[Bug 202228] [proposal] x11/nvidia-driver: Patch for new file hierarchy starting with NVVERION >= 355.006

[bugzilla-noreply at freebsd.org]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=202228

            Bug ID: 202228
           Summary: [proposal] x11/nvidia-driver: Patch for new file
                    hierarchy starting with NVVERION >= 355.006
           Product: Ports & Packages
           Version: Latest
          Hardware: Any
                OS: Any
            Status: New
          Severity: Affects Many People
          Priority: ---
         Component: Individual Port(s)
          Assignee: danfe at FreeBSD.org
          Reporter: ohartman at zedat.fu-berlin.de
             Flags: maintainer-feedback?(danfe at FreeBSD.org)
          Assignee: danfe at FreeBSD.org

Created attachment 159739
  --> https://bugs.freebsd.org/bugzilla/attachment.cgi?id=159739&action=edit
x11/nvidia-driver patchset >= 355.06

nVidia started a new dir hierarchy introduced with driver BETA 355.06 (aka
355.006), which differs slightly from the previous dir hierarchy.nVidia also
introduces a new build structure.

It seems on a sneak look that the patches supplied so far for FreeBSD still
apply and I tried to come along with the introduced dir hierarchy via a simple
and mechanical patch, see attached.

I intriduced a variable NVSRC in Makefile, which is set to "nvidia", if
NVVERSION >= 355.006, and empty else. With that variable, some paths get
resembled like ${WRKSRC}/src/${NVSERC}/nvidia_linux.c, for instance, so if the
old sources for the driver are used nothing changes, but with the newer driver
sources, the new paths get introduced.

I didn't find an elegant solution for the patch files located in files. I
simply duplicated the old ones and changes the diff header to accomplish the
new directory hierarchy. I thought this might be the most suitable way even if
some redundancy has been introduced.

I tested (an run so far successfully) the BETA driver 355.06 on CURRENT
(FreeBSD 11.0-CURRENT #51 r286580: Mon Aug 10 16:52:52 CEST 2015 amd64) so far
and it seems to be stable.

Could someone test it for 9.X and 10.X?

Regards,
oh

-- 
You are receiving this mail because:
You are the assignee for the bug.