Re: [PATCH] perf record: Add support to collect callchains from kernel or user space only.

[Arnaldo Carvalho de Melo <arnaldo.melo@gmail.com>]

Em Thu, Jun 06, 2019 at 11:26:44AM -0300, Arnaldo Carvalho de Melo escreveu:
> Em Thu, May 30, 2019 at 02:29:22PM +0100, ufo19890607 escreveu:
> > From: yuzhoujian <yuzhoujian@didichuxing.com>
> > 
> > One can just record callchains in the kernel or user space with
> > this new options. We can use it together with "--all-kernel" options.
> > This two options is used just like print_stack(sys) or print_ustack(usr)
> > for systemtap.
> > 
> > Show below is the usage of this new option combined with "--all-kernel"
> > options.
> > 	1. Configure all used events to run in kernel space and just
> > collect kernel callchains.
> > 	$ perf record -a -g --all-kernel --kernel-callchains
> > 	2. Configure all used events to run in kernel space and just
> > collect user callchains.
> > 	$ perf record -a -g --all-kernel --user-callchains
> > 
> > Signed-off-by: yuzhoujian <yuzhoujian@didichuxing.com>
> > ---
> >  tools/perf/Documentation/perf-record.txt | 6 ++++++
> >  tools/perf/builtin-record.c              | 4 ++++
> >  tools/perf/perf.h                        | 2 ++
> >  tools/perf/util/evsel.c                  | 4 ++++
> >  4 files changed, 16 insertions(+)
> > 
> > diff --git a/tools/perf/Documentation/perf-record.txt b/tools/perf/Documentation/perf-record.txt
> > index de269430720a..b647eb3db0c6 100644
> > --- a/tools/perf/Documentation/perf-record.txt
> > +++ b/tools/perf/Documentation/perf-record.txt
> > @@ -490,6 +490,12 @@ Configure all used events to run in kernel space.
> >  --all-user::
> >  Configure all used events to run in user space.
> >  
> > +--kernel-callchains::
> > +Collect callchains from kernel space.
> 
> Ok, changing this to:
> 
> Collect callchains only from kernel space. I.e. this option sets
> perf_event_attr.exclude_callchain_user to 1,
> perf_event_attr.exclude_callchain_kernel to 0.
> 
> > +
> > +--user-callchains::
> > +Collect callchains from user space.
> 
> And this one to:
> Collect callchains only from user space. I.e. this option sets
> 
> perf_event_attr.exclude_callchain_kernel to 1,
> perf_event_attr.exclude_callchain_user to 0.

Yeah, each of this options just sets the exclude bit for the undesired
callchains, not setting 0 for the desired, so I'm fixing up the doc I
suggested accordingly, my comment below remains valid tho:

> 
> So that the user don't try using:
> 
>     pref record --user-callchains --kernel-callchains
> 
> expecting to get both user and kernel callchains and instead gets
> nothing.
> 
> Ok?
> 
> - Arnaldo
> 
> > +
> >  --timestamp-filename
> >  Append timestamp to output file name.
> >  
> > diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
> > index e2c3a585a61e..dca55997934e 100644
> > --- a/tools/perf/builtin-record.c
> > +++ b/tools/perf/builtin-record.c
> > @@ -2191,6 +2191,10 @@ static struct option __record_options[] = {
> >  	OPT_BOOLEAN_FLAG(0, "all-user", &record.opts.all_user,
> >  			 "Configure all used events to run in user space.",
> >  			 PARSE_OPT_EXCLUSIVE),
> > +	OPT_BOOLEAN(0, "kernel-callchains", &record.opts.kernel_callchains,
> > +		    "collect kernel callchains"),
> > +	OPT_BOOLEAN(0, "user-callchains", &record.opts.user_callchains,
> > +		    "collect user callchains"),
> >  	OPT_STRING(0, "clang-path", &llvm_param.clang_path, "clang path",
> >  		   "clang binary to use for compiling BPF scriptlets"),
> >  	OPT_STRING(0, "clang-opt", &llvm_param.clang_opt, "clang options",
> > diff --git a/tools/perf/perf.h b/tools/perf/perf.h
> > index d59dee61b64d..711e009381ec 100644
> > --- a/tools/perf/perf.h
> > +++ b/tools/perf/perf.h
> > @@ -61,6 +61,8 @@ struct record_opts {
> >  	bool	     record_switch_events;
> >  	bool	     all_kernel;
> >  	bool	     all_user;
> > +	bool	     kernel_callchains;
> > +	bool	     user_callchains;
> >  	bool	     tail_synthesize;
> >  	bool	     overwrite;
> >  	bool	     ignore_missing_thread;
> > diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c
> > index a6f572a40deb..a606b2833e27 100644
> > --- a/tools/perf/util/evsel.c
> > +++ b/tools/perf/util/evsel.c
> > @@ -680,6 +680,10 @@ static void __perf_evsel__config_callchain(struct perf_evsel *evsel,
> >  
> >  	attr->sample_max_stack = param->max_stack;
> >  
> > +	if (opts->kernel_callchains)
> > +		attr->exclude_callchain_user = 1;
> > +	if (opts->user_callchains)
> > +		attr->exclude_callchain_kernel = 1;
> >  	if (param->record_mode == CALLCHAIN_LBR) {
> >  		if (!opts->branch_stack) {
> >  			if (attr->exclude_user) {
> > -- 
> > 2.14.1
> 
> -- 
> 
> - Arnaldo

-- 

- Arnaldo