[PATCH 0/5] Fixes for MST (daisy-chain and 4k tiles)

2016-02-04 Thread Dave Airlie
On 3 February 2016 at 07:20, Alex Deucher  wrote:
> On Mon, Jan 25, 2016 at 1:55 PM, Wentland, Harry  
> wrote:
>> Hi Dave,
>>
>> I've been running with lockdep with these changes for over a week now. Just 
>> ran another test with our daisy-chain displays and the 4k tiled display with 
>> no deadlocks or lockdep prints (other than "RCU lockdep checking is 
>> enabled.").
>>
>
> Dave, do you want to pick these up directly or should I pull them into
> my -fixes tree?
>
I'm going to pull them in myself for this round I think, there are a
few other MST bits around I want to test before pushing.

Dave.


[PATCH 0/5] Fixes for MST (daisy-chain and 4k tiles)

2016-02-02 Thread Alex Deucher
On Mon, Jan 25, 2016 at 1:55 PM, Wentland, Harry  
wrote:
> Hi Dave,
>
> I've been running with lockdep with these changes for over a week now. Just 
> ran another test with our daisy-chain displays and the 4k tiled display with 
> no deadlocks or lockdep prints (other than "RCU lockdep checking is 
> enabled.").
>

Dave, do you want to pick these up directly or should I pull them into
my -fixes tree?

Alex

> Thanks,
> Harry
>
> 
> From: Dave Airlie 
> Sent: Sunday, January 24, 2016 11:53 PM
> To: Wentland, Harry
> Cc: dri-devel
> Subject: Re: [PATCH 0/5] Fixes for MST (daisy-chain and 4k tiles)
>
> On 23 January 2016 at 08:07, Harry Wentland  wrote:
>> A couple of MST fixes to bugs in the framework that we encountered when
>> testing with
>>   - three-display daisy-chain configurations
>>   - 4k tiled displays
>
> Hi Harry,
>
> these all look pretty good to me, have you tested them with lockdep enabled?
>
> I had started to fix the CSN/GUID one here a little while back but it's 
> sitting
> on a laptop I can't access at the moment.
>
> But if you can test with lockdep, and it doesn't deadlock, then I'm happy
> to accept these..
>
> Reviewed-by: Dave Airlie 
>
> Dave.
>>
>> Andrey Grodzovsky (1):
>>   drm/dp/mst: Reverse order of MST enable and clearing VC payload table.
>>
>> Harry Wentland (2):
>>   drm: Add drm_fixp_from_fraction and drm_fixp2int_ceil
>>   drm/dp/mst: Calculate MST PBN with 31.32 fixed point
>>
>> Hersen Wu (1):
>>   drm/dp/mst: move GUID storage from mgr, port to only mst branch
>>
>> Mykola Lysenko (1):
>>   drm/dp/mst: change MST detection scheme
>>
>>  drivers/gpu/drm/drm_dp_mst_topology.c | 180 
>> +-
>>  include/drm/drm_dp_mst_helper.h   |  25 ++---
>>  include/drm/drm_fixed.h   |  54 +-
>>  3 files changed, 154 insertions(+), 105 deletions(-)
>>
>> --
>> 2.1.4
>>
>> ___
>> dri-devel mailing list
>> dri-devel at lists.freedesktop.org
>> http://lists.freedesktop.org/mailman/listinfo/dri-devel
> ___
> dri-devel mailing list
> dri-devel at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/dri-devel


[PATCH 0/5] Fixes for MST (daisy-chain and 4k tiles)

2016-01-25 Thread Wentland, Harry
Hi Dave,

I've been running with lockdep with these changes for over a week now. Just ran 
another test with our daisy-chain displays and the 4k tiled display with no 
deadlocks or lockdep prints (other than "RCU lockdep checking is enabled.").

Thanks,
Harry


From: Dave Airlie <airl...@gmail.com>
Sent: Sunday, January 24, 2016 11:53 PM
To: Wentland, Harry
Cc: dri-devel
Subject: Re: [PATCH 0/5] Fixes for MST (daisy-chain and 4k tiles)

On 23 January 2016 at 08:07, Harry Wentland  wrote:
> A couple of MST fixes to bugs in the framework that we encountered when
> testing with
>   - three-display daisy-chain configurations
>   - 4k tiled displays

Hi Harry,

these all look pretty good to me, have you tested them with lockdep enabled?

I had started to fix the CSN/GUID one here a little while back but it's sitting
on a laptop I can't access at the moment.

But if you can test with lockdep, and it doesn't deadlock, then I'm happy
to accept these..

Reviewed-by: Dave Airlie 

Dave.
>
> Andrey Grodzovsky (1):
>   drm/dp/mst: Reverse order of MST enable and clearing VC payload table.
>
> Harry Wentland (2):
>   drm: Add drm_fixp_from_fraction and drm_fixp2int_ceil
>   drm/dp/mst: Calculate MST PBN with 31.32 fixed point
>
> Hersen Wu (1):
>   drm/dp/mst: move GUID storage from mgr, port to only mst branch
>
> Mykola Lysenko (1):
>   drm/dp/mst: change MST detection scheme
>
>  drivers/gpu/drm/drm_dp_mst_topology.c | 180 
> +-
>  include/drm/drm_dp_mst_helper.h   |  25 ++---
>  include/drm/drm_fixed.h   |  54 +-
>  3 files changed, 154 insertions(+), 105 deletions(-)
>
> --
> 2.1.4
>
> ___
> dri-devel mailing list
> dri-devel at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/dri-devel


[PATCH 0/5] Fixes for MST (daisy-chain and 4k tiles)

2016-01-25 Thread Dave Airlie
On 23 January 2016 at 08:07, Harry Wentland  wrote:
> A couple of MST fixes to bugs in the framework that we encountered when
> testing with
>   - three-display daisy-chain configurations
>   - 4k tiled displays

Hi Harry,

these all look pretty good to me, have you tested them with lockdep enabled?

I had started to fix the CSN/GUID one here a little while back but it's sitting
on a laptop I can't access at the moment.

But if you can test with lockdep, and it doesn't deadlock, then I'm happy
to accept these..

Reviewed-by: Dave Airlie 

Dave.
>
> Andrey Grodzovsky (1):
>   drm/dp/mst: Reverse order of MST enable and clearing VC payload table.
>
> Harry Wentland (2):
>   drm: Add drm_fixp_from_fraction and drm_fixp2int_ceil
>   drm/dp/mst: Calculate MST PBN with 31.32 fixed point
>
> Hersen Wu (1):
>   drm/dp/mst: move GUID storage from mgr, port to only mst branch
>
> Mykola Lysenko (1):
>   drm/dp/mst: change MST detection scheme
>
>  drivers/gpu/drm/drm_dp_mst_topology.c | 180 
> +-
>  include/drm/drm_dp_mst_helper.h   |  25 ++---
>  include/drm/drm_fixed.h   |  54 +-
>  3 files changed, 154 insertions(+), 105 deletions(-)
>
> --
> 2.1.4
>
> ___
> dri-devel mailing list
> dri-devel at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/dri-devel


[PATCH 0/5] Fixes for MST (daisy-chain and 4k tiles)

2016-01-22 Thread Harry Wentland
A couple of MST fixes to bugs in the framework that we encountered when
testing with 
  - three-display daisy-chain configurations
  - 4k tiled displays

Andrey Grodzovsky (1):
  drm/dp/mst: Reverse order of MST enable and clearing VC payload table.

Harry Wentland (2):
  drm: Add drm_fixp_from_fraction and drm_fixp2int_ceil
  drm/dp/mst: Calculate MST PBN with 31.32 fixed point

Hersen Wu (1):
  drm/dp/mst: move GUID storage from mgr, port to only mst branch

Mykola Lysenko (1):
  drm/dp/mst: change MST detection scheme

 drivers/gpu/drm/drm_dp_mst_topology.c | 180 +-
 include/drm/drm_dp_mst_helper.h   |  25 ++---
 include/drm/drm_fixed.h   |  54 +-
 3 files changed, 154 insertions(+), 105 deletions(-)

-- 
2.1.4