[PATCH] ocfs2: make ocfs2_calc_xattr_init() return void

[PATCH] ocfs2: make ocfs2_calc_xattr_init() return void

[Joseph Qi]

ocfs2_calc_xattr_init() used to read the default ACL off the parent
inode itself, so it could return an error from ocfs2_xattr_get_nolock().
Commit bd7c05fb4a47 ("ocfs2: fix circular locking dependency in
ocfs2_init_acl()") moved that lookup before the transaction starts and
deleted the error path, but left the now vestigial 'int ret = 0'
declaration and both 'return ret' statements behind, along with an
unreachable error branch in ocfs2_mknod().

Drop the leftover variable and convert the return type to void, so the
callee states that it always succeeds and the caller no longer carries
a check that can never trigger.

No functional change.

Fixes: bd7c05fb4a47 ("ocfs2: fix circular locking dependency in ocfs2_init_acl()")
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202609040247.8B3lmoqX-lkp@intel.com/
Signed-off-by: Joseph Qi <joseph.qi@linux.alibaba.com>
---
 fs/ocfs2/namei.c |  9 ++-------
 fs/ocfs2/xattr.c | 13 +++++--------
 fs/ocfs2/xattr.h |  8 ++++----
 3 files changed, 11 insertions(+), 19 deletions(-)

diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c
index fea889b9f577..ef03f90e7265 100644
--- a/fs/ocfs2/namei.c
+++ b/fs/ocfs2/namei.c
@@ -335,13 +335,8 @@ static int ocfs2_mknod(struct mnt_idmap *idmap,
 		goto leave;
 
 	/* calculate meta data/clusters for setting security and acl xattr */
-	status = ocfs2_calc_xattr_init(dir, mode, &si, &want_clusters,
-				       &xattr_credits, &want_meta,
-				       &acl_state);
-	if (status < 0) {
-		mlog_errno(status);
-		goto leave;
-	}
+	ocfs2_calc_xattr_init(dir, mode, &si, &want_clusters, &xattr_credits,
+			      &want_meta, &acl_state);
 
 	/* Reserve a cluster if creating an extent based directory. */
 	if (S_ISDIR(mode) && !ocfs2_supports_inline_data(osb)) {
diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c
index f30faabf601b..f2b7a53ed969 100644
--- a/fs/ocfs2/xattr.c
+++ b/fs/ocfs2/xattr.c
@@ -682,13 +682,12 @@ int ocfs2_calc_security_init(struct inode *dir,
 	return ret;
 }
 
-int ocfs2_calc_xattr_init(struct inode *dir, umode_t mode,
-			  struct ocfs2_security_xattr_info *si,
-			  int *want_clusters, int *xattr_credits,
-			  int *want_meta, struct ocfs2_acl_state *acl_state)
+void ocfs2_calc_xattr_init(struct inode *dir, umode_t mode,
+			   struct ocfs2_security_xattr_info *si,
+			   int *want_clusters, int *xattr_credits,
+			   int *want_meta, struct ocfs2_acl_state *acl_state)
 {
 	int i;
-	int ret = 0;
 	struct ocfs2_super *osb = OCFS2_SB(dir->i_sb);
 	int s_size = 0, a_size = 0, acl_len = 0, new_clusters;
 
@@ -713,7 +712,7 @@ int ocfs2_calc_xattr_init(struct inode *dir, umode_t mode,
 	}
 
 	if (!(s_size + a_size))
-		return ret;
+		return;
 
 	/*
 	 * The max space of security xattr taken inline is
@@ -781,8 +780,6 @@ int ocfs2_calc_xattr_init(struct inode *dir, umode_t mode,
 			}
 		}
 	}
-
-	return ret;
 }
 
 static int ocfs2_xattr_extend_allocation(struct inode *inode,
diff --git a/fs/ocfs2/xattr.h b/fs/ocfs2/xattr.h
index 70a6a9dab71e..1dd08f979e3c 100644
--- a/fs/ocfs2/xattr.h
+++ b/fs/ocfs2/xattr.h
@@ -65,10 +65,10 @@ int ocfs2_calc_security_init(struct inode *,
 			     int *, int *, struct ocfs2_alloc_context **);
 
 struct ocfs2_acl_state;
-int ocfs2_calc_xattr_init(struct inode *dir, umode_t mode,
-			  struct ocfs2_security_xattr_info *si,
-			  int *want_clusters, int *xattr_credits,
-			  int *want_meta, struct ocfs2_acl_state *acl_state);
+void ocfs2_calc_xattr_init(struct inode *dir, umode_t mode,
+			   struct ocfs2_security_xattr_info *si,
+			   int *want_clusters, int *xattr_credits,
+			   int *want_meta, struct ocfs2_acl_state *acl_state);
 
 /*
  * xattrs can live inside an inode, as part of an external xattr block,
-- 
2.39.3

Re: [PATCH] ocfs2: make ocfs2_calc_xattr_init() return void

[Andrew Morton]

On Fri,  4 Sep 2026 10:37:51 +0800 Joseph Qi <joseph.qi@linux.alibaba.com> wrote:

> ocfs2_calc_xattr_init() used to read the default ACL off the parent
> inode itself, so it could return an error from ocfs2_xattr_get_nolock().
> Commit bd7c05fb4a47 ("ocfs2: fix circular locking dependency in
> ocfs2_init_acl()") moved that lookup before the transaction starts and
> deleted the error path, but left the now vestigial 'int ret = 0'
> declaration and both 'return ret' statements behind, along with an
> unreachable error branch in ocfs2_mknod().
> 
> Drop the leftover variable and convert the return type to void, so the
> callee states that it always succeeds and the caller no longer carries
> a check that can never trigger.
> 
> No functional change.

Thanks.

> Fixes: bd7c05fb4a47 ("ocfs2: fix circular locking dependency in ocfs2_init_acl()")
> Reported-by: kernel test robot <lkp@intel.com>
> Closes: https://lore.kernel.org/oe-kbuild-all/202609040247.8B3lmoqX-lkp@intel.com/

Gee, that's a bit picky, isn't it?

Re: [PATCH] ocfs2: make ocfs2_calc_xattr_init() return void

[Joseph Qi]

On 9/4/26 12:36 PM, Andrew Morton wrote:
> On Fri,  4 Sep 2026 10:37:51 +0800 Joseph Qi <joseph.qi@linux.alibaba.com> wrote:
> 
>> ocfs2_calc_xattr_init() used to read the default ACL off the parent
>> inode itself, so it could return an error from ocfs2_xattr_get_nolock().
>> Commit bd7c05fb4a47 ("ocfs2: fix circular locking dependency in
>> ocfs2_init_acl()") moved that lookup before the transaction starts and
>> deleted the error path, but left the now vestigial 'int ret = 0'
>> declaration and both 'return ret' statements behind, along with an
>> unreachable error branch in ocfs2_mknod().
>>
>> Drop the leftover variable and convert the return type to void, so the
>> callee states that it always succeeds and the caller no longer carries
>> a check that can never trigger.
>>
>> No functional change.
> 
> Thanks.
> 
>> Fixes: bd7c05fb4a47 ("ocfs2: fix circular locking dependency in ocfs2_init_acl()")
>> Reported-by: kernel test robot <lkp@intel.com>
>> Closes: https://lore.kernel.org/oe-kbuild-all/202609040247.8B3lmoqX-lkp@intel.com/
> 
> Gee, that's a bit picky, isn't it?

It's indeed a trivial fix.
I've followed the report mail to add the tags and I think it may help
kernel test robot tracking.

Thanks,
Joseph