OpenVZ / Virtuozzo kernel development (devel@openvz.org)
 help / color / mirror / Atom feed
* [Devel] [PATCH hci-8.1 v6 0/5] Fix kernel crash in overlayfs
@ 2026-11-16 16:06 Vladimir Riabchun
  2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 1/5] ve: Remove STOPPED state Vladimir Riabchun
                   ` (4 more replies)
  0 siblings, 5 replies; 13+ messages in thread
From: Vladimir Riabchun @ 2026-11-16 16:06 UTC (permalink / raw)
  To: vasileios.almpanis, ptikhomirov, Konstantin Khorenko; +Cc: devel

When testing CRIU with containers with docker images,
we found an easily reproducible kernel crash:

[ 2300.870181] BUG: kernel NULL pointer dereference, address: 0000000000000005
[ 2300.871537] #PF: supervisor read access in kernel mode
[ 2300.872577] #PF: error_code(0x0000) - not-present page
[ 2300.873595] PGD 800000010a18f067 P4D 800000010a18f067 PUD 10730c067 PMD 0
[ 2300.874900] Oops: Oops: 0000 [#1] SMP PTI
[ 2300.875742] CPU: 1 UID: 0 PID: 5419 Comm: criu ve: 0 Kdump: loaded Not tainted 6.12.0-211.39.1.16.6.vz10.x86_64 #1 16.6 PREEMPT(voluntary)
[ 2300.878057] Hardware name: Acronis OpenStack Compute/Virtuozzo, BIOS 1.16.1-1.vz9.2 04/01/2014
[ 2300.879684] RIP: 0010:ovl_get_index_name_fh+0x18/0x70 [overlay]
[ 2300.880859] Code: ...
[ 2300.884128] RSP: 0018:ffffd32582e0f7c0 EFLAGS: 00010246
[ 2300.895807] Call Trace:
[ 2300.896397]  <TASK>
[ 2300.904435]  ovl_create_index+0x82/0x260 [overlay]
[ 2300.905406]  ovl_copy_up_workdir+0x30a/0x330 [overlay]
[ 2300.906427]  ovl_do_copy_up+0x10d/0x2e0 [overlay]
[ 2300.907387]  ovl_copy_up_one+0x456/0x720 [overlay]
[ 2300.910736]  ovl_copy_up_flags+0x81/0x110 [overlay]
[ 2300.911740]  ovl_create_object+0x51/0x130 [overlay]
[ 2300.912719]  vfs_mkdir+0x194/0x250
[ 2300.913470]  do_mkdirat+0x14c/0x180
[ 2300.914223]  __x64_sys_mkdir+0x46/0x70
[ 2300.915019]  do_syscall_64+0x7d/0x160
[ 2300.922814]  entry_SYSCALL_64_after_hwframe+0x76/0x7e

The cause is that overlayfs does some feature checks on superblock
creation using capable/ve_capable and it stores current credentials.
But ve_capable check result depends on both current creds and
current VE. Later, overlayfs switches credentials only, leading
to unexpected feature disappearance.

1. Patch 1 removes separate STOPPED state - VE goes directly to DEAD.
2. Patch 2 adds helpers for overriding tasks VE.
3. Patch 3 delays ve cred drop until final destruction of VE.
4. Patch 4 adds VE owner to overlay superblock.
5. Patch 5 fixes the bug by switching VE together with creds in overlayfs.

As a side effect of this series - CT1 overlayfs mounts won't work
in CT2 and this is a good thing.
There's still a gap - creds may be overrided in other places and
we should handle them - a separate VSTOR-155347 task is for it.

https://virtuozzo.atlassian.net/browse/VSTOR-143259 - non-CRIU repro
https://virtuozzo.atlassian.net/browse/VSTOR-142405

Signed-off-by: Vladimir Riabchun <vladimir.riabchun@virtuozzo.com>
---
Changes in v6:
- Fixed various typos spotted by Kostya.
- Updated comments in commit messages and code.
- Rebased on rh10-6.12.0-211.75.1.24.1.vz10, overlayfs got one more
  place with creds override ovl_copyfile.
- Patch 2: helper always returns -EPERM for non-host switches.
- Patch 5: CT -> Host switch is handled by overlayfs now, code
  updated with Kostya's suggestions.
- Link to v5: http://lore.virtuozzo.com/20261110-ovl-use-orig-ve-v5-0-f3f2bdb29e23@virtuozzo.com

Changes in v5:
- Patch 2: report non-host VE changes, also treat VEx -> VE0 and VEx -> VEy
  differently. VEx -> VE0 might be allowed, VEx -> VEy - never.
  Replaced most WARN_ON with normal warnings or WARN_ON_ONCE.
- Patch 3: updated commit name.
- Patch 5: allow VEx -> VE0 case. Now creds are switched without VE.
- Link to v4: http://lore.virtuozzo.com/20261109-ovl-use-orig-ve-v4-0-88e65484dc5b@virtuozzo.com

Changes in v4:
- Patch 2: clean non-NULL overrided_ve in fork and report it,
  Report revert_ve(ERR_PTR) usages and reject them.
- Patch 5: changed ovl_override_creds signature so it returns
  error. Added checks for it in all places to avoid.
- Link to v3: http://lore.virtuozzo.com/20261103-ovl-use-orig-ve-v3-0-add042cdd087@virtuozzo.com

Changes in v3:
- Patch 2: added a separate field for ve_override to fix problems Pasha spotted.
- Patch 3: new one, delays init_cred destroy so dead CT can still use it.
- Link to v2: http://lore.virtuozzo.com/20261026-ovl-use-orig-ve-v2-0-eab58bfc29ca@virtuozzo.com

Changes in v2:
- Patch 1: updated description as Sashiko suggested..
- Patch 2: moved stubs outside CONFIG_X86 section,
  added task_lock protection in helpers. Verified via lockdep,
  no splats after vzctl tests, C/R and crash reproducer.
- Patch 4: fixed typo in comment.
- Link to v1: http://lore.virtuozzo.com/20261022-ovl-use-orig-ve-v1-0-da84c00c8065@virtuozzo.com

---
Vladimir Riabchun (5):
      ve: Remove STOPPED state
      ve: Introduce override_ve/revert_ve helpers
      ve: Defer ve creds drop until ve is completely destroyed
      ovl: Keep creator VE in overlayfs superblock
      ve/overlayfs: Override VE together with creds

 fs/overlayfs/copy_up.c   |  6 ++--
 fs/overlayfs/dir.c       | 59 ++++++++++++++++++------------
 fs/overlayfs/file.c      | 71 +++++++++++++++++++++---------------
 fs/overlayfs/inode.c     | 94 +++++++++++++++++++++++++++++++-----------------
 fs/overlayfs/namei.c     | 39 +++++++++++++-------
 fs/overlayfs/overlayfs.h | 10 +++++-
 fs/overlayfs/ovl_entry.h |  1 +
 fs/overlayfs/params.c    |  3 ++
 fs/overlayfs/readdir.c   | 36 ++++++++++++-------
 fs/overlayfs/super.c     |  3 ++
 fs/overlayfs/util.c      | 52 ++++++++++++++++++++++-----
 fs/overlayfs/xattrs.c    | 35 +++++++++++-------
 include/linux/sched.h    |  1 +
 include/linux/ve.h       | 34 ++++++++++++++++--
 kernel/capability.c      | 10 +++---
 kernel/fork.c            |  2 ++
 kernel/ve/ve.c           | 10 +++---
 17 files changed, 322 insertions(+), 144 deletions(-)
---
base-commit: 7f638aced7fd165f1a711b8bbd90f63ffaa8dd9e
change-id: 20260918-ovl-use-orig-ve-2dc365976202

Best regards,
--  
Vladimir Riabchun <vladimir.riabchun@virtuozzo.com>

_______________________________________________
Devel mailing list
Devel@openvz.org
https://lists.openvz.org/mailman/listinfo/devel

^ permalink raw reply	[flat|nested] 13+ messages in thread

* [Devel] [PATCH hci-8.1 v6 1/5] ve: Remove STOPPED state
  2026-11-16 16:06 [Devel] [PATCH hci-8.1 v6 0/5] Fix kernel crash in overlayfs Vladimir Riabchun
@ 2026-11-16 16:06 ` Vladimir Riabchun
  2026-11-16 16:10   ` no-reply
  2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 2/5] ve: Introduce override_ve/revert_ve helpers Vladimir Riabchun
                   ` (3 subsequent siblings)
  4 siblings, 1 reply; 13+ messages in thread
From: Vladimir Riabchun @ 2026-11-16 16:06 UTC (permalink / raw)
  To: vasileios.almpanis, ptikhomirov, Konstantin Khorenko; +Cc: devel

We don't need extra STOPPED state. It makes us feel that CT
could be started again from it, but it is actually not true.

Separate STOPPED state doesn't bring any value. Drop it and
transition VE to DEAD state directly on the stop.

https://virtuozzo.atlassian.net/browse/VSTOR-142405
Reviewed-by: Pavel Tikhomirov <ptikhomirov@virtuozzo.com>
Signed-off-by: Vladimir Riabchun <vladimir.riabchun@virtuozzo.com>
---
 include/linux/ve.h | 3 +--
 kernel/ve/ve.c     | 6 ++----
 2 files changed, 3 insertions(+), 6 deletions(-)

diff --git a/include/linux/ve.h b/include/linux/ve.h
index a400d21593b2..717567775f58 100644
--- a/include/linux/ve.h
+++ b/include/linux/ve.h
@@ -40,8 +40,7 @@ struct vfsmount;
 #define VE_STATE_STARTING	0
 #define VE_STATE_RUNNING	1
 #define VE_STATE_STOPPING	2
-#define VE_STATE_STOPPED	3
-#define VE_STATE_DEAD		4
+#define VE_STATE_DEAD		3
 
 #define VE_IS_RUNNING(ve) ((ve)->state == VE_STATE_RUNNING)
 
diff --git a/kernel/ve/ve.c b/kernel/ve/ve.c
index 6cdfaaa4e703..b3988a8d590f 100644
--- a/kernel/ve/ve.c
+++ b/kernel/ve/ve.c
@@ -608,7 +608,7 @@ static int ve_start_container(struct ve_struct *ve)
 err_mark_ve:
 	ve_list_del(ve);
 err_list:
-	ve_set_state(ve, VE_STATE_STOPPED);
+	ve_set_state(ve, VE_STATE_DEAD);
 	ve_drop_context(ve);
 	return err;
 }
@@ -684,7 +684,7 @@ void ve_exit_ns(struct pid_namespace *pid_ns)
 	ve_hook_iterate_fini(VE_SS_CHAIN, ve);
 	ve_list_del(ve);
 	ve_drop_context(ve);
-	ve_set_state(ve, VE_STATE_STOPPED);
+	ve_set_state(ve, VE_STATE_DEAD);
 	up_write(&ve->op_sem);
 
 	printk(KERN_INFO "CT: %s: stopped\n", ve_name(ve));
@@ -939,7 +939,6 @@ static void ve_destroy(struct cgroup_subsys_state *css)
 	ve_free_vdso(ve);
 	mntput(ve->devtmpfs_mnt);
 	free_percpu(ve->sched_lat_ve.cur);
-	ve_set_state(ve, VE_STATE_DEAD);
 	kmem_cache_free(ve_cachep, ve);
 }
 
@@ -959,7 +958,6 @@ static int ve_state_show(struct seq_file *sf, void *v)
 	case VE_STATE_STOPPING:
 		seq_puts(sf, "STOPPING");
 		break;
-	case VE_STATE_STOPPED:
 	case VE_STATE_DEAD:
 		seq_puts(sf, "STOPPED");
 		break;

-- 
2.52.0

_______________________________________________
Devel mailing list
Devel@openvz.org
https://lists.openvz.org/mailman/listinfo/devel

^ permalink raw reply	[flat|nested] 13+ messages in thread

* [Devel] [PATCH hci-8.1 v6 2/5] ve: Introduce override_ve/revert_ve helpers
  2026-11-16 16:06 [Devel] [PATCH hci-8.1 v6 0/5] Fix kernel crash in overlayfs Vladimir Riabchun
  2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 1/5] ve: Remove STOPPED state Vladimir Riabchun
@ 2026-11-16 16:06 ` Vladimir Riabchun
  2026-11-16 16:12   ` no-reply
  2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 3/5] ve: Defer ve creds drop until ve is completely destroyed Vladimir Riabchun
                   ` (2 subsequent siblings)
  4 siblings, 1 reply; 13+ messages in thread
From: Vladimir Riabchun @ 2026-11-16 16:06 UTC (permalink / raw)
  To: vasileios.almpanis, ptikhomirov, Konstantin Khorenko; +Cc: devel

These helpers mirror override_creds/revert_creds and are supposed to
be used together in cases when VE creds may be stored and later used
by the different VE task.

To keep task_ve consistent for other users, introduce a separate
task->overrided_ve field. Nobody should race on it since its
updates are supposed to go with override_creds, and non-current
processes shouldn't do this.

This field is not supposed to be inherited via fork, so a check
for non-NULL overrided_ve was added in fork path.

https://virtuozzo.atlassian.net/browse/VSTOR-142405
Signed-off-by: Vladimir Riabchun <vladimir.riabchun@virtuozzo.com>
---
 include/linux/sched.h |  1 +
 include/linux/ve.h    | 31 +++++++++++++++++++++++++++++++
 kernel/capability.c   |  6 ++++--
 kernel/fork.c         |  2 ++
 4 files changed, 38 insertions(+), 2 deletions(-)

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 10545e470d3d..7f4e6b1fde38 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1065,6 +1065,7 @@ struct task_struct {
 #ifdef CONFIG_VE
 	struct ve_struct		*task_ve;
 	struct ve_namespace		*ve_ns;
+	struct ve_struct		*overrided_ve; /* NULL when no overrides */
 #endif
 #ifdef CONFIG_PREEMPT_RT
 	struct netdev_xmit		net_xmit;
diff --git a/include/linux/ve.h b/include/linux/ve.h
index 717567775f58..f5260bf25c00 100644
--- a/include/linux/ve.h
+++ b/include/linux/ve.h
@@ -295,6 +295,30 @@ static inline struct vdso_image *ve_get_vdso_64(struct ve_struct *ve)
 	return ve->vdso_64;
 }
 
+static inline struct ve_struct *override_ve(struct ve_struct *new_ve)
+{
+	struct ve_struct *old_ve = current->overrided_ve;
+
+	/* CT tasks are not supposed to switch VE context, this is a breach */
+	if (new_ve != get_exec_env() && !ve_is_super(get_exec_env())) {
+		pr_err_ratelimited("Invalid VE override: %s -> %s\n",
+			ve_name(get_exec_env()),
+			VE_IS_RUNNING(new_ve) ? ve_name(new_ve) : "(dead VE)");
+		return ERR_PTR(-EPERM);
+	}
+
+	current->overrided_ve = new_ve;
+	return old_ve;
+}
+
+static inline void revert_ve(struct ve_struct *orig_ve)
+{
+	/* Somebody forgot to check override_ve return value? */
+	if (WARN_ON_ONCE(IS_ERR(orig_ve)))
+		return;
+	current->overrided_ve = orig_ve;
+}
+
 #else	/* CONFIG_VE */
 #include <linux/init_task.h>
 #include <net/net_namespace.h>
@@ -405,6 +429,13 @@ static inline struct vdso_image *ve_get_vdso_64(struct ve_struct *ve)
 }
 #endif /* CONFIG_X86 */
 
+static inline struct ve_struct *override_ve(struct ve_struct *new_ve)
+{
+	return NULL;
+}
+
+static inline void revert_ve(struct ve_struct *orig_ve) {}
+
 #endif	/* CONFIG_VE */
 
 struct seq_file;
diff --git a/kernel/capability.c b/kernel/capability.c
index 34716847c6b8..6de4dcba2283 100644
--- a/kernel/capability.c
+++ b/kernel/capability.c
@@ -389,10 +389,11 @@ EXPORT_SYMBOL(ns_capable);
 bool ve_capable(int cap)
 {
 	struct cred *cred;
+	struct ve_struct *ve = current->overrided_ve ?: get_exec_env();
 	bool ret;
 
 	rcu_read_lock();
-	cred = get_exec_env()->init_cred;
+	cred = ve->init_cred;
 
 	if (cred == NULL) /* ve isn't running */
 		cred = ve0.init_cred;
@@ -406,10 +407,11 @@ bool ve_capable(int cap)
 bool ve_capable_noaudit(int cap)
 {
 	struct cred *cred;
+	struct ve_struct *ve = current->overrided_ve ?: get_exec_env();
 	bool ret;
 
 	rcu_read_lock();
-	cred = get_exec_env()->init_cred;
+	cred = ve->init_cred;
 
 	if (cred == NULL) /* ve isn't running */
 		cred = ve0.init_cred;
diff --git a/kernel/fork.c b/kernel/fork.c
index 25e87cdbeb07..0b4c4554bf80 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -1161,6 +1161,8 @@ static struct task_struct *dup_task_struct(struct task_struct *orig, int node)
 
 #ifdef CONFIG_VE
 	memset(tsk->alloc_lat, 0, sizeof(tsk->alloc_lat));
+	WARN_ON_ONCE(orig->overrided_ve);
+	tsk->overrided_ve = NULL;
 #endif
 
 	/*

-- 
2.52.0

_______________________________________________
Devel mailing list
Devel@openvz.org
https://lists.openvz.org/mailman/listinfo/devel

^ permalink raw reply	[flat|nested] 13+ messages in thread

* [Devel] [PATCH hci-8.1 v6 3/5] ve: Defer ve creds drop until ve is completely destroyed
  2026-11-16 16:06 [Devel] [PATCH hci-8.1 v6 0/5] Fix kernel crash in overlayfs Vladimir Riabchun
  2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 1/5] ve: Remove STOPPED state Vladimir Riabchun
  2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 2/5] ve: Introduce override_ve/revert_ve helpers Vladimir Riabchun
@ 2026-11-16 16:06 ` Vladimir Riabchun
  2026-11-16 16:14   ` no-reply
  2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 4/5] ovl: Keep creator VE in overlayfs superblock Vladimir Riabchun
  2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 5/5] ve/overlayfs: Override VE together with creds Vladimir Riabchun
  4 siblings, 1 reply; 13+ messages in thread
From: Vladimir Riabchun @ 2026-11-16 16:06 UTC (permalink / raw)
  To: vasileios.almpanis, ptikhomirov, Konstantin Khorenko; +Cc: devel

We may have a case, when VE is dead, but there are some users of it,
overlayfs superblock for example.

This user may want to use VE creds, so we need to keep ve->init_cred alive.
Drop creds in ve_destroy. We need to drop ve_nsproxy early because
ploop waits for refcount going to zero, dropping ve_nsproxy in destroy
causes CT suspend/resume hangs.

https://virtuozzo.atlassian.net/browse/VSTOR-142405
Signed-off-by: Vladimir Riabchun <vladimir.riabchun@virtuozzo.com>
---
 kernel/capability.c | 4 ++--
 kernel/ve/ve.c      | 4 ++++
 2 files changed, 6 insertions(+), 2 deletions(-)

diff --git a/kernel/capability.c b/kernel/capability.c
index 6de4dcba2283..68c3d5288ba7 100644
--- a/kernel/capability.c
+++ b/kernel/capability.c
@@ -395,7 +395,7 @@ bool ve_capable(int cap)
 	rcu_read_lock();
 	cred = ve->init_cred;
 
-	if (cred == NULL) /* ve isn't running */
+	if (cred == NULL) /* ve isn't started */
 		cred = ve0.init_cred;
 
 	ret = ns_capable(cred->user_ns, cap);
@@ -413,7 +413,7 @@ bool ve_capable_noaudit(int cap)
 	rcu_read_lock();
 	cred = ve->init_cred;
 
-	if (cred == NULL) /* ve isn't running */
+	if (cred == NULL) /* ve isn't started */
 		cred = ve0.init_cred;
 
 	ret = ns_capable_noaudit(cred->user_ns, cap);
diff --git a/kernel/ve/ve.c b/kernel/ve/ve.c
index b3988a8d590f..1e0c092f080e 100644
--- a/kernel/ve/ve.c
+++ b/kernel/ve/ve.c
@@ -378,7 +378,10 @@ static void ve_drop_context(struct ve_struct *ve)
 	rcu_assign_pointer(ve->ve_nsproxy, NULL);
 	synchronize_rcu();
 	put_nsproxy(ve_nsproxy);
+}
 
+static void ve_drop_cred(struct ve_struct *ve)
+{
 	put_cred(ve->init_cred);
 	ve->init_cred = NULL;
 }
@@ -929,6 +932,7 @@ static void ve_destroy(struct cgroup_subsys_state *css)
 {
 	struct ve_struct *ve = css_to_ve(css);
 
+	ve_drop_cred(ve);
 	free_ve_devmnts(ve);
 
 	kmapset_unlink(&ve->sysfs_perms_key, &sysfs_ve_perms_set);

-- 
2.52.0

_______________________________________________
Devel mailing list
Devel@openvz.org
https://lists.openvz.org/mailman/listinfo/devel

^ permalink raw reply	[flat|nested] 13+ messages in thread

* [Devel] [PATCH hci-8.1 v6 4/5] ovl: Keep creator VE in overlayfs superblock
  2026-11-16 16:06 [Devel] [PATCH hci-8.1 v6 0/5] Fix kernel crash in overlayfs Vladimir Riabchun
                   ` (2 preceding siblings ...)
  2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 3/5] ve: Defer ve creds drop until ve is completely destroyed Vladimir Riabchun
@ 2026-11-16 16:06 ` Vladimir Riabchun
  2026-11-16 16:13   ` no-reply
  2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 5/5] ve/overlayfs: Override VE together with creds Vladimir Riabchun
  4 siblings, 1 reply; 13+ messages in thread
From: Vladimir Riabchun @ 2026-11-16 16:06 UTC (permalink / raw)
  To: vasileios.almpanis, ptikhomirov, Konstantin Khorenko; +Cc: devel

Overlayfs superblock stores creator creds in it, but these creds
should go together with the VE they were taken from.

Store the VE in the superblock. We can't use mount->owner_ve,
because several mounts can reference one superblock.

Feature: ve/overlayfs
https://virtuozzo.atlassian.net/browse/VSTOR-142405
Signed-off-by: Vladimir Riabchun <vladimir.riabchun@virtuozzo.com>
---
 fs/overlayfs/ovl_entry.h | 1 +
 fs/overlayfs/params.c    | 3 +++
 fs/overlayfs/super.c     | 3 +++
 3 files changed, 7 insertions(+)

diff --git a/fs/overlayfs/ovl_entry.h b/fs/overlayfs/ovl_entry.h
index ab78001635e7..a6ad8b8de48c 100644
--- a/fs/overlayfs/ovl_entry.h
+++ b/fs/overlayfs/ovl_entry.h
@@ -77,6 +77,7 @@ struct ovl_fs {
 	struct ovl_config config;
 	/* creds of process who forced instantiation of super block */
 	const struct cred *creator_cred;
+	struct ve_struct *creator_ve;
 	bool tmpfile;
 	bool noxattr;
 	bool nofh;
diff --git a/fs/overlayfs/params.c b/fs/overlayfs/params.c
index 72a522d5fa17..b394e30741f4 100644
--- a/fs/overlayfs/params.c
+++ b/fs/overlayfs/params.c
@@ -8,6 +8,7 @@
 #include <linux/posix_acl_xattr.h>
 #include <linux/seq_file.h>
 #include <linux/xattr.h>
+#include <linux/ve.h>
 #include "overlayfs.h"
 #include "params.h"
 
@@ -758,6 +759,8 @@ void ovl_free_fs(struct ovl_fs *ofs)
 	kfree(ofs->config.workdir);
 	if (ofs->creator_cred)
 		put_cred(ofs->creator_cred);
+	if (ofs->creator_ve)
+		put_ve(ofs->creator_ve);
 	kfree(ofs);
 }
 
diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c
index 8153362caad1..beb08ef2f156 100644
--- a/fs/overlayfs/super.c
+++ b/fs/overlayfs/super.c
@@ -18,6 +18,7 @@
 #include <linux/file.h>
 #include <linux/fs_context.h>
 #include <linux/fs_parser.h>
+#include <linux/ve.h>
 #include "overlayfs.h"
 #include "params.h"
 
@@ -1328,6 +1329,8 @@ int ovl_fill_super(struct super_block *sb, struct fs_context *fc)
 
 	sb->s_d_op = &ovl_dentry_operations;
 
+	ofs->creator_ve = get_ve(get_exec_env());
+
 	err = -ENOMEM;
 	ofs->creator_cred = cred = prepare_creds();
 	if (!cred)

-- 
2.52.0

_______________________________________________
Devel mailing list
Devel@openvz.org
https://lists.openvz.org/mailman/listinfo/devel

^ permalink raw reply	[flat|nested] 13+ messages in thread

* [Devel] [PATCH hci-8.1 v6 5/5] ve/overlayfs: Override VE together with creds
  2026-11-16 16:06 [Devel] [PATCH hci-8.1 v6 0/5] Fix kernel crash in overlayfs Vladimir Riabchun
                   ` (3 preceding siblings ...)
  2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 4/5] ovl: Keep creator VE in overlayfs superblock Vladimir Riabchun
@ 2026-11-16 16:06 ` Vladimir Riabchun
  2026-11-16 16:19   ` no-reply
  4 siblings, 1 reply; 13+ messages in thread
From: Vladimir Riabchun @ 2026-11-16 16:06 UTC (permalink / raw)
  To: vasileios.almpanis, ptikhomirov, Konstantin Khorenko; +Cc: devel

Host users might want to access overlayfs owned by CT. In such case
ve_capable checks will fail - they will perform VE cred checks against
host user_ns.

Modify ovl_override_creds so that it returns both old cred and old VE.
This will also protect us from future rebase issues - all new users of
old_cred and ovl_override_creds will fail with compilation error.

We don't allow cross-CT overlayfs usages - container creds will not
work inside other CT, which will trigger the bug (see description
below). Such accesses will fail with -EPERM.

But we want to allow usage of host-owned overlayfs by CT.
In this case we keep CT VE and override creds only. Host creds are
expected to work inside container.

This fixes an observed crash in overlayfs when checkpointing
a container with a docker running inside:

[ 2300.870181] BUG: kernel NULL pointer dereference, address: 0000000000000005
[ 2300.871537] #PF: supervisor read access in kernel mode
[ 2300.872577] #PF: error_code(0x0000) - not-present page
[ 2300.873595] PGD 800000010a18f067 P4D 800000010a18f067 PUD 10730c067 PMD 0
[ 2300.874900] Oops: Oops: 0000 [#1] SMP PTI
[ 2300.875742] CPU: 1 UID: 0 PID: 5419 Comm: criu ve: 0 Kdump: loaded Not tainted 6.12.0-211.39.1.16.6.vz10.x86_64 #1 16.6 PREEMPT(voluntary)
[ 2300.878057] Hardware name: Acronis OpenStack Compute/Virtuozzo, BIOS 1.16.1-1.vz9.2 04/01/2014
[ 2300.879684] RIP: 0010:ovl_get_index_name_fh+0x18/0x70 [overlay]
[ 2300.880859] Code: ...
[ 2300.884128] RSP: 0018:ffffd32582e0f7c0 EFLAGS: 00010246
[ 2300.885153] RAX: ...
[ 2300.895807] Call Trace:
[ 2300.896397]
[ 2300.903298] ? ovl_get_index_name_fh+0x18/0x70 [overlay]
[ 2300.904435] ovl_create_index+0x82/0x260 [overlay]
[ 2300.905406] ovl_copy_up_workdir+0x30a/0x330 [overlay]
[ 2300.906427] ovl_do_copy_up+0x10d/0x2e0 [overlay]
[ 2300.907387] ovl_copy_up_one+0x456/0x720 [overlay]
[ 2300.910736] ovl_copy_up_flags+0x81/0x110 [overlay]
[ 2300.911740] ovl_create_object+0x51/0x130 [overlay]
[ 2300.912719] vfs_mkdir+0x194/0x250
[ 2300.913470] do_mkdirat+0x14c/0x180
[ 2300.914223] __x64_sys_mkdir+0x46/0x70
[ 2300.915019] do_syscall_64+0x7d/0x160
[ 2300.922814] entry_SYSCALL_64_after_hwframe+0x76/0x7e
[ 2300.923823] RIP: 0033:0x7f5a2f8de3eb

Overlayfs was created inside CT, so ovl_fill_super stored
in-container credentials to ofs->creator_cred.

Those creds were later used in ovl_override_creds to temporarily
switch creds and perform checks using them.

One of these checks happened in ovl_get_origin_fh/ovl_can_decode_fh.
On fs creation this check passed because we had container VE
and container creds, so overlayfs thought that we support
file handles (ovl_can_decode_fh call in ovl_make_workdir).

When we tried to do something with the overlayfs from host VE,
ve_capable check was performed with credentials from CT
and with host VE.

ovl_get_origin_fh returned unexpected NULL and all exploded.

Storing ve_capable check result on creation is a possible fix,
but it just masks the problem. We have other paths from overlay to
ve_capable checks:

ovl_fileattr_set
  ovl_override_creds
  ovl_real_fileattr_set
    vfs_fileattr_set
      fileattr_set_prepare
        ve_capable

And storing capability check result will make rebase dangerous - we
would have to track new added code - not funny.

https://virtuozzo.atlassian.net/browse/VSTOR-142405
Fixes: ec01a90d0ef9 ("ovl: replace capable by ve_capable for indexing feature")
Signed-off-by: Vladimir Riabchun <vladimir.riabchun@virtuozzo.com>
---
 fs/overlayfs/copy_up.c   |  6 ++--
 fs/overlayfs/dir.c       | 59 ++++++++++++++++++------------
 fs/overlayfs/file.c      | 71 +++++++++++++++++++++---------------
 fs/overlayfs/inode.c     | 94 +++++++++++++++++++++++++++++++-----------------
 fs/overlayfs/namei.c     | 39 +++++++++++++-------
 fs/overlayfs/overlayfs.h | 10 +++++-
 fs/overlayfs/readdir.c   | 36 ++++++++++++-------
 fs/overlayfs/util.c      | 52 ++++++++++++++++++++++-----
 fs/overlayfs/xattrs.c    | 35 +++++++++++-------
 9 files changed, 268 insertions(+), 134 deletions(-)

diff --git a/fs/overlayfs/copy_up.c b/fs/overlayfs/copy_up.c
index 4388004a319d..f478d1b51af9 100644
--- a/fs/overlayfs/copy_up.c
+++ b/fs/overlayfs/copy_up.c
@@ -1215,7 +1215,7 @@ static int ovl_copy_up_one(struct dentry *parent, struct dentry *dentry,
 static int ovl_copy_up_flags(struct dentry *dentry, int flags)
 {
 	int err = 0;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	bool disconnected = (dentry->d_flags & DCACHE_DISCONNECTED);
 
 	/*
@@ -1235,7 +1235,7 @@ static int ovl_copy_up_flags(struct dentry *dentry, int flags)
 	if (err)
 		return err;
 
-	old_cred = ovl_override_creds(dentry->d_sb);
+	err = ovl_override_creds(dentry->d_sb, &old_cred);
 	while (!err) {
 		struct dentry *next;
 		struct dentry *parent = NULL;
@@ -1260,7 +1260,7 @@ static int ovl_copy_up_flags(struct dentry *dentry, int flags)
 		dput(parent);
 		dput(next);
 	}
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 
 	return err;
 }
diff --git a/fs/overlayfs/dir.c b/fs/overlayfs/dir.c
index 1c8009bf194b..23897fabe9d5 100644
--- a/fs/overlayfs/dir.c
+++ b/fs/overlayfs/dir.c
@@ -554,7 +554,7 @@ static int ovl_create_over_whiteout(struct dentry *dentry, struct inode *inode,
 }
 
 static int ovl_setup_cred_for_create(struct dentry *dentry, struct inode *inode,
-				     umode_t mode, const struct cred *old_cred)
+				     umode_t mode, const struct ovl_creds_desc *old_cred)
 {
 	int err;
 	struct cred *override_cred;
@@ -566,7 +566,7 @@ static int ovl_setup_cred_for_create(struct dentry *dentry, struct inode *inode,
 	override_cred->fsuid = inode->i_uid;
 	override_cred->fsgid = inode->i_gid;
 	err = security_dentry_create_files_as(dentry, mode, &dentry->d_name,
-					      old_cred, override_cred);
+					      old_cred->cred, override_cred);
 	if (err) {
 		put_cred(override_cred);
 		return err;
@@ -581,10 +581,12 @@ static int ovl_create_or_link(struct dentry *dentry, struct inode *inode,
 			      struct ovl_cattr *attr, bool origin)
 {
 	int err;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	struct dentry *parent = dentry->d_parent;
 
-	old_cred = ovl_override_creds(dentry->d_sb);
+	err = ovl_override_creds(dentry->d_sb, &old_cred);
+	if (err)
+		return err;
 
 	/*
 	 * When linking a file with copy up origin into a new parent, mark the
@@ -610,7 +612,7 @@ static int ovl_create_or_link(struct dentry *dentry, struct inode *inode,
 		 * create a new inode, so just use the ovl mounter's
 		 * fs{u,g}id.
 		 */
-		err = ovl_setup_cred_for_create(dentry, inode, attr->mode, old_cred);
+		err = ovl_setup_cred_for_create(dentry, inode, attr->mode, &old_cred);
 		if (err)
 			goto out_revert_creds;
 	}
@@ -621,7 +623,7 @@ static int ovl_create_or_link(struct dentry *dentry, struct inode *inode,
 		err = ovl_create_over_whiteout(dentry, inode, attr);
 
 out_revert_creds:
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 	return err;
 }
 
@@ -697,12 +699,15 @@ static int ovl_symlink(struct mnt_idmap *idmap, struct inode *dir,
 
 static int ovl_set_link_redirect(struct dentry *dentry)
 {
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	int err;
 
-	old_cred = ovl_override_creds(dentry->d_sb);
+	err = ovl_override_creds(dentry->d_sb, &old_cred);
+	if (err)
+		return err;
+
 	err = ovl_set_redirect(dentry, false);
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 
 	return err;
 }
@@ -888,7 +893,7 @@ static void ovl_drop_nlink(struct dentry *dentry)
 static int ovl_do_remove(struct dentry *dentry, bool is_dir)
 {
 	int err;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	bool lower_positive = ovl_lower_positive(dentry);
 	LIST_HEAD(list);
 
@@ -907,12 +912,15 @@ static int ovl_do_remove(struct dentry *dentry, bool is_dir)
 	if (err)
 		goto out;
 
-	old_cred = ovl_override_creds(dentry->d_sb);
-	if (!lower_positive)
-		err = ovl_remove_upper(dentry, is_dir, &list);
-	else
-		err = ovl_remove_and_whiteout(dentry, &list);
-	revert_creds(old_cred);
+	err = ovl_override_creds(dentry->d_sb, &old_cred);
+	if (!err) {
+		if (!lower_positive)
+			err = ovl_remove_upper(dentry, is_dir, &list);
+		else
+			err = ovl_remove_and_whiteout(dentry, &list);
+		ovl_revert_creds(&old_cred);
+	}
+
 	if (!err) {
 		if (is_dir)
 			clear_nlink(dentry->d_inode);
@@ -1095,7 +1103,7 @@ static int ovl_rename(struct mnt_idmap *idmap, struct inode *olddir,
 	bool new_is_dir = d_is_dir(new);
 	bool samedir = olddir == newdir;
 	struct dentry *opaquedir = NULL;
-	const struct cred *old_cred = NULL;
+	struct ovl_creds_desc old_cred = {};
 	struct ovl_fs *ofs = OVL_FS(old->d_sb);
 	LIST_HEAD(list);
 
@@ -1159,7 +1167,9 @@ static int ovl_rename(struct mnt_idmap *idmap, struct inode *olddir,
 			goto out;
 	}
 
-	old_cred = ovl_override_creds(old->d_sb);
+	err = ovl_override_creds(old->d_sb, &old_cred);
+	if (err)
+		goto out_revert_creds;
 
 	if (!list_empty(&list)) {
 		opaquedir = ovl_clear_empty(new, &list);
@@ -1292,7 +1302,7 @@ static int ovl_rename(struct mnt_idmap *idmap, struct inode *olddir,
 out_unlock:
 	unlock_rename(new_upperdir, old_upperdir);
 out_revert_creds:
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 	if (update_nlink)
 		ovl_nlink_end(new);
 	else
@@ -1306,7 +1316,7 @@ static int ovl_rename(struct mnt_idmap *idmap, struct inode *olddir,
 static int ovl_create_tmpfile(struct file *file, struct dentry *dentry,
 			      struct inode *inode, umode_t mode)
 {
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	struct path realparentpath;
 	struct file *realfile;
 	struct dentry *newdentry;
@@ -1314,8 +1324,11 @@ static int ovl_create_tmpfile(struct file *file, struct dentry *dentry,
 	int flags = file->f_flags | OVL_OPEN_FLAGS;
 	int err;
 
-	old_cred = ovl_override_creds(dentry->d_sb);
-	err = ovl_setup_cred_for_create(dentry, inode, mode, old_cred);
+	err = ovl_override_creds(dentry->d_sb, &old_cred);
+	if (err)
+		return err;
+
+	err = ovl_setup_cred_for_create(dentry, inode, mode, &old_cred);
 	if (err)
 		goto out_revert_creds;
 
@@ -1337,7 +1350,7 @@ static int ovl_create_tmpfile(struct file *file, struct dentry *dentry,
 		fput(realfile);
 	}
 out_revert_creds:
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 	return err;
 }
 
diff --git a/fs/overlayfs/file.c b/fs/overlayfs/file.c
index 2531433f47bf..464699e51695 100644
--- a/fs/overlayfs/file.c
+++ b/fs/overlayfs/file.c
@@ -31,7 +31,7 @@ static struct file *ovl_open_realfile(const struct file *file,
 	struct inode *inode = file_inode(file);
 	struct mnt_idmap *real_idmap;
 	struct file *realfile;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	int flags = file->f_flags | OVL_OPEN_FLAGS;
 	int acc_mode = ACC_MODE(flags);
 	int err;
@@ -39,7 +39,10 @@ static struct file *ovl_open_realfile(const struct file *file,
 	if (flags & O_APPEND)
 		acc_mode |= MAY_APPEND;
 
-	old_cred = ovl_override_creds(inode->i_sb);
+	err = ovl_override_creds(inode->i_sb, &old_cred);
+	if (err)
+		return ERR_PTR(err);
+
 	real_idmap = mnt_idmap(realpath->mnt);
 	err = inode_permission(real_idmap, realinode, MAY_OPEN | acc_mode);
 	if (err) {
@@ -51,7 +54,7 @@ static struct file *ovl_open_realfile(const struct file *file,
 		realfile = backing_file_open(file, flags, realpath,
 					     current_cred());
 	}
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 
 	pr_debug("open(%p[%pD2/%c], 0%o) -> (%p, 0%o)\n",
 		 file, file, ovl_whatisit(inode, realinode), file->f_flags,
@@ -184,7 +187,7 @@ static loff_t ovl_llseek(struct file *file, loff_t offset, int whence)
 {
 	struct inode *inode = file_inode(file);
 	struct fd real;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	loff_t ret;
 
 	/*
@@ -213,9 +216,10 @@ static loff_t ovl_llseek(struct file *file, loff_t offset, int whence)
 	ovl_inode_lock(inode);
 	fd_file(real)->f_pos = file->f_pos;
 
-	old_cred = ovl_override_creds(inode->i_sb);
-	ret = vfs_llseek(fd_file(real), offset, whence);
-	revert_creds(old_cred);
+	ret = ovl_override_creds(inode->i_sb, &old_cred);
+	if (!ret)
+		ret = vfs_llseek(fd_file(real), offset, whence);
+	ovl_revert_creds(&old_cred);
 
 	file->f_pos = fd_file(real)->f_pos;
 	ovl_inode_unlock(inode);
@@ -392,7 +396,7 @@ static ssize_t ovl_splice_write(struct pipe_inode_info *pipe, struct file *out,
 static int ovl_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct fd real;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	int ret;
 
 	ret = ovl_sync_status(OVL_FS(file_inode(file)->i_sb));
@@ -405,9 +409,10 @@ static int ovl_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 
 	/* Don't sync lower file for fear of receiving EROFS error */
 	if (file_inode(fd_file(real)) == ovl_inode_upper(file_inode(file))) {
-		old_cred = ovl_override_creds(file_inode(file)->i_sb);
-		ret = vfs_fsync_range(fd_file(real), start, end, datasync);
-		revert_creds(old_cred);
+		ret = ovl_override_creds(file_inode(file)->i_sb, &old_cred);
+		if (!ret)
+			ret = vfs_fsync_range(fd_file(real), start, end, datasync);
+		ovl_revert_creds(&old_cred);
 	}
 
 	fdput(real);
@@ -431,7 +436,7 @@ static long ovl_fallocate(struct file *file, int mode, loff_t offset, loff_t len
 {
 	struct inode *inode = file_inode(file);
 	struct fd real;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	int ret;
 
 	inode_lock(inode);
@@ -445,9 +450,10 @@ static long ovl_fallocate(struct file *file, int mode, loff_t offset, loff_t len
 	if (ret)
 		goto out_unlock;
 
-	old_cred = ovl_override_creds(file_inode(file)->i_sb);
-	ret = vfs_fallocate(fd_file(real), mode, offset, len);
-	revert_creds(old_cred);
+	ret = ovl_override_creds(file_inode(file)->i_sb, &old_cred);
+	if (!ret)
+		ret = vfs_fallocate(fd_file(real), mode, offset, len);
+	ovl_revert_creds(&old_cred);
 
 	/* Update size */
 	ovl_file_modified(file);
@@ -463,16 +469,17 @@ static long ovl_fallocate(struct file *file, int mode, loff_t offset, loff_t len
 static int ovl_fadvise(struct file *file, loff_t offset, loff_t len, int advice)
 {
 	struct fd real;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	int ret;
 
 	ret = ovl_real_fdget(file, &real);
 	if (ret)
 		return ret;
 
-	old_cred = ovl_override_creds(file_inode(file)->i_sb);
-	ret = vfs_fadvise(fd_file(real), offset, len, advice);
-	revert_creds(old_cred);
+	ret = ovl_override_creds(file_inode(file)->i_sb, &old_cred);
+	if (!ret)
+		ret = vfs_fadvise(fd_file(real), offset, len, advice);
+	ovl_revert_creds(&old_cred);
 
 	fdput(real);
 
@@ -492,7 +499,7 @@ static loff_t ovl_copyfile(struct file *file_in, loff_t pos_in,
 	struct inode *inode_in = file_inode(file_in);
 	struct inode *inode_out = file_inode(file_out);
 	struct fd real_in, real_out;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	loff_t ret;
 
 	inode_lock(inode_out);
@@ -520,14 +527,19 @@ static loff_t ovl_copyfile(struct file *file_in, loff_t pos_in,
 	 * verify read access with the source mounter creds.
 	 */
 	if (unlikely(inode_in->i_sb != inode_out->i_sb)) {
-		old_cred = ovl_override_creds(inode_in->i_sb);
+		ret = ovl_override_creds(inode_in->i_sb, &old_cred);
+		if (ret)
+			goto out_unlock;
 		ret = rw_verify_area(READ, fd_file(real_in), &pos_in, len);
-		revert_creds(old_cred);
+		ovl_revert_creds(&old_cred);
 		if (unlikely(ret))
 			goto out_unlock;
 	}
 
-	old_cred = ovl_override_creds(inode_out->i_sb);
+	ret = ovl_override_creds(file_inode(file_out)->i_sb, &old_cred);
+	if (ret)
+		goto skip_switch;
+
 	switch (op) {
 	case OVL_COPY:
 		ret = vfs_copy_file_range(fd_file(real_in), pos_in,
@@ -545,7 +557,9 @@ static loff_t ovl_copyfile(struct file *file_in, loff_t pos_in,
 						flags);
 		break;
 	}
-	revert_creds(old_cred);
+
+skip_switch:
+	ovl_revert_creds(&old_cred);
 
 	/* Update size */
 	ovl_file_modified(file_out);
@@ -597,7 +611,7 @@ static loff_t ovl_remap_file_range(struct file *file_in, loff_t pos_in,
 static int ovl_flush(struct file *file, fl_owner_t id)
 {
 	struct fd real;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	int err;
 
 	err = ovl_real_fdget(file, &real);
@@ -605,9 +619,10 @@ static int ovl_flush(struct file *file, fl_owner_t id)
 		return err;
 
 	if (fd_file(real)->f_op->flush) {
-		old_cred = ovl_override_creds(file_inode(file)->i_sb);
-		err = fd_file(real)->f_op->flush(fd_file(real), id);
-		revert_creds(old_cred);
+		err = ovl_override_creds(file_inode(file)->i_sb, &old_cred);
+		if (!err)
+			err = fd_file(real)->f_op->flush(fd_file(real), id);
+		ovl_revert_creds(&old_cred);
 	}
 	fdput(real);
 
diff --git a/fs/overlayfs/inode.c b/fs/overlayfs/inode.c
index 8b31f44c12cd..2d27bafad986 100644
--- a/fs/overlayfs/inode.c
+++ b/fs/overlayfs/inode.c
@@ -25,7 +25,7 @@ int ovl_setattr(struct mnt_idmap *idmap, struct dentry *dentry,
 	struct ovl_fs *ofs = OVL_FS(dentry->d_sb);
 	bool full_copy_up = false;
 	struct dentry *upperdentry;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 
 	err = setattr_prepare(&nop_mnt_idmap, dentry, attr);
 	if (err)
@@ -78,9 +78,10 @@ int ovl_setattr(struct mnt_idmap *idmap, struct dentry *dentry,
 			goto out_put_write;
 
 		inode_lock(upperdentry->d_inode);
-		old_cred = ovl_override_creds(dentry->d_sb);
-		err = ovl_do_notify_change(ofs, upperdentry, attr);
-		revert_creds(old_cred);
+		err = ovl_override_creds(dentry->d_sb, &old_cred);
+		if (!err)
+			err = ovl_do_notify_change(ofs, upperdentry, attr);
+		ovl_revert_creds(&old_cred);
 		if (!err)
 			ovl_copyattr(dentry->d_inode);
 		inode_unlock(upperdentry->d_inode);
@@ -159,7 +160,7 @@ int ovl_getattr(struct mnt_idmap *idmap, const struct path *path,
 	struct dentry *dentry = path->dentry;
 	enum ovl_path_type type;
 	struct path realpath;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	struct inode *inode = d_inode(dentry);
 	bool is_dir = S_ISDIR(inode->i_mode);
 	int fsid = 0;
@@ -169,8 +170,10 @@ int ovl_getattr(struct mnt_idmap *idmap, const struct path *path,
 	metacopy_blocks = ovl_is_metacopy_dentry(dentry);
 
 	type = ovl_path_real(dentry, &realpath);
-	old_cred = ovl_override_creds(dentry->d_sb);
-	err = vfs_getattr_nosec(&realpath, stat, request_mask, flags);
+
+	err = ovl_override_creds(dentry->d_sb, &old_cred);
+	if (!err)
+		err = vfs_getattr_nosec(&realpath, stat, request_mask, flags);
 	if (err)
 		goto out;
 
@@ -280,7 +283,7 @@ int ovl_getattr(struct mnt_idmap *idmap, const struct path *path,
 		stat->nlink = dentry->d_inode->i_nlink;
 
 out:
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 
 	return err;
 }
@@ -291,7 +294,7 @@ int ovl_permission(struct mnt_idmap *idmap,
 	struct inode *upperinode = ovl_inode_upper(inode);
 	struct inode *realinode;
 	struct path realpath;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	int err;
 
 	/* Careful in RCU walk mode */
@@ -309,7 +312,10 @@ int ovl_permission(struct mnt_idmap *idmap,
 	if (err)
 		return err;
 
-	old_cred = ovl_override_creds(inode->i_sb);
+	err = ovl_override_creds(inode->i_sb, &old_cred);
+	if (err)
+		return err;
+
 	if (!upperinode &&
 	    !special_file(realinode->i_mode) && mask & MAY_WRITE) {
 		mask &= ~(MAY_WRITE | MAY_APPEND);
@@ -317,7 +323,7 @@ int ovl_permission(struct mnt_idmap *idmap,
 		mask |= MAY_READ;
 	}
 	err = inode_permission(mnt_idmap(realpath.mnt), realinode, mask);
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 
 	return err;
 }
@@ -326,15 +332,19 @@ static const char *ovl_get_link(struct dentry *dentry,
 				struct inode *inode,
 				struct delayed_call *done)
 {
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	const char *p;
+	int err;
 
 	if (!dentry)
 		return ERR_PTR(-ECHILD);
 
-	old_cred = ovl_override_creds(dentry->d_sb);
+	err = ovl_override_creds(dentry->d_sb, &old_cred);
+	if (err)
+		return ERR_PTR(err);
+
 	p = vfs_get_link(ovl_dentry_real(dentry), done);
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 	return p;
 }
 
@@ -465,11 +475,15 @@ struct posix_acl *do_ovl_get_acl(struct mnt_idmap *idmap,
 
 		acl = get_cached_acl_rcu(realinode, type);
 	} else {
-		const struct cred *old_cred;
+		struct ovl_creds_desc old_cred;
+		int err;
+
+		err = ovl_override_creds(inode->i_sb, &old_cred);
+		if (err)
+			return ERR_PTR(err);
 
-		old_cred = ovl_override_creds(inode->i_sb);
 		acl = ovl_get_acl_path(&realpath, posix_acl_xattr_name(type), noperm);
-		revert_creds(old_cred);
+		ovl_revert_creds(&old_cred);
 	}
 
 	return acl;
@@ -481,7 +495,7 @@ static int ovl_set_or_remove_acl(struct dentry *dentry, struct inode *inode,
 	int err;
 	struct path realpath;
 	const char *acl_name;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	struct ovl_fs *ofs = OVL_FS(dentry->d_sb);
 	struct dentry *upperdentry = ovl_dentry_upper(dentry);
 	struct dentry *realdentry = upperdentry ?: ovl_dentry_lower(dentry);
@@ -495,10 +509,14 @@ static int ovl_set_or_remove_acl(struct dentry *dentry, struct inode *inode,
 		struct posix_acl *real_acl;
 
 		ovl_path_lower(dentry, &realpath);
-		old_cred = ovl_override_creds(dentry->d_sb);
+
+		err = ovl_override_creds(dentry->d_sb, &old_cred);
+		if (err)
+			return err;
+
 		real_acl = vfs_get_acl(mnt_idmap(realpath.mnt), realdentry,
 				       acl_name);
-		revert_creds(old_cred);
+		ovl_revert_creds(&old_cred);
 		if (IS_ERR(real_acl)) {
 			err = PTR_ERR(real_acl);
 			goto out;
@@ -518,12 +536,17 @@ static int ovl_set_or_remove_acl(struct dentry *dentry, struct inode *inode,
 	if (err)
 		goto out;
 
-	old_cred = ovl_override_creds(dentry->d_sb);
+	err = ovl_override_creds(dentry->d_sb, &old_cred);
+	if (err)
+		goto out_revert;
+
 	if (acl)
 		err = ovl_do_set_acl(ofs, realdentry, acl_name, acl);
 	else
 		err = ovl_do_remove_acl(ofs, realdentry, acl_name);
-	revert_creds(old_cred);
+
+out_revert:
+	ovl_revert_creds(&old_cred);
 	ovl_drop_write(dentry);
 
 	/* copy c/mtime */
@@ -590,7 +613,7 @@ static int ovl_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 {
 	int err;
 	struct inode *realinode = ovl_inode_realdata(inode);
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 
 	if (!realinode)
 		return -EIO;
@@ -598,9 +621,10 @@ static int ovl_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	if (!realinode->i_op->fiemap)
 		return -EOPNOTSUPP;
 
-	old_cred = ovl_override_creds(inode->i_sb);
-	err = realinode->i_op->fiemap(realinode, fieinfo, start, len);
-	revert_creds(old_cred);
+	err = ovl_override_creds(inode->i_sb, &old_cred);
+	if (!err)
+		err = realinode->i_op->fiemap(realinode, fieinfo, start, len);
+	ovl_revert_creds(&old_cred);
 
 	return err;
 }
@@ -648,7 +672,7 @@ int ovl_fileattr_set(struct mnt_idmap *idmap,
 {
 	struct inode *inode = d_inode(dentry);
 	struct path upperpath;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	unsigned int flags;
 	int err;
 
@@ -660,7 +684,7 @@ int ovl_fileattr_set(struct mnt_idmap *idmap,
 		if (err)
 			goto out;
 
-		old_cred = ovl_override_creds(inode->i_sb);
+		err = ovl_override_creds(inode->i_sb, &old_cred);
 		/*
 		 * Store immutable/append-only flags in xattr and clear them
 		 * in upper fileattr (in case they were set by older kernel)
@@ -668,10 +692,11 @@ int ovl_fileattr_set(struct mnt_idmap *idmap,
 		 * "ovl-immutable" hardlinks could be copied up.
 		 * Clear xattr when flags are cleared.
 		 */
-		err = ovl_set_protattr(inode, upperpath.dentry, fa);
+		if (!err)
+			err = ovl_set_protattr(inode, upperpath.dentry, fa);
 		if (!err)
 			err = ovl_real_fileattr_set(&upperpath, fa);
-		revert_creds(old_cred);
+		ovl_revert_creds(&old_cred);
 		ovl_drop_write(dentry);
 
 		/*
@@ -725,15 +750,18 @@ int ovl_fileattr_get(struct dentry *dentry, struct fileattr *fa)
 {
 	struct inode *inode = d_inode(dentry);
 	struct path realpath;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	int err;
 
 	ovl_path_real(dentry, &realpath);
 
-	old_cred = ovl_override_creds(inode->i_sb);
+	err = ovl_override_creds(inode->i_sb, &old_cred);
+	if (err)
+		return err;
+
 	err = ovl_real_fileattr_get(&realpath, fa);
 	ovl_fileattr_prot_flags(inode, fa);
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 
 	return err;
 }
diff --git a/fs/overlayfs/namei.c b/fs/overlayfs/namei.c
index efa2a072aa69..edc11a7b27ee 100644
--- a/fs/overlayfs/namei.c
+++ b/fs/overlayfs/namei.c
@@ -969,15 +969,19 @@ static int ovl_maybe_validate_verity(struct dentry *dentry)
 		return err;
 
 	if (!ovl_test_flag(OVL_VERIFIED_DIGEST, inode)) {
-		const struct cred *old_cred;
+		struct ovl_creds_desc old_cred;
 
-		old_cred = ovl_override_creds(dentry->d_sb);
+		err = ovl_override_creds(dentry->d_sb, &old_cred);
+		if (err) {
+			ovl_inode_unlock(inode);
+			return err;
+		}
 
 		err = ovl_validate_verity(ofs, &metapath, &datapath);
 		if (err == 0)
 			ovl_set_flag(OVL_VERIFIED_DIGEST, inode);
 
-		revert_creds(old_cred);
+		ovl_revert_creds(&old_cred);
 	}
 
 	ovl_inode_unlock(inode);
@@ -991,7 +995,7 @@ static int ovl_maybe_lookup_lowerdata(struct dentry *dentry)
 	struct inode *inode = d_inode(dentry);
 	const char *redirect = ovl_lowerdata_redirect(inode);
 	struct ovl_path datapath = {};
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	int err;
 
 	if (!redirect || ovl_dentry_lowerdata(dentry))
@@ -1009,9 +1013,12 @@ static int ovl_maybe_lookup_lowerdata(struct dentry *dentry)
 	if (ovl_dentry_lowerdata(dentry))
 		goto out;
 
-	old_cred = ovl_override_creds(dentry->d_sb);
+	err = ovl_override_creds(dentry->d_sb, &old_cred);
+	if (err)
+		goto out_err;
+
 	err = ovl_lookup_data_layers(dentry, redirect, &datapath);
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 	if (err)
 		goto out_err;
 
@@ -1071,7 +1078,7 @@ struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry,
 			  unsigned int flags)
 {
 	struct ovl_entry *oe = NULL;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	struct ovl_fs *ofs = OVL_FS(dentry->d_sb);
 	struct ovl_entry *poe = OVL_E(dentry->d_parent);
 	struct ovl_entry *roe = OVL_E(dentry->d_sb->s_root);
@@ -1104,7 +1111,10 @@ struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry,
 	if (dentry->d_name.len > ofs->namelen)
 		return ERR_PTR(-ENAMETOOLONG);
 
-	old_cred = ovl_override_creds(dentry->d_sb);
+	err = ovl_override_creds(dentry->d_sb, &old_cred);
+	if (err)
+		return ERR_PTR(err);
+
 	upperdir = ovl_dentry_upper(dentry->d_parent);
 	if (upperdir) {
 		d.layer = &ofs->layers[0];
@@ -1378,7 +1388,7 @@ struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry,
 
 	ovl_dentry_init_reval(dentry, upperdentry, OVL_I_E(inode));
 
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 	if (origin_path) {
 		dput(origin_path->dentry);
 		kfree(origin_path);
@@ -1402,7 +1412,7 @@ struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry,
 	kfree(d.upperredirect);
 out:
 	kfree(d.redirect);
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 	return ERR_PTR(err);
 }
 
@@ -1410,7 +1420,7 @@ bool ovl_lower_positive(struct dentry *dentry)
 {
 	struct ovl_entry *poe = OVL_E(dentry->d_parent);
 	const struct qstr *name = &dentry->d_name;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	unsigned int i;
 	bool positive = false;
 	bool done = false;
@@ -1426,7 +1436,10 @@ bool ovl_lower_positive(struct dentry *dentry)
 	if (!ovl_dentry_upper(dentry))
 		return true;
 
-	old_cred = ovl_override_creds(dentry->d_sb);
+	/* Ugh, hard - we can't pass an error from here. */
+	if (ovl_override_creds(dentry->d_sb, &old_cred))
+		return false;
+
 	/* Positive upper -> have to look up lower to see whether it exists */
 	for (i = 0; !done && !positive && i < ovl_numlower(poe); i++) {
 		struct dentry *this;
@@ -1459,7 +1472,7 @@ bool ovl_lower_positive(struct dentry *dentry)
 			dput(this);
 		}
 	}
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 
 	return positive;
 }
diff --git a/fs/overlayfs/overlayfs.h b/fs/overlayfs/overlayfs.h
index 2331bd55302c..28c54c018f45 100644
--- a/fs/overlayfs/overlayfs.h
+++ b/fs/overlayfs/overlayfs.h
@@ -164,6 +164,13 @@ struct ovl_metacopy {
 	u8 digest[FS_VERITY_MAX_DIGEST_SIZE];  /* Only the used part on disk */
 } __packed;
 
+/* Credential information used by overlay */
+struct ovl_creds_desc {
+	const struct cred *cred;
+	struct ve_struct *ve;
+	bool ve_overridden;
+};
+
 #define OVL_METACOPY_MAX_SIZE (sizeof(struct ovl_metacopy))
 #define OVL_METACOPY_MIN_SIZE (OVL_METACOPY_MAX_SIZE - FS_VERITY_MAX_DIGEST_SIZE)
 #define OVL_METACOPY_INIT { 0, OVL_METACOPY_MIN_SIZE }
@@ -420,7 +427,8 @@ void ovl_end_write(struct dentry *dentry);
 int ovl_want_write(struct dentry *dentry);
 void ovl_drop_write(struct dentry *dentry);
 struct dentry *ovl_workdir(struct dentry *dentry);
-const struct cred *ovl_override_creds(struct super_block *sb);
+int ovl_override_creds(struct super_block *sb, struct ovl_creds_desc *desc);
+void ovl_revert_creds(struct ovl_creds_desc *old_creds);
 
 static inline const struct cred *ovl_creds(struct super_block *sb)
 {
diff --git a/fs/overlayfs/readdir.c b/fs/overlayfs/readdir.c
index 0ca8af060b0c..478b0a32ff64 100644
--- a/fs/overlayfs/readdir.c
+++ b/fs/overlayfs/readdir.c
@@ -273,9 +273,11 @@ static int ovl_check_whiteouts(const struct path *path, struct ovl_readdir_data
 	int err;
 	struct ovl_cache_entry *p;
 	struct dentry *dentry, *dir = path->dentry;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 
-	old_cred = ovl_override_creds(rdd->dentry->d_sb);
+	err = ovl_override_creds(rdd->dentry->d_sb, &old_cred);
+	if (err)
+		return err;
 
 	err = down_write_killable(&dir->d_inode->i_rwsem);
 	if (!err) {
@@ -290,7 +292,7 @@ static int ovl_check_whiteouts(const struct path *path, struct ovl_readdir_data
 		}
 		inode_unlock(dir->d_inode);
 	}
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 
 	return err;
 }
@@ -753,10 +755,13 @@ static int ovl_iterate(struct file *file, struct dir_context *ctx)
 	struct dentry *dentry = file->f_path.dentry;
 	struct ovl_fs *ofs = OVL_FS(dentry->d_sb);
 	struct ovl_cache_entry *p;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	int err;
 
-	old_cred = ovl_override_creds(dentry->d_sb);
+	err = ovl_override_creds(dentry->d_sb, &old_cred);
+	if (err)
+		return err;
+
 	if (!ctx->pos)
 		ovl_dir_reset(file);
 
@@ -808,7 +813,7 @@ static int ovl_iterate(struct file *file, struct dir_context *ctx)
 	}
 	err = 0;
 out:
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 	return err;
 }
 
@@ -856,11 +861,15 @@ static struct file *ovl_dir_open_realfile(const struct file *file,
 					  const struct path *realpath)
 {
 	struct file *res;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
+	int err;
+
+	err = ovl_override_creds(file_inode(file)->i_sb, &old_cred);
+	if (err)
+		return ERR_PTR(err);
 
-	old_cred = ovl_override_creds(file_inode(file)->i_sb);
 	res = ovl_path_open(realpath, O_RDONLY | (file->f_flags & O_LARGEFILE));
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 
 	return res;
 }
@@ -983,11 +992,14 @@ int ovl_check_empty_dir(struct dentry *dentry, struct list_head *list)
 	int err;
 	struct ovl_cache_entry *p, *n;
 	struct rb_root root = RB_ROOT;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
+
+	err = ovl_override_creds(dentry->d_sb, &old_cred);
+	if (err)
+		return err;
 
-	old_cred = ovl_override_creds(dentry->d_sb);
 	err = ovl_dir_read_merged(dentry, list, &root);
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 	if (err)
 		return err;
 
diff --git a/fs/overlayfs/util.c b/fs/overlayfs/util.c
index 817563215001..273b2e2e17eb 100644
--- a/fs/overlayfs/util.c
+++ b/fs/overlayfs/util.c
@@ -63,11 +63,38 @@ struct dentry *ovl_workdir(struct dentry *dentry)
 	return ofs->workdir;
 }
 
-const struct cred *ovl_override_creds(struct super_block *sb)
+/* On error no revert is needed. */
+int ovl_override_creds(struct super_block *sb, struct ovl_creds_desc *desc)
 {
 	struct ovl_fs *ofs = OVL_FS(sb);
 
-	return override_creds(ofs->creator_cred);
+	/* Cleanup after previous users */
+	*desc = (struct ovl_creds_desc) {};
+
+	if (ve_is_super(ofs->creator_ve) && !ve_is_super(get_exec_env())) {
+		/* Host overlayfs used from a CT: host creds work there. */
+		desc->ve = NULL;
+		desc->ve_overridden = false;
+	} else {
+		desc->ve = override_ve(ofs->creator_ve);
+		if (IS_ERR(desc->ve))
+			return PTR_ERR(desc->ve);
+		desc->ve_overridden = true;
+	}
+
+	desc->cred = override_creds(ofs->creator_cred);
+	return 0;
+}
+
+/* Accepts desc from a failed ovl_override_creds and does nothing. */
+void ovl_revert_creds(struct ovl_creds_desc *old_creds)
+{
+	if (IS_ERR(old_creds->ve))
+		return;
+
+	revert_creds(old_creds->cred);
+	if (old_creds->ve_overridden)
+		revert_ve(old_creds->ve);
 }
 
 /*
@@ -1162,7 +1189,7 @@ static void ovl_cleanup_index(struct dentry *dentry)
 int ovl_nlink_start(struct dentry *dentry)
 {
 	struct inode *inode = d_inode(dentry);
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	int err;
 
 	if (WARN_ON(!inode))
@@ -1199,7 +1226,10 @@ int ovl_nlink_start(struct dentry *dentry)
 	if (d_is_dir(dentry) || !ovl_test_flag(OVL_INDEX, inode))
 		return 0;
 
-	old_cred = ovl_override_creds(dentry->d_sb);
+	err = ovl_override_creds(dentry->d_sb, &old_cred);
+	if (err)
+		goto out_drop_write;
+
 	/*
 	 * The overlay inode nlink should be incremented/decremented IFF the
 	 * upper operation succeeds, along with nlink change of upper inode.
@@ -1207,7 +1237,7 @@ int ovl_nlink_start(struct dentry *dentry)
 	 * value relative to the upper inode nlink in an upper inode xattr.
 	 */
 	err = ovl_set_nlink_upper(dentry);
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 	if (err)
 		goto out_drop_write;
 
@@ -1228,11 +1258,15 @@ void ovl_nlink_end(struct dentry *dentry)
 	ovl_drop_write(dentry);
 
 	if (ovl_test_flag(OVL_INDEX, inode) && inode->i_nlink == 0) {
-		const struct cred *old_cred;
+		struct ovl_creds_desc old_cred;
+		int err;
 
-		old_cred = ovl_override_creds(dentry->d_sb);
-		ovl_cleanup_index(dentry);
-		revert_creds(old_cred);
+		err = ovl_override_creds(dentry->d_sb, &old_cred);
+		if (err)
+			pr_err("%s: skipping ovl_cleanup_index.\n", __func__);
+		else
+			ovl_cleanup_index(dentry);
+		ovl_revert_creds(&old_cred);
 	}
 
 	ovl_inode_unlock(inode);
diff --git a/fs/overlayfs/xattrs.c b/fs/overlayfs/xattrs.c
index 0ed2b049056d..7528ec76868d 100644
--- a/fs/overlayfs/xattrs.c
+++ b/fs/overlayfs/xattrs.c
@@ -41,13 +41,16 @@ static int ovl_xattr_set(struct dentry *dentry, struct inode *inode, const char
 	struct dentry *upperdentry = ovl_i_dentry_upper(inode);
 	struct dentry *realdentry = upperdentry ?: ovl_dentry_lower(dentry);
 	struct path realpath;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 
 	if (!value && !upperdentry) {
 		ovl_path_lower(dentry, &realpath);
-		old_cred = ovl_override_creds(dentry->d_sb);
+		err = ovl_override_creds(dentry->d_sb, &old_cred);
+		if (err)
+			return err;
+
 		err = vfs_getxattr(mnt_idmap(realpath.mnt), realdentry, name, NULL, 0);
-		revert_creds(old_cred);
+		ovl_revert_creds(&old_cred);
 		if (err < 0)
 			goto out;
 	}
@@ -64,7 +67,12 @@ static int ovl_xattr_set(struct dentry *dentry, struct inode *inode, const char
 	if (err)
 		goto out;
 
-	old_cred = ovl_override_creds(dentry->d_sb);
+	err = ovl_override_creds(dentry->d_sb, &old_cred);
+	if (err) {
+		ovl_drop_write(dentry);
+		return err;
+	}
+
 	if (value) {
 		err = ovl_do_setxattr(ofs, realdentry, name, value, size,
 				      flags);
@@ -72,7 +80,7 @@ static int ovl_xattr_set(struct dentry *dentry, struct inode *inode, const char
 		WARN_ON(flags != XATTR_REPLACE);
 		err = ovl_do_removexattr(ofs, realdentry, name);
 	}
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 	ovl_drop_write(dentry);
 
 	/* copy c/mtime */
@@ -85,13 +93,14 @@ static int ovl_xattr_get(struct dentry *dentry, struct inode *inode, const char
 			 void *value, size_t size)
 {
 	ssize_t res;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	struct path realpath;
 
 	ovl_i_path_real(inode, &realpath);
-	old_cred = ovl_override_creds(dentry->d_sb);
-	res = vfs_getxattr(mnt_idmap(realpath.mnt), realpath.dentry, name, value, size);
-	revert_creds(old_cred);
+	res = ovl_override_creds(dentry->d_sb, &old_cred);
+	if (!res)
+		res = vfs_getxattr(mnt_idmap(realpath.mnt), realpath.dentry, name, value, size);
+	ovl_revert_creds(&old_cred);
 	return res;
 }
 
@@ -116,12 +125,14 @@ ssize_t ovl_listxattr(struct dentry *dentry, char *list, size_t size)
 	ssize_t res;
 	size_t len;
 	char *s;
-	const struct cred *old_cred;
+	struct ovl_creds_desc old_cred;
 	size_t prefix_len, name_len;
 
-	old_cred = ovl_override_creds(dentry->d_sb);
+	res = ovl_override_creds(dentry->d_sb, &old_cred);
+	if (res)
+		return res;
 	res = vfs_listxattr(realdentry, list, size);
-	revert_creds(old_cred);
+	ovl_revert_creds(&old_cred);
 	if (res <= 0 || size == 0)
 		return res;
 

-- 
2.52.0

_______________________________________________
Devel mailing list
Devel@openvz.org
https://lists.openvz.org/mailman/listinfo/devel

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [Devel] [PATCH hci-8.1 v6 1/5] ve: Remove STOPPED state
  2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 1/5] ve: Remove STOPPED state Vladimir Riabchun
@ 2026-11-16 16:10   ` no-reply
  2026-11-16 16:25     ` Vladimir Riabchun
  0 siblings, 1 reply; 13+ messages in thread
From: no-reply @ 2026-11-16 16:10 UTC (permalink / raw)
  To: Vladimir Riabchun, vasileios.almpanis, ptikhomirov, Konstantin Khorenko
  Cc: devel

Thank you for your contribution! Sashiko AI review found 1 potential
issue to consider:
- [Low] ve: commit message omits removal of DEAD transition in
  ve_destroy()
--

commit 6a34f90f8fbb9ff68c8aa5b754aa9b1d014675fb
Author: Vladimir Riabchun <vladimir.riabchun@virtuozzo.com>

ve: Remove STOPPED state

This commit drops the separate VE_STATE_STOPPED state and moves the VE directly to VE_STATE_DEAD when it is stopped, both on the ve_start_container() error path and in ve_exit_ns().

> diff --git a/kernel/ve/ve.c b/kernel/ve/ve.c
> index 6cdfaaa4e703..b3988a8d590f 100644
> --- a/kernel/ve/ve.c
> +++ b/kernel/ve/ve.c

[ ... ]

> @@ -939,7 +939,6 @@ static void ve_destroy(struct cgroup_subsys_state *css)
>  	ve_free_vdso(ve);
>  	mntput(ve->devtmpfs_mnt);
>  	free_percpu(ve->sched_lat_ve.cur);
> -	ve_set_state(ve, VE_STATE_DEAD);
>  	kmem_cache_free(ve_cachep, ve);
>  }

[Severity: Low]
The commit message does not mention that this DEAD transition in ve_destroy() is removed. Could it say so, and why it is safe?

The removal looks necessary. DEAD is now set in ve_exit_ns() and in the ve_start_container() error path, so keeping this call would hit the "already DEAD" warning in ve_set_state(). The old call was also a store to an object freed on the next line. As a result, a VE destroyed while still in STARTING never reaches DEAD, which the message could also state.

-- 
Sashiko AI review · https://sashiko.vzint.dev/#/patchset/20261116-ovl-use-orig-ve-v6-0-7cfb41c8c973@virtuozzo.com?part=1

_______________________________________________
Devel mailing list
Devel@openvz.org
https://lists.openvz.org/mailman/listinfo/devel

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [Devel] [PATCH hci-8.1 v6 2/5] ve: Introduce override_ve/revert_ve helpers
  2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 2/5] ve: Introduce override_ve/revert_ve helpers Vladimir Riabchun
@ 2026-11-16 16:12   ` no-reply
  0 siblings, 0 replies; 13+ messages in thread
From: no-reply @ 2026-11-16 16:12 UTC (permalink / raw)
  To: Vladimir Riabchun, vasileios.almpanis, ptikhomirov, Konstantin Khorenko
  Cc: devel

Thank you for your contribution! Sashiko AI review found 2 potential
issues to consider:
- [Low] ve: commit message omits ve_capable() change and override_ve()
  rules
- [Low] ve: override_ve() error path reads ve_name(new_ve) with no pin
  against ve_offline()
--

commit 1ff80fe27a47be85961240ab0b7e9287052ee68d
Author: Vladimir Riabchun <vladimir.riabchun@virtuozzo.com>

ve: Introduce override_ve/revert_ve helpers

This commit adds override_ve() and revert_ve() helpers, which mirror override_creds() and revert_creds(). It also adds a task->overrided_ve field. ve_capable() and ve_capable_noaudit() now use that field, and dup_task_struct() clears it on fork.

> diff --git a/include/linux/ve.h b/include/linux/ve.h
> --- a/include/linux/ve.h
> +++ b/include/linux/ve.h
> @@ -295,6 +295,30 @@ static inline struct vdso_image *ve_get_vdso_64(struct ve_struct *ve)
>  	return ve->vdso_64;
>  }
>  
> +static inline struct ve_struct *override_ve(struct ve_struct *new_ve)
> +{
> +	struct ve_struct *old_ve = current->overrided_ve;
> +
> +	/* CT tasks are not supposed to switch VE context, this is a breach */
> +	if (new_ve != get_exec_env() && !ve_is_super(get_exec_env())) {
> +		pr_err_ratelimited("Invalid VE override: %s -> %s\n",
> +			ve_name(get_exec_env()),
> +			VE_IS_RUNNING(new_ve) ? ve_name(new_ve) : "(dead VE)");

[Severity: Low]
As reported by Konstantin Khorenko on v5 [1], can this read freed memory?
ve_name(new_ve) is read with no lock, RCU section or reference. The
VE_IS_RUNNING() check is racy, and ve_offline() frees ve->ve_name with a
plain kfree() and no RCU. The get_ve() reference held by the caller keeps
the ve_struct alive, but not the name string.

The message is ratelimited and the window is narrow. Would it be better
to stop printing ve_name(new_ve) here, as requested on v5?

> +		return ERR_PTR(-EPERM);
> +	}
> +
> +	current->overrided_ve = new_ve;
> +	return old_ve;
> +}

[ ... ]

> diff --git a/kernel/capability.c b/kernel/capability.c
> --- a/kernel/capability.c
> +++ b/kernel/capability.c
> @@ -389,10 +389,11 @@ EXPORT_SYMBOL(ns_capable);
>  bool ve_capable(int cap)
>  {
>  	struct cred *cred;
> +	struct ve_struct *ve = current->overrided_ve ?: get_exec_env();
>  	bool ret;

[Severity: Low]
As reported by Konstantin Khorenko on v5 [1], the commit message never
mentions that ve_capable() and ve_capable_noaudit() now check capabilities
against current->overrided_ve instead of get_exec_env(). It also does not
say which VE switches override_ve() allows, and that it refuses a switch
by a CT task to a different VE with -EPERM.

Could the message describe these behaviour changes?

[1] https://lore.kernel.org/r/179448942112.4060562.12042466360723083159.b4-review@b4

-- 
Sashiko AI review · https://sashiko.vzint.dev/#/patchset/20261116-ovl-use-orig-ve-v6-0-7cfb41c8c973@virtuozzo.com?part=2

_______________________________________________
Devel mailing list
Devel@openvz.org
https://lists.openvz.org/mailman/listinfo/devel

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [Devel] [PATCH hci-8.1 v6 4/5] ovl: Keep creator VE in overlayfs superblock
  2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 4/5] ovl: Keep creator VE in overlayfs superblock Vladimir Riabchun
@ 2026-11-16 16:13   ` no-reply
  0 siblings, 0 replies; 13+ messages in thread
From: no-reply @ 2026-11-16 16:13 UTC (permalink / raw)
  To: Vladimir Riabchun, vasileios.almpanis, ptikhomirov, Konstantin Khorenko
  Cc: devel

> Overlayfs superblock stores creator creds in it, but these creds
> should go together with the VE they were taken from.
> 
> Store the VE in the superblock. We can't use mount->owner_ve,
> because several mounts can reference one superblock.
> 
> Feature: ve/overlayfs
> https://virtuozzo.atlassian.net/browse/VSTOR-142405
> Signed-off-by: Vladimir Riabchun <vladimir.riabchun@virtuozzo.com>

Sashiko has reviewed this patch and found no issues.

-- 
Sashiko AI review · https://sashiko.vzint.dev/#/patchset/20261116-ovl-use-orig-ve-v6-0-7cfb41c8c973@virtuozzo.com?part=4


_______________________________________________
Devel mailing list
Devel@openvz.org
https://lists.openvz.org/mailman/listinfo/devel

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [Devel] [PATCH hci-8.1 v6 3/5] ve: Defer ve creds drop until ve is completely destroyed
  2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 3/5] ve: Defer ve creds drop until ve is completely destroyed Vladimir Riabchun
@ 2026-11-16 16:14   ` no-reply
  0 siblings, 0 replies; 13+ messages in thread
From: no-reply @ 2026-11-16 16:14 UTC (permalink / raw)
  To: Vladimir Riabchun, vasileios.almpanis, ptikhomirov, Konstantin Khorenko
  Cc: devel

Thank you for your contribution! Sashiko AI review found 1 potential
issue to consider:
- [Low] ve: stale 've isn't running' comment in netlink_ve_capable()
  after init_cred change
--

commit 0ce43e99309aa75e8847e7c9d6dca465801086bf
Author: Vladimir Riabchun <vladimir.riabchun@virtuozzo.com>

ve: Defer ve creds drop until ve is completely destroyed

The patch moves put_cred() and the init_cred = NULL store out of
ve_drop_context() and into a new ve_drop_cred(), which ve_destroy() now
calls. This keeps ve->init_cred alive while a dead VE still has users, such
as an overlayfs superblock.

> diff --git a/kernel/capability.c b/kernel/capability.c
> index 6de4dcba2283..68c3d5288ba7 100644
> --- a/kernel/capability.c
> +++ b/kernel/capability.c
> @@ -395,7 +395,7 @@ bool ve_capable(int cap)
>  	rcu_read_lock();
>  	cred = ve->init_cred;
>  
> -	if (cred == NULL) /* ve isn't running */
> +	if (cred == NULL) /* ve isn't started */
>  		cred = ve0.init_cred;

[Severity: Low]
This isn't a bug, but the same NULL-init_cred fallback in
netlink_ve_capable() in net/netlink/af_netlink.c still says "ve isn't
running". Since init_cred now stays set after the VE stops or dies, and NULL
means only that the VE was never started, should that comment be reworded
to match?

The commit message also doesn't say that capability checks against a dead VE
no longer fall back to ve0.init_cred. That looks like the intended effect, so
could it be mentioned?

-- 
Sashiko AI review · https://sashiko.vzint.dev/#/patchset/20261116-ovl-use-orig-ve-v6-0-7cfb41c8c973@virtuozzo.com?part=3

_______________________________________________
Devel mailing list
Devel@openvz.org
https://lists.openvz.org/mailman/listinfo/devel

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [Devel] [PATCH hci-8.1 v6 5/5] ve/overlayfs: Override VE together with creds
  2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 5/5] ve/overlayfs: Override VE together with creds Vladimir Riabchun
@ 2026-11-16 16:19   ` no-reply
  2026-11-16 16:52     ` Vladimir Riabchun
  0 siblings, 1 reply; 13+ messages in thread
From: no-reply @ 2026-11-16 16:19 UTC (permalink / raw)
  To: Vladimir Riabchun, vasileios.almpanis, ptikhomirov, Konstantin Khorenko
  Cc: devel

Thank you for your contribution! Sashiko AI review found 4 potential
issues to consider:
- [Medium] ovl: ovl_copyfile() leaks real_in/real_out refs when override
  fails
- [Medium] ve/overlayfs: backing-file I/O switches creds without the
  matching VE override
- [Low] ovl: ovl_lower_positive() returns false when
  ovl_override_creds() fails
- [Low] ovl: ovl_nlink_end() silently skips index cleanup on override
  failure
--

commit 9665af403c75ade64577a4c77c18a22fb272209d
Author: Vladimir Riabchun <vladimir.riabchun@virtuozzo.com>

ve/overlayfs: Override VE together with creds

ovl_override_creds() now returns an error and fills a struct ovl_creds_desc holding both the old cred and the old VE. ovl_revert_creds() restores both. Every caller is converted, so overlayfs switches the VE together with the creator creds. Access to an overlayfs created in a different CT is rejected with -EPERM.

> diff --git a/fs/overlayfs/file.c b/fs/overlayfs/file.c
> index 2531433f47bf..464699e51695 100644
> --- a/fs/overlayfs/file.c
> +++ b/fs/overlayfs/file.c

[ ... ]

> @@ -520,14 +527,19 @@ static loff_t ovl_copyfile(struct file *file_in, loff_t pos_in,
>  	 * verify read access with the source mounter creds.
>  	 */
>  	if (unlikely(inode_in->i_sb != inode_out->i_sb)) {
> -		old_cred = ovl_override_creds(inode_in->i_sb);
> +		ret = ovl_override_creds(inode_in->i_sb, &old_cred);
> +		if (ret)
> +			goto out_unlock;
>  		ret = rw_verify_area(READ, fd_file(real_in), &pos_in, len);
> -		revert_creds(old_cred);
> +		ovl_revert_creds(&old_cred);
>  		if (unlikely(ret))
>  			goto out_unlock;
>  	}

[Severity: Medium]
Does this new error exit leak the real_in and real_out file references?

Both were taken by ovl_real_fdget() earlier in ovl_copyfile(), but out_unlock only calls inode_unlock(inode_out), so fdput() is never reached. Before this patch the override could not fail, so the leak is a regression. A task in one CT doing copy_file_range() between its own overlayfs and one created by another CT gets -EPERM from override_VE() and leaks both files on every attempt. The second ovl_override_creds() failure below already uses skip_switch, which reaches the fdput() calls.

[ ... ]

> diff --git a/fs/overlayfs/namei.c b/fs/overlayfs/namei.c
> index efa2a072aa69..edc11a7b27ee 100644
> --- a/fs/overlayfs/namei.c
> +++ b/fs/overlayfs/namei.c

[ ... ]

> @@ -1426,7 +1436,10 @@ bool ovl_lower_positive(struct dentry *dentry)
>  	if (!ovl_dentry_upper(dentry))
>  		return true;
>  
> -	old_cred = ovl_override_creds(dentry->d_sb);
> +	/* Ugh, hard - we can't pass an error from here. */
> +	if (ovl_override_creds(dentry->d_sb, &old_cred))
> +		return false;
> +

[Severity: Low]
Is false the safe answer when the override fails?

Returning false means "no lower entry exists", so ovl_do_remove() would choose ovl_remove_upper() instead of ovl_remove_and_whiteout(), and ovl_rename() would skip RENAME_WHITEOUT. A lower file could then reappear after unlink. In practice the later override calls in the same operation fail the same way, but correctness then depends on call ordering.

The lookup-error branch of this function already assumes "something is there" and sets positive = true. Would returning true here be consistent with that?

[ ... ]

> diff --git a/fs/overlayfs/util.c b/fs/overlayfs/util.c
> index 817563215001..273b2e2e17eb 100644
> --- a/fs/overlayfs/util.c
> +++ b/fs/overlayfs/util.c
> @@ -63,11 +63,38 @@ struct dentry *ovl_workdir(struct dentry *dentry)
>  	return ofs->workdir;
>  }
>  
> -const struct cred *ovl_override_creds(struct super_block *sb)
> +/* On error no revert is needed. */
> +int ovl_override_creds(struct super_block *sb, struct ovl_creds_desc *desc)
>  {
>  	struct ovl_fs *ofs = OVL_FS(sb);
>  
> -	return override_creds(ofs->creator_cred);
> +	/* Cleanup after previous users */
> +	*desc = (struct ovl_creds_desc) {};
> +
> +	if (ve_is_super(ofs->creator_ve) && !ve_is_super(get_exec_env())) {
> +		/* Host overlayfs used from a CT: host creds work there. */
> +		desc->ve = NULL;
> +		desc->ve_overridden = false;
> +	} else {
> +		desc->ve = override_ve(ofs->creator_ve);
> +		if (IS_ERR(desc->ve))
> +			return PTR_ERR(desc->ve);
> +		desc->ve_overridden = true;
> +	}
> +
> +	desc->cred = override_creds(ofs->creator_cred);
> +	return 0;
> +}

[Severity: Medium]
As reported by Pavel Tikhomirov on v2 [1], the conversion is incomplete. Are the users of ovl_creds() still left with the cred/VE mismatch this patch describes?

ovl_read_iter(), ovl_write_iter(), ovl_splice_read(), ovl_splice_write() and ovl_mmap() set backing_file_ctx.cred = ovl_creds(sb). The helpers in fs/backing-file.c then call only override_creds(ctx->cred), with no override_ve(). ve_capable() uses current->overrided_ve ?: get_exec_env(), so on these paths the creator creds are still checked against the caller's VE. A host task reading a file on a CT overlayfs is one such case. Cross-CT access is also not rejected there with -EPERM, unlike the paths converted here. The commit message suggests all users of the old cred are covered, which is not the case.

> +
> +/* Accepts desc from a failed ovl_override_creds and does nothing. */
> +void ovl_revert_creds(struct ovl_creds_desc *old_creds)
> +{
> +	if (IS_ERR(old_creds->ve))
> +		return;
> +
> +	revert_creds(old_creds->cred);
> +	if (old_creds->ve_overridden)
> +		revert_ve(old_creds->ve);
>  }

[ ... ]

> @@ -1228,11 +1258,15 @@ void ovl_nlink_end(struct dentry *dentry)
>  	ovl_drop_write(dentry);
>  
>  	if (ovl_test_flag(OVL_INDEX, inode) && inode->i_nlink == 0) {
> -		const struct cred *old_cred;
> +		struct ovl_creds_desc old_cred;
> +		int err;
>  
> -		old_cred = ovl_override_creds(dentry->d_sb);
> -		ovl_cleanup_index(dentry);
> -		revert_creds(old_cred);
> +		err = ovl_override_creds(dentry->d_sb, &old_cred);
> +		if (err)
> +			pr_err("%s: skipping ovl_cleanup_index.\n", __func__);
> +		else
> +			ovl_cleanup_index(dentry);
> +		ovl_revert_creds(&old_cred);
>  	}

[Severity: Low]
Is it acceptable to silently skip ovl_cleanup_index() here?

ovl_nlink_end() is void and runs after the upper unlink or rename has already succeeded. If the override fails, an orphan index entry is left behind with i_nlink at 0, and only a pr_err() records it. The old code could not fail at this point. ovl_nlink_start() would normally have failed first for the same task, sb and VE, so this looks mostly unreachable. But it is an error path that now leaves stale state.

[1] https://lore.kernel.org/r/179318868950.3559520.6473416824609223980.b4-review@b4

-- 
Sashiko AI review · https://sashiko.vzint.dev/#/patchset/20261116-ovl-use-orig-ve-v6-0-7cfb41c8c973@virtuozzo.com?part=5

_______________________________________________
Devel mailing list
Devel@openvz.org
https://lists.openvz.org/mailman/listinfo/devel

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [Devel] [PATCH hci-8.1 v6 1/5] ve: Remove STOPPED state
  2026-11-16 16:10   ` no-reply
@ 2026-11-16 16:25     ` Vladimir Riabchun
  0 siblings, 0 replies; 13+ messages in thread
From: Vladimir Riabchun @ 2026-11-16 16:25 UTC (permalink / raw)
  To: vasileios.almpanis, ptikhomirov, Konstantin Khorenko; +Cc: devel



On 11/16/26 17:10, no-reply@sashiko.virtuozzo.com wrote:
> Thank you for your contribution! Sashiko AI review found 1 potential
> issue to consider:
> - [Low] ve: commit message omits removal of DEAD transition in
>    ve_destroy()
> --
> 
> commit 6a34f90f8fbb9ff68c8aa5b754aa9b1d014675fb
> Author: Vladimir Riabchun <vladimir.riabchun@virtuozzo.com>
> 
> ve: Remove STOPPED state
> 
> This commit drops the separate VE_STATE_STOPPED state and moves the VE directly to VE_STATE_DEAD when it is stopped, both on the ve_start_container() error path and in ve_exit_ns().
> 
>> diff --git a/kernel/ve/ve.c b/kernel/ve/ve.c
>> index 6cdfaaa4e703..b3988a8d590f 100644
>> --- a/kernel/ve/ve.c
>> +++ b/kernel/ve/ve.c
> 
> [ ... ]
> 
>> @@ -939,7 +939,6 @@ static void ve_destroy(struct cgroup_subsys_state *css)
>>   	ve_free_vdso(ve);
>>   	mntput(ve->devtmpfs_mnt);
>>   	free_percpu(ve->sched_lat_ve.cur);
>> -	ve_set_state(ve, VE_STATE_DEAD);
>>   	kmem_cache_free(ve_cachep, ve);
>>   }
> 
> [Severity: Low]
> The commit message does not mention that this DEAD transition in ve_destroy() is removed. Could it say so, and why it is safe?

No comments about it because this is clear from the code and
commit description.

> 
> The removal looks necessary. DEAD is now set in ve_exit_ns() and in the ve_start_container() error path, so keeping this call would hit the "already DEAD" warning in ve_set_state(). The old call was also a store to an object freed on the next line. As a result, a VE destroyed while still in STARTING never reaches DEAD, which the message could also state.
> 

-- 
Best regards, Riabchun Vladimir
Linux Kernel Developer, Virtuozzo

_______________________________________________
Devel mailing list
Devel@openvz.org
https://lists.openvz.org/mailman/listinfo/devel

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [Devel] [PATCH hci-8.1 v6 5/5] ve/overlayfs: Override VE together with creds
  2026-11-16 16:19   ` no-reply
@ 2026-11-16 16:52     ` Vladimir Riabchun
  0 siblings, 0 replies; 13+ messages in thread
From: Vladimir Riabchun @ 2026-11-16 16:52 UTC (permalink / raw)
  To: no-reply, vasileios.almpanis, ptikhomirov, Konstantin Khorenko; +Cc: devel



On 11/16/26 17:19, no-reply@sashiko.virtuozzo.com wrote:
> Thank you for your contribution! Sashiko AI review found 4 potential
> issues to consider:
> - [Medium] ovl: ovl_copyfile() leaks real_in/real_out refs when override
>    fails
> - [Medium] ve/overlayfs: backing-file I/O switches creds without the
>    matching VE override
> - [Low] ovl: ovl_lower_positive() returns false when
>    ovl_override_creds() fails
> - [Low] ovl: ovl_nlink_end() silently skips index cleanup on override
>    failure
> --
> 
> commit 9665af403c75ade64577a4c77c18a22fb272209d
> Author: Vladimir Riabchun <vladimir.riabchun@virtuozzo.com>
> 
> ve/overlayfs: Override VE together with creds
> 
> ovl_override_creds() now returns an error and fills a struct ovl_creds_desc holding both the old cred and the old VE. ovl_revert_creds() restores both. Every caller is converted, so overlayfs switches the VE together with the creator creds. Access to an overlayfs created in a different CT is rejected with -EPERM.
> 
>> diff --git a/fs/overlayfs/file.c b/fs/overlayfs/file.c
>> index 2531433f47bf..464699e51695 100644
>> --- a/fs/overlayfs/file.c
>> +++ b/fs/overlayfs/file.c
> 
> [ ... ]
> 
>> @@ -520,14 +527,19 @@ static loff_t ovl_copyfile(struct file *file_in, loff_t pos_in,
>>   	 * verify read access with the source mounter creds.
>>   	 */
>>   	if (unlikely(inode_in->i_sb != inode_out->i_sb)) {
>> -		old_cred = ovl_override_creds(inode_in->i_sb);
>> +		ret = ovl_override_creds(inode_in->i_sb, &old_cred);
>> +		if (ret)
>> +			goto out_unlock;
>>   		ret = rw_verify_area(READ, fd_file(real_in), &pos_in, len);
>> -		revert_creds(old_cred);
>> +		ovl_revert_creds(&old_cred);
>>   		if (unlikely(ret))
>>   			goto out_unlock;
>>   	}
> 
> [Severity: Medium]
> Does this new error exit leak the real_in and real_out file references?
> 
> Both were taken by ovl_real_fdget() earlier in ovl_copyfile(), but out_unlock only calls inode_unlock(inode_out), so fdput() is never reached. Before this patch the override could not fail, so the leak is a regression. A task in one CT doing copy_file_range() between its own overlayfs and one created by another CT gets -EPERM from override_VE() and leaks both files on every attempt. The second ovl_override_creds() failure below already uses skip_switch, which reaches the fdput() calls.
> 

There is a leak and this is a rebase bug.
https://github.com/torvalds/linux/commit/a1e0eb8f55cfe09bb31a202a388babc411292656

We don't have with_ovl_creds helper.
I'll send a fix for this in v7.

> [ ... ]
> 
>> diff --git a/fs/overlayfs/namei.c b/fs/overlayfs/namei.c
>> index efa2a072aa69..edc11a7b27ee 100644
>> --- a/fs/overlayfs/namei.c
>> +++ b/fs/overlayfs/namei.c
> 
> [ ... ]
> 
>> @@ -1426,7 +1436,10 @@ bool ovl_lower_positive(struct dentry *dentry)
>>   	if (!ovl_dentry_upper(dentry))
>>   		return true;
>>   
>> -	old_cred = ovl_override_creds(dentry->d_sb);
>> +	/* Ugh, hard - we can't pass an error from here. */
>> +	if (ovl_override_creds(dentry->d_sb, &old_cred))
>> +		return false;
>> +
> 
> [Severity: Low]
> Is false the safe answer when the override fails?
> 
> Returning false means "no lower entry exists", so ovl_do_remove() would choose ovl_remove_upper() instead of ovl_remove_and_whiteout(), and ovl_rename() would skip RENAME_WHITEOUT. A lower file could then reappear after unlink. In practice the later override calls in the same operation fail the same way, but correctness then depends on call ordering.
> 
> The lookup-error branch of this function already assumes "something is there" and sets positive = true. Would returning true here be consistent with that?
> 
> [ ... ]
> 
>> diff --git a/fs/overlayfs/util.c b/fs/overlayfs/util.c
>> index 817563215001..273b2e2e17eb 100644
>> --- a/fs/overlayfs/util.c
>> +++ b/fs/overlayfs/util.c
>> @@ -63,11 +63,38 @@ struct dentry *ovl_workdir(struct dentry *dentry)
>>   	return ofs->workdir;
>>   }
>>   
>> -const struct cred *ovl_override_creds(struct super_block *sb)
>> +/* On error no revert is needed. */
>> +int ovl_override_creds(struct super_block *sb, struct ovl_creds_desc *desc)
>>   {
>>   	struct ovl_fs *ofs = OVL_FS(sb);
>>   
>> -	return override_creds(ofs->creator_cred);
>> +	/* Cleanup after previous users */
>> +	*desc = (struct ovl_creds_desc) {};
>> +
>> +	if (ve_is_super(ofs->creator_ve) && !ve_is_super(get_exec_env())) {
>> +		/* Host overlayfs used from a CT: host creds work there. */
>> +		desc->ve = NULL;
>> +		desc->ve_overridden = false;
>> +	} else {
>> +		desc->ve = override_ve(ofs->creator_ve);
>> +		if (IS_ERR(desc->ve))
>> +			return PTR_ERR(desc->ve);
>> +		desc->ve_overridden = true;
>> +	}
>> +
>> +	desc->cred = override_creds(ofs->creator_cred);
>> +	return 0;
>> +}
> 
> [Severity: Medium]
> As reported by Pavel Tikhomirov on v2 [1], the conversion is incomplete. Are the users of ovl_creds() still left with the cred/VE mismatch this patch describes?

Yes. https://virtuozzo.atlassian.net/browse/VSTOR-155347

> 
> ovl_read_iter(), ovl_write_iter(), ovl_splice_read(), ovl_splice_write() and ovl_mmap() set backing_file_ctx.cred = ovl_creds(sb). The helpers in fs/backing-file.c then call only override_creds(ctx->cred), with no override_ve(). ve_capable() uses current->overrided_ve ?: get_exec_env(), so on these paths the creator creds are still checked against the caller's VE. A host task reading a file on a CT overlayfs is one such case. Cross-CT access is also not rejected there with -EPERM, unlike the paths converted here. The commit message suggests all users of the old cred are covered, which is not the case.
> 
>> +
>> +/* Accepts desc from a failed ovl_override_creds and does nothing. */
>> +void ovl_revert_creds(struct ovl_creds_desc *old_creds)
>> +{
>> +	if (IS_ERR(old_creds->ve))
>> +		return;
>> +
>> +	revert_creds(old_creds->cred);
>> +	if (old_creds->ve_overridden)
>> +		revert_ve(old_creds->ve);
>>   }
> 
> [ ... ]
> 
>> @@ -1228,11 +1258,15 @@ void ovl_nlink_end(struct dentry *dentry)
>>   	ovl_drop_write(dentry);
>>   
>>   	if (ovl_test_flag(OVL_INDEX, inode) && inode->i_nlink == 0) {
>> -		const struct cred *old_cred;
>> +		struct ovl_creds_desc old_cred;
>> +		int err;
>>   
>> -		old_cred = ovl_override_creds(dentry->d_sb);
>> -		ovl_cleanup_index(dentry);
>> -		revert_creds(old_cred);
>> +		err = ovl_override_creds(dentry->d_sb, &old_cred);
>> +		if (err)
>> +			pr_err("%s: skipping ovl_cleanup_index.\n", __func__);
>> +		else
>> +			ovl_cleanup_index(dentry);
>> +		ovl_revert_creds(&old_cred);
>>   	}
> 
> [Severity: Low]
> Is it acceptable to silently skip ovl_cleanup_index() here?
> 
> ovl_nlink_end() is void and runs after the upper unlink or rename has already succeeded. If the override fails, an orphan index entry is left behind with i_nlink at 0, and only a pr_err() records it. The old code could not fail at this point. ovl_nlink_start() would normally have failed first for the same task, sb and VE, so this looks mostly unreachable. But it is an error path that now leaves stale state.

Nothing we can do here. Everything is bad already. And we shouldn't be
able to get there anyway.

> 
> [1] https://lore.kernel.org/r/179318868950.3559520.6473416824609223980.b4-review@b4
> 

-- 
Best regards, Riabchun Vladimir
Linux Kernel Developer, Virtuozzo

_______________________________________________
Devel mailing list
Devel@openvz.org
https://lists.openvz.org/mailman/listinfo/devel

^ permalink raw reply	[flat|nested] 13+ messages in thread

end of thread, other threads:[~2026-11-16 16:54 UTC | newest]

Thread overview: 13+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2026-11-16 16:06 [Devel] [PATCH hci-8.1 v6 0/5] Fix kernel crash in overlayfs Vladimir Riabchun
2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 1/5] ve: Remove STOPPED state Vladimir Riabchun
2026-11-16 16:10   ` no-reply
2026-11-16 16:25     ` Vladimir Riabchun
2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 2/5] ve: Introduce override_ve/revert_ve helpers Vladimir Riabchun
2026-11-16 16:12   ` no-reply
2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 3/5] ve: Defer ve creds drop until ve is completely destroyed Vladimir Riabchun
2026-11-16 16:14   ` no-reply
2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 4/5] ovl: Keep creator VE in overlayfs superblock Vladimir Riabchun
2026-11-16 16:13   ` no-reply
2026-11-16 16:06 ` [Devel] [PATCH hci-8.1 v6 5/5] ve/overlayfs: Override VE together with creds Vladimir Riabchun
2026-11-16 16:19   ` no-reply
2026-11-16 16:52     ` Vladimir Riabchun

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox