Commit graph

485 commits

Author SHA1 Message Date
kd-11 31b07f2c5c rsx: Tweaks
- Optimize get_surface_subresource
- Add check_program_status time to draw call setup statistics. It can slow down games significantly
2017-11-02 14:35:19 +03:00
kd-11 361e80f7dc rsx: Tag cache blocks returned on access violation to validate data passed
to flush_all is up to date. Should prevent recursive exceptions

Partially revert Jarves' fix to invalidate cache on tile unbind. This will
need alot more work. Fixes hangs
2017-10-29 15:25:17 +03:00
kd-11 ce2b790d8f rsx/qt: GsFrame improvements
- Identify minimize/restore events as separate from regular resize and do not react to them
- Enable message queue consumption after loading the shaders cache. Also hides the frame in this step
-- This fixes the 'start fullscreen' bug when running vulkan
2017-10-29 15:25:17 +03:00
kd-11 055f0e2e4a rsx: Export more information about affected cache sections when handling violations
- This allows for better handling of deferred flushes.
-- There's still no guarantee that cache contents will have changed between the set acquisition and following flush operation
-- Hopefully this is rare enough that it doesnt cause serious issues for now
2017-10-28 13:26:16 +03:00
kd-11 bf234dc668 rsx: Implement memory tags for strict mode to validate render target memory 2017-10-28 13:26:16 +03:00
Jake e0d1ac676e rsx: invalidate surface store address when tile is unbound 2017-10-28 12:46:20 +03:00
kd-11 d53673c6f7 rsx: Disable draw call batching when strict mode is enabled
- Minor bugs have been observed with it always enabled
2017-10-26 00:35:45 +03:00
kd-11 e9f293f522 rsx: Improve separate treatment of write exceptions vs read exceptions
- Optimizes search functionality and avoids thrashing valid sections
2017-10-24 22:59:09 +03:00
kd-11 7a5c88a8d1 rsx: Only transfer framebuffer contents if memory contents are compatible 2017-10-24 22:59:09 +03:00
kd-11 0de0dded53 rsx: Texture fixes continued
- Fix buffer invalidate behaviour (wcb)
- Disable auto rebuild with only framebuffer storage getting rebuilt
- Fix vulkan subresource sampling
2017-10-24 22:59:09 +03:00
kd-11 5e58cf6079 rsx: Restructuring [WIP]
- Refactor invalidate memory functions into one function
- Add cached object rebuilding functionality to avoid throwing away useful memory on an invalidate
- Added debug monitoring of texture unit VRAM usage
2017-10-24 22:59:09 +03:00
kd-11 a8e1754fa4 rsx: Allow clear_surface command for depth-only passes. Removes invalidate cache hack 2017-10-19 12:22:52 +03:00
kd-11 5db45c3699 rsx: More fixes
- Workaround for AMD glMultiDrawArrays bug
- Disable disjoint command submission when multidraw support is disabled
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 58860614e3 rsx/qt: Implement native window hooks for win32 windows to allow communication between WndProc thread and rsx::thread
- This communication is important in communicating window events. Helps properly synchronize swapchain management on vulkan and stops nvidia crashing
- Do not block the message queue lest the driver detect window as not responding
2017-10-12 13:51:29 +03:00
kd-11 5a03db7d2e rsx: Improve memory protection behavior when strict mode is off 2017-10-10 12:10:41 +03:00
kd-11 12ab03b0b5 rsx/gl: Implement resolution scaling
rsx: Revise wpos calculation to take resolution scale into account
2017-10-09 20:25:41 +03:00
kd-11 393e3b702f rsx: Clean up debug overlays. Add unreleased textures metric to track texture memory 2017-09-23 16:46:41 +03:00
kd-11 3499d089e7 rsx: Texture cache fixes and improvements
rsx: Conditional lock hack removed
vulkan - Fixes
- Remove unused texture class
- Fix native pitch calculation (WCB)
rsx: Catch hanging begin/end pairs when flushing deferred draw calls
vulkan: Register DXT compressed formats
vulkan: Register depth formats
gl: Workaround for 'texture stitching' when gathering flip surface
- TODO: Add a proper flip hack option
rsx: Fix texture memory size calculation
- DXT textures dont have real pitch. Since pitch is used to calculate memory size, make sure it always evaluates to rsx_size
rsx: Fix cpu copy detection
rsx: Validate blit dst surface and dont make assumptions about region blit order
- Also relax restrictions on memory owned by the blit engine if strict rendering is not enabled
rsx: Fix depth texture detection
rsx: Do not manually offset into dst. The overlapped range check does so automatically
rsx: Minor optimizations
rsx: Minor fixes
- Fix to detect incompatible formats when using GPU texture scaling and show message
- Better 'is_depth_texture' algorithm to eliminate false positives
2017-09-21 16:17:06 +03:00
kd-11 6b96a2022a rsx: Add support for non-projective shadow sampling
- Fixes missing shadows in persona 5

vk: Enable polygon depth bias a.k.a polygonOffset
- Fixes shadow acne in persona 5
2017-09-21 16:17:06 +03:00
kd-11 3836b40bf7 rsx: Fixups 2017-09-21 16:17:06 +03:00
kd-11 45d0e821dc gl: Minor optimizations
rsx: Texture cache - improvements to locking
rsx: Minor optimizations to get_current_vertex_program and begin-end batch flushes
rsx: Optimize texture cache storage
- Manages storage in blocks of 16MB
rsx/vk/gl: Fix swizzled texture input
gl: Hotfix for compressed texture formats
2017-09-21 16:17:06 +03:00
kd-11 e37a2a8f7d rsx: Texture cache fixes and improvments
gl/vk/rsx: Refactoring; unify texture cache code
gl: Fixups
- Removes rsx::gl::texture class and leave gl::texture intact
- Simplify texture create and upload mechanisms
- Re-enable texture uploads with the new texture cache mechanism
rsx: texture cache - check if bit region fits into dst texture before attempting to copy
gl/vk: Cleanup
- Set initial texture layout to DST_OPTIMAL since it has no data in it anyway at the start
- Move structs outside of classes to avoid clutter
2017-09-21 16:17:06 +03:00
kd-11 07c83f6e44 gl: cleanup; fix program linkage on mesa using GL_ARB_explicit_uniform_location, also make use of ARB_multidraw 2017-09-21 16:17:06 +03:00
kd-11 061824a7ec rsx: Add support for batched multidraw
gl: Fix multidraw [WIP]
rsx: Ignore vertex base when data source is generated using arithmetic
vk: Check pending flag before doing fence poke
vk/gl: Fix for inlined array and immediate draws
rsx: Collapse joined draws when batching
2017-09-21 16:17:06 +03:00
kd-11 2d0f1f27a8 esx: Fixes to the texture cache
rsx: Blit engine improvements
- Always handle blits to and from framebuffers through the GPU
- Handle depth surfaces properly when using GL
- Check for format mismatches when blitting to the surface store [WIP]
2017-09-21 16:17:06 +03:00
kd-11 2033f3f7dc rsx/vk/gl: Refactoring and reimplementation of blit engine
Fix rsx offscreen-render-to-display-buffer-blit surface reads
- Also, properly scale display output height if reading from compressed tile

gl: Fix broken dst height computation
- The extra padding is only there to force power-of-2 sizes and isnt used

gl: Ignore compression scaling if output is rendered to in a renderpass

rsx/gl/vk: Cleanup for GPU texture scaling. Initial impl [WIP]
- TODO: Refactor more shared code into RSX/common
2017-09-21 16:17:06 +03:00
kd-11 e56d8db6a3 gl: Do not modify bound depth/stencil buffers bound to zcull memory
- Fixes some rare cases where zcull always fails
2017-08-26 21:53:54 +03:00
kd-11 e76b6e3f27 gl: Workaround for restrictive z range 2017-08-26 21:53:54 +03:00
kd-11 f71f67c4ff rsx: Make fragment state dynamic to reduce shader permutations 2017-08-26 21:53:54 +03:00
kd-11 8358bda133 gl/rsx: Fixes to zcull pixel counting 2017-08-26 21:53:54 +03:00
kd-11 b86745ee4e gl: Fix vsync; always use adaptive vsync instead of double-buffered 2017-08-16 23:58:30 +03:00
kd-11 3c9bab8563 ui: Move disable vertex cache to debug. Also add an option to disable zcull emulation in the debugger 2017-08-16 23:58:30 +03:00
kd-11 00c6a589a5 rsx/util: Add simple consistent hash function
rsx/vk/shaders_cache: Move vp control mask to dynamic state

rsx/vk/gl: adds a shader cache for GL. Also Separates pipeline storage for each backend

rsx: Add more texture state variables to the cache
2017-08-16 23:58:30 +03:00
kd-11 1da732bbf5 rsx/gl/vk: Invalidate texture regions when memory is unmapped
- Free GPU resources immediately if mappings change to avoid leaking VRAM
2017-08-16 23:58:30 +03:00
kd-11 d54c2dd39a gl: Move vertex processing to the GPU
- Significant gains from greatly reduced CPU work
- Also reorders command submission in end() to improve throughput

- Refactors most of the vertex buffer handling
- All vertex processing is moved GPU side
2017-08-16 23:58:30 +03:00
kd-11 1e46cd255b gl: Fix inconsistent state tracking 2017-08-10 00:16:20 +03:00
kd-11 e5e6af41a6 gl: Fix occlusion statistics clearing 2017-08-10 00:16:20 +03:00
kd-11 4c019c55d2 rsx/gl: Fix zcull queries and log conditional render modes
- Fixes a situation where a query readback is requested while zcull render is still active
2017-08-10 00:16:20 +03:00
kd-11 fcb7072fee Implement hardware zcull emulation
rsx/gl: Support s1 immediate values; ogl minor refactoring
2017-08-06 14:29:42 +03:00
Jake d9a693019b rsx/gcm: Implement rsx dma. Refactor gcm/rsx to not be as codependent 2017-08-02 01:33:12 +03:00
kd-11 f613901eac gl/vk: Do not trust gcm buffer size to match internal rendering resolution
- Should fix games with strange scaling artifacts due to upscaling from lower resolutions
2017-07-27 14:33:30 +03:00
kd-11 4cd5624fa7 rsx/vk/gl: Refactoring - Also adds a vertex cache to openGL as well 2017-07-27 14:33:30 +03:00
kd-11 cb194e06ce gl: Fix framebuffer validity checks 2017-07-24 16:52:42 +03:00
kd-11 2526626646 rsx: Surface cache bug fixes
- Properly handle data 'transfer' when recycling frame buffer images
- Clear 'recycled' surfaces before use
2017-07-19 23:28:33 +03:00
kd-11 05ffb50037 vk/rsx: Bug fixes and improvements
- Improvements to framebuffer usage; Avoid creating new resources every frame
- Handle null fragment program properly
- Collect vertex upload statistics

- vk: Pre-initialize 'unused' varying registers in the vertex shader in case it gets matched with a fs that consumes it
 -- Fixes a crash about fog_c not being declared

gl/dx12/vk: Handle null fragment program

- cleanup - use yield semantic instead of sleep(0) as yield is more cross-platform
 -- sleep(0) is a windows specific scheduler hint
2017-07-19 23:28:33 +03:00
raven02 2e47c42367 Add Stretch to Display Area option (#2961) 2017-07-08 16:06:11 +03:00
kd-11 9e7a42d057 rsx: Minor bug fixes
- vk: Do not select first available format when choosing a swapchain format
- gl/vk: Ignore rendering zero sized framebuffers/scissors
- fp: Re-enable range clamp on fp16 registers; fix fx12 clamping [-2, 2]
2017-07-08 14:52:16 +03:00
kd-11 3d935b64f2 rsx/gl/vk: Enable contents transfer when a new framebuffer is created and not cleared 2017-07-08 14:52:16 +03:00
kd-11 e9b8f94fb1 rsx/gl/vk: Enable frame skipping 2017-07-08 14:52:16 +03:00
kd-11 b95ffaf4dd rsx: Implement skip draw. Also, start working on MT vertex upload 2017-07-08 14:52:16 +03:00
kd-11 72edf26fb8 rsx/gui: Add a strict mode toggle. Also reorganizes the graphics tab a bit
code cleanup
2017-06-22 23:36:15 +03:00
kd-11 590bb7cbe4 rsx: Bug fixes
rsx: Give more info when ring buffer allocations fail
2017-06-22 23:36:15 +03:00
kd-11 798f90dac2 rsx/ui: Make a few settings configurable via the GUI 2017-06-22 23:36:15 +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 86adc42739 rsx/gl/vk: Invert default stencil values
- Fixes stencil volume shadows in a few games. This is part of the surface invalidation workarounds
2017-06-22 23:36:15 +03:00
kd-11 98cf72e0fb rsx: Fix clip space computations 2017-06-22 23:36:15 +03:00
kd-11 db1a90d828 rsx: Discard surface store contents once per frame (temp workaround)
Need to find the proper command issued to discard all surfaces
2017-06-22 23:36:15 +03:00
kd-11 69d3d47901 gl: Fix clip-space -> depth conversion. Fixes remaining depth read issues
- Also set some default values for samplers in a cleaner way using their 'natural' float values
2017-06-22 23:36:15 +03:00
kd-11 5f66d0b996 rsx/wip: Fix depth surface reuse and clearing (fixes shadows) 2017-06-22 23:36:15 +03:00
kd-11 860b76452f vulkan bringup on linux
cleanup: drop unused stuff
2017-06-08 19:08:44 +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
Nekotekina f010b5b235 Configuration simplified 2017-05-20 16:01:48 +03:00
kd-11 c5975d5f66 rsx: Vertex program output fixes 2017-05-12 20:10:03 +03:00
Jake 60ce85f840 [Render] Userclip for d12/vk/ogl (#2719) 2017-04-25 18:32:39 +08:00
scribam 30eb680c9d Check variable before to execute remove, in GLGSRender::on_exit (#2682) 2017-04-18 00:30:34 +03:00
kd-11 ac3b22902a gl: Autodetect supported OpenGL extensions
cleanup
2017-04-10 17:24:30 +03:00
raven02 9b48074389 GL: fix wireframe in Catherine (#2660) 2017-04-10 09:28:46 +08:00
kd-11 6d6d0e4e36 gl: Use the GPU to scale textures; use ARB_sampler_object
Improve scaling and separate sampler state from texture state

gl: Unify all texture cache objects under one structure separate by use case
gl: Texture cache fixes

- Acquire lock when finding matching textures
- Account for swizzled surfaces when deciding whether to cpu memcpy
- Handle swizzled images on the GPU
2017-04-08 23:12:09 +03:00
kd-11 3ec1fe9ee7 rsx: Fix depth clears on z16 surfaces without stencil
remove some debug code
2017-03-26 16:22:53 +03:00
kd-11 79d114cc06 rsx: Support immediate mode rendering 2017-03-26 16:22:53 +03:00
kd-11 8fc281e034 gl: Separate scale offset, fragment constants and transform constants uploads 2017-03-13 23:40:34 +03:00
kd-11 2093134420 gl: Do not attempt to re-flush an address if another thread flushes it first 2017-03-13 23:40:34 +03:00
kd-11 1fd33f611c gl: Fix texture cache bugs
Fix endianness bug
Fix r/w when real pitch is <= 64
2017-03-10 19:11:53 +03:00
kd-11 7062efeb3e gl: Fix a fence deadlock in fence::wait_for_signal
gl: Fix potential race condition when executing local work queue
2017-03-01 00:38:22 +03:00
kd-11 ef8e92cbe3 gl: Change clear timing to immediate
gl: Handle cell write to locked RTT region

addendum: Fix buffer locking (personal 4)
2017-03-01 00:16:55 +03:00
kd-11 7c73c3b75c rsx/gl: Minor refactoring; prepare vulkan backend 2017-03-01 00:16:55 +03:00
kd-11 96561c16ad gl: Rewrite texture cache
gl: scale downloaded RTT images
2017-03-01 00:16:55 +03:00
kd-11 9263999ad1 [rsx/vp] Improve BRB opcode implementation
fix merge issues
2017-02-26 10:17:34 +03:00
kd-11 d6159a35aa gl/vk/dx12: Fix texture scaling on unnormalized rtt access 2017-02-11 15:45:59 +03:00
O1L a809f33418 GL: try to skip incorrectly prepared FBO's. 2017-02-04 19:32:54 +03:00
Ryan Gonzalez 11e2965b5c Switch to always using a steady clock (closes #2200) 2017-01-29 03:07:08 +03:00
O1L 778199ceb1 Added missing stencil op 2017-01-23 23:49:17 +03:00
kd-11 973bf5abdf gl: Disable overlay if required extension is not supported (#2212) 2016-10-20 11:20:45 +08:00
kd-11 2c803dbe66 gl/vk: Bug fixes and improvements (#2206)
* gl: Only bind attrib textures on thread startup

* gl: Persistent mapped buffers

* gl: Fix emulated primitives in an inlined array

* gl: Do not re-update program information every draw call

* gl/vk: s1 type is signed normalized not unsigned normalized

* gl/rsx: Allow disabling of persistent buffers for debugging

gl: Large heap size is more practical

gl: Fix a bug with legacy opengl buffers

* gl/rsx: Allow emulation of unsupported attribute formats

* gl: Fix typos and remove dprints

gl: cleanup debug prints

* ui: Move the GL legacy buffer toggle to the left pane

* vk/gl: Fix cmp type, its range is [-1,1] not [0,1] SNORM_INT
2016-10-18 15:57:28 +08:00
kd-11 8454949eea gl/vk/rsx: Add a cross-platform overlay text; Minor perf improvements and rsx bugfixes (#2196)
* gl/rsx: Implement platform-agnostic text overlays

gl: Restore performance metrics using new text out helper

gl/rsx: Refactor text generation class

* vk: Enable text overlay

gl/vk: Polish overlay counters implementation

gl: Better resource shutdown for text writer

* gl: Optimization, do not rebind TIUs every frame. Speedup

* gl: Optimizations and improvements to vertex upload code

* gl/vk: Texture format swizzles

vk: Texture format fix

vk: Fix YX format swizzles

* rsx: Decode vertex texture index
2016-10-11 08:55:42 +08:00
kd-11 7a9841bb2a gl/vk/dx12: Fix depth reconstruction bug; Fix sampler parameters (#2188)
vulkan: Fix duplicated code bug

fix rebase
2016-10-04 21:51:40 +08:00
kd-11 5430e1d310 rsx/gl/vk/dx12: Add emulated texture fetch for depth read (#2173)
* rsx/gl/vk/dx12: Add emulated texture fetch for depth read

gl/vk/dx12: Simplify reinterpretation equation

* gl: Remove unnecessary re-swizzle

* glsl: explicitly cast uint to float
2016-09-29 14:54:32 +08:00
kd-11 4f74cbe510 gl: Reverse cull face when origin == top (#2161) 2016-09-27 15:34:01 +08:00
kd-11 7884356e90 gl/vk: Bug fixes and improvements (#2155)
* gl: Fix broken buffer reserve computation

* gl: Texture format fixes

* gl: Two sided lighting

* gl: Always update glsl output registers

* gl: Simplify vertex input declaration

* vk: Always write output registers

* vk/gl: swizzle component read on depth textures

* gl/vk: Use proper MVP matrix

gl: fix broken mvp when window_origin=top

* vk/gl: Move fragment operations block into the proxy function
2016-09-26 20:21:17 +08:00
raven02 fcb853792d Typo fix (#2153) 2016-09-25 16:53:57 +08:00
Oil b72f6da853 Porting of DH's user clip implementation in gl (#2139)
* Porting of DH's user clip implementation in gl

* Tweak clang-format

* rsx: Move inline array to draw_clause structure.

* rsx: Use variant based draw commands.

* rsx: Dump DRAW_ARRAYS and DRAW_INDEX_ARRAYS.

* GL: old recompiler clean up (#2142)

* RSX texture refactor (#2144)

* gl/vk: Enable vertex texture fetch (#2127)

* gl: Enable vertex textures

* rsx: use textureLod instead of generic texture sample

* rsx: handle uploading of W32_X32_Y32_Z32

* gl: Re-enable proper shader logging

remove old logging method that overwrites single file

* gl: Declare texture_coord_scale for vertex samplers

* gl: texture remap fixes; enable remap for vertex textures

* gl: offset texture indices to base layer 16

* rsx: Fix W32_Z32_Y32_X32_FLOAT subresource layout

* vk: Enable vertex textures

* rsx: define special calls for vertex texture fetch

* gl: improved vertex texture fetch setup

* vk: Fix texture formats and component mapping

* vk: Implement vertex texture fetch functions properly

* vk/gl: proper fix for primitive restart index

revert inadvertent decompiler update

* gl: Disable filtering for vertex textures

* Hopefully fix appveyor build (#2148)

* GL/Vulkan: Bug fixes and improvements; alphakill for vulkan (#2146)

* vk: Zero-initialize some more structs

* gl: Clean up fragment program generation code

* vk: Enable alpha kill

* vk: Fix surface clear; redirect output for surface_type:b

* vk: Tie renderpass to program object to avoid incompatible passes

* vk: Properly compute descriptor pool size (#2150)

* rsx: Set default attribute format to float.

* rsx: vertex attribute size is 1 again for CMP, let backend handles this formats properties.

* rsx: Move printing function in a separate header/cpp

* Porting of DH's user clip implementation in gl
2016-09-24 18:21:41 +08:00
kd-11 867e9210d7 gl/vk: Enable vertex texture fetch (#2127)
* gl: Enable vertex textures

* rsx: use textureLod instead of generic texture sample

* rsx: handle uploading of W32_X32_Y32_Z32

* gl: Re-enable proper shader logging

remove old logging method that overwrites single file

* gl: Declare texture_coord_scale for vertex samplers

* gl: texture remap fixes; enable remap for vertex textures

* gl: offset texture indices to base layer 16

* rsx: Fix W32_Z32_Y32_X32_FLOAT subresource layout

* vk: Enable vertex textures

* rsx: define special calls for vertex texture fetch

* gl: improved vertex texture fetch setup

* vk: Fix texture formats and component mapping

* vk: Implement vertex texture fetch functions properly

* vk/gl: proper fix for primitive restart index

revert inadvertent decompiler update

* gl: Disable filtering for vertex textures
2016-09-20 22:23:56 +08:00
raven02 77f8ce503d RSX texture refactor (#2144) 2016-09-19 09:25:49 +08:00
vlj a64053fd68 rsx: Remove some unused code. 2016-08-27 15:40:41 +02:00
raven02 bb66b97251 GL: minor fixes (#2105)
* Minor fixes

* temporary disable 2-sided lighting

* Disable user clip planes until they are properly handled
2016-08-27 14:12:44 +08:00
raven02 a270ac7f02 GL: re-use common fp/vp decompiler (#2100) 2016-08-26 22:23:23 +08:00
Vincent Lejeune b8721b6475 gl: Refactor vertex_buffer.cpp 2016-08-11 19:47:14 +02:00
Nekotekina a7e808b35b EXCEPTION macro removed
fmt::throw_exception<> implemented
::narrow improved
Minor fixes
2016-08-08 19:19:32 +03:00
Vincent Lejeune eb1d4811de rsx: Use a "draw clause" object in rsx_state. 2016-08-05 23:33:40 +02:00
Nekotekina 5a36c57c57 Formatting system improved
`unveil<>` renamed to `fmt_unveil<>`, now packs args to u64 imitating va_args
`bijective...` removed, `cfg::enum_entry` now uses formatting system
`fmt_class_string<>` added, providing type-specific "%s" handler function
Added `fmt::append`, removed `fmt::narrow` (too obscure)
Utilities/cfmt.h: C-style format template function (WIP)
Minor formatting fixes and cleanup
2016-08-04 21:34:00 +03:00
kd-11 74fec275c7 gl: Clear depth surfaces before use; Vulkan cleanup (#1957)
* vk: Clean up a few trap messages

vk: silence a few compile errors

* gl: Clear new depth surfaces to max_depth

gl: use glClear to clear depth buffers
2016-07-22 07:31:58 +08:00
Nekotekina ae634bb87e RSX exception fix
VBlank thread management fix
2016-07-20 15:16:19 +03:00
raven02 e1ff3f4674 rsx: use fragment_textures_count (#1948)
* rsx:  use fragment_textures_count

* Typo: unknow -> unknown
2016-07-19 22:50:40 +08:00
Nekotekina ceb4cb59ac Typo fix: comparaison->comparison 2016-07-19 14:17:25 +03:00
Vincent Lejeune c6ed5a8e6b gl: Fix direct call to rsx::method_registers array. 2016-07-17 17:31:53 +02:00
kd-11 ce56351554 gl: Fix fragment constants streaming (#1907) 2016-07-15 02:25:38 +08:00
kd-11 6401eefb26 gl: reset texture scaling factor during bind
reset vertex textures as well
2016-07-13 21:01:16 +03:00
kd-11 ea6f3ff5ed gl: Properly determine front face winding 2016-07-13 21:01:16 +03:00
kd-11 97581d8635 gl: Fix primitive type checks 2016-07-13 21:01:16 +03:00
Vincent Lejeune 772706ca4c Factorize rsx state 2016-07-07 21:38:57 +02:00
raven02 c3bac7c92b GL: front face regression fix (#1854) 2016-07-01 20:47:56 +08:00
raven02 246b593783 Vulkan/GL: add default case for front face in case of garbage value (#1844) 2016-06-30 12:46:25 +08:00
DH 989f954432 Added WIP vertex textures support 2016-06-28 12:58:44 +03:00
DH 44879dd9f3 Implemented alpha kill and fog 2016-06-27 01:52:08 +03:00
DH 6ae54ae27b RSX: Added legacy non-array vertex attributes support (if count of elements > 1)
Fixed ps1ght games
2016-06-26 21:32:50 +03:00
DH bf8a20c4b9 Avoid using flip hack if it's possible
Also added some documentation about it
2016-06-26 21:32:49 +03:00
DH e296f81a37 Shaders decompiler: support non 2D textures
Do not validate programs with undefined textures uniforms
Minor fix
2016-06-26 21:32:48 +03:00
DH 714e07b1dc OpenGL rendered: use textures cache 2016-06-23 01:01:29 +03:00
DH f30d71da6c OpenGL renderer: improved vertex attributes setup
Minor fixes
2016-06-22 22:46:47 +03:00
DHrpcs3 cbf880fb71 OpenGL renderer: Proper clipping implementation 2016-06-21 19:56:07 +03:00
DHrpcs3 8fd3064905 OpenGL renderer: check for gl errors in release builds.
Ignore null-sized textures
Cleanup
2016-06-21 19:56:06 +03:00
DHrpcs3 3b5cd4845e OpenGL renderer: use correct MVP matrix. Cleanup
Simplified gl::ring_buffer helper
2016-06-21 19:56:05 +03:00
O1L 67fe5e1107 Fill fragment constants 2016-06-21 19:56:02 +03:00
O1L 1778113b71 Use sampler buffers in vertex shaders 2016-06-21 19:56:01 +03:00
O1L 8637754d6a Missed changes 2016-06-21 19:56:00 +03:00
O1L 083c4fc855 Try to use new shaders decompiler in OpenGL backend 2016-06-21 19:56:00 +03:00
raven02 9f1ee5bd28 OGL: disable pixel processing test before flip 2016-06-15 21:51:51 +08:00
kd-11 3a63b62486 gl: dynamically determine texture buffer offset alignment
fix alignment issues for gpus where align < 16 is ok
2016-06-14 00:08:23 +03:00
kd-11 6761d4a30c gl: use streaming buffers for uniform & elem buffer as well
gl: stream uniform data using stream buffer

gl: vertex streaming improvements and bugfixes

gl: add basic timing info check for profiling

gl: ebo streaming fixes and enhancements
2016-06-14 00:08:23 +03:00
kd-11 ed94626411 gl: use unsynchronized buffer streaming for attribs
fix linux build
2016-06-14 00:08:23 +03:00
kd-11 c4102f3b18 vk/gl: Enable alpha test in shaders (#1743) 2016-06-05 20:31:23 +03:00
kd-11 2bcc755fca Ignore clear if surface target is set to CELL_GCM_SURFACE_TARGET_NONE 2016-06-03 17:07:48 +03:00
Ivan aafcf44581 Header optimizations (#1684)
Shouldn't break anything. I hope.
2016-04-27 01:27:24 +03:00
Ivan da7472fe81 Optimizations (#1680)
* Optimizations

1) Some headers simplified for better compilation time
2) Some templates simplified for smaller executable size
3) Eliminate std::future to fix compilation for mingw64
4) PKG installation can be cancelled now
5) cellGame fixes
6) XAudio2 fix for mingw64
7) PPUInterpreter bug fixed (Clang)

* any_pod<> implemented

Aliases: any16, any32, any64
rsx::make_command fixed
2016-04-25 13:49:12 +03:00
Nekotekina b85a68e8a1 Partial commit: RSX 2016-04-15 19:22:36 +03:00
kd-11 1a61ef0440 Remove deprecated features from core profile 2016-04-06 11:16:49 +03:00
Vincent Lejeune bad2d7f121 gl: Move vertex setting in a separate file. 2016-04-02 18:27:49 +02:00
Vincent Lejeune 51a6862bf4 gl: Use right texture target. 2016-03-31 00:15:17 +02:00
Vincent Lejeune 855d6935d9 gl: Enable core context and debug output. 2016-03-27 03:45:30 +02:00
Vincent Lejeune 23a0116975 gl: Expand index for quads/quad_strip/triangle_fan 2016-03-22 21:39:00 +01:00
kd-11 ee0eb70de4 Clean up invalidated resources 2016-03-09 21:11:22 +03:00
Vincent Lejeune 81fcadac38 gl: Use shared surface store (updated)
Update surface cache optimization to master; minor fixes

Fix mixed EOL

Remove unused include; change null bindings to 0
2016-03-05 12:21:45 +03:00
kd-11 7eb5da73fe Allow attrib padding for types that require expansion 2016-03-02 01:09:51 +03:00
Vincent Lejeune 32434dd848 rsx/common/d3d12/gl: Support for fog mode.
Fix hitman 2
2016-02-29 16:31:18 +01:00
Vincent Lejeune a6ba47265f rsx/common/gl: s32k is actually signed short unormalized.
gl fix
2016-02-27 00:21:12 +01:00
kd-11 c5b3359c4b Remove normalization for natively normalized formats; fix cmp attribute type 2016-02-25 17:41:03 +03:00
kd-11 974ea68cf9 Always bind null for unused shader inputs
Properly rename index offsets to improve readability
Handle indexed and other array cases
2016-02-25 02:29:35 +03:00
Vincent Lejeune 5a14644cd4 rsx/common/d3d12/gl: Use span in vertex upload function. 2016-02-22 20:22:47 +01:00
kd-11 a451bb574e Enable texture and render target caching for OpenGL 2016-02-17 16:00:23 +03:00
kd-11 3813c09be6 Properly handle inlined vertex data
Fix initial array offsets for dx12
2016-02-16 15:39:18 +03:00
kd-11 2230975e55 Speed up program constants upload by switching to glBufferSubData over glMapBuffer
Fix initialization for fragment constants buffer

Orphan buffers before update. Remove redundant binding
2016-02-15 12:53:55 +03:00
kd-11 ddf5b52360 Implement rectangle textures natively on OpenGL 2016-02-07 08:57:04 +03:00
kd-11 d3405da5c4 Fix LLVM build by changing variable to appropriate type 2016-02-05 18:57:57 +03:00
kd-11 137821c866 Move program validation from link-time to draw-time to compy with spec 2016-02-04 11:19:13 +03:00
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 5f35f2ac7d rsx/common/d3d12: Support for texture 1d too.
They are used in after burner climax
2016-01-30 01:13:15 +01:00
Vincent Lejeune 149fa9d750 rsx/common: Make RSXFragmentProgram key and not just pointer. 2016-01-27 23:16:06 +01:00
Vincent Lejeune 3bd2114815 rsx-debug: Dump windows origin/pixel center. 2016-01-27 22:41:19 +01:00
Nekotekina 7417033d7f GLGSRender fix 2016-01-27 18:14:39 +03:00
Vincent Lejeune 24255f7883 rsx/common/d3d12/gl: Add some texture info to RSXFragmentProgram 2016-01-26 17:56:01 +01:00
Vincent Lejeune 4ce4cf5242 rsx: Add vertex input and output in RSXVertexProgram. 2016-01-22 01:24:54 +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 3b4339c8e0 rsx: Use enum class for surface related regs 2016-01-15 23:44:27 +01:00
Vincent Lejeune 69272edba6 gl: Fix warnings and enable trat warning as error. 2016-01-13 23:53:11 +01:00
Nekotekina 38531459df Logging system rewritten
GUI doesn't freeze anymore
Some things simplified
2016-01-13 18:54:57 +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 bab52c132d rsx/common/d3d12/gl: Clean ProgramStateCache
Use a_b_c format.
Use using =
Use tuple as output
Use RAII to delete program safely
Ensure const correctness.
2016-01-11 19:21:57 +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 3ac9e0933f gl: fixed nv4097_clear_surface & front face selection
added window shader and clip plane constants to GCM.h
2016-01-06 13:30:25 +02:00
DHrpcs3 48919330d7 rsx methods moved from rsx thread 2016-01-06 13:30:24 +02:00
DHrpcs3 ba12c489ec gl: using tiled region for read/write color buffers and flip
gl: fixed flip buffer row length
compilation fixes
2016-01-06 13:30:23 +02:00
DHrpcs3 dbccf5fbad gl: fixed multiple textures binding 2016-01-06 13:30:23 +02:00
DHrpcs3 509bbddac1 OpenGL renderer: use pitch as image row length
Fixed rsx_utils code style
2016-01-06 13:30:23 +02:00
Vincent Lejeune cae2d929ab rsx/common/d3d12/gl: Implement draw inline array support 2016-01-05 00:31:47 +01:00
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
DHrpcs3 ae83ab5436 GLTexture excluded from GLGSRender and renamed to rsx::gl::texture 2015-12-21 04:14:56 +02:00
Nekotekina 321e6d3a86 fs::get_config_dir, fs::get_executable_dir 2015-12-20 15:41:06 +03:00
Vincent Lejeune 6221fecf3b common/d3d12/gl: Start implementing cubemap sampling 2015-12-16 20:36:34 +01:00
Nekotekina f3311bb5e5 GL render crash fix 2015-12-04 23:37:36 +03:00
Nekotekina ca6783ba9a Threads improved, ID manager improved 2015-12-04 23:37:34 +03:00
Jake 178bcfc8df rsx: Improve NV3089_IMAGE_IN_SIZE and use faster loop for swizzle conversions 2015-12-02 04:22:18 -06:00
Jake 8edf35a9e3 gl: remove incorrect/unneeded glenable enum 2015-12-02 04:22:16 -06:00
Vincent Lejeune 1e9d1c1dd9 gl: Uncomment viewport setting code 2015-11-29 18:22:53 +01:00
Vincent Lejeune 7a996648bb gl: Use fill_gragment_constans_buffer 2015-11-28 20:41:30 +01:00
Vincent Lejeune 5447521ce5 gl: Use uniform buffer for transform constants. 2015-11-19 20:26:14 +01:00
Vincent Lejeune e2d0a63b98 gl: Use factored scale offset function call 2015-11-15 22:52:39 +01:00
Nekotekina a4db58f5f2 Added fs::file_ptr, fom::rewrite, cleanup 2015-11-14 20:26:40 +03:00
O1L c0255208cc Old ini-manager only partially used in VFS. Global configuration used as defualt. 2015-10-27 01:09:31 +04:00
Nekotekina a974ee009e vm::var improved, cleanup
Mostly vm::var initialization introduced.
Added vm::make_var function.
2015-10-14 18:17:37 +03:00
DH a443682aaf OpenGL renderer fixes
Fixed attributes loading
Fixed scale-offset matrix
Fixed attribute debug option
2015-10-14 14:34:55 +03:00
DH 4a55ba3067 OpenGL renderer improvements
Flush program cache at thread exit
Use cached locations
2015-10-14 03:16:39 +03:00
DH 37a3800597 rsx:🧵 disable all vertex data at end command
Fixed OpenGL program attributes cache flush
2015-10-14 01:49:47 +03:00
DH 737080274a Improved NV4097_BACK_END_WRITE_SEMAPHORE_RELEASE and NV4097_TEXTURE_READ_SEMAPHORE_RELEASE commands handling 2015-10-13 20:39:36 +03:00
DH e4f6eb04e8 Added new config system
Added emulator state class
Added emulator events
Minor improvements
2015-10-13 20:32:08 +03:00
Vincent Lejeune fc40b0ed6b RSX: Let backend customize semaphore_*_release 2015-10-13 14:28:58 +02:00
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 4154a1026f Rename RSXThread 2015-10-13 00:04:19 +02:00
Vincent Lejeune dd000caad7 Remove context_dma* 2015-10-13 00:04:18 +02:00
Vincent Lejeune 1e431edf09 Remove fragment program 2015-10-13 00:04:15 +02:00
Vincent Lejeune 67f1f0fea3 Remove fragment constants 2015-10-13 00:04:15 +02:00
Vincent Lejeune beabc88f79 Remove transform constants 2015-10-13 00:04:14 +02:00
Vincent Lejeune e086b7a522 Remove cull/front face 2015-10-13 00:04:13 +02:00
Vincent Lejeune dd1afe85ae Remove alpha func 2015-10-13 00:04:13 +02:00
Vincent Lejeune 60bccf0f10 Remove RSXVertexArray 2015-10-13 00:04:12 +02:00
Vincent Lejeune 73535e0c04 remove m_surface* 2015-10-13 00:04:12 +02:00
Vincent Lejeune 537187687d Remove blend 2015-10-13 00:04:10 +02:00
Vincent Lejeune 403f585a19 Remove color mask 2015-10-13 00:04:10 +02:00
Vincent Lejeune 8dd19d1446 Remove color target 2015-10-13 00:04:09 +02:00
Vincent Lejeune 71552f7ffe Remove logic Op 2015-10-13 00:04:09 +02:00
Vincent Lejeune b2a7f0fd11 Remove stencil 2015-10-13 00:04:08 +02:00
Vincent Lejeune 529efd6ac8 Remove depth 2015-10-13 00:04:07 +02:00
Vincent Lejeune ed61023dd3 Remove clear_surface_color 2015-10-13 00:04:07 +02:00
Vincent Lejeune 28cdfa9feb Rename some methods. 2015-10-13 00:04:06 +02:00
Vincent Lejeune bf474f0cdb remove m_ for textures 2015-10-13 00:04:06 +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 3de47c201c RSX: Create a rsx namespace.
Put get_address inside.
2015-10-13 00:04:04 +02:00
Vincent Lejeune e3e5b46cbf Remove m_ prefix on some members of RSXThread 2015-10-13 00:04:03 +02:00
Vincent Lejeune ceab44b9ab d3d12: Remove D3DGSFrame 2015-10-13 00:04:02 +02:00
Vincent Lejeune 0138a95567 RSX: Remove using namespace ps3 in header 2015-10-13 00:04:02 +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 cc0c3fc98d Implemented fragment constants loading (OpenGL renderer)
Fixed nv308a::color
Minor improvements
2015-10-07 17:36:26 +03:00
DH 86720f90cd rsx:🧵 initialize flip_status
Fixed clear_surface (OpenGL)
Enabled some RSXDebugger features (disasm and textures/buffers view)
Minor improvements
2015-10-06 01:22:21 +03:00
DH 3721941ffb Fixed OpenGL renderer crash
Temporary downgraded OpenGL requirements to OpenGL 3.1 (GLSL 1.40)
Fixed some D3D12 Renderer compilation errors
2015-10-05 18:40:22 +03:00
DH 4fdeeace66 D3D12Renderer: fixed some compilation errors
Removed GSFrameBase2 and D3DGSFrame.
Added frame for NullRender.
Minor improvements and fixes
2015-10-05 13:03:23 +03:00
DH 1c890f85c5 OpenGL: do not align buffers 2015-10-05 12:56:05 +03:00
DH 6cd62a9fd0 OpenGL: fixed nv4097_clear_surface implementation
minor improvements
2015-10-05 12:56:05 +03:00
DH 1e7ded2163 rsx::thread moved from rsx2 branch
TODO: cellResc module & RsxDebugger disabled, DX12 renderer not compilable
2015-10-05 12:56:05 +03:00
Vincent Lejeune 62d7bf2159 Common: Move generic upload texture code from d3d12 2015-10-03 18:25:18 +02:00
Nekotekina 8ae3401ffa Some things improved
shared_mutex_t implemented
GUI Emu Callbacks rewritten
fxm::import, fxm::import_always implemented
cellMsgDialog rewritten
Emu.CallAfter improved (returns std::future)
2015-09-22 16:48:21 +03:00
Raul Tambre 8204737efa Implement GetLicenseArea, category improvements
Also fixed newlines, CellSailEvent, fixed CellMusic changes, fixed
console_write, fixed L10n changes and removed the login dialog for now,
until a more proper implementation.
2015-09-13 10:26:01 +03:00
Nekotekina 5e14310071 noexcept usage fixed
thread_t renamed to named_thread_t
2015-08-24 21:22:48 +03:00
Nekotekina ce494f8847 fmt::by_value, fmt::Format removed 2015-08-24 21:22:42 +03:00
Nekotekina 73b108765e fs:: const renaming, fs::g_tls_error stub 2015-08-24 21:22:39 +03:00
Nekotekina 405d7cb70c Revert "RSX: Fix NV3089"
This reverts commit a86e44deb4.
Reasons: it's broken and not strictly related with d3d12 PR.
2015-08-16 18:37:20 +03:00
raven02 a86e44deb4 RSX: Fix NV3089
Make convert_swizzle.elf works
2015-08-12 00:26:50 +02:00
vlj ca90c05b0a gl : refix 2015-08-12 00:26:32 +02:00
vlj 744b56b12d gl: Fix texture unswizzling
Fix multi texture test
2015-08-12 00:26:32 +02:00
Nekotekina 8f9e1100c8 SPU fixes, various fixes 2015-07-10 04:31:25 +03:00
Nekotekina 6f1e76198a Some fixes 2015-07-10 04:31:18 +03:00
Nekotekina 3aefa2b4e1 ThreadBase rewritten (wip) 2015-07-10 04:31:16 +03:00
vlj f4efeabcf3 RSX: Rename and document function member. 2015-07-03 17:08:06 +02:00
vlj 5abeec85cc GL/RSX: Let backend call LoadVertex themselves. 2015-07-02 19:01:19 +02:00
vlj 288581cb43 RSX/GL/Null: Move semaphore handling to backend 2015-07-02 19:01:18 +02:00
raven02 e6df1ddcc0 Fix CELL_GCM_TEXTURE_CLAMP
It should be GL_CLAMP
2015-05-29 21:59:51 +08:00
vlj 2a7c65c019 GL: Use pipeline cache 2015-05-19 17:26:06 +02:00
Nekotekina 74b3580b69 rfile_t renamed 2015-04-25 00:38:11 +03:00