Thanks for catching the typo.
> On 11/18/2016 01:43 PM, Xiang, Haihao wrote:
> > This function can be used on GEN8 too
>
> The patch looks good to me.
>
> But please change the title to "gen8_gpe_mi_store_register_mem"
> instead
> of gen8_gpe_mi_load_register_mem
>
> Thanks.
> >
> > Signed-o
On 11/18/2016 01:43 PM, Xiang, Haihao wrote:
Signed-off-by: Xiang, Haihao
This looks good to me.
Thanks
---
src/i965_gpe_utils.c | 16
src/i965_gpe_utils.h | 19 +++
src/intel_driver.h | 6 ++
3 files changed, 41 insertions(+)
diff --git a/src/i96
On 11/18/2016 01:43 PM, Xiang, Haihao wrote:
Signed-off-by: Xiang, Haihao
This looks good to me.
Thanks
---
src/i965_gpe_utils.c | 73
src/i965_gpe_utils.h | 5
2 files changed, 78 insertions(+)
diff --git a/src/i965_gpe_utils.c
On 11/18/2016 01:43 PM, Xiang, Haihao wrote:
v2: Add support for override_offset, clean up the code for 2D surface.
This looks good to me.
Thanks
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 217 +++
src/i965_gpe_utils.h | 5 ++
On 11/18/2016 01:43 PM, Xiang, Haihao wrote:
This looks good to me.
Thanks
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 18 +-
src/i965_gpe_utils.h | 3 ++-
2 files changed, 19 insertions(+), 2 deletions(-)
diff --git a/src/i965_gpe_utils.c b/src/i965_gpe_utils.
On 11/18/2016 01:43 PM, Xiang, Haihao wrote:
This function can be used on GEN8 too
The patch looks good to me.
But please change the title to "gen8_gpe_mi_store_register_mem" instead
of gen8_gpe_mi_load_register_mem
Thanks.
Signed-off-by: Xiang, Haihao
---
src/gen9_vdenc.c | 10 ++
On 11/18/2016 01:43 PM, Xiang, Haihao wrote:
This function can be used on GEN8 too
This looks good to me.
Thanks
Signed-off-by: Xiang, Haihao
---
src/gen9_vdenc.c | 4 ++--
src/gen9_vp9_encoder.c | 4 ++--
src/i965_gpe_utils.c | 2 +-
src/i965_gpe_utils.h | 2 +-
4 files cha
On 11/18/2016 01:43 PM, Xiang, Haihao wrote:
This function can be used on GEN8 too
This looks good to me.
Signed-off-by: Xiang, Haihao
---
src/gen9_vdenc.c | 2 +-
src/gen9_vp9_encoder.c | 6 +++---
src/i965_gpe_utils.c | 2 +-
src/i965_gpe_utils.h | 2 +-
4 files changed, 6 i
On 11/18/2016 01:43 PM, Xiang, Haihao wrote:
This function can be used on GEN8 too
This looks good to me.
Thanks
Signed-off-by: Xiang, Haihao
---
src/gen9_post_processing.c | 2 +-
src/gen9_vp9_encoder.c | 2 +-
src/i965_gpe_utils.c | 2 +-
src/i965_gpe_utils.h | 2 +-
On 11/18/2016 01:43 PM, Xiang, Haihao wrote:
Like as i965_dri_object_to_buffer_gpe_resource(), use i965_ instead of
i965_gpe_ as prefix
This looks good to me.
Thanks
Signed-off-by: Xiang, Haihao
---
src/gen9_post_processing.c | 2 +-
src/gen9_vp9_encoder.c | 2 +-
src/i965_gpe_ut
On 11/18/2016 01:43 PM, Xiang, Haihao wrote:
gpe_context->curbe.bo always points to the curbe buffer now, and the two
functions
can be used on all platforms
This looks good to me.
Thanks
Yakui
Signed-off-by: Xiang, Haihao
---
src/gen9_post_processing.c | 4 ++--
src/gen9_vp9_encoder.
On 11/18/2016 01:43 PM, Xiang, Haihao wrote:
It is multiple of 64 bytes
This looks good to me.
Add: Reviewed-by: Zhao Yakui
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/i965_gpe_utils.c b/src/i965_gpe_utils
On 11/18/2016 01:43 PM, Xiang, Haihao wrote:
Make sure the size for each part in dynamic state buffer is multiple of 64
bytes.
This looks good to me.
Add: Reviewed-by: Zhao Yakui
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 9 +
1 file changed, 5 insertions(+), 4 del
On 11/18/2016 01:43 PM, Xiang, Haihao wrote:
User can set sampler entry size and the number of sampler entries now.
sampler.bo is
always set even if the sampler state is a part of the dynamic state buffer,
hence we
can use the corresponding settings no matter sampler state is a part of the
dyn
On 11/18/2016 01:43 PM, Xiang, Haihao wrote:
This patch delete idrt_size and use (idrt.max_entries * idrt.entry_size)
instead.
idrt.bo is always set even if the interface descriptor remapping table is a
part of
the dynamic state buffer, hence we can use the corresponding settings no matter
thi
On 11/18/2016 01:43 PM, Xiang, Haihao wrote:
To avoid confusion between curbe.length and curbe_size, this patch uses
curbe.length only. curbe.bo is always set even if curbe is a part of the
dynamic state buffer, hence we can use curbe related settings no matter
it is a part of the dynamic state b
On 11/18/2016 01:43 PM, Xiang, Haihao wrote:
Make sure the size is multiple of 64 bytes
This looks good to me.
Add: Reviewed-by: Zhao Yakui
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/src/i965_gpe_utils.c
And it is treated as non-tiling for I010 surfaces, which is like I420.
Signed-off-by: Zhao Yakui
---
src/i965_drv_video.c | 23 +++
src/i965_fourcc.h| 4
2 files changed, 27 insertions(+)
diff --git a/src/i965_drv_video.c b/src/i965_drv_video.c
index 04670f4..fbb64
I010 format is another kind of 10-bit surface. And its layout is similar to
I420.
Signed-off-by: Zhao Yakui
---
src/gen75_picture_process.c | 45 ++---
1 file changed, 42 insertions(+), 3 deletions(-)
diff --git a/src/gen75_picture_process.c b/src/gen75_
If it is not supported, it will fall back to the original config.
Signed-off-by: Zhao Yakui
Reviewed-by: Charles Daniel
---
src/gen8_mfc.c | 6 +-
src/gen8_post_processing.c | 5 -
src/gen8_vme.c | 7 ++-
src/gen9_post_processing.c | 12
src/g
This info can be used to configure the max EU threads of GPU device.
>eu_total * 6
V1->V2: Remove the extra bit_field of has_eu_flag. Instead the eu_total > 0
can be used.
Signed-off-by: Zhao Yakui
Reviewed-by: Charles Daniel
---
src/intel_driver.c | 11 +++
src/intel_driver.h | 2 ++
This function can be used on GEN8 too
Signed-off-by: Xiang, Haihao
---
src/gen9_vdenc.c | 10 +-
src/gen9_vp9_encoder.c | 4 ++--
src/i965_gpe_utils.c | 2 +-
src/i965_gpe_utils.h | 2 +-
4 files changed, 9 insertions(+), 9 deletions(-)
diff --git a/src/gen9_vdenc.c b/src/g
It is multiple of 64 bytes
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/i965_gpe_utils.c b/src/i965_gpe_utils.c
index 139ab1b..a6d539b 100644
--- a/src/i965_gpe_utils.c
+++ b/src/i965_gpe_utils.c
@@ -1066,7 +1066,7
This function can be used on GEN8 too
Signed-off-by: Xiang, Haihao
---
src/gen9_vdenc.c | 2 +-
src/gen9_vp9_encoder.c | 6 +++---
src/i965_gpe_utils.c | 2 +-
src/i965_gpe_utils.h | 2 +-
4 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/src/gen9_vdenc.c b/src/gen9_vdenc.
This patch delete idrt_size and use (idrt.max_entries * idrt.entry_size)
instead.
idrt.bo is always set even if the interface descriptor remapping table is a
part of
the dynamic state buffer, hence we can use the corresponding settings no matter
this
table is a part of the dynamic state buffer o
Make sure the size is multiple of 64 bytes
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/src/i965_gpe_utils.c b/src/i965_gpe_utils.c
index 5d4ca5c..c5a8935 100644
--- a/src/i965_gpe_utils.c
+++ b/src/i965_gpe_utils.
To avoid confusion between curbe.length and curbe_size, this patch uses
curbe.length only. curbe.bo is always set even if curbe is a part of the
dynamic state buffer, hence we can use curbe related settings no matter
it is a part of the dynamic state buffer or not.
Signed-off-by: Xiang, Haihao
--
Like as i965_dri_object_to_buffer_gpe_resource(), use i965_ instead of
i965_gpe_ as prefix
Signed-off-by: Xiang, Haihao
---
src/gen9_post_processing.c | 2 +-
src/gen9_vp9_encoder.c | 2 +-
src/i965_gpe_utils.c | 10 +-
src/i965_gpe_utils.h | 10 +-
4 files cha
User can set sampler entry size and the number of sampler entries now.
sampler.bo is
always set even if the sampler state is a part of the dynamic state buffer,
hence we
can use the corresponding settings no matter sampler state is a part of the
dynamic
state buffer or not.
Signed-off-by: Xiang
This function can be used on GEN8 too
Signed-off-by: Xiang, Haihao
---
src/gen9_vdenc.c | 4 ++--
src/gen9_vp9_encoder.c | 4 ++--
src/i965_gpe_utils.c | 2 +-
src/i965_gpe_utils.h | 2 +-
4 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/src/gen9_vdenc.c b/src/gen9_vdenc.
This function can be used on GEN8 too
Signed-off-by: Xiang, Haihao
---
src/gen9_vdenc.c | 10 +-
src/gen9_vp9_encoder.c | 10 +-
src/i965_gpe_utils.c | 2 +-
src/i965_gpe_utils.h | 2 +-
4 files changed, 12 insertions(+), 12 deletions(-)
diff --git a/src/gen9_vdenc.c
The patch series cleans up some gpe settings, renames some gen9 fucntions for
gen8
and addes some new gpe functions for gen8.
v2: add support for override_offset when adding a new surface
Xiang, Haihao (17):
Change the size calculation of the required memory for all GPE kernels
Move all cur
Make sure the size for each part in dynamic state buffer is multiple of 64
bytes.
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 9 +
1 file changed, 5 insertions(+), 4 deletions(-)
diff --git a/src/i965_gpe_utils.c b/src/i965_gpe_utils.c
index d7286fc..139ab1b 100644
--- a/sr
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 16
src/i965_gpe_utils.h | 19 +++
src/intel_driver.h | 6 ++
3 files changed, 41 insertions(+)
diff --git a/src/i965_gpe_utils.c b/src/i965_gpe_utils.c
index 058e6b6..46960e6 100644
--- a/src/i965_gp
v2: Add support for override_offset, clean up the code for 2D surface.
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 217 +++
src/i965_gpe_utils.h | 5 ++
2 files changed, 222 insertions(+)
diff --git a/src/i965_gpe_utils.c b/src/i965_
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 73
src/i965_gpe_utils.h | 5
2 files changed, 78 insertions(+)
diff --git a/src/i965_gpe_utils.c b/src/i965_gpe_utils.c
index 46960e6..ca8c86a 100644
--- a/src/i965_gpe_utils.c
+++
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 18 +-
src/i965_gpe_utils.h | 3 ++-
2 files changed, 19 insertions(+), 2 deletions(-)
diff --git a/src/i965_gpe_utils.c b/src/i965_gpe_utils.c
index 4c0384e..d1b1941 100644
--- a/src/i965_gpe_utils.c
+++ b/src/i965_gpe_uti
gpe_context->curbe.bo always points to the curbe buffer now, and the two
functions
can be used on all platforms
Signed-off-by: Xiang, Haihao
---
src/gen9_post_processing.c | 4 ++--
src/gen9_vp9_encoder.c | 20 ++--
src/i965_gpe_utils.c | 4 ++--
src/i965_gpe_utils.h
This function can be used on GEN8 too
Signed-off-by: Xiang, Haihao
---
src/gen9_post_processing.c | 2 +-
src/gen9_vp9_encoder.c | 2 +-
src/i965_gpe_utils.c | 2 +-
src/i965_gpe_utils.h | 2 +-
4 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/src/gen9_post_processi
On 11/18/2016 09:42 AM, Xiang, Haihao wrote:
On 11/18/2016 12:58 AM, Xiang, Haihao wrote:
-Original Message-
From: Zhao, Yakui
Sent: Thursday, November 17, 2016 8:42 PM
To: Xiang, Haihao
Cc: libva@lists.freedesktop.org
Subject: Re: [Libva] [Libva-intel-driver][PATCH 02/17] Move all
> On Thu, 2016-11-17 at 08:15 +, Xiang, Haihao wrote:
> > But in Lukas's comment:
> >
> > "Had to revert it because of number of hard lockups."
> >
> > Did you have other patch to fix the lockup issue in other component
> > or
> > does the issue has gone away? If not, We have to revert the
>
> On 11/18/2016 12:58 AM, Xiang, Haihao wrote:
> >
> >
> > > -Original Message-
> > > From: Zhao, Yakui
> > > Sent: Thursday, November 17, 2016 8:42 PM
> > > To: Xiang, Haihao
> > > Cc: libva@lists.freedesktop.org
> > > Subject: Re: [Libva] [Libva-intel-driver][PATCH 02/17] Move all
> >
> On 11/18/2016 08:47 AM, Xiang, Haihao wrote:
> > On Wed, 2016-11-16 at 21:06 -0500, Zhao Yakui wrote:
> > > This info can be used to configure the max EU threads of GPU
> > > device.
> > > > eu_total * 6
> > >
> > > Signed-off-by: Zhao Yakui
> > > ---
> > > src/intel_driver.c | 13 +++
I010 is another kind of 10-bit format. Its layout is similar to IYUV(I420).
And the lower 10-bit is used.
Signed-off-by: Zhao Yakui
---
va/va.h | 5 +
1 file changed, 5 insertions(+)
diff --git a/va/va.h b/va/va.h
index 88628a8..153a89b 100644
--- a/va/va.h
+++ b/va/va.h
@@ -2545,6 +2545,11
On 11/18/2016 08:47 AM, Xiang, Haihao wrote:
On Wed, 2016-11-16 at 21:06 -0500, Zhao Yakui wrote:
This info can be used to configure the max EU threads of GPU device.
eu_total * 6
Signed-off-by: Zhao Yakui
---
src/intel_driver.c | 13 +
src/intel_driver.h | 3 +++
2 files cha
On 11/18/2016 09:00 AM, Xiang, Haihao wrote:
Could you add the new FOURCC in libva as well ?
Sure.
Thanks
Haihao
And it is treated as non-tiling for I010 surfaces, which is like
I420.
Signed-off-by: Zhao Yakui
---
src/i965_drv_video.c | 23 +++
src/i965_fourcc.h
On 11/18/2016 12:22 AM, Xiang, Haihao wrote:
-Original Message-
From: Zhao, Yakui
Sent: Thursday, November 17, 2016 8:47 PM
To: Xiang, Haihao
Cc: libva@lists.freedesktop.org
Subject: Re: [Libva] [Libva-intel-driver][PATCH 06/17] Fix curbe length in
CMD_MEDIA_CURBE_LOAD on GEN8+
On 11/
Could you add the new FOURCC in libva as well ?
Thanks
Haihao
> And it is treated as non-tiling for I010 surfaces, which is like
> I420.
>
> Signed-off-by: Zhao Yakui
> ---
> src/i965_drv_video.c | 23 +++
> src/i965_fourcc.h| 4
> 2 files changed, 27 insertions
On 11/17/2016 11:48 PM, Xiang, Haihao wrote:
-Original Message-
From: Zhao, Yakui
Sent: Thursday, November 17, 2016 9:14 PM
To: Xiang, Haihao
Cc: libva@lists.freedesktop.org
Subject: Re: [Libva] [Libva-intel-driver][PATCH 15/17] Add a new gpe function
gen8_gpe_context_add_surface() to
On 11/18/2016 12:12 AM, Xiang, Haihao wrote:
-Original Message-
From: Libva [mailto:libva-boun...@lists.freedesktop.org] On Behalf Of Zhao
Yakui
Sent: Thursday, November 17, 2016 8:59 PM
To: libva@lists.freedesktop.org
Subject: Re: [Libva] [Libva-intel-driver][PATCH 09/17] Rename
gen9_
On 11/18/2016 12:58 AM, Xiang, Haihao wrote:
-Original Message-
From: Zhao, Yakui
Sent: Thursday, November 17, 2016 8:42 PM
To: Xiang, Haihao
Cc: libva@lists.freedesktop.org
Subject: Re: [Libva] [Libva-intel-driver][PATCH 02/17] Move all curbe related
settings to the inner structure in
On Wed, 2016-11-16 at 21:06 -0500, Zhao Yakui wrote:
> This info can be used to configure the max EU threads of GPU device.
> > eu_total * 6
>
> Signed-off-by: Zhao Yakui
> ---
> src/intel_driver.c | 13 +
> src/intel_driver.h | 3 +++
> 2 files changed, 16 insertions(+)
>
> diff -
On Tue, 2016-11-15 at 15:24 +, Emil Velikov wrote:
> If we do not use a render node we must authenticate. Doing the extra
> GetClient calls/ioctls does not help much, so don't bother.
>
> Cc: David Herrmann
> Cc: Daniel Vetter
> Signed-off-by: Emil Velikov
After some discussion in this thr
On Thu, 2016-11-17 at 18:17 +, Xiang, Haihao wrote:
> > -Original Message-
> > From: Sean V Kelley [mailto:sea...@posteo.de]
> > Sent: Friday, November 18, 2016 12:47 AM
> > To: Emil Velikov ; Xiang, Haihao
> >
> > Cc: Daniel Vetter ; libva@lists.freedesktop
> > .org; David
> > Herrman
On Wed, Nov 16, 2016 at 6:06 PM, Zhao Yakui wrote:
> This will help to make good use of HW EU resources.
> If it is not supported, it will fall back to the original config.
>
>
> Signed-off-by: Zhao Yakui
> ---
> src/gen8_mfc.c | 6 +-
> src/gen8_post_processing.c | 5 -
>
On Wed, Nov 16, 2016 at 6:06 PM, Zhao Yakui wrote:
> This info can be used to configure the max EU threads of GPU device.
>>eu_total * 6
>
> Signed-off-by: Zhao Yakui
> ---
> src/intel_driver.c | 13 +
> src/intel_driver.h | 3 +++
> 2 files changed, 16 insertions(+)
>
> diff --git
>-Original Message-
>From: Sean V Kelley [mailto:sea...@posteo.de]
>Sent: Friday, November 18, 2016 12:47 AM
>To: Emil Velikov ; Xiang, Haihao
>
>Cc: Daniel Vetter ; libva@lists.freedesktop.org; David
>Herrmann
>Subject: Re: [Libva] [PATCH] drm: remove va_drm_is_authenticated check
>
>On
>-Original Message-
>From: Zhao, Yakui
>Sent: Thursday, November 17, 2016 8:42 PM
>To: Xiang, Haihao
>Cc: libva@lists.freedesktop.org
>Subject: Re: [Libva] [Libva-intel-driver][PATCH 02/17] Move all curbe related
>settings to the inner structure in i965_gpe_context
>
>On 11/17/2016 04:35
On Thu, 2016-11-17 at 08:15 +, Xiang, Haihao wrote:
> But in Lukas's comment:
>
> "Had to revert it because of number of hard lockups."
>
> Did you have other patch to fix the lockup issue in other component
> or
> does the issue has gone away? If not, We have to revert the
> patch
> otherwi
On Thu, 2016-11-17 at 16:42 +, Emil Velikov wrote:
> On 17 November 2016 at 15:38, Xiang, Haihao
> wrote:
> > > From: Libva [mailto:libva-boun...@lists.freedesktop.org] On
> > > Behalf Of Emil
> > > Velikov
> > > Sent: Tuesday, November 15, 2016 11:24 PM
> > > To: libva@lists.freedesktop.org
>
On 17 November 2016 at 15:38, Xiang, Haihao wrote:
>>From: Libva [mailto:libva-boun...@lists.freedesktop.org] On Behalf Of Emil
>>Velikov
>>Sent: Tuesday, November 15, 2016 11:24 PM
>>To: libva@lists.freedesktop.org
>>Cc: Daniel Vetter ; David Herrmann
>>
>>Subject: [Libva] [PATCH] drm: remove va_
On Thu, 2016-11-17 at 15:38 +, Xiang, Haihao wrote:
> > From: Libva [mailto:libva-boun...@lists.freedesktop.org] On Behalf
> > Of Emil
> > Velikov
> > Sent: Tuesday, November 15, 2016 11:24 PM
> > To: libva@lists.freedesktop.org
> > Cc: Daniel Vetter ; David Herrmann
> >
> > Subject: [Libva] [
>-Original Message-
>From: Zhao, Yakui
>Sent: Thursday, November 17, 2016 8:47 PM
>To: Xiang, Haihao
>Cc: libva@lists.freedesktop.org
>Subject: Re: [Libva] [Libva-intel-driver][PATCH 06/17] Fix curbe length in
>CMD_MEDIA_CURBE_LOAD on GEN8+
>
>On 11/17/2016 04:35 PM, Xiang, Haihao wrote:
>-Original Message-
>From: Libva [mailto:libva-boun...@lists.freedesktop.org] On Behalf Of Zhao
>Yakui
>Sent: Thursday, November 17, 2016 8:59 PM
>To: libva@lists.freedesktop.org
>Subject: Re: [Libva] [Libva-intel-driver][PATCH 09/17] Rename
>gen9_gpe_media_object_walker() to gen8_gpe_med
>-Original Message-
>From: Zhao, Yakui
>Sent: Thursday, November 17, 2016 9:14 PM
>To: Xiang, Haihao
>Cc: libva@lists.freedesktop.org
>Subject: Re: [Libva] [Libva-intel-driver][PATCH 15/17] Add a new gpe function
>gen8_gpe_context_add_surface() to set surface state on GEN8
>
>On 11/17/20
>From: Libva [mailto:libva-boun...@lists.freedesktop.org] On Behalf Of Emil
>Velikov
>Sent: Tuesday, November 15, 2016 11:24 PM
>To: libva@lists.freedesktop.org
>Cc: Daniel Vetter ; David Herrmann
>
>Subject: [Libva] [PATCH] drm: remove va_drm_is_authenticated check
>
>If we do not use a render no
On 11/17/2016 04:35 PM, Xiang, Haihao wrote:
Signed-off-by: Xiang, Haihao
Can the gen8_gpe_context_add_surface follow the similar logic in
gen9_gpe_context_add_surface?
>2d_surface && override_offset
>2d && is_uv
>2d
>adv
>buffer
The 2d_surface && override_offset is especia
On 11/17/2016 04:35 PM, Xiang, Haihao wrote:
This function can be used on GEN8 too
Based on HW spec, Broadwell and Gen9 have some difference about the
MEDIA_OBJECT_WALKER commands.
So it will be better to add a new wrapper of media_object_walker for gen8.
Thanks
Signed-off-by: Xiang, Haih
On 11/17/2016 04:35 PM, Xiang, Haihao wrote:
It is multiple of 64 bytes
Can we align the size of curbe_buffer when calling the
gen8_gpe_context_init?
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/i965_gpe_ut
On 11/17/2016 04:35 PM, Xiang, Haihao wrote:
To avoid confusion between curbe.length and curbe_size, this patch uses
curbe.length only. curbe.bo is always set even if curbe is a part of the
dynamic state buffer, hence we can use curbe related settings no matter
it is a part of the dynamic state b
I010 format is another kind of 10-bit surface. And its layout is similar to
I420.
Signed-off-by: Zhao Yakui
---
src/gen75_picture_process.c | 45 ++---
1 file changed, 42 insertions(+), 3 deletions(-)
diff --git a/src/gen75_picture_process.c b/src/gen75_
This will help to make good use of HW EU resources.
If it is not supported, it will fall back to the original config.
Signed-off-by: Zhao Yakui
---
src/gen8_mfc.c | 6 +-
src/gen8_post_processing.c | 5 -
src/gen8_vme.c | 7 ++-
src/gen9_post_processing.c
This info can be used to configure the max EU threads of GPU device.
>eu_total * 6
Signed-off-by: Zhao Yakui
---
src/intel_driver.c | 13 +
src/intel_driver.h | 3 +++
2 files changed, 16 insertions(+)
diff --git a/src/intel_driver.c b/src/intel_driver.c
index bb19401..a2c8c71 1006
And it is treated as non-tiling for I010 surfaces, which is like I420.
Signed-off-by: Zhao Yakui
---
src/i965_drv_video.c | 23 +++
src/i965_fourcc.h| 4
2 files changed, 27 insertions(+)
diff --git a/src/i965_drv_video.c b/src/i965_drv_video.c
index 04670f4..fbb64
This function can be used on GEN8 too
Signed-off-by: Xiang, Haihao
---
src/gen9_vdenc.c | 2 +-
src/gen9_vp9_encoder.c | 6 +++---
src/i965_gpe_utils.c | 2 +-
src/i965_gpe_utils.h | 2 +-
4 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/src/gen9_vdenc.c b/src/gen9_vdenc.
gpe_context->curbe.bo always points to the curbe buffer now, and the two
functions
can be used on all platforms
Signed-off-by: Xiang, Haihao
---
src/gen9_post_processing.c | 4 ++--
src/gen9_vp9_encoder.c | 20 ++--
src/i965_gpe_utils.c | 4 ++--
src/i965_gpe_utils.h
This function can be used on GEN8 too
Signed-off-by: Xiang, Haihao
---
src/gen9_vdenc.c | 10 +-
src/gen9_vp9_encoder.c | 4 ++--
src/i965_gpe_utils.c | 2 +-
src/i965_gpe_utils.h | 2 +-
4 files changed, 9 insertions(+), 9 deletions(-)
diff --git a/src/gen9_vdenc.c b/src/g
To avoid confusion between curbe.length and curbe_size, this patch uses
curbe.length only. curbe.bo is always set even if curbe is a part of the
dynamic state buffer, hence we can use curbe related settings no matter
it is a part of the dynamic state buffer or not.
Signed-off-by: Xiang, Haihao
--
The patch series cleans up some gpe settings, renames some gen9 fucntions for
gen8
and addes some new gpe functions for gen8.
Xiang, Haihao (17):
Change the size calculation of the required memory for all GPE kernels
Move all curbe related settings to the inner structure in
i965_gpe_conte
This function can be used on GEN8 too
Signed-off-by: Xiang, Haihao
---
src/gen9_post_processing.c | 2 +-
src/gen9_vp9_encoder.c | 2 +-
src/i965_gpe_utils.c | 2 +-
src/i965_gpe_utils.h | 2 +-
4 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/src/gen9_post_processi
This patch delete idrt_size and use (idrt.max_entries * idrt.entry_size)
instead.
idrt.bo is always set even if the interface descriptor remapping table is a
part of
the dynamic state buffer, hence we can use the corresponding settings no matter
this
table is a part of the dynamic state buffer o
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 73
src/i965_gpe_utils.h | 5
2 files changed, 78 insertions(+)
diff --git a/src/i965_gpe_utils.c b/src/i965_gpe_utils.c
index 4ac7ab7..2e54be3 100644
--- a/src/i965_gpe_utils.c
+++
Like as i965_dri_object_to_buffer_gpe_resource(), use i965_ instead of
i965_gpe_ as prefix
Signed-off-by: Xiang, Haihao
---
src/gen9_post_processing.c | 2 +-
src/gen9_vp9_encoder.c | 2 +-
src/i965_gpe_utils.c | 10 +-
src/i965_gpe_utils.h | 10 +-
4 files cha
It is multiple of 64 bytes
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/i965_gpe_utils.c b/src/i965_gpe_utils.c
index 139ab1b..a6d539b 100644
--- a/src/i965_gpe_utils.c
+++ b/src/i965_gpe_utils.c
@@ -1066,7 +1066,7
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 16
src/i965_gpe_utils.h | 19 +++
src/intel_driver.h | 6 ++
3 files changed, 41 insertions(+)
diff --git a/src/i965_gpe_utils.c b/src/i965_gpe_utils.c
index e7f1c25..4ac7ab7 100644
--- a/src/i965_gp
Make sure the size for each part in dynamic state buffer is multiple of 64
bytes.
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 9 +
1 file changed, 5 insertions(+), 4 deletions(-)
diff --git a/src/i965_gpe_utils.c b/src/i965_gpe_utils.c
index d7286fc..139ab1b 100644
--- a/sr
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 231 +++
src/i965_gpe_utils.h | 5 ++
2 files changed, 236 insertions(+)
diff --git a/src/i965_gpe_utils.c b/src/i965_gpe_utils.c
index d1b1941..e7f1c25 100644
--- a/src/i965_gpe_utils.c
+++
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 18 +-
src/i965_gpe_utils.h | 3 ++-
2 files changed, 19 insertions(+), 2 deletions(-)
diff --git a/src/i965_gpe_utils.c b/src/i965_gpe_utils.c
index 4c0384e..d1b1941 100644
--- a/src/i965_gpe_utils.c
+++ b/src/i965_gpe_uti
This function can be used on GEN8 too
Signed-off-by: Xiang, Haihao
---
src/gen9_vdenc.c | 4 ++--
src/gen9_vp9_encoder.c | 4 ++--
src/i965_gpe_utils.c | 2 +-
src/i965_gpe_utils.h | 2 +-
4 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/src/gen9_vdenc.c b/src/gen9_vdenc.
User can set sampler entry size and the number of sampler entries now.
sampler.bo is
always set even if the sampler state is a part of the dynamic state buffer,
hence we
can use the corresponding settings no matter sampler state is a part of the
dynamic
state buffer or not.
Signed-off-by: Xiang
This function can be used on GEN8 too
Signed-off-by: Xiang, Haihao
---
src/gen9_vdenc.c | 10 +-
src/gen9_vp9_encoder.c | 10 +-
src/i965_gpe_utils.c | 2 +-
src/i965_gpe_utils.h | 2 +-
4 files changed, 12 insertions(+), 12 deletions(-)
diff --git a/src/gen9_vdenc.c
Make sure the size is multiple of 64 bytes
Signed-off-by: Xiang, Haihao
---
src/i965_gpe_utils.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/src/i965_gpe_utils.c b/src/i965_gpe_utils.c
index 5d4ca5c..c5a8935 100644
--- a/src/i965_gpe_utils.c
+++ b/src/i965_gpe_utils.
But in Lukas's comment:
"Had to revert it because of number of hard lockups."
Did you have other patch to fix the lockup issue in other component or
does the issue has gone away? If not, We have to revert the patch
otherwise user will suffer lockup issue with libva.
Thanks
Haihao
> On Thu, No
93 matches
Mail list logo