Summary: | add 'staging' support to comms/uarduno | ||||||||
---|---|---|---|---|---|---|---|---|---|
Product: | Ports & Packages | Reporter: | Bob Frazier <bobf> | ||||||
Component: | Individual Port(s) | Assignee: | Rene Ladan <rene> | ||||||
Status: | Closed FIXED | ||||||||
Severity: | Affects Only Me | ||||||||
Priority: | Normal | ||||||||
Version: | Latest | ||||||||
Hardware: | Any | ||||||||
OS: | Any | ||||||||
Attachments: |
|
Description
Bob Frazier
2014-05-10 19:10:00 UTC
Class Changed From-To: sw-bug->maintainer-update Fix category (submitter is maintainer) (via the GNATS Auto Assign Tool) Responsible Changed From-To: freebsd-ports-bugs->rene Take I made some changes to your patch: - bump PORTREVISION because ids.txt changes - remove parts of Makefile which are already handled by USES=kmod: - CATEGORIES += kld - setting SSP_UNSAFE - adding KMODDIR to MAKE_ENV - running kldxref upon (de)installation - remove NO_MTREE, works fine with USES=kmod out of the box - allowing packaging as regular user by adding USES=uidfix and making the kernel object writable - setting the LICENSE to BSD2CLAUSE and moving it to the correct location in the Makefile - remove support for FreeBSD < 8.0 - other portlint fixes Two questions left: - is BSD 2 clause the correct license? - upon installation, kldxref dumps core on my 10.0-amd64 box, claiming that the .ko just built has too many sections and is not dynamically linked. Can you investigate? Thanks, René State Changed From-To: open->feedback Feedback requested. Ok, the kldxref problem seems to be 64-bit specific. It works fine when installing the package in a 10-i386 poudriere jail. René Created attachment 143543 [details]
Do not strip the .ko file, additional cleanups (unloading is done by USES=kmod, and all supported FreeBSD versions have usb_core.c)
Not stripping the .ko file does not help on amd64, so maybe the driver does something wrong? (In reply to Rene Ladan from comment #7) > Not stripping the .ko file does not help on amd64, so maybe the driver does > something wrong? How about committing the patch as-is and marking it broken on amd64 for now? A commit references this bug: Author: rene Date: Sun Jun 22 21:11:42 UTC 2014 New revision: 358829 URL: http://svnweb.freebsd.org/changeset/ports/358829 Log: comms/uarduno: modernize Makefile and add device IDs. - Use kmod, so add staging support. - Fix license support (BSD2CLAUSE) - Fix packaging as user - Add support for device IDs 0x0010, 0x0042, 0x0043 - Remove superfluous Makefile variables (NO_PACKAGE, SSP_UNSAFE, MAKE_ENV, PREFIX, SRCPREFIX, NO_MTREE) - Remove support for FreeBSD < 8 - Fix WWW line in pkg-descr - Bump PORTREVISION PR: ports/189621 Submitted by: maintainer (Bob Frazier, <bobf@mrp3.com>) Changes: head/comms/uarduno/Makefile head/comms/uarduno/files/ids.txt head/comms/uarduno/pkg-descr |