[syzbot] [fs?] BUG: unable to handle kernel NULL pointer dereference in path_from_stashed

[syzbot] [fs?] BUG: unable to handle kernel NULL pointer dereference in path_from_stashed

[syzbot]

Hello,

syzbot found the following issue on:

HEAD commit:    51835949dda3 Merge tag 'net-next-6.11' of git://git.kernel..
git tree:       upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=1325e60d980000
kernel config:  https://syzkaller.appspot.com/x/.config?x=c1c1b0a8065e216
dashboard link: https://syzkaller.appspot.com/bug?extid=34a0ee986f61f15da35d
compiler:       arm-linux-gnueabi-gcc (Debian 12.2.0-14) 12.2.0, GNU ld (GNU Binutils for Debian) 2.40
userspace arch: arm
syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=11bd9a5e980000
C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=15726e95980000

Downloadable assets:
disk image (non-bootable): https://storage.googleapis.com/syzbot-assets/8ead8862021c/non_bootable_disk-51835949.raw.xz
vmlinux: https://storage.googleapis.com/syzbot-assets/a129ae4ab997/vmlinux-51835949.xz
kernel image: https://storage.googleapis.com/syzbot-assets/9339fe082652/zImage-51835949.xz

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+34a0ee986f61f15da35d@syzkaller.appspotmail.com

8<--- cut here ---
Unable to handle kernel NULL pointer dereference at virtual address 00000008 when read
[00000008] *pgd=844c0003, *pmd=fe12e003
Internal error: Oops: 205 [#1] PREEMPT SMP ARM
Modules linked in:
CPU: 0 PID: 3011 Comm: syz-executor103 Not tainted 6.10.0-syzkaller #0
Hardware name: ARM-Versatile Express
PC is at path_from_stashed+0x1c/0x308 fs/libfs.c:2204
LR is at open_namespace+0x44/0xbc fs/nsfs.c:102
pc : [<8053af54>]    lr : [<8054d6f8>]    psr: 80000013
sp : df959e80  ip : 84183000  fp : df959ec4
r10: 84183000  r9 : 00000003  r8 : 843f3300
r7 : 82caa250  r6 : 84183000  r5 : 00000000  r4 : 82625878
r3 : df959ecc  r2 : 00000008  r1 : 82c95800  r0 : 00000008
Flags: Nzcv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment user
Control: 30c5387d  Table: 843f1a00  DAC: fffffffd
Register r0 information: non-paged memory
Register r1 information: slab kmalloc-1k start 82c95800 pointer offset 0 size 1024
Register r2 information: non-paged memory
Register r3 information: 2-page vmalloc region starting at 0xdf958000 allocated at kernel_clone+0xac/0x3e4 kernel/fork.c:2780
Register r4 information: non-slab/vmalloc memory
Register r5 information: NULL pointer
Register r6 information: slab task_struct start 84183000 pointer offset 0 size 3072
Register r7 information: slab mnt_cache start 82caa240 pointer offset 16 size 184
Register r8 information: slab filp start 843f3300 pointer offset 0 size 160
Register r9 information: non-paged memory
Register r10 information: slab task_struct start 84183000 pointer offset 0 size 3072
Register r11 information: 2-page vmalloc region starting at 0xdf958000 allocated at kernel_clone+0xac/0x3e4 kernel/fork.c:2780
Register r12 information: slab task_struct start 84183000 pointer offset 0 size 3072
Process syz-executor103 (pid: 3011, stack limit = 0xdf958000)
Stack: (0xdf959e80 to 0xdf95a000)
9e80: 841b2400 00000008 df959eb4 df959e98 806e6544 804065ac 00000009 82625878
9ea0: 00000000 84183000 841b2c80 843f3300 00000003 84183000 df959ef4 df959ec8
9ec0: 8054d6f8 8053af44 df959ef4 00000000 00000000 f9244696 82625878 82625878
9ee0: 841b2400 00000008 df959f14 df959ef8 8055271c 8054d6c0 0000ff07 00000000
9f00: 843f3300 00000000 df959fa4 df959f18 8051a7f0 805525a4 000001b2 8020029c
9f20: 84183000 000001b2 df959fac df959f38 8020ba70 8042c724 83f01500 df959f80
9f40: 00000000 843f3300 00000003 82cb0800 df959f7c df959f60 805283b8 8027aebc
9f60: 83f01500 00000003 83f01500 00000003 df959fa4 f9244696 8026b8b0 00000000
9f80: 00000000 0008e058 00000036 8020029c 84183000 00000036 00000000 df959fa8
9fa0: 80200060 8051a6c8 00000000 00000000 00000003 0000ff07 00000000 00000000
9fc0: 00000000 00000000 0008e058 00000036 7ee54e0c 00000000 00000001 00000000
9fe0: 7ee54c70 7ee54c60 0001064c 0002e7a0 00000010 00000003 00000000 00000000
Call trace: 
[<8053af38>] (path_from_stashed) from [<8054d6f8>] (open_namespace+0x44/0xbc fs/nsfs.c:102)
 r10:84183000 r9:00000003 r8:843f3300 r7:841b2c80 r6:84183000 r5:00000000
 r4:82625878
[<8054d6b4>] (open_namespace) from [<8055271c>] (pidfd_ioctl+0x184/0x4c4 fs/pidfs.c:196)
 r6:00000008 r5:841b2400 r4:82625878
[<80552598>] (pidfd_ioctl) from [<8051a7f0>] (vfs_ioctl fs/ioctl.c:51 [inline])
[<80552598>] (pidfd_ioctl) from [<8051a7f0>] (do_vfs_ioctl fs/ioctl.c:861 [inline])
[<80552598>] (pidfd_ioctl) from [<8051a7f0>] (__do_sys_ioctl fs/ioctl.c:905 [inline])
[<80552598>] (pidfd_ioctl) from [<8051a7f0>] (sys_ioctl+0x134/0xda4 fs/ioctl.c:893)
 r7:00000000 r6:843f3300 r5:00000000 r4:0000ff07
[<8051a6bc>] (sys_ioctl) from [<80200060>] (ret_fast_syscall+0x0/0x1c arch/arm/mm/proc-v7.S:67)
Exception stack(0xdf959fa8 to 0xdf959ff0)
9fa0:                   00000000 00000000 00000003 0000ff07 00000000 00000000
9fc0: 00000000 00000000 0008e058 00000036 7ee54e0c 00000000 00000001 00000000
9fe0: 7ee54c70 7ee54c60 0001064c 0002e7a0
 r10:00000036 r9:84183000 r8:8020029c r7:00000036 r6:0008e058 r5:00000000
 r4:00000000
Code: e24dd01c e1a07001 e5911004 ee1dcf70 (e5905000) 
---[ end trace 0000000000000000 ]---
----------------
Code disassembly (best guess):
   0:	e24dd01c 	sub	sp, sp, #28
   4:	e1a07001 	mov	r7, r1
   8:	e5911004 	ldr	r1, [r1, #4]
   c:	ee1dcf70 	mrc	15, 0, ip, cr13, cr0, {3}
* 10:	e5905000 	ldr	r5, [r0] <-- trapping instruction


---
This report is generated by a bot. It may contain errors.
See https://goo.gl/tpsmEJ for more information about syzbot.
syzbot engineers can be reached at syzkaller@googlegroups.com.

syzbot will keep track of this issue. See:
https://goo.gl/tpsmEJ#status for how to communicate with syzbot.

If the report is already addressed, let syzbot know by replying with:
#syz fix: exact-commit-title

If you want syzbot to run the reproducer, reply with:
#syz test: git://repo/address.git branch-or-commit-hash
If you attach or paste a git patch, syzbot will apply it before testing.

If you want to overwrite report's subsystems, reply with:
#syz set subsystems: new-subsystem
(See the list of subsystem names on the web dashboard)

If the report is a duplicate of another one, reply with:
#syz dup: exact-subject-of-another-report

If you want to undo deduplication, reply with:
#syz undup

[PATCH] fs/pidfs: when time ns disabled add check for ioctl

[Edward Adam Davis]

syzbot call pidfd_ioctl() with cmd "PIDFD_GET_TIME_NAMESPACE" and disabled
CONFIG_TIME_NS, since time_ns is NULL, it will make NULL ponter deref in
open_namespace.

Reported-and-tested-by: syzbot+34a0ee986f61f15da35d@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=34a0ee986f61f15da35d
Signed-off-by: Edward Adam Davis <eadavis@qq.com>
---
 fs/pidfs.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/fs/pidfs.c b/fs/pidfs.c
index c9cb14181def..fe0ddab48f57 100644
--- a/fs/pidfs.c
+++ b/fs/pidfs.c
@@ -168,6 +168,8 @@ static long pidfd_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 	case PIDFD_GET_TIME_NAMESPACE:
 		get_time_ns(nsp->time_ns);
 		ns_common = to_ns_common(nsp->time_ns);
+		if (!nsp->time_ns)
+			return -EINVAL;
 		break;
 	case PIDFD_GET_TIME_FOR_CHILDREN_NAMESPACE:
 		get_time_ns(nsp->time_ns_for_children);
-- 
2.43.0

Re: [PATCH] fs/pidfs: when time ns disabled add check for ioctl

[Matthew Wilcox]

On Sun, Jul 21, 2024 at 02:23:12PM +0800, Edward Adam Davis wrote:
> syzbot call pidfd_ioctl() with cmd "PIDFD_GET_TIME_NAMESPACE" and disabled
> CONFIG_TIME_NS, since time_ns is NULL, it will make NULL ponter deref in
> open_namespace.

what about PIDFD_GET_TIME_FOR_CHILDREN_NAMESPACE?

Re: [PATCH] fs/pidfs: when time ns disabled add check for ioctl

[Christian Brauner]

On Sun, Jul 21, 2024 at 02:23:12PM GMT, Edward Adam Davis wrote:
> syzbot call pidfd_ioctl() with cmd "PIDFD_GET_TIME_NAMESPACE" and disabled
> CONFIG_TIME_NS, since time_ns is NULL, it will make NULL ponter deref in
> open_namespace.
> 
> Reported-and-tested-by: syzbot+34a0ee986f61f15da35d@syzkaller.appspotmail.com
> Closes: https://syzkaller.appspot.com/bug?extid=34a0ee986f61f15da35d
> Signed-off-by: Edward Adam Davis <eadavis@qq.com>
> ---

Hm, nsproxy is really messy in that regard. Some namespaces will always
be set to init_<type>_ns and others will be set to NULL.

That's an invitation for bugs such as this. Imho the correct fix is to
change nsproxy to always set nsp-><type>_ns to init_<type>_ns and no
code ever needs to worry about dereferencing NULL.

But that'll require more changes so this seems an appropriate fix for now.

[PATCH 1/2] pidfs: handle kernels without namespaces cleanly

[Christian Brauner]

The nsproxy structure contains nearly all of the namespaces associated
with a task. When a given namespace type is not supported by this kernel
the rules whether the corresponding pointer in struct nsproxy is NULL or
always init_<ns_type>_ns differ per namespace. Ideally, that wouldn't be
the case and for all namespace types we'd always set it to
init_<ns_type>_ns when the corresponding namespace type isn't supported.

Make sure we handle all namespaces where the pointer in struct nsproxy
can be NULL when the namespace type isn't supported.

Fixes: 5b08bd408534 ("pidfs: allow retrieval of namespace file descriptors") # mainline only
Signed-off-by: Christian Brauner <brauner@kernel.org>
---
 fs/pidfs.c | 65 +++++++++++++++++++++++++++++++++++-------------------
 1 file changed, 42 insertions(+), 23 deletions(-)

diff --git a/fs/pidfs.c b/fs/pidfs.c
index fe0ddab48f57..7ffdc88dfb52 100644
--- a/fs/pidfs.c
+++ b/fs/pidfs.c
@@ -119,7 +119,7 @@ static long pidfd_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 	struct task_struct *task __free(put_task) = NULL;
 	struct nsproxy *nsp __free(put_nsproxy) = NULL;
 	struct pid *pid = pidfd_pid(file);
-	struct ns_common *ns_common;
+	struct ns_common *ns_common = NULL;
 
 	if (arg)
 		return -EINVAL;
@@ -146,54 +146,73 @@ static long pidfd_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 	switch (cmd) {
 	/* Namespaces that hang of nsproxy. */
 	case PIDFD_GET_CGROUP_NAMESPACE:
-		get_cgroup_ns(nsp->cgroup_ns);
-		ns_common = to_ns_common(nsp->cgroup_ns);
+		if (IS_ENABLED(CONFIG_CGROUPS)) {
+			get_cgroup_ns(nsp->cgroup_ns);
+			ns_common = to_ns_common(nsp->cgroup_ns);
+		}
 		break;
 	case PIDFD_GET_IPC_NAMESPACE:
-		get_ipc_ns(nsp->ipc_ns);
-		ns_common = to_ns_common(nsp->ipc_ns);
+		if (IS_ENABLED(CONFIG_IPC_NS)) {
+			get_ipc_ns(nsp->ipc_ns);
+			ns_common = to_ns_common(nsp->ipc_ns);
+		}
 		break;
 	case PIDFD_GET_MNT_NAMESPACE:
 		get_mnt_ns(nsp->mnt_ns);
 		ns_common = to_ns_common(nsp->mnt_ns);
 		break;
 	case PIDFD_GET_NET_NAMESPACE:
-		ns_common = to_ns_common(nsp->net_ns);
-		get_net_ns(ns_common);
+		if (IS_ENABLED(CONFIG_NET_NS)) {
+			ns_common = to_ns_common(nsp->net_ns);
+			get_net_ns(ns_common);
+		}
 		break;
 	case PIDFD_GET_PID_FOR_CHILDREN_NAMESPACE:
-		get_pid_ns(nsp->pid_ns_for_children);
-		ns_common = to_ns_common(nsp->pid_ns_for_children);
+		if (IS_ENABLED(CONFIG_PID_NS)) {
+			get_pid_ns(nsp->pid_ns_for_children);
+			ns_common = to_ns_common(nsp->pid_ns_for_children);
+		}
 		break;
 	case PIDFD_GET_TIME_NAMESPACE:
-		get_time_ns(nsp->time_ns);
-		ns_common = to_ns_common(nsp->time_ns);
-		if (!nsp->time_ns)
-			return -EINVAL;
+		if (IS_ENABLED(CONFIG_TIME_NS)) {
+			get_time_ns(nsp->time_ns);
+			ns_common = to_ns_common(nsp->time_ns);
+		}
 		break;
 	case PIDFD_GET_TIME_FOR_CHILDREN_NAMESPACE:
-		get_time_ns(nsp->time_ns_for_children);
-		ns_common = to_ns_common(nsp->time_ns_for_children);
+		if (IS_ENABLED(CONFIG_TIME_NS)) {
+			get_time_ns(nsp->time_ns_for_children);
+			ns_common = to_ns_common(nsp->time_ns_for_children);
+		}
 		break;
 	case PIDFD_GET_UTS_NAMESPACE:
-		get_uts_ns(nsp->uts_ns);
-		ns_common = to_ns_common(nsp->uts_ns);
+		if (IS_ENABLED(CONFIG_UTS_NS)) {
+			get_uts_ns(nsp->uts_ns);
+			ns_common = to_ns_common(nsp->uts_ns);
+		}
 		break;
 	/* Namespaces that don't hang of nsproxy. */
 	case PIDFD_GET_USER_NAMESPACE:
-		rcu_read_lock();
-		ns_common = to_ns_common(get_user_ns(task_cred_xxx(task, user_ns)));
-		rcu_read_unlock();
+		if (IS_ENABLED(CONFIG_USER_NS)) {
+			rcu_read_lock();
+			ns_common = to_ns_common(get_user_ns(task_cred_xxx(task, user_ns)));
+			rcu_read_unlock();
+		}
 		break;
 	case PIDFD_GET_PID_NAMESPACE:
-		rcu_read_lock();
-		ns_common = to_ns_common(get_pid_ns(task_active_pid_ns(task)));
-		rcu_read_unlock();
+		if (IS_ENABLED(CONFIG_PID_NS)) {
+			rcu_read_lock();
+			ns_common = to_ns_common( get_pid_ns(task_active_pid_ns(task)));
+			rcu_read_unlock();
+		}
 		break;
 	default:
 		return -ENOIOCTLCMD;
 	}
 
+	if (!ns_common)
+		return -EOPNOTSUPP;
+
 	/* open_namespace() unconditionally consumes the reference */
 	return open_namespace(ns_common);
 }
-- 
2.43.0

[PATCH 2/2] pidfs: add selftests for new namespace ioctls

[Christian Brauner]

Add selftests to verify that deriving namespace file descriptors from
pidfd file descriptors works correctly.

Signed-off-by: Christian Brauner <brauner@kernel.org>
---
 .../selftests/pidfd/pidfd_setns_test.c        | 258 +++++++++++++++---
 1 file changed, 227 insertions(+), 31 deletions(-)

diff --git a/tools/testing/selftests/pidfd/pidfd_setns_test.c b/tools/testing/selftests/pidfd/pidfd_setns_test.c
index 47746b0c6acd..7c2a4349170a 100644
--- a/tools/testing/selftests/pidfd/pidfd_setns_test.c
+++ b/tools/testing/selftests/pidfd/pidfd_setns_test.c
@@ -16,11 +16,56 @@
 #include <unistd.h>
 #include <sys/socket.h>
 #include <sys/stat.h>
+#include <linux/ioctl.h>
 
 #include "pidfd.h"
 #include "../clone3/clone3_selftests.h"
 #include "../kselftest_harness.h"
 
+#ifndef PIDFS_IOCTL_MAGIC
+#define PIDFS_IOCTL_MAGIC 0xFF
+#endif
+
+#ifndef PIDFD_GET_CGROUP_NAMESPACE
+#define PIDFD_GET_CGROUP_NAMESPACE            _IO(PIDFS_IOCTL_MAGIC, 1)
+#endif
+
+#ifndef PIDFD_GET_IPC_NAMESPACE
+#define PIDFD_GET_IPC_NAMESPACE               _IO(PIDFS_IOCTL_MAGIC, 2)
+#endif
+
+#ifndef PIDFD_GET_MNT_NAMESPACE
+#define PIDFD_GET_MNT_NAMESPACE               _IO(PIDFS_IOCTL_MAGIC, 3)
+#endif
+
+#ifndef PIDFD_GET_NET_NAMESPACE
+#define PIDFD_GET_NET_NAMESPACE               _IO(PIDFS_IOCTL_MAGIC, 4)
+#endif
+
+#ifndef PIDFD_GET_PID_NAMESPACE
+#define PIDFD_GET_PID_NAMESPACE               _IO(PIDFS_IOCTL_MAGIC, 5)
+#endif
+
+#ifndef PIDFD_GET_PID_FOR_CHILDREN_NAMESPACE
+#define PIDFD_GET_PID_FOR_CHILDREN_NAMESPACE  _IO(PIDFS_IOCTL_MAGIC, 6)
+#endif
+
+#ifndef PIDFD_GET_TIME_NAMESPACE
+#define PIDFD_GET_TIME_NAMESPACE              _IO(PIDFS_IOCTL_MAGIC, 7)
+#endif
+
+#ifndef PIDFD_GET_TIME_FOR_CHILDREN_NAMESPACE
+#define PIDFD_GET_TIME_FOR_CHILDREN_NAMESPACE _IO(PIDFS_IOCTL_MAGIC, 8)
+#endif
+
+#ifndef PIDFD_GET_USER_NAMESPACE
+#define PIDFD_GET_USER_NAMESPACE              _IO(PIDFS_IOCTL_MAGIC, 9)
+#endif
+
+#ifndef PIDFD_GET_UTS_NAMESPACE
+#define PIDFD_GET_UTS_NAMESPACE               _IO(PIDFS_IOCTL_MAGIC, 10)
+#endif
+
 enum {
 	PIDFD_NS_USER,
 	PIDFD_NS_MNT,
@@ -31,22 +76,25 @@ enum {
 	PIDFD_NS_CGROUP,
 	PIDFD_NS_PIDCLD,
 	PIDFD_NS_TIME,
+	PIDFD_NS_TIMECLD,
 	PIDFD_NS_MAX
 };
 
 const struct ns_info {
 	const char *name;
 	int flag;
+	unsigned int pidfd_ioctl;
 } ns_info[] = {
-	[PIDFD_NS_USER]   = { "user",             CLONE_NEWUSER,   },
-	[PIDFD_NS_MNT]    = { "mnt",              CLONE_NEWNS,     },
-	[PIDFD_NS_PID]    = { "pid",              CLONE_NEWPID,    },
-	[PIDFD_NS_UTS]    = { "uts",              CLONE_NEWUTS,    },
-	[PIDFD_NS_IPC]    = { "ipc",              CLONE_NEWIPC,    },
-	[PIDFD_NS_NET]    = { "net",              CLONE_NEWNET,    },
-	[PIDFD_NS_CGROUP] = { "cgroup",           CLONE_NEWCGROUP, },
-	[PIDFD_NS_PIDCLD] = { "pid_for_children", 0,               },
-	[PIDFD_NS_TIME]	  = { "time",             CLONE_NEWTIME,   },
+	[PIDFD_NS_USER]    = { "user",              CLONE_NEWUSER,   PIDFD_GET_USER_NAMESPACE,              },
+	[PIDFD_NS_MNT]     = { "mnt",               CLONE_NEWNS,     PIDFD_GET_MNT_NAMESPACE,               },
+	[PIDFD_NS_PID]     = { "pid",               CLONE_NEWPID,    PIDFD_GET_PID_NAMESPACE,               },
+	[PIDFD_NS_UTS]     = { "uts",               CLONE_NEWUTS,    PIDFD_GET_UTS_NAMESPACE,               },
+	[PIDFD_NS_IPC]     = { "ipc",               CLONE_NEWIPC,    PIDFD_GET_IPC_NAMESPACE,               },
+	[PIDFD_NS_NET]     = { "net",               CLONE_NEWNET,    PIDFD_GET_NET_NAMESPACE,               },
+	[PIDFD_NS_CGROUP]  = { "cgroup",            CLONE_NEWCGROUP, PIDFD_GET_CGROUP_NAMESPACE,            },
+	[PIDFD_NS_TIME]	   = { "time",              CLONE_NEWTIME,   PIDFD_GET_TIME_NAMESPACE,              },
+	[PIDFD_NS_PIDCLD]  = { "pid_for_children",  0,               PIDFD_GET_PID_FOR_CHILDREN_NAMESPACE,  },
+	[PIDFD_NS_TIMECLD] = { "time_for_children", 0,               PIDFD_GET_TIME_FOR_CHILDREN_NAMESPACE, },
 };
 
 FIXTURE(current_nsset)
@@ -54,6 +102,7 @@ FIXTURE(current_nsset)
 	pid_t pid;
 	int pidfd;
 	int nsfds[PIDFD_NS_MAX];
+	int child_pidfd_derived_nsfds[PIDFD_NS_MAX];
 
 	pid_t child_pid_exited;
 	int child_pidfd_exited;
@@ -61,10 +110,12 @@ FIXTURE(current_nsset)
 	pid_t child_pid1;
 	int child_pidfd1;
 	int child_nsfds1[PIDFD_NS_MAX];
+	int child_pidfd_derived_nsfds1[PIDFD_NS_MAX];
 
 	pid_t child_pid2;
 	int child_pidfd2;
 	int child_nsfds2[PIDFD_NS_MAX];
+	int child_pidfd_derived_nsfds2[PIDFD_NS_MAX];
 };
 
 static int sys_waitid(int which, pid_t pid, int options)
@@ -128,9 +179,12 @@ FIXTURE_SETUP(current_nsset)
 	char c;
 
 	for (i = 0; i < PIDFD_NS_MAX; i++) {
-		self->nsfds[i]		= -EBADF;
-		self->child_nsfds1[i]	= -EBADF;
-		self->child_nsfds2[i]	= -EBADF;
+		self->nsfds[i]				= -EBADF;
+		self->child_nsfds1[i]			= -EBADF;
+		self->child_nsfds2[i]			= -EBADF;
+		self->child_pidfd_derived_nsfds[i]	= -EBADF;
+		self->child_pidfd_derived_nsfds1[i]	= -EBADF;
+		self->child_pidfd_derived_nsfds2[i]	= -EBADF;
 	}
 
 	proc_fd = open("/proc/self/ns", O_DIRECTORY | O_CLOEXEC);
@@ -139,6 +193,11 @@ FIXTURE_SETUP(current_nsset)
 	}
 
 	self->pid = getpid();
+	self->pidfd = sys_pidfd_open(self->pid, 0);
+	EXPECT_GT(self->pidfd, 0) {
+		TH_LOG("%m - Failed to open pidfd for process %d", self->pid);
+	}
+
 	for (i = 0; i < PIDFD_NS_MAX; i++) {
 		const struct ns_info *info = &ns_info[i];
 		self->nsfds[i] = openat(proc_fd, info->name, O_RDONLY | O_CLOEXEC);
@@ -148,20 +207,27 @@ FIXTURE_SETUP(current_nsset)
 				       info->name, self->pid);
 			}
 		}
-	}
 
-	self->pidfd = sys_pidfd_open(self->pid, 0);
-	EXPECT_GT(self->pidfd, 0) {
-		TH_LOG("%m - Failed to open pidfd for process %d", self->pid);
+		self->child_pidfd_derived_nsfds[i] = ioctl(self->pidfd, info->pidfd_ioctl, 0);
+		if (self->child_pidfd_derived_nsfds[i] < 0) {
+			EXPECT_EQ(errno, EOPNOTSUPP) {
+				TH_LOG("%m - Failed to derive %s namespace from pidfd of process %d",
+				       info->name, self->pid);
+			}
+		}
 	}
 
 	/* Create task that exits right away. */
-	self->child_pid_exited = create_child(&self->child_pidfd_exited,
-					      CLONE_NEWUSER | CLONE_NEWNET);
+	self->child_pid_exited = create_child(&self->child_pidfd_exited, 0);
 	EXPECT_GE(self->child_pid_exited, 0);
 
-	if (self->child_pid_exited == 0)
+	if (self->child_pid_exited == 0) {
+		if (self->nsfds[PIDFD_NS_USER] >= 0 && unshare(CLONE_NEWUSER) < 0)
+			_exit(EXIT_FAILURE);
+		if (self->nsfds[PIDFD_NS_NET] >= 0 && unshare(CLONE_NEWNET) < 0)
+			_exit(EXIT_FAILURE);
 		_exit(EXIT_SUCCESS);
+	}
 
 	ASSERT_EQ(sys_waitid(P_PID, self->child_pid_exited, WEXITED | WNOWAIT), 0);
 
@@ -174,18 +240,43 @@ FIXTURE_SETUP(current_nsset)
 	EXPECT_EQ(ret, 0);
 
 	/* Create tasks that will be stopped. */
-	self->child_pid1 = create_child(&self->child_pidfd1,
-					CLONE_NEWUSER | CLONE_NEWNS |
-					CLONE_NEWCGROUP | CLONE_NEWIPC |
-					CLONE_NEWUTS | CLONE_NEWPID |
-					CLONE_NEWNET);
+	if (self->nsfds[PIDFD_NS_USER] >= 0 && self->nsfds[PIDFD_NS_PID] >= 0)
+		self->child_pid1 = create_child(&self->child_pidfd1, CLONE_NEWUSER | CLONE_NEWPID);
+	else if (self->nsfds[PIDFD_NS_PID] >= 0)
+		self->child_pid1 = create_child(&self->child_pidfd1, CLONE_NEWPID);
+	else if (self->nsfds[PIDFD_NS_USER] >= 0)
+		self->child_pid1 = create_child(&self->child_pidfd1, CLONE_NEWUSER);
+	else
+		self->child_pid1 = create_child(&self->child_pidfd1, 0);
 	EXPECT_GE(self->child_pid1, 0);
 
 	if (self->child_pid1 == 0) {
 		close(ipc_sockets[0]);
 
-		if (!switch_timens())
+		if (self->nsfds[PIDFD_NS_MNT] >= 0 && unshare(CLONE_NEWNS) < 0) {
+			TH_LOG("%m - Failed to unshare mount namespace for process %d", self->pid);
 			_exit(EXIT_FAILURE);
+		}
+		if (self->nsfds[PIDFD_NS_CGROUP] >= 0 && unshare(CLONE_NEWCGROUP) < 0) {
+			TH_LOG("%m - Failed to unshare cgroup namespace for process %d", self->pid);
+			_exit(EXIT_FAILURE);
+		}
+		if (self->nsfds[PIDFD_NS_IPC] >= 0 && unshare(CLONE_NEWIPC) < 0) {
+			TH_LOG("%m - Failed to unshare ipc namespace for process %d", self->pid);
+			_exit(EXIT_FAILURE);
+		}
+		if (self->nsfds[PIDFD_NS_UTS] >= 0 && unshare(CLONE_NEWUTS) < 0) {
+			TH_LOG("%m - Failed to unshare uts namespace for process %d", self->pid);
+			_exit(EXIT_FAILURE);
+		}
+		if (self->nsfds[PIDFD_NS_NET] >= 0 && unshare(CLONE_NEWNET) < 0) {
+			TH_LOG("%m - Failed to unshare net namespace for process %d", self->pid);
+			_exit(EXIT_FAILURE);
+		}
+		if (self->nsfds[PIDFD_NS_TIME] >= 0 && !switch_timens()) {
+			TH_LOG("%m - Failed to unshare time namespace for process %d", self->pid);
+			_exit(EXIT_FAILURE);
+		}
 
 		if (write_nointr(ipc_sockets[1], "1", 1) < 0)
 			_exit(EXIT_FAILURE);
@@ -203,18 +294,43 @@ FIXTURE_SETUP(current_nsset)
 	ret = socketpair(AF_LOCAL, SOCK_STREAM | SOCK_CLOEXEC, 0, ipc_sockets);
 	EXPECT_EQ(ret, 0);
 
-	self->child_pid2 = create_child(&self->child_pidfd2,
-					CLONE_NEWUSER | CLONE_NEWNS |
-					CLONE_NEWCGROUP | CLONE_NEWIPC |
-					CLONE_NEWUTS | CLONE_NEWPID |
-					CLONE_NEWNET);
+	if (self->nsfds[PIDFD_NS_USER] >= 0 && self->nsfds[PIDFD_NS_PID] >= 0)
+		self->child_pid2 = create_child(&self->child_pidfd2, CLONE_NEWUSER | CLONE_NEWPID);
+	else if (self->nsfds[PIDFD_NS_PID] >= 0)
+		self->child_pid2 = create_child(&self->child_pidfd2, CLONE_NEWPID);
+	else if (self->nsfds[PIDFD_NS_USER] >= 0)
+		self->child_pid2 = create_child(&self->child_pidfd2, CLONE_NEWUSER);
+	else
+		self->child_pid2 = create_child(&self->child_pidfd2, 0);
 	EXPECT_GE(self->child_pid2, 0);
 
 	if (self->child_pid2 == 0) {
 		close(ipc_sockets[0]);
 
-		if (!switch_timens())
+		if (self->nsfds[PIDFD_NS_MNT] >= 0 && unshare(CLONE_NEWNS) < 0) {
+			TH_LOG("%m - Failed to unshare mount namespace for process %d", self->pid);
+			_exit(EXIT_FAILURE);
+		}
+		if (self->nsfds[PIDFD_NS_CGROUP] >= 0 && unshare(CLONE_NEWCGROUP) < 0) {
+			TH_LOG("%m - Failed to unshare cgroup namespace for process %d", self->pid);
 			_exit(EXIT_FAILURE);
+		}
+		if (self->nsfds[PIDFD_NS_IPC] >= 0 && unshare(CLONE_NEWIPC) < 0) {
+			TH_LOG("%m - Failed to unshare ipc namespace for process %d", self->pid);
+			_exit(EXIT_FAILURE);
+		}
+		if (self->nsfds[PIDFD_NS_UTS] >= 0 && unshare(CLONE_NEWUTS) < 0) {
+			TH_LOG("%m - Failed to unshare uts namespace for process %d", self->pid);
+			_exit(EXIT_FAILURE);
+		}
+		if (self->nsfds[PIDFD_NS_NET] >= 0 && unshare(CLONE_NEWNET) < 0) {
+			TH_LOG("%m - Failed to unshare net namespace for process %d", self->pid);
+			_exit(EXIT_FAILURE);
+		}
+		if (self->nsfds[PIDFD_NS_TIME] >= 0 && !switch_timens()) {
+			TH_LOG("%m - Failed to unshare time namespace for process %d", self->pid);
+			_exit(EXIT_FAILURE);
+		}
 
 		if (write_nointr(ipc_sockets[1], "1", 1) < 0)
 			_exit(EXIT_FAILURE);
@@ -267,6 +383,22 @@ FIXTURE_SETUP(current_nsset)
 				       info->name, self->child_pid1);
 			}
 		}
+
+		self->child_pidfd_derived_nsfds1[i] = ioctl(self->child_pidfd1, info->pidfd_ioctl, 0);
+		if (self->child_pidfd_derived_nsfds1[i] < 0) {
+			EXPECT_EQ(errno, EOPNOTSUPP) {
+				TH_LOG("%m - Failed to derive %s namespace from pidfd of process %d",
+				       info->name, self->child_pid1);
+			}
+		}
+
+		self->child_pidfd_derived_nsfds2[i] = ioctl(self->child_pidfd2, info->pidfd_ioctl, 0);
+		if (self->child_pidfd_derived_nsfds2[i] < 0) {
+			EXPECT_EQ(errno, EOPNOTSUPP) {
+				TH_LOG("%m - Failed to derive %s namespace from pidfd of process %d",
+				       info->name, self->child_pid2);
+			}
+		}
 	}
 
 	close(proc_fd);
@@ -288,6 +420,12 @@ FIXTURE_TEARDOWN(current_nsset)
 			close(self->child_nsfds1[i]);
 		if (self->child_nsfds2[i] >= 0)
 			close(self->child_nsfds2[i]);
+		if (self->child_pidfd_derived_nsfds[i] >= 0)
+			close(self->child_pidfd_derived_nsfds[i]);
+		if (self->child_pidfd_derived_nsfds1[i] >= 0)
+			close(self->child_pidfd_derived_nsfds1[i]);
+		if (self->child_pidfd_derived_nsfds2[i] >= 0)
+			close(self->child_pidfd_derived_nsfds2[i]);
 	}
 
 	if (self->child_pidfd1 >= 0)
@@ -446,6 +584,42 @@ TEST_F(current_nsset, nsfd_incremental_setns)
 	}
 }
 
+TEST_F(current_nsset, pidfd_derived_nsfd_incremental_setns)
+{
+	int i;
+	pid_t pid;
+
+	pid = getpid();
+	for (i = 0; i < PIDFD_NS_MAX; i++) {
+		const struct ns_info *info = &ns_info[i];
+		int nsfd;
+
+		if (self->child_pidfd_derived_nsfds1[i] < 0)
+			continue;
+
+		if (info->flag) {
+			ASSERT_EQ(setns(self->child_pidfd_derived_nsfds1[i], info->flag), 0) {
+				TH_LOG("%m - Failed to setns to %s namespace of %d via nsfd %d",
+				       info->name, self->child_pid1,
+				       self->child_pidfd_derived_nsfds1[i]);
+			}
+		}
+
+		/* Verify that we have changed to the correct namespaces. */
+		if (info->flag == CLONE_NEWPID)
+			nsfd = self->child_pidfd_derived_nsfds[i];
+		else
+			nsfd = self->child_pidfd_derived_nsfds1[i];
+		ASSERT_EQ(in_same_namespace(nsfd, pid, info->name), 1) {
+			TH_LOG("setns failed to place us correctly into %s namespace of %d via nsfd %d",
+			       info->name, self->child_pid1,
+			       self->child_pidfd_derived_nsfds1[i]);
+		}
+		TH_LOG("Managed to correctly setns to %s namespace of %d via nsfd %d",
+		       info->name, self->child_pid1, self->child_pidfd_derived_nsfds1[i]);
+	}
+}
+
 TEST_F(current_nsset, pidfd_one_shot_setns)
 {
 	unsigned flags = 0;
@@ -542,6 +716,28 @@ TEST_F(current_nsset, no_foul_play)
 		       info->name, self->child_pid2,
 		       self->child_nsfds2[i]);
 	}
+
+	/*
+	 * Can't setns to a user namespace outside of our hierarchy since we
+	 * don't have caps in there and didn't create it. That means that under
+	 * no circumstances should we be able to setns to any of the other
+	 * ones since they aren't owned by our user namespace.
+	 */
+	for (i = 0; i < PIDFD_NS_MAX; i++) {
+		const struct ns_info *info = &ns_info[i];
+
+		if (self->child_pidfd_derived_nsfds2[i] < 0 || !info->flag)
+			continue;
+
+		ASSERT_NE(setns(self->child_pidfd_derived_nsfds2[i], info->flag), 0) {
+			TH_LOG("Managed to setns to %s namespace of %d via nsfd %d",
+			       info->name, self->child_pid2,
+			       self->child_pidfd_derived_nsfds2[i]);
+		}
+		TH_LOG("%m - Correctly failed to setns to %s namespace of %d via nsfd %d",
+		       info->name, self->child_pid2,
+		       self->child_pidfd_derived_nsfds2[i]);
+	}
 }
 
 TEST(setns_einval)
-- 
2.43.0