linux-stable/kernel/trace/trace_events_filter.c

1964 lines
41 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/module.h>
#include <linux/ctype.h>
#include <linux/mutex.h>
#include <linux/perf_event.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 08:04:11 +00:00
#include <linux/slab.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
#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
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
enum filter_op_ids
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
{
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
OP_OR,
OP_AND,
OP_GLOB,
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
OP_NE,
OP_EQ,
OP_LT,
OP_LE,
OP_GT,
OP_GE,
OP_NONE,
OP_OPEN_PAREN,
};
struct filter_op {
int id;
char *string;
int precedence;
};
static struct filter_op filter_ops[] = {
{ OP_OR, "||", 1 },
{ OP_AND, "&&", 2 },
{ OP_GLOB, "~", 4 },
{ OP_NE, "!=", 4 },
{ OP_EQ, "==", 4 },
{ OP_LT, "<", 5 },
{ OP_LE, "<=", 5 },
{ OP_GT, ">", 5 },
{ OP_GE, ">=", 5 },
{ OP_NONE, "OP_NONE", 0 },
{ OP_OPEN_PAREN, "(", 0 },
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
};
enum {
FILT_ERR_NONE,
FILT_ERR_INVALID_OP,
FILT_ERR_UNBALANCED_PAREN,
FILT_ERR_TOO_MANY_OPERANDS,
FILT_ERR_OPERAND_TOO_LONG,
FILT_ERR_FIELD_NOT_FOUND,
FILT_ERR_ILLEGAL_FIELD_OP,
FILT_ERR_ILLEGAL_INTVAL,
FILT_ERR_BAD_SUBSYS_FILTER,
FILT_ERR_TOO_MANY_PREDS,
FILT_ERR_MISSING_FIELD,
FILT_ERR_INVALID_FILTER,
};
static char *err_text[] = {
"No error",
"Invalid operator",
"Unbalanced parens",
"Too many operands",
"Operand too long",
"Field not found",
"Illegal operation for field type",
"Illegal integer value",
"Couldn't find or set field in one of a subsystem's events",
"Too many terms in predicate expression",
"Missing field name and/or value",
"Meaningless filter expression",
};
struct opstack_op {
int op;
struct list_head list;
};
struct postfix_elt {
int op;
char *operand;
struct list_head list;
};
struct filter_parse_state {
struct filter_op *ops;
struct list_head opstack;
struct list_head postfix;
int lasterr;
int lasterr_pos;
struct {
char *string;
unsigned int cnt;
unsigned int tail;
} infix;
struct {
char string[MAX_FILTER_STR_VAL];
int pos;
unsigned int tail;
} operand;
};
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
struct pred_stack {
struct filter_pred **preds;
int index;
};
#define DEFINE_COMPARISON_PRED(type) \
static int filter_pred_##type(struct filter_pred *pred, void *event) \
{ \
type *addr = (type *)(event + pred->offset); \
type val = (type)pred->val; \
int match = 0; \
\
switch (pred->op) { \
case OP_LT: \
match = (*addr < val); \
break; \
case OP_LE: \
match = (*addr <= val); \
break; \
case OP_GT: \
match = (*addr > val); \
break; \
case OP_GE: \
match = (*addr >= val); \
break; \
default: \
break; \
} \
\
return match; \
}
#define DEFINE_EQUALITY_PRED(size) \
static int filter_pred_##size(struct filter_pred *pred, void *event) \
{ \
u##size *addr = (u##size *)(event + pred->offset); \
u##size val = (u##size)pred->val; \
int match; \
\
match = (val == *addr) ^ pred->not; \
\
return match; \
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
DEFINE_COMPARISON_PRED(s64);
DEFINE_COMPARISON_PRED(u64);
DEFINE_COMPARISON_PRED(s32);
DEFINE_COMPARISON_PRED(u32);
DEFINE_COMPARISON_PRED(s16);
DEFINE_COMPARISON_PRED(u16);
DEFINE_COMPARISON_PRED(s8);
DEFINE_COMPARISON_PRED(u8);
DEFINE_EQUALITY_PRED(64);
DEFINE_EQUALITY_PRED(32);
DEFINE_EQUALITY_PRED(16);
DEFINE_EQUALITY_PRED(8);
/* Filter predicate for fixed sized arrays of characters */
static int filter_pred_string(struct filter_pred *pred, void *event)
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 *addr = (char *)(event + pred->offset);
int cmp, match;
cmp = pred->regex.match(addr, &pred->regex, pred->regex.field_len);
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
match = cmp ^ 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
return match;
}
/* Filter predicate for char * pointers */
static int filter_pred_pchar(struct filter_pred *pred, void *event)
{
char **addr = (char **)(event + pred->offset);
int cmp, match;
int len = strlen(*addr) + 1; /* including tailing '\0' */
cmp = pred->regex.match(*addr, &pred->regex, len);
match = cmp ^ pred->not;
return match;
}
/*
* Filter predicate for dynamic sized arrays of characters.
* These are implemented through a list of strings at the end
* of the entry.
* Also each of these strings have a field in the entry which
* contains its offset from the beginning of the entry.
* We have then first to get this field, dereference it
* and add it to the address of the entry, and at last we have
* the address of the string.
*/
static int filter_pred_strloc(struct filter_pred *pred, void *event)
{
u32 str_item = *(u32 *)(event + pred->offset);
int str_loc = str_item & 0xffff;
int str_len = str_item >> 16;
char *addr = (char *)(event + str_loc);
int cmp, match;
cmp = pred->regex.match(addr, &pred->regex, str_len);
match = cmp ^ pred->not;
return match;
}
static int filter_pred_none(struct filter_pred *pred, void *event)
tracing/filters: allow on-the-fly filter switching This patch allows event filters to be safely removed or switched on-the-fly while avoiding the use of rcu or the suspension of tracing of previous versions. It does it by adding a new filter_pred_none() predicate function which does nothing and by never deallocating either the predicates or any of the filter_pred members used in matching; the predicate lists are allocated and initialized during ftrace_event_calls initialization. Whenever a filter is removed or replaced, the filter_pred_* functions currently in use by the affected ftrace_event_call are immediately switched over to to the filter_pred_none() function, while the rest of the filter_pred members are left intact, allowing any currently executing filter_pred_* functions to finish up, using the values they're currently using. In the case of filter replacement, the new predicate values are copied into the old predicates after the above step, and the filter_pred_none() functions are replaced by the filter_pred_* functions for the new filter. In this case, it is possible though very unlikely that a previous filter_pred_* is still running even after the filter_pred_none() switch and the switch to the new filter_pred_*. In that case, however, because nothing has been deallocated in the filter_pred, the worst that can happen is that the old filter_pred_* function sees the new values and as a result produces either a false positive or a false negative, depending on the values it finds. So one downside to this method is that rarely, it can produce a bad match during the filter switch, but it should be possible to live with that, IMHO. The other downside is that at least in this patch the predicate lists are always pre-allocated, taking up memory from the start. They could probably be allocated on first-use, and de-allocated when tracing is completely stopped - if this patch makes sense, I could create another one to do that later on. Oh, and it also places a restriction on the size of __arrays in events, currently set to 128, since they can't be larger than the now embedded str_val arrays in the filter_pred struct. Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: paulmck@linux.vnet.ibm.com LKML-Reference: <1239610670.6660.49.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-13 08:17:50 +00:00
{
return 0;
}
/*
* regex_match_foo - Basic regex callbacks
*
* @str: the string to be searched
* @r: the regex structure containing the pattern string
* @len: the length of the string to be searched (including '\0')
*
* Note:
* - @str might not be NULL-terminated if it's of type DYN_STRING
* or STATIC_STRING
*/
static int regex_match_full(char *str, struct regex *r, int len)
{
if (strncmp(str, r->pattern, len) == 0)
return 1;
return 0;
}
static int regex_match_front(char *str, struct regex *r, int len)
{
if (strncmp(str, r->pattern, r->len) == 0)
return 1;
return 0;
}
static int regex_match_middle(char *str, struct regex *r, int len)
{
if (strnstr(str, r->pattern, len))
return 1;
return 0;
}
static int regex_match_end(char *str, struct regex *r, int len)
{
int strlen = len - 1;
if (strlen >= r->len &&
memcmp(str + strlen - r->len, r->pattern, r->len) == 0)
return 1;
return 0;
}
/**
* filter_parse_regex - parse a basic regex
* @buff: the raw regex
* @len: length of the regex
* @search: will point to the beginning of the string to compare
* @not: tell whether the match will have to be inverted
*
* This passes in a buffer containing a regex and this function will
* set search to point to the search part of the buffer and
* return the type of search it is (see enum above).
* This does modify buff.
*
* Returns enum type.
* search returns the pointer to use for comparison.
* not returns 1 if buff started with a '!'
* 0 otherwise.
*/
enum regex_type filter_parse_regex(char *buff, int len, char **search, int *not)
{
int type = MATCH_FULL;
int i;
if (buff[0] == '!') {
*not = 1;
buff++;
len--;
} else
*not = 0;
*search = buff;
for (i = 0; i < len; i++) {
if (buff[i] == '*') {
if (!i) {
*search = buff + 1;
type = MATCH_END_ONLY;
} else {
if (type == MATCH_END_ONLY)
type = MATCH_MIDDLE_ONLY;
else
type = MATCH_FRONT_ONLY;
buff[i] = 0;
break;
}
}
}
return type;
}
static void filter_build_regex(struct filter_pred *pred)
{
struct regex *r = &pred->regex;
char *search;
enum regex_type type = MATCH_FULL;
int not = 0;
if (pred->op == OP_GLOB) {
type = filter_parse_regex(r->pattern, r->len, &search, &not);
r->len = strlen(search);
memmove(r->pattern, search, r->len+1);
}
switch (type) {
case MATCH_FULL:
r->match = regex_match_full;
break;
case MATCH_FRONT_ONLY:
r->match = regex_match_front;
break;
case MATCH_MIDDLE_ONLY:
r->match = regex_match_middle;
break;
case MATCH_END_ONLY:
r->match = regex_match_end;
break;
}
pred->not ^= not;
}
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
enum move_type {
MOVE_DOWN,
MOVE_UP_FROM_LEFT,
MOVE_UP_FROM_RIGHT
};
static struct filter_pred *
get_pred_parent(struct filter_pred *pred, struct filter_pred *preds,
int index, enum move_type *move)
{
if (pred->parent & FILTER_PRED_IS_RIGHT)
*move = MOVE_UP_FROM_RIGHT;
else
*move = MOVE_UP_FROM_LEFT;
pred = &preds[pred->parent & ~FILTER_PRED_IS_RIGHT];
return pred;
}
enum walk_return {
WALK_PRED_ABORT,
WALK_PRED_PARENT,
WALK_PRED_DEFAULT,
};
typedef int (*filter_pred_walkcb_t) (enum move_type move,
struct filter_pred *pred,
int *err, void *data);
static int walk_pred_tree(struct filter_pred *preds,
struct filter_pred *root,
filter_pred_walkcb_t cb, void *data)
{
struct filter_pred *pred = root;
enum move_type move = MOVE_DOWN;
int done = 0;
if (!preds)
return -EINVAL;
do {
int err = 0, ret;
ret = cb(move, pred, &err, data);
if (ret == WALK_PRED_ABORT)
return err;
if (ret == WALK_PRED_PARENT)
goto get_parent;
switch (move) {
case MOVE_DOWN:
if (pred->left != FILTER_PRED_INVALID) {
pred = &preds[pred->left];
continue;
}
goto get_parent;
case MOVE_UP_FROM_LEFT:
pred = &preds[pred->right];
move = MOVE_DOWN;
continue;
case MOVE_UP_FROM_RIGHT:
get_parent:
if (pred == root)
break;
pred = get_pred_parent(pred, preds,
pred->parent,
&move);
continue;
}
done = 1;
} while (!done);
/* We are fine. */
return 0;
}
/*
* A series of AND or ORs where found together. Instead of
* climbing up and down the tree branches, an array of the
* ops were made in order of checks. We can just move across
* the array and short circuit if needed.
*/
static int process_ops(struct filter_pred *preds,
struct filter_pred *op, void *rec)
{
struct filter_pred *pred;
int match = 0;
int type;
int i;
/*
* Micro-optimization: We set type to true if op
* is an OR and false otherwise (AND). Then we
* just need to test if the match is equal to
* the type, and if it is, we can short circuit the
* rest of the checks:
*
* if ((match && op->op == OP_OR) ||
* (!match && op->op == OP_AND))
* return match;
*/
type = op->op == OP_OR;
for (i = 0; i < op->val; i++) {
pred = &preds[op->ops[i]];
match = pred->fn(pred, rec);
if (!!match == type)
return match;
}
return match;
}
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 1 if event matches, 0 otherwise (discard) */
int filter_match_preds(struct event_filter *filter, void *rec)
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
{
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
int match = -1;
enum move_type move = MOVE_DOWN;
struct filter_pred *preds;
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
struct filter_pred *pred;
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
struct filter_pred *root;
int n_preds;
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
int done = 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
/* no filter is considered a match */
if (!filter)
return 1;
n_preds = filter->n_preds;
if (!n_preds)
return 1;
/*
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
* n_preds, root and filter->preds are protect with preemption disabled.
*/
preds = rcu_dereference_sched(filter->preds);
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
root = rcu_dereference_sched(filter->root);
if (!root)
return 1;
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
pred = root;
/* match is currently meaningless */
match = -1;
do {
switch (move) {
case MOVE_DOWN:
/* only AND and OR have children */
if (pred->left != FILTER_PRED_INVALID) {
/* If ops is set, then it was folded. */
if (!pred->ops) {
/* keep going to down the left side */
pred = &preds[pred->left];
continue;
}
/* We can treat folded ops as a leaf node */
match = process_ops(preds, pred, rec);
} else
match = pred->fn(pred, rec);
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
/* If this pred is the only pred */
if (pred == root)
break;
pred = get_pred_parent(pred, preds,
pred->parent, &move);
continue;
case MOVE_UP_FROM_LEFT:
/*
* Check for short circuits.
*
* Optimization: !!match == (pred->op == OP_OR)
* is the same as:
* if ((match && pred->op == OP_OR) ||
* (!match && pred->op == OP_AND))
*/
if (!!match == (pred->op == OP_OR)) {
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
if (pred == root)
break;
pred = get_pred_parent(pred, preds,
pred->parent, &move);
continue;
}
/* now go down the right side of the tree. */
pred = &preds[pred->right];
move = MOVE_DOWN;
continue;
case MOVE_UP_FROM_RIGHT:
/* We finished this equation. */
if (pred == root)
break;
pred = get_pred_parent(pred, preds,
pred->parent, &move);
tracing/filters: allow on-the-fly filter switching This patch allows event filters to be safely removed or switched on-the-fly while avoiding the use of rcu or the suspension of tracing of previous versions. It does it by adding a new filter_pred_none() predicate function which does nothing and by never deallocating either the predicates or any of the filter_pred members used in matching; the predicate lists are allocated and initialized during ftrace_event_calls initialization. Whenever a filter is removed or replaced, the filter_pred_* functions currently in use by the affected ftrace_event_call are immediately switched over to to the filter_pred_none() function, while the rest of the filter_pred members are left intact, allowing any currently executing filter_pred_* functions to finish up, using the values they're currently using. In the case of filter replacement, the new predicate values are copied into the old predicates after the above step, and the filter_pred_none() functions are replaced by the filter_pred_* functions for the new filter. In this case, it is possible though very unlikely that a previous filter_pred_* is still running even after the filter_pred_none() switch and the switch to the new filter_pred_*. In that case, however, because nothing has been deallocated in the filter_pred, the worst that can happen is that the old filter_pred_* function sees the new values and as a result produces either a false positive or a false negative, depending on the values it finds. So one downside to this method is that rarely, it can produce a bad match during the filter switch, but it should be possible to live with that, IMHO. The other downside is that at least in this patch the predicate lists are always pre-allocated, taking up memory from the start. They could probably be allocated on first-use, and de-allocated when tracing is completely stopped - if this patch makes sense, I could create another one to do that later on. Oh, and it also places a restriction on the size of __arrays in events, currently set to 128, since they can't be larger than the now embedded str_val arrays in the filter_pred struct. Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: paulmck@linux.vnet.ibm.com LKML-Reference: <1239610670.6660.49.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-13 08:17:50 +00:00
continue;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
}
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
done = 1;
} while (!done);
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
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
return match;
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
}
EXPORT_SYMBOL_GPL(filter_match_preds);
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
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
static void parse_error(struct filter_parse_state *ps, int err, int pos)
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
{
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
ps->lasterr = err;
ps->lasterr_pos = pos;
}
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
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
static void remove_filter_string(struct event_filter *filter)
{
if (!filter)
return;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
kfree(filter->filter_string);
filter->filter_string = NULL;
}
static int replace_filter_string(struct event_filter *filter,
char *filter_string)
{
kfree(filter->filter_string);
filter->filter_string = kstrdup(filter_string, GFP_KERNEL);
if (!filter->filter_string)
return -ENOMEM;
return 0;
}
static int append_filter_string(struct event_filter *filter,
char *string)
{
int newlen;
char *new_filter_string;
BUG_ON(!filter->filter_string);
newlen = strlen(filter->filter_string) + strlen(string) + 1;
new_filter_string = kmalloc(newlen, GFP_KERNEL);
if (!new_filter_string)
return -ENOMEM;
strcpy(new_filter_string, filter->filter_string);
strcat(new_filter_string, string);
kfree(filter->filter_string);
filter->filter_string = new_filter_string;
return 0;
}
static void append_filter_err(struct filter_parse_state *ps,
struct event_filter *filter)
{
int pos = ps->lasterr_pos;
char *buf, *pbuf;
buf = (char *)__get_free_page(GFP_TEMPORARY);
if (!buf)
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
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
append_filter_string(filter, "\n");
memset(buf, ' ', PAGE_SIZE);
if (pos > PAGE_SIZE - 128)
pos = 0;
buf[pos] = '^';
pbuf = &buf[pos] + 1;
sprintf(pbuf, "\nparse_error: %s\n", err_text[ps->lasterr]);
append_filter_string(filter, buf);
free_page((unsigned long) buf);
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
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
void print_event_filter(struct ftrace_event_call *call, struct trace_seq *s)
{
struct event_filter *filter;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
mutex_lock(&event_mutex);
filter = call->filter;
if (filter && filter->filter_string)
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
trace_seq_printf(s, "%s\n", filter->filter_string);
else
trace_seq_printf(s, "none\n");
mutex_unlock(&event_mutex);
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
void print_subsystem_event_filter(struct event_subsystem *system,
struct trace_seq *s)
{
struct event_filter *filter;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
mutex_lock(&event_mutex);
filter = system->filter;
if (filter && filter->filter_string)
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
trace_seq_printf(s, "%s\n", filter->filter_string);
else
trace_seq_printf(s, "none\n");
mutex_unlock(&event_mutex);
}
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 struct ftrace_event_field *
__find_event_field(struct list_head *head, char *name)
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
{
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
tracing: Move fields from event to class structure Move the defined fields from the event to the class structure. Since the fields of the event are defined by the class they belong to, it makes sense to have the class hold the information instead of the individual events. The events of the same class would just hold duplicate information. After this change the size of the kernel dropped another 3K: text data bss dec hex filename 4913961 1088356 861512 6863829 68bbd5 vmlinux.orig 4900252 1057412 861512 6819176 680d68 vmlinux.regs 4900375 1053380 861512 6815267 67fe23 vmlinux.fields Although the text increased, this was mainly due to the C files having to adapt to the change. This is a constant increase, where new tracepoints will not increase the Text. But the big drop is in the data size (as well as needed allocations to hold the fields). This will give even more savings as more tracepoints are created. Note, if just TRACE_EVENT()s are used and not DECLARE_EVENT_CLASS() with several DEFINE_EVENT()s, then the savings will be lost. But we are pushing developers to consolidate events with DEFINE_EVENT() so this should not be an issue. The kprobes define a unique class to every new event, but are dynamic so it should not be a issue. The syscalls however have a single class but the fields for the individual events are different. The syscalls use a metadata to define the fields. I moved the fields list from the event to the metadata and added a "get_fields()" function to the class. This function is used to find the fields. For normal events and kprobes, get_fields() just returns a pointer to the fields list_head in the class. For syscall events, it returns the fields list_head in the metadata for the event. v2: Fixed the syscall fields. The syscall metadata needs a list of fields for both enter and exit. Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Acked-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Acked-by: Masami Hiramatsu <mhiramat@redhat.com> Cc: Tom Zanussi <tzanussi@gmail.com> Cc: Peter Zijlstra <peterz@infradead.org> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2010-04-22 14:35:55 +00:00
list_for_each_entry(field, head, 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;
}
static struct ftrace_event_field *
find_event_field(struct ftrace_event_call *call, char *name)
{
struct ftrace_event_field *field;
struct list_head *head;
field = __find_event_field(&ftrace_common_fields, name);
if (field)
return field;
head = trace_get_fields(call);
return __find_event_field(head, name);
}
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
static int __alloc_pred_stack(struct pred_stack *stack, int n_preds)
{
stack->preds = kzalloc(sizeof(*stack->preds)*(n_preds + 1), GFP_KERNEL);
if (!stack->preds)
return -ENOMEM;
stack->index = n_preds;
return 0;
}
static void __free_pred_stack(struct pred_stack *stack)
{
kfree(stack->preds);
stack->index = 0;
}
static int __push_pred_stack(struct pred_stack *stack,
struct filter_pred *pred)
{
int index = stack->index;
if (WARN_ON(index == 0))
return -ENOSPC;
stack->preds[--index] = pred;
stack->index = index;
return 0;
}
static struct filter_pred *
__pop_pred_stack(struct pred_stack *stack)
{
struct filter_pred *pred;
int index = stack->index;
pred = stack->preds[index++];
if (!pred)
return NULL;
stack->index = index;
return pred;
}
static int filter_set_pred(struct event_filter *filter,
int idx,
struct pred_stack *stack,
struct filter_pred *src)
tracing/filters: allow on-the-fly filter switching This patch allows event filters to be safely removed or switched on-the-fly while avoiding the use of rcu or the suspension of tracing of previous versions. It does it by adding a new filter_pred_none() predicate function which does nothing and by never deallocating either the predicates or any of the filter_pred members used in matching; the predicate lists are allocated and initialized during ftrace_event_calls initialization. Whenever a filter is removed or replaced, the filter_pred_* functions currently in use by the affected ftrace_event_call are immediately switched over to to the filter_pred_none() function, while the rest of the filter_pred members are left intact, allowing any currently executing filter_pred_* functions to finish up, using the values they're currently using. In the case of filter replacement, the new predicate values are copied into the old predicates after the above step, and the filter_pred_none() functions are replaced by the filter_pred_* functions for the new filter. In this case, it is possible though very unlikely that a previous filter_pred_* is still running even after the filter_pred_none() switch and the switch to the new filter_pred_*. In that case, however, because nothing has been deallocated in the filter_pred, the worst that can happen is that the old filter_pred_* function sees the new values and as a result produces either a false positive or a false negative, depending on the values it finds. So one downside to this method is that rarely, it can produce a bad match during the filter switch, but it should be possible to live with that, IMHO. The other downside is that at least in this patch the predicate lists are always pre-allocated, taking up memory from the start. They could probably be allocated on first-use, and de-allocated when tracing is completely stopped - if this patch makes sense, I could create another one to do that later on. Oh, and it also places a restriction on the size of __arrays in events, currently set to 128, since they can't be larger than the now embedded str_val arrays in the filter_pred struct. Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: paulmck@linux.vnet.ibm.com LKML-Reference: <1239610670.6660.49.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-13 08:17:50 +00:00
{
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
struct filter_pred *dest = &filter->preds[idx];
struct filter_pred *left;
struct filter_pred *right;
tracing/filters: allow on-the-fly filter switching This patch allows event filters to be safely removed or switched on-the-fly while avoiding the use of rcu or the suspension of tracing of previous versions. It does it by adding a new filter_pred_none() predicate function which does nothing and by never deallocating either the predicates or any of the filter_pred members used in matching; the predicate lists are allocated and initialized during ftrace_event_calls initialization. Whenever a filter is removed or replaced, the filter_pred_* functions currently in use by the affected ftrace_event_call are immediately switched over to to the filter_pred_none() function, while the rest of the filter_pred members are left intact, allowing any currently executing filter_pred_* functions to finish up, using the values they're currently using. In the case of filter replacement, the new predicate values are copied into the old predicates after the above step, and the filter_pred_none() functions are replaced by the filter_pred_* functions for the new filter. In this case, it is possible though very unlikely that a previous filter_pred_* is still running even after the filter_pred_none() switch and the switch to the new filter_pred_*. In that case, however, because nothing has been deallocated in the filter_pred, the worst that can happen is that the old filter_pred_* function sees the new values and as a result produces either a false positive or a false negative, depending on the values it finds. So one downside to this method is that rarely, it can produce a bad match during the filter switch, but it should be possible to live with that, IMHO. The other downside is that at least in this patch the predicate lists are always pre-allocated, taking up memory from the start. They could probably be allocated on first-use, and de-allocated when tracing is completely stopped - if this patch makes sense, I could create another one to do that later on. Oh, and it also places a restriction on the size of __arrays in events, currently set to 128, since they can't be larger than the now embedded str_val arrays in the filter_pred struct. Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: paulmck@linux.vnet.ibm.com LKML-Reference: <1239610670.6660.49.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-13 08:17:50 +00:00
*dest = *src;
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
dest->index = idx;
tracing/filters: allow on-the-fly filter switching This patch allows event filters to be safely removed or switched on-the-fly while avoiding the use of rcu or the suspension of tracing of previous versions. It does it by adding a new filter_pred_none() predicate function which does nothing and by never deallocating either the predicates or any of the filter_pred members used in matching; the predicate lists are allocated and initialized during ftrace_event_calls initialization. Whenever a filter is removed or replaced, the filter_pred_* functions currently in use by the affected ftrace_event_call are immediately switched over to to the filter_pred_none() function, while the rest of the filter_pred members are left intact, allowing any currently executing filter_pred_* functions to finish up, using the values they're currently using. In the case of filter replacement, the new predicate values are copied into the old predicates after the above step, and the filter_pred_none() functions are replaced by the filter_pred_* functions for the new filter. In this case, it is possible though very unlikely that a previous filter_pred_* is still running even after the filter_pred_none() switch and the switch to the new filter_pred_*. In that case, however, because nothing has been deallocated in the filter_pred, the worst that can happen is that the old filter_pred_* function sees the new values and as a result produces either a false positive or a false negative, depending on the values it finds. So one downside to this method is that rarely, it can produce a bad match during the filter switch, but it should be possible to live with that, IMHO. The other downside is that at least in this patch the predicate lists are always pre-allocated, taking up memory from the start. They could probably be allocated on first-use, and de-allocated when tracing is completely stopped - if this patch makes sense, I could create another one to do that later on. Oh, and it also places a restriction on the size of __arrays in events, currently set to 128, since they can't be larger than the now embedded str_val arrays in the filter_pred struct. Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: paulmck@linux.vnet.ibm.com LKML-Reference: <1239610670.6660.49.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-13 08:17:50 +00:00
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
if (dest->op == OP_OR || dest->op == OP_AND) {
right = __pop_pred_stack(stack);
left = __pop_pred_stack(stack);
if (!left || !right)
return -EINVAL;
/*
* If both children can be folded
* and they are the same op as this op or a leaf,
* then this op can be folded.
*/
if (left->index & FILTER_PRED_FOLD &&
(left->op == dest->op ||
left->left == FILTER_PRED_INVALID) &&
right->index & FILTER_PRED_FOLD &&
(right->op == dest->op ||
right->left == FILTER_PRED_INVALID))
dest->index |= FILTER_PRED_FOLD;
dest->left = left->index & ~FILTER_PRED_FOLD;
dest->right = right->index & ~FILTER_PRED_FOLD;
left->parent = dest->index & ~FILTER_PRED_FOLD;
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
right->parent = dest->index | FILTER_PRED_IS_RIGHT;
} else {
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
/*
* Make dest->left invalid to be used as a quick
* way to know this is a leaf node.
*/
dest->left = FILTER_PRED_INVALID;
/* All leafs allow folding the parent ops. */
dest->index |= FILTER_PRED_FOLD;
}
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
return __push_pred_stack(stack, dest);
tracing/filters: allow on-the-fly filter switching This patch allows event filters to be safely removed or switched on-the-fly while avoiding the use of rcu or the suspension of tracing of previous versions. It does it by adding a new filter_pred_none() predicate function which does nothing and by never deallocating either the predicates or any of the filter_pred members used in matching; the predicate lists are allocated and initialized during ftrace_event_calls initialization. Whenever a filter is removed or replaced, the filter_pred_* functions currently in use by the affected ftrace_event_call are immediately switched over to to the filter_pred_none() function, while the rest of the filter_pred members are left intact, allowing any currently executing filter_pred_* functions to finish up, using the values they're currently using. In the case of filter replacement, the new predicate values are copied into the old predicates after the above step, and the filter_pred_none() functions are replaced by the filter_pred_* functions for the new filter. In this case, it is possible though very unlikely that a previous filter_pred_* is still running even after the filter_pred_none() switch and the switch to the new filter_pred_*. In that case, however, because nothing has been deallocated in the filter_pred, the worst that can happen is that the old filter_pred_* function sees the new values and as a result produces either a false positive or a false negative, depending on the values it finds. So one downside to this method is that rarely, it can produce a bad match during the filter switch, but it should be possible to live with that, IMHO. The other downside is that at least in this patch the predicate lists are always pre-allocated, taking up memory from the start. They could probably be allocated on first-use, and de-allocated when tracing is completely stopped - if this patch makes sense, I could create another one to do that later on. Oh, and it also places a restriction on the size of __arrays in events, currently set to 128, since they can't be larger than the now embedded str_val arrays in the filter_pred struct. Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: paulmck@linux.vnet.ibm.com LKML-Reference: <1239610670.6660.49.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-13 08:17:50 +00:00
}
static void __free_preds(struct event_filter *filter)
{
if (filter->preds) {
kfree(filter->preds);
filter->preds = NULL;
}
filter->a_preds = 0;
filter->n_preds = 0;
}
static void filter_disable(struct ftrace_event_call *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
{
call->flags &= ~TRACE_EVENT_FL_FILTERED;
tracing/filters: allow on-the-fly filter switching This patch allows event filters to be safely removed or switched on-the-fly while avoiding the use of rcu or the suspension of tracing of previous versions. It does it by adding a new filter_pred_none() predicate function which does nothing and by never deallocating either the predicates or any of the filter_pred members used in matching; the predicate lists are allocated and initialized during ftrace_event_calls initialization. Whenever a filter is removed or replaced, the filter_pred_* functions currently in use by the affected ftrace_event_call are immediately switched over to to the filter_pred_none() function, while the rest of the filter_pred members are left intact, allowing any currently executing filter_pred_* functions to finish up, using the values they're currently using. In the case of filter replacement, the new predicate values are copied into the old predicates after the above step, and the filter_pred_none() functions are replaced by the filter_pred_* functions for the new filter. In this case, it is possible though very unlikely that a previous filter_pred_* is still running even after the filter_pred_none() switch and the switch to the new filter_pred_*. In that case, however, because nothing has been deallocated in the filter_pred, the worst that can happen is that the old filter_pred_* function sees the new values and as a result produces either a false positive or a false negative, depending on the values it finds. So one downside to this method is that rarely, it can produce a bad match during the filter switch, but it should be possible to live with that, IMHO. The other downside is that at least in this patch the predicate lists are always pre-allocated, taking up memory from the start. They could probably be allocated on first-use, and de-allocated when tracing is completely stopped - if this patch makes sense, I could create another one to do that later on. Oh, and it also places a restriction on the size of __arrays in events, currently set to 128, since they can't be larger than the now embedded str_val arrays in the filter_pred struct. Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: paulmck@linux.vnet.ibm.com LKML-Reference: <1239610670.6660.49.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-13 08:17:50 +00:00
}
static void __free_filter(struct event_filter *filter)
{
if (!filter)
return;
__free_preds(filter);
kfree(filter->filter_string);
kfree(filter);
}
/*
* Called when destroying the ftrace_event_call.
* The call is being freed, so we do not need to worry about
* the call being currently used. This is for module code removing
* the tracepoints from within it.
*/
void destroy_preds(struct ftrace_event_call *call)
{
__free_filter(call->filter);
call->filter = NULL;
}
static struct event_filter *__alloc_filter(void)
tracing/filters: allow on-the-fly filter switching This patch allows event filters to be safely removed or switched on-the-fly while avoiding the use of rcu or the suspension of tracing of previous versions. It does it by adding a new filter_pred_none() predicate function which does nothing and by never deallocating either the predicates or any of the filter_pred members used in matching; the predicate lists are allocated and initialized during ftrace_event_calls initialization. Whenever a filter is removed or replaced, the filter_pred_* functions currently in use by the affected ftrace_event_call are immediately switched over to to the filter_pred_none() function, while the rest of the filter_pred members are left intact, allowing any currently executing filter_pred_* functions to finish up, using the values they're currently using. In the case of filter replacement, the new predicate values are copied into the old predicates after the above step, and the filter_pred_none() functions are replaced by the filter_pred_* functions for the new filter. In this case, it is possible though very unlikely that a previous filter_pred_* is still running even after the filter_pred_none() switch and the switch to the new filter_pred_*. In that case, however, because nothing has been deallocated in the filter_pred, the worst that can happen is that the old filter_pred_* function sees the new values and as a result produces either a false positive or a false negative, depending on the values it finds. So one downside to this method is that rarely, it can produce a bad match during the filter switch, but it should be possible to live with that, IMHO. The other downside is that at least in this patch the predicate lists are always pre-allocated, taking up memory from the start. They could probably be allocated on first-use, and de-allocated when tracing is completely stopped - if this patch makes sense, I could create another one to do that later on. Oh, and it also places a restriction on the size of __arrays in events, currently set to 128, since they can't be larger than the now embedded str_val arrays in the filter_pred struct. Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: paulmck@linux.vnet.ibm.com LKML-Reference: <1239610670.6660.49.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-13 08:17:50 +00:00
{
struct event_filter *filter;
tracing/filters: allow on-the-fly filter switching This patch allows event filters to be safely removed or switched on-the-fly while avoiding the use of rcu or the suspension of tracing of previous versions. It does it by adding a new filter_pred_none() predicate function which does nothing and by never deallocating either the predicates or any of the filter_pred members used in matching; the predicate lists are allocated and initialized during ftrace_event_calls initialization. Whenever a filter is removed or replaced, the filter_pred_* functions currently in use by the affected ftrace_event_call are immediately switched over to to the filter_pred_none() function, while the rest of the filter_pred members are left intact, allowing any currently executing filter_pred_* functions to finish up, using the values they're currently using. In the case of filter replacement, the new predicate values are copied into the old predicates after the above step, and the filter_pred_none() functions are replaced by the filter_pred_* functions for the new filter. In this case, it is possible though very unlikely that a previous filter_pred_* is still running even after the filter_pred_none() switch and the switch to the new filter_pred_*. In that case, however, because nothing has been deallocated in the filter_pred, the worst that can happen is that the old filter_pred_* function sees the new values and as a result produces either a false positive or a false negative, depending on the values it finds. So one downside to this method is that rarely, it can produce a bad match during the filter switch, but it should be possible to live with that, IMHO. The other downside is that at least in this patch the predicate lists are always pre-allocated, taking up memory from the start. They could probably be allocated on first-use, and de-allocated when tracing is completely stopped - if this patch makes sense, I could create another one to do that later on. Oh, and it also places a restriction on the size of __arrays in events, currently set to 128, since they can't be larger than the now embedded str_val arrays in the filter_pred struct. Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: paulmck@linux.vnet.ibm.com LKML-Reference: <1239610670.6660.49.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-13 08:17:50 +00:00
filter = kzalloc(sizeof(*filter), GFP_KERNEL);
return filter;
}
static int __alloc_preds(struct event_filter *filter, int n_preds)
{
struct filter_pred *pred;
int i;
if (filter->preds)
__free_preds(filter);
filter->preds =
kzalloc(sizeof(*filter->preds) * n_preds, GFP_KERNEL);
if (!filter->preds)
return -ENOMEM;
filter->a_preds = n_preds;
filter->n_preds = 0;
for (i = 0; i < n_preds; i++) {
pred = &filter->preds[i];
tracing/filters: allow on-the-fly filter switching This patch allows event filters to be safely removed or switched on-the-fly while avoiding the use of rcu or the suspension of tracing of previous versions. It does it by adding a new filter_pred_none() predicate function which does nothing and by never deallocating either the predicates or any of the filter_pred members used in matching; the predicate lists are allocated and initialized during ftrace_event_calls initialization. Whenever a filter is removed or replaced, the filter_pred_* functions currently in use by the affected ftrace_event_call are immediately switched over to to the filter_pred_none() function, while the rest of the filter_pred members are left intact, allowing any currently executing filter_pred_* functions to finish up, using the values they're currently using. In the case of filter replacement, the new predicate values are copied into the old predicates after the above step, and the filter_pred_none() functions are replaced by the filter_pred_* functions for the new filter. In this case, it is possible though very unlikely that a previous filter_pred_* is still running even after the filter_pred_none() switch and the switch to the new filter_pred_*. In that case, however, because nothing has been deallocated in the filter_pred, the worst that can happen is that the old filter_pred_* function sees the new values and as a result produces either a false positive or a false negative, depending on the values it finds. So one downside to this method is that rarely, it can produce a bad match during the filter switch, but it should be possible to live with that, IMHO. The other downside is that at least in this patch the predicate lists are always pre-allocated, taking up memory from the start. They could probably be allocated on first-use, and de-allocated when tracing is completely stopped - if this patch makes sense, I could create another one to do that later on. Oh, and it also places a restriction on the size of __arrays in events, currently set to 128, since they can't be larger than the now embedded str_val arrays in the filter_pred struct. Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: paulmck@linux.vnet.ibm.com LKML-Reference: <1239610670.6660.49.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-13 08:17:50 +00:00
pred->fn = filter_pred_none;
}
return 0;
}
static void filter_free_subsystem_preds(struct event_subsystem *system)
{
struct ftrace_event_call *call;
list_for_each_entry(call, &ftrace_events, list) {
if (strcmp(call->class->system, system->name) != 0)
continue;
filter_disable(call);
remove_filter_string(call->filter);
}
}
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 void filter_free_subsystem_filters(struct event_subsystem *system)
{
struct ftrace_event_call *call;
list_for_each_entry(call, &ftrace_events, list) {
if (strcmp(call->class->system, system->name) != 0)
continue;
__free_filter(call->filter);
call->filter = NULL;
}
}
static int filter_add_pred(struct filter_parse_state *ps,
struct event_filter *filter,
struct filter_pred *pred,
struct pred_stack *stack)
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 err;
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 (WARN_ON(filter->n_preds == filter->a_preds)) {
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
parse_error(ps, FILT_ERR_TOO_MANY_PREDS, 0);
tracing/filters: allow on-the-fly filter switching This patch allows event filters to be safely removed or switched on-the-fly while avoiding the use of rcu or the suspension of tracing of previous versions. It does it by adding a new filter_pred_none() predicate function which does nothing and by never deallocating either the predicates or any of the filter_pred members used in matching; the predicate lists are allocated and initialized during ftrace_event_calls initialization. Whenever a filter is removed or replaced, the filter_pred_* functions currently in use by the affected ftrace_event_call are immediately switched over to to the filter_pred_none() function, while the rest of the filter_pred members are left intact, allowing any currently executing filter_pred_* functions to finish up, using the values they're currently using. In the case of filter replacement, the new predicate values are copied into the old predicates after the above step, and the filter_pred_none() functions are replaced by the filter_pred_* functions for the new filter. In this case, it is possible though very unlikely that a previous filter_pred_* is still running even after the filter_pred_none() switch and the switch to the new filter_pred_*. In that case, however, because nothing has been deallocated in the filter_pred, the worst that can happen is that the old filter_pred_* function sees the new values and as a result produces either a false positive or a false negative, depending on the values it finds. So one downside to this method is that rarely, it can produce a bad match during the filter switch, but it should be possible to live with that, IMHO. The other downside is that at least in this patch the predicate lists are always pre-allocated, taking up memory from the start. They could probably be allocated on first-use, and de-allocated when tracing is completely stopped - if this patch makes sense, I could create another one to do that later on. Oh, and it also places a restriction on the size of __arrays in events, currently set to 128, since they can't be larger than the now embedded str_val arrays in the filter_pred struct. Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: paulmck@linux.vnet.ibm.com LKML-Reference: <1239610670.6660.49.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-13 08:17:50 +00:00
return -ENOSPC;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
}
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
err = filter_set_pred(filter, filter->n_preds, stack, pred);
tracing/filters: allow on-the-fly filter switching This patch allows event filters to be safely removed or switched on-the-fly while avoiding the use of rcu or the suspension of tracing of previous versions. It does it by adding a new filter_pred_none() predicate function which does nothing and by never deallocating either the predicates or any of the filter_pred members used in matching; the predicate lists are allocated and initialized during ftrace_event_calls initialization. Whenever a filter is removed or replaced, the filter_pred_* functions currently in use by the affected ftrace_event_call are immediately switched over to to the filter_pred_none() function, while the rest of the filter_pred members are left intact, allowing any currently executing filter_pred_* functions to finish up, using the values they're currently using. In the case of filter replacement, the new predicate values are copied into the old predicates after the above step, and the filter_pred_none() functions are replaced by the filter_pred_* functions for the new filter. In this case, it is possible though very unlikely that a previous filter_pred_* is still running even after the filter_pred_none() switch and the switch to the new filter_pred_*. In that case, however, because nothing has been deallocated in the filter_pred, the worst that can happen is that the old filter_pred_* function sees the new values and as a result produces either a false positive or a false negative, depending on the values it finds. So one downside to this method is that rarely, it can produce a bad match during the filter switch, but it should be possible to live with that, IMHO. The other downside is that at least in this patch the predicate lists are always pre-allocated, taking up memory from the start. They could probably be allocated on first-use, and de-allocated when tracing is completely stopped - if this patch makes sense, I could create another one to do that later on. Oh, and it also places a restriction on the size of __arrays in events, currently set to 128, since they can't be larger than the now embedded str_val arrays in the filter_pred struct. Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: paulmck@linux.vnet.ibm.com LKML-Reference: <1239610670.6660.49.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-13 08:17:50 +00:00
if (err)
return err;
filter->n_preds++;
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
tracing/filters: allow on-the-fly filter switching This patch allows event filters to be safely removed or switched on-the-fly while avoiding the use of rcu or the suspension of tracing of previous versions. It does it by adding a new filter_pred_none() predicate function which does nothing and by never deallocating either the predicates or any of the filter_pred members used in matching; the predicate lists are allocated and initialized during ftrace_event_calls initialization. Whenever a filter is removed or replaced, the filter_pred_* functions currently in use by the affected ftrace_event_call are immediately switched over to to the filter_pred_none() function, while the rest of the filter_pred members are left intact, allowing any currently executing filter_pred_* functions to finish up, using the values they're currently using. In the case of filter replacement, the new predicate values are copied into the old predicates after the above step, and the filter_pred_none() functions are replaced by the filter_pred_* functions for the new filter. In this case, it is possible though very unlikely that a previous filter_pred_* is still running even after the filter_pred_none() switch and the switch to the new filter_pred_*. In that case, however, because nothing has been deallocated in the filter_pred, the worst that can happen is that the old filter_pred_* function sees the new values and as a result produces either a false positive or a false negative, depending on the values it finds. So one downside to this method is that rarely, it can produce a bad match during the filter switch, but it should be possible to live with that, IMHO. The other downside is that at least in this patch the predicate lists are always pre-allocated, taking up memory from the start. They could probably be allocated on first-use, and de-allocated when tracing is completely stopped - if this patch makes sense, I could create another one to do that later on. Oh, and it also places a restriction on the size of __arrays in events, currently set to 128, since they can't be larger than the now embedded str_val arrays in the filter_pred struct. Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: paulmck@linux.vnet.ibm.com LKML-Reference: <1239610670.6660.49.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-13 08:17:50 +00:00
return 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
}
int filter_assign_type(const char *type)
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 (strstr(type, "__data_loc") && strstr(type, "char"))
return FILTER_DYN_STRING;
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 (strchr(type, '[') && strstr(type, "char"))
return FILTER_STATIC_STRING;
return FILTER_OTHER;
}
static bool is_string_field(struct ftrace_event_field *field)
{
return field->filter_type == FILTER_DYN_STRING ||
field->filter_type == FILTER_STATIC_STRING ||
field->filter_type == FILTER_PTR_STRING;
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
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
static int is_legal_op(struct ftrace_event_field *field, int op)
{
if (is_string_field(field) &&
(op != OP_EQ && op != OP_NE && op != OP_GLOB))
return 0;
if (!is_string_field(field) && op == OP_GLOB)
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
return 0;
return 1;
}
static filter_pred_fn_t select_comparison_fn(int op, int field_size,
int field_is_signed)
{
filter_pred_fn_t fn = NULL;
switch (field_size) {
case 8:
if (op == OP_EQ || op == OP_NE)
fn = filter_pred_64;
else if (field_is_signed)
fn = filter_pred_s64;
else
fn = filter_pred_u64;
break;
case 4:
if (op == OP_EQ || op == OP_NE)
fn = filter_pred_32;
else if (field_is_signed)
fn = filter_pred_s32;
else
fn = filter_pred_u32;
break;
case 2:
if (op == OP_EQ || op == OP_NE)
fn = filter_pred_16;
else if (field_is_signed)
fn = filter_pred_s16;
else
fn = filter_pred_u16;
break;
case 1:
if (op == OP_EQ || op == OP_NE)
fn = filter_pred_8;
else if (field_is_signed)
fn = filter_pred_s8;
else
fn = filter_pred_u8;
break;
}
return fn;
}
static int init_pred(struct filter_parse_state *ps,
struct ftrace_event_field *field,
struct filter_pred *pred)
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
{
filter_pred_fn_t fn = filter_pred_none;
unsigned long long val;
int ret;
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->offset = field->offset;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
if (!is_legal_op(field, pred->op)) {
parse_error(ps, FILT_ERR_ILLEGAL_FIELD_OP, 0);
return -EINVAL;
}
if (is_string_field(field)) {
filter_build_regex(pred);
if (field->filter_type == FILTER_STATIC_STRING) {
fn = filter_pred_string;
pred->regex.field_len = field->size;
} else if (field->filter_type == FILTER_DYN_STRING)
fn = filter_pred_strloc;
else
fn = filter_pred_pchar;
} else {
if (field->is_signed)
ret = strict_strtoll(pred->regex.pattern, 0, &val);
else
ret = strict_strtoull(pred->regex.pattern, 0, &val);
if (ret) {
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
parse_error(ps, FILT_ERR_ILLEGAL_INTVAL, 0);
return -EINVAL;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
}
pred->val = 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
fn = select_comparison_fn(pred->op, field->size,
field->is_signed);
if (!fn) {
parse_error(ps, FILT_ERR_INVALID_OP, 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
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
if (pred->op == OP_NE)
pred->not = 1;
pred->fn = fn;
return 0;
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
static void parse_init(struct filter_parse_state *ps,
struct filter_op *ops,
char *infix_string)
{
memset(ps, '\0', sizeof(*ps));
ps->infix.string = infix_string;
ps->infix.cnt = strlen(infix_string);
ps->ops = ops;
INIT_LIST_HEAD(&ps->opstack);
INIT_LIST_HEAD(&ps->postfix);
}
static char infix_next(struct filter_parse_state *ps)
{
ps->infix.cnt--;
return ps->infix.string[ps->infix.tail++];
}
static char infix_peek(struct filter_parse_state *ps)
{
if (ps->infix.tail == strlen(ps->infix.string))
return 0;
return ps->infix.string[ps->infix.tail];
}
static void infix_advance(struct filter_parse_state *ps)
{
ps->infix.cnt--;
ps->infix.tail++;
}
static inline int is_precedence_lower(struct filter_parse_state *ps,
int a, int b)
{
return ps->ops[a].precedence < ps->ops[b].precedence;
}
static inline int is_op_char(struct filter_parse_state *ps, char c)
{
int i;
for (i = 0; strcmp(ps->ops[i].string, "OP_NONE"); i++) {
if (ps->ops[i].string[0] == c)
return 1;
}
tracing/filters: allow on-the-fly filter switching This patch allows event filters to be safely removed or switched on-the-fly while avoiding the use of rcu or the suspension of tracing of previous versions. It does it by adding a new filter_pred_none() predicate function which does nothing and by never deallocating either the predicates or any of the filter_pred members used in matching; the predicate lists are allocated and initialized during ftrace_event_calls initialization. Whenever a filter is removed or replaced, the filter_pred_* functions currently in use by the affected ftrace_event_call are immediately switched over to to the filter_pred_none() function, while the rest of the filter_pred members are left intact, allowing any currently executing filter_pred_* functions to finish up, using the values they're currently using. In the case of filter replacement, the new predicate values are copied into the old predicates after the above step, and the filter_pred_none() functions are replaced by the filter_pred_* functions for the new filter. In this case, it is possible though very unlikely that a previous filter_pred_* is still running even after the filter_pred_none() switch and the switch to the new filter_pred_*. In that case, however, because nothing has been deallocated in the filter_pred, the worst that can happen is that the old filter_pred_* function sees the new values and as a result produces either a false positive or a false negative, depending on the values it finds. So one downside to this method is that rarely, it can produce a bad match during the filter switch, but it should be possible to live with that, IMHO. The other downside is that at least in this patch the predicate lists are always pre-allocated, taking up memory from the start. They could probably be allocated on first-use, and de-allocated when tracing is completely stopped - if this patch makes sense, I could create another one to do that later on. Oh, and it also places a restriction on the size of __arrays in events, currently set to 128, since they can't be larger than the now embedded str_val arrays in the filter_pred struct. Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: paulmck@linux.vnet.ibm.com LKML-Reference: <1239610670.6660.49.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-13 08:17:50 +00:00
return 0;
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
static int infix_get_op(struct filter_parse_state *ps, char firstc)
{
char nextc = infix_peek(ps);
char opstr[3];
int i;
opstr[0] = firstc;
opstr[1] = nextc;
opstr[2] = '\0';
for (i = 0; strcmp(ps->ops[i].string, "OP_NONE"); i++) {
if (!strcmp(opstr, ps->ops[i].string)) {
infix_advance(ps);
return ps->ops[i].id;
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
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
}
opstr[1] = '\0';
for (i = 0; strcmp(ps->ops[i].string, "OP_NONE"); i++) {
if (!strcmp(opstr, ps->ops[i].string))
return ps->ops[i].id;
}
return OP_NONE;
}
static inline void clear_operand_string(struct filter_parse_state *ps)
{
memset(ps->operand.string, '\0', MAX_FILTER_STR_VAL);
ps->operand.tail = 0;
}
static inline int append_operand_char(struct filter_parse_state *ps, char c)
{
if (ps->operand.tail == MAX_FILTER_STR_VAL - 1)
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
return -EINVAL;
ps->operand.string[ps->operand.tail++] = c;
return 0;
}
static int filter_opstack_push(struct filter_parse_state *ps, int op)
{
struct opstack_op *opstack_op;
opstack_op = kmalloc(sizeof(*opstack_op), GFP_KERNEL);
if (!opstack_op)
return -ENOMEM;
opstack_op->op = op;
list_add(&opstack_op->list, &ps->opstack);
return 0;
}
static int filter_opstack_empty(struct filter_parse_state *ps)
{
return list_empty(&ps->opstack);
}
static int filter_opstack_top(struct filter_parse_state *ps)
{
struct opstack_op *opstack_op;
if (filter_opstack_empty(ps))
return OP_NONE;
opstack_op = list_first_entry(&ps->opstack, struct opstack_op, list);
return opstack_op->op;
}
static int filter_opstack_pop(struct filter_parse_state *ps)
{
struct opstack_op *opstack_op;
int op;
if (filter_opstack_empty(ps))
return OP_NONE;
opstack_op = list_first_entry(&ps->opstack, struct opstack_op, list);
op = opstack_op->op;
list_del(&opstack_op->list);
kfree(opstack_op);
return op;
}
static void filter_opstack_clear(struct filter_parse_state *ps)
{
while (!filter_opstack_empty(ps))
filter_opstack_pop(ps);
}
static char *curr_operand(struct filter_parse_state *ps)
{
return ps->operand.string;
}
static int postfix_append_operand(struct filter_parse_state *ps, char *operand)
{
struct postfix_elt *elt;
elt = kmalloc(sizeof(*elt), GFP_KERNEL);
if (!elt)
return -ENOMEM;
elt->op = OP_NONE;
elt->operand = kstrdup(operand, GFP_KERNEL);
if (!elt->operand) {
kfree(elt);
return -ENOMEM;
}
list_add_tail(&elt->list, &ps->postfix);
return 0;
}
static int postfix_append_op(struct filter_parse_state *ps, int op)
{
struct postfix_elt *elt;
elt = kmalloc(sizeof(*elt), GFP_KERNEL);
if (!elt)
return -ENOMEM;
elt->op = op;
elt->operand = NULL;
list_add_tail(&elt->list, &ps->postfix);
return 0;
}
static void postfix_clear(struct filter_parse_state *ps)
{
struct postfix_elt *elt;
while (!list_empty(&ps->postfix)) {
elt = list_first_entry(&ps->postfix, struct postfix_elt, list);
list_del(&elt->list);
kfree(elt->operand);
kfree(elt);
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
}
}
static int filter_parse(struct filter_parse_state *ps)
{
int in_string = 0;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
int op, top_op;
char ch;
while ((ch = infix_next(ps))) {
if (ch == '"') {
in_string ^= 1;
continue;
}
if (in_string)
goto parse_operand;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
if (isspace(ch))
continue;
if (is_op_char(ps, ch)) {
op = infix_get_op(ps, ch);
if (op == OP_NONE) {
parse_error(ps, FILT_ERR_INVALID_OP, 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
return -EINVAL;
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
if (strlen(curr_operand(ps))) {
postfix_append_operand(ps, curr_operand(ps));
clear_operand_string(ps);
}
while (!filter_opstack_empty(ps)) {
top_op = filter_opstack_top(ps);
if (!is_precedence_lower(ps, top_op, op)) {
top_op = filter_opstack_pop(ps);
postfix_append_op(ps, top_op);
continue;
}
break;
}
filter_opstack_push(ps, op);
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
continue;
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
if (ch == '(') {
filter_opstack_push(ps, OP_OPEN_PAREN);
continue;
}
if (ch == ')') {
if (strlen(curr_operand(ps))) {
postfix_append_operand(ps, curr_operand(ps));
clear_operand_string(ps);
}
top_op = filter_opstack_pop(ps);
while (top_op != OP_NONE) {
if (top_op == OP_OPEN_PAREN)
break;
postfix_append_op(ps, top_op);
top_op = filter_opstack_pop(ps);
}
if (top_op == OP_NONE) {
parse_error(ps, FILT_ERR_UNBALANCED_PAREN, 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
}
continue;
}
parse_operand:
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
if (append_operand_char(ps, ch)) {
parse_error(ps, FILT_ERR_OPERAND_TOO_LONG, 0);
return -EINVAL;
}
}
if (strlen(curr_operand(ps)))
postfix_append_operand(ps, curr_operand(ps));
while (!filter_opstack_empty(ps)) {
top_op = filter_opstack_pop(ps);
if (top_op == OP_NONE)
break;
if (top_op == OP_OPEN_PAREN) {
parse_error(ps, FILT_ERR_UNBALANCED_PAREN, 0);
return -EINVAL;
}
postfix_append_op(ps, top_op);
}
return 0;
}
static struct filter_pred *create_pred(struct filter_parse_state *ps,
struct ftrace_event_call *call,
int op, char *operand1, char *operand2)
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
{
struct ftrace_event_field *field;
static struct filter_pred pred;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
memset(&pred, 0, sizeof(pred));
pred.op = op;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
if (op == OP_AND || op == OP_OR)
return &pred;
if (!operand1 || !operand2) {
parse_error(ps, FILT_ERR_MISSING_FIELD, 0);
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
return NULL;
}
field = find_event_field(call, operand1);
if (!field) {
parse_error(ps, FILT_ERR_FIELD_NOT_FOUND, 0);
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
return NULL;
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
strcpy(pred.regex.pattern, operand2);
pred.regex.len = strlen(pred.regex.pattern);
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
return init_pred(ps, field, &pred) ? NULL : &pred;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
}
static int check_preds(struct filter_parse_state *ps)
{
int n_normal_preds = 0, n_logical_preds = 0;
struct postfix_elt *elt;
list_for_each_entry(elt, &ps->postfix, list) {
if (elt->op == OP_NONE)
continue;
if (elt->op == OP_AND || elt->op == OP_OR) {
n_logical_preds++;
continue;
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
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
n_normal_preds++;
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
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
if (!n_normal_preds || n_logical_preds >= n_normal_preds) {
parse_error(ps, FILT_ERR_INVALID_FILTER, 0);
return -EINVAL;
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
return 0;
}
static int count_preds(struct filter_parse_state *ps)
{
struct postfix_elt *elt;
int n_preds = 0;
list_for_each_entry(elt, &ps->postfix, list) {
if (elt->op == OP_NONE)
continue;
n_preds++;
}
return n_preds;
}
struct check_pred_data {
int count;
int max;
};
static int check_pred_tree_cb(enum move_type move, struct filter_pred *pred,
int *err, void *data)
{
struct check_pred_data *d = data;
if (WARN_ON(d->count++ > d->max)) {
*err = -EINVAL;
return WALK_PRED_ABORT;
}
return WALK_PRED_DEFAULT;
}
/*
* The tree is walked at filtering of an event. If the tree is not correctly
* built, it may cause an infinite loop. Check here that the tree does
* indeed terminate.
*/
static int check_pred_tree(struct event_filter *filter,
struct filter_pred *root)
{
struct check_pred_data data = {
/*
* The max that we can hit a node is three times.
* Once going down, once coming up from left, and
* once coming up from right. This is more than enough
* since leafs are only hit a single time.
*/
.max = 3 * filter->n_preds,
.count = 0,
};
return walk_pred_tree(filter->preds, root,
check_pred_tree_cb, &data);
}
static int count_leafs_cb(enum move_type move, struct filter_pred *pred,
int *err, void *data)
{
int *count = data;
if ((move == MOVE_DOWN) &&
(pred->left == FILTER_PRED_INVALID))
(*count)++;
return WALK_PRED_DEFAULT;
}
static int count_leafs(struct filter_pred *preds, struct filter_pred *root)
{
int count = 0, ret;
ret = walk_pred_tree(preds, root, count_leafs_cb, &count);
WARN_ON(ret);
return count;
}
static int fold_pred(struct filter_pred *preds, struct filter_pred *root)
{
struct filter_pred *pred;
enum move_type move = MOVE_DOWN;
int count = 0;
int children;
int done = 0;
/* No need to keep the fold flag */
root->index &= ~FILTER_PRED_FOLD;
/* If the root is a leaf then do nothing */
if (root->left == FILTER_PRED_INVALID)
return 0;
/* count the children */
children = count_leafs(preds, &preds[root->left]);
children += count_leafs(preds, &preds[root->right]);
root->ops = kzalloc(sizeof(*root->ops) * children, GFP_KERNEL);
if (!root->ops)
return -ENOMEM;
root->val = children;
pred = root;
do {
switch (move) {
case MOVE_DOWN:
if (pred->left != FILTER_PRED_INVALID) {
pred = &preds[pred->left];
continue;
}
if (WARN_ON(count == children))
return -EINVAL;
pred->index &= ~FILTER_PRED_FOLD;
root->ops[count++] = pred->index;
pred = get_pred_parent(pred, preds,
pred->parent, &move);
continue;
case MOVE_UP_FROM_LEFT:
pred = &preds[pred->right];
move = MOVE_DOWN;
continue;
case MOVE_UP_FROM_RIGHT:
if (pred == root)
break;
pred = get_pred_parent(pred, preds,
pred->parent, &move);
continue;
}
done = 1;
} while (!done);
return 0;
}
/*
* To optimize the processing of the ops, if we have several "ors" or
* "ands" together, we can put them in an array and process them all
* together speeding up the filter logic.
*/
static int fold_pred_tree(struct event_filter *filter,
struct filter_pred *root)
{
struct filter_pred *preds;
struct filter_pred *pred;
enum move_type move = MOVE_DOWN;
int done = 0;
int err;
preds = filter->preds;
if (!preds)
return -EINVAL;
pred = root;
do {
switch (move) {
case MOVE_DOWN:
if (pred->index & FILTER_PRED_FOLD) {
err = fold_pred(preds, pred);
if (err)
return err;
/* Folded nodes are like leafs */
} else if (pred->left != FILTER_PRED_INVALID) {
pred = &preds[pred->left];
continue;
}
/* A leaf at the root is just a leaf in the tree */
if (pred == root)
break;
pred = get_pred_parent(pred, preds,
pred->parent, &move);
continue;
case MOVE_UP_FROM_LEFT:
pred = &preds[pred->right];
move = MOVE_DOWN;
continue;
case MOVE_UP_FROM_RIGHT:
if (pred == root)
break;
pred = get_pred_parent(pred, preds,
pred->parent, &move);
continue;
}
done = 1;
} while (!done);
return 0;
}
static int replace_preds(struct ftrace_event_call *call,
struct event_filter *filter,
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
struct filter_parse_state *ps,
char *filter_string,
bool dry_run)
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
{
char *operand1 = NULL, *operand2 = NULL;
struct filter_pred *pred;
struct filter_pred *root;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
struct postfix_elt *elt;
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
struct pred_stack stack = { }; /* init to NULL */
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
int err;
int n_preds = 0;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
n_preds = count_preds(ps);
if (n_preds >= MAX_FILTER_PRED) {
parse_error(ps, FILT_ERR_TOO_MANY_PREDS, 0);
return -ENOSPC;
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
err = check_preds(ps);
if (err)
return err;
if (!dry_run) {
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
err = __alloc_pred_stack(&stack, n_preds);
if (err)
return err;
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
err = __alloc_preds(filter, n_preds);
if (err)
goto fail;
}
n_preds = 0;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
list_for_each_entry(elt, &ps->postfix, list) {
if (elt->op == OP_NONE) {
if (!operand1)
operand1 = elt->operand;
else if (!operand2)
operand2 = elt->operand;
else {
parse_error(ps, FILT_ERR_TOO_MANY_OPERANDS, 0);
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
err = -EINVAL;
goto fail;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
}
continue;
}
if (WARN_ON(n_preds++ == MAX_FILTER_PRED)) {
parse_error(ps, FILT_ERR_TOO_MANY_PREDS, 0);
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
err = -ENOSPC;
goto fail;
}
pred = create_pred(ps, call, elt->op, operand1, operand2);
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
if (!pred) {
err = -EINVAL;
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
goto fail;
}
if (!dry_run) {
err = filter_add_pred(ps, filter, pred, &stack);
if (err)
goto fail;
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
operand1 = operand2 = NULL;
}
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
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
if (!dry_run) {
/* We should have one item left on the stack */
pred = __pop_pred_stack(&stack);
if (!pred)
return -EINVAL;
/* This item is where we start from in matching */
root = pred;
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
/* Make sure the stack is empty */
pred = __pop_pred_stack(&stack);
if (WARN_ON(pred)) {
err = -EINVAL;
filter->root = NULL;
goto fail;
}
err = check_pred_tree(filter, root);
if (err)
goto fail;
/* Optimize the tree */
err = fold_pred_tree(filter, root);
if (err)
goto fail;
/* We don't set root until we know it works */
barrier();
filter->root = root;
tracing/filter: Use a tree instead of stack for filter_match_preds() Currently the filter_match_preds() requires a stack to push and pop the preds to determine if the filter matches the record or not. This has two drawbacks: 1) It requires a stack to store state information. As this is done in fast paths we can't allocate the storage for this stack, and we can't use a global as it must be re-entrant. The stack is stored on the kernel stack and this greatly limits how many preds we may allow. 2) All conditions are calculated even when a short circuit exists. a || b will always calculate a and b even though a was determined to be true. Using a tree we can walk a constant structure that will save the state as we go. The algorithm is simply: pred = root; do { switch (move) { case MOVE_DOWN: if (OR or AND) { pred = left; continue; } if (pred == root) break; match = pred->fn(); pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; case MOVE_UP_FROM_LEFT: /* Only OR or AND can be a parent */ if (match && OR || !match && AND) { /* short circuit */ if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } pred = pred->right; move = MOVE_DOWN; continue; case MOVE_UP_FROM_RIGHT: if (pred == root) break; pred = pred->parent; move = left child ? MOVE_UP_FROM_LEFT : MOVE_UP_FROM_RIGHT; continue; } done = 1; } while (!done); This way there's no strict limit to how many preds we allow and it also will short circuit the logical operations when possible. Cc: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-01-28 03:54:33 +00:00
}
err = 0;
fail:
__free_pred_stack(&stack);
return err;
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
}
struct filter_list {
struct list_head list;
struct event_filter *filter;
};
static int replace_system_preds(struct event_subsystem *system,
struct filter_parse_state *ps,
char *filter_string)
{
struct ftrace_event_call *call;
struct filter_list *filter_item;
struct filter_list *tmp;
LIST_HEAD(filter_list);
bool fail = true;
int err;
list_for_each_entry(call, &ftrace_events, list) {
if (strcmp(call->class->system, system->name) != 0)
continue;
/*
* Try to see if the filter can be applied
* (filter arg is ignored on dry_run)
*/
err = replace_preds(call, NULL, ps, filter_string, true);
if (err)
goto fail;
}
list_for_each_entry(call, &ftrace_events, list) {
struct event_filter *filter;
if (strcmp(call->class->system, system->name) != 0)
continue;
filter_item = kzalloc(sizeof(*filter_item), GFP_KERNEL);
if (!filter_item)
goto fail_mem;
list_add_tail(&filter_item->list, &filter_list);
filter_item->filter = __alloc_filter();
if (!filter_item->filter)
goto fail_mem;
filter = filter_item->filter;
/* Can only fail on no memory */
err = replace_filter_string(filter, filter_string);
if (err)
goto fail_mem;
err = replace_preds(call, filter, ps, filter_string, false);
if (err) {
filter_disable(call);
parse_error(ps, FILT_ERR_BAD_SUBSYS_FILTER, 0);
append_filter_err(ps, filter);
} else
call->flags |= TRACE_EVENT_FL_FILTERED;
/*
* Regardless of if this returned an error, we still
* replace the filter for the call.
*/
filter = call->filter;
call->filter = filter_item->filter;
filter_item->filter = filter;
fail = false;
}
if (fail)
goto fail;
/*
* The calls can still be using the old filters.
* Do a synchronize_sched() to ensure all calls are
* done with them before we free them.
*/
synchronize_sched();
list_for_each_entry_safe(filter_item, tmp, &filter_list, list) {
__free_filter(filter_item->filter);
list_del(&filter_item->list);
kfree(filter_item);
}
return 0;
fail:
/* No call succeeded */
list_for_each_entry_safe(filter_item, tmp, &filter_list, list) {
list_del(&filter_item->list);
kfree(filter_item);
}
parse_error(ps, FILT_ERR_BAD_SUBSYS_FILTER, 0);
return -EINVAL;
fail_mem:
/* If any call succeeded, we still need to sync */
if (!fail)
synchronize_sched();
list_for_each_entry_safe(filter_item, tmp, &filter_list, list) {
__free_filter(filter_item->filter);
list_del(&filter_item->list);
kfree(filter_item);
}
return -ENOMEM;
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
int apply_event_filter(struct ftrace_event_call *call, char *filter_string)
{
struct filter_parse_state *ps;
struct event_filter *filter;
struct event_filter *tmp;
int err = 0;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
mutex_lock(&event_mutex);
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
if (!strcmp(strstrip(filter_string), "0")) {
filter_disable(call);
filter = call->filter;
if (!filter)
goto out_unlock;
call->filter = NULL;
/* Make sure the filter is not being used */
synchronize_sched();
__free_filter(filter);
goto out_unlock;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
}
err = -ENOMEM;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
ps = kzalloc(sizeof(*ps), GFP_KERNEL);
if (!ps)
goto out_unlock;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
filter = __alloc_filter();
if (!filter) {
kfree(ps);
goto out_unlock;
}
replace_filter_string(filter, filter_string);
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
parse_init(ps, filter_ops, filter_string);
err = filter_parse(ps);
if (err) {
append_filter_err(ps, filter);
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
goto out;
}
err = replace_preds(call, filter, ps, filter_string, false);
if (err) {
filter_disable(call);
append_filter_err(ps, filter);
} else
call->flags |= TRACE_EVENT_FL_FILTERED;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
out:
/*
* Always swap the call filter with the new filter
* even if there was an error. If there was an error
* in the filter, we disable the filter and show the error
* string
*/
tmp = call->filter;
call->filter = filter;
if (tmp) {
/* Make sure the call is done with the filter */
synchronize_sched();
__free_filter(tmp);
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
filter_opstack_clear(ps);
postfix_clear(ps);
kfree(ps);
out_unlock:
mutex_unlock(&event_mutex);
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
return err;
}
int apply_subsystem_event_filter(struct event_subsystem *system,
char *filter_string)
{
struct filter_parse_state *ps;
struct event_filter *filter;
int err = 0;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
mutex_lock(&event_mutex);
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
/* Make sure the system still has events */
if (!system->nr_events) {
err = -ENODEV;
goto out_unlock;
}
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
if (!strcmp(strstrip(filter_string), "0")) {
filter_free_subsystem_preds(system);
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
remove_filter_string(system->filter);
filter = system->filter;
system->filter = NULL;
/* Ensure all filters are no longer used */
synchronize_sched();
filter_free_subsystem_filters(system);
__free_filter(filter);
goto out_unlock;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
}
err = -ENOMEM;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
ps = kzalloc(sizeof(*ps), GFP_KERNEL);
if (!ps)
goto out_unlock;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
filter = __alloc_filter();
if (!filter)
goto out;
replace_filter_string(filter, filter_string);
/*
* No event actually uses the system filter
* we can free it without synchronize_sched().
*/
__free_filter(system->filter);
system->filter = filter;
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
parse_init(ps, filter_ops, filter_string);
err = filter_parse(ps);
if (err) {
append_filter_err(ps, system->filter);
goto out;
}
err = replace_system_preds(system, ps, filter_string);
if (err)
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
append_filter_err(ps, system->filter);
out:
filter_opstack_clear(ps);
postfix_clear(ps);
kfree(ps);
out_unlock:
mutex_unlock(&event_mutex);
tracing/filters: a better event parser Replace the current event parser hack with a better one. Filters are no longer specified predicate by predicate, but all at once and can use parens and any of the following operators: numeric fields: ==, !=, <, <=, >, >= string fields: ==, != predicates can be combined with the logical operators: &&, || examples: "common_preempt_count > 4" > filter "((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter If there was an error, the erroneous string along with an error message can be seen by looking at the filter e.g.: ((sig >= 10 && sig < 15) || dsig == 17) && comm != bash ^ parse_error: Field not found Currently the caret for an error always appears at the beginning of the filter; a real position should be used, but the error message should be useful even without it. To clear a filter, '0' can be written to the filter file. Filters can also be set or cleared for a complete subsystem by writing the same filter as would be written to an individual event to the filter file at the root of the subsytem. Note however, that if any event in the subsystem lacks a field specified in the filter being set, the set will fail and all filters in the subsytem are automatically cleared. This change from the previous version was made because using only the fields that happen to exist for a given event would most likely result in a meaningless filter. Because the logical operators are now implemented as predicates, the maximum number of predicates in a filter was increased from 8 to 16. [ Impact: add new, extended trace-filter implementation ] Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com> LKML-Reference: <1240905899.6416.121.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-04-28 08:04:59 +00:00
return err;
}
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
#ifdef CONFIG_PERF_EVENTS
void ftrace_profile_free_filter(struct perf_event *event)
{
struct event_filter *filter = event->filter;
event->filter = NULL;
__free_filter(filter);
}
int ftrace_profile_set_filter(struct perf_event *event, int event_id,
char *filter_str)
{
int err;
struct event_filter *filter;
struct filter_parse_state *ps;
struct ftrace_event_call *call;
mutex_lock(&event_mutex);
call = event->tp_event;
err = -EINVAL;
if (!call)
goto out_unlock;
err = -EEXIST;
if (event->filter)
goto out_unlock;
filter = __alloc_filter();
if (!filter) {
err = PTR_ERR(filter);
goto out_unlock;
}
err = -ENOMEM;
ps = kzalloc(sizeof(*ps), GFP_KERNEL);
if (!ps)
goto free_filter;
parse_init(ps, filter_ops, filter_str);
err = filter_parse(ps);
if (err)
goto free_ps;
err = replace_preds(call, filter, ps, filter_str, false);
if (!err)
event->filter = filter;
free_ps:
filter_opstack_clear(ps);
postfix_clear(ps);
kfree(ps);
free_filter:
if (err)
__free_filter(filter);
out_unlock:
mutex_unlock(&event_mutex);
return err;
}
#endif /* CONFIG_PERF_EVENTS */