Issues with Debian Stretch Image on BeagleBoard-X15

GoogleGroups
1

I have two issues working with the latest BeagleBoard Stretch image on the BeagleBoard X15.

( I used image under both https://rcn-ee.net/rootfs/bb.org/testing/2017-06-22/stretch-lxqt/ and https://rcn-ee.net/rootfs/bb.org/testing/2017-06-18/stretch-lxqt/ )

  1. The bootup from microSD card containing the image happens only after roughly 3-4 minutes, the board is unresponsive until then. I don’t know why this is happening but this did not happen with a Jessie image ( the board boots immediately )

  2. The syslog is getting spammed with log messages like this, and it fills up the entire space on the microSD card, to the point I can no longer do anything:

[ 4906.387892] alloc_contig_range: [fceb1, fceb2) PFNs busy
[ 4906.387951] alloc_contig_range: [fceb2, fceb3) PFNs busy
[ 4906.388082] alloc_contig_range: [fceb3, fceb4) PFNs busy
[ 4906.388152] alloc_contig_range: [fceb4, fceb5) PFNs busy
[ 4906.388251] alloc_contig_range: [fceb5, fceb6) PFNs busy
[ 4906.388307] alloc_contig_range: [fceb6, fceb7) PFNs busy
[ 4906.540999] alloc_contig_range: [fceb0, fceb1) PFNs busy
[ 4906.541076] alloc_contig_range: [fceb1, fceb2) PFNs busy
[ 4906.541135] alloc_contig_range: [fceb2, fceb3) PFNs busy
[ 4906.541195] alloc_contig_range: [fceb3, fceb4) PFNs busy
[ 4906.541253] alloc_contig_range: [fceb4, fceb5) PFNs busy
[ 4906.541311] alloc_contig_range: [fceb5, fceb6) PFNs busy
[ 4906.541369] alloc_contig_range: [fceb6, fceb7) PFNs busy
[ 4906.549392] alloc_contig_range: [fceb0, fceb1) PFNs busy
[ 4906.549539] alloc_contig_range: [fceb1, fceb2) PFNs busy
[ 4906.549599] alloc_contig_range: [fceb2, fceb3) PFNs busy
[ 4906.549654] alloc_contig_range: [fceb3, fceb4) PFNs busy
[ 4906.549708] alloc_contig_range: [fceb4, fceb5) PFNs busy
[ 4906.549762] alloc_contig_range: [fceb5, fceb6) PFNs busy
[ 4906.549816] alloc_contig_range: [fceb6, fceb7) PFNs busy

I wanted to bring these to your notice.

Thanks
Kumar Abhishek

2

Hi Kumar, what do you have plugged in? Which rev of the x15? I had tested this image on Friday, it was working for me on the new x15 (sr 2.0 silicon)

Regards,

3

I have Rev A2 of the X15, and the kernel logs say that it is a DRA752 ES1.1 silicon.

4
  1. Using a serial console, I get the following output:
5
  1. Update: Switching to a different SD card solves the boot time being long
  2. Doing ‘sudo service lightdm stop’ seems to stop the “alloc_contig_range” spam in the kernel log, suggesting it might have something to do with the X server (?)
6

I found some other people report the same issue ‘alloc_contig_range’ spamming the system logs:

http://lists-archives.com/linux-kernel/28714307-problem-persists-dmesg-spam-alloc_contig_range-xx-yy-pfns-busy.html

http://lists-archives.com/linux-kernel/28715065-drm-radeon-spamming-alloc_contig_range-xxx-yyy-pfns-busy-busy.html

http://forums.fedoraforum.org/showthread.php?t=314123

https://bugzilla.redhat.com/show_bug.cgi?id=1387732

One of the suggested fix is to suppress printing of this message (example as was done in a raspberry Pi tree)

7

Argh.

I was hoping to test the bbx15-debian-9.0-lxqt-armhf-2017-06-22 image on my 572xEVM, by burning a 4GB SD card, but for some reason, my board seems to boot the SPL 9 times (rinse and repeat fashion) and then appears to get stuck in the SPL. Looks like my SOC revision is different than your’s…

.
.
.

U-Boot SPL 2017.01-00341-gb2aad42503 (Jun 21 2017 - 10:56:05)
DRA752-GP ES2.0
Trying to boot from MMC1

** Unable to use mmc 0:1 for loading the env **
Using default environment
.
.
.
U-Boot SPL 2017.01-00341-gb2aad42503 (Jun 21 2017 - 10:56:05)
DRA752-GP ES2.0

8

Will TRY to muck around with this, but will likely have to wait a week…