ports/127889: portupgrade detects spurious failures and skips ports unnecessarily

Nate Eldredge neldredge at math.ucsd.edu
Sun Oct 5 23:20:04 UTC 2008


>Number:         127889
>Category:       ports
>Synopsis:       portupgrade detects spurious failures and skips ports unnecessarily
>Confidential:   no
>Severity:       non-critical
>Priority:       low
>Responsible:    freebsd-ports-bugs
>State:          open
>Quarter:        
>Keywords:       
>Date-Required:
>Class:          sw-bug
>Submitter-Id:   current-users
>Arrival-Date:   Sun Oct 05 23:20:04 UTC 2008
>Closed-Date:
>Last-Modified:
>Originator:     Nate Eldredge
>Release:        7.0-RELEASE-p4
>Organization:
>Environment:
FreeBSD vulcan.lan 7.0-RELEASE-p4 FreeBSD 7.0-RELEASE-p4 #13: Wed Sep  3 16:13:10 PDT 2008     nate at vulcan.lan:/usr/obj/usr/src/sys/VULCAN  amd64
>Description:
When running portupgrade -a, if ports A and B are to be upgraded and B depends on A, it sometimes happens that A is upgraded first, successfully, but portupgrade for some reason thinks it fails and skips upgrading B.

A script session with portupgrade is posted at http://math.ucsd.edu/~neldredg/script-200810051055 (too large to attach).  Note, at line 53711, that textproc/rarian was skipped due to the ostensible failure of libxslt.  However, just before this we see that libxslt in fact installed successfully.  If portupgrade is run a second time, rarian will upgrade successfully.
>How-To-Repeat:
It is a bit awkward because it requires that you have ports that are out of date, and once you run portupgrade, they are no longer out of date.  I tried to put together an easy testcase, but did not succeed.  But hopefully there is enough information here to let someone else make one.
>Fix:
As a workaround, you can just run portupgrade -a repeatedly until no more packages are skipped, but this is a pain.

>Release-Note:
>Audit-Trail:
>Unformatted:



More information about the freebsd-ports-bugs mailing list