[PATCH v3 1/2] erofs: set error to bio if file-backed IO fails

[PATCH v3 1/2] erofs: set error to bio if file-backed IO fails

[Sheng Yong]

From: Sheng Yong <shengyong1@xiaomi.com>

If a file-backed IO fails before submitting the bio to the lower
filesystem, an error is returned, but the bio->bi_status is not
marked as an error. However, the error information should be passed
to the end_io handler. Otherwise, the IO request will be treated as
successful.

Signed-off-by: Sheng Yong <shengyong1@xiaomi.com>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
---
 fs/erofs/fileio.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/fs/erofs/fileio.c b/fs/erofs/fileio.c
index bec4b56b3826..4fa0a0121288 100644
--- a/fs/erofs/fileio.c
+++ b/fs/erofs/fileio.c
@@ -32,6 +32,8 @@ static void erofs_fileio_ki_complete(struct kiocb *iocb, long ret)
 		ret = 0;
 	}
 	if (rq->bio.bi_end_io) {
+		if (ret < 0 && !rq->bio.bi_status)
+			rq->bio.bi_status = errno_to_blk_status(ret);
 		rq->bio.bi_end_io(&rq->bio);
 	} else {
 		bio_for_each_folio_all(fi, &rq->bio) {
-- 
2.43.0

[PATCH v3 2/2] erofs: add 'offset' mount option for file-backed & bdev-based mounts

[Sheng Yong]

From: Sheng Yong <shengyong1@xiaomi.com>

When attempting to use an archive file, such as APEX on android,
as a file-backed mount source, it fails because EROFS image within
the archive file does not start at offset 0. As a result, a loop
device is still needed to attach the image file at an appropriate
offset first. Similarly, if an EROFS image within a block device
does not start at offset 0, it cannot be mounted directly either.

To address this issue, this patch adds a new mount option `offset=x'
to accept a start offset for both file-backed and bdev-based mounts.
The offset should be aligned to block size. EROFS will add this offset
before performing read requests.

Signed-off-by: Sheng Yong <shengyong1@xiaomi.com>
Signed-off-by: Wang Shuai <wangshuai12@xiaomi.com>
---
 Documentation/filesystems/erofs.rst |  1 +
 fs/erofs/data.c                     |  8 ++++++--
 fs/erofs/fileio.c                   |  4 +++-
 fs/erofs/internal.h                 |  2 ++
 fs/erofs/super.c                    | 24 +++++++++++++++++++++++-
 fs/erofs/zdata.c                    | 22 ++++++++++++++--------
 6 files changed, 49 insertions(+), 12 deletions(-)
---
v3: * rename `offs' to `off'
    * parse offset using fsparam_u64 and validate it in fill_super
    * update bi_sector inline

v2: * add a new mount option `offset=X' for start offset, and offset
       should be aligned to PAGE_SIZE
    * add start offset for both file-backed and bdev-based mounts
    https://lore.kernel.org/linux-erofs/0725c2ec-528c-42a8-9557-4713e7e35153@linux.alibaba.com

v1: https://lore.kernel.org/all/20250324022849.2715578-1-shengyong1@xiaomi.com/

diff --git a/Documentation/filesystems/erofs.rst b/Documentation/filesystems/erofs.rst
index c293f8e37468..44dbfa6cffb1 100644
--- a/Documentation/filesystems/erofs.rst
+++ b/Documentation/filesystems/erofs.rst
@@ -128,6 +128,7 @@ device=%s              Specify a path to an extra device to be used together.
 fsid=%s                Specify a filesystem image ID for Fscache back-end.
 domain_id=%s           Specify a domain ID in fscache mode so that different images
                        with the same blobs under a given domain ID can share storage.
+offset=%s              Specify image offset for file-backed or bdev-based mounts.
 ===================    =========================================================
 
 Sysfs Entries
diff --git a/fs/erofs/data.c b/fs/erofs/data.c
index 2409d2ab0c28..7da503480f4d 100644
--- a/fs/erofs/data.c
+++ b/fs/erofs/data.c
@@ -27,9 +27,12 @@ void erofs_put_metabuf(struct erofs_buf *buf)
 
 void *erofs_bread(struct erofs_buf *buf, erofs_off_t offset, bool need_kmap)
 {
-	pgoff_t index = offset >> PAGE_SHIFT;
+	pgoff_t index;
 	struct folio *folio = NULL;
 
+	offset += buf->off;
+	index = offset >> PAGE_SHIFT;
+
 	if (buf->page) {
 		folio = page_folio(buf->page);
 		if (folio_file_page(folio, index) != buf->page)
@@ -54,6 +57,7 @@ void erofs_init_metabuf(struct erofs_buf *buf, struct super_block *sb)
 	struct erofs_sb_info *sbi = EROFS_SB(sb);
 
 	buf->file = NULL;
+	buf->off = sbi->dif0.off;
 	if (erofs_is_fileio_mode(sbi)) {
 		buf->file = sbi->dif0.file;	/* some fs like FUSE needs it */
 		buf->mapping = buf->file->f_mapping;
@@ -299,7 +303,7 @@ static int erofs_iomap_begin(struct inode *inode, loff_t offset, loff_t length,
 		iomap->private = buf.base;
 	} else {
 		iomap->type = IOMAP_MAPPED;
-		iomap->addr = mdev.m_pa;
+		iomap->addr = EROFS_SB(sb)->dif0.off + mdev.m_pa;
 		if (flags & IOMAP_DAX)
 			iomap->addr += mdev.m_dif->dax_part_off;
 	}
diff --git a/fs/erofs/fileio.c b/fs/erofs/fileio.c
index 4fa0a0121288..2c003cbb0fbb 100644
--- a/fs/erofs/fileio.c
+++ b/fs/erofs/fileio.c
@@ -52,7 +52,9 @@ static void erofs_fileio_rq_submit(struct erofs_fileio_rq *rq)
 
 	if (!rq)
 		return;
-	rq->iocb.ki_pos = rq->bio.bi_iter.bi_sector << SECTOR_SHIFT;
+
+	rq->iocb.ki_pos = EROFS_SB(rq->sb)->dif0.off +
+				(rq->bio.bi_iter.bi_sector << SECTOR_SHIFT);
 	rq->iocb.ki_ioprio = get_current_ioprio();
 	rq->iocb.ki_complete = erofs_fileio_ki_complete;
 	if (test_opt(&EROFS_SB(rq->sb)->opt, DIRECT_IO) &&
diff --git a/fs/erofs/internal.h b/fs/erofs/internal.h
index 4ac188d5d894..10656bd986bd 100644
--- a/fs/erofs/internal.h
+++ b/fs/erofs/internal.h
@@ -43,6 +43,7 @@ struct erofs_device_info {
 	char *path;
 	struct erofs_fscache *fscache;
 	struct file *file;
+	loff_t off;
 	struct dax_device *dax_dev;
 	u64 dax_part_off;
 
@@ -199,6 +200,7 @@ enum {
 struct erofs_buf {
 	struct address_space *mapping;
 	struct file *file;
+	loff_t off;
 	struct page *page;
 	void *base;
 };
diff --git a/fs/erofs/super.c b/fs/erofs/super.c
index cadec6b1b554..bd2a2c634f1d 100644
--- a/fs/erofs/super.c
+++ b/fs/erofs/super.c
@@ -356,7 +356,7 @@ static void erofs_default_options(struct erofs_sb_info *sbi)
 
 enum {
 	Opt_user_xattr, Opt_acl, Opt_cache_strategy, Opt_dax, Opt_dax_enum,
-	Opt_device, Opt_fsid, Opt_domain_id, Opt_directio,
+	Opt_device, Opt_fsid, Opt_domain_id, Opt_directio, Opt_offset,
 	Opt_err
 };
 
@@ -384,6 +384,7 @@ static const struct fs_parameter_spec erofs_fs_parameters[] = {
 	fsparam_string("fsid",		Opt_fsid),
 	fsparam_string("domain_id",	Opt_domain_id),
 	fsparam_flag_no("directio",	Opt_directio),
+	fsparam_u64("offset",		Opt_offset),
 	{}
 };
 
@@ -507,6 +508,9 @@ static int erofs_fc_parse_param(struct fs_context *fc,
 		errorfc(fc, "%s option not supported", erofs_fs_parameters[opt].name);
 #endif
 		break;
+	case Opt_offset:
+		sbi->dif0.off = result.uint_64;
+		break;
 	}
 	return 0;
 }
@@ -600,6 +604,22 @@ static int erofs_fc_fill_super(struct super_block *sb, struct fs_context *fc)
 				&sbi->dif0.dax_part_off, NULL, NULL);
 	}
 
+	if (sbi->dif0.off) {
+		loff_t devsz;
+
+		if (sbi->dif0.off & ((1 << sbi->blkszbits) - 1))
+			return invalfc(fc, "offset %lld not aligned to block size",
+				       sbi->dif0.off);
+		if (sb->s_bdev)
+			devsz = bdev_nr_bytes(sb->s_bdev);
+		else if (erofs_is_fileio_mode(sbi))
+			devsz = i_size_read(file_inode(sbi->dif0.file));
+		else
+			return invalfc(fc, "offset only supports file or bdev backing");
+		if (sbi->dif0.off + (1 << sbi->blkszbits) > devsz)
+			return invalfc(fc, "offset exceeds device size");
+	}
+
 	err = erofs_read_superblock(sb);
 	if (err)
 		return err;
@@ -948,6 +968,8 @@ static int erofs_show_options(struct seq_file *seq, struct dentry *root)
 	if (sbi->domain_id)
 		seq_printf(seq, ",domain_id=%s", sbi->domain_id);
 #endif
+	if (sbi->dif0.off)
+		seq_printf(seq, ",offset=%lld", sbi->dif0.off);
 	return 0;
 }
 
diff --git a/fs/erofs/zdata.c b/fs/erofs/zdata.c
index 0671184d9cf1..671527b63c6d 100644
--- a/fs/erofs/zdata.c
+++ b/fs/erofs/zdata.c
@@ -1624,7 +1624,8 @@ static void z_erofs_submit_queue(struct z_erofs_frontend *f,
 				 bool *force_fg, bool readahead)
 {
 	struct super_block *sb = f->inode->i_sb;
-	struct address_space *mc = MNGD_MAPPING(EROFS_SB(sb));
+	struct erofs_sb_info *sbi = EROFS_SB(sb);
+	struct address_space *mc = MNGD_MAPPING(sbi);
 	struct z_erofs_pcluster **qtail[NR_JOBQUEUES];
 	struct z_erofs_decompressqueue *q[NR_JOBQUEUES];
 	struct z_erofs_pcluster *pcl, *next;
@@ -1673,12 +1674,15 @@ static void z_erofs_submit_queue(struct z_erofs_frontend *f,
 			if (bio && (cur != last_pa ||
 				    bio->bi_bdev != mdev.m_bdev)) {
 drain_io:
-				if (erofs_is_fileio_mode(EROFS_SB(sb)))
+				if (erofs_is_fileio_mode(sbi)) {
 					erofs_fileio_submit_bio(bio);
-				else if (erofs_is_fscache_mode(sb))
+				} else if (erofs_is_fscache_mode(sb)) {
 					erofs_fscache_submit_bio(bio);
-				else
+				} else {
+					bio->bi_iter.bi_sector +=
+						sbi->dif0.off >> SECTOR_SHIFT;
 					submit_bio(bio);
+				}
 
 				if (memstall) {
 					psi_memstall_leave(&pflags);
@@ -1703,7 +1707,7 @@ static void z_erofs_submit_queue(struct z_erofs_frontend *f,
 			}
 
 			if (!bio) {
-				if (erofs_is_fileio_mode(EROFS_SB(sb)))
+				if (erofs_is_fileio_mode(sbi))
 					bio = erofs_fileio_bio_alloc(&mdev);
 				else if (erofs_is_fscache_mode(sb))
 					bio = erofs_fscache_bio_alloc(&mdev);
@@ -1732,12 +1736,14 @@ static void z_erofs_submit_queue(struct z_erofs_frontend *f,
 	} while (next != Z_EROFS_PCLUSTER_TAIL);
 
 	if (bio) {
-		if (erofs_is_fileio_mode(EROFS_SB(sb)))
+		if (erofs_is_fileio_mode(sbi)) {
 			erofs_fileio_submit_bio(bio);
-		else if (erofs_is_fscache_mode(sb))
+		} else if (erofs_is_fscache_mode(sb)) {
 			erofs_fscache_submit_bio(bio);
-		else
+		} else {
+			bio->bi_iter.bi_sector += sbi->dif0.off >> SECTOR_SHIFT;
 			submit_bio(bio);
+		}
 	}
 	if (memstall)
 		psi_memstall_leave(&pflags);
-- 
2.43.0

Re: [PATCH v3 2/2] erofs: add 'offset' mount option for file-backed & bdev-based mounts

[Karel Zak]

On Mon, Apr 07, 2025 at 07:05:51PM +0800, Sheng Yong wrote:
> From: Sheng Yong <shengyong1@xiaomi.com>
> 
> When attempting to use an archive file, such as APEX on android,
> as a file-backed mount source, it fails because EROFS image within
> the archive file does not start at offset 0. As a result, a loop
> device is still needed to attach the image file at an appropriate
> offset first. Similarly, if an EROFS image within a block device
> does not start at offset 0, it cannot be mounted directly either.

Does it work with mount(8)? The mount option offset= has been defined
for decades as userspace-only and is used for loop devices. If I
remember correctly, libmount does not send the option to the kernel at
all. The option also triggers loop device usage by mount(8).

In recent years, we use the "X-" prefix for userspace options.
Unfortunately, loop=, offset=, and sizelimit= are older than any
currently used convention (I see the option in mount code from year
1998).

We can improve it in libmount and add any if-erofs hack there, but my
suggestion is to select a better name for the mount option. For
example, erofsoff=, erostart=, fsoffset=, start=, or similar.

    Karel


-- 
 Karel Zak  <kzak@redhat.com>
 http://karelzak.blogspot.com

Re: [PATCH v3 2/2] erofs: add 'offset' mount option for file-backed & bdev-based mounts

[Gao Xiang]

Hi Karel,

On 2025/4/7 19:40, Karel Zak wrote:
> On Mon, Apr 07, 2025 at 07:05:51PM +0800, Sheng Yong wrote:
>> From: Sheng Yong <shengyong1@xiaomi.com>
>>
>> When attempting to use an archive file, such as APEX on android,
>> as a file-backed mount source, it fails because EROFS image within
>> the archive file does not start at offset 0. As a result, a loop
>> device is still needed to attach the image file at an appropriate
>> offset first. Similarly, if an EROFS image within a block device
>> does not start at offset 0, it cannot be mounted directly either.
> 
> Does it work with mount(8)? The mount option offset= has been defined
> for decades as userspace-only and is used for loop devices. If I
> remember correctly, libmount does not send the option to the kernel at
> all. The option also triggers loop device usage by mount(8).
> 
> In recent years, we use the "X-" prefix for userspace options.
> Unfortunately, loop=, offset=, and sizelimit= are older than any
> currently used convention (I see the option in mount code from year
> 1998).
> 
> We can improve it in libmount and add any if-erofs hack there, but my
> suggestion is to select a better name for the mount option. For
> example, erofsoff=, erostart=, fsoffset=, start=, or similar.

Thanks for your suggestion!

it's somewhat weird to use erofsprefix here, I think fsoffset
may be fine.

Thanks,
Gao Xiang

> 
>      Karel
> 
> 

Re: [PATCH v3 2/2] erofs: add 'offset' mount option for file-backed & bdev-based mounts

[Karel Zak]

On Mon, Apr 07, 2025 at 11:49:31PM +0800, Gao Xiang wrote:
> On 2025/4/7 19:40, Karel Zak wrote:
> > We can improve it in libmount and add any if-erofs hack there, but my
> > suggestion is to select a better name for the mount option. For
> > example, erofsoff=, erostart=, fsoffset=, start=, or similar.
> 
> Thanks for your suggestion!
> 
> it's somewhat weird to use erofsprefix here, I think fsoffset
> may be fine.

Yes, fsoffset sounds good. I anticipate more filesystems will support
file-backed mounts in the future, making this option reusable.

    Karel

-- 
 Karel Zak  <kzak@redhat.com>
 http://karelzak.blogspot.com

Re: [PATCH v3 2/2] erofs: add 'offset' mount option for file-backed & bdev-based mounts

[Gao Xiang]

On 2025/4/8 16:46, Karel Zak wrote:
> On Mon, Apr 07, 2025 at 11:49:31PM +0800, Gao Xiang wrote:
>> On 2025/4/7 19:40, Karel Zak wrote:
>>> We can improve it in libmount and add any if-erofs hack there, but my
>>> suggestion is to select a better name for the mount option. For
>>> example, erofsoff=, erostart=, fsoffset=, start=, or similar.
>>
>> Thanks for your suggestion!
>>
>> it's somewhat weird to use erofsprefix here, I think fsoffset
>> may be fine.
> 
> Yes, fsoffset sounds good. I anticipate more filesystems will support
> file-backed mounts in the future, making this option reusable.

 From my own kernel perspective, it's not possible for the majority
of read-write fses especially have nested transaction (like
`current->journal_info`) in addition to many deadlock factors
without extra workqueue contexts.  But for erofs it's safe.

Thanks,
Gao Xiang