[Webkit-unassigned] [Bug 267261] New: Regression: Crashes through a non-local X11 connection
bugzilla-daemon at webkit.org
bugzilla-daemon at webkit.org
Tue Jan 9 00:01:50 PST 2024
https://bugs.webkit.org/show_bug.cgi?id=267261
Bug ID: 267261
Summary: Regression: Crashes through a non-local X11 connection
Product: WebKit
Version: Other
Hardware: Unspecified
OS: Linux
Status: NEW
Severity: Normal
Priority: P2
Component: WebKitGTK
Assignee: webkit-unassigned at lists.webkit.org
Reporter: m at tfiu.de
CC: bugs-noreply at webkitgtk.org
Somewhere between 2.40.3 and 2.42.4, webkit stopped working via non-local X11. Actually (and that might be relevant), it was sort-of broken before, where it required WEBKIT_DISABLE_COMPOSITING_MODE=1 to work. I have set up a test system where the 4.0 API corresponds to webkit 2.42.4 and the 4.1 API corresponds to 2.40.3 (but the behaviour is independent of the API version, it just depends on the webkit version).
With this, consider:
# 2.40.3 without env hack
$ ssh -X localhost /usr/lib/i386-linux-gnu/webkit2gtk-4.1/MiniBrowser
libEGL warning: DRI2: failed to authenticate
(WebKitWebProcess:2590): Gdk-WARNING **: 08:19:55.350: The program 'WebKitWebProcess' received an X Window System error.
This probably reflects a bug in the program.
The error was 'BadRequest (invalid request code or no such operation)'.
(Details: serial 169 error_code 1 request_code 155 (unknown) minor_code 1)
9[...]
** (MiniBrowser:2565): WARNING **: 08:19:55.354: WebProcess CRASHED
# 2.40.3 with env hack
$ ssh -X localhost "env WEBKIT_DISABLE_COMPOSITING_MODE=1 /usr/lib/i386-linux-gnu/webkit2gtk-4.1/MiniBrowser"
(works)
# 2.42.4
$ ssh -X localhost "env WEBKIT_DISABLE_COMPOSITING_MODE=1 /usr/lib/i386-linux-gnu/webkit2gtk-4.0/MiniBrowser"
libEGL warning: DRI2: failed to authenticate
libEGL warning: failed to open /dev/dri/renderD128: Permission denied
(MiniBrowser:2912): Gdk-WARNING **: 08:25:17.975: The program 'MiniBrowser' received an X Window System error.
This probably reflects a bug in the program.
The error was 'BadRequest (invalid request code or no such operation)'.
(Details: serial 158 error_code 1 request_code 155 (unknown) minor_code 1)
[...]
... and the window goes down. The behaviour and the request code are identical without disabling the compositing mode, except the serial reproducibly is 149 then on my box (the 158 is reproducible, too).
The crashes occur in the backend process, so I'm asking for forgiveness for not providing a traceback (I keep forgetting how to do that, but I'll look it up if the crash isn't immediately reproducible).
[ancillary information in case this doesn't immediately happen elsewhere: this is a Debian bookworm box with an i386 userland and Haswell i915 graphics]
--
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/20240109/78b3579e/attachment-0001.htm>
More information about the webkit-unassigned
mailing list