xanatos, I don't reread posts in unless I try to find something. Did not realize there were more info when you updated. > So the stock kernel seems to be much much faster > (at least with btrfs). I did not pay much attention to btrfs. So not sure what can be improved in the kernel or user space. > Made some tests with Ext4. The Ext4 partition is > at the end of thby bodhi - Debian
user45 Wrote: ------------------------------------------------------- > OK have my usb ready to go. Cool! > It doesnt boot to debian because I think I need to > interrupt u-boot count down as per > https://forum.doozan.com/read.php?2,102054,105374#msg-105374 > So I guess that means resolder my serial port > connections? Yes. If you want to boot Debian with stockby bodhi - Off-Topic
user45, That installation method is quite old. See here for installing latest Debian and u-boot: https://forum.doozan.com/read.php?3,134562,134565#msg-134565by bodhi - Off-Topic
Quote> docker: Error response from daemon: failed to > create task for container: failed to create shi m > task: OCI runtime create failed: runc create > failed: unable to start container process: err or > during container init: error setting cgroup config > for procHooks process: bpf_prog_query(B > PF_CGROUP_DEVICE) failed This docker error is not related to bootargsby bodhi - Debian
Quotedid I do something wrong? Yes. The last initramfs is for the older kernel. > ls -lart /boot ...... -rw-r--r-- 1 root root 10065543 Jun 4 08:41 initrd.img-5.13.8-mvebu-tld-1 The docker installation populated this: > ls -lart /lib/modules lrwxrwxrwx 1 root docker 18 Jun 4 22:19 6.3.5-mvebu-tld-1 -> /path/to/module.ko So you need to reinstall kernel 6.3.5-mvebby bodhi - Debian
Chris, Yes. It should be there. depmod: ERROR: could not open directory /lib/modules/6.3.5-mvebu-tld-1: No such file or directory The error above error indicated that you have not installed the kernel 6.3.5-mvebu-tld-1 in the rootfs. Do uname -a ls -lart /lib/modules ls -lart /boot mount lsblk df -hby bodhi - Debian
How To Shrink initramfs The approach described here in this tutorial works perfectly for the kernel and rootfs I releases here: Kirkwood and MVEBU. In general, it should work for any Debian kernel. However, it aso depends on how a particular kernel is configured. IOW, whether the necessary mouldes were already built into the kernel, or some other loadable modules must be loaded early. Pleaby bodhi - Debian
Chris, > So now I have the newest kernel with Debian, > on the stick, only the uImage and uInitrd. > I know this is not a complete Linux and kernel, That's incorrect. 1. The stick has the kernel files only (not a rootfs), and you are running the Debian-5.13.6-kirkwood-tld-1-rootfs-bodhi.tar.bz2 rootfs on HDD. This rootfs is a complete Debian system, with the kernel runnby bodhi - Debian
Chris Wrote: ------------------------------------------------------- > But I can remain with the uImage and uIntrd on the > stick till then, yes? > > Thank You! For now, boot with both EXT3 USB rootfs and HDD rootfs. The kernel will load the HDD rootfs. And then we'll update a bit more logistics to ensure that will happens everytime.by bodhi - Debian
OK, we'll pick it up tomorrow.by bodhi - Debian
run sata_init sleep 5 Do it a few timesby bodhi - Debian
bodhi Wrote: ------------------------------------------------------- > Ah! I forgot. > > > run sata_init > sleep 5 > ext2ls scsi 0:1 /boot > > If the files are listed correctly then do, > > ext2load scsi 0:1 0x00800000 boot/uImage; ext2load > scsi 0:1 0x02100000 boot/uInitrd; bootm 0x00800000 > 0x02100000' > boot > See above.by bodhi - Debian
Ah! I forgot. run sata_init sleep 5 ext2ls scsi 0:1 /boot If the files are listed correctly then do, ext2load scsi 0:1 0x00800000 boot/uImage; ext2load scsi 0:1 0x02100000 boot/uInitrd; bootm 0x00800000 0x02100000' bootby bodhi - Debian
Try scsi initby bodhi - Debian
How about: scsi reset sleep 5 ext2ls scsi 0:1 /boot or if scsi reset failed: scsi init If the files are listed correctly then do, ext2load scsi 0:1 0x00800000 boot/uImage; ext2load scsi 0:1 0x02100000 boot/uInitrd; bootm 0x00800000 0x02100000' bootby bodhi - Debian
Chris, I've corrected the typos (thanks bluzfanmr1).by bodhi - Debian
> boot successful :) Cool! > Next step? > I'm starting to be sleepy, here is 23 ... Come back tomorrow and try these next steps (before I'll wake up in PST timezone :) Remove the USB roofs, and plug in the HDD rootfs. Power up, interrupt u-boot countdown and, ide reset If the output shows that the HDD is initialized with capacity,... then do, ext2ls ide 0:by bodhi - Debian
I see, boot_debian_usb=usb start; sleep 5; fatload usb 0:1 0x00800000 /boot/uImage; fatload usb 0:1 0x02100000 /boot/uRamdisk; bootm 0x00800000 0x02100000 bootcmd=run set_bootargs_debian; run boot_debian_usb; reset So you are still booting with the FAT USB drive using fatload. See this Step 4 in this thread. https://forum.doozan.com/read.php?2,133718,133722#msg-133722 The new u-by bodhi - Debian
Chris, > But I am booting the uImage and uInitrd from the > stick. > Can I boot somehow direct from HDD? Yes you can. I forgot that's where you are. > It didn't work because uInitrd is too big, no? Not that reason. You have updated u-boot so the initrd size is irrelevant. We need to changes some envs to make it boot HDD rootfs. Also make it boot USB if the HDDby bodhi - Debian
Chris, > do I need to copy now the uImage and uInitrd to > the stick? If you intend to use the stick as a rescue system then you should install the new kernel on the USB rootfs. Remove the HDD, and boot back to USB, and then do the kernel installation. If you don't think you would use the USB rootfs at all, then copying the kernel files only would be OK. As I mentionedby bodhi - Debian
Looks like sending PM is broken. Thanks bluzfanmr1 for pointing out the typos in the release threads! bluzfanmr1 or Chris, please try sending PM to me (test if it's really broken).by bodhi - Off-Topic
Chris, > now I've deledet mdo and changed only to /dev/sda1 > to rootfs and it's booting That's what I would do. > another question the update to newr kernel doesn't > work > > 2023-06-02 22:48:46 (15.4 MB/s) - > 'linux-6.2.8-mvebu-tld-1-bodhi.tar.bz2' saved > [37044194/37044194] > > root@debian:/boot# tar -xjf >by bodhi - Debian
Mischif, > Still no luck with emmc, same pattern every time - > when I boot the device after a long while the > kernel sees the mmc, but I get an error every time > I try to interact with it; when I reboot either > with the same kernel or an updated one I get the > "Card stuck being busy" error. I think a capacitor > or something is blown out on my board.by bodhi - Debian
Usually, the rootfs is on md0.by bodhi - Debian
Chris, [ 8.244490][ T1817] md/raid1:md0: active with 2 out of 2 mirrors [ 8.251369][ T1817] md0: detected capacity change from 0 to 2240512 [ 8.257932][ T1822] md/raid1:md2: not clean -- starting background reconstrun [ 8.267977][ T1832] md/raid1:md1: active with 2 out of 2 mirrors [ 8.268578][ T1822] md/raid1:md2: active with 2 out of 2 mirrors [ 8.274362][ T1832] md1by bodhi - Debian
Hi Mischif , I've released kernel 6.3.5-mvebu-tld-1. QuoteKernel 6.3.5-mvebu-tld-1 package has been uploaded. New/Updated in this release: - General kernel upgrade. Some code and DTS rebase were needed in the patch. - Add support for Mavell MMP3 SoC (CONFIG_ARCH_MMP and assiocated configs). - Add support for Dell Wyse 3020 (works on eMMC in still in progress). -Add CONFIGby bodhi - Debian
Kernel 6.3.5-mvebu-tld-1 package has been uploaded. See 1st post for download link. Please create a new thread for any installation problem or questions.by bodhi - Debian
Kernel linux-6.3.5-kirkwood-tld-1 package has been uploaded. See 1st post for download link. And please create a new thread for any installation problem or question.by bodhi - Debian
> done that, Can you show the log of the boot with USB rootfs and then how you run "mdadm --assemble --scan", and mount the rootfs partition.by bodhi - Debian