All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Ville Syrjälä" <ville.syrjala@linux.intel.com>
To: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: intel-gfx@lists.freedesktop.org, "Runyan,
	Arthur J" <arthur.j.runyan@intel.com>
Subject: Re: [PATCH 8/8] drm/i915: Adjust system agent voltage on CNL if required by DDI ports
Date: Fri, 20 Oct 2017 14:11:52 +0300	[thread overview]
Message-ID: <20171020111152.GC10981@intel.com> (raw)
In-Reply-To: <20171019235456.wagrvqrvuw4m3bas@intel.com>

On Thu, Oct 19, 2017 at 04:54:56PM -0700, Rodrigo Vivi wrote:
> 
> On Wed, Oct 18, 2017 at 08:48:25PM +0000, Ville Syrjala wrote:
> > From: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > 
> > On CNL we may need to bump up the system agent voltage not only due
> > to CDCLK but also when driving DDI port with a sufficiently high clock.
> > To that end start tracking the minimum acceptable voltage for each crtc.
> > We do the tracking via crtcs because we don't have any kind of encoder
> > state. Also there's no downside to doing it this way, and it matches how
> > we track cdclk requirements on account of pixel rate.
> > 
> > Cc: Mika Kahola <mika.kahola@intel.com>
> > Cc: Manasi Navare <manasi.d.navare@intel.com>
> > Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
> > Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> 
> 
> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
> 
> Tested-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
> (Although I didn't find cases where I could force a higher voltage level,
> everything works well on CNL with this.)

I went and actually read the spec now :) so I can now see that this
doesn't really match the sequence listed in the spec.

For example, let's assume we have a modeset where we have to disable
a DPLL, change CDCLK, and finally enable a DPLL again.

What the spec says we should do is something like this:
1.  DVFS pre sequence
2.  disable DPLL
3.  DVFS post sequence
4.  disable DPLL power
...
5.  DVFS pre sequence
6.  configure CDCLK_CTL
7.  DVFS post sequence
...
8.  enable DPLL power
9.  DVFS pre sequence
10. enable DPLL
11. DVFS post sequence

With my code what we'd end up doing is this instead:
1. disable DPLL
2. disable DPLL power
...
3. DVFS pre sequence
4. configure CDCLK_CTL
5. DVFS post sequence
...
6. enable DPLL power
7. enable DPLL

So my way always results in running the DVFS sequences at most once.
With the way the spec has things listed we might have to run the
sequences multiple times. 

Art, is there a good reason why we'd actually have to run the
DVFS sequences around each DPLL_ENABLE write, instead of just
doing it once at any point between disabling and enabling
DPLLs?

-- 
Ville Syrjälä
Intel OTC
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

  reply	other threads:[~2017-10-20 11:11 UTC|newest]

Thread overview: 58+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-10-18 20:48 [PATCH 0/8] drm/i915: CNL DVFS thing Ville Syrjala
2017-10-18 20:48 ` [PATCH 1/8] drm/i915: Clean up some cdclk switch statements Ville Syrjala
2017-10-19  7:20   ` Mika Kahola
2017-10-18 20:48 ` [PATCH 2/8] drm/i915: Start tracking voltage level in the cdclk state Ville Syrjala
2017-10-19 23:32   ` Rodrigo Vivi
2017-10-20 14:01   ` Ville Syrjälä
2017-10-20 20:43     ` Rodrigo Vivi
2017-10-23 12:13       ` Ville Syrjälä
2017-10-23 17:14         ` Rodrigo Vivi
2017-10-18 20:48 ` [PATCH 3/8] drm/i915: USe cdclk_state->voltage on VLV/CHV Ville Syrjala
2017-10-19 17:43   ` [PATCH v2 3/8] drm/i915: Use " Ville Syrjala
2017-10-19 23:42     ` Rodrigo Vivi
2017-10-20 16:20       ` Ville Syrjälä
2017-10-20 17:03   ` [PATCH v3 " Ville Syrjala
2017-10-18 20:48 ` [PATCH 4/8] drm/i915: Use cdclk_state->voltage on BDW Ville Syrjala
2017-10-19 23:44   ` Rodrigo Vivi
2017-10-20 16:14     ` Ville Syrjälä
2017-10-20 17:03   ` [PATCH v2 " Ville Syrjala
2017-10-20 20:47     ` Rodrigo Vivi
2017-10-18 20:48 ` [PATCH 5/8] drm/i915: Use cdclk_state->voltage on SKL/KBL/CFL Ville Syrjala
2017-10-19 23:47   ` Rodrigo Vivi
2017-10-20 11:18     ` Ville Syrjälä
2017-10-20 20:45       ` Rodrigo Vivi
2017-10-18 20:48 ` [PATCH 6/8] drm/i915: Use cdclk_state->voltage on BXT/GLK Ville Syrjala
2017-10-20 20:51   ` Rodrigo Vivi
2017-10-18 20:48 ` [PATCH 7/8] drm/i915: Use cdclk_state->voltage on CNL Ville Syrjala
2017-10-18 21:50   ` Rodrigo Vivi
2017-10-18 22:43     ` Rodrigo Vivi
2017-10-19 10:48     ` Ville Syrjälä
2017-10-19 10:56       ` Mika Kahola
2017-10-19 12:19         ` Ville Syrjälä
2017-10-19 23:52       ` Rodrigo Vivi
2017-10-23 18:29   ` Rodrigo Vivi
2017-10-18 20:48 ` [PATCH 8/8] drm/i915: Adjust system agent voltage on CNL if required by DDI ports Ville Syrjala
2017-10-19 23:54   ` Rodrigo Vivi
2017-10-20 11:11     ` Ville Syrjälä [this message]
2017-10-20 17:48       ` Runyan, Arthur J
2017-10-20 20:07         ` Ville Syrjälä
2017-10-20 20:36           ` Rodrigo Vivi
2017-10-20 21:44             ` Runyan, Arthur J
2017-10-23 12:03               ` Ville Syrjälä
2017-10-23 11:48             ` Ville Syrjälä
2017-10-20 14:18     ` Ville Syrjälä
2017-10-20 16:11       ` Ville Syrjälä
2017-10-20 16:09   ` [PATCH v2 " Ville Syrjala
2017-10-20 16:52     ` Ville Syrjälä
2017-10-20 17:05   ` [PATCH v3 " Ville Syrjala
2017-10-23 18:39     ` Rodrigo Vivi
2017-10-18 21:07 ` ✗ Fi.CI.BAT: warning for drm/i915: CNL DVFS thing Patchwork
2017-10-19 17:31   ` Ville Syrjälä
2017-10-19 18:17 ` ✗ Fi.CI.BAT: failure for drm/i915: CNL DVFS thing (rev2) Patchwork
2017-10-19 18:52 ` Patchwork
2017-10-19 20:07 ` ✗ Fi.CI.BAT: warning " Patchwork
2017-10-19 23:27 ` ✓ Fi.CI.BAT: success " Patchwork
2017-10-20  0:22 ` ✓ Fi.CI.IGT: " Patchwork
2017-10-20 16:28 ` ✓ Fi.CI.BAT: success for drm/i915: CNL DVFS thing (rev3) Patchwork
2017-10-20 17:48 ` ✓ Fi.CI.BAT: success for drm/i915: CNL DVFS thing (rev6) Patchwork
2017-10-20 19:19 ` ✓ Fi.CI.IGT: " Patchwork

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=20171020111152.GC10981@intel.com \
    --to=ville.syrjala@linux.intel.com \
    --cc=arthur.j.runyan@intel.com \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=rodrigo.vivi@intel.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.