Re: [Mesa-dev] [PATCH 4/4] radv: initialize the per-queue descriptor BO only once

2019-01-18 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen for the series. On Thu, Jan 17, 2019 at 6:08 PM Samuel Pitoiset wrote: > > Totally useless to write the descriptors inside the loop. > > Signed-off-by: Samuel Pitoiset > --- > src/amd/vulkan/radv_device.c | 47 ++--

Re: [Mesa-dev] [PATCH] radv: avoid context rolls when binding graphics pipelines

2019-01-18 Thread Bas Nieuwenhuizen
Besides CTS, I'd appreciate if you can test with Talos, as that was the msot affected by bugs in this code. Otherwise, Reviewed-by: Bas Nieuwenhuizen On Mon, Jan 14, 2019 at 5:02 PM Rhys Perry wrote: > > It's common in some applications to bind a new graphics pipeline without >

Re: [Mesa-dev] [PATCH] radv: avoid context rolls when binding graphics pipelines

2019-01-15 Thread Bas Nieuwenhuizen
On Mon, Jan 14, 2019 at 5:12 PM Rhys Perry wrote: > > I did and found small improvements in Rise of the Tomb Raider. I > measured framerates ~104.3% that of without the changes for the > Geothermal Valley scene, ~101.2% for Spine of the Mountain and ~102.3% > for Prophets Tomb. My main question

Re: [Mesa-dev] [PATCH] radv: prevent dirtying of dynamic state when it does not change

2019-01-15 Thread Bas Nieuwenhuizen
On Tue, Jan 15, 2019 at 10:59 PM Rhys Perry wrote: > > DXVK often sets dynamic state without actually changing it. > > Signed-off-by: Rhys Perry > --- > src/amd/vulkan/radv_cmd_buffer.c | 92 ++-- > 1 file changed, 76 insertions(+), 16 deletions(-) > > diff --git

Re: [Mesa-dev] [PATCH] radeonsi: also apply the GS hang workaround to draws without tessellation

2019-01-14 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Tue, Jan 15, 2019 at 12:33 AM Marek Olšák wrote: > > From: Marek Olšák > > ported from AMDVLK. > > Cc: 18.3 > --- > src/gallium/drivers/radeonsi/si_state_draw.c | 25 +++- > 1 file changed, 14 insertions(+), 11

Re: [Mesa-dev] [PATCH v2] ac: add missing 16-bit types to glsl_base_to_llvm_type()

2019-01-14 Thread Bas Nieuwenhuizen
On Mon, Jan 14, 2019 at 5:55 PM Samuel Pitoiset wrote: > > Fix crashes with > dEQP-VK.spirv_assembly.instruction.compute.workgroup_memory.*16 > > v2: - add INT16/UINT16 too > - update commit description > > Signed-off-by: Samuel Pitoiset > Reviewed-by: Bas Nieuw

Re: [Mesa-dev] [PATCH] ac: add missing GLSL_TYPE_FLOAT16 to glsl_base_to_llvm_type()

2019-01-14 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Mon, Jan 14, 2019 at 1:49 PM Samuel Pitoiset wrote: > > Fix a crash with > dEQP-VK.spirv_assembly.instruction.compute.workgroup_memory.float16 > > Signed-off-by: Samuel Pitoiset > --- > src/amd/common/ac_nir_to_llvm.c | 2 ++ > 1 fi

Re: [Mesa-dev] [PATCH v3 3/3] radv: add support for VK_EXT_memory_budget

2019-01-14 Thread Bas Nieuwenhuizen
On Wed, Jan 9, 2019, 2:37 PM Samuel Pitoiset A simple Vulkan extension that allows apps to query size and > usage of all exposed memory heaps. > > The different usage values are not really accurate because > they are per drm-fd, but they should be close enough. > > v3: - use atomic operations in

Re: [Mesa-dev] [PATCH 6/6] ac: use llvm.amdgcn.s.buffer.load

2019-01-13 Thread Bas Nieuwenhuizen
I think this is done with https://patchwork.freedesktop.org/series/55025/ ? On Sat, Jan 12, 2019 at 12:53 AM Marek Olšák wrote: > > From: Marek Olšák > > --- > src/amd/common/ac_llvm_build.c | 18 +- > .../drivers/radeonsi/si_shader_tgsi_mem.c | 4 ++-- >

Re: [Mesa-dev] [PATCH] radv: allow secondary command buffers to inherit unknown framebuffers

2019-01-12 Thread Bas Nieuwenhuizen
On Thu, Dec 20, 2018 at 8:05 PM Rhys Perry wrote: > > Fixes: f4e499ec79 ('radv: add initial non-conformant radv vulkan driver') > Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107986 > Signed-off-by: Rhys Perry > --- > src/amd/vulkan/radv_cmd_buffer.c | 59

Re: [Mesa-dev] [PATCH 2/2] amd/common/vi+: enable SMEM loads with GLC=1

2019-01-12 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen for both. On Thu, Jan 10, 2019 at 11:26 PM Nicolai Hähnle wrote: > > From: Nicolai Hähnle > > Only on LLVM 8.0+, which supports the new intrinsic. > --- > src/amd/common/ac_llvm_build.c | 10 +++--- > 1 file changed, 7 inser

Re: [Mesa-dev] [PATCH v2] radv: skip draws with instance_count == 0

2019-01-09 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Wed, Jan 9, 2019 at 10:03 AM Samuel Pitoiset wrote: > > Loosely based on RadeonSI. > > v2: - do not check for indexed draws > > Signed-off-by: Samuel Pitoiset > --- > src/amd/vulkan/radv_cmd_buffer.c | 13 + > 1

Re: [Mesa-dev] [PATCH] st/mesa: don't leak pipe_surface if pipe_context is not current

2019-01-08 Thread Bas Nieuwenhuizen
On Tue, Jan 8, 2019 at 5:17 PM Marek Olšák wrote: > > From: Marek Olšák > > We have found some pipe_surface leaks internally. > > This is the same code as surface_destroy in radeonsi. > Ideally, surface_destroy would be in pipe_screen. > No, pipe_surfaces are not context objects. > > Cc: 18.3

Re: [Mesa-dev] [PATCH] ac/nir_to_llvm: add missing parentheses in get_inst_tessfactor_writemask()

2019-01-07 Thread Bas Nieuwenhuizen
I think this got fixed by https://gitlab.freedesktop.org/mesa/mesa/commit/be6cee51c06dc72ac159bd75b4201c61952515bd already? On Tue, Jan 8, 2019 at 12:28 AM Timothy Arceri wrote: > > Cc: Marek Olšák > --- > src/amd/common/ac_nir_to_llvm.c | 4 ++-- > 1 file changed, 2 insertions(+), 2

Re: [Mesa-dev] [PATCH 3/3] radv: add support for VK_EXT_memory_budget

2019-01-07 Thread Bas Nieuwenhuizen
On Mon, Jan 7, 2019 at 6:20 PM Samuel Pitoiset wrote: > > > On 1/7/19 6:06 PM, Alex Smith wrote: > > Hi Samuel, > > Thanks for implementing this - I've been wanting this extension for a while > so it's good it's finally available. > > This is just reporting the total heap sizes as the budget,

Re: [Mesa-dev] [PATCH] radv: skip draws with instance_count == 0

2019-01-07 Thread Bas Nieuwenhuizen
On Mon, Jan 7, 2019 at 2:45 PM Samuel Pitoiset wrote: > > Loosely based on RadeonSI. > > Signed-off-by: Samuel Pitoiset > --- > src/amd/vulkan/radv_cmd_buffer.c | 14 ++ > 1 file changed, 14 insertions(+) > > diff --git a/src/amd/vulkan/radv_cmd_buffer.c >

Re: [Mesa-dev] [PATCH] anv/android: handle storage images in vkGetSwapchainGrallocUsageANDROID

2019-01-07 Thread Bas Nieuwenhuizen
On Mon, Jan 7, 2019 at 1:23 PM Tapani Pälli wrote: > > > > On 1/7/19 1:28 PM, Bas Nieuwenhuizen wrote: > > On Mon, Jan 7, 2019 at 11:54 AM Tapani Pälli wrote: > >> > >> > >> > >> On 1/7/19 11:56 AM, Bas Nieuwenhuizen wrote: > >

Re: [Mesa-dev] [PATCH] anv/android: handle storage images in vkGetSwapchainGrallocUsageANDROID

2019-01-07 Thread Bas Nieuwenhuizen
On Mon, Jan 7, 2019 at 11:54 AM Tapani Pälli wrote: > > > > On 1/7/19 11:56 AM, Bas Nieuwenhuizen wrote: > > On Wed, Dec 5, 2018 at 1:05 PM Tapani Pälli wrote: > >> > >> > >> > >> On 12/5/18 2:00 PM, Bas Nieuwenhuizen wrote: > >&

Re: [Mesa-dev] [PATCH] anv/android: handle storage images in vkGetSwapchainGrallocUsageANDROID

2019-01-07 Thread Bas Nieuwenhuizen
On Wed, Dec 5, 2018 at 1:05 PM Tapani Pälli wrote: > > > > On 12/5/18 2:00 PM, Bas Nieuwenhuizen wrote: > > On Wed, Dec 5, 2018 at 12:51 PM Tapani Pälli wrote: > >> > >> > >> > >> On 12/5/18 1:44 PM, Bas Nieuwenhuizen wrote: > >&

Re: [Mesa-dev] [PATCH] nir: avoid uninitialized variable warning

2019-01-06 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Mon, Jan 7, 2019 at 12:35 AM Timothy Arceri wrote: > > Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=109231 > --- > src/compiler/nir/nir_loop_analyze.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --g

Re: [Mesa-dev] NIR constant problem for GPU which doesn't have native integer support

2019-01-03 Thread Bas Nieuwenhuizen
On Thu, Jan 3, 2019 at 6:59 PM Jason Ekstrand wrote: > > On Thu, Jan 3, 2019 at 3:39 AM Erik Faye-Lund > wrote: >> >> On Wed, 2019-01-02 at 10:16 -0600, Jason Ekstrand wrote: >> > On Wed, Jan 2, 2019 at 9:43 AM Ilia Mirkin >> > wrote: >> > > Have a look at the first 4 patches in the series

Re: [Mesa-dev] [PATCH] meson, anv: Add inc_vulkan to include directories

2018-12-29 Thread Bas Nieuwenhuizen
On Sat, Dec 29, 2018 at 5:23 PM Jan Vesely wrote: > > On Wed, 2018-12-26 at 15:26 +, Eric Engestrom wrote: > > On Tuesday, 2018-12-25 23:09:53 +0100, Jan Vesely wrote: > > > Guess my meson-fu is still pretty weak. > > > Now I see the build failure again: > > > In file included from

[Mesa-dev] [PATCH] radv: Implement buffer stores with less than 4 components.

2018-12-24 Thread Bas Nieuwenhuizen
We started using it in the btoi paths for r32g32b32, and the LLVM IR checker will complain about it because we end up with intrinsics with the wrong type extension in the name. Fixes: 593996bc02 ("radv: implement buffer to image operations for R32G32B32") --- src/amd/common/ac_nir_to_llvm.c | 19

Re: [Mesa-dev] [PATCH] radv: compute optimal VM alignment for imported buffers

2018-12-20 Thread Bas Nieuwenhuizen
Happy to have this patch, but all descriptions and comments say this is an optimization Can we get some comment that this avoids hangs? Otherwise Reviewed-by: Bas Nieuwenhuizen On Thu, Dec 20, 2018 at 3:23 PM Samuel Pitoiset wrote: > > This fixes GPU hangs on GFX9 with

Re: [Mesa-dev] [PATCH] radv: fix subpass image transitions with multiviews

2018-12-20 Thread Bas Nieuwenhuizen
Is not ideal but will have to do for now Reviewed-by: Bas Nieuwenhuizen On Thu, Dec 20, 2018 at 12:01 PM Samuel Pitoiset wrote: > > The driver needs to decompress all image layers if a fast > depth/color clear has been performed. > > Signed-off-by: Samuel Pitoiset > ---

Re: [Mesa-dev] [PATCH 2/5] ac/nir: restrict fmask lookup to image load intrinsics

2018-12-20 Thread Bas Nieuwenhuizen
On Tue, Dec 18, 2018, 9:57 PM Samuel Pitoiset We don't ever want to do the fmask lookup on a atomic or > store, the fmask should have been decompressed if the > surface has been moved to IMAGE_LAYOUT. > Clarify the description to be a *storage* image layout? > > Original patch by Dave Airlie. >

Re: [Mesa-dev] [PATCH] radv: drop the amdgpu-skip-threshold=1 workaround for LLVM 8

2018-12-20 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Wed, Dec 19, 2018, 6:13 PM Samuel Pitoiset This workaround has been introduced by 135e4d434f6 for fixing > DXVK GPU hangs with many games. It is no longer needed since > LLVM r345718. > > Signed-off-by: Samuel Pitoiset > --- > src/amd/vu

[Mesa-dev] [PATCH v2] radv: Work around non-renderable 128bpp compressed 3d textures on GFX9.

2018-12-20 Thread Bas Nieuwenhuizen
Exactly what title says, the new addrlib does not allow the above with certain dimensions that the CTS seems to hit. Work around it by not allowing the app to render to it via compat with other 128bpp formats and do not render to it ourselves during copies. Fixes: 776b9113656 "amd/addrlib:

Re: [Mesa-dev] [PATCH] amd/surface: fix setting of ADDR2_SURFACE_FLAGS::color

2018-12-20 Thread Bas Nieuwenhuizen
cating a non-renderable surface without driver changes goes against actual usage? > > Marek > > On Tue, Dec 18, 2018 at 12:50 PM Haehnle, Nicolai > wrote: >> >> On 18.12.18 18:36, Bas Nieuwenhuizen wrote: >> > Hi Nicolai, >> > >> > I happened to b

Re: [Mesa-dev] [PATCH v2 2/2] radv/query: Use 1-bit booleans in query shaders

2018-12-19 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Wed, Dec 19, 2018 at 11:05 PM Jason Ekstrand wrote: > > Fixes: 44227453ec03f "nir: Switch to using 1-bit Booleans for almost..." > Reviewed-by: Rhys Perry > --- > src/amd/vulkan/radv_query.c | 42 ++

Re: [Mesa-dev] [PATCH 1/2] radv/query: Add a nir_flag_set helper

2018-12-19 Thread Bas Nieuwenhuizen
ice *device) > return num_db; > } > > + > +static nir_ssa_def *nir_flag_set(nir_builder *b, nir_ssa_def *flags, > uint32_t flag) Can I ask you to rename this to something like nir_test_flag? nir_flag_set would first make me think of the verb in imperative, which would result in

Re: [Mesa-dev] [PATCH] amd/surface: fix setting of ADDR2_SURFACE_FLAGS::color

2018-12-18 Thread Bas Nieuwenhuizen
Hi Nicolai, I happened to be writing something similar which also fixes up radv to never render to those surfaces as r32g32b32a32: https://patchwork.freedesktop.org/series/54172/ I can split out the radv specific stuff and this one is r-b after than. Thanks, Bas On Tue, Dec 18, 2018 at 5:37 PM

[Mesa-dev] [PATCH] radv: Work around non-renderable 128bpp compressed 3d textures on GFX9.

2018-12-17 Thread Bas Nieuwenhuizen
Exactly what title says, the new addrlib does not allow the above with certain dimensions that the CTS seems to hit. Work around it by not allowing the app to render to it via compat with other 128bpp formats and do not render to it ourselves during copies. Fixes: 776b9113656 "amd/addrlib:

Re: [Mesa-dev] [PATCH mesa] drop autotools

2018-12-17 Thread Bas Nieuwenhuizen
On Mon, Dec 17, 2018 at 6:33 PM Juan A. Suarez Romero wrote: > > On Mon, 2018-12-03 at 10:21 +, Eric Engestrom wrote: > > Cc: Emil Velikov > > Cc: Andres Gomez > > Cc: Juan A. Suarez Romero > > Cc: Dylan Baker > > Signed-off-by: Eric Engestrom > > --- > > This patch depends on the

Re: [Mesa-dev] [PATCH] radv: Fix a stupid if in gather_intrinsic_info

2018-12-16 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Sun, Dec 16, 2018 at 7:59 AM Jason Ekstrand wrote: > > --- > src/amd/vulkan/radv_shader_info.c | 18 +- > 1 file changed, 9 insertions(+), 9 deletions(-) > > diff --git a/src/amd/vulkan/radv_shader_info.c > b/src/amd/vulk

Re: [Mesa-dev] [PATCH] radv: don't set surf_index for stencil-only images

2018-12-15 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Fri, Dec 14, 2018 at 7:32 PM Rhys Perry wrote: > > Fixes: f8d5b377c8b ('radv: set cb base tile swizzles for MRT speedups (v4)') > Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=108116 > Signed-off-by: Rhys Perry > --- > Unfortunat

Re: [Mesa-dev] [PATCH v2] docs: Document GitLab merge request process (email alternative)

2018-12-13 Thread Bas Nieuwenhuizen
On Thu, Dec 13, 2018 at 4:52 PM Alex Deucher wrote: > > On Wed, Dec 12, 2018 at 3:42 AM Samuel Pitoiset > wrote: > > > > Personally, I will continue to use the list, at least for a simplicity > > point of view. I'm not sure if using a new tool will improve quality and > > code review process. >

Re: [Mesa-dev] [PATCH 2/2] radv: don't check if format is depth in radv_image_can_enable_hile()

2018-12-12 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen for both. On Wed, Dec 12, 2018 at 2:13 PM Samuel Pitoiset wrote: > > This is always TRUE if htile_size is not 0. > > Signed-off-by: Samuel Pitoiset > --- > src/amd/vulkan/radv_image.c | 1 - > 1 file changed, 1 deletion(-) > >

Re: [Mesa-dev] last call for autotools

2018-12-12 Thread Bas Nieuwenhuizen
On Wed, Dec 12, 2018 at 8:59 PM Marek Olšák wrote: > > There are 2 issues with meson: > * -DDEBUG is not present in debugoptimized builds. Do people expect -DDEBUG for debugoptimized? I would think that debug optimized would be an optimized build with debug symbols, but not expensive checks &

Re: [Mesa-dev] [PATCH v2] docs: Document GitLab merge request process (email alternative)

2018-12-11 Thread Bas Nieuwenhuizen
On Wed, Dec 12, 2018 at 1:06 AM Jason Ekstrand wrote: > > Ping? > > I see about 5 acks/reviews, 3 of which are from Intel which doesn't exactly > seem like overwhelming consensus. However, we also haven't had any debate in > a while. FWIW, I'm fine with it too. Acked-by:

Re: [Mesa-dev] [PATCH] radv: implement VK_EXT_sample_locations

2018-12-08 Thread Bas Nieuwenhuizen
On Sat, Dec 8, 2018 at 7:03 PM Rhys Perry wrote: > > A small number of questions/concerns: > > - sampleLocationCoordinateRange[1] should probably be set to 0.9375, > because of how the sample locations are encoded > - gl_SamplePosition doesn't seem like it would return the new sample >

Re: [Mesa-dev] [PATCH mesa] radv: drop unused variable

2018-12-05 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Wed, Dec 5, 2018 at 4:44 PM Eric Engestrom wrote: > > Added in 824cfc1ee5e0aba15b676 "radv: rework the TC-compat HTILE > hardware bug with COND_EXEC", but it is unused. > > Signed-off-by: Eric Engestrom > --- > src/amd/vulkan/ra

Re: [Mesa-dev] [PATCH] radv: expose VK_EXT_scalar_block_layout

2018-12-05 Thread Bas Nieuwenhuizen
On Wed, Dec 5, 2018 at 2:14 PM Samuel Pitoiset wrote: > > Nothing to do, the compiler already handles that. > > All new dEQP.VK.ubo.* and dEQP.VK.ssbo.* pass, except some > 16-bit tests that are quite related to fdo bug #108114. > > Signed-off-by: Samuel Pitoiset > --- >

Re: [Mesa-dev] [PATCH] anv/android: handle storage images in vkGetSwapchainGrallocUsageANDROID

2018-12-05 Thread Bas Nieuwenhuizen
On Wed, Dec 5, 2018 at 12:51 PM Tapani Pälli wrote: > > > > On 12/5/18 1:44 PM, Bas Nieuwenhuizen wrote: > > On Wed, Dec 5, 2018 at 12:37 PM Tapani Pälli wrote: > >> > >> > >> > >> On 12/5/18 1:22 PM, Bas Nieuwenhuizen wrote: > >&

Re: [Mesa-dev] [PATCH] anv/android: handle storage images in vkGetSwapchainGrallocUsageANDROID

2018-12-05 Thread Bas Nieuwenhuizen
On Wed, Dec 5, 2018 at 12:37 PM Tapani Pälli wrote: > > > > On 12/5/18 1:22 PM, Bas Nieuwenhuizen wrote: > > On Wed, Dec 5, 2018 at 12:15 PM Tapani Pälli wrote: > >> > >> > >> > >> On 12/5/18 1:01 PM, Bas Nieuwenhuizen wrote: > >>

Re: [Mesa-dev] [PATCH] anv/android: handle storage images in vkGetSwapchainGrallocUsageANDROID

2018-12-05 Thread Bas Nieuwenhuizen
On Wed, Dec 5, 2018 at 12:15 PM Tapani Pälli wrote: > > > > On 12/5/18 1:01 PM, Bas Nieuwenhuizen wrote: > > On Fri, Sep 7, 2018 at 12:54 AM Kevin Strasser > > wrote: > >> > >> Android P and earlier expect that the surface supports storage image

Re: [Mesa-dev] [PATCH] anv/android: handle storage images in vkGetSwapchainGrallocUsageANDROID

2018-12-05 Thread Bas Nieuwenhuizen
On Fri, Sep 7, 2018 at 12:54 AM Kevin Strasser wrote: > > Android P and earlier expect that the surface supports storage images, and > so many of the tests fail when the framework checks for that support. The > framework also includes various image format and usage combinations that are > invalid

Re: [Mesa-dev] [PATCH] radv: wait on the high 32 bits of timestamp queries

2018-12-05 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Wed, Dec 5, 2018 at 11:43 AM Samuel Pitoiset wrote: > > In case we are unlucky if the low part is 0x. > > Fixes: 5d6a560a29 ("radv: do not use the availability bit for timestamp > queries") > Signed-off-by: Samuel Pitoi

[Mesa-dev] [PATCH] anv/android: Do not reject storage images.

2018-12-05 Thread Bas Nieuwenhuizen
We do the ImageFormatProperties check already, and rejecting an usage flag when both ImageFormatProperties and the WSI (which is Android) support it is not allowed. Intel does support storage for some of the support WSI formats, such as R8G8B8A8_UNORM, and looking at the

Re: [Mesa-dev] [PATCH] radv: fix vkCmdCopyQueryoolResults() for timestamp queries

2018-12-04 Thread Bas Nieuwenhuizen
On Tue, Dec 4, 2018 at 4:52 PM Samuel Pitoiset wrote: > > Because WAIT_REG_MEM can only wait for a 32-bit value, it's not > safe to use it for timestamp queries. If we only wait on the low > 32 bits of a timestamp query we could be unlucky and the GPU > might hang. > > One possible fix is to emit

Re: [Mesa-dev] [AppVeyor] mesa master #9462 failed

2018-12-03 Thread Bas Nieuwenhuizen
This just broke for me locally: ../mesa/src/mesa/state_tracker/st_glsl_to_tgsi.cpp: In member function ‘virtual void glsl_to_tgsi_visitor::visit(ir_call*)’: ../mesa/src/mesa/state_tracker/st_glsl_to_tgsi.cpp:4075:9: error: ‘ir_intrinsic_begin_fragment_shader_ordering’ was not declared in this

Re: [Mesa-dev] [PATCH] radv: rework the TC-compat HTILE hardware bug with COND_EXEC

2018-12-03 Thread Bas Nieuwenhuizen
ce->instance->debug_flags & > RADV_DEBUG_NO_HIZ)) { > - radv_image_alloc_htile(image); > image->tc_compatible_htile = > image->surface.flags & RADEON_SURF_TC_C

Re: [Mesa-dev] [PATCH mesa] drop autotools

2018-12-03 Thread Bas Nieuwenhuizen
On Mon, Dec 3, 2018 at 7:30 PM Jan Vesely wrote: > > On Mon, 2018-12-03 at 10:21 +, Eric Engestrom wrote: > > Cc: Emil Velikov > > Cc: Andres Gomez > > Cc: Juan A. Suarez Romero > > Cc: Dylan Baker > > Signed-off-by: Eric Engestrom > > --- > > This patch depends on the releasing

Re: [Mesa-dev] [PATCH 1/2] nir: add a compiler option for disabling float comparison simplifications

2018-11-30 Thread Bas Nieuwenhuizen
On Fri, Nov 30, 2018 at 10:29 PM Jason Ekstrand wrote: > > On Fri, Nov 30, 2018 at 3:18 PM Ian Romanick wrote: >> >> On 11/29/2018 07:47 AM, Connor Abbott wrote: >> > On Thu, Nov 29, 2018 at 4:22 PM Jason Ekstrand >> > wrote: >> >> >> >> Can you provide some context for this? Those rules are

Re: [Mesa-dev] [PATCH 3/3] radeonsi: add memory management stress tests for GDS

2018-11-28 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen for the series. Any plans with the newly gotten GDS/OA support? On Tue, Nov 27, 2018 at 2:57 AM Marek Olšák wrote: > > From: Marek Olšák > > --- > src/gallium/drivers/radeonsi/si_pipe.c | 46 ++ > src/gallium/drivers/

Re: [Mesa-dev] [PATCH 4/7] winsys/amdgpu: clean up code around BO VM alignment

2018-11-28 Thread Bas Nieuwenhuizen
On Mon, Nov 26, 2018 at 8:54 PM Marek Olšák wrote: > > On Mon, Nov 26, 2018 at 8:46 AM Bas Nieuwenhuizen wrote: >> >> patches 4-7 are >> >> Reviewed-by: Bas Nieuwenhuizen >> >> though I agree with Christian that it would be nice to get a case &

Re: [Mesa-dev] [PATCH] radv: Align large buffers to the fragment size.

2018-11-26 Thread Bas Nieuwenhuizen
On Mon, 26 Nov 2018, 20:48 Marek Olšák You increase both virtual and physical alignment. Are you sure it's > necessary to increase the physical alignment? > Hence the v2 which should not do that. > > Marek > > On Sun, Nov 25, 2018 at 11:13 PM Bas Nieuwenhuizen < >

[Mesa-dev] [PATCH 3/3] radv: Check for shareable images in central place.

2018-11-26 Thread Bas Nieuwenhuizen
One place to put the logic makes things easier to change. --- src/amd/vulkan/radv_image.c | 10 +- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/src/amd/vulkan/radv_image.c b/src/amd/vulkan/radv_image.c index 7c0ecb89661..c459e81379b 100644 ---

[Mesa-dev] [PATCH 2/3] radv/android: Use buffer metadata to determine scanout compat.

2018-11-26 Thread Bas Nieuwenhuizen
These days we don't always allocate scanout compatible textures anymore. That does mean we have to fix the radv android WSI though. Fixes: b1444c9ccb0 "radv: Implement VK_ANDROID_native_buffer." --- src/amd/vulkan/radv_android.c | 73 ++-

[Mesa-dev] [PATCH 1/3] radv/android: Mark android WSI image as shareable.

2018-11-26 Thread Bas Nieuwenhuizen
Fixes: b1444c9ccb0 "radv: Implement VK_ANDROID_native_buffer." --- src/amd/vulkan/radv_android.c | 12 +++- 1 file changed, 11 insertions(+), 1 deletion(-) diff --git a/src/amd/vulkan/radv_android.c b/src/amd/vulkan/radv_android.c index f5d70825dd2..93799b87b8f 100644 ---

Re: [Mesa-dev] [PATCH 4/7] winsys/amdgpu: clean up code around BO VM alignment

2018-11-26 Thread Bas Nieuwenhuizen
patches 4-7 are Reviewed-by: Bas Nieuwenhuizen though I agree with Christian that it would be nice to get a case where 7 improves things before we submit it. For patches 1-3 I need to dive into the slab allocator before I have enough knowledge to review. On Sat, Nov 24, 2018 at 12:41 AM Marek

Re: [Mesa-dev] [PATCH] radeonsi: clean up primitive binning enablement

2018-11-26 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Sat, Nov 24, 2018 at 12:34 AM Marek Olšák wrote: > > From: Marek Olšák > > no change in behavior. > --- > src/gallium/drivers/radeonsi/si_pipe.c | 27 +++--- > 1 file changed, 16 insertions(+), 11 deletions(-) > &g

Re: [Mesa-dev] [PATCH mesa 2/2] build: stop defining unused VERSION

2018-11-26 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen for both. On Mon, Nov 26, 2018 at 5:14 AM Eric Engestrom wrote: > > Scons and autotools don't define it, and as of last commit nothing > uses it. > > `VERSION` is also a generic enough name that something somewhere will > eventually clash,

Re: [Mesa-dev] [PATCH mesa] anv: correctly use vulkan 1.0 by default

2018-11-26 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Mon, Nov 26, 2018 at 5:14 AM Lionel Landwerlin wrote: > > On 23/11/2018 19:47, Eric Engestrom wrote: > > Per chapter 3.2 "Instances": > >> Providing a NULL VkInstanceCreateInfo::pApplicationInfo or providing > >> an ap

Re: [Mesa-dev] [PATCH] radv: only allocate HTILE if depth/stencil attachment usage is used

2018-11-26 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Fri, Nov 23, 2018 at 5:04 PM Samuel Pitoiset wrote: > > Apparently some games allocate depth buffers without rendering > to them, this will avoid wasting memory if that image usage > flag is not set. > > Signed-off-by: Samuel Pitoiset >

[Mesa-dev] [PATCH v2] radv: Align large buffers to the fragment size.

2018-11-25 Thread Bas Nieuwenhuizen
Improves performance in Talos by about 15% (and significant improvements in RotR and possibly other but did not bench with final patch) on kernel 4.19 and earlier. On 4.20+ a similar effect comes from 433ca054949a "drm/amdgpu: try allocating VRAM as power of two" v2: Do not impact the alignment

[Mesa-dev] [PATCH] radv: Align large buffers to the fragment size.

2018-11-25 Thread Bas Nieuwenhuizen
Improves performance in Talos by about 15% (and significant improvements in RotR and possibly other but did not bench with final patch) on kernel 4.19 and earlier. On 4.20+ a similar effect comes from 433ca054949a "drm/amdgpu: try allocating VRAM as power of two" Given the simplicity of the

[Mesa-dev] [PATCH] radv: Clamp gfx9 image view extents to the allocated image extents.

2018-11-24 Thread Bas Nieuwenhuizen
Mirrors AMDVLK. Looks like if we go over the alignment of height we actually start to change the addressing. Seems like the extra miplevels actually work with this. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=108245 Fixes: f6cc15dccd5 "radv/gfx9: fix block compression texture views.

[Mesa-dev] [PATCH] radv: Fix opaque metadata descriptor last layer.

2018-11-24 Thread Bas Nieuwenhuizen
We used the layer count which results in an off by one error. Not sure this really affects anything. Fixes: f4e499ec791 "radv: add initial non-conformant radv vulkan driver" --- src/amd/vulkan/radv_image.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git

Re: [Mesa-dev] [PATCH 1/2] winsys/amdgpu: fix a buffer leak in amdgpu_bo_from_handle

2018-11-23 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen for both On Fri, Nov 23, 2018 at 7:53 PM Marek Olšák wrote: > > Ping > > On Mon, Nov 19, 2018 at 6:24 PM Marek Olšák wrote: >> >> From: Marek Olšák >> >> Cc: 18.2 18.3 >> --- >> src/gallium/winsys/amdgpu/

Re: [Mesa-dev] [PATCH 2/2] radv: correctly use vulkan 1.0 by default

2018-11-23 Thread Bas Nieuwenhuizen
Fixes: ffa15861ef7c924a33e1f "radv: UseEnumerateInstanceVersion for the > default version." > > Bas, I double-checked the spec and Niklas is right; why did you change > it to use the icd version instead? Was is fixing an issue somewhere? Copying behavior from ANV ;) No ob

Re: [Mesa-dev] [PATCH 2/2] radv: ignore subpass self-dependencies for CreateRenderPass() too

2018-11-23 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen for both. On Fri, Nov 23, 2018 at 9:45 AM Samuel Pitoiset wrote: > > We really need to refactor this... > > Signed-off-by: Samuel Pitoiset > --- > src/amd/vulkan/radv_pass.c | 10 ++ > 1 file changed, 10 insertions(+) > >

Re: [Mesa-dev] [PATCH 1/2] radv: add a predicate for reflecting DCC decompression state

2018-11-22 Thread Bas Nieuwenhuizen
So to complicate things, drawing with DCC enabled also compresses things again. (maybe set it on framebuffer CB emit time?) On Thu, Nov 22, 2018 at 1:47 PM Samuel Pitoiset wrote: > > It's somehow similar to the FCE predicate. > > Signed-off-by: Samuel Pitoiset > --- >

Re: [Mesa-dev] [RFC 2/5] nir: Add unordered comparisons and ordered fne

2018-11-22 Thread Bas Nieuwenhuizen
rc1") > +binop_compare("fne", tfloat, commutative, "!isnan(src0) && !isnan(src1) && > src0 == src1") > +binop_compare("fltu", tfloat, "", "!(src0 >= src1)") > +binop_compare("fgeu", tfloat, "", &q

Re: [Mesa-dev] [PATCH 1.5/2] ac/surface/gfx9: let addrlib choose the preferred swizzle kind

2018-11-21 Thread Bas Nieuwenhuizen
On Wed, Nov 21, 2018 at 9:33 PM Marek Olšák wrote: > > See what Gustaw said. Other than that: > > Reviewed-by: Marek Olšák With that change: Reviewed-by: Bas Nieuwenhuizen Tested-by: Bas Nieuwenhuizen > > Marek > > On Wed, Nov 21, 2018 at 10:21 AM Nicolai Hähnle wr

Re: [Mesa-dev] [PATCH] radv: remove useless sync after CmdClear{Color, DepthStencil}Image()

2018-11-21 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Wed, Nov 21, 2018 at 11:32 AM Samuel Pitoiset wrote: > > 'post_flush' is only set to NULL for the normal clear path > (ie. only vkCmdClearColorImage() and vkCmdClearDepthStencilImage() > are affected commands). > > Because these t

[Mesa-dev] [PATCH] vulkan: Allow storage images in the WSI.

2018-11-20 Thread Bas Nieuwenhuizen
Since apps also have to follow the ImageFormatProperties query, we can disallow formats that don't allow image stores (for AMD that would be SRGB formats). Note that this only affects anything if the app actually decides to use the flag. Had someone ask for this on IRC and at least on the AMD

Re: [Mesa-dev] [PATCH] radv: ignore subpass self-dependencies

2018-11-20 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Tue, Nov 20, 2018 at 4:41 PM Samuel Pitoiset wrote: > > Unnecessary as they allow the app to call vkCmdPipelineBarrier() > inside the render pass. > > Signed-off-by: Samuel Pitoiset > --- > src/amd/vulkan/radv_pass.c | 10 ++

Re: [Mesa-dev] [PATCH] ac/nir: fix intrinsic name string size

2018-11-20 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Tue, Nov 20, 2018 at 9:58 AM Samuel Pitoiset wrote: > > Fixes an assertion in SoTTR. > > Fixes: dd0172e865 ("radv: Use structured intrinsics instead of indexing > workaround for GFX9.") > Signed-off-by: Samuel Pitoiset > --- >

Re: [Mesa-dev] [PATCH 2/2] amd/addrlib: update Mesa's copy of addrlib

2018-11-20 Thread Bas Nieuwenhuizen
; >> > On Mon, Nov 19, 2018 at 7:15 PM Bas Nieuwenhuizen >> > wrote: >> >> >> >> So I tried to test this with radv and got a bunch of crashes in CTS, >> >> mostly around 3d image support: >> >> >> >> #3 0x771

Re: [Mesa-dev] [PATCH 2/2] amd/addrlib: update Mesa's copy of addrlib

2018-11-19 Thread Bas Nieuwenhuizen
So I tried to test this with radv and got a bunch of crashes in CTS, mostly around 3d image support: #3 0x771a9396 in __assert_fail () from /usr/lib/libc.so.6 #4 0x769da3b4 in Addr::V2::Gfx9Lib::HwlGetPreferredSurfaceSetting (this=0x57661b30, pIn=0x7fffd5f0,

Re: [Mesa-dev] [PATCH 1/2] vulkan: Update the XML and headers to 1.1.93

2018-11-19 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Mon, Nov 19, 2018 at 4:39 PM Jason Ekstrand wrote: > > --- > include/vulkan/vulkan_core.h | 35 --- > src/vulkan/registry/vk.xml | 84 +--- > 2 files changed, 98 insertions(+), 21 deletions(-)

Re: [Mesa-dev] [PATCH 00/14] freedreno: move all the things

2018-11-19 Thread Bas Nieuwenhuizen
With my limited knowledge of the parts up till now, this is Reviewed-by: Bas Nieuwenhuizen with the caveat that this might very well break the Android build. Have anyone to look at that or do I need to take a stab? On Mon, Nov 12, 2018 at 4:51 PM Rob Clark wrote: > > Moves drm and t

Re: [Mesa-dev] [PATCH] radv: don't force clearing the stencil aspect for LOAD_OP_DONT_CARE

2018-11-19 Thread Bas Nieuwenhuizen
On Mon, Nov 19, 2018 at 12:21 PM Samuel Pitoiset wrote: > > > > On 11/19/18 11:56 AM, Bas Nieuwenhuizen wrote: > > On Thu, Nov 15, 2018 at 10:51 AM Samuel Pitoiset > > wrote: > >> > >> This has been initially added for a Sascha demo (ie. deferredshado

Re: [Mesa-dev] [PATCH] radv: don't force clearing the stencil aspect for LOAD_OP_DONT_CARE

2018-11-19 Thread Bas Nieuwenhuizen
On Thu, Nov 15, 2018 at 10:51 AM Samuel Pitoiset wrote: > > This has been initially added for a Sascha demo (ie. deferredshadows), > but as far I can tell this seems to no longer be needed. > > No CTS changes on GFX9 and I didn't find a game that is affected. > > Signed-off-by: Samuel Pitoiset >

Re: [Mesa-dev] [PATCH] radv: Use structured intrinsics instead of indexing workaround for GFX9.

2018-11-16 Thread Bas Nieuwenhuizen
On Fri, Nov 16, 2018 at 9:34 AM Samuel Pitoiset wrote: > > > > On 11/12/18 10:51 PM, Bas Nieuwenhuizen wrote: > > These force the index to be used in the instruction so we don't need the > > workaround. > > > > Totals: > > SGPRS: 1321642 -> 13218

Re: [Mesa-dev] [PATCH 1/2] radv/winsys: remove the max IBs per submit limit for the fallback path

2018-11-15 Thread Bas Nieuwenhuizen
Nice, seems to halve CPU cost of submitting 100 command buffers in 1 submit with the simultaneous use flag set. Reviewed-by: Bas Nieuwenhuizen for the series. On Thu, Nov 15, 2018 at 11:27 AM Samuel Pitoiset wrote: > > The chained submission is the fastest path and it should now > be

Re: [Mesa-dev] [PATCH 2/2] radv: enable primitive binning by default

2018-11-15 Thread Bas Nieuwenhuizen
gt; + (!(device->instance->debug_flags & > RADV_DEBUG_NOBINNING) || > device->physical_device->rad_info.family == > CHIP_RAVEN || > device->physical_device->rad_info.family == > CHIP_RAVEN

Re: [Mesa-dev] [PATCH 2/2] radv: force-enable primitive binning for The Talos Principle

2018-11-15 Thread Bas Nieuwenhuizen
On Thu, Nov 15, 2018 at 9:06 AM Samuel Pitoiset wrote: > > > > On 11/15/18 12:16 AM, Bas Nieuwenhuizen wrote: > > On Wed, Nov 14, 2018 at 5:32 PM Samuel Pitoiset > > wrote: > >> > >> > >> > >> On 11/14/18 5:28 PM, Bas Nieuwenhuizen

Re: [Mesa-dev] [PATCH 2/2] radv: force-enable primitive binning for The Talos Principle

2018-11-14 Thread Bas Nieuwenhuizen
On Wed, Nov 14, 2018 at 5:32 PM Samuel Pitoiset wrote: > > > > On 11/14/18 5:28 PM, Bas Nieuwenhuizen wrote: > > On Wed, Nov 14, 2018 at 5:26 PM Samuel Pitoiset > > wrote: > >> > >> +5% in ultra settings for both 4k and 1080p. > >> > >

Re: [Mesa-dev] [PATCH 2/2] radv: force-enable primitive binning for The Talos Principle

2018-11-14 Thread Bas Nieuwenhuizen
On Wed, Nov 14, 2018 at 5:26 PM Samuel Pitoiset wrote: > > +5% in ultra settings for both 4k and 1080p. > > Signed-off-by: Samuel Pitoiset > --- > src/amd/vulkan/radv_device.c | 3 +++ > 1 file changed, 3 insertions(+) > > diff --git a/src/amd/vulkan/radv_device.c b/src/amd/vulkan/radv_device.c

Re: [Mesa-dev] [PATCH] radeonsi: fix video APIs on Raven2

2018-11-14 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Wed, Nov 14, 2018 at 12:51 AM Marek Olšák wrote: > > From: Marek Olšák > > This was missed when I added the new enum. > > Cc: 18.3 > --- > src/gallium/drivers/radeonsi/si_get.c | 9 ++--- > src/gallium/drivers/radeonsi/si_uvd.c

Re: [Mesa-dev] [PATCH 1/2] nir: add new nir_remove_dead_barriers() pass

2018-11-14 Thread Bas Nieuwenhuizen
On Wed, Nov 14, 2018 at 1:03 PM Timothy Arceri wrote: > > On 14/11/18 10:22 pm, Bas Nieuwenhuizen wrote: > > On Wed, Nov 14, 2018 at 6:30 AM Timothy Arceri > > wrote: > >> > >> Link time opts can remove unused outputs leaving behind useless > >>

Re: [Mesa-dev] [PATCH 1/2] nir: add new nir_remove_dead_barriers() pass

2018-11-14 Thread Bas Nieuwenhuizen
On Wed, Nov 14, 2018 at 6:30 AM Timothy Arceri wrote: > > Link time opts can remove unused outputs leaving behind useless > barriers. This pass is intended to be called after linking so > I've added it to nir_linking_helpers.c > > This pass removes some barriers from Witcher 3 (DXVK) Vulkan >

Re: [Mesa-dev] [PATCH 6/6] radv: implement fast HTILE clears for depth or stencil only on GFX9

2018-11-12 Thread Bas Nieuwenhuizen
On Mon, Nov 12, 2018 at 10:46 PM Connor Abbott wrote: > > On Mon, Nov 12, 2018 at 9:34 PM Bas Nieuwenhuizen > wrote: > > > > On Mon, Nov 12, 2018 at 5:55 PM Samuel Pitoiset > > wrote: > > > > > > This allows to fast clear the depth part (or the s

[Mesa-dev] [PATCH] radv: Use structured intrinsics instead of indexing workaround for GFX9.

2018-11-12 Thread Bas Nieuwenhuizen
These force the index to be used in the instruction so we don't need the workaround. Totals: SGPRS: 1321642 -> 1321802 (0.01 %) VGPRS: 943664 -> 943788 (0.01 %) Spilled SGPRs: 28468 -> 28480 (0.04 %) Spilled VGPRs: 88 -> 89 (1.14 %) Private memory VGPRs: 0 -> 0 (0.00 %) Scratch size: 80 -> 80

Re: [Mesa-dev] [PATCH 6/6] radv: implement fast HTILE clears for depth or stencil only on GFX9

2018-11-12 Thread Bas Nieuwenhuizen
On Mon, Nov 12, 2018 at 5:55 PM Samuel Pitoiset wrote: > > This allows to fast clear the depth part (or the stencil part) > of a depth+stencil surface when HTILE is enabled. I didn't test > on GFX8, so it's disabled currently. > > This gives a very nice boost, for example when clearing the depth

Re: [Mesa-dev] [PATCH v2] radv: set optimal OVERWRITE_COMBINER_WATERMARK on GFX9

2018-11-12 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Mon, Nov 12, 2018 at 9:44 AM Samuel Pitoiset wrote: > > Ported from RadeonSI. > > v2: - use the current subpass instead of the whole fb > > Signed-off-by: Samuel Pitoiset > --- > src/amd/vulkan/radv_cmd_buffer.c | 21 +++

Re: [Mesa-dev] [PATCH 1/2] radv: cleanup and document a Hawaii bug with offchip buffers

2018-11-12 Thread Bas Nieuwenhuizen
Not sure I agree this is a cleanup. The code here is independent of hardware, while the initialization of tess_offchip_block_dw_size in radv_device.c has the specific hardware this applies to. We add unnecessary asserts here and I'd argue the hardware bug should be explained when

Re: [Mesa-dev] [PATCH 2/2] radv: clean up setting partial_es_wave for distributed tess on VI

2018-11-12 Thread Bas Nieuwenhuizen
Reviewed-by: Bas Nieuwenhuizen On Thu, Nov 8, 2018 at 1:58 PM Samuel Pitoiset wrote: > > Only needed when the pipeline actually uses tessellation. I don't > think that changes anything, except improving readability. > > Signed-off-by: Samuel Pitoiset > --- > src/amd/vulkan

  1   2   3   4   5   6   7   8   9   10   >