* [PATCH 13/34] ext4: Convert ext4_fc_submit_bh() to bh_submit()
From: Matthew Wilcox (Oracle) @ 2026-05-25 17:19 UTC (permalink / raw)
To: Jan Kara
Cc: Matthew Wilcox (Oracle), Christian Brauner, Christoph Hellwig,
linux-fsdevel, linux-ext4
In-Reply-To: <20260525171931.4144395-1-willy@infradead.org>
Avoid an extra indirect function call by converting
ext4_end_buffer_io_sync() from bh_end_io_t to bio_end_io_t and
calling bh_submit().
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: linux-ext4@vger.kernel.org
---
fs/ext4/fast_commit.c | 8 +++++---
1 file changed, 5 insertions(+), 3 deletions(-)
diff --git a/fs/ext4/fast_commit.c b/fs/ext4/fast_commit.c
index b3c22636251d..d52c64adf416 100644
--- a/fs/ext4/fast_commit.c
+++ b/fs/ext4/fast_commit.c
@@ -184,8 +184,11 @@
#include <trace/events/ext4.h>
static struct kmem_cache *ext4_fc_dentry_cachep;
-static void ext4_end_buffer_io_sync(struct buffer_head *bh, int uptodate)
+static void ext4_end_buffer_io_sync(struct bio *bio)
{
+ bool uptodate = bio->bi_status == BLK_STS_OK;
+ struct buffer_head *bh = bio_endio_bh(bio);
+
BUFFER_TRACE(bh, "");
if (uptodate) {
ext4_debug("%s: Block %lld up-to-date",
@@ -659,8 +662,7 @@ static void ext4_fc_submit_bh(struct super_block *sb, bool is_tail)
lock_buffer(bh);
set_buffer_dirty(bh);
set_buffer_uptodate(bh);
- bh->b_end_io = ext4_end_buffer_io_sync;
- submit_bh(REQ_OP_WRITE | write_flags, bh);
+ bh_submit(bh, REQ_OP_WRITE | write_flags, ext4_end_buffer_io_sync);
EXT4_SB(sb)->s_fc_bh = NULL;
}
--
2.47.3
^ permalink raw reply related
* [PATCH 15/34] ext4: Convert ext4_commit_super() to bh_submit()
From: Matthew Wilcox (Oracle) @ 2026-05-25 17:19 UTC (permalink / raw)
To: Jan Kara
Cc: Matthew Wilcox (Oracle), Christian Brauner, Christoph Hellwig,
linux-fsdevel, linux-ext4
In-Reply-To: <20260525171931.4144395-1-willy@infradead.org>
Avoid an extra indirect function call by using bh_submit() instead of
submit_bh().
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: linux-ext4@vger.kernel.org
---
fs/ext4/super.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index fbe175951e01..905d66cbe3f2 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -6320,9 +6320,8 @@ static int ext4_commit_super(struct super_block *sb)
get_bh(sbh);
/* Clear potential dirty bit if it was journalled update */
clear_buffer_dirty(sbh);
- sbh->b_end_io = end_buffer_write_sync;
- submit_bh(REQ_OP_WRITE | REQ_SYNC |
- (test_opt(sb, BARRIER) ? REQ_FUA : 0), sbh);
+ bh_submit(sbh, REQ_OP_WRITE | REQ_SYNC |
+ (test_opt(sb, BARRIER) ? REQ_FUA : 0), bh_end_write);
wait_on_buffer(sbh);
if (buffer_write_io_error(sbh)) {
ext4_msg(sb, KERN_ERR, "I/O error while writing "
--
2.47.3
^ permalink raw reply related
* [PATCH 14/34] ext4: Convert write_mmp_block_thawed() to bh_submit()
From: Matthew Wilcox (Oracle) @ 2026-05-25 17:19 UTC (permalink / raw)
To: Jan Kara
Cc: Matthew Wilcox (Oracle), Christian Brauner, Christoph Hellwig,
linux-fsdevel, linux-ext4
In-Reply-To: <20260525171931.4144395-1-willy@infradead.org>
Avoid an extra indirect function call by using bh_submit() instead of
submit_bh().
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: linux-ext4@vger.kernel.org
---
fs/ext4/mmp.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/fs/ext4/mmp.c b/fs/ext4/mmp.c
index 6f57c181ff77..493528fbed75 100644
--- a/fs/ext4/mmp.c
+++ b/fs/ext4/mmp.c
@@ -46,9 +46,9 @@ static int write_mmp_block_thawed(struct super_block *sb,
ext4_mmp_csum_set(sb, mmp);
lock_buffer(bh);
- bh->b_end_io = end_buffer_write_sync;
get_bh(bh);
- submit_bh(REQ_OP_WRITE | REQ_SYNC | REQ_META | REQ_PRIO, bh);
+ bh_submit(bh, REQ_OP_WRITE | REQ_SYNC | REQ_META | REQ_PRIO,
+ bh_end_write);
wait_on_buffer(bh);
if (unlikely(!buffer_uptodate(bh)))
return -EIO;
--
2.47.3
^ permalink raw reply related
* [PATCH 12/34] ext4; Convert __ext4_read_bh() to bh_submit()
From: Matthew Wilcox (Oracle) @ 2026-05-25 17:19 UTC (permalink / raw)
To: Jan Kara
Cc: Matthew Wilcox (Oracle), Christian Brauner, Christoph Hellwig,
linux-fsdevel, linux-ext4
In-Reply-To: <20260525171931.4144395-1-willy@infradead.org>
Avoid an extra indirect function call by converting
ext4_end_bitmap_read() from bh_end_io_t to bio_end_io_t and
calling bh_submit().
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: linux-ext4@vger.kernel.org
---
fs/ext4/ext4.h | 10 +++++-----
fs/ext4/ialloc.c | 5 ++++-
fs/ext4/super.c | 11 ++++++-----
3 files changed, 15 insertions(+), 11 deletions(-)
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 94283a991e5c..6af11f0ff1c5 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -2959,7 +2959,7 @@ extern unsigned long ext4_count_dirs(struct super_block *);
extern void ext4_mark_bitmap_end(int start_bit, int end_bit, char *bitmap);
extern int ext4_init_inode_table(struct super_block *sb,
ext4_group_t group, int barrier);
-extern void ext4_end_bitmap_read(struct buffer_head *bh, int uptodate);
+void ext4_end_bitmap_read(struct bio *bio);
/* fast_commit.c */
int ext4_fc_info_show(struct seq_file *seq, void *v);
@@ -3184,10 +3184,10 @@ extern struct buffer_head *ext4_sb_bread_unmovable(struct super_block *sb,
sector_t block);
extern struct buffer_head *ext4_sb_bread_nofail(struct super_block *sb,
sector_t block);
-extern void ext4_read_bh_nowait(struct buffer_head *bh, blk_opf_t op_flags,
- bh_end_io_t *end_io, bool simu_fail);
-extern int ext4_read_bh(struct buffer_head *bh, blk_opf_t op_flags,
- bh_end_io_t *end_io, bool simu_fail);
+void ext4_read_bh_nowait(struct buffer_head *bh, blk_opf_t op_flags,
+ bio_end_io_t end_io, bool simu_fail);
+int ext4_read_bh(struct buffer_head *bh, blk_opf_t op_flags,
+ bio_end_io_t end_io, bool simu_fail);
extern int ext4_read_bh_lock(struct buffer_head *bh, blk_opf_t op_flags, bool wait);
extern void ext4_sb_breadahead_unmovable(struct super_block *sb, sector_t block);
extern int ext4_seq_options_show(struct seq_file *seq, void *offset);
diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c
index 3fd8f0099852..2db68b1bf855 100644
--- a/fs/ext4/ialloc.c
+++ b/fs/ext4/ialloc.c
@@ -66,8 +66,11 @@ void ext4_mark_bitmap_end(int start_bit, int end_bit, char *bitmap)
memset(bitmap + (i >> 3), 0xff, (end_bit - i) >> 3);
}
-void ext4_end_bitmap_read(struct buffer_head *bh, int uptodate)
+void ext4_end_bitmap_read(struct bio *bio)
{
+ bool uptodate = bio->bi_status == BLK_STS_OK;
+ struct buffer_head *bh = bio_endio_bh(bio);
+
if (uptodate) {
set_buffer_uptodate(bh);
set_bitmap_uptodate(bh);
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 6a77db4d3124..fbe175951e01 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -161,7 +161,7 @@ MODULE_ALIAS("ext3");
static inline void __ext4_read_bh(struct buffer_head *bh, blk_opf_t op_flags,
- bh_end_io_t *end_io, bool simu_fail)
+ bio_end_io_t end_io, bool simu_fail)
{
if (simu_fail) {
clear_buffer_uptodate(bh);
@@ -176,13 +176,14 @@ static inline void __ext4_read_bh(struct buffer_head *bh, blk_opf_t op_flags,
*/
clear_buffer_verified(bh);
- bh->b_end_io = end_io ? end_io : end_buffer_read_sync;
+ if (!end_io)
+ end_io = bh_end_read;
get_bh(bh);
- submit_bh(REQ_OP_READ | op_flags, bh);
+ bh_submit(bh, REQ_OP_READ | op_flags, end_io);
}
void ext4_read_bh_nowait(struct buffer_head *bh, blk_opf_t op_flags,
- bh_end_io_t *end_io, bool simu_fail)
+ bio_end_io_t end_io, bool simu_fail)
{
BUG_ON(!buffer_locked(bh));
@@ -194,7 +195,7 @@ void ext4_read_bh_nowait(struct buffer_head *bh, blk_opf_t op_flags,
}
int ext4_read_bh(struct buffer_head *bh, blk_opf_t op_flags,
- bh_end_io_t *end_io, bool simu_fail)
+ bio_end_io_t end_io, bool simu_fail)
{
BUG_ON(!buffer_locked(bh));
--
2.47.3
^ permalink raw reply related
* Re: [PATCH 04/17] nilfs2: replace get_zeroed_page() with kzalloc()
From: Viacheslav Dubeyko @ 2026-05-25 17:07 UTC (permalink / raw)
To: Mike Rapoport (Microsoft), Jan Kara, Mark Fasheh, Joel Becker,
Joseph Qi, Ryusuke Konishi, Viacheslav Dubeyko, Trond Myklebust,
Anna Schumaker, Chuck Lever, Jeff Layton, NeilBrown,
Olga Kornievskaia, Dai Ngo, Tom Talpey, Alexander Viro,
Christian Brauner, Jan Kara, Dave Kleikamp, Theodore Ts'o,
Miklos Szeredi, Andreas Hindborg, Breno Leitao, Kees Cook,
Tigran A. Aivazian
Cc: linux-kernel, linux-fsdevel, ocfs2-devel, linux-nilfs, linux-nfs,
jfs-discussion, linux-ext4, linux-mm
In-Reply-To: <20260523-b4-fs-v1-4-275e36a83f0e@kernel.org>
On Sat, 2026-05-23 at 20:54 +0300, Mike Rapoport (Microsoft) wrote:
> nilfs_ioctl_wrap_copy() allocates a temporary buffer with
> get_zeroed_page().
>
> kzalloc() is a better API for such use and it also provides better
> scalability and more debugging possibilities.
>
> Replace use of get_zeroed_page() with kzalloc().
>
> Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
> ---
> fs/nilfs2/ioctl.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/fs/nilfs2/ioctl.c b/fs/nilfs2/ioctl.c
> index e0a606643e87..b73f2c5d10f0 100644
> --- a/fs/nilfs2/ioctl.c
> +++ b/fs/nilfs2/ioctl.c
> @@ -69,7 +69,7 @@ static int nilfs_ioctl_wrap_copy(struct the_nilfs *nilfs,
> if (argv->v_index > ~(__u64)0 - argv->v_nmembs)
> return -EINVAL;
>
> - buf = (void *)get_zeroed_page(GFP_NOFS);
> + buf = kzalloc(PAGE_SIZE, GFP_NOFS);
> if (unlikely(!buf))
> return -ENOMEM;
> maxmembs = PAGE_SIZE / argv->v_size;
> @@ -107,7 +107,7 @@ static int nilfs_ioctl_wrap_copy(struct the_nilfs *nilfs,
> }
> argv->v_nmembs = total;
>
> - free_pages((unsigned long)buf, 0);
> + kfree(buf);
> return ret;
> }
>
Makes sense to me.
Reviewed-by: Viacheslav Dubeyko <slava@dubeyko.com>
Thanks,
Slava.
^ permalink raw reply
* Re: [PATCH v2] ext2: Remove deprecated DAX support
From: Jan Kara @ 2026-05-25 16:30 UTC (permalink / raw)
To: Ashwin Gundarapu; +Cc: jack, linux-ext4, linux-kernel
In-Reply-To: <19e587e5143.434d5026117452.233431645221906532@zohomail.in>
On Sun 24-05-26 11:08:53, Ashwin Gundarapu wrote:
>
> DAX support in ext2 was deprecated in commit d5a2693f93e4
> ("ext2: Deprecate DAX") with a removal deadline of end of 2025.
> Remove all DAX code from ext2 as scheduled.
>
> This removes the DAX mount option, IOMAP DAX support, DAX file
> operations, DAX address_space_operations, and the DAX fault handler.
>
> Signed-off-by: Ashwin Gundarapu <linuxuser509@zohomail.in>
> ---
> v2: Removed unused sbi variable and fixed indentation as reported
> by kernel test robot.
Thanks for the patch. Some style nits below.
> static ssize_t ext2_file_read_iter(struct kiocb *iocb, struct iov_iter *to)
> {
> -#ifdef CONFIG_FS_DAX
> - if (IS_DAX(iocb->ki_filp->f_mapping->host))
> - return ext2_dax_read_iter(iocb, to);
> -#endif
> +
Stray empty line here.
> if (iocb->ki_flags & IOCB_DIRECT)
> return ext2_dio_read_iter(iocb, to);
>
> @@ -297,10 +188,7 @@ static ssize_t ext2_file_read_iter(struct kiocb *iocb, struct iov_iter *to)
>
> static ssize_t ext2_file_write_iter(struct kiocb *iocb, struct iov_iter *from)
> {
> -#ifdef CONFIG_FS_DAX
> - if (IS_DAX(iocb->ki_filp->f_mapping->host))
> - return ext2_dax_write_iter(iocb, from);
> -#endif
> +
... and here.
> if (iocb->ki_flags & IOCB_DIRECT)
> return ext2_dio_write_iter(iocb, from);
>
> @@ -321,7 +209,7 @@ const struct file_operations ext2_file_operations = {
> #ifdef CONFIG_COMPAT
> .compat_ioctl = ext2_compat_ioctl,
> #endif
> - .mmap_prepare = ext2_file_mmap_prepare,
> + .mmap_prepare = generic_file_mmap_prepare,
Please indent this with tab the same way as other methods.
> @@ -841,10 +818,7 @@ static int ext2_iomap_begin(struct inode *inode, loff_t offset, loff_t length,
>
> iomap->flags = 0;
> iomap->offset = (u64)first_block << blkbits;
> - if (flags & IOMAP_DAX)
> - iomap->dax_dev = sbi->s_daxdev;
> - else
> - iomap->bdev = inode->i_sb->s_bdev;
> + iomap->bdev = inode->i_sb->s_bdev;
Indented with spaces instead of tabs.
> @@ -1290,12 +1248,8 @@ static int ext2_setsize(struct inode *inode, loff_t newsize)
>
> inode_dio_wait(inode);
>
> - if (IS_DAX(inode))
> - error = dax_truncate_page(inode, newsize, NULL,
> - &ext2_iomap_ops);
> - else
> - error = block_truncate_page(inode->i_mapping,
> - newsize, ext2_get_block);
> + error = block_truncate_page(inode->i_mapping,
> + newsize, ext2_get_block);
Indented with spaces instead of tabs.
> if (error)
> return error;
>
...
> + case Opt_xip:
> + ext2_msg_fc(fc, KERN_ERR, "DAX support has been removed. Please use ext4 instead.");
> + return -EINVAL;
Indented with spaces instead of tabs.
> @@ -992,16 +974,8 @@ static int ext2_fill_super(struct super_block *sb, struct fs_context *fc)
> }
> blocksize = BLOCK_SIZE << le32_to_cpu(sbi->s_es->s_log_block_size);
>
> - if (test_opt(sb, DAX)) {
> - if (!sbi->s_daxdev) {
> - ext2_msg(sb, KERN_ERR,
> - "DAX unsupported by block device. Turning off DAX.");
> - clear_opt(sbi->s_mount_opt, DAX);
> - } else if (blocksize != PAGE_SIZE) {
> - ext2_msg(sb, KERN_ERR, "unsupported blocksize for DAX\n");
> - clear_opt(sbi->s_mount_opt, DAX);
> - }
> - }
> +
> +
Stray empty lines.
Honza
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply
* Re: [PATCH 15/17] configfs: replace __get_free_pages() with kzalloc()
From: Jan Kara @ 2026-05-25 16:22 UTC (permalink / raw)
To: Mike Rapoport (Microsoft)
Cc: Jan Kara, Mark Fasheh, Joel Becker, Joseph Qi, Ryusuke Konishi,
Viacheslav Dubeyko, Trond Myklebust, Anna Schumaker, Chuck Lever,
Jeff Layton, NeilBrown, Olga Kornievskaia, Dai Ngo, Tom Talpey,
Alexander Viro, Christian Brauner, Jan Kara, Dave Kleikamp,
Theodore Ts'o, Miklos Szeredi, Andreas Hindborg, Breno Leitao,
Kees Cook, Tigran A. Aivazian, linux-kernel, linux-fsdevel,
ocfs2-devel, linux-nilfs, linux-nfs, jfs-discussion, linux-ext4,
linux-mm
In-Reply-To: <20260523-b4-fs-v1-15-275e36a83f0e@kernel.org>
On Sat 23-05-26 20:54:27, Mike Rapoport (Microsoft) wrote:
> configfs allocates staging buffers __get_free_pages().
>
> kmalloc() is a better API for such use and it also provides better
> scalability and more debugging possibilities.
>
> Replace use of __get_free_pages() with kzalloc().
>
> Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Looks good. Feel free to add:
Reviewed-by: Jan Kara <jack@suse.cz>
Honza
> ---
> fs/configfs/file.c | 7 +++----
> 1 file changed, 3 insertions(+), 4 deletions(-)
>
> diff --git a/fs/configfs/file.c b/fs/configfs/file.c
> index ef8c3cd10cc6..a48cece775a3 100644
> --- a/fs/configfs/file.c
> +++ b/fs/configfs/file.c
> @@ -59,7 +59,7 @@ static int fill_read_buffer(struct file *file, struct configfs_buffer *buffer)
> ssize_t count = -ENOENT;
>
> if (!buffer->page)
> - buffer->page = (char *) get_zeroed_page(GFP_KERNEL);
> + buffer->page = kzalloc(PAGE_SIZE, GFP_KERNEL);
> if (!buffer->page)
> return -ENOMEM;
>
> @@ -184,7 +184,7 @@ static int fill_write_buffer(struct configfs_buffer *buffer,
> int copied;
>
> if (!buffer->page)
> - buffer->page = (char *)__get_free_pages(GFP_KERNEL, 0);
> + buffer->page = kmalloc(PAGE_SIZE, GFP_KERNEL);
> if (!buffer->page)
> return -ENOMEM;
>
> @@ -381,8 +381,7 @@ static int configfs_release(struct inode *inode, struct file *filp)
> struct configfs_buffer *buffer = filp->private_data;
>
> module_put(buffer->owner);
> - if (buffer->page)
> - free_page((unsigned long)buffer->page);
> + kfree(buffer->page);
> mutex_destroy(&buffer->mutex);
> kfree(buffer);
> return 0;
>
> --
> 2.53.0
>
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply
* Re: [PATCH 14/17] fs/namespace: use __getname() to allocate mntpath buffer
From: Jan Kara @ 2026-05-25 16:22 UTC (permalink / raw)
To: Mike Rapoport (Microsoft)
Cc: Jan Kara, Mark Fasheh, Joel Becker, Joseph Qi, Ryusuke Konishi,
Viacheslav Dubeyko, Trond Myklebust, Anna Schumaker, Chuck Lever,
Jeff Layton, NeilBrown, Olga Kornievskaia, Dai Ngo, Tom Talpey,
Alexander Viro, Christian Brauner, Jan Kara, Dave Kleikamp,
Theodore Ts'o, Miklos Szeredi, Andreas Hindborg, Breno Leitao,
Kees Cook, Tigran A. Aivazian, linux-kernel, linux-fsdevel,
ocfs2-devel, linux-nilfs, linux-nfs, jfs-discussion, linux-ext4,
linux-mm
In-Reply-To: <20260523-b4-fs-v1-14-275e36a83f0e@kernel.org>
On Sat 23-05-26 20:54:26, Mike Rapoport (Microsoft) wrote:
> mnt_warn_timestamp_expiry() allocates memory for a path with
> __get_free_page() although there is a dedicated helper for allocation of
> file paths: __getname().
>
> Replace __get_free_page() for allocation of a path buffer with __getname().
>
> Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
> ---
> fs/namespace.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/fs/namespace.c b/fs/namespace.c
> index fe919abd2f01..2ed9cd846a81 100644
> --- a/fs/namespace.c
> +++ b/fs/namespace.c
> @@ -3303,7 +3303,7 @@ static void mnt_warn_timestamp_expiry(const struct path *mountpoint,
> (ktime_get_real_seconds() + TIME_UPTIME_SEC_MAX > sb->s_time_max)) {
> char *buf, *mntpath;
>
> - buf = (char *)__get_free_page(GFP_KERNEL);
> + buf = __getname();
Fair but d_path() below should then get PATH_MAX and not PAGE_SIZE.
> if (buf)
> mntpath = d_path(mountpoint, buf, PAGE_SIZE);
> else
> @@ -3319,7 +3319,7 @@ static void mnt_warn_timestamp_expiry(const struct path *mountpoint,
>
> sb->s_iflags |= SB_I_TS_EXPIRY_WARNED;
> if (buf)
> - free_page((unsigned long)buf);
> + __putname(buf);
And __putname() is fine with NULL so no need for the if (buf) check here.
Honza
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply
* Re: [PATCH 13/17] fs/select: replace __get_free_page() with kmalloc()
From: Jan Kara @ 2026-05-25 16:19 UTC (permalink / raw)
To: Mike Rapoport (Microsoft)
Cc: Jan Kara, Mark Fasheh, Joel Becker, Joseph Qi, Ryusuke Konishi,
Viacheslav Dubeyko, Trond Myklebust, Anna Schumaker, Chuck Lever,
Jeff Layton, NeilBrown, Olga Kornievskaia, Dai Ngo, Tom Talpey,
Alexander Viro, Christian Brauner, Jan Kara, Dave Kleikamp,
Theodore Ts'o, Miklos Szeredi, Andreas Hindborg, Breno Leitao,
Kees Cook, Tigran A. Aivazian, linux-kernel, linux-fsdevel,
ocfs2-devel, linux-nilfs, linux-nfs, jfs-discussion, linux-ext4,
linux-mm
In-Reply-To: <20260523-b4-fs-v1-13-275e36a83f0e@kernel.org>
On Sat 23-05-26 20:54:25, Mike Rapoport (Microsoft) wrote:
> poll_get_entry() allocates new memory for poll_table entries using
> __get_free_page().
>
> kmalloc() is a better API for such use and it also provides better
> scalability and more debugging possibilities.
>
> Replace use of __get_free_page() with kmalloc().
>
> Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Looks good. Feel free to add:
Reviewed-by: Jan Kara <jack@suse.cz>
Honza
> ---
> fs/select.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/fs/select.c b/fs/select.c
> index 75978b18f48f..6fa63e48cdee 100644
> --- a/fs/select.c
> +++ b/fs/select.c
> @@ -150,7 +150,7 @@ void poll_freewait(struct poll_wqueues *pwq)
> } while (entry > p->entries);
> old = p;
> p = p->next;
> - free_page((unsigned long) old);
> + kfree(old);
> }
> }
> EXPORT_SYMBOL(poll_freewait);
> @@ -165,7 +165,7 @@ static struct poll_table_entry *poll_get_entry(struct poll_wqueues *p)
> if (!table || POLL_TABLE_FULL(table)) {
> struct poll_table_page *new_table;
>
> - new_table = (struct poll_table_page *) __get_free_page(GFP_KERNEL);
> + new_table = kmalloc(PAGE_SIZE, GFP_KERNEL);
> if (!new_table) {
> p->error = -ENOMEM;
> return NULL;
>
> --
> 2.53.0
>
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply
* Re: [PATCH 11/17] isofs: replace __get_free_page() with kmalloc()
From: Jan Kara @ 2026-05-25 16:17 UTC (permalink / raw)
To: Mike Rapoport (Microsoft)
Cc: Jan Kara, Mark Fasheh, Joel Becker, Joseph Qi, Ryusuke Konishi,
Viacheslav Dubeyko, Trond Myklebust, Anna Schumaker, Chuck Lever,
Jeff Layton, NeilBrown, Olga Kornievskaia, Dai Ngo, Tom Talpey,
Alexander Viro, Christian Brauner, Jan Kara, Dave Kleikamp,
Theodore Ts'o, Miklos Szeredi, Andreas Hindborg, Breno Leitao,
Kees Cook, Tigran A. Aivazian, linux-kernel, linux-fsdevel,
ocfs2-devel, linux-nilfs, linux-nfs, jfs-discussion, linux-ext4,
linux-mm
In-Reply-To: <20260523-b4-fs-v1-11-275e36a83f0e@kernel.org>
On Sat 23-05-26 20:54:23, Mike Rapoport (Microsoft) wrote:
> isofs_readdir() allocates a temporary buffer with __get_free_page().
>
> kmalloc() is a better API for such use and it also provides better
> scalability and more debugging possibilities.
>
> Replace use of __get_free_page() with kmalloc().
>
> Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Thanks. Added to my tree.
Honza
> ---
> fs/isofs/dir.c | 5 +++--
> 1 file changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/fs/isofs/dir.c b/fs/isofs/dir.c
> index 2fd9948d606e..6d220eab531e 100644
> --- a/fs/isofs/dir.c
> +++ b/fs/isofs/dir.c
> @@ -13,6 +13,7 @@
> */
> #include <linux/gfp.h>
> #include <linux/filelock.h>
> +#include <linux/slab.h>
> #include "isofs.h"
>
> int isofs_name_translate(struct iso_directory_record *de, char *new, struct inode *inode)
> @@ -255,7 +256,7 @@ static int isofs_readdir(struct file *file, struct dir_context *ctx)
> struct iso_directory_record *tmpde;
> struct inode *inode = file_inode(file);
>
> - tmpname = (char *)__get_free_page(GFP_KERNEL);
> + tmpname = kmalloc(PAGE_SIZE, GFP_KERNEL);
> if (tmpname == NULL)
> return -ENOMEM;
>
> @@ -263,7 +264,7 @@ static int isofs_readdir(struct file *file, struct dir_context *ctx)
>
> result = do_isofs_readdir(inode, file, ctx, tmpname, tmpde);
>
> - free_page((unsigned long) tmpname);
> + kfree(tmpname);
> return result;
> }
>
>
> --
> 2.53.0
>
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply
* Re: [PATCH 10/17] jbd2: replace __get_free_pages() with kmalloc()
From: Jan Kara @ 2026-05-25 16:17 UTC (permalink / raw)
To: Mike Rapoport (Microsoft)
Cc: Jan Kara, Mark Fasheh, Joel Becker, Joseph Qi, Ryusuke Konishi,
Viacheslav Dubeyko, Trond Myklebust, Anna Schumaker, Chuck Lever,
Jeff Layton, NeilBrown, Olga Kornievskaia, Dai Ngo, Tom Talpey,
Alexander Viro, Christian Brauner, Jan Kara, Dave Kleikamp,
Theodore Ts'o, Miklos Szeredi, Andreas Hindborg, Breno Leitao,
Kees Cook, Tigran A. Aivazian, linux-kernel, linux-fsdevel,
ocfs2-devel, linux-nilfs, linux-nfs, jfs-discussion, linux-ext4,
linux-mm
In-Reply-To: <20260523-b4-fs-v1-10-275e36a83f0e@kernel.org>
On Sat 23-05-26 20:54:22, Mike Rapoport (Microsoft) wrote:
> jbd2_alloc() falls back from kmem_cache_alloc() to __get_free_pages() for
> allocations larger than PAGE_SIZE.
> But kmalloc() can handle such cases with essentially the same fallback.
>
> Replace use of __get_free_pages() with kmalloc() and simplify
> jbd2_free() as both kmem_cache_alloc() and kmalloc() allocations can be
> freed with kfree().
>
> Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Looks good. Feel free to add:
Reviewed-by: Jan Kara <jack@suse.cz>
I'll just note that we allocate here fs block size large buffer so the same
kind of allocator as we use for folios would be even better. But that's a
different cleanup I guess.
Honza
> ---
> fs/jbd2/journal.c | 7 ++-----
> 1 file changed, 2 insertions(+), 5 deletions(-)
>
> diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
> index 4f397fcdb13c..1137b471e490 100644
> --- a/fs/jbd2/journal.c
> +++ b/fs/jbd2/journal.c
> @@ -2784,7 +2784,7 @@ void *jbd2_alloc(size_t size, gfp_t flags)
> if (size < PAGE_SIZE)
> ptr = kmem_cache_alloc(get_slab(size), flags);
> else
> - ptr = (void *)__get_free_pages(flags, get_order(size));
> + ptr = kmalloc(size, flags);
>
> /* Check alignment; SLUB has gotten this wrong in the past,
> * and this can lead to user data corruption! */
> @@ -2795,10 +2795,7 @@ void *jbd2_alloc(size_t size, gfp_t flags)
>
> void jbd2_free(void *ptr, size_t size)
> {
> - if (size < PAGE_SIZE)
> - kmem_cache_free(get_slab(size), ptr);
> - else
> - free_pages((unsigned long)ptr, get_order(size));
> + kfree(ptr);
> };
>
> /*
>
> --
> 2.53.0
>
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply
* Re: [PATCH 03/17] ocfs2/dlm: replace __get_free_page() with kmalloc()
From: Jan Kara @ 2026-05-25 16:13 UTC (permalink / raw)
To: Mike Rapoport (Microsoft)
Cc: Jan Kara, Mark Fasheh, Joel Becker, Joseph Qi, Ryusuke Konishi,
Viacheslav Dubeyko, Trond Myklebust, Anna Schumaker, Chuck Lever,
Jeff Layton, NeilBrown, Olga Kornievskaia, Dai Ngo, Tom Talpey,
Alexander Viro, Christian Brauner, Jan Kara, Dave Kleikamp,
Theodore Ts'o, Miklos Szeredi, Andreas Hindborg, Breno Leitao,
Kees Cook, Tigran A. Aivazian, linux-kernel, linux-fsdevel,
ocfs2-devel, linux-nilfs, linux-nfs, jfs-discussion, linux-ext4,
linux-mm
In-Reply-To: <20260523-b4-fs-v1-3-275e36a83f0e@kernel.org>
On Sat 23-05-26 20:54:15, Mike Rapoport (Microsoft) wrote:
> A few places in ocsfs2 allocate temporary buffers with __get_free_page() or
> get_zeroed_page().
>
> kmalloc() is a better API for such use and it also provides better
> scalability and more debugging possibilities.
>
> Replace use of __get_free_page() and get_zeroed_page() with kmalloc() and
> kzalloc() respectively.
>
> Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Looks good. Feel free to add:
Reviewed-by: Jan Kara <jack@suse.cz>
Honza
> ---
> fs/ocfs2/dlm/dlmdebug.c | 24 +++++++++---------------
> fs/ocfs2/dlm/dlmdomain.c | 8 +++++---
> fs/ocfs2/dlm/dlmmaster.c | 5 ++---
> fs/ocfs2/dlm/dlmrecovery.c | 4 ++--
> 4 files changed, 18 insertions(+), 23 deletions(-)
>
> diff --git a/fs/ocfs2/dlm/dlmdebug.c b/fs/ocfs2/dlm/dlmdebug.c
> index fe4fdd09bae3..6ca8b3b68eef 100644
> --- a/fs/ocfs2/dlm/dlmdebug.c
> +++ b/fs/ocfs2/dlm/dlmdebug.c
> @@ -260,10 +260,10 @@ void dlm_print_one_mle(struct dlm_master_list_entry *mle)
> {
> char *buf;
>
> - buf = (char *) get_zeroed_page(GFP_ATOMIC);
> + buf = kzalloc(PAGE_SIZE, GFP_ATOMIC);
> if (buf) {
> dump_mle(mle, buf, PAGE_SIZE - 1);
> - free_page((unsigned long)buf);
> + kfree(buf);
> }
> }
>
> @@ -280,7 +280,7 @@ static struct dentry *dlm_debugfs_root;
> /* begin - utils funcs */
> static int debug_release(struct inode *inode, struct file *file)
> {
> - free_page((unsigned long)file->private_data);
> + kfree(file->private_data);
> return 0;
> }
>
> @@ -327,17 +327,15 @@ static int debug_purgelist_open(struct inode *inode, struct file *file)
> struct dlm_ctxt *dlm = inode->i_private;
> char *buf = NULL;
>
> - buf = (char *) get_zeroed_page(GFP_NOFS);
> + buf = kzalloc(PAGE_SIZE, GFP_NOFS);
> if (!buf)
> - goto bail;
> + return -ENOMEM;
>
> i_size_write(inode, debug_purgelist_print(dlm, buf, PAGE_SIZE - 1));
>
> file->private_data = buf;
>
> return 0;
> -bail:
> - return -ENOMEM;
> }
>
> static const struct file_operations debug_purgelist_fops = {
> @@ -384,17 +382,15 @@ static int debug_mle_open(struct inode *inode, struct file *file)
> struct dlm_ctxt *dlm = inode->i_private;
> char *buf = NULL;
>
> - buf = (char *) get_zeroed_page(GFP_NOFS);
> + buf = kzalloc(PAGE_SIZE, GFP_NOFS);
> if (!buf)
> - goto bail;
> + return -ENOMEM;
>
> i_size_write(inode, debug_mle_print(dlm, buf, PAGE_SIZE - 1));
>
> file->private_data = buf;
>
> return 0;
> -bail:
> - return -ENOMEM;
> }
>
> static const struct file_operations debug_mle_fops = {
> @@ -775,17 +771,15 @@ static int debug_state_open(struct inode *inode, struct file *file)
> struct dlm_ctxt *dlm = inode->i_private;
> char *buf = NULL;
>
> - buf = (char *) get_zeroed_page(GFP_NOFS);
> + buf = kzalloc(PAGE_SIZE, GFP_NOFS);
> if (!buf)
> - goto bail;
> + return -ENOMEM;
>
> i_size_write(inode, debug_state_print(dlm, buf, PAGE_SIZE - 1));
>
> file->private_data = buf;
>
> return 0;
> -bail:
> - return -ENOMEM;
> }
>
> static const struct file_operations debug_state_fops = {
> diff --git a/fs/ocfs2/dlm/dlmdomain.c b/fs/ocfs2/dlm/dlmdomain.c
> index dc9da9133c8e..97bb9400e24b 100644
> --- a/fs/ocfs2/dlm/dlmdomain.c
> +++ b/fs/ocfs2/dlm/dlmdomain.c
> @@ -63,7 +63,7 @@ static inline void byte_copymap(u8 dmap[], unsigned long smap[],
> static void dlm_free_pagevec(void **vec, int pages)
> {
> while (pages--)
> - free_page((unsigned long)vec[pages]);
> + kfree(vec[pages]);
> kfree(vec);
> }
>
> @@ -75,9 +75,11 @@ static void **dlm_alloc_pagevec(int pages)
> if (!vec)
> return NULL;
>
> - for (i = 0; i < pages; i++)
> - if (!(vec[i] = (void *)__get_free_page(GFP_KERNEL)))
> + for (i = 0; i < pages; i++) {
> + vec[i] = kmalloc(PAGE_SIZE, GFP_KERNEL);
> + if (!vec[i])
> goto out_free;
> + }
>
> mlog(0, "Allocated DLM hash pagevec; %d pages (%lu expected), %lu buckets per page\n",
> pages, (unsigned long)DLM_HASH_PAGES,
> diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c
> index 93eff38fdadd..aee3b4c56dcc 100644
> --- a/fs/ocfs2/dlm/dlmmaster.c
> +++ b/fs/ocfs2/dlm/dlmmaster.c
> @@ -2548,7 +2548,7 @@ static int dlm_migrate_lockres(struct dlm_ctxt *dlm,
>
> /* preallocate up front. if this fails, abort */
> ret = -ENOMEM;
> - mres = (struct dlm_migratable_lockres *) __get_free_page(GFP_NOFS);
> + mres = kmalloc(PAGE_SIZE, GFP_NOFS);
> if (!mres) {
> mlog_errno(ret);
> goto leave;
> @@ -2725,8 +2725,7 @@ static int dlm_migrate_lockres(struct dlm_ctxt *dlm,
> if (wake)
> wake_up(&res->wq);
>
> - if (mres)
> - free_page((unsigned long)mres);
> + kfree(mres);
>
> dlm_put(dlm);
>
> diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c
> index 128872bd945d..9b97bf73df22 100644
> --- a/fs/ocfs2/dlm/dlmrecovery.c
> +++ b/fs/ocfs2/dlm/dlmrecovery.c
> @@ -837,7 +837,7 @@ int dlm_request_all_locks_handler(struct o2net_msg *msg, u32 len, void *data,
> }
>
> /* this will get freed by dlm_request_all_locks_worker */
> - buf = (char *) __get_free_page(GFP_NOFS);
> + buf = kmalloc(PAGE_SIZE, GFP_NOFS);
> if (!buf) {
> kfree(item);
> dlm_put(dlm);
> @@ -933,7 +933,7 @@ static void dlm_request_all_locks_worker(struct dlm_work_item *item, void *data)
> }
> }
> leave:
> - free_page((unsigned long)data);
> + kfree(data);
> }
>
>
>
> --
> 2.53.0
>
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply
* Re: [PATCH 02/17] proc: replace __get_free_page() with kmalloc()
From: Jan Kara @ 2026-05-25 16:11 UTC (permalink / raw)
To: Mike Rapoport (Microsoft)
Cc: Jan Kara, Mark Fasheh, Joel Becker, Joseph Qi, Ryusuke Konishi,
Viacheslav Dubeyko, Trond Myklebust, Anna Schumaker, Chuck Lever,
Jeff Layton, NeilBrown, Olga Kornievskaia, Dai Ngo, Tom Talpey,
Alexander Viro, Christian Brauner, Jan Kara, Dave Kleikamp,
Theodore Ts'o, Miklos Szeredi, Andreas Hindborg, Breno Leitao,
Kees Cook, Tigran A. Aivazian, linux-kernel, linux-fsdevel,
ocfs2-devel, linux-nilfs, linux-nfs, jfs-discussion, linux-ext4,
linux-mm
In-Reply-To: <20260523-b4-fs-v1-2-275e36a83f0e@kernel.org>
On Sat 23-05-26 20:54:14, Mike Rapoport (Microsoft) wrote:
> A few functions in fs/proc/base.c use __get_free_page() to allocate a
> temporary buffer.
>
> kmalloc() is a better API for such use and it also provides better
> scalability and more debugging possibilities.
>
> Replace use of __get_free_page() with kmalloc().
>
> Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Looks good. Feel free to add:
Reviewed-by: Jan Kara <jack@suse.cz>
Honza
> ---
> fs/proc/base.c | 16 ++++++++--------
> 1 file changed, 8 insertions(+), 8 deletions(-)
>
> diff --git a/fs/proc/base.c b/fs/proc/base.c
> index d9acfa89c894..e129dc509b79 100644
> --- a/fs/proc/base.c
> +++ b/fs/proc/base.c
> @@ -261,7 +261,7 @@ static ssize_t get_mm_proctitle(struct mm_struct *mm, char __user *buf,
> if (pos >= PAGE_SIZE)
> return 0;
>
> - page = (char *)__get_free_page(GFP_KERNEL);
> + page = kmalloc(PAGE_SIZE, GFP_KERNEL);
> if (!page)
> return -ENOMEM;
>
> @@ -284,7 +284,7 @@ static ssize_t get_mm_proctitle(struct mm_struct *mm, char __user *buf,
> ret = len;
> }
> }
> - free_page((unsigned long)page);
> + kfree(page);
> return ret;
> }
>
> @@ -347,7 +347,7 @@ static ssize_t get_mm_cmdline(struct mm_struct *mm, char __user *buf,
> if (count > arg_end - pos)
> count = arg_end - pos;
>
> - page = (char *)__get_free_page(GFP_KERNEL);
> + page = kmalloc(PAGE_SIZE, GFP_KERNEL);
> if (!page)
> return -ENOMEM;
>
> @@ -371,7 +371,7 @@ static ssize_t get_mm_cmdline(struct mm_struct *mm, char __user *buf,
> count -= got;
> }
>
> - free_page((unsigned long)page);
> + kfree(page);
> return len;
> }
>
> @@ -908,7 +908,7 @@ static ssize_t mem_rw(struct file *file, char __user *buf,
> if (!mm)
> return 0;
>
> - page = (char *)__get_free_page(GFP_KERNEL);
> + page = kmalloc(PAGE_SIZE, GFP_KERNEL);
> if (!page)
> return -ENOMEM;
>
> @@ -949,7 +949,7 @@ static ssize_t mem_rw(struct file *file, char __user *buf,
>
> mmput(mm);
> free:
> - free_page((unsigned long) page);
> + kfree(page);
> return copied;
> }
>
> @@ -1016,7 +1016,7 @@ static ssize_t environ_read(struct file *file, char __user *buf,
> if (!mm || !mm->env_end)
> return 0;
>
> - page = (char *)__get_free_page(GFP_KERNEL);
> + page = kmalloc(PAGE_SIZE, GFP_KERNEL);
> if (!page)
> return -ENOMEM;
>
> @@ -1062,7 +1062,7 @@ static ssize_t environ_read(struct file *file, char __user *buf,
> mmput(mm);
>
> free:
> - free_page((unsigned long) page);
> + kfree(page);
> return ret;
> }
>
>
> --
> 2.53.0
>
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply
* Re: [PATCH 01/17] quota: allocate dquot_hash with kmalloc()
From: Jan Kara @ 2026-05-25 16:10 UTC (permalink / raw)
To: Mike Rapoport (Microsoft)
Cc: Jan Kara, Mark Fasheh, Joel Becker, Joseph Qi, Ryusuke Konishi,
Viacheslav Dubeyko, Trond Myklebust, Anna Schumaker, Chuck Lever,
Jeff Layton, NeilBrown, Olga Kornievskaia, Dai Ngo, Tom Talpey,
Alexander Viro, Christian Brauner, Jan Kara, Dave Kleikamp,
Theodore Ts'o, Miklos Szeredi, Andreas Hindborg, Breno Leitao,
Kees Cook, Tigran A. Aivazian, linux-kernel, linux-fsdevel,
ocfs2-devel, linux-nilfs, linux-nfs, jfs-discussion, linux-ext4,
linux-mm
In-Reply-To: <20260523-b4-fs-v1-1-275e36a83f0e@kernel.org>
On Sat 23-05-26 20:54:13, Mike Rapoport (Microsoft) wrote:
> dquot_init() allocates a single page for dquot_hash with
> __get_free_pages().
>
> kmalloc() is a better API for such use and it also provides better
> scalability and more debugging possibilities.
>
> Replace use of __get_free_pages() with kmalloc() and get rid of the order
> variable that remained 0 for more than 20 years.
>
> Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Thanks! I've added this patch to my tree.
Honza
> ---
> fs/quota/dquot.c | 11 +++++------
> 1 file changed, 5 insertions(+), 6 deletions(-)
>
> diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c
> index 64cf42721496..9850de3955d3 100644
> --- a/fs/quota/dquot.c
> +++ b/fs/quota/dquot.c
> @@ -3022,7 +3022,7 @@ static const struct ctl_table fs_dqstats_table[] = {
> static int __init dquot_init(void)
> {
> int i, ret;
> - unsigned long nr_hash, order;
> + unsigned long nr_hash;
> struct shrinker *dqcache_shrinker;
>
> printk(KERN_NOTICE "VFS: Disk quotas %s\n", __DQUOT_VERSION__);
> @@ -3035,8 +3035,7 @@ static int __init dquot_init(void)
> SLAB_PANIC),
> NULL);
>
> - order = 0;
> - dquot_hash = (struct hlist_head *)__get_free_pages(GFP_KERNEL, order);
> + dquot_hash = kmalloc(PAGE_SIZE, GFP_KERNEL);
> if (!dquot_hash)
> panic("Cannot create dquot hash table");
>
> @@ -3046,7 +3045,7 @@ static int __init dquot_init(void)
> panic("Cannot create dquot stat counters");
>
> /* Find power-of-two hlist_heads which can fit into allocation */
> - nr_hash = (1UL << order) * PAGE_SIZE / sizeof(struct hlist_head);
> + nr_hash = PAGE_SIZE / sizeof(struct hlist_head);
> dq_hash_bits = ilog2(nr_hash);
>
> nr_hash = 1UL << dq_hash_bits;
> @@ -3054,8 +3053,8 @@ static int __init dquot_init(void)
> for (i = 0; i < nr_hash; i++)
> INIT_HLIST_HEAD(dquot_hash + i);
>
> - pr_info("VFS: Dquot-cache hash table entries: %ld (order %ld,"
> - " %ld bytes)\n", nr_hash, order, (PAGE_SIZE << order));
> + pr_info("VFS: Dquot-cache hash table entries: %ld (%ld bytes)\n",
> + nr_hash, PAGE_SIZE);
>
> dqcache_shrinker = shrinker_alloc(0, "dquota-cache");
> if (!dqcache_shrinker)
>
> --
> 2.53.0
>
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply
* Re: [PATCH] jbd2: update outdated comment for jbd2_journal_try_to_free_buffers()
From: Jan Kara @ 2026-05-25 15:42 UTC (permalink / raw)
To: Zhang Yi
Cc: linux-ext4, linux-fsdevel, linux-kernel, tytso, adilger.kernel,
libaokun, jack, ojaswin, ritesh.list, yi.zhang, yizhang089,
yangerkun, yukuai
In-Reply-To: <20260522030540.3896201-1-yi.zhang@huaweicloud.com>
On Fri 22-05-26 11:05:40, Zhang Yi wrote:
> From: Zhang Yi <yi.zhang@huawei.com>
>
> jbd2_journal_try_to_free_buffers() currently only tries to remove
> checkpointed data buffers from the checkpoint list for data=journal
> mode, and bails out if any buffer is still attached to a transaction.
> For data=ordered and writeback modes, data buffers never have
> journal_heads, so the function degenerates to a plain
> try_to_free_buffers() call.
>
> Besides, The release of metadata buffers has been delegated to the jbd2
> journal shrinker in commit 4ba3fcdde7e3 ("jbd2,ext4: add a shrinker to
> release checkpointed buffers"). jbd2_journal_try_to_free_buffers() is
> not used for handling metadata buffers now.
>
> However, the comment above the function still references
> jbd2_journal_dirty_data(), __jbd2_journal_unfile_buffer(), t_datalist,
> BKL, and BUF_CLEAN, all of which were removed in commit 87c89c232c8f
> ("jbd2: Remove data=ordered mode support using jbd buffer heads").
>
> Replace it with a description of what the function actually does now.
>
> Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
Thanks for the update. Feel free to add:
Reviewed-by: Jan Kara <jack@suse.cz>
Honza
> ---
> fs/jbd2/transaction.c | 39 ++++++++++++---------------------------
> 1 file changed, 12 insertions(+), 27 deletions(-)
>
> diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c
> index 4885903bbd10..239bcf88ed1c 100644
> --- a/fs/jbd2/transaction.c
> +++ b/fs/jbd2/transaction.c
> @@ -2139,38 +2139,23 @@ static void __jbd2_journal_unfile_buffer(struct journal_head *jh)
> }
>
> /**
> - * jbd2_journal_try_to_free_buffers() - try to free page buffers.
> + * jbd2_journal_try_to_free_buffers() - try to free folio buffers.
> * @journal: journal for operation
> * @folio: Folio to detach data from.
> *
> - * For all the buffers on this page,
> - * if they are fully written out ordered data, move them onto BUF_CLEAN
> - * so try_to_free_buffers() can reap them.
> + * For each buffer_head on @folio, if the buffer has a journal_head but
> + * is not attached to a running or committing transaction, try to remove
> + * it from the checkpoint list. This is needed for data=journal mode
> + * where data buffers are journaled: once they are checkpointed, the
> + * journal_head can be detached and the buffer freed. If any buffer is
> + * still attached to a transaction, the folio cannot be released and we
> + * bail out. Otherwise we call try_to_free_buffers() to detach all
> + * buffer_heads from the folio.
> *
> - * This function returns non-zero if we wish try_to_free_buffers()
> - * to be called. We do this if the page is releasable by try_to_free_buffers().
> - * We also do it if the page has locked or dirty buffers and the caller wants
> - * us to perform sync or async writeout.
> + * For data=ordered and writeback modes, data buffers never have
> + * journal_heads, so this degenerates to a plain try_to_free_buffers().
> *
> - * This complicates JBD locking somewhat. We aren't protected by the
> - * BKL here. We wish to remove the buffer from its committing or
> - * running transaction's ->t_datalist via __jbd2_journal_unfile_buffer.
> - *
> - * This may *change* the value of transaction_t->t_datalist, so anyone
> - * who looks at t_datalist needs to lock against this function.
> - *
> - * Even worse, someone may be doing a jbd2_journal_dirty_data on this
> - * buffer. So we need to lock against that. jbd2_journal_dirty_data()
> - * will come out of the lock with the buffer dirty, which makes it
> - * ineligible for release here.
> - *
> - * Who else is affected by this? hmm... Really the only contender
> - * is do_get_write_access() - it could be looking at the buffer while
> - * journal_try_to_free_buffer() is changing its state. But that
> - * cannot happen because we never reallocate freed data as metadata
> - * while the data is part of a transaction. Yes?
> - *
> - * Return false on failure, true on success
> + * Return: true if the folio's buffers were freed, false otherwise
> */
> bool jbd2_journal_try_to_free_buffers(journal_t *journal, struct folio *folio)
> {
> --
> 2.52.0
>
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply
* Re: [PATCH 2/2] ext4: get ext4_group_desc in ext4_mb_prefetch only when necessary
From: Jan Kara @ 2026-05-25 15:32 UTC (permalink / raw)
To: Bohdan Trach
Cc: Theodore Ts'o, Andreas Dilger, Baokun Li, Jan Kara,
Ojaswin Mujoo, Ritesh Harjani (IBM), Zhang Yi, mchehab+huawei,
lilith.oberhauser, linux-ext4, linux-kernel
In-Reply-To: <20260521125931.16474-3-bohdan.trach@huaweicloud.com>
On Thu 21-05-26 14:59:29, Bohdan Trach wrote:
> Getting ext4_group_desc structure can contribute to the cost of
> ext4_mb_prefetch() without any need, as most groups fail the
> !EXT4_MB_GRP_TEST_AND_SET_READ check.
>
> Optimize ext4_mb_prefetch by getting the group description only when
> necessary.
>
> The result is further increase in performance of fallocate() system call
> path that triggers ext4_mb_prefetch() via a linear group scan.
>
> Signed-off-by: Bohdan Trach <bohdan.trach@huaweicloud.com>
Looks good. Just one nit below:
> @@ -2872,14 +2870,17 @@ ext4_group_t ext4_mb_prefetch(struct super_block *sb, ext4_group_t group,
> * prefetch once, so we avoid getblk() call, which can
> * be expensive.
> */
> - if (gdp && grp && !EXT4_MB_GRP_TEST_AND_SET_READ(grp) &&
> - EXT4_MB_GRP_NEED_INIT(grp) &&
> - ext4_free_group_clusters(sb, gdp) > 0 ) {
> - bh = ext4_read_block_bitmap_nowait(sb, group, true);
> - if (!IS_ERR_OR_NULL(bh)) {
> - if (!buffer_uptodate(bh) && cnt)
> - (*cnt)++;
> - brelse(bh);
> + if (group < ngroups && grp && !EXT4_MB_GRP_TEST_AND_SET_READ(grp) &&
The group < ngroup should be always true and is also implicitly contained
in the 'grp != NULL' check. So I'd remove that bit. Otherwise feel free to
add:
Reviewed-by: Jan Kara <jack@suse.cz>
Honza
> + EXT4_MB_GRP_NEED_INIT(grp)) {
> + struct ext4_group_desc *gdp = ext4_get_group_desc(sb, group, NULL);
> +
> + if (gdp && ext4_free_group_clusters(sb, gdp) > 0) {
> + bh = ext4_read_block_bitmap_nowait(sb, group, true);
> + if (!IS_ERR_OR_NULL(bh)) {
> + if (!buffer_uptodate(bh) && cnt)
> + (*cnt)++;
> + brelse(bh);
> + }
> }
> }
> if (++group >= ngroups)
> --
> 2.43.0
>
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply
* Re: [PATCH 1/2] ext4: avoid RWM atomic in EXT4_MB_GRP_TEST_AND_SET_READ
From: Jan Kara @ 2026-05-25 15:28 UTC (permalink / raw)
To: Bohdan Trach
Cc: Theodore Ts'o, Andreas Dilger, Baokun Li, Jan Kara,
Ojaswin Mujoo, Ritesh Harjani (IBM), Zhang Yi, mchehab+huawei,
lilith.oberhauser, linux-ext4, linux-kernel
In-Reply-To: <20260521125931.16474-2-bohdan.trach@huaweicloud.com>
On Thu 21-05-26 14:59:28, Bohdan Trach wrote:
> EXT4_MB_GRP_TEST_AND_SET_READ uses test_and_set_bit function which
> issues an atomic write. This can cause high overhead due to cache
> contention when multiple threads iterate over groups in a tight loop,
> as is the case for ext4_mb_prefetch(). We have seen this to be a
> problem for Kunpeng 920b CPUs which uses a single ARM LSE instruction
> for this purpose.
>
> This change significantly reduces costs of fallocate() operations which
> trigger linear group scans on large multicore machines where
> test_and_set_bit issues an atomic write operation unconditionally.
>
> Signed-off-by: Bohdan Trach <bohdan.trach@huaweicloud.com>
...
> diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
> index 56b82d4a15d7..0713207811a6 100644
> --- a/fs/ext4/ext4.h
> +++ b/fs/ext4/ext4.h
> @@ -3551,7 +3551,17 @@ struct ext4_group_info {
> #define EXT4_MB_GRP_CLEAR_TRIMMED(grp) \
> (clear_bit(EXT4_GROUP_INFO_WAS_TRIMMED_BIT, &((grp)->bb_state)))
> #define EXT4_MB_GRP_TEST_AND_SET_READ(grp) \
> - (test_and_set_bit(EXT4_GROUP_INFO_BBITMAP_READ_BIT, &((grp)->bb_state)))
> + (ext4_mb_grp_test_and_set_read((grp)))
> +
> +static inline int ext4_mb_grp_test_and_set_read(struct ext4_group_info *grp)
> +{
> + int r = test_bit_acquire(EXT4_GROUP_INFO_BBITMAP_READ_BIT, &grp->bb_state);
> +
> + if (!r)
> + return test_and_set_bit(EXT4_GROUP_INFO_BBITMAP_READ_BIT, &grp->bb_state);
> + else
> + return r;
> +}
Good idea but do we really need the 'acquire' barrier here? I don't see
anything that would really need this so I think
EXT4_MB_GRP_TEST_AND_SET_READ() can be just:
test_bit(EXT4_GROUP_INFO_BBITMAP_READ_BIT, &grp->bb_state) || \
test_and_set_bit(EXT4_GROUP_INFO_BBITMAP_READ_BIT, &grp->bb_state)
or am I missing something?
Honza
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply
* Re: [PATCH] ext4: convert legacy ext4_debug() to standard pr_debug()
From: Jan Kara @ 2026-05-25 15:16 UTC (permalink / raw)
To: lirongqing
Cc: Theodore Ts'o, Andreas Dilger, Baokun Li, Jan Kara,
Ojaswin Mujoo, Ritesh Harjani, Zhang Yi, linux-ext4, linux-kernel
In-Reply-To: <20260521074634.2697-1-lirongqing@baidu.com>
On Thu 21-05-26 03:46:34, lirongqing wrote:
> From: Li RongQing <lirongqing@baidu.com>
>
> The ext4 file system historically implemented its own debug logging macro
> ext4_debug() via EXT4FS_DEBUG conditional compilation. This legacy
> implementation suffers from two major drawbacks:
>
> 1. It makes two consecutive un-serialized printk() calls, which can
> lead to severe log interleaving and corruption under multi-core
> concurrent workloads.
> 2. It completely bypasses the standard modern kernel dynamic debug
> (CONFIG_DYNAMIC_DEBUG) infrastructure.
>
> Clean up the legacy implementation by leveraging pr_debug(). This squashes
> the multiple printk() calls into a single atomic execution, ensuring
> log integrity, while seamlessly hooking ext4 into the kernel's native
> dynamic debug framework.
>
> The redundant __FILE__ and __LINE__ macros are intentionally removed from
> the string format because the dynamic debug infrastructure can already
> append them automatically at runtime (via the '+fl' flags) if desired.
> This avoids redundancy and double-logging in modern production/debugging
> environments while keeping the macro clean and robust against dangling
> comma compiler errors.
>
> Signed-off-by: Li RongQing <lirongqing@baidu.com>
Nice! Feel free to add:
Reviewed-by: Jan Kara <jack@suse.cz>
Honza
> ---
> fs/ext4/ext4.h | 20 ++------------------
> 1 file changed, 2 insertions(+), 18 deletions(-)
>
> diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
> index 94283a9..39e86ff 100644
> --- a/fs/ext4/ext4.h
> +++ b/fs/ext4/ext4.h
> @@ -62,24 +62,8 @@
> */
> #define DOUBLE_CHECK__
>
> -/*
> - * Define EXT4FS_DEBUG to produce debug messages
> - */
> -#undef EXT4FS_DEBUG
> -
> -/*
> - * Debug code
> - */
> -#ifdef EXT4FS_DEBUG
> -#define ext4_debug(f, a...) \
> - do { \
> - printk(KERN_DEBUG "EXT4-fs DEBUG (%s, %d): %s:", \
> - __FILE__, __LINE__, __func__); \
> - printk(KERN_DEBUG f, ## a); \
> - } while (0)
> -#else
> -#define ext4_debug(fmt, ...) no_printk(fmt, ##__VA_ARGS__)
> -#endif
> +#define ext4_debug(fmt, ...) \
> + pr_debug("EXT4-fs DEBUG %s: " fmt, __func__, ##__VA_ARGS__)
>
> /*
> * Turn on EXT_DEBUG to enable ext4_ext_show_path/leaf/move in extents.c
> --
> 2.9.4
>
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply
* Re: [PATCH 2/2] fs: jbd2: use clear_and_wake_up_bit() in journal_end_buffer_io_sync()
From: Jan Kara @ 2026-05-25 15:03 UTC (permalink / raw)
To: Agatha Isabelle Moreira
Cc: linux-ext4, linux-kernel, linux-fsdevel, Theodore Ts'o,
Jan Kara, shuo chen, linux-kernel-mentees, shuah, patch-reply
In-Reply-To: <ag4SrrOl7R2DcLLi@guidai>
On Wed 20-05-26 17:05:46, Agatha Isabelle Moreira wrote:
> Use `clear_and_wake_up_bit()` in `journal_end_buffer_io_sync()`, since
> the helper was introduced in 'commit 8236b0ae31c83 ("bdi: wake up
> concurrent wb_shutdown() callers.")' as a generic way of doing the same
> sequence of operations:
> clear_bit_unlock();
> smp_mb__after_atomic();
> wake_up_bit();
>
> The helper was first implemented to avoid bugs caused by forgetting to
> call `wake_up_bit()` after `clear_bit_unlock()`.
>
> Since `journal_end_buffer_io_sync()` was first introduced by 'commit
> 470decc613ab2 ("jbd2: initial copy of files from jbd")' and last
> modified in this operation by 'commit 4e857c58efeb9 ("arch: Mass
> conversion of smp_mb__*()")', years before `clear_and_wake_up_bit()`, it
> still uses the open-coded sequence.
>
> Replace the open-coded sequence with the helper to avoid duplicate code
> and reduce code paths to maintain.
>
> Suggested-by: shuo chen <1289151713@qq.com>
> Link: https://lore.kernel.org/kernelnewbies/agzoqV835-co4kAN@guidai/T/#t
> Signed-off-by: Agatha Isabelle Moreira <code@agatha.dev>
Looks good. Feel free to add:
Reviewed-by: Jan Kara <jack@suse.cz>
Honza
> ---
> fs/jbd2/commit.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c
> index 8cf61e7185c4..b647fde76e49 100644
> --- a/fs/jbd2/commit.c
> +++ b/fs/jbd2/commit.c
> @@ -39,9 +39,7 @@ static void journal_end_buffer_io_sync(struct buffer_head *bh, int uptodate)
> else
> clear_buffer_uptodate(bh);
> if (orig_bh) {
> - clear_bit_unlock(BH_Shadow, &orig_bh->b_state);
> - smp_mb__after_atomic();
> - wake_up_bit(&orig_bh->b_state, BH_Shadow);
> + clear_and_wake_up_bit(BH_Shadow, &orig_bh->b_state);
> }
> unlock_buffer(bh);
> }
> --
> 2.53.0
>
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply
* Re: [PATCH 1/2] fs: buffer: use clear_and_wake_up_bit() in unlock_buffer()
From: Jan Kara @ 2026-05-25 15:02 UTC (permalink / raw)
To: Agatha Isabelle Moreira
Cc: linux-fsdevel, linux-ext4, linux-kernel, Alexander Viro,
Christian Brauner, Jan Kara, shuo chen, linux-kernel-mentees,
shuah, patch-reply
In-Reply-To: <ag4SD-mkmn5IbuN7@guidai>
On Wed 20-05-26 16:58:16, Agatha Isabelle Moreira wrote:
> Use `clear_and_wake_up_bit()` in `unlock_buffer()`, since the helper was
> introduced in 'commit 8236b0ae31c83 ("bdi: wake up concurrent
> wb_shutdown() callers.")' as a generic way of doing the same sequence of
> operations:
> clear_bit_unlock();
> smp_mb__after_atomic();
> wake_up_bit();
>
> The helper was implemented to avoid bugs caused by forgetting to call
> `wake_up_bit()` after `clear_bit_unlock()`.
>
> Since `unlock_buffer()` predates git and was last modified in
> 'commit 4e857c58efeb9 ("arch: Mass conversion of smp_mb__*()")', years
> before `clear_and_wake_up_bit()`, it still uses the open-coded sequence.
>
> Replace the open-coded sequence with the helper to avoid duplicate code
> and reduce code paths to maintain.
>
> Suggested-by: shuo chen <1289151713@qq.com>
> Link: https://lore.kernel.org/kernelnewbies/agzoqV835-co4kAN@guidai/T/#t
> Signed-off-by: Agatha Isabelle Moreira <code@agatha.dev>
Looks good. Feel free to add:
Reviewed-by: Jan Kara <jack@suse.cz>
Honza
> ---
> fs/buffer.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/fs/buffer.c b/fs/buffer.c
> index b0b3792b1496..4348b240bd97 100644
> --- a/fs/buffer.c
> +++ b/fs/buffer.c
> @@ -74,9 +74,7 @@ EXPORT_SYMBOL(__lock_buffer);
>
> void unlock_buffer(struct buffer_head *bh)
> {
> - clear_bit_unlock(BH_Lock, &bh->b_state);
> - smp_mb__after_atomic();
> - wake_up_bit(&bh->b_state, BH_Lock);
> + clear_and_wake_up_bit(BH_Lock, &bh->b_state);
> }
> EXPORT_SYMBOL(unlock_buffer);
>
> --
> 2.53.0
>
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply
* Re: [PATCH v2 01/10] affs: Drop support for metadata bh tracking
From: David Sterba @ 2026-05-25 10:06 UTC (permalink / raw)
To: Jan Kara
Cc: linux-fsdevel, Christian Brauner, aivazian.tigran, Ted Tso,
linux-ext4, OGAWA Hirofumi, David Sterba
In-Reply-To: <20260525085821.769119-11-jack@suse.cz>
On Mon, May 25, 2026 at 10:58:07AM +0200, Jan Kara wrote:
> AFFS did all the hard work of tracking metadata bhs dirtied for an inode
> but it actually never used this information as affs_file_fsync() just
> calls sync_blockdev() to writeback all filesystem metadata bhs. After a
> discussion with AFFS maintainer nobody cares about AFFS performance
> so let's keep this affs_file_fsync() behavior and just drop all the
> pointless tracking from AFFS.
>
> CC: David Sterba <dsterba@suse.com>
> Signed-off-by: Jan Kara <jack@suse.cz>
Acked-by: David Sterba <dsterba@suse.com>
^ permalink raw reply
* [PATCH v2 10/10] fs: Fix missed inode writeback when racing with __writeback_single_inode
From: Jan Kara @ 2026-05-25 8:58 UTC (permalink / raw)
To: linux-fsdevel
Cc: Christian Brauner, aivazian.tigran, Ted Tso, linux-ext4,
OGAWA Hirofumi, Jan Kara
In-Reply-To: <20260525085035.12891-1-jack@suse.cz>
When mmb_fsync_noflush() or simple_fsync_noflush() race with another
writeback of the same inode, they can see inode dirty bits are already
clear and skip inode writeback although the racing
__writeback_single_inode() didn't yet get to writing anything. This can
result in fsync(2) returning without properly persisting the inode.
We already have I_SYNC bit for this synchronization and
writeback_single_inode() properly uses it so just fix
mmb_fsync_noflush() and simple_fsync_noflush() to take it into account
as well.
Signed-off-by: Jan Kara <jack@suse.cz>
---
fs/buffer.c | 5 +++--
fs/libfs.c | 5 +++--
2 files changed, 6 insertions(+), 4 deletions(-)
diff --git a/fs/buffer.c b/fs/buffer.c
index f83fb3cdc6ac..01e17a287a42 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -662,9 +662,10 @@ int mmb_fsync_noflush(struct file *file, struct mapping_metadata_bhs *mmb,
if (err)
return err;
- if (!(inode_state_read_once(inode) & I_DIRTY_ALL))
+ if (!(inode_state_read_once(inode) & (I_DIRTY_ALL | I_SYNC)))
goto sync_buffers;
- if (datasync && !(inode_state_read_once(inode) & I_DIRTY_DATASYNC))
+ if (datasync &&
+ !(inode_state_read_once(inode) & (I_DIRTY_DATASYNC | I_SYNC)))
goto sync_buffers;
ret = sync_inode_metadata(inode, 1);
diff --git a/fs/libfs.c b/fs/libfs.c
index 1bbea5e7bae3..3a98dacd81b2 100644
--- a/fs/libfs.c
+++ b/fs/libfs.c
@@ -1559,9 +1559,10 @@ int simple_fsync_noflush(struct file *file, loff_t start, loff_t end,
if (err)
return err;
- if (!(inode_state_read_once(inode) & I_DIRTY_ALL))
+ if (!(inode_state_read_once(inode) & (I_DIRTY_ALL | I_SYNC)))
goto out;
- if (datasync && !(inode_state_read_once(inode) & I_DIRTY_DATASYNC))
+ if (datasync &&
+ !(inode_state_read_once(inode) & (I_DIRTY_DATASYNC | I_SYNC)))
goto out;
ret = sync_inode_metadata(inode, 1);
--
2.51.0
^ permalink raw reply related
* [PATCH v2 07/10] minix: Fix possibly missing inode write on fsync(2)
From: Jan Kara @ 2026-05-25 8:58 UTC (permalink / raw)
To: linux-fsdevel
Cc: Christian Brauner, aivazian.tigran, Ted Tso, linux-ext4,
OGAWA Hirofumi, Jan Kara
In-Reply-To: <20260525085035.12891-1-jack@suse.cz>
Use mmb inode buffer writeout infrastructure to reliably write out
inode's buffer on fsync(2).
Signed-off-by: Jan Kara <jack@suse.cz>
---
fs/minix/inode.c | 12 ++----------
1 file changed, 2 insertions(+), 10 deletions(-)
diff --git a/fs/minix/inode.c b/fs/minix/inode.c
index 9c6bac248907..401056fb682e 100644
--- a/fs/minix/inode.c
+++ b/fs/minix/inode.c
@@ -649,7 +649,7 @@ static struct buffer_head * V1_minix_update_inode(struct inode * inode)
raw_inode->i_zone[0] = old_encode_dev(inode->i_rdev);
else for (i = 0; i < 9; i++)
raw_inode->i_zone[i] = minix_inode->u.i1_data[i];
- mark_buffer_dirty(bh);
+ mmb_mark_inode_buffer_dirty(bh, &minix_i(inode)->i_metadata_bhs);
return bh;
}
@@ -678,7 +678,7 @@ static struct buffer_head * V2_minix_update_inode(struct inode * inode)
raw_inode->i_zone[0] = old_encode_dev(inode->i_rdev);
else for (i = 0; i < 10; i++)
raw_inode->i_zone[i] = minix_inode->u.i2_data[i];
- mark_buffer_dirty(bh);
+ mmb_mark_inode_buffer_dirty(bh, &minix_i(inode)->i_metadata_bhs);
return bh;
}
@@ -693,14 +693,6 @@ static int minix_write_inode(struct inode *inode, struct writeback_control *wbc)
bh = V2_minix_update_inode(inode);
if (!bh)
return -EIO;
- if (wbc->sync_mode == WB_SYNC_ALL && buffer_dirty(bh)) {
- sync_dirty_buffer(bh);
- if (buffer_req(bh) && !buffer_uptodate(bh)) {
- printk("IO error syncing minix inode [%s:%08llx]\n",
- inode->i_sb->s_id, inode->i_ino);
- err = -EIO;
- }
- }
brelse (bh);
return err;
}
--
2.51.0
^ permalink raw reply related
* [PATCH v2 09/10] ext4: Use mmb infrastructure for inode buffer writeout
From: Jan Kara @ 2026-05-25 8:58 UTC (permalink / raw)
To: linux-fsdevel
Cc: Christian Brauner, aivazian.tigran, Ted Tso, linux-ext4,
OGAWA Hirofumi, Jan Kara
In-Reply-To: <20260525085035.12891-1-jack@suse.cz>
Use mmb inode buffer writeout infrastructure to reliably write out
inode's inode table block on fsync(2) in nojournal mode (from
ext4_sync_parent() and ext4_fsync_nojournal()). This significantly
simplifies the code as we don't have to explicitely handle inode buffer
writeback in ext4_write_inode() and thus we can also remove
sync_inode_metadata() calls from ext4_sync_parent() and
ext4_write_inode() call from ext4_fsync_nojournal().
Signed-off-by: Jan Kara <jack@suse.cz>
---
fs/ext4/ext4_jbd2.c | 2 +-
fs/ext4/ext4_jbd2.h | 2 ++
fs/ext4/fsync.c | 12 ------------
fs/ext4/inode.c | 33 ++++++++++-----------------------
4 files changed, 13 insertions(+), 36 deletions(-)
diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index 752326f3b653..3d6b8f494a76 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -350,7 +350,7 @@ int __ext4_journal_get_create_access(const char *where, unsigned int line,
return 0;
}
-static void ext4_inode_attach_mmb(struct inode *inode)
+void ext4_inode_attach_mmb(struct inode *inode)
{
struct mapping_metadata_bhs *mmb;
diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
index 63d17c5201b5..2a01b8279c88 100644
--- a/fs/ext4/ext4_jbd2.h
+++ b/fs/ext4/ext4_jbd2.h
@@ -122,6 +122,8 @@
#define EXT4_HT_EXT_CONVERT 11
#define EXT4_HT_MAX 12
+void ext4_inode_attach_mmb(struct inode *inode);
+
int
ext4_mark_iloc_dirty(handle_t *handle,
struct inode *inode,
diff --git a/fs/ext4/fsync.c b/fs/ext4/fsync.c
index c104f55a0242..1cb613478386 100644
--- a/fs/ext4/fsync.c
+++ b/fs/ext4/fsync.c
@@ -75,9 +75,6 @@ static int ext4_sync_parent(struct inode *inode)
if (ret)
break;
}
- ret = sync_inode_metadata(inode, 1);
- if (ret)
- break;
}
dput(dentry);
return ret;
@@ -87,10 +84,6 @@ static int ext4_fsync_nojournal(struct file *file, loff_t start, loff_t end,
int datasync, bool *needs_barrier)
{
struct inode *inode = file->f_inode;
- struct writeback_control wbc = {
- .sync_mode = WB_SYNC_ALL,
- .nr_to_write = 0,
- };
int ret;
ret = mmb_fsync_noflush(file, READ_ONCE(EXT4_I(inode)->i_metadata_bhs),
@@ -98,11 +91,6 @@ static int ext4_fsync_nojournal(struct file *file, loff_t start, loff_t end,
if (ret)
return ret;
- /* Force writeout of inode table buffer to disk */
- ret = ext4_write_inode(inode, &wbc);
- if (ret)
- return ret;
-
ret = ext4_sync_parent(inode);
if (test_opt(inode->i_sb, BARRIER))
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 3e66e9510909..651201849667 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -5696,10 +5696,16 @@ static int ext4_do_update_inode(handle_t *handle,
ext4_update_other_inodes_time(inode->i_sb, inode->i_ino,
bh->b_data);
- BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata");
- err = ext4_handle_dirty_metadata(handle, NULL, bh);
- if (err)
- goto out_error;
+ if (!ext4_handle_valid(handle)) {
+ if (!ei->i_metadata_bhs)
+ ext4_inode_attach_mmb(inode);
+ mmb_mark_inode_buffer_dirty(iloc->bh, ei->i_metadata_bhs);
+ } else {
+ BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata");
+ err = ext4_handle_dirty_metadata(handle, NULL, bh);
+ if (err)
+ goto out_error;
+ }
ext4_clear_inode_state(inode, EXT4_STATE_NEW);
if (set_large_file) {
BUFFER_TRACE(EXT4_SB(sb)->s_sbh, "get write access");
@@ -5786,24 +5792,6 @@ int ext4_write_inode(struct inode *inode, struct writeback_control *wbc)
err = ext4_fc_commit(EXT4_SB(inode->i_sb)->s_journal,
EXT4_I(inode)->i_sync_tid);
- } else {
- struct ext4_iloc iloc;
-
- err = __ext4_get_inode_loc_noinmem(inode, &iloc);
- if (err)
- return err;
- /*
- * sync(2) will flush the whole buffer cache. No need to do
- * it here separately for each inode.
- */
- if (wbc->sync_mode == WB_SYNC_ALL && !wbc->for_sync)
- sync_dirty_buffer(iloc.bh);
- if (buffer_req(iloc.bh) && !buffer_uptodate(iloc.bh)) {
- ext4_error_inode_block(inode, iloc.bh->b_blocknr, EIO,
- "IO error syncing inode");
- err = -EIO;
- }
- brelse(iloc.bh);
}
return err;
}
@@ -6348,7 +6336,6 @@ int ext4_mark_iloc_dirty(handle_t *handle,
/* the do_update_inode consumes one bh->b_count */
get_bh(iloc->bh);
-
/* ext4_do_update_inode() does jbd2_journal_dirty_metadata */
err = ext4_do_update_inode(handle, inode, iloc);
put_bh(iloc->bh);
--
2.51.0
^ permalink raw reply related
* [PATCH v2 05/10] udf: Fix possibly missing inode write on fsync(2)
From: Jan Kara @ 2026-05-25 8:58 UTC (permalink / raw)
To: linux-fsdevel
Cc: Christian Brauner, aivazian.tigran, Ted Tso, linux-ext4,
OGAWA Hirofumi, Jan Kara
In-Reply-To: <20260525085035.12891-1-jack@suse.cz>
Use mmb inode buffer writeout infrastructure to reliably write out
inode's block on fsync(2).
Signed-off-by: Jan Kara <jack@suse.cz>
---
fs/udf/inode.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/fs/udf/inode.c b/fs/udf/inode.c
index 67bcf83758c8..861400152261 100644
--- a/fs/udf/inode.c
+++ b/fs/udf/inode.c
@@ -1707,7 +1707,7 @@ void udf_update_extra_perms(struct inode *inode, umode_t mode)
int udf_write_inode(struct inode *inode, struct writeback_control *wbc)
{
- return udf_update_inode(inode, wbc->sync_mode == WB_SYNC_ALL);
+ return udf_update_inode(inode, 0);
}
static int udf_sync_inode(struct inode *inode)
@@ -1936,7 +1936,7 @@ static int udf_update_inode(struct inode *inode, int do_sync)
unlock_buffer(bh);
/* write the data blocks */
- mark_buffer_dirty(bh);
+ mmb_mark_inode_buffer_dirty(bh, &iinfo->i_metadata_bhs);
if (do_sync) {
sync_dirty_buffer(bh);
if (buffer_write_io_error(bh)) {
--
2.51.0
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox