On Sat, Feb 05, 2022 at 11:44:03AM +0100, Mark Kettenis wrote:
> We want to get rid of the uvm_km_valloc() interfaces in favour of
> km_alloc(). This changes the calls in drm(4) over. The kv_physwait
> struct is made static to prevent collission with a symbol in
> vm_machdep.c on some
On Sat, Feb 05, 2022 at 02:26:48PM +0100, Mark Kettenis wrote:
> > Date: Sat, 5 Feb 2022 22:34:10 +1100
> > From: Jonathan Gray
> >
> > On Sat, Feb 05, 2022 at 11:44:03AM +0100, Mark Kettenis wrote:
> > > We want to get rid of the uvm_km_valloc() interfaces in favour of
> > > km_alloc(). This
> Date: Sat, 5 Feb 2022 22:34:10 +1100
> From: Jonathan Gray
>
> On Sat, Feb 05, 2022 at 11:44:03AM +0100, Mark Kettenis wrote:
> > We want to get rid of the uvm_km_valloc() interfaces in favour of
> > km_alloc(). This changes the calls in drm(4) over. The kv_physwait
> > struct is made static
On Sat, Feb 05, 2022 at 11:44:03AM +0100, Mark Kettenis wrote:
> We want to get rid of the uvm_km_valloc() interfaces in favour of
> km_alloc(). This changes the calls in drm(4) over. The kv_physwait
> struct is made static to prevent collission with a symbol in
> vm_machdep.c on some
We want to get rid of the uvm_km_valloc() interfaces in favour of
km_alloc(). This changes the calls in drm(4) over. The kv_physwait
struct is made static to prevent collission with a symbol in
vm_machdep.c on some architectures. The goal is to move this into
uvm/uvm_km.c eventually.
Just to