[Intel-gfx] [PATCH] drm/i915: Force 2*96 MHz cdclk on glk/cnl when audio power is enabled

2018-04-29 Thread Abhay Kumar
CDCLK has to be at least twice the BLCK regardless of audio. Audio driver has to probe using this hook and increase the clock even in absence of any display. Signed-off-by: Ville Syrjälä Signed-off-by: Abhay Kumar ---

Re: [Intel-gfx] [PATCH] x86: Mark up large pm4/5 constants with UL

2018-04-29 Thread Kirill A. Shutemov
On Sun, Apr 29, 2018 at 12:48:32PM +0100, Chris Wilson wrote: > To silence sparse while maintaining compatibility with the assembly, use > _UL which conditionally only appends the UL suffix for C code. http://lkml.kernel.org/r/nycvar.yfh.7.76.1804121437350.28...@cbobk.fhfr.pm -- Kirill A.

[Intel-gfx] [PATCH] drm/i915: Wait for PSR exit before checking for vblank evasion for an atomic update

2018-04-29 Thread Tarun Vyas
From: Tarun The PIPEDSL freezes on PSR entry and if PSR hasn't fully exited, then the pipe_update_start call schedules itself out to check back later. On ChromeOS-4.4 kernel, which is fairly up-to-date w.r.t drm/i915 but lags w.r.t core kernel code, hot plugging an

Re: [Intel-gfx] [PATCH 04/17] dma-fence: Allow wait_any_timeout for all fences

2018-04-29 Thread Christian König
Am 27.04.2018 um 08:17 schrieb Daniel Vetter: When this was introduced in commit a519435a96597d8cd96123246fea4ae5a6c90b02 Author: Christian König Date: Tue Oct 20 16:34:16 2015 +0200 dma-buf/fence: add fence_wait_any_timeout function v2 there was a

Re: [Intel-gfx] [PATCH 13/17] drm/radeon: Remove custom dma_fence_ops->wait implementation

2018-04-29 Thread Christian König
NAK, there is a subtitle but major difference: - if (rdev->needs_reset) { - t = -EDEADLK; - break; - } Without that the whole radeon GPU reset code breaks. Regards, Christian. Am 27.04.2018 um 08:17 schrieb Daniel

Re: [Intel-gfx] [PATCH 00/17] dma-fence doc polish and small cleanup

2018-04-29 Thread Christian König
Patches #1-#5 in this series are Reviewed-by: Christian König Regards, Christian. Am 27.04.2018 um 08:17 schrieb Daniel Vetter: Hi all, Somewhat motivated by me looking at the v3d patch I went and dug around in the dma-fence code a bit. Result was a bit of doc

[Intel-gfx] [PATCH] x86: Mark up large pm4/5 constants with UL

2018-04-29 Thread Chris Wilson
To silence sparse while maintaining compatibility with the assembly, use _UL which conditionally only appends the UL suffix for C code. Fixes: a7412546d8cb ("x86/mm: Adjust vmalloc base and size at boot-time") Signed-off-by: Chris Wilson Cc: Kirill A. Shutemov

Re: [Intel-gfx] [PATCH 06/17] drm/amdgpu: Remove unecessary dma_fence_ops

2018-04-29 Thread Christian König
Am 27.04.2018 um 08:17 schrieb Daniel Vetter: dma_fence_default_wait is the default now. Signed-off-by: Daniel Vetter Cc: Alex Deucher Cc: "Christian König" Cc: Monk Liu Cc: pding