From: Gabriel krisman Bertazi <[email protected]>
This patchset enables negative dentries for case-insensitive directories
in ext4/f2fs. It solves the corner cases for this feature, including
those already tested by fstests (generic/556). It also solves an
existing bug with the existing implementation where old negative
dentries are left behind after a directory conversion to
case-insensitive.
Testing-wise, I ran sanity checks to show it properly uses the created
negative dentries, observed the expected performance increase of the
dentry cache hit, and showed it survives the quick group in fstests on
both f2fs and ext4 without regressions.
* Background
Negative dentries have always been disabled in case-insensitive
directories because, in their current form, they can't provide enough
assurances that all the case variations of a filename won't exist in a
directory, and the name-preserving case-insenstive semantics
during file creation prevents some negative dentries from being
instantiated unmodified.
Nevertheless, for the general case, the existing implementation would
already work with negative dentries, even though they are fully
disabled. That is: if the original lookup that created the dentry was
done in a case-insensitive way, the negative dentry can usually be
validated, since it assures that no other dcache entry exists, *and*
that no variation of the file exists on disk (since the lookup
failed). A following lookup would then be executed with the
case-insensitive-aware d_hash and d_lookup, which would find the right
negative dentry and use it.
The first corner case arises when a case-insensitive directory has
negative dentries that were created before the directory was flipped to
case-insensitive. A directory must be empty to be converted, but it
doesn't mean the directory doesn't have negative dentry children. If
that happens, the dangling dentries left behind can't assure that no
case-variation of the name exists. They only mean the exact name
doesn't exist. A further lookup would incorrectly validate them.
The code below demonstrates the problem. In this example $1 and $2 are
two strings, where:
(i) $1 != $2
(ii) casefold($1) == casefold($2)
(iii) hash($1) == hash($2) == hash(casefold($1))
Then, the following sequence could potentially return a ENOENT, even
though the case-insensitive lookup should exist:
mkdir d <- Case-sensitive directory
touch d/$1
touch d/$2
unlink d/$1 <- leaves negative dentry behind.
unlink d/$2 <- leaves *another* negative dentry behind.
chattr +F d <- make 'd' case-insensitive.
touch d/$1 <- Both negative dentries could match. finds one of them,
and instantiate
access d/$1 <- Find the other negative dentry, get -ENOENT.
In fact, this is a problem even on the current implementation, where
negative dentries for CI are disabled. There was a bug reported by Al
Viro in 2020, where a directory might end up with dangling negative
dentries created during a case-sensitive lookup, because they existed
before the +F attribute was set.
It is hard to trigger the issue, because condition (iii) is hard to test
on an unmodified kernel. By hacking the kernel to force the hash
collision, there are a few ways we can trigger this bizarre behavior in
case-insensitive directories through the insertion of negative dentries.
Another problem exists when turning a negative dentry to positive. If
the negative dentry has a different case than what is currently being
used for lookup, the dentry cannot be reused without changing its name,
in order to guarantee filename-preserving semantics to userspace. We
need to either change the name or invalidate the dentry. This issue is
currently avoided in mainline, since the negative dentry mechanism is
disabled.
* Proposal
The main idea is to differentiate negative dentries created in a
case-insensitive context from those created during a case-sensitive
lookup via a new dentry flag, D_CASEFOLD_LOOKUP, set by the filesystem
the d_lookup hook. Since the former can be used (except for the
name-preserving issue), d_revalidate will just check the flag to
quickly accept or reject the dentry.
A different solution would be to guarantee no negative dentry exists
during the case-sensitive to case-insensitive directory conversion (the
other direction is safe). It has the following problems:
1) It is not trivial to implement a race-free mechanism to ensure
negative dentries won't be recreated immediately after invalidation
while converting the directory.
2) The knowledge whether the negative dentry is valid (i.e. comes from
a case-insensitive lookup) is implicit on the fact that we are
correctly invalidating dentries when converting the directory.
Having a D_CASEFOLD_LOOKUP avoids both issues, and seems to be a cheap
solution to the problem.
But, as explained above, due to the filename preserving semantics, we
cannot just validate based on D_CASEFOLD_LOOKUP.
For that, one solution would be to invalidate the negative dentry when
it is decided to turn it positive, instead of reusing it. I implemented
that in the past (2018) but Al Viro made it clear we don't want to incur
costs on the VFS critical path for filesystems who don't care about
case-insensitiveness.
Instead, this patch invalidates negative dentries in casefold
directories in d_revalidate during creation lookups, iff the lookup name
is not exactly what is cached. Other kinds of lookups wouldn't need
this limitation.
* caveats
1) Encryption
Negative dentries on case-insensitive encrypted directories are also
disabled. No semantic change for them is intended in
this patchset; we just bypass the revalidation directly to fscrypt, for
positive dentries. Encryption support is future work.
2) revalidate the cached dentry using the name under lookup
Validating based on the lookup name is strange for a cache. the new
semantic is implemented by d_revalidate, to stay out of the critical
path of filesystems who don't care about case-insensitiveness, as much
as possible. The only change is the addition of a new flavor of
d_revalidate.
* Tests
There are a tests in place for most of the corner cases in generic/556.
They mainly verify the name-preserving semantics. The invalidation when
converting the directory is harder to test, because it is hard to force
the invalidation of specific cached dentries that occlude a dangling
invalid dentry. I tested it with forcing the positive dentries to be
removed, but I'm not sure how to write an upstreamable test.
It also survives fstests quick group regression testing on both ext4 and
f2fs.
* Performance
The latency of lookups of non-existing files is obviously improved, as
would be expected. The following numbers compare the execution time of 10^6
lookups of a non-existing file in a case-insensitive directory
pre-populated with 100k files in ext4.
Without the patch: 10.363s / 0.349s / 9.920s (real/user/sys)
With the patch: 1.752s / 0.276s / 1.472s (real/user/sys)
* patchset
Patch 1 introduces a new flavor of d_revalidate to provide the
filesystem with the name under lookup; Patch 2 introduces the new flag
to signal the dentry creation context; Patch 3 introduces a libfs helper
to revalidate negative dentries on case-insensitive directories; Patch 4
deals with encryption; Patch 5 cleans up the now redundant dentry
operations for case-insensitive with and without encryption; Finally,
Patch 6 and 7 enable support on case-insensitive directories
for ext4 and f2fs, respectively.
Gabriel Krisman Bertazi (7):
fs: Expose name under lookup to d_revalidate hook
fs: Add DCACHE_CASEFOLD_LOOKUP flag
libfs: Validate negative dentries in case-insensitive directories
libfs: Support revalidation of encrypted case-insensitive dentries
libfs: Merge encrypted_ci_dentry_ops and ci_dentry_ops
ext4: Enable negative dentries on case-insensitive lookup
f2fs: Enable negative dentries on case-insensitive lookup
fs/dcache.c | 10 +++++-
fs/ext4/namei.c | 34 ++------------------
fs/f2fs/namei.c | 23 ++------------
fs/libfs.c | 72 ++++++++++++++++++++++++------------------
fs/namei.c | 23 ++++++++------
include/linux/dcache.h | 9 ++++++
6 files changed, 78 insertions(+), 93 deletions(-)
--
2.35.3
From: Gabriel Krisman Bertazi <[email protected]>
Negative dentries support on case-insensitive ext4/f2fs will require
access to the name under lookup to ensure it matches the dentry. This
adds an optional new flavor of cached dentry revalidation hook to expose
this extra parameter.
I'm fine with extending d_revalidate instead of adding a new hook, if
it is considered cleaner and the approach is accepted. I wrote a new
hook to simplify reviewing.
Signed-off-by: Gabriel Krisman Bertazi <[email protected]>
---
fs/dcache.c | 2 +-
fs/namei.c | 23 ++++++++++++++---------
include/linux/dcache.h | 1 +
3 files changed, 16 insertions(+), 10 deletions(-)
diff --git a/fs/dcache.c b/fs/dcache.c
index 52e6d5fdab6b..98521862e58a 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -1928,7 +1928,7 @@ void d_set_d_op(struct dentry *dentry, const struct dentry_operations *op)
dentry->d_flags |= DCACHE_OP_HASH;
if (op->d_compare)
dentry->d_flags |= DCACHE_OP_COMPARE;
- if (op->d_revalidate)
+ if (op->d_revalidate || op->d_revalidate_name)
dentry->d_flags |= DCACHE_OP_REVALIDATE;
if (op->d_weak_revalidate)
dentry->d_flags |= DCACHE_OP_WEAK_REVALIDATE;
diff --git a/fs/namei.c b/fs/namei.c
index 309ae6fc8c99..525d31cc3b85 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -850,11 +850,16 @@ static bool try_to_unlazy_next(struct nameidata *nd, struct dentry *dentry)
return false;
}
-static inline int d_revalidate(struct dentry *dentry, unsigned int flags)
+static inline int d_revalidate(struct dentry *dentry,
+ const struct qstr *name,
+ unsigned int flags)
{
- if (unlikely(dentry->d_flags & DCACHE_OP_REVALIDATE))
+
+ if (unlikely(dentry->d_flags & DCACHE_OP_REVALIDATE)) {
+ if (dentry->d_op->d_revalidate_name)
+ return dentry->d_op->d_revalidate_name(dentry, name, flags);
return dentry->d_op->d_revalidate(dentry, flags);
- else
+ } else
return 1;
}
@@ -1562,7 +1567,7 @@ static struct dentry *lookup_dcache(const struct qstr *name,
{
struct dentry *dentry = d_lookup(dir, name);
if (dentry) {
- int error = d_revalidate(dentry, flags);
+ int error = d_revalidate(dentry, name, flags);
if (unlikely(error <= 0)) {
if (!error)
d_invalidate(dentry);
@@ -1631,19 +1636,19 @@ static struct dentry *lookup_fast(struct nameidata *nd)
if (read_seqcount_retry(&parent->d_seq, nd->seq))
return ERR_PTR(-ECHILD);
- status = d_revalidate(dentry, nd->flags);
+ status = d_revalidate(dentry, &nd->last, nd->flags);
if (likely(status > 0))
return dentry;
if (!try_to_unlazy_next(nd, dentry))
return ERR_PTR(-ECHILD);
if (status == -ECHILD)
/* we'd been told to redo it in non-rcu mode */
- status = d_revalidate(dentry, nd->flags);
+ status = d_revalidate(dentry, &nd->last, nd->flags);
} else {
dentry = __d_lookup(parent, &nd->last);
if (unlikely(!dentry))
return NULL;
- status = d_revalidate(dentry, nd->flags);
+ status = d_revalidate(dentry, &nd->last, nd->flags);
}
if (unlikely(status <= 0)) {
if (!status)
@@ -1671,7 +1676,7 @@ static struct dentry *__lookup_slow(const struct qstr *name,
if (IS_ERR(dentry))
return dentry;
if (unlikely(!d_in_lookup(dentry))) {
- int error = d_revalidate(dentry, flags);
+ int error = d_revalidate(dentry, name, flags);
if (unlikely(error <= 0)) {
if (!error) {
d_invalidate(dentry);
@@ -3342,7 +3347,7 @@ static struct dentry *lookup_open(struct nameidata *nd, struct file *file,
if (d_in_lookup(dentry))
break;
- error = d_revalidate(dentry, nd->flags);
+ error = d_revalidate(dentry, &nd->last, nd->flags);
if (likely(error > 0))
break;
if (error)
diff --git a/include/linux/dcache.h b/include/linux/dcache.h
index 6b351e009f59..b6188f2e8950 100644
--- a/include/linux/dcache.h
+++ b/include/linux/dcache.h
@@ -127,6 +127,7 @@ enum dentry_d_lock_class
struct dentry_operations {
int (*d_revalidate)(struct dentry *, unsigned int);
+ int (*d_revalidate_name)(struct dentry *, const struct qstr *, unsigned int);
int (*d_weak_revalidate)(struct dentry *, unsigned int);
int (*d_hash)(const struct dentry *, struct qstr *);
int (*d_compare)(const struct dentry *,
--
2.35.3
From: Gabriel Krisman Bertazi <[email protected]>
Introduce a dentry revalidation helper to be used by case-insensitive
filesystems to check if it is safe to reuse a negative dentry.
A negative dentry is safe to be reused on a case-insensitive lookup if
it was created during a case-insensitive lookup and this is not a lookup
that will instantiate a dentry. If this is a creation lookup, we also
need to make sure the name matches sensitively the name under lookup in
order to assure the name preserving semantics.
Signed-off-by: Gabriel Krisman Bertazi <[email protected]>
---
fs/libfs.c | 24 ++++++++++++++++++++++++
1 file changed, 24 insertions(+)
diff --git a/fs/libfs.c b/fs/libfs.c
index aada4e7c8713..e3daca88d1d3 100644
--- a/fs/libfs.c
+++ b/fs/libfs.c
@@ -1467,9 +1467,33 @@ static int generic_ci_d_hash(const struct dentry *dentry, struct qstr *str)
return 0;
}
+static inline int generic_ci_d_revalidate(struct dentry *dentry,
+ const struct qstr *name,
+ unsigned int flags)
+{
+ int is_creation = flags & (LOOKUP_CREATE | LOOKUP_RENAME_TARGET);
+
+ if (d_is_negative(dentry)) {
+ const struct dentry *parent = READ_ONCE(dentry->d_parent);
+ const struct inode *dir = READ_ONCE(parent->d_inode);
+
+ if (dir && needs_casefold(dir)) {
+ if (!d_is_casefold_lookup(dentry))
+ return 0;
+
+ if (is_creation &&
+ (dentry->d_name.len != name->len ||
+ memcmp(dentry->d_name.name, name->name, name->len)))
+ return 0;
+ }
+ }
+ return 1;
+}
+
static const struct dentry_operations generic_ci_dentry_ops = {
.d_hash = generic_ci_d_hash,
.d_compare = generic_ci_d_compare,
+ .d_revalidate_name = generic_ci_d_revalidate,
};
#endif
--
2.35.3
From: Gabriel Krisman Bertazi <[email protected]>
Instead of invalidating negative dentries during case-insensitive
lookups, mark them as such and let them be added to the dcache.
d_ci_revalidate is able to properly filter them out if necessary based
on the dentry casefold flag.
Signed-off-by: Gabriel Krisman Bertazi <[email protected]>
---
fs/ext4/namei.c | 34 +++-------------------------------
1 file changed, 3 insertions(+), 31 deletions(-)
diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index dd28453d6ea3..36d6683ff616 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -1851,16 +1851,9 @@ static struct dentry *ext4_lookup(struct inode *dir, struct dentry *dentry, unsi
}
}
-#if IS_ENABLED(CONFIG_UNICODE)
- if (!inode && IS_CASEFOLDED(dir)) {
- /* Eventually we want to call d_add_ci(dentry, NULL)
- * for negative dentries in the encoding case as
- * well. For now, prevent the negative dentry
- * from being cached.
- */
- return NULL;
- }
-#endif
+ if (IS_ENABLED(CONFIG_UNICODE) && IS_CASEFOLDED(dir))
+ d_set_casefold_lookup(dentry);
+
return d_splice_alias(inode, dentry);
}
@@ -3186,17 +3179,6 @@ static int ext4_rmdir(struct inode *dir, struct dentry *dentry)
ext4_fc_track_unlink(handle, dentry);
retval = ext4_mark_inode_dirty(handle, dir);
-#if IS_ENABLED(CONFIG_UNICODE)
- /* VFS negative dentries are incompatible with Encoding and
- * Case-insensitiveness. Eventually we'll want avoid
- * invalidating the dentries here, alongside with returning the
- * negative dentries at ext4_lookup(), when it is better
- * supported by the VFS for the CI case.
- */
- if (IS_CASEFOLDED(dir))
- d_invalidate(dentry);
-#endif
-
end_rmdir:
brelse(bh);
if (handle)
@@ -3297,16 +3279,6 @@ static int ext4_unlink(struct inode *dir, struct dentry *dentry)
goto out_trace;
retval = __ext4_unlink(dir, &dentry->d_name, d_inode(dentry), dentry);
-#if IS_ENABLED(CONFIG_UNICODE)
- /* VFS negative dentries are incompatible with Encoding and
- * Case-insensitiveness. Eventually we'll want avoid
- * invalidating the dentries here, alongside with returning the
- * negative dentries at ext4_lookup(), when it is better
- * supported by the VFS for the CI case.
- */
- if (IS_CASEFOLDED(dir))
- d_invalidate(dentry);
-#endif
out_trace:
trace_ext4_unlink_exit(dentry, retval);
--
2.35.3
From: Gabriel Krisman Bertazi <[email protected]>
Now that casefold needs d_revalidate and calls fscrypt_d_revalidate
itself, generic_encrypt_ci_dentry_ops and generic_ci_dentry_ops are now
equivalent. Merge them together and simplify the setup code.
Signed-off-by: Gabriel Krisman Bertazi <[email protected]>
---
fs/libfs.c | 44 +++++++++++++-------------------------------
1 file changed, 13 insertions(+), 31 deletions(-)
diff --git a/fs/libfs.c b/fs/libfs.c
index 26a06fd5f5a1..96934c7e54ab 100644
--- a/fs/libfs.c
+++ b/fs/libfs.c
@@ -1494,7 +1494,7 @@ static inline int generic_ci_d_revalidate(struct dentry *dentry,
return fscrypt_d_revalidate(dentry, flags);
}
-static const struct dentry_operations generic_ci_dentry_ops = {
+static const struct dentry_operations generic_encrypted_ci_dentry_ops = {
.d_hash = generic_ci_d_hash,
.d_compare = generic_ci_d_compare,
.d_revalidate_name = generic_ci_d_revalidate,
@@ -1507,26 +1507,20 @@ static const struct dentry_operations generic_encrypted_dentry_ops = {
};
#endif
-#if defined(CONFIG_FS_ENCRYPTION) && IS_ENABLED(CONFIG_UNICODE)
-static const struct dentry_operations generic_encrypted_ci_dentry_ops = {
- .d_hash = generic_ci_d_hash,
- .d_compare = generic_ci_d_compare,
- .d_revalidate_name = generic_ci_d_revalidate,
-};
-#endif
-
/**
* generic_set_encrypted_ci_d_ops - helper for setting d_ops for given dentry
* @dentry: dentry to set ops on
*
- * Casefolded directories need d_hash and d_compare set, so that the dentries
- * contained in them are handled case-insensitively. Note that these operations
- * are needed on the parent directory rather than on the dentries in it, and
- * while the casefolding flag can be toggled on and off on an empty directory,
- * dentry_operations can't be changed later. As a result, if the filesystem has
- * casefolding support enabled at all, we have to give all dentries the
- * casefolding operations even if their inode doesn't have the casefolding flag
- * currently (and thus the casefolding ops would be no-ops for now).
+ * Casefolded directories need d_hash, d_compare and d_revalidate set, so
+ * that the dentries contained in them are handled case-insensitively,
+ * but implement support for fs_encryption. Note that these operations
+ * are needed on the parent directory rather than on the dentries in it,
+ * and while the casefolding flag can be toggled on and off on an empty
+ * directory, dentry_operations can't be changed later. As a result, if
+ * the filesystem has casefolding support enabled at all, we have to
+ * give all dentries the casefolding operations even if their inode
+ * doesn't have the casefolding flag currently (and thus the casefolding
+ * ops would be no-ops for now).
*
* Encryption works differently in that the only dentry operation it needs is
* d_revalidate, which it only needs on dentries that have the no-key name flag.
@@ -1539,30 +1533,18 @@ static const struct dentry_operations generic_encrypted_ci_dentry_ops = {
*/
void generic_set_encrypted_ci_d_ops(struct dentry *dentry)
{
-#ifdef CONFIG_FS_ENCRYPTION
- bool needs_encrypt_ops = dentry->d_flags & DCACHE_NOKEY_NAME;
-#endif
#if IS_ENABLED(CONFIG_UNICODE)
- bool needs_ci_ops = dentry->d_sb->s_encoding;
-#endif
-#if defined(CONFIG_FS_ENCRYPTION) && IS_ENABLED(CONFIG_UNICODE)
- if (needs_encrypt_ops && needs_ci_ops) {
+ if (dentry->d_sb->s_encoding) {
d_set_d_op(dentry, &generic_encrypted_ci_dentry_ops);
return;
}
#endif
#ifdef CONFIG_FS_ENCRYPTION
- if (needs_encrypt_ops) {
+ if (dentry->d_flags & DCACHE_NOKEY_NAME) {
d_set_d_op(dentry, &generic_encrypted_dentry_ops);
return;
}
#endif
-#if IS_ENABLED(CONFIG_UNICODE)
- if (needs_ci_ops) {
- d_set_d_op(dentry, &generic_ci_dentry_ops);
- return;
- }
-#endif
}
EXPORT_SYMBOL(generic_set_encrypted_ci_d_ops);
--
2.35.3
From: Gabriel Krisman Bertazi <[email protected]>
Preserve the existing behavior for encrypted directories, by rejecting
negative dentries of encrypted+casefolded directories. This allows
generic_ci_d_revalidate to be used by filesystems with both features
enabled, as long as the directory is either casefolded or encrypted, but
not both at the same time.
Signed-off-by: Gabriel Krisman Bertazi <[email protected]>
---
fs/libfs.c | 8 ++++++--
1 file changed, 6 insertions(+), 2 deletions(-)
diff --git a/fs/libfs.c b/fs/libfs.c
index e3daca88d1d3..26a06fd5f5a1 100644
--- a/fs/libfs.c
+++ b/fs/libfs.c
@@ -1478,6 +1478,9 @@ static inline int generic_ci_d_revalidate(struct dentry *dentry,
const struct inode *dir = READ_ONCE(parent->d_inode);
if (dir && needs_casefold(dir)) {
+ if (IS_ENCRYPTED(dir))
+ return 0;
+
if (!d_is_casefold_lookup(dentry))
return 0;
@@ -1487,7 +1490,8 @@ static inline int generic_ci_d_revalidate(struct dentry *dentry,
return 0;
}
}
- return 1;
+
+ return fscrypt_d_revalidate(dentry, flags);
}
static const struct dentry_operations generic_ci_dentry_ops = {
@@ -1507,7 +1511,7 @@ static const struct dentry_operations generic_encrypted_dentry_ops = {
static const struct dentry_operations generic_encrypted_ci_dentry_ops = {
.d_hash = generic_ci_d_hash,
.d_compare = generic_ci_d_compare,
- .d_revalidate = fscrypt_d_revalidate,
+ .d_revalidate_name = generic_ci_d_revalidate,
};
#endif
--
2.35.3
From: Gabriel Krisman Bertazi <[email protected]>
This flag marks a negative or positive dentry as being created after a
case-insensitive lookup operation. It is useful to differentiate
dentries this way to detect whether the negative dentry can be trusted
during a case-insensitive lookup.
Signed-off-by: Gabriel Krisman Bertazi <[email protected]>
---
fs/dcache.c | 8 ++++++++
include/linux/dcache.h | 8 ++++++++
2 files changed, 16 insertions(+)
diff --git a/fs/dcache.c b/fs/dcache.c
index 98521862e58a..05e4c7019e17 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -1958,6 +1958,14 @@ void d_set_fallthru(struct dentry *dentry)
}
EXPORT_SYMBOL(d_set_fallthru);
+void d_set_casefold_lookup(struct dentry *dentry)
+{
+ spin_lock(&dentry->d_lock);
+ dentry->d_flags |= DCACHE_CASEFOLD_LOOKUP;
+ spin_unlock(&dentry->d_lock);
+}
+EXPORT_SYMBOL(d_set_casefold_lookup);
+
static unsigned d_flags_for_inode(struct inode *inode)
{
unsigned add_flags = DCACHE_REGULAR_TYPE;
diff --git a/include/linux/dcache.h b/include/linux/dcache.h
index b6188f2e8950..457345123cb6 100644
--- a/include/linux/dcache.h
+++ b/include/linux/dcache.h
@@ -209,6 +209,7 @@ struct dentry_operations {
#define DCACHE_FALLTHRU 0x01000000 /* Fall through to lower layer */
#define DCACHE_NOKEY_NAME 0x02000000 /* Encrypted name encoded without key */
#define DCACHE_OP_REAL 0x04000000
+#define DCACHE_CASEFOLD_LOOKUP 0x08000000 /* Dentry comes from a casefold directory */
#define DCACHE_PAR_LOOKUP 0x10000000 /* being looked up (with parent locked shared) */
#define DCACHE_DENTRY_CURSOR 0x20000000
@@ -497,6 +498,13 @@ static inline bool d_is_fallthru(const struct dentry *dentry)
return dentry->d_flags & DCACHE_FALLTHRU;
}
+extern void d_set_casefold_lookup(struct dentry *dentry);
+
+static inline bool d_is_casefold_lookup(const struct dentry *dentry)
+{
+ return dentry->d_flags & DCACHE_CASEFOLD_LOOKUP;
+}
+
extern int sysctl_vfs_cache_pressure;
--
2.35.3
From: Gabriel Krisman Bertazi <[email protected]>
Instead of invalidating negative dentries during case-insensitive
lookups, mark them as such and let them be added to the dcache.
d_ci_revalidate is able to properly filter them out if necessary based
on the dentry casefold flag.
Signed-off-by: Gabriel Krisman Bertazi <[email protected]>
---
fs/f2fs/namei.c | 23 ++---------------------
1 file changed, 2 insertions(+), 21 deletions(-)
diff --git a/fs/f2fs/namei.c b/fs/f2fs/namei.c
index 6032589099ce..558cef6e25fc 100644
--- a/fs/f2fs/namei.c
+++ b/fs/f2fs/namei.c
@@ -564,17 +564,8 @@ static struct dentry *f2fs_lookup(struct inode *dir, struct dentry *dentry,
goto out_iput;
}
out_splice:
-#if IS_ENABLED(CONFIG_UNICODE)
- if (!inode && IS_CASEFOLDED(dir)) {
- /* Eventually we want to call d_add_ci(dentry, NULL)
- * for negative dentries in the encoding case as
- * well. For now, prevent the negative dentry
- * from being cached.
- */
- trace_f2fs_lookup_end(dir, dentry, ino, err);
- return NULL;
- }
-#endif
+ if (IS_ENABLED(CONFIG_UNICODE) && IS_CASEFOLDED(dir))
+ d_set_casefold_lookup(dentry);
new = d_splice_alias(inode, dentry);
err = PTR_ERR_OR_ZERO(new);
trace_f2fs_lookup_end(dir, dentry, ino, !new ? -ENOENT : err);
@@ -627,16 +618,6 @@ static int f2fs_unlink(struct inode *dir, struct dentry *dentry)
f2fs_delete_entry(de, page, dir, inode);
f2fs_unlock_op(sbi);
-#if IS_ENABLED(CONFIG_UNICODE)
- /* VFS negative dentries are incompatible with Encoding and
- * Case-insensitiveness. Eventually we'll want avoid
- * invalidating the dentries here, alongside with returning the
- * negative dentries at f2fs_lookup(), when it is better
- * supported by the VFS for the CI case.
- */
- if (IS_CASEFOLDED(dir))
- d_invalidate(dentry);
-#endif
if (IS_DIRSYNC(dir))
f2fs_sync_fs(sbi->sb, 1);
fail:
--
2.35.3