of all known maps, and is a negative @@ -281,32 +464,149 @@ try_again: * "[vdso]" dso, but for now lets use the old trick of looking * in the whole kernel symbol list. */ - if ((long long)al->addr < 0 && thread != kthread) { - thread = kthread; + if ((long long)al->addr < 0 && mg != &session->kmaps) { + mg = &session->kmaps; goto try_again; } - al->sym = NULL; - } else { + } else al->addr = al->map->map_ip(al->map, al->addr); +} + +void thread__find_addr_location(struct thread *self, + struct perf_session *session, u8 cpumode, + enum map_type type, u64 addr, + struct addr_location *al, + symbol_filter_t filter) +{ + thread__find_addr_map(self, session, cpumode, type, addr, al); + if (al->map != NULL) al->sym = map__find_symbol(al->map, al->addr, filter); + else + al->sym = NULL; +} + +static void dso__calc_col_width(struct dso *self) +{ + if (!symbol_conf.col_width_list_str && !symbol_conf.field_sep && + (!symbol_conf.dso_list || + strlist__has_entry(symbol_conf.dso_list, self->name))) { + unsigned int slen = strlen(self->name); + if (slen > dsos__col_width) + dsos__col_width = slen; } + + self->slen_calculated = 1; } -int event__preprocess_sample(const event_t *self, struct addr_location *al, - symbol_filter_t filter) +int event__preprocess_sample(const event_t *self, struct perf_session *session, + struct addr_location *al, symbol_filter_t filter) { u8 cpumode = self->header.misc & PERF_RECORD_MISC_CPUMODE_MASK; - struct thread *thread = threads__findnew(self->ip.pid); + struct thread *thread = perf_session__findnew(session, self->ip.pid); if (thread == NULL) return -1; + if (symbol_conf.comm_list && + !strlist__has_entry(symbol_conf.comm_list, thread->comm)) + goto out_filtered; + dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid); - thread__find_addr_location(thread, cpumode, MAP__FUNCTION, + thread__find_addr_location(thread, session, cpumode, MAP__FUNCTION, self->ip.ip, al, filter); dump_printf(" ...... dso: %s\n", al->map ? al->map->dso->long_name : al->level == 'H' ? "[hypervisor]" : "<not found>"); + /* + * We have to do this here as we may have a dso with no symbol hit that + * has a name longer than the ones with symbols sampled. + */ + if (al->map && !sort_dso.elide && !al->map->dso->slen_calculated) + dso__calc_col_width(al->map->dso); + + if (symbol_conf.dso_list && + (!al->map || !al->map->dso || + !(strlist__has_entry(symbol_conf.dso_list, al->map->dso->short_name) || + (al->map->dso->short_name != al->map->dso->long_name && + strlist__has_entry(symbol_conf.dso_list, al->map->dso->long_name))))) + goto out_filtered; + + if (symbol_conf.sym_list && al->sym && + !strlist__has_entry(symbol_conf.sym_list, al->sym->name)) + goto out_filtered; + + al->filtered = false; + return 0; + +out_filtered: + al->filtered = true; + return 0; +} + +int event__parse_sample(event_t *event, u64 type, struct sample_data *data) +{ + u64 *array = event->sample.array; + + if (type & PERF_SAMPLE_IP) { + data->ip = event->ip.ip; + array++; + } + + if (type & PERF_SAMPLE_TID) { + u32 *p = (u32 *)array; + data->pid = p[0]; + data->tid = p[1]; + array++; + } + + if (type & PERF_SAMPLE_TIME) { + data->time = *array; + array++; + } + + if (type & PERF_SAMPLE_ADDR) { + data->addr = *array; + array++; + } + + if (type & PERF_SAMPLE_ID) { + data->id = *array; + array++; + } + + if (type & PERF_SAMPLE_STREAM_ID) { + data->stream_id = *array; + array++; + } + + if (type & PERF_SAMPLE_CPU) { + u32 *p = (u32 *)array; + data->cpu = *p; + array++; + } + + if (type & PERF_SAMPLE_PERIOD) { + data->period = *array; + array++; + } + + if (type & PERF_SAMPLE_READ) { + pr_debug("PERF_SAMPLE_READ is unsuported for now\n"); + return -1; + } + + if (type & PERF_SAMPLE_CALLCHAIN) { + data->callchain = (struct ip_callchain *)array; + array += 1 + data->callchain->nr; + } + + if (type & PERF_SAMPLE_RAW) { + u32 *p = (u32 *)array; + data->raw_size = *p; + p++; + data->raw_data = p; + } + return 0; } diff --git a/tools/perf/util/event.h b/tools/perf/util/event.h index a4cc8105cf67..50a7132887f5 100644 --- a/tools/perf/util/event.h +++ b/tools/perf/util/event.h @@ -1,10 +1,10 @@ #ifndef __PERF_RECORD_H #define __PERF_RECORD_H +#include <limits.h> + #include "../perf.h" -#include "util.h" -#include <linux/list.h> -#include <linux/rbtree.h> +#include "map.h" /* * PERF_SAMPLE_IP | PERF_SAMPLE_TID | * @@ -56,11 +56,25 @@ struct read_event { u64 id; }; -struct sample_event{ +struct sample_event { struct perf_event_header header; u64 array[]; }; +struct sample_data { + u64 ip; + u32 pid, tid; + u64 time; + u64 addr; + u64 id; + u64 stream_id; + u32 cpu; + u64 period; + struct ip_callchain *callchain; + u32 raw_size; + void *raw_data; +}; + #define BUILD_ID_SIZE 20 struct build_id_event { @@ -81,79 +95,34 @@ typedef union event_union { } event_t; struct events_stats { - unsigned long total; - unsigned long lost; + u64 total; + u64 lost; }; void event__print_totals(void); -enum map_type { - MAP__FUNCTION = 0, +struct perf_session; - MAP__NR_TYPES, -}; +typedef int (*event__handler_t)(event_t *event, struct perf_session *session); -struct map { - union { - struct rb_node rb_node; - struct list_head node; - }; - u64 start; - u64 end; - enum map_type type; - u64 pgoff; - u64 (*map_ip)(struct map *, u64); - u64 (*unmap_ip)(struct map *, u64); - struct dso *dso; -}; +int event__synthesize_thread(pid_t pid, event__handler_t process, + struct perf_session *session); +void event__synthesize_threads(event__handler_t process, + struct perf_session *session); +int event__synthesize_kernel_mmap(event__handler_t process, + struct perf_session *session, + const char *symbol_name); +int event__synthesize_modules(event__handler_t process, + struct perf_session *session); -static inline u64 map__map_ip(struct map *map, u64 ip) -{ - return ip - map->start + map->pgoff; -} - -static inline u64 map__unmap_ip(struct map *map, u64 ip) -{ - return ip + map->start - map->pgoff; -} - -static inline u64 identity__map_ip(struct map *map __used, u64 ip) -{ - return ip; -} - -struct symbol; - -typedef int (*symbol_filter_t)(struct map *map, struct symbol *sym); - -void map__init(struct map *self, enum map_type type, - u64 start, u64 end, u64 pgoff, struct dso *dso); -struct map *map__new(struct mmap_event *event, enum map_type, - char *cwd, int cwdlen); -void map__delete(struct map *self); -struct map *map__clone(struct map *self); -int map__overlap(struct map *l, struct map *r); -size_t map__fprintf(struct map *self, FILE *fp); -struct symbol *map__find_symbol(struct map *self, u64 addr, - symbol_filter_t filter); -void map__fixup_start(struct map *self); -void map__fixup_end(struct map *self); - -int event__synthesize_thread(pid_t pid, int (*process)(event_t *event)); -void event__synthesize_threads(int (*process)(event_t *event)); - -extern char *event__cwd; -extern int event__cwdlen; -extern struct events_stats event__stats; -extern unsigned long event__total[PERF_RECORD_MAX]; - -int event__process_comm(event_t *self); -int event__process_lost(event_t *self); -int event__process_mmap(event_t *self); -int event__process_task(event_t *self); +int event__process_comm(event_t *self, struct perf_session *session); +int event__process_lost(event_t *self, struct perf_session *session); +int event__process_mmap(event_t *self, struct perf_session *session); +int event__process_task(event_t *self, struct perf_session *session); struct addr_location; -int event__preprocess_sample(const event_t *self, struct addr_location *al, - symbol_filter_t filter); +int event__preprocess_sample(const event_t *self, struct perf_session *session, + struct addr_location *al, symbol_filter_t filter); +int event__parse_sample(event_t *event, u64 type, struct sample_data *data); #endif /* __PERF_RECORD_H */ diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index 4805e6dfd23c..6c9aa16ee51f 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -1,15 +1,19 @@ +#define _FILE_OFFSET_BITS 64 + #include <sys/types.h> +#include <byteswap.h> #include <unistd.h> #include <stdio.h> #include <stdlib.h> #include <linux/list.h> +#include <linux/kernel.h> #include "util.h" #include "header.h" #include "../perf.h" #include "trace-event.h" +#include "session.h" #include "symbol.h" -#include "data_map.h" #include "debug.h" /* @@ -58,35 +62,19 @@ int perf_header_attr__add_id(struct perf_header_attr *self, u64 id) return 0; } -/* - * Create new perf.data header: - */ -struct perf_header *perf_header__new(void) +int perf_header__init(struct perf_header *self) { - struct perf_header *self = zalloc(sizeof(*self)); - - if (self != NULL) { - self->size = 1; - self->attr = malloc(sizeof(void *)); - - if (self->attr == NULL) { - free(self); - self = NULL; - } - } - - return self; + self->size = 1; + self->attr = malloc(sizeof(void *)); + return self->attr == NULL ? -ENOMEM : 0; } -void perf_header__delete(struct perf_header *self) +void perf_header__exit(struct perf_header *self) { int i; - for (i = 0; i < self->attrs; ++i) - perf_header_attr__delete(self->attr[i]); - + perf_header_attr__delete(self->attr[i]); free(self->attr); - free(self); } int perf_header__add_attr(struct perf_header *self, @@ -121,24 +109,28 @@ struct perf_trace_event_type { static int event_count; static struct perf_trace_event_type *events; -void perf_header__push_event(u64 id, const char *name) +int perf_header__push_event(u64 id, const char *name) { if (strlen(name) > MAX_EVENT_NAME) pr_warning("Event %s will be truncated\n", name); if (!events) { events = malloc(sizeof(struct perf_trace_event_type)); - if (!events) - die("nomem"); + if (events == NULL) + return -ENOMEM; } else { - events = realloc(events, (event_count + 1) * sizeof(struct perf_trace_event_type)); - if (!events) - die("nomem"); + struct perf_trace_event_type *nevents; + + nevents = realloc(events, (event_count + 1) * sizeof(*events)); + if (nevents == NULL) + return -ENOMEM; + events = nevents; } memset(&events[event_count], 0, sizeof(struct perf_trace_event_type)); events[event_count].event_id = id; strncpy(events[event_count].name, name, MAX_EVENT_NAME - 1); event_count++; + return 0; } char *perf_header__find_event(u64 id) @@ -185,26 +177,48 @@ static int do_write(int fd, const void *buf, size_t size) return 0; } -static int __dsos__write_buildid_table(struct list_head *head, int fd) +#define NAME_ALIGN 64 + +static int write_padded(int fd, const void *bf, size_t count, + size_t count_aligned) +{ + static const char zero_buf[NAME_ALIGN]; + int err = do_write(fd, bf, count); + + if (!err) + err = do_write(fd, zero_buf, count_aligned - count); + + return err; +} + +#define dsos__for_each_with_build_id(pos, head) \ + list_for_each_entry(pos, head, node) \ + if (!pos->has_build_id) \ + continue; \ + else + +static int __dsos__write_buildid_table(struct list_head *head, u16 misc, int fd) { struct dso *pos; - list_for_each_entry(pos, head, node) { + dsos__for_each_with_build_id(pos, head) { int err; struct build_id_event b; size_t len; - if (!pos->has_build_id) + if (!pos->hit) continue; len = pos->long_name_len + 1; - len = ALIGN(len, 64); + len = ALIGN(len, NAME_ALIGN); memset(&b, 0, sizeof(b)); memcpy(&b.build_id, pos->build_id, sizeof(pos->build_id)); + b.header.misc = misc; b.header.size = sizeof(b) + len; err = do_write(fd, &b, sizeof(b)); if (err < 0) return err; - err = do_write(fd, pos->long_name, len); + err = write_padded(fd, pos->long_name, + pos->long_name_len + 1, len); if (err < 0) return err; } @@ -214,12 +228,143 @@ static int __dsos__write_buildid_table(struct list_head *head, int fd) static int dsos__write_buildid_table(int fd) { - int err = __dsos__write_buildid_table(&dsos__kernel, fd); + int err = __dsos__write_buildid_table(&dsos__kernel, + PERF_RECORD_MISC_KERNEL, fd); if (err == 0) - err = __dsos__write_buildid_table(&dsos__user, fd); + err = __dsos__write_buildid_table(&dsos__user, + PERF_RECORD_MISC_USER, fd); + return err; +} + +int build_id_cache__add_s(const char *sbuild_id, const char *debugdir, + const char *name, bool is_kallsyms) +{ + const size_t size = PATH_MAX; + char *filename = malloc(size), + *linkname = malloc(size), *targetname; + int len, err = -1; + + if (filename == NULL || linkname == NULL) + goto out_free; + + len = snprintf(filename, size, "%s%s%s", + debugdir, is_kallsyms ? "/" : "", name); + if (mkdir_p(filename, 0755)) + goto out_free; + + snprintf(filename + len, sizeof(filename) - len, "/%s", sbuild_id); + + if (access(filename, F_OK)) { + if (is_kallsyms) { + if (copyfile("/proc/kallsyms", filename)) + goto out_free; + } else if (link(name, filename) && copyfile(name, filename)) + goto out_free; + } + + len = snprintf(linkname, size, "%s/.build-id/%.2s", + debugdir, sbuild_id); + + if (access(linkname, X_OK) && mkdir_p(linkname, 0755)) + goto out_free; + + snprintf(linkname + len, size - len, "/%s", sbuild_id + 2); + targetname = filename + strlen(debugdir) - 5; + memcpy(targetname, "../..", 5); + + if (symlink(targetname, linkname) == 0) + err = 0; +out_free: + free(filename); + free(linkname); + return err; +} + +static int build_id_cache__add_b(const u8 *build_id, size_t build_id_size, + const char *name, const char *debugdir, + bool is_kallsyms) +{ + char sbuild_id[BUILD_ID_SIZE * 2 + 1]; + + build_id__sprintf(build_id, build_id_size, sbuild_id); + + return build_id_cache__add_s(sbuild_id, debugdir, name, is_kallsyms); +} + +int build_id_cache__remove_s(const char *sbuild_id, const char *debugdir) +{ + const size_t size = PATH_MAX; + char *filename = malloc(size), + *linkname = malloc(size); + int err = -1; + + if (filename == NULL || linkname == NULL) + goto out_free; + + snprintf(linkname, size, "%s/.build-id/%.2s/%s", + debugdir, sbuild_id, sbuild_id + 2); + + if (access(linkname, F_OK)) + goto out_free; + + if (readlink(linkname, filename, size) < 0) + goto out_free; + + if (unlink(linkname)) + goto out_free; + + /* + * Since the link is relative, we must make it absolute: + */ + snprintf(linkname, size, "%s/.build-id/%.2s/%s", + debugdir, sbuild_id, filename); + + if (unlink(linkname)) + goto out_free; + + err = 0; +out_free: + free(filename); + free(linkname); return err; } +static int dso__cache_build_id(struct dso *self, const char *debugdir) +{ + bool is_kallsyms = self->kernel && self->long_name[0] != '/'; + + return build_id_cache__add_b(self->build_id, sizeof(self->build_id), + self->long_name, debugdir, is_kallsyms); +} + +static int __dsos__cache_build_ids(struct list_head *head, const char *debugdir) +{ + struct dso *pos; + int err = 0; + + dsos__for_each_with_build_id(pos, head) + if (dso__cache_build_id(pos, debugdir)) + err = -1; + + return err; +} + +static int dsos__cache_build_ids(void) +{ + int err_kernel, err_user; + char debugdir[PATH_MAX]; + + snprintf(debugdir, sizeof(debugdir), "%s/%s", getenv("HOME"), + DEBUG_CACHE_DIR); + + if (mkdir(debugdir, 0755) != 0 && errno != EEXIST) + return -1; + + err_kernel = __dsos__cache_build_ids(&dsos__kernel, debugdir); + err_user = __dsos__cache_build_ids(&dsos__user, debugdir); + return err_kernel || err_user ? -1 : 0; +} + static int perf_header__adds_write(struct perf_header *self, int fd) { int nr_sections; @@ -228,7 +373,7 @@ static int perf_header__adds_write(struct perf_header *self, int fd) u64 sec_start; int idx = 0, err; - if (dsos__read_build_ids()) + if (dsos__read_build_ids(true)) perf_header__set_feat(self, HEADER_BUILD_ID); nr_sections = bitmap_weight(self->adds_features, HEADER_FEAT_BITS); @@ -268,7 +413,9 @@ static int perf_header__adds_write(struct perf_header *self, int fd) pr_debug("failed to write buildid table\n"); goto out_free; } - buildid_sec->size = lseek(fd, 0, SEEK_CUR) - buildid_sec->offset; + buildid_sec->size = lseek(fd, 0, SEEK_CUR) - + buildid_sec->offset; + dsos__cache_build_ids(); } lseek(fd, sec_start, SEEK_SET); @@ -371,30 +518,43 @@ int perf_header__write(struct perf_header *self, int fd, bool at_exit) return 0; } -static void do_read(int fd, void *buf, size_t size) +static int do_read(int fd, void *buf, size_t size) { while (size) { int ret = read(fd, buf, size); - if (ret < 0) - die("failed to read"); - if (ret == 0) - die("failed to read: missing data"); + if (ret <= 0) + return -1; size -= ret; buf += ret; } + + return 0; +} + +static int perf_header__getbuffer64(struct perf_header *self, + int fd, void *buf, size_t size) +{ + if (do_read(fd, buf, size)) + return -1; + + if (self->needs_swap) + mem_bswap_64(buf, size); + + return 0; } int perf_header__process_sections(struct perf_header *self, int fd, int (*process)(struct perf_file_section *self, + struct perf_header *ph, int feat, int fd)) { struct perf_file_section *feat_sec; int nr_sections; int sec_size; int idx = 0; - int err = 0, feat = 1; + int err = -1, feat = 1; nr_sections = bitmap_weight(self->adds_features, HEADER_FEAT_BITS); if (!nr_sections) @@ -408,33 +568,45 @@ int perf_header__process_sections(struct perf_header *self, int fd, lseek(fd, self->data_offset + self->data_size, SEEK_SET); - do_read(fd, feat_sec, sec_size); + if (perf_header__getbuffer64(self, fd, feat_sec, sec_size)) + goto out_free; + err = 0; while (idx < nr_sections && feat < HEADER_LAST_FEATURE) { if (perf_header__has_feat(self, feat)) { struct perf_file_section *sec = &feat_sec[idx++]; - err = process(sec, feat, fd); + err = process(sec, self, feat, fd); if (err < 0) break; } ++feat; } - +out_free: free(feat_sec); return err; -}; +} int perf_file_header__read(struct perf_file_header *self, struct perf_header *ph, int fd) { lseek(fd, 0, SEEK_SET); - do_read(fd, self, sizeof(*self)); - if (self->magic != PERF_MAGIC || - self->attr_size != sizeof(struct perf_file_attr)) + if (do_read(fd, self, sizeof(*self)) || + memcmp(&self->magic, __perf_magic, sizeof(self->magic))) return -1; + if (self->attr_size != sizeof(struct perf_file_attr)) { + u64 attr_size = bswap_64(self->attr_size); + + if (attr_size != sizeof(struct perf_file_attr)) + return -1; + + mem_bswap_64(self, offsetof(struct perf_file_header, + adds_features)); + ph->needs_swap = true; + } + if (self->size != sizeof(*self)) { /* Support the previous format */ if (self->size == offsetof(typeof(*self), adds_features)) @@ -444,19 +616,31 @@ int perf_file_header__read(struct perf_file_header *self, } memcpy(&ph->adds_features, &self->adds_features, - sizeof(self->adds_features)); + sizeof(ph->adds_features)); + /* + * FIXME: hack that assumes that if we need swap the perf.data file + * may be coming from an arch with a different word-size, ergo different + * DEFINE_BITMAP format, investigate more later, but for now its mostly + * safe to assume that we have a build-id section. Trace files probably + * have several other issues in this realm anyway... + */ + if (ph->needs_swap) { + memset(&ph->adds_features, 0, sizeof(ph->adds_features)); + perf_header__set_feat(ph, HEADER_BUILD_ID); + } ph->event_offset = self->event_types.offset; - ph->event_size = self->event_types.size; - ph->data_offset = self->data.offset; + ph->event_size = self->event_types.size; + ph->data_offset = self->data.offset; ph->data_size = self->data.size; return 0; } static int perf_file_section__process(struct perf_file_section *self, + struct perf_header *ph, int feat, int fd) { - if (lseek(fd, self->offset, SEEK_SET) < 0) { + if (lseek(fd, self->offset, SEEK_SET) == (off_t)-1) { pr_debug("Failed to lseek to %Ld offset for feature %d, " "continuing...\n", self->offset, feat); return 0; @@ -468,7 +652,7 @@ static int perf_file_section__process(struct perf_file_section *self, break; case HEADER_BUILD_ID: - if (perf_header__read_build_ids(fd, self->offset, self->size)) + if (perf_header__read_build_ids(ph, fd, self->offset, self->size)) pr_debug("Failed to read buildids, continuing...\n"); break; default: @@ -480,7 +664,7 @@ static int perf_file_section__process(struct perf_file_section *self, int perf_header__read(struct perf_header *self, int fd) { - struct perf_file_header f_header; + struct perf_file_header f_header; struct perf_file_attr f_attr; u64 f_id; int nr_attrs, nr_ids, i, j; @@ -497,7 +681,9 @@ int perf_header__read(struct perf_header *self, int fd) struct perf_header_attr *attr; off_t tmp; - do_read(fd, &f_attr, sizeof(f_attr)); + if (perf_header__getbuffer64(self, fd, &f_attr, sizeof(f_attr))) + goto out_errno; + tmp = lseek(fd, 0, SEEK_CUR); attr = perf_header_attr__new(&f_attr.attr); @@ -508,7 +694,8 @@ int perf_header__read(struct perf_header *self, int fd) lseek(fd, f_attr.ids.offset, SEEK_SET); for (j = 0; j < nr_ids; j++) { - do_read(fd, &f_id, sizeof(f_id)); + if (perf_header__getbuffer64(self, fd, &f_id, sizeof(f_id))) + goto out_errno; if (perf_header_attr__add_id(attr, f_id) < 0) { perf_header_attr__delete(attr); @@ -528,7 +715,9 @@ int perf_header__read(struct perf_header *self, int fd) events = malloc(f_header.event_types.size); if (events == NULL) return -ENOMEM; - do_read(fd, events, f_header.event_types.size); + if (perf_header__getbuffer64(self, fd, events, + f_header.event_types.size)) + goto out_errno; event_count = f_header.event_types.size / sizeof(struct perf_trace_event_type); } @@ -538,6 +727,8 @@ int perf_header__read(struct perf_header *self, int fd) self->frozen = 1; return 0; +out_errno: + return -errno; } u64 perf_header__sample_type(struct perf_header *header) diff --git a/tools/perf/util/header.h b/tools/perf/util/header.h index d1dbe2b79c42..82a6af72d4cc 100644 --- a/tools/perf/util/header.h +++ b/tools/perf/util/header.h @@ -5,6 +5,7 @@ #include <sys/types.h> #include <stdbool.h> #include "types.h" +#include "event.h" #include <linux/bitmap.h> @@ -52,11 +53,12 @@ struct perf_header { u64 data_size; u64 event_offset; u64 event_size; + bool needs_swap; DECLARE_BITMAP(adds_features, HEADER_FEAT_BITS); }; -struct perf_header *perf_header__new(void); -void perf_header__delete(struct perf_header *self); +int perf_header__init(struct perf_header *self); +void perf_header__exit(struct perf_header *self); int perf_header__read(struct perf_header *self, int fd); int perf_header__write(struct perf_header *self, int fd, bool at_exit); @@ -64,7 +66,7 @@ int perf_header__write(struct perf_header *self, int fd, bool at_exit); int perf_header__add_attr(struct perf_header *self, struct perf_header_attr *attr); -void perf_header__push_event(u64 id, const char *name); +int perf_header__push_event(u64 id, const char *name); char *perf_header__find_event(u64 id); struct perf_header_attr *perf_header_attr__new(struct perf_event_attr *attr); @@ -80,6 +82,11 @@ bool perf_header__has_feat(const struct perf_header *self, int feat); int perf_header__process_sections(struct perf_header *self, int fd, int (*process)(struct perf_file_section *self, + struct perf_header *ph, int feat, int fd)); +int build_id_cache__add_s(const char *sbuild_id, const char *debugdir, + const char *name, bool is_kallsyms); +int build_id_cache__remove_s(const char *sbuild_id, const char *debugdir); + #endif /* __PERF_HEADER_H */ diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c index 0ebf6ee16caa..e8daf5ca6fd2 100644 --- a/tools/perf/util/hist.c +++ b/tools/perf/util/hist.c @@ -1,9 +1,7 @@ #include "hist.h" - -struct rb_root hist; -struct rb_root collapse_hists; -struct rb_root output_hists; -int callchain; +#include "session.h" +#include "sort.h" +#include <math.h> struct callchain_param callchain_param = { .mode = CHAIN_GRAPH_REL, @@ -14,11 +12,12 @@ struct callchain_param callchain_param = { * histogram, sorted on item, collects counts */ -struct hist_entry *__hist_entry__add(struct addr_location *al, - struct symbol *sym_parent, - u64 count, bool *hit) +struct hist_entry *__perf_session__add_hist_entry(struct perf_session *self, + struct addr_location *al, + struct symbol *sym_parent, + u64 count, bool *hit) { - struct rb_node **p = &hist.rb_node; + struct rb_node **p = &self->hists.rb_node; struct rb_node *parent = NULL; struct hist_entry *he; struct hist_entry entry = { @@ -54,7 +53,7 @@ struct hist_entry *__hist_entry__add(struct addr_location *al, return NULL; *he = entry; rb_link_node(&he->rb_node, parent, p); - rb_insert_color(&he->rb_node, &hist); + rb_insert_color(&he->rb_node, &self->hists); *hit = false; return he; } @@ -102,9 +101,9 @@ void hist_entry__free(struct hist_entry *he) * collapse the histogram */ -void collapse__insert_entry(struct hist_entry *he) +static void collapse__insert_entry(struct rb_root *root, struct hist_entry *he) { - struct rb_node **p = &collapse_hists.rb_node; + struct rb_node **p = &root->rb_node; struct rb_node *parent = NULL; struct hist_entry *iter; int64_t cmp; @@ -128,38 +127,45 @@ void collapse__insert_entry(struct hist_entry *he) } rb_link_node(&he->rb_node, parent, p); - rb_insert_color(&he->rb_node, &collapse_hists); + rb_insert_color(&he->rb_node, root); } -void collapse__resort(void) +void perf_session__collapse_resort(struct perf_session *self) { + struct rb_root tmp; struct rb_node *next; struct hist_entry *n; if (!sort__need_collapse) return; - next = rb_first(&hist); + tmp = RB_ROOT; + next = rb_first(&self->hists); + while (next) { n = rb_entry(next, struct hist_entry, rb_node); next = rb_next(&n->rb_node); - rb_erase(&n->rb_node, &hist); - collapse__insert_entry(n); + rb_erase(&n->rb_node, &self->hists); + collapse__insert_entry(&tmp, n); } + + self->hists = tmp; } /* * reverse the map, sort on count. */ -void output__insert_entry(struct hist_entry *he, u64 min_callchain_hits) +static void perf_session__insert_output_hist_entry(struct rb_root *root, + struct hist_entry *he, + u64 min_callchain_hits) { - struct rb_node **p = &output_hists.rb_node; + struct rb_node **p = &root->rb_node; struct rb_node *parent = NULL; struct hist_entry *iter; - if (callchain) + if (symbol_conf.use_callchain) callchain_param.sort(&he->sorted_chain, &he->callchain, min_callchain_hits, &callchain_param); @@ -174,29 +180,483 @@ void output__insert_entry(struct hist_entry *he, u64 min_callchain_hits) } rb_link_node(&he->rb_node, parent, p); - rb_insert_color(&he->rb_node, &output_hists); + rb_insert_color(&he->rb_node, root); } -void output__resort(u64 total_samples) +void perf_session__output_resort(struct perf_session *self, u64 total_samples) { + struct rb_root tmp; struct rb_node *next; struct hist_entry *n; - struct rb_root *tree = &hist; u64 min_callchain_hits; min_callchain_hits = total_samples * (callchain_param.min_percent / 100); - if (sort__need_collapse) - tree = &collapse_hists; - - next = rb_first(tree); + tmp = RB_ROOT; + next = rb_first(&self->hists); while (next) { n = rb_entry(next, struct hist_entry, rb_node); next = rb_next(&n->rb_node); - rb_erase(&n->rb_node, tree); - output__insert_entry(n, min_callchain_hits); + rb_erase(&n->rb_node, &self->hists); + perf_session__insert_output_hist_entry(&tmp, n, + min_callchain_hits); + } + + self->hists = tmp; +} + +static size_t callchain__fprintf_left_margin(FILE *fp, int left_margin) +{ + int i; + int ret = fprintf(fp, " "); + + for (i = 0; i < left_margin; i++) + ret += fprintf(fp, " "); + + return ret; +} + +static size_t ipchain__fprintf_graph_line(FILE *fp, int depth, int depth_mask, + int left_margin) +{ + int i; + size_t ret = callchain__fprintf_left_margin(fp, left_margin); + + for (i = 0; i < depth; i++) + if (depth_mask & (1 << i)) + ret += fprintf(fp, "| "); + else + ret += fprintf(fp, " "); + + ret += fprintf(fp, "\n"); + + return ret; +} + +static size_t ipchain__fprintf_graph(FILE *fp, struct callchain_list *chain, + int depth, int depth_mask, int count, + u64 total_samples, int hits, + int left_margin) +{ + int i; + size_t ret = 0; + + ret += callchain__fprintf_left_margin(fp, left_margin); + for (i = 0; i < depth; i++) { + if (depth_mask & (1 << i)) + ret += fprintf(fp, "|"); + else + ret += fprintf(fp, " "); + if (!count && i == depth - 1) { + double percent; + + percent = hits * 100.0 / total_samples; + ret += percent_color_fprintf(fp, "--%2.2f%%-- ", percent); + } else + ret += fprintf(fp, "%s", " "); + } + if (chain->sym) + ret += fprintf(fp, "%s\n", chain->sym->name); + else + ret += fprintf(fp, "%p\n", (void *)(long)chain->ip); + + return ret; +} + +static struct symbol *rem_sq_bracket; +static struct callchain_list rem_hits; + +static void init_rem_hits(void) +{ + rem_sq_bracket = malloc(sizeof(*rem_sq_bracket) + 6); + if (!rem_sq_bracket) { + fprintf(stderr, "Not enough memory to display remaining hits\n"); + return; + } + + strcpy(rem_sq_bracket->name, "[...]"); + rem_hits.sym = rem_sq_bracket; +} + +static size_t __callchain__fprintf_graph(FILE *fp, struct callchain_node *self, + u64 total_samples, int depth, + int depth_mask, int left_margin) +{ + struct rb_node *node, *next; + struct callchain_node *child; + struct callchain_list *chain; + int new_depth_mask = depth_mask; + u64 new_total; + u64 remaining; + size_t ret = 0; + int i; + + if (callchain_param.mode == CHAIN_GRAPH_REL) + new_total = self->children_hit; + else + new_total = total_samples; + + remaining = new_total; + + node = rb_first(&self->rb_root); + while (node) { + u64 cumul; + + child = rb_entry(node, struct callchain_node, rb_node); + cumul = cumul_hits(child); + remaining -= cumul; + + /* + * The depth mask manages the output of pipes that show + * the depth. We don't want to keep the pipes of the current + * level for the last child of this depth. + * Except if we have remaining filtered hits. They will + * supersede the last child + */ + next = rb_next(node); + if (!next && (callchain_param.mode != CHAIN_GRAPH_REL || !remaining)) + new_depth_mask &= ~(1 << (depth - 1)); + + /* + * But we keep the older depth mask for the line seperator + * to keep the level link until we reach the last child + */ + ret += ipchain__fprintf_graph_line(fp, depth, depth_mask, + left_margin); + i = 0; + list_for_each_entry(chain, &child->val, list) { + if (chain->ip >= PERF_CONTEXT_MAX) + continue; + ret += ipchain__fprintf_graph(fp, chain, depth, + new_depth_mask, i++, + new_total, + cumul, + left_margin); + } + ret += __callchain__fprintf_graph(fp, child, new_total, + depth + 1, + new_depth_mask | (1 << depth), + left_margin); + node = next; + } + + if (callchain_param.mode == CHAIN_GRAPH_REL && + remaining && remaining != new_total) { + + if (!rem_sq_bracket) + return ret; + + new_depth_mask &= ~(1 << (depth - 1)); + + ret += ipchain__fprintf_graph(fp, &rem_hits, depth, + new_depth_mask, 0, new_total, + remaining, left_margin); + } + + return ret; +} + +static size_t callchain__fprintf_graph(FILE *fp, struct callchain_node *self, + u64 total_samples, int left_margin) +{ + struct callchain_list *chain; + bool printed = false; + int i = 0; + int ret = 0; + + list_for_each_entry(chain, &self->val, list) { + if (chain->ip >= PERF_CONTEXT_MAX) + continue; + + if (!i++ && sort__first_dimension == SORT_SYM) + continue; + + if (!printed) { + ret += callchain__fprintf_left_margin(fp, left_margin); + ret += fprintf(fp, "|\n"); + ret += callchain__fprintf_left_margin(fp, left_margin); + ret += fprintf(fp, "---"); + + left_margin += 3; + printed = true; + } else + ret += callchain__fprintf_left_margin(fp, left_margin); + + if (chain->sym) + ret += fprintf(fp, " %s\n", chain->sym->name); + else + ret += fprintf(fp, " %p\n", (void *)(long)chain->ip); + } + + ret += __callchain__fprintf_graph(fp, self, total_samples, 1, 1, left_margin); + + return ret; +} + +static size_t callchain__fprintf_flat(FILE *fp, struct callchain_node *self, + u64 total_samples) +{ + struct callchain_list *chain; + size_t ret = 0; + + if (!self) + return 0; + + ret += callchain__fprintf_flat(fp, self->parent, total_samples); + + + list_for_each_entry(chain, &self->val, list) { + if (chain->ip >= PERF_CONTEXT_MAX) + continue; + if (chain->sym) + ret += fprintf(fp, " %s\n", chain->sym->name); + else + ret += fprintf(fp, " %p\n", + (void *)(long)chain->ip); + } + + return ret; +} + +static size_t hist_entry_callchain__fprintf(FILE *fp, struct hist_entry *self, + u64 total_samples, int left_margin) +{ + struct rb_node *rb_node; + struct callchain_node *chain; + size_t ret = 0; + + rb_node = rb_first(&self->sorted_chain); + while (rb_node) { + double percent; + + chain = rb_entry(rb_node, struct callchain_node, rb_node); + percent = chain->hit * 100.0 / total_samples; + switch (callchain_param.mode) { + case CHAIN_FLAT: + ret += percent_color_fprintf(fp, " %6.2f%%\n", + percent); + ret += callchain__fprintf_flat(fp, chain, total_samples); + break; + case CHAIN_GRAPH_ABS: /* Falldown */ + case CHAIN_GRAPH_REL: + ret += callchain__fprintf_graph(fp, chain, total_samples, + left_margin); + case CHAIN_NONE: + default: + break; + } + ret += fprintf(fp, "\n"); + rb_node = rb_next(rb_node); + } + + return ret; +} + +static size_t hist_entry__fprintf(struct hist_entry *self, + struct perf_session *session, + struct perf_session *pair_session, + bool show_displacement, + long displacement, FILE *fp) +{ + struct sort_entry *se; + u64 count, total; + const char *sep = symbol_conf.field_sep; + size_t ret; + + if (symbol_conf.exclude_other && !self->parent) + return 0; + + if (pair_session) { + count = self->pair ? self->pair->count : 0; + total = pair_session->events_stats.total; + } else { + count = self->count; + total = session->events_stats.total; + } + + if (total) + ret = percent_color_fprintf(fp, sep ? "%.2f" : " %6.2f%%", + (count * 100.0) / total); + else + ret = fprintf(fp, sep ? "%lld" : "%12lld ", count); + + if (symbol_conf.show_nr_samples) { + if (sep) + fprintf(fp, "%c%lld", *sep, count); + else + fprintf(fp, "%11lld", count); + } + + if (pair_session) { + char bf[32]; + double old_percent = 0, new_percent = 0, diff; + + if (total > 0) + old_percent = (count * 100.0) / total; + if (session->events_stats.total > 0) + new_percent = (self->count * 100.0) / session->events_stats.total; + + diff = new_percent - old_percent; + + if (fabs(diff) >= 0.01) + snprintf(bf, sizeof(bf), "%+4.2F%%", diff); + else + snprintf(bf, sizeof(bf), " "); + + if (sep) + ret += fprintf(fp, "%c%s", *sep, bf); + else + ret += fprintf(fp, "%11.11s", bf); + + if (show_displacement) { + if (displacement) + snprintf(bf, sizeof(bf), "%+4ld", displacement); + else + snprintf(bf, sizeof(bf), " "); + + if (sep) + fprintf(fp, "%c%s", *sep, bf); + else + fprintf(fp, "%6.6s", bf); + } + } + + list_for_each_entry(se, &hist_entry__sort_list, list) { + if (se->elide) + continue; + + fprintf(fp, "%s", sep ?: " "); + ret += se->print(fp, self, se->width ? *se->width : 0); + } + + ret += fprintf(fp, "\n"); + + if (symbol_conf.use_callchain) { + int left_margin = 0; + + if (sort__first_dimension == SORT_COMM) { + se = list_first_entry(&hist_entry__sort_list, typeof(*se), + list); + left_margin = se->width ? *se->width : 0; + left_margin -= thread__comm_len(self->thread); + } + + hist_entry_callchain__fprintf(fp, self, session->events_stats.total, + left_margin); + } + + return ret; +} + +size_t perf_session__fprintf_hists(struct perf_session *self, + struct perf_session *pair, + bool show_displacement, FILE *fp) +{ + struct sort_entry *se; + struct rb_node *nd; + size_t ret = 0; + unsigned long position = 1; + long displacement = 0; + unsigned int width; + const char *sep = symbol_conf.field_sep; + char *col_width = symbol_conf.col_width_list_str; + + init_rem_hits(); + + fprintf(fp, "# %s", pair ? "Baseline" : "Overhead"); + + if (symbol_conf.show_nr_samples) { + if (sep) + fprintf(fp, "%cSamples", *sep); + else + fputs(" Samples ", fp); + } + + if (pair) { + if (sep) + ret += fprintf(fp, "%cDelta", *sep); + else + ret += fprintf(fp, " Delta "); + + if (show_displacement) { + if (sep) + ret += fprintf(fp, "%cDisplacement", *sep); + else + ret += fprintf(fp, " Displ"); + } + } + + list_for_each_entry(se, &hist_entry__sort_list, list) { + if (se->elide) + continue; + if (sep) { + fprintf(fp, "%c%s", *sep, se->header); + continue; + } + width = strlen(se->header); + if (se->width) { + if (symbol_conf.col_width_list_str) { + if (col_width) { + *se->width = atoi(col_width); + col_width = strchr(col_width, ','); + if (col_width) + ++col_width; + } + } + width = *se->width = max(*se->width, width); + } + fprintf(fp, " %*s", width, se->header); + } + fprintf(fp, "\n"); + + if (sep) + goto print_entries; + + fprintf(fp, "# ........"); + if (symbol_conf.show_nr_samples) + fprintf(fp, " .........."); + if (pair) { + fprintf(fp, " .........."); + if (show_displacement) + fprintf(fp, " ....."); + } + list_for_each_entry(se, &hist_entry__sort_list, list) { + unsigned int i; + + if (se->elide) + continue; + + fprintf(fp, " "); + if (se->width) + width = *se->width; + else + width = strlen(se->header); + for (i = 0; i < width; i++) + fprintf(fp, "."); + } + + fprintf(fp, "\n#\n"); + +print_entries: + for (nd = rb_first(&self->hists); nd; nd = rb_next(nd)) { + struct hist_entry *h = rb_entry(nd, struct hist_entry, rb_node); + + if (show_displacement) { + if (h->pair != NULL) + displacement = ((long)h->pair->position - + (long)position); + else + displacement = 0; + ++position; + } + ret += hist_entry__fprintf(h, self, pair, show_displacement, + displacement, fp); } + + free(rem_sq_bracket); + + return ret; } diff --git a/tools/perf/util/hist.h b/tools/perf/util/hist.h index 3020db0c9292..e5f99b24048b 100644 --- a/tools/perf/util/hist.h +++ b/tools/perf/util/hist.h @@ -1,50 +1,27 @@ #ifndef __PERF_HIST_H #define __PERF_HIST_H -#include "../builtin.h" -#include "util.h" - -#include "color.h" -#include <linux/list.h> -#include "cache.h" -#include <linux/rbtree.h> -#include "symbol.h" -#include "string.h" +#include <linux/types.h> #include "callchain.h" -#include "strlist.h" -#include "values.h" - -#include "../perf.h" -#include "debug.h" -#include "header.h" - -#include "parse-options.h" -#include "parse-events.h" -#include "thread.h" -#include "sort.h" - -extern struct rb_root hist; -extern struct rb_root collapse_hists; -extern struct rb_root output_hists; -extern int callchain; extern struct callchain_param callchain_param; -extern unsigned long total; -extern unsigned long total_mmap; -extern unsigned long total_comm; -extern unsigned long total_fork; -extern unsigned long total_unknown; -extern unsigned long total_lost; -struct hist_entry *__hist_entry__add(struct addr_location *al, - struct symbol *parent, - u64 count, bool *hit); +struct perf_session; +struct hist_entry; +struct addr_location; +struct symbol; + +struct hist_entry *__perf_session__add_hist_entry(struct perf_session *self, + struct addr_location *al, + struct symbol *parent, + u64 count, bool *hit); extern int64_t hist_entry__cmp(struct hist_entry *, struct hist_entry *); extern int64_t hist_entry__collapse(struct hist_entry *, struct hist_entry *); -extern void hist_entry__free(struct hist_entry *); -extern void collapse__insert_entry(struct hist_entry *); -extern void collapse__resort(void); -extern void output__insert_entry(struct hist_entry *, u64); -extern void output__resort(u64); +void hist_entry__free(struct hist_entry *); +void perf_session__output_resort(struct perf_session *self, u64 total_samples); +void perf_session__collapse_resort(struct perf_session *self); +size_t perf_session__fprintf_hists(struct perf_session *self, + struct perf_session *pair, + bool show_displacement, FILE *fp); #endif /* __PERF_HIST_H */ diff --git a/tools/perf/util/include/linux/hash.h b/tools/perf/util/include/linux/hash.h new file mode 100644 index 000000000000..201f57397997 --- /dev/null +++ b/tools/perf/util/include/linux/hash.h @@ -0,0 +1,5 @@ +#include "../../../../include/linux/hash.h" + +#ifndef PERF_HASH_H +#define PERF_HASH_H +#endif diff --git a/tools/perf/util/include/linux/kernel.h b/tools/perf/util/include/linux/kernel.h index 21c0274c02fa..f2611655ab51 100644 --- a/tools/perf/util/include/linux/kernel.h +++ b/tools/perf/util/include/linux/kernel.h @@ -101,5 +101,6 @@ simple_strtoul(const char *nptr, char **endptr, int base) eprintf(n, pr_fmt(fmt), ##__VA_ARGS__) #define pr_debug2(fmt, ...) pr_debugN(2, pr_fmt(fmt), ##__VA_ARGS__) #define pr_debug3(fmt, ...) pr_debugN(3, pr_fmt(fmt), ##__VA_ARGS__) +#define pr_debug4(fmt, ...) pr_debugN(4, pr_fmt(fmt), ##__VA_ARGS__) #endif diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c index 69f94fe9db20..e509cd59c67d 100644 --- a/tools/perf/util/map.c +++ b/tools/perf/util/map.c @@ -5,6 +5,11 @@ #include <stdio.h> #include "debug.h" +const char *map_type__name[MAP__NR_TYPES] = { + [MAP__FUNCTION] = "Functions", + [MAP__VARIABLE] = "Variables", +}; + static inline int is_anon_memory(const char *filename) { return strcmp(filename, "//anon") == 0; @@ -68,8 +73,13 @@ struct map *map__new(struct mmap_event *event, enum map_type type, map__init(self, type, event->start, event->start + event->len, event->pgoff, dso); - if (self->dso == vdso || anon) + if (anon) { +set_identity: self->map_ip = self->unmap_ip = identity__map_ip; + } else if (strcmp(filename, "[vdso]") == 0) { + dso__set_loaded(dso, self->type); + goto set_identity; + } } return self; out_delete: @@ -104,43 +114,74 @@ void map__fixup_end(struct map *self) #define DSO__DELETED "(deleted)" -struct symbol *map__find_symbol(struct map *self, u64 addr, - symbol_filter_t filter) +int map__load(struct map *self, symbol_filter_t filter) { - if (!dso__loaded(self->dso, self->type)) { - int nr = dso__load(self->dso, self, filter); - - if (nr < 0) { - if (self->dso->has_build_id) { - char sbuild_id[BUILD_ID_SIZE * 2 + 1]; - - build_id__sprintf(self->dso->build_id, - sizeof(self->dso->build_id), - sbuild_id); - pr_warning("%s with build id %s not found", - self->dso->long_name, sbuild_id); - } else - pr_warning("Failed to open %s", - self->dso->long_name); - pr_warning(", continuing without symbols\n"); - return NULL; - } else if (nr == 0) { - const char *name = self->dso->long_name; - const size_t len = strlen(name); - const size_t real_len = len - sizeof(DSO__DELETED); - - if (len > sizeof(DSO__DELETED) && - strcmp(name + real_len + 1, DSO__DELETED) == 0) { - pr_warning("%.*s was updated, restart the long running apps that use it!\n", - (int)real_len, name); - } else { - pr_warning("no symbols found in %s, maybe install a debug package?\n", name); - } - return NULL; + const char *name = self->dso->long_name; + int nr; + + if (dso__loaded(self->dso, self->type)) + return 0; + + nr = dso__load(self->dso, self, filter); + if (nr < 0) { + if (self->dso->has_build_id) { + char sbuild_id[BUILD_ID_SIZE * 2 + 1]; + + build_id__sprintf(self->dso->build_id, + sizeof(self->dso->build_id), + sbuild_id); + pr_warning("%s with build id %s not found", + name, sbuild_id); + } else + pr_warning("Failed to open %s", name); + + pr_warning(", continuing without symbols\n"); + return -1; + } else if (nr == 0) { + const size_t len = strlen(name); + const size_t real_len = len - sizeof(DSO__DELETED); + + if (len > sizeof(DSO__DELETED) && + strcmp(name + real_len + 1, DSO__DELETED) == 0) { + pr_warning("%.*s was updated, restart the long " + "running apps that use it!\n", + (int)real_len, name); + } else { + pr_warning("no symbols found in %s, maybe install " + "a debug package?\n", name); } + + return -1; } + /* + * Only applies to the kernel, as its symtabs aren't relative like the + * module ones. + */ + if (self->dso->kernel) + map__reloc_vmlinux(self); - return self->dso->find_symbol(self->dso, self->type, addr); + return 0; +} + +struct symbol *map__find_symbol(struct map *self, u64 addr, + symbol_filter_t filter) +{ + if (map__load(self, filter) < 0) + return NULL; + + return dso__find_symbol(self->dso, self->type, addr); +} + +struct symbol *map__find_symbol_by_name(struct map *self, const char *name, + symbol_filter_t filter) +{ + if (map__load(self, filter) < 0) + return NULL; + + if (!dso__sorted_by_name(self->dso, self->type)) + dso__sort_by_name(self->dso, self->type); + + return dso__find_symbol_by_name(self->dso, self->type, name); } struct map *map__clone(struct map *self) @@ -174,3 +215,23 @@ size_t map__fprintf(struct map *self, FILE *fp) return fprintf(fp, " %Lx-%Lx %Lx %s\n", self->start, self->end, self->pgoff, self->dso->name); } + +/* + * objdump wants/reports absolute IPs for ET_EXEC, and RIPs for ET_DYN. + * map->dso->adjust_symbols==1 for ET_EXEC-like cases. + */ +u64 map__rip_2objdump(struct map *map, u64 rip) +{ + u64 addr = map->dso->adjust_symbols ? + map->unmap_ip(map, rip) : /* RIP -> IP */ + rip; + return addr; +} + +u64 map__objdump_2ip(struct map *map, u64 addr) +{ + u64 ip = map->dso->adjust_symbols ? + addr : + map->unmap_ip(map, addr); /* RIP -> IP */ + return ip; +} diff --git a/tools/perf/util/map.h b/tools/perf/util/map.h new file mode 100644 index 000000000000..b756368076c6 --- /dev/null +++ b/tools/perf/util/map.h @@ -0,0 +1,94 @@ +#ifndef __PERF_MAP_H +#define __PERF_MAP_H + +#include <linux/compiler.h> +#include <linux/list.h> +#include <linux/rbtree.h> +#include <linux/types.h> + +enum map_type { + MAP__FUNCTION = 0, + MAP__VARIABLE, +}; + +#define MAP__NR_TYPES (MAP__VARIABLE + 1) + +extern const char *map_type__name[MAP__NR_TYPES]; + +struct dso; +struct ref_reloc_sym; +struct map_groups; + +struct map { + union { + struct rb_node rb_node; + struct list_head node; + }; + u64 start; + u64 end; + enum map_type type; + u64 pgoff; + + /* ip -> dso rip */ + u64 (*map_ip)(struct map *, u64); + /* dso rip -> ip */ + u64 (*unmap_ip)(struct map *, u64); + + struct dso *dso; +}; + +struct kmap { + struct ref_reloc_sym *ref_reloc_sym; + struct map_groups *kmaps; +}; + +static inline struct kmap *map__kmap(struct map *self) +{ + return (struct kmap *)(self + 1); +} + +static inline u64 map__map_ip(struct map *map, u64 ip) +{ + return ip - map->start + map->pgoff; +} + +static inline u64 map__unmap_ip(struct map *map, u64 ip) +{ + return ip + map->start - map->pgoff; +} + +static inline u64 identity__map_ip(struct map *map __used, u64 ip) +{ + return ip; +} + + +/* rip/ip <-> addr suitable for passing to `objdump --start-address=` */ +u64 map__rip_2objdump(struct map *map, u64 rip); +u64 map__objdump_2ip(struct map *map, u64 addr); + +struct symbol; +struct mmap_event; + +typedef int (*symbol_filter_t)(struct map *map, struct symbol *sym); + +void map__init(struct map *self, enum map_type type, + u64 start, u64 end, u64 pgoff, struct dso *dso); +struct map *map__new(struct mmap_event *event, enum map_type, + char *cwd, int cwdlen); +void map__delete(struct map *self); +struct map *map__clone(struct map *self); +int map__overlap(struct map *l, struct map *r); +size_t map__fprintf(struct map *self, FILE *fp); + +int map__load(struct map *self, symbol_filter_t filter); +struct symbol *map__find_symbol(struct map *self, + u64 addr, symbol_filter_t filter); +struct symbol *map__find_symbol_by_name(struct map *self, const char *name, + symbol_filter_t filter); +void map__fixup_start(struct map *self); +void map__fixup_end(struct map *self); + +void map__reloc_vmlinux(struct map *self); + +#endif /* __PERF_MAP_H */ diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c index 9e5dbd66d34d..05d0c5c2030c 100644 --- a/tools/perf/util/parse-events.c +++ b/tools/perf/util/parse-events.c @@ -197,7 +197,7 @@ struct tracepoint_path *tracepoint_id_to_path(u64 config) if (id == config) { closedir(evt_dir); closedir(sys_dir); - path = zalloc(sizeof(path)); + path = zalloc(sizeof(*path)); path->system = malloc(MAX_EVENT_LENGTH); if (!path->system) { free(path); @@ -450,7 +450,8 @@ parse_single_tracepoint_event(char *sys_name, /* sys + ':' + event + ':' + flags*/ #define MAX_EVOPT_LEN (MAX_EVENT_LENGTH * 2 + 2 + 128) static enum event_result -parse_subsystem_tracepoint_event(char *sys_name, char *flags) +parse_multiple_tracepoint_event(char *sys_name, const char *evt_exp, + char *flags) { char evt_path[MAXPATHLEN]; struct dirent *evt_ent; @@ -467,7 +468,6 @@ parse_subsystem_tracepoint_event(char *sys_name, char *flags) while ((evt_ent = readdir(evt_dir))) { char event_opt[MAX_EVOPT_LEN + 1]; int len; - unsigned int rem = MAX_EVOPT_LEN; if (!strcmp(evt_ent->d_name, ".") || !strcmp(evt_ent->d_name, "..") @@ -475,20 +475,15 @@ parse_subsystem_tracepoint_event(char *sys_name, char *flags) || !strcmp(evt_ent->d_name, "filter")) continue; - len = snprintf(event_opt, MAX_EVOPT_LEN, "%s:%s", sys_name, - evt_ent->d_name); + if (!strglobmatch(evt_ent->d_name, evt_exp)) + continue; + + len = snprintf(event_opt, MAX_EVOPT_LEN, "%s:%s%s%s", sys_name, + evt_ent->d_name, flags ? ":" : "", + flags ?: ""); if (len < 0) return EVT_FAILED; - rem -= len; - if (flags) { - if (rem < strlen(flags) + 1) - return EVT_FAILED; - - strcat(event_opt, ":"); - strcat(event_opt, flags); - } - if (parse_events(NULL, event_opt, 0)) return EVT_FAILED; } @@ -531,9 +526,10 @@ static enum event_result parse_tracepoint_event(const char **strp, if (evt_length >= MAX_EVENT_LENGTH) return EVT_FAILED; - if (!strcmp(evt_name, "*")) { + if (strpbrk(evt_name, "*?")) { *strp = evt_name + evt_length; - return parse_subsystem_tracepoint_event(sys_name, flags); + return parse_multiple_tracepoint_event(sys_name, evt_name, + flags); } else return parse_single_tracepoint_event(sys_name, evt_name, evt_length, flags, @@ -762,11 +758,11 @@ modifier: return ret; } -static void store_event_type(const char *orgname) +static int store_event_type(const char *orgname) { char filename[PATH_MAX], *c; FILE *file; - int id; + int id, n; sprintf(filename, "%s/", debugfs_path); strncat(filename, orgname, strlen(orgname)); @@ -778,11 +774,14 @@ static void store_event_type(const char *orgname) file = fopen(filename, "r"); if (!file) - return; - if (fscanf(file, "%i", &id) < 1) - die("cannot store event ID"); + return 0; + n = fscanf(file, "%i", &id); fclose(file); - perf_header__push_event(id, orgname); + if (n < 1) { + pr_err("cannot store event ID\n"); + return -EINVAL; + } + return perf_header__push_event(id, orgname); } int parse_events(const struct option *opt __used, const char *str, int unset __used) @@ -791,7 +790,8 @@ int parse_events(const struct option *opt __used, const char *str, int unset __u enum event_result ret; if (strchr(str, ':')) - store_event_type(str); + if (store_event_type(str) < 0) + return -1; for (;;) { if (nr_counters == MAX_COUNTERS) @@ -844,11 +844,12 @@ int parse_filter(const struct option *opt __used, const char *str, } static const char * const event_type_descriptors[] = { - "", "Hardware event", "Software event", "Tracepoint event", "Hardware cache event", + "Raw hardware event descriptor", + "Hardware breakpoint", }; /* @@ -881,7 +882,7 @@ static void print_tracepoint_events(void) snprintf(evt_path, MAXPATHLEN, "%s:%s", sys_dirent.d_name, evt_dirent.d_name); printf(" %-42s [%s]\n", evt_path, - event_type_descriptors[PERF_TYPE_TRACEPOINT+1]); + event_type_descriptors[PERF_TYPE_TRACEPOINT]); } closedir(evt_dir); } @@ -901,9 +902,7 @@ void print_events(void) printf("List of pre-defined events (to be used in -e):\n"); for (i = 0; i < ARRAY_SIZE(event_symbols); i++, syms++) { - type = syms->type + 1; - if (type >= ARRAY_SIZE(event_type_descriptors)) - type = 0; + type = syms->type; if (type != prev_type) printf("\n"); @@ -928,17 +927,19 @@ void print_events(void) for (i = 0; i < PERF_COUNT_HW_CACHE_RESULT_MAX; i++) { printf(" %-42s [%s]\n", event_cache_name(type, op, i), - event_type_descriptors[4]); + event_type_descriptors[PERF_TYPE_HW_CACHE]); } } } printf("\n"); - printf(" %-42s [raw hardware event descriptor]\n", - "rNNN"); + printf(" %-42s [%s]\n", + "rNNN", event_type_descriptors[PERF_TYPE_RAW]); printf("\n"); - printf(" %-42s [hardware breakpoint]\n", "mem:<addr>[:access]"); + printf(" %-42s [%s]\n", + "mem:<addr>[:access]", + event_type_descriptors[PERF_TYPE_BREAKPOINT]); printf("\n"); print_tracepoint_events(); diff --git a/tools/perf/util/parse-options.c b/tools/perf/util/parse-options.c index 6d8af48c925e..efebd5b476b3 100644 --- a/tools/perf/util/parse-options.c +++ b/tools/perf/util/parse-options.c @@ -430,6 +430,9 @@ int usage_with_options_internal(const char * const *usagestr, pos = fprintf(stderr, " "); if (opts->short_name) pos += fprintf(stderr, "-%c", opts->short_name); + else + pos += fprintf(stderr, " "); + if (opts->long_name && opts->short_name) pos += fprintf(stderr, ", "); if (opts->long_name) diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c index cd7fbda5e2a5..8f0568849691 100644 --- a/tools/perf/util/probe-event.c +++ b/tools/perf/util/probe-event.c @@ -37,6 +37,8 @@ #include "string.h" #include "strlist.h" #include "debug.h" +#include "cache.h" +#include "color.h" #include "parse-events.h" /* For debugfs_path */ #include "probe-event.h" @@ -48,6 +50,9 @@ /* If there is no space to write, returns -E2BIG. */ static int e_snprintf(char *str, size_t size, const char *format, ...) + __attribute__((format(printf, 3, 4))); + +static int e_snprintf(char *str, size_t size, const char *format, ...) { int ret; va_list ap; @@ -59,6 +64,54 @@ static int e_snprintf(char *str, size_t size, const char *format, ...) return ret; } +void parse_line_range_desc(const char *arg, struct line_range *lr) +{ + const char *ptr; + char *tmp; + /* + * <Syntax> + * SRC:SLN[+NUM|-ELN] + * FUNC[:SLN[+NUM|-ELN]] + */ + ptr = strchr(arg, ':'); + if (ptr) { + lr->start = (unsigned int)strtoul(ptr + 1, &tmp, 0); + if (*tmp == '+') + lr->end = lr->start + (unsigned int)strtoul(tmp + 1, + &tmp, 0); + else if (*tmp == '-') + lr->end = (unsigned int)strtoul(tmp + 1, &tmp, 0); + else + lr->end = 0; + pr_debug("Line range is %u to %u\n", lr->start, lr->end); + if (lr->end && lr->start > lr->end) + semantic_error("Start line must be smaller" + " than end line."); + if (*tmp != '\0') + semantic_error("Tailing with invalid character '%d'.", + *tmp); + tmp = strndup(arg, (ptr - arg)); + } else + tmp = strdup(arg); + + if (strchr(tmp, '.')) + lr->file = tmp; + else + lr->function = tmp; +} + +/* Check the name is good for event/group */ +static bool check_event_name(const char *name) +{ + if (!isalpha(*name) && *name != '_') + return false; + while (*++name != '\0') { + if (!isalpha(*name) && !isdigit(*name) && *name != '_') + return false; + } + return true; +} + /* Parse probepoint definition. */ static void parse_perf_probe_probepoint(char *arg, struct probe_point *pp) { @@ -66,10 +119,26 @@ static void parse_perf_probe_probepoint(char *arg, struct probe_point *pp) char c, nc = 0; /* * <Syntax> - * perf probe SRC:LN - * perf probe FUNC[+OFFS|%return][@SRC] + * perf probe [EVENT=]SRC:LN + * perf probe [EVENT=]FUNC[+OFFS|%return][@SRC] + * + * TODO:Group name support */ + ptr = strchr(arg, '='); + if (ptr) { /* Event name */ + *ptr = '\0'; + tmp = ptr + 1; + ptr = strchr(arg, ':'); + if (ptr) /* Group name is not supported yet. */ + semantic_error("Group name is not supported yet."); + if (!check_event_name(arg)) + semantic_error("%s is bad for event name -it must " + "follow C symbol-naming rule.", arg); + pp->event = strdup(arg); + arg = tmp; + } + ptr = strpbrk(arg, ":+@%"); if (ptr) { nc = *ptr; @@ -147,10 +216,13 @@ static void parse_perf_probe_probepoint(char *arg, struct probe_point *pp) } /* Parse perf-probe event definition */ -int parse_perf_probe_event(const char *str, struct probe_point *pp) +void parse_perf_probe_event(const char *str, struct probe_point *pp, + bool *need_dwarf) { char **argv; - int argc, i, need_dwarf = 0; + int argc, i; + + *need_dwarf = false; argv = argv_split(str, &argc); if (!argv) @@ -161,7 +233,7 @@ int parse_perf_probe_event(const char *str, struct probe_point *pp) /* Parse probe point */ parse_perf_probe_probepoint(argv[0], pp); if (pp->file || pp->line) - need_dwarf = 1; + *need_dwarf = true; /* Copy arguments and ensure return probe has no C argument */ pp->nr_args = argc - 1; @@ -174,17 +246,15 @@ int parse_perf_probe_event(const char *str, struct probe_point *pp) if (pp->retprobe) semantic_error("You can't specify local" " variable for kretprobe"); - need_dwarf = 1; + *need_dwarf = true; } } argv_free(argv); - return need_dwarf; } /* Parse kprobe_events event into struct probe_point */ -void parse_trace_kprobe_event(const char *str, char **group, char **event, - struct probe_point *pp) +void parse_trace_kprobe_event(const char *str, struct probe_point *pp) { char pr; char *p; @@ -200,18 +270,17 @@ void parse_trace_kprobe_event(const char *str, char **group, char **event, /* Scan event and group name. */ ret = sscanf(argv[0], "%c:%a[^/ \t]/%a[^ \t]", - &pr, (float *)(void *)group, (float *)(void *)event); + &pr, (float *)(void *)&pp->group, + (float *)(void *)&pp->event); if (ret != 3) semantic_error("Failed to parse event name: %s", argv[0]); - pr_debug("Group:%s Event:%s probe:%c\n", *group, *event, pr); - - if (!pp) - goto end; + pr_debug("Group:%s Event:%s probe:%c\n", pp->group, pp->event, pr); pp->retprobe = (pr == 'r'); /* Scan function name and offset */ - ret = sscanf(argv[1], "%a[^+]+%d", (float *)(void *)&pp->function, &pp->offset); + ret = sscanf(argv[1], "%a[^+]+%d", (float *)(void *)&pp->function, + &pp->offset); if (ret == 1) pp->offset = 0; @@ -230,17 +299,18 @@ void parse_trace_kprobe_event(const char *str, char **group, char **event, die("Failed to copy argument."); } -end: argv_free(argv); } -int synthesize_perf_probe_event(struct probe_point *pp) +/* Synthesize only probe point (not argument) */ +int synthesize_perf_probe_point(struct probe_point *pp) { char *buf; char offs[64] = "", line[64] = ""; - int i, len, ret; + int ret; pp->probes[0] = buf = zalloc(MAX_CMDLEN); + pp->found = 1; if (!buf) die("Failed to allocate memory by zalloc."); if (pp->offset) { @@ -258,11 +328,26 @@ int synthesize_perf_probe_event(struct probe_point *pp) ret = e_snprintf(buf, MAX_CMDLEN, "%s%s%s%s", pp->function, offs, pp->retprobe ? "%return" : "", line); else - ret = e_snprintf(buf, MAX_CMDLEN, "%s%s%s%s", pp->file, line); - if (ret <= 0) - goto error; - len = ret; + ret = e_snprintf(buf, MAX_CMDLEN, "%s%s", pp->file, line); + if (ret <= 0) { +error: + free(pp->probes[0]); + pp->probes[0] = NULL; + pp->found = 0; + } + return ret; +} +int synthesize_perf_probe_event(struct probe_point *pp) +{ + char *buf; + int i, len, ret; + + len = synthesize_perf_probe_point(pp); + if (len < 0) + return 0; + + buf = pp->probes[0]; for (i = 0; i < pp->nr_args; i++) { ret = e_snprintf(&buf[len], MAX_CMDLEN - len, " %s", pp->args[i]); @@ -275,6 +360,7 @@ int synthesize_perf_probe_event(struct probe_point *pp) return pp->found; error: free(pp->probes[0]); + pp->probes[0] = NULL; return ret; } @@ -304,6 +390,7 @@ int synthesize_trace_kprobe_event(struct probe_point *pp) return pp->found; error: free(pp->probes[0]); + pp->probes[0] = NULL; return ret; } @@ -321,7 +408,7 @@ static int open_kprobe_events(int flags, int mode) if (ret < 0) { if (errno == ENOENT) die("kprobe_events file does not exist -" - " please rebuild with CONFIG_KPROBE_TRACER."); + " please rebuild with CONFIG_KPROBE_EVENT."); else die("Could not open kprobe_events file: %s", strerror(errno)); @@ -363,6 +450,10 @@ static void clear_probe_point(struct probe_point *pp) { int i; + if (pp->event) + free(pp->event); + if (pp->group) + free(pp->group); if (pp->function) free(pp->function); if (pp->file) @@ -373,30 +464,50 @@ static void clear_probe_point(struct probe_point *pp) free(pp->args); for (i = 0; i < pp->found; i++) free(pp->probes[i]); - memset(pp, 0, sizeof(pp)); + memset(pp, 0, sizeof(*pp)); +} + +/* Show an event */ +static void show_perf_probe_event(const char *event, const char *place, + struct probe_point *pp) +{ + int i, ret; + char buf[128]; + + ret = e_snprintf(buf, 128, "%s:%s", pp->group, event); + if (ret < 0) + die("Failed to copy event: %s", strerror(-ret)); + printf(" %-40s (on %s", buf, place); + + if (pp->nr_args > 0) { + printf(" with"); + for (i = 0; i < pp->nr_args; i++) + printf(" %s", pp->args[i]); + } + printf(")\n"); } /* List up current perf-probe events */ void show_perf_probe_events(void) { - unsigned int i; int fd; - char *group, *event; struct probe_point pp; struct strlist *rawlist; struct str_node *ent; + setup_pager(); + + memset(&pp, 0, sizeof(pp)); fd = open_kprobe_events(O_RDONLY, 0); rawlist = get_trace_kprobe_event_rawlist(fd); close(fd); - for (i = 0; i < strlist__nr_entries(rawlist); i++) { - ent = strlist__entry(rawlist, i); - parse_trace_kprobe_event(ent->s, &group, &event, &pp); - synthesize_perf_probe_event(&pp); - printf("[%s:%s]\t%s\n", group, event, pp.probes[0]); - free(group); - free(event); + strlist__for_each(ent, rawlist) { + parse_trace_kprobe_event(ent->s, &pp); + /* Synthesize only event probe point */ + synthesize_perf_probe_point(&pp); + /* Show an event */ + show_perf_probe_event(pp.event, pp.probes[0], &pp); clear_probe_point(&pp); } @@ -404,21 +515,27 @@ void show_perf_probe_events(void) } /* Get current perf-probe event names */ -static struct strlist *get_perf_event_names(int fd) +static struct strlist *get_perf_event_names(int fd, bool include_group) { - unsigned int i; - char *group, *event; + char buf[128]; struct strlist *sl, *rawlist; struct str_node *ent; + struct probe_point pp; + memset(&pp, 0, sizeof(pp)); rawlist = get_trace_kprobe_event_rawlist(fd); - sl = strlist__new(false, NULL); - for (i = 0; i < strlist__nr_entries(rawlist); i++) { - ent = strlist__entry(rawlist, i); - parse_trace_kprobe_event(ent->s, &group, &event, NULL); - strlist__add(sl, event); - free(group); + sl = strlist__new(true, NULL); + strlist__for_each(ent, rawlist) { + parse_trace_kprobe_event(ent->s, &pp); + if (include_group) { + if (e_snprintf(buf, 128, "%s:%s", pp.group, + pp.event) < 0) + die("Failed to copy group:event name."); + strlist__add(sl, buf); + } else + strlist__add(sl, pp.event); + clear_probe_point(&pp); } strlist__delete(rawlist); @@ -426,24 +543,36 @@ static struct strlist *get_perf_event_names(int fd) return sl; } -static int write_trace_kprobe_event(int fd, const char *buf) +static void write_trace_kprobe_event(int fd, const char *buf) { int ret; + pr_debug("Writing event: %s\n", buf); ret = write(fd, buf, strlen(buf)); if (ret <= 0) - die("Failed to create event."); - else - printf("Added new event: %s\n", buf); - - return ret; + die("Failed to write event: %s", strerror(errno)); } static void get_new_event_name(char *buf, size_t len, const char *base, - struct strlist *namelist) + struct strlist *namelist, bool allow_suffix) { int i, ret; - for (i = 0; i < MAX_EVENT_INDEX; i++) { + + /* Try no suffix */ + ret = e_snprintf(buf, len, "%s", base); + if (ret < 0) + die("snprintf() failed: %s", strerror(-ret)); + if (!strlist__has_entry(namelist, buf)) + return; + + if (!allow_suffix) { + pr_warning("Error: event \"%s\" already exists. " + "(Use -f to force duplicates.)\n", base); + die("Can't add new event."); + } + + /* Try to add suffix */ + for (i = 1; i < MAX_EVENT_INDEX; i++) { ret = e_snprintf(buf, len, "%s_%d", base, i); if (ret < 0) die("snprintf() failed: %s", strerror(-ret)); @@ -454,31 +583,201 @@ static void get_new_event_name(char *buf, size_t len, const char *base, die("Too many events are on the same function."); } -void add_trace_kprobe_events(struct probe_point *probes, int nr_probes) +void add_trace_kprobe_events(struct probe_point *probes, int nr_probes, + bool force_add) { int i, j, fd; struct probe_point *pp; char buf[MAX_CMDLEN]; char event[64]; struct strlist *namelist; + bool allow_suffix; fd = open_kprobe_events(O_RDWR, O_APPEND); /* Get current event names */ - namelist = get_perf_event_names(fd); + namelist = get_perf_event_names(fd, false); for (j = 0; j < nr_probes; j++) { pp = probes + j; + if (!pp->event) + pp->event = strdup(pp->function); + if (!pp->group) + pp->group = strdup(PERFPROBE_GROUP); + DIE_IF(!pp->event || !pp->group); + /* If force_add is true, suffix search is allowed */ + allow_suffix = force_add; for (i = 0; i < pp->found; i++) { /* Get an unused new event name */ - get_new_event_name(event, 64, pp->function, namelist); + get_new_event_name(event, 64, pp->event, namelist, + allow_suffix); snprintf(buf, MAX_CMDLEN, "%c:%s/%s %s\n", pp->retprobe ? 'r' : 'p', - PERFPROBE_GROUP, event, + pp->group, event, pp->probes[i]); write_trace_kprobe_event(fd, buf); + printf("Added new event:\n"); + /* Get the first parameter (probe-point) */ + sscanf(pp->probes[i], "%s", buf); + show_perf_probe_event(event, buf, pp); /* Add added event name to namelist */ strlist__add(namelist, event); + /* + * Probes after the first probe which comes from same + * user input are always allowed to add suffix, because + * there might be several addresses corresponding to + * one code line. + */ + allow_suffix = true; + } + } + /* Show how to use the event. */ + printf("\nYou can now use it on all perf tools, such as:\n\n"); + printf("\tperf record -e %s:%s -a sleep 1\n\n", PERFPROBE_GROUP, event); + + strlist__delete(namelist); + close(fd); +} + +static void __del_trace_kprobe_event(int fd, struct str_node *ent) +{ + char *p; + char buf[128]; + + /* Convert from perf-probe event to trace-kprobe event */ + if (e_snprintf(buf, 128, "-:%s", ent->s) < 0) + die("Failed to copy event."); + p = strchr(buf + 2, ':'); + if (!p) + die("Internal error: %s should have ':' but not.", ent->s); + *p = '/'; + + write_trace_kprobe_event(fd, buf); + printf("Remove event: %s\n", ent->s); +} + +static void del_trace_kprobe_event(int fd, const char *group, + const char *event, struct strlist *namelist) +{ + char buf[128]; + struct str_node *ent, *n; + int found = 0; + + if (e_snprintf(buf, 128, "%s:%s", group, event) < 0) + die("Failed to copy event."); + + if (strpbrk(buf, "*?")) { /* Glob-exp */ + strlist__for_each_safe(ent, n, namelist) + if (strglobmatch(ent->s, buf)) { + found++; + __del_trace_kprobe_event(fd, ent); + strlist__remove(namelist, ent); + } + } else { + ent = strlist__find(namelist, buf); + if (ent) { + found++; + __del_trace_kprobe_event(fd, ent); + strlist__remove(namelist, ent); + } + } + if (found == 0) + pr_info("Info: event \"%s\" does not exist, could not remove it.\n", buf); +} + +void del_trace_kprobe_events(struct strlist *dellist) +{ + int fd; + const char *group, *event; + char *p, *str; + struct str_node *ent; + struct strlist *namelist; + + fd = open_kprobe_events(O_RDWR, O_APPEND); + /* Get current event names */ + namelist = get_perf_event_names(fd, true); + + strlist__for_each(ent, dellist) { + str = strdup(ent->s); + if (!str) + die("Failed to copy event."); + pr_debug("Parsing: %s\n", str); + p = strchr(str, ':'); + if (p) { + group = str; + *p = '\0'; + event = p + 1; + } else { + group = "*"; + event = str; } + pr_debug("Group: %s, Event: %s\n", group, event); + del_trace_kprobe_event(fd, group, event, namelist); + free(str); } + strlist__delete(namelist); close(fd); } + +#define LINEBUF_SIZE 256 + +static void show_one_line(FILE *fp, unsigned int l, bool skip, bool show_num) +{ + char buf[LINEBUF_SIZE]; + const char *color = PERF_COLOR_BLUE; + + if (fgets(buf, LINEBUF_SIZE, fp) == NULL) + goto error; + if (!skip) { + if (show_num) + fprintf(stdout, "%7u %s", l, buf); + else + color_fprintf(stdout, color, " %s", buf); + } + + while (strlen(buf) == LINEBUF_SIZE - 1 && + buf[LINEBUF_SIZE - 2] != '\n') { + if (fgets(buf, LINEBUF_SIZE, fp) == NULL) + goto error; + if (!skip) { + if (show_num) + fprintf(stdout, "%s", buf); + else + color_fprintf(stdout, color, "%s", buf); + } + } + return; +error: + if (feof(fp)) + die("Source file is shorter than expected."); + else + die("File read error: %s", strerror(errno)); +} + +void show_line_range(struct line_range *lr) +{ + unsigned int l = 1; + struct line_node *ln; + FILE *fp; + + setup_pager(); + + if (lr->function) + fprintf(stdout, "<%s:%d>\n", lr->function, + lr->start - lr->offset); + else + fprintf(stdout, "<%s:%d>\n", lr->file, lr->start); + + fp = fopen(lr->path, "r"); + if (fp == NULL) + die("Failed to open %s: %s", lr->path, strerror(errno)); + /* Skip to starting line number */ + while (l < lr->start) + show_one_line(fp, l++, true, false); + + list_for_each_entry(ln, &lr->line_list, list) { + while (ln->line > l) + show_one_line(fp, (l++) - lr->offset, false, false); + show_one_line(fp, (l++) - lr->offset, false, true); + } + fclose(fp); +} diff --git a/tools/perf/util/probe-event.h b/tools/perf/util/probe-event.h index 0c6fe56fe38a..711287d4baea 100644 --- a/tools/perf/util/probe-event.h +++ b/tools/perf/util/probe-event.h @@ -1,16 +1,22 @@ #ifndef _PROBE_EVENT_H #define _PROBE_EVENT_H +#include <stdbool.h> #include "probe-finder.h" #include "strlist.h" -extern int parse_perf_probe_event(const char *str, struct probe_point *pp); +extern void parse_line_range_desc(const char *arg, struct line_range *lr); +extern void parse_perf_probe_event(const char *str, struct probe_point *pp, + bool *need_dwarf); +extern int synthesize_perf_probe_point(struct probe_point *pp); extern int synthesize_perf_probe_event(struct probe_point *pp); -extern void parse_trace_kprobe_event(const char *str, char **group, - char **event, struct probe_point *pp); +extern void parse_trace_kprobe_event(const char *str, struct probe_point *pp); extern int synthesize_trace_kprobe_event(struct probe_point *pp); -extern void add_trace_kprobe_events(struct probe_point *probes, int nr_probes); +extern void add_trace_kprobe_events(struct probe_point *probes, int nr_probes, + bool force_add); +extern void del_trace_kprobe_events(struct strlist *dellist); extern void show_perf_probe_events(void); +extern void show_line_range(struct line_range *lr); /* Maximum index number of event-name postfix */ #define MAX_EVENT_INDEX 1024 diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c index 293cdfc1b8ca..1b2124d12f68 100644 --- a/tools/perf/util/probe-finder.c +++ b/tools/perf/util/probe-finder.c @@ -106,7 +106,7 @@ static int strtailcmp(const char *s1, const char *s2) { int i1 = strlen(s1); int i2 = strlen(s2); - while (--i1 > 0 && --i2 > 0) { + while (--i1 >= 0 && --i2 >= 0) { if (s1[i1] != s2[i2]) return s1[i1] - s2[i2]; } @@ -140,6 +140,31 @@ static Dwarf_Unsigned cu_find_fileno(Dwarf_Die cu_die, const char *fname) return found; } +static int cu_get_filename(Dwarf_Die cu_die, Dwarf_Unsigned fno, char **buf) +{ + Dwarf_Signed cnt, i; + char **srcs; + int ret = 0; + + if (!buf || !fno) + return -EINVAL; + + ret = dwarf_srcfiles(cu_die, &srcs, &cnt, &__dw_error); + if (ret == DW_DLV_OK) { + if ((Dwarf_Unsigned)cnt > fno - 1) { + *buf = strdup(srcs[fno - 1]); + ret = 0; + pr_debug("found filename: %s\n", *buf); + } else + ret = -ENOENT; + for (i = 0; i < cnt; i++) + dwarf_dealloc(__dw_debug, srcs[i], DW_DLA_STRING); + dwarf_dealloc(__dw_debug, srcs, DW_DLA_LIST); + } else + ret = -EINVAL; + return ret; +} + /* Compare diename and tname */ static int die_compare_name(Dwarf_Die dw_die, const char *tname) { @@ -402,11 +427,11 @@ static void show_location(Dwarf_Loc *loc, struct probe_finder *pf) } else if (op == DW_OP_regx) { regn = loc->lr_number; } else - die("Dwarf_OP %d is not supported.\n", op); + die("Dwarf_OP %d is not supported.", op); regs = get_arch_regstr(regn); if (!regs) - die("%lld exceeds max register number.\n", regn); + die("%lld exceeds max register number.", regn); if (deref) ret = snprintf(pf->buf, pf->len, @@ -438,7 +463,7 @@ static void show_variable(Dwarf_Die vr_die, struct probe_finder *pf) return ; error: die("Failed to find the location of %s at this address.\n" - " Perhaps, it has been optimized out.\n", pf->var); + " Perhaps, it has been optimized out.", pf->var); } static int variable_callback(struct die_link *dlink, void *data) @@ -476,7 +501,7 @@ static void find_variable(Dwarf_Die sp_die, struct probe_finder *pf) /* Search child die for local variables and parameters. */ ret = search_die_from_children(sp_die, variable_callback, pf); if (!ret) - die("Failed to find '%s' in this function.\n", pf->var); + die("Failed to find '%s' in this function.", pf->var); } /* Get a frame base on the address */ @@ -567,7 +592,7 @@ static int probeaddr_callback(struct die_link *dlink, void *data) } /* Find probe point from its line number */ -static void find_by_line(struct probe_finder *pf) +static void find_probe_point_by_line(struct probe_finder *pf) { Dwarf_Signed cnt, i, clm; Dwarf_Line *lines; @@ -602,7 +627,7 @@ static void find_by_line(struct probe_finder *pf) ret = search_die_from_children(pf->cu_die, probeaddr_callback, pf); if (ret == 0) - die("Probe point is not found in subprograms.\n"); + die("Probe point is not found in subprograms."); /* Continuing, because target line might be inlined. */ } dwarf_srclines_dealloc(__dw_debug, lines, cnt); @@ -626,7 +651,7 @@ static int probefunc_callback(struct die_link *dlink, void *data) pf->fno = die_get_decl_file(dlink->die); pf->lno = die_get_decl_line(dlink->die) + pp->line; - find_by_line(pf); + find_probe_point_by_line(pf); return 1; } if (die_inlined_subprogram(dlink->die)) { @@ -661,7 +686,7 @@ static int probefunc_callback(struct die_link *dlink, void *data) !die_inlined_subprogram(lk->die)) goto found; } - die("Failed to find real subprogram.\n"); + die("Failed to find real subprogram."); found: /* Get offset from subprogram */ ret = die_within_subprogram(lk->die, pf->addr, &offs); @@ -673,7 +698,7 @@ found: return 0; } -static void find_by_func(struct probe_finder *pf) +static void find_probe_point_by_func(struct probe_finder *pf) { search_die_from_children(pf->cu_die, probefunc_callback, pf); } @@ -687,10 +712,8 @@ int find_probepoint(int fd, struct probe_point *pp) struct probe_finder pf = {.pp = pp}; ret = dwarf_init(fd, DW_DLC_READ, 0, 0, &__dw_debug, &__dw_error); - if (ret != DW_DLV_OK) { - pr_warning("No dwarf info found in the vmlinux - please rebuild with CONFIG_DEBUG_INFO.\n"); + if (ret != DW_DLV_OK) return -ENOENT; - } pp->found = 0; while (++cu_number) { @@ -716,10 +739,10 @@ int find_probepoint(int fd, struct probe_point *pp) if (ret == DW_DLV_NO_ENTRY) pf.cu_base = 0; if (pp->function) - find_by_func(&pf); + find_probe_point_by_func(&pf); else { pf.lno = pp->line; - find_by_line(&pf); + find_probe_point_by_line(&pf); } } dwarf_dealloc(__dw_debug, pf.cu_die, DW_DLA_DIE); @@ -730,3 +753,159 @@ int find_probepoint(int fd, struct probe_point *pp) return pp->found; } + +static void line_range_add_line(struct line_range *lr, unsigned int line) +{ + struct line_node *ln; + struct list_head *p; + + /* Reverse search, because new line will be the last one */ + list_for_each_entry_reverse(ln, &lr->line_list, list) { + if (ln->line < line) { + p = &ln->list; + goto found; + } else if (ln->line == line) /* Already exist */ + return ; + } + /* List is empty, or the smallest entry */ + p = &lr->line_list; +found: + pr_debug("Debug: add a line %u\n", line); + ln = zalloc(sizeof(struct line_node)); + DIE_IF(ln == NULL); + ln->line = line; + INIT_LIST_HEAD(&ln->list); + list_add(&ln->list, p); +} + +/* Find line range from its line number */ +static void find_line_range_by_line(struct line_finder *lf) +{ + Dwarf_Signed cnt, i; + Dwarf_Line *lines; + Dwarf_Unsigned lineno = 0; + Dwarf_Unsigned fno; + Dwarf_Addr addr; + int ret; + + ret = dwarf_srclines(lf->cu_die, &lines, &cnt, &__dw_error); + DIE_IF(ret != DW_DLV_OK); + + for (i = 0; i < cnt; i++) { + ret = dwarf_line_srcfileno(lines[i], &fno, &__dw_error); + DIE_IF(ret != DW_DLV_OK); + if (fno != lf->fno) + continue; + + ret = dwarf_lineno(lines[i], &lineno, &__dw_error); + DIE_IF(ret != DW_DLV_OK); + if (lf->lno_s > lineno || lf->lno_e < lineno) + continue; + + /* Filter line in the function address range */ + if (lf->addr_s && lf->addr_e) { + ret = dwarf_lineaddr(lines[i], &addr, &__dw_error); + DIE_IF(ret != DW_DLV_OK); + if (lf->addr_s > addr || lf->addr_e <= addr) + continue; + } + line_range_add_line(lf->lr, (unsigned int)lineno); + } + dwarf_srclines_dealloc(__dw_debug, lines, cnt); + if (!list_empty(&lf->lr->line_list)) + lf->found = 1; +} + +/* Search function from function name */ +static int linefunc_callback(struct die_link *dlink, void *data) +{ + struct line_finder *lf = (struct line_finder *)data; + struct line_range *lr = lf->lr; + Dwarf_Half tag; + int ret; + + ret = dwarf_tag(dlink->die, &tag, &__dw_error); + DIE_IF(ret == DW_DLV_ERROR); + if (tag == DW_TAG_subprogram && + die_compare_name(dlink->die, lr->function) == 0) { + /* Get the address range of this function */ + ret = dwarf_highpc(dlink->die, &lf->addr_e, &__dw_error); + if (ret == DW_DLV_OK) + ret = dwarf_lowpc(dlink->die, &lf->addr_s, &__dw_error); + DIE_IF(ret == DW_DLV_ERROR); + if (ret == DW_DLV_NO_ENTRY) { + lf->addr_s = 0; + lf->addr_e = 0; + } + + lf->fno = die_get_decl_file(dlink->die); + lr->offset = die_get_decl_line(dlink->die);; + lf->lno_s = lr->offset + lr->start; + if (!lr->end) + lf->lno_e = (Dwarf_Unsigned)-1; + else + lf->lno_e = lr->offset + lr->end; + lr->start = lf->lno_s; + lr->end = lf->lno_e; + find_line_range_by_line(lf); + /* If we find a target function, this should be end. */ + lf->found = 1; + return 1; + } + return 0; +} + +static void find_line_range_by_func(struct line_finder *lf) +{ + search_die_from_children(lf->cu_die, linefunc_callback, lf); +} + +int find_line_range(int fd, struct line_range *lr) +{ + Dwarf_Half addr_size = 0; + Dwarf_Unsigned next_cuh = 0; + int ret; + struct line_finder lf = {.lr = lr}; + + ret = dwarf_init(fd, DW_DLC_READ, 0, 0, &__dw_debug, &__dw_error); + if (ret != DW_DLV_OK) + return -ENOENT; + + while (!lf.found) { + /* Search CU (Compilation Unit) */ + ret = dwarf_next_cu_header(__dw_debug, NULL, NULL, NULL, + &addr_size, &next_cuh, &__dw_error); + DIE_IF(ret == DW_DLV_ERROR); + if (ret == DW_DLV_NO_ENTRY) + break; + + /* Get the DIE(Debugging Information Entry) of this CU */ + ret = dwarf_siblingof(__dw_debug, 0, &lf.cu_die, &__dw_error); + DIE_IF(ret != DW_DLV_OK); + + /* Check if target file is included. */ + if (lr->file) + lf.fno = cu_find_fileno(lf.cu_die, lr->file); + + if (!lr->file || lf.fno) { + if (lr->function) + find_line_range_by_func(&lf); + else { + lf.lno_s = lr->start; + if (!lr->end) + lf.lno_e = (Dwarf_Unsigned)-1; + else + lf.lno_e = lr->end; + find_line_range_by_line(&lf); + } + /* Get the real file path */ + if (lf.found) + cu_get_filename(lf.cu_die, lf.fno, &lr->path); + } + dwarf_dealloc(__dw_debug, lf.cu_die, DW_DLA_DIE); + } + ret = dwarf_finish(__dw_debug, &__dw_error); + DIE_IF(ret != DW_DLV_OK); + return lf.found; +} + diff --git a/tools/perf/util/probe-finder.h b/tools/perf/util/probe-finder.h index bdebca6697d2..972b386116f1 100644 --- a/tools/perf/util/probe-finder.h +++ b/tools/perf/util/probe-finder.h @@ -1,9 +1,11 @@ #ifndef _PROBE_FINDER_H #define _PROBE_FINDER_H -#define MAX_PATH_LEN 256 -#define MAX_PROBE_BUFFER 1024 -#define MAX_PROBES 128 +#include "util.h" + +#define MAX_PATH_LEN 256 +#define MAX_PROBE_BUFFER 1024 +#define MAX_PROBES 128 static inline int is_c_varname(const char *name) { @@ -12,46 +14,85 @@ static inline int is_c_varname(const char *name) } struct probe_point { + char *event; /* Event name */ + char *group; /* Event group */ + /* Inputs */ - char *file; /* File name */ - int line; /* Line number */ + char *file; /* File name */ + int line; /* Line number */ - char *function; /* Function name */ - int offset; /* Offset bytes */ + char *function; /* Function name */ + int offset; /* Offset bytes */ - int nr_args; /* Number of arguments */ - char **args; /* Arguments */ + int nr_args; /* Number of arguments */ + char **args; /* Arguments */ - int retprobe; /* Return probe */ + int retprobe; /* Return probe */ /* Output */ - int found; /* Number of found probe points */ - char *probes[MAX_PROBES]; /* Output buffers (will be allocated)*/ + int found; /* Number of found probe points */ + char *probes[MAX_PROBES]; /* Output buffers (will be allocated)*/ +}; + +/* Line number container */ +struct line_node { + struct list_head list; + unsigned int line; +}; + +/* Line range */ +struct line_range { + char *file; /* File name */ + char *function; /* Function name */ + unsigned int start; /* Start line number */ + unsigned int end; /* End line number */ + unsigned int offset; /* Start line offset */ + char *path; /* Real path name */ + struct list_head line_list; /* Visible lines */ }; #ifndef NO_LIBDWARF extern int find_probepoint(int fd, struct probe_point *pp); +extern int find_line_range(int fd, struct line_range *lr); -#include <libdwarf/dwarf.h> -#include <libdwarf/libdwarf.h> +/* Workaround for undefined _MIPS_SZLONG bug in libdwarf.h: */ +#ifndef _MIPS_SZLONG +# define _MIPS_SZLONG 0 +#endif + +#include <dwarf.h> +#include <libdwarf.h> struct probe_finder { - struct probe_point *pp; /* Target probe point */ + struct probe_point *pp; /* Target probe point */ /* For function searching */ - Dwarf_Addr addr; /* Address */ - Dwarf_Unsigned fno; /* File number */ - Dwarf_Unsigned lno; /* Line number */ - Dwarf_Off inl_offs; /* Inline offset */ - Dwarf_Die cu_die; /* Current CU */ + Dwarf_Addr addr; /* Address */ + Dwarf_Unsigned fno; /* File number */ + Dwarf_Unsigned lno; /* Line number */ + Dwarf_Off inl_offs; /* Inline offset */ + Dwarf_Die cu_die; /* Current CU */ /* For variable searching */ - Dwarf_Addr cu_base; /* Current CU base address */ - Dwarf_Locdesc fbloc; /* Location of Current Frame Base */ - const char *var; /* Current variable name */ - char *buf; /* Current output buffer */ - int len; /* Length of output buffer */ + Dwarf_Addr cu_base; /* Current CU base address */ + Dwarf_Locdesc fbloc; /* Location of Current Frame Base */ + const char *var; /* Current variable name */ + char *buf; /* Current output buffer */ + int len; /* Length of output buffer */ }; + +struct line_finder { + struct line_range *lr; /* Target line range */ + + Dwarf_Unsigned fno; /* File number */ + Dwarf_Unsigned lno_s; /* Start line number */ + Dwarf_Unsigned lno_e; /* End line number */ + Dwarf_Addr addr_s; /* Start address */ + Dwarf_Addr addr_e; /* End address */ + Dwarf_Die cu_die; /* Current CU */ + int found; +}; + #endif /* NO_LIBDWARF */ #endif /*_PROBE_FINDER_H */ diff --git a/tools/perf/util/trace-event-perl.c b/tools/perf/util/scripting-engines/trace-event-perl.c index 51e833fd58c3..5376378e0cfc 100644 --- a/tools/perf/util/trace-event-perl.c +++ b/tools/perf/util/scripting-engines/trace-event-perl.c @@ -25,15 +25,18 @@ #include <ctype.h> #include <errno.h> -#include "../perf.h" -#include "util.h" -#include "trace-event.h" -#include "trace-event-perl.h" +#include "../../perf.h" +#include "../util.h" +#include "../trace-event.h" -void xs_init(pTHX); +#include <EXTERN.h> +#include <perl.h> void boot_Perf__Trace__Context(pTHX_ CV *cv); void boot_DynaLoader(pTHX_ CV *cv); +typedef PerlInterpreter * INTERP; + +void xs_init(pTHX); void xs_init(pTHX) { @@ -52,7 +55,7 @@ INTERP my_perl; struct event *events[FTRACE_MAX_EVENT]; -static struct scripting_context *scripting_context; +extern struct scripting_context *scripting_context; static char *cur_field_name; static int zero_flag_atom; @@ -242,35 +245,8 @@ static inline struct event *find_cache_event(int type) return event; } -int common_pc(struct scripting_context *context) -{ - int pc; - - pc = parse_common_pc(context->event_data); - - return pc; -} - -int common_flags(struct scripting_context *context) -{ - int flags; - - flags = parse_common_flags(context->event_data); - - return flags; -} - -int common_lock_depth(struct scripting_context *context) -{ - int lock_depth; - - lock_depth = parse_common_lock_depth(context->event_data); - - return lock_depth; -} - static void perl_process_event(int cpu, void *data, - int size __attribute((unused)), + int size __unused, unsigned long long nsecs, char *comm) { struct format_field *field; @@ -362,28 +338,46 @@ static void run_start_sub(void) /* * Start trace script */ -static int perl_start_script(const char *script) +static int perl_start_script(const char *script, int argc, const char **argv) { - const char *command_line[2] = { "", NULL }; + const char **command_line; + int i, err = 0; + command_line = malloc((argc + 2) * sizeof(const char *)); + command_line[0] = ""; command_line[1] = script; + for (i = 2; i < argc + 2; i++) + command_line[i] = argv[i - 2]; my_perl = perl_alloc(); perl_construct(my_perl); - if (perl_parse(my_perl, xs_init, 2, (char **)command_line, - (char **)NULL)) - return -1; + if (perl_parse(my_perl, xs_init, argc + 2, (char **)command_line, + (char **)NULL)) { + err = -1; + goto error; + } - perl_run(my_perl); - if (SvTRUE(ERRSV)) - return -1; + if (perl_run(my_perl)) { + err = -1; + goto error; + } + + if (SvTRUE(ERRSV)) { + err = -1; + goto error; + } run_start_sub(); + free(command_line); fprintf(stderr, "perf trace started with Perl script %s\n\n", script); - return 0; +error: + perl_free(my_perl); + free(command_line); + + return err; } /* @@ -572,27 +566,3 @@ struct scripting_ops perl_scripting_ops = { .process_event = perl_process_event, .generate_script = perl_generate_script, }; - -#ifdef NO_LIBPERL -void setup_perl_scripting(void) -{ - fprintf(stderr, "Perl scripting not supported." - " Install libperl and rebuild perf to enable it. e.g. " - "apt-get install libperl-dev (ubuntu), yum install " - "perl-ExtUtils-Embed (Fedora), etc.\n"); -} -#else -void setup_perl_scripting(void) -{ - int err; - err = script_spec_register("Perl", &perl_scripting_ops); - if (err) - die("error registering Perl script extension"); - - err = script_spec_register("pl", &perl_scripting_ops); - if (err) - die("error registering pl script extension"); - - scripting_context = malloc(sizeof(struct scripting_context)); -} -#endif diff --git a/tools/perf/util/scripting-engines/trace-event-python.c b/tools/perf/util/scripting-engines/trace-event-python.c new file mode 100644 index 000000000000..33a414bbba3e --- /dev/null +++ b/tools/perf/util/scripting-engines/trace-event-python.c @@ -0,0 +1,573 @@ +/* + * trace-event-python. Feed trace events to an embedded Python interpreter. + * + * Copyright (C) 2010 Tom Zanussi <tzanussi@gmail.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + */ + +#include <Python.h> + +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <ctype.h> +#include <errno.h> + +#include "../../perf.h" +#include "../util.h" +#include "../trace-event.h" + +PyMODINIT_FUNC initperf_trace_context(void); + +#define FTRACE_MAX_EVENT \ + ((1 << (sizeof(unsigned short) * 8)) - 1) + +struct event *events[FTRACE_MAX_EVENT]; + +#define MAX_FIELDS 64 +#define N_COMMON_FIELDS 7 + +extern struct scripting_context *scripting_context; + +static char *cur_field_name; +static int zero_flag_atom; + +static PyObject *main_module, *main_dict; + +static void handler_call_die(const char *handler_name) +{ + PyErr_Print(); + Py_FatalError("problem in Python trace event handler"); +} + +static void define_value(enum print_arg_type field_type, + const char *ev_name, + const char *field_name, + const char *field_value, + const char *field_str) +{ + const char *handler_name = "define_flag_value"; + PyObject *handler, *t, *retval; + unsigned long long value; + unsigned n = 0; + + if (field_type == PRINT_SYMBOL) + handler_name = "define_symbolic_value"; + + t = PyTuple_New(4); + if (!t) + Py_FatalError("couldn't create Python tuple"); + + value = eval_flag(field_value); + + PyTuple_SetItem(t, n++, PyString_FromString(ev_name)); + PyTuple_SetItem(t, n++, PyString_FromString(field_name)); + PyTuple_SetItem(t, n++, PyInt_FromLong(value)); + PyTuple_SetItem(t, n++, PyString_FromString(field_str)); + + handler = PyDict_GetItemString(main_dict, handler_name); + if (handler && PyCallable_Check(handler)) { + retval = PyObject_CallObject(handler, t); + if (retval == NULL) + handler_call_die(handler_name); + } + + Py_DECREF(t); +} + +static void define_values(enum print_arg_type field_type, + struct print_flag_sym *field, + const char *ev_name, + const char *field_name) +{ + define_value(field_type, ev_name, field_name, field->value, + field->str); + + if (field->next) + define_values(field_type, field->next, ev_name, field_name); +} + +static void define_field(enum print_arg_type field_type, + const char *ev_name, + const char *field_name, + const char *delim) +{ + const char *handler_name = "define_flag_field"; + PyObject *handler, *t, *retval; + unsigned n = 0; + + if (field_type == PRINT_SYMBOL) + handler_name = "define_symbolic_field"; + + if (field_type == PRINT_FLAGS) + t = PyTuple_New(3); + else + t = PyTuple_New(2); + if (!t) + Py_FatalError("couldn't create Python tuple"); + + PyTuple_SetItem(t, n++, PyString_FromString(ev_name)); + PyTuple_SetItem(t, n++, PyString_FromString(field_name)); + if (field_type == PRINT_FLAGS) + PyTuple_SetItem(t, n++, PyString_FromString(delim)); + + handler = PyDict_GetItemString(main_dict, handler_name); + if (handler && PyCallable_Check(handler)) { + retval = PyObject_CallObject(handler, t); + if (retval == NULL) + handler_call_die(handler_name); + } + + Py_DECREF(t); +} + +static void define_event_symbols(struct event *event, + const char *ev_name, + struct print_arg *args) +{ + switch (args->type) { + case PRINT_NULL: + break; + case PRINT_ATOM: + define_value(PRINT_FLAGS, ev_name, cur_field_name, "0", + args->atom.atom); + zero_flag_atom = 0; + break; + case PRINT_FIELD: + if (cur_field_name) + free(cur_field_name); + cur_field_name = strdup(args->field.name); + break; + case PRINT_FLAGS: + define_event_symbols(event, ev_name, args->flags.field); + define_field(PRINT_FLAGS, ev_name, cur_field_name, + args->flags.delim); + define_values(PRINT_FLAGS, args->flags.flags, ev_name, + cur_field_name); + break; + case PRINT_SYMBOL: + define_event_symbols(event, ev_name, args->symbol.field); + define_field(PRINT_SYMBOL, ev_name, cur_field_name, NULL); + define_values(PRINT_SYMBOL, args->symbol.symbols, ev_name, + cur_field_name); + break; + case PRINT_STRING: + break; + case PRINT_TYPE: + define_event_symbols(event, ev_name, args->typecast.item); + break; + case PRINT_OP: + if (strcmp(args->op.op, ":") == 0) + zero_flag_atom = 1; + define_event_symbols(event, ev_name, args->op.left); + define_event_symbols(event, ev_name, args->op.right); + break; + default: + /* we should warn... */ + return; + } + + if (args->next) + define_event_symbols(event, ev_name, args->next); +} + +static inline struct event *find_cache_event(int type) +{ + static char ev_name[256]; + struct event *event; + + if (events[type]) + return events[type]; + + events[type] = event = trace_find_event(type); + if (!event) + return NULL; + + sprintf(ev_name, "%s__%s", event->system, event->name); + + define_event_symbols(event, ev_name, event->print_fmt.args); + + return event; +} + +static void python_process_event(int cpu, void *data, + int size __unused, + unsigned long long nsecs, char *comm) +{ + PyObject *handler, *retval, *context, *t; + static char handler_name[256]; + struct format_field *field; + unsigned long long val; + unsigned long s, ns; + struct event *event; + unsigned n = 0; + int type; + int pid; + + t = PyTuple_New(MAX_FIELDS); + if (!t) + Py_FatalError("couldn't create Python tuple"); + + type = trace_parse_common_type(data); + + event = find_cache_event(type); + if (!event) + die("ug! no event found for type %d", type); + + pid = trace_parse_common_pid(data); + + sprintf(handler_name, "%s__%s", event->system, event->name); + + s = nsecs / NSECS_PER_SEC; + ns = nsecs - s * NSECS_PER_SEC; + + scripting_context->event_data = data; + + context = PyCObject_FromVoidPtr(scripting_context, NULL); + + PyTuple_SetItem(t, n++, PyString_FromString(handler_name)); + PyTuple_SetItem(t, n++, + PyCObject_FromVoidPtr(scripting_context, NULL)); + PyTuple_SetItem(t, n++, PyInt_FromLong(cpu)); + PyTuple_SetItem(t, n++, PyInt_FromLong(s)); + PyTuple_SetItem(t, n++, PyInt_FromLong(ns)); + PyTuple_SetItem(t, n++, PyInt_FromLong(pid)); + PyTuple_SetItem(t, n++, PyString_FromString(comm)); + + for (field = event->format.fields; field; field = field->next) { + if (field->flags & FIELD_IS_STRING) { + int offset; + if (field->flags & FIELD_IS_DYNAMIC) { + offset = *(int *)(data + field->offset); + offset &= 0xffff; + } else + offset = field->offset; + PyTuple_SetItem(t, n++, + PyString_FromString((char *)data + offset)); + } else { /* FIELD_IS_NUMERIC */ + val = read_size(data + field->offset, field->size); + if (field->flags & FIELD_IS_SIGNED) { + PyTuple_SetItem(t, n++, PyInt_FromLong(val)); + } else { + PyTuple_SetItem(t, n++, PyInt_FromLong(val)); + } + } + } + + if (_PyTuple_Resize(&t, n) == -1) + Py_FatalError("error resizing Python tuple"); + + handler = PyDict_GetItemString(main_dict, handler_name); + if (handler && PyCallable_Check(handler)) { + retval = PyObject_CallObject(handler, t); + if (retval == NULL) + handler_call_die(handler_name); + } else { + handler = PyDict_GetItemString(main_dict, "trace_unhandled"); + if (handler && PyCallable_Check(handler)) { + if (_PyTuple_Resize(&t, N_COMMON_FIELDS) == -1) + Py_FatalError("error resizing Python tuple"); + + retval = PyObject_CallObject(handler, t); + if (retval == NULL) + handler_call_die("trace_unhandled"); + } + } + + Py_DECREF(t); +} + +static int run_start_sub(void) +{ + PyObject *handler, *retval; + int err = 0; + + main_module = PyImport_AddModule("__main__"); + if (main_module == NULL) + return -1; + Py_INCREF(main_module); + + main_dict = PyModule_GetDict(main_module); + if (main_dict == NULL) { + err = -1; + goto error; + } + Py_INCREF(main_dict); + + handler = PyDict_GetItemString(main_dict, "trace_begin"); + if (handler == NULL || !PyCallable_Check(handler)) + goto out; + + retval = PyObject_CallObject(handler, NULL); + if (retval == NULL) + handler_call_die("trace_begin"); + + Py_DECREF(retval); + return err; +error: + Py_XDECREF(main_dict); + Py_XDECREF(main_module); +out: + return err; +} + +/* + * Start trace script + */ +static int python_start_script(const char *script, int argc, const char **argv) +{ + const char **command_line; + char buf[PATH_MAX]; + int i, err = 0; + FILE *fp; + + command_line = malloc((argc + 1) * sizeof(const char *)); + command_line[0] = script; + for (i = 1; i < argc + 1; i++) + command_line[i] = argv[i - 1]; + + Py_Initialize(); + + initperf_trace_context(); + + PySys_SetArgv(argc + 1, (char **)command_line); + + fp = fopen(script, "r"); + if (!fp) { + sprintf(buf, "Can't open python script \"%s\"", script); + perror(buf); + err = -1; + goto error; + } + + err = PyRun_SimpleFile(fp, script); + if (err) { + fprintf(stderr, "Error running python script %s\n", script); + goto error; + } + + err = run_start_sub(); + if (err) { + fprintf(stderr, "Error starting python script %s\n", script); + goto error; + } + + free(command_line); + fprintf(stderr, "perf trace started with Python script %s\n\n", + script); + + return err; +error: + Py_Finalize(); + free(command_line); + + return err; +} + +/* + * Stop trace script + */ +static int python_stop_script(void) +{ + PyObject *handler, *retval; + int err = 0; + + handler = PyDict_GetItemString(main_dict, "trace_end"); + if (handler == NULL || !PyCallable_Check(handler)) + goto out; + + retval = PyObject_CallObject(handler, NULL); + if (retval == NULL) + handler_call_die("trace_end"); + else + Py_DECREF(retval); +out: + Py_XDECREF(main_dict); + Py_XDECREF(main_module); + Py_Finalize(); + + fprintf(stderr, "\nperf trace Python script stopped\n"); + + return err; +} + +static int python_generate_script(const char *outfile) +{ + struct event *event = NULL; + struct format_field *f; + char fname[PATH_MAX]; + int not_first, count; + FILE *ofp; + + sprintf(fname, "%s.py", outfile); + ofp = fopen(fname, "w"); + if (ofp == NULL) { + fprintf(stderr, "couldn't open %s\n", fname); + return -1; + } + fprintf(ofp, "# perf trace event handlers, " + "generated by perf trace -g python\n"); + + fprintf(ofp, "# Licensed under the terms of the GNU GPL" + " License version 2\n\n"); + + fprintf(ofp, "# The common_* event handler fields are the most useful " + "fields common to\n"); + + fprintf(ofp, "# all events. They don't necessarily correspond to " + "the 'common_*' fields\n"); + + fprintf(ofp, "# in the format files. Those fields not available as " + "handler params can\n"); + + fprintf(ofp, "# be retrieved using Python functions of the form " + "common_*(context).\n"); + + fprintf(ofp, "# See the perf-trace-python Documentation for the list " + "of available functions.\n\n"); + + fprintf(ofp, "import os\n"); + fprintf(ofp, "import sys\n\n"); + + fprintf(ofp, "sys.path.append(os.environ['PERF_EXEC_PATH'] + \\\n"); + fprintf(ofp, "\t'/scripts/python/Perf-Trace-Util/lib/Perf/Trace')\n"); + fprintf(ofp, "\nfrom perf_trace_context import *\n"); + fprintf(ofp, "from Core import *\n\n\n"); + + fprintf(ofp, "def trace_begin():\n"); + fprintf(ofp, "\tprint \"in trace_begin\"\n\n"); + + fprintf(ofp, "def trace_end():\n"); + fprintf(ofp, "\tprint \"in trace_end\"\n\n"); + + while ((event = trace_find_next_event(event))) { + fprintf(ofp, "def %s__%s(", event->system, event->name); + fprintf(ofp, "event_name, "); + fprintf(ofp, "context, "); + fprintf(ofp, "common_cpu,\n"); + fprintf(ofp, "\tcommon_secs, "); + fprintf(ofp, "common_nsecs, "); + fprintf(ofp, "common_pid, "); + fprintf(ofp, "common_comm,\n\t"); + + not_first = 0; + count = 0; + + for (f = event->format.fields; f; f = f->next) { + if (not_first++) + fprintf(ofp, ", "); + if (++count % 5 == 0) + fprintf(ofp, "\n\t"); + + fprintf(ofp, "%s", f->name); + } + fprintf(ofp, "):\n"); + + fprintf(ofp, "\t\tprint_header(event_name, common_cpu, " + "common_secs, common_nsecs,\n\t\t\t" + "common_pid, common_comm)\n\n"); + + fprintf(ofp, "\t\tprint \""); + + not_first = 0; + count = 0; + + for (f = event->format.fields; f; f = f->next) { + if (not_first++) + fprintf(ofp, ", "); + if (count && count % 3 == 0) { + fprintf(ofp, "\" \\\n\t\t\""); + } + count++; + + fprintf(ofp, "%s=", f->name); + if (f->flags & FIELD_IS_STRING || + f->flags & FIELD_IS_FLAG || + f->flags & FIELD_IS_SYMBOLIC) + fprintf(ofp, "%%s"); + else if (f->flags & FIELD_IS_SIGNED) + fprintf(ofp, "%%d"); + else + fprintf(ofp, "%%u"); + } + + fprintf(ofp, "\\n\" %% \\\n\t\t("); + + not_first = 0; + count = 0; + + for (f = event->format.fields; f; f = f->next) { + if (not_first++) + fprintf(ofp, ", "); + + if (++count % 5 == 0) + fprintf(ofp, "\n\t\t"); + + if (f->flags & FIELD_IS_FLAG) { + if ((count - 1) % 5 != 0) { + fprintf(ofp, "\n\t\t"); + count = 4; + } + fprintf(ofp, "flag_str(\""); + fprintf(ofp, "%s__%s\", ", event->system, + event->name); + fprintf(ofp, "\"%s\", %s)", f->name, + f->name); + } else if (f->flags & FIELD_IS_SYMBOLIC) { + if ((count - 1) % 5 != 0) { + fprintf(ofp, "\n\t\t"); + count = 4; + } + fprintf(ofp, "symbol_str(\""); + fprintf(ofp, "%s__%s\", ", event->system, + event->name); + fprintf(ofp, "\"%s\", %s)", f->name, + f->name); + } else + fprintf(ofp, "%s", f->name); + } + + fprintf(ofp, "),\n\n"); + } + + fprintf(ofp, "def trace_unhandled(event_name, context, " + "common_cpu, common_secs, common_nsecs,\n\t\t" + "common_pid, common_comm):\n"); + + fprintf(ofp, "\t\tprint_header(event_name, common_cpu, " + "common_secs, common_nsecs,\n\t\tcommon_pid, " + "common_comm)\n\n"); + + fprintf(ofp, "def print_header(" + "event_name, cpu, secs, nsecs, pid, comm):\n" + "\tprint \"%%-20s %%5u %%05u.%%09u %%8u %%-20s \" %% \\\n\t" + "(event_name, cpu, secs, nsecs, pid, comm),\n"); + + fclose(ofp); + + fprintf(stderr, "generated Python script: %s\n", fname); + + return 0; +} + +struct scripting_ops python_scripting_ops = { + .name = "Python", + .start_script = python_start_script, + .stop_script = python_stop_script, + .process_event = python_process_event, + .generate_script = python_generate_script, +}; diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c new file mode 100644 index 000000000000..0de7258e70a5 --- /dev/null +++ b/tools/perf/util/session.c @@ -0,0 +1,573 @@ +#define _FILE_OFFSET_BITS 64 + +#include <linux/kernel.h> + +#include <byteswap.h> +#include <unistd.h> +#include <sys/types.h> + +#include "session.h" +#include "sort.h" +#include "util.h" + +static int perf_session__open(struct perf_session *self, bool force) +{ + struct stat input_stat; + + self->fd = open(self->filename, O_RDONLY); + if (self->fd < 0) { + pr_err("failed to open file: %s", self->filename); + if (!strcmp(self->filename, "perf.data")) + pr_err(" (try 'perf record' first)"); + pr_err("\n"); + return -errno; + } + + if (fstat(self->fd, &input_stat) < 0) + goto out_close; + + if (!force && input_stat.st_uid && (input_stat.st_uid != geteuid())) { + pr_err("file %s not owned by current user or root\n", + self->filename); + goto out_close; + } + + if (!input_stat.st_size) { + pr_info("zero-sized file (%s), nothing to do!\n", + self->filename); + goto out_close; + } + + if (perf_header__read(&self->header, self->fd) < 0) { + pr_err("incompatible file format"); + goto out_close; + } + + self->size = input_stat.st_size; + return 0; + +out_close: + close(self->fd); + self->fd = -1; + return -1; +} + +static inline int perf_session__create_kernel_maps(struct perf_session *self) +{ + return map_groups__create_kernel_maps(&self->kmaps, self->vmlinux_maps); +} + +struct perf_session *perf_session__new(const char *filename, int mode, bool force) +{ + size_t len = filename ? strlen(filename) + 1 : 0; + struct perf_session *self = zalloc(sizeof(*self) + len); + + if (self == NULL) + goto out; + + if (perf_header__init(&self->header) < 0) + goto out_free; + + memcpy(self->filename, filename, len); + self->threads = RB_ROOT; + self->last_match = NULL; + self->mmap_window = 32; + self->cwd = NULL; + self->cwdlen = 0; + self->unknown_events = 0; + map_groups__init(&self->kmaps); + + if (mode == O_RDONLY) { + if (perf_session__open(self, force) < 0) + goto out_delete; + } else if (mode == O_WRONLY) { + /* + * In O_RDONLY mode this will be performed when reading the + * kernel MMAP event, in event__process_mmap(). + */ + if (perf_session__create_kernel_maps(self) < 0) + goto out_delete; + } + + self->sample_type = perf_header__sample_type(&self->header); +out: + return self; +out_free: + free(self); + return NULL; +out_delete: + perf_session__delete(self); + return NULL; +} + +void perf_session__delete(struct perf_session *self) +{ + perf_header__exit(&self->header); + close(self->fd); + free(self->cwd); + free(self); +} + +static bool symbol__match_parent_regex(struct symbol *sym) +{ + if (sym->name && !regexec(&parent_regex, sym->name, 0, NULL, 0)) + return 1; + + return 0; +} + +struct symbol **perf_session__resolve_callchain(struct perf_session *self, + struct thread *thread, + struct ip_callchain *chain, + struct symbol **parent) +{ + u8 cpumode = PERF_RECORD_MISC_USER; + struct symbol **syms = NULL; + unsigned int i; + + if (symbol_conf.use_callchain) { + syms = calloc(chain->nr, sizeof(*syms)); + if (!syms) { + fprintf(stderr, "Can't allocate memory for symbols\n"); + exit(-1); + } + } + + for (i = 0; i < chain->nr; i++) { + u64 ip = chain->ips[i]; + struct addr_location al; + + if (ip >= PERF_CONTEXT_MAX) { + switch (ip) { + case PERF_CONTEXT_HV: + cpumode = PERF_RECORD_MISC_HYPERVISOR; break; + case PERF_CONTEXT_KERNEL: + cpumode = PERF_RECORD_MISC_KERNEL; break; + case PERF_CONTEXT_USER: + cpumode = PERF_RECORD_MISC_USER; break; + default: + break; + } + continue; + } + + thread__find_addr_location(thread, self, cpumode, + MAP__FUNCTION, ip, &al, NULL); + if (al.sym != NULL) { + if (sort__has_parent && !*parent && + symbol__match_parent_regex(al.sym)) + *parent = al.sym; + if (!symbol_conf.use_callchain) + break; + syms[i] = al.sym; + } + } + + return syms; +} + +static int process_event_stub(event_t *event __used, + struct perf_session *session __used) +{ + dump_printf(": unhandled!\n"); + return 0; +} + +static void perf_event_ops__fill_defaults(struct perf_event_ops *handler) +{ + if (handler->sample == NULL) + handler->sample = process_event_stub; + if (handler->mmap == NULL) + handler->mmap = process_event_stub; + if (handler->comm == NULL) + handler->comm = process_event_stub; + if (handler->fork == NULL) + handler->fork = process_event_stub; + if (handler->exit == NULL) + handler->exit = process_event_stub; + if (handler->lost == NULL) + handler->lost = process_event_stub; + if (handler->read == NULL) + handler->read = process_event_stub; + if (handler->throttle == NULL) + handler->throttle = process_event_stub; + if (handler->unthrottle == NULL) + handler->unthrottle = process_event_stub; +} + +static const char *event__name[] = { + [0] = "TOTAL", + [PERF_RECORD_MMAP] = "MMAP", + [PERF_RECORD_LOST] = "LOST", + [PERF_RECORD_COMM] = "COMM", + [PERF_RECORD_EXIT] = "EXIT", + [PERF_RECORD_THROTTLE] = "THROTTLE", + [PERF_RECORD_UNTHROTTLE] = "UNTHROTTLE", + [PERF_RECORD_FORK] = "FORK", + [PERF_RECORD_READ] = "READ", + [PERF_RECORD_SAMPLE] = "SAMPLE", +}; + +unsigned long event__total[PERF_RECORD_MAX]; + +void event__print_totals(void) +{ + int i; + for (i = 0; i < PERF_RECORD_MAX; ++i) + pr_info("%10s events: %10ld\n", + event__name[i], event__total[i]); +} + +void mem_bswap_64(void *src, int byte_size) +{ + u64 *m = src; + + while (byte_size > 0) { + *m = bswap_64(*m); + byte_size -= sizeof(u64); + ++m; + } +} + +static void event__all64_swap(event_t *self) +{ + struct perf_event_header *hdr = &self->header; + mem_bswap_64(hdr + 1, self->header.size - sizeof(*hdr)); +} + +static void event__comm_swap(event_t *self) +{ + self->comm.pid = bswap_32(self->comm.pid); + self->comm.tid = bswap_32(self->comm.tid); +} + +static void event__mmap_swap(event_t *self) +{ + self->mmap.pid = bswap_32(self->mmap.pid); + self->mmap.tid = bswap_32(self->mmap.tid); + self->mmap.start = bswap_64(self->mmap.start); + self->mmap.len = bswap_64(self->mmap.len); + self->mmap.pgoff = bswap_64(self->mmap.pgoff); +} + +static void event__task_swap(event_t *self) +{ + self->fork.pid = bswap_32(self->fork.pid); + self->fork.tid = bswap_32(self->fork.tid); + self->fork.ppid = bswap_32(self->fork.ppid); + self->fork.ptid = bswap_32(self->fork.ptid); + self->fork.time = bswap_64(self->fork.time); +} + +static void event__read_swap(event_t *self) +{ + self->read.pid = bswap_32(self->read.pid); + self->read.tid = bswap_32(self->read.tid); + self->read.value = bswap_64(self->read.value); + self->read.time_enabled = bswap_64(self->read.time_enabled); + self->read.time_running = bswap_64(self->read.time_running); + self->read.id = bswap_64(self->read.id); +} + +typedef void (*event__swap_op)(event_t *self); + +static event__swap_op event__swap_ops[] = { + [PERF_RECORD_MMAP] = event__mmap_swap, + [PERF_RECORD_COMM] = event__comm_swap, + [PERF_RECORD_FORK] = event__task_swap, + [PERF_RECORD_EXIT] = event__task_swap, + [PERF_RECORD_LOST] = event__all64_swap, + [PERF_RECORD_READ] = event__read_swap, + [PERF_RECORD_SAMPLE] = event__all64_swap, + [PERF_RECORD_MAX] = NULL, +}; + +static int perf_session__process_event(struct perf_session *self, + event_t *event, + struct perf_event_ops *ops, + u64 offset, u64 head) +{ + trace_event(event); + + if (event->header.type < PERF_RECORD_MAX) { + dump_printf("%#Lx [%#x]: PERF_RECORD_%s", + offset + head, event->header.size, + event__name[event->header.type]); + ++event__total[0]; + ++event__total[event->header.type]; + } + + if (self->header.needs_swap && event__swap_ops[event->header.type]) + event__swap_ops[event->header.type](event); + + switch (event->header.type) { + case PERF_RECORD_SAMPLE: + return ops->sample(event, self); + case PERF_RECORD_MMAP: + return ops->mmap(event, self); + case PERF_RECORD_COMM: + return ops->comm(event, self); + case PERF_RECORD_FORK: + return ops->fork(event, self); + case PERF_RECORD_EXIT: + return ops->exit(event, self); + case PERF_RECORD_LOST: + return ops->lost(event, self); + case PERF_RECORD_READ: + return ops->read(event, self); + case PERF_RECORD_THROTTLE: + return ops->throttle(event, self); + case PERF_RECORD_UNTHROTTLE: + return ops->unthrottle(event, self); + default: + self->unknown_events++; + return -1; + } +} + +void perf_event_header__bswap(struct perf_event_header *self) +{ + self->type = bswap_32(self->type); + self->misc = bswap_16(self->misc); + self->size = bswap_16(self->size); +} + +int perf_header__read_build_ids(struct perf_header *self, + int input, u64 offset, u64 size) +{ + struct build_id_event bev; + char filename[PATH_MAX]; + u64 limit = offset + size; + int err = -1; + + while (offset < limit) { + struct dso *dso; + ssize_t len; + struct list_head *head = &dsos__user; + + if (read(input, &bev, sizeof(bev)) != sizeof(bev)) + goto out; + + if (self->needs_swap) + perf_event_header__bswap(&bev.header); + + len = bev.header.size - sizeof(bev); + if (read(input, filename, len) != len) + goto out; + + if (bev.header.misc & PERF_RECORD_MISC_KERNEL) + head = &dsos__kernel; + + dso = __dsos__findnew(head, filename); + if (dso != NULL) { + dso__set_build_id(dso, &bev.build_id); + if (head == &dsos__kernel && filename[0] == '[') + dso->kernel = 1; + } + + offset += bev.header.size; + } + err = 0; +out: + return err; +} + +static struct thread *perf_session__register_idle_thread(struct perf_session *self) +{ + struct thread *thread = perf_session__findnew(self, 0); + + if (thread == NULL || thread__set_comm(thread, "swapper")) { + pr_err("problem inserting idle task.\n"); + thread = NULL; + } + + return thread; +} + +int __perf_session__process_events(struct perf_session *self, + u64 data_offset, u64 data_size, + u64 file_size, struct perf_event_ops *ops) +{ + int err, mmap_prot, mmap_flags; + u64 head, shift; + u64 offset = 0; + size_t page_size; + event_t *event; + uint32_t size; + char *buf; + + perf_event_ops__fill_defaults(ops); + + page_size = sysconf(_SC_PAGESIZE); + + head = data_offset; + shift = page_size * (head / page_size); + offset += shift; + head -= shift; + + mmap_prot = PROT_READ; + mmap_flags = MAP_SHARED; + + if (self->header.needs_swap) { + mmap_prot |= PROT_WRITE; + mmap_flags = MAP_PRIVATE; + } +remap: + buf = mmap(NULL, page_size * self->mmap_window, mmap_prot, + mmap_flags, self->fd, offset); + if (buf == MAP_FAILED) { + pr_err("failed to mmap file\n"); + err = -errno; + goto out_err; + } + +more: + event = (event_t *)(buf + head); + + if (self->header.needs_swap) + perf_event_header__bswap(&event->header); + size = event->header.size; + if (size == 0) + size = 8; + + if (head + event->header.size >= page_size * self->mmap_window) { + int munmap_ret; + + shift = page_size * (head / page_size); + + munmap_ret = munmap(buf, page_size * self->mmap_window); + assert(munmap_ret == 0); + + offset += shift; + head -= shift; + goto remap; + } + + size = event->header.size; + + dump_printf("\n%#Lx [%#x]: event: %d\n", + offset + head, event->header.size, event->header.type); + + if (size == 0 || + perf_session__process_event(self, event, ops, offset, head) < 0) { + dump_printf("%#Lx [%#x]: skipping unknown header type: %d\n", + offset + head, event->header.size, + event->header.type); + /* + * assume we lost track of the stream, check alignment, and + * increment a single u64 in the hope to catch on again 'soon'. + */ + if (unlikely(head & 7)) + head &= ~7ULL; + + size = 8; + } + + head += size; + + if (offset + head >= data_offset + data_size) + goto done; + + if (offset + head < file_size) + goto more; +done: + err = 0; +out_err: + return err; +} + +int perf_session__process_events(struct perf_session *self, + struct perf_event_ops *ops) +{ + int err; + + if (perf_session__register_idle_thread(self) == NULL) + return -ENOMEM; + + if (!symbol_conf.full_paths) { + char bf[PATH_MAX]; + + if (getcwd(bf, sizeof(bf)) == NULL) { + err = -errno; +out_getcwd_err: + pr_err("failed to get the current directory\n"); + goto out_err; + } + self->cwd = strdup(bf); + if (self->cwd == NULL) { + err = -ENOMEM; + goto out_getcwd_err; + } + self->cwdlen = strlen(self->cwd); + } + + err = __perf_session__process_events(self, self->header.data_offset, + self->header.data_size, + self->size, ops); +out_err: + return err; +} + +bool perf_session__has_traces(struct perf_session *self, const char *msg) +{ + if (!(self->sample_type & PERF_SAMPLE_RAW)) { + pr_err("No trace sample to read. Did you call 'perf %s'?\n", msg); + return false; + } + + return true; +} + +int perf_session__set_kallsyms_ref_reloc_sym(struct perf_session *self, + const char *symbol_name, + u64 addr) +{ + char *bracket; + enum map_type i; + + self->ref_reloc_sym.name = strdup(symbol_name); + if (self->ref_reloc_sym.name == NULL) + return -ENOMEM; + + bracket = strchr(self->ref_reloc_sym.name, ']'); + if (bracket) + *bracket = '\0'; + + self->ref_reloc_sym.addr = addr; + + for (i = 0; i < MAP__NR_TYPES; ++i) { + struct kmap *kmap = map__kmap(self->vmlinux_maps[i]); + kmap->ref_reloc_sym = &self->ref_reloc_sym; + } + + return 0; +} + +static u64 map__reloc_map_ip(struct map *map, u64 ip) +{ + return ip + (s64)map->pgoff; +} + +static u64 map__reloc_unmap_ip(struct map *map, u64 ip) +{ + return ip - (s64)map->pgoff; +} + +void map__reloc_vmlinux(struct map *self) +{ + struct kmap *kmap = map__kmap(self); + s64 reloc; + + if (!kmap->ref_reloc_sym || !kmap->ref_reloc_sym->unrelocated_addr) + return; + + reloc = (kmap->ref_reloc_sym->unrelocated_addr - + kmap->ref_reloc_sym->addr); + + if (!reloc) + return; + + self->map_ip = map__reloc_map_ip; + self->unmap_ip = map__reloc_unmap_ip; + self->pgoff = reloc; +} diff --git a/tools/perf/util/session.h b/tools/perf/util/session.h new file mode 100644 index 000000000000..31950fcd8a4d --- /dev/null +++ b/tools/perf/util/session.h @@ -0,0 +1,88 @@ +#ifndef __PERF_SESSION_H +#define __PERF_SESSION_H + +#include "event.h" +#include "header.h" +#include "symbol.h" +#include "thread.h" +#include <linux/rbtree.h> +#include "../../../include/linux/perf_event.h" + +struct ip_callchain; +struct thread; + +struct perf_session { + struct perf_header header; + unsigned long size; + unsigned long mmap_window; + struct map_groups kmaps; + struct rb_root threads; + struct thread *last_match; + struct map *vmlinux_maps[MAP__NR_TYPES]; + struct events_stats events_stats; + unsigned long event_total[PERF_RECORD_MAX]; + unsigned long unknown_events; + struct rb_root hists; + u64 sample_type; + struct ref_reloc_sym ref_reloc_sym; + int fd; + int cwdlen; + char *cwd; + char filename[0]; +}; + +typedef int (*event_op)(event_t *self, struct perf_session *session); + +struct perf_event_ops { + event_op sample, + mmap, + comm, + fork, + exit, + lost, + read, + throttle, + unthrottle; +}; + +struct perf_session *perf_session__new(const char *filename, int mode, bool force); +void perf_session__delete(struct perf_session *self); + +void perf_event_header__bswap(struct perf_event_header *self); + +int __perf_session__process_events(struct perf_session *self, + u64 data_offset, u64 data_size, u64 size, + struct perf_event_ops *ops); +int perf_session__process_events(struct perf_session *self, + struct perf_event_ops *event_ops); + +struct symbol **perf_session__resolve_callchain(struct perf_session *self, + struct thread *thread, + struct ip_callchain *chain, + struct symbol **parent); + +bool perf_session__has_traces(struct perf_session *self, const char *msg); + +int perf_header__read_build_ids(struct perf_header *self, int input, + u64 offset, u64 file_size); + +int perf_session__set_kallsyms_ref_reloc_sym(struct perf_session *self, + const char *symbol_name, + u64 addr); + +void mem_bswap_64(void *src, int byte_size); + +static inline int __perf_session__create_kernel_maps(struct perf_session *self, + struct dso *kernel) +{ + return __map_groups__create_kernel_maps(&self->kmaps, + self->vmlinux_maps, kernel); +} + +static inline struct map * + perf_session__new_module_map(struct perf_session *self, + u64 start, const char *filename) +{ + return map_groups__new_module(&self->kmaps, start, filename); +} +#endif /* __PERF_SESSION_H */ diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c index b490354d1b23..cb0f327de9e8 100644 --- a/tools/perf/util/sort.c +++ b/tools/perf/util/sort.c @@ -288,3 +288,29 @@ int sort_dimension__add(const char *tok) return -ESRCH; } + +void setup_sorting(const char * const usagestr[], const struct option *opts) +{ + char *tmp, *tok, *str = strdup(sort_order); + + for (tok = strtok_r(str, ", ", &tmp); + tok; tok = strtok_r(NULL, ", ", &tmp)) { + if (sort_dimension__add(tok) < 0) { + error("Unknown --sort key: `%s'", tok); + usage_with_options(usagestr, opts); + } + } + + free(str); +} + +void sort_entry__setup_elide(struct sort_entry *self, struct strlist *list, + const char *list_name, FILE *fp) +{ + if (list && strlist__nr_entries(list) == 1) { + if (fp != NULL) + fprintf(fp, "# %s: %s\n", list_name, + strlist__entry(list, 0)->s); + self->elide = true; + } +} diff --git a/tools/perf/util/sort.h b/tools/perf/util/sort.h index 333e664ff45f..753f9ea99fb0 100644 --- a/tools/perf/util/sort.h +++ b/tools/perf/util/sort.h @@ -49,9 +49,13 @@ struct hist_entry { struct symbol *sym; u64 ip; char level; - struct symbol *parent; + struct symbol *parent; struct callchain_node callchain; - struct rb_root sorted_chain; + union { + unsigned long position; + struct hist_entry *pair; + struct rb_root sorted_chain; + }; }; enum sort_type { @@ -81,6 +85,8 @@ struct sort_entry { extern struct sort_entry sort_thread; extern struct list_head hist_entry__sort_list; +void setup_sorting(const char * const usagestr[], const struct option *opts); + extern int repsep_fprintf(FILE *fp, const char *fmt, ...); extern size_t sort__thread_print(FILE *, struct hist_entry *, unsigned int); extern size_t sort__comm_print(FILE *, struct hist_entry *, unsigned int); @@ -95,5 +101,7 @@ extern int64_t sort__sym_cmp(struct hist_entry *, struct hist_entry *); extern int64_t sort__parent_cmp(struct hist_entry *, struct hist_entry *); extern size_t sort__parent_print(FILE *, struct hist_entry *, unsigned int); extern int sort_dimension__add(const char *); +void sort_entry__setup_elide(struct sort_entry *self, struct strlist *list, + const char *list_name, FILE *fp); #endif /* __PERF_SORT_H */ diff --git a/tools/perf/util/string.c b/tools/perf/util/string.c index f24a8cc933d5..c397d4f6f748 100644 --- a/tools/perf/util/string.c +++ b/tools/perf/util/string.c @@ -226,3 +226,85 @@ fail: argv_free(argv); return NULL; } + +/* Character class matching */ +static bool __match_charclass(const char *pat, char c, const char **npat) +{ + bool complement = false, ret = true; + + if (*pat == '!') { + complement = true; + pat++; + } + if (*pat++ == c) /* First character is special */ + goto end; + + while (*pat && *pat != ']') { /* Matching */ + if (*pat == '-' && *(pat + 1) != ']') { /* Range */ + if (*(pat - 1) <= c && c <= *(pat + 1)) + goto end; + if (*(pat - 1) > *(pat + 1)) + goto error; + pat += 2; + } else if (*pat++ == c) + goto end; + } + if (!*pat) + goto error; + ret = false; + +end: + while (*pat && *pat != ']') /* Searching closing */ + pat++; + if (!*pat) + goto error; + *npat = pat + 1; + return complement ? !ret : ret; + +error: + return false; +} + +/** + * strglobmatch - glob expression pattern matching + * @str: the target string to match + * @pat: the pattern string to match + * + * This returns true if the @str matches @pat. @pat can includes wildcards + * ('*','?') and character classes ([CHARS], complementation and ranges are + * also supported). Also, this supports escape character ('\') to use special + * characters as normal character. + * + * Note: if @pat syntax is broken, this always returns false. + */ +bool strglobmatch(const char *str, const char *pat) +{ + while (*str && *pat && *pat != '*') { + if (*pat == '?') { /* Matches any single character */ + str++; + pat++; + continue; + } else if (*pat == '[') /* Character classes/Ranges */ + if (__match_charclass(pat + 1, *str, &pat)) { + str++; + continue; + } else + return false; + else if (*pat == '\\') /* Escaped char match as normal char */ + pat++; + if (*str++ != *pat++) + return false; + } + /* Check wild card */ + if (*pat == '*') { + while (*pat == '*') + pat++; + if (!*pat) /* Tail wild card matches all */ + return true; + while (*str) + if (strglobmatch(str++, pat)) + return true; + } + return !*str && !*pat; +} + diff --git a/tools/perf/util/string.h b/tools/perf/util/string.h index bfecec265a1a..02ede58c54b4 100644 --- a/tools/perf/util/string.h +++ b/tools/perf/util/string.h @@ -1,6 +1,7 @@ #ifndef __PERF_STRING_H_ #define __PERF_STRING_H_ +#include <stdbool.h> #include "types.h" int hex2u64(const char *ptr, u64 *val); @@ -8,6 +9,7 @@ char *strxfrchar(char *s, char from, char to); s64 perf_atoll(const char *str); char **argv_split(const char *str, int *argcp); void argv_free(char **argv); +bool strglobmatch(const char *str, const char *pat); #define _STR(x) #x #define STR(x) _STR(x) diff --git a/tools/perf/util/strlist.c b/tools/perf/util/strlist.c index 7ad38171dc2b..6783a2043555 100644 --- a/tools/perf/util/strlist.c +++ b/tools/perf/util/strlist.c @@ -102,7 +102,7 @@ void strlist__remove(struct strlist *self, struct str_node *sn) str_node__delete(sn, self->dupstr); } -bool strlist__has_entry(struct strlist *self, const char *entry) +struct str_node *strlist__find(struct strlist *self, const char *entry) { struct rb_node **p = &self->entries.rb_node; struct rb_node *parent = NULL; @@ -120,10 +120,10 @@ bool strlist__has_entry(struct strlist *self, const char *entry) else if (rc < 0) p = &(*p)->rb_right; else - return true; + return sn; } - return false; + return NULL; } static int strlist__parse_list_entry(struct strlist *self, const char *s) diff --git a/tools/perf/util/strlist.h b/tools/perf/util/strlist.h index cb4659306d7b..3ba839007d2c 100644 --- a/tools/perf/util/strlist.h +++ b/tools/perf/util/strlist.h @@ -23,7 +23,12 @@ int strlist__load(struct strlist *self, const char *filename); int strlist__add(struct strlist *self, const char *str); struct str_node *strlist__entry(const struct strlist *self, unsigned int idx); -bool strlist__has_entry(struct strlist *self, const char *entry); +struct str_node *strlist__find(struct strlist *self, const char *entry); + +static inline bool strlist__has_entry(struct strlist *self, const char *entry) +{ + return strlist__find(self, entry) != NULL; +} static inline bool strlist__empty(const struct strlist *self) { @@ -35,5 +40,39 @@ static inline unsigned int strlist__nr_entries(const struct strlist *self) return self->nr_entries; } +/* For strlist iteration */ +static inline struct str_node *strlist__first(struct strlist *self) +{ + struct rb_node *rn = rb_first(&self->entries); + return rn ? rb_entry(rn, struct str_node, rb_node) : NULL; +} +static inline struct str_node *strlist__next(struct str_node *sn) +{ + struct rb_node *rn; + if (!sn) + return NULL; + rn = rb_next(&sn->rb_node); + return rn ? rb_entry(rn, struct str_node, rb_node) : NULL; +} + +/** + * strlist_for_each - iterate over a strlist + * @pos: the &struct str_node to use as a loop cursor. + * @self: the &struct strlist for loop. + */ +#define strlist__for_each(pos, self) \ + for (pos = strlist__first(self); pos; pos = strlist__next(pos)) + +/** + * strlist_for_each_safe - iterate over a strlist safe against removal of + * str_node + * @pos: the &struct str_node to use as a loop cursor. + * @n: another &struct str_node to use as temporary storage. + * @self: the &struct strlist for loop. + */ +#define strlist__for_each_safe(pos, n, self) \ + for (pos = strlist__first(self), n = strlist__next(pos); pos;\ + pos = n, n = strlist__next(n)) + int strlist__parse_list(struct strlist *self, const char *s); #endif /* __PERF_STRLIST_H */ diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index fffcb937cdcb..323c0aea0a91 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -1,5 +1,6 @@ #include "util.h" #include "../perf.h" +#include "sort.h" #include "string.h" #include "symbol.h" #include "thread.h" @@ -20,6 +21,7 @@ enum dso_origin { DSO__ORIG_KERNEL = 0, DSO__ORIG_JAVA_JIT, + DSO__ORIG_BUILD_ID_CACHE, DSO__ORIG_FEDORA, DSO__ORIG_UBUNTU, DSO__ORIG_BUILDID, @@ -29,31 +31,43 @@ enum dso_origin { }; static void dsos__add(struct list_head *head, struct dso *dso); -static struct map *thread__find_map_by_name(struct thread *self, char *name); static struct map *map__new2(u64 start, struct dso *dso, enum map_type type); -struct symbol *dso__find_symbol(struct dso *self, enum map_type type, u64 addr); static int dso__load_kernel_sym(struct dso *self, struct map *map, - struct thread *thread, symbol_filter_t filter); -unsigned int symbol__priv_size; + symbol_filter_t filter); static int vmlinux_path__nr_entries; static char **vmlinux_path; -static struct symbol_conf symbol_conf__defaults = { +struct symbol_conf symbol_conf = { + .exclude_other = true, .use_modules = true, .try_vmlinux_path = true, }; -static struct thread kthread_mem; -struct thread *kthread = &kthread_mem; - bool dso__loaded(const struct dso *self, enum map_type type) { return self->loaded & (1 << type); } -static void dso__set_loaded(struct dso *self, enum map_type type) +bool dso__sorted_by_name(const struct dso *self, enum map_type type) +{ + return self->sorted_by_name & (1 << type); +} + +static void dso__set_sorted_by_name(struct dso *self, enum map_type type) { - self->loaded |= (1 << type); + self->sorted_by_name |= (1 << type); +} + +bool symbol_type__is_a(char symbol_type, enum map_type map_type) +{ + switch (map_type) { + case MAP__FUNCTION: + return symbol_type == 'T' || symbol_type == 'W'; + case MAP__VARIABLE: + return symbol_type == 'D' || symbol_type == 'd'; + default: + return false; + } } static void symbols__fixup_end(struct rb_root *self) @@ -79,7 +93,7 @@ static void symbols__fixup_end(struct rb_root *self) curr->end = roundup(curr->start, 4096); } -static void __thread__fixup_maps_end(struct thread *self, enum map_type type) +static void __map_groups__fixup_end(struct map_groups *self, enum map_type type) { struct map *prev, *curr; struct rb_node *nd, *prevnd = rb_first(&self->maps[type]); @@ -102,37 +116,37 @@ static void __thread__fixup_maps_end(struct thread *self, enum map_type type) curr->end = ~0UL; } -static void thread__fixup_maps_end(struct thread *self) +static void map_groups__fixup_end(struct map_groups *self) { int i; for (i = 0; i < MAP__NR_TYPES; ++i) - __thread__fixup_maps_end(self, i); + __map_groups__fixup_end(self, i); } static struct symbol *symbol__new(u64 start, u64 len, const char *name) { size_t namelen = strlen(name) + 1; - struct symbol *self = zalloc(symbol__priv_size + + struct symbol *self = zalloc(symbol_conf.priv_size + sizeof(*self) + namelen); if (self == NULL) return NULL; - if (symbol__priv_size) - self = ((void *)self) + symbol__priv_size; + if (symbol_conf.priv_size) + self = ((void *)self) + symbol_conf.priv_size; self->start = start; self->end = len ? start + len - 1 : start; - pr_debug3("%s: %s %#Lx-%#Lx\n", __func__, name, start, self->end); + pr_debug4("%s: %s %#Lx-%#Lx\n", __func__, name, start, self->end); memcpy(self->name, name, namelen); return self; } -static void symbol__delete(struct symbol *self) +void symbol__delete(struct symbol *self) { - free(((void *)self) - symbol__priv_size); + free(((void *)self) - symbol_conf.priv_size); } static size_t symbol__fprintf(struct symbol *self, FILE *fp) @@ -141,7 +155,7 @@ static size_t symbol__fprintf(struct symbol *self, FILE *fp) self->start, self->end, self->name); } -static void dso__set_long_name(struct dso *self, char *name) +void dso__set_long_name(struct dso *self, char *name) { if (name == NULL) return; @@ -156,7 +170,7 @@ static void dso__set_basename(struct dso *self) struct dso *dso__new(const char *name) { - struct dso *self = malloc(sizeof(*self) + strlen(name) + 1); + struct dso *self = zalloc(sizeof(*self) + strlen(name) + 1); if (self != NULL) { int i; @@ -164,11 +178,11 @@ struct dso *dso__new(const char *name) dso__set_long_name(self, self->name); self->short_name = self->name; for (i = 0; i < MAP__NR_TYPES; ++i) - self->symbols[i] = RB_ROOT; - self->find_symbol = dso__find_symbol; + self->symbols[i] = self->symbol_names[i] = RB_ROOT; self->slen_calculated = 0; self->origin = DSO__ORIG_NOT_FOUND; self->loaded = 0; + self->sorted_by_name = 0; self->has_build_id = 0; } @@ -246,15 +260,89 @@ static struct symbol *symbols__find(struct rb_root *self, u64 ip) return NULL; } -struct symbol *dso__find_symbol(struct dso *self, enum map_type type, u64 addr) +struct symbol_name_rb_node { + struct rb_node rb_node; + struct symbol sym; +}; + +static void symbols__insert_by_name(struct rb_root *self, struct symbol *sym) +{ + struct rb_node **p = &self->rb_node; + struct rb_node *parent = NULL; + struct symbol_name_rb_node *symn = ((void *)sym) - sizeof(*parent), *s; + + while (*p != NULL) { + parent = *p; + s = rb_entry(parent, struct symbol_name_rb_node, rb_node); + if (strcmp(sym->name, s->sym.name) < 0) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + } + rb_link_node(&symn->rb_node, parent, p); + rb_insert_color(&symn->rb_node, self); +} + +static void symbols__sort_by_name(struct rb_root *self, struct rb_root *source) +{ + struct rb_node *nd; + + for (nd = rb_first(source); nd; nd = rb_next(nd)) { + struct symbol *pos = rb_entry(nd, struct symbol, rb_node); + symbols__insert_by_name(self, pos); + } +} + +static struct symbol *symbols__find_by_name(struct rb_root *self, const char *name) +{ + struct rb_node *n; + + if (self == NULL) + return NULL; + + n = self->rb_node; + + while (n) { + struct symbol_name_rb_node *s; + int cmp; + + s = rb_entry(n, struct symbol_name_rb_node, rb_node); + cmp = strcmp(name, s->sym.name); + + if (cmp < 0) + n = n->rb_left; + else if (cmp > 0) + n = n->rb_right; + else + return &s->sym; + } + + return NULL; +} + +struct symbol *dso__find_symbol(struct dso *self, + enum map_type type, u64 addr) { return symbols__find(&self->symbols[type], addr); } -int build_id__sprintf(u8 *self, int len, char *bf) +struct symbol *dso__find_symbol_by_name(struct dso *self, enum map_type type, + const char *name) +{ + return symbols__find_by_name(&self->symbol_names[type], name); +} + +void dso__sort_by_name(struct dso *self, enum map_type type) +{ + dso__set_sorted_by_name(self, type); + return symbols__sort_by_name(&self->symbol_names[type], + &self->symbols[type]); +} + +int build_id__sprintf(const u8 *self, int len, char *bf) { char *bid = bf; - u8 *raw = self; + const u8 *raw = self; int i; for (i = 0; i < len; ++i) { @@ -279,6 +367,10 @@ size_t dso__fprintf(struct dso *self, enum map_type type, FILE *fp) struct rb_node *nd; size_t ret = fprintf(fp, "dso: %s (", self->short_name); + if (self->short_name != self->long_name) + ret += fprintf(fp, "%s, ", self->long_name); + ret += fprintf(fp, "%s, %sloaded, ", map_type__name[type], + self->loaded ? "" : "NOT "); ret += dso__fprintf_buildid(self, fp); ret += fprintf(fp, ")\n"); for (nd = rb_first(&self->symbols[type]); nd; nd = rb_next(nd)) { @@ -289,24 +381,20 @@ size_t dso__fprintf(struct dso *self, enum map_type type, FILE *fp) return ret; } -/* - * Loads the function entries in /proc/kallsyms into kernel_map->dso, - * so that we can in the next step set the symbol ->end address and then - * call kernel_maps__split_kallsyms. - */ -static int dso__load_all_kallsyms(struct dso *self, struct map *map) +int kallsyms__parse(const char *filename, void *arg, + int (*process_symbol)(void *arg, const char *name, + char type, u64 start)) { char *line = NULL; size_t n; - struct rb_root *root = &self->symbols[map->type]; - FILE *file = fopen("/proc/kallsyms", "r"); + int err = 0; + FILE *file = fopen(filename, "r"); if (file == NULL) goto out_failure; while (!feof(file)) { u64 start; - struct symbol *sym; int line_len, len; char symbol_type; char *symbol_name; @@ -327,46 +415,72 @@ static int dso__load_all_kallsyms(struct dso *self, struct map *map) continue; symbol_type = toupper(line[len]); - /* - * We're interested only in code ('T'ext) - */ - if (symbol_type != 'T' && symbol_type != 'W') - continue; - symbol_name = line + len + 2; - /* - * Will fix up the end later, when we have all symbols sorted. - */ - sym = symbol__new(start, 0, symbol_name); - if (sym == NULL) - goto out_delete_line; - /* - * We will pass the symbols to the filter later, in - * map__split_kallsyms, when we have split the maps per module - */ - symbols__insert(root, sym); + err = process_symbol(arg, symbol_name, symbol_type, start); + if (err) + break; } free(line); fclose(file); + return err; - return 0; - -out_delete_line: - free(line); out_failure: return -1; } +struct process_kallsyms_args { + struct map *map; + struct dso *dso; +}; + +static int map__process_kallsym_symbol(void *arg, const char *name, + char type, u64 start) +{ + struct symbol *sym; + struct process_kallsyms_args *a = arg; + struct rb_root *root = &a->dso->symbols[a->map->type]; + + if (!symbol_type__is_a(type, a->map->type)) + return 0; + + /* + * Will fix up the end later, when we have all symbols sorted. + */ + sym = symbol__new(start, 0, name); + + if (sym == NULL) + return -ENOMEM; + /* + * We will pass the symbols to the filter later, in + * map__split_kallsyms, when we have split the maps per module + */ + symbols__insert(root, sym); + return 0; +} + +/* + * Loads the function entries in /proc/kallsyms into kernel_map->dso, + * so that we can in the next step set the symbol ->end address and then + * call kernel_maps__split_kallsyms. + */ +static int dso__load_all_kallsyms(struct dso *self, const char *filename, + struct map *map) +{ + struct process_kallsyms_args args = { .map = map, .dso = self, }; + return kallsyms__parse(filename, &args, map__process_kallsym_symbol); +} + /* * Split the symbols into maps, making sure there are no overlaps, i.e. the * kernel range is broken in several maps, named [kernel].N, as we don't have * the original ELF section names vmlinux have. */ -static int dso__split_kallsyms(struct dso *self, struct map *map, struct thread *thread, +static int dso__split_kallsyms(struct dso *self, struct map *map, symbol_filter_t filter) { + struct map_groups *kmaps = map__kmap(map)->kmaps; struct map *curr_map = map; struct symbol *pos; int count = 0; @@ -382,18 +496,22 @@ static int dso__split_kallsyms(struct dso *self, struct map *map, struct thread module = strchr(pos->name, '\t'); if (module) { - if (!thread->use_modules) + if (!symbol_conf.use_modules) goto discard_symbol; *module++ = '\0'; - if (strcmp(self->name, module)) { - curr_map = thread__find_map_by_name(thread, module); + if (strcmp(curr_map->dso->short_name, module)) { + curr_map = map_groups__find_by_name(kmaps, map->type, module); if (curr_map == NULL) { pr_debug("/proc/{kallsyms,modules} " - "inconsistency!\n"); + "inconsistency while looking " + "for \"%s\" module!\n", module); return -1; } + + if (curr_map->dso->loaded) + goto discard_symbol; } /* * So that we look just like we get from .ko files, @@ -413,13 +531,13 @@ static int dso__split_kallsyms(struct dso *self, struct map *map, struct thread return -1; curr_map = map__new2(pos->start, dso, map->type); - if (map == NULL) { + if (curr_map == NULL) { dso__delete(dso); return -1; } curr_map->map_ip = curr_map->unmap_ip = identity__map_ip; - __thread__insert_map(thread, curr_map); + map_groups__insert(kmaps, curr_map); ++kernel_range; } @@ -438,24 +556,16 @@ discard_symbol: rb_erase(&pos->rb_node, root); return count; } - -static int dso__load_kallsyms(struct dso *self, struct map *map, - struct thread *thread, symbol_filter_t filter) +int dso__load_kallsyms(struct dso *self, const char *filename, + struct map *map, symbol_filter_t filter) { - if (dso__load_all_kallsyms(self, map) < 0) + if (dso__load_all_kallsyms(self, filename, map) < 0) return -1; symbols__fixup_end(&self->symbols[map->type]); self->origin = DSO__ORIG_KERNEL; - return dso__split_kallsyms(self, map, thread, filter); -} - -size_t kernel_maps__fprintf(FILE *fp) -{ - size_t printed = fprintf(fp, "Kernel maps:\n"); - printed += thread__fprintf_maps(kthread, fp); - return printed + fprintf(fp, "END kernel maps\n"); + return dso__split_kallsyms(self, map, filter); } static int dso__load_perf_map(struct dso *self, struct map *map, @@ -544,6 +654,13 @@ static inline int elf_sym__is_function(const GElf_Sym *sym) sym->st_shndx != SHN_UNDEF; } +static inline bool elf_sym__is_object(const GElf_Sym *sym) +{ + return elf_sym__type(sym) == STT_OBJECT && + sym->st_name != 0 && + sym->st_shndx != SHN_UNDEF; +} + static inline int elf_sym__is_label(const GElf_Sym *sym) { return elf_sym__type(sym) == STT_NOTYPE && @@ -564,6 +681,12 @@ static inline int elf_sec__is_text(const GElf_Shdr *shdr, return strstr(elf_sec__name(shdr, secstrs), "text") != NULL; } +static inline bool elf_sec__is_data(const GElf_Shdr *shdr, + const Elf_Data *secstrs) +{ + return strstr(elf_sec__name(shdr, secstrs), "data") != NULL; +} + static inline const char *elf_sym__name(const GElf_Sym *sym, const Elf_Data *symstrs) { @@ -744,10 +867,34 @@ out: return 0; } -static int dso__load_sym(struct dso *self, struct map *map, - struct thread *thread, const char *name, int fd, - symbol_filter_t filter, int kernel, int kmodule) +static bool elf_sym__is_a(GElf_Sym *self, enum map_type type) +{ + switch (type) { + case MAP__FUNCTION: + return elf_sym__is_function(self); + case MAP__VARIABLE: + return elf_sym__is_object(self); + default: + return false; + } +} + +static bool elf_sec__is_a(GElf_Shdr *self, Elf_Data *secstrs, enum map_type type) +{ + switch (type) { + case MAP__FUNCTION: + return elf_sec__is_text(self, secstrs); + case MAP__VARIABLE: + return elf_sec__is_data(self, secstrs); + default: + return false; + } +} + +static int dso__load_sym(struct dso *self, struct map *map, const char *name, + int fd, symbol_filter_t filter, int kmodule) { + struct kmap *kmap = self->kernel ? map__kmap(map) : NULL; struct map *curr_map = map; struct dso *curr_dso = self; size_t dso_name_len = strlen(self->short_name); @@ -804,7 +951,7 @@ static int dso__load_sym(struct dso *self, struct map *map, nr_syms = shdr.sh_size / shdr.sh_entsize; memset(&sym, 0, sizeof(sym)); - if (!kernel) { + if (!self->kernel) { self->adjust_symbols = (ehdr.e_type == ET_EXEC || elf_section_by_name(elf, &ehdr, &shdr, ".gnu.prelink_undo", @@ -813,12 +960,16 @@ static int dso__load_sym(struct dso *self, struct map *map, elf_symtab__for_each_symbol(syms, nr_syms, idx, sym) { struct symbol *f; - const char *elf_name; + const char *elf_name = elf_sym__name(&sym, symstrs); char *demangled = NULL; int is_label = elf_sym__is_label(&sym); const char *section_name; - if (!is_label && !elf_sym__is_function(&sym)) + if (kmap && kmap->ref_reloc_sym && kmap->ref_reloc_sym->name && + strcmp(elf_name, kmap->ref_reloc_sym->name) == 0) + kmap->ref_reloc_sym->unrelocated_addr = sym.st_value; + + if (!is_label && !elf_sym__is_a(&sym, map->type)) continue; sec = elf_getscn(elf, sym.st_shndx); @@ -827,13 +978,12 @@ static int dso__load_sym(struct dso *self, struct map *map, gelf_getshdr(sec, &shdr); - if (is_label && !elf_sec__is_text(&shdr, secstrs)) + if (is_label && !elf_sec__is_a(&shdr, secstrs, map->type)) continue; - elf_name = elf_sym__name(&sym, symstrs); section_name = elf_sec__name(&shdr, secstrs); - if (kernel || kmodule) { + if (self->kernel || kmodule) { char dso_name[PATH_MAX]; if (strcmp(section_name, @@ -849,7 +999,7 @@ static int dso__load_sym(struct dso *self, struct map *map, snprintf(dso_name, sizeof(dso_name), "%s%s", self->short_name, section_name); - curr_map = thread__find_map_by_name(thread, dso_name); + curr_map = map_groups__find_by_name(kmap->kmaps, map->type, dso_name); if (curr_map == NULL) { u64 start = sym.st_value; @@ -860,7 +1010,7 @@ static int dso__load_sym(struct dso *self, struct map *map, if (curr_dso == NULL) goto out_elf_end; curr_map = map__new2(start, curr_dso, - MAP__FUNCTION); + map->type); if (curr_map == NULL) { dso__delete(curr_dso); goto out_elf_end; @@ -868,8 +1018,9 @@ static int dso__load_sym(struct dso *self, struct map *map, curr_map->map_ip = identity__map_ip; curr_map->unmap_ip = identity__map_ip; curr_dso->origin = DSO__ORIG_KERNEL; - __thread__insert_map(kthread, curr_map); + map_groups__insert(kmap->kmaps, curr_map); dsos__add(&dsos__kernel, curr_dso); + dso__set_loaded(curr_dso, map->type); } else curr_dso = curr_map->dso; @@ -877,9 +1028,10 @@ static int dso__load_sym(struct dso *self, struct map *map, } if (curr_dso->adjust_symbols) { - pr_debug2("adjusting symbol: st_value: %Lx sh_addr: " - "%Lx sh_offset: %Lx\n", (u64)sym.st_value, - (u64)shdr.sh_addr, (u64)shdr.sh_offset); + pr_debug4("%s: adjusting symbol: st_value: %#Lx " + "sh_addr: %#Lx sh_offset: %#Lx\n", __func__, + (u64)sym.st_value, (u64)shdr.sh_addr, + (u64)shdr.sh_offset); sym.st_value -= shdr.sh_addr - shdr.sh_offset; } /* @@ -907,8 +1059,16 @@ new_symbol: /* * For misannotated, zeroed, ASM function sizes. */ - if (nr > 0) + if (nr > 0) { symbols__fixup_end(&self->symbols[map->type]); + if (kmap) { + /* + * We need to fixup this here too because we create new + * maps here, for things like vsyscall sections. + */ + __map_groups__fixup_end(kmap->kmaps, map->type); + } + } err = nr; out_elf_end: elf_end(elf); @@ -921,25 +1081,29 @@ static bool dso__build_id_equal(const struct dso *self, u8 *build_id) return memcmp(self->build_id, build_id, sizeof(self->build_id)) == 0; } -static bool __dsos__read_build_ids(struct list_head *head) +static bool __dsos__read_build_ids(struct list_head *head, bool with_hits) { bool have_build_id = false; struct dso *pos; - list_for_each_entry(pos, head, node) + list_for_each_entry(pos, head, node) { + if (with_hits && !pos->hit) + continue; if (filename__read_build_id(pos->long_name, pos->build_id, sizeof(pos->build_id)) > 0) { have_build_id = true; pos->has_build_id = true; } + } return have_build_id; } -bool dsos__read_build_ids(void) +bool dsos__read_build_ids(bool with_hits) { - return __dsos__read_build_ids(&dsos__kernel) || - __dsos__read_build_ids(&dsos__user); + bool kbuildids = __dsos__read_build_ids(&dsos__kernel, with_hits), + ubuildids = __dsos__read_build_ids(&dsos__user, with_hits); + return kbuildids || ubuildids; } /* @@ -1070,6 +1234,7 @@ char dso__symtab_origin(const struct dso *self) static const char origin[] = { [DSO__ORIG_KERNEL] = 'k', [DSO__ORIG_JAVA_JIT] = 'j', + [DSO__ORIG_BUILD_ID_CACHE] = 'B', [DSO__ORIG_FEDORA] = 'f', [DSO__ORIG_UBUNTU] = 'u', [DSO__ORIG_BUILDID] = 'b', @@ -1087,13 +1252,14 @@ int dso__load(struct dso *self, struct map *map, symbol_filter_t filter) int size = PATH_MAX; char *name; u8 build_id[BUILD_ID_SIZE]; + char build_id_hex[BUILD_ID_SIZE * 2 + 1]; int ret = -1; int fd; dso__set_loaded(self, map->type); if (self->kernel) - return dso__load_kernel_sym(self, map, kthread, filter); + return dso__load_kernel_sym(self, map, filter); name = malloc(size); if (!name) @@ -1108,8 +1274,16 @@ int dso__load(struct dso *self, struct map *map, symbol_filter_t filter) return ret; } - self->origin = DSO__ORIG_FEDORA - 1; + self->origin = DSO__ORIG_BUILD_ID_CACHE; + if (self->has_build_id) { + build_id__sprintf(self->build_id, sizeof(self->build_id), + build_id_hex); + snprintf(name, size, "%s/%s/.build-id/%.2s/%s", + getenv("HOME"), DEBUG_CACHE_DIR, + build_id_hex, build_id_hex + 2); + goto open_file; + } more: do { self->origin++; @@ -1125,8 +1299,6 @@ more: case DSO__ORIG_BUILDID: if (filename__read_build_id(self->long_name, build_id, sizeof(build_id))) { - char build_id_hex[BUILD_ID_SIZE * 2 + 1]; - build_id__sprintf(build_id, sizeof(build_id), build_id_hex); snprintf(name, size, @@ -1154,11 +1326,11 @@ compare_build_id: if (!dso__build_id_equal(self, build_id)) goto more; } - +open_file: fd = open(name, O_RDONLY); } while (fd < 0); - ret = dso__load_sym(self, map, NULL, name, fd, filter, 0, 0); + ret = dso__load_sym(self, map, name, fd, filter, 0); close(fd); /* @@ -1179,21 +1351,42 @@ out: return ret; } -static struct map *thread__find_map_by_name(struct thread *self, char *name) +struct map *map_groups__find_by_name(struct map_groups *self, + enum map_type type, const char *name) { struct rb_node *nd; - for (nd = rb_first(&self->maps[MAP__FUNCTION]); nd; nd = rb_next(nd)) { + for (nd = rb_first(&self->maps[type]); nd; nd = rb_next(nd)) { struct map *map = rb_entry(nd, struct map, rb_node); - if (map->dso && strcmp(map->dso->name, name) == 0) + if (map->dso && strcmp(map->dso->short_name, name) == 0) return map; } return NULL; } -static int dsos__set_modules_path_dir(char *dirname) +static int dso__kernel_module_get_build_id(struct dso *self) +{ + char filename[PATH_MAX]; + /* + * kernel module short names are of the form "[module]" and + * we need just "module" here. + */ + const char *name = self->short_name + 1; + + snprintf(filename, sizeof(filename), + "/sys/module/%.*s/notes/.note.gnu.build-id", + (int)strlen(name - 1), name); + + if (sysfs__read_build_id(filename, self->build_id, + sizeof(self->build_id)) == 0) + self->has_build_id = true; + + return 0; +} + +static int map_groups__set_modules_path_dir(struct map_groups *self, char *dirname) { struct dirent *dent; DIR *dir = opendir(dirname); @@ -1213,7 +1406,7 @@ static int dsos__set_modules_path_dir(char *dirname) snprintf(path, sizeof(path), "%s/%s", dirname, dent->d_name); - if (dsos__set_modules_path_dir(path) < 0) + if (map_groups__set_modules_path_dir(self, path) < 0) goto failure; } else { char *dot = strrchr(dent->d_name, '.'), @@ -1227,7 +1420,7 @@ static int dsos__set_modules_path_dir(char *dirname) (int)(dot - dent->d_name), dent->d_name); strxfrchar(dso_name, '-', '_'); - map = thread__find_map_by_name(kthread, dso_name); + map = map_groups__find_by_name(self, MAP__FUNCTION, dso_name); if (map == NULL) continue; @@ -1238,6 +1431,7 @@ static int dsos__set_modules_path_dir(char *dirname) if (long_name == NULL) goto failure; dso__set_long_name(map->dso, long_name); + dso__kernel_module_get_build_id(map->dso); } } @@ -1247,7 +1441,7 @@ failure: return -1; } -static int dsos__set_modules_path(void) +static int map_groups__set_modules_path(struct map_groups *self) { struct utsname uts; char modules_path[PATH_MAX]; @@ -1258,7 +1452,7 @@ static int dsos__set_modules_path(void) snprintf(modules_path, sizeof(modules_path), "/lib/modules/%s/kernel", uts.release); - return dsos__set_modules_path_dir(modules_path); + return map_groups__set_modules_path_dir(self, modules_path); } /* @@ -1268,8 +1462,8 @@ static int dsos__set_modules_path(void) */ static struct map *map__new2(u64 start, struct dso *dso, enum map_type type) { - struct map *self = malloc(sizeof(*self)); - + struct map *self = zalloc(sizeof(*self) + + (dso->kernel ? sizeof(struct kmap) : 0)); if (self != NULL) { /* * ->end will be filled after we load all the symbols @@ -1280,7 +1474,25 @@ static struct map *map__new2(u64 start, struct dso *dso, enum map_type type) return self; } -static int thread__create_module_maps(struct thread *self) +struct map *map_groups__new_module(struct map_groups *self, u64 start, + const char *filename) +{ + struct map *map; + struct dso *dso = __dsos__findnew(&dsos__kernel, filename); + + if (dso == NULL) + return NULL; + + map = map__new2(start, dso, MAP__FUNCTION); + if (map == NULL) + return NULL; + + dso->origin = DSO__ORIG_KMODULE; + map_groups__insert(self, map); + return map; +} + +static int map_groups__create_modules(struct map_groups *self) { char *line = NULL; size_t n; @@ -1293,7 +1505,6 @@ static int thread__create_module_maps(struct thread *self) while (!feof(file)) { char name[PATH_MAX]; u64 start; - struct dso *dso; char *sep; int line_len; @@ -1319,32 +1530,16 @@ static int thread__create_module_maps(struct thread *self) *sep = '\0'; snprintf(name, sizeof(name), "[%s]", line); - dso = dso__new(name); - - if (dso == NULL) - goto out_delete_line; - - map = map__new2(start, dso, MAP__FUNCTION); - if (map == NULL) { - dso__delete(dso); + map = map_groups__new_module(self, start, name); + if (map == NULL) goto out_delete_line; - } - - snprintf(name, sizeof(name), - "/sys/module/%s/notes/.note.gnu.build-id", line); - if (sysfs__read_build_id(name, dso->build_id, - sizeof(dso->build_id)) == 0) - dso->has_build_id = true; - - dso->origin = DSO__ORIG_KMODULE; - __thread__insert_map(self, map); - dsos__add(&dsos__kernel, dso); + dso__kernel_module_get_build_id(map->dso); } free(line); fclose(file); - return dsos__set_modules_path(); + return map_groups__set_modules_path(self); out_delete_line: free(line); @@ -1352,7 +1547,7 @@ out_failure: return -1; } -static int dso__load_vmlinux(struct dso *self, struct map *map, struct thread *thread, +static int dso__load_vmlinux(struct dso *self, struct map *map, const char *vmlinux, symbol_filter_t filter) { int err = -1, fd; @@ -1386,51 +1581,124 @@ static int dso__load_vmlinux(struct dso *self, struct map *map, struct thread *t return -1; dso__set_loaded(self, map->type); - err = dso__load_sym(self, map, thread, self->long_name, fd, filter, 1, 0); + err = dso__load_sym(self, map, vmlinux, fd, filter, 0); close(fd); + if (err > 0) + pr_debug("Using %s for symbols\n", vmlinux); + + return err; +} + +int dso__load_vmlinux_path(struct dso *self, struct map *map, + symbol_filter_t filter) +{ + int i, err = 0; + + pr_debug("Looking at the vmlinux_path (%d entries long)\n", + vmlinux_path__nr_entries); + + for (i = 0; i < vmlinux_path__nr_entries; ++i) { + err = dso__load_vmlinux(self, map, vmlinux_path[i], filter); + if (err > 0) { + dso__set_long_name(self, strdup(vmlinux_path[i])); + break; + } + } + return err; } static int dso__load_kernel_sym(struct dso *self, struct map *map, - struct thread *thread, symbol_filter_t filter) + symbol_filter_t filter) { int err; - bool is_kallsyms; + const char *kallsyms_filename = NULL; + char *kallsyms_allocated_filename = NULL; + /* + * Step 1: if the user specified a vmlinux filename, use it and only + * it, reporting errors to the user if it cannot be used. + * + * For instance, try to analyse an ARM perf.data file _without_ a + * build-id, or if the user specifies the wrong path to the right + * vmlinux file, obviously we can't fallback to another vmlinux (a + * x86_86 one, on the machine where analysis is being performed, say), + * or worse, /proc/kallsyms. + * + * If the specified file _has_ a build-id and there is a build-id + * section in the perf.data file, we will still do the expected + * validation in dso__load_vmlinux and will bail out if they don't + * match. + */ + if (symbol_conf.vmlinux_name != NULL) { + err = dso__load_vmlinux(self, map, + symbol_conf.vmlinux_name, filter); + goto out_try_fixup; + } if (vmlinux_path != NULL) { - int i; - pr_debug("Looking at the vmlinux_path (%d entries long)\n", - vmlinux_path__nr_entries); - for (i = 0; i < vmlinux_path__nr_entries; ++i) { - err = dso__load_vmlinux(self, map, thread, - vmlinux_path[i], filter); - if (err > 0) { - pr_debug("Using %s for symbols\n", - vmlinux_path[i]); - dso__set_long_name(self, - strdup(vmlinux_path[i])); - goto out_fixup; + err = dso__load_vmlinux_path(self, map, filter); + if (err > 0) + goto out_fixup; + } + + /* + * Say the kernel DSO was created when processing the build-id header table, + * we have a build-id, so check if it is the same as the running kernel, + * using it if it is. + */ + if (self->has_build_id) { + u8 kallsyms_build_id[BUILD_ID_SIZE]; + char sbuild_id[BUILD_ID_SIZE * 2 + 1]; + + if (sysfs__read_build_id("/sys/kernel/notes", kallsyms_build_id, + sizeof(kallsyms_build_id)) == 0) { + if (dso__build_id_equal(self, kallsyms_build_id)) { + kallsyms_filename = "/proc/kallsyms"; + goto do_kallsyms; } } - } + /* + * Now look if we have it on the build-id cache in + * $HOME/.debug/[kernel.kallsyms]. + */ + build_id__sprintf(self->build_id, sizeof(self->build_id), + sbuild_id); - is_kallsyms = self->long_name[0] == '['; - if (is_kallsyms) - goto do_kallsyms; + if (asprintf(&kallsyms_allocated_filename, + "%s/.debug/[kernel.kallsyms]/%s", + getenv("HOME"), sbuild_id) == -1) { + pr_err("Not enough memory for kallsyms file lookup\n"); + return -1; + } - err = dso__load_vmlinux(self, map, thread, self->long_name, filter); - if (err <= 0) { - pr_info("The file %s cannot be used, " - "trying to use /proc/kallsyms...", self->long_name); -do_kallsyms: - err = dso__load_kallsyms(self, map, thread, filter); - if (err > 0 && !is_kallsyms) - dso__set_long_name(self, strdup("[kernel.kallsyms]")); + kallsyms_filename = kallsyms_allocated_filename; + + if (access(kallsyms_filename, F_OK)) { + pr_err("No kallsyms or vmlinux with build-id %s " + "was found\n", sbuild_id); + free(kallsyms_allocated_filename); + return -1; + } + } else { + /* + * Last resort, if we don't have a build-id and couldn't find + * any vmlinux file, try the running kernel kallsyms table. + */ + kallsyms_filename = "/proc/kallsyms"; } +do_kallsyms: + err = dso__load_kallsyms(self, kallsyms_filename, map, filter); + if (err > 0) + pr_debug("Using %s for symbols\n", kallsyms_filename); + free(kallsyms_allocated_filename); + +out_try_fixup: if (err > 0) { out_fixup: + if (kallsyms_filename != NULL) + dso__set_long_name(self, strdup("[kernel.kallsyms]")); map__fixup_start(map); map__fixup_end(map); } @@ -1440,7 +1708,6 @@ out_fixup: LIST_HEAD(dsos__user); LIST_HEAD(dsos__kernel); -struct dso *vdso; static void dsos__add(struct list_head *head, struct dso *dso) { @@ -1452,19 +1719,19 @@ static struct dso *dsos__find(struct list_head *head, const char *name) struct dso *pos; list_for_each_entry(pos, head, node) - if (strcmp(pos->name, name) == 0) + if (strcmp(pos->long_name, name) == 0) return pos; return NULL; } -struct dso *dsos__findnew(const char *name) +struct dso *__dsos__findnew(struct list_head *head, const char *name) { - struct dso *dso = dsos__find(&dsos__user, name); + struct dso *dso = dsos__find(head, name); if (!dso) { dso = dso__new(name); if (dso != NULL) { - dsos__add(&dsos__user, dso); + dsos__add(head, dso); dso__set_basename(dso); } } @@ -1489,60 +1756,80 @@ void dsos__fprintf(FILE *fp) __dsos__fprintf(&dsos__user, fp); } -static size_t __dsos__fprintf_buildid(struct list_head *head, FILE *fp) +static size_t __dsos__fprintf_buildid(struct list_head *head, FILE *fp, + bool with_hits) { struct dso *pos; size_t ret = 0; list_for_each_entry(pos, head, node) { + if (with_hits && !pos->hit) + continue; ret += dso__fprintf_buildid(pos, fp); ret += fprintf(fp, " %s\n", pos->long_name); } return ret; } -size_t dsos__fprintf_buildid(FILE *fp) +size_t dsos__fprintf_buildid(FILE *fp, bool with_hits) { - return (__dsos__fprintf_buildid(&dsos__kernel, fp) + - __dsos__fprintf_buildid(&dsos__user, fp)); + return (__dsos__fprintf_buildid(&dsos__kernel, fp, with_hits) + + __dsos__fprintf_buildid(&dsos__user, fp, with_hits)); } -static int thread__create_kernel_map(struct thread *self, const char *vmlinux) +struct dso *dso__new_kernel(const char *name) { - struct map *kmap; - struct dso *kernel = dso__new(vmlinux ?: "[kernel.kallsyms]"); + struct dso *self = dso__new(name ?: "[kernel.kallsyms]"); - if (kernel == NULL) - return -1; + if (self != NULL) { + self->short_name = "[kernel]"; + self->kernel = 1; + } - kmap = map__new2(0, kernel, MAP__FUNCTION); - if (kmap == NULL) - goto out_delete_kernel_dso; + return self; +} - kmap->map_ip = kmap->unmap_ip = identity__map_ip; - kernel->short_name = "[kernel]"; - kernel->kernel = 1; +void dso__read_running_kernel_build_id(struct dso *self) +{ + if (sysfs__read_build_id("/sys/kernel/notes", self->build_id, + sizeof(self->build_id)) == 0) + self->has_build_id = true; +} - vdso = dso__new("[vdso]"); - if (vdso == NULL) - goto out_delete_kernel_map; - dso__set_loaded(vdso, MAP__FUNCTION); +static struct dso *dsos__create_kernel(const char *vmlinux) +{ + struct dso *kernel = dso__new_kernel(vmlinux); - if (sysfs__read_build_id("/sys/kernel/notes", kernel->build_id, - sizeof(kernel->build_id)) == 0) - kernel->has_build_id = true; + if (kernel != NULL) { + dso__read_running_kernel_build_id(kernel); + dsos__add(&dsos__kernel, kernel); + } - __thread__insert_map(self, kmap); - dsos__add(&dsos__kernel, kernel); - dsos__add(&dsos__user, vdso); + return kernel; +} - return 0; +int __map_groups__create_kernel_maps(struct map_groups *self, + struct map *vmlinux_maps[MAP__NR_TYPES], + struct dso *kernel) +{ + enum map_type type; -out_delete_kernel_map: - map__delete(kmap); -out_delete_kernel_dso: - dso__delete(kernel); - return -1; + for (type = 0; type < MAP__NR_TYPES; ++type) { + struct kmap *kmap; + + vmlinux_maps[type] = map__new2(0, kernel, type); + if (vmlinux_maps[type] == NULL) + return -1; + + vmlinux_maps[type]->map_ip = + vmlinux_maps[type]->unmap_ip = identity__map_ip; + + kmap = map__kmap(vmlinux_maps[type]); + kmap->kmaps = self; + map_groups__insert(self, vmlinux_maps[type]); + } + + return 0; } static void vmlinux_path__exit(void) @@ -1600,29 +1887,72 @@ out_fail: return -1; } -int symbol__init(struct symbol_conf *conf) +static int setup_list(struct strlist **list, const char *list_str, + const char *list_name) { - const struct symbol_conf *pconf = conf ?: &symbol_conf__defaults; + if (list_str == NULL) + return 0; + *list = strlist__new(true, list_str); + if (!*list) { + pr_err("problems parsing %s list\n", list_name); + return -1; + } + return 0; +} + +int symbol__init(void) +{ elf_version(EV_CURRENT); - symbol__priv_size = pconf->priv_size; - thread__init(kthread, 0); + if (symbol_conf.sort_by_name) + symbol_conf.priv_size += (sizeof(struct symbol_name_rb_node) - + sizeof(struct symbol)); - if (pconf->try_vmlinux_path && vmlinux_path__init() < 0) + if (symbol_conf.try_vmlinux_path && vmlinux_path__init() < 0) return -1; - if (thread__create_kernel_map(kthread, pconf->vmlinux_name) < 0) { - vmlinux_path__exit(); + if (symbol_conf.field_sep && *symbol_conf.field_sep == '.') { + pr_err("'.' is the only non valid --field-separator argument\n"); return -1; } - kthread->use_modules = pconf->use_modules; - if (pconf->use_modules && thread__create_module_maps(kthread) < 0) - pr_debug("Failed to load list of modules in use, " - "continuing...\n"); + if (setup_list(&symbol_conf.dso_list, + symbol_conf.dso_list_str, "dso") < 0) + return -1; + + if (setup_list(&symbol_conf.comm_list, + symbol_conf.comm_list_str, "comm") < 0) + goto out_free_dso_list; + + if (setup_list(&symbol_conf.sym_list, + symbol_conf.sym_list_str, "symbol") < 0) + goto out_free_comm_list; + + return 0; + +out_free_dso_list: + strlist__delete(symbol_conf.dso_list); +out_free_comm_list: + strlist__delete(symbol_conf.comm_list); + return -1; +} + +int map_groups__create_kernel_maps(struct map_groups *self, + struct map *vmlinux_maps[MAP__NR_TYPES]) +{ + struct dso *kernel = dsos__create_kernel(symbol_conf.vmlinux_name); + + if (kernel == NULL) + return -1; + + if (__map_groups__create_kernel_maps(self, vmlinux_maps, kernel) < 0) + return -1; + + if (symbol_conf.use_modules && map_groups__create_modules(self) < 0) + pr_debug("Problems creating module maps, continuing anyway...\n"); /* * Now that we have all the maps created, just set the ->end of them: */ - thread__fixup_maps_end(kthread); + map_groups__fixup_end(self); return 0; } diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h index 17003efa0b39..280dadd32a08 100644 --- a/tools/perf/util/symbol.h +++ b/tools/perf/util/symbol.h @@ -8,6 +8,8 @@ #include <linux/rbtree.h> #include "event.h" +#define DEBUG_CACHE_DIR ".debug" + #ifdef HAVE_CPLUS_DEMANGLE extern char *cplus_demangle(const char *, int); @@ -49,38 +51,63 @@ struct symbol { char name[0]; }; +void symbol__delete(struct symbol *self); + +struct strlist; + struct symbol_conf { unsigned short priv_size; bool try_vmlinux_path, - use_modules; - const char *vmlinux_name; + use_modules, + sort_by_name, + show_nr_samples, + use_callchain, + exclude_other, + full_paths; + const char *vmlinux_name, + *field_sep; + char *dso_list_str, + *comm_list_str, + *sym_list_str, + *col_width_list_str; + struct strlist *dso_list, + *comm_list, + *sym_list; }; -extern unsigned int symbol__priv_size; +extern struct symbol_conf symbol_conf; static inline void *symbol__priv(struct symbol *self) { - return ((void *)self) - symbol__priv_size; + return ((void *)self) - symbol_conf.priv_size; } +struct ref_reloc_sym { + const char *name; + u64 addr; + u64 unrelocated_addr; +}; + struct addr_location { struct thread *thread; struct map *map; struct symbol *sym; u64 addr; char level; + bool filtered; }; struct dso { struct list_head node; struct rb_root symbols[MAP__NR_TYPES]; - struct symbol *(*find_symbol)(struct dso *self, - enum map_type type, u64 addr); + struct rb_root symbol_names[MAP__NR_TYPES]; u8 adjust_symbols:1; u8 slen_calculated:1; u8 has_build_id:1; u8 kernel:1; + u8 hit:1; unsigned char origin; + u8 sorted_by_name; u8 loaded; u8 build_id[BUILD_ID_SIZE]; u16 long_name_len; @@ -90,31 +117,55 @@ struct dso { }; struct dso *dso__new(const char *name); +struct dso *dso__new_kernel(const char *name); void dso__delete(struct dso *self); bool dso__loaded(const struct dso *self, enum map_type type); +bool dso__sorted_by_name(const struct dso *self, enum map_type type); + +static inline void dso__set_loaded(struct dso *self, enum map_type type) +{ + self->loaded |= (1 << type); +} + +void dso__sort_by_name(struct dso *self, enum map_type type); + +extern struct list_head dsos__user, dsos__kernel; + +struct dso *__dsos__findnew(struct list_head *head, const char *name); + +static inline struct dso *dsos__findnew(const char *name) +{ + return __dsos__findnew(&dsos__user, name); +} -struct dso *dsos__findnew(const char *name); int dso__load(struct dso *self, struct map *map, symbol_filter_t filter); +int dso__load_vmlinux_path(struct dso *self, struct map *map, + symbol_filter_t filter); +int dso__load_kallsyms(struct dso *self, const char *filename, struct map *map, + symbol_filter_t filter); void dsos__fprintf(FILE *fp); -size_t dsos__fprintf_buildid(FILE *fp); +size_t dsos__fprintf_buildid(FILE *fp, bool with_hits); size_t dso__fprintf_buildid(struct dso *self, FILE *fp); size_t dso__fprintf(struct dso *self, enum map_type type, FILE *fp); char dso__symtab_origin(const struct dso *self); +void dso__set_long_name(struct dso *self, char *name); void dso__set_build_id(struct dso *self, void *build_id); +void dso__read_running_kernel_build_id(struct dso *self); +struct symbol *dso__find_symbol(struct dso *self, enum map_type type, u64 addr); +struct symbol *dso__find_symbol_by_name(struct dso *self, enum map_type type, + const char *name); int filename__read_build_id(const char *filename, void *bf, size_t size); int sysfs__read_build_id(const char *filename, void *bf, size_t size); -bool dsos__read_build_ids(void); -int build_id__sprintf(u8 *self, int len, char *bf); +bool dsos__read_build_ids(bool with_hits); +int build_id__sprintf(const u8 *self, int len, char *bf); +int kallsyms__parse(const char *filename, void *arg, + int (*process_symbol)(void *arg, const char *name, + char type, u64 start)); -size_t kernel_maps__fprintf(FILE *fp); +int symbol__init(void); +bool symbol_type__is_a(char symbol_type, enum map_type map_type); -int symbol__init(struct symbol_conf *conf); - -struct thread; -struct thread *kthread; -extern struct list_head dsos__user, dsos__kernel; -extern struct dso *vdso; #endif /* __PERF_SYMBOL */ diff --git a/tools/perf/util/thread.c b/tools/perf/util/thread.c index 603f5610861b..21b92162282b 100644 --- a/tools/perf/util/thread.c +++ b/tools/perf/util/thread.c @@ -2,18 +2,14 @@ #include <stdlib.h> #include <stdio.h> #include <string.h> +#include "session.h" #include "thread.h" #include "util.h" #include "debug.h" -static struct rb_root threads; -static struct thread *last_match; - -void thread__init(struct thread *self, pid_t pid) +void map_groups__init(struct map_groups *self) { int i; - self->pid = pid; - self->comm = NULL; for (i = 0; i < MAP__NR_TYPES; ++i) { self->maps[i] = RB_ROOT; INIT_LIST_HEAD(&self->removed_maps[i]); @@ -25,7 +21,8 @@ static struct thread *thread__new(pid_t pid) struct thread *self = zalloc(sizeof(*self)); if (self != NULL) { - thread__init(self, pid); + map_groups__init(&self->mg); + self->pid = pid; self->comm = malloc(32); if (self->comm) snprintf(self->comm, 32, ":%d", self->pid); @@ -34,12 +31,41 @@ static struct thread *thread__new(pid_t pid) return self; } +static void map_groups__flush(struct map_groups *self) +{ + int type; + + for (type = 0; type < MAP__NR_TYPES; type++) { + struct rb_root *root = &self->maps[type]; + struct rb_node *next = rb_first(root); + + while (next) { + struct map *pos = rb_entry(next, struct map, rb_node); + next = rb_next(&pos->rb_node); + rb_erase(&pos->rb_node, root); + /* + * We may have references to this map, for + * instance in some hist_entry instances, so + * just move them to a separate list. + */ + list_add_tail(&pos->node, &self->removed_maps[pos->type]); + } + } +} + int thread__set_comm(struct thread *self, const char *comm) { + int err; + if (self->comm) free(self->comm); self->comm = strdup(comm); - return self->comm ? 0 : -ENOMEM; + err = self->comm == NULL ? -ENOMEM : 0; + if (!err) { + self->comm_set = true; + map_groups__flush(&self->mg); + } + return err; } int thread__comm_len(struct thread *self) @@ -53,12 +79,8 @@ int thread__comm_len(struct thread *self) return self->comm_len; } -static const char *map_type__name[MAP__NR_TYPES] = { - [MAP__FUNCTION] = "Functions", -}; - -static size_t __thread__fprintf_maps(struct thread *self, - enum map_type type, FILE *fp) +static size_t __map_groups__fprintf_maps(struct map_groups *self, + enum map_type type, FILE *fp) { size_t printed = fprintf(fp, "%s:\n", map_type__name[type]); struct rb_node *nd; @@ -76,16 +98,16 @@ static size_t __thread__fprintf_maps(struct thread *self, return printed; } -size_t thread__fprintf_maps(struct thread *self, FILE *fp) +size_t map_groups__fprintf_maps(struct map_groups *self, FILE *fp) { size_t printed = 0, i; for (i = 0; i < MAP__NR_TYPES; ++i) - printed += __thread__fprintf_maps(self, i, fp); + printed += __map_groups__fprintf_maps(self, i, fp); return printed; } -static size_t __thread__fprintf_removed_maps(struct thread *self, - enum map_type type, FILE *fp) +static size_t __map_groups__fprintf_removed_maps(struct map_groups *self, + enum map_type type, FILE *fp) { struct map *pos; size_t printed = 0; @@ -101,25 +123,30 @@ static size_t __thread__fprintf_removed_maps(struct thread *self, return printed; } -static size_t thread__fprintf_removed_maps(struct thread *self, FILE *fp) +static size_t map_groups__fprintf_removed_maps(struct map_groups *self, FILE *fp) { size_t printed = 0, i; for (i = 0; i < MAP__NR_TYPES; ++i) - printed += __thread__fprintf_removed_maps(self, i, fp); + printed += __map_groups__fprintf_removed_maps(self, i, fp); return printed; } -static size_t thread__fprintf(struct thread *self, FILE *fp) +static size_t map_groups__fprintf(struct map_groups *self, FILE *fp) { - size_t printed = fprintf(fp, "Thread %d %s\n", self->pid, self->comm); - printed += thread__fprintf_removed_maps(self, fp); + size_t printed = map_groups__fprintf_maps(self, fp); printed += fprintf(fp, "Removed maps:\n"); - return printed + thread__fprintf_removed_maps(self, fp); + return printed + map_groups__fprintf_removed_maps(self, fp); +} + +static size_t thread__fprintf(struct thread *self, FILE *fp) +{ + return fprintf(fp, "Thread %d %s\n", self->pid, self->comm) + + map_groups__fprintf(&self->mg, fp); } -struct thread *threads__findnew(pid_t pid) +struct thread *perf_session__findnew(struct perf_session *self, pid_t pid) { - struct rb_node **p = &threads.rb_node; + struct rb_node **p = &self->threads.rb_node; struct rb_node *parent = NULL; struct thread *th; @@ -128,15 +155,15 @@ struct thread *threads__findnew(pid_t pid) * so most of the time we dont have to look up * the full rbtree: */ - if (last_match && last_match->pid == pid) - return last_match; + if (self->last_match && self->last_match->pid == pid) + return self->last_match; while (*p != NULL) { parent = *p; th = rb_entry(parent, struct thread, rb_node); if (th->pid == pid) { - last_match = th; + self->last_match = th; return th; } @@ -149,26 +176,15 @@ struct thread *threads__findnew(pid_t pid) th = thread__new(pid); if (th != NULL) { rb_link_node(&th->rb_node, parent, p); - rb_insert_color(&th->rb_node, &threads); - last_match = th; + rb_insert_color(&th->rb_node, &self->threads); + self->last_match = th; } return th; } -struct thread *register_idle_thread(void) -{ - struct thread *thread = threads__findnew(0); - - if (!thread || thread__set_comm(thread, "swapper")) { - fprintf(stderr, "problem inserting idle task.\n"); - exit(-1); - } - - return thread; -} - -static void thread__remove_overlappings(struct thread *self, struct map *map) +static void map_groups__remove_overlappings(struct map_groups *self, + struct map *map) { struct rb_root *root = &self->maps[map->type]; struct rb_node *next = rb_first(root); @@ -238,12 +254,15 @@ struct map *maps__find(struct rb_root *maps, u64 ip) void thread__insert_map(struct thread *self, struct map *map) { - thread__remove_overlappings(self, map); - maps__insert(&self->maps[map->type], map); + map_groups__remove_overlappings(&self->mg, map); + map_groups__insert(&self->mg, map); } -static int thread__clone_maps(struct thread *self, struct thread *parent, - enum map_type type) +/* + * XXX This should not really _copy_ te maps, but refcount them. + */ +static int map_groups__clone(struct map_groups *self, + struct map_groups *parent, enum map_type type) { struct rb_node *nd; for (nd = rb_first(&parent->maps[type]); nd; nd = rb_next(nd)) { @@ -251,7 +270,7 @@ static int thread__clone_maps(struct thread *self, struct thread *parent, struct map *new = map__clone(map); if (new == NULL) return -ENOMEM; - thread__insert_map(self, new); + map_groups__insert(self, new); } return 0; } @@ -260,24 +279,27 @@ int thread__fork(struct thread *self, struct thread *parent) { int i; - if (self->comm) - free(self->comm); - self->comm = strdup(parent->comm); - if (!self->comm) - return -ENOMEM; + if (parent->comm_set) { + if (self->comm) + free(self->comm); + self->comm = strdup(parent->comm); + if (!self->comm) + return -ENOMEM; + self->comm_set = true; + } for (i = 0; i < MAP__NR_TYPES; ++i) - if (thread__clone_maps(self, parent, i) < 0) + if (map_groups__clone(&self->mg, &parent->mg, i) < 0) return -ENOMEM; return 0; } -size_t threads__fprintf(FILE *fp) +size_t perf_session__fprintf(struct perf_session *self, FILE *fp) { size_t ret = 0; struct rb_node *nd; - for (nd = rb_first(&threads); nd; nd = rb_next(nd)) { + for (nd = rb_first(&self->threads); nd; nd = rb_next(nd)) { struct thread *pos = rb_entry(nd, struct thread, rb_node); ret += thread__fprintf(pos, fp); @@ -286,11 +308,11 @@ size_t threads__fprintf(FILE *fp) return ret; } -struct symbol *thread__find_symbol(struct thread *self, - enum map_type type, u64 addr, - symbol_filter_t filter) +struct symbol *map_groups__find_symbol(struct map_groups *self, + enum map_type type, u64 addr, + symbol_filter_t filter) { - struct map *map = thread__find_map(self, type, addr); + struct map *map = map_groups__find(self, type, addr); if (map != NULL) return map__find_symbol(map, map->map_ip(map, addr), filter); diff --git a/tools/perf/util/thread.h b/tools/perf/util/thread.h index 686d6e914d9e..0a28f39de545 100644 --- a/tools/perf/util/thread.h +++ b/tools/perf/util/thread.h @@ -5,52 +5,80 @@ #include <unistd.h> #include "symbol.h" -struct thread { - struct rb_node rb_node; +struct map_groups { struct rb_root maps[MAP__NR_TYPES]; struct list_head removed_maps[MAP__NR_TYPES]; +}; + +struct thread { + struct rb_node rb_node; + struct map_groups mg; pid_t pid; - bool use_modules; char shortname[3]; + bool comm_set; char *comm; int comm_len; }; -void thread__init(struct thread *self, pid_t pid); +void map_groups__init(struct map_groups *self); int thread__set_comm(struct thread *self, const char *comm); int thread__comm_len(struct thread *self); -struct thread *threads__findnew(pid_t pid); -struct thread *register_idle_thread(void); +struct thread *perf_session__findnew(struct perf_session *self, pid_t pid); void thread__insert_map(struct thread *self, struct map *map); int thread__fork(struct thread *self, struct thread *parent); -size_t thread__fprintf_maps(struct thread *self, FILE *fp); -size_t threads__fprintf(FILE *fp); +size_t map_groups__fprintf_maps(struct map_groups *self, FILE *fp); +size_t perf_session__fprintf(struct perf_session *self, FILE *fp); void maps__insert(struct rb_root *maps, struct map *map); struct map *maps__find(struct rb_root *maps, u64 addr); -static inline struct map *thread__find_map(struct thread *self, +static inline void map_groups__insert(struct map_groups *self, struct map *map) +{ + maps__insert(&self->maps[map->type], map); +} + +static inline struct map *map_groups__find(struct map_groups *self, enum map_type type, u64 addr) { - return self ? maps__find(&self->maps[type], addr) : NULL; + return maps__find(&self->maps[type], addr); } -static inline void __thread__insert_map(struct thread *self, struct map *map) +static inline struct map *thread__find_map(struct thread *self, + enum map_type type, u64 addr) { - maps__insert(&self->maps[map->type], map); + return self ? map_groups__find(&self->mg, type, addr) : NULL; } -void thread__find_addr_location(struct thread *self, u8 cpumode, +void thread__find_addr_map(struct thread *self, + struct perf_session *session, u8 cpumode, + enum map_type type, u64 addr, + struct addr_location *al); + +void thread__find_addr_location(struct thread *self, + struct perf_session *session, u8 cpumode, enum map_type type, u64 addr, struct addr_location *al, symbol_filter_t filter); -struct symbol *thread__find_symbol(struct thread *self, - enum map_type type, u64 addr, - symbol_filter_t filter); +struct symbol *map_groups__find_symbol(struct map_groups *self, + enum map_type type, u64 addr, + symbol_filter_t filter); -static inline struct symbol * -thread__find_function(struct thread *self, u64 addr, symbol_filter_t filter) +static inline struct symbol *map_groups__find_function(struct map_groups *self, + u64 addr, + symbol_filter_t filter) { - return thread__find_symbol(self, MAP__FUNCTION, addr, filter); + return map_groups__find_symbol(self, MAP__FUNCTION, addr, filter); } + +struct map *map_groups__find_by_name(struct map_groups *self, + enum map_type type, const char *name); + +int __map_groups__create_kernel_maps(struct map_groups *self, + struct map *vmlinux_maps[MAP__NR_TYPES], + struct dso *kernel); +int map_groups__create_kernel_maps(struct map_groups *self, + struct map *vmlinux_maps[MAP__NR_TYPES]); + +struct map *map_groups__new_module(struct map_groups *self, u64 start, + const char *filename); #endif /* __PERF_THREAD_H */ diff --git a/tools/perf/util/trace-event-info.c b/tools/perf/util/trace-event-info.c index cace35595530..5ea8973ad331 100644 --- a/tools/perf/util/trace-event-info.c +++ b/tools/perf/util/trace-event-info.c @@ -20,6 +20,7 @@ */ #define _GNU_SOURCE #include <dirent.h> +#include <mntent.h> #include <stdio.h> #include <stdlib.h> #include <string.h> @@ -37,6 +38,7 @@ #include "../perf.h" #include "trace-event.h" +#include "debugfs.h" #define VERSION "0.5" @@ -101,32 +103,12 @@ void *malloc_or_die(unsigned int size) static const char *find_debugfs(void) { - static char debugfs[MAX_PATH+1]; - static int debugfs_found; - char type[100]; - FILE *fp; - - if (debugfs_found) - return debugfs; - - if ((fp = fopen("/proc/mounts","r")) == NULL) - die("Can't open /proc/mounts for read"); - - while (fscanf(fp, "%*s %" - STR(MAX_PATH) - "s %99s %*s %*d %*d\n", - debugfs, type) == 2) { - if (strcmp(type, "debugfs") == 0) - break; - } - fclose(fp); - - if (strcmp(type, "debugfs") != 0) - die("debugfs not mounted, please mount"); + const char *path = debugfs_mount(NULL); - debugfs_found = 1; + if (!path) + die("Your kernel not support debugfs filesystem"); - return debugfs; + return path; } /* @@ -271,6 +253,8 @@ static void read_header_files(void) write_or_die("header_page", 12); write_or_die(&size, 8); check_size = copy_file_fd(fd); + close(fd); + if (size != check_size) die("wrong size for '%s' size=%lld read=%lld", path, size, check_size); @@ -289,6 +273,7 @@ static void read_header_files(void) if (size != check_size) die("wrong size for '%s'", path); put_tracing_file(path); + close(fd); } static bool name_in_tp_list(char *sys, struct tracepoint_path *tps) @@ -317,7 +302,8 @@ static void copy_event_system(const char *sys, struct tracepoint_path *tps) die("can't read directory '%s'", sys); while ((dent = readdir(dir))) { - if (strcmp(dent->d_name, ".") == 0 || + if (dent->d_type != DT_DIR || + strcmp(dent->d_name, ".") == 0 || strcmp(dent->d_name, "..") == 0 || !name_in_tp_list(dent->d_name, tps)) continue; @@ -334,7 +320,8 @@ static void copy_event_system(const char *sys, struct tracepoint_path *tps) rewinddir(dir); while ((dent = readdir(dir))) { - if (strcmp(dent->d_name, ".") == 0 || + if (dent->d_type != DT_DIR || + strcmp(dent->d_name, ".") == 0 || strcmp(dent->d_name, "..") == 0 || !name_in_tp_list(dent->d_name, tps)) continue; @@ -353,6 +340,7 @@ static void copy_event_system(const char *sys, struct tracepoint_path *tps) free(format); } + closedir(dir); } static void read_ftrace_files(struct tracepoint_path *tps) @@ -394,26 +382,21 @@ static void read_event_files(struct tracepoint_path *tps) die("can't read directory '%s'", path); while ((dent = readdir(dir))) { - if (strcmp(dent->d_name, ".") == 0 || + if (dent->d_type != DT_DIR || + strcmp(dent->d_name, ".") == 0 || strcmp(dent->d_name, "..") == 0 || strcmp(dent->d_name, "ftrace") == 0 || !system_in_tp_list(dent->d_name, tps)) continue; - sys = malloc_or_die(strlen(path) + strlen(dent->d_name) + 2); - sprintf(sys, "%s/%s", path, dent->d_name); - ret = stat(sys, &st); - free(sys); - if (ret < 0) - continue; - if (S_ISDIR(st.st_mode)) - count++; + count++; } write_or_die(&count, 4); rewinddir(dir); while ((dent = readdir(dir))) { - if (strcmp(dent->d_name, ".") == 0 || + if (dent->d_type != DT_DIR || + strcmp(dent->d_name, ".") == 0 || strcmp(dent->d_name, "..") == 0 || strcmp(dent->d_name, "ftrace") == 0 || !system_in_tp_list(dent->d_name, tps)) @@ -422,14 +405,13 @@ static void read_event_files(struct tracepoint_path *tps) sprintf(sys, "%s/%s", path, dent->d_name); ret = stat(sys, &st); if (ret >= 0) { - if (S_ISDIR(st.st_mode)) { - write_or_die(dent->d_name, strlen(dent->d_name) + 1); - copy_event_system(sys, tps); - } + write_or_die(dent->d_name, strlen(dent->d_name) + 1); + copy_event_system(sys, tps); } free(sys); } + closedir(dir); put_tracing_file(path); } @@ -533,7 +515,7 @@ int read_tracing_data(int fd, struct perf_event_attr *pattrs, int nb_events) write_or_die(buf, 1); /* save page_size */ - page_size = getpagesize(); + page_size = sysconf(_SC_PAGESIZE); write_or_die(&page_size, 4); read_header_files(); diff --git a/tools/perf/util/trace-event-parse.c b/tools/perf/util/trace-event-parse.c index 0302405aa2ca..9b3c20f42f98 100644 --- a/tools/perf/util/trace-event-parse.c +++ b/tools/perf/util/trace-event-parse.c @@ -177,7 +177,7 @@ void parse_proc_kallsyms(char *file, unsigned int size __unused) func_count++; } - func_list = malloc_or_die(sizeof(*func_list) * func_count + 1); + func_list = malloc_or_die(sizeof(*func_list) * (func_count + 1)); i = 0; while (list) { @@ -1477,7 +1477,7 @@ process_fields(struct event *event, struct print_flag_sym **list, char **tok) goto out_free; field = malloc_or_die(sizeof(*field)); - memset(field, 0, sizeof(field)); + memset(field, 0, sizeof(*field)); value = arg_eval(arg); field->value = strdup(value); @@ -1925,6 +1925,15 @@ void *raw_field_ptr(struct event *event, const char *name, void *data) if (!field) return NULL; + if (field->flags & FIELD_IS_STRING) { + int offset; + + offset = *(int *)(data + field->offset); + offset &= 0xffff; + + return data + offset; + } + return data + field->offset; } @@ -3277,3 +3286,18 @@ void parse_set_info(int nr_cpus, int long_sz) cpus = nr_cpus; long_size = long_sz; } + +int common_pc(struct scripting_context *context) +{ + return parse_common_pc(context->event_data); +} + +int common_flags(struct scripting_context *context) +{ + return parse_common_flags(context->event_data); +} + +int common_lock_depth(struct scripting_context *context) +{ + return parse_common_lock_depth(context->event_data); +} diff --git a/tools/perf/util/trace-event-perl.h b/tools/perf/util/trace-event-perl.h deleted file mode 100644 index 8fe0d866fe1a..000000000000 --- a/tools/perf/util/trace-event-perl.h +++ /dev/null @@ -1,51 +0,0 @@ -#ifndef __PERF_TRACE_EVENT_PERL_H -#define __PERF_TRACE_EVENT_PERL_H -#ifdef NO_LIBPERL -typedef int INTERP; -#define dSP -#define ENTER -#define SAVETMPS -#define PUTBACK -#define SPAGAIN -#define FREETMPS -#define LEAVE -#define SP -#define ERRSV -#define G_SCALAR (0) -#define G_DISCARD (0) -#define G_NOARGS (0) -#define PUSHMARK(a) -#define SvTRUE(a) (0) -#define XPUSHs(s) -#define sv_2mortal(a) -#define newSVpv(a,b) -#define newSVuv(a) -#define newSViv(a) -#define get_cv(a,b) (0) -#define call_pv(a,b) (0) -#define perl_alloc() (0) -#define perl_construct(a) (0) -#define perl_parse(a,b,c,d,e) (0) -#define perl_run(a) (0) -#define perl_destruct(a) (0) -#define perl_free(a) (0) -#define pTHX void -#define CV void -#define dXSUB_SYS -#define pTHX_ -static inline void newXS(const char *a, void *b, const char *c) {} -#else -#include <EXTERN.h> -#include <perl.h> -typedef PerlInterpreter * INTERP; -#endif - -struct scripting_context { - void *event_data; -}; - -int common_pc(struct scripting_context *context); -int common_flags(struct scripting_context *context); -int common_lock_depth(struct scripting_context *context); - -#endif /* __PERF_TRACE_EVENT_PERL_H */ diff --git a/tools/perf/util/trace-event-read.c b/tools/perf/util/trace-event-read.c index 342dfdd43f87..7cd1193918c7 100644 --- a/tools/perf/util/trace-event-read.c +++ b/tools/perf/util/trace-event-read.c @@ -18,7 +18,7 @@ * * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ */ -#define _LARGEFILE64_SOURCE +#define _FILE_OFFSET_BITS 64 #include <dirent.h> #include <stdio.h> @@ -83,7 +83,7 @@ static char *read_string(void) char *str = NULL; int size = 0; int i; - int r; + off_t r; for (;;) { r = read(input_fd, buf, BUFSIZ); @@ -118,7 +118,7 @@ static char *read_string(void) /* move the file descriptor to the end of the string */ r = lseek(input_fd, -(r - i), SEEK_CUR); - if (r < 0) + if (r == (off_t)-1) die("lseek"); if (str) { @@ -145,8 +145,9 @@ static void read_proc_kallsyms(void) if (!size) return; - buf = malloc_or_die(size); + buf = malloc_or_die(size + 1); read_or_die(buf, size); + buf[size] = '\0'; parse_proc_kallsyms(buf, size); @@ -281,8 +282,8 @@ static void update_cpu_data_index(int cpu) static void get_next_page(int cpu) { - off64_t save_seek; - off64_t ret; + off_t save_seek; + off_t ret; if (!cpu_data[cpu].page) return; @@ -297,17 +298,17 @@ static void get_next_page(int cpu) update_cpu_data_index(cpu); /* other parts of the code may expect the pointer to not move */ - save_seek = lseek64(input_fd, 0, SEEK_CUR); + save_seek = lseek(input_fd, 0, SEEK_CUR); - ret = lseek64(input_fd, cpu_data[cpu].offset, SEEK_SET); - if (ret < 0) + ret = lseek(input_fd, cpu_data[cpu].offset, SEEK_SET); + if (ret == (off_t)-1) die("failed to lseek"); ret = read(input_fd, cpu_data[cpu].page, page_size); if (ret < 0) die("failed to read page"); /* reset the file pointer back */ - lseek64(input_fd, save_seek, SEEK_SET); + lseek(input_fd, save_seek, SEEK_SET); return; } diff --git a/tools/perf/util/trace-event-scripting.c b/tools/perf/util/trace-event-scripting.c new file mode 100644 index 000000000000..7ea983acfaea --- /dev/null +++ b/tools/perf/util/trace-event-scripting.c @@ -0,0 +1,167 @@ +/* + * trace-event-scripting. Scripting engine common and initialization code. + * + * Copyright (C) 2009-2010 Tom Zanussi <tzanussi@gmail.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + */ + +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <ctype.h> +#include <errno.h> + +#include "../perf.h" +#include "util.h" +#include "trace-event.h" + +struct scripting_context *scripting_context; + +static int stop_script_unsupported(void) +{ + return 0; +} + +static void process_event_unsupported(int cpu __unused, + void *data __unused, + int size __unused, + unsigned long long nsecs __unused, + char *comm __unused) +{ +} + +static void print_python_unsupported_msg(void) +{ + fprintf(stderr, "Python scripting not supported." + " Install libpython and rebuild perf to enable it.\n" + "For example:\n # apt-get install python-dev (ubuntu)" + "\n # yum install python-devel (Fedora)" + "\n etc.\n"); +} + +static int python_start_script_unsupported(const char *script __unused, + int argc __unused, + const char **argv __unused) +{ + print_python_unsupported_msg(); + + return -1; +} + +static int python_generate_script_unsupported(const char *outfile __unused) +{ + print_python_unsupported_msg(); + + return -1; +} + +struct scripting_ops python_scripting_unsupported_ops = { + .name = "Python", + .start_script = python_start_script_unsupported, + .stop_script = stop_script_unsupported, + .process_event = process_event_unsupported, + .generate_script = python_generate_script_unsupported, +}; + +static void register_python_scripting(struct scripting_ops *scripting_ops) +{ + int err; + err = script_spec_register("Python", scripting_ops); + if (err) + die("error registering Python script extension"); + + err = script_spec_register("py", scripting_ops); + if (err) + die("error registering py script extension"); + + scripting_context = malloc(sizeof(struct scripting_context)); +} + +#ifdef NO_LIBPYTHON +void setup_python_scripting(void) +{ + register_python_scripting(&python_scripting_unsupported_ops); +} +#else +struct scripting_ops python_scripting_ops; + +void setup_python_scripting(void) +{ + register_python_scripting(&python_scripting_ops); +} +#endif + +static void print_perl_unsupported_msg(void) +{ + fprintf(stderr, "Perl scripting not supported." + " Install libperl and rebuild perf to enable it.\n" + "For example:\n # apt-get install libperl-dev (ubuntu)" + "\n # yum install 'perl(ExtUtils::Embed)' (Fedora)" + "\n etc.\n"); +} + +static int perl_start_script_unsupported(const char *script __unused, + int argc __unused, + const char **argv __unused) +{ + print_perl_unsupported_msg(); + + return -1; +} + +static int perl_generate_script_unsupported(const char *outfile __unused) +{ + print_perl_unsupported_msg(); + + return -1; +} + +struct scripting_ops perl_scripting_unsupported_ops = { + .name = "Perl", + .start_script = perl_start_script_unsupported, + .stop_script = stop_script_unsupported, + .process_event = process_event_unsupported, + .generate_script = perl_generate_script_unsupported, +}; + +static void register_perl_scripting(struct scripting_ops *scripting_ops) +{ + int err; + err = script_spec_register("Perl", scripting_ops); + if (err) + die("error registering Perl script extension"); + + err = script_spec_register("pl", scripting_ops); + if (err) + die("error registering pl script extension"); + + scripting_context = malloc(sizeof(struct scripting_context)); +} + +#ifdef NO_LIBPERL +void setup_perl_scripting(void) +{ + register_perl_scripting(&perl_scripting_unsupported_ops); +} +#else +struct scripting_ops perl_scripting_ops; + +void setup_perl_scripting(void) +{ + register_perl_scripting(&perl_scripting_ops); +} +#endif diff --git a/tools/perf/util/trace-event.h b/tools/perf/util/trace-event.h index 81698d5e6503..c3269b937db4 100644 --- a/tools/perf/util/trace-event.h +++ b/tools/perf/util/trace-event.h @@ -270,7 +270,7 @@ enum trace_flag_type { struct scripting_ops { const char *name; - int (*start_script) (const char *); + int (*start_script) (const char *script, int argc, const char **argv); int (*stop_script) (void); void (*process_event) (int cpu, void *data, int size, unsigned long long nsecs, char *comm); @@ -279,7 +279,15 @@ struct scripting_ops { int script_spec_register(const char *spec, struct scripting_ops *ops); -extern struct scripting_ops perl_scripting_ops; void setup_perl_scripting(void); +void setup_python_scripting(void); + +struct scripting_context { + void *event_data; +}; + +int common_pc(struct scripting_context *context); +int common_flags(struct scripting_context *context); +int common_lock_depth(struct scripting_context *context); #endif /* __PERF_TRACE_EVENTS_H */ diff --git a/tools/perf/util/util.c b/tools/perf/util/util.c new file mode 100644 index 000000000000..f9b890fde681 --- /dev/null +++ b/tools/perf/util/util.c @@ -0,0 +1,94 @@ +#include "util.h" +#include <sys/mman.h> + +int mkdir_p(char *path, mode_t mode) +{ + struct stat st; + int err; + char *d = path; + + if (*d != '/') + return -1; + + if (stat(path, &st) == 0) + return 0; + + while (*++d == '/'); + + while ((d = strchr(d, '/'))) { + *d = '\0'; + err = stat(path, &st) && mkdir(path, mode); + *d++ = '/'; + if (err) + return -1; + while (*d == '/') + ++d; + } + return (stat(path, &st) && mkdir(path, mode)) ? -1 : 0; +} + +static int slow_copyfile(const char *from, const char *to) +{ + int err = 0; + char *line = NULL; + size_t n; + FILE *from_fp = fopen(from, "r"), *to_fp; + + if (from_fp == NULL) + goto out; + + to_fp = fopen(to, "w"); + if (to_fp == NULL) + goto out_fclose_from; + + while (getline(&line, &n, from_fp) > 0) + if (fputs(line, to_fp) == EOF) + goto out_fclose_to; + err = 0; +out_fclose_to: + fclose(to_fp); + free(line); +out_fclose_from: + fclose(from_fp); +out: + return err; +} + +int copyfile(const char *from, const char *to) +{ + int fromfd, tofd; + struct stat st; + void *addr; + int err = -1; + + if (stat(from, &st)) + goto out; + + if (st.st_size == 0) /* /proc? do it slowly... */ + return slow_copyfile(from, to); + + fromfd = open(from, O_RDONLY); + if (fromfd < 0) + goto out; + + tofd = creat(to, 0755); + if (tofd < 0) + goto out_close_from; + + addr = mmap(NULL, st.st_size, PROT_READ, MAP_PRIVATE, fromfd, 0); + if (addr == MAP_FAILED) + goto out_close_to; + + if (write(tofd, addr, st.st_size) == st.st_size) + err = 0; + + munmap(addr, st.st_size); +out_close_to: + close(tofd); + if (err) + unlink(to); +out_close_from: + close(fromfd); +out: + return err; +} diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h index c673d8825883..0f5b2a6f1080 100644 --- a/tools/perf/util/util.h +++ b/tools/perf/util/util.h @@ -403,4 +403,7 @@ void git_qsort(void *base, size_t nmemb, size_t size, #endif #endif +int mkdir_p(char *path, mode_t mode); +int copyfile(const char *from, const char *to); + #endif diff --git a/tools/perf/util/values.c b/tools/perf/util/values.c index 1c15e39f99e3..cfa55d686e3b 100644 --- a/tools/perf/util/values.c +++ b/tools/perf/util/values.c @@ -169,6 +169,7 @@ static void perf_read_values__display_pretty(FILE *fp, counterwidth[j], values->value[i][j]); fprintf(fp, "\n"); } + free(counterwidth); } static void perf_read_values__display_raw(FILE *fp, |