Debian Jessie dCore

[netnomad]

hi jason,

for me waitusb with label is still unstable, but i have no problem to change my bootcode to waitusb=5 without label.
wheezy  and all other cuts work fine with waitusb=5:LABEL=tc.

never mind, thank you for your help and commitment.

[Jason W]

What is the md5sum of your dCore-jessie.gz?


Just to be sure.

And do you see the "Waiting as requested.." line when booting dCore-jessie at all?

And what is the output of the below command once dCore-jessie has booted:

blkid -it LABEL=tc

[netnomad]

What is the md5sum of your dCore-jessie.gz?

i renamed dCore-jessie.gz to dCore.gz:
tc@box:~/.MNT/sdb1/live/dcore-testing/boot$ md5sum dCore.gz
1b2e5d9ba7ff976652f48698a3a44664  dCore.gz
that md5sum should be correct.

blkid -it LABEL=tc

tc@box:~/.MNT/sdb1/live/dcore-testing/boot$ blkid -it LABEL=tc
The low-level probing mode requires a device

tc@box:~$ blkid -s LABEL /dev/sdb2
/dev/sdb2: LABEL="tc"

tc@box:~$ sudo lsblk -o name,label,mountpoint /dev/sdb
NAME   LABEL   MOUNTPOINT
sdb           
|-sdb1 MYSTICK /mnt/sdb1
`-sdb2 tc      /mnt/sdb2

i attached the photos of my two boot-versions,
first with label the failed boot process
second without label and a successful boot process

[Jason W]

The command is "blkid -lt LABEL=TC", I made a typo.

Oh, and I assume you have renamed vmlinuz-jessie to vmlinuz?

One thing that may help is to start with only the tce directory being specified on the boot line, and if that works start adding each one back one at a time and see at what point it fails.

[netnomad]

The command is "blkid -lt LABEL=TC", I made a typo.

tc@box:~$ blkid -lt LABEL=TC
tc@box:~$ sudo blkid -lt LABEL=TC
tc@box:~$

there is no echo or result to this command.

Oh, and I assume you have renamed vmlinuz-jessie to vmlinuz?

yes, you're right.

One thing that may help is to start with only the tce directory being specified on the boot line, and if that works start adding each one back one at a time and see at what point it fails.

my whole bootcode works fine, only waitusb with label-mode waitusb=5:LABEL=tc makes problems:

kernel /live/dcore-testing/boot/vmlinuz waitusb=5 protect restore=LABEL=tc tce=LABEL=tc/tce lst=sceboottest.lst vga=795 psmouse.proto=imps settime tz=CET-1CEST,M3.5.0,M10.5.0/3 laptop kmap=de multivt lang=de_DE.UTF-8 mydata=dcoret blacklist=snd_usb_audio intel_rng.no_fwh_detect=-1 printk.time=1 cron desktop=fluxbox log=3 acpi_sleep=s3_bios,s3_mode showapps
append initrd=/live/dcore-testing/boot/dCore.gz

[Jason W]

Using lowercase tc with the command should work.  If not, then that is the issue as the blkid command is used to check whether the device is available and then exists the waitusb loop.

[netnomad]

tc@box:~$ blkid -lt LABEL=tc
/dev/sdb2: LABEL="tc" UUID="6cc7058b-ac89-40a1-8700-e43e47539ce3" SEC_TYPE="ext2" TYPE="ext3"

in my common bootcode i use lowercase, too:
waitusb=5:LABEL=tc

[Jason W]

Ok, I will try my setup with a two digit label and ext3, now I have fat32 with label SCE.  It shouldn't make a difference, but there is something making my setup work and yours not.

[tinypoodle]

tce=LABEL=tc/tce

Why not

Code: [Select]

tce=LABEL=tc
?

[netnomad]

hi tinypoodle,

you are right that there is no need for
tce=LABEL=tc/tce
so i tried also
tce=LABEL=tc

there is no difference for me, i have tried both.

the faq describes the option to use a path to the tce-directory, f.e. tce=LABEL=tc/path/tce,
so both should work.

[Jason W]

I have found that there seems to be an issue when using an ext2 formatted usb drive.

I will look into it.

[Jason W]

This was a udev issue in Debian Jessie that needed a simple udev fix.  Thank you netnomad, JLS, and any others that tested and reported this issue, for helping to make dCore/dCore-jessie better than it was yesterday.  The fix is also applicable or at least not damaging to dCore Wheezy, also updated in the release_candidates area.

[netnomad]

hi jason,

yeah, you solved it

thank you for your great work.

[netnomad]

hi jason,

sorry, but there are still minor issues, so the last cut is still a bit unstable.

i guess a warm reboot of dcore-jessie could initiate errors,
rebooting from tinycore 4.7.7 to dcore-jessie works better.
a cold boot process seems to be more stable.
there could be a problem with the reboot of unclean-ext3-filesystems, too?
or is it just the problem that a warm reboot causes issues with the init of the usb-bus?

you improved the situation, but some reboots still fail

[Jason W]

Does your usb drive have an unclean file system?  Fixing the fs would remove one variable that could certainly cause issue.

What errors do you see?

EDIT:  I am still having LABEL boot issues with ext2 usb, I will look into it.