From: Arnaldo Carvalho de Melo <acme@kernel.org>
To: Ingo Molnar <mingo@kernel.org>
Cc: linux-kernel@vger.kernel.org, Waiman Long <Waiman.Long@hp.com>,
Adrian Hunter <adrian.hunter@intel.com>,
Don Zickus <dzickus@redhat.com>,
Douglas Hatch <doug.hatch@hp.com>, Ingo Molnar <mingo@redhat.com>,
Jiri Olsa <jolsa@kernel.org>, Namhyung Kim <namhyung@kernel.org>,
Paul Mackerras <paulus@samba.org>,
Peter Zijlstra <peterz@infradead.org>,
Scott J Norton <scott.norton@hp.com>,
Arnaldo Carvalho de Melo <acme@redhat.com>
Subject: [PATCH 6/8] perf symbols: Improve DSO long names lookup speed with rbtree
Date: Wed, 1 Oct 2014 16:50:41 -0300 [thread overview]
Message-ID: <1412193043-15493-7-git-send-email-acme@kernel.org> (raw)
In-Reply-To: <1412193043-15493-1-git-send-email-acme@kernel.org>
From: Waiman Long <Waiman.Long@hp.com>
With workload that spawns and destroys many threads and processes, it
was found that perf-mem could took a long time to post-process the perf
data after the target workload had completed its operation.
The performance bottleneck was found to be the lookup and insertion of
the new DSO structures (thousands of them in this case).
In a dual-socket Ivy-Bridge E7-4890 v2 machine (30-core, 60-thread), the
perf profile below shows what perf was doing after the profiled AIM7
shared workload completed:
- 83.94% perf libc-2.11.3.so [.] __strcmp_sse42
- __strcmp_sse42
- 99.82% map__new
machine__process_mmap_event
perf_session_deliver_event
perf_session__process_event
__perf_session__process_events
cmd_record
cmd_mem
run_builtin
main
__libc_start_main
- 13.17% perf perf [.] __dsos__findnew
__dsos__findnew
map__new
machine__process_mmap_event
perf_session_deliver_event
perf_session__process_event
__perf_session__process_events
cmd_record
cmd_mem
run_builtin
main
__libc_start_main
So about 97% of CPU times were spent in the map__new() function trying
to insert new DSO entry into the DSO linked list. The whole
post-processing step took about 9 minutes.
The DSO structures are currently searched linearly. So the total
processing time will be proportional to n^2.
To overcome this performance problem, the DSO code is modified to also
put the DSO structures in a RB tree sorted by its long name in
additional to being in a simple linked list. With this change, the
processing time will become proportional to n*log(n) which will be much
quicker for large n. However, the short name will still be searched
using the old linear searching method. With that patch in place, the
same perf-mem post-processing step took less than 30 seconds to
complete.
Signed-off-by: Waiman Long <Waiman.Long@hp.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Don Zickus <dzickus@redhat.com>
Cc: Douglas Hatch <doug.hatch@hp.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Scott J Norton <scott.norton@hp.com>
Link: http://lkml.kernel.org/r/1412098575-27863-3-git-send-email-Waiman.Long@hp.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
tools/perf/util/dso.c | 70 ++++++++++++++++++++++++++++++++++++++++++++---
tools/perf/util/dso.h | 5 +++-
tools/perf/util/machine.c | 1 +
3 files changed, 71 insertions(+), 5 deletions(-)
diff --git a/tools/perf/util/dso.c b/tools/perf/util/dso.c
index 901a58fa3f22..0247acfdfaca 100644
--- a/tools/perf/util/dso.c
+++ b/tools/perf/util/dso.c
@@ -653,6 +653,65 @@ struct dso *dso__kernel_findnew(struct machine *machine, const char *name,
return dso;
}
+/*
+ * Find a matching entry and/or link current entry to RB tree.
+ * Either one of the dso or name parameter must be non-NULL or the
+ * function will not work.
+ */
+static struct dso *dso__findlink_by_longname(struct rb_root *root,
+ struct dso *dso, const char *name)
+{
+ struct rb_node **p = &root->rb_node;
+ struct rb_node *parent = NULL;
+
+ if (!name)
+ name = dso->long_name;
+ /*
+ * Find node with the matching name
+ */
+ while (*p) {
+ struct dso *this = rb_entry(*p, struct dso, rb_node);
+ int rc = strcmp(name, this->long_name);
+
+ parent = *p;
+ if (rc == 0) {
+ /*
+ * In case the new DSO is a duplicate of an existing
+ * one, print an one-time warning & put the new entry
+ * at the end of the list of duplicates.
+ */
+ if (!dso || (dso == this))
+ return this; /* Find matching dso */
+ /*
+ * The core kernel DSOs may have duplicated long name.
+ * In this case, the short name should be different.
+ * Comparing the short names to differentiate the DSOs.
+ */
+ rc = strcmp(dso->short_name, this->short_name);
+ if (rc == 0) {
+ pr_err("Duplicated dso name: %s\n", name);
+ return NULL;
+ }
+ }
+ if (rc < 0)
+ p = &parent->rb_left;
+ else
+ p = &parent->rb_right;
+ }
+ if (dso) {
+ /* Add new node and rebalance tree */
+ rb_link_node(&dso->rb_node, parent, p);
+ rb_insert_color(&dso->rb_node, root);
+ }
+ return NULL;
+}
+
+static inline struct dso *
+dso__find_by_longname(const struct rb_root *root, const char *name)
+{
+ return dso__findlink_by_longname((struct rb_root *)root, NULL, name);
+}
+
void dso__set_long_name(struct dso *dso, const char *name, bool name_allocated)
{
if (name == NULL)
@@ -755,6 +814,7 @@ struct dso *dso__new(const char *name)
dso->a2l_fails = 1;
dso->kernel = DSO_TYPE_USER;
dso->needs_swap = DSO_SWAP__UNSET;
+ RB_CLEAR_NODE(&dso->rb_node);
INIT_LIST_HEAD(&dso->node);
INIT_LIST_HEAD(&dso->data.open_entry);
}
@@ -765,6 +825,10 @@ struct dso *dso__new(const char *name)
void dso__delete(struct dso *dso)
{
int i;
+
+ if (!RB_EMPTY_NODE(&dso->rb_node))
+ pr_err("DSO %s is still in rbtree when being deleted!\n",
+ dso->long_name);
for (i = 0; i < MAP__NR_TYPES; ++i)
symbols__delete(&dso->symbols[i]);
@@ -854,6 +918,7 @@ bool __dsos__read_build_ids(struct list_head *head, bool with_hits)
void dsos__add(struct dsos *dsos, struct dso *dso)
{
list_add_tail(&dso->node, &dsos->head);
+ dso__findlink_by_longname(&dsos->root, dso, NULL);
}
struct dso *dsos__find(const struct dsos *dsos, const char *name,
@@ -867,10 +932,7 @@ struct dso *dsos__find(const struct dsos *dsos, const char *name,
return pos;
return NULL;
}
- list_for_each_entry(pos, &dsos->head, node)
- if (strcmp(pos->long_name, name) == 0)
- return pos;
- return NULL;
+ return dso__find_by_longname(&dsos->root, name);
}
struct dso *__dsos__findnew(struct dsos *dsos, const char *name)
diff --git a/tools/perf/util/dso.h b/tools/perf/util/dso.h
index b63dc98ad71d..acb651acc7fd 100644
--- a/tools/perf/util/dso.h
+++ b/tools/perf/util/dso.h
@@ -91,14 +91,17 @@ struct dso_cache {
};
/*
- * DSOs are put into a list for fast iteration.
+ * DSOs are put into both a list for fast iteration and rbtree for fast
+ * long name lookup.
*/
struct dsos {
struct list_head head;
+ struct rb_root root; /* rbtree root sorted by long name */
};
struct dso {
struct list_head node;
+ struct rb_node rb_node; /* rbtree node sorted by long name */
struct rb_root symbols[MAP__NR_TYPES];
struct rb_root symbol_names[MAP__NR_TYPES];
void *a2l;
diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c
index 49a75ec4c47b..b7d477fbda02 100644
--- a/tools/perf/util/machine.c
+++ b/tools/perf/util/machine.c
@@ -77,6 +77,7 @@ static void dsos__delete(struct dsos *dsos)
struct dso *pos, *n;
list_for_each_entry_safe(pos, n, &dsos->head, node) {
+ RB_CLEAR_NODE(&pos->rb_node);
list_del(&pos->node);
dso__delete(pos);
}
--
1.9.3
next prev parent reply other threads:[~2014-10-01 19:51 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-10-01 19:50 [GIT PULL 0/8] perf/core improvements and fixes Arnaldo Carvalho de Melo
2014-10-01 19:50 ` [PATCH 1/8] perf tools: Refactor unit and scale function parameters Arnaldo Carvalho de Melo
2014-10-01 19:50 ` [PATCH 2/8] perf trace: Fix mmap return address truncation to 32-bit Arnaldo Carvalho de Melo
2014-10-01 19:50 ` [PATCH 3/8] perf bench futex: Support operations for shared futexes Arnaldo Carvalho de Melo
2014-10-01 19:50 ` [PATCH 4/8] perf bench futex: Sanitize -q option in requeue Arnaldo Carvalho de Melo
2014-10-01 19:50 ` [PATCH 5/8] perf symbols: Encapsulate dsos list head into struct dsos Arnaldo Carvalho de Melo
2014-10-01 19:50 ` Arnaldo Carvalho de Melo [this message]
2014-10-14 9:09 ` [PATCH 6/8] perf symbols: Improve DSO long names lookup speed with rbtree Jiri Olsa
2014-10-14 17:34 ` Arnaldo Carvalho de Melo
2014-10-14 18:03 ` Arnaldo Carvalho de Melo
2014-10-15 10:05 ` [tip:perf/urgent] perf machine: Add missing dsos-> root rbtree root initialization tip-bot for Arnaldo Carvalho de Melo
2014-10-01 19:50 ` [PATCH 7/8] perf tools: Fix build breakage on arm64 targets Arnaldo Carvalho de Melo
2014-10-01 19:50 ` [PATCH 8/8] perf record: Fix error message for --filter option not coming after tracepoint Arnaldo Carvalho de Melo
2014-10-03 3:31 ` [GIT PULL 0/8] perf/core improvements and fixes Ingo Molnar
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1412193043-15493-7-git-send-email-acme@kernel.org \
--to=acme@kernel.org \
--cc=Waiman.Long@hp.com \
--cc=acme@redhat.com \
--cc=adrian.hunter@intel.com \
--cc=doug.hatch@hp.com \
--cc=dzickus@redhat.com \
--cc=jolsa@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=mingo@redhat.com \
--cc=namhyung@kernel.org \
--cc=paulus@samba.org \
--cc=peterz@infradead.org \
--cc=scott.norton@hp.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
Powered by JetHome