[Slackbuilds-users] Updates - 20201107.1
majek at w7i.pl
Sat Nov 14 20:51:15 UTC 2020
I had also recently some Intel related problems with Chromium (Alien's
build) and also Opera. I didn't test Vivaldi, but it's also
I think that there were some changes in Chromium engine which broke all
these browsers on Intel cards.
Browsers started ok, but full-screen didn't work and after some use of
GL, application stopped redraw itself (it worked and responded to my
clicks, but nothing changed on screen). I tried kernel upgrade (nothing
changed), Mesa upgrade (Mesa 13.0.6 can be built on stock 14.2) - i have
a feeling that this made things little better. Of course I tried
disabling hardware acceleration in browser and it helped a
lot, but in long term it didn't fixed issue.
What fixed in my setup was creating configuration for Intel card for X11
(maybe together with new Mesa, but I didn't try to downgrade it) in
/etc/X11/xorg.conf.d/20-intel.conf to force enabling DRI 3:
Identifier "Intel Graphics"
Option "DRI" "3"
"If you want something done...do yourself!"
Jean-Baptiste Emmanuel Zorg
On Tue, 10 Nov 2020, Spycrowsoft wrote:
> I can confirm that I had a similar experience with Vivaldi.
> Except in my case, the profile folder had been mutilated to such an extent by
> the new version that even the previous version of Vivaldi would just crash
> and burn when I tried to start it.
> Starting over with a clean profile folder and hardware acceleration disabled
> was easy because I'm using the sync feature.
> I've tried re-enabling hardware acceleration afterwards, which lead to the
> same result.
> I've also been running the latest versions of Vivaldi on SBo with hardware
> acceleration disabled on Slackware 14.2 without any issues since then.
> For what it's worth: I've seen the same thing happening on a system with
> NVIDIA Optimus disabled, and a system which only has an Intel embedded
> graphics card. So it seems to be some kind of Intel-related issue.
> Given that this is a pretty common setup, it might be feasible to make a note
> of this in the README file, or even to take precautions in the
> slackbuild-script itself.
> However, given that the general philosophy of Slackware is to assume that the
> user is smart and that other hardware configurations are not guaranteed to
> fail, such precautions should be optional.
> Therefore I think that a note in the README and a variable that can
> optionally be set to apply your fix directly from the slackbuild with
> something like "DEFAULT_DISABLE_GPU_ACCELERATION=yes ./vivaldi.Slackbuild",
> might be the most elegant solution here.
> Kind regards,
> Op 10-11-2020 om 01:55 schreef Luiz Carlos Ramos via SlackBuilds-users:
>> just posting about a thing that happened here with vivaldi (the
>> When it was first upgraded to version 3.4, it became broken due to some
>> anoying problems with the GUI (windows seemed to be opened but didn't
>> react to mouse clicks; minimizing the window seemed to work but the
>> screen didn't change, etc.).
>> At that time I went back to the version 2022.47 (from the end of
>> Setember and the last one from 3.3 series) and it ran flawlessly again.
>> This happened again for two or three times, for each slackbuild version
>> posted so far in the 3.4 series.
>> Last week (in another frustating upgrade) I tried to investigate a
>> little bit more about this and ultimately found an answer in one post on
>> this thread:
>> which is to run vivaldi from the CLI this way:
>> $ vivaldi --disable-gpu-compositing
>> another way to solve the problem is to disable GPU acceleration in the
>> settings page (but to do this one must bring to the "settings" screen,
>> which seems to be impossible when the problem happens).
>> Here I managed to change /usr/bin/vivaldi inserting that option
>> hardcoded in the shell script which starts the real vivaldi binary. This
>> way I can run vivaldi from the GUI as well.
>> The issue seems to be related to the specifics of the GPU used and I
>> would not risk talking about it as it seems to require a deep knowledge of
>> the inner bits.
>> A curiosity of mine is if anybody else has experienced such problem.
>> Anyway (and supposing this isn't a problem specific of myself), a
>> suggestion would be writing some notes in the README file about it. I
>> would volunteer doing so if there are no objections.
>> Many thanks,
>> Luiz Carlos Ramos
>> Sao Paulo - Brazil
>> lramos dot prof at yahoo dot com dot br
More information about the SlackBuilds-users