From: Byungchul Park <byungchul.park@lge.com>
To: torvalds@linux-foundation.org
Cc: damien.lemoal@opensource.wdc.com, linux-ide@vger.kernel.org,
adilger.kernel@dilger.ca, linux-ext4@vger.kernel.org,
mingo@redhat.com, linux-kernel@vger.kernel.org,
peterz@infradead.org, will@kernel.org, tglx@linutronix.de,
rostedt@goodmis.org, joel@joelfernandes.org, sashal@kernel.org,
daniel.vetter@ffwll.ch, chris@chris-wilson.co.uk,
duyuyang@gmail.com, johannes.berg@intel.com, tj@kernel.org,
tytso@mit.edu, willy@infradead.org, david@fromorbit.com,
amir73il@gmail.com, bfields@fieldses.org,
gregkh@linuxfoundation.org, kernel-team@lge.com,
linux-mm@kvack.org, akpm@linux-foundation.org, mhocko@kernel.org,
minchan@kernel.org, hannes@cmpxchg.org, vdavydov.dev@gmail.com,
sj@kernel.org, jglisse@redhat.com, dennis@kernel.org,
cl@linux.com, penberg@kernel.org, rientjes@google.com,
vbabka@suse.cz, ngupta@vflare.org, linux-block@vger.kernel.org,
paolo.valente@linaro.org, josef@toxicpanda.com,
linux-fsdevel@vger.kernel.org, viro@zeniv.linux.org.uk,
jack@suse.cz, jack@suse.com, jlayton@kernel.org,
dan.j.williams@intel.com, hch@infradead.org, djwong@kernel.org,
dri-devel@lists.freedesktop.org, airlied@linux.ie,
rodrigosiqueiramelo@gmail.com, melissa.srw@gmail.com,
hamohammed.sa@gmail.com, 42.hyeyoo@gmail.com
Subject: [PATCH RFC v6 10/21] dept: Introduce split map concept and new APIs for them
Date: Wed, 4 May 2022 17:17:38 +0900 [thread overview]
Message-ID: <1651652269-15342-11-git-send-email-byungchul.park@lge.com> (raw)
In-Reply-To: <1651652269-15342-1-git-send-email-byungchul.park@lge.com>
There is a case where total maps for its wait/event is so large in size.
For instance, struct page for PG_locked and PG_writeback is the case.
The additional memory size for the maps would be 'the # of pages *
sizeof(struct dept_map)' if each struct page keeps its map all the way,
which might be too big to accept in some systems.
It'd better have split map. One is for each instance and the other is
for what is commonly used. So split map and added new APIs for them.
Signed-off-by: Byungchul Park <byungchul.park@lge.com>
---
include/linux/dept.h | 78 ++++++++++++++++++-------
kernel/dependency/dept.c | 146 +++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 203 insertions(+), 21 deletions(-)
diff --git a/include/linux/dept.h b/include/linux/dept.h
index c498060..9698134 100644
--- a/include/linux/dept.h
+++ b/include/linux/dept.h
@@ -367,6 +367,30 @@ struct dept_map {
.nocheck = false, \
}
+struct dept_map_each {
+ /*
+ * wait timestamp associated to this map
+ */
+ unsigned int wgen;
+};
+
+struct dept_map_common {
+ const char *name;
+ struct dept_key *keys;
+ int sub_usr;
+
+ /*
+ * It's local copy for fast acces to the associated classes. And
+ * Also used for dept_key instance for statically defined map.
+ */
+ struct dept_key keys_local;
+
+ /*
+ * whether this map should be going to be checked or not
+ */
+ bool nocheck;
+};
+
struct dept_task {
/*
* all event contexts that have entered and before exiting
@@ -468,6 +492,11 @@ struct dept_task {
extern void dept_ask_event(struct dept_map *m);
extern void dept_event(struct dept_map *m, unsigned long e_f, unsigned long ip, const char *e_fn);
extern void dept_ecxt_exit(struct dept_map *m, unsigned long e_f, unsigned long ip);
+extern void dept_split_map_each_init(struct dept_map_each *me);
+extern void dept_split_map_common_init(struct dept_map_common *mc, struct dept_key *k, const char *n);
+extern void dept_wait_split_map(struct dept_map_each *me, struct dept_map_common *mc, unsigned long ip, const char *w_fn, int ne);
+extern void dept_event_split_map(struct dept_map_each *me, struct dept_map_common *mc, unsigned long ip, const char *e_fn);
+extern void dept_ask_event_split_map(struct dept_map_each *me, struct dept_map_common *mc);
static inline void dept_ecxt_enter_nokeep(struct dept_map *m)
{
@@ -490,32 +519,39 @@ static inline void dept_ecxt_enter_nokeep(struct dept_map *m)
#else /* !CONFIG_DEPT */
struct dept_key { };
struct dept_map { };
+struct dept_map_each { };
+struct dept_map_commmon { };
struct dept_task { };
#define DEPT_MAP_INITIALIZER(n) { }
#define DEPT_TASK_INITIALIZER(t) { }
-#define dept_on() do { } while (0)
-#define dept_off() do { } while (0)
-#define dept_init() do { } while (0)
-#define dept_task_init(t) do { } while (0)
-#define dept_task_exit(t) do { } while (0)
-#define dept_free_range(s, sz) do { } while (0)
-#define dept_map_init(m, k, s, n) do { (void)(n); (void)(k); } while (0)
-#define dept_map_reinit(m) do { } while (0)
-#define dept_map_nocheck(m) do { } while (0)
-
-#define dept_wait(m, w_f, ip, w_fn, ne) do { (void)(w_fn); } while (0)
-#define dept_stage_wait(m, w_f, w_fn, ne) do { (void)(w_fn); } while (0)
-#define dept_ask_event_wait_commit(ip) do { } while (0)
-#define dept_clean_stage() do { } while (0)
-#define dept_ecxt_enter(m, e_f, ip, c_fn, e_fn, ne) do { (void)(c_fn); (void)(e_fn); } while (0)
-#define dept_ask_event(m) do { } while (0)
-#define dept_event(m, e_f, ip, e_fn) do { (void)(e_fn); } while (0)
-#define dept_ecxt_exit(m, e_f, ip) do { } while (0)
-#define dept_ecxt_enter_nokeep(m) do { } while (0)
-#define dept_key_init(k) do { (void)(k); } while (0)
-#define dept_key_destroy(k) do { (void)(k); } while (0)
+#define dept_on() do { } while (0)
+#define dept_off() do { } while (0)
+#define dept_init() do { } while (0)
+#define dept_task_init(t) do { } while (0)
+#define dept_task_exit(t) do { } while (0)
+#define dept_free_range(s, sz) do { } while (0)
+#define dept_map_init(m, k, s, n) do { (void)(n); (void)(k); } while (0)
+#define dept_map_reinit(m) do { } while (0)
+#define dept_map_nocheck(m) do { } while (0)
+
+#define dept_wait(m, w_f, ip, w_fn, ne) do { (void)(w_fn); } while (0)
+#define dept_stage_wait(m, w_f, w_fn, ne) do { (void)(w_fn); } while (0)
+#define dept_ask_event_wait_commit(ip) do { } while (0)
+#define dept_clean_stage() do { } while (0)
+#define dept_ecxt_enter(m, e_f, ip, c_fn, e_fn, ne) do { (void)(c_fn); (void)(e_fn); } while (0)
+#define dept_ask_event(m) do { } while (0)
+#define dept_event(m, e_f, ip, e_fn) do { (void)(e_fn); } while (0)
+#define dept_ecxt_exit(m, e_f, ip) do { } while (0)
+#define dept_split_map_each_init(me) do { } while (0)
+#define dept_split_map_common_init(mc, k, n) do { (void)(n); (void)(k); } while (0)
+#define dept_wait_split_map(me, mc, ip, w_fn, ne) do { } while (0)
+#define dept_event_split_map(me, mc, ip, e_fn) do { } while (0)
+#define dept_ask_event_split_map(me, mc) do { } while (0)
+#define dept_ecxt_enter_nokeep(m) do { } while (0)
+#define dept_key_init(k) do { (void)(k); } while (0)
+#define dept_key_destroy(k) do { (void)(k); } while (0)
#define dept_softirq_enter() do { } while (0)
#define dept_hardirq_enter() do { } while (0)
diff --git a/kernel/dependency/dept.c b/kernel/dependency/dept.c
index 4670eec..a0413f1 100644
--- a/kernel/dependency/dept.c
+++ b/kernel/dependency/dept.c
@@ -2427,6 +2427,152 @@ void dept_ecxt_exit(struct dept_map *m, unsigned long e_f,
}
EXPORT_SYMBOL_GPL(dept_ecxt_exit);
+void dept_split_map_each_init(struct dept_map_each *me)
+{
+ unsigned long flags;
+
+ if (unlikely(READ_ONCE(dept_stop) || in_nmi()))
+ return;
+
+ /*
+ * Allow recursive entrance.
+ */
+ flags = dept_enter_recursive();
+
+ me->wgen = 0U;
+
+ dept_exit_recursive(flags);
+}
+EXPORT_SYMBOL_GPL(dept_split_map_each_init);
+
+void dept_split_map_common_init(struct dept_map_common *mc,
+ struct dept_key *k, const char *n)
+{
+ unsigned long flags;
+
+ if (unlikely(READ_ONCE(dept_stop) || in_nmi()))
+ return;
+
+ /*
+ * Allow recursive entrance.
+ */
+ flags = dept_enter_recursive();
+
+ clean_classes_cache(&mc->keys_local);
+
+ /*
+ * sub_usr is not used with split map.
+ */
+ mc->sub_usr = 0;
+ mc->keys = k;
+ mc->name = n;
+ mc->nocheck = false;
+
+ dept_exit_recursive(flags);
+}
+EXPORT_SYMBOL_GPL(dept_split_map_common_init);
+
+void dept_wait_split_map(struct dept_map_each *me,
+ struct dept_map_common *mc,
+ unsigned long ip, const char *w_fn, int ne)
+{
+ struct dept_task *dt = dept_task();
+ struct dept_class *c;
+ struct dept_key *k;
+ unsigned long flags;
+
+ if (unlikely(READ_ONCE(dept_stop) || in_nmi()))
+ return;
+
+ if (dt->recursive)
+ return;
+
+ if (mc->nocheck)
+ return;
+
+ flags = dept_enter();
+
+ k = mc->keys ?: &mc->keys_local;
+ c = check_new_class(&mc->keys_local, k, 0, mc->name);
+ if (c)
+ add_wait(c, ip, w_fn, ne);
+
+ dept_exit(flags);
+}
+EXPORT_SYMBOL_GPL(dept_wait_split_map);
+
+void dept_ask_event_split_map(struct dept_map_each *me,
+ struct dept_map_common *mc)
+{
+ unsigned int wg;
+ unsigned long flags;
+
+ if (unlikely(READ_ONCE(dept_stop) || in_nmi()))
+ return;
+
+ if (mc->nocheck)
+ return;
+
+ /*
+ * Allow recursive entrance.
+ */
+ flags = dept_enter_recursive();
+
+ /*
+ * Avoid zero wgen.
+ */
+ wg = atomic_inc_return(&wgen) ?: atomic_inc_return(&wgen);
+ WRITE_ONCE(me->wgen, wg);
+
+ dept_exit_recursive(flags);
+}
+EXPORT_SYMBOL_GPL(dept_ask_event_split_map);
+
+void dept_event_split_map(struct dept_map_each *me,
+ struct dept_map_common *mc,
+ unsigned long ip, const char *e_fn)
+{
+ struct dept_task *dt = dept_task();
+ struct dept_class *c;
+ struct dept_key *k;
+ unsigned long flags;
+
+ if (unlikely(READ_ONCE(dept_stop) || in_nmi()))
+ return;
+
+ if (dt->recursive) {
+ /*
+ * Dept won't work with this map even though anyway an
+ * event has been just triggered. Don't make it confused
+ * at that time handling the next event. Disable it
+ * until the next real case.
+ */
+ WRITE_ONCE(me->wgen, 0U);
+ return;
+ }
+
+ if (mc->nocheck)
+ return;
+
+ flags = dept_enter();
+
+ k = mc->keys ?: &mc->keys_local;
+ c = check_new_class(&mc->keys_local, k, 0, mc->name);
+
+ if (c && add_ecxt((void *)me, c, 0UL, NULL, e_fn, 0)) {
+ do_event((void *)me, c, READ_ONCE(me->wgen), ip);
+ pop_ecxt((void *)me, c);
+ }
+
+ /*
+ * Keep the map diabled until the next sleep.
+ */
+ WRITE_ONCE(me->wgen, 0U);
+
+ dept_exit(flags);
+}
+EXPORT_SYMBOL_GPL(dept_event_split_map);
+
void dept_task_exit(struct task_struct *t)
{
struct dept_task *dt = &t->dept_task;
--
1.9.1
next prev parent reply other threads:[~2022-05-04 8:19 UTC|newest]
Thread overview: 52+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-05-04 8:17 [PATCH RFC v6 00/21] DEPT(Dependency Tracker) Byungchul Park
2022-05-04 8:17 ` [PATCH RFC v6 01/21] llist: Move llist_{head,node} definition to types.h Byungchul Park
2022-05-04 8:17 ` [PATCH RFC v6 02/21] dept: Implement Dept(Dependency Tracker) Byungchul Park
2022-05-21 3:24 ` Hyeonggon Yoo
2022-05-04 8:17 ` [PATCH RFC v6 03/21] dept: Apply Dept to spinlock Byungchul Park
2022-05-04 8:17 ` [PATCH RFC v6 04/21] dept: Apply Dept to mutex families Byungchul Park
2022-05-04 8:17 ` [PATCH RFC v6 05/21] dept: Apply Dept to rwlock Byungchul Park
2022-05-04 8:17 ` [PATCH RFC v6 06/21] dept: Apply Dept to wait_for_completion()/complete() Byungchul Park
2022-05-04 8:17 ` [PATCH RFC v6 07/21] dept: Apply Dept to seqlock Byungchul Park
2022-05-21 5:25 ` Hyeonggon Yoo
2022-05-24 6:00 ` Byungchul Park
2022-05-04 8:17 ` [PATCH RFC v6 08/21] dept: Apply Dept to rwsem Byungchul Park
2022-05-04 8:17 ` [PATCH RFC v6 09/21] dept: Add proc knobs to show stats and dependency graph Byungchul Park
2022-05-04 8:17 ` Byungchul Park [this message]
2022-05-04 8:17 ` [PATCH RFC v6 11/21] dept: Apply Dept to wait/event of PG_{locked,writeback} Byungchul Park
2022-05-04 8:17 ` [PATCH RFC v6 12/21] dept: Apply SDT to swait Byungchul Park
2022-05-04 8:17 ` [PATCH RFC v6 13/21] dept: Apply SDT to wait(waitqueue) Byungchul Park
2022-05-04 8:17 ` [PATCH RFC v6 14/21] locking/lockdep, cpu/hotplus: Use a weaker annotation in AP thread Byungchul Park
2022-05-04 8:17 ` [PATCH RFC v6 15/21] dept: Distinguish each syscall context from another Byungchul Park
2022-05-04 8:17 ` [PATCH RFC v6 16/21] dept: Distinguish each work " Byungchul Park
2022-05-04 11:23 ` Sergey Shtylyov
2022-05-04 8:17 ` [PATCH RFC v6 17/21] dept: Disable Dept within the wait_bit layer by default Byungchul Park
2022-05-04 8:17 ` [PATCH RFC v6 18/21] dept: Disable Dept on struct crypto_larval's completion for now Byungchul Park
2022-05-04 8:17 ` [PATCH RFC v6 19/21] dept: Differentiate onstack maps from others of different tasks in class Byungchul Park
2022-05-04 8:17 ` [PATCH RFC v6 20/21] dept: Do not add dependencies between events within scheduler and sleeps Byungchul Park
2022-05-04 8:17 ` [PATCH RFC v6 21/21] dept: Unstage wait when tagging a normal sleep wait Byungchul Park
2022-05-04 18:17 ` [PATCH RFC v6 00/21] DEPT(Dependency Tracker) Linus Torvalds
2022-05-06 0:11 ` Byungchul Park
2022-05-07 7:20 ` Hyeonggon Yoo
2022-05-09 0:16 ` Byungchul Park
2022-05-09 20:47 ` Steven Rostedt
2022-05-09 23:38 ` Byungchul Park
2022-05-10 14:12 ` Steven Rostedt
2022-05-10 23:26 ` Byungchul Park
2022-05-10 11:18 ` Hyeonggon Yoo
2022-05-10 23:39 ` Byungchul Park
2022-05-11 10:04 ` Hyeonggon Yoo
2022-05-19 10:11 ` Catalin Marinas
2022-05-23 2:43 ` Byungchul Park
2022-05-09 1:22 ` Byungchul Park
2022-05-09 21:05 ` Theodore Ts'o
2022-05-09 22:28 ` Theodore Ts'o
2022-05-10 0:32 ` Byungchul Park
2022-05-10 1:32 ` Theodore Ts'o
2022-05-10 5:37 ` Byungchul Park
2022-05-11 1:16 ` Byungchul Park
2022-05-12 5:25 ` [REPORT] syscall reboot + umh + firmware fallback Byungchul Park
2022-05-12 9:15 ` Tejun Heo
2022-05-12 11:18 ` Byungchul Park
2022-05-12 13:56 ` Theodore Ts'o
2022-05-23 1:10 ` Byungchul Park
2022-05-12 16:41 ` Tejun Heo
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=1651652269-15342-11-git-send-email-byungchul.park@lge.com \
--to=byungchul.park@lge.com \
--cc=42.hyeyoo@gmail.com \
--cc=adilger.kernel@dilger.ca \
--cc=airlied@linux.ie \
--cc=akpm@linux-foundation.org \
--cc=amir73il@gmail.com \
--cc=bfields@fieldses.org \
--cc=chris@chris-wilson.co.uk \
--cc=cl@linux.com \
--cc=damien.lemoal@opensource.wdc.com \
--cc=dan.j.williams@intel.com \
--cc=daniel.vetter@ffwll.ch \
--cc=david@fromorbit.com \
--cc=dennis@kernel.org \
--cc=djwong@kernel.org \
--cc=dri-devel@lists.freedesktop.org \
--cc=duyuyang@gmail.com \
--cc=gregkh@linuxfoundation.org \
--cc=hamohammed.sa@gmail.com \
--cc=hannes@cmpxchg.org \
--cc=hch@infradead.org \
--cc=jack@suse.com \
--cc=jack@suse.cz \
--cc=jglisse@redhat.com \
--cc=jlayton@kernel.org \
--cc=joel@joelfernandes.org \
--cc=johannes.berg@intel.com \
--cc=josef@toxicpanda.com \
--cc=kernel-team@lge.com \
--cc=linux-block@vger.kernel.org \
--cc=linux-ext4@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-ide@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=melissa.srw@gmail.com \
--cc=mhocko@kernel.org \
--cc=minchan@kernel.org \
--cc=mingo@redhat.com \
--cc=ngupta@vflare.org \
--cc=paolo.valente@linaro.org \
--cc=penberg@kernel.org \
--cc=peterz@infradead.org \
--cc=rientjes@google.com \
--cc=rodrigosiqueiramelo@gmail.com \
--cc=rostedt@goodmis.org \
--cc=sashal@kernel.org \
--cc=sj@kernel.org \
--cc=tglx@linutronix.de \
--cc=tj@kernel.org \
--cc=torvalds@linux-foundation.org \
--cc=tytso@mit.edu \
--cc=vbabka@suse.cz \
--cc=vdavydov.dev@gmail.com \
--cc=viro@zeniv.linux.org.uk \
--cc=will@kernel.org \
--cc=willy@infradead.org \
/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).