Ran fine this time... root@m300-2:/boot# ls -lh linux-headers-5.5.1-kirkwood-tld-1_1.0_armel.deb -rw-r--r-- 1 root root 11M Feb 2 01:03 linux-headers-5.5.1-kirkwood-tld-1_1.0_armel.deb root@m300-2:/boot# dpkg -i linux-headers-5.5.1-kirkwood-tld-1_1.0_armel.deb (Reading database ... 15907 files and directories currently installed.) Preparing to unpack linux-headers-5.5.1-kirkwood-tld-1_1.0by mikeh49 - Debian
To be clear, I should install the headers file with dkpg also? If so, there must be a problem as I tried that and it did not complete, complaining of a problem with the tar. The md5 of the full tarball checked good. I see if I have the complete log and update. However, the M300 seems to be running fine.by mikeh49 - Debian
I'm confused about the kirkwood kernel update process. I updated a Pogo E02 OK, but am running into problems with a Dell M300. The kernel tarball contains 2 .deb files, but only the image file is installed by the instructions. Where is the headers file used? At the mkimage step, the instructions state: "Generate the uImage and uInitrd (the kernel files vmlinuz-5.5.1-kirkwood-tld-1by mikeh49 - Debian
I thought I followed your directions in the kernel thread. Can I rerun the dpkg of the .deb to get the modules? Or extract them from the .deb and put them in the right places? The deb file is in the /boot folder. Is this OK, or should it be in /? Looks like too many files to do manually -> dpkg -c > many, many files!by mikeh49 - Debian
Tried to update the Debian 10 (Buster) file system on my Dell M300. Did apt-get update followed by apt-get upgrade and ran into this: Processing triggers for initramfs-tools (0.133+deb10u1) ... update-initramfs: Generating /boot/initrd.img-5.5.1-kirkwood-tld-1 W: missing /lib/modules/5.5.1-kirkwood-tld-1 W: Ensure all necessary drivers are built into the linux image! depmod: ERROR: couldby mikeh49 - Debian
A user on the LMS forum has compiled Perl 5.28 for LMS and posted it here: https://github.com/mw9/slimserver/commit/46c22a3c7e69187a2ead9392c53f1c90790a03fa I could not get the buildme script to work, this kind soul stepped up on the forum and saved me. LMS has been running great on the M300 for about 5 months with a internal 120GB drive plugged into the SATA port. I have been happy wiby mikeh49 - Debian
Good suggestions, thanks. I'll do #1. I forgot to include the u boot environments. Here they are: ethact=egiga0 bootdelay=3 baudrate=115200 mainlineLinux=yes console=ttyS0,115200 led_init=green blinking led_exit=green off led_error=orange blinking mtdparts=mtdparts=orion_nand:1M(u-boot),4M(uImage),32M(rootfs),-(data) mtdids=nand0=orion_nand partition=nand0,2 stdin=serialby mikeh49 - Debian
I have a Pogo E02 that has been out of service for a few years and want to update it and configure it as a NAS. It has an old u boot: U-Boot 2011.12 (Feb 20 2012 - 21:21:59) BTW, is this the orignal Pogo u boot. or perhaps Jeff's first Kirkwood u boot? and is currently running Bodhi's 4.4.0 file system. I'm thinking I should update the u boot before updating to the latestby mikeh49 - Debian
OK, thanks. I was thinking about the chkdsk, but forgot to try it before posting. I'll report back.by mikeh49 - Debian
I'm trying to get my second M300 going. I have installed the Debian filesystem (5.9.2 with M300 DTB) on a 4GB partition of the 16GB SSD card that comes in the M300. The remaining space on the card is formatted EXT 3 to be used for data. I had this card running on my frist M300 back in September with no significant problems that I can recall. After installing it in the new M300, making thby mikeh49 - Debian
I got more curious about this and googled some and found that Stretch does not require the persistent-rules file (but will use it if present). The SUBSYSTEM line in Stretch on the Kace M300 is commented and there is no regenerator script in /lib/udev. It sounds like this may now be plug-and-play. I plan some more research and testing when I get the second M300 going. Short of time at the momeby mikeh49 - Debian
QuoteWhen you use an existing rootfs to boot another identical box (or another Kirkwood box in this case), you will need to make sure the udev rules for persistent-net are removed or modified. This usually causes network problem because of conflicting MAC address. To make it really simple, remove these 2 files (or rename them to *.save so that they will be regenerated by udev using the new MAC adby mikeh49 - Debian
I'd like to use a single USB drive on more than one box, i.e., plug and play, so to speak. This would be, for example, several Kirkwood boxes with the same DTS file such as Pogo E02 or Dell Kace M300. I can move a USB to another box by deleting the appropriate udev rules files. In another thread, Bodhi mentioned a way to make this easier. So, I'm looking for some guidance or instrucby mikeh49 - Debian
The M300 went into production today with the 120GB SSD and the cover installed. It's had a week or two of burn-in and several successful boots. No haveged, I just wait. The SSD is secured with zip ties through a couple of fittings installed in the tapped holes in the enclosure on each side of the SSD. Working great. Thanks Bodhi!by mikeh49 - Debian
I've done a couple of apt-get updates with no problem. Would it be worth starting over with a fresh download of the filesystem and kernel, and rebuilding with the latest DTB? Mine has been solid for 24 hours with12 hours with the music server serving.by mikeh49 - Debian
OK, thanks. I searched for echo, would have been smarter to search for printenv. Never mind.by mikeh49 - Debian
@renojim - I'm wondering why your M300 prints the environment variables one or more times in each boot. Especially in the middle of setting up the uImage and uInitrd. Mine doesn't do that. I looked for an environment that would echo them, but didn't see it. Did you put something like that in? My M300 has stalled several times while doing the scripts as well. Another sby mikeh49 - Debian
bootcmd runs bootcmd_usb before bootcmd_ide so does that mean it checks USB and if it fails, moves on to ide? The Boot USB. echo comes from bootcmd_usb which runs before bootcmd_ide. If removing haveged fixes it, I can live with slower boot. I usually reboot this server 3 times a year.by mikeh49 - Debian
Three boots without haveged (dpkg -r haveged) and all successful. About 1 minute to start the SSH server. I think I first tried haveged on the 16GB stock drive and I don't think there was ever a boot fail there (at least from this reason). The 16GB is noticeably slower to boot, especially loading the images, than the new 120GB drive; no surprise there. Per bohdi's request: Boby mikeh49 - Debian
All tests with SSD connected One test with debian USB attached also, boot OK from Debian as it should, i.e., prefer USB Three failures to boot with only SSD connected: All 3 choked right after the Loading Essential Drivers step when it starts running various scripts. 2 at running /scripts/local-block One after /scripts/init-bottom Does that help?by mikeh49 - Debian
Good point, I gave it a minute at the most, probably more like 30 secs. I'm doing a burn-in now, I'll give it a try when there's a good opportunity.by mikeh49 - Debian
Further testing now with the 120GB SSD installed, with label sata_roots and sata_rootfs in fstab. First boot failed, stuck in loop around running /scripts/local-block, reference to a .so file. Was not able to capture. Second boot stuck in loop below: Begin: Running /scripts/local-block ... done. /init: line 329: blkid: not found /init: line 96: sleep: not found /init: line 97: caby mikeh49 - Debian
Interesting. My experience has been very good with booting from SATA, both with the stock flash card (formatted, partitioned and set up with bohdi's file system and tld-2 kernel) and with a 120GB SSD. There have been 3 instances (I think) where it got into a loop trying to do something, but, of course, I didn't keep the info. I think all have been after installing the Logitech Mediaby mikeh49 - Debian
Thanks, good info. I get it. Two things from this Kace experience: serial is pretty useful/important; a modern u-boot is a thing of beauty, especially environments in a file. Actually 3 things: I've learned a lot.by mikeh49 - Debian
Thanks for the info, now I get it. The problem will be keeping it. I've concluded that I need to document all of this stuff in my own way. It's available, but in various threads here and pretty spotty on the web. Thanks for all you do, bohdi. I think I'm good for now.by mikeh49 - Debian
yep... root@Dell-M300:~# cat /etc/fw_env.config # MTD device name Device offset Env. size Flash sector size Number of sectors #/dev/mtd0 0xc0000 0x20000 0x20000 /dev/mtd1 0x0 00001000 0x00001000 Updated above.by mikeh49 - Debian
I got a successful SSD boot correcting several environment variable at the u-boot Marvell> prompt. I had several errors. But, I'm having trouble getting the settings to "stick". I have not used saveenv in the u-boot shell. Bohdi seems to prefer waiting for the system to boot and then using fw_setenv in debian. I finally remembered that settings made in the u-boot shell do nby mikeh49 - Debian
Thanks @renojim, good eye. There was another problem on that line as well. Copy-paste difficulties and looked at these lines too many times yesterday. Fixed them and now back to the original problem of booting from a device with LABEL=sata_rootfs. Below is the serial log where it has the problem. It sticks at running /scripts/local-block. Quote[ 7.823410][ T1] Run /init as init proceby mikeh49 - Debian
Not good. Quote Marvell>> printenv baudrate=115200 loads_echo=0 rootpath=/mnt/ARM_FS/ kernel2bootup=no usiversion=kacemini-1.5 console=console=ttyS0,115200 mtdparts=spi_flash:0x7f000@0(uboot),0x1000@0x7f000(u-boot-env) CASset=min MALLOC_len=1 ethprime=egiga0 bootargs_root=root=/dev/nfs rw bootargs_end=:::DB88FXX81:eth0:none image_name=uImage standalone=fsload 0x2000000 $(imaby mikeh49 - Debian
Children should not be allowed to play with this stuff. I tried to set up the SSD sata_rootfs and USB rootfs boot scheme and ran into a problem. It sounds similare to renojim in that the USB boots but the SSD does not. I changed the label on the SSD to sata-rootfs and made the changes above to setup bootargs_usb and bootargs_sata. USB booted fine. SSD delayed for a while at the end of boby mikeh49 - Debian