Bug 384585 – media-libs/libvpx sandbox violation in command 'php -v'

• Status: RESOLVED FIXED
• Product: Gentoo Linux

[Description Geoff Madden 2011-09-27 03:20:43 UTC]

Created attachment 287859 [details]
sandbox log

according to sandbox log summary php is denied access

[Comment 1 Geoff Madden 2011-09-27 03:23:02 UTC]

Created attachment 287861 [details]
build log for x86 athlon-xp

[Comment 2 Geoff Madden 2011-09-27 03:28:50 UTC]

Created attachment 287863 [details]
emerge info +(emerge pqv)

[Comment 3 Christoph Junghans (RETIRED) 2011-09-27 09:43:01 UTC]

Same problem here!

[Comment 4 Luca Barbato 2011-09-27 09:45:21 UTC]

I'm afraid the bug is in php, not libvpx...

[Comment 5 Christoph Junghans (RETIRED) 2011-09-27 09:55:20 UTC]

(In reply to comment #4)
> I'm afraid the bug is in php, not libvpx...

Your are right.

[Comment 6 Christoph Junghans (RETIRED) 2011-09-27 09:59:07 UTC]

However, there is also a small issue in libvpx as php is a doc dependency and it also fails with USE=-doc.

[Comment 7 Ole Markus With (RETIRED) 2011-09-27 11:10:00 UTC]

Anything using php during install must addpredict that file. It is a general issue with snmp. More details can be found in bug 324739.

[Comment 8 Christoph Junghans (RETIRED) 2011-09-27 12:30:25 UTC]

If the media-video herd gives its okay I will add the line
"addpredict /usr/share/snmp/mibs/.index" to src_install!

[Comment 9 Luca Barbato 2011-09-27 12:40:50 UTC]

fine for me.

[Comment 10 Daniel Savard 2013-04-16 21:22:36 UTC]

This bug is recurring in libvpx-1.1.0 even though there is a line in the ebuild file with the addpredict /usr/share/snmp/mibs/.index which pretend to solve exactly this bug. The comment line before the addpredict command is referring to this bug report.

So, seems the bug hasn't been resolved completely or with a workaround which isn't working for subsequent versions.

[Comment 11 Samuli Suominen (RETIRED) 2013-04-17 08:18:14 UTC]

(In reply to comment #10)
> This bug is recurring in libvpx-1.1.0 even though there is a line in the
> ebuild file with the addpredict /usr/share/snmp/mibs/.index which pretend to
> solve exactly this bug. The comment line before the addpredict command is
> referring to this bug report.
> 
> So, seems the bug hasn't been resolved completely or with a workaround which
> isn't working for subsequent versions.

Attach the complete build log here. I suspect the addpredict was added to wrong function, like should have been on top of src_configure() instead of src_install()
But that's only guessing, the log is really required if you want action to be taken

[Comment 12 Samuli Suominen (RETIRED) 2013-04-17 08:18:46 UTC]

(In reply to comment #11)
> (In reply to comment #10)
> > This bug is recurring in libvpx-1.1.0 even though there is a line in the
> > ebuild file with the addpredict /usr/share/snmp/mibs/.index which pretend to
> > solve exactly this bug. The comment line before the addpredict command is
> > referring to this bug report.
> > 
> > So, seems the bug hasn't been resolved completely or with a workaround which
> > isn't working for subsequent versions.
> 
> Attach the complete build log here. I suspect the addpredict was added to
> wrong function, like should have been on top of src_configure() instead of
> src_install()
> But that's only guessing, the log is really required if you want action to
> be taken

... even better, try moving the line yourself :)

[Comment 13 Daniel Savard 2013-04-18 14:45:02 UTC]

Here is a snippet from the ebuild. The command is attached to the proper function.

src_configure() {
	replace-flags -ggdb3 -g #402825

	unset CODECS #357487

	# let the build system decide which AS to use (it honours $AS but
	# then feeds it with yasm flags without checking...) #345161
	local a
	tc-export AS
	for a in {amd64,x86}{,-{fbsd,linux}} ; do
		use ${a} && unset AS
	done

	# build verbose by default
	MAKEOPTS="${MAKEOPTS} verbose=yes"

	# http://bugs.gentoo.org/show_bug.cgi?id=384585
	addpredict /usr/share/snmp/mibs/.index
...

[Comment 14 Mark Wright 2013-04-20 03:23:11 UTC]

Since the problem re-occurs, I raised a new bug:
https://bugs.gentoo.org/show_bug.cgi?id=466522