thunderbolt: Fix NULL dereference in tb_remove_work()grafted

There is a slight race between tb_remove_work() and tb_domain_remove()
which leads to dereferencing a NULL tb->root_switch pointer inside
tb_free_unplugged_xdomains():

        Thread A                            Thread B

  tb_remove_work()
                                    tb_domain_remove()
                                      mutex_lock(&tb->lock)
                                      tb_stop()
                                        /* doesn't cancel a running callback */
                                        cancel_delayed_work(&tcm->remove_work)
                                        ...
                                        tb_switch_remove(tb->root_switch)
                                        tb->root_switch = NULL
                                      mutex_unlock(&tb->lock)
    mutex_lock(&tb->lock)
    ...
    /* without checking ->root_switch */
    tb_free_unplugged_xdomains(tb->root_switch)
    mutex_unlock(&tb->lock)

Commit a8937f35cf39 ("thunderbolt: Remove XDomain from the bus without
holding tb->lock") doesn't seem right to move tb_free_unplugged_xdomains()
out of the &tb->lock section and the check for tb->root_switch, in
particular.  It states:

  For this reason separate removing the XDomain from the topology data
  structures (where we need the lock) from unregistering the device from
  the bus (where remove callbacks of the drivers are being called).

tb_free_unplugged_xdomains() belongs to the former group of functions
requiring the lock.  And it also calls tb_xdomain_remove() which should
only be called with &tb->lock held.

Found by Linux Verification Center (linuxtesting.org) with Svace static
analysis tool.

Fixes: a8937f35cf39 ("thunderbolt: Remove XDomain from the bus without holding tb->lock")
Cc: stable@vger.kernel.org
Signed-off-by: Fedor Pchelkin <pchelkin@ispras.ru>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>

1 files changed, 57 insertions, 0 deletions

diff --git a/include/trace/events/capability.h b/include/trace/events/capability.h
new file mode 100644
index 000000000..173402579
--- /dev/null
+++ b/include/trace/events/capability.h
@@ -0,0 +1,57 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM capability
+
+#if !defined(_TRACE_CAPABILITY_H) || defined(TRACE_HEADER_MULTI_READ)
+#define _TRACE_CAPABILITY_H
+
+#include <linux/cred.h>
+#include <linux/tracepoint.h>
+#include <linux/user_namespace.h>
+
+/**
+ * cap_capable - called after it's determined if a task has a particular
+ * effective capability
+ *
+ * @cred: The credentials used
+ * @target_ns: The user namespace of the resource being accessed
+ * @capable_ns: The user namespace in which the credential provides the
+ *              capability to access the targeted resource.
+ *              This will be NULL if ret is not 0.
+ * @cap: The capability to check for
+ * @ret: The return value of the check: 0 if it does, -ve if it does not
+ *
+ * Allows to trace calls to cap_capable in commoncap.c
+ */
+TRACE_EVENT(cap_capable,
+
+	TP_PROTO(const struct cred *cred, struct user_namespace *target_ns,
+		const struct user_namespace *capable_ns, int cap, int ret),
+
+	TP_ARGS(cred, target_ns, capable_ns, cap, ret),
+
+	TP_STRUCT__entry(
+		__field(const struct cred *, cred)
+		__field(struct user_namespace *, target_ns)
+		__field(const struct user_namespace *, capable_ns)
+		__field(int, cap)
+		__field(int, ret)
+	),
+
+	TP_fast_assign(
+		__entry->cred       = cred;
+		__entry->target_ns    = target_ns;
+		__entry->capable_ns = ret == 0 ? capable_ns : NULL;
+		__entry->cap        = cap;
+		__entry->ret        = ret;
+	),
+
+	TP_printk("cred %p, target_ns %p, capable_ns %p, cap %d, ret %d",
+		__entry->cred, __entry->target_ns, __entry->capable_ns, __entry->cap,
+		__entry->ret)
+);
+
+#endif /* _TRACE_CAPABILITY_H */
+
+/* This part must be outside protection */
+#include <trace/define_trace.h>