WelcomeWelcome | FAQFAQ | DownloadsDownloads | WikiWiki

Author Topic: tce-update bug  (Read 2877 times)

Offline bmarkus

  • Administrator
  • Hero Member
  • *****
  • Posts: 7035
    • My Community Forum
tce-update bug
« on: November 03, 2012, 05:09:53 PM »
I have three extensions to update, Xprogs, flwm_topside and htop. Their total size is 216k, however tce-update reports there is no enough free disk space for easy mode. In fact, free space is 44M in the disk where /tce is located and 183M free RAM.

Same in 4.6.3 and 4.7

When system is up to date, tce-update still prints a full screen claiming disk space, advicing boot with base norestore text and so on and at the end says my system is up to date :)

In X, 'Your system is up to date' message is missing, only 'Scan for updates completed' printed (4.7)
« Last Edit: November 03, 2012, 05:18:21 PM by bmarkus »
Béla
Ham Radio callsign: HA5DI

"Amateur Radio: The First Technology-Based Social Network."

Offline roberts

  • Administrator
  • Hero Member
  • *****
  • Posts: 7361
  • Founder Emeritus
Re: tce-update bug
« Reply #1 on: November 03, 2012, 09:25:36 PM »
Appears to be a busybox bug that also affects x86 Core.

The default for df is stated to be in k. See df --help

Yet using df /dev/sda1 the result is in human readable form, as if the -h flag was used.

Now using the same command as:
sudo df /dev/sda1 the result is in the default k
10+ Years Contributing to Linux Open Source Projects.

Offline tinypoodle

  • Hero Member
  • *****
  • Posts: 3857
Re: tce-update bug
« Reply #2 on: November 03, 2012, 10:19:43 PM »
Appears to be a busybox bug that also affects x86 Core.

The default for df is stated to be in k. See df --help

Yet using df /dev/sda1 the result is in human readable form, as if the -h flag was used.

I suspect that might be a shell alias rather than a busybox bug.

Code: [Select]
'df'would tell

Note: Quotes are literally meant to be included in this very case (=unalias)!
"Software gets slower faster than hardware gets faster." Niklaus Wirth - A Plea for Lean Software (1995)

Offline curaga

  • Administrator
  • Hero Member
  • *****
  • Posts: 10274
Re: tce-update bug
« Reply #3 on: November 04, 2012, 02:39:50 AM »
Yes, it's due to the default alias being "df -h".
The only barriers that can stop you are the ones you create yourself.

Offline roberts

  • Administrator
  • Hero Member
  • *****
  • Posts: 7361
  • Founder Emeritus
Re: tce-update bug
« Reply #4 on: November 04, 2012, 05:11:50 AM »
My bad. However the alias does not even come into play with the script's calculation.
Using tce-update on my pi, I forced an update to be required my editing md5s. It worked as expected. Calculation is fine.
But if you always see the "base norestore" message then that would suggest that there is no tce directive on the command line.
Use showbootcodes to check.
10+ Years Contributing to Linux Open Source Projects.

Offline bmarkus

  • Administrator
  • Hero Member
  • *****
  • Posts: 7035
    • My Community Forum
Re: tce-update bug
« Reply #5 on: November 04, 2012, 05:52:17 AM »
There are no tce directeve used. tce dir is in the root of /dev/sda1, found and used correctly by the system. How tce direcive is influencing operation of tce-update?
Béla
Ham Radio callsign: HA5DI

"Amateur Radio: The First Technology-Based Social Network."

Offline roberts

  • Administrator
  • Hero Member
  • *****
  • Posts: 7361
  • Founder Emeritus
Re: tce-update bug
« Reply #6 on: November 04, 2012, 06:27:28 AM »
It is faster to use a boot code of tce= than to have the system search for one. I always use such a directive.
tce-update is not new nor has any changes been recently made. Using the tce boot code should suffice.
I will consider an update to support the slower method of system scanning all drive to find tce directory.
10+ Years Contributing to Linux Open Source Projects.