Summary: | media-libs/netpbm fails parallel make | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Jiri Tyr <jiri.tyr> |
Component: | New packages | Assignee: | Gentoo Graphics Project <graphics+disabled> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | drizzt, flameeyes, jimmy.rosen, kierans777, wendallc |
Priority: | High | ||
Version: | 2006.0 | ||
Hardware: | x86 | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: |
2649-netpbm-10.34.log
Build log |
Description
Jiri Tyr
2006-10-02 05:26:59 UTC
post a full log ... what you posted doesnt contain an error and i can build these things with -j on my amd64 ... Created attachment 98608 [details]
2649-netpbm-10.34.log
Full compilation log of the netpbm 10.34.
Now I have tried to reemerge media-libs/netpbm-10.34 with -j3 and it didn't failed. I don't know what was wrong before. that's the fun thing with race conditions ;) ln: creating symbolic link `importinc/wordaccess.h' to `/var/tmp/portage/netpbm-10.34/work/netpbm-10.34/lib/util/wordaccess.h': File exists I just had the same problem, with -jN where N > 1. Is there a patch for the race condition?? *** Bug 160947 has been marked as a duplicate of this bug. *** It fails also on Gentoo/FreeBSD wtf does this have to do with the release process I've forced -j1 in the ebuild (both stable and ~arch), it's better than having this package fail for three months now. that doesnt fix anything Then please fix, in the mean time at least users won't get it to die, and releng is not forced to run the build at -j1... I'm removing the snapshot bug from this one, since there is a workaround in place. probably fixed in 10.51.00 Not fixed yet… Created attachment 248131 [details]
Build log
use the right bug: 337984 |