For many use cases (e.g. container images are just fetched from remote),
performance will be impacted if underlay page cache is up-to-date but
direct i/o flushes dirty pages first.
Instead, let's use buffered I/O by default to keep in sync with loop
devices and add a (re)mount option to explicitly give a try to use
direct I/O if supported by the underlying files.
The container startup time is improved as below:
[workload] docker.io/library/workpress:latest
unpack 1st run non-1st runs
EROFS snapshotter buffered I/O file 4.586404265s 0.308s 0.198s
EROFS snapshotter direct I/O file 4.581742849s 2.238s 0.222s
EROFS snapshotter loop 4.596023152s 0.346s 0.201s
Overlayfs snapshotter 5.382851037s 0.206s 0.214s
Fixes: fb176750266a ("erofs: add file-backed mount support")
Cc: Derek McGowan <de...@mcg.dev>
Reviewed-by: Chao Yu <c...@kernel.org>
Signed-off-by: Gao Xiang <hsiang...@linux.alibaba.com>
---
fs/erofs/fileio.c | 7 +++++--
fs/erofs/internal.h | 1 +
fs/erofs/super.c | 23 +++++++++++++++--------
3 files changed, 21 insertions(+), 10 deletions(-)
diff --git a/fs/erofs/fileio.c b/fs/erofs/fileio.c
index a61b8faec651..33f8539dda4a 100644
--- a/fs/erofs/fileio.c
+++ b/fs/erofs/fileio.c
@@ -9,6 +9,7 @@ struct erofs_fileio_rq {
struct bio_vec bvecs[BIO_MAX_VECS];
struct bio bio;
struct kiocb iocb;
+ struct super_block *sb;
};
struct erofs_fileio {
@@ -52,8 +53,9 @@ static void erofs_fileio_rq_submit(struct erofs_fileio_rq *rq)
rq->iocb.ki_pos = rq->bio.bi_iter.bi_sector << SECTOR_SHIFT;
rq->iocb.ki_ioprio = get_current_ioprio();
rq->iocb.ki_complete = erofs_fileio_ki_complete;
- rq->iocb.ki_flags = (rq->iocb.ki_filp->f_mode & FMODE_CAN_ODIRECT) ?
- IOCB_DIRECT : 0;
+ if (test_opt(&EROFS_SB(rq->sb)->opt, DIRECT_IO) &&
+ rq->iocb.ki_filp->f_mode & FMODE_CAN_ODIRECT)
+ rq->iocb.ki_flags = IOCB_DIRECT;
iov_iter_bvec(&iter, ITER_DEST, rq->bvecs, rq->bio.bi_vcnt,
rq->bio.bi_iter.bi_size);
ret = vfs_iocb_iter_read(rq->iocb.ki_filp, &rq->iocb, &iter);
@@ -68,6 +70,7 @@ static struct erofs_fileio_rq *erofs_fileio_rq_alloc(struct
erofs_map_dev *mdev)
bio_init(&rq->bio, NULL, rq->bvecs, BIO_MAX_VECS, REQ_OP_READ);
rq->iocb.ki_filp = mdev->m_dif->file;
+ rq->sb = mdev->m_sb;
return rq;
}
diff --git a/fs/erofs/internal.h b/fs/erofs/internal.h
index 7cc8e1be04e8..686d835eb533 100644
--- a/fs/erofs/internal.h
+++ b/fs/erofs/internal.h
@@ -176,6 +176,7 @@ struct erofs_sb_info {
#define EROFS_MOUNT_POSIX_ACL 0x00000020
#define EROFS_MOUNT_DAX_ALWAYS 0x00000040
#define EROFS_MOUNT_DAX_NEVER 0x00000080
+#define EROFS_MOUNT_DIRECT_IO 0x00000100
#define clear_opt(opt, option) ((opt)->mount_opt &= ~EROFS_MOUNT_##option)
#define set_opt(opt, option) ((opt)->mount_opt |= EROFS_MOUNT_##option)
diff --git a/fs/erofs/super.c b/fs/erofs/super.c
index 9044907354e1..f5956474bfde 100644
--- a/fs/erofs/super.c
+++ b/fs/erofs/super.c
@@ -364,14 +364,8 @@ 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_user_xattr, Opt_acl, Opt_cache_strategy, Opt_dax, Opt_dax_enum,
+ Opt_device, Opt_fsid, Opt_domain_id, Opt_directio,
Opt_err
};
@@ -398,6 +392,7 @@ static const struct fs_parameter_spec erofs_fs_parameters[]
= {
fsparam_string("device", Opt_device),
fsparam_string("fsid", Opt_fsid),
fsparam_string("domain_id", Opt_domain_id),
+ fsparam_flag_no("directio", Opt_directio),
{}
};
@@ -511,6 +506,16 @@ static int erofs_fc_parse_param(struct fs_context *fc,
errorfc(fc, "%s option not supported",
erofs_fs_parameters[opt].name);
break;
#endif
+ case Opt_directio:
+#ifdef CONFIG_EROFS_FS_BACKED_BY_FILE
+ if (result.boolean)
+ set_opt(&sbi->opt, DIRECT_IO);
+ else
+ clear_opt(&sbi->opt, DIRECT_IO);
+#else
+ errorfc(fc, "%s option not supported",
erofs_fs_parameters[opt].name);
+#endif
+ break;
default:
return -ENOPARAM;
}
@@ -948,6 +953,8 @@ static int erofs_show_options(struct seq_file *seq, struct
dentry *root)
seq_puts(seq, ",dax=always");
if (test_opt(opt, DAX_NEVER))
seq_puts(seq, ",dax=never");
+ if (erofs_is_fileio_mode(sbi) && test_opt(opt, DIRECT_IO))
+ seq_puts(seq, ",directio");
#ifdef CONFIG_EROFS_FS_ONDEMAND
if (sbi->fsid)
seq_printf(seq, ",fsid=%s", sbi->fsid);
--
2.43.5
