From: "Ville Syrjälä" <ville.syrjala@linux.intel.com>
To: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Cc: intel-gfx@lists.freedesktop.org
Subject: Re: [PATCH] drm/i915: Simplify the way BC bifurcation state consistency is kept
Date: Wed, 11 Mar 2015 14:24:02 +0200 [thread overview]
Message-ID: <20150311122402.GR11371@intel.com> (raw)
In-Reply-To: <1426073743-30724-1-git-send-email-ander.conselvan.de.oliveira@intel.com>
On Wed, Mar 11, 2015 at 01:35:43PM +0200, Ander Conselvan de Oliveira wrote:
> Remove the global modeset resource function that would disable the
> bifurcation bit, and instead enable/disable it when enabling the pch
> transcoder. The mode set consistency check should prevent us from
> disabling the bit if pipe C is enabled so the change should be safe.
>
> Note that this doens't affect the logic that prevents the bit being
> set while a pipe is active, since the patch retains the behavior of
> only chaging the bit if necessary. Because of the checks during mode
> set, the first change would necessarily happen with both pipes B and
> C disabled, and any subsequent write would be skipped.
>
> v2: Only change the bit during pch trancoder enable. (Ville)
Actually what I had inb mind was something like this:
pch_enable()
{
if (pipe == B && fdi_lanes > 2)
disable_bifurcation()
...
}
pch_disable()
{
...
if (pipe == B && fdi_lanes > 2)
enable_bifurcation()
}
So we only change it when enabling/disabling pipe B, never for pipe C.
Hence it remains enabled whenever pipe B doesn't need >2 FDI lanes.
> ---
> drivers/gpu/drm/i915/intel_display.c | 46 ++++++++----------------------------
> 1 file changed, 10 insertions(+), 36 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
> index 4008bf4..bfbd829 100644
> --- a/drivers/gpu/drm/i915/intel_display.c
> +++ b/drivers/gpu/drm/i915/intel_display.c
> @@ -3153,32 +3153,6 @@ static bool pipe_has_enabled_pch(struct intel_crtc *crtc)
> return crtc->base.state->enable && crtc->config->has_pch_encoder;
> }
>
> -static void ivb_modeset_global_resources(struct drm_device *dev)
> -{
> - struct drm_i915_private *dev_priv = dev->dev_private;
> - struct intel_crtc *pipe_B_crtc =
> - to_intel_crtc(dev_priv->pipe_to_crtc_mapping[PIPE_B]);
> - struct intel_crtc *pipe_C_crtc =
> - to_intel_crtc(dev_priv->pipe_to_crtc_mapping[PIPE_C]);
> - uint32_t temp;
> -
> - /*
> - * When everything is off disable fdi C so that we could enable fdi B
> - * with all lanes. Note that we don't care about enabled pipes without
> - * an enabled pch encoder.
> - */
> - if (!pipe_has_enabled_pch(pipe_B_crtc) &&
> - !pipe_has_enabled_pch(pipe_C_crtc)) {
> - WARN_ON(I915_READ(FDI_RX_CTL(PIPE_B)) & FDI_RX_ENABLE);
> - WARN_ON(I915_READ(FDI_RX_CTL(PIPE_C)) & FDI_RX_ENABLE);
> -
> - temp = I915_READ(SOUTH_CHICKEN1);
> - temp &= ~FDI_BC_BIFURCATION_SELECT;
> - DRM_DEBUG_KMS("disabling fdi C rx\n");
> - I915_WRITE(SOUTH_CHICKEN1, temp);
> - }
> -}
> -
> /* The FDI link training functions for ILK/Ibexpeak. */
> static void ironlake_fdi_link_train(struct drm_crtc *crtc)
> {
> @@ -3834,20 +3808,23 @@ static void ironlake_pch_transcoder_set_timings(struct intel_crtc *crtc,
> I915_READ(VSYNCSHIFT(cpu_transcoder)));
> }
>
> -static void cpt_enable_fdi_bc_bifurcation(struct drm_device *dev)
> +static void cpt_set_fdi_bc_bifurcation(struct drm_device *dev, bool enable)
> {
> struct drm_i915_private *dev_priv = dev->dev_private;
> uint32_t temp;
>
> temp = I915_READ(SOUTH_CHICKEN1);
> - if (temp & FDI_BC_BIFURCATION_SELECT)
> + if (!!(temp & FDI_BC_BIFURCATION_SELECT) == enable)
> return;
>
> WARN_ON(I915_READ(FDI_RX_CTL(PIPE_B)) & FDI_RX_ENABLE);
> WARN_ON(I915_READ(FDI_RX_CTL(PIPE_C)) & FDI_RX_ENABLE);
>
> - temp |= FDI_BC_BIFURCATION_SELECT;
> - DRM_DEBUG_KMS("enabling fdi C rx\n");
> + temp &= ~FDI_BC_BIFURCATION_SELECT;
> + if (enable)
> + temp |= FDI_BC_BIFURCATION_SELECT;
> +
> + DRM_DEBUG_KMS("%sabling fdi C rx\n", enable ? "en" : "dis");
> I915_WRITE(SOUTH_CHICKEN1, temp);
> POSTING_READ(SOUTH_CHICKEN1);
> }
> @@ -3855,20 +3832,19 @@ static void cpt_enable_fdi_bc_bifurcation(struct drm_device *dev)
> static void ivybridge_update_fdi_bc_bifurcation(struct intel_crtc *intel_crtc)
> {
> struct drm_device *dev = intel_crtc->base.dev;
> - struct drm_i915_private *dev_priv = dev->dev_private;
>
> switch (intel_crtc->pipe) {
> case PIPE_A:
> break;
> case PIPE_B:
> if (intel_crtc->config->fdi_lanes > 2)
> - WARN_ON(I915_READ(SOUTH_CHICKEN1) & FDI_BC_BIFURCATION_SELECT);
> + cpt_set_fdi_bc_bifurcation(dev, false);
> else
> - cpt_enable_fdi_bc_bifurcation(dev);
> + cpt_set_fdi_bc_bifurcation(dev, true);
>
> break;
> case PIPE_C:
> - cpt_enable_fdi_bc_bifurcation(dev);
> + cpt_set_fdi_bc_bifurcation(dev, true);
>
> break;
> default:
> @@ -13056,8 +13032,6 @@ static void intel_init_display(struct drm_device *dev)
> } else if (IS_IVYBRIDGE(dev)) {
> /* FIXME: detect B0+ stepping and use auto training */
> dev_priv->display.fdi_link_train = ivb_manual_fdi_link_train;
> - dev_priv->display.modeset_global_resources =
> - ivb_modeset_global_resources;
> } else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
> dev_priv->display.fdi_link_train = hsw_fdi_link_train;
> } else if (IS_VALLEYVIEW(dev)) {
> --
> 2.1.0
--
Ville Syrjälä
Intel OTC
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
next prev parent reply other threads:[~2015-03-11 12:24 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-03-09 8:59 [PATCH] drm/i915: Ignore pipe B active state when enabling pipe C Ander Conselvan de Oliveira
2015-03-09 9:24 ` Jani Nikula
2015-03-09 9:33 ` Ander Conselvan De Oliveira
2015-03-09 16:21 ` Daniel Vetter
2015-03-10 12:32 ` [PATCH] drm/i915: Simplify the way BC bifurcation state consistency is kept Ander Conselvan de Oliveira
2015-03-10 12:35 ` [PATCH] tests: Add test for pipe B and C interactions in IVB Ander Conselvan de Oliveira
2015-03-10 19:05 ` Daniel Vetter
2015-03-11 11:33 ` [PATCH igt 1/2] lib/kms: Add a way to override an output's mode Ander Conselvan de Oliveira
2015-03-11 11:33 ` [PATCH igt 2/2] tests: Add test for pipe B and C interactions in IVB Ander Conselvan de Oliveira
2015-03-27 13:35 ` Thomas Wood
2015-03-11 13:26 ` [PATCH igt 1/2] lib/kms: Add a way to override an output's mode Damien Lespiau
2015-03-11 13:48 ` Ander Conselvan De Oliveira
2015-03-11 14:26 ` Damien Lespiau
2015-03-27 13:30 ` Thomas Wood
2015-03-10 13:03 ` [PATCH] drm/i915: Simplify the way BC bifurcation state consistency is kept Ville Syrjälä
2015-03-10 19:14 ` Daniel Vetter
2015-03-11 11:35 ` Ander Conselvan de Oliveira
2015-03-11 11:37 ` Conselvan De Oliveira, Ander
2015-03-11 16:58 ` Ville Syrjälä
2015-03-11 20:42 ` Daniel Vetter
2015-03-11 12:24 ` Ville Syrjälä [this message]
2015-03-11 13:10 ` Ville Syrjälä
2015-03-11 13:23 ` Conselvan De Oliveira, Ander
2015-03-11 20:12 ` shuang.he
2015-03-10 19:10 ` Daniel Vetter
2015-03-10 20:40 ` shuang.he
2015-03-09 12:17 ` [PATCH] drm/i915: Ignore pipe B active state when enabling pipe C shuang.he
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20150311122402.GR11371@intel.com \
--to=ville.syrjala@linux.intel.com \
--cc=ander.conselvan.de.oliveira@intel.com \
--cc=intel-gfx@lists.freedesktop.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox