I now seem to have captured a call trace which may be related:
INFO: task kworker/0:0:7395 blocked for more than 120 seconds. [ 1088.815089] Not tainted 6.5.0-3-amd64 #1 Debian 6.5.8-1 [ 1088.815093] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [ 1088.815095] task:kworker/0:0 state:D stack:0 pid:7395 ppid:2 flags:0x00004000 [ 1088.815105] Workqueue: usb_hub_wq hub_event [usbcore] [ 1088.815183] Call Trace: [ 1088.815186] <TASK> [ 1088.815192] __schedule+0x3df/0xb80 [ 1088.815205] schedule+0x61/0xe0 [ 1088.815211] schedule_timeout+0x98/0x160 [ 1088.815218] ? __pfx_process_timeout+0x10/0x10 [ 1088.815227] wait_for_completion_timeout+0x83/0x170 [ 1088.815236] usb_start_wait_urb+0xa7/0x180 [usbcore] [ 1088.815299] usb_control_msg+0xef/0x150 [usbcore] [ 1088.815359] get_bMaxPacketSize0+0x5f/0xc0 [usbcore] [ 1088.815410] hub_port_init+0x418/0xff0 [usbcore] [ 1088.815465] hub_event+0x1207/0x1c10 [usbcore] [ 1088.815521] ? __schedule+0x3e7/0xb80 [ 1088.815528] process_one_work+0x1e1/0x3f0 [ 1088.815537] worker_thread+0x51/0x390 [ 1088.815542] ? _raw_spin_lock_irqsave+0x27/0x60 [ 1088.815549] ? __pfx_worker_thread+0x10/0x10 [ 1088.815554] kthread+0xf7/0x130 [ 1088.815562] ? __pfx_kthread+0x10/0x10 [ 1088.815570] ret_from_fork+0x34/0x50 [ 1088.815578] ? __pfx_kthread+0x10/0x10 [ 1088.815584] ret_from_fork_asm+0x1b/0x30 [ 1088.815597] </TASK> [ 1209.646327] INFO: task kworker/0:0:7395 blocked for more than 241 seconds. [ 1209.646341] Not tainted 6.5.0-3-amd64 #1 Debian 6.5.8-1 [ 1209.646345] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [ 1209.646348] task:kworker/0:0 state:D stack:0 pid:7395 ppid:2 flags:0x00004000 [ 1209.646363] Workqueue: usb_hub_wq hub_event [usbcore] [ 1209.646465] Call Trace: [ 1209.646469] <TASK> [ 1209.646476] __schedule+0x3df/0xb80 [ 1209.646493] schedule+0x61/0xe0 [ 1209.646501] schedule_timeout+0x98/0x160 [ 1209.646509] ? __pfx_process_timeout+0x10/0x10 [ 1209.646520] wait_for_completion_timeout+0x83/0x170 [ 1209.646532] usb_start_wait_urb+0xa7/0x180 [usbcore] [ 1209.646620] usb_control_msg+0xef/0x150 [usbcore] [ 1209.646703] get_bMaxPacketSize0+0x5f/0xc0 [usbcore] [ 1209.646772] hub_port_init+0x418/0xff0 [usbcore] [ 1209.646849] hub_event+0x1207/0x1c10 [usbcore] [ 1209.646925] ? __schedule+0x3e7/0xb80 [ 1209.646935] process_one_work+0x1e1/0x3f0 [ 1209.646946] worker_thread+0x51/0x390 [ 1209.646954] ? _raw_spin_lock_irqsave+0x27/0x60 [ 1209.646962] ? __pfx_worker_thread+0x10/0x10 [ 1209.646969] kthread+0xf7/0x130 [ 1209.646980] ? __pfx_kthread+0x10/0x10 [ 1209.646990] ret_from_fork+0x34/0x50 [ 1209.647002] ? __pfx_kthread+0x10/0x10 [ 1209.647011] ret_from_fork_asm+0x1b/0x30 [ 1209.647028] </TASK> [ 1330.478037] INFO: task kworker/0:0:7395 blocked for more than 362 seconds. [ 1330.478048] Not tainted 6.5.0-3-amd64 #1 Debian 6.5.8-1 [ 1330.478052] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [ 1330.478055] task:kworker/0:0 state:D stack:0 pid:7395 ppid:2 flags:0x00004000 [ 1330.478066] Workqueue: usb_hub_wq hub_event [usbcore] [ 1330.478149] Call Trace: [ 1330.478152] <TASK> [ 1330.478158] __schedule+0x3df/0xb80 [ 1330.478172] schedule+0x61/0xe0 [ 1330.478179] schedule_timeout+0x98/0x160 [ 1330.478186] ? __pfx_process_timeout+0x10/0x10 [ 1330.478196] wait_for_completion_timeout+0x83/0x170 [ 1330.478206] usb_start_wait_urb+0xa7/0x180 [usbcore] [ 1330.478276] usb_control_msg+0xef/0x150 [usbcore] [ 1330.478342] get_bMaxPacketSize0+0x5f/0xc0 [usbcore] [ 1330.478397] hub_port_init+0x418/0xff0 [usbcore] [ 1330.478458] hub_event+0x1207/0x1c10 [usbcore] [ 1330.478520] ? __schedule+0x3e7/0xb80 [ 1330.478527] process_one_work+0x1e1/0x3f0 [ 1330.478536] worker_thread+0x51/0x390 [ 1330.478542] ? _raw_spin_lock_irqsave+0x27/0x60 [ 1330.478549] ? __pfx_worker_thread+0x10/0x10 [ 1330.478555] kthread+0xf7/0x130 [ 1330.478564] ? __pfx_kthread+0x10/0x10 [ 1330.478572] ret_from_fork+0x34/0x50 [ 1330.478580] ? __pfx_kthread+0x10/0x10 [ 1330.478588] ret_from_fork_asm+0x1b/0x30 [ 1330.478601] </TASK> However this was on hotplugging a garmin Etrex 20 which is always very slow and after increasing /sys/module/usbcore/parameters/initial_descriptor_timeout to a very large value (600000 !), so it may not be related.
