Apalis iMX6 Linux freeze/hangs on kernel start

Hi,

We’ve just started development with the Apalis iMX6 quad on the evaluation board and the unit has started to randomly hang on kernel start (see below for debug at hang point)…

I’m using a slightly changed version of V2.5 and was experimenting with moving the console uart using “setenv console ttymxc1,115200n8” and I don’t know if this is coincidence but it seems to have started hanging from this point (maybe after I did a saveenv). I have now gone back to the straight downloaded V2.5 image but the problem still seems to happen.

Here are two instances (below) showing a similar point in the kernel startup where it freezes… worryingly the processor seems to start getting hotter than normal at this point so I haven’t left it for a long time to see if it eventually carries on…

(Freeze 1)

[    3.615899] TCP: cubic registered
[    3.616398] NET: Registered protocol family 10
[    3.617286] sit: IPv6 over IPv4 tunneling driver
[    3.617899] NET: Registered protocol family 17
[    3.617914] can: controller area network core (rev 20120528 abi 9)
[    3.617969] NET: Registered protocol family 29
[    3.617994] can: raw protocol (rev 20120528)

(Freeze 2)

[    2.722899] platform caam_sm: [0024] 0c 73 21 92 74 ac e4 89
[    2.723166] snvs-secvio 20cc000.caam-snvs: violation handlers armed - non-secure state
[    2.723435] hidraw: raw HID events driver (C) Jiri Kosina
[    2.723591] usbcore: registered new interface driver usbhid
[    2.723593] usbhid: USB HID core driver
[    2.729196] stmpe-adc stmpe-adc: Initialized
[    2.731071] sgtl5000 2-000a: sgtl5000 revision 0x11
[    2.890000] 2-000a: 1200 mV normal
[    2.890062] sgtl5000 2-000a: Using internal LDO instead of VDDD
[    3.442436] imx-sgtl5000 sound.31: sgtl5000 <-> 2028000.ssi mapping ok
[    3.479908] imx-spdif sound-spdif.33: snd-soc-dummy-dai <-> 2004000.spdif mapping ok
[    3.483725] imx-audio-hdmi sound-hdmi.32: hdmi-hifi <-> hdmi_audio.3 mapping ok
[    3.484583] TCP: cubic registered
[    3.485074] NET: Registered protocol family 10

I’ve attached a file showing the complete session from commanded reboot to hang…

I have another module so I can try the same image on this (it’s straight from the factory at the moment V2.4) at the moment this doesn;t seem to exhibit the problem and I don’t really want to make this go wrong as well…

I’m wondering if it is a hardware problem (because the cpu seems to get hot) or is it possible something can persists even through doing a complete new image update (via run setupdate and run update in the bootloader).

Best Regards,
Paul.

link text

Hi Paul,

Is DVI/HDMI connected? Does it still hang with DVI/HDMI disconnected?

Hi Brandon, Sorry didn’t spot you reply until I came to the same thoughts…
I’ve been trying to track this down by changing modules and standard images… and I’m coming to the conclusion it is image version related (probably the kernel version) and yes does seem to be related to the DVI… it’s quite random so you have to be careful not to jump to conclusions… pretty sure V2.4 doesn’t do it with DVI in or out… 2.5 does show problem but I think only with DVI connected… cheers

Same thing here … hangs quite reliably if the DVI is connected but never hangs if the DVI is not connected. Using iMX6 v1.1B on Apalis eval. board v1.1A.
More often than not it gets stack in the exact same place as in examples above:

[    3.348093] imx-sgtl5000 sound.31: sgtl5000 <-> 2028000.ssi mapping ok
[    3.384646] imx-spdif sound-spdif.33: snd-soc-dummy-dai <-> 2004000.spdif mapping ok
[    3.388156] imx-audio-hdmi sound-hdmi.32: hdmi-hifi <-> hdmi_audio.3 mapping ok
[    3.389015] TCP: cubic registered
[    3.389556] NET: Registered protocol family 10
[    3.390541] sit: IPv6 over IPv4 tunneling driver
[    3.391214] NET: Registered protocol family 17
[    3.391230] can: controller area network core (rev 20120528 abi 9)
[    3.391289] NET: Registered protocol family 29
[    3.391316] can: raw protocol (rev 20120528)

Please note that I am not using OpenEmbedded but a buildroot generated image.

# uname -a
Linux buildroot 3.14.28 #1 SMP Wed Feb 15 13:47:27 GMT 2017 armv7l GNU/Linux

Please note that we long since stopped supporting V2.5 beta BSPs which only ever where meant to be run on the sample hardware of that time and we recommend running one of our latest stable or alternatively the latest beta BSP. At the time of this writing this would be either BSP V2.6 or 2.7b2.