499202 – Sandbox violations when XDG_CONFIG_HOME is set in the calling environment

Bug 499202 - Sandbox violations when XDG_CONFIG_HOME is set in the calling environment

Summary: Sandbox violations when XDG_CONFIG_HOME is set in the calling environment

Status:	RESOLVED FIXED

Alias:	None

Product:	Gentoo Linux
Classification:	Unclassified
Component:	Current packages (show other bugs)
Hardware:	All Linux

Importance:	Normal normal (vote)
Assignee:	Python Gentoo Team

URL:
Whiteboard:
Keywords:

Depends on:	481822 488378 499068 499288
Blocks:
	Show dependency tree

Reported:	2014-01-25 09:57 UTC by Justin Lecher (RETIRED)
Modified:	2018-04-24 15:41 UTC (History)
CC List:	2 users (show)

See Also:
Package list:
Runtime testing required:	---

Attachments
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.

Description Justin Lecher (RETIRED) gentoo-dev

2014-01-25 09:57:03 UTC

As seen in many bugs the presents of ~/.config/matplotlib for the user you are compiling the package with, causes sandbox violations as soon as matplotlib comes into play.

This can be worked around by setting XDG_CONFIG_HOME to e.g. $T

How about proactivaly setting this in the python*r1 eclasses instead of waiting for a user filing a bug when hitting this.

Additionally we could also set MPLCONFIGDIR to avoid checking in users directory.

Comment 1 Mike Gilbert gentoo-dev

2014-01-25 16:13:58 UTC

I sent a note to gentoo-dev asking for advice.

Comment 2 Mike Gilbert gentoo-dev

2014-01-26 00:27:36 UTC

Ideally this would be dealt with by Portage, but it looks like that's going to get caught up in the usual EAPI lag time.

Comment 3 Bignaux Ronan 2015-02-06 15:28:37 UTC

same problem with mpv

Comment 4 Pacho Ramos gentoo-dev

2016-02-08 16:03:09 UTC

xdg.eclass was added by Gilles some weeks ago for this kind of issue (and yes, I would also like to see this done at PM level for not needing to keep adding xdg.eclass to random ebuilds when bugs are reported but, until that is done...)

Comment 5 Michał Górny archtester

2018-04-24 15:41:37 UTC

This looks tracker-ish, and it seems that all tracked bugs have been fixed.  The only remaining point is to have ENV_UNSET from EAPI 7 implemented and in use.