Re: [PATCH] drm/i915/skl: Buffer translation improvements

From: David Weinehall <david.weinehall@linux.intel.com>
To: Chris Wilson <chris@chris-wilson.co.uk>,
	intel-gfx <intel-gfx@lists.freedesktop.org>
Subject: Re: [PATCH] drm/i915/skl: Buffer translation improvements
Date: Tue, 23 Jun 2015 13:28:33 +0300	[thread overview]
Message-ID: <20150623102833.GJ14570@boom> (raw)
In-Reply-To: <20150618105934.GI24012@nuc-i3427.alporthouse.com>

On Thu, Jun 18, 2015 at 11:59:34AM +0100, Chris Wilson wrote:
> On Thu, Jun 18, 2015 at 01:47:30PM +0300, David Weinehall wrote:
> > On Thu, Jun 18, 2015 at 11:10:13AM +0100, Chris Wilson wrote:
> > > On Thu, Jun 18, 2015 at 12:50:33PM +0300, David Weinehall wrote:
> > > > +static const struct ddi_buf_trans *skl_get_buf_trans_dp(struct drm_device *dev,
> > > 
> > > struct drm_i915_private not struct drm_device!
> > 
> > The device uses both dev and dev_priv; only passing in
> > drm_i915_private wouldn't provide access to dev,
> > or am I missing something?
> 
> Hmm, Oh, you didn't set bits in intel_info for ult/ulx. Instead you have
> large if chains hidden in macros.
> 
> So another task is to translate IS_*_UL? over to using a field in
> intel_info to reduce the code size (at the expense of more intel_info
> data).
> 
> #define IS_SKL_ULT(P) (IS_SKL(P) && INTEL_INFO(P)->is_ult)
> #define IS_SKL_ULX(P) (IS_SKL(P) && INTEL_INFO(P)->is_ulx)

I just followed the pattern of the existing IS_xxx_yyy macros.
Cleanup sounds sensible; this patch merely aims to add a missing
feature.

Kind regards, David
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx