papoteur, this is your baby

CC: (none) => fri
Assignee: bugsquad => yves.brungard

From bug 34877, comment 22:

> @katnatek, are you installing onto a USB flash drive that you've previously
> used with an ISO image? If so, your problem is most likely that the is9660
> filesystem signature is still present on the drive. Try using wipefs to
> examine that drive and to remove any filesystem signatures that shouldn't be
> there.

Isodumper could do, or learn to do, that on any target devices that you use the tool for.

But your error occurred when you tried to install an already dumped iso to a USB key, previously used for a Live ISO, right? So that USB key was at that time a target of the dumped iso, but not of Isodumper. How can isodumper fix that at that time?

CC: (none) => marja11

@katnatek, did you format the target for *system installation* using Isodumper?

(In reply to Morgan Leijström from comment #3)
> @katnatek, did you format the target for *system installation* using
> Isodumper?

I use the usb to make a live USB ,after play sometime format with isodumper and then try to install
If not, perhaps I directly try to install in the usb and after the 1st fail format with isodumper

I not want to mention win2 tools but one I use did make something like this, but if I remember well with 3rd party tool and not let any trace of the previous usage in the usb.

Sorry i do not understand clearly.

(In reply to katnatek from comment #0)
> When you try to make a Classic install to usb device previously used to
> create a LiveUsb, the installation of grub could fail due the presence of
> the iso9660 filesystem

That USB device where Grub is about to install, did you format that USB using Isodumper before running the installer?

Also, are you sure you selected the correct device for configuring boot?
(I ask because i have selected wrong myself before...)

(In reply to Morgan Leijström from comment #5)
> Sorry i do not understand clearly.
> 
> (In reply to katnatek from comment #0)
> > When you try to make a Classic install to usb device previously used to
> > create a LiveUsb, the installation of grub could fail due the presence of
> > the iso9660 filesystem
> 
> That USB device where Grub is about to install, did you format that USB
> using Isodumper before running the installer?

Short answer, yes

> 
> Also, are you sure you selected the correct device for configuring boot?
> (I ask because i have selected wrong myself before...)

Short answer yes
I disconnect the hdd to avoid confusion

(In reply to katnatek from comment #6)
> (In reply to Morgan Leijström from comment #5)
> > Also, are you sure you selected the correct device for configuring boot?
> > (I ask because i have selected wrong myself before...)
> 
> Short answer yes
> I disconnect the hdd to avoid confusion

And you are sure you did not select the device from which installer is booted (where you dumped the .iso) ?

(In reply to Morgan Leijström from comment #7)
> (In reply to katnatek from comment #6)
> > (In reply to Morgan Leijström from comment #5)
> > > Also, are you sure you selected the correct device for configuring boot?
> > > (I ask because i have selected wrong myself before...)
> > 
> > Short answer yes
> > I disconnect the hdd to avoid confusion
> 
> And you are sure you did not select the device from which installer is
> booted (where you dumped the .iso) ?

Yes, after the fail I think I did that but selecting the other device end with grub in the ventoy usb and have to recreate

commit ee7bf6ddbca0423f7a6a799ec34136896903fd58
Author: Papoteur <papoteur@...>
Date:   Thu Jan 1 15:30:50 2026 +0100

    Fix running do_format in magiback (mga#34877 mga#34867)
    
    Restore erasing MBR when formatting in one partition
    Fix messages about partition, not only persistent partition.
---
 Commit Link:
   https://gitweb.mageia.org/software/isodumper/commit/?id=ee7bf6ddbca0423f7a6a799ec34136896903fd58

 Bug links:
   Mageia
      https://bugs.mageia.org/34877
      https://bugs.mageia.org/34867

commit 9ddaf70a5cfa851433ea003be92f1830f4d5dcd1
Author: Papoteur <papoteur@...>
Date:   Thu Jan 1 15:51:01 2026 +0100

    1.58
    
    mga#34877 mga#34867
---
 Commit Link:
   https://gitweb.mageia.org/software/isodumper/commit/?id=9ddaf70a5cfa851433ea003be92f1830f4d5dcd1

 Bug links:
   Mageia
      https://bugs.mageia.org/34877
      https://bugs.mageia.org/34867

A version 1.59 is to test in cauldron.

Before format

wipefs /dev/sdc
DEVICE OFFSET TYPE    UUID                   LABEL
sdc    0x8001 iso9660 2025-12-20-14-23-01-00 Mageia-10-a1-Live-Xfce-x86_64
sdc    0x1fe  dos       

After the format

wipefs /dev/sdc
DEVICE OFFSET TYPE UUID             LABEL
sdc    0x3    ntfs 45EC2D1239C09E1F 
sdc    0x1fe  dos               

Add to the success dialog a suggesting to unplug & replug the usb device will be fine, because if not is done still see the first output

(In reply to katnatek from comment #12)
> Add to the success dialog a suggesting to unplug & replug the usb device
> will be fine, because if not is done still see the first output

Even better if the program can perform that programatically.

Keywords: (none) => feedback

(In reply to katnatek from comment #12)
> Before format
> 
> wipefs /dev/sdc
> DEVICE OFFSET TYPE    UUID                   LABEL
> sdc    0x8001 iso9660 2025-12-20-14-23-01-00 Mageia-10-a1-Live-Xfce-x86_64
> sdc    0x1fe  dos       
> 
> After the format
> 
> wipefs /dev/sdc
> DEVICE OFFSET TYPE UUID             LABEL
> sdc    0x3    ntfs 45EC2D1239C09E1F 
> sdc    0x1fe  dos               
> 
> Add to the success dialog a suggesting to unplug & replug the usb device
> will be fine, because if not is done still see the first output

And can this change be backported to current mageia 9 version?

commit 5d7bf470f90156197b27ce5c47fd9e9ed0a02325
Author: Papoteur <papoteur@...>
Date:   Thu Jan 1 15:51:01 2026 +0100

    1.59
    
    mga#34877 mga#34867
---
 Commit Link:
   https://gitweb.mageia.org/software/isodumper/commit/?id=5d7bf470f90156197b27ce5c47fd9e9ed0a02325

 Bug links:
   Mageia
      https://bugs.mageia.org/34877
      https://bugs.mageia.org/34867

Still in 1.60-1.1 per Bug 33657 Comment 9

Summary: isodumper should remove iso9660 filesystem when format LiveUsb device => isodumper should remove iso9660 filesystem when formatting a LiveUsb device

commit bf8628182b86cfa52089903461b4e0ba87bba450
Author: Papoteur <papoteur@...>
Date:   Mon Jan 5 08:54:06 2026 +0100

    Format: erase also iso 9660 descriptor before formatting (mga#34877)
    
    Improve progress bar
---
 Commit Link:
   https://gitweb.mageia.org/software/isodumper/commit/?id=bf8628182b86cfa52089903461b4e0ba87bba450

A new version is available in Cauldron and MGA9 testing.

RPMS:
=====
isodumper-gtk-1.61-1.mga9.noarch
isodumper-qt-1.61-1.mga9.noarch
isodumper-1.61-1.mga9.noarch

SRPMS:
======
isodumper-1.61-1.mga9

Assignee: yves.brungard => qa-bugs

Seems OK here :-)

gparted now say "[Encrypted]" for that partition after I told isodumper-1.61-1.mga9 to make an encrypted FAT32 partition.

Also, the progress bar is now really visible (not greyed)

The encrypted partition works.

Great.
I did not know that FAT32 could be encrypted with luks.

installing isodumper-1.61-1.mga9.noarch.rpm isodumper-gtk-1.61-1.mga9.noarch.rpm isodumper-qt-1.61-1.mga9.noarch.rpm from //home/katnatek/qa-testing/x86_64
Preparing...                     #####################################################################################
      1/3: isodumper-gtk         #####################################################################################
      2/3: isodumper             #####################################################################################
      3/3: isodumper-qt          #####################################################################################
      1/3: removing isodumper-qt-1.56-1.mga9.noarch
                                 #####################################################################################
      2/3: removing isodumper-1.56-1.mga9.noarch
                                 #####################################################################################
      3/3: removing isodumper-gtk-1.56-1.mga9.noarch
                                 #####################################################################################

Created a LiveUsb OK
After unplug & plug the usb

wipefs /dev/sdc
DEVICE OFFSET TYPE    UUID                   LABEL
sdc    0x8001 iso9660 2025-12-20-14-03-12-00 Mageia-10-a1-Live-Plasma-x86_64
sdc    0x1fe  dos

Dispositivo de destino: Kingston DataTraveler 2.0 (/dev/sdc) 14.538 GiB
Contiene esta(s) partición(es)
/dev/sdc1: Tipo=iso9660, Etiqueta=Mageia-10-a1-Live-Plasma-x86_64
/dev/sdc2: Tipo=vfat, Etiqueta=MGAISO-ESP
Ha ocurrido un error mientras se creaba la partición.

Or translated

Destination device: Kingston DataTraveler 2.0 (/dev/sdc) 14.538 GiB
Contains this partition(s)
/dev/sdc1: Type=iso9660, Label=Mageia-10-a1-Live-Plasma-x86_64
/dev/sdc2: Type=vfat, Label=MGAISO-ESP
An error occurred while creating the partition.

Hi Katnatek,
Please provide pertinent parts of journal and /var/log/magiback.log

(In reply to katnatek from comment #21)
That's my fail, I don't check the devices mounted in the filemanager

wipefs /dev/sdc
DEVICE OFFSET TYPE UUID LABEL
sdc    0x1fe  dos       

So I think is good

(In reply to katnatek from comment #23)
> So I think is good
Be sure!

(In reply to papoteur from comment #24)
> (In reply to katnatek from comment #23)
> > So I think is good
> Be sure!

That was the cause
BTW after a successful image dump the dialog suggest you can unplug the memory
But if you do it then a dialog about no memory is present and when you click on cancel application crash

But as all the other bugs are fixed I think we can go

One thing i wonder, why do isodumper translate lowercase letters given for partition label to uppercase?

I have always used mixed case since i started using linux and never hit a problem...

A quick browsing around say specification for FAT32 allows mixed case but it may be treated case insensitive by the OS:es like filenames are handled.   Other tools allows it lowercase, (even some on Windows) some warn.

For the other filesytsems i think there is no limitation, and i think isodumper should allow lowercase for all filesystems types. (possibly warn for FAT32, but that is entirely optional, IMO)

(In reply to katnatek from comment #25)
> But if you do it then a dialog about no memory is present and when you click
> on cancel application crash
I just tested, it does not crash, but close.
What beheviour do you expect? There is no device, thus nothing to do. Bye.

(In reply to Morgan Leijström from comment #26)
> One thing i wonder, why do isodumper translate lowercase letters given for
> partition label to uppercase?
> 
> I have always used mixed case since i started using linux and never hit a
> problem...
> 
> A quick browsing around say specification for FAT32 allows mixed case but it
> may be treated case insensitive by the OS:es like filenames are handled.  
> Other tools allows it lowercase, (even some on Windows) some warn.
> 
> For the other filesytsems i think there is no limitation, and i think
> isodumper should allow lowercase for all filesystems types. (possibly warn
> for FAT32, but that is entirely optional, IMO)
Upper case is means to be sure that no unexpected behaviour occurs. This is for FAT32 and exFAT in isodumper. The field is also limited to 11 characters.
The FAT norm does not allow lower case in file names, thus also in volume name which is a file name.
https://academy.cba.mit.edu/classes/networking_communications/SD/FAT.pdf p24

I am not sure I understand all this. Two issues addressed in one bug??
Anyway, choosing a simple testcase:
writing M10Plasma live iso to USB stick om my desktop M9 with isodumper 1.56.
Completes successfully, install the updated packages on Compaq laptop, insert the USB and write M10Plasma classic iso to it.
Output of isodumper in its report:
Target Device: ADATA USB Flash Drive (/dev/sdb) 7.241 GiB
Contains this/these partition(s)
/dev/sdb1: Type=iso9660, Label=Mageia-10-a1-Live-Plasma-x86_64
/dev/sdb2: Type=vfat, Label=MGAISO-ESP
Signature file /home/tester9/Downloads/Mageia-10-alpha1-x86_64.iso.sha3.gpg not found

ISO Image to copy: /home/tester9/Downloads/Mageia-10-alpha1-x86_64.iso
Executing copy from /home/tester9/Downloads/Mageia-10-alpha1-x86_64.iso to /dev/sdb
Image Mageia-10-alpha1-x86_64.iso written to /dev/sdb
Bytes written: 5388920832
SHA3 sum: 6B951A765D31E6C9A0AC41210733968B960E17333683D2E2476461C8DEBFDFEC3CD42D3DE6001E43E10F11A713C5FFD487267269372D4A4DFF3939D5B8D298F4

Plug out the USB and plug in again, mount and
$ ls -als /run/media/tester9/Mageia-10-alpha1-x86_64/
total 13
2 drwxr-xr-x  1 tester9 tester9 2048 Dec 20 13:44 ./
0 drwxr-x---+ 3 root    root      60 Jan  6 17:11 ../
1 -rw-r--r--  1 tester9 tester9   80 May 27  2011 autorun.inf
2 drwxr-xr-x  1 tester9 tester9 2048 Dec 20 13:44 boot/
2 -r--r--r--  1 tester9 tester9 2048 Dec 20 13:44 boot.catalog
1 -rw-r--r--  1 tester9 tester9   32 Dec 20 13:44 DATE.txt
2 drwxr-xr-x  1 tester9 tester9 2048 Feb 16  2021 dosutils/
2 drwxr-xr-x  1 tester9 tester9 2048 Dec 20 13:44 EFI/
2 drwxr-xr-x  1 tester9 tester9 2048 Dec 20 13:44 x86_64/
Now trying to boot the USB.

CC: (none) => herman.viaene

USB stick boots successfully into classic installer.

(In reply to papoteur from comment #27)
> (In reply to katnatek from comment #25)
> > But if you do it then a dialog about no memory is present and when you click
> > on cancel application crash
> I just tested, it does not crash, but close.
> What beheviour do you expect? There is no device, thus nothing to do. Bye.

I think the behaviour is correct.

Optimally, to show it better to the user, the button could read "Exit" or "Close Isodumper" 


> (In reply to Morgan Leijström from comment #26)
> > One thing i wonder, why do isodumper translate lowercase letters given for
> > partition label to uppercase?
-
> Upper case is means to be sure that no unexpected behaviour occurs. This is
> for FAT32 and exFAT in isodumper. The field is also limited to 11 characters.
> The FAT norm does not allow lower case in file names, thus also in volume
> name which is a file name.
> https://academy.cba.mit.edu/classes/networking_communications/SD/FAT.pdf p24

OK good then :-)
Some dont follow specifications, for example...
https://www.winhelponline.com/blog/cant-rename-files-upper-lowercase-usb-drives/

---

(In reply to Herman Viaene from comment #28)
> I am not sure I understand all this. Two issues addressed in one bug??
Actually we hammer also
  Bug 33657 - Unable to encrypt a disk with Isodumper 
all in one update.
Both problems by that and this bug titles are now fixed.

Plus we have a more visible progress bar.

We could release now I think, or maybe also rename that exiting key.
I also mentionned i would like a way to see a way to unhide the set encryption key.  But those two can be discussed later instead.

As you wish.

Whiteboard: (none) => MGA9-64-OK

(In reply to Morgan Leijström from comment #30)
> (In reply to papoteur from comment #27)
> > (In reply to katnatek from comment #25)
> > > But if you do it then a dialog about no memory is present and when you click
> > > on cancel application crash
> > I just tested, it does not crash, but close.
> > What beheviour do you expect? There is no device, thus nothing to do. Bye.
> 
> I think the behaviour is correct.
> 
> Optimally, to show it better to the user, the button could read "Exit" or
> "Close Isodumper" 

Both are ignoring the possibility that the user want to insert other usb device and perform other task.

The actual behavior force to open again the application in that case

But That will be other bug

papoteur In the comprobation step, where in the translation files is the checking (Verifying?) <partition> part?

(In reply to katnatek from comment #32)
> The actual behavior force to open again the application in that case

Insert USB, wait two seconds, click update, and you can select it in the dropdown.

(In reply to Morgan Leijström from comment #34)
> (In reply to katnatek from comment #32)
> > The actual behavior force to open again the application in that case
> 
> Insert USB, wait two seconds, click update, and you can select it in the
> dropdown.

What if I not want to walk to the usb port?
I have an extension cable where I connect the usb devices that I use with more frequency

(In reply to katnatek from comment #35)
> What if I not want to walk to the usb port?
> I have an extension cable where I connect the usb devices that I use with
> more frequency

I use that trick too, cable on table. Also avoids wearing out the port on the computer.

We seem to release 10alpha1 next day or so.
We should have this fixed Isodumper out to our Mageia 9 users so they can use it for the ISOs.

Keywords: (none) => validated_update
CC: (none) => sysadmin-bugs

updated to latest version.

confirmed iso9660 signature removed

CC: (none) => westel

An update for this issue has been pushed to the Mageia Updates repository.

https://advisories.mageia.org/MGAA-2026-0003.html

Status: NEW => RESOLVED
Resolution: (none) => FIXED