diff options
author | Tadao Tanikawa <tanikawa.tadao@jp.panasonic.com> | 2017-01-03 01:33:11 +0900 |
---|---|---|
committer | Jan-Simon Moeller <jsmoeller@linuxfoundation.org> | 2017-01-04 09:39:13 +0000 |
commit | b708c9814ecfdb9209b2d183350141c3fa2ac635 (patch) | |
tree | 7f67c5e06d06b5b1ef801fd6888e364bd0d51a42 /recipes-qt/qt5/qtbase_%.bbappend | |
parent | 87f999160a602748a247ed1237634b24870d05b3 (diff) |
Revert "Change QWindow::visible when surface visibility in ivi shell changes"
This reverts commit d0f95885030e5f425c0423a7a29414a204aaa341.
This is root cause of issue that screen doesn't change
from blank screen after booting.
ivi-surface's visibility must not link to QWindow's visibility
because changing it's visibility to 0 means destroying wayland
surface (--> eglsurface --> ivi-surface) immmediatly.
On the other hand, ivi-surface is created with
visibility == 0. Therefore, creating ivi-surface triggers
self destruction of ivi-surface itself immediately.
And to make matters worse, Qt5.7's hiding(visibility == 0)
window method, wl.attach(nil, 0, 0) was wrong, it cause
unpredictable operation of weston (In our case, it drops some
events between WindowManager and Weston).
For linking ivi-surface's visibility and visibility in QML,
we should consider another method. At least, commit d0f958's
approach is incorrect in Qt5.7 so far.
Change-Id: Ibe90f9dae634aa5cdb3c166caf4db02dfd71965b
Signed-off-by: Tadao Tanikawa <tanikawa.tadao@jp.panasonic.com>
Diffstat (limited to 'recipes-qt/qt5/qtbase_%.bbappend')
0 files changed, 0 insertions, 0 deletions