Eric Anholt wrote:
> I would like to see a form of OS-templated code get into CVS. If the
> rest of the DRI developers are interested in having this done, this is
> what I would like to happen:
>
> Step 1:
> - repocopy most of linux/drm/kernel/ to shared/drm/kernel (see below)
> - make the build symlink stuff in shared/drm/kernel to <os>/drm/kernel
> - cvs rm the repocopied stuff from linux/drm/kernel
> This would have to happen simultaneously in the branches to preserve
> history.
>
> Step 2:
> In the bsd branch (bsd-3-0-0), create the drm_os_* headers and commit
> the DRM_OS_ERR and DRM_OS_IOCTL changes. These cover a large number of
> the bytes of the diff. Quickly test and merge it, as they should be
> harmless.
>
> Step 3:
> Start applying the BSD changes to the bsd branch. These will be tested
> on Linux as they are applied.
>
> Once that bsd branch is brought up to the point of being better than
> FreeBSD's DRM, I would merge it back to FreeBSD and hopefully merge to
> trunk. Then I could work on adding the new stuff I've worked on
> (PCIGART, i8x0) then work on new things (NetBSD, which looks like it
> wouldn't be very difficult at all).
>
> Most of the *.[ch] from linux/drm/kernel would be repocopied. Currently
> the BSD branch has its own drm_vm.h and drm_agpsupport.h, and does not
> use drm_proc.h and drm_stub.h, (which would be left in
> linux/drm/kernel). Additionally I think we would split drm_fops.h for
> the two OSes, because the shared code is so limited it makes reading the
> file tough.
>
> What do people think about the code, and if they like it, what about the
> process for getting it into CVS?
Eric,
The watermark I would apply for measuring the success of OS-templating
is device indepedence. If all device specific code can go into the
shared/drm/kernel directory and only device independent code is found in
the OS specific directories <os>/drm/kernel then this will be a huge
step forward for the DRI. It's likely you'll end up with some OS
specific #ifdef's in the shared directory, but if you can completely
remove these ifdef's while supporting three or more OS's then we can be
confident we have a great template interface, and you can focus more of
your efforts on keeping up to date with general DRM functionality and
less on each device specific implementation.
Regards,
Jens
-- /\
Jens Owen / \/\ _
[EMAIL PROTECTED] / \ \ \ Steamboat Springs, Colorado
_______________________________________________________________
Have big pipes? SourceForge.net is looking for download mirrors. We supply
the hardware. You get the recognition. Email Us: [EMAIL PROTECTED]
_______________________________________________
Dri-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/dri-devel
