592 Commits

Author	SHA1	Message	Date
Marcin Mikołajczyk	139a253edc	Stub PM4 0x8E opcode (#2998)	2025-05-27 12:09:03 -07:00
Lander Gallastegi	10d09ac977	Added back the "Attempted to track non-GPU memory" assert. (#2980) ... * Fix log message * Add non-GPU memory assert back	2025-05-24 01:48:40 +03:00
squidbus	e5c6c88835	texture_cache: Handle overlap with equal address and different tiling mode (#2981)	2025-05-23 10:56:46 -07:00
Marcin Mikołajczyk	e518a7062c	Fix image extent in buffer copy to image (#2961)	2025-05-22 12:17:34 -07:00
Lander Gallastegi	f9bbde9c79	video_core: Implement DMA. (#2819) ... * Import memory * 64K pages and fix memory mapping * Queue coverage * Buffer syncing, faulted readback adn BDA in Buffer * Base DMA implementation * Preparations for implementing SPV DMA access * Base impl (pending 16K pages and getbuffersize) * 16K pages and stack overflow fix * clang-format * clang-format but for real this time * Try to fix macOS build * Correct decltype * Add testing log * Fix stride and patch phi node blocks * No need to check if it is a deleted buffer * Clang format once more * Offset in bytes * Removed host buffers (may do it in another PR) Also some random barrier fixes * Add IR dumping from my read-const branch * clang-format * Correct size insteed of end * Fix incorrect assert * Possible fix for NieR deadlock * Copy to avoid deadlock * Use 2 mutexes insteed of copy * Attempt to range sync error * Revert "Attempt to range sync error" This reverts commit dd287b48682b50f215680bb0956e39c2809bf3fe. * Fix size truncated when syncing range And memory barrier * Some fixes (and async testing (doesn't work)) * Use compute to parse fault buffer * Process faults on submit * Only sync in the first time we see a readconst Thsi is partialy wrong. We need to save the state into the submission context itself, not the rasterizer since we can yield and process another sumission (if im not understanding wrong). * Use spec const and 32 bit atomic * 32 bit counter * Fix store_index * Better sync (WIP, breaks PR now) * Fixes for better sync * Better sync * Remove memory coveragte logic * Point sirit to upstream * Less waiting and barriers * Correctly checkout moltenvk * Bring back applying pending operations in wait * Sync the whole buffer insteed of only the range * Implement recursive shared/scoped locks * Iterators * Faster syncing with ranges * Some alignment fixes * fixed clang format * Fix clang-format again * Port page_manager from readbacks-poc * clang-format * Defer memory protect * Remove RENDERER_TRACE * Experiment: only sync on first readconst * Added profiling (will be removed) * Don't sync entire buffers * Added logging for testing * Updated temporary workaround to use 4k pages * clang.-format * Cleanup part 1 * Make ReadConst a SPIR-V function --------- Co-authored-by: georgemoralis <giorgosmrls@gmail.com>	2025-05-22 21:00:15 +03:00
squidbus	3f949d2b6c	amdgpu: Handle 32-bit Unorm formats. (#2974)	2025-05-22 03:16:20 -07:00
squidbus	99eaba7c96	liverpool: Lower SetQueueReg to warning log. (#2930)	2025-05-13 16:08:25 -07:00
Marcin Mikołajczyk	647b1d3ee4	Handle VgtStreamoutFlush event (#2929)	2025-05-13 14:34:22 -07:00
squidbus	073f931729	fix: PM4CmdWaitRegMem memory address	2025-05-13 14:14:11 -07:00
Marcin Mikołajczyk	e5b675d607	Handle IT_WAIT_REG_MEM with Register argument (#2927)	2025-05-13 13:56:20 -07:00
squidbus	4cd13ea9d8	fix: Disable emitting bounds checks until fixed.	2025-05-12 23:10:48 -07:00
squidbus	6bbb424c28	vk_instance: Enable robustImageAccess2 (#2922)	2025-05-12 18:03:55 -07:00
squidbus	b23f6fdc1d	buffer_cache: Split updateBuffer calls into 65536 byte chunks. (#2915)	2025-05-12 15:27:47 -07:00
Randomuser8219	9baa58dd92	renderer_vulkan: Properly enable dualSrcBlend feature. (#2921) ... * renderer_vulkan: Properly enable dualSrcBlend feature * whoops	2025-05-12 15:26:55 -07:00
squidbus	02d3ed4973	liverpool: Log more information on SetQueueReg. (#2912)	2025-05-11 20:27:43 -07:00
squidbus	a1439b15cf	gnm: Implement sceGnmDrawIndexIndirectMulti (#2889)	2025-05-09 10:04:37 -07:00
squidbus	b130fe6ed5	vulkan: Handle incompatible depth format using null binding. (#2892) ... Co-authored-by: kalaposfos13 <153381648+kalaposfos13@users.noreply.github.com>	2025-05-09 08:43:20 -07:00
oltolm	9a22185ab7	vulkan: do not use VK_EXT_extended_dynamic_state (#2880) ... fixes Bloodborne crashing on RX 580	2025-05-04 03:11:02 -07:00
squidbus	eb09c4ccce	vk_presenter: Use correct format for output frame image and view. (#2871)	2025-05-01 20:10:42 -07:00
Mahmoud Adel	b0e4e87ff3	Implement SnormNz conversion (#2841) ... * + * + * Unpack Snorm 2x16 * + * SintToSnormNz * all is broken ig.... * review changes * my stupid ass messed all while trying to resolve the conflicts.. * + * + * fix rebase * clang-format fix (1) * clang-format fix (2) --------- Co-authored-by: squidbus <175574877+squidbus@users.noreply.github.com>	2025-05-01 02:12:15 -07:00
squidbus	5fd5b62539	shader_recompiler: Few fixes for buffer number conversions. (#2869) ... * liverpool: Pass correct color buffer number type for conversion mapping. * shader_recompiler: Apply number conversion to vertex inputs.	2025-04-30 20:46:16 -07:00
Marcin Mikołajczyk	c08f92aca1	Implement IMAGE_ATOMIC_FMIN and IMAGE_ATOMIC_FMAX for 32bit floats (#2820) ... * Implement IMAGE_ATOMIC_FMIN and IMAGE_ATOMIC_FMAX for 32bit floats * Handle missing VK_EXT_shader_atomic_float2	2025-04-30 11:42:08 -07:00
squidbus	5e3157a82c	vk_rasterizer: Fix updating wrong color attachment when skipped by mask. (#2859)	2025-04-29 16:29:35 -07:00
squidbus	81fa9b7fff	shader_recompiler: Add lowering pass for when 64-bit float is unsupported. (#2858) ... * shader_recompiler: Add lowering pass for when 64-bit float is unsupported. * shader_recompiler: Fix PackDouble2x32/UnpackDouble2x32 type. * shader_recompiler: Remove extra bit cast implementations.	2025-04-28 00:04:16 -07:00
squidbus	59d060bc16	fix: gcc compile	2025-04-27 21:06:10 -07:00
squidbus	83fd0683fa	fix: Properly enable depthBounds feature.	2025-04-27 20:57:04 -07:00
squidbus	81ad31ce31	pp_pass: Use correct surface format. (#2860)	2025-04-27 20:56:17 -07:00
squidbus	ce3aded3e5	fix: Intel crash on startup.	2025-04-23 11:09:04 -07:00
squidbus	aeee7706ee	renderer_vulkan: Restore Vulkan version to 1.3 (#2827) ... Co-authored-by: georgemoralis <giorgosmrls@gmail.com>	2025-04-23 13:28:31 +03:00
squidbus	53ca64f6ff	renderer_vulkan: Support loading Vulkan layers on macOS SDL build. (#2832)	2025-04-23 01:10:03 -07:00
squidbus	ee7fe305c9	fix: Make sure right MoltenVK is loaded for macOS SDL.	2025-04-22 23:18:39 -07:00
squidbus	0297aee3f4	texture_cache: Relax mismatched image type from assert to cache miss. (#2830)	2025-04-22 15:03:43 -07:00
squidbus	5b0205bc59	fix: Texture cache image type assert.	2025-04-22 14:00:08 -07:00
Dmugetsu	cda421434b	Adding missmatch info to Texture cache assert (#2828)	2025-04-22 22:49:29 +03:00
squidbus	df89241eb8	vk_rasterizer: Improve viewport depth calculations. (#2799)	2025-04-17 13:04:48 -07:00
squidbus	ddf3df7f56	fix: clang-format	2025-04-16 10:48:17 -07:00
squidbus	04e4ce0775	fix: Bad Intel drivers.	2025-04-16 09:52:18 -07:00
squidbus	3bc876ca78	renderer_vulkan: Improve handling of required vs optional extensions. (#2792) ... * renderer_vulkan: Improve handling of required vs optional extensions. * documents: Update quickstart Vulkan requirements.	2025-04-15 22:31:58 -07:00
squidbus	247e935353	externals: Update MoltenVK (#2791)	2025-04-15 17:13:19 -07:00
squidbus	683a223c1b	vk_rasterizer: Control mapped_ranges access with shared lock. (#2788)	2025-04-14 20:58:33 -07:00
squidbus	37d4cd091c	renderer_vulkan: Make some primitive state dynamic. (#2764) ... * renderer_vulkan: Make some primitive state dynamic. * renderer_vulkan: Silence MoltenVK primitive restart warning spam.	2025-04-10 16:20:01 -07:00
Dmugetsu	5abec2a291	Enabling Depth Bias Explicity (#2766)	2025-04-09 17:06:54 -07:00
squidbus	03b1fef331	renderer_vulkan: Only update dynamic state when changed. (#2751)	2025-04-08 07:22:16 -07:00
squidbus	b8f6ef1c0b	externals: Update MoltenVK (#2754)	2025-04-07 12:23:04 +03:00
squidbus	0c6f2b470f	renderer_vulkan: Use more depth-stencil dynamic state. (#2749)	2025-04-05 23:14:27 +03:00
squidbus	afd0251dd2	shader_recompiler: Use VK_AMD_shader_trinary_minmax when available. (#2739) ... * shader_recompiler: Use VK_AMD_shader_trinary_minmax when available. * shader_recompiler: Simplify signed/unsigned trinary instruction variants.	2025-04-02 23:36:54 +03:00
squidbus	f85d8df71e	vulkan: Lower list primitive restart warning to debug log. (#2725)	2025-03-30 22:02:33 +03:00
kalaposfos13	5caab76a45	Implement DmaDataSrc::MemoryUsingL2 and DmaDataDst::MemoryUsingL2 (#2680) ... * Implement DmaDataSrc::MemoryUsingL2 and DmaDataDst::MemoryUsingL2 * Add L2 handling to the other place it's used	2025-03-26 23:03:50 +02:00
kalaposfos13	a1ec8b0a88	Handle compute packets that are split between the ends of two command buffers (#2476) ... * Squashed initial implementation * Logging for checking if buffers are memory contiguous * Add check to see if first instruction is valid in the next buffer to avoid false positives * Oof * Replace old code with IndecisiveTurtle's new, better implementation * Add `unlikely` keyword to the split packet handling branches Co-authored-by: TheTurtle <47210458+raphaelthegreat@users.noreply.github.com> --------- Co-authored-by: IndecisiveTurtle <47210458+raphaelthegreat@users.noreply.github.com>	2025-03-26 00:01:21 +02:00
Vinicius Rangel	5691046dcc	Renderer fixes (Splash + Aspect Ratio) (#2645) ... * rewrite splash removed Splash class rewrite using imgui texture manager fix crashes & old validation error * handle games with abnormal aspect ratios	2025-03-13 18:10:24 +02:00