Skip to content

Conversation

@TurboGit
Copy link
Member

We setup the tabs using GROUP_SLIDERS(). The signal switch_page is raised calling _channel_tabs_switch_callback. In this callback we are accessing gui_date (g->param_size, etc) but this widget is setup after calling GROUP_SLIDERS() which crashes.

Connect to the signal at the end of gui_init() when all widgets are created.

We setup the tabs using GROUP_SLIDERS(). The signal switch_page
is raised calling _channel_tabs_switch_callback. In this callback
we are accessing gui_date (g->param_size, etc) but this widget
is setup after calling GROUP_SLIDERS() which crashes.

Connect to the signal at the end of gui_init() when all widgets
are created.
@TurboGit TurboGit added this to the 5.4.1 milestone Jan 19, 2026
@TurboGit TurboGit added bugfix pull request fixing a bug priority: high core features are broken and not usable at all, software crashes scope: image processing correcting pixels release notes: pending labels Jan 19, 2026
@TurboGit
Copy link
Member Author

For unknown reason Darktable started to crashes every ten minutes while editing. After this patch all is back to normal on my side.

@TurboGit TurboGit merged commit 1492ecd into master Jan 19, 2026
5 of 6 checks passed
@TurboGit TurboGit deleted the po/fix-coloreq branch January 19, 2026 18:31
@dterrahe
Copy link
Member

Are you sure this really solves a problem? darktable.gui->reset is TRUE during gui_init and in that case the notebook callback immediately returns. So this PR should not make a difference (except for saving one signal handler call). It would also be surprising if a crash triggered in gui_init didn't happen deterministically at startup 100%.

So my suspicion is that if you had a crash before, you still have it...

@TurboGit
Copy link
Member Author

I don't have the backtrace anymore, but it was clearly from gui_init() to a null widget in the callback. I'll watch if it happens again, but up to now I don't have the issue.

@TurboGit
Copy link
Member Author

But looking at the code you're right (of course), so no explanation for now. Just for the record this was not when starting Darktable but it was after doing some editing (Agx, Local Contrast, Exposure...) and switching to the next image. Also I was using the synch of the last edit amongst selected image on the filmstrip. Maybe one of those have a memory corruption...

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

bugfix pull request fixing a bug priority: high core features are broken and not usable at all, software crashes scope: image processing correcting pixels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants