sporadic system freeze

mipko · December 9, 2014, 8:42pm

Hello,

I have sporadic system freeze that can be resolved only by reseting the board.
This is fresh L4T21.1 install with booting from SSD. Beside that I have added this line into rc.local to prevent USB dropouts:

echo -1 > /sys/module/usbcore/parameters/autosuspend

I have logitech wireless kbd with touchpad and the unified receiver is the only device connected to USB 3.0 port.

I have also modified extlinux.conf to get USB 3.0 support.

System sporadically freezes - in a way that keyboard or mouse or both are dead. The same scenario happens sporadically when USB 2.0 HUB with usb mouse and keyboard is attached to USB 3.0 port. When this happens only thing that I can do is to reset the board.

These are the lines from system log that correspond to system freeze.

Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.329787] ------------[ cut here ]------------
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.330150] WARNING: at /dvs/git/dirty/git-master_linux/kernel/drivers/gpu/nvgpu/gk20a/channel_gk20a.c:1636 gk20a_channel_finish+0x88/0x17c()
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.330273] Modules linked in: joydev dm_crypt dm_mod bnep rfcomm bluetooth rfkill nvhost_vi
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.331255] CPU: 0 PID: 3857 Comm: Watchdog Tainted: G        W    3.10.40-g8c4516e #1
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.331600] [<c00168e0>] (unwind_backtrace+0x0/0x140) from [<c0013234>] (show_stack+0x18/0x1c)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.332103] [<c0013234>] (show_stack+0x18/0x1c) from [<c0067150>] (warn_slowpath_common+0x54/0x70)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.332357] [<c0067150>] (warn_slowpath_common+0x54/0x70) from [<c0067190>] (warn_slowpath_null+0x24/0x2c)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.332599] [<c0067190>] (warn_slowpath_null+0x24/0x2c) from [<c03dcaec>] (gk20a_channel_finish+0x88/0x17c)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.333052] [<c03dcaec>] (gk20a_channel_finish+0x88/0x17c) from [<c03dcc44>] (gk20a_disable_channel+0x64/0x80)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.333283] [<c03dcc44>] (gk20a_disable_channel+0x64/0x80) from [<c03dcd28>] (gk20a_free_channel+0xc8/0x3c0)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.333515] [<c03dcd28>] (gk20a_free_channel+0xc8/0x3c0) from [<c03dd074>] (gk20a_channel_release+0x54/0x110)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.333972] [<c03dd074>] (gk20a_channel_release+0x54/0x110) from [<c01501b0>] (__fput+0x94/0x200)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.334221] [<c01501b0>] (__fput+0x94/0x200) from [<c008aac8>] (task_work_run+0xcc/0x108)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.334446] [<c008aac8>] (task_work_run+0xcc/0x108) from [<c006d1a4>] (do_exit+0x15c/0x3a4)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.334835] [<c006d1a4>] (do_exit+0x15c/0x3a4) from [<c006d5ec>] (do_group_exit+0x48/0xb8)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.335056] [<c006d5ec>] (do_group_exit+0x48/0xb8) from [<c007c144>] (get_signal_to_deliver+0x21c/0x434)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.335324] [<c007c144>] (get_signal_to_deliver+0x21c/0x434) from [<c0012644>] (do_signal+0x84/0x168)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.335584] [<c0012644>] (do_signal+0x84/0x168) from [<c0012a70>] (do_work_pending+0x6c/0xb4)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.351870] [<c0012a70>] (do_work_pending+0x6c/0xb4) from [<c000f1c0>] (work_pending+0xc/0x20)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.351898] ---[ end trace 83181803af49d533 ]---
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.351921] gk20a gk20a.0: timed out waiting for gk20a channel to finish
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.351930] ------------[ cut here ]------------
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.351951] WARNING: at /dvs/git/dirty/git-master_linux/kernel/drivers/gpu/nvgpu/gk20a/channel_gk20a.c:453 gk20a_disable_channel+0x78/0x80()
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.351960] Modules linked in: joydev dm_crypt dm_mod bnep rfcomm bluetooth rfkill nvhost_vi
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.352003] CPU: 0 PID: 3857 Comm: Watchdog Tainted: G        W    3.10.40-g8c4516e #1
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.352022] [<c00168e0>] (unwind_backtrace+0x0/0x140) from [<c0013234>] (show_stack+0x18/0x1c)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.352043] [<c0013234>] (show_stack+0x18/0x1c) from [<c0067150>] (warn_slowpath_common+0x54/0x70)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.352060] [<c0067150>] (warn_slowpath_common+0x54/0x70) from [<c0067190>] (warn_slowpath_null+0x24/0x2c)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.352073] [<c0067190>] (warn_slowpath_null+0x24/0x2c) from [<c03dcc58>] (gk20a_disable_channel+0x78/0x80)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.352089] [<c03dcc58>] (gk20a_disable_channel+0x78/0x80) from [<c03dcd28>] (gk20a_free_channel+0xc8/0x3c0)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.352101] [<c03dcd28>] (gk20a_free_channel+0xc8/0x3c0) from [<c03dd074>] (gk20a_channel_release+0x54/0x110)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.352119] [<c03dd074>] (gk20a_channel_release+0x54/0x110) from [<c01501b0>] (__fput+0x94/0x200)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.352138] [<c01501b0>] (__fput+0x94/0x200) from [<c008aac8>] (task_work_run+0xcc/0x108)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.352154] [<c008aac8>] (task_work_run+0xcc/0x108) from [<c006d1a4>] (do_exit+0x15c/0x3a4)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.352164] [<c006d1a4>] (do_exit+0x15c/0x3a4) from [<c006d5ec>] (do_group_exit+0x48/0xb8)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.352179] [<c006d5ec>] (do_group_exit+0x48/0xb8) from [<c007c144>] (get_signal_to_deliver+0x21c/0x434)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.352195] [<c007c144>] (get_signal_to_deliver+0x21c/0x434) from [<c0012644>] (do_signal+0x84/0x168)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.352208] [<c0012644>] (do_signal+0x84/0x168) from [<c0012a70>] (do_work_pending+0x6c/0xb4)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.352220] [<c0012a70>] (do_work_pending+0x6c/0xb4) from [<c000f1c0>] (work_pending+0xc/0x20)
Dec  9 20:23:55 tegra-ubuntu kernel: [ 1033.352227] ---[ end trace 83181803af49d534 ]---

Any help would be appreciated.

many thanks
Mirko

mipko · December 10, 2014, 10:24am

update: I have removed Logitech unified receiver and connected USB kbd and USB mouse to USB 2.0 HUB. Jetson was stable for more than 3 hours. I have left it overnight to download some stuff and in the morning it was frozen. It went to sleep or suspend or whatever and was unable to wake up. Fan was spinning and green led was on.

linuxdev · December 10, 2014, 8:28pm

I don’t know if this is the exact issue, but R21.2 was just released and there was some mention of USB issues (seemed to be USB3 only). You might want to look at this R21.2 release doc and see if this shows the issue:
http://developer.download.nvidia.com/mobile/tegra/l4t/r21.2.0/Tegra_Linux_Driver_Package_Release_Notes_R21.2.pdf

mipko · December 11, 2014, 9:43am

Linuxdev thanks for the input.

I have installed L4T 21.2 and only change I made is that I enabled USB 3.0 ports. USB auto disconnect seems disabled out of the box ?
System is stable for now, no lock ups, no deep sleep lock ups. External USB 3.0 HDD is much more responsive in terms of mounting, and file operations.
So, far so good. I will keep it up&running for next 24h and if lock up is gone I will mark this as solved.