Commit graph

218 commits

Author SHA1 Message Date
kd-11 f0bd0b5a7c rsx: Conditional render sync optimization
- ZCULL queue was updated to one-per-cb but the conditional render sync hint was not updated.
- Do not unconditionally flush the queue unless the upcoming ref is contained in the active CB.
- This avoids spamming queue flush, which frees up resources and improves performance
2019-07-30 21:13:42 +03:00
Eladash fcc75c8b0f rsx: Write atomically semaphore updates and fix zcull timestamp 2019-07-26 21:27:55 +03:00
Eladash 85b1152e29 Timers scaling and fixes 2019-07-23 00:09:01 +01:00
kd-11 9a7c2784f0 rsx: Do not clip scissor to viewport when doing buffer clear 2019-07-20 16:39:32 +03:00
kd-11 e2574ff100 rsx: Support CSAA transparency without multiple rasterization samples enabled 2019-07-19 15:49:08 +03:00
kd-11 b5a2f0df68 rsx: Implement separate viewport raster clipping
- Merge viewport raster window and scissor into one clipping region
- Viewport raster clip is different from viewport geometry clipping in
hardware as the latter is configurable separately
2019-07-19 14:21:19 +03:00
Eladash c4d8ef4340 rsx: Allow to configure vblank rate
Removed "HLE protection" hack from sys_rsx_context_attribute
2019-07-12 00:19:56 +03:00
Eladash 1ee7b91646 Refactoring (#6143)
Prefer vm::ptr<>::ptr over vm::get_addr.
    Prefer vm::_ptr/base over vm::g_base_addr with offset.
    Added methods atomic_t<>::bts and atomic_t<>::btr .
    Removed obsolute rsx:🧵:Read/WriteIO32 methods.
    Removed wrong check in semaphore_release.
    Added handling for PUTRx commands for RawSPU MFC proxy.
    Prefer overloaded methods of v128 instead of _mm_... in VPKSHUS ppu interpreter precise.
    Fixed more potential overflows that may result in wrong behaviour.
    Added io/size alignment check for sys_rsx_context_iounmap.
    Added rsx::constants::local_mem_base which represents RSX local memory base address.
    Removed obsolute rsx:🧵:main_mem_addr/ioSize/ioAddress members.
2019-06-29 01:27:49 +03:00
kd-11 d26b25816d rsx: Improve profiling setup
- Avoid spamming QPC when not needed
- Free performance when debug overlay is not enabled
2019-06-25 20:50:54 +03:00
kd-11 b893a75002 rsx: Rework RSX offloading
- Use a lockless queue
- Do not enqueue small transfers
2019-06-25 20:50:54 +03:00
kd-11 0fa3bcc336 rsx: Asynchronous data transfer 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
Lassi Hämäläinen a070a414a6 Move rsx::constants and rsx::limits to rsx_utils.h 2019-06-25 17:11:10 +03:00
scribam 1e327ad31b rsx: Apply Clang-Tidy fix "readability-avoid-const-params-in-decls" 2019-06-12 15:11:52 +03:00
scribam 44265aa27d rsx: Apply Clang-Tidy fix "modernize-use-equals-default" 2019-06-12 15:11:52 +03:00
scribam 801fa0113f rsx: Apply Clang-Tidy fix "readability-inconsistent-declaration-parameter-name" 2019-06-12 15:11:52 +03:00
scribam db926ee671 rsx: Apply Clang-Tidy fix "performance-unnecessary-value-param" 2019-06-12 15:11:52 +03:00
scribam 09c9996f31 Use empty() instead of comparing size() with 0
Recommendation from Clang-Tidy: https://clang.llvm.org/extra/clang-tidy/checks/readability-container-size-empty.html
2019-06-01 22:59:23 +03:00
scribam 78c7ef3039 rsx: Use clear() instead of resize(0)
The result is the same but clear [1] has slightly less code than resize [2] and signals better the intent IMHO.

[1] fb7fb646fa/libstdc%2B%2B-v3/include/bits/stl_vector.h (L1495)
[2] fb7fb646fa/libstdc%2B%2B-v3/include/bits/stl_vector.h (L934)
2019-06-01 22:59:23 +03:00
kd-11 8009e53642 rsx: Fix upload block range optimization
- The 'max' index should take the first assigned ID; fixes problems with divisors
2019-05-19 17:33:21 +03:00
scribam 6c5ea068c9 Remove redundant semicolons
Fix "-Wextra-semi" warnings
2019-05-12 18:32:11 +03:00
eladash 6f76e34104 rsx: Fix race on clearing native_ui vs emu_requested flag 2019-04-20 01:04:41 +03:00
elad fc253165e2 Correctness fix for RSXIOMem
- Make RSXIOMem volatile.
- Hint the compiler to check only once the address returned.
2019-03-08 23:44:46 +03:00
kd-11 fb778e4821 rsx: Reimplement attrib divisor 2019-01-25 14:34:22 +03:00
kd-11 64a8829614 rsx: Minor cleanup 2019-01-06 10:44:40 +03:00
kd-11 15488eb247 rsx: Avoid unnecessarily touching framebuffer memory
- Do not bind companion framebuffer when clearing single aspect; let the
  contest mechanism sort it out instead
- Do not prematurely tag framebuffers, instead only do so at
  write-confirmation time. Should avoid false tagging if setup does not
  allow a render to occur.
2019-01-06 10:44:40 +03:00
eladash 45ed58cdaf Fix rsx capture replay
Allow to capture non-increment cmd flag that was missing in command.reg
2018-12-15 19:40:18 +03:00
eladash 87988e9da8 rsx fifo: Stability improvements
* Restore stack in fifo error handling

* Update get register after the cmd execution

* Fix put pause in the middle of command

* Add restore points when branching to self

* Precise nopcmd detection

* Test all invalid cmds for early treatment of queue corruption
2018-12-15 19:40:18 +03:00
kd-11 504ab5a6d4 rsx: Minor cleanup to silence stupid compiler warnings 2018-12-03 20:01:23 +03:00
kd-11 5b6e1420f3 rsx: Pipeline barriers fixed up
- Ensure barriers are invoked even if no draw occurs!
-- Ensures that deferred commands are executed eventually
2018-11-30 23:51:25 +03:00
kd-11 1d19f71a46 rsx: Re-enable fifo error reset 2018-11-30 23:51:25 +03:00
kd-11 5193c99973 rsx: Enable dynamic FIFO preprocessing
- Tries to detect when FIFO preprocessing is beneficial and only enables optimizations if the benefit outweighs the cost
- Current threshold is at least 500 draw calls saved at over 2000 draw calls to justify the overhead
- TODO: More tuning for other CPUs
2018-11-30 23:51:25 +03:00
kd-11 54ec363e88 rsx: Critical pipeline fixes
- Fix scissor and viewport binding behavior
- Fixes recovery if empty scissor is specified and then 'fixed' later
- Optimizes state binding a bit
2018-11-30 23:51:25 +03:00
kd-11 1ad76ad331 rsx: Restructure programs
- Also re-enable pipeline optimizations
2018-11-30 23:51:25 +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
Nekotekina 1b37e775be Migration to named_thread<>
Add atomic_t<>::try_dec instead of fetch_dec_sat
Add atomic_t<>::try_inc
GDBDebugServer is broken (needs rewrite)
Removed old_thread class (former named_thread)
Removed storing/rethrowing exceptions from thread
Emu.Stop doesn't inject an exception anymore
task_stack helper class removed
thread_base simplified (no shared_from_this)
thread_ctrl::spawn simplified (creates detached thread)
Implemented overrideable thread detaching logic
Disabled cellAdec, cellDmux, cellFsAio
SPUThread renamed to spu_thread
RawSPUThread removed, spu_thread used instead
Disabled deriving from ppu_thread
Partial support for thread renaming
lv2_timer... simplified, screw it
idm/fxm: butchered support for on_stop/on_init
vm: improved allocation structure (added size)
2018-10-19 22:22:35 +03:00
Nekotekina da6ce80f4f Make vm::get_super_ptr return contiguous memory
Cleanup RSX code complexity
2018-09-27 23:37:13 +03:00
Nekotekina 306f95a9ae New named_thread template (preview)
Old class named_thread renamed to old_thread
It's too hard to move in a single commit
2018-09-27 14:04:16 +03:00
kd-11 bdf85ed900 rsx: Fixup for vertex attrib parsing
- POS does not have to be fetched from ATTR[0]
  - Confirmed with UC1 that uses WEIGHT for positions
  - At least one POS stream has to exist to feed the position attribute which cannot repeat for a single triangle
2018-09-26 19:41:50 +03:00
kd-11 f72157bcec rsx: Fix vertex attrib parsing 2018-09-25 22:03:35 +03:00
kd-11 a3d44b5e1f rsx: Cleanup changes for the flip patch 2018-09-24 16:44:02 +03:00
Jake 699eadc84f rsx: Move render flip from rsx queue command to flip command 2018-09-24 16:44:02 +03:00
Rui Pinheiro 35139ebf5d Texture cache cleanup, refactoring and fixes 2018-09-24 15:26:40 +03:00
Rui Pinheiro f3029b2b42 Change Cell->RSX map/unmap notifications
This allows for further flexibility on the RSX side, allowing us to fix
some bugs and crashes in later commits.
2018-09-24 15:26:40 +03:00
Lassi Hämäläinen 7aef811ff7 CMake: Refactor CMake build (#5032)
* CMake: Refactor build to multiple libraries

- Refactor CMake build system by creating separate libraries for
  different components
- Create interface libraries for most dependencies and add 3rdparty::*
  ALIAS targets for ease of use and use them to try specifying correct
  dependencies for each target
- Prefer 3rdparty:: ALIAS when linking dependencies
- Exclude xxHash subdirectory from ALL build target
- Add USE_SYSTEM_ZLIB option to select between using included ZLib and
  the ZLib in CMake search path

* Add cstring include to Log.cpp

* CMake: Add 3rdparty::glew interface target

* Add Visual Studio CMakeSettings.json to gitignore

* CMake: Move building and finding LLVM to 3rdparty/llvm.cmake script

- LLVM is now built under 3rdparty/ directory in the binary directory

* CMake: Move finding Qt5 to 3rdparty/qt5.cmake script

- Script has to be included in rpcs3/CMakeLists.txt because it defines
  Qt5::moc target which isn't available in that folder if it is
  included in 3rdparty directory
- Set AUTOMOC and AUTOUIC properties for targets requiring them (rpcs3
  and rpcs3_ui) instead of setting CMAKE_AUTOMOC and CMAKE_AUTOUIC so
  those properties are not defined for all targets under rpcs3 dir

* CMake: Remove redundant code from rpcs3/CMakeLists.txt

* CMake: Add BUILD_LLVM_SUBMODULE option instead of hardcoded check

- Add BUILD_LLVM_SUBMODULE option (defaults to ON) to allow controlling
  usage of the LLVM submodule.
- Move option definitions to root CMakeLists

* CMake: Remove separate Emu subtargets

- Based on discussion in pull request #5032, I decided to combine
  subtargets under Emu folder back to a single rpcs3_emu target

* CMake: Remove utilities, loader and crypto targets: merge them to Emu

- Removed separate targets and merged them into rpcs3_emu target as
  recommended in pull request (#5032) conversations. Separating targets
  probably later in a separate pull request

* Fix relative includes in pad_thread.cpp

* Fix Travis-CI cloning all submodules needlessly
2018-09-18 13:07:33 +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
scribam bf89b709cb Remove useless #include 2018-08-31 20:13:54 +04:00
Lassi Hämäläinen 79cf2832ae Remove Utilities/variant.hpp and use C++17 variant
- Remove also Utilities/variant_visitor.hpp
- Fix variant and variant_visitor usages and #includes
2018-08-31 17:49:59 +04:00
eladash 158019b50f Rsx: fix translation when address is negative
move address shift to where it should be, extend io table to catch all possible values.
2018-08-22 12:20:31 +03:00