Bug 8397 - ROOTLESS_X should be used as default for Poky BSP
Summary: ROOTLESS_X should be used as default for Poky BSP
Status: RESOLVED NOTABUG
Alias: None
Product: BSPs
Classification: Build System, Metadata & Runtime
Component: bsps-configuration (show other bugs)
Version: unspecified
Hardware: x86 Multiple
: Medium normal
Target Milestone: 2.2 M1
Assignee: Alejandro Hernandez
QA Contact:
URL:
Whiteboard:
Depends on:
Blocks:
 
Reported: 2015-09-25 10:35 UTC by Jussi Kukkonen
Modified: 2016-06-09 15:44 UTC (History)
6 users (show)

See Also:
OS type for building Yocto: ---
Type of Regression: Regression (Used to work)
Verified:
Documentation change: Don't know


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Jussi Kukkonen 2015-09-25 10:35:19 UTC
Filing this based on chat discussion with Ross, Richard.

Currently no meta-intel bsp uses ROOTLESS_X to run X as "xuser" instead of root. This should be done where possible (and obviously bugs preventing it should be fixed).
Comment 1 Saul Wold 2015-10-01 21:23:45 UTC
Currently we don't enable ROOTLESS_X by default in Poky (the Yocto Project Distro) at all, so why should we make a policy decision to enable ROOTLESS_X for meta-intel specifically, this would mean different BSPs have different distro policies. 

The enabling of ROOTLESS_X should be done at a Distro Level not BSP level.

Are there specific issues that need to be addressed?  I just completed a ROOTLESS_X = "1" build and booted with no issues.

I guess I would like to understand the background of the discussion that was had with Ross and Richard.
Comment 2 Jussi Kukkonen 2015-10-02 11:51:57 UTC
(In reply to comment #1)
> Are there specific issues that need to be addressed?  I just completed a
> ROOTLESS_X = "1" build and booted with no issues.

If we don't test it then it will be broken. If not now, then later. The last use of ROOTLESS_X in meta-intel I can find was in fact removed because it broke X.

> I guess I would like to understand the background of the discussion that was
> had with Ross and Richard.

This was a short discussion on connman that was changing the build based on ROOTLESS_X quite needlessly: as an offshoot of that it was discovered that nothing actually tests ROOTLESS_X anymore. This was seen as Bad Thing so I promised to file a bug so we don't forget.

The impression that it should be a bsp-specific decision was mine and could very well be wrong: I just knew that's how it was done before.
Comment 3 Alejandro Hernandez 2016-06-09 15:44:34 UTC
(In reply to comment #1)
> Currently we don't enable ROOTLESS_X by default in Poky (the Yocto Project
> Distro) at all, so why should we make a policy decision to enable ROOTLESS_X
> for meta-intel specifically, this would mean different BSPs have different
> distro policies. 
> 
> The enabling of ROOTLESS_X should be done at a Distro Level not BSP level.
> 
> Are there specific issues that need to be addressed?  I just completed a
> ROOTLESS_X = "1" build and booted with no issues.
> 
> I guess I would like to understand the background of the discussion that was
> had with Ross and Richard.

After some discussion, it was decided that this should be done at a Distro level, as Saul was stating, so I am closing this.