kd-11
aea1efd5f2
rsx: Fix instancing constants lookup layout when running interpreter
2025-03-16 16:33:39 +03:00
kd-11
5ae05e26c4
vk: Implement instancing when running in interpreter mode
2025-03-16 16:33:39 +03:00
kd-11
023221bd3b
gl: Reimplement driver capabilities detection in a more sane way
2025-03-16 16:33:39 +03:00
kd-11
14575f3efd
rsx: Always initialize CC register in FP interpreter
2025-03-16 16:33:39 +03:00
kd-11
9f3eb33d80
gl: Double instancing buffer size to improve frametimes
2025-03-16 16:33:39 +03:00
kd-11
225d0b9482
gl: Support transform instancing in the shader interpreter
2025-03-16 16:33:39 +03:00
kd-11
65c0d3d425
gl: Add support for hardware instancing
2025-03-16 16:33:39 +03:00
kd-11
3d3fc2f3cd
rsx: Fix cache invalidation when instancing state changes
2025-03-16 16:33:39 +03:00
kd-11
b2ed787c3a
rsx: Invalidate program cache hint if the current vertex or fragment program is re-evaluated
2025-03-12 15:36:29 +03:00
DH
db945f6aed
android: do not invalidate swapchain on VK_SUBOPTIMAL_KHR
...
Workaround
2025-03-10 21:09:27 +01:00
DH
95d0cb18e4
Coding style issues fixes
...
Thanks @Megamouse
2025-03-10 21:09:27 +01:00
DH
cac068dad9
rsx: implement android swapchain
2025-03-10 21:09:27 +01:00
DH
7520c09087
rsx: reduce shaders_cache::unpacked_type size for android
2025-03-10 21:09:27 +01:00
DH
2ebf257f84
vm: removed c_page_size, it cannot be used by globals
2025-03-10 21:09:27 +01:00
DH
798c194025
android stuff
2025-03-10 21:09:27 +01:00
kd-11
11e214f332
rsx: Handle out-of-bounds read/writes from FIFO register IO
2025-03-10 16:07:26 +03:00
kd-11
2dfbab457a
rsx: Reupload surface if the surface cache denies knowledge of it
2025-03-09 22:01:49 +03:00
kd-11
e6b3c2061e
rsx: Fix boundary seek logic for FIFO addresses
2025-03-09 01:36:26 +03:00
kd-11
3be1e0b6ff
rsx: Clean up and simplify texture register manipulation tests
2025-03-08 22:10:55 +03:00
kd-11
93899bf5a9
rsx/FIFO: Reimplement grabbing command args as a contiguous pointer
...
- Allows grabbing indefinite size of buffer as long as the data is contiguously mapped.
2025-03-08 22:10:55 +03:00
kd-11
ec120d1cbb
rsx: Add stubs for FP rehash checks on barrier acquire
2025-03-08 22:10:55 +03:00
kd-11
df7a2ab467
rsx: Unify fragment texture config changes dirty bit tracking
...
- Reverts unnecessary changes to dirty tracking.
- The GPU keeps a separate set of details, so all dirty bits are treated equal.
2025-03-08 22:10:55 +03:00
kd-11
a39a743abe
rsx/prog: Fix compiler warning
2025-03-08 22:10:55 +03:00
kd-11
eca86ad449
rsx: Add program cache lookup ellision rate to overlay
2025-03-08 22:10:55 +03:00
kd-11
a1c8f3a528
rsx/prog: Use a proper cache hint key instead of disjointed counters
2025-03-08 22:10:55 +03:00
Elad
32df7315be
RSX: Do not invalidate FP analysis on same register writes
2025-03-08 14:21:57 +02:00
Elad
7995181235
RSX: Return FP shader as is if no changes were made
2025-03-08 14:21:57 +02:00
Elad
75dc2a12e2
RSX: Return VP shader as is if no changes were made
2025-03-08 14:21:57 +02:00
Elad
ae39c5b8cb
RSX: Optimize vertex program ucode invalidation
2025-03-08 14:21:57 +02:00
Whatcookie
bfb9dfea7e
RSX: ProgramStateCache loops optimizations
2025-03-07 09:54:12 +02:00
Elad
8ed2089070
RSX: Fix UB in FIFO_control::get_current_arg_ptr()
2025-03-07 09:41:52 +02:00
kd-11
59280a133c
vk: Do not force strict query scopes by default
2025-03-05 14:40:35 +03:00
Megamouse
ca3d02fc61
static analysis: ensure is always false
2025-03-05 00:12:30 +01:00
Megamouse
a49c4c7682
static analysis: duplicate assignment
2025-03-05 00:12:30 +01:00
Megamouse
7a6a96254c
static analysis: remove unused overrides
2025-03-05 00:12:30 +01:00
Megamouse
6abb863a54
static analysis: std::move
2025-03-05 00:12:30 +01:00
Megamouse
815a4e7dc1
static analysis: less references
2025-03-05 00:12:30 +01:00
Megamouse
df2f68da44
static analysis: more const ref
2025-03-05 00:12:30 +01:00
Megamouse
a8fc3f5f74
static analysis: more const
2025-03-05 00:12:30 +01:00
Megamouse
cb659474d9
overlays: fix friends/trophy list input on paused emulation
2025-03-03 18:21:35 +01:00
Malcolm Jestadt
1920aee415
RSX: Fix issue where linux builds could hit illegal instruction on machines without AVX-512
...
- Place avx-512 function attributes in their own functions
2025-03-03 15:58:55 +03:00
Megamouse
bd41774960
overlays/osk: implement horizontal scroll
2025-03-02 23:44:08 +01:00
Megamouse
4aa25285a3
overlays: use thread local static vector in get_glyph_data
2025-03-02 23:44:08 +01:00
Megamouse
72ee5ed10d
overlays: add ability to hide hidden trophies
2025-03-02 23:02:22 +01:00
Megamouse
048a44cbe5
overlays: add getter for selected entry
2025-03-02 23:02:22 +01:00
Megamouse
6f5f7ece06
overlays: fix and optimize layout exit loops
2025-03-02 23:02:22 +01:00
Malcolm Jestadt
6bd1ab576c
RSX: Use AVX-512-ICL code in get_vertecx_program_ucode_hash and in vertex_program_compare
...
- Code is about 4 times as fast on my zen4 machine
- Should be twice as fast on zen5 machines with full width AVX-512
2025-03-02 09:07:07 +02:00
kd-11
0aca5e222f
vk: Fix MacOS build
2025-03-01 18:02:32 +03:00
kd-11
b8411a90cd
vk/swapchain: Fix WSI swapchain configuration overrides
2025-03-01 18:02:32 +03:00
kd-11
4e9365f76b
vk/swapchain: Consolidate native swapchain implementation for platforms where it is not implemented
...
- NATIVE swapchain only matters in headless scenarios or where we otherwise cannot access WSI properly.
- It is now optional, with a stub provided when WSI is not available.
2025-03-01 18:02:32 +03:00