Resguardo inicial

linux-7.2.3/tools/testing/selftests/namespaces/.gitignore

@@ -0,0 +1,12 @@
+nsid_test
+file_handle_test
+init_ino_test
+ns_active_ref_test
+listns_test
+listns_permissions_test
+listns_efault_test
+siocgskns_test
+cred_change_test
+stress_test
+listns_pagination_bug
+regression_pidfd_setns_test

linux-7.2.3/tools/testing/selftests/namespaces/Makefile

@@ -0,0 +1,29 @@
+# SPDX-License-Identifier: GPL-2.0-only
+CFLAGS += -Wall -O0 -g $(KHDR_INCLUDES) $(TOOLS_INCLUDES)
+LDLIBS += -lcap
+
+TEST_GEN_PROGS := nsid_test \
+		  file_handle_test \
+		  init_ino_test \
+		  ns_active_ref_test \
+		  listns_test \
+		  listns_permissions_test \
+		  listns_efault_test \
+		  siocgskns_test \
+		  cred_change_test \
+		  stress_test \
+		  listns_pagination_bug \
+		  regression_pidfd_setns_test
+
+include ../lib.mk
+
+$(OUTPUT)/ns_active_ref_test: ../filesystems/utils.c
+$(OUTPUT)/listns_test: ../filesystems/utils.c
+$(OUTPUT)/listns_permissions_test: ../filesystems/utils.c
+$(OUTPUT)/listns_efault_test: ../filesystems/utils.c
+$(OUTPUT)/siocgskns_test: ../filesystems/utils.c
+$(OUTPUT)/cred_change_test: ../filesystems/utils.c
+$(OUTPUT)/stress_test: ../filesystems/utils.c
+$(OUTPUT)/listns_pagination_bug: ../filesystems/utils.c
+$(OUTPUT)/regression_pidfd_setns_test: ../filesystems/utils.c
+

linux-7.2.3/tools/testing/selftests/namespaces/config

@@ -0,0 +1,7 @@
+CONFIG_UTS_NS=y
+CONFIG_TIME_NS=y
+CONFIG_IPC_NS=y
+CONFIG_USER_NS=y
+CONFIG_PID_NS=y
+CONFIG_NET_NS=y
+CONFIG_CGROUPS=y

linux-7.2.3/tools/testing/selftests/namespaces/cred_change_test.c

@@ -0,0 +1,814 @@
+// SPDX-License-Identifier: GPL-2.0
+#define _GNU_SOURCE
+#include <errno.h>
+#include <fcntl.h>
+#include <limits.h>
+#include <sched.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/capability.h>
+#include <sys/ioctl.h>
+#include <sys/stat.h>
+#include <sys/syscall.h>
+#include <sys/types.h>
+#include <sys/wait.h>
+#include <unistd.h>
+#include <linux/nsfs.h>
+#include "../kselftest_harness.h"
+#include "../filesystems/utils.h"
+#include "wrappers.h"
+
+/*
+ * Test credential changes and their impact on namespace active references.
+ */
+
+/*
+ * Test setuid() in a user namespace properly swaps active references.
+ * Create a user namespace with multiple UIDs mapped, then setuid() between them.
+ * Verify that the user namespace remains active throughout.
+ */
+TEST(setuid_preserves_active_refs)
+{
+	pid_t pid;
+	int status;
+	__u64 userns_id;
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = CLONE_NEWUSER,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 ns_ids[256];
+	ssize_t ret;
+	int i;
+	bool found = false;
+	int pipefd[2];
+
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		/* Child process */
+		int fd, userns_fd;
+		__u64 child_userns_id;
+		uid_t orig_uid = getuid();
+		int setuid_count;
+
+		close(pipefd[0]);
+
+		/* Create new user namespace with multiple UIDs mapped (0-9) */
+		userns_fd = get_userns_fd(0, orig_uid, 10);
+		if (userns_fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (setns(userns_fd, CLONE_NEWUSER) < 0) {
+			close(userns_fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(userns_fd);
+
+		/* Get user namespace ID */
+		fd = open("/proc/self/ns/user", O_RDONLY);
+		if (fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (ioctl(fd, NS_GET_ID, &child_userns_id) < 0) {
+			close(fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(fd);
+
+		/* Send namespace ID to parent */
+		write(pipefd[1], &child_userns_id, sizeof(child_userns_id));
+
+		/*
+		 * Perform multiple setuid() calls.
+		 * Each setuid() triggers commit_creds() which should properly
+		 * swap active references via switch_cred_namespaces().
+		 */
+		for (setuid_count = 0; setuid_count < 50; setuid_count++) {
+			uid_t target_uid = (setuid_count % 10);
+			if (setuid(target_uid) < 0) {
+				if (errno != EPERM) {
+					close(pipefd[1]);
+					exit(1);
+				}
+			}
+		}
+
+		close(pipefd[1]);
+		exit(0);
+	}
+
+	/* Parent process */
+	close(pipefd[1]);
+
+	if (read(pipefd[0], &userns_id, sizeof(userns_id)) != sizeof(userns_id)) {
+		close(pipefd[0]);
+		kill(pid, SIGKILL);
+		waitpid(pid, NULL, 0);
+		SKIP(return, "Failed to get namespace ID from child");
+	}
+	close(pipefd[0]);
+
+	TH_LOG("Child user namespace ID: %llu", (unsigned long long)userns_id);
+
+	/* Verify namespace is active while child is running */
+	ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+	if (ret < 0) {
+		kill(pid, SIGKILL);
+		waitpid(pid, NULL, 0);
+		if (errno == ENOSYS)
+			SKIP(return, "listns() not supported");
+		ASSERT_GE(ret, 0);
+	}
+
+	for (i = 0; i < ret; i++) {
+		if (ns_ids[i] == userns_id) {
+			found = true;
+			break;
+		}
+	}
+	ASSERT_TRUE(found);
+
+	waitpid(pid, &status, 0);
+	ASSERT_TRUE(WIFEXITED(status));
+	ASSERT_EQ(WEXITSTATUS(status), 0);
+
+	/* Verify namespace becomes inactive after child exits */
+	ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+	ASSERT_GE(ret, 0);
+
+	found = false;
+	for (i = 0; i < ret; i++) {
+		if (ns_ids[i] == userns_id) {
+			found = true;
+			break;
+		}
+	}
+
+	ASSERT_FALSE(found);
+	TH_LOG("setuid() correctly preserved active references (no leak)");
+}
+
+/*
+ * Test setgid() in a user namespace properly handles active references.
+ */
+TEST(setgid_preserves_active_refs)
+{
+	pid_t pid;
+	int status;
+	__u64 userns_id;
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = CLONE_NEWUSER,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 ns_ids[256];
+	ssize_t ret;
+	int i;
+	bool found = false;
+	int pipefd[2];
+
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		/* Child process */
+		int fd, userns_fd;
+		__u64 child_userns_id;
+		uid_t orig_uid = getuid();
+		int setgid_count;
+
+		close(pipefd[0]);
+
+		/* Create new user namespace with multiple GIDs mapped */
+		userns_fd = get_userns_fd(0, orig_uid, 10);
+		if (userns_fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (setns(userns_fd, CLONE_NEWUSER) < 0) {
+			close(userns_fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(userns_fd);
+
+		/* Get user namespace ID */
+		fd = open("/proc/self/ns/user", O_RDONLY);
+		if (fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (ioctl(fd, NS_GET_ID, &child_userns_id) < 0) {
+			close(fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(fd);
+
+		write(pipefd[1], &child_userns_id, sizeof(child_userns_id));
+
+		/* Perform multiple setgid() calls */
+		for (setgid_count = 0; setgid_count < 50; setgid_count++) {
+			gid_t target_gid = (setgid_count % 10);
+			if (setgid(target_gid) < 0) {
+				if (errno != EPERM) {
+					close(pipefd[1]);
+					exit(1);
+				}
+			}
+		}
+
+		close(pipefd[1]);
+		exit(0);
+	}
+
+	/* Parent process */
+	close(pipefd[1]);
+
+	if (read(pipefd[0], &userns_id, sizeof(userns_id)) != sizeof(userns_id)) {
+		close(pipefd[0]);
+		kill(pid, SIGKILL);
+		waitpid(pid, NULL, 0);
+		SKIP(return, "Failed to get namespace ID from child");
+	}
+	close(pipefd[0]);
+
+	waitpid(pid, &status, 0);
+	ASSERT_TRUE(WIFEXITED(status));
+	ASSERT_EQ(WEXITSTATUS(status), 0);
+
+	/* Verify namespace becomes inactive */
+	ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+	if (ret < 0) {
+		if (errno == ENOSYS)
+			SKIP(return, "listns() not supported");
+		ASSERT_GE(ret, 0);
+	}
+
+	for (i = 0; i < ret; i++) {
+		if (ns_ids[i] == userns_id) {
+			found = true;
+			break;
+		}
+	}
+
+	ASSERT_FALSE(found);
+	TH_LOG("setgid() correctly preserved active references (no leak)");
+}
+
+/*
+ * Test setresuid() which changes real, effective, and saved UIDs.
+ * This should properly swap active references via commit_creds().
+ */
+TEST(setresuid_preserves_active_refs)
+{
+	pid_t pid;
+	int status;
+	__u64 userns_id;
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = CLONE_NEWUSER,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 ns_ids[256];
+	ssize_t ret;
+	int i;
+	bool found = false;
+	int pipefd[2];
+
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		/* Child process */
+		int fd, userns_fd;
+		__u64 child_userns_id;
+		uid_t orig_uid = getuid();
+		int setres_count;
+
+		close(pipefd[0]);
+
+		/* Create new user namespace */
+		userns_fd = get_userns_fd(0, orig_uid, 10);
+		if (userns_fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (setns(userns_fd, CLONE_NEWUSER) < 0) {
+			close(userns_fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(userns_fd);
+
+		/* Get user namespace ID */
+		fd = open("/proc/self/ns/user", O_RDONLY);
+		if (fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (ioctl(fd, NS_GET_ID, &child_userns_id) < 0) {
+			close(fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(fd);
+
+		write(pipefd[1], &child_userns_id, sizeof(child_userns_id));
+
+		/* Perform multiple setresuid() calls */
+		for (setres_count = 0; setres_count < 30; setres_count++) {
+			uid_t uid1 = (setres_count % 5);
+			uid_t uid2 = ((setres_count + 1) % 5);
+			uid_t uid3 = ((setres_count + 2) % 5);
+
+			if (setresuid(uid1, uid2, uid3) < 0) {
+				if (errno != EPERM) {
+					close(pipefd[1]);
+					exit(1);
+				}
+			}
+		}
+
+		close(pipefd[1]);
+		exit(0);
+	}
+
+	/* Parent process */
+	close(pipefd[1]);
+
+	if (read(pipefd[0], &userns_id, sizeof(userns_id)) != sizeof(userns_id)) {
+		close(pipefd[0]);
+		kill(pid, SIGKILL);
+		waitpid(pid, NULL, 0);
+		SKIP(return, "Failed to get namespace ID from child");
+	}
+	close(pipefd[0]);
+
+	waitpid(pid, &status, 0);
+	ASSERT_TRUE(WIFEXITED(status));
+	ASSERT_EQ(WEXITSTATUS(status), 0);
+
+	/* Verify namespace becomes inactive */
+	ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+	if (ret < 0) {
+		if (errno == ENOSYS)
+			SKIP(return, "listns() not supported");
+		ASSERT_GE(ret, 0);
+	}
+
+	for (i = 0; i < ret; i++) {
+		if (ns_ids[i] == userns_id) {
+			found = true;
+			break;
+		}
+	}
+
+	ASSERT_FALSE(found);
+	TH_LOG("setresuid() correctly preserved active references (no leak)");
+}
+
+/*
+ * Test credential changes across multiple user namespaces.
+ * Create nested user namespaces and verify active reference tracking.
+ */
+TEST(cred_change_nested_userns)
+{
+	pid_t pid;
+	int status;
+	__u64 parent_userns_id, child_userns_id;
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = CLONE_NEWUSER,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 ns_ids[256];
+	ssize_t ret;
+	int i;
+	bool found_parent = false, found_child = false;
+	int pipefd[2];
+
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		/* Child process */
+		int fd, userns_fd;
+		__u64 parent_id, child_id;
+		uid_t orig_uid = getuid();
+
+		close(pipefd[0]);
+
+		/* Create first user namespace */
+		userns_fd = get_userns_fd(0, orig_uid, 1);
+		if (userns_fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (setns(userns_fd, CLONE_NEWUSER) < 0) {
+			close(userns_fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(userns_fd);
+
+		/* Get first namespace ID */
+		fd = open("/proc/self/ns/user", O_RDONLY);
+		if (fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (ioctl(fd, NS_GET_ID, &parent_id) < 0) {
+			close(fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(fd);
+
+		/* Create nested user namespace */
+		userns_fd = get_userns_fd(0, 0, 1);
+		if (userns_fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (setns(userns_fd, CLONE_NEWUSER) < 0) {
+			close(userns_fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(userns_fd);
+
+		/* Get nested namespace ID */
+		fd = open("/proc/self/ns/user", O_RDONLY);
+		if (fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (ioctl(fd, NS_GET_ID, &child_id) < 0) {
+			close(fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(fd);
+
+		/* Send both IDs to parent */
+		write(pipefd[1], &parent_id, sizeof(parent_id));
+		write(pipefd[1], &child_id, sizeof(child_id));
+
+		/* Perform some credential changes in nested namespace */
+		setuid(0);
+		setgid(0);
+
+		close(pipefd[1]);
+		exit(0);
+	}
+
+	/* Parent process */
+	close(pipefd[1]);
+
+	/* Read both namespace IDs */
+	if (read(pipefd[0], &parent_userns_id, sizeof(parent_userns_id)) != sizeof(parent_userns_id)) {
+		close(pipefd[0]);
+		kill(pid, SIGKILL);
+		waitpid(pid, NULL, 0);
+		SKIP(return, "Failed to get parent namespace ID");
+	}
+
+	if (read(pipefd[0], &child_userns_id, sizeof(child_userns_id)) != sizeof(child_userns_id)) {
+		close(pipefd[0]);
+		kill(pid, SIGKILL);
+		waitpid(pid, NULL, 0);
+		SKIP(return, "Failed to get child namespace ID");
+	}
+	close(pipefd[0]);
+
+	TH_LOG("Parent userns: %llu, Child userns: %llu",
+	       (unsigned long long)parent_userns_id,
+	       (unsigned long long)child_userns_id);
+
+	/* Verify both namespaces are active */
+	ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+	if (ret < 0) {
+		kill(pid, SIGKILL);
+		waitpid(pid, NULL, 0);
+		if (errno == ENOSYS)
+			SKIP(return, "listns() not supported");
+		ASSERT_GE(ret, 0);
+	}
+
+	for (i = 0; i < ret; i++) {
+		if (ns_ids[i] == parent_userns_id)
+			found_parent = true;
+		if (ns_ids[i] == child_userns_id)
+			found_child = true;
+	}
+
+	ASSERT_TRUE(found_parent);
+	ASSERT_TRUE(found_child);
+
+	/* Wait for child */
+	waitpid(pid, &status, 0);
+	ASSERT_TRUE(WIFEXITED(status));
+	ASSERT_EQ(WEXITSTATUS(status), 0);
+
+	/* Verify both namespaces become inactive */
+	ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+	ASSERT_GE(ret, 0);
+
+	found_parent = false;
+	found_child = false;
+	for (i = 0; i < ret; i++) {
+		if (ns_ids[i] == parent_userns_id)
+			found_parent = true;
+		if (ns_ids[i] == child_userns_id)
+			found_child = true;
+	}
+
+	ASSERT_FALSE(found_parent);
+	ASSERT_FALSE(found_child);
+	TH_LOG("Nested user namespace credential changes preserved active refs (no leak)");
+}
+
+/*
+ * Test rapid credential changes don't cause refcount imbalances.
+ * This stress-tests the switch_cred_namespaces() logic.
+ */
+TEST(rapid_cred_changes_no_leak)
+{
+	pid_t pid;
+	int status;
+	__u64 userns_id;
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = CLONE_NEWUSER,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 ns_ids[256];
+	ssize_t ret;
+	int i;
+	bool found = false;
+	int pipefd[2];
+
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		/* Child process */
+		int fd, userns_fd;
+		__u64 child_userns_id;
+		uid_t orig_uid = getuid();
+		int change_count;
+
+		close(pipefd[0]);
+
+		/* Create new user namespace with wider range of UIDs/GIDs */
+		userns_fd = get_userns_fd(0, orig_uid, 100);
+		if (userns_fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (setns(userns_fd, CLONE_NEWUSER) < 0) {
+			close(userns_fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(userns_fd);
+
+		/* Get user namespace ID */
+		fd = open("/proc/self/ns/user", O_RDONLY);
+		if (fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (ioctl(fd, NS_GET_ID, &child_userns_id) < 0) {
+			close(fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(fd);
+
+		write(pipefd[1], &child_userns_id, sizeof(child_userns_id));
+
+		/*
+		 * Perform many rapid credential changes.
+		 * Mix setuid, setgid, setreuid, setregid, setresuid, setresgid.
+		 */
+		for (change_count = 0; change_count < 200; change_count++) {
+			switch (change_count % 6) {
+			case 0:
+				setuid(change_count % 50);
+				break;
+			case 1:
+				setgid(change_count % 50);
+				break;
+			case 2:
+				setreuid(change_count % 50, (change_count + 1) % 50);
+				break;
+			case 3:
+				setregid(change_count % 50, (change_count + 1) % 50);
+				break;
+			case 4:
+				setresuid(change_count % 50, (change_count + 1) % 50, (change_count + 2) % 50);
+				break;
+			case 5:
+				setresgid(change_count % 50, (change_count + 1) % 50, (change_count + 2) % 50);
+				break;
+			}
+		}
+
+		close(pipefd[1]);
+		exit(0);
+	}
+
+	/* Parent process */
+	close(pipefd[1]);
+
+	if (read(pipefd[0], &userns_id, sizeof(userns_id)) != sizeof(userns_id)) {
+		close(pipefd[0]);
+		kill(pid, SIGKILL);
+		waitpid(pid, NULL, 0);
+		SKIP(return, "Failed to get namespace ID from child");
+	}
+	close(pipefd[0]);
+
+	TH_LOG("Testing with user namespace ID: %llu", (unsigned long long)userns_id);
+
+	waitpid(pid, &status, 0);
+	ASSERT_TRUE(WIFEXITED(status));
+	ASSERT_EQ(WEXITSTATUS(status), 0);
+
+	/* Verify namespace becomes inactive (no leaked active refs) */
+	ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+	if (ret < 0) {
+		if (errno == ENOSYS)
+			SKIP(return, "listns() not supported");
+		ASSERT_GE(ret, 0);
+	}
+
+	for (i = 0; i < ret; i++) {
+		if (ns_ids[i] == userns_id) {
+			found = true;
+			break;
+		}
+	}
+
+	ASSERT_FALSE(found);
+	TH_LOG("200 rapid credential changes completed with no active ref leak");
+}
+
+/*
+ * Test setfsuid/setfsgid which change filesystem UID/GID.
+ * These also trigger credential changes but may have different code paths.
+ */
+TEST(setfsuid_preserves_active_refs)
+{
+	pid_t pid;
+	int status;
+	__u64 userns_id;
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = CLONE_NEWUSER,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 ns_ids[256];
+	ssize_t ret;
+	int i;
+	bool found = false;
+	int pipefd[2];
+
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		/* Child process */
+		int fd, userns_fd;
+		__u64 child_userns_id;
+		uid_t orig_uid = getuid();
+		int change_count;
+
+		close(pipefd[0]);
+
+		/* Create new user namespace */
+		userns_fd = get_userns_fd(0, orig_uid, 10);
+		if (userns_fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (setns(userns_fd, CLONE_NEWUSER) < 0) {
+			close(userns_fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(userns_fd);
+
+		/* Get user namespace ID */
+		fd = open("/proc/self/ns/user", O_RDONLY);
+		if (fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (ioctl(fd, NS_GET_ID, &child_userns_id) < 0) {
+			close(fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(fd);
+
+		write(pipefd[1], &child_userns_id, sizeof(child_userns_id));
+
+		/* Perform multiple setfsuid/setfsgid calls */
+		for (change_count = 0; change_count < 50; change_count++) {
+			setfsuid(change_count % 10);
+			setfsgid(change_count % 10);
+		}
+
+		close(pipefd[1]);
+		exit(0);
+	}
+
+	/* Parent process */
+	close(pipefd[1]);
+
+	if (read(pipefd[0], &userns_id, sizeof(userns_id)) != sizeof(userns_id)) {
+		close(pipefd[0]);
+		kill(pid, SIGKILL);
+		waitpid(pid, NULL, 0);
+		SKIP(return, "Failed to get namespace ID from child");
+	}
+	close(pipefd[0]);
+
+	waitpid(pid, &status, 0);
+	ASSERT_TRUE(WIFEXITED(status));
+	ASSERT_EQ(WEXITSTATUS(status), 0);
+
+	/* Verify namespace becomes inactive */
+	ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+	if (ret < 0) {
+		if (errno == ENOSYS)
+			SKIP(return, "listns() not supported");
+		ASSERT_GE(ret, 0);
+	}
+
+	for (i = 0; i < ret; i++) {
+		if (ns_ids[i] == userns_id) {
+			found = true;
+			break;
+		}
+	}
+
+	ASSERT_FALSE(found);
+	TH_LOG("setfsuid/setfsgid correctly preserved active references (no leak)");
+}
+
+TEST_HARNESS_MAIN

linux-7.2.3/tools/testing/selftests/namespaces/init_ino_test.c

@@ -0,0 +1,61 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+// Copyright (c) 2025 Christian Brauner <brauner@kernel.org>
+
+#define _GNU_SOURCE
+#include <fcntl.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <sys/stat.h>
+#include <unistd.h>
+#include <errno.h>
+#include <string.h>
+#include <linux/nsfs.h>
+
+#include "kselftest_harness.h"
+
+struct ns_info {
+	const char *name;
+	const char *proc_path;
+	unsigned int expected_ino;
+};
+
+static struct ns_info namespaces[] = {
+	{ "ipc", "/proc/1/ns/ipc", IPC_NS_INIT_INO },
+	{ "uts", "/proc/1/ns/uts", UTS_NS_INIT_INO },
+	{ "user", "/proc/1/ns/user", USER_NS_INIT_INO },
+	{ "pid", "/proc/1/ns/pid", PID_NS_INIT_INO },
+	{ "cgroup", "/proc/1/ns/cgroup", CGROUP_NS_INIT_INO },
+	{ "time", "/proc/1/ns/time", TIME_NS_INIT_INO },
+	{ "net", "/proc/1/ns/net", NET_NS_INIT_INO },
+	{ "mnt", "/proc/1/ns/mnt", MNT_NS_INIT_INO },
+};
+
+TEST(init_namespace_inodes)
+{
+	struct stat st;
+
+	for (int i = 0; i < sizeof(namespaces) / sizeof(namespaces[0]); i++) {
+		int ret = stat(namespaces[i].proc_path, &st);
+
+		/* Some namespaces might not be available (e.g., time namespace on older kernels) */
+		if (ret < 0) {
+			if (errno == ENOENT) {
+				ksft_test_result_skip("%s namespace not available\n",
+						      namespaces[i].name);
+				continue;
+			}
+			ASSERT_GE(ret, 0)
+			TH_LOG("Failed to stat %s: %s",
+			       namespaces[i].proc_path, strerror(errno));
+		}
+
+		ASSERT_EQ(st.st_ino, namespaces[i].expected_ino)
+			TH_LOG("Namespace %s has inode 0x%lx, expected 0x%x",
+			       namespaces[i].name, st.st_ino, namespaces[i].expected_ino);
+
+		ksft_print_msg("Namespace %s: inode 0x%lx matches expected 0x%x\n",
+			       namespaces[i].name, st.st_ino, namespaces[i].expected_ino);
+	}
+}
+
+TEST_HARNESS_MAIN

linux-7.2.3/tools/testing/selftests/namespaces/listns_efault_test.c

@@ -0,0 +1,550 @@
+// SPDX-License-Identifier: GPL-2.0
+#define _GNU_SOURCE
+#include <errno.h>
+#include <fcntl.h>
+#include <limits.h>
+#include <sched.h>
+#include <signal.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <linux/nsfs.h>
+#include <sys/ioctl.h>
+#include <sys/mman.h>
+#include <sys/mount.h>
+#include <sys/socket.h>
+#include <sys/stat.h>
+#include <sys/syscall.h>
+#include <sys/types.h>
+#include <sys/wait.h>
+#include <unistd.h>
+#include "../kselftest_harness.h"
+#include "../pidfd/pidfd.h"
+#include "wrappers.h"
+
+/*
+ * Test listns() error handling with invalid buffer addresses.
+ *
+ * When the buffer pointer is invalid (e.g., crossing page boundaries
+ * into unmapped memory), listns() returns EINVAL.
+ *
+ * This test also creates mount namespaces that get destroyed during
+ * iteration, testing that namespace cleanup happens outside the RCU
+ * read lock.
+ */
+TEST(listns_partial_fault_with_ns_cleanup)
+{
+	void *map;
+	__u64 *ns_ids;
+	ssize_t ret;
+	long page_size;
+	pid_t pid, iter_pid, ns_pids[5];
+	int pidfds[5];
+	int sv[5][2];
+	int iter_pidfd;
+	int i, status;
+	char c;
+
+	page_size = sysconf(_SC_PAGESIZE);
+	ASSERT_GT(page_size, 0);
+
+	/*
+	 * Map two pages:
+	 * - First page: readable and writable
+	 * - Second page: will be unmapped to trigger EFAULT
+	 */
+	map = mmap(NULL, page_size * 2, PROT_READ | PROT_WRITE,
+		   MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
+	ASSERT_NE(map, MAP_FAILED);
+
+	/* Unmap the second page */
+	ret = munmap((char *)map + page_size, page_size);
+	ASSERT_EQ(ret, 0);
+
+	/*
+	 * Position the buffer pointer so there's room for exactly one u64
+	 * before the page boundary. The second u64 would fall into the
+	 * unmapped page.
+	 */
+	ns_ids = ((__u64 *)((char *)map + page_size)) - 1;
+
+	/*
+	 * Create a separate process to run listns() in a loop concurrently
+	 * with namespace creation and destruction.
+	 */
+	iter_pid = create_child(&iter_pidfd, 0);
+	ASSERT_NE(iter_pid, -1);
+
+	if (iter_pid == 0) {
+		struct ns_id_req req = {
+			.size = sizeof(req),
+			.spare = 0,
+			.ns_id = 0,
+			.ns_type = 0,  /* All types */
+			.spare2 = 0,
+			.user_ns_id = 0,  /* Global listing */
+		};
+		int iter_ret;
+
+		/*
+		 * Loop calling listns() until killed.
+		 * The kernel should:
+		 * 1. Successfully write the first namespace ID (within valid page)
+		 * 2. Fail with EFAULT when trying to write the second ID (unmapped page)
+		 * 3. Handle concurrent namespace destruction without deadlock
+		 */
+		while (1) {
+			iter_ret = sys_listns(&req, ns_ids, 2, 0);
+
+			if (iter_ret == -1 && errno == ENOSYS)
+				_exit(PIDFD_SKIP);
+		}
+	}
+
+	/* Small delay to let iterator start looping */
+	usleep(50000);
+
+	/*
+	 * Create several child processes, each in its own mount namespace.
+	 * These will be destroyed while the iterator is running listns().
+	 */
+	for (i = 0; i < 5; i++) {
+		/* Create socketpair for synchronization */
+		ASSERT_EQ(socketpair(AF_UNIX, SOCK_STREAM, 0, sv[i]), 0);
+
+		pid = create_child(&pidfds[i], CLONE_NEWNS);
+		ASSERT_NE(pid, -1);
+		ns_pids[i] = pid;
+
+		if (pid == 0) {
+			close(sv[i][0]); /* Close parent end */
+
+			if (mount(NULL, "/", NULL, MS_REC | MS_PRIVATE, 0))
+				_exit(1);
+
+			/* Child: create a couple of tmpfs mounts */
+			if (mkdir("/tmp/test_mnt1", 0755) == -1 && errno != EEXIST)
+				_exit(1);
+			if (mkdir("/tmp/test_mnt2", 0755) == -1 && errno != EEXIST)
+				_exit(1);
+
+			if (mount("tmpfs", "/tmp/test_mnt1", "tmpfs", 0, NULL) == -1)
+				_exit(1);
+			if (mount("tmpfs", "/tmp/test_mnt2", "tmpfs", 0, NULL) == -1)
+				_exit(1);
+
+			/* Signal parent that setup is complete */
+			if (write_nointr(sv[i][1], "R", 1) != 1)
+				_exit(1);
+
+			/* Wait for parent to signal us to exit */
+			if (read_nointr(sv[i][1], &c, 1) != 1)
+				_exit(1);
+
+			close(sv[i][1]);
+			_exit(0);
+		}
+
+		close(sv[i][1]); /* Close child end */
+	}
+
+	/* Wait for all children to finish setup */
+	for (i = 0; i < 5; i++) {
+		ret = read_nointr(sv[i][0], &c, 1);
+		ASSERT_EQ(ret, 1);
+		ASSERT_EQ(c, 'R');
+	}
+
+	/*
+	 * Signal children to exit. This will destroy their mount namespaces
+	 * while listns() is iterating the namespace tree.
+	 * This tests that cleanup happens outside the RCU read lock.
+	 */
+	for (i = 0; i < 5; i++)
+		write_nointr(sv[i][0], "X", 1);
+
+	/* Wait for all mount namespace children to exit and cleanup */
+	for (i = 0; i < 5; i++) {
+		waitpid(ns_pids[i], NULL, 0);
+		close(sv[i][0]);
+		close(pidfds[i]);
+	}
+
+	/* Kill iterator and wait for it */
+	sys_pidfd_send_signal(iter_pidfd, SIGKILL, NULL, 0);
+	ret = waitpid(iter_pid, &status, 0);
+	ASSERT_EQ(ret, iter_pid);
+	close(iter_pidfd);
+
+	/* If listns() is not supported the iterator exits cleanly via ENOSYS */
+	if (WIFEXITED(status) && WEXITSTATUS(status) == PIDFD_SKIP) {
+		munmap(map, page_size);
+		SKIP(return, "listns() not supported");
+	}
+
+	/* Should have been killed */
+	ASSERT_TRUE(WIFSIGNALED(status));
+	ASSERT_EQ(WTERMSIG(status), SIGKILL);
+
+	/* Clean up */
+	munmap(map, page_size);
+}
+
+/*
+ * Test listns() error handling when the entire buffer is invalid.
+ * This is a sanity check that basic invalid pointer detection works.
+ */
+TEST(listns_complete_fault)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = 0,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 *ns_ids;
+	ssize_t ret;
+
+	/* Use a clearly invalid pointer */
+	ns_ids = (__u64 *)0xdeadbeef;
+
+	ret = sys_listns(&req, ns_ids, 10, 0);
+
+	if (ret == -1 && errno == ENOSYS)
+		SKIP(return, "listns() not supported");
+
+	/* Should fail with EFAULT */
+	ASSERT_EQ(ret, -1);
+	ASSERT_EQ(errno, EFAULT);
+}
+
+/*
+ * Test listns() error handling when the buffer is NULL.
+ */
+TEST(listns_null_buffer)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = 0,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	ssize_t ret;
+
+	/* NULL buffer with non-zero count should fail */
+	ret = sys_listns(&req, NULL, 10, 0);
+
+	if (ret == -1 && errno == ENOSYS)
+		SKIP(return, "listns() not supported");
+
+	/* Should fail with EFAULT */
+	ASSERT_EQ(ret, -1);
+	ASSERT_EQ(errno, EFAULT);
+}
+
+/*
+ * Test listns() with a buffer that becomes invalid mid-iteration
+ * (after several successful writes), combined with mount namespace
+ * destruction to test RCU cleanup logic.
+ */
+TEST(listns_late_fault_with_ns_cleanup)
+{
+	void *map;
+	__u64 *ns_ids;
+	ssize_t ret;
+	long page_size;
+	pid_t pid, iter_pid, ns_pids[10];
+	int pidfds[10];
+	int sv[10][2];
+	int iter_pidfd;
+	int i, status;
+	char c;
+
+	page_size = sysconf(_SC_PAGESIZE);
+	ASSERT_GT(page_size, 0);
+
+	/* Map two pages */
+	map = mmap(NULL, page_size * 2, PROT_READ | PROT_WRITE,
+		   MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
+	ASSERT_NE(map, MAP_FAILED);
+
+	/* Unmap the second page */
+	ret = munmap((char *)map + page_size, page_size);
+	ASSERT_EQ(ret, 0);
+
+	/*
+	 * Position buffer so we can write several u64s successfully
+	 * before hitting the page boundary.
+	 */
+	ns_ids = ((__u64 *)((char *)map + page_size)) - 5;
+
+	/*
+	 * Create a separate process to run listns() concurrently.
+	 */
+	iter_pid = create_child(&iter_pidfd, 0);
+	ASSERT_NE(iter_pid, -1);
+
+	if (iter_pid == 0) {
+		struct ns_id_req req = {
+			.size = sizeof(req),
+			.spare = 0,
+			.ns_id = 0,
+			.ns_type = 0,
+			.spare2 = 0,
+			.user_ns_id = 0,
+		};
+		int iter_ret;
+
+		/*
+		 * Loop calling listns() until killed.
+		 * Request 10 namespace IDs while namespaces are being destroyed.
+		 * This tests:
+		 * 1. EFAULT handling when buffer becomes invalid
+		 * 2. Namespace cleanup outside RCU read lock during iteration
+		 */
+		while (1) {
+			iter_ret = sys_listns(&req, ns_ids, 10, 0);
+
+			if (iter_ret == -1 && errno == ENOSYS)
+				_exit(PIDFD_SKIP);
+		}
+	}
+
+	/* Small delay to let iterator start looping */
+	usleep(50000);
+
+	/*
+	 * Create more children with mount namespaces to increase the
+	 * likelihood that namespace cleanup happens during iteration.
+	 */
+	for (i = 0; i < 10; i++) {
+		/* Create socketpair for synchronization */
+		ASSERT_EQ(socketpair(AF_UNIX, SOCK_STREAM, 0, sv[i]), 0);
+
+		pid = create_child(&pidfds[i], CLONE_NEWNS);
+		ASSERT_NE(pid, -1);
+		ns_pids[i] = pid;
+
+		if (pid == 0) {
+			close(sv[i][0]); /* Close parent end */
+
+			if (mount(NULL, "/", NULL, MS_REC | MS_PRIVATE, 0))
+				_exit(1);
+
+			/* Child: create tmpfs mounts */
+			if (mkdir("/tmp/test_mnt1", 0755) == -1 && errno != EEXIST)
+				_exit(1);
+			if (mkdir("/tmp/test_mnt2", 0755) == -1 && errno != EEXIST)
+				_exit(1);
+
+			if (mount("tmpfs", "/tmp/test_mnt1", "tmpfs", 0, NULL) == -1)
+				_exit(1);
+			if (mount("tmpfs", "/tmp/test_mnt2", "tmpfs", 0, NULL) == -1)
+				_exit(1);
+
+			/* Signal parent that setup is complete */
+			if (write_nointr(sv[i][1], "R", 1) != 1)
+				_exit(1);
+
+			/* Wait for parent to signal us to exit */
+			if (read_nointr(sv[i][1], &c, 1) != 1)
+				_exit(1);
+
+			close(sv[i][1]);
+			_exit(0);
+		}
+
+		close(sv[i][1]); /* Close child end */
+	}
+
+	/* Wait for all children to finish setup */
+	for (i = 0; i < 10; i++) {
+		ret = read_nointr(sv[i][0], &c, 1);
+		ASSERT_EQ(ret, 1);
+		ASSERT_EQ(c, 'R');
+	}
+
+	/* Kill half the children */
+	for (i = 0; i < 5; i++)
+		write_nointr(sv[i][0], "X", 1);
+
+	/* Small delay to let some exit */
+	usleep(10000);
+
+	/* Kill remaining children */
+	for (i = 5; i < 10; i++)
+		write_nointr(sv[i][0], "X", 1);
+
+	/* Wait for all children and cleanup */
+	for (i = 0; i < 10; i++) {
+		waitpid(ns_pids[i], NULL, 0);
+		close(sv[i][0]);
+		close(pidfds[i]);
+	}
+
+	/* Kill iterator and wait for it */
+	sys_pidfd_send_signal(iter_pidfd, SIGKILL, NULL, 0);
+	ret = waitpid(iter_pid, &status, 0);
+	ASSERT_EQ(ret, iter_pid);
+	close(iter_pidfd);
+
+	/* If listns() is not supported the iterator exits cleanly via ENOSYS */
+	if (WIFEXITED(status) && WEXITSTATUS(status) == PIDFD_SKIP) {
+		munmap(map, page_size);
+		SKIP(return, "listns() not supported");
+	}
+
+	/* Should have been killed */
+	ASSERT_TRUE(WIFSIGNALED(status));
+	ASSERT_EQ(WTERMSIG(status), SIGKILL);
+
+	/* Clean up */
+	munmap(map, page_size);
+}
+
+/*
+ * Test specifically focused on mount namespace cleanup during EFAULT.
+ * Filter for mount namespaces only.
+ */
+TEST(listns_mnt_ns_cleanup_on_fault)
+{
+	void *map;
+	__u64 *ns_ids;
+	ssize_t ret;
+	long page_size;
+	pid_t pid, iter_pid, ns_pids[8];
+	int pidfds[8];
+	int sv[8][2];
+	int iter_pidfd;
+	int i, status;
+	char c;
+
+	page_size = sysconf(_SC_PAGESIZE);
+	ASSERT_GT(page_size, 0);
+
+	/* Set up partial fault buffer */
+	map = mmap(NULL, page_size * 2, PROT_READ | PROT_WRITE,
+		   MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
+	ASSERT_NE(map, MAP_FAILED);
+
+	ret = munmap((char *)map + page_size, page_size);
+	ASSERT_EQ(ret, 0);
+
+	/* Position for 3 successful writes, then fault */
+	ns_ids = ((__u64 *)((char *)map + page_size)) - 3;
+
+	/*
+	 * Create a separate process to run listns() concurrently.
+	 */
+	iter_pid = create_child(&iter_pidfd, 0);
+	ASSERT_NE(iter_pid, -1);
+
+	if (iter_pid == 0) {
+		struct ns_id_req req = {
+			.size = sizeof(req),
+			.spare = 0,
+			.ns_id = 0,
+			.ns_type = CLONE_NEWNS,  /* Only mount namespaces */
+			.spare2 = 0,
+			.user_ns_id = 0,
+		};
+		int iter_ret;
+
+		/*
+		 * Loop calling listns() until killed.
+		 * Call listns() to race with namespace destruction.
+		 */
+		while (1) {
+			iter_ret = sys_listns(&req, ns_ids, 10, 0);
+
+			if (iter_ret == -1 && errno == ENOSYS)
+				_exit(PIDFD_SKIP);
+		}
+	}
+
+	/* Small delay to let iterator start looping */
+	usleep(50000);
+
+	/* Create children with mount namespaces */
+	for (i = 0; i < 8; i++) {
+		/* Create socketpair for synchronization */
+		ASSERT_EQ(socketpair(AF_UNIX, SOCK_STREAM, 0, sv[i]), 0);
+
+		pid = create_child(&pidfds[i], CLONE_NEWNS);
+		ASSERT_NE(pid, -1);
+		ns_pids[i] = pid;
+
+		if (pid == 0) {
+			close(sv[i][0]); /* Close parent end */
+
+			if (mount(NULL, "/", NULL, MS_REC | MS_PRIVATE, 0))
+				_exit(1);
+
+			/* Do some mount operations to make cleanup more interesting */
+			if (mkdir("/tmp/test_mnt1", 0755) == -1 && errno != EEXIST)
+				_exit(1);
+			if (mkdir("/tmp/test_mnt2", 0755) == -1 && errno != EEXIST)
+				_exit(1);
+
+			if (mount("tmpfs", "/tmp/test_mnt1", "tmpfs", 0, NULL) == -1)
+				_exit(1);
+			if (mount("tmpfs", "/tmp/test_mnt2", "tmpfs", 0, NULL) == -1)
+				_exit(1);
+
+			/* Signal parent that setup is complete */
+			if (write_nointr(sv[i][1], "R", 1) != 1)
+				_exit(1);
+
+			/* Wait for parent to signal us to exit */
+			if (read_nointr(sv[i][1], &c, 1) != 1)
+				_exit(1);
+
+			close(sv[i][1]);
+			_exit(0);
+		}
+
+		close(sv[i][1]); /* Close child end */
+	}
+
+	/* Wait for all children to finish setup */
+	for (i = 0; i < 8; i++) {
+		ret = read_nointr(sv[i][0], &c, 1);
+		ASSERT_EQ(ret, 1);
+		ASSERT_EQ(c, 'R');
+	}
+
+	/* Kill children to trigger namespace destruction during iteration */
+	for (i = 0; i < 8; i++)
+		write_nointr(sv[i][0], "X", 1);
+
+	/* Wait for children and cleanup */
+	for (i = 0; i < 8; i++) {
+		waitpid(ns_pids[i], NULL, 0);
+		close(sv[i][0]);
+		close(pidfds[i]);
+	}
+
+	/* Kill iterator and wait for it */
+	sys_pidfd_send_signal(iter_pidfd, SIGKILL, NULL, 0);
+	ret = waitpid(iter_pid, &status, 0);
+	ASSERT_EQ(ret, iter_pid);
+	close(iter_pidfd);
+
+	/* If listns() is not supported the iterator exits cleanly via ENOSYS */
+	if (WIFEXITED(status) && WEXITSTATUS(status) == PIDFD_SKIP) {
+		munmap(map, page_size);
+		SKIP(return, "listns() not supported");
+	}
+
+	/* Should have been killed */
+	ASSERT_TRUE(WIFSIGNALED(status));
+	ASSERT_EQ(WTERMSIG(status), SIGKILL);
+
+	munmap(map, page_size);
+}
+
+TEST_HARNESS_MAIN

linux-7.2.3/tools/testing/selftests/namespaces/listns_pagination_bug.c

@@ -0,0 +1,138 @@
+// SPDX-License-Identifier: GPL-2.0
+#define _GNU_SOURCE
+#include <errno.h>
+#include <sched.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <sys/socket.h>
+#include <sys/wait.h>
+#include <unistd.h>
+#include "../kselftest_harness.h"
+#include "../filesystems/utils.h"
+#include "wrappers.h"
+
+/*
+ * Minimal test case to reproduce KASAN out-of-bounds in listns pagination.
+ *
+ * The bug occurs when:
+ * 1. Filtering by a specific namespace type (e.g., CLONE_NEWUSER)
+ * 2. Using pagination (req.ns_id != 0)
+ * 3. The lookup_ns_id_at() call in do_listns() passes ns_type=0 instead of
+ *    the filtered type, causing it to search the unified tree and potentially
+ *    return a namespace of the wrong type.
+ */
+TEST(pagination_with_type_filter)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = CLONE_NEWUSER,  /* Filter by user namespace */
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	pid_t pids[10];
+	int num_children = 10;
+	int i;
+	int sv[2];
+	__u64 first_batch[3];
+	ssize_t ret;
+
+	ASSERT_EQ(socketpair(AF_UNIX, SOCK_STREAM, 0, sv), 0);
+
+	/* Create children with user namespaces */
+	for (i = 0; i < num_children; i++) {
+		pids[i] = fork();
+		ASSERT_GE(pids[i], 0);
+
+		if (pids[i] == 0) {
+			char c;
+			close(sv[0]);
+
+			if (setup_userns() < 0) {
+				close(sv[1]);
+				exit(1);
+			}
+
+			/* Signal parent we're ready */
+			if (write(sv[1], &c, 1) != 1) {
+				close(sv[1]);
+				exit(1);
+			}
+
+			/* Wait for parent signal to exit */
+			if (read(sv[1], &c, 1) != 1) {
+				close(sv[1]);
+				exit(1);
+			}
+
+			close(sv[1]);
+			exit(0);
+		}
+	}
+
+	close(sv[1]);
+
+	/* Wait for all children to signal ready */
+	for (i = 0; i < num_children; i++) {
+		char c;
+		if (read(sv[0], &c, 1) != 1) {
+			close(sv[0]);
+			for (int j = 0; j < num_children; j++)
+				kill(pids[j], SIGKILL);
+			for (int j = 0; j < num_children; j++)
+				waitpid(pids[j], NULL, 0);
+			ASSERT_TRUE(false);
+		}
+	}
+
+	/* First batch - this should work */
+	ret = sys_listns(&req, first_batch, 3, 0);
+	if (ret < 0) {
+		if (errno == ENOSYS) {
+			close(sv[0]);
+			for (i = 0; i < num_children; i++)
+				kill(pids[i], SIGKILL);
+			for (i = 0; i < num_children; i++)
+				waitpid(pids[i], NULL, 0);
+			SKIP(return, "listns() not supported");
+		}
+		ASSERT_GE(ret, 0);
+	}
+
+	TH_LOG("First batch returned %zd entries", ret);
+
+	if (ret == 3) {
+		__u64 second_batch[3];
+
+		/* Second batch - pagination triggers the bug */
+		req.ns_id = first_batch[2];  /* Continue from last ID */
+		ret = sys_listns(&req, second_batch, 3, 0);
+
+		TH_LOG("Second batch returned %zd entries", ret);
+		ASSERT_GE(ret, 0);
+	}
+
+	/* Signal all children to exit */
+	for (i = 0; i < num_children; i++) {
+		char c = 'X';
+		if (write(sv[0], &c, 1) != 1) {
+			close(sv[0]);
+			for (int j = i; j < num_children; j++)
+				kill(pids[j], SIGKILL);
+			for (int j = 0; j < num_children; j++)
+				waitpid(pids[j], NULL, 0);
+			ASSERT_TRUE(false);
+		}
+	}
+
+	close(sv[0]);
+
+	/* Cleanup */
+	for (i = 0; i < num_children; i++) {
+		int status;
+		waitpid(pids[i], &status, 0);
+	}
+}
+
+TEST_HARNESS_MAIN

linux-7.2.3/tools/testing/selftests/namespaces/listns_permissions_test.c

@@ -0,0 +1,759 @@
+// SPDX-License-Identifier: GPL-2.0
+#define _GNU_SOURCE
+#include <errno.h>
+#include <fcntl.h>
+#include <limits.h>
+#include <sched.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <linux/nsfs.h>
+#include <sys/capability.h>
+#include <sys/ioctl.h>
+#include <sys/prctl.h>
+#include <sys/stat.h>
+#include <sys/syscall.h>
+#include <sys/types.h>
+#include <sys/wait.h>
+#include <unistd.h>
+#include "../kselftest_harness.h"
+#include "../filesystems/utils.h"
+#include "wrappers.h"
+
+/*
+ * Test that unprivileged users can only see namespaces they're currently in.
+ * Create a namespace, drop privileges, verify we can only see our own namespaces.
+ */
+TEST(listns_unprivileged_current_only)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = CLONE_NEWNET,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 ns_ids[100];
+	ssize_t ret;
+	int pipefd[2];
+	pid_t pid;
+	int status;
+	bool found_ours;
+	int unexpected_count;
+
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		int fd;
+		__u64 our_netns_id;
+		bool found_ours;
+		int unexpected_count;
+
+		close(pipefd[0]);
+
+		/* Create user namespace to be unprivileged */
+		if (setup_userns() < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		/* Create a network namespace */
+		if (unshare(CLONE_NEWNET) < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		/* Get our network namespace ID */
+		fd = open("/proc/self/ns/net", O_RDONLY);
+		if (fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (ioctl(fd, NS_GET_ID, &our_netns_id) < 0) {
+			close(fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(fd);
+
+		/* Now we're unprivileged - list all network namespaces */
+		ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+		if (ret < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		/* We should only see our own network namespace */
+		found_ours = false;
+		unexpected_count = 0;
+
+		for (ssize_t i = 0; i < ret; i++) {
+			if (ns_ids[i] == our_netns_id) {
+				found_ours = true;
+			} else {
+				/* This is either init_net (which we can see) or unexpected */
+				unexpected_count++;
+			}
+		}
+
+		/* Send results to parent */
+		write(pipefd[1], &found_ours, sizeof(found_ours));
+		write(pipefd[1], &unexpected_count, sizeof(unexpected_count));
+		close(pipefd[1]);
+		exit(0);
+	}
+
+	/* Parent */
+	close(pipefd[1]);
+
+	found_ours = false;
+	unexpected_count = 0;
+	read(pipefd[0], &found_ours, sizeof(found_ours));
+	read(pipefd[0], &unexpected_count, sizeof(unexpected_count));
+	close(pipefd[0]);
+
+	waitpid(pid, &status, 0);
+	ASSERT_TRUE(WIFEXITED(status));
+	ASSERT_EQ(WEXITSTATUS(status), 0);
+
+	/* Child should have seen its own namespace */
+	ASSERT_TRUE(found_ours);
+
+	TH_LOG("Unprivileged child saw its own namespace, plus %d others (likely init_net)",
+			unexpected_count);
+}
+
+/*
+ * Test that users with CAP_SYS_ADMIN in a user namespace can see
+ * all namespaces owned by that user namespace.
+ */
+TEST(listns_cap_sys_admin_in_userns)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = 0,  /* All types */
+		.spare2 = 0,
+		.user_ns_id = 0,  /* Will be set to our created user namespace */
+	};
+	__u64 ns_ids[100];
+	int pipefd[2];
+	pid_t pid;
+	int status;
+	bool success;
+	ssize_t count;
+
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		int fd;
+		__u64 userns_id;
+		ssize_t ret;
+		int min_expected;
+		bool success;
+
+		close(pipefd[0]);
+
+		/* Create user namespace - we'll have CAP_SYS_ADMIN in it */
+		if (setup_userns() < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		/* Get the user namespace ID */
+		fd = open("/proc/self/ns/user", O_RDONLY);
+		if (fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (ioctl(fd, NS_GET_ID, &userns_id) < 0) {
+			close(fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(fd);
+
+		/* Create several namespaces owned by this user namespace */
+		unshare(CLONE_NEWNET);
+		unshare(CLONE_NEWUTS);
+		unshare(CLONE_NEWIPC);
+
+		/* List namespaces owned by our user namespace */
+		req.user_ns_id = userns_id;
+		ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+		if (ret < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		/*
+		 * We have CAP_SYS_ADMIN in this user namespace,
+		 * so we should see all namespaces owned by it.
+		 * That includes: net, uts, ipc, and the user namespace itself.
+		 */
+		min_expected = 4;
+		success = (ret >= min_expected);
+
+		write(pipefd[1], &success, sizeof(success));
+		write(pipefd[1], &ret, sizeof(ret));
+		close(pipefd[1]);
+		exit(0);
+	}
+
+	/* Parent */
+	close(pipefd[1]);
+
+	success = false;
+	count = 0;
+	read(pipefd[0], &success, sizeof(success));
+	read(pipefd[0], &count, sizeof(count));
+	close(pipefd[0]);
+
+	waitpid(pid, &status, 0);
+	ASSERT_TRUE(WIFEXITED(status));
+	ASSERT_EQ(WEXITSTATUS(status), 0);
+
+	ASSERT_TRUE(success);
+	TH_LOG("User with CAP_SYS_ADMIN saw %zd namespaces owned by their user namespace",
+			count);
+}
+
+/*
+ * Test that users cannot see namespaces from unrelated user namespaces.
+ * Create two sibling user namespaces, verify they can't see each other's
+ * owned namespaces.
+ */
+TEST(listns_cannot_see_sibling_userns_namespaces)
+{
+	int pipefd[2];
+	pid_t pid1, pid2;
+	int status;
+	__u64 netns_a_id;
+	int pipefd2[2];
+	bool found_sibling_netns;
+
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	/* Fork first child - creates user namespace A */
+	pid1 = fork();
+	ASSERT_GE(pid1, 0);
+
+	if (pid1 == 0) {
+		int fd;
+		__u64 netns_a_id;
+		char buf;
+
+		close(pipefd[0]);
+
+		/* Create user namespace A */
+		if (setup_userns() < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		/* Create network namespace owned by user namespace A */
+		if (unshare(CLONE_NEWNET) < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		/* Get network namespace ID */
+		fd = open("/proc/self/ns/net", O_RDONLY);
+		if (fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (ioctl(fd, NS_GET_ID, &netns_a_id) < 0) {
+			close(fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(fd);
+
+		/* Send namespace ID to parent */
+		write(pipefd[1], &netns_a_id, sizeof(netns_a_id));
+
+		/* Keep alive for sibling to check */
+		read(pipefd[1], &buf, 1);
+		close(pipefd[1]);
+		exit(0);
+	}
+
+	/* Parent reads namespace A ID */
+	close(pipefd[1]);
+	netns_a_id = 0;
+	read(pipefd[0], &netns_a_id, sizeof(netns_a_id));
+
+	TH_LOG("User namespace A created network namespace with ID %llu",
+	       (unsigned long long)netns_a_id);
+
+	/* Fork second child - creates user namespace B */
+	ASSERT_EQ(pipe(pipefd2), 0);
+
+	pid2 = fork();
+	ASSERT_GE(pid2, 0);
+
+	if (pid2 == 0) {
+		struct ns_id_req req = {
+			.size = sizeof(req),
+			.spare = 0,
+			.ns_id = 0,
+			.ns_type = CLONE_NEWNET,
+			.spare2 = 0,
+			.user_ns_id = 0,
+		};
+		__u64 ns_ids[100];
+		ssize_t ret;
+		bool found_sibling_netns;
+
+		close(pipefd[0]);
+		close(pipefd2[0]);
+
+		/* Create user namespace B (sibling to A) */
+		if (setup_userns() < 0) {
+			close(pipefd2[1]);
+			exit(1);
+		}
+
+		/* Try to list all network namespaces */
+		ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+
+		found_sibling_netns = false;
+		if (ret > 0) {
+			for (ssize_t i = 0; i < ret; i++) {
+				if (ns_ids[i] == netns_a_id) {
+					found_sibling_netns = true;
+					break;
+				}
+			}
+		}
+
+		/* We should NOT see the sibling's network namespace */
+		write(pipefd2[1], &found_sibling_netns, sizeof(found_sibling_netns));
+		close(pipefd2[1]);
+		exit(0);
+	}
+
+	/* Parent reads result from second child */
+	close(pipefd2[1]);
+	found_sibling_netns = false;
+	read(pipefd2[0], &found_sibling_netns, sizeof(found_sibling_netns));
+	close(pipefd2[0]);
+
+	/* Signal first child to exit */
+	close(pipefd[0]);
+
+	/* Wait for both children */
+	waitpid(pid2, &status, 0);
+	ASSERT_TRUE(WIFEXITED(status));
+
+	waitpid(pid1, &status, 0);
+	ASSERT_TRUE(WIFEXITED(status));
+
+	/* Second child should NOT have seen first child's namespace */
+	ASSERT_FALSE(found_sibling_netns);
+	TH_LOG("User namespace B correctly could not see sibling namespace A's network namespace");
+}
+
+/*
+ * Test permission checking with LISTNS_CURRENT_USER.
+ * Verify that listing with LISTNS_CURRENT_USER respects permissions.
+ */
+TEST(listns_current_user_permissions)
+{
+	int pipefd[2];
+	pid_t pid;
+	int status;
+	bool success;
+	ssize_t count;
+
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		struct ns_id_req req = {
+			.size = sizeof(req),
+			.spare = 0,
+			.ns_id = 0,
+			.ns_type = 0,
+			.spare2 = 0,
+			.user_ns_id = LISTNS_CURRENT_USER,
+		};
+		__u64 ns_ids[100];
+		ssize_t ret;
+		bool success;
+
+		close(pipefd[0]);
+
+		/* Create user namespace */
+		if (setup_userns() < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		/* Create some namespaces owned by this user namespace */
+		if (unshare(CLONE_NEWNET) < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (unshare(CLONE_NEWUTS) < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		/* List with LISTNS_CURRENT_USER - should see our owned namespaces */
+		ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+
+		success = (ret >= 3);  /* At least user, net, uts */
+		write(pipefd[1], &success, sizeof(success));
+		write(pipefd[1], &ret, sizeof(ret));
+		close(pipefd[1]);
+		exit(0);
+	}
+
+	/* Parent */
+	close(pipefd[1]);
+
+	success = false;
+	count = 0;
+	read(pipefd[0], &success, sizeof(success));
+	read(pipefd[0], &count, sizeof(count));
+	close(pipefd[0]);
+
+	waitpid(pid, &status, 0);
+	ASSERT_TRUE(WIFEXITED(status));
+	ASSERT_EQ(WEXITSTATUS(status), 0);
+
+	ASSERT_TRUE(success);
+	TH_LOG("LISTNS_CURRENT_USER returned %zd namespaces", count);
+}
+
+/*
+ * Test that CAP_SYS_ADMIN in parent user namespace allows seeing
+ * child user namespace's owned namespaces.
+ */
+TEST(listns_parent_userns_cap_sys_admin)
+{
+	int pipefd[2];
+	pid_t pid;
+	int status;
+	bool found_child_userns;
+	ssize_t count;
+
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		int fd;
+		__u64 parent_userns_id;
+		__u64 child_userns_id;
+		struct ns_id_req req;
+		__u64 ns_ids[100];
+		ssize_t ret;
+		bool found_child_userns;
+
+		close(pipefd[0]);
+
+		/* Create parent user namespace - we have CAP_SYS_ADMIN in it */
+		if (setup_userns() < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		/* Get parent user namespace ID */
+		fd = open("/proc/self/ns/user", O_RDONLY);
+		if (fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (ioctl(fd, NS_GET_ID, &parent_userns_id) < 0) {
+			close(fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(fd);
+
+		/* Create child user namespace */
+		if (setup_userns() < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		/* Get child user namespace ID */
+		fd = open("/proc/self/ns/user", O_RDONLY);
+		if (fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (ioctl(fd, NS_GET_ID, &child_userns_id) < 0) {
+			close(fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(fd);
+
+		/* Create namespaces owned by child user namespace */
+		if (unshare(CLONE_NEWNET) < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		/* List namespaces owned by parent user namespace */
+		req.size = sizeof(req);
+		req.spare = 0;
+		req.ns_id = 0;
+		req.ns_type = 0;
+		req.spare2 = 0;
+		req.user_ns_id = parent_userns_id;
+
+		ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+
+		/* Should see child user namespace in the list */
+		found_child_userns = false;
+		if (ret > 0) {
+			for (ssize_t i = 0; i < ret; i++) {
+				if (ns_ids[i] == child_userns_id) {
+					found_child_userns = true;
+					break;
+				}
+			}
+		}
+
+		write(pipefd[1], &found_child_userns, sizeof(found_child_userns));
+		write(pipefd[1], &ret, sizeof(ret));
+		close(pipefd[1]);
+		exit(0);
+	}
+
+	/* Parent */
+	close(pipefd[1]);
+
+	found_child_userns = false;
+	count = 0;
+	read(pipefd[0], &found_child_userns, sizeof(found_child_userns));
+	read(pipefd[0], &count, sizeof(count));
+	close(pipefd[0]);
+
+	waitpid(pid, &status, 0);
+	ASSERT_TRUE(WIFEXITED(status));
+	ASSERT_EQ(WEXITSTATUS(status), 0);
+
+	ASSERT_TRUE(found_child_userns);
+	TH_LOG("Process with CAP_SYS_ADMIN in parent user namespace saw child user namespace (total: %zd)",
+			count);
+}
+
+/*
+ * Test that we can see user namespaces we have CAP_SYS_ADMIN inside of.
+ * This is different from seeing namespaces owned by a user namespace.
+ */
+TEST(listns_cap_sys_admin_inside_userns)
+{
+	int pipefd[2];
+	pid_t pid;
+	int status;
+	bool found_ours;
+
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		int fd;
+		__u64 our_userns_id;
+		struct ns_id_req req;
+		__u64 ns_ids[100];
+		ssize_t ret;
+		bool found_ours;
+
+		close(pipefd[0]);
+
+		/* Create user namespace - we have CAP_SYS_ADMIN inside it */
+		if (setup_userns() < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		/* Get our user namespace ID */
+		fd = open("/proc/self/ns/user", O_RDONLY);
+		if (fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (ioctl(fd, NS_GET_ID, &our_userns_id) < 0) {
+			close(fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(fd);
+
+		/* List all user namespaces globally */
+		req.size = sizeof(req);
+		req.spare = 0;
+		req.ns_id = 0;
+		req.ns_type = CLONE_NEWUSER;
+		req.spare2 = 0;
+		req.user_ns_id = 0;
+
+		ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+
+		/* We should be able to see our own user namespace */
+		found_ours = false;
+		if (ret > 0) {
+			for (ssize_t i = 0; i < ret; i++) {
+				if (ns_ids[i] == our_userns_id) {
+					found_ours = true;
+					break;
+				}
+			}
+		}
+
+		write(pipefd[1], &found_ours, sizeof(found_ours));
+		close(pipefd[1]);
+		exit(0);
+	}
+
+	/* Parent */
+	close(pipefd[1]);
+
+	found_ours = false;
+	read(pipefd[0], &found_ours, sizeof(found_ours));
+	close(pipefd[0]);
+
+	waitpid(pid, &status, 0);
+	ASSERT_TRUE(WIFEXITED(status));
+	ASSERT_EQ(WEXITSTATUS(status), 0);
+
+	ASSERT_TRUE(found_ours);
+	TH_LOG("Process can see user namespace it has CAP_SYS_ADMIN inside of");
+}
+
+/*
+ * Test that dropping CAP_SYS_ADMIN restricts what we can see.
+ */
+TEST(listns_drop_cap_sys_admin)
+{
+	cap_t caps;
+	cap_value_t cap_list[1] = { CAP_SYS_ADMIN };
+
+	/* This test needs to start with CAP_SYS_ADMIN */
+	caps = cap_get_proc();
+	if (!caps) {
+		SKIP(return, "Cannot get capabilities");
+	}
+
+	cap_flag_value_t cap_val;
+	if (cap_get_flag(caps, CAP_SYS_ADMIN, CAP_EFFECTIVE, &cap_val) < 0) {
+		cap_free(caps);
+		SKIP(return, "Cannot check CAP_SYS_ADMIN");
+	}
+
+	if (cap_val != CAP_SET) {
+		cap_free(caps);
+		SKIP(return, "Test needs CAP_SYS_ADMIN to start");
+	}
+	cap_free(caps);
+
+	int pipefd[2];
+	pid_t pid;
+	int status;
+	bool correct;
+	ssize_t count_before, count_after;
+
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		struct ns_id_req req = {
+			.size = sizeof(req),
+			.spare = 0,
+			.ns_id = 0,
+			.ns_type = CLONE_NEWNET,
+			.spare2 = 0,
+			.user_ns_id = LISTNS_CURRENT_USER,
+		};
+		__u64 ns_ids_before[100];
+		ssize_t count_before;
+		__u64 ns_ids_after[100];
+		ssize_t count_after;
+		bool correct;
+
+		close(pipefd[0]);
+
+		/* Create user namespace */
+		if (setup_userns() < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		/* Count namespaces with CAP_SYS_ADMIN */
+		count_before = sys_listns(&req, ns_ids_before, ARRAY_SIZE(ns_ids_before), 0);
+
+		/* Drop CAP_SYS_ADMIN */
+		caps = cap_get_proc();
+		if (caps) {
+			cap_set_flag(caps, CAP_EFFECTIVE, 1, cap_list, CAP_CLEAR);
+			cap_set_flag(caps, CAP_PERMITTED, 1, cap_list, CAP_CLEAR);
+			cap_set_proc(caps);
+			cap_free(caps);
+		}
+
+		/* Ensure we can't regain the capability */
+		prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0);
+
+		/* Count namespaces without CAP_SYS_ADMIN */
+		count_after = sys_listns(&req, ns_ids_after, ARRAY_SIZE(ns_ids_after), 0);
+
+		/* Without CAP_SYS_ADMIN, we should see same or fewer namespaces */
+		correct = (count_after <= count_before);
+
+		write(pipefd[1], &correct, sizeof(correct));
+		write(pipefd[1], &count_before, sizeof(count_before));
+		write(pipefd[1], &count_after, sizeof(count_after));
+		close(pipefd[1]);
+		exit(0);
+	}
+
+	/* Parent */
+	close(pipefd[1]);
+
+	correct = false;
+	count_before = 0;
+	count_after = 0;
+	read(pipefd[0], &correct, sizeof(correct));
+	read(pipefd[0], &count_before, sizeof(count_before));
+	read(pipefd[0], &count_after, sizeof(count_after));
+	close(pipefd[0]);
+
+	waitpid(pid, &status, 0);
+	ASSERT_TRUE(WIFEXITED(status));
+	ASSERT_EQ(WEXITSTATUS(status), 0);
+
+	ASSERT_TRUE(correct);
+	TH_LOG("With CAP_SYS_ADMIN: %zd namespaces, without: %zd namespaces",
+			count_before, count_after);
+}
+
+TEST_HARNESS_MAIN

linux-7.2.3/tools/testing/selftests/namespaces/listns_test.c

@@ -0,0 +1,679 @@
+// SPDX-License-Identifier: GPL-2.0
+#define _GNU_SOURCE
+#include <errno.h>
+#include <fcntl.h>
+#include <limits.h>
+#include <sched.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <linux/nsfs.h>
+#include <sys/ioctl.h>
+#include <sys/socket.h>
+#include <sys/stat.h>
+#include <sys/syscall.h>
+#include <sys/types.h>
+#include <sys/wait.h>
+#include <unistd.h>
+#include "../kselftest_harness.h"
+#include "../filesystems/utils.h"
+#include "wrappers.h"
+
+/*
+ * Test basic listns() functionality with the unified namespace tree.
+ * List all active namespaces globally.
+ */
+TEST(listns_basic_unified)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = 0,  /* All types */
+		.spare2 = 0,
+		.user_ns_id = 0,  /* Global listing */
+	};
+	__u64 ns_ids[100];
+	ssize_t ret;
+
+	ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+	if (ret < 0) {
+		if (errno == ENOSYS)
+			SKIP(return, "listns() not supported");
+		TH_LOG("listns failed: %s (errno=%d)", strerror(errno), errno);
+		ASSERT_TRUE(false);
+	}
+
+	/* Should find at least the initial namespaces */
+	ASSERT_GT(ret, 0);
+	TH_LOG("Found %zd active namespaces", ret);
+
+	/* Verify all returned IDs are non-zero */
+	for (ssize_t i = 0; i < ret; i++) {
+		ASSERT_NE(ns_ids[i], 0);
+		TH_LOG("  [%zd] ns_id: %llu", i, (unsigned long long)ns_ids[i]);
+	}
+}
+
+/*
+ * Test listns() with type filtering.
+ * List only network namespaces.
+ */
+TEST(listns_filter_by_type)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = CLONE_NEWNET,  /* Only network namespaces */
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 ns_ids[100];
+	ssize_t ret;
+
+	ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+	if (ret < 0) {
+		if (errno == ENOSYS)
+			SKIP(return, "listns() not supported");
+		TH_LOG("listns failed: %s (errno=%d)", strerror(errno), errno);
+		ASSERT_TRUE(false);
+	}
+	ASSERT_GE(ret, 0);
+
+	/* Should find at least init_net */
+	ASSERT_GT(ret, 0);
+	TH_LOG("Found %zd active network namespaces", ret);
+
+	/* Verify we can open each namespace and it's actually a network namespace */
+	for (ssize_t i = 0; i < ret && i < 5; i++) {
+		struct nsfs_file_handle nsfh = {
+			.ns_id = ns_ids[i],
+			.ns_type = CLONE_NEWNET,
+			.ns_inum = 0,
+		};
+		struct file_handle *fh;
+		int fd;
+
+		fh = (struct file_handle *)malloc(sizeof(*fh) + sizeof(nsfh));
+		ASSERT_NE(fh, NULL);
+		fh->handle_bytes = sizeof(nsfh);
+		fh->handle_type = 0;
+		memcpy(fh->f_handle, &nsfh, sizeof(nsfh));
+
+		fd = open_by_handle_at(-10003, fh, O_RDONLY);
+		free(fh);
+
+		if (fd >= 0) {
+			int ns_type;
+			/* Verify it's a network namespace via ioctl */
+			ns_type = ioctl(fd, NS_GET_NSTYPE);
+			if (ns_type >= 0) {
+				ASSERT_EQ(ns_type, CLONE_NEWNET);
+			}
+			close(fd);
+		}
+	}
+}
+
+/*
+ * Test listns() pagination.
+ * List namespaces in batches.
+ */
+TEST(listns_pagination)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = 0,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 batch1[2], batch2[2];
+	ssize_t ret1, ret2;
+
+	/* Get first batch */
+	ret1 = sys_listns(&req, batch1, ARRAY_SIZE(batch1), 0);
+	if (ret1 < 0) {
+		if (errno == ENOSYS)
+			SKIP(return, "listns() not supported");
+		TH_LOG("listns failed: %s (errno=%d)", strerror(errno), errno);
+		ASSERT_TRUE(false);
+	}
+	ASSERT_GE(ret1, 0);
+
+	if (ret1 == 0)
+		SKIP(return, "No namespaces found");
+
+	TH_LOG("First batch: %zd namespaces", ret1);
+
+	/* Get second batch using last ID from first batch */
+	if (ret1 == ARRAY_SIZE(batch1)) {
+		req.ns_id = batch1[ret1 - 1];
+		ret2 = sys_listns(&req, batch2, ARRAY_SIZE(batch2), 0);
+		ASSERT_GE(ret2, 0);
+
+		TH_LOG("Second batch: %zd namespaces (after ns_id=%llu)",
+		       ret2, (unsigned long long)req.ns_id);
+
+		/* If we got more results, verify IDs are monotonically increasing */
+		if (ret2 > 0) {
+			ASSERT_GT(batch2[0], batch1[ret1 - 1]);
+			TH_LOG("Pagination working: %llu > %llu",
+			       (unsigned long long)batch2[0],
+			       (unsigned long long)batch1[ret1 - 1]);
+		}
+	} else {
+		TH_LOG("All namespaces fit in first batch");
+	}
+}
+
+/*
+ * Test listns() with LISTNS_CURRENT_USER.
+ * List namespaces owned by current user namespace.
+ */
+TEST(listns_current_user)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = 0,
+		.spare2 = 0,
+		.user_ns_id = LISTNS_CURRENT_USER,
+	};
+	__u64 ns_ids[100];
+	ssize_t ret;
+
+	ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+	if (ret < 0) {
+		if (errno == ENOSYS)
+			SKIP(return, "listns() not supported");
+		TH_LOG("listns failed: %s (errno=%d)", strerror(errno), errno);
+		ASSERT_TRUE(false);
+	}
+	ASSERT_GE(ret, 0);
+
+	/* Should find at least the initial namespaces if we're in init_user_ns */
+	TH_LOG("Found %zd namespaces owned by current user namespace", ret);
+
+	for (ssize_t i = 0; i < ret; i++)
+		TH_LOG("  [%zd] ns_id: %llu", i, (unsigned long long)ns_ids[i]);
+}
+
+/*
+ * Test that listns() only returns active namespaces.
+ * Create a namespace, let it become inactive, verify it's not listed.
+ */
+TEST(listns_only_active)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = CLONE_NEWNET,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 ns_ids_before[100], ns_ids_after[100];
+	ssize_t ret_before, ret_after;
+	int pipefd[2];
+	pid_t pid;
+	__u64 new_ns_id = 0;
+	int status;
+
+	/* Get initial list */
+	ret_before = sys_listns(&req, ns_ids_before, ARRAY_SIZE(ns_ids_before), 0);
+	if (ret_before < 0) {
+		if (errno == ENOSYS)
+			SKIP(return, "listns() not supported");
+		TH_LOG("listns failed: %s (errno=%d)", strerror(errno), errno);
+		ASSERT_TRUE(false);
+	}
+	ASSERT_GE(ret_before, 0);
+
+	TH_LOG("Before: %zd active network namespaces", ret_before);
+
+	/* Create a new namespace in a child process and get its ID */
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		int fd;
+		__u64 ns_id;
+
+		close(pipefd[0]);
+
+		/* Create new network namespace */
+		if (unshare(CLONE_NEWNET) < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		/* Get its ID */
+		fd = open("/proc/self/ns/net", O_RDONLY);
+		if (fd < 0) {
+			close(pipefd[1]);
+			exit(1);
+		}
+
+		if (ioctl(fd, NS_GET_ID, &ns_id) < 0) {
+			close(fd);
+			close(pipefd[1]);
+			exit(1);
+		}
+		close(fd);
+
+		/* Send ID to parent */
+		write(pipefd[1], &ns_id, sizeof(ns_id));
+		close(pipefd[1]);
+
+		/* Keep namespace active briefly */
+		usleep(100000);
+		exit(0);
+	}
+
+	/* Parent reads the new namespace ID */
+	{
+		int bytes;
+
+		close(pipefd[1]);
+		bytes = read(pipefd[0], &new_ns_id, sizeof(new_ns_id));
+		close(pipefd[0]);
+
+		if (bytes == sizeof(new_ns_id)) {
+			__u64 ns_ids_during[100];
+			int ret_during;
+
+			TH_LOG("Child created namespace with ID %llu", (unsigned long long)new_ns_id);
+
+			/* List namespaces while child is still alive - should see new one */
+			ret_during = sys_listns(&req, ns_ids_during, ARRAY_SIZE(ns_ids_during), 0);
+			ASSERT_GE(ret_during, 0);
+			TH_LOG("During: %d active network namespaces", ret_during);
+
+			/* Should have more namespaces than before */
+			ASSERT_GE(ret_during, ret_before);
+		}
+	}
+
+	/* Wait for child to exit */
+	waitpid(pid, &status, 0);
+
+	/* Give time for namespace to become inactive */
+	usleep(100000);
+
+	/* List namespaces after child exits - should not see new one */
+	ret_after = sys_listns(&req, ns_ids_after, ARRAY_SIZE(ns_ids_after), 0);
+	ASSERT_GE(ret_after, 0);
+	TH_LOG("After: %zd active network namespaces", ret_after);
+
+	/* Verify the new namespace ID is not in the after list */
+	if (new_ns_id != 0) {
+		bool found = false;
+
+		for (ssize_t i = 0; i < ret_after; i++) {
+			if (ns_ids_after[i] == new_ns_id) {
+				found = true;
+				break;
+			}
+		}
+		ASSERT_FALSE(found);
+	}
+}
+
+/*
+ * Test listns() with specific user namespace ID.
+ * Create a user namespace and list namespaces it owns.
+ */
+TEST(listns_specific_userns)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = 0,
+		.spare2 = 0,
+		.user_ns_id = 0,  /* Will be filled with created userns ID */
+	};
+	__u64 ns_ids[100];
+	int sv[2];
+	pid_t pid;
+	int status;
+	__u64 user_ns_id = 0;
+	int bytes;
+	ssize_t ret;
+
+	ASSERT_EQ(socketpair(AF_UNIX, SOCK_STREAM, 0, sv), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		int fd;
+		__u64 ns_id;
+		char buf;
+
+		close(sv[0]);
+
+		/* Create new user namespace */
+		if (setup_userns() < 0) {
+			close(sv[1]);
+			exit(1);
+		}
+
+		/* Get user namespace ID */
+		fd = open("/proc/self/ns/user", O_RDONLY);
+		if (fd < 0) {
+			close(sv[1]);
+			exit(1);
+		}
+
+		if (ioctl(fd, NS_GET_ID, &ns_id) < 0) {
+			close(fd);
+			close(sv[1]);
+			exit(1);
+		}
+		close(fd);
+
+		/* Send ID to parent */
+		if (write(sv[1], &ns_id, sizeof(ns_id)) != sizeof(ns_id)) {
+			close(sv[1]);
+			exit(1);
+		}
+
+		/* Create some namespaces owned by this user namespace */
+		unshare(CLONE_NEWNET);
+		unshare(CLONE_NEWUTS);
+
+		/* Wait for parent signal */
+		if (read(sv[1], &buf, 1) != 1) {
+			close(sv[1]);
+			exit(1);
+		}
+		close(sv[1]);
+		exit(0);
+	}
+
+	/* Parent */
+	close(sv[1]);
+	bytes = read(sv[0], &user_ns_id, sizeof(user_ns_id));
+
+	if (bytes != sizeof(user_ns_id)) {
+		close(sv[0]);
+		kill(pid, SIGKILL);
+		waitpid(pid, NULL, 0);
+		SKIP(return, "Failed to get user namespace ID from child");
+	}
+
+	TH_LOG("Child created user namespace with ID %llu", (unsigned long long)user_ns_id);
+
+	/* List namespaces owned by this user namespace */
+	req.user_ns_id = user_ns_id;
+	ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+
+	if (ret < 0) {
+		TH_LOG("listns failed: %s (errno=%d)", strerror(errno), errno);
+		close(sv[0]);
+		kill(pid, SIGKILL);
+		waitpid(pid, NULL, 0);
+		if (errno == ENOSYS) {
+			SKIP(return, "listns() not supported");
+		}
+		ASSERT_GE(ret, 0);
+	}
+
+	TH_LOG("Found %zd namespaces owned by user namespace %llu", ret,
+	       (unsigned long long)user_ns_id);
+
+	/* Should find at least the network and UTS namespaces we created */
+	if (ret > 0) {
+		for (ssize_t i = 0; i < ret && i < 10; i++)
+			TH_LOG("  [%zd] ns_id: %llu", i, (unsigned long long)ns_ids[i]);
+	}
+
+	/* Signal child to exit */
+	if (write(sv[0], "X", 1) != 1) {
+		close(sv[0]);
+		kill(pid, SIGKILL);
+		waitpid(pid, NULL, 0);
+		ASSERT_TRUE(false);
+	}
+	close(sv[0]);
+	waitpid(pid, &status, 0);
+}
+
+/*
+ * Test listns() with multiple namespace types filter.
+ */
+TEST(listns_multiple_types)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = CLONE_NEWNET | CLONE_NEWUTS,  /* Network and UTS */
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 ns_ids[100];
+	ssize_t ret;
+
+	ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+	if (ret < 0) {
+		if (errno == ENOSYS)
+			SKIP(return, "listns() not supported");
+		TH_LOG("listns failed: %s (errno=%d)", strerror(errno), errno);
+		ASSERT_TRUE(false);
+	}
+	ASSERT_GE(ret, 0);
+
+	TH_LOG("Found %zd active network/UTS namespaces", ret);
+
+	for (ssize_t i = 0; i < ret; i++)
+		TH_LOG("  [%zd] ns_id: %llu", i, (unsigned long long)ns_ids[i]);
+}
+
+/*
+ * Test that hierarchical active reference propagation keeps parent
+ * user namespaces visible in listns().
+ */
+TEST(listns_hierarchical_visibility)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = CLONE_NEWUSER,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 parent_ns_id = 0, child_ns_id = 0;
+	int sv[2];
+	pid_t pid;
+	int status;
+	int bytes;
+	__u64 ns_ids[100];
+	ssize_t ret;
+	bool found_parent, found_child;
+
+	ASSERT_EQ(socketpair(AF_UNIX, SOCK_STREAM, 0, sv), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		int fd;
+		char buf;
+
+		close(sv[0]);
+
+		/* Create parent user namespace */
+		if (setup_userns() < 0) {
+			close(sv[1]);
+			exit(1);
+		}
+
+		fd = open("/proc/self/ns/user", O_RDONLY);
+		if (fd < 0) {
+			close(sv[1]);
+			exit(1);
+		}
+
+		if (ioctl(fd, NS_GET_ID, &parent_ns_id) < 0) {
+			close(fd);
+			close(sv[1]);
+			exit(1);
+		}
+		close(fd);
+
+		/* Create child user namespace */
+		if (setup_userns() < 0) {
+			close(sv[1]);
+			exit(1);
+		}
+
+		fd = open("/proc/self/ns/user", O_RDONLY);
+		if (fd < 0) {
+			close(sv[1]);
+			exit(1);
+		}
+
+		if (ioctl(fd, NS_GET_ID, &child_ns_id) < 0) {
+			close(fd);
+			close(sv[1]);
+			exit(1);
+		}
+		close(fd);
+
+		/* Send both IDs to parent */
+		if (write(sv[1], &parent_ns_id, sizeof(parent_ns_id)) != sizeof(parent_ns_id)) {
+			close(sv[1]);
+			exit(1);
+		}
+		if (write(sv[1], &child_ns_id, sizeof(child_ns_id)) != sizeof(child_ns_id)) {
+			close(sv[1]);
+			exit(1);
+		}
+
+		/* Wait for parent signal */
+		if (read(sv[1], &buf, 1) != 1) {
+			close(sv[1]);
+			exit(1);
+		}
+		close(sv[1]);
+		exit(0);
+	}
+
+	/* Parent */
+	close(sv[1]);
+
+	/* Read both namespace IDs */
+	bytes = read(sv[0], &parent_ns_id, sizeof(parent_ns_id));
+	bytes += read(sv[0], &child_ns_id, sizeof(child_ns_id));
+
+	if (bytes != (int)(2 * sizeof(__u64))) {
+		close(sv[0]);
+		kill(pid, SIGKILL);
+		waitpid(pid, NULL, 0);
+		SKIP(return, "Failed to get namespace IDs from child");
+	}
+
+	TH_LOG("Parent user namespace ID: %llu", (unsigned long long)parent_ns_id);
+	TH_LOG("Child user namespace ID: %llu", (unsigned long long)child_ns_id);
+
+	/* List all user namespaces */
+	ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+
+	if (ret < 0 && errno == ENOSYS) {
+		close(sv[0]);
+		kill(pid, SIGKILL);
+		waitpid(pid, NULL, 0);
+		SKIP(return, "listns() not supported");
+	}
+
+	ASSERT_GE(ret, 0);
+	TH_LOG("Found %zd active user namespaces", ret);
+
+	/* Both parent and child should be visible (active due to child process) */
+	found_parent = false;
+	found_child = false;
+	for (ssize_t i = 0; i < ret; i++) {
+		if (ns_ids[i] == parent_ns_id)
+			found_parent = true;
+		if (ns_ids[i] == child_ns_id)
+			found_child = true;
+	}
+
+	TH_LOG("Parent namespace %s, child namespace %s",
+	       found_parent ? "found" : "NOT FOUND",
+	       found_child ? "found" : "NOT FOUND");
+
+	ASSERT_TRUE(found_child);
+	/* With hierarchical propagation, parent should also be active */
+	ASSERT_TRUE(found_parent);
+
+	/* Signal child to exit */
+	if (write(sv[0], "X", 1) != 1) {
+		close(sv[0]);
+		kill(pid, SIGKILL);
+		waitpid(pid, NULL, 0);
+		ASSERT_TRUE(false);
+	}
+	close(sv[0]);
+	waitpid(pid, &status, 0);
+}
+
+/*
+ * Test error cases for listns().
+ */
+TEST(listns_error_cases)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = 0,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 ns_ids[10];
+	int ret;
+
+	/* Test with invalid flags */
+	ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0xFFFF);
+	if (errno == ENOSYS) {
+		/* listns() not supported, skip this check */
+	} else {
+		ASSERT_LT(ret, 0);
+		ASSERT_EQ(errno, EINVAL);
+	}
+
+	/* Test with NULL ns_ids array */
+	ret = sys_listns(&req, NULL, 10, 0);
+	ASSERT_LT(ret, 0);
+
+	/* Test with invalid spare field */
+	req.spare = 1;
+	ret = sys_listns(&req, ns_ids, ARRAY_SIZE(ns_ids), 0);
+	if (errno == ENOSYS) {
+		/* listns() not supported, skip this check */
+	} else {
+		ASSERT_LT(ret, 0);
+		ASSERT_EQ(errno, EINVAL);
+	}
+	req.spare = 0;
+
+	/* Test with huge nr_ns_ids */
+	ret = sys_listns(&req, ns_ids, 2000000, 0);
+	if (errno == ENOSYS) {
+		/* listns() not supported, skip this check */
+	} else {
+		ASSERT_LT(ret, 0);
+	}
+}
+
+TEST_HARNESS_MAIN

linux-7.2.3/tools/testing/selftests/namespaces/nsid_test.c

@@ -0,0 +1,993 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include <assert.h>
+#include <fcntl.h>
+#include <inttypes.h>
+#include <libgen.h>
+#include <limits.h>
+#include <pthread.h>
+#include <signal.h>
+#include <string.h>
+#include <sys/mount.h>
+#include <poll.h>
+#include <sys/epoll.h>
+#include <sys/resource.h>
+#include <sys/stat.h>
+#include <sys/socket.h>
+#include <sys/un.h>
+#include <sys/wait.h>
+#include <unistd.h>
+#include <linux/fs.h>
+#include <linux/limits.h>
+#include <linux/nsfs.h>
+#include "kselftest_harness.h"
+
+/* Fixture for tests that create child processes */
+FIXTURE(nsid) {
+	pid_t child_pid;
+	pid_t grandchild_pid;
+};
+
+FIXTURE_SETUP(nsid) {
+	self->child_pid = 0;
+	self->grandchild_pid = 0;
+}
+
+FIXTURE_TEARDOWN(nsid) {
+	/*
+	 * Kill grandchild first: timens_separate and pidns_separate fork a
+	 * grandchild that calls pause().  It is reparented to init on child
+	 * exit and keeps the test runner's tap pipe open, hanging the runner.
+	 */
+	if (self->grandchild_pid > 0) {
+		kill(self->grandchild_pid, SIGKILL);
+		waitpid(self->grandchild_pid, NULL, 0);
+	}
+	if (self->child_pid > 0) {
+		kill(self->child_pid, SIGKILL);
+		waitpid(self->child_pid, NULL, 0);
+	}
+}
+
+TEST(nsid_mntns_basic)
+{
+	__u64 mnt_ns_id = 0;
+	int fd_mntns;
+	int ret;
+
+	/* Open the current mount namespace */
+	fd_mntns = open("/proc/self/ns/mnt", O_RDONLY);
+	ASSERT_GE(fd_mntns, 0);
+
+	/* Get the mount namespace ID */
+	ret = ioctl(fd_mntns, NS_GET_MNTNS_ID, &mnt_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(mnt_ns_id, 0);
+
+	/* Verify we can get the same ID again */
+	__u64 mnt_ns_id2 = 0;
+	ret = ioctl(fd_mntns, NS_GET_ID, &mnt_ns_id2);
+	ASSERT_EQ(ret, 0);
+	ASSERT_EQ(mnt_ns_id, mnt_ns_id2);
+
+	close(fd_mntns);
+}
+
+TEST_F(nsid, mntns_separate)
+{
+	__u64 parent_mnt_ns_id = 0;
+	__u64 child_mnt_ns_id = 0;
+	int fd_parent_mntns, fd_child_mntns;
+	int ret;
+	pid_t pid;
+	int pipefd[2];
+
+	/* Get parent's mount namespace ID */
+	fd_parent_mntns = open("/proc/self/ns/mnt", O_RDONLY);
+	ASSERT_GE(fd_parent_mntns, 0);
+	ret = ioctl(fd_parent_mntns, NS_GET_ID, &parent_mnt_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(parent_mnt_ns_id, 0);
+
+	/* Create a pipe for synchronization */
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		/* Child process */
+		close(pipefd[0]);
+
+		/* Create new mount namespace */
+		ret = unshare(CLONE_NEWNS);
+		if (ret != 0) {
+			/* Skip test if we don't have permission */
+			if (errno == EPERM || errno == EACCES) {
+				write(pipefd[1], "S", 1); /* Signal skip */
+				_exit(0);
+			}
+			_exit(1);
+		}
+
+		/* Signal success */
+		write(pipefd[1], "Y", 1);
+		close(pipefd[1]);
+
+		/* Keep namespace alive */
+		pause();
+		_exit(0);
+	}
+
+	/* Track child for cleanup */
+	self->child_pid = pid;
+
+	/* Parent process */
+	close(pipefd[1]);
+
+	char buf;
+	ASSERT_EQ(read(pipefd[0], &buf, 1), 1);
+	close(pipefd[0]);
+
+	if (buf == 'S') {
+		/* Child couldn't create namespace, skip test */
+		close(fd_parent_mntns);
+		SKIP(return, "No permission to create mount namespace");
+	}
+
+	ASSERT_EQ(buf, 'Y');
+
+	/* Open child's mount namespace */
+	char path[256];
+	snprintf(path, sizeof(path), "/proc/%d/ns/mnt", pid);
+	fd_child_mntns = open(path, O_RDONLY);
+	ASSERT_GE(fd_child_mntns, 0);
+
+	/* Get child's mount namespace ID */
+	ret = ioctl(fd_child_mntns, NS_GET_ID, &child_mnt_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(child_mnt_ns_id, 0);
+
+	/* Parent and child should have different mount namespace IDs */
+	ASSERT_NE(parent_mnt_ns_id, child_mnt_ns_id);
+
+	close(fd_parent_mntns);
+	close(fd_child_mntns);
+}
+
+TEST(nsid_cgroupns_basic)
+{
+	__u64 cgroup_ns_id = 0;
+	int fd_cgroupns;
+	int ret;
+
+	/* Open the current cgroup namespace */
+	fd_cgroupns = open("/proc/self/ns/cgroup", O_RDONLY);
+	ASSERT_GE(fd_cgroupns, 0);
+
+	/* Get the cgroup namespace ID */
+	ret = ioctl(fd_cgroupns, NS_GET_ID, &cgroup_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(cgroup_ns_id, 0);
+
+	/* Verify we can get the same ID again */
+	__u64 cgroup_ns_id2 = 0;
+	ret = ioctl(fd_cgroupns, NS_GET_ID, &cgroup_ns_id2);
+	ASSERT_EQ(ret, 0);
+	ASSERT_EQ(cgroup_ns_id, cgroup_ns_id2);
+
+	close(fd_cgroupns);
+}
+
+TEST_F(nsid, cgroupns_separate)
+{
+	__u64 parent_cgroup_ns_id = 0;
+	__u64 child_cgroup_ns_id = 0;
+	int fd_parent_cgroupns, fd_child_cgroupns;
+	int ret;
+	pid_t pid;
+	int pipefd[2];
+
+	/* Get parent's cgroup namespace ID */
+	fd_parent_cgroupns = open("/proc/self/ns/cgroup", O_RDONLY);
+	ASSERT_GE(fd_parent_cgroupns, 0);
+	ret = ioctl(fd_parent_cgroupns, NS_GET_ID, &parent_cgroup_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(parent_cgroup_ns_id, 0);
+
+	/* Create a pipe for synchronization */
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		/* Child process */
+		close(pipefd[0]);
+
+		/* Create new cgroup namespace */
+		ret = unshare(CLONE_NEWCGROUP);
+		if (ret != 0) {
+			/* Skip test if we don't have permission */
+			if (errno == EPERM || errno == EACCES) {
+				write(pipefd[1], "S", 1); /* Signal skip */
+				_exit(0);
+			}
+			_exit(1);
+		}
+
+		/* Signal success */
+		write(pipefd[1], "Y", 1);
+		close(pipefd[1]);
+
+		/* Keep namespace alive */
+		pause();
+		_exit(0);
+	}
+
+	/* Track child for cleanup */
+	self->child_pid = pid;
+
+	/* Parent process */
+	close(pipefd[1]);
+
+	char buf;
+	ASSERT_EQ(read(pipefd[0], &buf, 1), 1);
+	close(pipefd[0]);
+
+	if (buf == 'S') {
+		/* Child couldn't create namespace, skip test */
+		close(fd_parent_cgroupns);
+		SKIP(return, "No permission to create cgroup namespace");
+	}
+
+	ASSERT_EQ(buf, 'Y');
+
+	/* Open child's cgroup namespace */
+	char path[256];
+	snprintf(path, sizeof(path), "/proc/%d/ns/cgroup", pid);
+	fd_child_cgroupns = open(path, O_RDONLY);
+	ASSERT_GE(fd_child_cgroupns, 0);
+
+	/* Get child's cgroup namespace ID */
+	ret = ioctl(fd_child_cgroupns, NS_GET_ID, &child_cgroup_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(child_cgroup_ns_id, 0);
+
+	/* Parent and child should have different cgroup namespace IDs */
+	ASSERT_NE(parent_cgroup_ns_id, child_cgroup_ns_id);
+
+	close(fd_parent_cgroupns);
+	close(fd_child_cgroupns);
+}
+
+TEST(nsid_ipcns_basic)
+{
+	__u64 ipc_ns_id = 0;
+	int fd_ipcns;
+	int ret;
+
+	/* Open the current IPC namespace */
+	fd_ipcns = open("/proc/self/ns/ipc", O_RDONLY);
+	ASSERT_GE(fd_ipcns, 0);
+
+	/* Get the IPC namespace ID */
+	ret = ioctl(fd_ipcns, NS_GET_ID, &ipc_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(ipc_ns_id, 0);
+
+	/* Verify we can get the same ID again */
+	__u64 ipc_ns_id2 = 0;
+	ret = ioctl(fd_ipcns, NS_GET_ID, &ipc_ns_id2);
+	ASSERT_EQ(ret, 0);
+	ASSERT_EQ(ipc_ns_id, ipc_ns_id2);
+
+	close(fd_ipcns);
+}
+
+TEST_F(nsid, ipcns_separate)
+{
+	__u64 parent_ipc_ns_id = 0;
+	__u64 child_ipc_ns_id = 0;
+	int fd_parent_ipcns, fd_child_ipcns;
+	int ret;
+	pid_t pid;
+	int pipefd[2];
+
+	/* Get parent's IPC namespace ID */
+	fd_parent_ipcns = open("/proc/self/ns/ipc", O_RDONLY);
+	ASSERT_GE(fd_parent_ipcns, 0);
+	ret = ioctl(fd_parent_ipcns, NS_GET_ID, &parent_ipc_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(parent_ipc_ns_id, 0);
+
+	/* Create a pipe for synchronization */
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		/* Child process */
+		close(pipefd[0]);
+
+		/* Create new IPC namespace */
+		ret = unshare(CLONE_NEWIPC);
+		if (ret != 0) {
+			/* Skip test if we don't have permission */
+			if (errno == EPERM || errno == EACCES) {
+				write(pipefd[1], "S", 1); /* Signal skip */
+				_exit(0);
+			}
+			_exit(1);
+		}
+
+		/* Signal success */
+		write(pipefd[1], "Y", 1);
+		close(pipefd[1]);
+
+		/* Keep namespace alive */
+		pause();
+		_exit(0);
+	}
+
+	/* Track child for cleanup */
+	self->child_pid = pid;
+
+	/* Parent process */
+	close(pipefd[1]);
+
+	char buf;
+	ASSERT_EQ(read(pipefd[0], &buf, 1), 1);
+	close(pipefd[0]);
+
+	if (buf == 'S') {
+		/* Child couldn't create namespace, skip test */
+		close(fd_parent_ipcns);
+		SKIP(return, "No permission to create IPC namespace");
+	}
+
+	ASSERT_EQ(buf, 'Y');
+
+	/* Open child's IPC namespace */
+	char path[256];
+	snprintf(path, sizeof(path), "/proc/%d/ns/ipc", pid);
+	fd_child_ipcns = open(path, O_RDONLY);
+	ASSERT_GE(fd_child_ipcns, 0);
+
+	/* Get child's IPC namespace ID */
+	ret = ioctl(fd_child_ipcns, NS_GET_ID, &child_ipc_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(child_ipc_ns_id, 0);
+
+	/* Parent and child should have different IPC namespace IDs */
+	ASSERT_NE(parent_ipc_ns_id, child_ipc_ns_id);
+
+	close(fd_parent_ipcns);
+	close(fd_child_ipcns);
+}
+
+TEST(nsid_utsns_basic)
+{
+	__u64 uts_ns_id = 0;
+	int fd_utsns;
+	int ret;
+
+	/* Open the current UTS namespace */
+	fd_utsns = open("/proc/self/ns/uts", O_RDONLY);
+	ASSERT_GE(fd_utsns, 0);
+
+	/* Get the UTS namespace ID */
+	ret = ioctl(fd_utsns, NS_GET_ID, &uts_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(uts_ns_id, 0);
+
+	/* Verify we can get the same ID again */
+	__u64 uts_ns_id2 = 0;
+	ret = ioctl(fd_utsns, NS_GET_ID, &uts_ns_id2);
+	ASSERT_EQ(ret, 0);
+	ASSERT_EQ(uts_ns_id, uts_ns_id2);
+
+	close(fd_utsns);
+}
+
+TEST_F(nsid, utsns_separate)
+{
+	__u64 parent_uts_ns_id = 0;
+	__u64 child_uts_ns_id = 0;
+	int fd_parent_utsns, fd_child_utsns;
+	int ret;
+	pid_t pid;
+	int pipefd[2];
+
+	/* Get parent's UTS namespace ID */
+	fd_parent_utsns = open("/proc/self/ns/uts", O_RDONLY);
+	ASSERT_GE(fd_parent_utsns, 0);
+	ret = ioctl(fd_parent_utsns, NS_GET_ID, &parent_uts_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(parent_uts_ns_id, 0);
+
+	/* Create a pipe for synchronization */
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		/* Child process */
+		close(pipefd[0]);
+
+		/* Create new UTS namespace */
+		ret = unshare(CLONE_NEWUTS);
+		if (ret != 0) {
+			/* Skip test if we don't have permission */
+			if (errno == EPERM || errno == EACCES) {
+				write(pipefd[1], "S", 1); /* Signal skip */
+				_exit(0);
+			}
+			_exit(1);
+		}
+
+		/* Signal success */
+		write(pipefd[1], "Y", 1);
+		close(pipefd[1]);
+
+		/* Keep namespace alive */
+		pause();
+		_exit(0);
+	}
+
+	/* Track child for cleanup */
+	self->child_pid = pid;
+
+	/* Parent process */
+	close(pipefd[1]);
+
+	char buf;
+	ASSERT_EQ(read(pipefd[0], &buf, 1), 1);
+	close(pipefd[0]);
+
+	if (buf == 'S') {
+		/* Child couldn't create namespace, skip test */
+		close(fd_parent_utsns);
+		SKIP(return, "No permission to create UTS namespace");
+	}
+
+	ASSERT_EQ(buf, 'Y');
+
+	/* Open child's UTS namespace */
+	char path[256];
+	snprintf(path, sizeof(path), "/proc/%d/ns/uts", pid);
+	fd_child_utsns = open(path, O_RDONLY);
+	ASSERT_GE(fd_child_utsns, 0);
+
+	/* Get child's UTS namespace ID */
+	ret = ioctl(fd_child_utsns, NS_GET_ID, &child_uts_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(child_uts_ns_id, 0);
+
+	/* Parent and child should have different UTS namespace IDs */
+	ASSERT_NE(parent_uts_ns_id, child_uts_ns_id);
+
+	close(fd_parent_utsns);
+	close(fd_child_utsns);
+}
+
+TEST(nsid_userns_basic)
+{
+	__u64 user_ns_id = 0;
+	int fd_userns;
+	int ret;
+
+	/* Open the current user namespace */
+	fd_userns = open("/proc/self/ns/user", O_RDONLY);
+	ASSERT_GE(fd_userns, 0);
+
+	/* Get the user namespace ID */
+	ret = ioctl(fd_userns, NS_GET_ID, &user_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(user_ns_id, 0);
+
+	/* Verify we can get the same ID again */
+	__u64 user_ns_id2 = 0;
+	ret = ioctl(fd_userns, NS_GET_ID, &user_ns_id2);
+	ASSERT_EQ(ret, 0);
+	ASSERT_EQ(user_ns_id, user_ns_id2);
+
+	close(fd_userns);
+}
+
+TEST_F(nsid, userns_separate)
+{
+	__u64 parent_user_ns_id = 0;
+	__u64 child_user_ns_id = 0;
+	int fd_parent_userns, fd_child_userns;
+	int ret;
+	pid_t pid;
+	int pipefd[2];
+
+	/* Get parent's user namespace ID */
+	fd_parent_userns = open("/proc/self/ns/user", O_RDONLY);
+	ASSERT_GE(fd_parent_userns, 0);
+	ret = ioctl(fd_parent_userns, NS_GET_ID, &parent_user_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(parent_user_ns_id, 0);
+
+	/* Create a pipe for synchronization */
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		/* Child process */
+		close(pipefd[0]);
+
+		/* Create new user namespace */
+		ret = unshare(CLONE_NEWUSER);
+		if (ret != 0) {
+			/* Skip test if we don't have permission */
+			if (errno == EPERM || errno == EACCES) {
+				write(pipefd[1], "S", 1); /* Signal skip */
+				_exit(0);
+			}
+			_exit(1);
+		}
+
+		/* Signal success */
+		write(pipefd[1], "Y", 1);
+		close(pipefd[1]);
+
+		/* Keep namespace alive */
+		pause();
+		_exit(0);
+	}
+
+	/* Track child for cleanup */
+	self->child_pid = pid;
+
+	/* Parent process */
+	close(pipefd[1]);
+
+	char buf;
+	ASSERT_EQ(read(pipefd[0], &buf, 1), 1);
+	close(pipefd[0]);
+
+	if (buf == 'S') {
+		/* Child couldn't create namespace, skip test */
+		close(fd_parent_userns);
+		SKIP(return, "No permission to create user namespace");
+	}
+
+	ASSERT_EQ(buf, 'Y');
+
+	/* Open child's user namespace */
+	char path[256];
+	snprintf(path, sizeof(path), "/proc/%d/ns/user", pid);
+	fd_child_userns = open(path, O_RDONLY);
+	ASSERT_GE(fd_child_userns, 0);
+
+	/* Get child's user namespace ID */
+	ret = ioctl(fd_child_userns, NS_GET_ID, &child_user_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(child_user_ns_id, 0);
+
+	/* Parent and child should have different user namespace IDs */
+	ASSERT_NE(parent_user_ns_id, child_user_ns_id);
+
+	close(fd_parent_userns);
+	close(fd_child_userns);
+}
+
+TEST(nsid_timens_basic)
+{
+	__u64 time_ns_id = 0;
+	int fd_timens;
+	int ret;
+
+	/* Open the current time namespace */
+	fd_timens = open("/proc/self/ns/time", O_RDONLY);
+	if (fd_timens < 0) {
+		SKIP(return, "Time namespaces not supported");
+	}
+
+	/* Get the time namespace ID */
+	ret = ioctl(fd_timens, NS_GET_ID, &time_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(time_ns_id, 0);
+
+	/* Verify we can get the same ID again */
+	__u64 time_ns_id2 = 0;
+	ret = ioctl(fd_timens, NS_GET_ID, &time_ns_id2);
+	ASSERT_EQ(ret, 0);
+	ASSERT_EQ(time_ns_id, time_ns_id2);
+
+	close(fd_timens);
+}
+
+TEST_F(nsid, timens_separate)
+{
+	__u64 parent_time_ns_id = 0;
+	__u64 child_time_ns_id = 0;
+	int fd_parent_timens, fd_child_timens;
+	int ret;
+	pid_t pid;
+	int pipefd[2];
+
+	/* Open the current time namespace */
+	fd_parent_timens = open("/proc/self/ns/time", O_RDONLY);
+	if (fd_parent_timens < 0) {
+		SKIP(return, "Time namespaces not supported");
+	}
+
+	/* Get parent's time namespace ID */
+	ret = ioctl(fd_parent_timens, NS_GET_ID, &parent_time_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(parent_time_ns_id, 0);
+
+	/* Create a pipe for synchronization */
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		/* Child process */
+		close(pipefd[0]);
+
+		/* Create new time namespace */
+		ret = unshare(CLONE_NEWTIME);
+		if (ret != 0) {
+			/* Skip test if we don't have permission */
+			if (errno == EPERM || errno == EACCES || errno == EINVAL) {
+				write(pipefd[1], "S", 1); /* Signal skip */
+				_exit(0);
+			}
+			_exit(1);
+		}
+
+		/* Fork a grandchild to actually enter the new namespace */
+		pid_t grandchild = fork();
+		if (grandchild == 0) {
+			/* Grandchild is in the new namespace */
+			write(pipefd[1], "Y", 1);
+			close(pipefd[1]);
+			pause();
+			_exit(0);
+		} else if (grandchild > 0) {
+			/* Child writes grandchild PID and waits */
+			write(pipefd[1], "Y", 1);
+			write(pipefd[1], &grandchild, sizeof(grandchild));
+			close(pipefd[1]);
+			pause(); /* Keep the parent alive to maintain the grandchild */
+			_exit(0);
+		} else {
+			_exit(1);
+		}
+	}
+
+	/* Track child for cleanup */
+	self->child_pid = pid;
+
+	/* Parent process */
+	close(pipefd[1]);
+
+	char buf;
+	ASSERT_EQ(read(pipefd[0], &buf, 1), 1);
+
+	if (buf == 'S') {
+		/* Child couldn't create namespace, skip test */
+		close(fd_parent_timens);
+		close(pipefd[0]);
+		SKIP(return, "Cannot create time namespace");
+	}
+
+	ASSERT_EQ(buf, 'Y');
+
+	pid_t grandchild_pid;
+	ASSERT_EQ(read(pipefd[0], &grandchild_pid, sizeof(grandchild_pid)), sizeof(grandchild_pid));
+	self->grandchild_pid = grandchild_pid;
+	close(pipefd[0]);
+
+	/* Open grandchild's time namespace */
+	char path[256];
+	snprintf(path, sizeof(path), "/proc/%d/ns/time", grandchild_pid);
+	fd_child_timens = open(path, O_RDONLY);
+	ASSERT_GE(fd_child_timens, 0);
+
+	/* Get child's time namespace ID */
+	ret = ioctl(fd_child_timens, NS_GET_ID, &child_time_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(child_time_ns_id, 0);
+
+	/* Parent and child should have different time namespace IDs */
+	ASSERT_NE(parent_time_ns_id, child_time_ns_id);
+
+	close(fd_parent_timens);
+	close(fd_child_timens);
+}
+
+TEST(nsid_pidns_basic)
+{
+	__u64 pid_ns_id = 0;
+	int fd_pidns;
+	int ret;
+
+	/* Open the current PID namespace */
+	fd_pidns = open("/proc/self/ns/pid", O_RDONLY);
+	ASSERT_GE(fd_pidns, 0);
+
+	/* Get the PID namespace ID */
+	ret = ioctl(fd_pidns, NS_GET_ID, &pid_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(pid_ns_id, 0);
+
+	/* Verify we can get the same ID again */
+	__u64 pid_ns_id2 = 0;
+	ret = ioctl(fd_pidns, NS_GET_ID, &pid_ns_id2);
+	ASSERT_EQ(ret, 0);
+	ASSERT_EQ(pid_ns_id, pid_ns_id2);
+
+	close(fd_pidns);
+}
+
+TEST_F(nsid, pidns_separate)
+{
+	__u64 parent_pid_ns_id = 0;
+	__u64 child_pid_ns_id = 0;
+	int fd_parent_pidns, fd_child_pidns;
+	int ret;
+	pid_t pid;
+	int pipefd[2];
+
+	/* Get parent's PID namespace ID */
+	fd_parent_pidns = open("/proc/self/ns/pid", O_RDONLY);
+	ASSERT_GE(fd_parent_pidns, 0);
+	ret = ioctl(fd_parent_pidns, NS_GET_ID, &parent_pid_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(parent_pid_ns_id, 0);
+
+	/* Create a pipe for synchronization */
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		/* Child process */
+		close(pipefd[0]);
+
+		/* Create new PID namespace */
+		ret = unshare(CLONE_NEWPID);
+		if (ret != 0) {
+			/* Skip test if we don't have permission */
+			if (errno == EPERM || errno == EACCES) {
+				write(pipefd[1], "S", 1); /* Signal skip */
+				_exit(0);
+			}
+			_exit(1);
+		}
+
+		/* Fork a grandchild to actually enter the new namespace */
+		pid_t grandchild = fork();
+		if (grandchild == 0) {
+			/* Grandchild is in the new namespace */
+			write(pipefd[1], "Y", 1);
+			close(pipefd[1]);
+			pause();
+			_exit(0);
+		} else if (grandchild > 0) {
+			/* Child writes grandchild PID and waits */
+			write(pipefd[1], "Y", 1);
+			write(pipefd[1], &grandchild, sizeof(grandchild));
+			close(pipefd[1]);
+			pause(); /* Keep the parent alive to maintain the grandchild */
+			_exit(0);
+		} else {
+			_exit(1);
+		}
+	}
+
+	/* Track child for cleanup */
+	self->child_pid = pid;
+
+	/* Parent process */
+	close(pipefd[1]);
+
+	char buf;
+	ASSERT_EQ(read(pipefd[0], &buf, 1), 1);
+
+	if (buf == 'S') {
+		/* Child couldn't create namespace, skip test */
+		close(fd_parent_pidns);
+		close(pipefd[0]);
+		SKIP(return, "No permission to create PID namespace");
+	}
+
+	ASSERT_EQ(buf, 'Y');
+
+	pid_t grandchild_pid;
+	ASSERT_EQ(read(pipefd[0], &grandchild_pid, sizeof(grandchild_pid)), sizeof(grandchild_pid));
+	self->grandchild_pid = grandchild_pid;
+	close(pipefd[0]);
+
+	/* Open grandchild's PID namespace */
+	char path[256];
+	snprintf(path, sizeof(path), "/proc/%d/ns/pid", grandchild_pid);
+	fd_child_pidns = open(path, O_RDONLY);
+	ASSERT_GE(fd_child_pidns, 0);
+
+	/* Get child's PID namespace ID */
+	ret = ioctl(fd_child_pidns, NS_GET_ID, &child_pid_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(child_pid_ns_id, 0);
+
+	/* Parent and child should have different PID namespace IDs */
+	ASSERT_NE(parent_pid_ns_id, child_pid_ns_id);
+
+	close(fd_parent_pidns);
+	close(fd_child_pidns);
+}
+
+TEST(nsid_netns_basic)
+{
+	__u64 net_ns_id = 0;
+	__u64 netns_cookie = 0;
+	int fd_netns;
+	int sock;
+	socklen_t optlen;
+	int ret;
+
+	/* Open the current network namespace */
+	fd_netns = open("/proc/self/ns/net", O_RDONLY);
+	ASSERT_GE(fd_netns, 0);
+
+	/* Get the network namespace ID via ioctl */
+	ret = ioctl(fd_netns, NS_GET_ID, &net_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(net_ns_id, 0);
+
+	/* Create a socket to get the SO_NETNS_COOKIE */
+	sock = socket(AF_UNIX, SOCK_STREAM, 0);
+	ASSERT_GE(sock, 0);
+
+	/* Get the network namespace cookie via socket option */
+	optlen = sizeof(netns_cookie);
+	ret = getsockopt(sock, SOL_SOCKET, SO_NETNS_COOKIE, &netns_cookie, &optlen);
+	ASSERT_EQ(ret, 0);
+	ASSERT_EQ(optlen, sizeof(netns_cookie));
+
+	/* The namespace ID and cookie should be identical */
+	ASSERT_EQ(net_ns_id, netns_cookie);
+
+	/* Verify we can get the same ID again */
+	__u64 net_ns_id2 = 0;
+	ret = ioctl(fd_netns, NS_GET_ID, &net_ns_id2);
+	ASSERT_EQ(ret, 0);
+	ASSERT_EQ(net_ns_id, net_ns_id2);
+
+	close(sock);
+	close(fd_netns);
+}
+
+TEST_F(nsid, netns_separate)
+{
+	__u64 parent_net_ns_id = 0;
+	__u64 parent_netns_cookie = 0;
+	__u64 child_net_ns_id = 0;
+	__u64 child_netns_cookie = 0;
+	int fd_parent_netns, fd_child_netns;
+	int parent_sock, child_sock;
+	socklen_t optlen;
+	int ret;
+	pid_t pid;
+	int pipefd[2];
+
+	/* Get parent's network namespace ID */
+	fd_parent_netns = open("/proc/self/ns/net", O_RDONLY);
+	ASSERT_GE(fd_parent_netns, 0);
+	ret = ioctl(fd_parent_netns, NS_GET_ID, &parent_net_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(parent_net_ns_id, 0);
+
+	/* Get parent's network namespace cookie */
+	parent_sock = socket(AF_UNIX, SOCK_STREAM, 0);
+	ASSERT_GE(parent_sock, 0);
+	optlen = sizeof(parent_netns_cookie);
+	ret = getsockopt(parent_sock, SOL_SOCKET, SO_NETNS_COOKIE, &parent_netns_cookie, &optlen);
+	ASSERT_EQ(ret, 0);
+
+	/* Verify parent's ID and cookie match */
+	ASSERT_EQ(parent_net_ns_id, parent_netns_cookie);
+
+	/* Create a pipe for synchronization */
+	ASSERT_EQ(pipe(pipefd), 0);
+
+	pid = fork();
+	ASSERT_GE(pid, 0);
+
+	if (pid == 0) {
+		/* Child process */
+		close(pipefd[0]);
+
+		/* Create new network namespace */
+		ret = unshare(CLONE_NEWNET);
+		if (ret != 0) {
+			/* Skip test if we don't have permission */
+			if (errno == EPERM || errno == EACCES) {
+				write(pipefd[1], "S", 1); /* Signal skip */
+				_exit(0);
+			}
+			_exit(1);
+		}
+
+		/* Signal success */
+		write(pipefd[1], "Y", 1);
+		close(pipefd[1]);
+
+		/* Keep namespace alive */
+		pause();
+		_exit(0);
+	}
+
+	/* Track child for cleanup */
+	self->child_pid = pid;
+
+	/* Parent process */
+	close(pipefd[1]);
+
+	char buf;
+	ASSERT_EQ(read(pipefd[0], &buf, 1), 1);
+	close(pipefd[0]);
+
+	if (buf == 'S') {
+		/* Child couldn't create namespace, skip test */
+		close(fd_parent_netns);
+		close(parent_sock);
+		SKIP(return, "No permission to create network namespace");
+	}
+
+	ASSERT_EQ(buf, 'Y');
+
+	/* Open child's network namespace */
+	char path[256];
+	snprintf(path, sizeof(path), "/proc/%d/ns/net", pid);
+	fd_child_netns = open(path, O_RDONLY);
+	ASSERT_GE(fd_child_netns, 0);
+
+	/* Get child's network namespace ID */
+	ret = ioctl(fd_child_netns, NS_GET_ID, &child_net_ns_id);
+	ASSERT_EQ(ret, 0);
+	ASSERT_NE(child_net_ns_id, 0);
+
+	/* Create socket in child's namespace to get cookie */
+	ret = setns(fd_child_netns, CLONE_NEWNET);
+	if (ret == 0) {
+		child_sock = socket(AF_UNIX, SOCK_STREAM, 0);
+		ASSERT_GE(child_sock, 0);
+
+		optlen = sizeof(child_netns_cookie);
+		ret = getsockopt(child_sock, SOL_SOCKET, SO_NETNS_COOKIE, &child_netns_cookie, &optlen);
+		ASSERT_EQ(ret, 0);
+
+		/* Verify child's ID and cookie match */
+		ASSERT_EQ(child_net_ns_id, child_netns_cookie);
+
+		close(child_sock);
+
+		/* Return to parent namespace */
+		setns(fd_parent_netns, CLONE_NEWNET);
+	}
+
+	/* Parent and child should have different network namespace IDs */
+	ASSERT_NE(parent_net_ns_id, child_net_ns_id);
+	if (child_netns_cookie != 0) {
+		ASSERT_NE(parent_netns_cookie, child_netns_cookie);
+	}
+
+	close(fd_parent_netns);
+	close(fd_child_netns);
+	close(parent_sock);
+}
+
+TEST_HARNESS_MAIN

linux-7.2.3/tools/testing/selftests/namespaces/regression_pidfd_setns_test.c

@@ -0,0 +1,113 @@
+// SPDX-License-Identifier: GPL-2.0
+#define _GNU_SOURCE
+#include <errno.h>
+#include <sched.h>
+#include <signal.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/socket.h>
+#include <unistd.h>
+#include "../pidfd/pidfd.h"
+#include "../kselftest_harness.h"
+
+/*
+ * Regression tests for the setns(pidfd) active reference counting bug.
+ *
+ * These tests are based on the reproducers that triggered the race condition
+ * fixed by commit 1c465d0518dc ("ns: handle setns(pidfd, ...) cleanly").
+ *
+ * The bug: When using setns() with a pidfd, if the target task exits between
+ * prepare_nsset() and commit_nsset(), the namespaces would become inactive.
+ * Then ns_ref_active_get() would increment from 0 without properly resurrecting
+ * the owner chain, causing active reference count underflows.
+ */
+
+/*
+ * Simple pidfd setns test using create_child()+unshare().
+ *
+ * Without the fix, this would trigger active refcount warnings when the
+ * parent exits after doing setns(pidfd) on a child that has already exited.
+ */
+TEST(simple_pidfd_setns)
+{
+	pid_t child_pid;
+	int pidfd = -1;
+	int ret;
+	int sv[2];
+	char c;
+
+	/* Ignore SIGCHLD for autoreap */
+	ASSERT_NE(signal(SIGCHLD, SIG_IGN), SIG_ERR);
+
+	ASSERT_EQ(socketpair(AF_UNIX, SOCK_STREAM, 0, sv), 0);
+
+	/* Create a child process without namespaces initially */
+	child_pid = create_child(&pidfd, 0);
+	ASSERT_GE(child_pid, 0);
+
+	if (child_pid == 0) {
+		close(sv[0]);
+
+		if (unshare(CLONE_NEWUTS | CLONE_NEWIPC | CLONE_NEWNET | CLONE_NEWUSER) < 0) {
+			close(sv[1]);
+			_exit(1);
+		}
+
+		/* Signal parent that namespaces are ready */
+		if (write_nointr(sv[1], "1", 1) < 0) {
+			close(sv[1]);
+			_exit(1);
+		}
+
+		close(sv[1]);
+		_exit(0);
+	}
+	ASSERT_GE(pidfd, 0);
+	EXPECT_EQ(close(sv[1]), 0);
+
+	ret = read_nointr(sv[0], &c, 1);
+	ASSERT_EQ(ret, 1);
+	EXPECT_EQ(close(sv[0]), 0);
+
+	/* Set to child's namespaces via pidfd */
+	ret = setns(pidfd, CLONE_NEWUTS | CLONE_NEWIPC);
+	TH_LOG("setns() returned %d", ret);
+	close(pidfd);
+}
+
+/*
+ * Simple pidfd setns test using create_child().
+ *
+ * This variation uses create_child() with namespace flags directly.
+ * Namespaces are created immediately at clone time.
+ */
+TEST(simple_pidfd_setns_clone)
+{
+	pid_t child_pid;
+	int pidfd = -1;
+	int ret;
+
+	/* Ignore SIGCHLD for autoreap */
+	ASSERT_NE(signal(SIGCHLD, SIG_IGN), SIG_ERR);
+
+	/* Create a child process with new namespaces using create_child() */
+	child_pid = create_child(&pidfd, CLONE_NEWUSER | CLONE_NEWUTS | CLONE_NEWIPC | CLONE_NEWNET);
+	ASSERT_GE(child_pid, 0);
+
+	if (child_pid == 0) {
+		/* Child: sleep for a while so parent can setns to us */
+		sleep(2);
+		_exit(0);
+	}
+
+	/* Parent: pidfd was already created by create_child() */
+	ASSERT_GE(pidfd, 0);
+
+	/* Set to child's namespaces via pidfd */
+	ret = setns(pidfd, CLONE_NEWUTS | CLONE_NEWIPC);
+	close(pidfd);
+	TH_LOG("setns() returned %d", ret);
+}
+
+TEST_HARNESS_MAIN

linux-7.2.3/tools/testing/selftests/namespaces/stress_test.c

@@ -0,0 +1,626 @@
+// SPDX-License-Identifier: GPL-2.0
+#define _GNU_SOURCE
+#include <errno.h>
+#include <fcntl.h>
+#include <limits.h>
+#include <sched.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/ioctl.h>
+#include <sys/socket.h>
+#include <sys/stat.h>
+#include <sys/syscall.h>
+#include <sys/types.h>
+#include <sys/wait.h>
+#include <unistd.h>
+#include <linux/nsfs.h>
+#include "../kselftest_harness.h"
+#include "../filesystems/utils.h"
+#include "wrappers.h"
+
+/*
+ * Stress tests for namespace active reference counting.
+ *
+ * These tests validate that the active reference counting system can handle
+ * high load scenarios including rapid namespace creation/destruction, large
+ * numbers of concurrent namespaces, and various edge cases under stress.
+ */
+
+/*
+ * Test rapid creation and destruction of user namespaces.
+ * Create and destroy namespaces in quick succession to stress the
+ * active reference tracking and ensure no leaks occur.
+ */
+TEST(rapid_namespace_creation_destruction)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = CLONE_NEWUSER,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 ns_ids_before[256], ns_ids_after[256];
+	ssize_t ret_before, ret_after;
+	int i;
+
+	/* Get baseline count of active user namespaces */
+	ret_before = sys_listns(&req, ns_ids_before, ARRAY_SIZE(ns_ids_before), 0);
+	if (ret_before < 0) {
+		if (errno == ENOSYS)
+			SKIP(return, "listns() not supported");
+		ASSERT_GE(ret_before, 0);
+	}
+
+	TH_LOG("Baseline: %zd active user namespaces", ret_before);
+
+	/* Rapidly create and destroy 100 user namespaces */
+	for (i = 0; i < 100; i++) {
+		pid_t pid = fork();
+		ASSERT_GE(pid, 0);
+
+		if (pid == 0) {
+			/* Child: create user namespace and immediately exit */
+			if (setup_userns() < 0)
+				exit(1);
+			exit(0);
+		}
+
+		/* Parent: wait for child */
+		int status;
+		waitpid(pid, &status, 0);
+		ASSERT_TRUE(WIFEXITED(status));
+		ASSERT_EQ(WEXITSTATUS(status), 0);
+	}
+
+	/* Verify we're back to baseline (no leaked namespaces) */
+	ret_after = sys_listns(&req, ns_ids_after, ARRAY_SIZE(ns_ids_after), 0);
+	ASSERT_GE(ret_after, 0);
+
+	TH_LOG("After 100 rapid create/destroy cycles: %zd active user namespaces", ret_after);
+	ASSERT_EQ(ret_before, ret_after);
+}
+
+/*
+ * Test creating many concurrent namespaces.
+ * Verify that listns() correctly tracks all of them and that they all
+ * become inactive after processes exit.
+ */
+TEST(many_concurrent_namespaces)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = CLONE_NEWUSER,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 ns_ids_before[512], ns_ids_during[512], ns_ids_after[512];
+	ssize_t ret_before, ret_during, ret_after;
+	pid_t pids[50];
+	int num_children = 50;
+	int i;
+	int sv[2];
+
+	/* Get baseline */
+	ret_before = sys_listns(&req, ns_ids_before, ARRAY_SIZE(ns_ids_before), 0);
+	if (ret_before < 0) {
+		if (errno == ENOSYS)
+			SKIP(return, "listns() not supported");
+		ASSERT_GE(ret_before, 0);
+	}
+
+	TH_LOG("Baseline: %zd active user namespaces", ret_before);
+
+	ASSERT_EQ(socketpair(AF_UNIX, SOCK_STREAM, 0, sv), 0);
+
+	/* Create many children, each with their own user namespace */
+	for (i = 0; i < num_children; i++) {
+		pids[i] = fork();
+		ASSERT_GE(pids[i], 0);
+
+		if (pids[i] == 0) {
+			/* Child: create user namespace and wait for parent signal */
+			char c;
+
+			close(sv[0]);
+
+			if (setup_userns() < 0) {
+				close(sv[1]);
+				exit(1);
+			}
+
+			/* Signal parent we're ready */
+			if (write(sv[1], &c, 1) != 1) {
+				close(sv[1]);
+				exit(1);
+			}
+
+			/* Wait for parent signal to exit */
+			if (read(sv[1], &c, 1) != 1) {
+				close(sv[1]);
+				exit(1);
+			}
+
+			close(sv[1]);
+			exit(0);
+		}
+	}
+
+	close(sv[1]);
+
+	/* Wait for all children to signal ready */
+	for (i = 0; i < num_children; i++) {
+		char c;
+		if (read(sv[0], &c, 1) != 1) {
+			/* If we fail to read, kill all children and exit */
+			close(sv[0]);
+			for (int j = 0; j < num_children; j++)
+				kill(pids[j], SIGKILL);
+			for (int j = 0; j < num_children; j++)
+				waitpid(pids[j], NULL, 0);
+			ASSERT_TRUE(false);
+		}
+	}
+
+	/* List namespaces while all children are running */
+	ret_during = sys_listns(&req, ns_ids_during, ARRAY_SIZE(ns_ids_during), 0);
+	ASSERT_GE(ret_during, 0);
+
+	TH_LOG("With %d children running: %zd active user namespaces", num_children, ret_during);
+
+	/* Should have at least num_children more namespaces than baseline */
+	ASSERT_GE(ret_during, ret_before + num_children);
+
+	/* Signal all children to exit */
+	for (i = 0; i < num_children; i++) {
+		char c = 'X';
+		if (write(sv[0], &c, 1) != 1) {
+			/* If we fail to write, kill remaining children */
+			close(sv[0]);
+			for (int j = i; j < num_children; j++)
+				kill(pids[j], SIGKILL);
+			for (int j = 0; j < num_children; j++)
+				waitpid(pids[j], NULL, 0);
+			ASSERT_TRUE(false);
+		}
+	}
+
+	close(sv[0]);
+
+	/* Wait for all children */
+	for (i = 0; i < num_children; i++) {
+		int status;
+		waitpid(pids[i], &status, 0);
+		ASSERT_TRUE(WIFEXITED(status));
+	}
+
+	/* Verify we're back to baseline */
+	ret_after = sys_listns(&req, ns_ids_after, ARRAY_SIZE(ns_ids_after), 0);
+	ASSERT_GE(ret_after, 0);
+
+	TH_LOG("After all children exit: %zd active user namespaces", ret_after);
+	ASSERT_EQ(ret_before, ret_after);
+}
+
+/*
+ * Test rapid namespace creation with different namespace types.
+ * Create multiple types of namespaces rapidly to stress the tracking system.
+ */
+TEST(rapid_mixed_namespace_creation)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = 0,  /* All types */
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 ns_ids_before[512], ns_ids_after[512];
+	ssize_t ret_before, ret_after;
+	int i;
+
+	/* Get baseline count */
+	ret_before = sys_listns(&req, ns_ids_before, ARRAY_SIZE(ns_ids_before), 0);
+	if (ret_before < 0) {
+		if (errno == ENOSYS)
+			SKIP(return, "listns() not supported");
+		ASSERT_GE(ret_before, 0);
+	}
+
+	TH_LOG("Baseline: %zd active namespaces (all types)", ret_before);
+
+	/* Rapidly create and destroy namespaces with multiple types */
+	for (i = 0; i < 50; i++) {
+		pid_t pid = fork();
+		ASSERT_GE(pid, 0);
+
+		if (pid == 0) {
+			/* Child: create multiple namespace types */
+			if (setup_userns() < 0)
+				exit(1);
+
+			/* Create additional namespace types */
+			if (unshare(CLONE_NEWNET) < 0)
+				exit(1);
+			if (unshare(CLONE_NEWUTS) < 0)
+				exit(1);
+			if (unshare(CLONE_NEWIPC) < 0)
+				exit(1);
+
+			exit(0);
+		}
+
+		/* Parent: wait for child */
+		int status;
+		waitpid(pid, &status, 0);
+		ASSERT_TRUE(WIFEXITED(status));
+	}
+
+	/* Verify we're back to baseline */
+	ret_after = sys_listns(&req, ns_ids_after, ARRAY_SIZE(ns_ids_after), 0);
+	ASSERT_GE(ret_after, 0);
+
+	TH_LOG("After 50 rapid mixed namespace cycles: %zd active namespaces", ret_after);
+	ASSERT_EQ(ret_before, ret_after);
+}
+
+/*
+ * Test nested namespace creation under stress.
+ * Create deeply nested namespace hierarchies and verify proper cleanup.
+ */
+TEST(nested_namespace_stress)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = CLONE_NEWUSER,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 ns_ids_before[512], ns_ids_after[512];
+	ssize_t ret_before, ret_after;
+	int i;
+
+	/* Get baseline */
+	ret_before = sys_listns(&req, ns_ids_before, ARRAY_SIZE(ns_ids_before), 0);
+	if (ret_before < 0) {
+		if (errno == ENOSYS)
+			SKIP(return, "listns() not supported");
+		ASSERT_GE(ret_before, 0);
+	}
+
+	TH_LOG("Baseline: %zd active user namespaces", ret_before);
+
+	/* Create 20 processes, each with nested user namespaces */
+	for (i = 0; i < 20; i++) {
+		pid_t pid = fork();
+		ASSERT_GE(pid, 0);
+
+		if (pid == 0) {
+			int userns_fd;
+			uid_t orig_uid = getuid();
+			int depth;
+
+			/* Create nested user namespaces (up to 5 levels) */
+			for (depth = 0; depth < 5; depth++) {
+				userns_fd = get_userns_fd(0, (depth == 0) ? orig_uid : 0, 1);
+				if (userns_fd < 0)
+					exit(1);
+
+				if (setns(userns_fd, CLONE_NEWUSER) < 0) {
+					close(userns_fd);
+					exit(1);
+				}
+				close(userns_fd);
+			}
+
+			exit(0);
+		}
+
+		/* Parent: wait for child */
+		int status;
+		waitpid(pid, &status, 0);
+		ASSERT_TRUE(WIFEXITED(status));
+	}
+
+	/* Verify we're back to baseline */
+	ret_after = sys_listns(&req, ns_ids_after, ARRAY_SIZE(ns_ids_after), 0);
+	ASSERT_GE(ret_after, 0);
+
+	TH_LOG("After 20 nested namespace hierarchies: %zd active user namespaces", ret_after);
+	ASSERT_EQ(ret_before, ret_after);
+}
+
+/*
+ * Test listns() pagination under stress.
+ * Create many namespaces and verify pagination works correctly.
+ */
+TEST(listns_pagination_stress)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = CLONE_NEWUSER,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	pid_t pids[30];
+	int num_children = 30;
+	int i;
+	int sv[2];
+	__u64 all_ns_ids[512];
+	int total_found = 0;
+
+	ASSERT_EQ(socketpair(AF_UNIX, SOCK_STREAM, 0, sv), 0);
+
+	/* Create many children with user namespaces */
+	for (i = 0; i < num_children; i++) {
+		pids[i] = fork();
+		ASSERT_GE(pids[i], 0);
+
+		if (pids[i] == 0) {
+			char c;
+			close(sv[0]);
+
+			if (setup_userns() < 0) {
+				close(sv[1]);
+				exit(1);
+			}
+
+			/* Signal parent we're ready */
+			if (write(sv[1], &c, 1) != 1) {
+				close(sv[1]);
+				exit(1);
+			}
+
+			/* Wait for parent signal to exit */
+			if (read(sv[1], &c, 1) != 1) {
+				close(sv[1]);
+				exit(1);
+			}
+
+			close(sv[1]);
+			exit(0);
+		}
+	}
+
+	close(sv[1]);
+
+	/* Wait for all children to signal ready */
+	for (i = 0; i < num_children; i++) {
+		char c;
+		if (read(sv[0], &c, 1) != 1) {
+			/* If we fail to read, kill all children and exit */
+			close(sv[0]);
+			for (int j = 0; j < num_children; j++)
+				kill(pids[j], SIGKILL);
+			for (int j = 0; j < num_children; j++)
+				waitpid(pids[j], NULL, 0);
+			ASSERT_TRUE(false);
+		}
+	}
+
+	/* Paginate through all namespaces using small batch sizes */
+	req.ns_id = 0;
+	while (1) {
+		__u64 batch[5];  /* Small batch size to force pagination */
+		ssize_t ret;
+
+		ret = sys_listns(&req, batch, ARRAY_SIZE(batch), 0);
+		if (ret < 0) {
+			if (errno == ENOSYS) {
+				close(sv[0]);
+				for (i = 0; i < num_children; i++)
+					kill(pids[i], SIGKILL);
+				for (i = 0; i < num_children; i++)
+					waitpid(pids[i], NULL, 0);
+				SKIP(return, "listns() not supported");
+			}
+			ASSERT_GE(ret, 0);
+		}
+
+		if (ret == 0)
+			break;
+
+		/* Store results */
+		for (i = 0; i < ret && total_found < 512; i++) {
+			all_ns_ids[total_found++] = batch[i];
+		}
+
+		/* Update cursor for next batch */
+		if (ret == ARRAY_SIZE(batch))
+			req.ns_id = batch[ret - 1];
+		else
+			break;
+	}
+
+	TH_LOG("Paginated through %d user namespaces", total_found);
+
+	/* Verify no duplicates in pagination */
+	for (i = 0; i < total_found; i++) {
+		for (int j = i + 1; j < total_found; j++) {
+			if (all_ns_ids[i] == all_ns_ids[j]) {
+				TH_LOG("Found duplicate ns_id: %llu at positions %d and %d",
+				       (unsigned long long)all_ns_ids[i], i, j);
+				ASSERT_TRUE(false);
+			}
+		}
+	}
+
+	/* Signal all children to exit */
+	for (i = 0; i < num_children; i++) {
+		char c = 'X';
+		if (write(sv[0], &c, 1) != 1) {
+			close(sv[0]);
+			for (int j = i; j < num_children; j++)
+				kill(pids[j], SIGKILL);
+			for (int j = 0; j < num_children; j++)
+				waitpid(pids[j], NULL, 0);
+			ASSERT_TRUE(false);
+		}
+	}
+
+	close(sv[0]);
+
+	/* Wait for all children */
+	for (i = 0; i < num_children; i++) {
+		int status;
+		waitpid(pids[i], &status, 0);
+	}
+}
+
+/*
+ * Test concurrent namespace operations.
+ * Multiple processes creating, querying, and destroying namespaces concurrently.
+ */
+TEST(concurrent_namespace_operations)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = 0,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 ns_ids_before[512], ns_ids_after[512];
+	ssize_t ret_before, ret_after;
+	pid_t pids[20];
+	int num_workers = 20;
+	int i;
+
+	/* Get baseline */
+	ret_before = sys_listns(&req, ns_ids_before, ARRAY_SIZE(ns_ids_before), 0);
+	if (ret_before < 0) {
+		if (errno == ENOSYS)
+			SKIP(return, "listns() not supported");
+		ASSERT_GE(ret_before, 0);
+	}
+
+	TH_LOG("Baseline: %zd active namespaces", ret_before);
+
+	/* Create worker processes that do concurrent operations */
+	for (i = 0; i < num_workers; i++) {
+		pids[i] = fork();
+		ASSERT_GE(pids[i], 0);
+
+		if (pids[i] == 0) {
+			/* Each worker: create namespaces, list them, repeat */
+			int iterations;
+
+			for (iterations = 0; iterations < 10; iterations++) {
+				int userns_fd;
+				__u64 temp_ns_ids[100];
+				ssize_t ret;
+
+				/* Create a user namespace */
+				userns_fd = get_userns_fd(0, getuid(), 1);
+				if (userns_fd < 0)
+					continue;
+
+				/* List namespaces */
+				ret = sys_listns(&req, temp_ns_ids, ARRAY_SIZE(temp_ns_ids), 0);
+				(void)ret;
+
+				close(userns_fd);
+
+				/* Small delay */
+				usleep(1000);
+			}
+
+			exit(0);
+		}
+	}
+
+	/* Wait for all workers */
+	for (i = 0; i < num_workers; i++) {
+		int status;
+		waitpid(pids[i], &status, 0);
+		ASSERT_TRUE(WIFEXITED(status));
+		ASSERT_EQ(WEXITSTATUS(status), 0);
+	}
+
+	/* Verify we're back to baseline */
+	ret_after = sys_listns(&req, ns_ids_after, ARRAY_SIZE(ns_ids_after), 0);
+	ASSERT_GE(ret_after, 0);
+
+	TH_LOG("After concurrent operations: %zd active namespaces", ret_after);
+	ASSERT_EQ(ret_before, ret_after);
+}
+
+/*
+ * Test namespace churn - continuous creation and destruction.
+ * Simulates high-churn scenarios like container orchestration.
+ */
+TEST(namespace_churn)
+{
+	struct ns_id_req req = {
+		.size = sizeof(req),
+		.spare = 0,
+		.ns_id = 0,
+		.ns_type = CLONE_NEWUSER | CLONE_NEWNET | CLONE_NEWUTS,
+		.spare2 = 0,
+		.user_ns_id = 0,
+	};
+	__u64 ns_ids_before[512], ns_ids_after[512];
+	ssize_t ret_before, ret_after;
+	int cycle;
+
+	/* Get baseline */
+	ret_before = sys_listns(&req, ns_ids_before, ARRAY_SIZE(ns_ids_before), 0);
+	if (ret_before < 0) {
+		if (errno == ENOSYS)
+			SKIP(return, "listns() not supported");
+		ASSERT_GE(ret_before, 0);
+	}
+
+	TH_LOG("Baseline: %zd active namespaces", ret_before);
+
+	/* Simulate churn: batches of namespaces created and destroyed */
+	for (cycle = 0; cycle < 10; cycle++) {
+		pid_t batch_pids[10];
+		int i;
+
+		/* Create batch */
+		for (i = 0; i < 10; i++) {
+			batch_pids[i] = fork();
+			ASSERT_GE(batch_pids[i], 0);
+
+			if (batch_pids[i] == 0) {
+				/* Create multiple namespace types */
+				if (setup_userns() < 0)
+					exit(1);
+				if (unshare(CLONE_NEWNET) < 0)
+					exit(1);
+				if (unshare(CLONE_NEWUTS) < 0)
+					exit(1);
+
+				/* Keep namespaces alive briefly */
+				usleep(10000);
+				exit(0);
+			}
+		}
+
+		/* Wait for batch to complete */
+		for (i = 0; i < 10; i++) {
+			int status;
+			waitpid(batch_pids[i], &status, 0);
+		}
+	}
+
+	/* Verify we're back to baseline */
+	ret_after = sys_listns(&req, ns_ids_after, ARRAY_SIZE(ns_ids_after), 0);
+	ASSERT_GE(ret_after, 0);
+
+	TH_LOG("After 10 churn cycles (100 namespace sets): %zd active namespaces", ret_after);
+	ASSERT_EQ(ret_before, ret_after);
+}
+
+TEST_HARNESS_MAIN

linux-7.2.3/tools/testing/selftests/namespaces/wrappers.h

@@ -0,0 +1,35 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include <linux/nsfs.h>
+#include <linux/types.h>
+#include <sys/syscall.h>
+#include <unistd.h>
+
+#ifndef __SELFTESTS_NAMESPACES_WRAPPERS_H__
+#define __SELFTESTS_NAMESPACES_WRAPPERS_H__
+
+#ifndef __NR_listns
+	#if defined __alpha__
+		#define __NR_listns 580
+	#elif defined _MIPS_SIM
+		#if _MIPS_SIM == _MIPS_SIM_ABI32	/* o32 */
+			#define __NR_listns 4470
+		#endif
+		#if _MIPS_SIM == _MIPS_SIM_NABI32	/* n32 */
+			#define __NR_listns 6470
+		#endif
+		#if _MIPS_SIM == _MIPS_SIM_ABI64	/* n64 */
+			#define __NR_listns 5470
+		#endif
+	#else
+		#define __NR_listns 470
+	#endif
+#endif
+
+static inline int sys_listns(const struct ns_id_req *req, __u64 *ns_ids,
+			     size_t nr_ns_ids, unsigned int flags)
+{
+	return syscall(__NR_listns, req, ns_ids, nr_ns_ids, flags);
+}
+
+#endif /* __SELFTESTS_NAMESPACES_WRAPPERS_H__ */