bluzfanmr1 Wrote: ------------------------------------------------------- > dhargens Wrote: > ------------------------------------------------------- > > I'm having a problem with upgrading the Debian > > kernel from 5.13 to 6.3.5 and using it in one > > of the rear USB 3 ports. > > I had the same issue and had to go back to the > earlier kernel toby dhargens - uBoot
bodhi Wrote: ------------------------------------------------------- > When u-boot crashes and reset (due to wrong envs > settings), netconsole artifacts cause a big > problem. I think that's what dhargens saw a few > posts ago, and had to kwboot the 2017.07 version > to recover. Don’t know if this helps, but I’m not enabling netconsole because I have hardwired aby dhargens - uBoot
I'm having a problem with upgrading the Debian kernel from 5.13 to 6.3.5 and using it in one of the rear USB 3 ports. I've created a USB 3 drive with rootfs & kernel 5.13, install it on the V4 and then update it to 6.3.5. U-Boot 2023.04-tld-1 (May 10) is installed on this. It boots fine from the rear port when booting in 5.13, but when the installation to 6.3.5 completes andby dhargens - uBoot
bodhi Wrote: ------------------------------------------------------- > Look at /etc/rc.local. I have some example in this initscript. > This /etc/rc.local is the last script Debian runs > during booting. So you can echo value cpu > to the trigger for the status LED. Ah, yes, I remember seeing that before. Thanks for the reminder. > For example, to make the box boot faby dhargens - uBoot
bodhi Wrote: ------------------------------------------------------- > That's the differrence between the mainline Debian > installation and a Debian derivative distro. Each > Debian version, eg. bullseye and bookworm, is > pegged to a Linux kernel version. So there is a > misconception is that you have to run a Linux > kernel that is consistent with Debian version.by dhargens - uBoot
Good to know. I'd hoped I could follow instructions - though I'd skipped some before... Now I've got solid steps to update the couple of boxes I've been messing with. Thanks for the your help. It's appreciated.by dhargens - uBoot
bodhi Wrote: ------------------------------------------------------- > We were testing. The versioning scheme only > applies after it was formally release. > > So you are running an interim version. So how do I get to the current version? I downloaded & re-installed (I think) the latest uboot using the instructions, and still ended up with this interim version. In theby dhargens - uBoot
bodhi Wrote: ------------------------------------------------------- > Furthermore, the version tld-1 is the unique extra > version. > > U-Boot 2023.04-tld-1 (May 25 ....) > > It means my patch version number is tld-1 (tld > stands for Tony's Linux Distribution). The > timestamp is really redundant. If I release > another 2023.04 u-boot for this box thby dhargens - uBoot
I installed the latest uboot in the Post 1 instructions, and it displays the following at the beginning of the boot: U-Boot 2023.04-tld-1 (May 10 2023 - 23:00:14 -0700) Pogoplug V4 much as I would expect it to do. However, I don't know if it's really the latest version or not. How can I tell if I'm up to date on uboot versioning? Also, bodhi Wrote: -------------------------by dhargens - uBoot
I updated a previously updated 2023.04 with the newest version from Post 1, and it boots fine. I don't have a serial connection for another week, so I can only see dmesg output, and it looks fine. I corrected the 'ethaddr' to what's on the V4's label, but otherwise all vars are from the update. I see there's no 'cpu' in the LED trigger file, though I'mby dhargens - uBoot
bodhi Wrote: ------------------------------------------------------- > I think we might not be done yet (my mistake in the envs image). So stay tume. I'll be glad to help when I return on the 5th or so.by dhargens - uBoot
bodhi Wrote: ------------------------------------------------------- > I migth be able to make them less noisy. But that > will be some day when I can revisit this subject. No worries. As long as it's expected, then I'm good with it. Was trying to point out a difference between boots, which didn't really make sense to me. Seems it's USB hardware, and when I get bacby dhargens - uBoot
bodhi Wrote: ------------------------------------------------------- > dhargens, > > I don't see any error in this log. > > Loading envs from usb 1... > > ** Bad device specification usb 1 ** > > Couldn't find partition usb 1:1 > > Can't set block device > > Loading envs from usb 2... > > ** Bad device specification usb 2 **by dhargens - uBoot
bodhi Wrote: ------------------------------------------------------- > That truncated log is not good! > > When you post the good log, please also do this in Debian. > cat /proc/mtd Here's that output on my live system: $ cat /proc/mtd dev: size erasesize name mtd0: 00200000 00020000 "u-boot" mtd1: 00300000 00020000 "uImage" mtd2: 00300000by dhargens - uBoot
One difference between logs 1, 2 and 3 might be that I'm using different USB drives on each boot - some older Sandisk Cruzer Fits (2.0) and a new Sandisk Ultra Fit (3.2). The older Cruzer has an led on it, while the new ones don't - and neither does the Ultra. Didn't think there'd be a difference between them, but maybe they've changed the hardware between 'modelsby dhargens - uBoot
And here's a non-truncated boot from the rear port using a USB 3 drive. It too has errors in the same place as above. U-Boot 2023.04-tld-1 (May 10 2023 - 23:00:14 -0700) Pogoplug V4 SoC: Kirkwood 88F6192_A1 Model: Cloud Engines PogoPlug Series 4 DRAM: 128 MiB Core: 20 devices, 16 uclasses, devicetree: separate NAND: 128 MiB MMC: mvsdio@90000: 0 Loading Environment from NANDby dhargens - uBoot
Here's a non-truncated log of a boot with a USB 2 drive in the top port. I note that there's more error messages ("Bad device specification") when loading envs from usb & sata - but I don't know why. U-Boot 2023.04-tld-1 (May 10 2023 - 23:00:14 -0700) Pogoplug V4 SoC: Kirkwood 88F6192_A1 Model: Cloud Engines PogoPlug Series 4 DRAM: 128 MiB Core: 20 devicesby dhargens - uBoot
bodhi Wrote: ------------------------------------------------------- > Please post the boot log. I'd like to double check a few things. Here's an early boot log, with some lines truncated due to capturing a minicom window. I'm not home to check these things until June 5 or 6, but it's a start. I'll post another log I recorded (not truncated) in the next post.by dhargens - uBoot
bodhi Wrote: ------------------------------------------------------- > Here is the updated boot script. I've had a wrong > mtdparts in the script before. > > Please try again. Yup, that fixed it. Boots fine, did BOTH parts of Step 6 (flash_erase & nandwrite), then removed the boot.scr because it no longer needed the crutch, and it booted as expected. Changed the MACby dhargens - uBoot
bodhi Wrote: ------------------------------------------------------- > Here is the updated boot script. I've had a wrong > mtdparts in the script before. It should be: setenv mtdparts "mtdparts=orion_nand:2M(u-boot),3M(uImage),3M(uImage2),8M(failsafe),112M(root)" > Please try again. That change makes sense. I'll get to it after midnight CT (in office now) anby dhargens - uBoot
bodhi Wrote: ------------------------------------------------------- > > So at least now it boots fine. Do I erase the NAND > > block next and retry the nandwrite of the image > > next, or do you have more I need to do? > > Yes, so back to Step 6 and onward. > Quote6. Flash default u-boot envs image So I tried: root@deb05:/tmp# /usr/sbin/flash_erase /dev/mtby dhargens - uBoot
It booted fine - found the uboot script /boot.scr and continued onward. Here's the console log to the boot prompt: U-Boot 2023.04-tld-1 (May 10 2023 - 23:00:14 -0700) Pogoplug V4 SoC: Kirkwood 88F6192_A1 Model: Cloud Engines PogoPlug Series 4 DRAM: 128 MiB Core: 20 devices, 16 uclasses, devicetree: separate NAND: 128 MiB MMC: mvsdio@90000: 0 Loading Environment from NAby dhargens - uBoot
bodhi Wrote: ------------------------------------------------------- > Did you erase the NAND block first? Quote6. Flash default u-boot envs image > > If you are upgrading from 2017.07-tld-1 u-boot, > you can skip this step. But some envs will need to > be updated. See Note section below after the > installation steps. > > Be extra careful with the next 2 comby dhargens - uBoot
Ok, after some hardware issues and other unexpected problems, I finally got things working. Here's my findings. Booted USB into Debian 5.13.6-kirkwood-tld-1, which has been updated to current. Connected by serial port. Steps: confirmed kwboot for this uboot works. It did the last time too, so I had no worries. stepped thru Sect B (Flashing) steps 1-7. No problems. 1. /etc/fw_env.cby dhargens - uBoot
bodhi Wrote: ------------------------------------------------------- > Are you ready to test the installation instruction? Yes, I have some time set aside for this evening and should be able to get started. Are there any other notes I need to know about before doing this? Pointers, maybe? (Not worried about it, just want to make sure I get the steps down correctly. And...I've sleptby dhargens - uBoot
bodhi Wrote: ------------------------------------------------------- > I've uploaded an updated version in the first > post. Great to hear! I'll get to it next week - busy weekend ahead, starting tomorrow.by dhargens - uBoot
bodhi Wrote: ------------------------------------------------------- > The CPU LED trigger is in the kernel. So you will > see it in the next kernel release. Way cool! > I would suggest a smaller file such as 50MB. This > box only has 128MB RAM. So if you use large file, > i.e 1GB, it will swap out to disk, and then the > test is invalid. Makes perfect sense. I knby dhargens - uBoot
I was successful kwbooting from the new 2023.04 image. Made a new rootfs with the 5.13.6-kirkwood kernel on a SanDisk 3.2Gen1 (Ultra Fit) and it worked fine. I didn't set any environment vars to do it, and it didn't fall apart when I upgraded the system & completed the install etc. I would consider that a success. Was hoping for the cpu setting for the LED too, but this is aby dhargens - uBoot
bodhi Wrote: ------------------------------------------------------- > No, it is not available in the current Kirkwood kernel here. I'll add it in the next release. It's > a better visual indicator than just a solid green health status! I'd agree - and I look forward to that! Also, while a few of the other options are obvious, is there a chart that explains what chooby dhargens - uBoot
Hello. I had originally installed Arch uboot in a few of my Pogo V4s. Now that I'm installing the doozan uboot to my next batch of V4s (and looking forward to the newest release with rear USBv3 booting! nudge nudge, wink wink) I was setting it up for similar operation. Because I don't want the green led to always be on, I liked setting it to show me CPU activity. On Arch uboot I cby dhargens - uBoot