Re: RPM %post dying - pthread, fork, execve

[Paul Nasrat <pnasrat redhat com>]

On Tue, Aug 10, 2004 at 08:05:26PM -0400, James Olin Oden wrote:
> On Tue, 10 Aug 2004, Paul Nasrat wrote:
> 
> > On Tue, Aug 10, 2004 at 10:55:03PM +0300, Panu Matilainen wrote:
> > > On Tue, 2004-08-10 at 22:44, David D. Hagood wrote:
> > > >   > Just guessing but are you calling rpmtsSetRootDir() before running the
> > > > > transaction? Even if the rootdir is set to NULL you have to initialize
> > > > > it on rpm >= 4.1, otherwise all scriptlets will fail.
> > > > 
> > > > That was it - thanks!
> > > 
> > > You're not the first and probably not the last either (been there... :)
> > > to miss that bit.
> > > 
> > > Jeff & co - how about a little comment in the sources about the
> > > necessity of calling rpmtsSetRootDir() before rpmtsRun() ?
> > 
> > Probably best as doxygen, if someone feels like knocking up a patch then that's great, else I'll try and look at tomorrow...
> >
> Paul I can do that if you have not already done it.  It will make feel 
> like I am contributing (-;  Been a bit busy and away from rpm these days.

Go for it.

Paul