Created attachment 1096086 [details] Close button with artifact (red arrow added to point on the artifact) Description of problem: The close button (and sometimes the maximize button) are very redrawn with artifacts (see attached screenshot for details). Version-Release number of selected component (if applicable): compiz-0.8.9-1.fc22.x86_64 ccsm-0.8.9-1.fc22.noarch emerald-0.8.9-1.fc22.x86_64 How reproducible: Always Steps to Reproduce: 1. Periodically minimize/maximize window, or normal work Actual results: After few minutes (at worst, but mostly much sooner) the close (and sometimes maximize) button is redrawn with artifacts. Expected results: No artifacts Additional info: I guess it is related to recent compiz/ccsm/emerald update, but I haven't time to prove it. I have been using compiz/emerald on different machines for years and this issue started appearing recently (few weeks ago). I am using f22, but I will also check on f23. I am reproducing it on Lenovo x240 with Intel integrated graphic card and Xfce desktop. It seems to be reproducible with all emerald themes. I also tried various ccsm settings, but without success. Filing against compiz, but it maybe problem in different component.
I tried downgrading to emerald-.8-13.fc22.x86_64.rpm, ccsm-0.8.4-16.fc22.noarch.rpm, compiz-0.8.8-29.fc22.x86_64.rpm and I will report few hours later (I cannot currently logout/restart DE).
I also reproduced the problem with: compiz-0.8.8-29.fc22.x86_64 emerald-0.8.8-13.fc22.x86_64 ccsm-0.8.4-16.fc22.noarch So it is probably not regression caused by the recent compiz rebase. Maybe it is related to the intel video driver update - I will try downgrading. But I am still not sure where the bug really is - it maybe compiz bug in the end. Any debugging hints/tips will be appreciated.
Well, after someone decided to power off old compiz git, i decided to give compiz-0.88 a new home at github. https://github.com/raveit65?tab=repositories Here i added the tons of patches from my spec file + a few improvements which i found in other git repos. This makes it easier to work on bugs and makes the spec file more simple. As you said in first comment "(few weeks ago)", it can't be related to new 0.89 versions. They are available since few days in repos. Im using nvidia on my main box, here i don't see the issue. I will test it again on my notebook with intel driver. But i'm using MATE and my expierience is that compiz + xfce isn't an ideal combination, ie. the problem that xfce workspace-switcher don't support virtual-workspaces from compiz. Well, i will install a minimal xfce version.
Btw, which emerald theme you are using? ...and if you like you can use now gwd decorator in xfce with 0.89.
i forgot, you need to set first a metacity theme for gwd. [rave@mother ~]$ gsettings get org.mate.Marco.general theme 'Blue-Submarine' Fusion-icon reads the marco theme for gwd.
Hi Wolfgang, thank you for quick response and for giving compiz new home, I really appreciate your work. IIRC I use Sky theme on legacy engine. But the problem seems to be there for other themes as well (legacy engine at least, I havent't tested whether other engines are also affected). I proved that this is not caused by compiz rebase. I am able to reproduce the problem with compiz-0.8.8 as well. This problem is probably Intel driver related. It works correctly for me with xorg-x11-drv-intel-2.99.917-6.20150211.fc22.x86_64 but the problem appeared after upgrade to: xorg-x11-drv-intel-2.99.917-6.20150211.fc22.x86_64 I am unable to reproduce the problem with Nvidia binary driver. But I am currently not sure whether the bug is in compiz or Intel driver (I will try to look deeper on it later, hopefully :). I will also check with the gwd.
(In reply to Jaroslav Škarvada from comment #6) > Hi Wolfgang, thank you for quick response and for giving compiz new home, I > really appreciate your work. > > IIRC I use Sky theme on legacy engine. But the problem seems to be there for > other themes as well (legacy engine at least, I havent't tested whether > other engines are also affected). > > I proved that this is not caused by compiz rebase. I am able to reproduce > the problem with compiz-0.8.8 as well. > > This problem is probably Intel driver related. It works correctly for me > with > > xorg-x11-drv-intel-2.99.917-6.20150211.fc22.x86_64 > > but the problem appeared after upgrade to: > > xorg-x11-drv-intel-2.99.917-6.20150211.fc22.x86_64 > > I am unable to reproduce the problem with Nvidia binary driver. > > But I am currently not sure whether the bug is in compiz or Intel driver (I > will try to look deeper on it later, hopefully :). > > I will also check with the gwd. Unaffected: xorg-x11-drv-intel-2.99.917-6.20150211.fc22.x86_64 Affected: xorg-x11-drv-intel-2.99.917-15.20150729.fc22.x86_64
I will try to bisect the driver.
It seems the following intel driver commit introduces the problem: http://cgit.freedesktop.org/xorg/driver/xf86-video-intel/commit/?id=ef24cb1b223486d7dfe57f33c28d8692f9857c98 I have been running latest intel-driver git snapshost with this commit reverted for cca. 24 hours and haven't reproduced the problem. Unfortunately it didn't help me much. I still don't know whether the problem is in the driver, cairo or compiz/emerald. I will try to prepare stripped down reproducer from the compiz/emerald code and let's see.
I installed xfce on my notebook with latest xorg-x11-drv-intel-2.99.917-15.20150729.fc22.x86_64 driver, but i couldn't reproduce the issue. [rave@satellite ~]$ inxi -G Graphics: Card-1: Intel Core Processor Integrated Graphics Controller Card-2: NVIDIA GF108M [GeForce GT 425M] Display Server: Fedora X.org 117.4 drivers: nouveau,intel (unloaded: fbdev,vesa) Resolution: 1600x900 GLX Renderer: Mesa DRI Intel Ironlake Mobile GLX Version: 2.1 Mesa 10.6.9 (git-8957b69) Intel card is in use. I recal there were redraw issues with intel driver for specific hardware. https://fedoramagazine.org/solution-graphics-issues-intel-graphics-chipsets-fedora-22/ Maybe the workaround helps. Maybe the issue is caused by the animation for minimize/restore. I use the animation plugin with that settings https://dl.dropboxusercontent.com/u/49862637/Mate-desktop/Test/animation
Thanks for info, my card: raphics: Card: Intel Haswell-ULT Integrated Graphics Controller Display Server: Fedora X.org 117.4 drivers: intel (unloaded: fbdev,vesa) Resolution: 1366x768 GLX Renderer: Mesa DRI Intel Haswell Mobile GLX Version: 3.0 Mesa 10.6.9 (git-8957b69) According to the intel driver commit, the change is in the SNA. I will check with the UXA and if not reproducible with UXA, I will reassign the bug to the intel-driver for further investigation.
It seems to work correctly in UXA with unpatched: xorg-x11-drv-intel-2.99.917-15.20150729.fc22.x86_64 It doesn't work in SNA with unpatched: xorg-x11-drv-intel-2.99.917-15.20150729.fc22.x86_64 and newer. It works correctly in SNA (even with intel-driver git snapshot 20151122) with the following commit reverted: http://cgit.freedesktop.org/xorg/driver/xf86-video-intel/commit/?id=ef24cb1b223486d7dfe57f33c28d8692f9857c98 Kernel: kernel-4.2.5-201.fc22.x86_64 Due to the look of the artifacts, I think it's due to some uninitialised buffer. But at least I think it should behave the same in UXA and SNA, thus reassigning to intel-driver for further investigation.
Upstream bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=93125
Fedora 22 changed to end-of-life (EOL) status on 2016-07-19. Fedora 22 is no longer maintained, which means that it will not receive any further security or bug fix updates. As a result we are closing this bug. If you can reproduce this bug against a currently maintained version of Fedora please feel free to reopen this bug against that version. If you are unable to reopen this bug, please file a new report against the current release. If you experience problems, please add a comment to this bug. Thank you for reporting this bug and we are sorry it could not be fixed.