Bug 108624 - [GTK] REGRESSION(r141319): Plugins never load and UI process hangs
Summary: [GTK] REGRESSION(r141319): Plugins never load and UI process hangs
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: WebKit2 (show other bugs)
Version: 528+ (Nightly build)
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Nobody
URL:
Keywords: Gtk
Depends on:
Blocks:
 
Reported: 2013-02-01 05:47 PST by Carlos Garcia Campos
Modified: 2013-02-01 06:20 PST (History)
3 users (show)

See Also:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Carlos Garcia Campos 2013-02-01 05:47:14 PST
After r141319 plugins never load, I think it's because the plugin process crashes and g_spawn_sync never returns making the UI process hang. I've confirmed that reverting the patch makes the plugins load and work again. I have no idea about those signals, so I don't know why it's failing or how to fix it.
Comment 1 Chris Dumez 2013-02-01 05:50:42 PST
Sorry about that.

It does not seem to affect the layout tests, does it?
Comment 2 Carlos Garcia Campos 2013-02-01 05:54:35 PST
I don't know if plugins tests are enabled in wk2, tbh. I can reproduce the issue just by running MiniBrowser, any page loading a plugin makes the UI process hang. It should affect the layout tests, because it happens when plugins are loaded in the UI process, the plugin process is launched with the -scanPlugin option for the first plugin found and it never returns. I have no idea why it crashes, though. Running it manually or without the patch works.
Comment 3 Chris Dumez 2013-02-01 05:59:26 PST
(In reply to comment #2)
> I don't know if plugins tests are enabled in wk2, tbh. I can reproduce the issue just by running MiniBrowser, any page loading a plugin makes the UI process hang. It should affect the layout tests, because it happens when plugins are loaded in the UI process, the plugin process is launched with the -scanPlugin option for the first plugin found and it never returns. I have no idea why it crashes, though. Running it manually or without the patch works.

Actually, you seem to have several plugins layout tests that time out on the GTK WK2 bot:
http://build.webkit.org/results/GTK%20Linux%2064-bit%20Release%20WK2%20(Tests)/r141461%20(4145)/results.html

Strangely, it does not seem to affect EFL port though.
Comment 4 Carlos Garcia Campos 2013-02-01 06:05:09 PST
Pugins worked for EFL before the patch that just fixes a runtime warning, right? In that case, what do you think about rolling out the patch until we figure out a solution?
Comment 5 Chris Dumez 2013-02-01 06:06:37 PST
(In reply to comment #4)
> Pugins worked for EFL before the patch that just fixes a runtime warning, right? In that case, what do you think about rolling out the patch until we figure out a solution?

Yes, I'm fine with that.
Comment 6 Carlos Garcia Campos 2013-02-01 06:07:05 PST
(In reply to comment #5)
> (In reply to comment #4)
> > Pugins worked for EFL before the patch that just fixes a runtime warning, right? In that case, what do you think about rolling out the patch until we figure out a solution?
> 
> Yes, I'm fine with that.

Thanks!
Comment 7 Chris Dumez 2013-02-01 06:20:19 PST
Should be fixed as the faulty patch was rolled out in Bug 108629.