[Webkit-unassigned] [Bug 243401] [GTK] webkit_web_context_get_default() crashes in Eclipse since webkit-gtk v2.36.5, v2.36.4 was fine
bugzilla-daemon at webkit.org
bugzilla-daemon at webkit.org
Mon Aug 1 08:12:34 PDT 2022
https://bugs.webkit.org/show_bug.cgi?id=243401
--- Comment #3 from Michael Haubenwallner <michael.haubenwallner at ssi-schaefer.com> ---
Agreed, such a change should not be shipped in a minor version.
Further thought: To allow for applications to recognize as early as possible, does initializeMainThread() assert for Thread 1 as well?
But then: Right now, Eclipse calls initializeMainThread() in Thread 2 here, and I would really be surprised if Eclipse ever has done that in Thread 1.
So while I understand/agree to use GTK/WebKit/SWT within one single thread only, I don't get the reason for why this has to be Thread 1.
Also, we do have some SWT/GTK application here in production for years, and the gdb-breakpoint in gtk_main_iteration_do() is fired in Thread 51 this time.
So yes, I would say GTK in no way does require the UI thread to be Thread 1.
(gdb) info threads
...
* 51 Thread 0x7fdb13bff640 (LWP 549155) "Thread-15" gtk_main_iteration_do (blocking=0) at ../gtk/gtkmain.c:1455
...
(gdb) frame
#0 gtk_main_iteration_do (blocking=0) at ../gtk/gtkmain.c:1455
1455 {
(gdb) bt
#0 gtk_main_iteration_do (blocking=0) at ../gtk/gtkmain.c:1455
#1 0x00007fdb13256dbc in Java_org_eclipse_swt_internal_gtk3_GTK3_gtk_1main_1iteration_1do ()
at /home/haubi/Workspaces/5_15/pd/WAMAS-5_15/.metadata/.plugins/org.eclipse.pde.core/WAMAS5_Mobile/org.eclipse.osgi/203/0/.cp/libswt-pi3-gtk-4948r9.so
#2 0x00007fdc23a154e3 in ()
#3 0x00007fdb13bfe908 in ()
#4 0x00007fdc23a15064 in ()
#5 0x00007fdb13bfe850 in ()
#6 0x00007fdb12b58700 in ()
#7 0x00007fdb13bfe8b8 in ()
#8 0x00007fdb12b604e0 in ()
#9 0x0000000000000000 in ()
(gdb)
--
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/20220801/e96a03cf/attachment.htm>
More information about the webkit-unassigned
mailing list