Summary: | sys-devel/llvm-3.9.1-r1: ninja-utils.eclass nonfatal is breaking DistCC. | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Simon-Pierre Dubé <shaedge> |
Component: | Eclasses | Assignee: | LLVM support project <llvm> |
Status: | RESOLVED INVALID | ||
Severity: | normal | CC: | jstein, mgorny, shaedge |
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
URL: | https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=5b38fedf | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: | Build log |
Description
Simon-Pierre Dubé
2017-05-31 19:29:16 UTC
Created attachment 474878 [details]
Build log
Nonfatal is completely irrelevant here. It can't find valgrind headers. I'd blame distcc-pump. It's half-working hack that breaks stuff randomly. Don't use it. (In reply to Michał Górny from comment #2) > Nonfatal is completely irrelevant here. It can't find valgrind headers. I'd > blame distcc-pump. It's half-working hack that breaks stuff randomly. Don't > use it. Thank you for your reply! I'll use it as less as possible now. And just to be clear, it usually gives little performance gain. I've seen it break Firefox in the past, and that considered, I wouldn't trust any executable produced by distcc-pump -- if it can pass incorrect includes, it can produce a broken object. (In reply to Michał Górny from comment #4) > And just to be clear, it usually gives little performance gain. I've seen it > break Firefox in the past, and that considered, I wouldn't trust any > executable produced by distcc-pump -- if it can pass incorrect includes, it > can produce a broken object. That's good to know, especially since i use a binhost to update my VMs and same architecture less powerful machines. |