Overview
Request 1030154 superseded
- Update to version 22.1.4
* xwayland: Aggregate scroll axis events to fix kinetic scrolling
* Forbid server grabs by non-WM on *rootless* XWayland
* xkb: Avoid length-check failure on empty strings.
* ci: remove redundant slash in libxcvt repository url
* dix: Skip more code in SetRootClip for ROOT_CLIP_INPUT_ONLY
* dix: Fix overzealous caching of ResourceClientBits()
* xwayland: Prevent Xserver grabs with rootless
* xwayland: Delay wl_surface destruction
* build: Bump wayland requirement to 1.18
* xwayland: set tag on our surfaces
* xwayland: Clear the "xwl-window" tag on unrealize
* xwayland: correct the type for the discrete scroll events
* xkb: fix some possible memleaks in XkbGetKbdByName
* xkb: length-check XkbGetKbdByName before accessing the fields
* xkb: length-check XkbListComponents before accessing the fields
* xkb: proof GetCountedString against request length attacks
- supersedes security patches:
* U_xkb-fix-some-possible-memleaks-in-XkbGetKbdByName.patch
* U_xkb-proof-GetCountedString-against-request-length-at.patch
- Created by sndirsch
- In state superseded
- Superseded by 1030894
- Open review for openSUSE:Factory:Staging:F
Request History
sndirsch created request
- Update to version 22.1.4
* xwayland: Aggregate scroll axis events to fix kinetic scrolling
* Forbid server grabs by non-WM on *rootless* XWayland
* xkb: Avoid length-check failure on empty strings.
* ci: remove redundant slash in libxcvt repository url
* dix: Skip more code in SetRootClip for ROOT_CLIP_INPUT_ONLY
* dix: Fix overzealous caching of ResourceClientBits()
* xwayland: Prevent Xserver grabs with rootless
* xwayland: Delay wl_surface destruction
* build: Bump wayland requirement to 1.18
* xwayland: set tag on our surfaces
* xwayland: Clear the "xwl-window" tag on unrealize
* xwayland: correct the type for the discrete scroll events
* xkb: fix some possible memleaks in XkbGetKbdByName
* xkb: length-check XkbGetKbdByName before accessing the fields
* xkb: length-check XkbListComponents before accessing the fields
* xkb: proof GetCountedString against request length attacks
- supersedes security patches:
* U_xkb-fix-some-possible-memleaks-in-XkbGetKbdByName.patch
* U_xkb-proof-GetCountedString-against-request-length-at.patch
dimstar_suse set openSUSE:Factory:Staging:D as a staging project
Being evaluated by staging project "openSUSE:Factory:Staging:D"
dimstar_suse accepted review
Picked "openSUSE:Factory:Staging:D"
factory-auto added opensuse-review-team as a reviewer
Please review sources
factory-auto accepted review
Check script succeeded
licensedigger accepted review
ok
dimstar accepted review
dimstar_suse added factory-staging as a reviewer
Being evaluated by group "factory-staging"
dimstar_suse accepted review
Unstaged from project "openSUSE:Factory:Staging:D"
dimstar_suse set openSUSE:Factory:Staging:M as a staging project
Being evaluated by staging project "openSUSE:Factory:Staging:M"
dimstar_suse accepted review
Picked "openSUSE:Factory:Staging:M"
dimstar_suse added factory-staging as a reviewer
Being evaluated by group "factory-staging"
dimstar_suse accepted review
Unstaged from project "openSUSE:Factory:Staging:M"
dimstar_suse set openSUSE:Factory:Staging:F as a staging project
Being evaluated by staging project "openSUSE:Factory:Staging:F"
dimstar_suse accepted review
Picked "openSUSE:Factory:Staging:F"
superseded by 1030894
https://openqa.opensuse.org/tests/2827176#step/sshxterm/5
This error happened repeatedly with this version of xwayland - and even moved from one staging to another when xwayland was moved (currently there is no other package in this staging, making it clearly an xwayland issue)
I don't understand this test. Sorry! I even don't know how openQA is testing Xwayland. I doubt that openQA is making any tests yet with Wayland, let alone with Xwayland running on top of Wayland for running X apps.
Actually, GNOME runs as Wayland compositor, xwayland usautostarted for xterm here.
This is confirmed in e.g https://openqa.opensuse.org/tests/2827176#step/x_vt/3 where we see xwayland running
Ok. but I still don't understand the test.
It's more an 'accident' that it shows an issue with xwayland...
Essentially it runs just xterm, and a 2nd xterm over ssh -X - and there we get nonsense on the keyboard buffer
openQA uses
Ctrl-D
to close xterm and the next xterm gets a stuckd
key. So I guess the key release event did somehow not get processed correctly.For reference, the test description:
I can reproduce this now outside from openQA. Bisecting now ...
Awesome news!
SR#1030894 includes the fix. :-)