[PATCH] drm/i915: reverse dp link param selection, prefer fast over wide again

[PATCH] drm/i915: reverse dp link param selection, prefer fast over wide again

[Daniel Vetter]

... it's this time of the year again. Originally we've frobbed this to
fix up some regressions, but maybe our DP code improved sufficiently
now that we can dare to do again what the spec recommends.

This reverts

commit 2514bc510d0c3aadcc5204056bb440fa36845147
Author: Jesse Barnes <jbarnes@virtuousgeek.org>
Date:   Thu Jun 21 15:13:50 2012 -0700

    drm/i915: prefer wide & slow to fast & narrow in DP configs

I'm pretty sure I'll regret this patch, but otoh I expect we won't
make progress here without poking the devil occasionally.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=73694
Cc: peter@colberg.org
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
Tested-by: Itai BEN YAACOV <candeb@free.fr>
Tested-by: David En <d.engraf@arcor.de>
Reported-and-Tested-by: Marcus Bergner <marcusbergner@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/intel_dp.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index 1ac4b11765c7..e50aadfc2184 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -909,8 +909,8 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 		mode_rate = intel_dp_link_required(adjusted_mode->crtc_clock,
 						   bpp);
 
-		for (clock = 0; clock <= max_clock; clock++) {
-			for (lane_count = 1; lane_count <= max_lane_count; lane_count <<= 1) {
+		for (lane_count = 1; lane_count <= max_lane_count; lane_count <<= 1) {
+			for (clock = 0; clock <= max_clock; clock++) {
 				link_clock = drm_dp_bw_code_to_link_rate(bws[clock]);
 				link_avail = intel_dp_max_data_rate(link_clock,
 								    lane_count);
-- 
1.8.5.2

Re: [PATCH] drm/i915: reverse dp link param selection, prefer fast over wide again

[Jani Nikula]

On Mon, 03 Mar 2014, Daniel Vetter <daniel.vetter@ffwll.ch> wrote:
> ... it's this time of the year again. Originally we've frobbed this to
> fix up some regressions, but maybe our DP code improved sufficiently
> now that we can dare to do again what the spec recommends.
>
> This reverts
>
> commit 2514bc510d0c3aadcc5204056bb440fa36845147
> Author: Jesse Barnes <jbarnes@virtuousgeek.org>
> Date:   Thu Jun 21 15:13:50 2012 -0700
>
>     drm/i915: prefer wide & slow to fast & narrow in DP configs
>
> I'm pretty sure I'll regret this patch, but otoh I expect we won't
> make progress here without poking the devil occasionally.

Reviewed-by: Jani Nikula <jani.nikula@intel.com>

I agree we should bite the bullet and do this. This is what the DP specs
tell us to do. I'm pretty sure there will be regressions, but I'm also
starting to believe there's a greater number of machines that are
currently either broken beyond repair or papered over with some hacks
than the number that will regress. We should just dig into them and fix
them instead of burying our heads in the sand.

BR,
Jani.

>
> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=73694
> Cc: peter@colberg.org
> Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
> Tested-by: Itai BEN YAACOV <candeb@free.fr>
> Tested-by: David En <d.engraf@arcor.de>
> Reported-and-Tested-by: Marcus Bergner <marcusbergner@gmail.com>
> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> ---
>  drivers/gpu/drm/i915/intel_dp.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
> index 1ac4b11765c7..e50aadfc2184 100644
> --- a/drivers/gpu/drm/i915/intel_dp.c
> +++ b/drivers/gpu/drm/i915/intel_dp.c
> @@ -909,8 +909,8 @@ intel_dp_compute_config(struct intel_encoder *encoder,
>  		mode_rate = intel_dp_link_required(adjusted_mode->crtc_clock,
>  						   bpp);
>  
> -		for (clock = 0; clock <= max_clock; clock++) {
> -			for (lane_count = 1; lane_count <= max_lane_count; lane_count <<= 1) {
> +		for (lane_count = 1; lane_count <= max_lane_count; lane_count <<= 1) {
> +			for (clock = 0; clock <= max_clock; clock++) {
>  				link_clock = drm_dp_bw_code_to_link_rate(bws[clock]);
>  				link_avail = intel_dp_max_data_rate(link_clock,
>  								    lane_count);
> -- 
> 1.8.5.2
>
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx

-- 
Jani Nikula, Intel Open Source Technology Center

Re: [PATCH] drm/i915: reverse dp link param selection, prefer fast over wide again

[Daniel Vetter]

On Mon, Mar 03, 2014 at 04:36:11PM +0200, Jani Nikula wrote:
> On Mon, 03 Mar 2014, Daniel Vetter <daniel.vetter@ffwll.ch> wrote:
> > ... it's this time of the year again. Originally we've frobbed this to
> > fix up some regressions, but maybe our DP code improved sufficiently
> > now that we can dare to do again what the spec recommends.
> >
> > This reverts
> >
> > commit 2514bc510d0c3aadcc5204056bb440fa36845147
> > Author: Jesse Barnes <jbarnes@virtuousgeek.org>
> > Date:   Thu Jun 21 15:13:50 2012 -0700
> >
> >     drm/i915: prefer wide & slow to fast & narrow in DP configs
> >
> > I'm pretty sure I'll regret this patch, but otoh I expect we won't
> > make progress here without poking the devil occasionally.
> 
> Reviewed-by: Jani Nikula <jani.nikula@intel.com>
> 
> I agree we should bite the bullet and do this. This is what the DP specs
> tell us to do. I'm pretty sure there will be regressions, but I'm also
> starting to believe there's a greater number of machines that are
> currently either broken beyond repair or papered over with some hacks
> than the number that will regress. We should just dig into them and fix
> them instead of burying our heads in the sand.

Queued for -next, thanks for the review.
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch

Re: [PATCH] drm/i915: reverse dp link param selection, prefer fast over wide again

[Jesse Barnes]

On Mon,  3 Mar 2014 11:18:10 +0100
Daniel Vetter <daniel.vetter@ffwll.ch> wrote:

> ... it's this time of the year again. Originally we've frobbed this to
> fix up some regressions, but maybe our DP code improved sufficiently
> now that we can dare to do again what the spec recommends.
> 
> This reverts
> 
> commit 2514bc510d0c3aadcc5204056bb440fa36845147
> Author: Jesse Barnes <jbarnes@virtuousgeek.org>
> Date:   Thu Jun 21 15:13:50 2012 -0700
> 
>     drm/i915: prefer wide & slow to fast & narrow in DP configs
> 
> I'm pretty sure I'll regret this patch, but otoh I expect we won't
> make progress here without poking the devil occasionally.

Acked-by: Jesse Barnes <jbarnes@virtuousgeek.org>