Commit graph

973 commits

Author SHA1 Message Date
B3n30 dc6a365337 Merge pull request #2951 from huwpascoe/perf-4
Optimized Morton
2017-09-25 08:28:55 +02:00
Huw Pascoe 903906da3b Optimized Float<M,E> multiplication
Before:

ucomiss xmm1, xmm1
jp      .L9
pxor    xmm2, xmm2
mov     edx, 1
ucomiss xmm0, xmm2
setp    al
cmovne  eax, edx
test    al, al
jne     .L9
.L3:
movaps  xmm0, xmm2
ret
.L9:
ucomiss xmm0, xmm0
jp      .L10
pxor    xmm2, xmm2
mov     edx, 1
ucomiss xmm1, xmm2
setp    al
cmovne  eax, edx
test    al, al
je      .L3

After:

movaps  xmm2, xmm1
mulss   xmm2, xmm0
ucomiss xmm2, xmm2
jnp     .L3
ucomiss xmm1, xmm0
jnp     .L11
.L3:
movaps  xmm0, xmm2
ret
.L11:
pxor    xmm2, xmm2
jmp     .L3
2017-09-25 00:54:02 +01:00
Huw Pascoe 876aa82c29 Optimized Morton 2017-09-24 22:27:14 +01:00
James Rowe 93930a966f Merge pull request #2921 from jroweboy/batch-fix-2
GPU: Add draw for immediate and batch modes
2017-09-24 07:57:16 -06:00
James Rowe 19d41dcc6e Remove pipeline.gpu_mode and fix minor issues 2017-09-23 09:28:20 -06:00
Yuri Kunde Schlesner a7758b0b36 Merge pull request #2928 from huwpascoe/master
Fixed framebuffer warning
2017-09-22 04:06:38 +02:00
Huw Pascoe a234e4c200 Improved performance of FromAttributeBuffer
Ternary operator is optimized by the compiler
whereas std::min() is meant to return a value.

I've noticed a 5%-10% emulation speed increase.
2017-09-17 15:56:36 +01:00
Huw Pascoe 6a110ac5f5 Fixed framebuffer warning 2017-09-17 11:57:06 +01:00
Yuri Kunde Schlesner 699c920991 Merge pull request #2900 from wwylele/clip-2
PICA: implement custom clip plane
2017-09-16 10:23:00 +02:00
James Rowe ad0b57f407 GPU: Add draw for immediate and batch modes
PR #1461 introduced a regression where some games would change configuration
even while in the poorly named "drawing" mode, which broke the heuristic
citra was using to determine when to draw the batch. This change adds
back in a draw call for batching, and also adds in a draw call in
immediate mode each time it adds a triangle.
2017-09-11 09:21:43 -06:00
bunnei 11baa40d75 Merge pull request #2865 from wwylele/gs++
PICA: implemented geometry shader
2017-09-07 23:02:59 -04:00
bunnei ff4941fb3a Merge pull request #2914 from wwylele/fresnel-fix
pica/lighting: only apply Fresnel factor for the last light
2017-09-05 10:00:49 -04:00
wwylele 12fbc8c8df pica/lighting: only apply Fresnel factor for the last light 2017-09-03 08:22:03 +03:00
wwylele e2c41a5891 video_core: report telemetry for gas mode 2017-08-31 12:54:17 +03:00
bunnei f0e461bf6f Merge pull request #2891 from wwylele/sw-bump
SwRasterizer/Lighting: implement bump mapping
2017-08-30 21:07:30 -04:00
Weiyi Wang 647f017c6d Merge pull request #2892 from Subv/warnings2
Warnings: Fixed a few missing-return warnings in video_core.
2017-08-28 03:21:51 -05:00
Subv da88f3b8f0 Warnings: Fixed a few missing-return warnings in video_core. 2017-08-26 11:58:22 -05:00
wwylele 417cb45e3f SwRasterizer/Clipper: flip the sign convention to match PICA and OpenGL 2017-08-25 07:26:45 +03:00
wwylele addbcd5784 gl_rasterizer: implement custom clip plane 2017-08-25 07:26:45 +03:00
wwylele ea51a3af26 SwRasterizer: implement custom clip plane 2017-08-24 15:34:27 +03:00
wwylele 17c6104d2a gl_rasterizer/lighting: more accurate CP formula 2017-08-22 09:34:44 +03:00
wwylele b5aa570354 SwRasterizer/Lighting: implement LUT input CP 2017-08-22 09:34:44 +03:00
wwylele 3e478ca131 SwRasterizer/Lighting: implement bump mapping 2017-08-22 09:34:44 +03:00
wwylele 63b6e802cd swrasterizer: remove invalid TODO
This function is called in clipping, before the pespective divide, and is not used in later rasterization. Thus it doesn't need perspective correction.
2017-08-21 08:03:07 +03:00
wwylele 72b26ac32f swrasterizer/clipper: remove tested TODO
hwtested. Current implementation is the correct behavior
2017-08-21 08:03:07 +03:00
wwylele 5a4af616c6 gl_shader_gen: simplify and clarify the depth transformation between vertex shader and fragment shader 2017-08-21 08:03:07 +03:00
wwylele 1eca380886 gl_rasterizer: add clipping plane z<=0 defined in PICA 2017-08-21 08:03:07 +03:00
Yuri Kunde Schlesner 46d1ca768d Merge pull request #2872 from wwylele/sw-geo-factor
SwRasterizer/Lighting: implement geometric factor
2017-08-20 17:49:42 -07:00
James Rowe 8afa81ac1b Merge pull request #2871 from wwylele/sw-spotlight
SwRasterizer/Lighting: implement spot light
2017-08-19 20:10:24 -06:00
wwylele 0f35755572 pica/command_processor: build geometry pipeline and run geometry shader
The geometry pipeline manages data transfer between VS, GS and primitive assembler. It has known four modes:
 - no GS mode: sends VS output directly to the primitive assembler (what citra currently does)
 - GS mode 0: sends VS output to GS input registers, and sends GS output to primitive assembler
 - GS mode 1: sends VS output to GS uniform registers, and sends GS output to primitive assembler. It also takes an index from the index buffer at the beginning of each primitive for determine the primitive size.
 - GS mode 2: similar to mode 1, but doesn't take the index and uses a fixed primitive size.
hwtest shows that immediate mode also supports GS (at least for mode 0), so the geometry pipeline gets refactored into its own class for supporting both drawing mode.
In the immediate mode, some games don't set the pipeline registers to a valid value until the first attribute input, so a geometry pipeline reset flag is set in `pipeline.vs_default_attributes_setup.index` trigger, and the actual pipeline reconfigure is triggered in the first attribute input.
In the normal drawing mode with index buffer, the vertex cache is a little bit modified to support the geometry pipeline. Instead of OutputVertex, it now holds AttributeBuffer, which is the input to the geometry pipeline. The AttributeBuffer->OutputVertex conversion is done inside the pipeline vertex handler. The actual hardware vertex cache is believed to be implemented in a similar way (because this is the only way that makes sense).
Both geometry pipeline and GS unit rely on states preservation across drawing call, so they are put into the global state. In the future, the other three vertex shader units should be also placed in the global state, and a scheduler should be implemented on top of the four units. Note that the current gs_unit already allows running VS on it in the future.
2017-08-19 10:13:20 +03:00
wwylele 8285ca4ad8 pica/shader/jit: implement SETEMIT and EMIT 2017-08-19 10:13:20 +03:00
wwylele 36981a5aa6 pica/primitive_assembly: Handle winding for GS primitive
hwtest shows that, although GS always emit a group of three vertices as one primitive, it still respects to the topology type, as if the three vertices are input into the primitive assembler independently and sequentially. It is also shown that the winding flag in SETEMIT only takes effect for Shader topology type, which is believed to be the actual difference between List and Shader (hence removed the TODO). However, only Shader topology type is observed in official games when GS is in use, so the other mode seems to be just unintended usage.
2017-08-19 10:13:20 +03:00
wwylele bb63ae3052 correct constness 2017-08-19 10:13:20 +03:00
wwylele 28128348f2 pica/shader/interpreter: implement SETEMIT and EMIT 2017-08-19 10:13:20 +03:00
wwylele 46c6973d2b pica/shader: extend UnitState for GS
Among four shader units in pica, a special unit can be configured to run both VS and GS program. GSUnitState represents this unit, which extends UnitState (which represents the other three normal units) with extra state for primitive emitting. It uses lots of raw pointers to represent internal structure in order to keep it standard layout type for JIT to access.
This unit doesn't handle triangle winding (inverting) itself; instead, it calls a WindingSetter handler. This will be explained in the following commits
2017-08-19 10:13:20 +03:00
wwylele 686fb3e78c gl_shader_gen: don't call SampleTexture when bump map is not used 2017-08-11 18:35:00 +03:00
wwylele 945f9a1b04 SwRasterizer/Lighting: implement spot light 2017-08-11 01:19:10 +03:00
wwylele 14ee32c46a SwRasterizer/Lighting: implement geometric factor 2017-08-11 01:18:43 +03:00
wwylele 5d9d42f0d0 SwRasterizer/Lighting: use make_tuple instead of constructor
implicit tuple constructor is a c++17 thing, which is not supported by some not-so-old libraries. Play safe for now
2017-08-10 12:19:58 +03:00
wwylele db309b2423 pica/regs: layout geometry shader configuration regs
All the register meanings are derived from ctrulib (3dbrew is outdated for most of them)
2017-08-10 01:53:08 +03:00
Weiyi Wang 792dee47a7 Merge pull request #2822 from wwylele/sw_lighting-2
Implement fragment lighting in the sw renderer (take 2)
2017-08-09 18:54:29 +03:00
wwylele baa24f4ea9 pica: upload shared shader code to both unit 2017-08-07 10:30:05 +03:00
wwylele 2252a63f80 SwRasterizer/Lighting: shorten file name 2017-08-03 13:51:22 +03:00
wwylele eda28266fb SwRasterizer/Lighting: move to its own file 2017-08-02 22:20:40 +03:00
wwylele 48b4105871 SwRasterizer/Lighting: reduce confusion 2017-08-02 22:07:15 +03:00
wwylele c59ed47608 SwRasterizer/Lighting: move quaternion normalization to the caller 2017-08-02 22:05:53 +03:00
wwylele c89f804a01 pica/shader_interpreter: fix off-by-one in LOOP 2017-07-27 13:48:27 +03:00
Sebastian Valle c6a2e519ef Merge pull request #2816 from wwylele/proctex-lutlutlut
gl_rasterizer: use texture buffer for proctex LUT
2017-07-22 23:03:48 -05:00
Sebastian Valle e646bd902d Merge pull request #2834 from wwylele/depth-enable-fix
gl_rasterizer_cache: fix using_depth_fb
2017-07-22 23:02:59 -05:00
bunnei df8b9863f9 telemetry: Log performance, configuration, and system data. 2017-07-17 21:32:28 -04:00
wwylele 4feff63ffa SwRasterizer/Lighting: dist atten lut input need to be clamp 2017-07-11 22:19:00 +03:00
wwylele 56e5425e59 SwRasterizer/Lighting: unify float suffix 2017-07-11 22:15:35 +03:00
wwylele e415558a4f SwRasterizer/Lighting: get rid of nested return 2017-07-11 22:15:35 +03:00
wwylele c6d1472513 SwRasterizer/Lighting: refactor GetLutValue into a function.
merging similar pattern. Also makes the code more similar to the gl one
2017-07-11 22:15:35 +03:00
wwylele f13cf506e0 SwRasterizer: only interpolate quat and view when lighting is enabled 2017-07-11 21:35:57 +03:00
wwylele efc655aec0 SwRasterizer/Lighting: pass lighting state as parameter 2017-07-11 20:06:26 +03:00
Subv 9906feefbd SwRasterizer/Lighting: Move the clamp highlight calculation to the end of the per-light loop body. 2017-07-11 19:39:15 +03:00
Subv 7526af5e52 SwRasterizer/Lighting: Move the lighting enable check outside the ComputeFragmentsColors function. 2017-07-11 19:39:15 +03:00
Subv b8229a7684 SwRasterizer/Lighting: Do not use global registers state in ComputeFragmentsColors. 2017-07-11 19:39:15 +03:00
Subv 7bc467e872 SwRasterizer/Lighting: Do not use global state in LookupLightingLut. 2017-07-11 19:39:15 +03:00
Subv 37ac2b6657 SwRasterizer/Lighting: Fixed a bug where the distance attenuation bias was being set to the dist atten scale. 2017-07-11 19:39:15 +03:00
Subv 6250f52e93 SwRasterizer: Fixed a few conversion warnings and moved per-light values into the per-light loop. 2017-07-11 19:39:15 +03:00
Subv 2d69a9b8bf SwRasterizer: Run clang-format 2017-07-11 19:39:15 +03:00
Subv 73566ff7a9 SwRasterizer: Flip the vertex quaternions before clipping (if necessary). 2017-07-11 19:39:15 +03:00
Subv 2a75837bc3 SwRasterizer: Corrected the light LUT lookups. 2017-07-11 19:39:15 +03:00
Subv f2d4d5c219 SwRasterizer: Corrected the light LUT lookups. 2017-07-11 19:39:15 +03:00
Subv 80b6fc592e SwRasterizer: Fixed the lighting lut lookup function. 2017-07-11 19:39:15 +03:00
Subv 10b0bea060 SwRasterizer: Calculate fresnel for fragment lighting. 2017-07-11 19:39:15 +03:00
Subv 46b8c8e1da SwRasterizer: Calculate specular_1 for fragment lighting. 2017-07-11 19:39:15 +03:00
Subv be25e78b07 SwRasterizer: Calculate specular_0 for fragment lighting. 2017-07-11 19:39:15 +03:00
Subv b2f472a2b1 SwRasterizer: Implement primary fragment color. 2017-07-11 19:39:15 +03:00
wwylele 8482933db8 gl_rasterizer: use texture buffer for proctex LUT 2017-07-01 11:02:48 +03:00
wwylele 8978ecb09c gl_rasterizer: use texture buffer for fog LUT 2017-06-22 20:41:00 +03:00
wwylele f1e377f57e gl_rasterizer: create the texture before applying the state
this is a rebasing error from #2792. It doesn't affect much though, because the later more Apply() call fixes/hides it
2017-06-22 17:47:46 +03:00
wwylele 457659fe01 gl_state: reset 1d textures 2017-06-21 23:13:06 +03:00
wwylele 42f7ca7412 gl_rasterizer: fix glGetUniformLocation type 2017-06-21 23:13:06 +03:00
wwylele be9e952bdc gl_rasterizer: manage texture ids in one place 2017-06-21 23:13:06 +03:00
wwylele ab60414122 gl_rasterizer/lighting: fix LUT interpolation 2017-06-21 23:13:06 +03:00
Yuri Kunde Schlesner d0888f8548 Merge pull request #2776 from wwylele/geo-factor
Fragment lighting: implement geometric factor
2017-06-18 14:18:48 -07:00
wwylele 5a454173a8 gl_rasterizer/lighting: use the formula from the paper for germetic factor 2017-06-18 10:29:02 +03:00
Yuri Kunde Schlesner f6715f98f5 Stop using reserved operator names (and/or/xor) with Xbyak
Also has the Dynarmic upgrade with the same change
2017-06-17 12:20:22 -07:00
wwylele 7052d43a67 gl_rasterizer/lighting: implement geometric factor 2017-06-15 14:59:01 +03:00
Yuri Kunde Schlesner da1bec121a Merge pull request #2762 from wwylele/light-cp-tangent
Fragment lighting: implement lut input 5 (CP) and tangent mapping
2017-06-14 20:08:26 -07:00
Yuri Kunde Schlesner 5fe5ccac42 Merge pull request #2743 from wwylele/wrap-fix
pica/rasterizer: implement/stub texture wrap mode 4-7
2017-06-13 21:28:12 -07:00
Yuri Kunde Schlesner 791cd14c8d Merge pull request #2767 from yuriks/quaternion-flip-comment
OpenGL: Update comment on AreQuaternionsOpposite with new information
2017-06-12 16:31:55 -07:00
wwylele 972548e3ee gl_rasterizer/lighting: Implement tangent mapping 2017-06-11 21:30:53 +03:00
wwylele 40b7d0bf3f gl_rasterizer/lighting: implement lut input 5 (CP) 2017-06-11 21:30:53 +03:00
Sebastian Valle 39c7c1f580 Merge pull request #2727 from wwylele/spot-light
Fragment lighting: implement spot light
2017-06-11 18:23:47 +00:00
wwylele b3b9468573 gl_rasterizer_cache: depth write is disabled if allow_depth_stencil_write is false 2017-06-10 15:10:34 +03:00
Yuri Kunde Schlesner ba01a8302a OpenGL: Update comment on AreQuaternionsOpposite with new information
While debugging the software renderer implementation, it was noticed
that this is actually exactly what the hardware does, upgrading the
status of this "hack" to being a proper implementation. And there was
much rejoicing.
2017-06-10 01:55:17 -07:00
wwylele 28d1e73d2f pica/rasterizer: implement/stub texture wrap mode 4-7 2017-06-04 09:47:25 +03:00
bunnei 54ea95cca7 Merge pull request #2721 from wwylele/texture-cube
swrasterizer: implemented TextureCube
2017-05-30 10:21:05 -04:00
wwylele 10906dceec gl_rasterizer: implement spot light 2017-05-30 10:54:58 +03:00
wwylele 686cbf3ac6 gl_rasterizer: sync spot light status 2017-05-30 10:54:58 +03:00
wwylele b5addf8fb8 pica: prepare registers for spotlight 2017-05-30 10:54:58 +03:00
Yuri Kunde Schlesner a4f88c7d7c Merge pull request #2734 from yuriks/cmake-imported-libs
CMake: Use CMake target properties for all libraries
2017-05-29 15:12:21 -07:00
wwylele 0b9bb082c3 swrasterizer: implement TextureCube 2017-05-29 22:28:48 +03:00
wwylele 077cc683e5 pica: add registers for texture cube 2017-05-29 22:03:08 +03:00
Yuri Kunde Schlesner 3df85a103a Merge pull request #2729 from yuriks/quaternion-fix
OpenGL: Improve accuracy of quaternion interpolation
2017-05-28 01:24:06 -07:00
Yuri Kunde Schlesner d736cca848 CMake: Create INTERFACE targets for microprofile and nihstro 2017-05-27 22:34:52 -07:00