Skip to contentSkip to conversation

fork: extend clone3() to support setting a PID

PATCH Superseded Adrian Reber 5 participants 12 messages Aug 12, 2019 – Aug 14, 2019

Subsystems (5)

v6 of 11 Superseded Latest is v11 →

lore
Share & tools
lore
#0 lore raw

PATCH v6 1/2 fork: extend clone3() to support setting a PID

The main motivation to add set_tid to clone3() is CRIU.

To restore a process with the same PID/TID CRIU currently uses /proc/sys/kernel/ns_last_pid. It writes the desired (PID - 1) to ns_last_pid and then (quickly) does a clone(). This works most of the time, but it is racy. It is also slow as it requires multiple syscalls.

Extending clone3() to support set_tid makes it possible restore a process using CRIU without accessing /proc/sys/kernel/ns_last_pid and race free (as long as the desired PID/TID is available).

This clone3() extension places the same restrictions (CAP_SYS_ADMIN) on clone3() with set_tid as they are currently in place for ns_last_pid.

---
v2:
- Removed (size < sizeof(struct clone_args)) as discussed with Christian and Dmitry
- Added comment to ((set_tid != 1) && idr_get_cursor() <= 1) (Oleg)
- Use idr_alloc() instead of idr_alloc_cyclic() (Oleg)

v3:
- Return EEXIST if PID is already in use (Christian)
- Drop CLONE_SET_TID (Christian and Oleg)
- Use idr_is_empty() instead of idr_get_cursor() (Oleg)
- Handle different struct clone_args sizes (Dmitry)

v4:
- Rework struct size check with defines (Christian)
- Reduce number of set_tid checks (Oleg)
- Less parentheses and more robust code (Oleg)
- Do ns_capable() on correct user_ns (Oleg, Christian)

v5:
- make set_tid checks earlier in alloc_pid() (Christian)
- remove unnecessary comment and struct size check (Christian)

v6:
- remove CLONE_SET_TID from description (Christian)
- add clone3() tests for different clone_args sizes (Christian)
- move more set_tid checks to alloc_pid() (Oleg)
- make all set_tid checks lockless (Oleg)

---
 include/linux/pid.h        |  2 +-
 include/linux/sched/task.h |  1 +
 include/uapi/linux/sched.h |  1 +
 kernel/fork.c              | 14 ++++++++++++--
 kernel/pid.c               | 37 ++++++++++++++++++++++++++++++-------
 5 files changed, 45 insertions(+), 10 deletions(-)
include/linux/pid.h +1 -1
index 2a83e434db9d..052000db0ced 100644
@@ -116,7 +116,7 @@ extern struct pid *find_vpid(int nr);
 extern struct pid *find_get_pid(int nr);
 extern struct pid *find_ge_pid(int nr, struct pid_namespace *);
 
-extern struct pid *alloc_pid(struct pid_namespace *ns);
+extern struct pid *alloc_pid(struct pid_namespace *ns, pid_t set_tid);
 extern void free_pid(struct pid *pid);
 extern void disable_pid_allocation(struct pid_namespace *ns);
 
include/linux/sched/task.h +1 -0
index 0497091e40c1..4f2a80564332 100644
@@ -26,6 +26,7 @@ struct kernel_clone_args {
 	unsigned long stack;
 	unsigned long stack_size;
 	unsigned long tls;
+	pid_t set_tid;
 };
 
 /*
include/uapi/linux/sched.h +1 -0
index b3105ac1381a..e1ce103a2c47 100644
@@ -45,6 +45,7 @@ struct clone_args {
 	__aligned_u64 stack;
 	__aligned_u64 stack_size;
 	__aligned_u64 tls;
+	__aligned_u64 set_tid;
 };
 
 /*
kernel/fork.c +12 -2
index 2852d0e76ea3..8317d408a8d6 100644
@@ -117,6 +117,11 @@
  */
 #define MAX_THREADS FUTEX_TID_MASK
 
+/*
+ * For different sizes of struct clone_args
+ */
+#define CLONE3_ARGS_SIZE_V0 64
+
 /*
  * Protected counters by write_lock_irq(&tasklist_lock)
  */
@@ -2031,7 +2036,7 @@ static __latent_entropy struct task_struct *copy_process(
 	stackleak_task_init(p);
 
 	if (pid != &init_struct_pid) {
-		pid = alloc_pid(p->nsproxy->pid_ns_for_children);
+		pid = alloc_pid(p->nsproxy->pid_ns_for_children, args->set_tid);
 		if (IS_ERR(pid)) {
 			retval = PTR_ERR(pid);
 			goto bad_fork_cleanup_thread;
@@ -2535,9 +2540,13 @@ noinline static int copy_clone_args_from_user(struct kernel_clone_args *kargs,
 	if (unlikely(size > PAGE_SIZE))
 		return -E2BIG;
 
-	if (unlikely(size < sizeof(struct clone_args)))
+	if (unlikely(size < CLONE3_ARGS_SIZE_V0))
 		return -EINVAL;
 
+	if (size < sizeof(struct clone_args))
+		memset((void *)&args + size, 0,
+				sizeof(struct clone_args) - size);
+
 	if (unlikely(!access_ok(uargs, size)))
 		return -EFAULT;
 
@@ -2571,6 +2580,7 @@ noinline static int copy_clone_args_from_user(struct kernel_clone_args *kargs,
 		.stack		= args.stack,
 		.stack_size	= args.stack_size,
 		.tls		= args.tls,
+		.set_tid	= args.set_tid,
 	};
 
 	return 0;
kernel/pid.c +30 -7
index 0a9f2e437217..5cdab73b9094 100644
@@ -157,7 +157,7 @@ void free_pid(struct pid *pid)
 	call_rcu(&pid->rcu, delayed_put_pid);
 }
 
-struct pid *alloc_pid(struct pid_namespace *ns)
+struct pid *alloc_pid(struct pid_namespace *ns, int set_tid)
 {
 	struct pid *pid;
 	enum pid_type type;
@@ -166,6 +166,16 @@ struct pid *alloc_pid(struct pid_namespace *ns)
 	struct upid *upid;
 	int retval = -ENOMEM;
 
+	if (set_tid) {
+		if (set_tid < 0 || set_tid >= pid_max)
+			return ERR_PTR(-EINVAL);
+		/* Also fail if a PID != 1 is requested and no PID 1 exists */
+		if (set_tid != 1 && !ns->child_reaper)
+			return ERR_PTR(-EINVAL);
+		if (!ns_capable(ns->user_ns, CAP_SYS_ADMIN))
+			return ERR_PTR(-EPERM);
+	}
+
 	pid = kmem_cache_alloc(ns->pid_cachep, GFP_KERNEL);
 	if (!pid)
 		return ERR_PTR(retval);
@@ -186,12 +196,25 @@ struct pid *alloc_pid(struct pid_namespace *ns)
 		if (idr_get_cursor(&tmp->idr) > RESERVED_PIDS)
 			pid_min = RESERVED_PIDS;
 
-		/*
-		 * Store a null pointer so find_pid_ns does not find
-		 * a partially initialized PID (see below).
-		 */
-		nr = idr_alloc_cyclic(&tmp->idr, NULL, pid_min,
-				      pid_max, GFP_ATOMIC);
+		if (set_tid) {
+			nr = idr_alloc(&tmp->idr, NULL, set_tid,
+					set_tid + 1, GFP_ATOMIC);
+			/*
+			 * If ENOSPC is returned it means that the PID is
+			 * alreay in use. Return EEXIST in that case.
+			 */
+			if (nr == -ENOSPC)
+				nr = -EEXIST;
+			/* Only use set_tid for one PID namespace. */
+			set_tid = 0;
+		} else {
+			/*
+			 * Store a null pointer so find_pid_ns does not find
+			 * a partially initialized PID (see below).
+			 */
+			nr = idr_alloc_cyclic(&tmp->idr, NULL, pid_min,
+					      pid_max, GFP_ATOMIC);
+		}
 		spin_unlock_irq(&pidmap_lock);
 		idr_preload_end();
 
Signature · 1 line
2.21.0
#1 lore raw

PATCH v6 2/2 selftests: add tests for clone3()

This tests clone3() with and without set_tid to see if all desired PIDs are working as expected. The test tries to clone3() with a set_tid of -1, 1, pid_max, a PID which is already in use and an unused PID. The same tests are also running in PID namespace.

In addition the clone3 test (without set_tid) tries to call clone3() with different sizes of clone_args.

---
 tools/testing/selftests/clone3/.gitignore     |   2 +
 tools/testing/selftests/clone3/Makefile       |  11 +
 tools/testing/selftests/clone3/clone3.c       | 231 ++++++++++++++++++
 .../testing/selftests/clone3/clone3_set_tid.c | 161 ++++++++++++
 4 files changed, 405 insertions(+)
 create mode 100644 tools/testing/selftests/clone3/.gitignore
 create mode 100644 tools/testing/selftests/clone3/Makefile
 create mode 100644 tools/testing/selftests/clone3/clone3.c
 create mode 100644 tools/testing/selftests/clone3/clone3_set_tid.c
tools/testing/selftests/clone3/.gitignore new file +2 -0
new file mode 100644
index 000000000000..c63c64a78ddf
@@ -0,0 +1,2 @@
+clone3_set_tid
+clone3
tools/testing/selftests/clone3/Makefile new file +11 -0
new file mode 100644
index 000000000000..4efcf45b995b
@@ -0,0 +1,11 @@
+# SPDX-License-Identifier: GPL-2.0
+uname_M := $(shell uname -m 2>/dev/null || echo not)
+ARCH ?= $(shell echo $(uname_M) | sed -e s/i.86/i386/)
+
+CFLAGS += -I../../../../usr/include/
+
+ifeq ($(ARCH),x86_64)
+	TEST_GEN_PROGS := clone3 clone3_set_tid
+endif
+
+include ../lib.mk
tools/testing/selftests/clone3/clone3.c new file +231 -0

231 lines not on this page · raw message

tools/testing/selftests/clone3/clone3_set_tid.c new file +161 -0

161 lines not on this page · raw message

Signature · 1 line
2.21.0
#2 lore raw

Re: [PATCH v6 2/2] selftests: add tests for clone3()

Quoted 9 lines · Adrian Reber

This tests clone3() with and without set_tid to see if all desired PIDs are working as expected. The test tries to clone3() with a set_tid of -1, 1, pid_max, a PID which is already in use and an unused PID. The same tests are also running in PID namespace.

In addition the clone3 test (without set_tid) tries to call clone3() with different sizes of clone_args.

This is missing a .gitignore file for the new binaries. But no need to resend I'll just add that myself.

#3 lore raw

Re: [PATCH v6 2/2] selftests: add tests for clone3()

Quoted 6 lines · Christian Brauner
Quoted 5 lines · Adrian Reber

This tests clone3() with and without set_tid to see if all desired PIDs are working as expected. The test tries to clone3() with a set_tid of -1, 1, pid_max, a PID which is already in use and an unused PID. The same tests are also running in PID namespace.

Quoted 7 lines
Quoted 4 lines

In addition the clone3 test (without set_tid) tries to call clone3() with different sizes of clone_args.

This is missing a .gitignore file for the new binaries. But no need to resend I'll just add that myself.

Oh, I see it now. :)

#4 lore raw
In reply to Adrian Reber
Quoted 151 lines · Adrian Reber

The main motivation to add set_tid to clone3() is CRIU.

To restore a process with the same PID/TID CRIU currently uses /proc/sys/kernel/ns_last_pid. It writes the desired (PID - 1) to ns_last_pid and then (quickly) does a clone(). This works most of the time, but it is racy. It is also slow as it requires multiple syscalls.

Extending clone3() to support set_tid makes it possible restore a process using CRIU without accessing /proc/sys/kernel/ns_last_pid and race free (as long as the desired PID/TID is available).

This clone3() extension places the same restrictions (CAP_SYS_ADMIN) on clone3() with set_tid as they are currently in place for ns_last_pid.

---
v2:
- Removed (size < sizeof(struct clone_args)) as discussed with Christian and Dmitry
- Added comment to ((set_tid != 1) && idr_get_cursor() <= 1) (Oleg)
- Use idr_alloc() instead of idr_alloc_cyclic() (Oleg)

v3:
- Return EEXIST if PID is already in use (Christian)
- Drop CLONE_SET_TID (Christian and Oleg)
- Use idr_is_empty() instead of idr_get_cursor() (Oleg)
- Handle different struct clone_args sizes (Dmitry)

v4:
- Rework struct size check with defines (Christian)
- Reduce number of set_tid checks (Oleg)
- Less parentheses and more robust code (Oleg)
- Do ns_capable() on correct user_ns (Oleg, Christian)

v5:
- make set_tid checks earlier in alloc_pid() (Christian)
- remove unnecessary comment and struct size check (Christian)

v6:
- remove CLONE_SET_TID from description (Christian)
- add clone3() tests for different clone_args sizes (Christian)
- move more set_tid checks to alloc_pid() (Oleg)
- make all set_tid checks lockless (Oleg)

---
 include/linux/pid.h        |  2 +-
 include/linux/sched/task.h |  1 +
 include/uapi/linux/sched.h |  1 +
 kernel/fork.c              | 14 ++++++++++++--
 kernel/pid.c               | 37 ++++++++++++++++++++++++++++++-------
 5 files changed, 45 insertions(+), 10 deletions(-)
include/linux/pid.h +1 -1
index 2a83e434db9d..052000db0ced 100644
@@ -116,7 +116,7 @@ extern struct pid *find_vpid(int nr);
 extern struct pid *find_get_pid(int nr);
 extern struct pid *find_ge_pid(int nr, struct pid_namespace *);
 
-extern struct pid *alloc_pid(struct pid_namespace *ns);
+extern struct pid *alloc_pid(struct pid_namespace *ns, pid_t set_tid);
 extern void free_pid(struct pid *pid);
 extern void disable_pid_allocation(struct pid_namespace *ns);
 
include/linux/sched/task.h +1 -0
index 0497091e40c1..4f2a80564332 100644
@@ -26,6 +26,7 @@ struct kernel_clone_args {
        unsigned long stack;
        unsigned long stack_size;
        unsigned long tls;
+       pid_t set_tid;
 };
 
 /*
include/uapi/linux/sched.h +1 -0
index b3105ac1381a..e1ce103a2c47 100644
@@ -45,6 +45,7 @@ struct clone_args {
        __aligned_u64 stack;
        __aligned_u64 stack_size;
        __aligned_u64 tls;
+       __aligned_u64 set_tid;
 };
 
 /*
kernel/fork.c +12 -2
index 2852d0e76ea3..8317d408a8d6 100644
@@ -117,6 +117,11 @@
  */
 #define MAX_THREADS FUTEX_TID_MASK
 
+/*
+ * For different sizes of struct clone_args
+ */
+#define CLONE3_ARGS_SIZE_V0 64
+
 /*
  * Protected counters by write_lock_irq(&tasklist_lock)
  */
@@ -2031,7 +2036,7 @@ static __latent_entropy struct task_struct *copy_process(
        stackleak_task_init(p);
 
        if (pid != &init_struct_pid) {
-               pid = alloc_pid(p->nsproxy->pid_ns_for_children);
+               pid = alloc_pid(p->nsproxy->pid_ns_for_children, args->set_tid);
                if (IS_ERR(pid)) {
                        retval = PTR_ERR(pid);
                        goto bad_fork_cleanup_thread;
@@ -2535,9 +2540,13 @@ noinline static int copy_clone_args_from_user(struct kernel_clone_args *kargs,
        if (unlikely(size > PAGE_SIZE))
                return -E2BIG;
 
-       if (unlikely(size < sizeof(struct clone_args)))
+       if (unlikely(size < CLONE3_ARGS_SIZE_V0))
                return -EINVAL;
 
+       if (size < sizeof(struct clone_args))
+               memset((void *)&args + size, 0,
+                               sizeof(struct clone_args) - size);
+
        if (unlikely(!access_ok(uargs, size)))
                return -EFAULT;
 
@@ -2571,6 +2580,7 @@ noinline static int copy_clone_args_from_user(struct kernel_clone_args *kargs,
                .stack          = args.stack,
                .stack_size     = args.stack_size,
                .tls            = args.tls,
+               .set_tid        = args.set_tid,
        };
 
        return 0;
kernel/pid.c +1 -1
index 0a9f2e437217..5cdab73b9094 100644
@@ -157,7 +157,7 @@ void free_pid(struct pid *pid)
        call_rcu(&pid->rcu, delayed_put_pid);
 }
 
-struct pid *alloc_pid(struct pid_namespace *ns)
+struct pid *alloc_pid(struct pid_namespace *ns, int set_tid)
 {
        struct pid *pid;
        enum pid_type type;

kernel/pid.c struct pid *alloc_pid(struct pid_namespace *ns) · line 174

Quoted 9 lines
kernel/pid.c +6 -0
@@ -166,6 +166,16 @@ struct pid *alloc_pid(struct pid_namespace *ns)
        struct upid *upid;
        int retval = -ENOMEM;
 
+       if (set_tid) {
+               if (set_tid < 0 || set_tid >= pid_max)
+                       return ERR_PTR(-EINVAL);
+               /* Also fail if a PID != 1 is requested and no PID 1 exists */
+               if (set_tid != 1 && !ns->child_reaper)
+                       return ERR_PTR(-EINVAL);

What will happen if I will create a new pid namespace (unsahre(CLONE_NEWPID)) and then call clone3 with tid which isn't equal to 1?

Will I get the init process with the 5 pid?...

Quoted 42 lines
kernel/pid.c +23 -7
+               if (!ns_capable(ns->user_ns, CAP_SYS_ADMIN))
+                       return ERR_PTR(-EPERM);
+       }
+
        pid = kmem_cache_alloc(ns->pid_cachep, GFP_KERNEL);
        if (!pid)
                return ERR_PTR(retval);
@@ -186,12 +196,25 @@ struct pid *alloc_pid(struct pid_namespace *ns)
                if (idr_get_cursor(&tmp->idr) > RESERVED_PIDS)
                        pid_min = RESERVED_PIDS;
 
-               /*
-                * Store a null pointer so find_pid_ns does not find
-                * a partially initialized PID (see below).
-                */
-               nr = idr_alloc_cyclic(&tmp->idr, NULL, pid_min,
-                                     pid_max, GFP_ATOMIC);
+               if (set_tid) {
+                       nr = idr_alloc(&tmp->idr, NULL, set_tid,
+                                       set_tid + 1, GFP_ATOMIC);
+                       /*
+                        * If ENOSPC is returned it means that the PID is
+                        * alreay in use. Return EEXIST in that case.
+                        */
+                       if (nr == -ENOSPC)
+                               nr = -EEXIST;
+                       /* Only use set_tid for one PID namespace. */
+                       set_tid = 0;
+               } else {
+                       /*
+                        * Store a null pointer so find_pid_ns does not find
+                        * a partially initialized PID (see below).
+                        */
+                       nr = idr_alloc_cyclic(&tmp->idr, NULL, pid_min,
+                                             pid_max, GFP_ATOMIC);
+               }
                spin_unlock_irq(&pidmap_lock);
                idr_preload_end();
 
--

2.21.0

#5 lore raw
Quoted 166 lines · Andrei Vagin
Quoted 151 lines · Adrian Reber

The main motivation to add set_tid to clone3() is CRIU.

To restore a process with the same PID/TID CRIU currently uses /proc/sys/kernel/ns_last_pid. It writes the desired (PID - 1) to ns_last_pid and then (quickly) does a clone(). This works most of the time, but it is racy. It is also slow as it requires multiple syscalls.

Extending clone3() to support set_tid makes it possible restore a process using CRIU without accessing /proc/sys/kernel/ns_last_pid and race free (as long as the desired PID/TID is available).

This clone3() extension places the same restrictions (CAP_SYS_ADMIN) on clone3() with set_tid as they are currently in place for ns_last_pid.

---
v2:
- Removed (size < sizeof(struct clone_args)) as discussed with Christian and Dmitry
- Added comment to ((set_tid != 1) && idr_get_cursor() <= 1) (Oleg)
- Use idr_alloc() instead of idr_alloc_cyclic() (Oleg)

v3:
- Return EEXIST if PID is already in use (Christian)
- Drop CLONE_SET_TID (Christian and Oleg)
- Use idr_is_empty() instead of idr_get_cursor() (Oleg)
- Handle different struct clone_args sizes (Dmitry)

v4:
- Rework struct size check with defines (Christian)
- Reduce number of set_tid checks (Oleg)
- Less parentheses and more robust code (Oleg)
- Do ns_capable() on correct user_ns (Oleg, Christian)

v5:
- make set_tid checks earlier in alloc_pid() (Christian)
- remove unnecessary comment and struct size check (Christian)

v6:
- remove CLONE_SET_TID from description (Christian)
- add clone3() tests for different clone_args sizes (Christian)
- move more set_tid checks to alloc_pid() (Oleg)
- make all set_tid checks lockless (Oleg)

---
 include/linux/pid.h        |  2 +-
 include/linux/sched/task.h |  1 +
 include/uapi/linux/sched.h |  1 +
 kernel/fork.c              | 14 ++++++++++++--
 kernel/pid.c               | 37 ++++++++++++++++++++++++++++++-------
 5 files changed, 45 insertions(+), 10 deletions(-)
include/linux/pid.h +1 -1
index 2a83e434db9d..052000db0ced 100644
@@ -116,7 +116,7 @@ extern struct pid *find_vpid(int nr);
 extern struct pid *find_get_pid(int nr);
 extern struct pid *find_ge_pid(int nr, struct pid_namespace *);
 
-extern struct pid *alloc_pid(struct pid_namespace *ns);
+extern struct pid *alloc_pid(struct pid_namespace *ns, pid_t set_tid);
 extern void free_pid(struct pid *pid);
 extern void disable_pid_allocation(struct pid_namespace *ns);
 
include/linux/sched/task.h +1 -0
index 0497091e40c1..4f2a80564332 100644
@@ -26,6 +26,7 @@ struct kernel_clone_args {
        unsigned long stack;
        unsigned long stack_size;
        unsigned long tls;
+       pid_t set_tid;
 };
 
 /*
include/uapi/linux/sched.h +1 -0
index b3105ac1381a..e1ce103a2c47 100644
@@ -45,6 +45,7 @@ struct clone_args {
        __aligned_u64 stack;
        __aligned_u64 stack_size;
        __aligned_u64 tls;
+       __aligned_u64 set_tid;
 };
 
 /*
kernel/fork.c +12 -2
index 2852d0e76ea3..8317d408a8d6 100644
@@ -117,6 +117,11 @@
  */
 #define MAX_THREADS FUTEX_TID_MASK
 
+/*
+ * For different sizes of struct clone_args
+ */
+#define CLONE3_ARGS_SIZE_V0 64
+
 /*
  * Protected counters by write_lock_irq(&tasklist_lock)
  */
@@ -2031,7 +2036,7 @@ static __latent_entropy struct task_struct *copy_process(
        stackleak_task_init(p);
 
        if (pid != &init_struct_pid) {
-               pid = alloc_pid(p->nsproxy->pid_ns_for_children);
+               pid = alloc_pid(p->nsproxy->pid_ns_for_children, args->set_tid);
                if (IS_ERR(pid)) {
                        retval = PTR_ERR(pid);
                        goto bad_fork_cleanup_thread;
@@ -2535,9 +2540,13 @@ noinline static int copy_clone_args_from_user(struct kernel_clone_args *kargs,
        if (unlikely(size > PAGE_SIZE))
                return -E2BIG;
 
-       if (unlikely(size < sizeof(struct clone_args)))
+       if (unlikely(size < CLONE3_ARGS_SIZE_V0))
                return -EINVAL;
 
+       if (size < sizeof(struct clone_args))
+               memset((void *)&args + size, 0,
+                               sizeof(struct clone_args) - size);
+
        if (unlikely(!access_ok(uargs, size)))
                return -EFAULT;
 
@@ -2571,6 +2580,7 @@ noinline static int copy_clone_args_from_user(struct kernel_clone_args *kargs,
                .stack          = args.stack,
                .stack_size     = args.stack_size,
                .tls            = args.tls,
+               .set_tid        = args.set_tid,
        };
 
        return 0;
kernel/pid.c +1 -1
index 0a9f2e437217..5cdab73b9094 100644
@@ -157,7 +157,7 @@ void free_pid(struct pid *pid)
        call_rcu(&pid->rcu, delayed_put_pid);
 }
 
-struct pid *alloc_pid(struct pid_namespace *ns)
+struct pid *alloc_pid(struct pid_namespace *ns, int set_tid)
 {
        struct pid *pid;
        enum pid_type type;

kernel/pid.c struct pid *alloc_pid(struct pid_namespace *ns) · line 174

Quoted 9 lines
kernel/pid.c +6 -0
@@ -166,6 +166,16 @@ struct pid *alloc_pid(struct pid_namespace *ns)
        struct upid *upid;
        int retval = -ENOMEM;
 
+       if (set_tid) {
+               if (set_tid < 0 || set_tid >= pid_max)
+                       return ERR_PTR(-EINVAL);
+               /* Also fail if a PID != 1 is requested and no PID 1 exists */
+               if (set_tid != 1 && !ns->child_reaper)
+                       return ERR_PTR(-EINVAL);

What will happen if I will create a new pid namespace (unsahre(CLONE_NEWPID)) and then call clone3 with tid which isn't equal to 1?

Quoted 1 line

Will I get the init process with the 5 pid?...

No, you will get -EINVAL. In the selftest exactly this is tested. A PID 1 is still required even with this patch.

Quoted 42 lines
Quoted 42 lines
kernel/pid.c +23 -7
+               if (!ns_capable(ns->user_ns, CAP_SYS_ADMIN))
+                       return ERR_PTR(-EPERM);
+       }
+
        pid = kmem_cache_alloc(ns->pid_cachep, GFP_KERNEL);
        if (!pid)
                return ERR_PTR(retval);
@@ -186,12 +196,25 @@ struct pid *alloc_pid(struct pid_namespace *ns)
                if (idr_get_cursor(&tmp->idr) > RESERVED_PIDS)
                        pid_min = RESERVED_PIDS;
 
-               /*
-                * Store a null pointer so find_pid_ns does not find
-                * a partially initialized PID (see below).
-                */
-               nr = idr_alloc_cyclic(&tmp->idr, NULL, pid_min,
-                                     pid_max, GFP_ATOMIC);
+               if (set_tid) {
+                       nr = idr_alloc(&tmp->idr, NULL, set_tid,
+                                       set_tid + 1, GFP_ATOMIC);
+                       /*
+                        * If ENOSPC is returned it means that the PID is
+                        * alreay in use. Return EEXIST in that case.
+                        */
+                       if (nr == -ENOSPC)
+                               nr = -EEXIST;
+                       /* Only use set_tid for one PID namespace. */
+                       set_tid = 0;
+               } else {
+                       /*
+                        * Store a null pointer so find_pid_ns does not find
+                        * a partially initialized PID (see below).
+                        */
+                       nr = idr_alloc_cyclic(&tmp->idr, NULL, pid_min,
+                                             pid_max, GFP_ATOMIC);
+               }
                spin_unlock_irq(&pidmap_lock);
                idr_preload_end();
 
--

2.21.0

#6 lore raw
Quoted 167 lines · Adrian Reber
Quoted 166 lines · Andrei Vagin
Quoted 151 lines · Adrian Reber

The main motivation to add set_tid to clone3() is CRIU.

To restore a process with the same PID/TID CRIU currently uses /proc/sys/kernel/ns_last_pid. It writes the desired (PID - 1) to ns_last_pid and then (quickly) does a clone(). This works most of the time, but it is racy. It is also slow as it requires multiple syscalls.

Extending clone3() to support set_tid makes it possible restore a process using CRIU without accessing /proc/sys/kernel/ns_last_pid and race free (as long as the desired PID/TID is available).

This clone3() extension places the same restrictions (CAP_SYS_ADMIN) on clone3() with set_tid as they are currently in place for ns_last_pid.

---
v2:
- Removed (size < sizeof(struct clone_args)) as discussed with Christian and Dmitry
- Added comment to ((set_tid != 1) && idr_get_cursor() <= 1) (Oleg)
- Use idr_alloc() instead of idr_alloc_cyclic() (Oleg)

v3:
- Return EEXIST if PID is already in use (Christian)
- Drop CLONE_SET_TID (Christian and Oleg)
- Use idr_is_empty() instead of idr_get_cursor() (Oleg)
- Handle different struct clone_args sizes (Dmitry)

v4:
- Rework struct size check with defines (Christian)
- Reduce number of set_tid checks (Oleg)
- Less parentheses and more robust code (Oleg)
- Do ns_capable() on correct user_ns (Oleg, Christian)

v5:
- make set_tid checks earlier in alloc_pid() (Christian)
- remove unnecessary comment and struct size check (Christian)

v6:
- remove CLONE_SET_TID from description (Christian)
- add clone3() tests for different clone_args sizes (Christian)
- move more set_tid checks to alloc_pid() (Oleg)
- make all set_tid checks lockless (Oleg)

---
 include/linux/pid.h        |  2 +-
 include/linux/sched/task.h |  1 +
 include/uapi/linux/sched.h |  1 +
 kernel/fork.c              | 14 ++++++++++++--
 kernel/pid.c               | 37 ++++++++++++++++++++++++++++++-------
 5 files changed, 45 insertions(+), 10 deletions(-)
include/linux/pid.h +1 -1
index 2a83e434db9d..052000db0ced 100644
@@ -116,7 +116,7 @@ extern struct pid *find_vpid(int nr);
 extern struct pid *find_get_pid(int nr);
 extern struct pid *find_ge_pid(int nr, struct pid_namespace *);
 
-extern struct pid *alloc_pid(struct pid_namespace *ns);
+extern struct pid *alloc_pid(struct pid_namespace *ns, pid_t set_tid);
 extern void free_pid(struct pid *pid);
 extern void disable_pid_allocation(struct pid_namespace *ns);
 
include/linux/sched/task.h +1 -0
index 0497091e40c1..4f2a80564332 100644
@@ -26,6 +26,7 @@ struct kernel_clone_args {
        unsigned long stack;
        unsigned long stack_size;
        unsigned long tls;
+       pid_t set_tid;
 };
 
 /*
include/uapi/linux/sched.h +1 -0
index b3105ac1381a..e1ce103a2c47 100644
@@ -45,6 +45,7 @@ struct clone_args {
        __aligned_u64 stack;
        __aligned_u64 stack_size;
        __aligned_u64 tls;
+       __aligned_u64 set_tid;
 };
 
 /*
kernel/fork.c +12 -2
index 2852d0e76ea3..8317d408a8d6 100644
@@ -117,6 +117,11 @@
  */
 #define MAX_THREADS FUTEX_TID_MASK
 
+/*
+ * For different sizes of struct clone_args
+ */
+#define CLONE3_ARGS_SIZE_V0 64
+
 /*
  * Protected counters by write_lock_irq(&tasklist_lock)
  */
@@ -2031,7 +2036,7 @@ static __latent_entropy struct task_struct *copy_process(
        stackleak_task_init(p);
 
        if (pid != &init_struct_pid) {
-               pid = alloc_pid(p->nsproxy->pid_ns_for_children);
+               pid = alloc_pid(p->nsproxy->pid_ns_for_children, args->set_tid);
                if (IS_ERR(pid)) {
                        retval = PTR_ERR(pid);
                        goto bad_fork_cleanup_thread;
@@ -2535,9 +2540,13 @@ noinline static int copy_clone_args_from_user(struct kernel_clone_args *kargs,
        if (unlikely(size > PAGE_SIZE))
                return -E2BIG;
 
-       if (unlikely(size < sizeof(struct clone_args)))
+       if (unlikely(size < CLONE3_ARGS_SIZE_V0))
                return -EINVAL;
 
+       if (size < sizeof(struct clone_args))
+               memset((void *)&args + size, 0,
+                               sizeof(struct clone_args) - size);
+
        if (unlikely(!access_ok(uargs, size)))
                return -EFAULT;
 
@@ -2571,6 +2580,7 @@ noinline static int copy_clone_args_from_user(struct kernel_clone_args *kargs,
                .stack          = args.stack,
                .stack_size     = args.stack_size,
                .tls            = args.tls,
+               .set_tid        = args.set_tid,
        };
 
        return 0;
kernel/pid.c +1 -1
index 0a9f2e437217..5cdab73b9094 100644
@@ -157,7 +157,7 @@ void free_pid(struct pid *pid)
        call_rcu(&pid->rcu, delayed_put_pid);
 }
 
-struct pid *alloc_pid(struct pid_namespace *ns)
+struct pid *alloc_pid(struct pid_namespace *ns, int set_tid)
 {
        struct pid *pid;
        enum pid_type type;

kernel/pid.c struct pid *alloc_pid(struct pid_namespace *ns) · line 174

Quoted 9 lines
kernel/pid.c +6 -0
@@ -166,6 +166,16 @@ struct pid *alloc_pid(struct pid_namespace *ns)
        struct upid *upid;
        int retval = -ENOMEM;
 
+       if (set_tid) {
+               if (set_tid < 0 || set_tid >= pid_max)
+                       return ERR_PTR(-EINVAL);
+               /* Also fail if a PID != 1 is requested and no PID 1 exists */
+               if (set_tid != 1 && !ns->child_reaper)
+                       return ERR_PTR(-EINVAL);

What will happen if I will create a new pid namespace (unsahre(CLONE_NEWPID)) and then call clone3 with tid which isn't equal to 1?

Quoted 4 lines
Quoted 1 line

Will I get the init process with the 5 pid?...

No, you will get -EINVAL. In the selftest exactly this is tested. A PID 1 is still required even with this patch.

Correct. unshare(CLONE_PIDNS) leaves ->child_reaper unset. ->child_reaper is set when pid 1 is created in the pid namespace. So you that check catches the scenarios afaict. Adrians test should also have caught that if it were possible. I hope he's running them. :)

#7 lore raw
In reply to Adrian Reber
Quoted 168 lines · Adrian Reber
Quoted 166 lines · Andrei Vagin
Quoted 151 lines · Adrian Reber

The main motivation to add set_tid to clone3() is CRIU.

To restore a process with the same PID/TID CRIU currently uses /proc/sys/kernel/ns_last_pid. It writes the desired (PID - 1) to ns_last_pid and then (quickly) does a clone(). This works most of the time, but it is racy. It is also slow as it requires multiple syscalls.

Extending clone3() to support set_tid makes it possible restore a process using CRIU without accessing /proc/sys/kernel/ns_last_pid and race free (as long as the desired PID/TID is available).

This clone3() extension places the same restrictions (CAP_SYS_ADMIN) on clone3() with set_tid as they are currently in place for ns_last_pid.

---
v2:
- Removed (size < sizeof(struct clone_args)) as discussed with Christian and Dmitry
- Added comment to ((set_tid != 1) && idr_get_cursor() <= 1) (Oleg)
- Use idr_alloc() instead of idr_alloc_cyclic() (Oleg)

v3:
- Return EEXIST if PID is already in use (Christian)
- Drop CLONE_SET_TID (Christian and Oleg)
- Use idr_is_empty() instead of idr_get_cursor() (Oleg)
- Handle different struct clone_args sizes (Dmitry)

v4:
- Rework struct size check with defines (Christian)
- Reduce number of set_tid checks (Oleg)
- Less parentheses and more robust code (Oleg)
- Do ns_capable() on correct user_ns (Oleg, Christian)

v5:
- make set_tid checks earlier in alloc_pid() (Christian)
- remove unnecessary comment and struct size check (Christian)

v6:
- remove CLONE_SET_TID from description (Christian)
- add clone3() tests for different clone_args sizes (Christian)
- move more set_tid checks to alloc_pid() (Oleg)
- make all set_tid checks lockless (Oleg)

---
 include/linux/pid.h        |  2 +-
 include/linux/sched/task.h |  1 +
 include/uapi/linux/sched.h |  1 +
 kernel/fork.c              | 14 ++++++++++++--
 kernel/pid.c               | 37 ++++++++++++++++++++++++++++++-------
 5 files changed, 45 insertions(+), 10 deletions(-)
include/linux/pid.h +1 -1
index 2a83e434db9d..052000db0ced 100644
@@ -116,7 +116,7 @@ extern struct pid *find_vpid(int nr);
 extern struct pid *find_get_pid(int nr);
 extern struct pid *find_ge_pid(int nr, struct pid_namespace *);
 
-extern struct pid *alloc_pid(struct pid_namespace *ns);
+extern struct pid *alloc_pid(struct pid_namespace *ns, pid_t set_tid);
 extern void free_pid(struct pid *pid);
 extern void disable_pid_allocation(struct pid_namespace *ns);
 
include/linux/sched/task.h +1 -0
index 0497091e40c1..4f2a80564332 100644
@@ -26,6 +26,7 @@ struct kernel_clone_args {
        unsigned long stack;
        unsigned long stack_size;
        unsigned long tls;
+       pid_t set_tid;
 };
 
 /*
include/uapi/linux/sched.h +1 -0
index b3105ac1381a..e1ce103a2c47 100644
@@ -45,6 +45,7 @@ struct clone_args {
        __aligned_u64 stack;
        __aligned_u64 stack_size;
        __aligned_u64 tls;
+       __aligned_u64 set_tid;
 };
 
 /*
kernel/fork.c +12 -2
index 2852d0e76ea3..8317d408a8d6 100644
@@ -117,6 +117,11 @@
  */
 #define MAX_THREADS FUTEX_TID_MASK
 
+/*
+ * For different sizes of struct clone_args
+ */
+#define CLONE3_ARGS_SIZE_V0 64
+
 /*
  * Protected counters by write_lock_irq(&tasklist_lock)
  */
@@ -2031,7 +2036,7 @@ static __latent_entropy struct task_struct *copy_process(
        stackleak_task_init(p);
 
        if (pid != &init_struct_pid) {
-               pid = alloc_pid(p->nsproxy->pid_ns_for_children);
+               pid = alloc_pid(p->nsproxy->pid_ns_for_children, args->set_tid);
                if (IS_ERR(pid)) {
                        retval = PTR_ERR(pid);
                        goto bad_fork_cleanup_thread;
@@ -2535,9 +2540,13 @@ noinline static int copy_clone_args_from_user(struct kernel_clone_args *kargs,
        if (unlikely(size > PAGE_SIZE))
                return -E2BIG;
 
-       if (unlikely(size < sizeof(struct clone_args)))
+       if (unlikely(size < CLONE3_ARGS_SIZE_V0))
                return -EINVAL;
 
+       if (size < sizeof(struct clone_args))
+               memset((void *)&args + size, 0,
+                               sizeof(struct clone_args) - size);
+
        if (unlikely(!access_ok(uargs, size)))
                return -EFAULT;
 
@@ -2571,6 +2580,7 @@ noinline static int copy_clone_args_from_user(struct kernel_clone_args *kargs,
                .stack          = args.stack,
                .stack_size     = args.stack_size,
                .tls            = args.tls,
+               .set_tid        = args.set_tid,
        };
 
        return 0;
kernel/pid.c +1 -1
index 0a9f2e437217..5cdab73b9094 100644
@@ -157,7 +157,7 @@ void free_pid(struct pid *pid)
        call_rcu(&pid->rcu, delayed_put_pid);
 }
 
-struct pid *alloc_pid(struct pid_namespace *ns)
+struct pid *alloc_pid(struct pid_namespace *ns, int set_tid)
 {
        struct pid *pid;
        enum pid_type type;

kernel/pid.c struct pid *alloc_pid(struct pid_namespace *ns) · line 174

Quoted 9 lines
kernel/pid.c +6 -0
@@ -166,6 +166,16 @@ struct pid *alloc_pid(struct pid_namespace *ns)
        struct upid *upid;
        int retval = -ENOMEM;
 
+       if (set_tid) {
+               if (set_tid < 0 || set_tid >= pid_max)
+                       return ERR_PTR(-EINVAL);
+               /* Also fail if a PID != 1 is requested and no PID 1 exists */
+               if (set_tid != 1 && !ns->child_reaper)
+                       return ERR_PTR(-EINVAL);

What will happen if I will create a new pid namespace (unsahre(CLONE_NEWPID)) and then call clone3 with tid which isn't equal to 1?

Quoted 4 lines
Quoted 1 line

Will I get the init process with the 5 pid?...

No, you will get -EINVAL. In the selftest exactly this is tested. A PID 1 is still required even with this patch.

Good. Thanks. I had to read the test.

Quoted 43 lines
Quoted 43 lines
Quoted 43 lines
kernel/pid.c +23 -7
+               if (!ns_capable(ns->user_ns, CAP_SYS_ADMIN))
+                       return ERR_PTR(-EPERM);
+       }
+
        pid = kmem_cache_alloc(ns->pid_cachep, GFP_KERNEL);
        if (!pid)
                return ERR_PTR(retval);
@@ -186,12 +196,25 @@ struct pid *alloc_pid(struct pid_namespace *ns)
                if (idr_get_cursor(&tmp->idr) > RESERVED_PIDS)
                        pid_min = RESERVED_PIDS;
 
-               /*
-                * Store a null pointer so find_pid_ns does not find
-                * a partially initialized PID (see below).
-                */
-               nr = idr_alloc_cyclic(&tmp->idr, NULL, pid_min,
-                                     pid_max, GFP_ATOMIC);
+               if (set_tid) {
+                       nr = idr_alloc(&tmp->idr, NULL, set_tid,
+                                       set_tid + 1, GFP_ATOMIC);
+                       /*
+                        * If ENOSPC is returned it means that the PID is
+                        * alreay in use. Return EEXIST in that case.
+                        */
+                       if (nr == -ENOSPC)
+                               nr = -EEXIST;
+                       /* Only use set_tid for one PID namespace. */
+                       set_tid = 0;
+               } else {
+                       /*
+                        * Store a null pointer so find_pid_ns does not find
+                        * a partially initialized PID (see below).
+                        */
+                       nr = idr_alloc_cyclic(&tmp->idr, NULL, pid_min,
+                                             pid_max, GFP_ATOMIC);
+               }
                spin_unlock_irq(&pidmap_lock);
                idr_preload_end();
 
--

2.21.0

Reviewed-by on v6 1/2
#8 lore raw
In reply to Adrian Reber
Quoted 7 lines · Adrian Reber

The main motivation to add set_tid to clone3() is CRIU.

To restore a process with the same PID/TID CRIU currently uses /proc/sys/kernel/ns_last_pid. It writes the desired (PID - 1) to ns_last_pid and then (quickly) does a clone(). This works most of the time, but it is racy. It is also slow as it requires multiple syscalls.

Quoted 8 lines

Extending clone3() to support set_tid makes it possible restore a process using CRIU without accessing /proc/sys/kernel/ns_last_pid and race free (as long as the desired PID/TID is available).

This clone3() extension places the same restrictions (CAP_SYS_ADMIN) on clone3() with set_tid as they are currently in place for ns_last_pid.

I'm fine with this version. Oleg, how do you feel about it now? :)

Quoted 186 lines

---
v2:
- Removed (size < sizeof(struct clone_args)) as discussed with Christian and Dmitry
- Added comment to ((set_tid != 1) && idr_get_cursor() <= 1) (Oleg)
- Use idr_alloc() instead of idr_alloc_cyclic() (Oleg)

v3:
- Return EEXIST if PID is already in use (Christian)
- Drop CLONE_SET_TID (Christian and Oleg)
- Use idr_is_empty() instead of idr_get_cursor() (Oleg)
- Handle different struct clone_args sizes (Dmitry)

v4:
- Rework struct size check with defines (Christian)
- Reduce number of set_tid checks (Oleg)
- Less parentheses and more robust code (Oleg)
- Do ns_capable() on correct user_ns (Oleg, Christian)

v5:
- make set_tid checks earlier in alloc_pid() (Christian)
- remove unnecessary comment and struct size check (Christian)

v6:
- remove CLONE_SET_TID from description (Christian)
- add clone3() tests for different clone_args sizes (Christian)
- move more set_tid checks to alloc_pid() (Oleg)
- make all set_tid checks lockless (Oleg)

---
 include/linux/pid.h        |  2 +-
 include/linux/sched/task.h |  1 +
 include/uapi/linux/sched.h |  1 +
 kernel/fork.c              | 14 ++++++++++++--
 kernel/pid.c               | 37 ++++++++++++++++++++++++++++++-------
 5 files changed, 45 insertions(+), 10 deletions(-)
include/linux/pid.h +1 -1
index 2a83e434db9d..052000db0ced 100644
@@ -116,7 +116,7 @@ extern struct pid *find_vpid(int nr);
 extern struct pid *find_get_pid(int nr);
 extern struct pid *find_ge_pid(int nr, struct pid_namespace *);
 
-extern struct pid *alloc_pid(struct pid_namespace *ns);
+extern struct pid *alloc_pid(struct pid_namespace *ns, pid_t set_tid);
 extern void free_pid(struct pid *pid);
 extern void disable_pid_allocation(struct pid_namespace *ns);
 
include/linux/sched/task.h +1 -0
index 0497091e40c1..4f2a80564332 100644
@@ -26,6 +26,7 @@ struct kernel_clone_args {
 	unsigned long stack;
 	unsigned long stack_size;
 	unsigned long tls;
+	pid_t set_tid;
 };
 
 /*
include/uapi/linux/sched.h +1 -0
index b3105ac1381a..e1ce103a2c47 100644
@@ -45,6 +45,7 @@ struct clone_args {
 	__aligned_u64 stack;
 	__aligned_u64 stack_size;
 	__aligned_u64 tls;
+	__aligned_u64 set_tid;
 };
 
 /*
kernel/fork.c +12 -2
index 2852d0e76ea3..8317d408a8d6 100644
@@ -117,6 +117,11 @@
  */
 #define MAX_THREADS FUTEX_TID_MASK
 
+/*
+ * For different sizes of struct clone_args
+ */
+#define CLONE3_ARGS_SIZE_V0 64
+
 /*
  * Protected counters by write_lock_irq(&tasklist_lock)
  */
@@ -2031,7 +2036,7 @@ static __latent_entropy struct task_struct *copy_process(
 	stackleak_task_init(p);
 
 	if (pid != &init_struct_pid) {
-		pid = alloc_pid(p->nsproxy->pid_ns_for_children);
+		pid = alloc_pid(p->nsproxy->pid_ns_for_children, args->set_tid);
 		if (IS_ERR(pid)) {
 			retval = PTR_ERR(pid);
 			goto bad_fork_cleanup_thread;
@@ -2535,9 +2540,13 @@ noinline static int copy_clone_args_from_user(struct kernel_clone_args *kargs,
 	if (unlikely(size > PAGE_SIZE))
 		return -E2BIG;
 
-	if (unlikely(size < sizeof(struct clone_args)))
+	if (unlikely(size < CLONE3_ARGS_SIZE_V0))
 		return -EINVAL;
 
+	if (size < sizeof(struct clone_args))
+		memset((void *)&args + size, 0,
+				sizeof(struct clone_args) - size);
+
 	if (unlikely(!access_ok(uargs, size)))
 		return -EFAULT;
 
@@ -2571,6 +2580,7 @@ noinline static int copy_clone_args_from_user(struct kernel_clone_args *kargs,
 		.stack		= args.stack,
 		.stack_size	= args.stack_size,
 		.tls		= args.tls,
+		.set_tid	= args.set_tid,
 	};
 
 	return 0;
kernel/pid.c +30 -7
index 0a9f2e437217..5cdab73b9094 100644
@@ -157,7 +157,7 @@ void free_pid(struct pid *pid)
 	call_rcu(&pid->rcu, delayed_put_pid);
 }
 
-struct pid *alloc_pid(struct pid_namespace *ns)
+struct pid *alloc_pid(struct pid_namespace *ns, int set_tid)
 {
 	struct pid *pid;
 	enum pid_type type;
@@ -166,6 +166,16 @@ struct pid *alloc_pid(struct pid_namespace *ns)
 	struct upid *upid;
 	int retval = -ENOMEM;
 
+	if (set_tid) {
+		if (set_tid < 0 || set_tid >= pid_max)
+			return ERR_PTR(-EINVAL);
+		/* Also fail if a PID != 1 is requested and no PID 1 exists */
+		if (set_tid != 1 && !ns->child_reaper)
+			return ERR_PTR(-EINVAL);
+		if (!ns_capable(ns->user_ns, CAP_SYS_ADMIN))
+			return ERR_PTR(-EPERM);
+	}
+
 	pid = kmem_cache_alloc(ns->pid_cachep, GFP_KERNEL);
 	if (!pid)
 		return ERR_PTR(retval);
@@ -186,12 +196,25 @@ struct pid *alloc_pid(struct pid_namespace *ns)
 		if (idr_get_cursor(&tmp->idr) > RESERVED_PIDS)
 			pid_min = RESERVED_PIDS;
 
-		/*
-		 * Store a null pointer so find_pid_ns does not find
-		 * a partially initialized PID (see below).
-		 */
-		nr = idr_alloc_cyclic(&tmp->idr, NULL, pid_min,
-				      pid_max, GFP_ATOMIC);
+		if (set_tid) {
+			nr = idr_alloc(&tmp->idr, NULL, set_tid,
+					set_tid + 1, GFP_ATOMIC);
+			/*
+			 * If ENOSPC is returned it means that the PID is
+			 * alreay in use. Return EEXIST in that case.
+			 */
+			if (nr == -ENOSPC)
+				nr = -EEXIST;
+			/* Only use set_tid for one PID namespace. */
+			set_tid = 0;
+		} else {
+			/*
+			 * Store a null pointer so find_pid_ns does not find
+			 * a partially initialized PID (see below).
+			 */
+			nr = idr_alloc_cyclic(&tmp->idr, NULL, pid_min,
+					      pid_max, GFP_ATOMIC);
+		}
 		spin_unlock_irq(&pidmap_lock);
 		idr_preload_end();
 
Signature · 1 line
2.21.0
Reviewed-by on v6 1/2
#9 lore raw
In reply to Adrian Reber
Quoted 8 lines · Adrian Reber

The main motivation to add set_tid to clone3() is CRIU.

To restore a process with the same PID/TID CRIU currently uses /proc/sys/kernel/ns_last_pid. It writes the desired (PID - 1) to ns_last_pid and then (quickly) does a clone(). This works most of the time, but it is racy. It is also slow as it requires multiple syscalls.

Quoted 8 lines

Extending clone3() to support set_tid makes it possible restore a process using CRIU without accessing /proc/sys/kernel/ns_last_pid and race free (as long as the desired PID/TID is available).

This clone3() extension places the same restrictions (CAP_SYS_ADMIN) on clone3() with set_tid as they are currently in place for ns_last_pid.

Acked-by Pavel Emelyanov on v6 1/2
#10 lore raw
In reply to Adrian Reber
Quoted 7 lines · Adrian Reber

The main motivation to add set_tid to clone3() is CRIU.

To restore a process with the same PID/TID CRIU currently uses /proc/sys/kernel/ns_last_pid. It writes the desired (PID - 1) to ns_last_pid and then (quickly) does a clone(). This works most of the time, but it is racy. It is also slow as it requires multiple syscalls.

Quoted 8 lines

Extending clone3() to support set_tid makes it possible restore a process using CRIU without accessing /proc/sys/kernel/ns_last_pid and race free (as long as the desired PID/TID is available).

This clone3() extension places the same restrictions (CAP_SYS_ADMIN) on clone3() with set_tid as they are currently in place for ns_last_pid.

Quoted 184 lines

---
v2:
- Removed (size < sizeof(struct clone_args)) as discussed with Christian and Dmitry
- Added comment to ((set_tid != 1) && idr_get_cursor() <= 1) (Oleg)
- Use idr_alloc() instead of idr_alloc_cyclic() (Oleg)

v3:
- Return EEXIST if PID is already in use (Christian)
- Drop CLONE_SET_TID (Christian and Oleg)
- Use idr_is_empty() instead of idr_get_cursor() (Oleg)
- Handle different struct clone_args sizes (Dmitry)

v4:
- Rework struct size check with defines (Christian)
- Reduce number of set_tid checks (Oleg)
- Less parentheses and more robust code (Oleg)
- Do ns_capable() on correct user_ns (Oleg, Christian)

v5:
- make set_tid checks earlier in alloc_pid() (Christian)
- remove unnecessary comment and struct size check (Christian)

v6:
- remove CLONE_SET_TID from description (Christian)
- add clone3() tests for different clone_args sizes (Christian)
- move more set_tid checks to alloc_pid() (Oleg)
- make all set_tid checks lockless (Oleg)

---
 include/linux/pid.h        |  2 +-
 include/linux/sched/task.h |  1 +
 include/uapi/linux/sched.h |  1 +
 kernel/fork.c              | 14 ++++++++++++--
 kernel/pid.c               | 37 ++++++++++++++++++++++++++++++-------
 5 files changed, 45 insertions(+), 10 deletions(-)
include/linux/pid.h +1 -1
index 2a83e434db9d..052000db0ced 100644
@@ -116,7 +116,7 @@ extern struct pid *find_vpid(int nr);
 extern struct pid *find_get_pid(int nr);
 extern struct pid *find_ge_pid(int nr, struct pid_namespace *);
 
-extern struct pid *alloc_pid(struct pid_namespace *ns);
+extern struct pid *alloc_pid(struct pid_namespace *ns, pid_t set_tid);
 extern void free_pid(struct pid *pid);
 extern void disable_pid_allocation(struct pid_namespace *ns);
 
include/linux/sched/task.h +1 -0
index 0497091e40c1..4f2a80564332 100644
@@ -26,6 +26,7 @@ struct kernel_clone_args {
 	unsigned long stack;
 	unsigned long stack_size;
 	unsigned long tls;
+	pid_t set_tid;
 };
 
 /*
include/uapi/linux/sched.h +1 -0
index b3105ac1381a..e1ce103a2c47 100644
@@ -45,6 +45,7 @@ struct clone_args {
 	__aligned_u64 stack;
 	__aligned_u64 stack_size;
 	__aligned_u64 tls;
+	__aligned_u64 set_tid;
 };
 
 /*
kernel/fork.c +12 -2
index 2852d0e76ea3..8317d408a8d6 100644
@@ -117,6 +117,11 @@
  */
 #define MAX_THREADS FUTEX_TID_MASK
 
+/*
+ * For different sizes of struct clone_args
+ */
+#define CLONE3_ARGS_SIZE_V0 64
+
 /*
  * Protected counters by write_lock_irq(&tasklist_lock)
  */
@@ -2031,7 +2036,7 @@ static __latent_entropy struct task_struct *copy_process(
 	stackleak_task_init(p);
 
 	if (pid != &init_struct_pid) {
-		pid = alloc_pid(p->nsproxy->pid_ns_for_children);
+		pid = alloc_pid(p->nsproxy->pid_ns_for_children, args->set_tid);
 		if (IS_ERR(pid)) {
 			retval = PTR_ERR(pid);
 			goto bad_fork_cleanup_thread;
@@ -2535,9 +2540,13 @@ noinline static int copy_clone_args_from_user(struct kernel_clone_args *kargs,
 	if (unlikely(size > PAGE_SIZE))
 		return -E2BIG;
 
-	if (unlikely(size < sizeof(struct clone_args)))
+	if (unlikely(size < CLONE3_ARGS_SIZE_V0))
 		return -EINVAL;
 
+	if (size < sizeof(struct clone_args))
+		memset((void *)&args + size, 0,
+				sizeof(struct clone_args) - size);
+
 	if (unlikely(!access_ok(uargs, size)))
 		return -EFAULT;
 
@@ -2571,6 +2580,7 @@ noinline static int copy_clone_args_from_user(struct kernel_clone_args *kargs,
 		.stack		= args.stack,
 		.stack_size	= args.stack_size,
 		.tls		= args.tls,
+		.set_tid	= args.set_tid,
 	};
 
 	return 0;
kernel/pid.c +30 -7
index 0a9f2e437217..5cdab73b9094 100644
@@ -157,7 +157,7 @@ void free_pid(struct pid *pid)
 	call_rcu(&pid->rcu, delayed_put_pid);
 }
 
-struct pid *alloc_pid(struct pid_namespace *ns)
+struct pid *alloc_pid(struct pid_namespace *ns, int set_tid)
 {
 	struct pid *pid;
 	enum pid_type type;
@@ -166,6 +166,16 @@ struct pid *alloc_pid(struct pid_namespace *ns)
 	struct upid *upid;
 	int retval = -ENOMEM;
 
+	if (set_tid) {
+		if (set_tid < 0 || set_tid >= pid_max)
+			return ERR_PTR(-EINVAL);
+		/* Also fail if a PID != 1 is requested and no PID 1 exists */
+		if (set_tid != 1 && !ns->child_reaper)
+			return ERR_PTR(-EINVAL);
+		if (!ns_capable(ns->user_ns, CAP_SYS_ADMIN))
+			return ERR_PTR(-EPERM);
+	}
+
 	pid = kmem_cache_alloc(ns->pid_cachep, GFP_KERNEL);
 	if (!pid)
 		return ERR_PTR(retval);
@@ -186,12 +196,25 @@ struct pid *alloc_pid(struct pid_namespace *ns)
 		if (idr_get_cursor(&tmp->idr) > RESERVED_PIDS)
 			pid_min = RESERVED_PIDS;
 
-		/*
-		 * Store a null pointer so find_pid_ns does not find
-		 * a partially initialized PID (see below).
-		 */
-		nr = idr_alloc_cyclic(&tmp->idr, NULL, pid_min,
-				      pid_max, GFP_ATOMIC);
+		if (set_tid) {
+			nr = idr_alloc(&tmp->idr, NULL, set_tid,
+					set_tid + 1, GFP_ATOMIC);
+			/*
+			 * If ENOSPC is returned it means that the PID is
+			 * alreay in use. Return EEXIST in that case.
+			 */
+			if (nr == -ENOSPC)
+				nr = -EEXIST;
+			/* Only use set_tid for one PID namespace. */
+			set_tid = 0;
+		} else {
+			/*
+			 * Store a null pointer so find_pid_ns does not find
+			 * a partially initialized PID (see below).
+			 */
+			nr = idr_alloc_cyclic(&tmp->idr, NULL, pid_min,
+					      pid_max, GFP_ATOMIC);
+		}
 		spin_unlock_irq(&pidmap_lock);
 		idr_preload_end();
 
#11 lore raw
Quoted 18 lines · Pavel Emelianov
Quoted 7 lines · Adrian Reber

The main motivation to add set_tid to clone3() is CRIU.

To restore a process with the same PID/TID CRIU currently uses /proc/sys/kernel/ns_last_pid. It writes the desired (PID - 1) to ns_last_pid and then (quickly) does a clone(). This works most of the time, but it is racy. It is also slow as it requires multiple syscalls.

Quoted 8 lines

Extending clone3() to support set_tid makes it possible restore a process using CRIU without accessing /proc/sys/kernel/ns_last_pid and race free (as long as the desired PID/TID is available).

This clone3() extension places the same restrictions (CAP_SYS_ADMIN) on clone3() with set_tid as they are currently in place for ns_last_pid.

Quoted 13 lines · Oleg Nesterov
Quoted 12 lines · Adrian Reber

The main motivation to add set_tid to clone3() is CRIU.

To restore a process with the same PID/TID CRIU currently uses /proc/sys/kernel/ns_last_pid. It writes the desired (PID - 1) to ns_last_pid and then (quickly) does a clone(). This works most of the time, but it is racy. It is also slow as it requires multiple syscalls.

Extending clone3() to support set_tid makes it possible restore a process using CRIU without accessing /proc/sys/kernel/ns_last_pid and race free (as long as the desired PID/TID is available).

Quoted 7 lines
Quoted 4 lines

This clone3() extension places the same restrictions (CAP_SYS_ADMIN) on clone3() with set_tid as they are currently in place for ns_last_pid.

Added-to:
https://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux.git/log/?h=pidfd

Merged-into:
https://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux.git/log/?h=for-next

Thanks!
Christian