Skip to contentSkip to conversation

nvfx: fix nv3x fallout from state validation changes

PATCH In review Lucas Stach 2 participants 3 messages Jan 24, 2012 – Jan 30, 2012 1 message not in the archive

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 nvfx: fix nv3x fallout from state validation changes

From c998f732d42da5e962fe5da294493132c3e8dc5f Mon Sep 17 00:00:00 2001
From: Lucas Stach <[email protected]>
Date: Tue, 24 Jan 2012 09:46:32 +0100
Subject: [PATCH] nvfx: fix nv3x fallout from state validation changes

Apparently nv3x needs some curde hacks to work properly. This is clearly not the right fix, but it's the behaviour of the old code and fixes regressions seen by users.

---
 src/gallium/drivers/nvfx/nvfx_state_emit.c |    6 +++++-
 1 files changed, 5 insertions(+), 1 deletions(-)
src/gallium/drivers/nvfx/nvfx_state_emit.c +5 -1
index a959015..60ed2d0 100644
@@ -244,6 +244,9 @@ nvfx_state_validate_common(struct nvfx_context *nvfx)
 	}
 
 	if(dirty & NVFX_NEW_SAMPLER) {
+		/* XXX: we take the big hammer here, I have no idea why this is needed
+		   to make this work properly */
+		nvfx->dirty &= ~NVFX_NEW_SAMPLER;
 		nvfx_fragtex_validate(nvfx);
 
 		// TODO: only set this if really necessary
@@ -303,7 +306,8 @@ nvfx_state_validate_common(struct nvfx_context *nvfx)
 	if(dirty & NVFX_NEW_SR)
 		nvfx_state_sr_validate(nvfx);
 
-	if(dirty & NVFX_NEW_VIEWPORT)
+	/* XXX: nv3x needs viewport revalidation after RAST or ZSA change */
+	if(dirty & (NVFX_NEW_VIEWPORT | NVFX_NEW_RAST | NVFX_NEW_ZSA))
 	{
 		nvfx_state_viewport_validate(nvfx);
 	}
Signature · 1 line
1.7.7.6
#1 lore raw
Quoted 8 lines · Le Tue, 24 Jan 2012 09:54:31 +0100 Lucas Stach…

From c998f732d42da5e962fe5da294493132c3e8dc5f Mon Sep 17 00:00:00 2001
From: Lucas Stach <[email protected]>
Date: Tue, 24 Jan 2012 09:46:32 +0100
Subject: [PATCH] nvfx: fix nv3x fallout from state validation changes

Apparently nv3x needs some curde hacks to work properly. This is clearly not the right fix, but it's the behaviour of the old code and fixes regressions seen by users.

Hello,

Sorry Lucas, but I finally found a test case I have forgotten, that would maybe need reverting your patches.

In mesa-demos/src/demos you'll find fbo_firecube test program that have the scene misrenderered (the scene part that is rendered and used as texture on the cube).

As far as I remember now, render temp stuff was used to handle fbo combinations supported on nv40 but not on nv30, or something like that.

Signature · 6 lines
Patrice Mandin
WWW: http://pmandin.atari.org/
Programmeur Linux, Atari
Spécialité: Développement, jeux

"who writes the code, decides"
#3 lore raw
In reply to a message not in the archive

Am Donnerstag, den 26.01.2012, 20:59 +0100 schrieb Patrice Mandin:

Quoted 17 lines
Quoted 8 lines · Le Tue, 24 Jan 2012 09:54:31 +0100 Lucas Stach <[email protected]> a écrit:

From c998f732d42da5e962fe5da294493132c3e8dc5f Mon Sep 17 00:00:00 2001
From: Lucas Stach <[email protected]>
Date: Tue, 24 Jan 2012 09:46:32 +0100
Subject: [PATCH] nvfx: fix nv3x fallout from state validation changes

Apparently nv3x needs some curde hacks to work properly. This is clearly not the right fix, but it's the behaviour of the old code and fixes regressions seen by users.

Hello,

Sorry Lucas, but I finally found a test case I have forgotten, that would maybe need reverting your patches.

Quoted 7 lines

In mesa-demos/src/demos you'll find fbo_firecube test program that have the scene misrenderered (the scene part that is rendered and used as texture on the cube).

As far as I remember now, render temp stuff was used to handle fbo combinations supported on nv40 but not on nv30, or something like that.

Thanks for the hint. Now that I finally own a nv35 myself, I can reproduce the issue and I'm investigating. I'm not yet sure why it misrenders, normally we should just hit an assert in case of invalid formats.

I will try to debug this, but I can't promise to get results before next week, because FOSDEM wants my attention.

Thanks,
Lucas

_______________________________________________
Nouveau mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/nouveau