This patch moves the call to alloc_empty_file() below the call to
path_init(). That changes is needed for the next patch, which adds
a cred override for alloc_empty_file(). The needed cred info is only
available after the call to path_init().
No functional changes are intended by that patch.
Signed-off-by: Stas Sergeev <[email protected]>
CC: Eric Biederman <[email protected]>
CC: Alexander Viro <[email protected]>
CC: Christian Brauner <[email protected]>
CC: Jan Kara <[email protected]>
CC: Andy Lutomirski <[email protected]>
CC: [email protected]
CC: [email protected]
---
fs/namei.c | 26 +++++++++++++++++---------
1 file changed, 17 insertions(+), 9 deletions(-)
diff --git a/fs/namei.c b/fs/namei.c
index c5b2a25be7d0..2fde2c320ae9 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -3782,22 +3782,30 @@ static struct file *path_openat(struct nameidata *nd,
struct file *file;
int error;
- file = alloc_empty_file(op->open_flag, current_cred());
- if (IS_ERR(file))
- return file;
-
- if (unlikely(file->f_flags & __O_TMPFILE)) {
+ if (unlikely(op->open_flag & __O_TMPFILE)) {
+ file = alloc_empty_file(op->open_flag, current_cred());
+ if (IS_ERR(file))
+ return file;
error = do_tmpfile(nd, flags, op, file);
- } else if (unlikely(file->f_flags & O_PATH)) {
+ } else if (unlikely(op->open_flag & O_PATH)) {
+ file = alloc_empty_file(op->open_flag, current_cred());
+ if (IS_ERR(file))
+ return file;
error = do_o_path(nd, flags, file);
} else {
const char *s = path_init(nd, flags);
- while (!(error = link_path_walk(s, nd)) &&
- (s = open_last_lookups(nd, file, op)) != NULL)
- ;
+ file = alloc_empty_file(op->open_flag, current_cred());
+ error = PTR_ERR_OR_ZERO(file);
+ if (!error) {
+ while (!(error = link_path_walk(s, nd)) &&
+ (s = open_last_lookups(nd, file, op)) != NULL)
+ ;
+ }
if (!error)
error = do_open(nd, file, op);
terminate_walk(nd);
+ if (IS_ERR(file))
+ return file;
}
if (likely(!error)) {
if (likely(file->f_mode & FMODE_OPENED))
--
2.44.0
24.04.2024 12:17, David Laight пишет:
> From: Stas Sergeev
>> Sent: 22 April 2024 09:45
> I seem to have 5 copies of this patch.....
Yep, its re-sent with every new iteartion,
but doesn't change by itself (the other
one changes).
Is there anything I can do to avoid
unneeded duplicates of an unchanged
patch?
Manually reduce Cc list until the patch
changes? That looks too much of a trouble
though.
> You probably ought to merge the two 'unlikely' tests.
Ok.
> Copying op->open_flag to a local may also generate better code.
Can't gcc deduce this on its own?
But ok, will do.