https://bugs.freedesktop.org/show_bug.cgi?id=101334
--- Comment #45 from Dave Airlie ---
https://patchwork.freedesktop.org/series/28535/
is a replacement for the cs flush, might be worth trying on master on its own.
and possibly with the hack patch on top (it might not apply cleanly though).
-
From: Dave Airlie
This seems like a workaround, but we don't see the bug on CIK/VI.
On SI with the dEQP-VK.memory.pipeline_barrier.host_read_transfer_dst.*
tests, when one tests complete, the first flush at the start of the next
test causes a VM fault as we've destroyed the VM, but we end up flu
Hi Thomas,
first of all, would it be possible to split these patches up a bit
further? At least patch 1 seems to contain several logical changes,
which makes it a bit difficult to review.
On 08/08/17 03:48 PM, Thomas Hellstrom wrote:
> Implement back-to-fake-front flips,
> Fix EGL_BUFFER_PRESE
On Wed, Aug 9, 2017 at 3:17 PM, Marathe, Yogesh
wrote:
> Tomasz,
>
>> -Original Message-
>> From: mesa-dev [mailto:mesa-dev-boun...@lists.freedesktop.org] On Behalf
>> Of Tomasz Figa
>> Sent: Tuesday, August 8, 2017 7:43 AM
>> To: Marathe, Yogesh
>> On Mon, Aug 7, 2017 at 2:19 PM, Marathe
Tomasz,
> -Original Message-
> From: mesa-dev [mailto:mesa-dev-boun...@lists.freedesktop.org] On Behalf
> Of Tomasz Figa
> Sent: Tuesday, August 8, 2017 7:43 AM
> To: Marathe, Yogesh
> On Mon, Aug 7, 2017 at 2:19 PM, Marathe, Yogesh
> wrote:
> > Tomasz,
> >
> >> -Original Message
https://bugs.freedesktop.org/show_bug.cgi?id=101334
--- Comment #44 from John ---
I've just tried on amd-staging 4.12, and without the hacky patch, and it still
froze the same (still heavy I/O when it did, but nothing in dmesg this time).
--
You are receiving this mail because:
You are the QA C
https://bugs.freedesktop.org/show_bug.cgi?id=102030
--- Comment #5 from Janpieter Sollie ---
Created attachment 133403
--> https://bugs.freedesktop.org/attachment.cgi?id=133403&action=edit
working program source
the source contains the following modifications:
-use of local memory
-insertion o
https://bugs.freedesktop.org/show_bug.cgi?id=102030
--- Comment #4 from Janpieter Sollie ---
I got it running on both pocl and clover. I will attach a working version with
a built_hints.txt file included, which also includes debug information to debug
memory contents.
what I concluded:
- atom_ad
On 08/08/2017 08:07 PM, Emil Velikov wrote:
On 8 August 2017 at 16:10, Eric Engestrom wrote:
On Saturday, 2017-08-05 00:25:49 +0100, Emil Velikov wrote:
From: Emil Velikov
As mentioned in previous commit the negative tests in dEQP expect the
arguments to be evaluated in particular order.
Th
On 08/08/2017 05:20 PM, Emil Velikov wrote:
On 8 August 2017 at 08:21, Tapani Pälli wrote:
On 08/05/2017 02:25 AM, Emil Velikov wrote:
From: Emil Velikov
As mentioned in previous commit the negative tests in dEQP expect the
arguments to be evaluated in particular order.
Namely - first the
This function differs from ralloc_strcat() and ralloc_strncat()
in that it does not do any strlen() calls which can become
costly on large strings.
---
src/util/ralloc.c | 19 +++
src/util/ralloc.h | 18 ++
2 files changed, 37 insertions(+)
diff --git a/src/util/r
The Deus Ex: Mankind Divided shaders go from spending ~20 seconds
in the GLSL IR compilers front-end down to ~18.5 seconds on a
Ryzen 1800X.
Tested by compiling once with shader-db then deleting the index file
from the shader cache and compiling again.
v2:
- fix rebasing issue in v1
---
src/com
f81ede469910d fixed a problem with shaders including IR that was
owned by builtins. However the approach of cloning the whole
function each time we referenced it lead to a significant
reduction in the GLSL IR compiler performance.
Everything was already cloned when inlining the function, as
far as
The cloning was introduced in f81ede469910d to fixed a problem with
shaders including IR that was owned by builtins.
However the approach of cloning the whole function each time we
reference a builtin lead to a significant reduction in the GLSL
IR compilers performance.
The previous patch fixes t
Also add a comment that this should only be used by the ir_reader
interface for testing purposes.
v2:
- fix grammar in comment
- use unreachable rather than assert
Reviewed-by: Thomas Helland
---
src/compiler/glsl/ir.cpp | 9 ++---
src/compiler/glsl/ir.h | 14 +++---
2 files ch
Reviewed-by: Thomas Helland
Tested-by: Dieter Nützel
---
src/compiler/glsl/ir.h | 5 -
1 file changed, 5 deletions(-)
diff --git a/src/compiler/glsl/ir.h b/src/compiler/glsl/ir.h
index ce4ade9e80..170759abeb 100644
--- a/src/compiler/glsl/ir.h
+++ b/src/compiler/glsl/ir.h
@@ -1595,32 +1595,
Extra validation is added to ir_validate to make sure this is
always updated to the correct numer of operands, as passes like
lower_instructions modify the instructions directly rather then
generating a new one.
The reduction in time is so small that it is not really
measurable. However callgrind
v2: Address various comments from Thomas. All minor but enough
to warrant a resend.
I've sent all but the last 2 already but they haven't received much
feedback yet.
This series reduces compile times of the Deus Ex shaders on my
Ryzen 1800X from 2m27s -> 2m8s with a cold cache on radeonsi.
Note
We are currently copying the name for each member dereference
but we can just share a single instance of the string provided
by the type.
This change also stops us recalculating the field index
repeatedly.
---
src/compiler/glsl/ast_array_index.cpp | 14 -
src/compiler/glsl/gl
On 09/08/17 05:59, Thomas Helland wrote:
I'm not quite sure if the increase in size of the base ir class
versus the reduced overhead and added code is worth it?
There is more code yes, but most of that is asserts and validation code.
This should actually make the code more robust than it is cu
https://bugs.freedesktop.org/show_bug.cgi?id=102125
Bug ID: 102125
Summary: [softpipe] piglit arb_texture_view-targets regression
Product: Mesa
Version: git
Hardware: x86-64 (AMD64)
OS: Linux (All)
Status: NEW
On 09/08/17 03:24 AM, Emil Velikov wrote:
> On 8 August 2017 at 08:23, Michel Dänzer wrote:
>> From: Michel Dänzer
>>
>> No need to manually look for the library files anymore with current
>> LLVM. This sidesteps the manual method failing when LLVM was built with
>> -DLLVM_APPEND_VC_REV=ON.
>>
>
Hi Nicolai,
I put this series through Intels CI system and it hit a couple of
issues. I haven't yet checked if these CTS test regress on radeonsi as
well or if its just an i965 thing.
Project: deqp-test
Test: dEQP-
GLES31.functional.shaders.multisample_interpolation.interpolate_at_ce
https://bugs.freedesktop.org/show_bug.cgi?id=102123
Bug ID: 102123
Summary: [llvmpipe] piglit
gl-3.2-layered-rendering-framebuffertexture regression
Product: Mesa
Version: git
Hardware: x86-64 (AMD64)
OS: Li
Matt Turner writes:
> The mixture of hardware encodings and logical types has caused lots of
> confusion. It's time to fix that.
fwiw, apart from the one comment about sizeof V/UV, series is
Reviewed-by: Scott D Phillips
> ___
> mesa-dev mailing lis
On 2017-08-08 15:45:25, Jason Ekstrand wrote:
> This little series adds support for the VK_KHR_external_fence family of
> extensions. Most of the real work in implementing these extensions is
> actually in the kernel. Once we have a DRM_SYNCOBJ_IOCTL_WAIT that does
> what we need, the userspace b
Matt Turner writes:
> The hardware encodings often mean different things depending on whether
> the source is an immediate.
> ---
> src/intel/compiler/brw_disasm.c | 46 ---
> src/intel/compiler/brw_eu_compact.c | 8 +--
> src/intel/compiler/brw_eu_defines.h | 48 +
Matt Turner writes:
> These vaguely corresponded to the hardware encodings, but that is purely
> historical at this point. Reorder them so we stop making things "almost
> work" when mixing enums.
>
> The ordering has been closen so that no enum value is the same as a
> compatible hardware encoding
https://bugs.freedesktop.org/show_bug.cgi?id=102122
Bug ID: 102122
Summary: [softpipe] piglit fdo10370 regression
Product: Mesa
Version: git
Hardware: x86-64 (AMD64)
OS: Linux (All)
Status: NEW
Keywords: b
---
src/intel/vulkan/anv_gem.c | 28 +
src/intel/vulkan/anv_gem_stubs.c | 13 ++
src/intel/vulkan/anv_private.h | 4 +++
src/intel/vulkan/anv_queue.c | 53 +++-
4 files changed, 87 insertions(+), 11 deletions(-)
diff --g
---
src/intel/vulkan/anv_batch_chain.c | 19 -
src/intel/vulkan/anv_extensions.py | 5 ++
src/intel/vulkan/anv_queue.c | 142 -
3 files changed, 161 insertions(+), 5 deletions(-)
diff --git a/src/intel/vulkan/anv_batch_chain.c
b/src/intel/vulkan/a
In order to implement VK_KHR_external_fence, we need to back our fences
with something that's shareable. Since the kernel wait interface for
sync objects already supports waiting for multiple fences in one go, it
makes anv_WaitForFences much simpler if we only have one type of fence.
---
src/inte
---
src/intel/vulkan/anv_wsi.c | 9 ++---
1 file changed, 6 insertions(+), 3 deletions(-)
diff --git a/src/intel/vulkan/anv_wsi.c b/src/intel/vulkan/anv_wsi.c
index 9369f26..00edb22 100644
--- a/src/intel/vulkan/anv_wsi.c
+++ b/src/intel/vulkan/anv_wsi.c
@@ -364,22 +364,25 @@ VkResult anv_Get
It only applies to legacy BO fences.
---
src/intel/vulkan/anv_batch_chain.c | 2 +-
src/intel/vulkan/anv_private.h | 10 +-
src/intel/vulkan/anv_queue.c | 24
3 files changed, 18 insertions(+), 18 deletions(-)
diff --git a/src/intel/vulkan/anv_batch_cha
---
include/drm-uapi/drm.h | 19 +++
1 file changed, 19 insertions(+)
diff --git a/include/drm-uapi/drm.h b/include/drm-uapi/drm.h
index bf3674a..06d1d0f 100644
--- a/include/drm-uapi/drm.h
+++ b/include/drm-uapi/drm.h
@@ -712,6 +712,23 @@ struct drm_syncobj_handle {
__u32
---
src/vulkan/util/vk_alloc.h | 14 ++
1 file changed, 14 insertions(+)
diff --git a/src/vulkan/util/vk_alloc.h b/src/vulkan/util/vk_alloc.h
index 2915021..f58a806 100644
--- a/src/vulkan/util/vk_alloc.h
+++ b/src/vulkan/util/vk_alloc.h
@@ -37,6 +37,20 @@ vk_alloc(const VkAllocationC
This commit changes fences to work a bit more like BO semaphores.
Instead of the fence being a batch, it's simply a BO that gets added
to the validation list for the last execbuf call in the QueueSubmit
operation. It's a bit annoying finding the last submit in the execbuf
but this allows us to avo
---
src/intel/vulkan/anv_gem.c | 61
src/intel/vulkan/anv_gem_stubs.c | 20 +
src/intel/vulkan/anv_private.h | 5
3 files changed, 86 insertions(+)
diff --git a/src/intel/vulkan/anv_gem.c b/src/intel/vulkan/anv_gem.c
index 9e6b2bb
This little series adds support for the VK_KHR_external_fence family of
extensions. Most of the real work in implementing these extensions is
actually in the kernel. Once we have a DRM_SYNCOBJ_IOCTL_WAIT that does
what we need, the userspace bits are fairly straightforward. This series
can be fo
This is just a refactor, similar to what we did for semaphores, in
preparation for handling VK_KHR_external_fence.
---
src/intel/vulkan/anv_batch_chain.c | 22 --
src/intel/vulkan/anv_private.h | 42 ++-
src/intel/vulkan/anv_queue.c | 144 ++-
On Tue, Aug 8, 2017 at 5:48 PM, Dieter Nützel wrote:
> Hello Alex,
>
> which is the 'right' way to switch to (from amd-staging-4.11)?
> 4.13-rx UP (like Phoronix 'found';-)) show nice improvements.
amd-staging-4.12 is our current internal development branch.
amd-staging-drm-next, is just the amd-
Hello Alex,
which is the 'right' way to switch to (from amd-staging-4.11)?
4.13-rx UP (like Phoronix 'found';-)) show nice improvements.
Sadly I'm hit by the BAD 'e1000e' nic regression.
NetworkManager switch it on and off over and over again, since 4.11 UP.
--- Daniel/Intel???
06:00.0 Ether
2017-08-07 2:18 GMT+00:00 Timothy Arceri :
> The Deus Ex: Mankind Divided shaders go from spending ~20 seconds
> in the GLSL IR compilers front-end down to ~18.5 seconds on a
> Ryzen 1800X.
>
> Tested by compiling once with shader-db then deleting the index file
> from the shader cache and compilin
2017-08-07 2:18 GMT+00:00 Timothy Arceri :
> Also add a comment that this should only be used by the ir_reader
> interface for testing purposes.
> ---
> src/compiler/glsl/ir.cpp | 8 ++--
> src/compiler/glsl/ir.h | 14 +++---
> 2 files changed, 9 insertions(+), 13 deletions(-)
>
> d
Ping
On 01.08.2017 12:49, Nicolai Hähnle wrote:
Hi all,
I sent a v1 of this around ~6 weeks ago. Since then, I brought some related
issues up with the OpenGL WG, and we now have a rule clarification in
GLSL 4.60 (which is intended to apply to earlier versions as well). This
updated series imple
I'm not quite sure if the increase in size of the base ir class
versus the reduced overhead and added code is worth it?
Some numbers would be nice. A comment below
2017-08-07 2:18 GMT+00:00 Timothy Arceri :
> Extra validation is added to ir_validate to make sure this is
> always updated to the cor
On Mon, Aug 7, 2017 at 6:32 PM, Connor Abbott wrote:
> From: Connor Abbott
>
> This series implements VK_AMD_shader_ballot for radv. This extension
> builds on VK_EXT_shader_subgroup_ballot and VK_EXT_shader_subgroup_vote
> by adding a number of reductions across a subgroup (or wavefront in AMD
>
On 2017-08-06 21:18:23, Iago Toral Quiroga wrote:
> The OpenGL 4.6 specs have been updated so that GetTextureParameter*
> with a texture object with an incompatible TEXTURE_TARGET should now
> report INVALID_OPERATION instead of INVALID_ENUM.
>
> Fixes:
> KHR-GL45.direct_state_access.textures_para
On Tue, 2017-08-08 at 19:11 +0100, Emil Velikov wrote:
> On 2 August 2017 at 21:24, Adam Jackson wrote:
> > The device extension string is expected to contain the name of the
> > extension defining what kind of device it is, so the caller can know
> > what kinds of operations it can perform with i
On Tuesday, August 8, 2017 12:26:41 AM PDT Kenneth Graunke wrote:
> We don't push UBOs on Gen6 currently, so there's no need for the
> larger alignment value.
>
> Cc: "17.2"
> ---
> src/mesa/drivers/dri/i965/brw_context.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/
On 8 August 2017 at 08:23, Michel Dänzer wrote:
> From: Michel Dänzer
>
> No need to manually look for the library files anymore with current
> LLVM. This sidesteps the manual method failing when LLVM was built with
> -DLLVM_APPEND_VC_REV=ON.
>
IIRC you recently spotted that the Suse packaging wa
On 2 August 2017 at 21:24, Adam Jackson wrote:
> The device extension string is expected to contain the name of the
> extension defining what kind of device it is, so the caller can know
> what kinds of operations it can perform with it. So that string had
> better be non-empty, hence this trivial
On Tue, Aug 8, 2017 at 6:01 PM, Ilia Mirkin wrote:
> On Tue, Aug 8, 2017 at 11:52 AM, Adam Jackson wrote:
>> The only difference from the EXT version is bumping the minmax to 16, so
>> just hit all the drivers at once.
>>
>> Signed-off-by: Adam Jackson
>> ---
>> docs/features.txt
Acked-by: Marek Olšák
Marek
On Tue, Aug 8, 2017 at 9:23 AM, Michel Dänzer wrote:
> From: Michel Dänzer
>
> No need to manually look for the library files anymore with current
> LLVM. This sidesteps the manual method failing when LLVM was built with
> -DLLVM_APPEND_VC_REV=ON.
>
> (This might al
Reviewed-by: Marek Olšák
Marek
On Mon, Aug 7, 2017 at 12:41 PM, Samuel Pitoiset
wrote:
> Signed-off-by: Samuel Pitoiset
> ---
> src/gallium/drivers/radeonsi/si_shader.c | 2 --
> 1 file changed, 2 deletions(-)
>
> diff --git a/src/gallium/drivers/radeonsi/si_shader.c
> b/src/gallium/drivers/
On Mon, 2017-08-07 at 09:24 +0100, Daniel Stone wrote:
> On 2 August 2017 at 21:24, Adam Jackson wrote:
> > +Overview
> > +
> > +A system may support multiple devices and multiple window systems. For
> > +example, a Wayland environment may drive multiple GPUs and support both
> > +GLX
On Tue, Aug 8, 2017 at 11:56 AM, Ilia Mirkin wrote:
> On Tue, Aug 8, 2017 at 12:50 PM, Rob Herring wrote:
>> Commit e794f8bf8bdb ("gallium: move loading of drirc to pipe-loader")
>> moved the option cache to the pipe_loader_device. However, the
>> screen->dev pointer is not set when dri_init_opti
On 8 August 2017 at 16:10, Eric Engestrom wrote:
> On Saturday, 2017-08-05 00:25:49 +0100, Emil Velikov wrote:
>> From: Emil Velikov
>>
>> As mentioned in previous commit the negative tests in dEQP expect the
>> arguments to be evaluated in particular order.
>
> The spec doesn't say that, so the
On 8 August 2017 at 17:34, Eric Engestrom wrote:
> On Saturday, 2017-08-05 00:25:52 +0100, Emil Velikov wrote:
>> From: Emil Velikov
>>
>> The function can be called only when the type is EGL_WINDOW_BIT.
>> Remove the unneeded switch statement.
>>
>> Signed-off-by: Emil Velikov
>> ---
>> src/eg
Currently, when the array is full it is resized but it can grow
over and over because we don't try to re-use descriptor slots.
v3: - use new idalloc gallium module
Signed-off-by: Samuel Pitoiset
---
src/gallium/drivers/radeonsi/si_descriptors.c | 57 +--
src/gallium/driv
Using VRAM address as bindless handles is not a good idea because
we have to use LLVMIntToPTr and the LLVM CSE pass can't optimize
because it has no information about the pointer.
Instead, use slots indexes like the existing descriptors. Note
that we use fixed 16-dword slots for both samplers and
Looks like it's useless to initialize that field when CE is
unused. This will also allow to declare more than 64 elements
for the array of bindless descriptors.
Signed-off-by: Samuel Pitoiset
Reviewed-by: Marek Olšák
---
src/gallium/drivers/radeonsi/si_descriptors.c | 7 ---
1 file changed,
The number of bindless descriptors is dynamic and we definitely
have to support more than 256 slots.
Signed-off-by: Samuel Pitoiset
Reviewed-by: Marek Olšák
---
src/gallium/drivers/radeonsi/si_state.h | 10 +-
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/src/gallium/dri
To share common code between rw buffers and bindless descriptors.
v3: - rename to si_emit_global_shader_pointers()
Signed-off-by: Samuel Pitoiset
---
src/gallium/drivers/radeonsi/si_descriptors.c | 57 +++
1 file changed, 31 insertions(+), 26 deletions(-)
diff --git a/s
A new pair of user SGPR is needed for loading the bindless
descriptors from shaders. Because the descriptors are global for
all stages, there is no need to add separate indices for GFX9.
v3: - fix merged shaders on GFX9
v2: - fix declaring new bindless parameter
Signed-off-by: Samuel Pitoiset
--
Will be used for allocating bindless descriptor slots for
RadeonSI.
Signed-off-by: Samuel Pitoiset
---
src/gallium/auxiliary/Makefile.sources | 1 +
src/gallium/auxiliary/util/u_idalloc.h | 103 +
2 files changed, 104 insertions(+)
create mode 100644 src/galliu
On Tue, Aug 8, 2017 at 12:50 PM, Rob Herring wrote:
> Commit e794f8bf8bdb ("gallium: move loading of drirc to pipe-loader")
> moved the option cache to the pipe_loader_device. However, the
> screen->dev pointer is not set when dri_init_options() is called. Move
> the call to after the pipe_loader_
Commit e794f8bf8bdb ("gallium: move loading of drirc to pipe-loader")
moved the option cache to the pipe_loader_device. However, the
screen->dev pointer is not set when dri_init_options() is called. Move
the call to after the pipe_loader_sw_probe_kms() call so screen->dev is
set. This mirrors the c
On Saturday, 2017-08-05 02:30:51 +0100, Emil Velikov wrote:
> On 5 August 2017 at 00:25, Emil Velikov wrote:
> > From: Emil Velikov
> >
> > The function can handle only window surfaces, so let's rename it
> > accordingly, killing the wrapper around it.
> >
> > Suggested-by: Eric Engestrom
> > Si
On Saturday, 2017-08-05 00:25:52 +0100, Emil Velikov wrote:
> From: Emil Velikov
>
> The function can be called only when the type is EGL_WINDOW_BIT.
> Remove the unneeded switch statement.
>
> Signed-off-by: Emil Velikov
> ---
> src/egl/drivers/dri2/platform_drm.c | 20 +++-
>
IMHO this should be implemented differently. There is no change of
functionality vis-a-vis the EXT version, just a name change. So a
single enable flag should be preserved, and a single function
entrypoint. I'd rename it to the plain name, and make the EXT one an
alias of the non-suffixed one.
Che
On Tue, Aug 8, 2017 at 11:52 AM, Adam Jackson wrote:
> Taken from c21e602b9fda1d3bbaecb08194592f67e6a0649b from
> OpenGL-Registry.
>
> Signed-off-by: Adam Jackson
Acked-by: Ilia Mirkin
I glanced through the diff, and looks like you carefully preserved
mesa's modifications to some of those file
On Tue, Aug 8, 2017 at 11:52 AM, Adam Jackson wrote:
> The only difference from the EXT version is bumping the minmax to 16, so
> just hit all the drivers at once.
>
> Signed-off-by: Adam Jackson
> ---
> docs/features.txt| 4 +++-
> src/glx/glxextensions.c
The only difference from the EXT version is bumping the minmax to 16, so
just hit all the drivers at once.
Signed-off-by: Adam Jackson
---
docs/features.txt| 4 +++-
src/glx/glxextensions.c | 1 +
src/glx/glxextensions.h | 1 +
Identical to the EXT version.
Signed-off-by: Adam Jackson
---
docs/features.txt| 2 +-
src/mapi/glapi/gen/GL4x.xml | 9 +
src/mesa/drivers/dri/i965/intel_extensions.c | 1 +
src/mesa/main/dlist.c| 20
Taken from c21e602b9fda1d3bbaecb08194592f67e6a0649b from
OpenGL-Registry.
Signed-off-by: Adam Jackson
---
include/GL/glcorearb.h | 66 ++-
include/GL/glext.h | 82 +++--
include/GL/glxext.h| 2 +-
include/GL/wglext.h| 2 +-
src/ma
I hadn't seen Plamena's patches for these before writing them, and I think
I got right most of the issues that were caught in review (namely: 16x
minmax for aniso, updated headers and xml for new dispatch for polygon
offset clamp), so in the interest of saving time...
- ajax
_
Am 08.08.2017 um 17:45 schrieb Ilia Mirkin:
> On Tue, Aug 8, 2017 at 11:34 AM, Roland Scheidegger
> wrote:
>> Am 08.08.2017 um 17:21 schrieb Ilia Mirkin:
>>> On Tue, Aug 8, 2017 at 11:16 AM, Adam Jackson wrote:
On Wed, 2017-06-28 at 16:50 +0300, Plamena Manolova wrote:
> ARB_texture_fil
Am 08.08.2017 um 17:21 schrieb Ilia Mirkin:
> On Tue, Aug 8, 2017 at 11:16 AM, Adam Jackson wrote:
>> On Wed, 2017-06-28 at 16:50 +0300, Plamena Manolova wrote:
>>> ARB_texture_filter_anisotropic is the ARB variation of
>>> EXT_texture_fitter_anisotropic and it operates in the
>>> same way, so the
On Tue, Aug 8, 2017 at 11:34 AM, Roland Scheidegger wrote:
> Am 08.08.2017 um 17:21 schrieb Ilia Mirkin:
>> On Tue, Aug 8, 2017 at 11:16 AM, Adam Jackson wrote:
>>> On Wed, 2017-06-28 at 16:50 +0300, Plamena Manolova wrote:
ARB_texture_filter_anisotropic is the ARB variation of
EXT_text
On Tue, Aug 8, 2017 at 11:16 AM, Adam Jackson wrote:
> On Wed, 2017-06-28 at 16:50 +0300, Plamena Manolova wrote:
>> ARB_texture_filter_anisotropic is the ARB variation of
>> EXT_texture_fitter_anisotropic and it operates in the
>> same way, so there's no reason not to advertise it.
>
> Nak. The A
On Wed, 2017-06-28 at 16:50 +0300, Plamena Manolova wrote:
> ARB_texture_filter_anisotropic is the ARB variation of
> EXT_texture_fitter_anisotropic and it operates in the
> same way, so there's no reason not to advertise it.
Nak. The ARB version bups the minmax anisotropy to 16x, see issue 2 in
t
On Saturday, 2017-08-05 00:25:49 +0100, Emil Velikov wrote:
> From: Emil Velikov
>
> As mentioned in previous commit the negative tests in dEQP expect the
> arguments to be evaluated in particular order.
The spec doesn't say that, so the test is wrong.
Changing it in Mesa doesn't hurt though, so
https://bugs.freedesktop.org/show_bug.cgi?id=102067
Emil Velikov changed:
What|Removed |Added
Resolution|--- |FIXED
Status|NEW
From: Emil Velikov
If we have an invalid display fed into the functions, the display lookup
will return NULL. Thus as we attempt to get the platform type, we'll
deref. it leading to a crash.
Keep in mind that this will not happen if Mesa is built without X11 or
when the legacy eglCreate*Surface
On Saturday, 2017-08-05 00:25:48 +0100, Emil Velikov wrote:
> From: Emil Velikov
>
> The basic (null) check is identical across all backends.
> Just move it to the top.
>
> v2:
> - Split the WINDOW vs PIXMAP into separate patches
> - Move check after the dpy and config - dEQP expects so
>
> C
On Tuesday, 2017-08-08 08:02:27 +0300, Tapani Pälli wrote:
>
>
> On 08/05/2017 02:25 AM, Emil Velikov wrote:
> > From: Emil Velikov
> >
> > The code in _eglCreatePixmapSurfaceCommon() already has a NULL check
>
> _eglCreateWindowSurfaceCommon()
>
> > which handles the condition. There's no po
On 8 August 2017 at 08:21, Tapani Pälli wrote:
>
>
> On 08/05/2017 02:25 AM, Emil Velikov wrote:
>>
>> From: Emil Velikov
>>
>> As mentioned in previous commit the negative tests in dEQP expect the
>> arguments to be evaluated in particular order.
>>
>> Namely - first the dpy, then the config, fo
https://bugs.freedesktop.org/show_bug.cgi?id=102077
Laurent carlier changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|---
On 7 August 2017 at 19:13, Vinson Lee wrote:
> On Mon, Aug 7, 2017 at 8:18 AM, Emil Velikov wrote:
>> On 5 August 2017 at 21:53, Vinson Lee wrote:
>>> Fixes build error on CentOS 6.9.
>>>
>>> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102052
>>> Fixes: 5c007203b73d ("configure.ac: dr
https://bugs.freedesktop.org/show_bug.cgi?id=102030
--- Comment #3 from Janpieter Sollie ---
confirmed on -RC3 (initial report on RC2)
--
You are receiving this mail because:
You are the assignee for the bug.
You are the QA Contact for the bug.___
mes
From: Zhongmin Wu
Add plumbing to allow creation of per display surface out fence.
Currently enabled only on android, since the system expects a valid
fd in ANativeWindow::{queue,cancel}Buffer. We pass a fd of -1 with
which native applications such as flatland fail. The patch enables
explicit sy
On 08/08/2017 12:00 PM, Eric Engestrom wrote:
On Tuesday, 2017-08-08 11:31:36 +0300, Tapani Pälli wrote:
Currently swrastGetDrawableInfo always initializes w and h, patch
refactors function as x11_get_drawable_info that returns success and
sets the values only if no error happened. Add swrastG
On 07/31/2017 10:59 PM, Samuel Pitoiset wrote:
This reverts commit fcbb93e860246375d03f280f927f79d3645a8988 and
also commit 7c5b204e38d8cae70f5bf26e7223da5bc448bb5c to avoid
compilation errors.
Basically, the parameter indexes look wrong when a non-bindless
sampler is declared inside a nested
On 8 August 2017 at 10:21, Samuel Pitoiset wrote:
>
>
> On 08/08/2017 07:34 AM, Timothy Arceri wrote:
>>
>> On 01/08/17 19:37, Timothy Arceri wrote:
>>>
>>> On 01/08/17 18:07, Samuel Pitoiset wrote:
Don't you think it's just safer to revert the bad commit for 17.2 and
fix it later o
On 8 August 2017 at 10:00, Eric Engestrom wrote:
> On Tuesday, 2017-08-08 11:31:36 +0300, Tapani Pälli wrote:
>> Currently swrastGetDrawableInfo always initializes w and h, patch
>> refactors function as x11_get_drawable_info that returns success and
>> sets the values only if no error happened. A
On Tue, Aug 8, 2017 at 12:58 AM, Rob Herring wrote:
> Use the common pipe_screen ref counting and fd hashing functions.
> The mutex can be dropped as the pipe loader serializes the
> create_screen() and destroy() calls.
>
> Signed-off-by: Rob Herring
> Cc: Christian Gmeiner
> Cc: Wladimir J. van
On 08/08/2017 07:34 AM, Timothy Arceri wrote:
On 01/08/17 19:37, Timothy Arceri wrote:
On 01/08/17 18:07, Samuel Pitoiset wrote:
Don't you think it's just safer to revert the bad commit for 17.2 and
fix it later on?
I'm not overly worried. If you really want to go that way we can, but
I do
On Tuesday, 2017-08-08 11:31:36 +0300, Tapani Pälli wrote:
> Currently swrastGetDrawableInfo always initializes w and h, patch
> refactors function as x11_get_drawable_info that returns success and
> sets the values only if no error happened. Add swrastGetDrawableInfo
> wrapper function as expected
Quoting Kenneth Graunke (2017-08-07 22:50:42)
> RELOC_NEEDS_GGTT is only meaningful on Sandybridge - it's skipped on
> other generations - so this has no purpose. Just use rw_bo().
Furthermore the w/a is for a pipecontrol, so we should be able to
trim it down further. Though it shouldn't make any
1 - 100 of 111 matches
Mail list logo