RPCSX/rpcsx
1
0
Fork 0
mirror of https://github.com/RPCSX/rpcsx.git synced 2026-01-08 01:30:13 +01:00
Code Issues Actions Packages Projects Releases Wiki Activity
16979 commits 5 branches 1 tag 150 MiB
Commit graph

133 commits

Author SHA1 Message Date
kd-11 8ca53f9c84 rsx: Remember to min-max the anchor indices of a polygon or triangle fan 2019-11-24 19:01:57 +03:00
Emmanuel Gil Peyrot f76720ceb0 Remove extraneous ::narrow<int>() calls 
GSL’s gsl::span didn’t use the correct type for its index_type, which is
why they were needed.
 2019-11-09 19:30:06 +01:00
Emmanuel Gil Peyrot ef368c5171 rsx: Replace gsl::byte with C++17’s std::byte 2019-11-09 19:30:05 +01:00
Nekotekina e3e7051ed3 Minor optimization in BufferUtils.cpp 
Don't use PSHUFB for horizontal operations.
Utilize PHMINPOSUW to compute max as well:
 + sse41_hmin_epu16
 + sse41_hmax_epu16
 2019-10-30 18:52:34 +03:00
Nekotekina b1968769b7 Minor cleanup in BufferUtils.cpp 
Replace inline asm with intrinsic using target attribute trick.
 2019-10-30 17:53:51 +03:00
linkmauve cfd5cf6bdb Optimise primitive_restart::upload_untouched() (#6881) 
* rsx: Optimise primitive_restart::upload_untouched() with SSE4.1

This optimisation is only applied when skip_restart is false.

I’ve only tested the u16 codepath, as it is the one used in NieR.

In some very unscientific profiling, this function used to take 2.76% of
the total frame time at the save point of the port town, it now takes
about 0.40%.

* rsx: Mark all SSE4.1 functions with attributes on gcc and clang

This assures the compiler we will take care of only calling these
functions after having checked that the CPU does support these
instructions.

* rsx: Add an AVX2 implementation of primitive restart ibo upload

* rsx: Remove redefinition of SSE4.1 instructions

Now that clang is aware that our functions are compiled with SSE4.1, it
lets us generate this code using its intrinsics.

* rsx: Optimise vector to scalar conversion

This is done using minpos and srli intrinsics and generate less code
than before.

Thanks Nekotekina for the suggestion!
 2019-10-30 16:42:44 +03:00
kd-11 c59cb1bdd3 rsx: Allow only sse4.1 capable CPUs to take the accelerated index path 
- Older sets lack the required min/max functionality
 2019-09-13 12:28:52 +03:00
kd-11 60845daf45 rsx: Improve use of CPU vector extensions 
- Allow use of intrinsics when SSSE3 and SSSE4.1 are not available in the build target environment
- Properly separate SSE4.1 code from SSSE3 code for some older proceessors without SSE4.1
 2019-09-12 14:08:21 +03:00
kd-11 652f18ebaa rsx/buffers: Warnings cleanup 2019-09-01 18:59:50 +03:00
JohnHolmesII ebb1ae6408 Properly ignore SIMD macros to avoid warning 2019-06-28 01:40:52 +03:00
kd-11 358169507c rsx: Use SSE to accelerate index buffer uploads 2019-06-25 20:50:54 +03:00
Lassi Hämäläinen c963c51a60 Remove unnecessary header includes 
- Manually removed lot of unneeded #includes to clean code and reduce
  compilation time
- Reordered some of the #includes to be in more logical order
 2019-06-25 17:11:10 +03:00
scribam db926ee671 rsx: Apply Clang-Tidy fix "performance-unnecessary-value-param" 2019-06-12 15:11:52 +03:00
Nekotekina dfd50d0185 Implement std::bit_cast<> 
Partial implementation of std::bit_cast from C++20.
Also fix most strict-aliasing rule break warnings (gcc).
 2019-06-02 23:22:16 +03:00
kd-11 417a2e6731 rsx: Refactor index buffers 
- Index offset is ignored anyway and only used to calculate vertex attribute divisor index
- Specialized optimization for untouched xfer without primitive restart
 2019-01-25 14:34:22 +03:00
kd-11 504ab5a6d4 rsx: Minor cleanup to silence stupid compiler warnings 2018-12-03 20:01:23 +03:00
kd-11 677b16f5c6 rsx: Fixups 
- Also fix visual corruption when using disjoint indexed draws

- Refactor draw call emit again (vk)

- Improve execution barrier resolve
  - Allow vertex/index rebase inside begin/end pair
  - Add ALPHA_TEST to list of excluded methods [TODO: defer raster state]

- gl bringup

- Simplify
  - using the simple_array gets back a few more fps :)
 2018-11-30 23:51:25 +03:00
kd-11 e01d2f08c9 rsx: Refactor FIFO 
- Removes fifo structures from common RSXThread
- Sets up a dedicated FIFO controller
- Allows for configurable queue optimizations
 2018-11-30 23:51:25 +03:00
eladash 83b6c98563 rsx: Fix u16 index arrays overflow 
Force u32 index array destinations to avoid overflows when adding vertex base index.
 2018-10-08 16:39:47 +03:00
eladash e361e0daa6 rsx: Fix restart index check for u16 index arrays 
Dont ignore upper bits of the restart index with u16 types
 2018-10-08 16:39:47 +03:00
kd-11 6399833182 rsx: Fix endianness order when immediate mode register is updated, but used as register lookup 
- Simplify the code by unifying all the register-backed memory
 2018-09-03 18:24:20 +03:00
Nekotekina 1c6c24f8ac Update GSL and yaml-cpp submodules 2018-08-25 01:15:47 +03:00
scribam 04ad49de4d typos 2018-05-14 21:14:39 +04:00
Jake 6d6d6fa827 dx12/vk/gl: implement use of vertex_data_base_index when calculating index 2018-03-30 13:30:04 +03:00
Nekotekina 61de20a633 RSX: remove SSSE3 dependency 2017-12-20 00:04:08 +03:00
Zion Nimchuk 3a9ae2df9e silence warnings in RSX stuff 2017-11-30 18:07:19 +03:00
kd-11 ddcacb8258 general fixes; Force u32 return type for index_count and add RX Vega to primitive restart blacklist 2017-10-19 12:22:52 +03:00
kd-11 89dcafbe41 rsx: Reimplement index buffer generation 
- Emulate primitive restart in software whenever we get the chance
- Ensure PRIMITIVE_RESTART is never active when LIST topologies are active
- Reimplement TRIANGLE_FAN, POLYGON and QUAD expansion
 2017-10-19 12:22:52 +03:00
kd-11 2e9405db4c rsx: Remove index expansion for quad strips 2017-08-26 21:53:54 +03:00
kd-11 fe5828cb47 rsx: Implement QUAD_STRIP 
- QUAD_STRIP evaluates to TRIANGLE_STRIP in memory. The memory layout is identical.
- The only difference between the two modes would be the primitive_ID but that doesnt matter on RSX
- Its worth noting that results will be different between the two modes if input vertices are non-coplanar for every set of N verts
 2017-08-26 21:53:54 +03:00
kd-11 a7c28f5827 rsx: Fix remainder/iteration computations in BufferUtils 2017-07-24 16:52:42 +03:00
kd-11 47e5074dc5 rsx: Emulated index buffers are based on vertex 0 with no disjoint ranges 
- Drop the 'first' argument as it is unused for now
 2017-06-29 13:13:19 +03:00
kd-11 b2e906f4cc rsx: Code cleanup. Fixes several dozen warnings 
- Wrap unused parameters as comments to prevent C1400
- Fix sized variable conversions with explicit casts
 2017-06-22 23:36:15 +03:00
kd-11 11317acdbe rsx: Handle non-zero base vertex better 
- Vertex buffer contents treat the base vertex as vertex 0 so we do the same for indices

rsx: Fix vertex base indexing

rsx: Properly fix non-zero offset indexed rendering
 2017-06-22 23:36:15 +03:00
kd-11 786bcb0d1b rsx: bugfix - avoid a divide by zero 2017-05-22 14:28:33 +03:00
kd-11 d4ddc40988 rsx: Add support for repeated data streams (broken attrib divisor?) 2017-05-22 14:28:33 +03:00
kd-11 e8b4d332eb rsx: Use faster upload path when conditions allow 
Fix aligned memory access (SSE)

rsx: BufferUtils; always use optimized paths
 2017-05-22 14:28:33 +03:00
kd-11 2d99f3556e rsx: Fix line_loop -> line_strip indexing 2017-04-03 13:50:58 +03:00
Zangetsu38 73906f9f29 d3d12: add x1r5g5b5_z1r5g5b5 and cleanup in D3D12Formats. 
Add info in BufferUtils for log.
 2017-02-10 22:04:45 +03:00
Vincent Lejeune 42b518cf7e rsx: use range for vertex buffer attribute. 2016-08-24 21:58:59 +02:00
Nekotekina 84d0d396ed EXPECTS usage removed 2016-08-15 16:29:38 +03:00
Nekotekina a7e808b35b EXCEPTION macro removed 
fmt::throw_exception<> implemented
::narrow improved
Minor fixes
 2016-08-08 19:19:32 +03:00
Vincent Lejeune fb47945930 rsx: Returns u32 instead of size_t for get_index_count/type_size 2016-08-06 00:25:23 +02:00
Vincent Lejeune 7a6f5b6ee5 rsx: Move index pointer generation in rsx::thread. 2016-08-05 17:54:44 +02:00
raven02 e1ff3f4674 rsx: use fragment_textures_count (#1948) 
* rsx:  use fragment_textures_count

* Typo: unknow -> unknown
 2016-07-19 22:50:40 +08:00
Vincent Lejeune 772706ca4c Factorize rsx state 2016-07-07 21:38:57 +02:00
DHrpcs3 3b5cd4845e OpenGL renderer: use correct MVP matrix. Cleanup 
Simplified gl::ring_buffer helper
 2016-06-21 19:56:05 +03:00
Nekotekina 266db1336d The rest 2016-05-23 16:22:25 +03:00
Nekotekina b85a68e8a1 Partial commit: RSX 2016-04-15 19:22:36 +03:00
Vincent Lejeune 3a3d264cb5 rsx/common/d3d12/gl/vulkan: Set dst stride in write_vertex_array_data_to_buffer. 2016-04-07 22:17:28 +02:00
Vincent Lejeune 2ae5a7ff39 rsx/common/d3d12/gl/vulkan: Use single overload for write_index_array_data_to_buffer. 2016-04-07 22:17:28 +02:00
Vincent Lejeune cbe119b457 rsx/common: Remove MIN2/MAX2 macro. 2016-04-07 22:17:28 +02:00
kd-11 0327e76320 Fix quad strip triangle winding 2016-03-24 10:52:35 +03:00
Vincent Lejeune 9cdb74efc7 rsx/common: Add supports for quads strip 
Used in Hitman 2
 2016-02-27 19:38:16 +01:00
Vincent Lejeune a6ba47265f rsx/common/gl: s32k is actually signed short unormalized. 
gl fix
 2016-02-27 00:21:12 +01:00
Vincent Lejeune 5ef7f8bf3e rsx/common: Fix handling of UB256 2016-02-27 00:21:06 +01:00
Vincent Lejeune 5a14644cd4 rsx/common/d3d12/gl: Use span in vertex upload function. 2016-02-22 20:22:47 +01:00
kd-11 7b889a10cc Add vertex texture buffers for VS input 
Support vertex instancing in vertex shader using VertexID

Relax OpenGL requirements by removing 4.5 features

Use EXT version of TexBufferRange; Implement buffer copy using TexBuffer

Apply travis workaround by danilaml

Fix vertex upload in in case of inlined array
 2016-02-03 13:38:23 +03:00
Vincent Lejeune acd384ae2d rsx/common: Base offset is actually correctly supported. 
Outrun uses it and cars are correctly displayed.
 2016-01-27 22:05:43 +01:00
DHrpcs3 19ce0cdc09 rsx methods constants moved to rsx namespace 
minor fix
 2016-01-20 20:12:48 +03:00
DHrpcs3 7972cb5bdc Code style fixes #1 2016-01-20 16:23:25 +03:00
Vincent Lejeune 689dee9944 rsx/common/d3d12: Consider separate index range as a whole. 
Fix Wolf of the Battlefield 3
 2016-01-13 00:28:48 +01:00
Vincent Lejeune d153575e59 rsx/common/d3d12/gl: Support for CMP/non pow of 2 size vertex formats. 
Also use class enum for base_vertex_type everywhere.
Fix Bomberman Ultra color and Cubixx HD geometry.
 2016-01-09 23:18:03 +01:00
DHrpcs3 48919330d7 rsx methods moved from rsx thread 2016-01-06 13:30:24 +02:00
Vincent Lejeune 5f12a4f7b5 rsx/common/d3d12/gl: Use separate vertex array/vertex register states. 2015-12-30 17:04:34 +01:00
Nekotekina 3ed603074c Changes done by [DH] rewritten 
Added rsx_program_decompiler submodule
Added fs::dir iterator
Added fmt::match
 2015-12-22 23:11:20 +03:00
Nekotekina aa811b6eef Cleanup (noexcept, unreachable) 
%x formatting fixes
 2015-12-20 15:41:07 +03:00
Vincent Lejeune 1cda2977bb common/d3d12: emulate polygon mode 2015-12-16 20:36:36 +01:00
Vincent Lejeune ae5d95d462 rsx/common: Take primitive restart index in account and turns it into -1. 2015-11-12 18:29:01 +01:00
Vincent Lejeune 2ad7051746 rsx/d3d12: Move vertex constants filling code to RSXThread 2015-11-06 20:08:41 +01:00
Vincent Lejeune 02ce78482c rsx/d3d12: Move scale offset buffer setting to RSXThread 2015-11-06 20:08:17 +01:00
Vincent Lejeune 1ec18bdf64 RSX/common: Clean BufferUtils code 
* Add noexcept
* Use a_b_c code style
* Use anonymous namespace
 2015-10-29 18:48:50 +01:00
Vincent Lejeune 9f49232cac d3d12: Avoid copying index data and use correct index range. 
This fixes Braid.
 2015-10-27 01:24:04 +01:00
Vincent Lejeune a2997a1109 d3d12: Avoid an extra vertex copy 2015-10-15 17:13:43 +02:00
Nekotekina a974ee009e vm::var improved, cleanup 
Mostly vm::var initialization introduced.
Added vm::make_var function.
 2015-10-14 18:17:37 +03:00
Vincent Lejeune a63fdf6c45 Use files from master 
- Drop smart vertex storage and use OpenGL's one instead.
 2015-10-13 14:27:17 +02:00
Vincent Lejeune 60bccf0f10 Remove RSXVertexArray 2015-10-13 00:04:12 +02:00
Vincent Lejeune d27f6c8fa7 Use rsx::limits values 2015-10-13 00:04:05 +02:00
Vincent Lejeune 6f71d04aa4 move linear to swizzle and get_size_type 
symbol undef though
 2015-10-13 00:04:04 +02:00
Vincent Lejeune f483c3b9ca Revert "Merge pull request #1245 from DHrpcs3/master" 
This reverts commit 5feba39ff7, reversing
changes made to ebf28f8da0.
 2015-10-09 20:04:20 +02:00
DH 6cb036d35f Fix for gcc/clang build 2015-10-08 00:05:04 +03:00
Vincent Lejeune d511153836 Common: Fix element count computation if addr is null (RSXVertexData) 2015-10-05 01:57:57 +02:00
Vincent Lejeune c7b7d1f71f Common: Move generic vertex buffer code from d3d12 backend 2015-10-03 18:25:19 +02:00