Skip to contentSkip to conversation

lib/rendercopy_gen9: Setup Push constant pointer before sending BTP commands

PATCH In review Arun Siluvery 5 participants 9 messages Aug 12, 2015 – Aug 19, 2015 Library code

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 lib/rendercopy_gen9: Setup Push constant pointer before sending BTP commands

From Gen9, by default push constant command is not committed to the shader unit untill the corresponding shader's BTP_* command is parsed. This is the behaviour when set shader is enabled. This patch updates the batch to follow this requirement otherwise it results in gpu hang.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89959

Set shader need to be disabled if legacy behaviour is required.

Cc Ben Widawsky <[email protected]>
Cc Joonas Lahtinen <[email protected]>
Cc Mika Kuoppala <[email protected]>

---

If this patch is applied then we don't need to disable set shader bit in kernel and the following can be discarded.

http://lists.freedesktop.org/archives/intel-gfx/2015-August/073425.html

 lib/rendercopy_gen9.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)
lib/rendercopy_gen9.c +7 -7
index 4a4a604..9537480 100644
@@ -590,13 +590,9 @@ gen8_emit_multisample(struct intel_batchbuffer *batch) {
 
 static void
 gen8_emit_vs(struct intel_batchbuffer *batch) {
-	OUT_BATCH(GEN7_3DSTATE_BINDING_TABLE_POINTERS_VS);
+	OUT_BATCH(GEN6_3DSTATE_CONSTANT_VS | (11-2));
 	OUT_BATCH(0);
-
-	OUT_BATCH(GEN7_3DSTATE_SAMPLER_STATE_POINTERS_VS);
 	OUT_BATCH(0);
-
-	OUT_BATCH(GEN6_3DSTATE_CONSTANT_VS | (11-2));
 	OUT_BATCH(0);
 	OUT_BATCH(0);
 	OUT_BATCH(0);
@@ -605,7 +601,11 @@ gen8_emit_vs(struct intel_batchbuffer *batch) {
 	OUT_BATCH(0);
 	OUT_BATCH(0);
 	OUT_BATCH(0);
+
+	OUT_BATCH(GEN7_3DSTATE_BINDING_TABLE_POINTERS_VS);
 	OUT_BATCH(0);
+
+	OUT_BATCH(GEN7_3DSTATE_SAMPLER_STATE_POINTERS_VS);
 	OUT_BATCH(0);
 
 	OUT_BATCH(GEN6_3DSTATE_VS | (9-2));
@@ -998,14 +998,14 @@ void gen9_render_copyfunc(struct intel_batchbuffer *batch,
 
 	gen8_emit_sf(batch);
 
+	gen8_emit_ps(batch, ps_kernel_off);
+
 	OUT_BATCH(GEN7_3DSTATE_BINDING_TABLE_POINTERS_PS);
 	OUT_BATCH(ps_binding_table);
 
 	OUT_BATCH(GEN7_3DSTATE_SAMPLER_STATE_POINTERS_PS);
 	OUT_BATCH(ps_sampler_state);
 
-	gen8_emit_ps(batch, ps_kernel_off);
-
 	OUT_BATCH(GEN6_3DSTATE_SCISSOR_STATE_POINTERS);
 	OUT_BATCH(scissor_state);
 
Signature · 6 lines
1.9.1

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Reviewed-by on v1 1/1
#1 lore raw
Quoted 8 lines · Arun Siluvery

From Gen9, by default push constant command is not committed to the shader unit
untill the corresponding shader's BTP_* command is parsed. This is the
behaviour when set shader is enabled. This patch updates the batch to follow
this requirement otherwise it results in gpu hang.

Quoted 8 lines

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89959

Set shader need to be disabled if legacy behaviour is required.

Cc Ben Widawsky <[email protected]>
Cc Joonas Lahtinen <[email protected]>
Cc Mika Kuoppala <[email protected]>
Quoted 51 lines

---

If this patch is applied then we don't need to disable set shader bit in kernel and the following can be discarded.

http://lists.freedesktop.org/archives/intel-gfx/2015-August/073425.ht
ml

 lib/rendercopy_gen9.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)
lib/rendercopy_gen9.c +1 -5
index 4a4a604..9537480 100644
@@ -590,13 +590,9 @@ gen8_emit_multisample(struct intel_batchbuffer 
 *batch) {
 
 static void
 gen8_emit_vs(struct intel_batchbuffer *batch) {
-	OUT_BATCH(GEN7_3DSTATE_BINDING_TABLE_POINTERS_VS);
+	OUT_BATCH(GEN6_3DSTATE_CONSTANT_VS | (11-2));
 	OUT_BATCH(0);
-
-	OUT_BATCH(GEN7_3DSTATE_SAMPLER_STATE_POINTERS_VS);
 	OUT_BATCH(0);
-
-	OUT_BATCH(GEN6_3DSTATE_CONSTANT_VS | (11-2));
 	OUT_BATCH(0);
 	OUT_BATCH(0);
lib/rendercopy_gen9.c +6 -0
 	OUT_BATCH(0);
@@ -605,7 +601,11 @@ gen8_emit_vs(struct intel_batchbuffer *batch) {
 	OUT_BATCH(0);
 	OUT_BATCH(0);
 	OUT_BATCH(0);
+
+	OUT_BATCH(GEN7_3DSTATE_BINDING_TABLE_POINTERS_VS);
 	OUT_BATCH(0);
+
+	OUT_BATCH(GEN7_3DSTATE_SAMPLER_STATE_POINTERS_VS);
 	OUT_BATCH(0);
 
 	OUT_BATCH(GEN6_3DSTATE_VS | (9-2));
@@ -998,14 +998,14 @@ void gen9_render_copyfunc(struct 
 intel_batchbuffer *batch,
 
 	gen8_emit_sf(batch);
 
+	gen8_emit_ps(batch, ps_kernel_off);
+

lib/rendercopy_gen9.c void gen9_render_copyfunc(struct · line 1011

Quoted 10 lines
lib/rendercopy_gen9.c +0 -2
 	OUT_BATCH(GEN7_3DSTATE_BINDING_TABLE_POINTERS_PS);
 	OUT_BATCH(ps_binding_table);
 
 	OUT_BATCH(GEN7_3DSTATE_SAMPLER_STATE_POINTERS_PS);
 	OUT_BATCH(ps_sampler_state);
 
-	gen8_emit_ps(batch, ps_kernel_off);
-
 	OUT_BATCH(GEN6_3DSTATE_SCISSOR_STATE_POINTERS);
 	OUT_BATCH(scissor_state);

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

#2 lore raw
Quoted 11 lines · Joonas Lahtinen
Quoted 10 lines · Arun Siluvery

From Gen9, by default push constant command is not committed to the shader unit
untill the corresponding shader's BTP_* command is parsed. This is the
behaviour when set shader is enabled. This patch updates the batch to follow
this requirement otherwise it results in gpu hang.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89959

Quoted 9 lines
Quoted 6 lines

Set shader need to be disabled if legacy behaviour is required.

Cc Ben Widawsky <[email protected]>
Cc Joonas Lahtinen <[email protected]>
Cc Mika Kuoppala <[email protected]>

Repeating what I said on the mesa thread:
Does anyone understand why this actually causes a hang on the IGT test? I certainly don't. The docs are pretty clear that the constant command is not committed until the BTP command, but I can't make any sense of how it related to a GPU hang.

[snip]

---
Ben Widawsky, Intel Open Source Technology Center
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

#3 lore raw

Hi,

Quoted 25 lines · Ben Widawsky
Quoted 14 lines · Joonas Lahtinen
Quoted 13 lines · Arun Siluvery

From Gen9, by default push constant command is not committed to the
shader unit
untill the corresponding shader's BTP_* command is parsed. This is
the
behaviour when set shader is enabled. This patch updates the batch to
follow
this requirement otherwise it results in gpu hang.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89959

Quoted 9 lines
Quoted 6 lines

Set shader need to be disabled if legacy behaviour is required.

Cc Ben Widawsky <[email protected]>
Cc Joonas Lahtinen <[email protected]>
Cc Mika Kuoppala <[email protected]>
Quoted 9 lines

Repeating what I said on the mesa thread:
Does anyone understand why this actually causes a hang on the IGT test? I
certainly don't. The docs are pretty clear that the constant command is not
committed until the BTP command, but I can't make any sense of how it related to
a GPU hang.

Changing the order makes the hang go away and come back for sure, we've all been experiencing that. System validation said it is a programming restriction, so I'm not sure how relevant it is what goes wrong if it's not followed. And the legacy mode bits were added to support the old behavior of having the order like it has been previously, so I do not see why question it without visibility to the actual RTL. And enabling the legacy bits makes the hang go away, too.

If I had the RTL sources, then it would be more relevant to take educated guesses as to why a set of hundreds of thousands of transistors doesn't work as it should :) Without that, if it gets stuck, it gets stuck.

Regards, Joonas

Quoted 4 lines

[snip]

---
Ben Widawsky, Intel Open Source Technology Center

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

#4 lore raw
Quoted 46 lines · Joonas Lahtinen

Hi,

Quoted 25 lines · Ben Widawsky
Quoted 14 lines · Joonas Lahtinen
Quoted 13 lines · Arun Siluvery

From Gen9, by default push constant command is not committed to the
shader unit
untill the corresponding shader's BTP_* command is parsed. This is
the
behaviour when set shader is enabled. This patch updates the batch to
follow
this requirement otherwise it results in gpu hang.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89959

Quoted 9 lines
Quoted 6 lines

Set shader need to be disabled if legacy behaviour is required.

Cc Ben Widawsky <[email protected]>
Cc Joonas Lahtinen <[email protected]>
Cc Mika Kuoppala <[email protected]>
Quoted 9 lines

Repeating what I said on the mesa thread:
Does anyone understand why this actually causes a hang on the IGT test? I
certainly don't. The docs are pretty clear that the constant command is not
committed until the BTP command, but I can't make any sense of how it related to
a GPU hang.

Changing the order makes the hang go away and come back for sure, we've all been experiencing that. System validation said it is a programming restriction, so I'm not sure how relevant it is what goes wrong if it's not followed. And the legacy mode bits were added to support the old behavior of having the order like it has been previously, so I do not see why question it without visibility to the actual RTL. And enabling the legacy bits makes the hang go away, too.

Quoted 7 lines

If I had the RTL sources, then it would be more relevant to take educated guesses as to why a set of hundreds of thousands of transistors doesn't work as it should :) Without that, if it gets stuck, it gets stuck.

Regards, Joonas

Let me start by saying I do believe that questioning this shouldn't prevent merging the patch.

<rant>
I absolutely disagree with you and think we should question these kind of things and get out of the mindset of, "well, it fixes a hang, let's move on." Understanding these kind of things is critical to writing stable drivers. If the programming guide/SV team said it can lead to a hang, that's one thing, but AFAICT, we do not understand why it is hanging nor does any of the documentation we do have suggest it should hang. Without clarification, next time we have a similar hang signature we're going to be right back here where we started.

It was one thing when there were a handful of us working on the stuff and we didn't have time to get to the bottom of bugs like this. I'm guilty of patches like this myself. I really do not see any excuse for this any more though. </rant>

Could you send me the reference for where SV said it was a "programming restriction"? To me it all sounds very much like an implementation detail, and I'd like to try to understand what I am missing.

Quoted 4 lines
Quoted 4 lines

[snip]

---
Ben Widawsky, Intel Open Source Technology Center

Signature · 5 lines
Ben Widawsky, Intel Open Source Technology Center
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
#5 lore raw
Quoted 67 lines · Ben Widawsky
Quoted 46 lines · Joonas Lahtinen

Hi,

Quoted 25 lines · Ben Widawsky
Quoted 14 lines · Joonas Lahtinen
Quoted 13 lines · Arun Siluvery

From Gen9, by default push constant command is not committed to the
shader unit
untill the corresponding shader's BTP_* command is parsed. This is
the
behaviour when set shader is enabled. This patch updates the batch to
follow
this requirement otherwise it results in gpu hang.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89959

Quoted 9 lines
Quoted 6 lines

Set shader need to be disabled if legacy behaviour is required.

Cc Ben Widawsky <[email protected]>
Cc Joonas Lahtinen <[email protected]>
Cc Mika Kuoppala <[email protected]>
Quoted 9 lines

Repeating what I said on the mesa thread:
Does anyone understand why this actually causes a hang on the IGT test? I
certainly don't. The docs are pretty clear that the constant command is not
committed until the BTP command, but I can't make any sense of how it related to
a GPU hang.

Changing the order makes the hang go away and come back for sure, we've all been experiencing that. System validation said it is a programming restriction, so I'm not sure how relevant it is what goes wrong if it's not followed. And the legacy mode bits were added to support the old behavior of having the order like it has been previously, so I do not see why question it without visibility to the actual RTL. And enabling the legacy bits makes the hang go away, too.

Quoted 7 lines

If I had the RTL sources, then it would be more relevant to take educated guesses as to why a set of hundreds of thousands of transistors doesn't work as it should :) Without that, if it gets stuck, it gets stuck.

Regards, Joonas

Let me start by saying I do believe that questioning this shouldn't prevent merging the patch.

<rant>
I absolutely disagree with you and think we should question these kind of things and get out of the mindset of, "well, it fixes a hang, let's move on." Understanding these kind of things is critical to writing stable drivers. If the programming guide/SV team said it can lead to a hang, that's one thing, but AFAICT, we do not understand why it is hanging nor does any of the documentation we do have suggest it should hang. Without clarification, next time we have a similar hang signature we're going to be right back here where we started.

Quoted 8 lines

It was one thing when there were a handful of us working on the stuff and we didn't have time to get to the bottom of bugs like this. I'm guilty of patches like this myself. I really do not see any excuse for this any more though. </rant>

Could you send me the reference for where SV said it was a "programming restriction"? To me it all sounds very much like an implementation detail, and I'd like to try to understand what I am missing.

Fully agree, we can't just ignore hangs but have to bottom out on them. And in the past we've had piles of examples where we discovered something and didn't chase it correctly, then a few months later massive panic in intel. At least this must be reflected in bspec.

Joonas, can you please work together with Ben to chase this down? If you don't get traction please escalate the shit out of this, we really must at least get docs to accurately reflect reality.

Thanks, Daniel

Signature · 7 lines
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
#6 lore raw

Hi,

Quoted 115 lines · Daniel Vetter
Quoted 100 lines · Ben Widawsky
Quoted 63 lines · Joonas Lahtinen

Hi,

Quoted 33 lines · Ben Widawsky
Quoted 21 lines · Joonas Lahtinen
Quoted 20 lines · Arun Siluvery

From Gen9, by default push constant command is not committed to
the
shader unit
untill the corresponding shader's BTP_* command is parsed. This
is
the
behaviour when set shader is enabled. This patch updates the
batch to
follow
this requirement otherwise it results in gpu hang.

Bugzilla:
https://bugs.freedesktop.org/show_bug.cgi?id=89959

Set shader need to be disabled if legacy behaviour is required.

Quoted 9 lines
Quoted 5 lines
Cc Ben Widawsky <[email protected]>
Cc Joonas Lahtinen <[email protected]>
Cc Mika Kuoppala <[email protected]>
Quoted 1 line
Quoted 1 line
Quoted 12 lines

Repeating what I said on the mesa thread:
Does anyone understand why this actually causes a hang on the IGT
test? I
certainly don't. The docs are pretty clear that the constant command
is not
committed until the BTP command, but I can't make any sense of how it
related to
a GPU hang.

Changing the order makes the hang go away and come back for sure, we've
all been experiencing that. System validation said it is a programming
restriction, so I'm not sure how relevant it is what goes wrong if it's
not followed. And the legacy mode bits were added to support the old
behavior of having the order like it has been previously, so I do not
see why question it without visibility to the actual RTL. And enabling
the legacy bits makes the hang go away, too.

Quoted 7 lines

If I had the RTL sources, then it would be more relevant to take educated guesses as to why a set of hundreds of thousands of transistors doesn't work as it should :) Without that, if it gets stuck, it gets stuck.

Regards, Joonas

Let me start by saying I do believe that questioning this shouldn't prevent
merging the patch.

<rant>
I absolutely disagree with you and think we should question these kind of things
and get out of the mindset of, "well, it fixes a hang, let's move on."
Understanding these kind of things is critical to writing stable drivers. If
the programming guide/SV team said it can lead to a hang, that's one thing, but
AFAICT, we do not understand why it is hanging nor does any of the documentation
we do have suggest it should hang. Without clarification, next time we have a
similar hang signature we're going to be right back here where we started.

It was one thing when there were a handful of us working on the stuff and we
didn't have time to get to the bottom of bugs like this. I'm guilty of patches
like this myself. I really do not see any excuse for this any more though.
</rant>

Quoted 5 lines

Could you send me the reference for where SV said it was a "programming
restriction"? To me it all sounds very much like an implementation detail, and
I'd like to try to understand what I am missing.

Fully agree, we can't just ignore hangs but have to bottom out on them.
And in the past we've had piles of examples where we discovered something
and didn't chase it correctly, then a few months later massive panic in
intel. At least this must be reflected in bspec.

Quoted 6 lines

Joonas, can you please work together with Ben to chase this down? If you
don't get traction please escalate the shit out of this, we really must
at least get docs to accurately reflect reality.

The latest reply from SV team to Arun's inquiry is that when the order is not followed, the previous push constant values get used (which will be all zeroes, so doesn't really make sense) and would lead to the hang or corruption. I'm checking them if the same applies for the BTP itself too, where the PS pointer keeps changing (and it's the only changing thing), and could explain it better.

Regards, Joonas

Quoted 1 line

Thanks, Daniel

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Reviewed-by
#7 lore raw
Quoted 125 lines · Joonas Lahtinen

Hi,

Quoted 115 lines · Daniel Vetter
Quoted 100 lines · Ben Widawsky
Quoted 63 lines · Joonas Lahtinen

Hi,

Quoted 33 lines · Ben Widawsky
Quoted 21 lines · Joonas Lahtinen
Quoted 20 lines · Arun Siluvery

From Gen9, by default push constant command is not committed to
the
shader unit
untill the corresponding shader's BTP_* command is parsed. This
is
the
behaviour when set shader is enabled. This patch updates the
batch to
follow
this requirement otherwise it results in gpu hang.

Bugzilla:
https://bugs.freedesktop.org/show_bug.cgi?id=89959

Set shader need to be disabled if legacy behaviour is required.

Quoted 9 lines
Quoted 5 lines
Cc Ben Widawsky <[email protected]>
Cc Joonas Lahtinen <[email protected]>
Cc Mika Kuoppala <[email protected]>
Quoted 1 line
Quoted 1 line
Quoted 12 lines

Repeating what I said on the mesa thread:
Does anyone understand why this actually causes a hang on the IGT
test? I
certainly don't. The docs are pretty clear that the constant command
is not
committed until the BTP command, but I can't make any sense of how it
related to
a GPU hang.

Changing the order makes the hang go away and come back for sure, we've
all been experiencing that. System validation said it is a programming
restriction, so I'm not sure how relevant it is what goes wrong if it's
not followed. And the legacy mode bits were added to support the old
behavior of having the order like it has been previously, so I do not
see why question it without visibility to the actual RTL. And enabling
the legacy bits makes the hang go away, too.

Quoted 7 lines

If I had the RTL sources, then it would be more relevant to take educated guesses as to why a set of hundreds of thousands of transistors doesn't work as it should :) Without that, if it gets stuck, it gets stuck.

Regards, Joonas

Let me start by saying I do believe that questioning this shouldn't prevent
merging the patch.

<rant>
I absolutely disagree with you and think we should question these kind of things
and get out of the mindset of, "well, it fixes a hang, let's move on."
Understanding these kind of things is critical to writing stable drivers. If
the programming guide/SV team said it can lead to a hang, that's one thing, but
AFAICT, we do not understand why it is hanging nor does any of the documentation
we do have suggest it should hang. Without clarification, next time we have a
similar hang signature we're going to be right back here where we started.

It was one thing when there were a handful of us working on the stuff and we
didn't have time to get to the bottom of bugs like this. I'm guilty of patches
like this myself. I really do not see any excuse for this any more though.
</rant>

Quoted 5 lines

Could you send me the reference for where SV said it was a "programming
restriction"? To me it all sounds very much like an implementation detail, and
I'd like to try to understand what I am missing.

Fully agree, we can't just ignore hangs but have to bottom out on them.
And in the past we've had piles of examples where we discovered something
and didn't chase it correctly, then a few months later massive panic in
intel. At least this must be reflected in bspec.

Quoted 6 lines

Joonas, can you please work together with Ben to chase this down? If you
don't get traction please escalate the shit out of this, we really must
at least get docs to accurately reflect reality.

Quoted 8 lines

The latest reply from SV team to Arun's inquiry is that when the order is not followed, the previous push constant values get used (which will be all zeroes, so doesn't really make sense) and would lead to the hang or corruption. I'm checking them if the same applies for the BTP itself too, where the PS pointer keeps changing (and it's the only changing thing), and could explain it better.

Regards, Joonas

Cool. It seems like an interesting conversation. I managed to convince myself corruption can happen, maybe - but not a hang. Now that you've kicked off that conversation, I'm fine with merging the patch (we've done our "due diligence" /me shudders).

Quoted 2 lines
Quoted 2 lines

Thanks, Daniel

Signature · 5 lines
Ben Widawsky, Intel Open Source Technology Center
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Tested-by on v1 1/1
#8 lore raw
In reply to Arun Siluvery
Quoted 7 lines · Arun Siluvery

From Gen9, by default push constant command is not committed to the shader unit untill the corresponding shader's BTP_* command is parsed. This is the behaviour when set shader is enabled. This patch updates the batch to follow this requirement otherwise it results in gpu hang.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89959

Quoted 8 lines

Set shader need to be disabled if legacy behaviour is required.

Cc Ben Widawsky <[email protected]>
Cc Joonas Lahtinen <[email protected]>
Cc Mika Kuoppala <[email protected]>

---

with skl-y,

Quoted 63 lines

If this patch is applied then we don't need to disable set shader bit in kernel and the following can be discarded.

http://lists.freedesktop.org/archives/intel-gfx/2015-August/073425.html

 lib/rendercopy_gen9.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)
lib/rendercopy_gen9.c +7 -7
index 4a4a604..9537480 100644
@@ -590,13 +590,9 @@ gen8_emit_multisample(struct intel_batchbuffer *batch) {
 
 static void
 gen8_emit_vs(struct intel_batchbuffer *batch) {
-	OUT_BATCH(GEN7_3DSTATE_BINDING_TABLE_POINTERS_VS);
+	OUT_BATCH(GEN6_3DSTATE_CONSTANT_VS | (11-2));
 	OUT_BATCH(0);
-
-	OUT_BATCH(GEN7_3DSTATE_SAMPLER_STATE_POINTERS_VS);
 	OUT_BATCH(0);
-
-	OUT_BATCH(GEN6_3DSTATE_CONSTANT_VS | (11-2));
 	OUT_BATCH(0);
 	OUT_BATCH(0);
 	OUT_BATCH(0);
@@ -605,7 +601,11 @@ gen8_emit_vs(struct intel_batchbuffer *batch) {
 	OUT_BATCH(0);
 	OUT_BATCH(0);
 	OUT_BATCH(0);
+
+	OUT_BATCH(GEN7_3DSTATE_BINDING_TABLE_POINTERS_VS);
 	OUT_BATCH(0);
+
+	OUT_BATCH(GEN7_3DSTATE_SAMPLER_STATE_POINTERS_VS);
 	OUT_BATCH(0);
 
 	OUT_BATCH(GEN6_3DSTATE_VS | (9-2));
@@ -998,14 +998,14 @@ void gen9_render_copyfunc(struct intel_batchbuffer *batch,
 
 	gen8_emit_sf(batch);
 
+	gen8_emit_ps(batch, ps_kernel_off);
+
 	OUT_BATCH(GEN7_3DSTATE_BINDING_TABLE_POINTERS_PS);
 	OUT_BATCH(ps_binding_table);
 
 	OUT_BATCH(GEN7_3DSTATE_SAMPLER_STATE_POINTERS_PS);
 	OUT_BATCH(ps_sampler_state);
 
-	gen8_emit_ps(batch, ps_kernel_off);
-
 	OUT_BATCH(GEN6_3DSTATE_SCISSOR_STATE_POINTERS);
 	OUT_BATCH(scissor_state);
 
Signature · 6 lines
1.9.1

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx