[Intel-gfx] [PATCH] drm/i915: Avoid using mappable space for relocation processing through the CPU

2011-11-29 Thread Chris Wilson
We try to avoid writing the relocations through the uncached GTT, if the buffer is currently in the CPU write domain and so will be flushed out to main memory afterwards anyway. Also on SandyBridge we can safely write to the pages in cacheable memory, so long as the buffer is LLC mapped. In either

[Intel-gfx] [PATCH] drm/i915: Avoid using mappable space for relocation processing through the CPU

2011-12-04 Thread Chris Wilson
We try to avoid writing the relocations through the uncached GTT, if the buffer is currently in the CPU write domain and so will be flushed out to main memory afterwards anyway. Also on SandyBridge we can safely write to the pages in cacheable memory, so long as the buffer is LLC mapped. In either

[Intel-gfx] [PATCH] drm/i915: Avoid using mappable space for relocation processing through the CPU

2012-03-26 Thread Daniel Vetter
From: Chris Wilson We try to avoid writing the relocations through the uncached GTT, if the buffer is currently in the CPU write domain and so will be flushed out to main memory afterwards anyway. Also on SandyBridge we can safely write to the pages in cacheable memory, so long as the buffer is L

Re: [Intel-gfx] [PATCH] drm/i915: Avoid using mappable space for relocation processing through the CPU

2011-11-29 Thread Daniel Vetter
On Tue, Nov 29, 2011 at 03:12:40PM +, Chris Wilson wrote: > We try to avoid writing the relocations through the uncached GTT, if the > buffer is currently in the CPU write domain and so will be flushed out to > main memory afterwards anyway. Also on SandyBridge we can safely write > to the page

Re: [Intel-gfx] [PATCH] drm/i915: Avoid using mappable space for relocation processing through the CPU

2011-11-29 Thread Chris Wilson
On Tue, 29 Nov 2011 16:34:41 +0100, Daniel Vetter wrote: > On Tue, Nov 29, 2011 at 03:12:40PM +, Chris Wilson wrote: > > We try to avoid writing the relocations through the uncached GTT, if the > > buffer is currently in the CPU write domain and so will be flushed out to > > main memory afterw

Re: [Intel-gfx] [PATCH] drm/i915: Avoid using mappable space for relocation processing through the CPU

2011-11-29 Thread Daniel Vetter
On Tue, Nov 29, 2011 at 04:48:15PM +, Chris Wilson wrote: > On Tue, 29 Nov 2011 16:34:41 +0100, Daniel Vetter wrote: > > On Tue, Nov 29, 2011 at 03:12:40PM +, Chris Wilson wrote: > > > We try to avoid writing the relocations through the uncached GTT, if the > > > buffer is currently in the

Re: [Intel-gfx] [PATCH] drm/i915: Avoid using mappable space for relocation processing through the CPU

2011-11-29 Thread Daniel Vetter
On Tue, Nov 29, 2011 at 06:03:53PM +0100, Daniel Vetter wrote: > On Tue, Nov 29, 2011 at 04:48:15PM +, Chris Wilson wrote: > > On Tue, 29 Nov 2011 16:34:41 +0100, Daniel Vetter wrote: > > > On Tue, Nov 29, 2011 at 03:12:40PM +, Chris Wilson wrote: > > > > We try to avoid writing the reloca

Re: [Intel-gfx] [PATCH] drm/i915: Avoid using mappable space for relocation processing through the CPU

2011-11-29 Thread Chris Wilson
On Tue, 29 Nov 2011 18:03:53 +0100, Daniel Vetter wrote: > On Tue, Nov 29, 2011 at 04:48:15PM +, Chris Wilson wrote: > > On Tue, 29 Nov 2011 16:34:41 +0100, Daniel Vetter wrote: > > > On Tue, Nov 29, 2011 at 03:12:40PM +, Chris Wilson wrote: > > > > We try to avoid writing the relocations

Re: [Intel-gfx] [PATCH] drm/i915: Avoid using mappable space for relocation processing through the CPU

2011-11-29 Thread Daniel Vetter
On Tue, Nov 29, 2011 at 06:17:21PM +0100, Daniel Vetter wrote: > On Tue, Nov 29, 2011 at 06:03:53PM +0100, Daniel Vetter wrote: > > On Tue, Nov 29, 2011 at 04:48:15PM +, Chris Wilson wrote: > > > On Tue, 29 Nov 2011 16:34:41 +0100, Daniel Vetter wrote: > > > > On Tue, Nov 29, 2011 at 03:12:40P

Re: [Intel-gfx] [PATCH] drm/i915: Avoid using mappable space for relocation processing through the CPU

2011-11-29 Thread Daniel Vetter
On Tue, Nov 29, 2011 at 03:12:40PM +, Chris Wilson wrote: > We try to avoid writing the relocations through the uncached GTT, if the > buffer is currently in the CPU write domain and so will be flushed out to > main memory afterwards anyway. Also on SandyBridge we can safely write > to the page

Re: [Intel-gfx] [PATCH] drm/i915: Avoid using mappable space for relocation processing through the CPU

2011-11-29 Thread Ben Widawsky
On Tue, Nov 29, 2011 at 03:12:40PM +, Chris Wilson wrote: > We try to avoid writing the relocations through the uncached GTT, if the > buffer is currently in the CPU write domain and so will be flushed out to > main memory afterwards anyway. Also on SandyBridge we can safely write > to the page

Re: [Intel-gfx] [PATCH] drm/i915: Avoid using mappable space for relocation processing through the CPU

2011-12-05 Thread Daniel Vetter
On Sun, Dec 04, 2011 at 12:26:41PM +, Chris Wilson wrote: > We try to avoid writing the relocations through the uncached GTT, if the > buffer is currently in the CPU write domain and so will be flushed out to > main memory afterwards anyway. Also on SandyBridge we can safely write > to the page

Re: [Intel-gfx] [PATCH] drm/i915: Avoid using mappable space for relocation processing through the CPU

2012-03-26 Thread Chris Wilson
On Mon, 26 Mar 2012 10:10:27 +0200, Daniel Vetter wrote: > From: Chris Wilson > > We try to avoid writing the relocations through the uncached GTT, if the > buffer is currently in the CPU write domain and so will be flushed out to > main memory afterwards anyway. Also on SandyBridge we can safe

Re: [Intel-gfx] [PATCH] drm/i915: Avoid using mappable space for relocation processing through the CPU

2012-03-27 Thread Daniel Vetter
On Mon, Mar 26, 2012 at 09:32:52AM +0100, Chris Wilson wrote: > On Mon, 26 Mar 2012 10:10:27 +0200, Daniel Vetter > wrote: > > From: Chris Wilson > > > > We try to avoid writing the relocations through the uncached GTT, if the > > buffer is currently in the CPU write domain and so will be flush