boundarydevices / meta-boundary Goto Github PK

It appears that the zeus branch was recently re-written. All commits since Oct 15 2019 have been updated.

Previously:

commit 13eade8 (m/zeus)
Author: Chris Dimich [email protected]
AuthorDate: Fri Feb 21 14:51:00 2020 -0800
Commit: Chris Dimich [email protected]
CommitDate: Fri Feb 21 14:51:00 2020 -0800

Remove xserver-xorg patch. This is apart of meta-freescale

... snip many commits...

commit 3360fe0
Author: Chris Dimich [email protected]
AuthorDate: Tue Oct 15 15:49:26 2019 -0700
Commit: Chris Dimich [email protected]
CommitDate: Tue Oct 15 15:49:26 2019 -0700

Initial commit for warrior branch

Now:

commit 03877f0 (HEAD, origin/zeus)
Author: Chris Dimich [email protected]
AuthorDate: Fri Mar 20 21:02:33 2020 -0700
Commit: Chris Dimich [email protected]
CommitDate: Fri Mar 20 21:02:33 2020 -0700

imx-mkimage: Fix recipe name syntax

Signed-off-by: Chris Dimich <[email protected]>

... snip many commits...

commit f4e096c
Author: Chris Dimich [email protected]
AuthorDate: Tue Oct 15 15:49:26 2019 -0700
Commit: Chris Dimich [email protected]
CommitDate: Fri Mar 13 16:22:40 2020 -0700

Initial commit for warrior branch

Note commit date of all commits since "Initial commit for warrior branch". Originally these commits were staggered since Oct 15. Now they all appear to have been committed on March 13th.

The master branch had been behind the zeus branch for a while and it looks like it was recently updated. Perhaps zeus was accidentally force-pushed to at this time?

Consider restoring the original state of the zeus branch--it will make this repository easier to work with.

Hi,

Is it possible to create a Thud branch (same as master)? In order to be more consistent during an integration.

For example, when integrating with Mender[1].

Thanks

/Pierre-Jean

[1] bdx-iot/meta-mender-community@48491f9#diff-889282f518cb02678a1ecfa1c8d58c9cR8

Why is the licensing listed as proprietary but publicly available on GitHub?

Every time I launch Chromium, I see an error like this:
ERROR:sandbox_linux.css(301) InitializeSandbox() called with multiple threads in process gpu-process.

Despite this error, Chrome launches fine. However, when visiting certain webpages (e.g. http://www.html5test.com), it crashes with a the following error:

line 11: 1105 Segmentation fault /usr/bin/chromium/chrome ${CHROME_EXTRA_ARGS} $@

Wonder if anyone has seen this and resolve this before?? I will capture a core dump to get more information.

If pam is enabled in DISTRO_FEATURES, dropbear refuses any password authentication.

The problem is apparently conflict between two files named creatively dropbear:
sources/poky/meta/recipes-core/dropbear/dropbear/dropbear - PAM config, goes to /etc/pam.d/
versus
sources/meta-boundary/recipes-core/dropbear/boundary/dropbear - goes to /etc/default/

File sources/meta-boundary/recipes-core/dropbear/boundary/dropbear overwrites PAM config during bitbake build.

I've built sato image with this layer. However, when I play a video in Chromium, it crashes. It also crashes if I attempt to visit http://html5test.com

What are the steps to follow to build Chromium with HTML5 video support?

In boundary-wayland.conf you comment there's no gstreamer-imx for iMX8 yet, but there actually is one in development.

It is being developed on the V2 branch. I'm not familiar with how Yocto builds work, but I would be interested in learning how to add this package to my board so I can take advantage of the imxg2dcompositor.

Currently on the iMX8 with the standard compositor it's horribly slow (I assume because it's using the CPU).