[PATCH] NFSv4: Always set NLINK even if the server doesn't support it

[PATCH] NFSv4: Always set NLINK even if the server doesn't support it

[Han Young]

fattr4_numlinks is a recommended attribute, so the client should emulate
it even if the server doesn't support it. In decode_attr_nlink function
in nfs4xdr.c, nlink is initialized to 1. However, this default value
isn't set to the inode due to the check in nfs_fhget.

So if the server doesn't support numlinks, inode's nlink will be zero,
the mount will fail with error "Stale file handle". Change the check in
nfs_fhget so that the nlink value is always set.

Signed-off-by: Han Young <hanyang.tony@bytedance.com>
---
 fs/nfs/inode.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index 119e447758b9..c19f135b5041 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -553,10 +553,11 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr)
 			inode->i_size = nfs_size_to_loff_t(fattr->size);
 		else
 			nfs_set_cache_invalid(inode, NFS_INO_INVALID_SIZE);
-		if (fattr->valid & NFS_ATTR_FATTR_NLINK)
-			set_nlink(inode, fattr->nlink);
-		else if (fattr_supported & NFS_ATTR_FATTR_NLINK)
+		if (!(fattr->valid & NFS_ATTR_FATTR_NLINK) &&
+			   fattr_supported & NFS_ATTR_FATTR_NLINK)
 			nfs_set_cache_invalid(inode, NFS_INO_INVALID_NLINK);
+		else
+			set_nlink(inode, fattr->nlink);
 		if (fattr->valid & NFS_ATTR_FATTR_OWNER)
 			inode->i_uid = fattr->uid;
 		else if (fattr_supported & NFS_ATTR_FATTR_OWNER)
-- 
2.48.1

Re: [PATCH] NFSv4: Always set NLINK even if the server doesn't support it

[Trond Myklebust]

On Fri, 2025-05-02 at 23:15 +0800, Han Young wrote:
> fattr4_numlinks is a recommended attribute, so the client should
> emulate
> it even if the server doesn't support it. In decode_attr_nlink
> function
> in nfs4xdr.c, nlink is initialized to 1. However, this default value
> isn't set to the inode due to the check in nfs_fhget.
> 
> So if the server doesn't support numlinks, inode's nlink will be
> zero,
> the mount will fail with error "Stale file handle". Change the check
> in
> nfs_fhget so that the nlink value is always set.
> 
> Signed-off-by: Han Young <hanyang.tony@bytedance.com>
> ---
>  fs/nfs/inode.c | 7 ++++---
>  1 file changed, 4 insertions(+), 3 deletions(-)
> 
> diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
> index 119e447758b9..c19f135b5041 100644
> --- a/fs/nfs/inode.c
> +++ b/fs/nfs/inode.c
> @@ -553,10 +553,11 @@ nfs_fhget(struct super_block *sb, struct nfs_fh
> *fh, struct nfs_fattr *fattr)
>  			inode->i_size = nfs_size_to_loff_t(fattr-
> >size);
>  		else
>  			nfs_set_cache_invalid(inode,
> NFS_INO_INVALID_SIZE);
> -		if (fattr->valid & NFS_ATTR_FATTR_NLINK)
> -			set_nlink(inode, fattr->nlink);
> -		else if (fattr_supported & NFS_ATTR_FATTR_NLINK)
> +		if (!(fattr->valid & NFS_ATTR_FATTR_NLINK) &&
> +			   fattr_supported & NFS_ATTR_FATTR_NLINK)
>  			nfs_set_cache_invalid(inode,
> NFS_INO_INVALID_NLINK);
> +		else
> +			set_nlink(inode, fattr->nlink);

No. You can't use the value of fattr->nlink when that value is
undefined.

>  		if (fattr->valid & NFS_ATTR_FATTR_OWNER)
>  			inode->i_uid = fattr->uid;
>  		else if (fattr_supported & NFS_ATTR_FATTR_OWNER)

-- 
Trond Myklebust
Linux NFS client maintainer, Hammerspace
trond.myklebust@hammerspace.com

[PATCH v2] NFSv4: Always set NLINK even if the server doesn't support it

[Han Young]

fattr4_numlinks is a recommended attribute, so the client should emulate
it even if the server doesn't support it. In decode_attr_nlink function
in nfs4xdr.c, nlink is initialized to 1. However, this default value
isn't set to the inode due to the check in nfs_fhget.

So if the server doesn't support numlinks, inode's nlink will be zero,
the mount will fail with error "Stale file handle". Set the nlink to 1
if the server doesn't support it.

Signed-off-by: Han Young <hanyang.tony@bytedance.com>
---
 fs/nfs/inode.c | 2 ++
 1 file changed, 2 insertions(+)

Changes since v1:
set the nlink to 1 in the else, instead of relying on fattr->nlink,
which is uninitialized.

diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index 119e447758b9..4695292378bb 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -557,6 +557,8 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr)
 			set_nlink(inode, fattr->nlink);
 		else if (fattr_supported & NFS_ATTR_FATTR_NLINK)
 			nfs_set_cache_invalid(inode, NFS_INO_INVALID_NLINK);
+		else
+			set_nlink(inode, 1);
 		if (fattr->valid & NFS_ATTR_FATTR_OWNER)
 			inode->i_uid = fattr->uid;
 		else if (fattr_supported & NFS_ATTR_FATTR_OWNER)
-- 
2.48.1