On Fri, Feb 07, 2020 at 12:44:28PM +0100, Andrzej Pietrasiewicz wrote:
Hi All again,
Hi Andrzej,
@malidp and komeda folks: I kindly ask you to have your say. If there is no interest from you (which is ok with me) I will resend the series dropping the komeda and malidp part.
I *am* trying to test your patches, unfortunately my colleagues in China have been kept out of the office for longer than I thought they will, so it is a bit difficult to coordinate.
As a backup, if I don't manage to give you feedback by end of Tuesday, please re-send the patches with malidp and rockchip and skip the komeda ones.
Best regards, Liviu
Andrzej
W dniu 30.01.2020 o 10:08, Andrzej Pietrasiewicz pisze:
Hi All,
A gentle reminder.
Please also see inline:
W dniu 17.12.2019 o 15:49, Andrzej Pietrasiewicz pisze:
This series adds AFBC support for Rockchip. It is inspired by:
https://chromium.googlesource.com/chromiumos/third_party/kernel/+/refs/heads...
This is the fifth iteration of the afbc series. Between v3 and v4 a lot of rework has been done, the main goal of which was to move all afbc-related checks to helpers, so that core does not deal with it.
A new struct drm_afbc_framebuffer is added, which stores afbc-related driver-specific data. Because of that, in drivers that wish to use this feature, the struct must be allocated directly in the driver code rather than inside helpers, so the first portion of the patchset does the necessary refactoring.
Then, there are 3 users of afbc: komeda, malidp and, finally, rockchip, the latter being the ultimate purpose of this work and the 3 subsequent portions of the patchset move komeda and malidp to generic helpers and add afbc support to rockchip.
If changes in komeda and malidp is too much to digest at a time I can focus on rockchip only. This would amount to patches 1-4 and 33-34. After all the ultimate purpose of this work and time spent on it is landing afbc support for rockchip.
Regards,
Andrzej