Gentoo Websites Logo
Go to: Gentoo Home Documentation Forums Lists Bugs Planet Store Wiki Get Gentoo!

Bug 44257

Summary: MAKEOPTS ignores the -l flag
Product: Portage Development Reporter: Chris Carlin <volkris>
Component: Core - Interface (emerge)Assignee: Portage team <dev-portage>
Status: RESOLVED INVALID    
Severity: normal    
Priority: High    
Version: unspecified   
Hardware: All   
OS: Linux   
Whiteboard:
Package list:
Runtime testing required: ---

Description Chris Carlin 2004-03-10 09:03:56 UTC
I can specify -j in the MAKEOPTS, but another very useful flag is -l, which tells make not to start a new process if the load is over a certain amount. This is a fantastic option for things like distcc where the load patterns are more complex.

-l doesn't seem to be applied, though. When I run make -l2.8 manually it works as expected, but not through emerge.

Reproducible: Always
Steps to Reproduce:
1. add -l<load> to a MAKEOPTS line that already includes a -j
2. emerge something
3. try it again but this time halt the merge after configuration and run "make -j<whatever> -l<load>" in the work directory

Actual Results:  
Going through emerge make will spawn the full number of processes specified by -j.

Going manually make will stop spawning new processes when load is higher than
<load>.

Expected Results:  
They should do the same thing if -l was passed.
Comment 1 Caleb Tennis (RETIRED) gentoo-dev 2004-03-10 10:38:06 UTC
MAKEOPTS is only used in ebuilds that call the "emake" command.  Were you perhaps testing this in an ebuild that didn't use that?
Comment 2 Chris Carlin 2004-03-10 11:14:32 UTC
I tried this in four or five ebuilds and found the weird behavior. I just checked two of them, php and sablotron, and saw make, not emake, in each package. This must have been the problem.

I'm fairly certain that the -j option was still being applied, though. There were still -j number of processes in some of these packages, they just weren't being throttled.

*shrug*
Comment 3 Chris Carlin 2004-05-24 06:12:31 UTC
(this bug can probably be closed...)
Comment 4 Alec Warner (RETIRED) archtester gentoo-dev Security 2005-08-13 08:31:51 UTC
(In reply to comment #3)
> (this bug can probably be closed...)

looks good to me ;)