https://bugs.kde.org/show_bug.cgi?id=523537

Méven <[email protected]> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
      Latest Commit|                            |https://invent.kde.org/util
                   |                            |ities/isoimagewriter/-/comm
                   |                            |it/72e925414b23363e2cd5def4
                   |                            |9a380f80b4b2f451
         Resolution|---                         |FIXED

--- Comment #3 from Méven <[email protected]> ---
Git commit 72e925414b23363e2cd5def49a380f80b4b2f451 by Méven Car.
Committed on 27/07/2026 at 17:09.
Pushed by meven into branch 'master'.

Say that the target device is in use instead of blaming the last block

The device is opened through UDisks2 with O_EXCL, so the open is turned down
while anything else still holds the device. That is what a live system does
to the medium it was booted from, and what any program still reading from
the stick does.

Neither the reply to OpenDevice nor the QFile::open that follows it were
looked at. The invalid reply gives a file descriptor of -1, opening a QFile
on it fails, and the write loop then ran against a closed file. Every write
returned -1, so the user was told "The last block was not fully written
(-1 of 1048576 bytes)" for the very first block, with nothing about the
device being busy, which is the part they can do something about.

Report what UDisks2 said about the open, and what to do about it. Report a
failed write with the reason as well, in place of the debug line that was
left where the throw had been commented out.

M  +13   -4    isoimagewriter/imagewriter.cpp

https://invent.kde.org/utilities/isoimagewriter/-/commit/72e925414b23363e2cd5def49a380f80b4b2f451

-- 
You are receiving this mail because:
You are watching all bug changes.

Reply via email to