[PATCH] Disable frame buffer compression on ironlake

[PATCH] Disable frame buffer compression on ironlake

[Alex,Shi]

 Open fbc will cause about 1 Watts power increase on ironlake
 for bltk-office/player benchmark. So just disable it and left
 code here to waiting for better solution.

Signed-off-by: Alex Shi <alex.shi@intel.com>
---
 drivers/gpu/drm/i915/i915_drv.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index 80745f8..029af7a 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -150,7 +150,7 @@ static const struct intel_device_info intel_ironlake_d_info = {
 
 static const struct intel_device_info intel_ironlake_m_info = {
 	.gen = 5, .is_mobile = 1,
-	.need_gfx_hws = 1, .has_fbc = 1, .has_rc6 = 1, .has_hotplug = 1,
+	.need_gfx_hws = 1, .has_fbc = 0, .has_rc6 = 1, .has_hotplug = 1,
 	.has_bsd_ring = 1,
 };
 
-- 
1.7.0

Re: [PATCH] Disable frame buffer compression on ironlake

[Chris Wilson]

On Fri, 19 Nov 2010 10:00:03 +0800, "Alex,Shi" <alex.shi@intel.com> wrote:
>  Open fbc will cause about 1 Watts power increase on ironlake
>  for bltk-office/player benchmark. So just disable it and left
>  code here to waiting for better solution.

FBC is disabled through the chicken bits, so I am really curious to know
where the power saving is coming from. I just want to verify that we have
not made a fundamental mistake in our FBC logic that is consuming more
power than is necessary across all platforms.

In terms of code clarity, we should leave a note as to why we explicitly
disable fbc, e.g.: .has_fbc = 0, /* hardware support is buggy */
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre

Re: [PATCH] Disable frame buffer compression on ironlake

[Jesse Barnes]

On Fri, 19 Nov 2010 08:47:27 +0000
Chris Wilson <chris@chris-wilson.co.uk> wrote:

> On Fri, 19 Nov 2010 10:00:03 +0800, "Alex,Shi" <alex.shi@intel.com> wrote:
> >  Open fbc will cause about 1 Watts power increase on ironlake
> >  for bltk-office/player benchmark. So just disable it and left
> >  code here to waiting for better solution.
> 
> FBC is disabled through the chicken bits, so I am really curious to know
> where the power saving is coming from. I just want to verify that we have
> not made a fundamental mistake in our FBC logic that is consuming more
> power than is necessary across all platforms.
> 
> In terms of code clarity, we should leave a note as to why we explicitly
> disable fbc, e.g.: .has_fbc = 0, /* hardware support is buggy */

Did we disable the whole thing with chicken bits?  I thought we only
disabled some aspects of FBC that were known buggy.  AFAIK it should
generally work on ILK.  You can assign the bug to me, I'll take a look.

Thanks,
-- 
Jesse Barnes, Intel Open Source Technology Center

Re: [PATCH] Disable frame buffer compression on ironlake

[Chris Wilson]

On Fri, 19 Nov 2010 10:00:03 +0800, "Alex,Shi" <alex.shi@intel.com> wrote:
>  Open fbc will cause about 1 Watts power increase on ironlake
>  for bltk-office/player benchmark. So just disable it and left
>  code here to waiting for better solution.

I applied a slightly modified patch to -fixes, thanks.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre