Bug 62263

Summary:	MAKEOPTS is changed by some ebuild, and its value is changed even in "emerge" process
Product:	Portage Development	Reporter:	Lluís Batlle i Rossell <viric>
Component:	Core - Interface (emerge)	Assignee:	Portage team <dev-portage>
Status:	VERIFIED NEEDINFO
Severity:	normal
Priority:	High
Version:	unspecified
Hardware:	All
OS:	Linux
Whiteboard:
Package list:		Runtime testing required:	---

Description Lluís Batlle i Rossell 2004-08-30 08:59:00 UTC

I set the MAKEOPTS env variable through environment, not using make.conf.
And I'm emerging LOTS of programs (in fact, I'm using catalyst).

the script "emerge" calls 'ebuild.sh' (I saw the lines in portage.py file) for merging each file. So, I think that "ebuild.sh" processes should ALWAYS have the "MAKEOPTS" env variable set as it was when "emerge" was called.

The only function I've found which changes MAKEOPTS is "get_number_of_jobs" in 'eutils.eclass' file, called by 'ebuild.sh', called by 'emerge'.

I added an echo line at start of 'ebuild.sh' which shows the content of MAKEOPTS variable. And BECAUSE OF SOMETHING I don't know, the contents of MAKEOPTS is "-j2" (the contents of MAKEOPTS variable before running emerge was "-j6").
ebuild.sh inherits that variable from the python script 'emerge', isn't it? and ebuild.sh cannot change the content of that variable of the 'emerge' process. So, what happened?

I cannot catch when the calls to 'ebuild.sh' begin to use -j2 and not my set -j6.
I added "MAKEOPTS=-j6" at start of ebuild.sh as a temporary solution.

Am I completely wrong in something? Am I missing something? Is the explanation clear enough?

Reproducible: Always
Steps to Reproduce:
Emerge a lot of packages, as catalyst does with grp building.

Comment 1 SpanKY gentoo-dev

2004-08-30 19:40:57 UTC

please narrow down how you reproduce this

some ebuilds actually override MAKEOPTS themselves (there are bugs open to handle that)

i'm not even sure what your bug is here ... are you saying that if one ebuild changes MAKEOPTS, all the other packages in the same emerge process are affected ?

what does catalyst have to do with this ?

Comment 2 Lluís Batlle i Rossell 2004-08-31 02:46:00 UTC

Oh... I found the solution to this. That was a catalyst bug.
Sorry. :(

Forget this.