diff mbox series

[trace-cmd] trace-cmd report: Ensure filter is applied to single input file

Message ID 20230317212430.12675-1-krisman@suse.de (mailing list archive)
State Superseded
Headers show
Series [trace-cmd] trace-cmd report: Ensure filter is applied to single input file | expand

Commit Message

Gabriel Krisman Bertazi March 17, 2023, 9:24 p.m. UTC
Since 955d05fc7aee ("trace-cmd report: Make filter arguments match their
files"), the -F filtering is silently ignored when a trace file is
provided with -i and the filter comes after -i .  The reason is that the
filter is now associated with input_files and not saved to the global
list only in this case, but process_filters still only checks the global
list when handles->input_file is not set.

Avoid this by checking last_input_file first, which always contains a
pointer to the correct filter in this case.

This was only lightly tested, using a single trace file.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=217038
Fixes: 955d05fc7aee ("trace-cmd report: Make filter arguments match their files")
Signed-off-by: Gabriel krisman Bertazi <krisman@suse.de>
---
 tracecmd/trace-read.c | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)

Comments

Steven Rostedt March 27, 2023, 10:11 p.m. UTC | #1
On Fri, 17 Mar 2023 18:24:30 -0300
Gabriel Krisman Bertazi <krisman@suse.de> wrote:

> Since 955d05fc7aee ("trace-cmd report: Make filter arguments match their
> files"), the -F filtering is silently ignored when a trace file is
> provided with -i and the filter comes after -i .  The reason is that the
> filter is now associated with input_files and not saved to the global
> list only in this case, but process_filters still only checks the global
> list when handles->input_file is not set.
> 
> Avoid this by checking last_input_file first, which always contains a
> pointer to the correct filter in this case.
> 
> This was only lightly tested, using a single trace file.
> 
> Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=217038
> Fixes: 955d05fc7aee ("trace-cmd report: Make filter arguments match their files")
> Signed-off-by: Gabriel krisman Bertazi <krisman@suse.de>
> ---
>  tracecmd/trace-read.c | 10 +++++++---
>  1 file changed, 7 insertions(+), 3 deletions(-)
> 
> diff --git a/tracecmd/trace-read.c b/tracecmd/trace-read.c
> index 52ba818..dbbd124 100644
> --- a/tracecmd/trace-read.c
> +++ b/tracecmd/trace-read.c
> @@ -567,11 +567,15 @@ static void process_filters(struct handle_list *handles)
>  	struct tracecmd_filter *trace_filter;
>  	struct filter_str *filter;
>  	int filters = 0;
> +	struct input_files *input_file = handles->input_file;

BTW, please keep the "upside-down x-mas tree" formatting of the above.

ie.

static void process_filters(struct handle_list *handles)
{
        struct input_files *input_file = handles->input_file;
        struct tracecmd_filter *trace_filter;
        struct filter_str *filter;
        int filters = 0;


>  
> -	make_pid_filter(handles->handle, handles->input_file);
> +	if (!input_file)
> +		input_file = last_input_file;
>  
> -	if (handles->input_file)
> -		filter = handles->input_file->filter_str;
> +	make_pid_filter(handles->handle, input_file);
> +
> +	if (input_file)
> +		filter = input_file->filter_str;
>  	else
>  		filter = filter_strings;
>  

So I was playing with this, and I actually needed something else.

If an -F comes before *any* file, I think it should be added to *all* files.

That is:

 -F sched_switch -i trace1.dat -i trace2.dat

Should filter sched_switch for both trace1.dat and trace2.dat.

I actually wanted this just a few minutes ago ;-)

Can you do this, or would you want me to?

-- Steve
Gabriel Krisman Bertazi March 28, 2023, 2:30 p.m. UTC | #2
Steven Rostedt <rostedt@goodmis.org> writes:

> On Fri, 17 Mar 2023 18:24:30 -0300
> Gabriel Krisman Bertazi <krisman@suse.de> wrote:
>
>> Since 955d05fc7aee ("trace-cmd report: Make filter arguments match their
>> files"), the -F filtering is silently ignored when a trace file is
>> provided with -i and the filter comes after -i .  The reason is that the
>> filter is now associated with input_files and not saved to the global
>> list only in this case, but process_filters still only checks the global
>> list when handles->input_file is not set.
>> 
>> Avoid this by checking last_input_file first, which always contains a
>> pointer to the correct filter in this case.
>> 
>> This was only lightly tested, using a single trace file.
>> 
>> Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=217038
>> Fixes: 955d05fc7aee ("trace-cmd report: Make filter arguments match their files")
>> Signed-off-by: Gabriel krisman Bertazi <krisman@suse.de>
>> ---
>>  tracecmd/trace-read.c | 10 +++++++---
>>  1 file changed, 7 insertions(+), 3 deletions(-)
>> 
>> diff --git a/tracecmd/trace-read.c b/tracecmd/trace-read.c
>> index 52ba818..dbbd124 100644
>> --- a/tracecmd/trace-read.c
>> +++ b/tracecmd/trace-read.c
>> @@ -567,11 +567,15 @@ static void process_filters(struct handle_list *handles)
>>  	struct tracecmd_filter *trace_filter;
>>  	struct filter_str *filter;
>>  	int filters = 0;
>> +	struct input_files *input_file = handles->input_file;
>
> BTW, please keep the "upside-down x-mas tree" formatting of the above.
>
> ie.

Hey Steve,

thanks for the review!

>
> static void process_filters(struct handle_list *handles)
> {
>         struct input_files *input_file = handles->input_file;
>         struct tracecmd_filter *trace_filter;
>         struct filter_str *filter;
>         int filters = 0;
>
>
>>  
>> -	make_pid_filter(handles->handle, handles->input_file);
>> +	if (!input_file)
>> +		input_file = last_input_file;
>>  
>> -	if (handles->input_file)
>> -		filter = handles->input_file->filter_str;
>> +	make_pid_filter(handles->handle, input_file);
>> +
>> +	if (input_file)
>> +		filter = input_file->filter_str;
>>  	else
>>  		filter = filter_strings;
>>  
>
> So I was playing with this, and I actually needed something else.
>
> If an -F comes before *any* file, I think it should be added to *all* files.
>
> That is:
>
>  -F sched_switch -i trace1.dat -i trace2.dat
>
> Should filter sched_switch for both trace1.dat and trace2.dat.
>
> I actually wanted this just a few minutes ago ;-)
>
> Can you do this, or would you want me to?

sure, i can do that.

Unless you are in a hurry, let me take a look later this week.
Steven Rostedt March 28, 2023, 2:44 p.m. UTC | #3
On Tue, 28 Mar 2023 11:30:03 -0300
Gabriel Krisman Bertazi <krisman@suse.de> wrote:

> > I actually wanted this just a few minutes ago ;-)
> >
> > Can you do this, or would you want me to?  
> 
> sure, i can do that.
> 
> Unless you are in a hurry, let me take a look later this week.

No hurry. Thanks.

-- Steve
diff mbox series

Patch

diff --git a/tracecmd/trace-read.c b/tracecmd/trace-read.c
index 52ba818..dbbd124 100644
--- a/tracecmd/trace-read.c
+++ b/tracecmd/trace-read.c
@@ -567,11 +567,15 @@  static void process_filters(struct handle_list *handles)
 	struct tracecmd_filter *trace_filter;
 	struct filter_str *filter;
 	int filters = 0;
+	struct input_files *input_file = handles->input_file;
 
-	make_pid_filter(handles->handle, handles->input_file);
+	if (!input_file)
+		input_file = last_input_file;
 
-	if (handles->input_file)
-		filter = handles->input_file->filter_str;
+	make_pid_filter(handles->handle, input_file);
+
+	if (input_file)
+		filter = input_file->filter_str;
 	else
 		filter = filter_strings;