linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: tip-bot for Davidlohr Bueso <tipbot@zytor.com>
To: linux-tip-commits@vger.kernel.org
Cc: acme@redhat.com, dbueso@suse.de, namhyung@kernel.org,
	dave@stgolabs.net, jolsa@kernel.org, mingo@kernel.org,
	linux-kernel@vger.kernel.org, hpa@zytor.com, tglx@linutronix.de
Subject: [tip:perf/core] perf machine: Use cached rbtrees
Date: Sat, 26 Jan 2019 02:01:55 -0800	[thread overview]
Message-ID: <tip-f3acb3a8a2081344801974ac5ec8e1b0d6f0ef36@git.kernel.org> (raw)
In-Reply-To: <20181206191819.30182-3-dave@stgolabs.net>

Commit-ID:  f3acb3a8a2081344801974ac5ec8e1b0d6f0ef36
Gitweb:     https://git.kernel.org/tip/f3acb3a8a2081344801974ac5ec8e1b0d6f0ef36
Author:     Davidlohr Bueso <dave@stgolabs.net>
AuthorDate: Thu, 6 Dec 2018 11:18:14 -0800
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Fri, 25 Jan 2019 15:12:09 +0100

perf machine: Use cached rbtrees

At the cost of an extra pointer, we can avoid the O(logN) cost of
finding the first element in the tree (smallest node), which is
something required for nearly every operation dealing with
machine->guests and threads->entries.

The conversion is straightforward, however, it's worth noticing that the
rb_erase_init() calls have been replaced by rb_erase_cached() which has
no _init() flavor, however, the node is explicitly cleared next anyway,
which was redundant until now.

Signed-off-by: Davidlohr Bueso <dbueso@suse.de>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: http://lkml.kernel.org/r/20181206191819.30182-3-dave@stgolabs.net
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-report.c |  3 ++-
 tools/perf/util/build-id.c  | 12 ++++++----
 tools/perf/util/machine.c   | 53 ++++++++++++++++++++++++++-------------------
 tools/perf/util/machine.h   | 12 +++++-----
 tools/perf/util/rb_resort.h |  6 ++---
 5 files changed, 50 insertions(+), 36 deletions(-)

diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index 794a0de5ac1d..c9ceaf88759c 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -753,7 +753,8 @@ static int tasks_print(struct report *rep, FILE *fp)
 	for (i = 0; i < THREADS__TABLE_SIZE; i++) {
 		struct threads *threads = &machine->threads[i];
 
-		for (nd = rb_first(&threads->entries); nd; nd = rb_next(nd)) {
+		for (nd = rb_first_cached(&threads->entries); nd;
+		     nd = rb_next(nd)) {
 			task = tasks + itask++;
 
 			task->thread = rb_entry(nd, struct thread, rb_node);
diff --git a/tools/perf/util/build-id.c b/tools/perf/util/build-id.c
index 07ef7fad689c..2ff802068f06 100644
--- a/tools/perf/util/build-id.c
+++ b/tools/perf/util/build-id.c
@@ -364,7 +364,8 @@ int perf_session__write_buildid_table(struct perf_session *session,
 	if (err)
 		return err;
 
-	for (nd = rb_first(&session->machines.guests); nd; nd = rb_next(nd)) {
+	for (nd = rb_first_cached(&session->machines.guests); nd;
+	     nd = rb_next(nd)) {
 		struct machine *pos = rb_entry(nd, struct machine, rb_node);
 		err = machine__write_buildid_table(pos, fd);
 		if (err)
@@ -397,7 +398,8 @@ int dsos__hit_all(struct perf_session *session)
 	if (err)
 		return err;
 
-	for (nd = rb_first(&session->machines.guests); nd; nd = rb_next(nd)) {
+	for (nd = rb_first_cached(&session->machines.guests); nd;
+	     nd = rb_next(nd)) {
 		struct machine *pos = rb_entry(nd, struct machine, rb_node);
 
 		err = machine__hit_all_dsos(pos);
@@ -850,7 +852,8 @@ int perf_session__cache_build_ids(struct perf_session *session)
 
 	ret = machine__cache_build_ids(&session->machines.host);
 
-	for (nd = rb_first(&session->machines.guests); nd; nd = rb_next(nd)) {
+	for (nd = rb_first_cached(&session->machines.guests); nd;
+	     nd = rb_next(nd)) {
 		struct machine *pos = rb_entry(nd, struct machine, rb_node);
 		ret |= machine__cache_build_ids(pos);
 	}
@@ -867,7 +870,8 @@ bool perf_session__read_build_ids(struct perf_session *session, bool with_hits)
 	struct rb_node *nd;
 	bool ret = machine__read_build_ids(&session->machines.host, with_hits);
 
-	for (nd = rb_first(&session->machines.guests); nd; nd = rb_next(nd)) {
+	for (nd = rb_first_cached(&session->machines.guests); nd;
+	     nd = rb_next(nd)) {
 		struct machine *pos = rb_entry(nd, struct machine, rb_node);
 		ret |= machine__read_build_ids(pos, with_hits);
 	}
diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c
index ae85106bb5bf..66f019fdc510 100644
--- a/tools/perf/util/machine.c
+++ b/tools/perf/util/machine.c
@@ -42,7 +42,7 @@ static void machine__threads_init(struct machine *machine)
 
 	for (i = 0; i < THREADS__TABLE_SIZE; i++) {
 		struct threads *threads = &machine->threads[i];
-		threads->entries = RB_ROOT;
+		threads->entries = RB_ROOT_CACHED;
 		init_rwsem(&threads->lock);
 		threads->nr = 0;
 		INIT_LIST_HEAD(&threads->dead);
@@ -180,7 +180,7 @@ void machine__delete_threads(struct machine *machine)
 	for (i = 0; i < THREADS__TABLE_SIZE; i++) {
 		struct threads *threads = &machine->threads[i];
 		down_write(&threads->lock);
-		nd = rb_first(&threads->entries);
+		nd = rb_first_cached(&threads->entries);
 		while (nd) {
 			struct thread *t = rb_entry(nd, struct thread, rb_node);
 
@@ -223,7 +223,7 @@ void machine__delete(struct machine *machine)
 void machines__init(struct machines *machines)
 {
 	machine__init(&machines->host, "", HOST_KERNEL_ID);
-	machines->guests = RB_ROOT;
+	machines->guests = RB_ROOT_CACHED;
 }
 
 void machines__exit(struct machines *machines)
@@ -235,9 +235,10 @@ void machines__exit(struct machines *machines)
 struct machine *machines__add(struct machines *machines, pid_t pid,
 			      const char *root_dir)
 {
-	struct rb_node **p = &machines->guests.rb_node;
+	struct rb_node **p = &machines->guests.rb_root.rb_node;
 	struct rb_node *parent = NULL;
 	struct machine *pos, *machine = malloc(sizeof(*machine));
+	bool leftmost = true;
 
 	if (machine == NULL)
 		return NULL;
@@ -252,12 +253,14 @@ struct machine *machines__add(struct machines *machines, pid_t pid,
 		pos = rb_entry(parent, struct machine, rb_node);
 		if (pid < pos->pid)
 			p = &(*p)->rb_left;
-		else
+		else {
 			p = &(*p)->rb_right;
+			leftmost = false;
+		}
 	}
 
 	rb_link_node(&machine->rb_node, parent, p);
-	rb_insert_color(&machine->rb_node, &machines->guests);
+	rb_insert_color_cached(&machine->rb_node, &machines->guests, leftmost);
 
 	return machine;
 }
@@ -268,7 +271,7 @@ void machines__set_comm_exec(struct machines *machines, bool comm_exec)
 
 	machines->host.comm_exec = comm_exec;
 
-	for (nd = rb_first(&machines->guests); nd; nd = rb_next(nd)) {
+	for (nd = rb_first_cached(&machines->guests); nd; nd = rb_next(nd)) {
 		struct machine *machine = rb_entry(nd, struct machine, rb_node);
 
 		machine->comm_exec = comm_exec;
@@ -277,7 +280,7 @@ void machines__set_comm_exec(struct machines *machines, bool comm_exec)
 
 struct machine *machines__find(struct machines *machines, pid_t pid)
 {
-	struct rb_node **p = &machines->guests.rb_node;
+	struct rb_node **p = &machines->guests.rb_root.rb_node;
 	struct rb_node *parent = NULL;
 	struct machine *machine;
 	struct machine *default_machine = NULL;
@@ -340,7 +343,7 @@ void machines__process_guests(struct machines *machines,
 {
 	struct rb_node *nd;
 
-	for (nd = rb_first(&machines->guests); nd; nd = rb_next(nd)) {
+	for (nd = rb_first_cached(&machines->guests); nd; nd = rb_next(nd)) {
 		struct machine *pos = rb_entry(nd, struct machine, rb_node);
 		process(pos, data);
 	}
@@ -353,7 +356,8 @@ void machines__set_id_hdr_size(struct machines *machines, u16 id_hdr_size)
 
 	machines->host.id_hdr_size = id_hdr_size;
 
-	for (node = rb_first(&machines->guests); node; node = rb_next(node)) {
+	for (node = rb_first_cached(&machines->guests); node;
+	     node = rb_next(node)) {
 		machine = rb_entry(node, struct machine, rb_node);
 		machine->id_hdr_size = id_hdr_size;
 	}
@@ -466,9 +470,10 @@ static struct thread *____machine__findnew_thread(struct machine *machine,
 						  pid_t pid, pid_t tid,
 						  bool create)
 {
-	struct rb_node **p = &threads->entries.rb_node;
+	struct rb_node **p = &threads->entries.rb_root.rb_node;
 	struct rb_node *parent = NULL;
 	struct thread *th;
+	bool leftmost = true;
 
 	th = threads__get_last_match(threads, machine, pid, tid);
 	if (th)
@@ -486,8 +491,10 @@ static struct thread *____machine__findnew_thread(struct machine *machine,
 
 		if (tid < th->tid)
 			p = &(*p)->rb_left;
-		else
+		else {
 			p = &(*p)->rb_right;
+			leftmost = false;
+		}
 	}
 
 	if (!create)
@@ -496,7 +503,7 @@ static struct thread *____machine__findnew_thread(struct machine *machine,
 	th = thread__new(pid, tid);
 	if (th != NULL) {
 		rb_link_node(&th->rb_node, parent, p);
-		rb_insert_color(&th->rb_node, &threads->entries);
+		rb_insert_color_cached(&th->rb_node, &threads->entries, leftmost);
 
 		/*
 		 * We have to initialize map_groups separately
@@ -507,7 +514,7 @@ static struct thread *____machine__findnew_thread(struct machine *machine,
 		 * leader and that would screwed the rb tree.
 		 */
 		if (thread__init_map_groups(th, machine)) {
-			rb_erase_init(&th->rb_node, &threads->entries);
+			rb_erase_cached(&th->rb_node, &threads->entries);
 			RB_CLEAR_NODE(&th->rb_node);
 			thread__put(th);
 			return NULL;
@@ -798,7 +805,7 @@ size_t machines__fprintf_dsos(struct machines *machines, FILE *fp)
 	struct rb_node *nd;
 	size_t ret = __dsos__fprintf(&machines->host.dsos.head, fp);
 
-	for (nd = rb_first(&machines->guests); nd; nd = rb_next(nd)) {
+	for (nd = rb_first_cached(&machines->guests); nd; nd = rb_next(nd)) {
 		struct machine *pos = rb_entry(nd, struct machine, rb_node);
 		ret += __dsos__fprintf(&pos->dsos.head, fp);
 	}
@@ -818,7 +825,7 @@ size_t machines__fprintf_dsos_buildid(struct machines *machines, FILE *fp,
 	struct rb_node *nd;
 	size_t ret = machine__fprintf_dsos_buildid(&machines->host, fp, skip, parm);
 
-	for (nd = rb_first(&machines->guests); nd; nd = rb_next(nd)) {
+	for (nd = rb_first_cached(&machines->guests); nd; nd = rb_next(nd)) {
 		struct machine *pos = rb_entry(nd, struct machine, rb_node);
 		ret += machine__fprintf_dsos_buildid(pos, fp, skip, parm);
 	}
@@ -858,7 +865,8 @@ size_t machine__fprintf(struct machine *machine, FILE *fp)
 
 		ret = fprintf(fp, "Threads: %u\n", threads->nr);
 
-		for (nd = rb_first(&threads->entries); nd; nd = rb_next(nd)) {
+		for (nd = rb_first_cached(&threads->entries); nd;
+		     nd = rb_next(nd)) {
 			struct thread *pos = rb_entry(nd, struct thread, rb_node);
 
 			ret += thread__fprintf(pos, fp);
@@ -1161,7 +1169,7 @@ failure:
 
 void machines__destroy_kernel_maps(struct machines *machines)
 {
-	struct rb_node *next = rb_first(&machines->guests);
+	struct rb_node *next = rb_first_cached(&machines->guests);
 
 	machine__destroy_kernel_maps(&machines->host);
 
@@ -1169,7 +1177,7 @@ void machines__destroy_kernel_maps(struct machines *machines)
 		struct machine *pos = rb_entry(next, struct machine, rb_node);
 
 		next = rb_next(&pos->rb_node);
-		rb_erase(&pos->rb_node, &machines->guests);
+		rb_erase_cached(&pos->rb_node, &machines->guests);
 		machine__delete(pos);
 	}
 }
@@ -1734,7 +1742,7 @@ static void __machine__remove_thread(struct machine *machine, struct thread *th,
 	BUG_ON(refcount_read(&th->refcnt) == 0);
 	if (lock)
 		down_write(&threads->lock);
-	rb_erase_init(&th->rb_node, &threads->entries);
+	rb_erase_cached(&th->rb_node, &threads->entries);
 	RB_CLEAR_NODE(&th->rb_node);
 	--threads->nr;
 	/*
@@ -2511,7 +2519,8 @@ int machine__for_each_thread(struct machine *machine,
 
 	for (i = 0; i < THREADS__TABLE_SIZE; i++) {
 		threads = &machine->threads[i];
-		for (nd = rb_first(&threads->entries); nd; nd = rb_next(nd)) {
+		for (nd = rb_first_cached(&threads->entries); nd;
+		     nd = rb_next(nd)) {
 			thread = rb_entry(nd, struct thread, rb_node);
 			rc = fn(thread, priv);
 			if (rc != 0)
@@ -2538,7 +2547,7 @@ int machines__for_each_thread(struct machines *machines,
 	if (rc != 0)
 		return rc;
 
-	for (nd = rb_first(&machines->guests); nd; nd = rb_next(nd)) {
+	for (nd = rb_first_cached(&machines->guests); nd; nd = rb_next(nd)) {
 		struct machine *machine = rb_entry(nd, struct machine, rb_node);
 
 		rc = machine__for_each_thread(machine, fn, priv);
diff --git a/tools/perf/util/machine.h b/tools/perf/util/machine.h
index 4ecd380ce1b4..e2f3df45410a 100644
--- a/tools/perf/util/machine.h
+++ b/tools/perf/util/machine.h
@@ -29,11 +29,11 @@ struct vdso_info;
 #define THREADS__TABLE_SIZE	(1 << THREADS__TABLE_BITS)
 
 struct threads {
-	struct rb_root	  entries;
-	struct rw_semaphore lock;
-	unsigned int	  nr;
-	struct list_head  dead;
-	struct thread	  *last_match;
+	struct rb_root_cached  entries;
+	struct rw_semaphore    lock;
+	unsigned int	       nr;
+	struct list_head       dead;
+	struct thread	       *last_match;
 };
 
 struct machine {
@@ -140,7 +140,7 @@ typedef void (*machine__process_t)(struct machine *machine, void *data);
 
 struct machines {
 	struct machine host;
-	struct rb_root guests;
+	struct rb_root_cached guests;
 };
 
 void machines__init(struct machines *machines);
diff --git a/tools/perf/util/rb_resort.h b/tools/perf/util/rb_resort.h
index a920f702a74d..f272e181d3d6 100644
--- a/tools/perf/util/rb_resort.h
+++ b/tools/perf/util/rb_resort.h
@@ -144,8 +144,8 @@ struct __name##_sorted *__name = __name##_sorted__new
 				  __ilist->rblist.nr_entries)
 
 /* For 'struct machine->threads' */
-#define DECLARE_RESORT_RB_MACHINE_THREADS(__name, __machine, hash_bucket)	\
-	DECLARE_RESORT_RB(__name)(&__machine->threads[hash_bucket].entries,	\
-				  __machine->threads[hash_bucket].nr)
+#define DECLARE_RESORT_RB_MACHINE_THREADS(__name, __machine, hash_bucket)    \
+ DECLARE_RESORT_RB(__name)(&__machine->threads[hash_bucket].entries.rb_root, \
+			   __machine->threads[hash_bucket].nr)
 
 #endif /* _PERF_RESORT_RB_H_ */

  reply	other threads:[~2019-01-26 10:02 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-12-06 19:18 [PATCH v2 -tip 0/7] tools/perf: Update rbtree implementation and optimize users Davidlohr Bueso
2018-12-06 19:18 ` [PATCH 1/7] tools/perf: Update rbtree implementation Davidlohr Bueso
2019-01-26 10:01   ` [tip:perf/core] tools: " tip-bot for Davidlohr Bueso
2018-12-06 19:18 ` [PATCH 2/7] perf machine: Use cached rbtrees Davidlohr Bueso
2019-01-26 10:01   ` tip-bot for Davidlohr Bueso [this message]
2018-12-06 19:18 ` [PATCH 3/7] perf callchain: " Davidlohr Bueso
2019-01-26 10:02   ` [tip:perf/core] " tip-bot for Davidlohr Bueso
2018-12-06 19:18 ` [PATCH 4/7] perf util: Use cached rbtree for rblists Davidlohr Bueso
2019-01-26 10:03   ` [tip:perf/core] " tip-bot for Davidlohr Bueso
2018-12-06 19:18 ` [PATCH 5/7] perf symbols: Use cached rbtrees Davidlohr Bueso
2019-01-26 10:03   ` [tip:perf/core] " tip-bot for Davidlohr Bueso
2018-12-06 19:18 ` [PATCH 6/7] perf hist: " Davidlohr Bueso
2019-01-22 13:59   ` Arnaldo Carvalho de Melo
2019-01-22 15:22     ` Davidlohr Bueso
2019-01-26 10:04   ` [tip:perf/core] " tip-bot for Davidlohr Bueso
2018-12-06 19:18 ` [PATCH 7/7] perf sched: " Davidlohr Bueso
2019-01-26 10:05   ` [tip:perf/core] " tip-bot for Davidlohr Bueso

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=tip-f3acb3a8a2081344801974ac5ec8e1b0d6f0ef36@git.kernel.org \
    --to=tipbot@zytor.com \
    --cc=acme@redhat.com \
    --cc=dave@stgolabs.net \
    --cc=dbueso@suse.de \
    --cc=hpa@zytor.com \
    --cc=jolsa@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-tip-commits@vger.kernel.org \
    --cc=mingo@kernel.org \
    --cc=namhyung@kernel.org \
    --cc=tglx@linutronix.de \
    /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;
as well as URLs for NNTP newsgroup(s).