Re: [PATCH v4 0/2] OMAP: DSS2: Framework to handle omap version specific DSS features

2010-10-07 Thread Tomi Valkeinen
On Wed, 2010-09-15 at 19:59 +0200, ext Tony Lindgren wrote:
> * arc...@ti.com  [100915 07:00]:
> > From: Archit Taneja 
> > 
> > This is a simple approach to prevent scattered cpu_is_omap checks
> > in DSS2 by bringing all omap version specific DSS features/values to one
> > single place, initialize them and expose a set of functions to DSS2 driver
> > files which return the value/existance of a feature.
> 
> Glad to hear. We should absolutely not use cpu_is_omap checks
> anywhere in the drivers. The drivers should be arch independent.
> Basically any driver using those is broken from Linux point of view.

Yep. This patch set won't accomplish that, but after they have been
moved to one central place, it'll be much easier to remove the
cpu_is_omap calls totally from the driver.

 Tomi


--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html


Re: [PATCH v4 0/2] OMAP: DSS2: Framework to handle omap version specific DSS features

2010-10-07 Thread Tomi Valkeinen
On Wed, 2010-09-15 at 16:09 +0200, ext arc...@ti.com wrote:
> From: Archit Taneja 
> 
> This is a simple approach to prevent scattered cpu_is_omap checks
> in DSS2 by bringing all omap version specific DSS features/values to one
> single place, initialize them and expose a set of functions to DSS2 driver
> files which return the value/existance of a feature.
> 
> This method provides the following:
>   -A set of functions which give the value of a feature.
>   -Functions which add and return the start and end bits of a register field
>which varies across omaps.
>   -A function which checks if a DSS2 feature exists or not on the present
>omap (omap_dss_has_feature)
> 
> Whenever a new DSS feature is introduced, its value/existance is specified
> for all omaps in omap_dss_feature_init().
> 
> This patch also fills up a minimal list of features in 
> omap_dss_features_init().
> The second patch in the series makes use of dss_features to remove some of the
> existing omap checks.

Thanks, applied to DSS2 tree.

 Tomi


--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html


Re: [PATCH v4 0/2] OMAP: DSS2: Framework to handle omap version specific DSS features

2010-09-15 Thread Tony Lindgren
* arc...@ti.com  [100915 07:00]:
> From: Archit Taneja 
> 
> This is a simple approach to prevent scattered cpu_is_omap checks
> in DSS2 by bringing all omap version specific DSS features/values to one
> single place, initialize them and expose a set of functions to DSS2 driver
> files which return the value/existance of a feature.

Glad to hear. We should absolutely not use cpu_is_omap checks
anywhere in the drivers. The drivers should be arch independent.
Basically any driver using those is broken from Linux point of view.

The current mess is:

$ grep -r cpu_is_omap drivers/ | wc -l
139

The right way to replace those is to pass feature flags from
the platform_data.

Tony
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html


[PATCH v4 0/2] OMAP: DSS2: Framework to handle omap version specific DSS features

2010-09-15 Thread arc...@ti.com
From: Archit Taneja 

This is a simple approach to prevent scattered cpu_is_omap checks
in DSS2 by bringing all omap version specific DSS features/values to one
single place, initialize them and expose a set of functions to DSS2 driver
files which return the value/existance of a feature.

This method provides the following:
  -A set of functions which give the value of a feature.
  -Functions which add and return the start and end bits of a register field
   which varies across omaps.
  -A function which checks if a DSS2 feature exists or not on the present
   omap (omap_dss_has_feature)

Whenever a new DSS feature is introduced, its value/existance is specified
for all omaps in omap_dss_feature_init().

This patch also fills up a minimal list of features in omap_dss_features_init().
The second patch in the series makes use of dss_features to remove some of the
existing omap checks.

Archit Taneja (2):
  OMAP: DSS2: Introduce dss_features files
  OMAP: DSS2: Use dss_features framework on DSS2 code

 arch/arm/plat-omap/include/plat/display.h |   31 -
 drivers/video/omap2/dss/Makefile  |2 +-
 drivers/video/omap2/dss/core.c|3 +
 drivers/video/omap2/dss/dispc.c   |   56 +
 drivers/video/omap2/dss/dss_features.c|  191 +
 drivers/video/omap2/dss/dss_features.h|   50 
 drivers/video/omap2/dss/manager.c |   33 +++---
 drivers/video/omap2/dss/overlay.c |   24 ++--
 8 files changed, 302 insertions(+), 88 deletions(-)
 create mode 100644 drivers/video/omap2/dss/dss_features.c
 create mode 100644 drivers/video/omap2/dss/dss_features.h
--
Version 4:
 - all dss_features members are made const, better naming of some
   vaiables
Version 3:
 - Naming of register fields enums is changed, dss_features struct
   has some members changed as pointers to static arrays.
Version 2:
 - Features are initialized statically as opposed to runtime before,
   function naming for dss_features files has changed.
Version 1:
https://patchwork.kernel.org/patch/134431/
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html