Re: [PATCH/RFC 00/11] expose btrfs subvols in mount table correctly

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 





On 2021/7/30 下午1:25, Qu Wenruo wrote:


On 2021/7/30 上午10:36, NeilBrown wrote:

I've been pondering all the excellent feedback, and what I have learnt
from examining the code in btrfs, and I have developed a different
perspective.

Great! Some new developers into the btrfs realm!


Maybe "subvol" is a poor choice of name because it conjures up
connections with the Volumes in LVM, and btrfs subvols are very different
things.  Btrfs subvols are really just subtrees that can be treated as a
unit for operations like "clone" or "destroy".

As such, they don't really deserve separate st_dev numbers.

Maybe the different st_dev numbers were introduced as a "cheap" way to
extend to size of the inode-number space.  Like many "cheap" things, it
has hidden costs.

Forgot another problem already caused by this st_dev method.

Since btrfs uses st_dev to distinguish them its inode name space, and
st_dev is allocated using anonymous bdev, and the anonymous bdev poor
has limited size (much smaller than btrfs subvolume id name space), it's
already causing problems like we can't allocate enough anonymous bdev
for each subvolume, and failed to create subvolume/snapshot.

Thus it's really a time to re-consider how we should export this info to
user space.

Thanks,
Qu


Maybe objects in different subvols should still be given different inode
numbers.  This would be problematic on 32bit systems, but much less so on
64bit systems.

The patch below, which is just a proof-of-concept, changes btrfs to
report a uniform st_dev, and different (64bit) st_ino in different
subvols.

It has problems:
  - it will break any 32bit readdir and 32bit stat.  I don't know how big
    a problem that is these days (ino_t in the kernel is "unsigned long",
    not "unsigned long long). That surprised me).
  - It might break some user-space expectations.  One thing I have learnt
    is not to make any assumption about what other people might expect.

Wouldn't any filesystem boundary check fail to stop at subvolume boundary?

Then it will go through the full btrfs subvolumes/snapshots, which can
be super slow.


However, it would be quite easy to make this opt-in (or opt-out) with a
mount option, so that people who need the current inode numbers and will
accept the current breakage can keep working.

I think this approach would be a net-win for NFS export, whether BTRFS
supports it directly or not.  I might post a patch which modifies NFS to
intuit improved inode numbers for btrfs exports....

Some extra ideas, but not familiar with VFS enough to be sure.

Can we generate "fake" superblock for each subvolume?
Like using the subolume UUID to replace the FSID of each subvolume.
Could that migrate the problem?

Thanks,
Qu


So: how would this break your use-case??

Thanks,
NeilBrown

diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 0117d867ecf8..8dc58c848502 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -6020,6 +6020,37 @@ static int btrfs_opendir(struct inode *inode,
struct file *file)
      return 0;
  }

+static u64 btrfs_make_inum(struct btrfs_key *root, struct btrfs_key
*ino)
+{
+    u64 rootid = root->objectid;
+    u64 inoid = ino->objectid;
+    int shift = 64-8;
+
+    if (ino->type == BTRFS_ROOT_ITEM_KEY) {
+        /* This is a subvol root found during readdir. */
+        rootid = inoid;
+        inoid = BTRFS_FIRST_FREE_OBJECTID;
+    }
+    if (rootid == BTRFS_FS_TREE_OBJECTID)
+        /* this is main vol, not subvol (I think) */
+        return inoid;
+    /* store the rootid in the high bits of the inum.  This
+     * will break if 32bit inums are required - we cannot know
+     */
+    while (rootid) {
+        inoid ^= (rootid & 0xff) << shift;
+        rootid >>= 8;
+        shift -= 8;
+    }
+    return inoid;
+}
+
+static inline u64 btrfs_ino_to_inum(struct inode *inode)
+{
+    return btrfs_make_inum(&BTRFS_I(inode)->root->root_key,
+                   &BTRFS_I(inode)->location);
+}
+
  struct dir_entry {
      u64 ino;
      u64 offset;
@@ -6045,6 +6076,49 @@ static int btrfs_filldir(void *addr, int
entries, struct dir_context *ctx)
      return 0;
  }

+static inline bool btrfs_dir_emit_dot(struct file *file,
+                      struct dir_context *ctx)
+{
+    return ctx->actor(ctx, ".", 1, ctx->pos,
+              btrfs_ino_to_inum(file->f_path.dentry->d_inode),
+              DT_DIR) == 0;
+}
+
+static inline ino_t btrfs_parent_ino(struct dentry *dentry)
+{
+    ino_t res;
+
+    /*
+     * Don't strictly need d_lock here? If the parent ino could change
+     * then surely we'd have a deeper race in the caller?
+     */
+    spin_lock(&dentry->d_lock);
+    res = btrfs_ino_to_inum(dentry->d_parent->d_inode);
+    spin_unlock(&dentry->d_lock);
+    return res;
+}
+
+static inline bool btrfs_dir_emit_dotdot(struct file *file,
+                     struct dir_context *ctx)
+{
+    return ctx->actor(ctx, "..", 2, ctx->pos,
+              btrfs_parent_ino(file->f_path.dentry), DT_DIR) == 0;
+}
+static inline bool btrfs_dir_emit_dots(struct file *file,
+                       struct dir_context *ctx)
+{
+    if (ctx->pos == 0) {
+        if (!btrfs_dir_emit_dot(file, ctx))
+            return false;
+        ctx->pos = 1;
+    }
+    if (ctx->pos == 1) {
+        if (!btrfs_dir_emit_dotdot(file, ctx))
+            return false;
+        ctx->pos = 2;
+    }
+    return true;
+}
  static int btrfs_real_readdir(struct file *file, struct dir_context
*ctx)
  {
      struct inode *inode = file_inode(file);
@@ -6067,7 +6141,7 @@ static int btrfs_real_readdir(struct file *file,
struct dir_context *ctx)
      bool put = false;
      struct btrfs_key location;

-    if (!dir_emit_dots(file, ctx))
+    if (!btrfs_dir_emit_dots(file, ctx))
          return 0;

      path = btrfs_alloc_path();
@@ -6136,7 +6210,8 @@ static int btrfs_real_readdir(struct file *file,
struct dir_context *ctx)
          put_unaligned(fs_ftype_to_dtype(btrfs_dir_type(leaf, di)),
                  &entry->type);
          btrfs_dir_item_key_to_cpu(leaf, di, &location);
-        put_unaligned(location.objectid, &entry->ino);
+        put_unaligned(btrfs_make_inum(&root->root_key, &location),
+                  &entry->ino);
          put_unaligned(found_key.offset, &entry->offset);
          entries++;
          addr += sizeof(struct dir_entry) + name_len;
@@ -9193,7 +9268,7 @@ static int btrfs_getattr(struct user_namespace
*mnt_userns,
                    STATX_ATTR_NODUMP);

      generic_fillattr(&init_user_ns, inode, stat);
-    stat->dev = BTRFS_I(inode)->root->anon_dev;
+    stat->ino = btrfs_ino_to_inum(inode);

      spin_lock(&BTRFS_I(inode)->lock);
      delalloc_bytes = BTRFS_I(inode)->new_delalloc_bytes;





[Index of Archives]     [Linux Filesystem Development]     [Linux USB Development]     [Linux Media Development]     [Video for Linux]     [Linux NILFS]     [Linux Audio Users]     [Yosemite Info]     [Linux SCSI]

  Powered by Linux