<html>
<head>
<meta content="text/html; charset=utf-8" http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<br>
<div class="moz-cite-prefix">On Wednesday 21 January 2015 08:33 PM,
Rodrigo Vivi wrote:<br>
</div>
<blockquote
cite="mid:1421852633.25196.4.camel@rdvivi-hillsboro.jf.intel.com"
type="cite">
<pre wrap="">On Wed, 2015-01-21 at 17:43 +0530, Ramalingam C wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi
On Friday 16 January 2015 04:41 AM, Rodrigo Vivi wrote:
</pre>
<blockquote type="cite">
<pre wrap="">On Fri, Jan 9, 2015 at 12:56 PM, Vandana Kannan
<a class="moz-txt-link-rfc2396E" href="mailto:vandana.kannan@intel.com"><vandana.kannan@intel.com></a> wrote:
</pre>
<blockquote type="cite">
<pre wrap="">From: Durgadoss R <a class="moz-txt-link-rfc2396E" href="mailto:durgadoss.r@intel.com"><durgadoss.r@intel.com></a>
This patch enables eDP DRRS for CHV by adding the
required IS_CHERRYVIEW() checks.
CHV uses the same register bit as VLV.
[Vandana]: Since CHV has 2 sets of M_N registers, it will follow the same code
path as gen < 8. Added CHV check in dp_set_m_n()
Signed-off-by: Durgadoss R <a class="moz-txt-link-rfc2396E" href="mailto:durgadoss.r@intel.com"><durgadoss.r@intel.com></a>
Signed-off-by: Vandana Kannan <a class="moz-txt-link-rfc2396E" href="mailto:vandana.kannan@intel.com"><vandana.kannan@intel.com></a>
---
drivers/gpu/drm/i915/intel_display.c | 4 ++--
drivers/gpu/drm/i915/intel_dp.c | 2 +-
2 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 25596ca..bb44fb9 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -5810,8 +5810,8 @@ void intel_cpu_transcoder_set_m_n(struct intel_crtc *crtc,
* for gen < 8) and if DRRS is supported (to make sure the
* registers are not unnecessarily accessed).
*/
- if (m2_n2 && INTEL_INFO(dev)->gen < 8 &&
- crtc->config.has_drrs) {
+ if (m2_n2 && (IS_CHERRYVIEW(dev) || INTEL_INFO(dev)->gen < 8)
+ && crtc->config.has_drrs) {
</pre>
</blockquote>
<pre wrap="">This change here doesn't seem safe. As I told on previous comment I'd
prefer changing intel_dp_set_m_n instead of re-using this intel_cpu
one...
</pre>
</blockquote>
<pre wrap="">Though I am rewriting the intel_dp_set_m_n() that will reuse
intel_cpu_transcoder_set_m_n() within.
But as a result, i am going to avoid the parallel usage of
intel_cpu_transcoder_set_m_n() and intel_dp_set_m_n().
So I am afraid this check for inclusion of cherryview for m2_n2
programming will be part of the newer code also.
Appending the RFC for the newer intel_dp_set_m_n() implementation
below. Please review.
RFC starts here:
drivers/gpu/drm/i915/intel_display.c | 19
++++++++++++++++---
drivers/gpu/drm/i915/intel_dp.c | 6
++----
drivers/gpu/drm/i915/intel_drv.h | 8
+++++++-
3 files changed, 25 insertions(+), 8 deletions(-)
diff --git a/drivers/gpu/drm/i915/intel_display.c
b/drivers/gpu/drm/i915/intel_display.c
index 061a253..59cc87f
100644
---
a/drivers/gpu/drm/i915/intel_display.c
+++
b/drivers/gpu/drm/i915/intel_display.c
@@ -5829,13 +5829,26 @@ void intel_cpu_transcoder_set_m_n(struct
intel_crtc *crtc,
}
}
-void intel_dp_set_m_n(struct intel_crtc
*crtc)
+void intel_dp_set_m_n(struct intel_crtc *crtc, enum link_m_n_set set
= M1_N1)
{
+ struct intel_link_m_n *dp_m_n,
*dp_m2_n2;
+
+ if (set == M1_N1)
{
+ dp_m_n =
&crtc->config.dp_m_n;
+ dp_m2_n2 =
&crtc->config.dp_m2_n2;
+ } else if (set == M2_N2)
{
+ /* Only one register programming is supported
*/
+ dp_m_n =
&crtc->config.dp_m_n;
+ dp_m2_n2 =
NULL;
+ } else
{
+ DRM_ERROR("Unsupported divider value
\n");
+
return;
+ }
+
if
(crtc->config.has_pch_encoder)
intel_pch_transcoder_set_m_n(crtc,
&crtc->config.dp_m_n);
else
- intel_cpu_transcoder_set_m_n(crtc,
&crtc->config.dp_m_n,
-
&crtc->config.dp_m2_n2);
+ intel_cpu_transcoder_set_m_n(crtc, dp_m_n,
dp_m2_n2);
}
static void vlv_update_pll(struct intel_crtc
*crtc,
diff --git a/drivers/gpu/drm/i915/intel_dp.c
b/drivers/gpu/drm/i915/intel_dp.c
index b315292..784b8dd
100644
---
a/drivers/gpu/drm/i915/intel_dp.c
+++
b/drivers/gpu/drm/i915/intel_dp.c
@@ -4817,11 +4817,10 @@ static void intel_dp_set_drrs_state(struct
drm_device *dev, int refresh_rate)
if (INTEL_INFO(dev)->gen >= 8 && !IS_CHERRYVIEW(dev))
{
switch(index)
{
case
DRRS_HIGH_RR:
-
intel_dp_set_m_n(intel_crtc);
+ intel_dp_set_m_n(intel_crtc,
M1_N1);
break;
case
DRRS_LOW_RR:
-
intel_cpu_transcoder_set_m_n(intel_crtc,
- &intel_crtc->config.dp_m2_n2,
NULL);
+ intel_dp_set_m_n(intel_crtc,
M2_N2);
break;
case
DRRS_MAX_RR:
default:
@@ -4835,7 +4834,6 @@ static void intel_dp_set_drrs_state(struct
drm_device *dev, int refresh_rate)
val |=
PIPECONF_EDP_RR_MODE_SWITCH_VLV;
else
val |=
PIPECONF_EDP_RR_MODE_SWITCH;
-
intel_dp_set_m_n(intel_crtc);
} else
{
if
(IS_VALLEYVIEW(dev))
val &=
~PIPECONF_EDP_RR_MODE_SWITCH_VLV;
diff --git a/drivers/gpu/drm/i915/intel_drv.h
b/drivers/gpu/drm/i915/intel_drv.h
index 86d31f2..910e613
100644
---
a/drivers/gpu/drm/i915/intel_drv.h
+++
b/drivers/gpu/drm/i915/intel_drv.h
@@ -595,6 +595,12 @@ struct intel_hdmi
{
struct
intel_dp_mst_encoder;
#define DP_MAX_DOWNSTREAM_PORTS
0x10
+enum link_m_n_set
{
+ M1_N1 =
0,
+
M2_N2,
+
DIVIDER_MAX
</pre>
</blockquote>
<pre wrap="">
do we need this divider_max here?</pre>
</blockquote>
<small>We dont need this. Removed it in the original patch.</small>
<br>
<blockquote
cite="mid:1421852633.25196.4.camel@rdvivi-hillsboro.jf.intel.com"
type="cite">
<pre wrap="">
</pre>
<blockquote type="cite">
<pre wrap="">+};
+
struct intel_dp
{
uint32_t
output_reg;
uint32_t
aux_ch_ctl_reg;
@@ -983,7 +989,7 @@ void hsw_enable_pc8(struct drm_i915_private
*dev_priv);
void hsw_disable_pc8(struct drm_i915_private
*dev_priv);
void intel_dp_get_m_n(struct intel_crtc
*crtc,
struct intel_crtc_config
*pipe_config);
-void intel_dp_set_m_n(struct intel_crtc
*crtc);
+void intel_dp_set_m_n(struct intel_crtc *crtc, enum link_m_n_set set
= M1_N1);
void intel_cpu_transcoder_set_m_n(struct intel_crtc
*crtc,
struct intel_link_m_n
*m_n,
struct intel_link_m_n
*m2_n2);
--
1.7.9.5
</pre>
<blockquote type="cite">
<pre wrap="">
</pre>
<blockquote type="cite">
<pre wrap=""> I915_WRITE(PIPE_DATA_M2(transcoder),
TU_SIZE(m2_n2->tu) | m2_n2->gmch_m);
I915_WRITE(PIPE_DATA_N2(transcoder), m2_n2->gmch_n);
diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index 3362d93..42195fe 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -4802,7 +4802,7 @@ static void intel_dp_set_drrs_state(struct drm_device *dev, int refresh_rate)
return;
}
- if (INTEL_INFO(dev)->gen >= 8) {
+ if (INTEL_INFO(dev)->gen >= 8 && !IS_CHERRYVIEW(dev)) {
switch(index) {
case DRRS_HIGH_RR:
intel_dp_set_m_n(intel_crtc);
--
2.0.1
_______________________________________________
Intel-gfx mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Intel-gfx@lists.freedesktop.org">Intel-gfx@lists.freedesktop.org</a>
<a class="moz-txt-link-freetext" href="http://lists.freedesktop.org/mailman/listinfo/intel-gfx">http://lists.freedesktop.org/mailman/listinfo/intel-gfx</a>
</pre>
</blockquote>
<pre wrap="">
</pre>
</blockquote>
<pre wrap="">--Ram
</pre>
</blockquote>
<pre wrap="">
But overall it is better and much more clear than previous version.</pre>
</blockquote>
<small>I have submitted this change for the intel_dp_set_m_n() as V2
of patch set 5/10. So this patch [7/10] doesn't have any change.</small><br>
<blockquote
cite="mid:1421852633.25196.4.camel@rdvivi-hillsboro.jf.intel.com"
type="cite">
<pre wrap="">
Thanks
</pre>
</blockquote>
--Ram<br>
</body>
</html>