usb: ulpi: ulpi_init should be used in subsys_initcall
PATCH 1/1 usb: ulpi: ulpi_init should be used in subsys_initcall
The intention of this change is to fix below kernel panic when USB_ULPI_BUS was configured as buildin.
[0.746856] kernel BUG at drivers/base/driver.c:153! [0.752418] invalid opcode: 0000 [#1] PREEMPT SMP [0.757804] Modules linked in: [0.893985] Call Trace: [0.896729] [<ffffffff81870cb1>] ? ulpi_register_driver+0x21/0x30 [0.903654] [<ffffffff823d5cac>] tusb1210_driver_init+0x10/0x12 [0.910386] [<ffffffff81000318>] do_one_initcall+0xd8/0x200 [0.916729] [<ffffffff8239b094>] kernel_init_freeable+0x196/0x21e [0.923655] [<ffffffff81affed0>] ? rest_init+0x90/0x90 [0.929509] [<ffffffff81affede>] kernel_init+0xe/0xf0 [0.935266] [<ffffffff81b0f442>] ret_from_fork+0x42/0x70 [0.941315] [<ffffffff81affed0>] ? rest_init+0x90/0x90
--- drivers/usb/common/ulpi.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/usb/common/ulpi.c +1 -1
@@ -242,7 +242,7 @@ static int __init ulpi_init(void) {return bus_register (&ulpi_bus); } -module_init (ulpi_init); +subsys_initcall (ulpi_init); static void __exit ulpi_exit(void) {
Signature · 1 line
2.1.4
Heikki Krogerus
maintainerQuoted 2 lines · Lu Baolu
The intention of this change is to fix below kernel panic when
USB_ULPI_BUSwas configured as buildin.
That is actually incorrect. Having the bus build-in does not cause this panic..
Quoted 4 lines
[0.746856] kernel BUG at drivers/base/driver.c:153! [0.752418] invalid opcode: 0000 [#1] PREEMPT SMP [0.757804] Modules linked in: [0.893985] Call Trace:
Quoted 8 lines
[0.896729] [<ffffffff81870cb1>] ? ulpi_register_driver+0x21/0x30 [0.903654] [<ffffffff823d5cac>] tusb1210_driver_init+0x10/0x12 [0.910386] [<ffffffff81000318>] do_one_initcall+0xd8/0x200 [0.916729] [<ffffffff8239b094>] kernel_init_freeable+0x196/0x21e [0.923655] [<ffffffff81affed0>] ? rest_init+0x90/0x90 [0.929509] [<ffffffff81affede>] kernel_init+0xe/0xf0 [0.935266] [<ffffffff81b0f442>] ret_from_fork+0x42/0x70 [0.941315] [<ffffffff81affed0>] ? rest_init+0x90/0x90
This happens if the PHY drivers are build-in. If the bus is build-in, but the drivers are modules, you are still fine.
Maybe it would have been better to explain that this is addressing an issue with the execution sequence, and consider Sasha's patch as the actual fix for panic.
Thanks guys,
Signature · 1 line
heikki
Quoted 18 lines · Heikki Krogerus
Quoted 2 lines · Lu Baolu
The intention of this change is to fix below kernel panic when
USB_ULPI_BUSwas configured as buildin.That is actually incorrect. Having the bus build-in does not cause this panic..
Quoted 4 lines
[0.746856] kernel BUG at drivers/base/driver.c:153! [0.752418] invalid opcode: 0000 [#1] PREEMPT SMP [0.757804] Modules linked in: [0.893985] Call Trace:Quoted 8 lines
[0.896729] [<ffffffff81870cb1>] ? ulpi_register_driver+0x21/0x30 [0.903654] [<ffffffff823d5cac>] tusb1210_driver_init+0x10/0x12 [0.910386] [<ffffffff81000318>] do_one_initcall+0xd8/0x200 [0.916729] [<ffffffff8239b094>] kernel_init_freeable+0x196/0x21e [0.923655] [<ffffffff81affed0>] ? rest_init+0x90/0x90 [0.929509] [<ffffffff81affede>] kernel_init+0xe/0xf0 [0.935266] [<ffffffff81b0f442>] ret_from_fork+0x42/0x70 [0.941315] [<ffffffff81affed0>] ? rest_init+0x90/0x90
Quoted 2 lines
This happens if the PHY drivers are build-in. If the bus is build-in, but the drivers are modules, you are still fine.
I ever tried the config of bus "=y" and driver "=n". It results in panic as well. The call trace looks like below.
1.058876] [<ffffffff8165a6d2>] device_add+0x402/0x640
[1.064829] [<ffffffff8165a92e>] device_register+0x1e/0x30 [1.071072] [<ffffffff81870bdc>] ulpi_register_interface+0x14c/0x1f0 [1.078291] [<ffffffff817c68d4>] dwc3_ulpi_init+0x24/0x60 [1.084437] [<ffffffff817be052>] dwc3_probe+0x802/0x1650 [1.090487] [<ffffffff8165fb54>] platform_drv_probe+0x34/0xa0 [1.097022] [<ffffffff8165d9e9>] driver_probe_device+0x209/0x4b0 [1.103850] [<ffffffff8165dc90>] ? driver_probe_device+0x4b0/0x4b0 [1.110871] [<ffffffff8165dccb>] __device_attach+0x3b/0x40 [1.117114] [<ffffffff8165b6d3>] bus_for_each_drv+0x63/0xa0 [1.123454] [<ffffffff8165d778>] device_attach+0x98/0xb0 [1.129503] [<ffffffff8165cb60>] bus_probe_device+0xa0/0xc0 [1.135843] [<ffffffff8165a73f>] device_add+0x46f/0x640 [1.141799] [<ffffffff81090fe0>] ? __insert_resource+0x60/0x150 [1.148530] [<ffffffff8165f780>] platform_device_add+0xd0/0x2b0 [1.155260] [<ffffffff817c7386>] dwc3_pci_probe+0xf6/0x2c0 [1.161505] [<ffffffff8149d86c>] pci_device_probe+0x8c/0xf0 [1.167846] [<ffffffff8165d9e9>] driver_probe_device+0x209/0x4b0 [1.174673] [<ffffffff8165dd6b>] __driver_attach+0x9b/0xa0 [1.180917] [<ffffffff8165dcd0>] ? __device_attach+0x40/0x40 [1.187354] [<ffffffff8165b60b>] bus_for_each_dev+0x6b/0xb0 [1.193694] [<ffffffff8165d2be>] driver_attach+0x1e/0x20 [1.199742] [<ffffffff8165ce80>] bus_add_driver+0x180/0x250 [1.206086] [<ffffffff823e453c>] ? ftrace_define_fields_dwc3_log_trb+0x126/0x126 [1.214474] [<ffffffff8165e5d4>] driver_register+0x64/0xf0 [1.220718] [<ffffffff8149c51b>] __pci_register_driver+0x4b/0x50 [1.227546] [<ffffffff823e4555>] dwc3_pci_driver_init+0x19/0x1b [1.234277] [<ffffffff81000318>] do_one_initcall+0xd8/0x200 [1.240619] [<ffffffff8239b094>] kernel_init_freeable+0x196/0x21e [1.247545] [<ffffffff81affbd0>] ? rest_init+0x90/0x90 [1.253399] [<ffffffff81affbde>] kernel_init+0xe/0xf0 [1.259156] [<ffffffff81b0f142>] ret_from_fork+0x42/0x70 [1.265205] [<ffffffff81affbd0>] ? rest_init+0x90/0x90
Quoted 4 lines
Maybe it would have been better to explain that this is addressing an issue with the execution sequence, and consider Sasha's patch as the actual fix for panic.
Fair enough. I will resend the patch.
Quoted 4 lines
Thanks guys,
Thanks,
Baolu