| Summary: | Window hang and crash after move and resize ntfs partition | ||
|---|---|---|---|
| Product: | [Applications] partitionmanager | Reporter: | cdalxndr |
| Component: | general | Assignee: | Andrius Štikonas <andrius> |
| Status: | RESOLVED DUPLICATE | ||
| Severity: | crash | CC: | jyn514 |
| Priority: | NOR | Keywords: | drkonqi |
| Version: | unspecified | ||
| Target Milestone: | --- | ||
| Platform: | Ubuntu | ||
| OS: | Linux | ||
| Latest Commit: | Version Fixed In: | ||
| Sentry Crash Report: | |||
| Attachments: | New crash information added by DrKonqi | ||
|
Description
cdalxndr
2018-09-06 04:25:13 UTC
Hard to tell from crash report what happened (especially since there are no debug symbols). Whether data is alright depends only on when crash occurred. If crash occurred during resize stage, it's much more likely that your data is alright, I would still try to run fsck. If it happened during move state, it's very tricky to recover data... In principle it's still there, but there will be an empty gap in your partition. Btw, I think at the moment rollback does not happen if you click cancel. I think running job has to finish... Rollback only happens if copying data function tells that copying failed. Also, note that in the next version of KPM, GUI crashes will not affect currently running job. I.e. even if GUI crashes, the job will successfully finish in the background. It seems the operation finished successfully, I didn't found any bad files until now and windown 'chkdsk' didn't found any errors. I can make some improvement suggestions: 1) Estimate the operation duration and display a warning that requires user confirmation before starting the operation 2) When use tries to cancel the operation, show a confirmation that the current operation has to finish before cancel/rollback (In reply to cdalxndr from comment #2) > It seems the operation finished successfully, I didn't found any bad files > until now and windown 'chkdsk' didn't found any errors. > I can make some improvement suggestions: > 1) Estimate the operation duration and display a warning that requires user > confirmation before starting the operation > 2) When use tries to cancel the operation, show a confirmation that the > current operation has to finish before cancel/rollback The only long jobs are I think moving data and shrinking partition. I think moving data job already shows estimated time (in details?) and percentage. And for shrinking fs we often don't have information about progress. Some tools (maybe ext4 resizing) can provide progress but for now KPM does not parse output for progress. There is already a TODO file in partitionmanager repository with this task but maybe it's better to move it to a wishlist bug (that TODO file is a bit outdated). And probably another wishlist bug for second suggestion. Created attachment 121954 [details]
New crash information added by DrKonqi
partitionmanager (3.3.1) using Qt 5.9.5
- What I was doing when the application crashed:
I was performing a 'move and resize partition' operation using KDE partition manager.
It completed the move and then showed an error performing the resize. I tried to save the log so I could look at it later, but when I clicked 'Save Log' the application crashed.
I don't remember what the error was and since the application crashed I can't go back to look.
-- Backtrace (Reduced):
#4 0x00007f0763ac7222 in PartWidget::paintEvent(QPaintEvent*) () at /usr/lib/x86_64-linux-gnu/libkpmcore.so.7
#5 0x00007f076142e038 in QWidget::event(QEvent*) () at /usr/lib/x86_64-linux-gnu/libQt5Widgets.so.5
#6 0x00007f07613ef82c in QApplicationPrivate::notify_helper(QObject*, QEvent*) () at /usr/lib/x86_64-linux-gnu/libQt5Widgets.so.5
#7 0x00007f07613f70f4 in QApplication::notify(QObject*, QEvent*) () at /usr/lib/x86_64-linux-gnu/libQt5Widgets.so.5
#8 0x00007f07606719a8 in QCoreApplication::notifyInternal2(QObject*, QEvent*) () at /usr/lib/x86_64-linux-gnu/libQt5Core.so.5
It's very likely to be fixed in kpmcore git master. If it still doesn't work in future releases (I guess 4.2.0) please reopen. *** This bug has been marked as a duplicate of bug 413418 *** |