BACKPORT: erofs: support to readahead dirent blocks in erofs_readdir()

This patch supports to readahead more blocks in erofs_readdir(), it can
enhance readdir performance in large direcotry.

readdir test in a large directory which contains 12000 sub-files.

files_per_second Before: 926385.54 After: 2380435.562

Meanwhile, let's introduces a new sysfs entry to control readahead bytes
to provide more flexible policy for readahead of readdir().
- location: /sys/fs/erofs/<disk>/dir_ra_bytes
- default value: 16384
- disable readahead: set the value to 0

Bug: 407861253
(cherry picked from commit df0ce6c)
Change-Id: I0b52529cebce61f598dda33244ae257d5f5e1320
[Chao: resolve the conflict]
Signed-off-by: Chao Yu <chao@kernel.org>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Link: https://lore.kernel.org/r/20250721021352.2495371-1-chao@kernel.org
[ Gao Xiang: minor styling adjustment. ]
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Signed-off-by: Chao Yu <chao@kernel.org>

Author: chaseyu

Documentation/ABI/testing/sysfs-fs-erofs

@@ -16,3 +16,11 @@ Description:	Control strategy of sync decompression:
 		  readahead on atomic contexts only.
 		- 1 (force on): enable for readpage and readahead.
 		- 2 (force off): disable for all situations.
+
+What:		/sys/fs/erofs/<disk>/dir_ra_bytes
+Date:		July 2025
+Contact:	"Chao Yu" <chao@kernel.org>
+Description:	Used to set or show readahead bytes during readdir(), by
+		default the value is 16384.
+
+		- 0: disable readahead.

fs/erofs/dir.c

@@ -47,8 +47,12 @@ static int erofs_readdir(struct file *f, struct dir_context *ctx)
 	struct inode *dir = file_inode(f);
 	struct erofs_buf buf = __EROFS_BUF_INITIALIZER;
 	struct super_block *sb = dir->i_sb;
+	struct file_ra_state *ra = &f->f_ra;
 	unsigned long bsz = sb->s_blocksize;
 	unsigned int ofs = erofs_blkoff(sb, ctx->pos);
+	pgoff_t ra_pages = DIV_ROUND_UP_POW2(
+			EROFS_I_SB(dir)->dir_ra_bytes, PAGE_SIZE);
+	pgoff_t nr_pages = DIV_ROUND_UP_POW2(dir->i_size, PAGE_SIZE);
 	int err = 0;
 	bool initial = true;
 
@@ -58,6 +62,16 @@ static int erofs_readdir(struct file *f, struct dir_context *ctx)
 		struct erofs_dirent *de;
 		unsigned int nameoff, maxsize;
 
+		/* readahead blocks to enhance performance for large directories */
+		if (ra_pages) {
+			pgoff_t idx = DIV_ROUND_UP_POW2(ctx->pos, PAGE_SIZE);
+			pgoff_t pages = min(nr_pages - idx, ra_pages);
+
+			if (pages > 1 && !ra_has_index(ra, idx))
+				page_cache_sync_readahead(dir->i_mapping, ra,
+							  f, idx, pages);
+		}
+
 		de = erofs_bread(&buf, dbstart, EROFS_KMAP);
 		if (IS_ERR(de)) {
 			erofs_err(sb, "fail to readdir of logical block %u of nid %llu",

fs/erofs/internal.h

@@ -166,6 +166,7 @@ struct erofs_sb_info {
 	/* sysfs support */
 	struct kobject s_kobj;		/* /sys/fs/erofs/<devname> */
 	struct completion s_kobj_unregister;
+	erofs_off_t dir_ra_bytes;
 
 	/* fscache support */
 	struct fscache_volume *volume;
@@ -252,6 +253,9 @@ EROFS_FEATURE_FUNCS(xattr_filter, compat, COMPAT_XATTR_FILTER)
 #define EROFS_I_BL_XATTR_BIT	(BITS_PER_LONG - 1)
 #define EROFS_I_BL_Z_BIT	(BITS_PER_LONG - 2)
 
+/* default readahead size of directories */
+#define EROFS_DIR_RA_BYTES	16384
+
 struct erofs_inode {
 	erofs_nid_t nid;
 

fs/erofs/super.c

@@ -742,6 +742,7 @@ static int erofs_fc_fill_super(struct super_block *sb, struct fs_context *fc)
 	if (err)
 		return err;
 
+	sbi->dir_ra_bytes = EROFS_DIR_RA_BYTES;
 	erofs_info(sb, "mounted with root inode @ nid %llu.", sbi->root_nid);
 	return 0;
 }

fs/erofs/sysfs.c

@@ -58,11 +58,13 @@ static struct erofs_attr erofs_attr_##_name = {			\
 #ifdef CONFIG_EROFS_FS_ZIP
 EROFS_ATTR_RW_UI(sync_decompress, erofs_mount_opts);
 #endif
+EROFS_ATTR_RW_UI(dir_ra_bytes, erofs_sb_info);
 
 static struct attribute *erofs_attrs[] = {
 #ifdef CONFIG_EROFS_FS_ZIP
 	ATTR_LIST(sync_decompress),
 #endif
+	ATTR_LIST(dir_ra_bytes),
 	NULL,
 };
 ATTRIBUTE_GROUPS(erofs);