mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Daeho Jeong <daeho43@gmail.com>
To: linux-kernel@vger.kernel.org,
	linux-f2fs-devel@lists.sourceforge.net, kernel-team@android.com
Cc: Daeho Jeong <daehojeong@google.com>
Subject: [PATCH 6/7] f2fs: create gc_no_zoned_gc_percent and gc_boost_zoned_gc_percent
Date: Thu, 29 Aug 2024 14:52:41 -0700	[thread overview]
Message-ID: <20240829215242.3641502-6-daeho43@gmail.com> (raw)
In-Reply-To: <20240829215242.3641502-1-daeho43@gmail.com>

From: Daeho Jeong <daehojeong@google.com>

Added control knobs for gc_no_zoned_gc_percent and
gc_boost_zoned_gc_percent.

Signed-off-by: Daeho Jeong <daehojeong@google.com>
---
 Documentation/ABI/testing/sysfs-fs-f2fs | 14 ++++++++++++++
 fs/f2fs/gc.c                            | 12 +++++++++---
 fs/f2fs/gc.h                            |  4 ++++
 fs/f2fs/sysfs.c                         |  4 ++++
 4 files changed, 31 insertions(+), 3 deletions(-)

diff --git a/Documentation/ABI/testing/sysfs-fs-f2fs b/Documentation/ABI/testing/sysfs-fs-f2fs
index 4fcd0f824bde..265baec879fd 100644
--- a/Documentation/ABI/testing/sysfs-fs-f2fs
+++ b/Documentation/ABI/testing/sysfs-fs-f2fs
@@ -769,3 +769,17 @@ Date:		August 2024
 Contact:	"Daeho Jeong" <daehojeong@google.com>
 Description:	In order to fine tune GC behavior, we can control the number of
 		reserved segments.
+
+What:		/sys/fs/f2fs/<disk>/gc_no_zoned_gc_percent
+Date:		August 2024
+Contact:	"Daeho Jeong" <daehojeong@google.com>
+Description:	If the percentage of free sections over total sections is above this
+		number, F2FS do not garbage collection for zoned devices through the
+		background GC thread. the default number is "60".
+
+What:		/sys/fs/f2fs/<disk>/gc_boost_zoned_gc_percent
+Date:		August 2024
+Contact:	"Daeho Jeong" <daehojeong@google.com>
+Description:	If the percentage of free sections over total sections is under this
+		number, F2FS boosts garbage collection for zoned devices through the
+		background GC thread. the default number is "25".
diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c
index 37b47a8d95f1..6b79c43a57e3 100644
--- a/fs/f2fs/gc.c
+++ b/fs/f2fs/gc.c
@@ -119,7 +119,8 @@ static int gc_thread_func(void *data)
 		}
 
 		if (f2fs_sb_has_blkzoned(sbi)) {
-			if (has_enough_free_blocks(sbi, LIMIT_NO_ZONED_GC)) {
+			if (has_enough_free_blocks(sbi,
+				gc_th->no_zoned_gc_percent)) {
 				wait_ms = gc_th->no_gc_sleep_time;
 				f2fs_up_write(&sbi->gc_lock);
 				goto next;
@@ -200,10 +201,14 @@ int f2fs_start_gc_thread(struct f2fs_sb_info *sbi)
 		gc_th->min_sleep_time = DEF_GC_THREAD_MIN_SLEEP_TIME_ZONED;
 		gc_th->max_sleep_time = DEF_GC_THREAD_MAX_SLEEP_TIME_ZONED;
 		gc_th->no_gc_sleep_time = DEF_GC_THREAD_NOGC_SLEEP_TIME_ZONED;
+		gc_th->no_zoned_gc_percent = LIMIT_NO_ZONED_GC;
+		gc_th->boost_zoned_gc_percent = LIMIT_BOOST_ZONED_GC;
 	} else {
 		gc_th->min_sleep_time = DEF_GC_THREAD_MIN_SLEEP_TIME;
 		gc_th->max_sleep_time = DEF_GC_THREAD_MAX_SLEEP_TIME;
 		gc_th->no_gc_sleep_time = DEF_GC_THREAD_NOGC_SLEEP_TIME;
+		gc_th->no_zoned_gc_percent = 0;
+		gc_th->boost_zoned_gc_percent = 0;
 	}
 
 	gc_th->gc_wake = false;
@@ -1740,8 +1745,9 @@ static int do_garbage_collect(struct f2fs_sb_info *sbi,
 
 			if (f2fs_sb_has_blkzoned(sbi) &&
 					!has_enough_free_blocks(sbi,
-					LIMIT_BOOST_ZONED_GC))
-				migration_granularity *= BOOST_GC_MULTIPLE;
+					sbi->gc_thread->boost_zoned_gc_percent))
+				migration_granularity *=
+					BOOST_GC_MULTIPLE;
 
 			end_segno = start_segno + migration_granularity;
 		}
diff --git a/fs/f2fs/gc.h b/fs/f2fs/gc.h
index 33848d550153..e8195eb4ca6d 100644
--- a/fs/f2fs/gc.h
+++ b/fs/f2fs/gc.h
@@ -61,6 +61,10 @@ struct f2fs_gc_kthread {
 						 * caller of f2fs_balance_fs()
 						 * will wait on this wait queue.
 						 */
+
+	/* for gc control for zoned devices */
+	unsigned int no_zoned_gc_percent;
+	unsigned int boost_zoned_gc_percent;
 };
 
 struct gc_inode_list {
diff --git a/fs/f2fs/sysfs.c b/fs/f2fs/sysfs.c
index 2ed773132f40..647f6660f4df 100644
--- a/fs/f2fs/sysfs.c
+++ b/fs/f2fs/sysfs.c
@@ -960,6 +960,8 @@ GC_THREAD_RW_ATTR(gc_urgent_sleep_time, urgent_sleep_time);
 GC_THREAD_RW_ATTR(gc_min_sleep_time, min_sleep_time);
 GC_THREAD_RW_ATTR(gc_max_sleep_time, max_sleep_time);
 GC_THREAD_RW_ATTR(gc_no_gc_sleep_time, no_gc_sleep_time);
+GC_THREAD_RW_ATTR(gc_no_zoned_gc_percent, no_zoned_gc_percent);
+GC_THREAD_RW_ATTR(gc_boost_zoned_gc_percent, boost_zoned_gc_percent);
 
 /* SM_INFO ATTR */
 SM_INFO_RW_ATTR(reclaim_segments, rec_prefree_segments);
@@ -1117,6 +1119,8 @@ static struct attribute *f2fs_attrs[] = {
 	ATTR_LIST(gc_min_sleep_time),
 	ATTR_LIST(gc_max_sleep_time),
 	ATTR_LIST(gc_no_gc_sleep_time),
+	ATTR_LIST(gc_no_zoned_gc_percent),
+	ATTR_LIST(gc_boost_zoned_gc_percent),
 	ATTR_LIST(gc_idle),
 	ATTR_LIST(gc_urgent),
 	ATTR_LIST(reclaim_segments),
-- 
2.46.0.469.g59c65b2a67-goog


  parent reply	other threads:[~2024-08-29 21:52 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-08-29 21:52 [PATCH 1/7] f2fs: make BG GC more aggressive for zoned devices Daeho Jeong
2024-08-29 21:52 ` [PATCH 2/7] f2fs: read summary blocks with the correct amount for migration_granularity Daeho Jeong
2024-09-06  2:56   ` [f2fs-dev] " Chao Yu
2024-09-06 20:23     ` Daeho Jeong
2024-09-09  8:18       ` Chao Yu
2024-08-29 21:52 ` [PATCH 3/7] f2fs: add reserved_segments sysfs node Daeho Jeong
2024-09-06  3:04   ` [f2fs-dev] " Chao Yu
2024-08-29 21:52 ` [PATCH 4/7] f2fs: increase BG GC migration granularity when boosted for zoned devices Daeho Jeong
2024-09-06  3:37   ` [f2fs-dev] " Chao Yu
2024-08-29 21:52 ` [PATCH 5/7] f2fs: do FG_GC when GC boosting is required " Daeho Jeong
2024-09-06  6:34   ` [f2fs-dev] " Chao Yu
2024-08-29 21:52 ` Daeho Jeong [this message]
2024-09-06  6:35   ` [f2fs-dev] [PATCH 6/7] f2fs: create gc_no_zoned_gc_percent and gc_boost_zoned_gc_percent Chao Yu
2024-08-29 21:52 ` [PATCH 7/7] f2fs: add valid block ratio not to do excessive GC for one time GC Daeho Jeong
2024-09-09  9:22   ` [f2fs-dev] " Chao Yu
2024-09-06  2:21 ` [f2fs-dev] [PATCH 1/7] f2fs: make BG GC more aggressive for zoned devices Chao Yu

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=20240829215242.3641502-6-daeho43@gmail.com \
    --to=daeho43@gmail.com \
    --cc=daehojeong@google.com \
    --cc=kernel-team@android.com \
    --cc=linux-f2fs-devel@lists.sourceforge.net \
    --cc=linux-kernel@vger.kernel.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

Powered by JetHome