[Webkit-unassigned] [Bug 233997] New: Connecting to context menu signal leads to segfault

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Wed Dec 8 02:25:19 PST 2021


https://bugs.webkit.org/show_bug.cgi?id=233997

            Bug ID: 233997
           Summary: Connecting to context menu signal leads to segfault
           Product: WebKit
           Version: Other
          Hardware: PC
                OS: Linux
            Status: NEW
          Severity: Normal
          Priority: P2
         Component: Platform
          Assignee: webkit-unassigned at lists.webkit.org
          Reporter: EvilTwin1 at gmx.de

I'm using the rust bindings to the Gtk4 version of webkit2gtk. I used them with both a local build of webkit as well as with the version in the org.gnome.Sdk flatpak runtime.
Connecting to the context-menu signal leads to a segmentation fault.

> GLib-GObject-WARNING **: 08:53:00.267: value_to_ffi_type: Unsupported fundamental type: GdkEvent Segmentation fault (core dumped)

Asking about it in the "Rust <3 Gnome" Matrix channel it was explained to me that this is an issue with webkit itself:

> slomo:
> value_to_ffi_type() is the libffi signal marshaller thingy from glib, and that can't handle custom fundamental types (i mean, that's what it says there �� )
> the Rust code is probably correct but webkitgtk will have to use a customer marshaller for that signal
> it needs to do things like here https://gitlab.gnome.org/GNOME/gtk/-/blob/master/gdk/gdksurface.c#L647-660
> important in that context is also the type annotation here https://gitlab.gnome.org/GNOME/gtk/-/blob/master/gdk/gdksurface.c#L641


> Emmanuele Bassi:
> Correct. Events are instance types, and there’s no marshaller for those because GType is stupid
> You need to provide a marshaller yourself
> Though I’d recommend not passing around event instances to begin with

I don't feel confident to prepare a patch to webkit myself. But I hope this report provides enough information.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20211208/55577573/attachment.htm>


More information about the webkit-unassigned mailing list