linux-ext4.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
To: linux-ext4@vger.kernel.org
Cc: tytso@mit.edu, Harshad Shirwadkar <harshadshirwadkar@gmail.com>,
	Jan Kara <jack@suse.cz>
Subject: [PATCH v2 08/22] jbd2: don't use state lock during commit path
Date: Thu,  5 Nov 2020 19:58:57 -0800	[thread overview]
Message-ID: <20201106035911.1942128-9-harshadshirwadkar@gmail.com> (raw)
In-Reply-To: <20201106035911.1942128-1-harshadshirwadkar@gmail.com>

Variables journal->j_fc_off, journal->j_fc_wbuf are accessed during
commit path. Since today we allow only one process to perform a fast
commit, there is no need take state lock before accessing these
variables. This patch removes these locks and adds comments to
describe this.

Suggested-by: Jan Kara <jack@suse.cz>
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
 fs/jbd2/journal.c    |  6 ------
 include/linux/jbd2.h | 10 ++++++----
 2 files changed, 6 insertions(+), 10 deletions(-)

diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
index 500152f0421a..778ea50fc8d1 100644
--- a/fs/jbd2/journal.c
+++ b/fs/jbd2/journal.c
@@ -865,7 +865,6 @@ int jbd2_fc_get_buf(journal_t *journal, struct buffer_head **bh_out)
 	int fc_off;
 
 	*bh_out = NULL;
-	write_lock(&journal->j_state_lock);
 
 	if (journal->j_fc_off + journal->j_fc_first < journal->j_fc_last) {
 		fc_off = journal->j_fc_off;
@@ -874,7 +873,6 @@ int jbd2_fc_get_buf(journal_t *journal, struct buffer_head **bh_out)
 	} else {
 		ret = -EINVAL;
 	}
-	write_unlock(&journal->j_state_lock);
 
 	if (ret)
 		return ret;
@@ -909,9 +907,7 @@ int jbd2_fc_wait_bufs(journal_t *journal, int num_blks)
 	struct buffer_head *bh;
 	int i, j_fc_off;
 
-	read_lock(&journal->j_state_lock);
 	j_fc_off = journal->j_fc_off;
-	read_unlock(&journal->j_state_lock);
 
 	/*
 	 * Wait in reverse order to minimize chances of us being woken up before
@@ -939,9 +935,7 @@ int jbd2_fc_release_bufs(journal_t *journal)
 	struct buffer_head *bh;
 	int i, j_fc_off;
 
-	read_lock(&journal->j_state_lock);
 	j_fc_off = journal->j_fc_off;
-	read_unlock(&journal->j_state_lock);
 
 	/*
 	 * Wait in reverse order to minimize chances of us being woken up before
diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
index b2caf7bbd8e5..5f0ef6380b0c 100644
--- a/include/linux/jbd2.h
+++ b/include/linux/jbd2.h
@@ -945,8 +945,9 @@ struct journal_s
 	/**
 	 * @j_fc_off:
 	 *
-	 * Number of fast commit blocks currently allocated.
-	 * [j_state_lock].
+	 * Number of fast commit blocks currently allocated. Accessed only
+	 * during fast commit. Currently only process can do fast commit, so
+	 * this field is not protected by any lock.
 	 */
 	unsigned long		j_fc_off;
 
@@ -1109,8 +1110,9 @@ struct journal_s
 	struct buffer_head	**j_wbuf;
 
 	/**
-	 * @j_fc_wbuf: Array of fast commit bhs for
-	 * jbd2_journal_commit_transaction.
+	 * @j_fc_wbuf: Array of fast commit bhs for fast commit. Accessed only
+	 * during a fast commit. Currently only process can do fast commit, so
+	 * this field is not protected by any lock.
 	 */
 	struct buffer_head	**j_fc_wbuf;
 
-- 
2.29.1.341.ge80a0c044ae-goog


  parent reply	other threads:[~2020-11-06  3:59 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-11-06  3:58 [PATCH v2 00/22] ext4 fast commit fixes Harshad Shirwadkar
2020-11-06  3:58 ` [PATCH v2 01/22] ext4: describe fast_commit feature flags Harshad Shirwadkar
2020-11-06  3:58 ` [PATCH v2 02/22] ext4: mark fc ineligible if inode gets evictied due to mem pressure Harshad Shirwadkar
2020-11-06  3:58 ` [PATCH v2 03/22] ext4: drop redundant calls ext4_fc_track_range Harshad Shirwadkar
2020-11-06  3:58 ` [PATCH v2 04/22] ext4: fixup ext4_fc_track_* functions' signature Harshad Shirwadkar
2020-11-06  3:58 ` [PATCH v2 05/22] jbd2: rename j_maxlen to j_total_len and add jbd2_journal_max_txn_bufs Harshad Shirwadkar
2020-11-06  3:58 ` [PATCH v2 06/22] ext4: clean up the JBD2 API that initializes fast commits Harshad Shirwadkar
2020-11-06  3:58 ` [PATCH v2 07/22] jbd2: drop jbd2_fc_init documentation Harshad Shirwadkar
2020-11-06  3:58 ` Harshad Shirwadkar [this message]
2020-11-06  3:58 ` [PATCH v2 09/22] jbd2: don't pass tid to jbd2_fc_end_commit_fallback() Harshad Shirwadkar
2020-11-06  3:58 ` [PATCH v2 10/22] jbd2: add todo for a fast commit performance optimization Harshad Shirwadkar
2020-11-06  3:59 ` [PATCH v2 11/22] jbd2: don't touch buffer state until it is filled Harshad Shirwadkar
2020-11-06  3:59 ` [PATCH v2 12/22] jbd2: don't read journal->j_commit_sequence without taking a lock Harshad Shirwadkar
2020-11-06  3:59 ` [PATCH v2 13/22] ext4: dedpulicate the code to wait on inode that's being committed Harshad Shirwadkar
2020-11-06  3:59 ` [PATCH v2 14/22] ext4: fix code documentatioon Harshad Shirwadkar
2020-11-06  3:59 ` [PATCH v2 15/22] ext4: mark buf dirty before submitting fast commit buffer Harshad Shirwadkar
2020-11-06  3:59 ` [PATCH v2 16/22] ext4: remove unnecessary fast commit calls from ext4_file_mmap Harshad Shirwadkar
2020-11-06  3:59 ` [PATCH v2 17/22] ext4: fix inode dirty check in case of fast commits Harshad Shirwadkar
2020-11-06  3:59 ` [PATCH v2 18/22] ext4: disable fast commit with data journalling Harshad Shirwadkar
2020-11-06  3:59 ` [PATCH v2 19/22] ext4: issue fsdev cache flush before starting fast commit Harshad Shirwadkar
2020-11-06  3:59 ` [PATCH v2 20/22] ext4: make s_mount_flags modifications atomic Harshad Shirwadkar
2020-11-06  3:59 ` [PATCH v2 21/22] jbd2: don't start fast commit on aborted journal Harshad Shirwadkar
2020-11-06  3:59 ` [PATCH v2 22/22] ext4: cleanup fast commit mount options Harshad Shirwadkar

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=20201106035911.1942128-9-harshadshirwadkar@gmail.com \
    --to=harshadshirwadkar@gmail.com \
    --cc=jack@suse.cz \
    --cc=linux-ext4@vger.kernel.org \
    --cc=tytso@mit.edu \
    /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).