svn commit: r338172 - in head: . sys/amd64/conf sys/dev/drm sys/dev/drm2 sys/dev/drm2/i915 sys/dev/drm2/radeon sys/dev/drm2/radeon/reg_srcs sys/dev/drm2/ttm sys/i386/conf sys/modules sys/modules/dr...

[Johannes Lundberg]

On Wed, Aug 22, 2018 at 8:51 AM John Baldwin <jhb at freebsd.org> wrote:

> On 8/22/18 8:41 AM, Hans Petter Selasky wrote:
> > On 8/22/18 9:35 AM, Matthew Macy wrote:
> >> Johannes - do you know off hand?
> >> -M
> >
> > Hi,
> >
> > This function is still not implemented. This is part of the LGPLv2 part.
> >
>

Yeah, I confused it with another function...

> Probably the function and the restore companion doesn't have to panic.
> >
> > --HPS
> >
> >> static inline void
> >> linux_pci_save_state(struct pci_dev *pdev)
> >> {
> >>
> >>         panic("implment me!!");
> >>         UNIMPLEMENTED();
> >> }
>
> Can't this just use pci_save_state() of the FreeBSD device_t?  If there
> are bits of PCI space that pci_save_state() isn't saving that we need to
> save/restore, we should probably just add those to the FreeBSD one
> (e.g. additional capability registers).
>

Yes, it seems i915 has been patched to do just this. We can have the
linux_* functions call the freebsd equivalent so that will fix amd/radeon.


> --
> John Baldwin
>