Move uid filtering to BPF filters
Subsystems (2)
Ian Rogers
reviewerv4 cover Move uid filtering to BPF filters
Rather than scanning /proc and skipping PIDs based on their UIDs, use BPF filters for uid filtering. The /proc scanning in thread_map is racy as the PID may exit before the perf_event_open causing perf to abort. BPF UID filters are more robust as they avoid the race. The /proc scanning also misses processes starting after the perf command. Add a helper for commands that support UID filtering and wire up. Remove the non-BPF UID filtering support given it doesn't work.
v4: Add a warning message on top of Namhyung's BPF filter error message:
https://lore.kernel.org/lkml/[email protected]/
in the parse_uid_filter helper. In TUI the warning is shown then the BPF error shown, with stdio the warning appears below the BPF errors.
v3: Add lengthier commit messages as requested by Arnaldo. Rebase on
tmp.perf-tools-next.
v2: Add a perf record uid test (Namhyung) and force setting
system-wide for perf trace and perf record (Namhyung). Ensure the uid filter isn't set on tracepoint evsels.
v1: https://lore.kernel.org/lkml/[email protected]/
Ian Rogers (10):
perf parse-events filter: Use evsel__find_pmu
perf target: Separate parse_uid into its own function perf parse-events: Add parse_uid_filter helper
perf record: Switch user option to use BPF filter
perf tests record: Add basic uid filtering test
perf top: Switch user option to use BPF filter
perf trace: Switch user option to use BPF filter
perf bench evlist-open-close: Switch user option to use BPF filter perf target: Remove uid from target
perf thread_map: Remove uid options
tools/perf/bench/evlist-open-close.c | 36 ++++++++------ tools/perf/builtin-ftrace.c | 1 - tools/perf/builtin-kvm.c | 2 - tools/perf/builtin-record.c | 27 ++++++----- tools/perf/builtin-stat.c | 4 +- tools/perf/builtin-top.c | 22 +++++---- tools/perf/builtin-trace.c | 27 +++++++---- tools/perf/tests/backward-ring-buffer.c | 1 - tools/perf/tests/event-times.c | 8 ++- tools/perf/tests/keep-tracking.c | 2 +- tools/perf/tests/mmap-basic.c | 2 +- tools/perf/tests/openat-syscall-all-cpus.c | 2 +- tools/perf/tests/openat-syscall-tp-fields.c | 1 - tools/perf/tests/openat-syscall.c | 2 +- tools/perf/tests/perf-record.c | 1 - tools/perf/tests/perf-time-to-tsc.c | 2 +- tools/perf/tests/shell/record.sh | 26 ++++++++++ tools/perf/tests/switch-tracking.c | 2 +- tools/perf/tests/task-exit.c | 1 - tools/perf/tests/thread-map.c | 2 +- tools/perf/util/bpf-filter.c | 2 +- tools/perf/util/evlist.c | 3 +- tools/perf/util/parse-events.c | 47 +++++++++++++----- tools/perf/util/parse-events.h | 1 + tools/perf/util/python.c | 10 ++-- tools/perf/util/target.c | 54 +++------------------ tools/perf/util/target.h | 15 ++---- tools/perf/util/thread_map.c | 32 ++---------- tools/perf/util/thread_map.h | 6 +-- tools/perf/util/top.c | 4 +- tools/perf/util/top.h | 1 + 31 files changed, 164 insertions(+), 182 deletions(-)
Signature · 1 line
2.50.0.rc0.604.gd4ff7b7c86-goog
Ian Rogers
reviewerPATCH v4 1/10 perf parse-events filter: Use evsel__find_pmu
Rather than manually scanning PMUs, use evsel__find_pmu that can use the PMU set during event parsing.
--- tools/perf/util/parse-events.c | 14 ++++---------- 1 file changed, 4 insertions(+), 10 deletions(-)
tools/perf/util/parse-events.c +4 -10
@@ -2564,9 +2564,8 @@ foreach_evsel_in_last_glob(struct evlist *evlist,static int set_filter (struct evsel *evsel,const void *arg) {const char *str = arg; -bool found = false;int nr_addr_filters =0 ; -struct perf_pmu *pmu =NULL ; +struct perf_pmu *pmu;if (evsel ==NULL ) { fprintf(stderr,
@@ -2584,16 +2583,11 @@ static int set_filter(struct evsel *evsel, const void *arg)return 0 ; } -while ((pmu =perf_pmus__scan (pmu)) !=NULL ) -if (pmu->type == evsel->core.attr.type) { - found = true; -break ; - } - -if (found) + pmu =evsel__find_pmu (evsel); +if (pmu) {perf_pmu__scan_file (pmu,"nr_addr_filters" ,"%d" , &nr_addr_filters); - + }if (!nr_addr_filters)return perf_bpf_filter__parse (&evsel->bpf_filters, str);
Signature · 1 line
2.50.0.rc0.604.gd4ff7b7c86-goog
Ian Rogers
reviewerPATCH v4 2/10 perf target: Separate parse_uid into its own function
In reply to Ian RogersAllow parse_uid to be called without a struct target. Rather than have two errors, remove TARGET_ERRNO__USER_NOT_FOUND and use TARGET_ERRNO__INVALID_UID as the handling is identical.
--- tools/perf/util/target.c | 22 ++++++++++++---------- tools/perf/util/target.h | 3 ++- 2 files changed, 14 insertions(+), 11 deletions(-)
tools/perf/util/target.c +12 -10
@@ -94,15 +94,13 @@ enum target_errno target__validate(struct target *target)return ret; } -enum target_errno target__parse_uid (struct target *target) +uid_t parse_uid (const char *str) {struct passwd pwd, *result;char buf[1024 ]; -const char *str = target->uid_str; - target->uid =UINT_MAX ;if (str ==NULL ) -return TARGET_ERRNO__SUCCESS ; +return UINT_MAX ;/* Try user name first */ getpwnam_r (str, &pwd, buf,sizeof (buf), &result);
@@ -115,16 +113,22 @@ enum target_errno target__parse_uid(struct target *target)int uid =strtol (str, &endptr,10 );if (*endptr !='\0' ) -return TARGET_ERRNO__INVALID_UID ; +return UINT_MAX ;getpwuid_r (uid, &pwd, buf,sizeof (buf), &result);if (result ==NULL ) -return TARGET_ERRNO__USER_NOT_FOUND ; +return UINT_MAX ; } - target->uid = result->pw_uid; -return TARGET_ERRNO__SUCCESS ; +return result->pw_uid; +} + +enum target_errno target__parse_uid (struct target *target) +{ + target->uid =parse_uid (target->uid_str); + +return target->uid !=UINT_MAX ?TARGET_ERRNO__SUCCESS :TARGET_ERRNO__INVALID_UID ; }/*
@@ -142,7 +146,6 @@ static const char *target__error_str[] = {"BPF switch overriding UID" ,"BPF switch overriding THREAD" , "Invalid User: %s", - "Problems obtaining information for user %s", }; int target__strerror(struct target *target, int errnum,
@@ -171,7 +174,6 @@ int target__strerror(struct target *target, int errnum,break ;case TARGET_ERRNO__INVALID_UID : -case TARGET_ERRNO__USER_NOT_FOUND :snprintf (buf, buflen, msg, target->uid_str);break ;
tools/perf/util/target.h +2 -1
@@ -48,12 +48,13 @@ enum target_errno {/* for target__parse_uid() */ TARGET_ERRNO__INVALID_UID , -TARGET_ERRNO__USER_NOT_FOUND , __TARGET_ERRNO__END, };enum target_errno target__validate (struct target *target); + +uid_t parse_uid (const char *str);enum target_errno target__parse_uid (struct target *target);int target__strerror (struct target *target,int errnum,char *buf,size_t buflen);
Signature · 1 line
2.50.0.rc0.604.gd4ff7b7c86-goog
Ian Rogers
reviewerPATCH v4 3/10 perf parse-events: Add parse_uid_filter helper
In reply to Ian RogersAdd parse_uid_filter filter as a helper to parse_filter, that constructs a uid filter string. As uid filters don't work with tracepoint filters, add a is_possible_tp_filter function so the tracepoint filter isn't attempted for tracepoint evsels.
--- tools/perf/util/parse-events.c | 33 ++++++++++++++++++++++++++++++++- tools/perf/util/parse-events.h | 1 + 2 files changed, 33 insertions(+), 1 deletion(-)
tools/perf/util/parse-events.c +32 -1
@@ -25,6 +25,7 @@#include "pmu.h" #include "pmus.h" #include "asm/bug.h" +#include "ui/ui.h" #include "util/parse-branch-options.h" #include "util/evsel_config.h" #include "util/event.h"
@@ -2561,6 +2562,12 @@ foreach_evsel_in_last_glob(struct evlist *evlist,return 0 ; } +/* Will a tracepoint filter work for str or should a BPF filter be used? */ +static bool is_possible_tp_filter (const char *str) +{ +return strstr (str,"uid" ) ==NULL ; +} +static int set_filter (struct evsel *evsel,const void *arg) {const char *str = arg;
@@ -2573,7 +2580,7 @@ static int set_filter(struct evsel *evsel, const void *arg)return -1 ; } -if (evsel->core.attr.type ==PERF_TYPE_TRACEPOINT ) { +if (evsel->core.attr.type ==PERF_TYPE_TRACEPOINT &&is_possible_tp_filter (str)) {if (evsel__append_tp_filter (evsel, str) <0 ) {fprintf (stderr,"not enough memory to hold filter string\n" );
@@ -2609,6 +2616,30 @@ int parse_filter(const struct option *opt, const char *str, (const void *)str); } +int parse_uid_filter (struct evlist *evlist,uid_t uid) +{ +struct option opt = { + .value = &evlist, + }; +char buf[128 ]; +int ret; + +snprintf (buf,sizeof (buf),"uid == %d" , uid); + ret =parse_filter (&opt, buf,/*unset=*/ 0 ); +if (ret) { +if (use_browser >=1 ) { +/* +* Use ui__warning so a pop up appears above the +* underlying BPF error message. +*/ +ui__warning ("Failed to add UID filtering that uses BPF filtering.\n" ); + }else { +fprintf (stderr,"Failed to add UID filtering that uses BPF filtering.\n" ); + } + } +return ret; +} +static int add_exclude_perf_filter (struct evsel *evsel,const void *arg__maybe_unused ) {
tools/perf/util/parse-events.h +1 -0
@@ -45,6 +45,7 @@ static inline int parse_events(struct evlist *evlist, const char *str,int parse_event (struct evlist *evlist,const char *str);int parse_filter (const struct option *opt,const char *str,int unset); +int parse_uid_filter (struct evlist *evlist,uid_t uid);int exclude_perf (const struct option *opt,const char *arg,int unset);enum parse_events__term_val_type {
Signature · 1 line
2.50.0.rc0.604.gd4ff7b7c86-goog
Namhyung Kim
maintainerRe: [PATCH v4 03/10] perf parse-events: Add parse_uid_filter helper
Hi Ian,
Quoted 46 lines · Ian Rogers
Add
parse_uid_filterfilter as a helper toparse_filter, that constructs a uid filter string. As uid filters don't work with tracepoint filters, add ais_possible_tp_filterfunction so the tracepoint filter isn't attempted for tracepoint evsels.Signed-off-by Ian Rogers <[email protected]>--- tools/perf/util/parse-events.c | 33 ++++++++++++++++++++++++++++++++- tools/perf/util/parse-events.h | 1 + 2 files changed, 33 insertions(+), 1 deletion(-)tools/perf/util/parse-events.c +8
-1@@ -25,6 +25,7 @@#include "pmu.h" #include "pmus.h" #include "asm/bug.h" +#include "ui/ui.h" #include "util/parse-branch-options.h" #include "util/evsel_config.h" #include "util/event.h" @@ -2561,6 +2562,12 @@ foreach_evsel_in_last_glob(struct evlist *evlist,return 0 ; } +/* Will a tracepoint filter work for str or should a BPF filter be used? */ +static bool is_possible_tp_filter (const char *str) +{ +return strstr (str,"uid" ) ==NULL ; +} +static int set_filter (struct evsel *evsel,const void *arg) {const char *str = arg;@@ -2573,7 +2580,7 @@ static int set_filter(struct evsel *evsel, const void *arg)return -1 ; } -if (evsel->core.attr.type ==PERF_TYPE_TRACEPOINT ) { +if (evsel->core.attr.type ==PERF_TYPE_TRACEPOINT &&is_possible_tp_filter (str)) {if (evsel__append_tp_filter (evsel, str) <0 ) {fprintf (stderr,"not enough memory to hold filter string\n" );
tools/perf/util/parse-events.c int parse_filter(const struct option *opt, const char *str, · line 2619
Quoted 4 lines
tools/perf/util/parse-events.c +1
-0@@ -2609,6 +2616,30 @@ int parse_filter(const struct option *opt, const char *str, (const void *)str); } +int parse_uid_filter (struct evlist *evlist,uid_t uid)
It failed to build on alpine 3.18.
util/parse-events.h:48:45: error: unknown type name 'uid_t'
48 | int parse_uid_filter(struct evlist *evlist, uid_t uid);
| ^~~~~
I'll add this.
Thanks,
Namhyung
---8<---
tools/perf/util/parse-events.h +1 -0
@@ -11,6 +11,7 @@#include <linux/perf_event.h> #include <stdio.h> #include <string.h> +#include <sys/types.h> struct evsel ;struct evlist ;
Quoted 41 lines
tools/perf/util/parse-events.c +23
-0+{ +struct option opt = { + .value = &evlist, + }; +char buf[128 ]; +int ret; + +snprintf (buf,sizeof (buf),"uid == %d" , uid); + ret =parse_filter (&opt, buf,/*unset=*/ 0 ); +if (ret) { +if (use_browser >=1 ) { +/* +* Use ui__warning so a pop up appears above the +* underlying BPF error message. +*/ +ui__warning ("Failed to add UID filtering that uses BPF filtering.\n" ); + }else { +fprintf (stderr,"Failed to add UID filtering that uses BPF filtering.\n" ); + } + } +return ret; +} +static int add_exclude_perf_filter (struct evsel *evsel,const void *arg__maybe_unused ) {tools/perf/util/parse-events.h +1
-0@@ -45,6 +45,7 @@ static inline int parse_events(struct evlist *evlist, const char *str,int parse_event (struct evlist *evlist,const char *str);int parse_filter (const struct option *opt,const char *str,int unset); +int parse_uid_filter (struct evlist *evlist,uid_t uid);int exclude_perf (const struct option *opt,const char *arg,int unset);enum parse_events__term_val_type {Signature · 1 line
2.50.0.rc0.604.gd4ff7b7c86-goog
Ian Rogers
reviewerRe: [PATCH v4 03/10] perf parse-events: Add parse_uid_filter helper
Quoted 57 lines · Namhyung Kim
Hi Ian,
Quoted 46 lines · Ian Rogers
Add
parse_uid_filterfilter as a helper toparse_filter, that constructs a uid filter string. As uid filters don't work with tracepoint filters, add ais_possible_tp_filterfunction so the tracepoint filter isn't attempted for tracepoint evsels.Signed-off-by Ian Rogers <[email protected]>--- tools/perf/util/parse-events.c | 33 ++++++++++++++++++++++++++++++++- tools/perf/util/parse-events.h | 1 + 2 files changed, 33 insertions(+), 1 deletion(-)tools/perf/util/parse-events.c +8
-1@@ -25,6 +25,7 @@#include "pmu.h" #include "pmus.h" #include "asm/bug.h" +#include "ui/ui.h" #include "util/parse-branch-options.h" #include "util/evsel_config.h" #include "util/event.h" @@ -2561,6 +2562,12 @@ foreach_evsel_in_last_glob(struct evlist *evlist,return 0 ; } +/* Will a tracepoint filter work for str or should a BPF filter be used? */ +static bool is_possible_tp_filter (const char *str) +{ +return strstr (str,"uid" ) ==NULL ; +} +static int set_filter (struct evsel *evsel,const void *arg) {const char *str = arg;@@ -2573,7 +2580,7 @@ static int set_filter(struct evsel *evsel, const void *arg)return -1 ; } -if (evsel->core.attr.type ==PERF_TYPE_TRACEPOINT ) { +if (evsel->core.attr.type ==PERF_TYPE_TRACEPOINT &&is_possible_tp_filter (str)) {if (evsel__append_tp_filter (evsel, str) <0 ) {fprintf (stderr,"not enough memory to hold filter string\n" );tools/perf/util/parse-events.c int parse_filter(const struct option *opt, const char *str, · line 2619
Quoted 4 lines
tools/perf/util/parse-events.c +1
-0@@ -2609,6 +2616,30 @@ int parse_filter(const struct option *opt, const char *str, (const void *)str); } +int parse_uid_filter (struct evlist *evlist,uid_t uid)It failed to build on alpine 3.18.
Quoted 5 lines
util/parse-events.h:48:45: error: unknown type name 'uid_t'
48 | intparse_uid_filter(struct evlist *evlist, uid_t uid);
| ^~~~~I'll add this.
Thanks Namhyung! I see this in tmp.perf-tools-next so I'll assume there's no need for a v5.
Ian
Quoted 61 lines
Thanks,
Namhyung---8<---
tools/perf/util/parse-events.h +1
-0@@ -11,6 +11,7 @@#include <linux/perf_event.h> #include <stdio.h> #include <string.h> +#include <sys/types.h> struct evsel ;struct evlist ;Quoted 44 lines
tools/perf/util/parse-events.c +23
-0+{ +struct option opt = { + .value = &evlist, + }; +char buf[128 ]; +int ret; + +snprintf (buf,sizeof (buf),"uid == %d" , uid); + ret =parse_filter (&opt, buf,/*unset=*/ 0 ); +if (ret) { +if (use_browser >=1 ) { +/* +* Use ui__warning so a pop up appears above the +* underlying BPF error message. +*/ +ui__warning ("Failed to add UID filtering that uses BPF filtering.\n" ); + }else { +fprintf (stderr,"Failed to add UID filtering that uses BPF filtering.\n" ); + } + } +return ret; +} +static int add_exclude_perf_filter (struct evsel *evsel,const void *arg__maybe_unused ) {tools/perf/util/parse-events.h +1
-1@@ -45,6 +45,7 @@ static inline int parse_events(struct evlist *evlist, const char *str,int parse_event (struct evlist *evlist,const char *str);int parse_filter (const struct option *opt,const char *str,int unset); +int parse_uid_filter (struct evlist *evlist,uid_t uid);int exclude_perf (const struct option *opt,const char *arg,int unset);enum parse_events__term_val_type { --2.50.0.rc0.604.gd4ff7b7c86-goog
Namhyung Kim
maintainerRe: [PATCH v4 03/10] perf parse-events: Add parse_uid_filter helper
Quoted 58 lines · Ian Rogers
Quoted 57 lines · Namhyung Kim
Hi Ian,
Quoted 46 lines · Ian Rogers
Add
parse_uid_filterfilter as a helper toparse_filter, that constructs a uid filter string. As uid filters don't work with tracepoint filters, add ais_possible_tp_filterfunction so the tracepoint filter isn't attempted for tracepoint evsels.Signed-off-by Ian Rogers <[email protected]>--- tools/perf/util/parse-events.c | 33 ++++++++++++++++++++++++++++++++- tools/perf/util/parse-events.h | 1 + 2 files changed, 33 insertions(+), 1 deletion(-)tools/perf/util/parse-events.c +8
-1@@ -25,6 +25,7 @@#include "pmu.h" #include "pmus.h" #include "asm/bug.h" +#include "ui/ui.h" #include "util/parse-branch-options.h" #include "util/evsel_config.h" #include "util/event.h" @@ -2561,6 +2562,12 @@ foreach_evsel_in_last_glob(struct evlist *evlist,return 0 ; } +/* Will a tracepoint filter work for str or should a BPF filter be used? */ +static bool is_possible_tp_filter (const char *str) +{ +return strstr (str,"uid" ) ==NULL ; +} +static int set_filter (struct evsel *evsel,const void *arg) {const char *str = arg;@@ -2573,7 +2580,7 @@ static int set_filter(struct evsel *evsel, const void *arg)return -1 ; } -if (evsel->core.attr.type ==PERF_TYPE_TRACEPOINT ) { +if (evsel->core.attr.type ==PERF_TYPE_TRACEPOINT &&is_possible_tp_filter (str)) {if (evsel__append_tp_filter (evsel, str) <0 ) {fprintf (stderr,"not enough memory to hold filter string\n" );tools/perf/util/parse-events.c int parse_filter(const struct option *opt, const char *str, · line 2619
Quoted 4 lines
tools/perf/util/parse-events.c +1
-0@@ -2609,6 +2616,30 @@ int parse_filter(const struct option *opt, const char *str, (const void *)str); } +int parse_uid_filter (struct evlist *evlist,uid_t uid)It failed to build on alpine 3.18.
Quoted 8 lines
Quoted 5 lines
util/parse-events.h:48:45: error: unknown type name 'uid_t'
48 | intparse_uid_filter(struct evlist *evlist, uid_t uid);
| ^~~~~I'll add this.
Thanks Namhyung! I see this in tmp.perf-tools-next so I'll assume there's no need for a v5.
Right, I've updated the branch with the fix.
Thanks,
Namhyung
Ian Rogers
reviewerPATCH v4 4/10 perf record: Switch user option to use BPF filter
In reply to Ian RogersFinding user processes by scanning /proc is inherently racy and results in perf_event_open failures. Use a BPF filter to drop samples where the uid doesn't match. Ensure adding the BPF filter forces system-wide.
--- tools/perf/builtin-record.c | 27 ++++++++++++++++----------- 1 file changed, 16 insertions(+), 11 deletions(-)
tools/perf/builtin-record.c +16 -11
@@ -175,6 +175,7 @@ struct record {bool timestamp_boundary;bool off_cpu;const char *filter_action; +const char *uid_str;struct switch_output switch_output;unsigned long long samples;unsigned long output_max_size;/* = 0: unlimited */
@@ -3513,8 +3514,7 @@ static struct option __record_options[] = { "or ranges of time to enable events e.g. '-D 10-20,30-40'", record__parse_event_enable_time),OPT_BOOLEAN (0 ,"kcore" , &record.opts.kcore,"copy /proc/kcore" ), -OPT_STRING ('u' ,"uid" , &record.opts.target.uid_str,"user" , -"user to profile" ), +OPT_STRING ('u' ,"uid" , &record.uid_str,"user" ,"user to profile" ), OPT_CALLBACK_NOOPT('b', "branch-any", &record.opts.branch_stack, "branch any", "sample any taken branches",
@@ -4256,19 +4256,24 @@ int cmd_record(int argc, const char **argv)ui__warning ("%s\n" , errbuf); } - err =target__parse_uid (&rec->opts.target); -if (err) { -int saved_errno = errno; +if (rec->uid_str) { +uid_t uid =parse_uid (rec->uid_str); -target__strerror (&rec->opts.target, err, errbuf,BUFSIZ ); -ui__error ("%s" , errbuf); +if (uid ==UINT_MAX ) { +ui__error ("Invalid User: %s" , rec->uid_str); + err = -EINVAL ; + goto out; + } + err =parse_uid_filter (rec->evlist, uid); +if (err) + goto out; - err = -saved_errno; - goto out; +/* User ID filtering implies system wide. */ + rec->opts.target.system_wide = true; } -/* Enable ignoring missing threads when -u/-p option is defined. */ - rec->opts.ignore_missing_thread = rec->opts.target.uid !=UINT_MAX || rec->opts.target.pid; +/* Enable ignoring missing threads when -p option is defined. */ + rec->opts.ignore_missing_thread = rec->opts.target.pid;evlist__warn_user_requested_cpus (rec->evlist, rec->opts.target.cpu_list);
Signature · 1 line
2.50.0.rc0.604.gd4ff7b7c86-goog
Ian Rogers
reviewerPATCH v4 5/10 perf tests record: Add basic uid filtering test
In reply to Ian RogersBased on the system-wide test with changes around how failure is handled as BPF permissions are a bigger issue than perf event paranoia.
--- tools/perf/tests/shell/record.sh | 26 ++++++++++++++++++++++++++ 1 file changed, 26 insertions(+)
tools/perf/tests/shell/record.sh +26 -0
@@ -231,6 +231,31 @@ test_cgroup() { echo "Cgroup sampling test [Success]" } +test_uid() { + echo "Uid sampling test" + if ! perf record -aB --synth=no --uid "$(id -u)" -o "${perfdata}" ${testprog} \ + > "${script_output}" 2>&1 + then + if grep -q "libbpf.*EPERM" "${script_output}" + then + echo "Uid sampling [Skipped permissions]" + return + else + echo "Uid sampling [Failed to record]" + err=1 + # cat "${script_output}" + return + fi + fi + if ! perf report -i "${perfdata}" -q | grep -q "${testsym}" + then + echo "Uid sampling [Failed missing output]" + err=1 + return + fi + echo "Uid sampling test [Success]" +} + test_leader_sampling() { echo "Basic leader sampling test" if ! perf record -o "${perfdata}" -e "{cycles,cycles}:Su" -- \
@@ -345,6 +370,7 @@ test_system_wide test_workload test_branch_counter test_cgroup +test_uid test_leader_sampling test_topdown_leader_sampling test_precise_max
Signature · 1 line
2.50.0.rc0.604.gd4ff7b7c86-goog
Ian Rogers
reviewerPATCH v4 6/10 perf top: Switch user option to use BPF filter
In reply to Ian RogersFinding user processes by scanning /proc is inherently racy and results in perf_event_open failures. Use a BPF filter to drop samples where the uid doesn't match.
--- tools/perf/builtin-top.c | 22 ++++++++++++---------- tools/perf/util/top.c | 4 ++-- tools/perf/util/top.h | 1 + 3 files changed, 15 insertions(+), 12 deletions(-)
tools/perf/builtin-top.c +12 -10
@@ -643,7 +643,7 @@ static void *display_thread_tui(void *arg)*/ evlist__for_each_entry (top->evlist, pos) {struct hists *hists =evsel__hists (pos); - hists->uid_filter_str = top->record_opts.target.uid_str; + hists->uid_filter_str = top->uid_str; } ret = evlist__tui_browse_hists(top->evlist, help, &hbt, top->min_percent,
@@ -1571,7 +1571,7 @@ int cmd_top(int argc, const char **argv) "Add prefix to source file path names in programs (with --prefix-strip)"),OPT_STRING (0 ,"prefix-strip" , &annotate_opts.prefix_strip,"N" ,"Strip first N entries of source file path name in programs (with --prefix)" ), -OPT_STRING ('u' ,"uid" , &target->uid_str,"user" ,"user to profile" ), +OPT_STRING ('u' ,"uid" , &top.uid_str,"user" ,"user to profile" ),OPT_CALLBACK (0 ,"percent-limit" , &top,"percent" ,"Don't show entries under that percent" , parse_percent_limit), OPT_CALLBACK(0, "percentage", NULL, "relative|absolute",
@@ -1762,15 +1762,17 @@ int cmd_top(int argc, const char **argv)ui__warning ("%s\n" , errbuf); } - status =target__parse_uid (target); -if (status) { -int saved_errno = errno; - -target__strerror (target, status, errbuf,BUFSIZ ); -ui__error ("%s\n" , errbuf); +if (top.uid_str) { +uid_t uid =parse_uid (top.uid_str); - status = -saved_errno; - goto out_delete_evlist; +if (uid ==UINT_MAX ) { +ui__error ("Invalid User: %s" , top.uid_str); + status = -EINVAL ; + goto out_delete_evlist; + } + status =parse_uid_filter (top.evlist, uid); +if (status) + goto out_delete_evlist; }if (target__none (target))
tools/perf/util/top.c +2 -2
@@ -88,9 +88,9 @@ size_t perf_top__header_snprintf(struct perf_top *top, char *bf, size_t size)else if (target->tid) ret +=SNPRINTF (bf + ret, size - ret," (target_tid: %s" , target->tid); -else if (target->uid_str !=NULL ) +else if (top->uid_str !=NULL ) ret +=SNPRINTF (bf + ret, size - ret," (uid: %s" , - target->uid_str); + top->uid_str);else ret +=SNPRINTF (bf + ret, size - ret," (all" );
tools/perf/util/top.h +1 -0
@@ -48,6 +48,7 @@ struct perf_top {const char *sym_filter;float min_percent;unsigned int nr_threads_synthesize; +const char *uid_str;struct {struct ordered_events *in;
Signature · 1 line
2.50.0.rc0.604.gd4ff7b7c86-goog
Ian Rogers
reviewerPATCH v4 7/10 perf trace: Switch user option to use BPF filter
In reply to Ian RogersFinding user processes by scanning /proc is inherently racy and results in perf_event_open failures. Use a BPF filter to drop samples where the uid doesn't match. Ensure adding the BPF filter forces system-wide.
--- tools/perf/builtin-trace.c | 26 +++++++++++++++++--------- 1 file changed, 17 insertions(+), 9 deletions(-)
tools/perf/builtin-trace.c +17 -9
@@ -236,6 +236,7 @@ struct trace {struct ordered_events data;u64 last; } oe; +const char *uid_str; };static void trace__load_vmlinux_btf (struct trace *trace__maybe_unused )
@@ -4412,8 +4413,8 @@ static int trace__run(struct trace *trace, int argc, const char **argv)evlist__add (evlist, pgfault_min); } -/* Enable ignoring missing threads when -u/-p option is defined. */ - trace->opts.ignore_missing_thread = trace->opts.target.uid !=UINT_MAX || trace->opts.target.pid; +/* Enable ignoring missing threads when -p option is defined. */ + trace->opts.ignore_missing_thread = trace->opts.target.pid;if (trace->sched &&evlist__add_newtp (evlist,"sched" ,"sched_stat_runtime" , trace__sched_stat_runtime))
@@ -5445,8 +5446,7 @@ int cmd_trace(int argc, const char **argv) "child tasks do not inherit counters"),OPT_CALLBACK ('m' ,"mmap-pages" , &trace.opts.mmap_pages,"pages" ,"number of mmap data pages" , evlist__parse_mmap_pages), -OPT_STRING ('u' ,"uid" , &trace.opts.target.uid_str,"user" , -"user to profile" ), +OPT_STRING ('u' ,"uid" , &trace.uid_str,"user" ,"user to profile" ),OPT_CALLBACK (0 ,"duration" , &trace,"float" ,"show only events with duration > N.M ms" , trace__set_duration),
@@ -5804,11 +5804,19 @@ int cmd_trace(int argc, const char **argv) goto out_close; } - err =target__parse_uid (&trace.opts.target); -if (err) { -target__strerror (&trace.opts.target, err, bf,sizeof (bf)); -fprintf (trace.output,"%s" , bf); - goto out_close; +if (trace.uid_str) { +uid_t uid =parse_uid (trace.uid_str); + +if (uid ==UINT_MAX ) { +ui__error ("Invalid User: %s" , trace.uid_str); + err = -EINVAL ; + goto out_close; + } + err =parse_uid_filter (trace.evlist, uid); +if (err) + goto out_close; + + trace.opts.target.system_wide = true; }if (!argc &&target__none (&trace.opts.target))
Signature · 1 line
2.50.0.rc0.604.gd4ff7b7c86-goog
Ian Rogers
reviewerPATCH v4 8/10 perf bench evlist-open-close: Switch user option to use BPF filter
In reply to Ian RogersFinding user processes by scanning /proc is inherently racy and results in perf_event_open failures. Use a BPF filter to drop samples where the uid doesn't match.
--- tools/perf/bench/evlist-open-close.c | 36 ++++++++++++++++------------ 1 file changed, 21 insertions(+), 15 deletions(-)
tools/perf/bench/evlist-open-close.c +21 -15
@@ -57,7 +57,7 @@ static int evlist__count_evsel_fds(struct evlist *evlist)return cnt; } -static struct evlist *bench__create_evlist (char *evstr) +static struct evlist *bench__create_evlist (char *evstr,const char *uid_str) {struct parse_events_error err;struct evlist *evlist =evlist__new ();
@@ -78,6 +78,18 @@ static struct evlist *bench__create_evlist(char *evstr) goto out_delete_evlist; }parse_events_error__exit (&err); +if (uid_str) { +uid_t uid =parse_uid (uid_str); + +if (uid ==UINT_MAX ) { +pr_err ("Invalid User: %s" , uid_str); + ret = -EINVAL ; + goto out_delete_evlist; + } + ret =parse_uid_filter (evlist, uid); +if (ret) + goto out_delete_evlist; + } ret =evlist__create_maps (evlist, &opts.target);if (ret <0 ) {pr_err ("Not enough memory to create thread/cpu maps\n" );
@@ -117,10 +129,10 @@ static int bench__do_evlist_open_close(struct evlist *evlist)return 0 ; } -static int bench_evlist_open_close__run (char *evstr) +static int bench_evlist_open_close__run (char *evstr,const char *uid_str) {// used to print statistics only -struct evlist *evlist =bench__create_evlist (evstr); +struct evlist *evlist =bench__create_evlist (evstr, uid_str);double time_average, time_stddev;struct timeval start, end, diff;struct stats time_stats;
@@ -142,7 +154,7 @@ static int bench_evlist_open_close__run(char *evstr)for (i =0 ; i < iterations; i++) {pr_debug ("Started iteration %d\n" , i); - evlist =bench__create_evlist (evstr); + evlist =bench__create_evlist (evstr, uid_str);if (!evlist)return -ENOMEM ;
@@ -206,6 +218,7 @@ static char *bench__repeat_event_string(const char *evstr, int n)int bench_evlist_open_close (int argc,const char **argv) { +const char *uid_str =NULL ;const struct option options[] = {OPT_STRING ('e' ,"event" , &event_string,"event" ,"event selector. use 'perf list' to list available events" ),
@@ -221,7 +234,7 @@ int bench_evlist_open_close(int argc, const char **argv) "record events on existing process id"),OPT_STRING ('t' ,"tid" , &opts.target.tid,"tid" ,"record events on existing thread id" ), -OPT_STRING ('u' ,"uid" , &opts.target.uid_str,"user" ,"user to profile" ), +OPT_STRING ('u' ,"uid" , &uid_str,"user" ,"user to profile" ),OPT_BOOLEAN (0 ,"per-thread" , &opts.target.per_thread,"use per-thread mmaps" ), OPT_END() };
@@ -245,15 +258,8 @@ int bench_evlist_open_close(int argc, const char **argv) goto out; } - err =target__parse_uid (&opts.target); -if (err) { -target__strerror (&opts.target, err, errbuf,sizeof (errbuf)); -pr_err ("%s" , errbuf); - goto out; - } - -/* Enable ignoring missing threads when -u/-p option is defined. */ - opts.ignore_missing_thread = opts.target.uid !=UINT_MAX || opts.target.pid; +/* Enable ignoring missing threads when -p option is defined. */ + opts.ignore_missing_thread = opts.target.pid; evstr =bench__repeat_event_string (event_string, nr_events);if (!evstr) {
@@ -261,7 +267,7 @@ int bench_evlist_open_close(int argc, const char **argv) goto out; } - err =bench_evlist_open_close__run (evstr); + err =bench_evlist_open_close__run (evstr, uid_str);free (evstr); out:
Signature · 1 line
2.50.0.rc0.604.gd4ff7b7c86-goog
Ian Rogers
reviewerPATCH v4 9/10 perf target: Remove uid from target
In reply to Ian RogersGathering threads with a uid by scanning /proc is inherently racy leading to perf_event_open failures that quit perf. All users of the functionality now use BPF filters, so remove uid and uid_str from target.
--- tools/perf/builtin-ftrace.c | 1 - tools/perf/builtin-kvm.c | 2 - tools/perf/builtin-stat.c | 4 +- tools/perf/builtin-trace.c | 1 - tools/perf/tests/backward-ring-buffer.c | 1 - tools/perf/tests/event-times.c | 4 +- tools/perf/tests/openat-syscall-tp-fields.c | 1 - tools/perf/tests/perf-record.c | 1 - tools/perf/tests/task-exit.c | 1 - tools/perf/util/bpf-filter.c | 2 +- tools/perf/util/evlist.c | 3 +- tools/perf/util/target.c | 46 +-------------------- tools/perf/util/target.h | 12 +----- 13 files changed, 6 insertions(+), 73 deletions(-)
tools/perf/builtin-ftrace.c +0 -1
@@ -1663,7 +1663,6 @@ int cmd_ftrace(int argc, const char **argv)int (*cmd_func)(struct perf_ftrace *) =NULL ;struct perf_ftrace ftrace = { .tracer =DEFAULT_TRACER , - .target = { .uid =UINT_MAX , }, };const struct option common_options[] = { OPT_STRING('p', "pid", &ftrace.target.pid, "pid",
tools/perf/builtin-kvm.c +0 -2
@@ -1871,8 +1871,6 @@ static int kvm_events_live(struct perf_kvm_stat *kvm, kvm->opts.user_interval =1 ; kvm->opts.mmap_pages =512 ; kvm->opts.target.uses_mmap = false; - kvm->opts.target.uid_str =NULL ; - kvm->opts.target.uid =UINT_MAX ;symbol__init (NULL );disable_buildid_cache ();
tools/perf/builtin-stat.c +1 -3
@@ -108,9 +108,7 @@ static struct parse_events_option_args parse_events_option_args = {static bool all_counters_use_bpf = true; -static struct target target = { - .uid =UINT_MAX , -}; +static struct target target;static volatile sig_atomic_t child_pid =-1 ;static int detailed_run =0 ;
tools/perf/builtin-trace.c +0 -1
@@ -5399,7 +5399,6 @@ int cmd_trace(int argc, const char **argv)struct trace trace = { .opts = { .target = { - .uid =UINT_MAX , .uses_mmap = true, }, .user_freq =UINT_MAX ,
tools/perf/tests/backward-ring-buffer.c +0 -1
@@ -91,7 +91,6 @@ static int test__backward_ring_buffer(struct test_suite *test __maybe_unused, instruct parse_events_error parse_error;struct record_opts opts = { .target = { - .uid =UINT_MAX , .uses_mmap = true, }, .freq =0 ,
tools/perf/tests/event-times.c +1 -3
@@ -17,9 +17,7 @@static int attach__enable_on_exec (struct evlist *evlist) {struct evsel *evsel =evlist__last (evlist); -struct target target = { - .uid =UINT_MAX , - }; +struct target target = {};const char *argv[] = {"true" ,NULL , };char sbuf[STRERR_BUFSIZE ];int err;
tools/perf/tests/openat-syscall-tp-fields.c +0 -1
@@ -28,7 +28,6 @@ static int test__syscall_openat_tp_fields(struct test_suite *test __maybe_unused {struct record_opts opts = { .target = { - .uid =UINT_MAX , .uses_mmap = true, }, .no_buffering = true,
tools/perf/tests/perf-record.c +0 -1
@@ -45,7 +45,6 @@ static int test__PERF_RECORD(struct test_suite *test __maybe_unused, int subtest {struct record_opts opts = { .target = { - .uid =UINT_MAX , .uses_mmap = true, }, .no_buffering = true,
tools/perf/tests/task-exit.c +0 -1
@@ -46,7 +46,6 @@ static int test__task_exit(struct test_suite *test __maybe_unused, int subtest _struct evsel *evsel;struct evlist *evlist;struct target target = { - .uid =UINT_MAX , .uses_mmap = true, };const char *argv[] = {"true" ,NULL };
tools/perf/util/bpf-filter.c +1 -1
@@ -450,7 +450,7 @@ int perf_bpf_filter__prepare(struct evsel *evsel, struct target *target)struct bpf_program *prog;struct bpf_link *link;struct perf_bpf_filter_entry *entry; -bool needs_idx_hash = !target__has_cpu (target) && !target->uid_str; +bool needs_idx_hash = !target__has_cpu (target); entry =calloc (MAX_FILTERS ,sizeof (*entry));if (entry ==NULL )
tools/perf/util/evlist.c +1 -2
@@ -1006,8 +1006,7 @@ int evlist__create_maps(struct evlist *evlist, struct target *target)* per-thread data. thread_map__new_str will call * thread_map__new_all_cpus to enumerate all threads. */ - threads =thread_map__new_str (target->pid, target->tid, target->uid, - all_threads); + threads =thread_map__new_str (target->pid, target->tid,UINT_MAX , all_threads);if (!threads)return -1 ;
tools/perf/util/target.c +1 -45
88 lines not on this page · raw message
tools/perf/util/target.h +1 -11
@@ -9,9 +9,7 @@ struct target {const char *pid;const char *tid;const char *cpu_list; -const char *uid_str;const char *bpf_str; -uid_t uid;bool system_wide;bool uses_mmap;bool default_per_cpu;
@@ -36,32 +34,24 @@ enum target_errno {/* for target__validate() */ TARGET_ERRNO__PID_OVERRIDE_CPU = __TARGET_ERRNO__START, -TARGET_ERRNO__PID_OVERRIDE_UID , -TARGET_ERRNO__UID_OVERRIDE_CPU ,TARGET_ERRNO__PID_OVERRIDE_SYSTEM , -TARGET_ERRNO__UID_OVERRIDE_SYSTEM ,TARGET_ERRNO__SYSTEM_OVERRIDE_THREAD ,TARGET_ERRNO__BPF_OVERRIDE_CPU ,TARGET_ERRNO__BPF_OVERRIDE_PID , -TARGET_ERRNO__BPF_OVERRIDE_UID ,TARGET_ERRNO__BPF_OVERRIDE_THREAD , -/* for target__parse_uid() */ -TARGET_ERRNO__INVALID_UID , - __TARGET_ERRNO__END, };enum target_errno target__validate (struct target *target);uid_t parse_uid (const char *str); -enum target_errno target__parse_uid (struct target *target);int target__strerror (struct target *target,int errnum,char *buf,size_t buflen);static inline bool target__has_task (struct target *target) { -return target->tid || target->pid || target->uid_str; +return target->tid || target->pid; }static inline bool target__has_cpu (struct target *target)
Signature · 1 line
2.50.0.rc0.604.gd4ff7b7c86-goog
Ian Rogers
reviewerPATCH v4 10/10 perf thread_map: Remove uid options
In reply to Ian RogersNow the target doesn't have a uid, it is handled through BPF filters, remove the uid options to thread_map creation. Tidy up the functions used in tests to avoid passing unused arguments.
--- tools/perf/tests/event-times.c | 4 +-- tools/perf/tests/keep-tracking.c | 2 +- tools/perf/tests/mmap-basic.c | 2 +- tools/perf/tests/openat-syscall-all-cpus.c | 2 +- tools/perf/tests/openat-syscall.c | 2 +- tools/perf/tests/perf-time-to-tsc.c | 2 +- tools/perf/tests/switch-tracking.c | 2 +- tools/perf/tests/thread-map.c | 2 +- tools/perf/util/evlist.c | 2 +- tools/perf/util/python.c | 10 +++---- tools/perf/util/thread_map.c | 32 ++-------------------- tools/perf/util/thread_map.h | 6 ++-- 12 files changed, 20 insertions(+), 48 deletions(-)
tools/perf/tests/event-times.c +2 -2
@@ -62,7 +62,7 @@ static int attach__current_disabled(struct evlist *evlist)pr_debug ("attaching to current thread as disabled\n" ); - threads =thread_map__new (-1 ,getpid (),UINT_MAX ); + threads =thread_map__new_by_tid (getpid ());if (threads ==NULL ) {pr_debug ("thread_map__new\n" );return -1 ;
@@ -88,7 +88,7 @@ static int attach__current_enabled(struct evlist *evlist)pr_debug ("attaching to current thread as enabled\n" ); - threads =thread_map__new (-1 ,getpid (),UINT_MAX ); + threads =thread_map__new_by_tid (getpid ());if (threads ==NULL ) {pr_debug ("failed to call thread_map__new\n" );return -1 ;
tools/perf/tests/keep-tracking.c +1 -1
@@ -78,7 +78,7 @@ static int test__keep_tracking(struct test_suite *test __maybe_unused, int subteint found, err =-1 ;const char *comm; - threads =thread_map__new (-1 ,getpid (),UINT_MAX ); + threads =thread_map__new_by_tid (getpid ());CHECK_NOT_NULL__ (threads); cpus =perf_cpu_map__new_online_cpus ();
tools/perf/tests/mmap-basic.c +1 -1
@@ -46,7 +46,7 @@ static int test__basic_mmap(struct test_suite *test __maybe_unused, int subtestchar sbuf[STRERR_BUFSIZE ];struct mmap *md; - threads =thread_map__new (-1 ,getpid (),UINT_MAX ); + threads =thread_map__new_by_tid (getpid ());if (threads ==NULL ) {pr_debug ("thread_map__new\n" );return -1 ;
tools/perf/tests/openat-syscall-all-cpus.c +1 -1
@@ -28,7 +28,7 @@ static int test__openat_syscall_event_on_all_cpus(struct test_suite *test __maybstruct evsel *evsel;unsigned int nr_openat_calls =111 , i;cpu_set_t cpu_set; -struct perf_thread_map *threads =thread_map__new (-1 ,getpid (),UINT_MAX ); +struct perf_thread_map *threads =thread_map__new_by_tid (getpid ());char sbuf[STRERR_BUFSIZE ];char errbuf[BUFSIZ ];
tools/perf/tests/openat-syscall.c +1 -1
@@ -20,7 +20,7 @@ static int test__openat_syscall_event(struct test_suite *test __maybe_unused,int err =TEST_FAIL , fd;struct evsel *evsel;unsigned int nr_openat_calls =111 , i; -struct perf_thread_map *threads =thread_map__new (-1 ,getpid (),UINT_MAX ); +struct perf_thread_map *threads =thread_map__new_by_tid (getpid ());char sbuf[STRERR_BUFSIZE ];char errbuf[BUFSIZ ];
tools/perf/tests/perf-time-to-tsc.c +1 -1
@@ -90,7 +90,7 @@ static int test__perf_time_to_tsc(struct test_suite *test __maybe_unused, int sustruct mmap *md; - threads =thread_map__new (-1 ,getpid (),UINT_MAX ); + threads =thread_map__new_by_tid (getpid ());CHECK_NOT_NULL__ (threads); cpus =perf_cpu_map__new_online_cpus ();
tools/perf/tests/switch-tracking.c +1 -1
@@ -351,7 +351,7 @@ static int test__switch_tracking(struct test_suite *test __maybe_unused, int subconst char *comm;int err =-1 ; - threads =thread_map__new (-1 ,getpid (),UINT_MAX ); + threads =thread_map__new_by_tid (getpid ());if (!threads) {pr_debug ("thread_map__new failed!\n" ); goto out_err;
tools/perf/tests/thread-map.c +1 -1
@@ -115,7 +115,7 @@ static int test__thread_map_remove(struct test_suite *test __maybe_unused, int sTEST_ASSERT_VAL ("failed to allocate map string" ,asprintf (&str,"%d,%d" ,getpid (),getppid ()) >=0 ); - threads =thread_map__new_str (str,NULL ,0 , false); + threads =thread_map__new_str (str,/*tid=*/ NULL ,/*all_threads=*/ false);free (str); TEST_ASSERT_VAL("failed to allocate thread_map",
tools/perf/util/evlist.c +1 -1
@@ -1006,7 +1006,7 @@ int evlist__create_maps(struct evlist *evlist, struct target *target)* per-thread data. thread_map__new_str will call * thread_map__new_all_cpus to enumerate all threads. */ - threads =thread_map__new_str (target->pid, target->tid,UINT_MAX , all_threads); + threads =thread_map__new_str (target->pid, target->tid, all_threads);if (!threads)return -1 ;
tools/perf/util/python.c +5 -5
@@ -566,14 +566,14 @@ struct pyrf_thread_map {static int pyrf_thread_map__init (struct pyrf_thread_map *pthreads,PyObject *args,PyObject *kwargs) { -static char *kwlist[] = {"pid" ,"tid" ,"uid" ,NULL }; -int pid =-1 , tid =-1 , uid =UINT_MAX ; +static char *kwlist[] = {"pid" ,"tid" ,NULL }; +int pid =-1 , tid =-1 ; -if (!PyArg_ParseTupleAndKeywords (args, kwargs,"|iii" , - kwlist, &pid, &tid, &uid)) +if (!PyArg_ParseTupleAndKeywords (args, kwargs,"|ii" , + kwlist, &pid, &tid))return -1 ; - pthreads->threads =thread_map__new (pid, tid, uid); + pthreads->threads =thread_map__new (pid, tid);if (pthreads->threads ==NULL )return -1 ;return 0 ;
tools/perf/util/thread_map.c +3 -29
@@ -72,7 +72,7 @@ struct perf_thread_map *thread_map__new_by_tid(pid_t tid)return threads; } -static struct perf_thread_map *__thread_map__new_all_cpus (uid_t uid) +static struct perf_thread_map *thread_map__new_all_cpus (void ) {DIR *proc;int max_threads =32 , items, i;
@@ -98,15 +98,6 @@ static struct perf_thread_map *__thread_map__new_all_cpus(uid_t uid)if (*end)/* only interested in proper numerical dirents */ continue ; -snprintf (path,sizeof (path),"/proc/%s" , dirent->d_name); - -if (uid !=UINT_MAX ) { -struct stat st; - -if (stat (path, &st) !=0 || st.st_uid != uid) -continue ; - } -snprintf (path,sizeof (path),"/proc/%d/task" , pid); items =scandir (path, &namelist, filter,NULL );if (items <=0 ) {
@@ -157,24 +148,11 @@ static struct perf_thread_map *__thread_map__new_all_cpus(uid_t uid) goto out_closedir; } -struct perf_thread_map *thread_map__new_all_cpus (void ) -{ -return __thread_map__new_all_cpus (UINT_MAX ); -} - -struct perf_thread_map *thread_map__new_by_uid (uid_t uid) -{ -return __thread_map__new_all_cpus (uid); -} - -struct perf_thread_map *thread_map__new (pid_t pid,pid_t tid,uid_t uid) +struct perf_thread_map *thread_map__new (pid_t pid,pid_t tid) {if (pid !=-1 )return thread_map__new_by_pid (pid); -if (tid ==-1 && uid !=UINT_MAX ) -return thread_map__new_by_uid (uid); -return thread_map__new_by_tid (tid); }
@@ -289,15 +267,11 @@ struct perf_thread_map *thread_map__new_by_tid_str(const char *tid_str) goto out; } -struct perf_thread_map *thread_map__new_str (const char *pid,const char *tid, -uid_t uid,bool all_threads) +struct perf_thread_map *thread_map__new_str (const char *pid,const char *tid,bool all_threads) {if (pid)return thread_map__new_by_pid_str (pid); -if (!tid && uid !=UINT_MAX ) -return thread_map__new_by_uid (uid); -if (all_threads)return thread_map__new_all_cpus ();
tools/perf/util/thread_map.h +2 -4
@@ -11,13 +11,11 @@ struct perf_record_thread_map;struct perf_thread_map *thread_map__new_dummy (void );struct perf_thread_map *thread_map__new_by_pid (pid_t pid);struct perf_thread_map *thread_map__new_by_tid (pid_t tid); -struct perf_thread_map *thread_map__new_by_uid (uid_t uid); -struct perf_thread_map *thread_map__new_all_cpus (void ); -struct perf_thread_map *thread_map__new (pid_t pid,pid_t tid,uid_t uid); +struct perf_thread_map *thread_map__new (pid_t pid,pid_t tid);struct perf_thread_map *thread_map__new_event (struct perf_record_thread_map *event);struct perf_thread_map *thread_map__new_str (const char *pid, -const char *tid,uid_t uid,bool all_threads); +const char *tid,bool all_threads);struct perf_thread_map *thread_map__new_by_tid_str (const char *tid_str);
Signature · 1 line
2.50.0.rc0.604.gd4ff7b7c86-goog
Namhyung Kim
maintainerHi Ian,
Quoted 23 lines · Ian Rogers
Rather than scanning /proc and skipping PIDs based on their UIDs, use BPF filters for uid filtering. The /proc scanning in
thread_mapis racy as the PID may exit before theperf_event_opencausing perf to abort. BPF UID filters are more robust as they avoid the race. The /proc scanning also misses processes starting after the perf command. Add a helper for commands that support UID filtering and wire up. Remove the non-BPF UID filtering support given it doesn't work.v4: Add a warning message on top of Namhyung's BPF filter error message:
https://lore.kernel.org/lkml/[email protected]/
in theparse_uid_filterhelper. In TUI the warning is shown then the BPF error shown, with stdio the warning appears below the BPF errors.v3: Add lengthier commit messages as requested by Arnaldo. Rebase on
tmp.perf-tools-next.v2: Add a perf record uid test (Namhyung) and force setting
system-wide for perf trace and perf record (Namhyung). Ensure the uid filter isn't set on tracepoint evsels.
Quoted 11 lines
Ian Rogers (10):
perf parse-events filter: Useevsel__find_pmu
perf target: Separateparse_uidinto its own function perf parse-events: Addparse_uid_filterhelper
perf record: Switch user option to use BPF filter
perf tests record: Add basic uid filtering test
perf top: Switch user option to use BPF filter
perf trace: Switch user option to use BPF filter
perf bench evlist-open-close: Switch user option to use BPF filter perf target: Remove uid from target
perfthread_map: Remove uid options
I've noticed two things.
* it takes a quite long time to load the BPF filter. Not sure what's the issue, but maybe annoying for users.
* normal users cannot use BPF filter even if the BPF is loaded and pinned already. It works fine for perf record.
I don't think the issues are from this change though.
Thanks,
Namhyung
Quoted 37 lines
tools/perf/bench/evlist-open-close.c | 36 ++++++++------ tools/perf/builtin-ftrace.c | 1 - tools/perf/builtin-kvm.c | 2 - tools/perf/builtin-record.c | 27 ++++++----- tools/perf/builtin-stat.c | 4 +- tools/perf/builtin-top.c | 22 +++++---- tools/perf/builtin-trace.c | 27 +++++++---- tools/perf/tests/backward-ring-buffer.c | 1 - tools/perf/tests/event-times.c | 8 ++- tools/perf/tests/keep-tracking.c | 2 +- tools/perf/tests/mmap-basic.c | 2 +- tools/perf/tests/openat-syscall-all-cpus.c | 2 +- tools/perf/tests/openat-syscall-tp-fields.c | 1 - tools/perf/tests/openat-syscall.c | 2 +- tools/perf/tests/perf-record.c | 1 - tools/perf/tests/perf-time-to-tsc.c | 2 +- tools/perf/tests/shell/record.sh | 26 ++++++++++ tools/perf/tests/switch-tracking.c | 2 +- tools/perf/tests/task-exit.c | 1 - tools/perf/tests/thread-map.c | 2 +- tools/perf/util/bpf-filter.c | 2 +- tools/perf/util/evlist.c | 3 +- tools/perf/util/parse-events.c | 47 +++++++++++++----- tools/perf/util/parse-events.h | 1 + tools/perf/util/python.c | 10 ++-- tools/perf/util/target.c | 54 +++------------------ tools/perf/util/target.h | 15 ++---- tools/perf/util/thread_map.c | 32 ++---------- tools/perf/util/thread_map.h | 6 +-- tools/perf/util/top.c | 4 +- tools/perf/util/top.h | 1 + 31 files changed, 164 insertions(+), 182 deletions(-)Signature · 1 line
2.50.0.rc0.604.gd4ff7b7c86-goog
Quoted 9 lines · Ian Rogers
Rather than scanning /proc and skipping PIDs based on their UIDs, use BPF filters for uid filtering. The /proc scanning in
thread_mapis racy as the PID may exit before theperf_event_opencausing perf to abort. BPF UID filters are more robust as they avoid the race. The /proc scanning also misses processes starting after the perf command. Add a helper for commands that support UID filtering and wire up. Remove the non-BPF UID filtering support given it doesn't work.[...]
Applied to perf-tools-next, thanks!
Best regards,
Namhyung