Re: [Intel-gfx] [PATCH 4/4] drm/i915: Move ioremap_wc tracking onto VMA

From: Daniel Vetter
Date: Thu Apr 21 2016 - 03:27:49 EST


On Wed, Apr 20, 2016 at 11:27:27PM +0200, Luis R. Rodriguez wrote:
> On Wed, Apr 20, 2016 at 01:17:30PM +0200, Daniel Vetter wrote:
> > On Wed, Apr 20, 2016 at 11:10:54AM +0200, Luis R. Rodriguez wrote:
> > > Reason I ask is since I noticed a while ago a lot of drivers
> > > were using info->fix.smem_start and info->fix.smem_len consistently
> > > for their ioremap'd areas it might make sense instead to let the
> > > internal framebuffer (register_framebuffer()) optionally manage the
> > > ioremap_wc() for drivers, given that this is pretty generic stuff.
> >
> > All that legacy fbdev stuff is just for legacy support, and I prefer to
> > have that as dumb as possible. There's been some discussion even around
> > lifting the "kick out firmware fb driver" out of fbdev, since we'd need it
> > to have a simple drm driver for e.g. uefi.
> >
> > But I definitely don't want a legacy horror show like fbdev to
> > automagically take care of device mappings for drivers.
>
> Makes sense, it also still begs the question if more modern APIs
> could manage the ioremap for you. Evidence shows people get
> sloppy and if things were done internally with helpers it may
> be easier to later make adjustments.

Real gpus generally have so much mmio space that you want to ioremap them
on demand. At least if you still care about 32bit support. And on-die gpus
on socs or similar tend to not have an mmio range to access the gfx
remapping range at all, but instead expect that to be done with gpu
pagetables.

So at least with gpus I don't see a real demand for this, and the existing
users are mostly old fbdev drivers that really no one should be touching
;-)

Cheers, Daniel
--
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch