ARM: 9485/1: mm: acquire mmap write lock around show_pte() for user faultsgrafted

When CONFIG_DEBUG_USER=y, and cmdline "user_debug=31" is set,
a user fault may trigger show_pte() without any lock.
If another thread in the same process concurrently calls munmap(),
the page table pages may be freed while show_pte() is still traversing
them, causing a use-after-free in show_pte().

If CONFIG_ARM_LPAE=y, this may cause a kernel panic if the pages table
of PMD are freed when show_pte() is running.

Acquire mmap_write_lock() around show_pte() for user faults to fix the
contention.

For user faults, additionally restrict that show_pte() is called only
when the addr is a user-space address (addr < TASK_SIZE). This is because
the lock of tsk->mm only protects the virtual memory of user address space,
furthermore, dumping the page tables of a kernel-space address for user
faults is unnecessary and may have security implications.

Keep everything unchanged for kernel faults, because the kernel is
already in the "oops" state, acquiring a lock may risk a deadlock.

Co-developed-by: Qi Xi <xiqi2@huawei.com>

Fixes: 6d021b724481 ("ARM: dump pgd, pmd and pte states on unhandled data abort faults")
Link: https://lore.kernel.org/20260716014022.2823-1-xieyuanbin1@huawei.com
Acked-by: Lorenzo Stoakes (ARM) <ljs@kernel.org>
Reviewed-by: Linus Walleij <linusw@kernel.org>
Signed-off-by: Qi Xi <xiqi2@huawei.com>
Signed-off-by: Xie Yuanbin <xieyuanbin1@huawei.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>

1 files changed, 84 insertions, 0 deletions

diff --git a/tools/perf/util/namespaces.h b/tools/perf/util/namespaces.h
new file mode 100644
index 000000000..e95c79b80
--- /dev/null
+++ b/tools/perf/util/namespaces.h
@@ -0,0 +1,84 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ *
+ * Copyright (C) 2017 Hari Bathini, IBM Corporation
+ */
+
+#ifndef __PERF_NAMESPACES_H
+#define __PERF_NAMESPACES_H
+
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <linux/stddef.h>
+#include <linux/perf_event.h>
+#include <linux/refcount.h>
+#include <linux/types.h>
+#include <internal/rc_check.h>
+
+#ifndef HAVE_SETNS_SUPPORT
+int setns(int fd, int nstype);
+#endif
+
+struct perf_record_namespaces;
+
+struct namespaces {
+	struct list_head list;
+	u64 end_time;
+	struct perf_ns_link_info link_info[];
+};
+
+struct namespaces *namespaces__new(struct perf_record_namespaces *event);
+void namespaces__free(struct namespaces *namespaces);
+
+DECLARE_RC_STRUCT(nsinfo) {
+	pid_t			pid;
+	pid_t			tgid;
+	pid_t			nstgid;
+	bool			need_setns;
+	bool			in_pidns;
+	char			*mntns_path;
+	refcount_t		refcnt;
+};
+
+struct nscookie {
+	int			oldns;
+	int			newns;
+	char			*oldcwd;
+};
+
+int nsinfo__init(struct nsinfo *nsi);
+struct nsinfo *nsinfo__new(pid_t pid);
+struct nsinfo *nsinfo__copy(const struct nsinfo *nsi);
+
+struct nsinfo *nsinfo__get(struct nsinfo *nsi);
+void nsinfo__put(struct nsinfo *nsi);
+
+bool nsinfo__need_setns(const struct nsinfo *nsi);
+void nsinfo__clear_need_setns(struct nsinfo *nsi);
+pid_t nsinfo__tgid(const struct nsinfo  *nsi);
+pid_t nsinfo__nstgid(const struct nsinfo  *nsi);
+pid_t nsinfo__pid(const struct nsinfo  *nsi);
+bool nsinfo__in_pidns(const struct nsinfo  *nsi);
+void nsinfo__set_in_pidns(struct nsinfo *nsi);
+
+void nsinfo__mountns_enter(struct nsinfo *nsi, struct nscookie *nc);
+void nsinfo__mountns_exit(struct nscookie *nc);
+
+char *nsinfo__realpath(const char *path, struct nsinfo *nsi);
+int nsinfo__stat(const char *filename, struct stat *st, struct nsinfo *nsi);
+
+bool nsinfo__is_in_root_namespace(void);
+
+static inline void __nsinfo__zput(struct nsinfo **nsip)
+{
+	if (nsip) {
+		nsinfo__put(*nsip);
+		*nsip = NULL;
+	}
+}
+
+#define nsinfo__zput(nsi) __nsinfo__zput(&nsi)
+
+const char *perf_ns__name(unsigned int id);
+
+#endif  /* __PERF_NAMESPACES_H */