atom feed16 messages in org.freebsd.freebsd-threadsSymbol versioning errors in libthr
FromSent OnAttachments
Dag-Erling SmørgravFeb 3, 2008 7:33 am 
Peter WemmFeb 3, 2008 8:24 am 
Dag-Erling SmørgravFeb 3, 2008 8:27 am 
Alexander KabaevFeb 3, 2008 8:27 am 
Kostik BelousovFeb 3, 2008 8:54 am 
Alexander KabaevFeb 3, 2008 9:11 am 
Daniel EischenFeb 3, 2008 9:14 am 
Daniel EischenFeb 3, 2008 9:14 am 
Dag-Erling SmørgravFeb 4, 2008 6:42 am.diff
Daniel EischenFeb 4, 2008 8:55 am 
Dag-Erling SmørgravFeb 4, 2008 9:04 am 
Daniel EischenFeb 4, 2008 9:10 am 
Dag-Erling SmørgravFeb 4, 2008 9:19 am 
Daniel EischenFeb 4, 2008 10:46 am 
David SchultzFeb 4, 2008 11:36 am 
Dag-Erling SmørgravFeb 5, 2008 12:41 am 
Subject:Symbol versioning errors in libthr
From:Daniel Eischen (deis@freebsd.org)
Date:Feb 4, 2008 9:10:24 am
List:org.freebsd.freebsd-threads

On Mon, 4 Feb 2008, Dag-Erling Sm?rgrav wrote:

Daniel Eischen <deis@freebsd.org> writes:

I question the bumping of FBSDprivate to 1.1. FBSDprivate is special, it is only for use in our base system, and our supported method of updating our system is buildworkd/buildkernel - so any changes to FBSDprivate should be unnoticed.

A whole lot of assumptions there. Bumping FBSDprivate costs nothing and might save our a***s down the road if one of those assumptions turn out to be false.

If there is a need for it, it can be bumped. There is no need as yet. The only need for bumping private is for an ABI change, and if that ABI change cannot be picked up in a buildworld. This is what FBSDprivate was designed for.

(unless you can show that it's actually harmful in some way?)

Please do not bump private, it was never meant to be bumped like this.

-- DE