From 032c59c7681c661b63123231824170c025afb7e7 Mon Sep 17 00:00:00 2001 From: Sven Peter Date: Sat, 29 Aug 2026 10:08:33 +0200 Subject: [PATCH 01/10] thunderbolt: Hold a router reference for each allocated HopID tb_stop() drops the reference to all DP tunnels but does not deactivate them, thus nothing cancels a dprx_work still in flight (which holds its own tunnel reference) and the tunnel can outlive tb_switch_remove(). The HopID releases in tb_path_free() then operate on freed IDAs and trigger warnings like ida_free called for id=8 which is not allocated. This can be triggered by unbinding the driver while a DP tunnel is still waiting for the DPRX capabilities read to finish. On the Apple NHI unplugging the cable runs into just that reliably because the read can never finish right now and because the unplug powers down the entire USB4 complex and removes the NHI device. Take a router reference whenever an input or output HopID is allocated and drop it again after the HopID is released. This keeps the ports and their HopID IDAs alive for as long as they are used. The KUnit tests allocate routers without ever registering their devices so initialize the embedded struct device there as well and drop its initial reference when the test is finished. Fixes: d6d458d42e1e ("thunderbolt: Handle DisplayPort tunnel activation asynchronously") Cc: stable@vger.kernel.org Signed-off-by: Sven Peter Signed-off-by: Mika Westerberg --- drivers/thunderbolt/switch.c | 9 ++++++++- drivers/thunderbolt/test.c | 21 +++++++++++++++++++++ 2 files changed, 29 insertions(+), 1 deletion(-) diff --git a/drivers/thunderbolt/switch.c b/drivers/thunderbolt/switch.c index 404c0693df50..cf571a7c98d0 100644 --- a/drivers/thunderbolt/switch.c +++ b/drivers/thunderbolt/switch.c @@ -774,6 +774,7 @@ static int tb_port_alloc_hopid(struct tb_port *port, bool in, int min_hopid, { int port_max_hopid; struct ida *ida; + int ret; if (in) { port_max_hopid = port->config.max_in_hop_id; @@ -793,7 +794,11 @@ static int tb_port_alloc_hopid(struct tb_port *port, bool in, int min_hopid, if (max_hopid < 0 || max_hopid > port_max_hopid) max_hopid = port_max_hopid; - return ida_alloc_range(ida, min_hopid, max_hopid, GFP_KERNEL); + ret = ida_alloc_range(ida, min_hopid, max_hopid, GFP_KERNEL); + if (ret >= 0) + tb_switch_get(port->sw); + + return ret; } /** @@ -832,6 +837,7 @@ int tb_port_alloc_out_hopid(struct tb_port *port, int min_hopid, int max_hopid) void tb_port_release_in_hopid(struct tb_port *port, int hopid) { ida_free(&port->in_hopids, hopid); + tb_switch_put(port->sw); } /** @@ -842,6 +848,7 @@ void tb_port_release_in_hopid(struct tb_port *port, int hopid) void tb_port_release_out_hopid(struct tb_port *port, int hopid) { ida_free(&port->out_hopids, hopid); + tb_switch_put(port->sw); } static inline bool tb_switch_is_reachable(const struct tb_switch *parent, diff --git a/drivers/thunderbolt/test.c b/drivers/thunderbolt/test.c index 05652ee82fbf..799ada97cbc9 100644 --- a/drivers/thunderbolt/test.c +++ b/drivers/thunderbolt/test.c @@ -33,6 +33,18 @@ static void kunit_ida_init(struct kunit *test, struct ida *ida) kunit_alloc_resource(test, __ida_init, __ida_destroy, GFP_KERNEL, ida); } +static void tb_test_switch_release(struct device *dev) +{ + /* The memory is owned by KUnit, nothing to do here */ +} + +static void tb_test_switch_put(void *data) +{ + struct tb_switch *sw = data; + + put_device(&sw->dev); +} + static struct tb_switch *alloc_switch(struct kunit *test, u64 route, u8 upstream_port, u8 max_port_number) { @@ -44,6 +56,15 @@ static struct tb_switch *alloc_switch(struct kunit *test, u64 route, if (!sw) return NULL; + /* + * HopID allocations take a reference to their routers and those devices + * have to be initialized for that to work. + */ + sw->dev.release = tb_test_switch_release; + device_initialize(&sw->dev); + if (kunit_add_action_or_reset(test, tb_test_switch_put, sw)) + return NULL; + sw->config.upstream_port_number = upstream_port; sw->config.depth = tb_route_length(route); sw->config.route_hi = upper_32_bits(route); From 12f5d8b85a66a0ac08d082517d92ce136f8c0d42 Mon Sep 17 00:00:00 2001 From: Sven Peter Date: Sat, 29 Aug 2026 10:08:34 +0200 Subject: [PATCH 02/10] thunderbolt: Make the DP tunnel activation callback mandatory tb_tunnel_alloc_dp() takes an optional callback which is run from dprx_work once the DPRX capabilities read has completed. Without that callback tb_dp_dprx_start() reads the capabilities synchronously and never queues the work. It however always takes a tunnel reference which is only dropped by dprx_work itself or by tb_dp_dprx_stop() when cancel_delayed_work() actually canceled that work. That reference is thus leaked for every tunnel without a callback. The only tunnels without one are those from tb_tunnel_discover_dp(), which are activated again when restoring from hibernation. Pass the callback to tb_tunnel_discover_dp() as well and drop the synchronous path such that the DPRX capabilities are always read from dprx_work. Hibernation restore then also no longer blocks for up to 12 seconds while waiting for that read to complete. Also fix up the KUnit tests. Fixes: d6d458d42e1e ("thunderbolt: Handle DisplayPort tunnel activation asynchronously") Cc: stable@vger.kernel.org Signed-off-by: Sven Peter Signed-off-by: Mika Westerberg --- drivers/thunderbolt/tb.c | 4 ++- drivers/thunderbolt/test.c | 37 +++++++++++++++++++--------- drivers/thunderbolt/tunnel.c | 47 ++++++++++++++++++++---------------- drivers/thunderbolt/tunnel.h | 8 +++--- 4 files changed, 60 insertions(+), 36 deletions(-) diff --git a/drivers/thunderbolt/tb.c b/drivers/thunderbolt/tb.c index 47753a5c0f2e..20f0f8f4bdf7 100644 --- a/drivers/thunderbolt/tb.c +++ b/drivers/thunderbolt/tb.c @@ -89,6 +89,7 @@ static void tb_dp_resource_unavailable(struct tb *tb, struct tb_port *port, const char *reason); static void tb_queue_dp_bandwidth_request(struct tb *tb, u64 route, u8 port, int retry, unsigned long delay); +static void tb_dp_tunnel_active(struct tb_tunnel *tunnel, void *data); static void tb_queue_hotplug(struct tb *tb, u64 route, u8 port, bool unplug) { @@ -385,7 +386,8 @@ static void tb_switch_discover_tunnels(struct tb_switch *sw, switch (port->config.type) { case TB_TYPE_DP_HDMI_IN: - tunnel = tb_tunnel_discover_dp(tb, port, alloc_hopids); + tunnel = tb_tunnel_discover_dp(tb, port, alloc_hopids, + tb_dp_tunnel_active, tb); tb_increase_tmu_accuracy(tunnel); break; diff --git a/drivers/thunderbolt/test.c b/drivers/thunderbolt/test.c index 799ada97cbc9..c8c648f31107 100644 --- a/drivers/thunderbolt/test.c +++ b/drivers/thunderbolt/test.c @@ -1407,6 +1407,10 @@ static void tb_test_tunnel_pcie(struct kunit *test) tb_tunnel_put(tunnel1); } +static void tb_test_dp_tunnel_active(struct tb_tunnel *tunnel, void *data) +{ +} + static void tb_test_tunnel_dp(struct kunit *test) { struct tb_switch *host, *dev; @@ -1427,7 +1431,8 @@ static void tb_test_tunnel_dp(struct kunit *test) in = &host->ports[5]; out = &dev->ports[13]; - tunnel = tb_tunnel_alloc_dp(NULL, in, out, 1, 0, 0, NULL, NULL); + tunnel = tb_tunnel_alloc_dp(NULL, in, out, 1, 0, 0, + tb_test_dp_tunnel_active, NULL); KUNIT_ASSERT_NOT_NULL(test, tunnel); KUNIT_EXPECT_EQ(test, tunnel->type, TB_TUNNEL_DP); KUNIT_EXPECT_PTR_EQ(test, tunnel->src_port, in); @@ -1473,7 +1478,8 @@ static void tb_test_tunnel_dp_chain(struct kunit *test) in = &host->ports[5]; out = &dev4->ports[14]; - tunnel = tb_tunnel_alloc_dp(NULL, in, out, 1, 0, 0, NULL, NULL); + tunnel = tb_tunnel_alloc_dp(NULL, in, out, 1, 0, 0, + tb_test_dp_tunnel_active, NULL); KUNIT_ASSERT_NOT_NULL(test, tunnel); KUNIT_EXPECT_EQ(test, tunnel->type, TB_TUNNEL_DP); KUNIT_EXPECT_PTR_EQ(test, tunnel->src_port, in); @@ -1523,7 +1529,8 @@ static void tb_test_tunnel_dp_tree(struct kunit *test) in = &dev2->ports[13]; out = &dev5->ports[13]; - tunnel = tb_tunnel_alloc_dp(NULL, in, out, 1, 0, 0, NULL, NULL); + tunnel = tb_tunnel_alloc_dp(NULL, in, out, 1, 0, 0, + tb_test_dp_tunnel_active, NULL); KUNIT_ASSERT_NOT_NULL(test, tunnel); KUNIT_EXPECT_EQ(test, tunnel->type, TB_TUNNEL_DP); KUNIT_EXPECT_PTR_EQ(test, tunnel->src_port, in); @@ -1588,7 +1595,8 @@ static void tb_test_tunnel_dp_max_length(struct kunit *test) in = &dev6->ports[13]; out = &dev12->ports[13]; - tunnel = tb_tunnel_alloc_dp(NULL, in, out, 1, 0, 0, NULL, NULL); + tunnel = tb_tunnel_alloc_dp(NULL, in, out, 1, 0, 0, + tb_test_dp_tunnel_active, NULL); KUNIT_ASSERT_NOT_NULL(test, tunnel); KUNIT_EXPECT_EQ(test, tunnel->type, TB_TUNNEL_DP); KUNIT_EXPECT_PTR_EQ(test, tunnel->src_port, in); @@ -1658,7 +1666,8 @@ static void tb_test_tunnel_3dp(struct kunit *test) out2 = &dev5->ports[13]; out3 = &dev4->ports[14]; - tunnel1 = tb_tunnel_alloc_dp(NULL, in1, out1, 1, 0, 0, NULL, NULL); + tunnel1 = tb_tunnel_alloc_dp(NULL, in1, out1, 1, 0, 0, + tb_test_dp_tunnel_active, NULL); KUNIT_ASSERT_TRUE(test, tunnel1 != NULL); KUNIT_EXPECT_EQ(test, tunnel1->type, TB_TUNNEL_DP); KUNIT_EXPECT_PTR_EQ(test, tunnel1->src_port, in1); @@ -1666,7 +1675,8 @@ static void tb_test_tunnel_3dp(struct kunit *test) KUNIT_ASSERT_EQ(test, tunnel1->npaths, 3); KUNIT_ASSERT_EQ(test, tunnel1->paths[0]->path_length, 3); - tunnel2 = tb_tunnel_alloc_dp(NULL, in2, out2, 1, 0, 0, NULL, NULL); + tunnel2 = tb_tunnel_alloc_dp(NULL, in2, out2, 1, 0, 0, + tb_test_dp_tunnel_active, NULL); KUNIT_ASSERT_TRUE(test, tunnel2 != NULL); KUNIT_EXPECT_EQ(test, tunnel2->type, TB_TUNNEL_DP); KUNIT_EXPECT_PTR_EQ(test, tunnel2->src_port, in2); @@ -1674,7 +1684,8 @@ static void tb_test_tunnel_3dp(struct kunit *test) KUNIT_ASSERT_EQ(test, tunnel2->npaths, 3); KUNIT_ASSERT_EQ(test, tunnel2->paths[0]->path_length, 4); - tunnel3 = tb_tunnel_alloc_dp(NULL, in3, out3, 1, 0, 0, NULL, NULL); + tunnel3 = tb_tunnel_alloc_dp(NULL, in3, out3, 1, 0, 0, + tb_test_dp_tunnel_active, NULL); KUNIT_ASSERT_TRUE(test, tunnel3 != NULL); KUNIT_EXPECT_EQ(test, tunnel3->type, TB_TUNNEL_DP); KUNIT_EXPECT_PTR_EQ(test, tunnel3->src_port, in3); @@ -1772,7 +1783,8 @@ static void tb_test_tunnel_port_on_path(struct kunit *test) in = &dev2->ports[13]; out = &dev5->ports[13]; - dp_tunnel = tb_tunnel_alloc_dp(NULL, in, out, 1, 0, 0, NULL, NULL); + dp_tunnel = tb_tunnel_alloc_dp(NULL, in, out, 1, 0, 0, + tb_test_dp_tunnel_active, NULL); KUNIT_ASSERT_NOT_NULL(test, dp_tunnel); KUNIT_EXPECT_TRUE(test, tb_tunnel_port_on_path(dp_tunnel, in)); @@ -2204,7 +2216,8 @@ static void tb_test_credit_alloc_dp(struct kunit *test) in = &host->ports[5]; out = &dev->ports[14]; - tunnel = tb_tunnel_alloc_dp(NULL, in, out, 1, 0, 0, NULL, NULL); + tunnel = tb_tunnel_alloc_dp(NULL, in, out, 1, 0, 0, + tb_test_dp_tunnel_active, NULL); KUNIT_ASSERT_NOT_NULL(test, tunnel); KUNIT_ASSERT_EQ(test, tunnel->npaths, (size_t)3); @@ -2440,7 +2453,8 @@ static struct tb_tunnel *TB_TEST_DP_TUNNEL1(struct kunit *test, in = &host->ports[5]; out = &dev->ports[13]; - dp_tunnel1 = tb_tunnel_alloc_dp(NULL, in, out, 1, 0, 0, NULL, NULL); + dp_tunnel1 = tb_tunnel_alloc_dp(NULL, in, out, 1, 0, 0, + tb_test_dp_tunnel_active, NULL); KUNIT_ASSERT_NOT_NULL(test, dp_tunnel1); KUNIT_ASSERT_EQ(test, dp_tunnel1->npaths, (size_t)3); @@ -2477,7 +2491,8 @@ static struct tb_tunnel *TB_TEST_DP_TUNNEL2(struct kunit *test, in = &host->ports[6]; out = &dev->ports[14]; - dp_tunnel2 = tb_tunnel_alloc_dp(NULL, in, out, 1, 0, 0, NULL, NULL); + dp_tunnel2 = tb_tunnel_alloc_dp(NULL, in, out, 1, 0, 0, + tb_test_dp_tunnel_active, NULL); KUNIT_ASSERT_NOT_NULL(test, dp_tunnel2); KUNIT_ASSERT_EQ(test, dp_tunnel2->npaths, (size_t)3); diff --git a/drivers/thunderbolt/tunnel.c b/drivers/thunderbolt/tunnel.c index 7e8284575dff..6ac44b371266 100644 --- a/drivers/thunderbolt/tunnel.c +++ b/drivers/thunderbolt/tunnel.c @@ -1109,8 +1109,7 @@ static void tb_dp_dprx_work(struct work_struct *work) mutex_unlock(&tb->lock); } - if (tunnel->callback) - tunnel->callback(tunnel, tunnel->callback_data); + tunnel->callback(tunnel, tunnel->callback_data); tb_tunnel_put(tunnel); } @@ -1123,15 +1122,10 @@ static int tb_dp_dprx_start(struct tb_tunnel *tunnel) tb_tunnel_get(tunnel); tunnel->dprx_started = true; + tunnel->dprx_timeout = dprx_timeout_to_ktime(dprx_timeout); + queue_delayed_work(tunnel->tb->wq, &tunnel->dprx_work, 0); - if (tunnel->callback) { - tunnel->dprx_timeout = dprx_timeout_to_ktime(dprx_timeout); - queue_delayed_work(tunnel->tb->wq, &tunnel->dprx_work, 0); - return -EINPROGRESS; - } - - return tb_dp_is_usb4(tunnel->src_port->sw) ? - tb_dp_wait_dprx(tunnel, dprx_timeout) : 0; + return -EINPROGRESS; } static void tb_dp_dprx_stop(struct tb_tunnel *tunnel) @@ -1582,20 +1576,28 @@ static void tb_dp_dump(struct tb_tunnel *tunnel) * @tb: Pointer to the domain structure * @in: DP in adapter * @alloc_hopid: Allocate HopIDs from visited ports + * @callback: Callback that is called when the DP tunnel is fully + * activated (or there is an error) + * @callback_data: Data for @callback * * If @in adapter is active, follows the tunnel to the DP out adapter * and back. Returns the discovered tunnel or %NULL if there was no - * tunnel. + * tunnel. See tb_tunnel_alloc_dp() for @callback. * * Return: Pointer to &struct tb_tunnel or %NULL if no tunnel found. */ struct tb_tunnel *tb_tunnel_discover_dp(struct tb *tb, struct tb_port *in, - bool alloc_hopid) + bool alloc_hopid, + void (*callback)(struct tb_tunnel *, void *), + void *callback_data) { struct tb_tunnel *tunnel; struct tb_port *port; struct tb_path *path; + if (WARN_ON(!callback)) + return NULL; + if (!tb_dp_port_is_enabled(in)) return NULL; @@ -1611,6 +1613,9 @@ struct tb_tunnel *tb_tunnel_discover_dp(struct tb *tb, struct tb_port *in, tunnel->alloc_bandwidth = tb_dp_alloc_bandwidth; tunnel->consumed_bandwidth = tb_dp_consumed_bandwidth; tunnel->src_port = in; + tunnel->callback = callback; + tunnel->callback_data = callback_data; + INIT_DELAYED_WORK(&tunnel->dprx_work, tb_dp_dprx_work); path = tb_path_discover(in, TB_DP_VIDEO_HOPID, NULL, -1, &tunnel->dst_port, "Video", alloc_hopid); @@ -1677,16 +1682,16 @@ err_free: * %0 if no available bandwidth. * @max_down: Maximum available downstream bandwidth for the DP tunnel. * %0 if no available bandwidth. - * @callback: Optional callback that is called when the DP tunnel is - * fully activated (or there is an error) - * @callback_data: Optional data for @callback + * @callback: Callback that is called when the DP tunnel is fully + * activated (or there is an error) + * @callback_data: Data for @callback * * Allocates a tunnel between @in and @out that is capable of tunneling - * Display Port traffic. If @callback is not %NULL it will be called - * after tb_tunnel_activate() once the tunnel has been fully activated. - * It can call tb_tunnel_is_active() to check if activation was - * successful (or if it returns %false there was some sort of issue). - * The @callback is called without @tb->lock held. + * Display Port traffic. The @callback is called after tb_tunnel_activate() + * once the tunnel has been fully activated. It can call + * tb_tunnel_is_active() to check if activation was successful (or if it + * returns %false there was some sort of issue). The @callback is called + * without @tb->lock held. * * Return: Pointer to @struct tb_tunnel or %NULL in case of failure. */ @@ -1701,7 +1706,7 @@ struct tb_tunnel *tb_tunnel_alloc_dp(struct tb *tb, struct tb_port *in, struct tb_path *path; bool pm_support; - if (WARN_ON(!in->cap_adap || !out->cap_adap)) + if (WARN_ON(!in->cap_adap || !out->cap_adap || !callback)) return NULL; tunnel = tb_tunnel_alloc(tb, 3, TB_TUNNEL_DP); diff --git a/drivers/thunderbolt/tunnel.h b/drivers/thunderbolt/tunnel.h index 4878763a82b3..7d1d255ab5a7 100644 --- a/drivers/thunderbolt/tunnel.h +++ b/drivers/thunderbolt/tunnel.h @@ -66,8 +66,8 @@ enum tb_tunnel_state { * @dprx_canceled: Was DPRX capabilities read poll canceled * @dprx_timeout: If set DPRX capabilities read poll work will timeout after this passes * @dprx_work: Worker that is scheduled to poll completion of DPRX capabilities read - * @callback: Optional callback called when DP tunnel is fully activated - * @callback_data: Optional data for @callback + * @callback: Callback called when DP tunnel is fully activated + * @callback_data: Data for @callback * @paths: All paths required by the tunnel */ struct tb_tunnel { @@ -117,7 +117,9 @@ struct tb_tunnel *tb_tunnel_alloc_pci(struct tb *tb, struct tb_port *up, bool tb_tunnel_reserved_pci(struct tb_port *port, int *reserved_up, int *reserved_down); struct tb_tunnel *tb_tunnel_discover_dp(struct tb *tb, struct tb_port *in, - bool alloc_hopid); + bool alloc_hopid, + void (*callback)(struct tb_tunnel *, void *), + void *callback_data); struct tb_tunnel *tb_tunnel_alloc_dp(struct tb *tb, struct tb_port *in, struct tb_port *out, int link_nr, int max_up, int max_down, From 1fd1f67c94d30889377bba774f032ec15e8b9299 Mon Sep 17 00:00:00 2001 From: Sven Peter Date: Sat, 29 Aug 2026 10:08:35 +0200 Subject: [PATCH 03/10] thunderbolt: Fix domain reference leak when DPRX read is canceled tb_tunnel_one_dp() takes a domain reference which is only dropped once tb_dp_tunnel_active() has run on the work queue. If that work is cancelled that reference is leaked. Since commit f5cc545f5969 ("thunderbolt: Wait for tb_domain_release() to complete when driver is removed") instead of just leaking memory this now also blocks in the completion wait forever when unbinding the driver. This can be triggered whenever a DP tunnel is torn down before the DPRX read has completed, e.g. by unplugging within the timeout, and then unbinding the driver. That reference only exists to keep the domain around while the DPRX work is scheduled so let the work itself own it: take it in tb_dp_dprx_start() and drop it in both places that end the work. Get/put are then paired inside the same file and it doesn't matter anymore if the callback ever runs. Fixes: d6d458d42e1e ("thunderbolt: Handle DisplayPort tunnel activation asynchronously") Cc: stable@vger.kernel.org Signed-off-by: Sven Peter Signed-off-by: Mika Westerberg --- drivers/thunderbolt/tb.c | 6 +----- drivers/thunderbolt/tunnel.c | 12 +++++++++--- 2 files changed, 10 insertions(+), 8 deletions(-) diff --git a/drivers/thunderbolt/tb.c b/drivers/thunderbolt/tb.c index 20f0f8f4bdf7..1c5e0c9159c4 100644 --- a/drivers/thunderbolt/tb.c +++ b/drivers/thunderbolt/tb.c @@ -1966,8 +1966,6 @@ static void tb_dp_tunnel_active(struct tb_tunnel *tunnel, void *data) tb_dp_resource_unavailable(tb, in, "DPRX negotiation failed"); } mutex_unlock(&tb->lock); - - tb_domain_put(tb); } static void tb_tunnel_one_dp(struct tb *tb, struct tb_port *in, @@ -2028,8 +2026,7 @@ static void tb_tunnel_one_dp(struct tb *tb, struct tb_port *in, available_up, available_down); tunnel = tb_tunnel_alloc_dp(tb, in, out, link_nr, available_up, - available_down, tb_dp_tunnel_active, - tb_domain_get(tb)); + available_down, tb_dp_tunnel_active, tb); if (!tunnel) { tb_port_dbg(out, "could not allocate DP tunnel\n"); goto err_reclaim_usb; @@ -2050,7 +2047,6 @@ err_free: tb_tunnel_put(tunnel); err_reclaim_usb: tb_reclaim_usb3_bandwidth(tb, in, out); - tb_domain_put(tb); err_detach_group: tb_detach_bandwidth_group(in); err_dealloc_dp: diff --git a/drivers/thunderbolt/tunnel.c b/drivers/thunderbolt/tunnel.c index 6ac44b371266..4f3a3717e37d 100644 --- a/drivers/thunderbolt/tunnel.c +++ b/drivers/thunderbolt/tunnel.c @@ -1111,15 +1111,17 @@ static void tb_dp_dprx_work(struct work_struct *work) tunnel->callback(tunnel, tunnel->callback_data); tb_tunnel_put(tunnel); + tb_domain_put(tb); } static int tb_dp_dprx_start(struct tb_tunnel *tunnel) { /* - * Bump up the reference to keep the tunnel around. It will be - * dropped in tb_dp_dprx_stop() once the tunnel is deactivated. + * Bump up the references to keep the tunnel and the domain around + * until the work has run or has been canceled. */ tb_tunnel_get(tunnel); + tb_domain_get(tunnel->tb); tunnel->dprx_started = true; tunnel->dprx_timeout = dprx_timeout_to_ktime(dprx_timeout); @@ -1130,11 +1132,15 @@ static int tb_dp_dprx_start(struct tb_tunnel *tunnel) static void tb_dp_dprx_stop(struct tb_tunnel *tunnel) { + struct tb *tb = tunnel->tb; + if (tunnel->dprx_started) { tunnel->dprx_started = false; tunnel->dprx_canceled = true; - if (cancel_delayed_work(&tunnel->dprx_work)) + if (cancel_delayed_work(&tunnel->dprx_work)) { tb_tunnel_put(tunnel); + tb_domain_put(tb); + } } } From 419fa32fa5bc2d7fb9d0d15d5630b1c1090808b3 Mon Sep 17 00:00:00 2001 From: Sven Peter Date: Sat, 29 Aug 2026 10:08:36 +0200 Subject: [PATCH 04/10] thunderbolt: Don't access a DP tunnel after its DPRX read was canceled tb_dp_dprx_work() checks ->dprx_canceled before it takes tb->lock so it misses a tb_dp_dprx_stop() that could not cancel the already running work. It then polls the DPRX capabilities and runs the callback for a tunnel that has already been torn down while the domain is suspending or going away. This can be hit by cancelling the DPRX read from outside the ordered tb->wq: During suspend tb_disconnect_and_release_dp() does just this and with a later patch tb_stop() will do it as well. The latter in combination with the Apple NHI where the DPRX read never completed is how I hit this. Check the flag with tb->lock held instead and check it again in tb_dp_tunnel_active() because the callback runs after the lock has been dropped again. Also clear the flag in tb_dp_dprx_start() so that it only ever describes the work that is currently in flight. Fixes: d6d458d42e1e ("thunderbolt: Handle DisplayPort tunnel activation asynchronously") Cc: stable@vger.kernel.org Signed-off-by: Sven Peter Signed-off-by: Mika Westerberg --- drivers/thunderbolt/tb.c | 12 ++++++++++++ drivers/thunderbolt/tunnel.c | 11 +++++++++-- 2 files changed, 21 insertions(+), 2 deletions(-) diff --git a/drivers/thunderbolt/tb.c b/drivers/thunderbolt/tb.c index 1c5e0c9159c4..70a1a585bd54 100644 --- a/drivers/thunderbolt/tb.c +++ b/drivers/thunderbolt/tb.c @@ -1912,6 +1912,18 @@ static void tb_dp_tunnel_active(struct tb_tunnel *tunnel, void *data) struct tb *tb = data; mutex_lock(&tb->lock); + + /* + * If the DPRX read was canceled the tunnel is already being torn + * down by whoever canceled it. Do not touch the adapters here + * because the routers may be gone by now. + */ + if (tunnel->dprx_canceled) { + tb_tunnel_dbg(tunnel, "DPRX read canceled, not activating\n"); + mutex_unlock(&tb->lock); + return; + } + if (tb_tunnel_is_active(tunnel)) { int consumed_up, consumed_down, ret; diff --git a/drivers/thunderbolt/tunnel.c b/drivers/thunderbolt/tunnel.c index 4f3a3717e37d..d015e54abde4 100644 --- a/drivers/thunderbolt/tunnel.c +++ b/drivers/thunderbolt/tunnel.c @@ -1093,8 +1093,14 @@ static void tb_dp_dprx_work(struct work_struct *work) struct tb_tunnel *tunnel = container_of(work, typeof(*tunnel), dprx_work.work); struct tb *tb = tunnel->tb; + /* + * The DPRX read can be canceled while this work is waiting for + * tb->lock. Check the flag only once it is held: while the lock is + * held the tunnel cannot be torn down under us and the adapters are + * safe to access. + */ + mutex_lock(&tb->lock); if (!tunnel->dprx_canceled) { - mutex_lock(&tb->lock); if (tb_dp_is_usb4(tunnel->src_port->sw) && tb_dp_wait_dprx(tunnel, TB_DPRX_WAIT_TIMEOUT)) { if (ktime_before(ktime_get(), tunnel->dprx_timeout)) { @@ -1106,8 +1112,8 @@ static void tb_dp_dprx_work(struct work_struct *work) } else { tb_tunnel_set_active(tunnel, true); } - mutex_unlock(&tb->lock); } + mutex_unlock(&tb->lock); tunnel->callback(tunnel, tunnel->callback_data); tb_tunnel_put(tunnel); @@ -1124,6 +1130,7 @@ static int tb_dp_dprx_start(struct tb_tunnel *tunnel) tb_domain_get(tunnel->tb); tunnel->dprx_started = true; + tunnel->dprx_canceled = false; tunnel->dprx_timeout = dprx_timeout_to_ktime(dprx_timeout); queue_delayed_work(tunnel->tb->wq, &tunnel->dprx_work, 0); From c222f80be55f6c17851af1c68a70ae4046c848e2 Mon Sep 17 00:00:00 2001 From: Sven Peter Date: Sat, 29 Aug 2026 10:08:37 +0200 Subject: [PATCH 05/10] thunderbolt: Mark discovered tunnels as active Discovered tunnels that have been activated by whatever was running before us stay in TB_TUNNEL_INACTIVE until hibernation restore such that anything depending on tb_tunnel_is_active() skips them. Mark them active during discovery instead. These now also emit TUNNEL_EVENT=activated uevents during discovery. Fixes: d6d458d42e1e ("thunderbolt: Handle DisplayPort tunnel activation asynchronously") Cc: stable@vger.kernel.org Signed-off-by: Sven Peter Signed-off-by: Mika Westerberg --- drivers/thunderbolt/tunnel.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/thunderbolt/tunnel.c b/drivers/thunderbolt/tunnel.c index d015e54abde4..ffd2d04b3bfa 100644 --- a/drivers/thunderbolt/tunnel.c +++ b/drivers/thunderbolt/tunnel.c @@ -510,6 +510,7 @@ struct tb_tunnel *tb_tunnel_discover_pci(struct tb *tb, struct tb_port *down, goto err_deactivate; } + tb_tunnel_set_active(tunnel, true); tb_tunnel_dbg(tunnel, "discovered\n"); return tunnel; @@ -1674,6 +1675,7 @@ struct tb_tunnel *tb_tunnel_discover_dp(struct tb *tb, struct tb_port *in, tb_dp_dump(tunnel); + tb_tunnel_set_active(tunnel, true); tb_tunnel_dbg(tunnel, "discovered\n"); return tunnel; @@ -2306,6 +2308,7 @@ struct tb_tunnel *tb_tunnel_discover_usb3(struct tb *tb, struct tb_port *down, tb_usb3_reclaim_available_bandwidth; } + tb_tunnel_set_active(tunnel, true); tb_tunnel_dbg(tunnel, "discovered\n"); return tunnel; From 0b457c6733f73421a3fb03786f13d67c618b5119 Mon Sep 17 00:00:00 2001 From: Sven Peter Date: Sat, 29 Aug 2026 10:08:38 +0200 Subject: [PATCH 06/10] thunderbolt: Tear down inactive DP tunnels when the domain is stopped tb_stop() only tears down DMA tunnels so a DP tunnel that is still waiting for dprx_work to complete keeps that work queued while the routers are removed and the control channel is stopped. The work only stops once the DPRX timeout has passed and because it requeues itself until then the flush_workqueue() in tb_domain_remove() won't wait for its final run. The callback then runs against a domain that is already torn down. A reference to that domain is kept so the completion waiting for that domain to disappear in unbind will block until the timeout is eventually reached. Tear down DP tunnels that are not active yet as well which also cancels that work. Tunnels for displays that are already alive are untouched and keep working. Fixes: d6d458d42e1e ("thunderbolt: Handle DisplayPort tunnel activation asynchronously") Cc: stable@vger.kernel.org Signed-off-by: Sven Peter Signed-off-by: Mika Westerberg --- drivers/thunderbolt/tb.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/thunderbolt/tb.c b/drivers/thunderbolt/tb.c index 70a1a585bd54..ba63610d2ce2 100644 --- a/drivers/thunderbolt/tb.c +++ b/drivers/thunderbolt/tb.c @@ -2960,11 +2960,12 @@ static void tb_stop(struct tb *tb) /* tunnels are only present after everything has been initialized */ list_for_each_entry_safe(tunnel, n, &tcm->tunnel_list, list) { /* - * DMA tunnels require the driver to be functional so we - * tear them down. Other protocol tunnels can be left - * intact. + * DMA tunnels and DP tunnels which are not yet active require + * the driver to be functional so we tear them down. + * Other protocol tunnels can be left intact. */ - if (tb_tunnel_is_dma(tunnel)) + if (tb_tunnel_is_dma(tunnel) || + (tb_tunnel_is_dp(tunnel) && !tb_tunnel_is_active(tunnel))) tb_tunnel_deactivate(tunnel); tb_tunnel_put(tunnel); } From 4310c6b8e75d6a47f7548e5948fbe6318aa4440a Mon Sep 17 00:00:00 2001 From: Fedor Pchelkin Date: Mon, 31 Aug 2026 10:58:09 +0300 Subject: [PATCH 07/10] thunderbolt: Fix NULL dereference in tb_remove_work() 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 Signed-off-by: Mika Westerberg --- drivers/thunderbolt/tb.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/thunderbolt/tb.c b/drivers/thunderbolt/tb.c index ba63610d2ce2..4e5d0578fd4b 100644 --- a/drivers/thunderbolt/tb.c +++ b/drivers/thunderbolt/tb.c @@ -3285,11 +3285,11 @@ static void tb_remove_work(struct work_struct *work) struct tb *tb = tcm_to_tb(tcm); mutex_lock(&tb->lock); - if (tb->root_switch) + if (tb->root_switch) { tb_free_unplugged_children(tb->root_switch); + tb_free_unplugged_xdomains(tb->root_switch); + } mutex_unlock(&tb->lock); - - tb_free_unplugged_xdomains(tb->root_switch); } static int tb_runtime_resume(struct tb *tb) From a02188ddc24b7872f0c5e0c5873f827318717803 Mon Sep 17 00:00:00 2001 From: Mika Westerberg Date: Thu, 7 May 2026 08:06:48 +0300 Subject: [PATCH 08/10] thunderbolt: Fix KASAN reported use-after-free when request is canceled Alan reported that when doing stress testing sometimes KASAN notices use-after-free during control channel operation (stripped down keeping the relevant parts): BUG: KASAN: slab-use-after-free in tb_cfg_request_sync+0x240/0x250 [thunderbolt] Read of size 24 at addr ffff88811067f290 by task kworker/u40:2/1760 tb_cfg_request_sync+0x240/0x250 [thunderbolt] tb_cfg_read_raw+0x367/0x510 [thunderbolt] tb_cfg_read+0xec/0x240 [thunderbolt] tb_port_get_link_generation+0x258/0x420 [thunderbolt] tb_usb3_consumed_bandwidth+0x1c1/0x2c0 [thunderbolt] tb_tunnel_consumed_bandwidth+0xfd/0x910 [thunderbolt] tb_available_bandwidth+0x5f2/0xeb0 [thunderbolt] tb_recalc_estimated_bandwidth+0x2a0/0x1bc0 [thunderbolt] tb_handle_dp_bandwidth_request+0x1897/0x5e20 [thunderbolt] process_one_work+0x675/0x1230 worker_thread+0x5e6/0xf70 kthread+0x365/0x470 ret_from_fork+0x54d/0x710 ret_from_fork_asm+0x1a/0x30 Allocated by task 1760: __kmalloc_cache_noprof+0x1ee/0x550 tb_cfg_read_raw+0x1d3/0x510 [thunderbolt] tb_cfg_read+0xec/0x240 [thunderbolt] tb_port_get_link_generation+0x258/0x420 [thunderbolt] tb_usb3_consumed_bandwidth+0x1c1/0x2c0 [thunderbolt] tb_tunnel_consumed_bandwidth+0xfd/0x910 [thunderbolt] tb_available_bandwidth+0x5f2/0xeb0 [thunderbolt] tb_recalc_estimated_bandwidth+0x2a0/0x1bc0 [thunderbolt] tb_handle_dp_bandwidth_request+0x1897/0x5e20 [thunderbolt] process_one_work+0x675/0x1230 worker_thread+0x5e6/0xf70 kthread+0x365/0x470 ret_from_fork+0x54d/0x710 ret_from_fork_asm+0x1a/0x30 Freed by task 926: kfree+0x18f/0x4a0 tb_cfg_request_put+0xb7/0xe0 [thunderbolt] tb_cfg_request_work+0x82/0x120 [thunderbolt] process_one_work+0x675/0x1230 worker_thread+0x5e6/0xf70 kthread+0x365/0x470 ret_from_fork+0x54d/0x710 ret_from_fork_asm+0x1a/0x30 Second to last potentially related work creation: __queue_work+0x575/0xd00 queue_work_on+0x77/0x80 tb_cfg_request_cancel+0xc7/0x260 [thunderbolt] tb_cfg_request_sync+0x1f6/0x250 [thunderbolt] tb_cfg_read_raw+0x367/0x510 [thunderbolt] tb_cfg_read+0xec/0x240 [thunderbolt] tb_port_get_link_generation+0x258/0x420 [thunderbolt] tb_usb3_consumed_bandwidth+0x1c1/0x2c0 [thunderbolt] tb_tunnel_consumed_bandwidth+0xfd/0x910 [thunderbolt] tb_available_bandwidth+0x5f2/0xeb0 [thunderbolt] tb_recalc_estimated_bandwidth+0x2a0/0x1bc0 [thunderbolt] tb_handle_dp_bandwidth_request+0x1897/0x5e20 [thunderbolt] process_one_work+0x675/0x1230 worker_thread+0x5e6/0xf70 kthread+0x365/0x470 ret_from_fork+0x54d/0x710 ret_from_fork_asm+0x1a/0x30 The last stack trace is helpful because it shows that we are cancelling a request and looking at tb_cfg_request_cancel() what might happen is that tb_cfg_request_work() completes right before tb_cfg_request_cancel() starts and because of this it will call schedule_work() queueing the same work to run again. However, it is already removed from the request_queue and reference count is dropped so when tb_cfg_request_work() triggers again it will access memory that is already released. Fix this so that we first make sure a cancelled request is not handed away from tb_cfg_request_find() or scheduled to run. Then instead of relying on the worker to clean up the request we will do it in tb_cfg_request_cancel() after the work is canceled from running. Make tb_cfg_request_dequeue() release the request only if it was actually removed from the queue. Reported-by: Alan Borzeszkowski Fixes: d7f781bfdbf4 ("thunderbolt: Rework control channel to be more reliable") Cc: stable@vger.kernel.org Signed-off-by: Mika Westerberg --- drivers/thunderbolt/ctl.c | 63 +++++++++++++++++++-------------------- 1 file changed, 31 insertions(+), 32 deletions(-) diff --git a/drivers/thunderbolt/ctl.c b/drivers/thunderbolt/ctl.c index cd47b627f97b..965988b18608 100644 --- a/drivers/thunderbolt/ctl.c +++ b/drivers/thunderbolt/ctl.c @@ -73,7 +73,6 @@ struct tb_ctl { #define tb_ctl_dbg_once(ctl, format, arg...) \ dev_dbg_once((ctl)->nhi->dev, format, ## arg) -static DECLARE_WAIT_QUEUE_HEAD(tb_cfg_request_cancel_queue); /* Serializes access to request kref_get/put */ static DEFINE_MUTEX(tb_cfg_request_lock); @@ -133,41 +132,42 @@ void tb_cfg_request_put(struct tb_cfg_request *req) static int tb_cfg_request_enqueue(struct tb_ctl *ctl, struct tb_cfg_request *req) { + tb_cfg_request_get(req); + WARN_ON(test_bit(TB_CFG_REQUEST_ACTIVE, &req->flags)); WARN_ON(req->ctl); - mutex_lock(&ctl->request_queue_lock); + guard(mutex)(&ctl->request_queue_lock); if (!ctl->running) { - mutex_unlock(&ctl->request_queue_lock); + tb_cfg_request_put(req); return -ENOTCONN; } req->ctl = ctl; list_add_tail(&req->list, &ctl->request_queue); set_bit(TB_CFG_REQUEST_ACTIVE, &req->flags); - mutex_unlock(&ctl->request_queue_lock); return 0; } +static bool tb_cfg_request_is_active(struct tb_cfg_request *req) +{ + return test_bit(TB_CFG_REQUEST_ACTIVE, &req->flags); +} + +static bool tb_cfg_request_is_canceled(struct tb_cfg_request *req) +{ + return test_bit(TB_CFG_REQUEST_CANCELED, &req->flags); +} + static void tb_cfg_request_dequeue(struct tb_cfg_request *req) { struct tb_ctl *ctl = req->ctl; - mutex_lock(&ctl->request_queue_lock); - if (!test_bit(TB_CFG_REQUEST_ACTIVE, &req->flags)) { - mutex_unlock(&ctl->request_queue_lock); - return; + guard(mutex)(&ctl->request_queue_lock); + if (tb_cfg_request_is_active(req)) { + list_del(&req->list); + clear_bit(TB_CFG_REQUEST_ACTIVE, &req->flags); + tb_cfg_request_put(req); } - - list_del(&req->list); - clear_bit(TB_CFG_REQUEST_ACTIVE, &req->flags); - if (test_bit(TB_CFG_REQUEST_CANCELED, &req->flags)) - wake_up(&tb_cfg_request_cancel_queue); - mutex_unlock(&ctl->request_queue_lock); -} - -static bool tb_cfg_request_is_active(struct tb_cfg_request *req) -{ - return test_bit(TB_CFG_REQUEST_ACTIVE, &req->flags); } static struct tb_cfg_request * @@ -178,7 +178,7 @@ tb_cfg_request_find(struct tb_ctl *ctl, struct ctl_pkg *pkg) mutex_lock(&pkg->ctl->request_queue_lock); list_for_each_entry(iter, &pkg->ctl->request_queue, list) { tb_cfg_request_get(iter); - if (iter->match(iter, pkg)) { + if (!tb_cfg_request_is_canceled(iter) && iter->match(iter, pkg)) { req = iter; break; } @@ -512,8 +512,11 @@ static void tb_ctl_rx_callback(struct tb_ring *ring, struct ring_frame *frame, trace_tb_rx(pkg->ctl->index, frame->eof, pkg->buffer, frame->size, !req); if (req) { - if (req->copy(req, pkg)) - schedule_work(&req->work); + scoped_guard(mutex, &pkg->ctl->request_queue_lock) { + if (!tb_cfg_request_is_canceled(req) && + req->copy(req, pkg)) + schedule_work(&req->work); + } tb_cfg_request_put(req); } @@ -525,11 +528,10 @@ static void tb_cfg_request_work(struct work_struct *work) { struct tb_cfg_request *req = container_of(work, typeof(*req), work); - if (!test_bit(TB_CFG_REQUEST_CANCELED, &req->flags)) + if (!tb_cfg_request_is_canceled(req)) req->callback(req->callback_data); tb_cfg_request_dequeue(req); - tb_cfg_request_put(req); } /** @@ -555,10 +557,9 @@ int tb_cfg_request(struct tb_ctl *ctl, struct tb_cfg_request *req, INIT_WORK(&req->work, tb_cfg_request_work); INIT_LIST_HEAD(&req->list); - tb_cfg_request_get(req); ret = tb_cfg_request_enqueue(ctl, req); if (ret) - goto err_put; + return ret; ret = tb_ctl_tx(ctl, req->request, req->request_size, req->request_type); @@ -572,9 +573,6 @@ int tb_cfg_request(struct tb_ctl *ctl, struct tb_cfg_request *req, err_dequeue: tb_cfg_request_dequeue(req); -err_put: - tb_cfg_request_put(req); - return ret; } @@ -588,9 +586,10 @@ err_put: */ void tb_cfg_request_cancel(struct tb_cfg_request *req, int err) { - set_bit(TB_CFG_REQUEST_CANCELED, &req->flags); - schedule_work(&req->work); - wait_event(tb_cfg_request_cancel_queue, !tb_cfg_request_is_active(req)); + scoped_guard(mutex, &req->ctl->request_queue_lock) + set_bit(TB_CFG_REQUEST_CANCELED, &req->flags); + cancel_work_sync(&req->work); + tb_cfg_request_dequeue(req); req->result.err = err; } From 96ff396e6e0eb754b73f1870ef7e72f2f80cdbe0 Mon Sep 17 00:00:00 2001 From: Mika Westerberg Date: Fri, 15 May 2026 10:28:40 +0300 Subject: [PATCH 09/10] thunderbolt: Use separate lock class for each ring When connected to another host and then unplugging cable lockdep triggers following: ====================================================== WARNING: possible circular locking dependency detected 7.1.0-rc2+ #1775 Tainted: G U ------------------------------------------------------ kworker/u16:6/312 is trying to acquire lock: ffff8881179c70a8 ((work_completion)(&ring->work)){+.+.}-{0:0}, at: __flush_work+0x3cf/0xd10 but task is already holding lock: ffff8881a8b810b0 (&net->connection_lock){+.+.}-{4:4}, at: tbnet_tear_down+0x110/0x720 [thunderbolt_net] which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #1 (&net->connection_lock){+.+.}-{4:4}: __mutex_lock+0x19a/0x2490 mutex_lock_nested+0x1b/0x30 tbnet_handle_packet+0x74c/0xd70 [thunderbolt_net] tb_xdomain_handle_request+0x37c/0x4b0 [thunderbolt] tb_domain_event_cb+0xc9/0x140 [thunderbolt] tb_ctl_handle_event+0xd6/0x2c0 [thunderbolt] tb_ctl_rx_callback+0x22c/0xa10 [thunderbolt] ring_work+0x715/0xcb0 [thunderbolt] process_one_work+0x902/0x1790 worker_thread+0x5cd/0xfe0 kthread+0x339/0x420 ret_from_fork+0x79a/0x9d0 ret_from_fork_asm+0x1a/0x30 -> #0 ((work_completion)(&ring->work)){+.+.}-{0:0}: __lock_acquire+0x1592/0x2640 lock_acquire+0x1a3/0x300 __flush_work+0x3e9/0xd10 flush_work+0x21/0x30 tb_ring_stop+0x240/0x840 [thunderbolt] tbnet_tear_down+0x2ff/0x720 [thunderbolt_net] tbnet_stop+0x47/0x1a0 [thunderbolt_net] __dev_close_many+0x19e/0x4e0 netif_close_many+0x1e8/0x640 unregister_netdevice_many_notify+0x6d3/0x22d0 unregister_netdevice_queue+0x2b9/0x3a0 unregister_netdev+0x1c/0x70 tbnet_remove+0x52/0xb0 [thunderbolt_net] tb_service_remove+0x8a/0xe0 [thunderbolt] device_remove+0xc5/0x190 device_release_driver_internal+0x3db/0x590 device_release_driver+0x12/0x20 bus_remove_device+0x2c1/0x580 device_del+0x3d9/0x9f0 device_unregister+0x17/0xc0 unregister_service+0x46/0x60 [thunderbolt] device_for_each_child_reverse+0xfa/0x180 tb_xdomain_unregister+0x57/0xe0 [thunderbolt] unregister_unplugged_xdomain+0x101/0x1a0 [thunderbolt] bus_for_each_dev+0x111/0x1a0 tb_domain_unregister_unplugged_xdomains+0x98/0xe0 [thunderbolt] tb_handle_hotplug+0xc3/0x2bb0 [thunderbolt] process_one_work+0x902/0x1790 worker_thread+0x5cd/0xfe0 kthread+0x339/0x420 ret_from_fork+0x79a/0x9d0 ret_from_fork_asm+0x1a/0x30 other info that might help us debug this: Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(&net->connection_lock); lock((work_completion)(&ring->work)); lock(&net->connection_lock); lock((work_completion)(&ring->work)); This in fact is false positive because they involve unrelated rings (and unrelated work structures). In the first one it is ring 0 which is used for control traffic and in the second it is dealing with another ring used for the high-speed traffic. Fix this by using separate lock class for each ring worker. Signed-off-by: Mika Westerberg --- drivers/thunderbolt/nhi.c | 5 +++++ include/linux/thunderbolt.h | 3 +++ 2 files changed, 8 insertions(+) diff --git a/drivers/thunderbolt/nhi.c b/drivers/thunderbolt/nhi.c index 5809809f64d4..d4d1efa2afa0 100644 --- a/drivers/thunderbolt/nhi.c +++ b/drivers/thunderbolt/nhi.c @@ -15,6 +15,7 @@ #include #include #include +#include #include #include #include @@ -560,6 +561,8 @@ static struct tb_ring *tb_ring_alloc(struct tb_nhi *nhi, u32 hop, int size, INIT_LIST_HEAD(&ring->in_flight); INIT_WORK(&ring->work, ring_work); init_waitqueue_head(&ring->wait); + lockdep_register_key(&ring->lock_key); + lockdep_init_map(&ring->work.lockdep_map, "ring.work", &ring->lock_key, 0); ring->nhi = nhi; ring->hop = hop; @@ -599,6 +602,7 @@ err_free_descs: ring->size * sizeof(*ring->descriptors), ring->descriptors, ring->descriptors_dma); err_free_ring: + lockdep_unregister_key(&ring->lock_key); kfree(ring); return NULL; @@ -848,6 +852,7 @@ void tb_ring_free(struct tb_ring *ring) * to finish before freeing the ring. */ flush_work(&ring->work); + lockdep_unregister_key(&ring->lock_key); kfree(ring); } EXPORT_SYMBOL_GPL(tb_ring_free); diff --git a/include/linux/thunderbolt.h b/include/linux/thunderbolt.h index d48623fda79b..b62dfa52b149 100644 --- a/include/linux/thunderbolt.h +++ b/include/linux/thunderbolt.h @@ -22,6 +22,7 @@ struct device; #include #include #include +#include #include #include #include @@ -565,6 +566,7 @@ struct tb_nhi { * @interval_nsec: Interval counter if interrupt throttling is to be * used with this ring (in ns) * @wait: Used to signal that the ring may be empty now + * @lock_key: Lock validator class key per-ring */ struct tb_ring { spinlock_t lock; @@ -590,6 +592,7 @@ struct tb_ring { void *poll_data; unsigned int interval_nsec; wait_queue_head_t wait; + struct lock_class_key lock_key; }; /* Leave ring interrupt enabled on suspend */ From e6df180e976fd08673214bc1305a46a8536aea9a Mon Sep 17 00:00:00 2001 From: Mika Westerberg Date: Fri, 14 Aug 2026 06:35:52 +0200 Subject: [PATCH 10/10] Revert "thunderbolt: xdomain: Notify peers after enumeration" This reverts commit e027dba038f0008df9bc9575f5c3e803e90636c6. Alan noticed that this causes the peers send change properties to each other continuously. Reported-by: Borzeszkowski, Alan Cc: Milo Chen Signed-off-by: Mika Westerberg --- drivers/thunderbolt/xdomain.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/thunderbolt/xdomain.c b/drivers/thunderbolt/xdomain.c index c179bd751fe4..05442df0e99c 100644 --- a/drivers/thunderbolt/xdomain.c +++ b/drivers/thunderbolt/xdomain.c @@ -1814,7 +1814,6 @@ static void tb_xdomain_state_work(struct work_struct *work) tb_xdomain_failed(xd); } else { xd->state = XDOMAIN_STATE_ENUMERATED; - tb_xdomain_queue_properties_changed(xd); } break;