All of lore.kernel.org
 help / color / mirror / Atom feed
From: Mathieu Othacehe <othacehe@gnu.org>
To: Theodore Ts'o <tytso@mit.edu>, Andreas Dilger <adilger.kernel@dilger.ca>
Cc: linux-ext4@vger.kernel.org, linux-kernel@vger.kernel.org,
	lukas.skupinski@landisgyr.com, anton.reding@landisgyr.com,
	Mathieu Othacehe <othacehe@gnu.org>
Subject: [PATCH 1/1] ext4: Prevent an infinite loop in the lazyinit thread.
Date: Wed,  6 Nov 2024 14:47:41 +0100	[thread overview]
Message-ID: <20241106134741.26948-2-othacehe@gnu.org> (raw)
In-Reply-To: <20241106134741.26948-1-othacehe@gnu.org>

Use ktime_get_ns instead of ktime_get_real_ns when computing the lr_timeout
not to be affected by system time jumps.

Use a boolean instead of the MAX_JIFFY_OFFSET value to determine whether
the next_wakeup value has been set. Comparing elr->lr_next_sched to
MAX_JIFFY_OFFSET can cause the lazyinit thread to loop indefinitely.

Co-developed-by: Lukas Skupinski <lukas.skupinski@landisgyr.com>
Signed-off-by: Lukas Skupinski <lukas.skupinski@landisgyr.com>
Signed-off-by: Mathieu Othacehe <othacehe@gnu.org>
---
 fs/ext4/super.c | 20 +++++++++++++-------
 1 file changed, 13 insertions(+), 7 deletions(-)

diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 9fcf44064c6a6..b4839ccd83ad5 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -3712,12 +3712,12 @@ static int ext4_run_li_request(struct ext4_li_request *elr)
 		ret = 1;
 
 	if (!ret) {
-		start_time = ktime_get_real_ns();
+		start_time = ktime_get_ns();
 		ret = ext4_init_inode_table(sb, group,
 					    elr->lr_timeout ? 0 : 1);
 		trace_ext4_lazy_itable_init(sb, group);
 		if (elr->lr_timeout == 0) {
-			elr->lr_timeout = nsecs_to_jiffies((ktime_get_real_ns() - start_time) *
+			elr->lr_timeout = nsecs_to_jiffies((ktime_get_ns() - start_time) *
 				EXT4_SB(elr->lr_super)->s_li_wait_mult);
 		}
 		elr->lr_next_sched = jiffies + elr->lr_timeout;
@@ -3777,8 +3777,9 @@ static int ext4_lazyinit_thread(void *arg)
 
 cont_thread:
 	while (true) {
-		next_wakeup = MAX_JIFFY_OFFSET;
+		bool next_wakeup_initialized = false;
 
+		next_wakeup = 0;
 		mutex_lock(&eli->li_list_mtx);
 		if (list_empty(&eli->li_request_list)) {
 			mutex_unlock(&eli->li_list_mtx);
@@ -3791,8 +3792,11 @@ static int ext4_lazyinit_thread(void *arg)
 					 lr_request);
 
 			if (time_before(jiffies, elr->lr_next_sched)) {
-				if (time_before(elr->lr_next_sched, next_wakeup))
+				if (!next_wakeup_initialized ||
+				    time_before(elr->lr_next_sched, next_wakeup)) {
 					next_wakeup = elr->lr_next_sched;
+					next_wakeup_initialized = true;
+				}
 				continue;
 			}
 			if (down_read_trylock(&elr->lr_super->s_umount)) {
@@ -3820,16 +3824,18 @@ static int ext4_lazyinit_thread(void *arg)
 				elr->lr_next_sched = jiffies +
 					get_random_u32_below(EXT4_DEF_LI_MAX_START_DELAY * HZ);
 			}
-			if (time_before(elr->lr_next_sched, next_wakeup))
+			if (!next_wakeup_initialized ||
+			    time_before(elr->lr_next_sched, next_wakeup)) {
 				next_wakeup = elr->lr_next_sched;
+				next_wakeup_initialized = true;
+			}
 		}
 		mutex_unlock(&eli->li_list_mtx);
 
 		try_to_freeze();
 
 		cur = jiffies;
-		if ((time_after_eq(cur, next_wakeup)) ||
-		    (MAX_JIFFY_OFFSET == next_wakeup)) {
+		if (!next_wakeup_initialized || time_after_eq(cur, next_wakeup)) {
 			cond_resched();
 			continue;
 		}
-- 
2.46.0


  reply	other threads:[~2024-11-06 13:52 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-11-06 13:47 [PATCH 0/1] ext4: Prevent an infinite loop in the lazyinit thread Mathieu Othacehe
2024-11-06 13:47 ` Mathieu Othacehe [this message]
2024-11-08 10:49   ` [PATCH 1/1] " Jan Kara
2024-11-14 13:53   ` Theodore Ts'o

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=20241106134741.26948-2-othacehe@gnu.org \
    --to=othacehe@gnu.org \
    --cc=adilger.kernel@dilger.ca \
    --cc=anton.reding@landisgyr.com \
    --cc=linux-ext4@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=lukas.skupinski@landisgyr.com \
    --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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.