Nouveau - Mar 2014 - [PATCH v2] disp/nvd0-: allow 540MHz data rate for nvd0+ devices

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=76319
Signed-off-by: Ilia Mirkin <imirkin at alum.mit.edu>
---

It's unclear to me whether GF11x devices actually support this, but
the disp "split" is at nvd0, so I went with that. The marketing docs
are fairly unclear. However most of them don't actually have DP in the
first place, so it may not be a huge issue.

I don't have the hardware for testing this myself, but hopefully
someone can verify that this doesn't break "regular" non-4K
setups.

v1 -> v2: nouveau_device -> nv_device

 nvkm/engine/disp/dport.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/nvkm/engine/disp/dport.c b/nvkm/engine/disp/dport.c
index 1bd4c63..1ea3e65 100644
--- a/nvkm/engine/disp/dport.c
+++ b/nvkm/engine/disp/dport.c
@@ -273,12 +273,15 @@ nouveau_dp_train(struct nouveau_disp *disp, const struct
nouveau_dp_func *func,
 		.outp = outp,
 		.head = head,
 	}, *dp = &_dp;
-	const u32 bw_list[] = { 270000, 162000, 0 };
+	const u32 bw_list[] = { 540000, 270000, 162000, 0 };
 	const u32 *link_bw = bw_list;
 	u8  hdr, cnt, len;
 	u32 data;
 	int ret;
 
+	if (nv_device(disp)->card_type < NV_D0)
+		link_bw++;
+
 	/* find the bios displayport data relevant to this output */
 	data = nvbios_dpout_match(bios, outp->hasht, outp->hashm,
&dp->version,
 				 &hdr, &cnt, &len, &dp->info);
-- 
1.8.3.2

On Thu, Mar 20, 2014 at 12:45 AM, Ilia Mirkin <imirkin at alum.mit.edu>
wrote:
> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=76319
> Signed-off-by: Ilia Mirkin <imirkin at alum.mit.edu>
> ---
>
> It's unclear to me whether GF11x devices actually support this, but
> the disp "split" is at nvd0, so I went with that. The marketing
docs
> are fairly unclear. However most of them don't actually have DP in the
> first place, so it may not be a huge issue.
>
> I don't have the hardware for testing this myself, but hopefully
> someone can verify that this doesn't break "regular" non-4K
setups.
>
> v1 -> v2: nouveau_device -> nv_device
I've pushed a couple of patches that make the NV_D0 check unnecessary,
should be good to go otherwise.

Ben.
>
>  nvkm/engine/disp/dport.c | 5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/nvkm/engine/disp/dport.c b/nvkm/engine/disp/dport.c
> index 1bd4c63..1ea3e65 100644
> --- a/nvkm/engine/disp/dport.c
> +++ b/nvkm/engine/disp/dport.c
> @@ -273,12 +273,15 @@ nouveau_dp_train(struct nouveau_disp *disp, const
struct nouveau_dp_func *func,
>                 .outp = outp,
>                 .head = head,
>         }, *dp = &_dp;
> -       const u32 bw_list[] = { 270000, 162000, 0 };
> +       const u32 bw_list[] = { 540000, 270000, 162000, 0 };
>         const u32 *link_bw = bw_list;
>         u8  hdr, cnt, len;
>         u32 data;
>         int ret;
>
> +       if (nv_device(disp)->card_type < NV_D0)
> +               link_bw++;
> +
>         /* find the bios displayport data relevant to this output */
>         data = nvbios_dpout_match(bios, outp->hasht, outp->hashm,
&dp->version,
>                                  &hdr, &cnt, &len,
&dp->info);
> --
> 1.8.3.2
>
> _______________________________________________
> Nouveau mailing list
> Nouveau at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/nouveau