linux/kernel/trace/trace_events_filter.c

434 lines
8.5 KiB
C
Raw Normal View History

tracing: add per-event filtering This patch adds per-event filtering to the event tracing subsystem. It adds a 'filter' debugfs file to each event directory. This file can be written to to set filters; reading from it will display the current set of filters set for that event. Basically, any field listed in the 'format' file for an event can be filtered on (including strings, but not yet other array types) using either matching ('==') or non-matching ('!=') 'predicates'. A 'predicate' can be either a single expression: # echo pid != 0 > filter # cat filter pid != 0 or a compound expression of up to 8 sub-expressions combined using '&&' or '||': # echo comm == Xorg > filter # echo "&& sig != 29" > filter # cat filter comm == Xorg && sig != 29 Only events having field values matching an expression will be available in the trace output; non-matching events are discarded. Note that a compound expression is built up by echoing each sub-expression separately - it's not the most efficient way to do things, but it keeps the parser simple and assumes that compound expressions will be relatively uncommon. In any case, a subsequent patch introducing a way to set filters for entire subsystems should mitigate any need to do this for lots of events. Setting a filter without an '&&' or '||' clears the previous filter completely and sets the filter to the new expression: # cat filter comm == Xorg && sig != 29 # echo comm != Xorg # cat filter comm != Xorg To clear a filter, echo 0 to the filter file: # echo 0 > filter # cat filter none The limit of 8 predicates for a compound expression is arbitrary - for efficiency, it's implemented as an array of pointers to predicates, and 8 seemed more than enough for any filter... Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <1237710665.7703.48.camel@charm-linux> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-22 08:31:04 +00:00
/*
* trace_events_filter - generic event filtering
*
* 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.
*
* Copyright (C) 2009 Tom Zanussi <tzanussi@gmail.com>
*/
#include <linux/debugfs.h>
#include <linux/uaccess.h>
#include <linux/module.h>
#include <linux/ctype.h>
#include "trace.h"
#include "trace_output.h"
tracing: add per-event filtering This patch adds per-event filtering to the event tracing subsystem. It adds a 'filter' debugfs file to each event directory. This file can be written to to set filters; reading from it will display the current set of filters set for that event. Basically, any field listed in the 'format' file for an event can be filtered on (including strings, but not yet other array types) using either matching ('==') or non-matching ('!=') 'predicates'. A 'predicate' can be either a single expression: # echo pid != 0 > filter # cat filter pid != 0 or a compound expression of up to 8 sub-expressions combined using '&&' or '||': # echo comm == Xorg > filter # echo "&& sig != 29" > filter # cat filter comm == Xorg && sig != 29 Only events having field values matching an expression will be available in the trace output; non-matching events are discarded. Note that a compound expression is built up by echoing each sub-expression separately - it's not the most efficient way to do things, but it keeps the parser simple and assumes that compound expressions will be relatively uncommon. In any case, a subsequent patch introducing a way to set filters for entire subsystems should mitigate any need to do this for lots of events. Setting a filter without an '&&' or '||' clears the previous filter completely and sets the filter to the new expression: # cat filter comm == Xorg && sig != 29 # echo comm != Xorg # cat filter comm != Xorg To clear a filter, echo 0 to the filter file: # echo 0 > filter # cat filter none The limit of 8 predicates for a compound expression is arbitrary - for efficiency, it's implemented as an array of pointers to predicates, and 8 seemed more than enough for any filter... Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <1237710665.7703.48.camel@charm-linux> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-22 08:31:04 +00:00
static int filter_pred_64(struct filter_pred *pred, void *event)
{
u64 *addr = (u64 *)(event + pred->offset);
u64 val = (u64)pred->val;
int match;
match = (val == *addr) ^ pred->not;
return match;
}
static int filter_pred_32(struct filter_pred *pred, void *event)
{
u32 *addr = (u32 *)(event + pred->offset);
u32 val = (u32)pred->val;
int match;
match = (val == *addr) ^ pred->not;
return match;
}
static int filter_pred_16(struct filter_pred *pred, void *event)
{
u16 *addr = (u16 *)(event + pred->offset);
u16 val = (u16)pred->val;
int match;
match = (val == *addr) ^ pred->not;
return match;
}
static int filter_pred_8(struct filter_pred *pred, void *event)
{
u8 *addr = (u8 *)(event + pred->offset);
u8 val = (u8)pred->val;
int match;
match = (val == *addr) ^ pred->not;
return match;
}
static int filter_pred_string(struct filter_pred *pred, void *event)
{
char *addr = (char *)(event + pred->offset);
int cmp, match;
cmp = strncmp(addr, pred->str_val, pred->str_len);
match = (!cmp) ^ pred->not;
return match;
}
/* return 1 if event matches, 0 otherwise (discard) */
int filter_match_preds(struct ftrace_event_call *call, void *rec)
{
int i, matched, and_failed = 0;
struct filter_pred *pred;
for (i = 0; i < MAX_FILTER_PRED; i++) {
if (call->preds[i]) {
pred = call->preds[i];
if (and_failed && !pred->or)
continue;
matched = pred->fn(pred, rec);
if (!matched && !pred->or) {
and_failed = 1;
continue;
} else if (matched && pred->or)
return 1;
} else
break;
}
if (and_failed)
return 0;
return 1;
}
void filter_print_preds(struct filter_pred **preds, struct trace_seq *s)
tracing: add per-event filtering This patch adds per-event filtering to the event tracing subsystem. It adds a 'filter' debugfs file to each event directory. This file can be written to to set filters; reading from it will display the current set of filters set for that event. Basically, any field listed in the 'format' file for an event can be filtered on (including strings, but not yet other array types) using either matching ('==') or non-matching ('!=') 'predicates'. A 'predicate' can be either a single expression: # echo pid != 0 > filter # cat filter pid != 0 or a compound expression of up to 8 sub-expressions combined using '&&' or '||': # echo comm == Xorg > filter # echo "&& sig != 29" > filter # cat filter comm == Xorg && sig != 29 Only events having field values matching an expression will be available in the trace output; non-matching events are discarded. Note that a compound expression is built up by echoing each sub-expression separately - it's not the most efficient way to do things, but it keeps the parser simple and assumes that compound expressions will be relatively uncommon. In any case, a subsequent patch introducing a way to set filters for entire subsystems should mitigate any need to do this for lots of events. Setting a filter without an '&&' or '||' clears the previous filter completely and sets the filter to the new expression: # cat filter comm == Xorg && sig != 29 # echo comm != Xorg # cat filter comm != Xorg To clear a filter, echo 0 to the filter file: # echo 0 > filter # cat filter none The limit of 8 predicates for a compound expression is arbitrary - for efficiency, it's implemented as an array of pointers to predicates, and 8 seemed more than enough for any filter... Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <1237710665.7703.48.camel@charm-linux> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-22 08:31:04 +00:00
{
char *field_name;
struct filter_pred *pred;
int i;
if (!preds) {
trace_seq_printf(s, "none\n");
return;
tracing: add per-event filtering This patch adds per-event filtering to the event tracing subsystem. It adds a 'filter' debugfs file to each event directory. This file can be written to to set filters; reading from it will display the current set of filters set for that event. Basically, any field listed in the 'format' file for an event can be filtered on (including strings, but not yet other array types) using either matching ('==') or non-matching ('!=') 'predicates'. A 'predicate' can be either a single expression: # echo pid != 0 > filter # cat filter pid != 0 or a compound expression of up to 8 sub-expressions combined using '&&' or '||': # echo comm == Xorg > filter # echo "&& sig != 29" > filter # cat filter comm == Xorg && sig != 29 Only events having field values matching an expression will be available in the trace output; non-matching events are discarded. Note that a compound expression is built up by echoing each sub-expression separately - it's not the most efficient way to do things, but it keeps the parser simple and assumes that compound expressions will be relatively uncommon. In any case, a subsequent patch introducing a way to set filters for entire subsystems should mitigate any need to do this for lots of events. Setting a filter without an '&&' or '||' clears the previous filter completely and sets the filter to the new expression: # cat filter comm == Xorg && sig != 29 # echo comm != Xorg # cat filter comm != Xorg To clear a filter, echo 0 to the filter file: # echo 0 > filter # cat filter none The limit of 8 predicates for a compound expression is arbitrary - for efficiency, it's implemented as an array of pointers to predicates, and 8 seemed more than enough for any filter... Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <1237710665.7703.48.camel@charm-linux> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-22 08:31:04 +00:00
}
for (i = 0; i < MAX_FILTER_PRED; i++) {
if (preds[i]) {
pred = preds[i];
field_name = pred->field_name;
if (i)
trace_seq_printf(s, pred->or ? "|| " : "&& ");
trace_seq_printf(s, "%s ", field_name);
trace_seq_printf(s, pred->not ? "!= " : "== ");
tracing: add per-event filtering This patch adds per-event filtering to the event tracing subsystem. It adds a 'filter' debugfs file to each event directory. This file can be written to to set filters; reading from it will display the current set of filters set for that event. Basically, any field listed in the 'format' file for an event can be filtered on (including strings, but not yet other array types) using either matching ('==') or non-matching ('!=') 'predicates'. A 'predicate' can be either a single expression: # echo pid != 0 > filter # cat filter pid != 0 or a compound expression of up to 8 sub-expressions combined using '&&' or '||': # echo comm == Xorg > filter # echo "&& sig != 29" > filter # cat filter comm == Xorg && sig != 29 Only events having field values matching an expression will be available in the trace output; non-matching events are discarded. Note that a compound expression is built up by echoing each sub-expression separately - it's not the most efficient way to do things, but it keeps the parser simple and assumes that compound expressions will be relatively uncommon. In any case, a subsequent patch introducing a way to set filters for entire subsystems should mitigate any need to do this for lots of events. Setting a filter without an '&&' or '||' clears the previous filter completely and sets the filter to the new expression: # cat filter comm == Xorg && sig != 29 # echo comm != Xorg # cat filter comm != Xorg To clear a filter, echo 0 to the filter file: # echo 0 > filter # cat filter none The limit of 8 predicates for a compound expression is arbitrary - for efficiency, it's implemented as an array of pointers to predicates, and 8 seemed more than enough for any filter... Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <1237710665.7703.48.camel@charm-linux> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-22 08:31:04 +00:00
if (pred->str_val)
trace_seq_printf(s, "%s\n", pred->str_val);
tracing: add per-event filtering This patch adds per-event filtering to the event tracing subsystem. It adds a 'filter' debugfs file to each event directory. This file can be written to to set filters; reading from it will display the current set of filters set for that event. Basically, any field listed in the 'format' file for an event can be filtered on (including strings, but not yet other array types) using either matching ('==') or non-matching ('!=') 'predicates'. A 'predicate' can be either a single expression: # echo pid != 0 > filter # cat filter pid != 0 or a compound expression of up to 8 sub-expressions combined using '&&' or '||': # echo comm == Xorg > filter # echo "&& sig != 29" > filter # cat filter comm == Xorg && sig != 29 Only events having field values matching an expression will be available in the trace output; non-matching events are discarded. Note that a compound expression is built up by echoing each sub-expression separately - it's not the most efficient way to do things, but it keeps the parser simple and assumes that compound expressions will be relatively uncommon. In any case, a subsequent patch introducing a way to set filters for entire subsystems should mitigate any need to do this for lots of events. Setting a filter without an '&&' or '||' clears the previous filter completely and sets the filter to the new expression: # cat filter comm == Xorg && sig != 29 # echo comm != Xorg # cat filter comm != Xorg To clear a filter, echo 0 to the filter file: # echo 0 > filter # cat filter none The limit of 8 predicates for a compound expression is arbitrary - for efficiency, it's implemented as an array of pointers to predicates, and 8 seemed more than enough for any filter... Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <1237710665.7703.48.camel@charm-linux> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-22 08:31:04 +00:00
else
trace_seq_printf(s, "%llu\n", pred->val);
tracing: add per-event filtering This patch adds per-event filtering to the event tracing subsystem. It adds a 'filter' debugfs file to each event directory. This file can be written to to set filters; reading from it will display the current set of filters set for that event. Basically, any field listed in the 'format' file for an event can be filtered on (including strings, but not yet other array types) using either matching ('==') or non-matching ('!=') 'predicates'. A 'predicate' can be either a single expression: # echo pid != 0 > filter # cat filter pid != 0 or a compound expression of up to 8 sub-expressions combined using '&&' or '||': # echo comm == Xorg > filter # echo "&& sig != 29" > filter # cat filter comm == Xorg && sig != 29 Only events having field values matching an expression will be available in the trace output; non-matching events are discarded. Note that a compound expression is built up by echoing each sub-expression separately - it's not the most efficient way to do things, but it keeps the parser simple and assumes that compound expressions will be relatively uncommon. In any case, a subsequent patch introducing a way to set filters for entire subsystems should mitigate any need to do this for lots of events. Setting a filter without an '&&' or '||' clears the previous filter completely and sets the filter to the new expression: # cat filter comm == Xorg && sig != 29 # echo comm != Xorg # cat filter comm != Xorg To clear a filter, echo 0 to the filter file: # echo 0 > filter # cat filter none The limit of 8 predicates for a compound expression is arbitrary - for efficiency, it's implemented as an array of pointers to predicates, and 8 seemed more than enough for any filter... Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <1237710665.7703.48.camel@charm-linux> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-22 08:31:04 +00:00
} else
break;
}
}
static struct ftrace_event_field *
find_event_field(struct ftrace_event_call *call, char *name)
{
struct ftrace_event_field *field;
tracing: add per-event filtering This patch adds per-event filtering to the event tracing subsystem. It adds a 'filter' debugfs file to each event directory. This file can be written to to set filters; reading from it will display the current set of filters set for that event. Basically, any field listed in the 'format' file for an event can be filtered on (including strings, but not yet other array types) using either matching ('==') or non-matching ('!=') 'predicates'. A 'predicate' can be either a single expression: # echo pid != 0 > filter # cat filter pid != 0 or a compound expression of up to 8 sub-expressions combined using '&&' or '||': # echo comm == Xorg > filter # echo "&& sig != 29" > filter # cat filter comm == Xorg && sig != 29 Only events having field values matching an expression will be available in the trace output; non-matching events are discarded. Note that a compound expression is built up by echoing each sub-expression separately - it's not the most efficient way to do things, but it keeps the parser simple and assumes that compound expressions will be relatively uncommon. In any case, a subsequent patch introducing a way to set filters for entire subsystems should mitigate any need to do this for lots of events. Setting a filter without an '&&' or '||' clears the previous filter completely and sets the filter to the new expression: # cat filter comm == Xorg && sig != 29 # echo comm != Xorg # cat filter comm != Xorg To clear a filter, echo 0 to the filter file: # echo 0 > filter # cat filter none The limit of 8 predicates for a compound expression is arbitrary - for efficiency, it's implemented as an array of pointers to predicates, and 8 seemed more than enough for any filter... Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <1237710665.7703.48.camel@charm-linux> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-22 08:31:04 +00:00
list_for_each_entry(field, &call->fields, link) {
tracing: add per-event filtering This patch adds per-event filtering to the event tracing subsystem. It adds a 'filter' debugfs file to each event directory. This file can be written to to set filters; reading from it will display the current set of filters set for that event. Basically, any field listed in the 'format' file for an event can be filtered on (including strings, but not yet other array types) using either matching ('==') or non-matching ('!=') 'predicates'. A 'predicate' can be either a single expression: # echo pid != 0 > filter # cat filter pid != 0 or a compound expression of up to 8 sub-expressions combined using '&&' or '||': # echo comm == Xorg > filter # echo "&& sig != 29" > filter # cat filter comm == Xorg && sig != 29 Only events having field values matching an expression will be available in the trace output; non-matching events are discarded. Note that a compound expression is built up by echoing each sub-expression separately - it's not the most efficient way to do things, but it keeps the parser simple and assumes that compound expressions will be relatively uncommon. In any case, a subsequent patch introducing a way to set filters for entire subsystems should mitigate any need to do this for lots of events. Setting a filter without an '&&' or '||' clears the previous filter completely and sets the filter to the new expression: # cat filter comm == Xorg && sig != 29 # echo comm != Xorg # cat filter comm != Xorg To clear a filter, echo 0 to the filter file: # echo 0 > filter # cat filter none The limit of 8 predicates for a compound expression is arbitrary - for efficiency, it's implemented as an array of pointers to predicates, and 8 seemed more than enough for any filter... Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <1237710665.7703.48.camel@charm-linux> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-22 08:31:04 +00:00
if (!strcmp(field->name, name))
return field;
}
return NULL;
}
void filter_free_pred(struct filter_pred *pred)
{
if (!pred)
return;
kfree(pred->field_name);
kfree(pred->str_val);
kfree(pred);
}
void filter_free_preds(struct ftrace_event_call *call)
{
int i;
if (call->preds) {
for (i = 0; i < MAX_FILTER_PRED; i++)
filter_free_pred(call->preds[i]);
kfree(call->preds);
call->preds = NULL;
}
}
void filter_free_subsystem_preds(struct event_subsystem *system)
{
struct ftrace_event_call *call = __start_ftrace_events;
int i;
if (system->preds) {
for (i = 0; i < MAX_FILTER_PRED; i++)
filter_free_pred(system->preds[i]);
kfree(system->preds);
system->preds = NULL;
}
events_for_each(call) {
tracing/filters: add run-time field descriptions to TRACE_EVENT_FORMAT events This patch adds run-time field descriptions to all the event formats exported using TRACE_EVENT_FORMAT. It also hooks up all the tracers that use them (i.e. the tracers in the 'ftrace subsystem') so they can also have their output filtered by the event-filtering mechanism. When I was testing this, there were a couple of things that fooled me into thinking the filters weren't working, when actually they were - I'll mention them here so others don't make the same mistakes (and file bug reports. ;-) One is that some of the tracers trace multiple events e.g. the sched_switch tracer uses the context_switch and wakeup events, and if you don't set filters on all of the traced events, the unfiltered output from the events without filters on them can make it look like the filtering as a whole isn't working properly, when actually it is doing what it was asked to do - it just wasn't asked to do the right thing. The other is that for the really high-volume tracers e.g. the function tracer, the volume of filtered events can be so high that it pushes the unfiltered events out of the ring buffer before they can be read so e.g. cat'ing the trace file repeatedly shows either no output, or once in awhile some output but that isn't there the next time you read the trace, which isn't what you normally expect when reading the trace file. If you read from the trace_pipe file though, you can catch them before they disappear. Changes from v1: As suggested by Frederic Weisbecker: - get rid of externs in functions - added unlikely() to filter_check_discard() Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <srostedt@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-31 05:48:49 +00:00
if (!call->define_fields)
continue;
if (!strcmp(call->system, system->name))
filter_free_preds(call);
}
}
tracing: add per-event filtering This patch adds per-event filtering to the event tracing subsystem. It adds a 'filter' debugfs file to each event directory. This file can be written to to set filters; reading from it will display the current set of filters set for that event. Basically, any field listed in the 'format' file for an event can be filtered on (including strings, but not yet other array types) using either matching ('==') or non-matching ('!=') 'predicates'. A 'predicate' can be either a single expression: # echo pid != 0 > filter # cat filter pid != 0 or a compound expression of up to 8 sub-expressions combined using '&&' or '||': # echo comm == Xorg > filter # echo "&& sig != 29" > filter # cat filter comm == Xorg && sig != 29 Only events having field values matching an expression will be available in the trace output; non-matching events are discarded. Note that a compound expression is built up by echoing each sub-expression separately - it's not the most efficient way to do things, but it keeps the parser simple and assumes that compound expressions will be relatively uncommon. In any case, a subsequent patch introducing a way to set filters for entire subsystems should mitigate any need to do this for lots of events. Setting a filter without an '&&' or '||' clears the previous filter completely and sets the filter to the new expression: # cat filter comm == Xorg && sig != 29 # echo comm != Xorg # cat filter comm != Xorg To clear a filter, echo 0 to the filter file: # echo 0 > filter # cat filter none The limit of 8 predicates for a compound expression is arbitrary - for efficiency, it's implemented as an array of pointers to predicates, and 8 seemed more than enough for any filter... Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <1237710665.7703.48.camel@charm-linux> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-22 08:31:04 +00:00
static int __filter_add_pred(struct ftrace_event_call *call,
struct filter_pred *pred)
{
int i;
if (call->preds && !pred->compound)
filter_free_preds(call);
if (!call->preds) {
call->preds = kzalloc(MAX_FILTER_PRED * sizeof(pred),
GFP_KERNEL);
if (!call->preds)
return -ENOMEM;
}
for (i = 0; i < MAX_FILTER_PRED; i++) {
if (!call->preds[i]) {
call->preds[i] = pred;
return 0;
}
}
return -ENOSPC;
tracing: add per-event filtering This patch adds per-event filtering to the event tracing subsystem. It adds a 'filter' debugfs file to each event directory. This file can be written to to set filters; reading from it will display the current set of filters set for that event. Basically, any field listed in the 'format' file for an event can be filtered on (including strings, but not yet other array types) using either matching ('==') or non-matching ('!=') 'predicates'. A 'predicate' can be either a single expression: # echo pid != 0 > filter # cat filter pid != 0 or a compound expression of up to 8 sub-expressions combined using '&&' or '||': # echo comm == Xorg > filter # echo "&& sig != 29" > filter # cat filter comm == Xorg && sig != 29 Only events having field values matching an expression will be available in the trace output; non-matching events are discarded. Note that a compound expression is built up by echoing each sub-expression separately - it's not the most efficient way to do things, but it keeps the parser simple and assumes that compound expressions will be relatively uncommon. In any case, a subsequent patch introducing a way to set filters for entire subsystems should mitigate any need to do this for lots of events. Setting a filter without an '&&' or '||' clears the previous filter completely and sets the filter to the new expression: # cat filter comm == Xorg && sig != 29 # echo comm != Xorg # cat filter comm != Xorg To clear a filter, echo 0 to the filter file: # echo 0 > filter # cat filter none The limit of 8 predicates for a compound expression is arbitrary - for efficiency, it's implemented as an array of pointers to predicates, and 8 seemed more than enough for any filter... Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <1237710665.7703.48.camel@charm-linux> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-22 08:31:04 +00:00
}
static int is_string_field(const char *type)
{
if (strchr(type, '[') && strstr(type, "char"))
return 1;
return 0;
}
int filter_add_pred(struct ftrace_event_call *call, struct filter_pred *pred)
{
struct ftrace_event_field *field;
field = find_event_field(call, pred->field_name);
if (!field)
return -EINVAL;
pred->offset = field->offset;
if (is_string_field(field->type)) {
if (!pred->str_val)
return -EINVAL;
tracing: add per-event filtering This patch adds per-event filtering to the event tracing subsystem. It adds a 'filter' debugfs file to each event directory. This file can be written to to set filters; reading from it will display the current set of filters set for that event. Basically, any field listed in the 'format' file for an event can be filtered on (including strings, but not yet other array types) using either matching ('==') or non-matching ('!=') 'predicates'. A 'predicate' can be either a single expression: # echo pid != 0 > filter # cat filter pid != 0 or a compound expression of up to 8 sub-expressions combined using '&&' or '||': # echo comm == Xorg > filter # echo "&& sig != 29" > filter # cat filter comm == Xorg && sig != 29 Only events having field values matching an expression will be available in the trace output; non-matching events are discarded. Note that a compound expression is built up by echoing each sub-expression separately - it's not the most efficient way to do things, but it keeps the parser simple and assumes that compound expressions will be relatively uncommon. In any case, a subsequent patch introducing a way to set filters for entire subsystems should mitigate any need to do this for lots of events. Setting a filter without an '&&' or '||' clears the previous filter completely and sets the filter to the new expression: # cat filter comm == Xorg && sig != 29 # echo comm != Xorg # cat filter comm != Xorg To clear a filter, echo 0 to the filter file: # echo 0 > filter # cat filter none The limit of 8 predicates for a compound expression is arbitrary - for efficiency, it's implemented as an array of pointers to predicates, and 8 seemed more than enough for any filter... Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <1237710665.7703.48.camel@charm-linux> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-22 08:31:04 +00:00
pred->fn = filter_pred_string;
pred->str_len = field->size;
return __filter_add_pred(call, pred);
} else {
if (pred->str_val)
return -EINVAL;
tracing: add per-event filtering This patch adds per-event filtering to the event tracing subsystem. It adds a 'filter' debugfs file to each event directory. This file can be written to to set filters; reading from it will display the current set of filters set for that event. Basically, any field listed in the 'format' file for an event can be filtered on (including strings, but not yet other array types) using either matching ('==') or non-matching ('!=') 'predicates'. A 'predicate' can be either a single expression: # echo pid != 0 > filter # cat filter pid != 0 or a compound expression of up to 8 sub-expressions combined using '&&' or '||': # echo comm == Xorg > filter # echo "&& sig != 29" > filter # cat filter comm == Xorg && sig != 29 Only events having field values matching an expression will be available in the trace output; non-matching events are discarded. Note that a compound expression is built up by echoing each sub-expression separately - it's not the most efficient way to do things, but it keeps the parser simple and assumes that compound expressions will be relatively uncommon. In any case, a subsequent patch introducing a way to set filters for entire subsystems should mitigate any need to do this for lots of events. Setting a filter without an '&&' or '||' clears the previous filter completely and sets the filter to the new expression: # cat filter comm == Xorg && sig != 29 # echo comm != Xorg # cat filter comm != Xorg To clear a filter, echo 0 to the filter file: # echo 0 > filter # cat filter none The limit of 8 predicates for a compound expression is arbitrary - for efficiency, it's implemented as an array of pointers to predicates, and 8 seemed more than enough for any filter... Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <1237710665.7703.48.camel@charm-linux> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-22 08:31:04 +00:00
}
switch (field->size) {
case 8:
pred->fn = filter_pred_64;
break;
case 4:
pred->fn = filter_pred_32;
break;
case 2:
pred->fn = filter_pred_16;
break;
case 1:
pred->fn = filter_pred_8;
break;
default:
return -EINVAL;
}
return __filter_add_pred(call, pred);
}
static struct filter_pred *copy_pred(struct filter_pred *pred)
{
struct filter_pred *new_pred = kmalloc(sizeof(*pred), GFP_KERNEL);
if (!new_pred)
return NULL;
memcpy(new_pred, pred, sizeof(*pred));
if (pred->field_name) {
new_pred->field_name = kstrdup(pred->field_name, GFP_KERNEL);
if (!new_pred->field_name) {
kfree(new_pred);
return NULL;
}
}
if (pred->str_val) {
new_pred->str_val = kstrdup(pred->str_val, GFP_KERNEL);
if (!new_pred->str_val) {
filter_free_pred(new_pred);
return NULL;
}
}
return new_pred;
}
int filter_add_subsystem_pred(struct event_subsystem *system,
struct filter_pred *pred)
{
struct ftrace_event_call *call = __start_ftrace_events;
struct filter_pred *event_pred;
int i;
if (system->preds && !pred->compound)
filter_free_subsystem_preds(system);
if (!system->preds) {
system->preds = kzalloc(MAX_FILTER_PRED * sizeof(pred),
GFP_KERNEL);
if (!system->preds)
return -ENOMEM;
}
for (i = 0; i < MAX_FILTER_PRED; i++) {
if (!system->preds[i]) {
system->preds[i] = pred;
break;
}
}
if (i == MAX_FILTER_PRED)
return -ENOSPC;
events_for_each(call) {
int err;
tracing/filters: add run-time field descriptions to TRACE_EVENT_FORMAT events This patch adds run-time field descriptions to all the event formats exported using TRACE_EVENT_FORMAT. It also hooks up all the tracers that use them (i.e. the tracers in the 'ftrace subsystem') so they can also have their output filtered by the event-filtering mechanism. When I was testing this, there were a couple of things that fooled me into thinking the filters weren't working, when actually they were - I'll mention them here so others don't make the same mistakes (and file bug reports. ;-) One is that some of the tracers trace multiple events e.g. the sched_switch tracer uses the context_switch and wakeup events, and if you don't set filters on all of the traced events, the unfiltered output from the events without filters on them can make it look like the filtering as a whole isn't working properly, when actually it is doing what it was asked to do - it just wasn't asked to do the right thing. The other is that for the really high-volume tracers e.g. the function tracer, the volume of filtered events can be so high that it pushes the unfiltered events out of the ring buffer before they can be read so e.g. cat'ing the trace file repeatedly shows either no output, or once in awhile some output but that isn't there the next time you read the trace, which isn't what you normally expect when reading the trace file. If you read from the trace_pipe file though, you can catch them before they disappear. Changes from v1: As suggested by Frederic Weisbecker: - get rid of externs in functions - added unlikely() to filter_check_discard() Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <srostedt@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-31 05:48:49 +00:00
if (!call->define_fields)
continue;
if (strcmp(call->system, system->name))
continue;
if (!find_event_field(call, pred->field_name))
continue;
event_pred = copy_pred(pred);
if (!event_pred)
goto oom;
err = filter_add_pred(call, event_pred);
if (err)
filter_free_pred(event_pred);
if (err == -ENOMEM)
goto oom;
}
return 0;
oom:
system->preds[i] = NULL;
return -ENOMEM;
}
tracing: add per-event filtering This patch adds per-event filtering to the event tracing subsystem. It adds a 'filter' debugfs file to each event directory. This file can be written to to set filters; reading from it will display the current set of filters set for that event. Basically, any field listed in the 'format' file for an event can be filtered on (including strings, but not yet other array types) using either matching ('==') or non-matching ('!=') 'predicates'. A 'predicate' can be either a single expression: # echo pid != 0 > filter # cat filter pid != 0 or a compound expression of up to 8 sub-expressions combined using '&&' or '||': # echo comm == Xorg > filter # echo "&& sig != 29" > filter # cat filter comm == Xorg && sig != 29 Only events having field values matching an expression will be available in the trace output; non-matching events are discarded. Note that a compound expression is built up by echoing each sub-expression separately - it's not the most efficient way to do things, but it keeps the parser simple and assumes that compound expressions will be relatively uncommon. In any case, a subsequent patch introducing a way to set filters for entire subsystems should mitigate any need to do this for lots of events. Setting a filter without an '&&' or '||' clears the previous filter completely and sets the filter to the new expression: # cat filter comm == Xorg && sig != 29 # echo comm != Xorg # cat filter comm != Xorg To clear a filter, echo 0 to the filter file: # echo 0 > filter # cat filter none The limit of 8 predicates for a compound expression is arbitrary - for efficiency, it's implemented as an array of pointers to predicates, and 8 seemed more than enough for any filter... Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <1237710665.7703.48.camel@charm-linux> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-22 08:31:04 +00:00
int filter_parse(char **pbuf, struct filter_pred *pred)
{
char *tmp, *tok, *val_str = NULL;
int tok_n = 0;
/* field ==/!= number, or/and field ==/!= number, number */
while ((tok = strsep(pbuf, " \n"))) {
if (tok_n == 0) {
if (!strcmp(tok, "0")) {
pred->clear = 1;
return 0;
} else if (!strcmp(tok, "&&")) {
pred->or = 0;
pred->compound = 1;
} else if (!strcmp(tok, "||")) {
pred->or = 1;
pred->compound = 1;
} else
pred->field_name = tok;
tok_n = 1;
continue;
}
if (tok_n == 1) {
if (!pred->field_name)
pred->field_name = tok;
else if (!strcmp(tok, "!="))
pred->not = 1;
else if (!strcmp(tok, "=="))
pred->not = 0;
else {
pred->field_name = NULL;
return -EINVAL;
}
tok_n = 2;
continue;
}
if (tok_n == 2) {
if (pred->compound) {
if (!strcmp(tok, "!="))
pred->not = 1;
else if (!strcmp(tok, "=="))
pred->not = 0;
else {
pred->field_name = NULL;
return -EINVAL;
}
} else {
val_str = tok;
break; /* done */
}
tok_n = 3;
continue;
}
if (tok_n == 3) {
val_str = tok;
break; /* done */
}
}
if (!val_str) {
pred->field_name = NULL;
return -EINVAL;
}
tracing: add per-event filtering This patch adds per-event filtering to the event tracing subsystem. It adds a 'filter' debugfs file to each event directory. This file can be written to to set filters; reading from it will display the current set of filters set for that event. Basically, any field listed in the 'format' file for an event can be filtered on (including strings, but not yet other array types) using either matching ('==') or non-matching ('!=') 'predicates'. A 'predicate' can be either a single expression: # echo pid != 0 > filter # cat filter pid != 0 or a compound expression of up to 8 sub-expressions combined using '&&' or '||': # echo comm == Xorg > filter # echo "&& sig != 29" > filter # cat filter comm == Xorg && sig != 29 Only events having field values matching an expression will be available in the trace output; non-matching events are discarded. Note that a compound expression is built up by echoing each sub-expression separately - it's not the most efficient way to do things, but it keeps the parser simple and assumes that compound expressions will be relatively uncommon. In any case, a subsequent patch introducing a way to set filters for entire subsystems should mitigate any need to do this for lots of events. Setting a filter without an '&&' or '||' clears the previous filter completely and sets the filter to the new expression: # cat filter comm == Xorg && sig != 29 # echo comm != Xorg # cat filter comm != Xorg To clear a filter, echo 0 to the filter file: # echo 0 > filter # cat filter none The limit of 8 predicates for a compound expression is arbitrary - for efficiency, it's implemented as an array of pointers to predicates, and 8 seemed more than enough for any filter... Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <1237710665.7703.48.camel@charm-linux> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-22 08:31:04 +00:00
pred->field_name = kstrdup(pred->field_name, GFP_KERNEL);
if (!pred->field_name)
return -ENOMEM;
pred->val = simple_strtoull(val_str, &tmp, 0);
tracing: add per-event filtering This patch adds per-event filtering to the event tracing subsystem. It adds a 'filter' debugfs file to each event directory. This file can be written to to set filters; reading from it will display the current set of filters set for that event. Basically, any field listed in the 'format' file for an event can be filtered on (including strings, but not yet other array types) using either matching ('==') or non-matching ('!=') 'predicates'. A 'predicate' can be either a single expression: # echo pid != 0 > filter # cat filter pid != 0 or a compound expression of up to 8 sub-expressions combined using '&&' or '||': # echo comm == Xorg > filter # echo "&& sig != 29" > filter # cat filter comm == Xorg && sig != 29 Only events having field values matching an expression will be available in the trace output; non-matching events are discarded. Note that a compound expression is built up by echoing each sub-expression separately - it's not the most efficient way to do things, but it keeps the parser simple and assumes that compound expressions will be relatively uncommon. In any case, a subsequent patch introducing a way to set filters for entire subsystems should mitigate any need to do this for lots of events. Setting a filter without an '&&' or '||' clears the previous filter completely and sets the filter to the new expression: # cat filter comm == Xorg && sig != 29 # echo comm != Xorg # cat filter comm != Xorg To clear a filter, echo 0 to the filter file: # echo 0 > filter # cat filter none The limit of 8 predicates for a compound expression is arbitrary - for efficiency, it's implemented as an array of pointers to predicates, and 8 seemed more than enough for any filter... Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <1237710665.7703.48.camel@charm-linux> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-22 08:31:04 +00:00
if (tmp == val_str) {
pred->str_val = kstrdup(val_str, GFP_KERNEL);
if (!pred->str_val)
return -ENOMEM;
} else if (*tmp != '\0')
return -EINVAL;
tracing: add per-event filtering This patch adds per-event filtering to the event tracing subsystem. It adds a 'filter' debugfs file to each event directory. This file can be written to to set filters; reading from it will display the current set of filters set for that event. Basically, any field listed in the 'format' file for an event can be filtered on (including strings, but not yet other array types) using either matching ('==') or non-matching ('!=') 'predicates'. A 'predicate' can be either a single expression: # echo pid != 0 > filter # cat filter pid != 0 or a compound expression of up to 8 sub-expressions combined using '&&' or '||': # echo comm == Xorg > filter # echo "&& sig != 29" > filter # cat filter comm == Xorg && sig != 29 Only events having field values matching an expression will be available in the trace output; non-matching events are discarded. Note that a compound expression is built up by echoing each sub-expression separately - it's not the most efficient way to do things, but it keeps the parser simple and assumes that compound expressions will be relatively uncommon. In any case, a subsequent patch introducing a way to set filters for entire subsystems should mitigate any need to do this for lots of events. Setting a filter without an '&&' or '||' clears the previous filter completely and sets the filter to the new expression: # cat filter comm == Xorg && sig != 29 # echo comm != Xorg # cat filter comm != Xorg To clear a filter, echo 0 to the filter file: # echo 0 > filter # cat filter none The limit of 8 predicates for a compound expression is arbitrary - for efficiency, it's implemented as an array of pointers to predicates, and 8 seemed more than enough for any filter... Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <1237710665.7703.48.camel@charm-linux> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-03-22 08:31:04 +00:00
return 0;
}