I also blindly flashed the new u-boot since I'm trying to clean things up and put this thing into production with a USB boot drive and my "eSATA" data drive (I ran a SATA cable out the slot between the top and the base of the M300). Everything looks good! (As usual. :-) I haven't played with the netconsole stuff yet. Is it still too early?by renojim - Debian
I wonder if there's something else going on with the flash. I was trying to get netconsole working and playing with env vars from u-boot and after adding a few I got a CRC error on the next boot. It's like I can change one and save it with no problem, but adding a few will eventually cause a CRC error.by renojim - Debian
Now that you mention it, I set bootdelay from the M300 prompt. Sure enough, if I try to use fw_setenv I get a CRC error. Definitely not a huge drawback in my opinion. I'd much rather have reliable booting than the ability to set environment variables from within Debian. :-)by renojim - Debian
bodhi, sorry for the false alarm about USB booting. I thought my uImage.orig was of dubious origins and was in fact from a different kernel version. I have to stop experimenting after midnight. I recreated uImage and can now boot from SSD or USB. Very nicely done! One thing, I don't see any CRC errors when I boot and I was able to change and save bootdelay. Maybe I'm misundersby renojim - Debian
Well it boots from SATA, but I can't get it to boot from USB. I need to take a break since I've watched this thing boot or not boot dozens of times today. Since the DTB is removed from uImage should there be a directive in uEnv.txt? I tried adding one, but it still kernel panics when I try to boot from USB. Without any dtb entry I got something like a machine ID error and somethinby renojim - Debian
Cloning the reliable SSD to the unreliable one definitely made things better. I did get this a couple of times: U-Boot 2021.10-tld-1 (Feb 22 2022 - 17:41:54 -0800) Dell KACE M300 SoC: Kirkwood 88F6282_A1 Model: Dell KACE M300 DRAM: 2 GiB Loading Environment from SPIFlash... SF: Detected mx25l4005a with page size 256 Bytes, erase size 4 KiB, total 512 KiB *** Warning - bad CRC, usby renojim - Debian
sudos Wrote: > does it still have the same no-boot behavior > unless something is plugged into USB with the > latest kernel/rootfs combo? One of my two SSDs still has that behavior. They're running two different debian versions (buster vs bullseye), so I don't know if that has anything to do with it. I'll have to check if they're also using different kernelby renojim - Debian
After booting many times with the new u-boot using my two stock SSDs one boots reliably while the other does the same old sometimes it boots, sometimes it kernel panics, sometimes it segmentation faults, and sometimes it spews error messages ad nauseam. I think it still boots reliably if I have something in the USB port. I'm ready to flash this anyway since it boots my USB flash drive reby renojim - Debian
sudos, welcome to the party! I mentioned it in another thread, but I've been meaning to mention here as well where maybe it will get more attention. I removed the CPU from one of my M300s and it's like nothing I've ever seen. I could trace the GPIOs if I could just a pinout for the 88F6282 in the FCBGA package, so if anyone has a source for Marvell data sheets let me know. -Jby renojim - Debian
LostFile, you skipped the part where bodhi said to interrupt u-boot and then type 'printenv'. printenv from within Linux is very different than printenv from u-boot.by renojim - Debian
It's late again, but this time I put uEnv.txt on the SATA SSD. It looks like SATA booting is still unreliable unless there's something in the USB port. On the other hand, USB booting seems to be more reliable. I have a flash drive that sometimes will boot with the old u-boot and sometimes won't, but I think it has never failed with the new u-boot. I tried to capture some logsby renojim - Debian
That was stupid of me! That's what I get for trying something new at 5am.by renojim - Debian
USB booting went well, but not SATA. I had two of the 16GB SSDs lying around, but I honestly don't know where I left off with them. I can get one to boot with the old u-boot if I have something in the USB port. With the new u-boot I get a kernel panic with or without something in the USB port. By the way, I had to add the -p flag for kwboot. Here's my printenv output: baudraby renojim - Debian
bodhi Wrote: ------------------------------------------------------- > Aslo, I guess somebody can spend the time to trace > the circut board to see where the GPIO pins go, > and if there is anything odds about them... I had my hot air tool out so I thought I'd pop off the CPU and see what I could find. That was probably a mistake. I was expecting something like the CPU onby renojim - Debian
Just out of curiosity, has anyone tried to disassemble any u-boot before? Not worth the trouble or too hard to follow the disassembled code?by renojim - Debian
Cool, I'm glad to hear it! Originally I wanted to use the SSD as the boot drive and use a USB hard drive for data, but I've changed my mind. I'm going to use a flash drive for the boot drive and the SATA port for a large eSATA drive, but it would be nice to know what the USB/SATA power issue is. I've only booted in that configuration a few times, but so far I haven't hby renojim - Debian
The new link works! I'm trying to get my M300 up to date. Thanks, bodhi!by renojim - Debian
Hi bodhi, The link for linux-5.16.5-kirkwood-tld-1-bodhi.tar.bz2 gives a 404 Thanks! -JTby renojim - Debian
RE: ECC errors I've been getting them for years on mtdblock0 & 1. It's always bothered me, but since everything seems to work I've never bothered with it. I don't think any amount of rebooting gets rid of them. [ 70.720733] __nand_correct_data: uncorrectable ECC error [ 70.726179] blk_update_request: I/O error, dev mtdblock0, sector 2040 [ 70.732987] __nby renojim - Debian
Thanks for that link! I had completely forgotten about archlinux having abandoned it years ago.by renojim - Debian
@davidalfa, wasting money? Nah, that's what hobbies are for! Thanks for the list. Most of that stuff I already have (or suitable substitutes) and the animal is more than willing, if not capable. :-) I'm going to have to give this a try. I have a spare Dockstar I acquired for just such a project. Thanks to your photo I can also maybe trace some of the unused MPP/GPIO pins (assuby renojim - Debian
That's awesome! I'd love to try this, but I don't know about reballing the CPU. I've only briefly looked into it. Any advice? What would I need to acquire and does it make sense financially for something I may only use once or twice?by renojim - Debian
I may have figured out what changed. My original lkm was built with gcc v4.9.2 and the one that requires the flag uses gcc v6.3.0. I don't know if that points to something wrong with my gcc install or if some defaults changed, but at least I know how to get around the error.by renojim - Debian
No, that's the thing. As I said, I'm long overdue for an update on my Dockstar (still playing with the M300) and I'm on 4.8.3-kirkwood-tld-1. It's weird since I don't think I've changed anything as far as the kernel is concerned. It just bugs me that it's suddenly necessary. I should just go ahead and get this thing up to date, but I haven't fully recovby renojim - Debian
Has anyone that builds loadable kernel modules run into the requirement for a -fno-pic compiler flag? I have a couple of lkms I created a few years ago and when I went to modify one I suddenly get "wx_lkm: Unknown symbol _GLOBAL_OFFSET_TABLE_ (err 0)" when I make and insmod it. I don't believe I've changed anything and I never needed the flag before, but searching for the eby renojim - Debian
@cratsil1979, I'm not going to claim to fully understand it, but it's because of the LetsEncrypt certificates that had a change on the first of the month. The easiest thing to do is to tell wget not to check certificates: wget --no-check-certificate -O "/etc/apt/trusted.gpg.d/openmediavault-archive-keyring.asc" https://packages.openmediavault.org/public/archive.key You migby renojim - Debian
Just reading along with this thread for my own curiosity I can see there's some confusion going on. @Bobby has kwboot'ed U-boot, but he hasn't booted into Debian yet - that's why there's confusion about fw_setenv, ls, and cp.by renojim - Debian
Last login: Wed Dec 31 16:01:59 1969 from Try setting the date first.by renojim - Debian
My thoughts (given my extreme hatred of anything java) are to find another server. :-) Having said that, I've fought this cacerts problem multiple times when trying to install the latest JRE without ever finding a solution (adding to my hatred of java). What I did finally get to work was installing an older version. I don't remember exactly what I did, but checking my history it appby renojim - Debian
Ok, so I'm finally sober enough to revisit this. :-) The TL;DR version is don't use NTFS to do disk tests. The SATA port is substantially faster than the USB port on the M300, which is what I had expected. Here's the data with the same Seagate 1TB drive formatted as ext4: USB read: root@kaceM300-2:/tmp/usb# sync ; dd if=zeros.bin of=/dev/null bs=4M count=1024 ; sync 1024by renojim - Debian