Re: [Mesa-dev] [PATCH 0/5] radeon: Write-combined CPU mappings of BOs in GTT

2014-07-23 Thread Alex Deucher
On Tue, Jul 22, 2014 at 11:54 PM, Michel Dänzer wrote: > On 21.07.2014 17:07, Christian König wrote: >> Am 19.07.2014 03:15, schrieb Michel Dänzer: >>> On 19.07.2014 00:47, Christian König wrote: Am 18.07.2014 05:07, schrieb Michel Dänzer: >>> [PATCH 5/5] drm/radeon: Use VRAM for indirect

Re: [Mesa-dev] [PATCH 0/5] radeon: Write-combined CPU mappings of BOs in GTT

2014-07-23 Thread Marek Olšák
On Wed, Jul 23, 2014 at 9:21 AM, Michel Dänzer wrote: > On 23.07.2014 15:42, Christian König wrote: >> Am 23.07.2014 05:54, schrieb Michel Dänzer: >>> On 21.07.2014 17:07, Christian König wrote: Am 19.07.2014 03:15, schrieb Michel Dänzer: > On 19.07.2014 00:47, Christian König wrote:

Re: [Mesa-dev] [PATCH 0/5] radeon: Write-combined CPU mappings of BOs in GTT

2014-07-23 Thread Christian König
Am 23.07.2014 09:21, schrieb Michel Dänzer: On 23.07.2014 15:42, Christian König wrote: Am 23.07.2014 05:54, schrieb Michel Dänzer: On 21.07.2014 17:07, Christian König wrote: Am 19.07.2014 03:15, schrieb Michel Dänzer: On 19.07.2014 00:47, Christian König wrote: Am 18.07.2014 05:07, schrieb

Re: [Mesa-dev] [PATCH 0/5] radeon: Write-combined CPU mappings of BOs in GTT

2014-07-23 Thread Michel Dänzer
On 23.07.2014 15:42, Christian König wrote: > Am 23.07.2014 05:54, schrieb Michel Dänzer: >> On 21.07.2014 17:07, Christian König wrote: >>> Am 19.07.2014 03:15, schrieb Michel Dänzer: On 19.07.2014 00:47, Christian König wrote: > Am 18.07.2014 05:07, schrieb Michel Dänzer: [PATCH

Re: [Mesa-dev] [PATCH 0/5] radeon: Write-combined CPU mappings of BOs in GTT

2014-07-22 Thread Christian König
Am 23.07.2014 05:54, schrieb Michel Dänzer: On 21.07.2014 17:07, Christian König wrote: Am 19.07.2014 03:15, schrieb Michel Dänzer: On 19.07.2014 00:47, Christian König wrote: Am 18.07.2014 05:07, schrieb Michel Dänzer: [PATCH 5/5] drm/radeon: Use VRAM for indirect buffers on >= SI I'm still

Re: [Mesa-dev] [PATCH 0/5] radeon: Write-combined CPU mappings of BOs in GTT

2014-07-22 Thread Michel Dänzer
On 21.07.2014 17:07, Christian König wrote: > Am 19.07.2014 03:15, schrieb Michel Dänzer: >> On 19.07.2014 00:47, Christian König wrote: >>> Am 18.07.2014 05:07, schrieb Michel Dänzer: >> [PATCH 5/5] drm/radeon: Use VRAM for indirect buffers on >= SI > I'm still not very keen with this chan

Re: [Mesa-dev] [PATCH 0/5] radeon: Write-combined CPU mappings of BOs in GTT

2014-07-21 Thread Christian König
Am 19.07.2014 03:15, schrieb Michel Dänzer: On 19.07.2014 00:47, Christian König wrote: Am 18.07.2014 05:07, schrieb Michel Dänzer: [PATCH 5/5] drm/radeon: Use VRAM for indirect buffers on >= SI I'm still not very keen with this change since I still don't understand the reason why it's faster

Re: [Mesa-dev] [PATCH 0/5] radeon: Write-combined CPU mappings of BOs in GTT

2014-07-18 Thread Michel Dänzer
On 19.07.2014 00:47, Christian König wrote: > Am 18.07.2014 05:07, schrieb Michel Dänzer: [PATCH 5/5] drm/radeon: Use VRAM for indirect buffers on >= SI >>> I'm still not very keen with this change since I still don't understand >>> the reason why it's faster than with GTT. Definitely needs mo

Re: [Mesa-dev] [PATCH 0/5] radeon: Write-combined CPU mappings of BOs in GTT

2014-07-18 Thread Marek Olšák
On Fri, Jul 18, 2014 at 7:47 PM, Marek Olšák wrote: > On Fri, Jul 18, 2014 at 5:47 PM, Christian König > wrote: >> Am 18.07.2014 05:07, schrieb Michel Dänzer: > > [PATCH 5/5] drm/radeon: Use VRAM for indirect buffers on >= SI I'm still not very keen with this change since I stil

Re: [Mesa-dev] [PATCH 0/5] radeon: Write-combined CPU mappings of BOs in GTT

2014-07-18 Thread Marek Olšák
On Fri, Jul 18, 2014 at 5:47 PM, Christian König wrote: > Am 18.07.2014 05:07, schrieb Michel Dänzer: [PATCH 5/5] drm/radeon: Use VRAM for indirect buffers on >= SI >>> >>> I'm still not very keen with this change since I still don't understand >>> the reason why it's faster than with GT

Re: [Mesa-dev] [PATCH 0/5] radeon: Write-combined CPU mappings of BOs in GTT

2014-07-18 Thread Dieter Nützel
Am 18.07.2014 05:07, schrieb Michel Dänzer: On 17.07.2014 19:09, Christian König wrote: Am 17.07.2014 12:01, schrieb Michel Dänzer: In order to try and improve X(Shm)PutImage performance with glamor, I implemented support for write-combined CPU mappings of BOs in GTT. This did provide a nice s

Re: [Mesa-dev] [PATCH 0/5] radeon: Write-combined CPU mappings of BOs in GTT

2014-07-18 Thread Christian König
Am 18.07.2014 05:07, schrieb Michel Dänzer: [PATCH 5/5] drm/radeon: Use VRAM for indirect buffers on >= SI I'm still not very keen with this change since I still don't understand the reason why it's faster than with GTT. Definitely needs more testing on a wider range of systems. Sure. If anyone

Re: [Mesa-dev] [PATCH 0/5] radeon: Write-combined CPU mappings of BOs in GTT

2014-07-18 Thread Michel Dänzer
On 18.07.2014 12:58, Dieter Nützel wrote: > Am 18.07.2014 05:07, schrieb Michel Dänzer: >> On 17.07.2014 19:09, Christian König wrote: >>> Am 17.07.2014 12:01, schrieb Michel Dänzer: [PATCH 5/5] drm/radeon: Use VRAM for indirect buffers on >= SI >>> >>> I'm still not very keen with this change

Re: [Mesa-dev] [PATCH 0/5] radeon: Write-combined CPU mappings of BOs in GTT

2014-07-17 Thread Michel Dänzer
On 17.07.2014 19:09, Christian König wrote: > Am 17.07.2014 12:01, schrieb Michel Dänzer: >> In order to try and improve X(Shm)PutImage performance with glamor, I >> implemented support for write-combined CPU mappings of BOs in GTT. >> >> This did provide a nice speedup, but to my surprise, using V

Re: [Mesa-dev] [PATCH 0/5] radeon: Write-combined CPU mappings of BOs in GTT

2014-07-17 Thread Alex Deucher
On Thu, Jul 17, 2014 at 6:01 AM, Michel Dänzer wrote: > In order to try and improve X(Shm)PutImage performance with glamor, I > implemented support for write-combined CPU mappings of BOs in GTT. > > This did provide a nice speedup, but to my surprise, using VRAM instead > of write-combined GTT tur

Re: [Mesa-dev] [PATCH 0/5] radeon: Write-combined CPU mappings of BOs in GTT

2014-07-17 Thread Marek Olšák
On Thu, Jul 17, 2014 at 12:01 PM, Michel Dänzer wrote: > Mesa patches: > > [PATCH 1/5] winsys/radeon: Use separate caching buffer managers for > [PATCH 2/5] r600g/radeonsi: Use write-combined CPU mappings of some > [PATCH 3/5] r600g/radeonsi: Prefer VRAM for CPU -> GPU streaming For these 3 patch

Re: [Mesa-dev] [PATCH 0/5] radeon: Write-combined CPU mappings of BOs in GTT

2014-07-17 Thread Christian König
Am 17.07.2014 12:01, schrieb Michel Dänzer: In order to try and improve X(Shm)PutImage performance with glamor, I implemented support for write-combined CPU mappings of BOs in GTT. This did provide a nice speedup, but to my surprise, using VRAM instead of write-combined GTT turned out to be even

[Mesa-dev] [PATCH 0/5] radeon: Write-combined CPU mappings of BOs in GTT

2014-07-17 Thread Michel Dänzer
In order to try and improve X(Shm)PutImage performance with glamor, I implemented support for write-combined CPU mappings of BOs in GTT. This did provide a nice speedup, but to my surprise, using VRAM instead of write-combined GTT turned out to be even faster in general on my Kaveri machine, both