proc: fix /proc/net/* after setns(2)

/proc entries under /proc/net/* can't be cached into dcache because
setns(2) can change current net namespace.

[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: avoid vim miscolorization]
[adobriyan@gmail.com: write test, add dummy ->d_revalidate hook: necessary if /proc/net/* is pinned at setns time]
Link: http://lkml.kernel.org/r/20190108192350.GA12034@avx2
Link: http://lkml.kernel.org/r/20190107162336.GA9239@avx2
Fixes: 1da4d377f943fe4194ffb9fb9c26cc58fad4dd24 ("proc: revalidate misc dentries")
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Reported-by: Mateusz Stępień <mateusz.stepien@netrounds.com>
Reported-by: Ahmad Fatoum <a.fatoum@pengutronix.de>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

diff --git a/fs/proc/generic.c b/fs/proc/generic.c
index 8ae109429a883939ed0241c1e122db7ba857ac1a..e39bac94dead0965733eb779758edfae31aacb81 100644 (file)
--- a/fs/proc/generic.c
+++ b/fs/proc/generic.c
@@ -256,7 +256,7 @@ struct dentry *proc_lookup_de(struct inode *dir, struct dentry *dentry,
                inode = proc_get_inode(dir->i_sb, de);
                if (!inode)
                        return ERR_PTR(-ENOMEM);
-               d_set_d_op(dentry, &proc_misc_dentry_ops);
+               d_set_d_op(dentry, de->proc_dops);
                return d_splice_alias(inode, dentry);
        }
        read_unlock(&proc_subdir_lock);
@@ -429,6 +429,8 @@ static struct proc_dir_entry *__proc_create(struct proc_dir_entry **parent,
        INIT_LIST_HEAD(&ent->pde_openers);
        proc_set_user(ent, (*parent)->uid, (*parent)->gid);
 
+       ent->proc_dops = &proc_misc_dentry_ops;
+
 out:
        return ent;
 }

diff --git a/fs/proc/internal.h b/fs/proc/internal.h
index 5185d7f6a51ee845fc12c6d49d4497d766af6c23..95b14196f28420934f9fe091854809151e5f39d3 100644 (file)
--- a/fs/proc/internal.h
+++ b/fs/proc/internal.h
@@ -44,6 +44,7 @@ struct proc_dir_entry {
        struct completion *pde_unload_completion;
        const struct inode_operations *proc_iops;
        const struct file_operations *proc_fops;
+       const struct dentry_operations *proc_dops;
        union {
                const struct seq_operations *seq_ops;
                int (*single_show)(struct seq_file *, void *);

diff --git a/fs/proc/proc_net.c b/fs/proc/proc_net.c
index d5e0fcb3439e91b8c722b7193c0d8a59e2ecbe98..a7b12435519e048e90abf6bd3250ce8e22da40d0 100644 (file)
--- a/fs/proc/proc_net.c
+++ b/fs/proc/proc_net.c
@@ -38,6 +38,22 @@ static struct net *get_proc_net(const struct inode *inode)
        return maybe_get_net(PDE_NET(PDE(inode)));
 }
 
+static int proc_net_d_revalidate(struct dentry *dentry, unsigned int flags)
+{
+       return 0;
+}
+
+static const struct dentry_operations proc_net_dentry_ops = {
+       .d_revalidate   = proc_net_d_revalidate,
+       .d_delete       = always_delete_dentry,
+};
+
+static void pde_force_lookup(struct proc_dir_entry *pde)
+{
+       /* /proc/net/ entries can be changed under us by setns(CLONE_NEWNET) */
+       pde->proc_dops = &proc_net_dentry_ops;
+}
+
 static int seq_open_net(struct inode *inode, struct file *file)
 {
        unsigned int state_size = PDE(inode)->state_size;
@@ -90,6 +106,7 @@ struct proc_dir_entry *proc_create_net_data(const char *name, umode_t mode,
        p = proc_create_reg(name, mode, &parent, data);
        if (!p)
                return NULL;
+       pde_force_lookup(p);
        p->proc_fops = &proc_net_seq_fops;
        p->seq_ops = ops;
        p->state_size = state_size;
@@ -133,6 +150,7 @@ struct proc_dir_entry *proc_create_net_data_write(const char *name, umode_t mode
        p = proc_create_reg(name, mode, &parent, data);
        if (!p)
                return NULL;
+       pde_force_lookup(p);
        p->proc_fops = &proc_net_seq_fops;
        p->seq_ops = ops;
        p->state_size = state_size;
@@ -181,6 +199,7 @@ struct proc_dir_entry *proc_create_net_single(const char *name, umode_t mode,
        p = proc_create_reg(name, mode, &parent, data);
        if (!p)
                return NULL;
+       pde_force_lookup(p);
        p->proc_fops = &proc_net_single_fops;
        p->single_show = show;
        return proc_register(parent, p);
@@ -223,6 +242,7 @@ struct proc_dir_entry *proc_create_net_single_write(const char *name, umode_t mo
        p = proc_create_reg(name, mode, &parent, data);
        if (!p)
                return NULL;
+       pde_force_lookup(p);
        p->proc_fops = &proc_net_single_fops;
        p->single_show = show;
        p->write = write;

diff --git a/tools/testing/selftests/proc/.gitignore b/tools/testing/selftests/proc/.gitignore
index 82121a81681f7720a16709d659c6e7c53e6a3fa3..29bac5ef9a93dda0b1964223ba77f39c23e75a72 100644 (file)
--- a/tools/testing/selftests/proc/.gitignore
+++ b/tools/testing/selftests/proc/.gitignore
@@ -10,4 +10,5 @@
 /proc-uptime-002
 /read
 /self
+/setns-dcache
 /thread-self

diff --git a/tools/testing/selftests/proc/Makefile b/tools/testing/selftests/proc/Makefile
index 1c12c34cf85d801b09c6adbfe1fb1067c55283e8..434d033ee06774aa4a14b75c8dfed7dacf2f75a4 100644 (file)
--- a/tools/testing/selftests/proc/Makefile
+++ b/tools/testing/selftests/proc/Makefile
@@ -14,6 +14,7 @@ TEST_GEN_PROGS += proc-uptime-001
 TEST_GEN_PROGS += proc-uptime-002
 TEST_GEN_PROGS += read
 TEST_GEN_PROGS += self
+TEST_GEN_PROGS += setns-dcache
 TEST_GEN_PROGS += thread-self
 
 include ../lib.mk

diff --git a/tools/testing/selftests/proc/setns-dcache.c b/tools/testing/selftests/proc/setns-dcache.c
new file mode 100644 (file)
index 0000000..60ab197
--- /dev/null
+++ b/tools/testing/selftests/proc/setns-dcache.c
@@ -0,0 +1,129 @@
+/*
+ * Copyright © 2019 Alexey Dobriyan <adobriyan@gmail.com>
+ *
+ * Permission to use, copy, modify, and distribute this software for any
+ * purpose with or without fee is hereby granted, provided that the above
+ * copyright notice and this permission notice appear in all copies.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+ * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+ * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+ * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+ * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+ * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+ * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+ */
+/*
+ * Test that setns(CLONE_NEWNET) points to new /proc/net content even
+ * if old one is in dcache.
+ *
+ * FIXME /proc/net/unix is under CONFIG_UNIX which can be disabled.
+ */
+#undef NDEBUG
+#include <assert.h>
+#include <errno.h>
+#include <sched.h>
+#include <signal.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <unistd.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+#include <sys/socket.h>
+
+static pid_t pid = -1;
+
+static void f(void)
+{
+       if (pid > 0) {
+               kill(pid, SIGTERM);
+       }
+}
+
+int main(void)
+{
+       int fd[2];
+       char _ = 0;
+       int nsfd;
+
+       atexit(f);
+
+       /* Check for priviledges and syscall availability straight away. */
+       if (unshare(CLONE_NEWNET) == -1) {
+               if (errno == ENOSYS || errno == EPERM) {
+                       return 4;
+               }
+               return 1;
+       }
+       /* Distinguisher between two otherwise empty net namespaces. */
+       if (socket(AF_UNIX, SOCK_STREAM, 0) == -1) {
+               return 1;
+       }
+
+       if (pipe(fd) == -1) {
+               return 1;
+       }
+
+       pid = fork();
+       if (pid == -1) {
+               return 1;
+       }
+
+       if (pid == 0) {
+               if (unshare(CLONE_NEWNET) == -1) {
+                       return 1;
+               }
+
+               if (write(fd[1], &_, 1) != 1) {
+                       return 1;
+               }
+
+               pause();
+
+               return 0;
+       }
+
+       if (read(fd[0], &_, 1) != 1) {
+               return 1;
+       }
+
+       {
+               char buf[64];
+               snprintf(buf, sizeof(buf), "/proc/%u/ns/net", pid);
+               nsfd = open(buf, O_RDONLY);
+               if (nsfd == -1) {
+                       return 1;
+               }
+       }
+
+       /* Reliably pin dentry into dcache. */
+       (void)open("/proc/net/unix", O_RDONLY);
+
+       if (setns(nsfd, CLONE_NEWNET) == -1) {
+               return 1;
+       }
+
+       kill(pid, SIGTERM);
+       pid = 0;
+
+       {
+               char buf[4096];
+               ssize_t rv;
+               int fd;
+
+               fd = open("/proc/net/unix", O_RDONLY);
+               if (fd == -1) {
+                       return 1;
+               }
+
+#define S "Num       RefCount Protocol Flags    Type St Inode Path\n"
+               rv = read(fd, buf, sizeof(buf));
+
+               assert(rv == strlen(S));
+               assert(memcmp(buf, S, strlen(S)) == 0);
+       }
+
+       return 0;
+}