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, 134 insertions, 0 deletions

diff --git a/include/trace/events/module.h b/include/trace/events/module.h
new file mode 100644
index 000000000..e5a006be9
--- /dev/null
+++ b/include/trace/events/module.h
@@ -0,0 +1,134 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Because linux/module.h has tracepoints in the header, and ftrace.h
+ * used to include this file, define_trace.h includes linux/module.h
+ * But we do not want the module.h to override the TRACE_SYSTEM macro
+ * variable that define_trace.h is processing, so we only set it
+ * when module events are being processed, which would happen when
+ * CREATE_TRACE_POINTS is defined.
+ */
+#ifdef CREATE_TRACE_POINTS
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM module
+#endif
+
+#if !defined(_TRACE_MODULE_H) || defined(TRACE_HEADER_MULTI_READ)
+#define _TRACE_MODULE_H
+
+#include <linux/tracepoint.h>
+
+#ifdef CONFIG_MODULES
+
+struct module;
+
+#define show_module_flags(flags) __print_flags(flags, "",	\
+	{ (1UL << TAINT_PROPRIETARY_MODULE),	"P" },		\
+	{ (1UL << TAINT_OOT_MODULE),		"O" },		\
+	{ (1UL << TAINT_FORCED_MODULE),		"F" },		\
+	{ (1UL << TAINT_CRAP),			"C" },		\
+	{ (1UL << TAINT_UNSIGNED_MODULE),	"E" })
+
+TRACE_EVENT(module_load,
+
+	TP_PROTO(struct module *mod),
+
+	TP_ARGS(mod),
+
+	TP_STRUCT__entry(
+		__field(	unsigned int,	taints		)
+		__string(	name,		mod->name	)
+	),
+
+	TP_fast_assign(
+		__entry->taints = mod->taints;
+		__assign_str(name);
+	),
+
+	TP_printk("%s %s", __get_str(name), show_module_flags(__entry->taints))
+);
+
+TRACE_EVENT(module_free,
+
+	TP_PROTO(struct module *mod),
+
+	TP_ARGS(mod),
+
+	TP_STRUCT__entry(
+		__string(	name,		mod->name	)
+	),
+
+	TP_fast_assign(
+		__assign_str(name);
+	),
+
+	TP_printk("%s", __get_str(name))
+);
+
+#ifdef CONFIG_MODULE_UNLOAD
+/* trace_module_get/put are only used if CONFIG_MODULE_UNLOAD is defined */
+
+DECLARE_EVENT_CLASS(module_refcnt,
+
+	TP_PROTO(struct module *mod, unsigned long ip),
+
+	TP_ARGS(mod, ip),
+
+	TP_STRUCT__entry(
+		__field(	unsigned long,	ip		)
+		__field(	int,		refcnt		)
+		__string(	name,		mod->name	)
+	),
+
+	TP_fast_assign(
+		__entry->ip	= ip;
+		__entry->refcnt	= atomic_read(&mod->refcnt);
+		__assign_str(name);
+	),
+
+	TP_printk("%s call_site=%ps refcnt=%d",
+		  __get_str(name), (void *)__entry->ip, __entry->refcnt)
+);
+
+DEFINE_EVENT(module_refcnt, module_get,
+
+	TP_PROTO(struct module *mod, unsigned long ip),
+
+	TP_ARGS(mod, ip)
+);
+
+DEFINE_EVENT(module_refcnt, module_put,
+
+	TP_PROTO(struct module *mod, unsigned long ip),
+
+	TP_ARGS(mod, ip)
+);
+#endif /* CONFIG_MODULE_UNLOAD */
+
+TRACE_EVENT(module_request,
+
+	TP_PROTO(char *name, bool wait, unsigned long ip),
+
+	TP_ARGS(name, wait, ip),
+
+	TP_STRUCT__entry(
+		__field(	unsigned long,	ip		)
+		__field(	bool,		wait		)
+		__string(	name,		name		)
+	),
+
+	TP_fast_assign(
+		__entry->ip	= ip;
+		__entry->wait	= wait;
+		__assign_str(name);
+	),
+
+	TP_printk("%s wait=%d call_site=%ps",
+		  __get_str(name), (int)__entry->wait, (void *)__entry->ip)
+);
+
+#endif /* CONFIG_MODULES */
+
+#endif /* _TRACE_MODULE_H */
+
+/* This part must be outside protection */
+#include <trace/define_trace.h>