Dietpi: XU4 | Kernel 4.9 testing

[Fourdee]

https://github.com/Fourdee/DietPi/issues/899#issuecomment-299480599

https://github.com/Fourdee/DietPi/issues/926#issuecomment-310918133

[Fourdee]

• 🈴 Cloudshell Screen | no display:

root@DietPi:~# lsmod
Module                  Size  Used by
exynos_gpiomem          3162  0
ads7846                11687  0
spidev                  6446  0
fbtft                  24420  0
spi_s3c64xx            11643  0
fuse                   75590  1
ipv6                  341326  28
uas                    11815  0

Missing fbtft_device, although its in /etc/modules

🈺 Works by running:

modprobe fbtft_device

[    3.399340] fbtft_device: module is from the staging directory, the quality is unknown, you have been warned.
[    3.402001] fbtft_device: failed to register SPI device
[   61.038750] fbtft_device: module is from the staging directory, the quality is unknown, you have been warned.
[   61.042413] fbtft_device: GPIOS used by 'hktft9340':
[   61.042431] fbtft_device: 'reset' = GPIO21
[   61.042450] fbtft_device: 'dc' = GPIO22
[   61.042470] fbtft_device: 'led' = GPIO18

🈺 Works: A little "hacky", two entries for fbtft_device in /etc/modules to load module twice during boot + enable screen:

---

- Cloudshell IR

- DietPi-Cloudshell

• 🈺 Kodi = https://github.com/Fourdee/DietPi/issues/899#issuecomment-299452995 +

wget http://oph.mdrjr.net/meveric/kernel/XU3/4.9.x/xf86-video-armsoc-odroid_20170419-1.4.1-1+deb8_armhf.deb -O package.deb
dpkg -i package.deb

[nickfox-taterli]

I test but not working with waveshare32.

───────────────────────────────────────
DietPi | 16:44 | Sat 06/05/17
───────────────────────────────────────
V149 | oDroid XU3/4 (armv7l)
───────────────────────────────────────
IP Address | 192.168.99.141
───────────────────────────────────────

Created by : Daniel Knight
Web : http://DietPi.com
Twitter : http://twitter.com/dietpi_
Donate : http://goo.gl/pzISt9

Device image possible thanks to: Meveric
DietPi's web hosting is powered by: MyVirtualServer.com

dietpi-launcher = All the DietPi programs in one place.
dietpi-config = Feature rich configuration tool for your device.
dietpi-software = Select optimized software for installation.
htop = Resource monitor.
cpu = Shows CPU information and stats.

dmesg | grep fbtft
[ 3.401338] fbtft: module is from the staging directory, the quality is unkno wn, you have been warned.
[ 3.410109] fbtft_device: module is from the staging directory, the quality i s unknown, you have been warned.
[ 3.414801] fbtft_device: display not supported: 'odroidc_tft32'

lsmod
Module Size Used by
ads7846 11687 0
spidev 6446 0
spi_s3c64xx 11643 0
exynos_gpiomem 3162 0
fbtft 24420 0
fuse 75590 1

modprobe fbtft_device
modprobe: ERROR: could not insert 'fbtft_device': Invalid argument

[ 3.401338] fbtft: module is from the staging directory, the quality is unknown, you have been warned.
[ 3.410109] fbtft_device: module is from the staging directory, the quality is unknown, you have been warned.
[ 3.414801] fbtft_device: display not supported: 'odroidc_tft32'
[ 8571.621657] fbtft_device: module is from the staging directory, the quality is unknown, you have been warned.
[ 8571.622475] fbtft_device: display not supported: 'odroidc_tft32'

or cloudshell screen

[ 3.509389] fbtft: module is from the staging directory, the quality is unknown, you have been warned.
[ 3.514321] fbtft_device: module is from the staging directory, the quality is unknown, you have been warned.
[ 3.514937] spi spi1.0: spidev spi1.0 1000kHz 8 bits mode=0x00
[ 3.514973] spi spi1.1: ads7846 spi1.1 1000kHz 8 bits mode=0x00
[ 3.515299] spi spi1.0: Deleting spi1.0

and exec modprobe fbtft_device

but noting output.

[Fourdee]

@nickfox-taterli

Thanks.

Did you enable Cloudshell/waveshare32 through dietpi-config > Display options > LCD panel?

[Fourdee]

Test possible fix for vsync tearing in Kodi video playback (thanks to Meveric):
advancedsettings.xml

<advancedsettings>
    <video>
        <adjustrefreshrate>
            <override>
                <fpsmin>23</fpsmin>
                <fpsmax>60</fpsmax>
                <refresh>60</refresh>
            </override>
        </adjustrefreshrate>
    </video>
</advancedsettings>

[justin0a0]

Tried testing with the image you posted, on first login I get:
[Failed] Insufficient free space on RootFS. 30 MB available, 500 MB required
[Failed] Aborting

I assume this is from you checking for free space before the updates but seeing as the partition was not yet resized I got this error.

[Fourdee]

@justin0a0

I assume this is from you checking for free space before the updates but seeing as the partition was not yet resized I got this error.

Thanks for report 👍
Might be an issue with mmcblk index's, was this on SD or EMMC? And any chance you can run/paste results:

[nickfox-taterli]

@Fourdee only cloudshell (hktft9340) install correct,i guess it miss " .controller_data = &hktft9340_controller_data, " at fbtft_device.c.

[justin0a0]

@Fourdee
It was while using an SD I have been trying to find time to put the SD back in but the house gets a bit upset when they can't access the media I'll give it another run this weekend if you still need the info

[joaofl]

@Fourdee
I've burned the testing image, which booted fine. But after running dietpi's tool to transfer the rootfs to the usb, the boot.ini file ended getting an the wrong UUID, which I had to set manually in order to boot. The one on fstab was correct.

[Fourdee]

@joaofl

Thanks for report 👍 I'll take a look.

boot.ini file ended getting an the wrong UUID, which I had to set manually in order to boot. The one on fstab was correct.

Strange, i've checked the code (v150), same UUID var is applied to fstab and boot.ini. I'll try to replicate on XU4.

• 🈯️ Unable to replicate on XU4 4.9 with DietPi v150. /boot/boot.ini and /etc/fstab UUID's match:

root@DietPi:~# cat /etc/fstab | grep UUID=dd568952-3621-42b7-8933-12085def278e
UUID=dd568952-3621-42b7-8933-12085def278e / auto defaults,noatime,discard 0 1

root@DietPi:~# cat /DietPi/boot.ini | grep root=
setenv bootrootfs "console=tty1 root=UUID=dd568952-3621-42b7-8933-12085def278e rootwait ro fsck.repair=yes"

• 🈴 Found a bug when changing format settings, fails to continue rootFS transfer afterwards.

[joaofl]

@Fourdee

Strange, i've checked the code (v150), same UUID var is applied to fstab and boot.ini. I'll try to replicate on XU4.

Maybe I messed it up, or maybe because I had 2 USB drives plugged in? I'll try to replicate it and I let you know if anything happens.

[joaofl]

@Fourdee
I actually reproduced the issue I was having, but it ended up not being due to the UUID I think. However I've observed a few things.

(1) After transferring the root partition to an external drive and rebooting, dietpi starts to auto-mount (line added to fstab) the old root partition of the sd card as an external drive. Something we actually don't want, right?

(2) When we transfer the root, you update the UUID in a couple of files, right? My idea was: Can't you update this files only on the new transfered rootfs, on the USB drive only (and not before rsync)? Because this way, if we want to go back to the previous scheme, we only need to modify the boot.ini file on the boot partition (maybe by uncommenting a line). This would allow for easy rollback in case the transfer fails.

(3) The problem I was having was actually due the spin-up delay of one of my external HDs on a cold boot. It means that if I simply reboot the system, it works fine. But the delay of spinning-up the disk before mouting it on a cold boot, ends up blocking the booting, even after successfully mounting the HD. Something like a deadlock. Have an idea on how to solve this?

Everything else seems fine so far with the 4.9.24 kernel. (4.9.28 is available by meveric, should I try to update?)

[joaofl]

@Fourdee
I actually just checked your code @ https://github.com/Fourdee/DietPi/blob/master/dietpi/dietpi-drive_manager and it actually already implements (2) and something regarding (1) as well. So I'm not sure... By the way, I'm running dietpi v150.

[Fourdee]

@joaofl

and it actually already implements (2)

Yep:
if Rsync fails with rootfs transfer, the original /etc/fstab is restored, then script exited to prevent further changes to boot.ini etc.:
https://github.com/Fourdee/DietPi/blob/testing/dietpi/dietpi-drive_manager#L497-L507

The problem I was having was actually due the spin-up delay of one of my external HDs on a cold boot. It means that if I simply reboot the system, it works fine.

On the rpi, adding rootdelay=10 to cmdline.txt helps resolve this. Try doing rootdelay=10 to boot.ini, in line setenv bootrootfs ""

After transferring the root partition to an external drive and rebooting, dietpi starts to auto-mount (line added to fstab) the old root partition of the sd card as an external drive. Something we actually don't want, right?

The existing entry should be removed.
https://github.com/Fourdee/DietPi/blob/testing/dietpi/dietpi-drive_manager#L481-L485

Any chance you can paste results of following after this occurs?

cat /etc/fstab
cat /DietPi/boot.ini

[joaofl]

@Fourdee

Try doing rootdelay=10 to boot.ini, in line setenv bootrootfs ""

It worked. Thanks!

the original /etc/fstab is restored, then script exited to prevent further changes to boot.ini

Thats true. But it did not work for me, because I had to interrupt the rsync with ctrl+c. This because it blocked during the copy, most probably because of some badblocks on the sd card. So maybe, if this part of the code comes after the rsync.... But anyway, thats not a big deal.

4.9.28 is available by meveric, should the update happen seamlessly?

Any chance you can paste results of following after this occurs?

blkid

/dev/mmcblk1: PTUUID="000f1766" PTTYPE="dos"
/dev/mmcblk1p1: SEC_TYPE="msdos" LABEL="boot" UUID="5886-8548" TYPE="vfat" PARTUUID="000f1766-01"
/dev/mmcblk1p2: LABEL="rootfs" UUID="a4f12232-1670-43f4-8eca-49bfbe4d01ee" TYPE="ext4" PARTUUID="000f1766-02"
/dev/sda1: LABEL="DATA" UUID="23504cb3-a823-422f-a1fa-61227456189c" TYPE="ext4" PARTLABEL="My Book" PARTUUID="69cd0e21-955d-4c89-925f-495a5395beb3"
/dev/sdb1: UUID="540cd7e8-4d07-4d36-9eb5-1fa8a31f534e" TYPE="ext4" PARTUUID="90991f47-01"

fstab

#Internal Drives---------------------------------------------------
proc            /proc           proc    defaults                0 0
UUID=5886-8548  /boot           auto    defaults,noatime,discard        0 2
UUID=540cd7e8-4d07-4d36-9eb5-1fa8a31f534e / auto defaults,noatime,discard 0 1
tmpfs           /tmp            tmpfs   defaults,noatime,nodev,nosuid,mode=1777             0 0
tmpfs           /var/log        tmpfs   defaults,size=20m,noatime,nodev,nosuid,mode=1777    0 0
tmpfs           /DietPi         tmpfs   defaults,size=10m,noatime,nodev,nosuid,mode=1777    0 0

#External Drives---------------------------------------------------
#NB: Please use dietpi-drive_manager to setup and control your external drives.

#Samba Client------------------------------------------------------
#/mnt/samba . Please use dietpi-config and the Networking Options: NAS menu to setup this mount

#FTP Client Mount--------------------------------------------------
#/mnt/ftp_client . Please use dietpi-config and the Networking Options: NAS menu to setup this mount

#NFS Client Mount--------------------------------------------------
#/mnt/nfs_client . Please use dietpi-config and the Networking Options: NAS menu to setup this mount
UUID=23504cb3-a823-422f-a1fa-61227456189c       /mnt/hd      auto     defaults,noatime,nofail,x-systemd.automount  0 0
#UUID=a4f12232-1670-43f4-8eca-49bfbe4d01ee       /mnt/a4f12232-1670-43f4-8eca-49bfbe4d01ee      auto     defaults,noatime,nofail,x-systemd.automount  0 0

boot.ini

#------------------------------------------------------------------------------------------------------
# Basic Ubuntu Setup. Don't touch unless you know what you are doing.
# --------------------------------
setenv bootrootfs "console=tty1 root=UUID=540cd7e8-4d07-4d36-9eb5-1fa8a31f534e rootwait ro fsck.repair=yes rootdelay=20"

# boot commands
setenv bootcmd "fatload mmc 0:1 0x40008000 zImage; fatload mmc 0:1 0x42000000 uInitrd; fatload mmc 0:1 0x44000000 exynos5422-odroidxu4.dtb; bootz 0x40008000 0x42000000 0x44000000"

[Fourdee]

@joaofl

Thats true. But it did not work for me, because I had to interrupt the rsync with ctrl+c. This because it blocked during the copy, most probably because of some badblocks on the sd card. So maybe, if this part of the code comes after the rsync.... But anyway, thats not a big deal.

Ah yes, I see, not much we can do about ctrl+c.

4.9.28 is available by meveric, should the update happen seamlessly?

For now, you'll need to install/update these manually. Meveric is due to update his repo, replacing the current linux packages to 4.9. Once done, apt-get upgrade would do it.

[Fourdee]

Meveric has provided apt packages for Kodi 17, needs testing:
https://forum.odroid.com/viewtopic.php?p=191780#p191780

[Fourdee]

Meveric is updating the 3.10 kernel to support EMMC 5.1. To avoid this kernel overwriting the 4.9 kernel, users will need to put kernel packages on hold:

apt-mark hold linux-image-armhf-odroid-xu3

[Fourdee]

• Need to check IR CS functionality with 4.9

After upgrading to 4.9 kernel the IR-remote didn't work anymore. I had to rerun the ./install_ir.sh script to fix the problem.
http://dietpi.com/phpbb/viewtopic.php?f=11&t=1937&p=7466#p7407

[joaofl]

@Fourdee
I've been running the 4.9.29 version for some days already, and had no major problems. It is quiet stable, and the performance upgrade on the ethernet is visible.

But I keep getting the following warning:
usb 4-1.1: reset SuperSpeed USB device number 3 using xhci-hcd

It happens specially when, from a remote machine, I upload and download files (simultaneously) from/to the hard-drive mounted on my XU4 USB3 on the network.

The problem does not seems to be related to the USB3 itself, since when working locally, I can read and write files from/to the hard drive with no warnings at all.

So, it seems to be related somehow to Ethernet and USB operations happening simultaneously.

I've searched around for similar issues. A guy had similar problem, that got solved but he doesn't know how.

Do you have any clue what it can be or where to look for answers?
Thanks.

[Fourdee]

@joaofl

I've been running the 4.9.29 version for some days already, and had no major problems. It is quiet stable, and the performance upgrade on the ethernet is visible.

Yep same here, running 24/7 NAS + PiHole. Runs great.

But I keep getting the following warning:
usb 4-1.1: reset SuperSpeed USB device number 3 using xhci-hcd

Could be a compatibility issue with USB host/adapter/kernel.

It happens specially when, from a remote machine, I upload and download files (simultaneously) from/to the hard-drive mounted on my XU4 USB3 on the network.

Might be power related if USB drive is host powered (lack of under load), which PSU are you running ?

[joaofl]

@Fourdee

Might be power related if USB drive is host powered (lack of under load), which PSU are you running ?

The hard drive is externally powered, and I'm not actually draining much current from the XU4 at all, and the power supply I use can supply +10A, so It should not be an issue. And it never happened with the kernel version 3.

As you said, It may be a compatibility problem? Not sure how to debug that....

This is the HD I'm currently using:
https://www.wdc.com/products/wd-recertified/my-book.html

[Fourdee]

@joaofl

RECERTIFIED

Could also be an outstanding issue with the unit, that didn't get fixed (eg: USB > SATA controller).

[joaofl]

@Fourdee
The closest I found was this thread:
http://linux-scsi.vger.kernel.narkive.com/B3jfin3j/xhci-issues-wd-mybook-1230-reset-superspeed-usb-device
Which seems to be kind of a deeper issue, on the device driver itself... Not sure what to do. There is no firmware upgrades for the HD so far.

[filemoon]

Kernel 4.9 (29+)

Only performance governor is present.
Do you have a plan to add other governors (I'm looking at on-demnd and conservative)

Regards Chris

[Fourdee]

@filemoon

Hi Chris, please see this post for more info:
http://dietpi.com/phpbb/viewtopic.php?f=9&t=1927&p=7388&hilit=ondemand#p7388

[Fourdee]

Update from Meveric, testing and feedback required:

today I've uploaded a new version of the Kernel for 4.9 with updated meta packages.
They will remove all currently installed 3.10 Kernels (but keep the headers, since they don't really matter).
Packages are as mentioned before: linux-image-4.9-armhf-odroid-xu3 and linux-headers-4.9-armhf-odroid-xu3
It would be nice if you or some of your community members could try it and give a little feedback how it works.
If you do an update from 3.10.XXX to 4.9 it will complain that it may be insecure to remove the currently loaded Kernel.
It will ask if it should STOP the removal.
So the correct answer here is NO.
Looking forward to some feedback.

---

Upgrade procedure for 4.9 Kernel:

• Please note, this kernel is still in devolpment and may be unstable. We recommend you backup your system, before hand: dietpi-backup
• Removal of temp XU4 DTB (package now contains this, must be removed prior to update):

rm /boot/exynos5422-odroidxu4.dtb

• Commands to update:

apt-get update
apt-get install linux-image-4.9-armhf-odroid-xu3

• 🈯️ Needs testing on 3.x image (as upgrade)
• 🈯️ Needs testing on 4.x image

[filemoon]

Freshly noticed issue:

HMP not active for ARM big.LITTLE (only big cores are in use):

• only 4 (big) cores available (another 4 small cores not in use)
• high temperature
• slower performance (vs 8 cores)
• high power consumption

I will restore the backup and wait for another kernel 4.9 build with HMP activated and more governors.

[studioai]

http://odroid.com/dokuwiki/doku.php?id=en:xu4_upgrade_from_3.10_to_4.9

Hardkernel has updated their wiki page and repository.

[Fourdee]

@studioai

http://odroid.com/dokuwiki/doku.php?id=en:xu4_upgrade_from_3.10_to_4.9
Hardkernel has updated their wiki page and repository.

This does not apply to our image as we are using Meveric's repo, which contains 4.9 kernel packages. The documentation may work, but I would avoid using that, instead use the Meveric repo packages as per below:
https://github.com/Fourdee/DietPi/issues/926#issuecomment-310918133

[rncwnd79]

Needs testing on 3.x image (as upgrade)

I tried Meverics exact steps @Fourdee and get this (on 3.x image, already running manually installed 4.9):
Fehl http://fuzon.co.uk/meveric/ all/xu3 linux-image-4.9-armhf-odroid-xu3 armhf 4.9.34-1 404 Not Found
E: Fehlschlag beim Holen von http://fuzon.co.uk/meveric/pool/xu3/l/linux-image-4.9-armhf-odroid-xu3/linux-image-4.9-armhf-odroid-xu3_4.9.34-1_armhf.deb 404 Not Found

E: Einige Archive konnten nicht heruntergeladen werden; vielleicht »apt-get update« ausführen oder mit »--fix-missing« probieren?

[joaofl]

@Fourdee
Just to understand, whats the difference between meverics compile and others?

I keep track of his releases, which was last updated end of may (v4.9.29), while the odroid repo is already at version 45?
(https://oph.mdrjr.net/meveric/kernel/XU3/4.9.x/)
(https://github.com/hardkernel/linux/tree/odroidxu4-4.9.y)

[Fourdee]

@rncwnd79

E: Fehlschlag beim Holen von http://fuzon.co.uk/meveric/pool/xu3/l/linux-image-4.9-armhf-odroid-xu3/linux-image-4.9-armhf-odroid-xu3_4.9.34-1_armhf.deb 404 Not Found

Confirmed earlier today.

Its functional now, please re-run:

apt-get update
apt-get install linux-image-4.9-armhf-odroid-xu3

Most likely Meveric was syncing our mirror server and its only just completed (fuzon.co.uk)

root@DietPi:~# apt-get install linux-image-4.9-armhf-odroid-xu3
Reading package lists... Done
Building dependency tree
Reading state information... Done
The following packages were automatically installed and are no longer required:
  libevent-2.0-5 libhunspell-1.3-0 linux-image-4.9.33+
Use 'apt-get autoremove' to remove them.
The following extra packages will be installed:
  linux-image-4.9.35+
Suggested packages:
  fdutils linux-doc-4.9.35+ linux-source-4.9.35+ linux-image-4.9.35+-dbg
  linux-manual-4.9.35+
Recommended packages:
  kernel-common
The following NEW packages will be installed:
  linux-image-4.9.35+
The following packages will be upgraded:
  linux-image-4.9-armhf-odroid-xu3
1 upgraded, 1 newly installed, 0 to remove and 22 not upgraded.
Need to get 14.2 MB of archives.
After this operation, 44.6 MB of additional disk space will be used.
Do you want to continue? [Y/n]
Get:1 http://fuzon.co.uk/meveric/ all/xu3 linux-image-4.9.35+ armhf 4.9.35-20170704-XU3 [14.2 MB]
Get:2 http://fuzon.co.uk/meveric/ all/xu3 linux-image-4.9-armhf-odroid-xu3 armhf 4.9.35-1 [1,378 B]
Fetched 14.2 MB in 3s (4,472 kB/s)
Preconfiguring packages ...
Selecting previously unselected package linux-image-4.9.35+.
(Reading database ... 36834 files and directories currently installed.)
Preparing to unpack .../linux-image-4.9.35+_4.9.35-20170704-XU3_armhf.deb ...
Examining /etc/kernel/preinst.d/
run-parts: executing /etc/kernel/preinst.d/remove-old-on-upgrade 4.9.35+ /boot/vmlinuz-4.9.35+
Done.
Unpacking linux-image-4.9.35+ (4.9.35-20170704-XU3) ...
Replacing files in old package linux-image-4.9.33+ (4.9.33-20170620-XU3) ...
Preparing to unpack .../linux-image-4.9-armhf-odroid-xu3_4.9.35-1_armhf.deb ...
Unpacking linux-image-4.9-armhf-odroid-xu3 (4.9.35-1) over (4.9.33-1) ...
Setting up linux-image-4.9.35+ (4.9.35-20170704-XU3) ...

 Hmm. There is a symbolic link /lib/modules/4.9.35+/build
 However, I can not read it: No such file or directory
 Therefore, I am deleting /lib/modules/4.9.35+/build


 Hmm. The package shipped with a symbolic link /lib/modules/4.9.35+/source
 However, I can not read the target: No such file or directory
 Therefore, I am deleting /lib/modules/4.9.35+/source

Running depmod.
Examining /etc/kernel/postinst.d.
run-parts: executing /etc/kernel/postinst.d/apt-auto-removal 4.9.35+ /boot/vmlinuz-4.9.35+
run-parts: executing /etc/kernel/postinst.d/copy_dtb 4.9.35+ /boot/vmlinuz-4.9.35+
run-parts: executing /etc/kernel/postinst.d/initramfs-tools 4.9.35+ /boot/vmlinuz-4.9.35+
update-initramfs: Generating /boot/initrd.img-4.9.35+
Image Name:   uInitrd
Created:      Tue Jul  4 17:03:05 2017
Image Type:   ARM Linux RAMDisk Image (uncompressed)
Data Size:    4186917 Bytes = 4088.79 kB = 3.99 MB
Load Address: 00000000
Entry Point:  00000000
run-parts: executing /etc/kernel/postinst.d/uInitrd 4.9.35+ /boot/vmlinuz-4.9.35+
run-parts: executing /etc/kernel/postinst.d/zImage 4.9.35+ /boot/vmlinuz-4.9.35+
Setting up linux-image-4.9-armhf-odroid-xu3 (4.9.35-1) ...

[Fourdee]

@joaofl

I keep track of his releases, which was last updated end of may (v4.9.29), while the odroid repo is already at version 45?

Could be outdated, or the x.x.XX code reflects the build count in Meverics builds. So each time Meveric updates his kernel with latest, his build count number will be unique to how many times hes done it.

I could be wrong, depends on his setup and numbering system, or, he has hes own custom fork of the kernel, but you'd need to ask Meveric on Odroid forums to confirm.

Just to understand, whats the difference between meverics compile and others?

As far as i'am aware, they are identical, minus any tweaks by Meveric (eg: module options), but again, you'd need to confirm with Meveric.

Its worth noting, Meveric's repo and his continued maintenance of it, allows for everyone to easily update their kernel via APT, without compiling each time its updated.

[studioai]

@Fourdee

Its functional now, please re-run:

apt-get update
apt-get install linux-image-4.9-armhf-odroid-xu3

I tried those command on 3.10 kernel image and I got

root@DietPi:~# apt-get install linux-image-4.9-armhf-odroid-xu3
Reading package lists... Done
Building dependency tree       
Reading state information... Done
E: Unable to locate package linux-image-4.9-armhf-odroid-xu3
E: Couldn't find any package by regex 'linux-image-4.9-armhf-odroid-xu3'

Do I need to edit source.list?

Here's contents of my /etc/apt/source.list.d/meveric-jessie.list

deb http://fuzon.co.uk/meveric jessie main backports

[Fourdee]

@studioai

Strange, do you have all the lists (mainly the XU3 one)?

root@DietPi:~# ls -lha /etc/apt/sources.list.d/
total 24K
drwxr-xr-x 2 root root 4.0K Jun 20 11:44 .
drwxr-xr-x 6 root root 4.0K Jun 20 11:44 ..
-rw-r--r-- 1 root root   48 May 19 13:04 meveric-all.list
-rw-r--r-- 1 root root   39 May 19 13:04 meveric-all-XU3.list
-rw-r--r-- 1 root root   53 May 19 13:04 meveric-jessie.list

root@DietPi:~# apt-cache showpkg linux-image-4.9-armhf-odroid-xu3
Package: linux-image-4.9-armhf-odroid-xu3
Versions:
4.9.35-1 (/var/lib/apt/lists/fuzon.co.uk_meveric_dists_all_xu3_binary-armhf_Packages) (/var/lib/dpkg/status)
 Description Language:
                 File: /var/lib/apt/lists/fuzon.co.uk_meveric_dists_all_xu3_binary-armhf_Packages
                  MD5: 9c9c3754dda52eaadfc61ec69d5a7d6d

And apt-get update ran before hand?

[studioai]

@Fourdee

Strange, do you have all the lists (mainly the XU3 one)?

Nope. There's only meveric-jessie.list.

root@DietPi:~# ls -lha /etc/apt/sources.list.d/
total 5.0K
drwxr-xr-x 2 root root 1.0K Mar  5 23:15 .
drwxr-xr-x 6 root root 1.0K Mar  5 23:15 ..
-rw-r--r-- 1 root root   55 Mar  5 23:15 deb-multimedia.list
-rw-r--r-- 1 root root   53 Mar  5 23:15 meveric-jessie.list
-rw-r--r-- 1 root root   40 Mar  5 23:15 plex.list

I didn't touched lists at all since I installed dietpi last year.

And apt-get update ran before hand?

Yes. I did apt-get update for sure.

Anyway I'll try again after downloading those lists.

[studioai]

@Fourdee

Adding meveric-all-XU3.list to /etc/apt/sources.list.d/ solved my problem. Thank you!

[rncwnd79]

Trying right now.
edit: It worked this time.

Btw.
--fix-missing
worked after cleaning /boot.

[studioai]

OnDemand Governor seems to does not work properly.

CPU4~8 Frequency fixed on 2000MHz and never go down.

[Fourdee]

@studioai

CPU4~8 Frequency fixed on 2000MHz and never go down.

Our CPU script will tend to cause the ondemand throttle up to trigger. There is nothing we can do about this inside bash.

To get the current frequency, without adding too much processing which may cause a throttle up:

cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_cur_freq

Results in:

root@DietPi:~# cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_cur_freq
200000
200000
200000
200000
700000
700000
700000
700000

---

You can also check the times spent in each cpu clock with:

cat /sys/devices/system/cpu/cpu*/cpufreq/stats/time_in_state

[studioai]

@Fourdee
cat /sys/devices/system/cpu/cpu*/cpufreq/stats/time_in_state

That gives me reasonable results. look good.
Thanks!

[filemoon]

[studioai]

[Fourdee]

[Fourdee]

[studioai]

[filemoon]

[rncwnd79]

[1989gironimo]

[joaofl]

[secpri]

[1989gironimo]

[Fourdee]

[joaofl]

[1989gironimo]

[Fourdee]