Commit Graph

1623 Commits

Author SHA1 Message Date
Billy Laws
2ce146e28f Don't crash on the Grp0SetSubDevMask TertOp
Used by Vulkan games to set the SLI mask, not applicable to the switch.
2022-11-02 17:46:07 +00:00
Billy Laws
1d83dadefb Drop size restruction bypass for frequently synced buffers
In cases where large buffers are updated every draw this could seriously increase memory usage beyond 3GB in the megabuffer.
2022-11-02 17:46:07 +00:00
Billy Laws
1088ed514c Introduce texture usage system to ensure RPs are split when necessary
Vulkan doesn't allow sampling a texture and using it as an RT in the same RP, by tracking the texture usage status and splitting RPs when this occurs we can avoid such potential sync errors.
2022-11-02 17:46:07 +00:00
Billy Laws
2dd4698441 Adjust texture matching hacks 2022-11-02 17:46:07 +00:00
Billy Laws
4f5c9047ef Add some additional texture formats used by Vulkan games 2022-11-02 17:46:07 +00:00
Billy Laws
6a830dfac5 Use shader-compiler side {S,U}Scaled format emulation 2022-11-02 17:46:07 +00:00
Billy Laws
579fd04117 Fixup ReadTextureType shader compiler callback 2022-11-02 17:46:07 +00:00
Billy Laws
b04d18eba5 Add support for split mappings to I2M uploads
Used by Super Mario Sunshine and other Vulkan games.
2022-11-02 17:46:07 +00:00
Billy Laws
db5e208379 Clear images even when aspects mismatch 2022-11-02 17:46:07 +00:00
Billy Laws
3c8df327f1 Fixup subpass barriers and flags 2022-11-02 17:46:07 +00:00
Billy Laws
5ab80901c6 Drop some debug code 2022-11-02 17:46:07 +00:00
Billy Laws
4de89c8839 GPU NEW MARGEBAC 2022-11-02 17:46:07 +00:00
Billy Laws
7670c83405 Ensure textures are clean before paging them out 2022-11-02 17:46:07 +00:00
Billy Laws
1a2351386d Add u64 iova ctor 2022-11-02 17:46:07 +00:00
Billy Laws
93d43e0115 Fully fill in swizzle component mappings
Avoids the rest being default initialised to identity, which would break the intended effect of them.
2022-11-02 17:46:07 +00:00
Billy Laws
37ff0ab814 Add buffer manager support for accelerated copies
These will be sequenced on the GPU/CPU depending on what's optimal and avoid any serialisation
2022-11-02 17:46:07 +00:00
Billy Laws
cac287d9fd Implement accelerated uploads/copies through buffer manager
Previously, both I2M uploads and DMA copies would force GPU serialisation if they happened to hit a trap or were used to copy GPU dirty buffers. By using the buffer manager to implement them on the host GPU we can avoid such slowdowns entiely.
2022-11-02 17:46:07 +00:00
Billy Laws
c5ec484d9a Avoid redundantly passing executor in ctors when it's already in ChannelCtx 2022-11-02 17:46:07 +00:00
Billy Laws
463394ba72 Pass correct size for XFB buffers 2022-11-02 17:46:07 +00:00
Billy Laws
bd976676f4 Fix SNorm vertex formats 2022-11-02 17:46:07 +00:00
Billy Laws
b74098570f Zero-out unused XFB varyings before passing to hades 2022-11-02 17:46:07 +00:00
Billy Laws
22f3ba6b93 Mark XFB buffers as GPU dirty 2022-11-02 17:46:07 +00:00
Billy Laws
26aeeaecf5 Add constant buffer GPU write pipeline barrier 2022-11-02 17:46:07 +00:00
Billy Laws
0b5d9308c4 Be more careful about potentially-unneeded GPU->CPU syncs
These can be especially expensive so should be avoided as much as possible.
2022-11-02 17:46:07 +00:00
Billy Laws
e6530e2386 Delete graphics_context
F
2022-11-02 17:46:07 +00:00
Billy Laws
ac2e6c125b Switch to Roboto for Korean font 2022-11-02 17:46:07 +00:00
Billy Laws
b24a8465da Don't require depthClamp 2022-11-02 17:46:07 +00:00
Billy Laws
9055c98e09 Only enable debug/verbose logs in (rel)debug builds 2022-11-02 17:46:07 +00:00
Billy Laws
0ebdbcf0ff Don't lock stateMutex when updating buffer cycle 2022-11-02 17:46:07 +00:00
Billy Laws
dd360b8f75 Pass correct wait semaphore array size to queue submit 2022-11-02 17:46:07 +00:00
Billy Laws
c78a4b9699 Fixup buffer recreation to avoid deadlock when waiting on srcs 2022-11-02 17:46:07 +00:00
Billy Laws
d236bfe454 Enable depthClamp VK device feature 2022-11-02 17:46:07 +00:00
Billy Laws
95d849e1f6 Check FenceCycle signalled flag immediately before waiting
The lock release within the wait for submission means that another thread could end up signalling the cycle and then the VK wait still happen after when the lock has been reacquired.
2022-11-02 17:46:07 +00:00
Billy Laws
1a23b929a7 Avoid chaining cycles in buffer recreation
This had a chance of creating circular chains which obviously caused issues, just do a wait instead for now.
2022-11-02 17:46:07 +00:00
Billy Laws
a15db9cb06 Update hades submodule 2022-11-02 17:46:07 +00:00
Billy Laws
cfc55e60b0 Add robin map submodule 2022-11-02 17:46:07 +00:00
Billy Laws
6c0f084aae Introduce hack to ignore frequently read-back textures
Readback can be especially slow on mobile due to the varying load pattern it creates which often prevents the CPU/GPU from clocking up. Since some games perform texture readback but don't actually use it for anything significant implement a hack to skip it and significantly improve performance in such cases.
2022-11-02 17:46:07 +00:00
Billy Laws
e45e7546c8 Redesign buffer megabuffering
Due to the frequency at which is is called megabuffering performance is critical to the performance of the entire emulator, especially in high-drawcall-count scenarios. After the view redesign, megabuffering on a per-view level was no longer possible nor desirable, and thus megabuffering was modified to just copy for every usage of a view. This worked great at the time since there were other bottlenecks, however gpu-new has since removed almost all of them and megabuffering is now a major sore point. Fix this by megabuffering small chunks and storing them in a page-table like structure within the buffer, these chunks can be referenced by multiple views and will be smartly invalidated whenever the sequence number or execution number changes to avoid any sequencing issues. In addition to this, to help the case where almost the whole buffer is read every single frame across a set of multiple views, an optimisation to skip the chunked tracking and use one large single megabuffer allocation and one single memcpy has been introduced. This reduces the overall amount of time spent in memcpy since large memcpys are quicker.
2022-11-02 17:46:07 +00:00
Billy Laws
7ea9aa52f5 Speed up reported guest GPU time
Avoids triggering DRS in games in cases where it wouldn't actually benefit anything due to being CPU bottlenecked.
2022-11-02 17:46:07 +00:00
Billy Laws
31c2fb7d7a Fixup IDirectory read 2022-11-02 17:46:07 +00:00
Billy Laws
7491178a9e Pass base array layer to texture views 2022-11-02 17:46:07 +00:00
Billy Laws
ff57d2fbbf Enforce stronger format and weaker dimension texture compat checks
Rather than using just bpb for format compat, additionally check that the exact component bit layout matches since many games end up reusing RTs for unrelated textures. The texture size requirements have also been weaked to only check the resulting layer size as opposed to width/height - this is somewhat hacky but it gets around the problem of blocklinear alignment.
2022-11-02 17:46:07 +00:00
Billy Laws
14af383238 Only allow submitting swapchainImageCount images for host present at a time
Prevents situations where nothing would otherwise be waiting on the GPU and since presentation no longer blocks too many images would be submitted for presentation.
2022-11-02 17:46:07 +00:00
Billy Laws
bcd96ac77d Fixup A8R8G8B8 TIC format mapping
8-bit formats are inverted in TICs compared to Vulkan
2022-11-02 17:46:07 +00:00
Billy Laws
90466b8830 Implement depth clamp rasterisation state
Used in SMO for shadows.
2022-11-02 17:46:07 +00:00
Billy Laws
1cfc4278f9 Disable preserve buffer/texture attachment opt for now
Causes several issues and crashes in Pokemon without an obvious cause.
2022-11-02 17:46:07 +00:00
Billy Laws
e483cf9634 Use shader memory mirror when reading guest shaders
Avoids triggering any traps that may be present on the region
2022-11-02 17:46:07 +00:00
Billy Laws
f6e4328b5a Ensure blit src/dst textures are attached as execution cycle dependencies
Since they're not in the TIC pool they would otherwise be freed
2022-11-02 17:46:07 +00:00
Billy Laws
77a131df60 Support using in-app renderdoc API to capture individual executions 2022-11-02 17:46:07 +00:00
Billy Laws
576bc6f37e Add CommandExecutor slot count setting 2022-11-02 17:46:07 +00:00
Billy Laws
1a0819fb76 Use semaphores for presentation engine frame synchronisation
Avoids waits on the CPU which can be costly and confuse the scheduler, also reduces latency significantly.
2022-11-02 17:46:07 +00:00
Billy Laws
0670e0e0dc Support using Vulkan semaphores with fence cycles
In some cases like presentation, it may be possible to avoid waiting on the CPU by using a semaphore to indicate GPU completion. Due to the binary nature of Vulkan semaphores this requires a fair bit of code as we need to ensure semaphores are always unsignalled before they are waited on and signalled again. This is achieved with a special kind of chained cycle that can be added even after guest GPFIFO processing for a given cycle, the main cycle's semaphore can be waited and then the cycle for the wait attached to the main cycle and it will be waited on before signalling.
2022-11-02 17:46:07 +00:00
Billy Laws
5b72be88c3 Stub ldn:u service 2022-11-02 17:46:07 +00:00
Billy Laws
77d76ed05a Batch contiguous GMMU ranges into one 2022-11-02 17:46:07 +00:00
Billy Laws
e52dbf202f Pass more Maxwell3D registers into interconnect 2022-11-02 17:46:07 +00:00
Billy Laws
83c7ed314e Setup KThread pthread handle in StartThread
Avoids a race with starting the thread and the handle not being set yet
2022-11-02 17:46:07 +00:00
Billy Laws
9784ae23e9 Skip checking affinity before taking load-balance WaitScheduler path
The affinity mask may be set after the wait has began
2022-11-02 17:46:07 +00:00
Billy Laws
ad3195e06f Split out guest texture layer size calcs into a seperate func 2022-11-02 17:46:07 +00:00
Billy Laws
8fa83fdf13 Fix deswizzling non-pow2 block size formats
We need to use DivideCeil to avoid rounding off part of the texture.
Fixes texture in Nier Automata: Game of the YoRHa edition.
2022-11-02 17:46:07 +00:00
Billy Laws
27de42f8df Use surfaceClip as a hint for the underlying rendertarget size
TIC sizes may not be aligned to block linear dimensions whereas RT sizes are and then limited by the surface clip. By using this to determine surface size we are more likely to get a match in texture manager for any future usages.
2022-11-02 17:46:07 +00:00
Billy Laws
297597f697 Fix texture manager depth compat comparison 2022-11-02 17:46:07 +00:00
Billy Laws
500f817a28 Synchronize all non-matching textures back to host before recreation 2022-11-02 17:46:07 +00:00
Billy Laws
05581f2230 Remove now redundant buffer/texture/megabuffer manager locks
They have been superseeded by the global channel lock
2022-11-02 17:46:07 +00:00
Billy Laws
f5a141a621 Add dirty resource operator* 2022-11-02 17:46:07 +00:00
Billy Laws
b72720e8db Finish off transform feedback implementation 2022-11-02 17:46:07 +00:00
Billy Laws
36fd885b49 Pack all draw state into a struct to avoid std::function allocations 2022-11-02 17:46:07 +00:00
Billy Laws
b5d0060c3f Only use scissor for clear rect when enabled 2022-11-02 17:46:07 +00:00
Billy Laws
f93df35e6c Only set line width when wideLines feature is supported 2022-11-02 17:46:07 +00:00
Billy Laws
4cebdfc8d3 Pass texture and cbuf state into pipeline manager for hades callbacks 2022-11-02 17:46:07 +00:00
Billy Laws
9ce848d4e0 Implement descriptor update batching and push descriptors
Batching helps to avoid the need to attach so many objects to the fence cycle, which ends up taking a fair bit of time due to the allocation required.
2022-11-02 17:46:07 +00:00
Billy Laws
62a165b51e Reformat maxwell3d interconnect codebase 2022-11-02 17:46:07 +00:00
Billy Laws
3766be59e7 Zero out vertex attribute state when disabled to avoid creating redundant pipelines 2022-11-02 17:46:07 +00:00
Billy Laws
751e3356e1 Keep shader trap lock held for the duration of an execution
Avoids constant relocking on the GPFIFO thread (~0.5% of total time)
2022-11-02 17:46:07 +00:00
Billy Laws
314a9bccbc Allow megabuffering readonly SSBOs 2022-11-02 17:46:07 +00:00
Billy Laws
4c2db0ba01 Implement ReadCbufValue and ReadTextureType hades callbacks
Used for bindless and BRX instruction emulation.
2022-11-02 17:46:07 +00:00
Billy Laws
2163f8cde6 Implement alpha test pipeline state 2022-11-02 17:46:07 +00:00
Billy Laws
c86ad638c4 Keep track of transform feedback varyings pipeline state 2022-11-02 17:46:07 +00:00
Billy Laws
7ad2d94345 Zero out blend state when disabled to avoid creating redundant pipelines 2022-11-02 17:46:07 +00:00
Billy Laws
4052a93051 Force non-pushdescriptors for blit helper shader 2022-11-02 17:46:07 +00:00
Billy Laws
cb2a8c6d24 Enable wideLines Vulkan feature 2022-11-02 17:46:07 +00:00
Billy Laws
a3369637a9 Don't entirely wipe out per-index TIC cache efter each execution
Keep a copy of the old TIC entry and view even after purge caches and use the execution number to check validity instead, if that doesn't match then just memcmp can be used as opposed to a full hash and map lookup.
2022-11-02 17:46:07 +00:00
Billy Laws
98c0cc3e7f Impl preserve attached buffers/textures to avoid GPFIFO lock thrashing
When profiling SMO, it became obvious that the constant locking of textures and buffers in SyncDescriptors took up a large amount of CPU time (3-5%), a precious resource in intensive areas like Metro. This commit implements somewhat of a workaround to avoid constant relocking, if a buffer is frequently attached on the GPU and almost never used on the CPU we can keep the lock held between executions. Of course it's not that simple though, if the guest tries to lock a texture for the first time which has already been locked as preserve on the GPFIFO we need to avoid a deadlock. This is acheived through a combination of two things: first we periodically clear the locked attachments every 2*SlotCount submissions, preventing a complete deadlock on the CPU (just a long wait instead) and meaning that the next time the resource is attached on the GPU it will not be marked for preservation due to having been locked on the guest before; second, we always need to unlock everything when the GPU thread runs out of work, as the perioding clearing will not execute in this case which would otherwise leave the textures locked on the GPFIFO thread forever (if guest was waiting on a lock to submit work). It should be noted that we don't clear preserve attached resources in the latter scenario, only unlock them and then relock when more work is available.
2022-11-02 17:46:07 +00:00
Billy Laws
0e8ccf1e99 Use memory_order_release for new descriptor set allocations 2022-11-02 17:46:07 +00:00
Billy Laws
34db5097da Avoid using a shared_ptr reference to cycle for command buffer submission
Somehow without this we can sometimes get crashes during appending to circular queue
2022-11-02 17:46:07 +00:00
Billy Laws
0428e8c7da Support forcing regular descriptor sets in VK pipeline cache 2022-11-02 17:46:07 +00:00
Billy Laws
0f394d516b Lock FenceCycle inbetween waiting on chained cycles and checking signalled
Avoids one race where we would end up hogging all the locks of chained cycles and ourself when waiting for submission of previous cycles and prevent any forward progress due to another thread locking one of the chained cycles.
2022-11-02 17:46:07 +00:00
Billy Laws
a015fe753d Only write npad controllerInfo entry on the HID thread if it is valid 2022-11-02 17:46:07 +00:00
Billy Laws
b5446846f7 Stub IsSixAxisSensorAtRest 2022-11-02 17:46:07 +00:00
Billy Laws
6719572b3b Keep track of how often textures/buffers are locked on the CPU
For the upcoming preserve attachment optimisation, which will keep buffers/textures locked on the GPU between executions, we don't want to preserve any which are frequently locked on the CPU as that would result in lots of needless waiting for a resource to be unlocked by the GPU when it occasionally frees all preserve attachments when it could have been done much sooner.  By checking if a resource has ever been locked on the CPU and using that to choose whether we preserve it we can avoid such waiting.
2022-11-02 17:46:07 +00:00
Billy Laws
993ffb56f4 Avoid waiting on texture/buffer fence with trapMutex locked
This could cause deadlocks under certain circumstances by preventing the GPFIFO from making any progress while also waiting on it at the same time.
2022-11-02 17:46:07 +00:00
Billy Laws
3e8bd26978 Add a global gm20b channel lock
Allowing for parallel execution of channels never really benefitted many games and prevented optimisations such as keeping frequently used resources always locked to avoid the constant overhead of locking on the hot path.
2022-11-02 17:46:07 +00:00
Billy Laws
57a4699bd1 Add IOCTL trace events 2022-11-02 17:46:07 +00:00
Billy Laws
7861968c05 Fix memory::Buffer move constructor 2022-11-02 17:46:07 +00:00
Billy Laws
ef0ae30667 Implement Maxwell3D texture pool management and view creation
Ontop of the TIC cache from previous code a simple index based lookup has been added which vastly speeds things up by avoding the need to hash the TIC structure every time.
2022-11-02 17:46:07 +00:00
Billy Laws
5542459c75 Use a SpinLock for guest shader code cache trap mutex 2022-11-02 17:46:07 +00:00
Billy Laws
3e12cde4d5 Make active Vulkan pipeline public 2022-11-02 17:46:07 +00:00
Billy Laws
2556966ec5 Don't attach textures to the active cycle in AttachTexture 2022-11-02 17:46:07 +00:00
Billy Laws
7dc3dde815 Introduce support for waiting for submission to FenceCycle
Introducing async record resulted in breaking the assumption that any work submitted through command scheduler would be submitted in order with graphics submits. Since async record now unlocks the texture before it's submitted a seperate mechanism is needed to ensure ordering of submits. This is achieved by building support into fence cycle itself, with a conditional variable that is waited on for submission before any fence waits occur.
2022-11-02 17:46:07 +00:00
Billy Laws
54b85583ae Fix layout transition in Texture::CopyFrom 2022-11-02 17:46:07 +00:00
Billy Laws
0f7c04ffb4 Use target format bpb when calculating linear mip level size 2022-11-02 17:46:07 +00:00
Billy Laws
849184452c Add function to check if any guest texture mappings are unmapped 2022-11-02 17:46:07 +00:00
Billy Laws
98cb94ca6c Bind an empty uniform buffer in place of unbound constant buffers 2022-11-02 17:46:07 +00:00
Billy Laws
55b85d0691 Implement combined image samplers and make descriptor code common between quick/normal updates 2022-11-02 17:46:07 +00:00
Billy Laws
ccf2d59351 Fixup input rate reading from packed pipeline state 2022-11-02 17:46:07 +00:00
Billy Laws
13970a5644 Refresh pipeline cached storage buffer bindings after each execution 2022-11-02 17:46:07 +00:00
Billy Laws
6bb2853ca0 Keep track of combined image samplers for quick bind 2022-11-02 17:46:07 +00:00
Billy Laws
040db37a28 Fix descriptor copies to be one per descriptor type 2022-11-02 17:46:07 +00:00
Billy Laws
d482e0ea98 Fix shader stage iteration to not miss the pixel stage 2022-11-02 17:46:07 +00:00
Billy Laws
0b808cc22b Use Sint/Uint attribute type in place of Sscaled/Uscaled
Scaled formats are not supported by any mobile GPUs.
2022-11-02 17:46:07 +00:00
Billy Laws
92ce220d3a Ignore constant buffer selector size for updates
Clustertruck performs a giant (0x3000 byte) update with a selector size of only 0x500, without this the update would only partially go through
2022-11-02 17:46:07 +00:00
Billy Laws
33f16ca26e Handle unmapped blocks in CachedMappedBufferView 2022-11-02 17:46:07 +00:00
Billy Laws
5c0e4a839d Fix SW BC2 decoding pitch 2022-11-02 17:46:07 +00:00
Billy Laws
60863fa162 Make viewports fallback to viewport 0 when their dimensions are invalid
OpenGL games use a zero width/height for viewports 1-15 when they're unused
2022-11-02 17:46:07 +00:00
Billy Laws
586f872655 Update indexed quad conversion for new API 2022-11-02 17:46:07 +00:00
Billy Laws
498b4966d3 Avoid crashing on unmapped buffers
Just log a warning instead
2022-11-02 17:46:07 +00:00
Billy Laws
9f2b20443b Implement Maxwell3D draws 2022-11-02 17:46:07 +00:00
Billy Laws
5020478ace Always rebind pipeline if it has changed from the previous draw 2022-11-02 17:46:07 +00:00
Billy Laws
af1b4ca4f8 Skip clears if attachments are invalid 2022-11-02 17:46:07 +00:00
Billy Laws
01a5e95ce1 Implement non-indexed quad conversion support in new GPU
Uses memory::Buffer directly to allow for cleaner recreation and allow for removing host-only buffers from Buffer in the future.
2022-11-02 17:46:07 +00:00
Billy Laws
d42814bdc1 Add callback for when non-Maxwell3D engines alter pipeline state
This is required so that Maxwell3D knows to restore all dynamic state before the next draw
2022-11-02 17:46:07 +00:00
Billy Laws
7133c5d6b3 Drop exclusiveSubpass in favour of only ending RPs when attachments change
Significantly helps Mali performance by avoiding creating an excessive number of RPs.
2022-11-02 17:46:07 +00:00
Billy Laws
a3f38c0cf7 Add perfetto tracepoints to async record 2022-11-02 17:46:07 +00:00
Billy Laws
6dfef095e8 Up default record slot count to 6 2022-11-02 17:46:07 +00:00
Billy Laws
7d3a117a6f Use a spinlock for descriptor set allocator 2022-11-02 17:46:07 +00:00
Billy Laws
78ddd03d1f Mark newly allocated descriptor slots as active 2022-11-02 17:46:07 +00:00
Billy Laws
0867c593be Support binding pipelines in state updater 2022-11-02 17:46:07 +00:00
Billy Laws
f42a0df72c Use ctSelect register for colour targets and impl null color attachments
Some games have invalid colour attachments sandwiched inbetween valid ones, these need to be skipped in Vulkan with UNUSED_ATTACHMENT
2022-11-02 17:46:07 +00:00
Billy Laws
38aad21d29 Share single flag variable for Maxwell3D batch draw/constant buffer update
Slightly cheaper
2022-11-02 17:46:07 +00:00
Billy Laws
bd7eee8e2b Optimise GPFIFO command processing
GPFIFO code is very high throughput due to the sheer number of commands used for rendering. Adjust some types and switch to a if statement with hints to slightly increase processing speed.
2022-11-02 17:46:07 +00:00
Billy Laws
2cdf6c1fe6 Add branch hint attributes to a couple branches 2022-11-02 17:46:07 +00:00
Billy Laws
b310b99bdc Handle unmapped ranges in TranslateRange 2022-11-02 17:46:07 +00:00
Billy Laws
acfa58ea8c State uopdater MERGEBACK desC 2022-11-02 17:46:07 +00:00
Billy Laws
68b6b20f78 Bunch of cleanup/bugfixes for initial pipeline desc set impl
Cleans up code slightly, fixes keeping track of per-stage descriptor counts and fixes storage buffer caching + more random stuff.
2022-11-02 17:46:07 +00:00
Billy Laws
eb4a9bab11 Mark freshly allocated descriptor slots as active 2022-11-02 17:46:07 +00:00
Billy Laws
f3184cdff1 Reorder active descriptor set slots to end of list
Speeds up allocations by significantly reducing the number of needed atomic ops per alloc. Could be optimised further in the future if needed.
2022-11-02 17:46:07 +00:00
Billy Laws
128b68d8b2 Avoid resetting command buffers manually it's implicit
Somewhat costly on adreno, and with the one time submit flag the reset is implicit for the next beginCommandBuffers call.
2022-11-02 17:46:07 +00:00
Billy Laws
e1717ed811 Implement Maxwell samplers 2022-11-02 17:46:07 +00:00
Billy Laws
f1600f5ad0 Support allocating into spans in the linear allocator 2022-11-02 17:46:07 +00:00
Billy Laws
04cea9239f Implement descriptor set updating through StateUpdater
Will automatically resolve buffer views at record-time into descriptor write/copy structures then apply the write and bind the set.
2022-11-02 17:46:07 +00:00
Billy Laws
d174ca950b Revert "Reset executor command buffers asynchronously"
This reverts commit fc7956df4ff56fdb2afc4b2bb0bbca82196179ca.
2022-11-02 17:46:07 +00:00
Billy Laws
2bbe975ea7 Reset executor command buffers asynchronously
This took a little while to do on qcom drivers, moving it to the cycle waiter thread gives a tiny speedup.
2022-11-02 17:46:07 +00:00
Billy Laws
054d32567d Allow mutation of input data by callback in CircularQueue::AppendTranform 2022-11-02 17:46:07 +00:00
Billy Laws
7c9212743c Implement asynchronous command recording
Recording of command nodes into Vulkan command buffers is very easily parallelisable as it can effectively be treated as part of the GPU execution, which is inherently async. By moving it to a seperate thread we can shave off about 20% of GPFIFO execution time. It should be noted that the command scheduler command buffer infra is no longer used, since we need to record texture updates on the GPFIFO thread (while another slot is being recorded on the record thread) and then use the same command buffer on the record thread later. This ends up requiring a pool per slot, which is reasonable considering we only have four slots by default.
2022-11-02 17:46:07 +00:00
Billy Laws
a197dd2b28 Allow for creating signalled fence cycles 2022-11-02 17:46:07 +00:00
Billy Laws
542651232b Add a mutex to allow preventing buffer recreation 2022-11-02 17:46:07 +00:00
Billy Laws
379b4f163d Implement popping from CircularQueue 2022-11-02 17:46:07 +00:00
Billy Laws
6d9dc9c6fb Implement some more of Draw
Now performs VK state updates on execution and takes advantage of quick descriptor binding.
2022-11-02 17:46:07 +00:00
Billy Laws
3b26f4f48a Expose way to check inter-pipeline descriptor compatibility
Allows users to skip/use quick descriptor sync even after switching pipelines.
2022-11-02 17:46:07 +00:00
Billy Laws
943a38e168 Implement StateUpdater for rapid recording of VK state updates
Using command executor for each state individual update was found to be infeasible due to the shear number of state updates per draw and it relying on per-node heap allocations. Instead this commit takes advantage of each state update being used only once to implement a system of linearly-allocated state update commands that are linked together. After setting up all draw state with StateUpdateBuilder, the built StateUpdater can then be used in the execution phase to record all of the draw state into the command buffer with almost zero ovehead.
2022-11-02 17:46:07 +00:00
Billy Laws
7b4da52445 Add a fast binding sync path for when only one cbuf has changed
SMO implements instanced draws by repeating the same draw just with a different constant buffer bound. Reduce the cost of this significantly by detecting such cases and instead of processing every descriptor, copy the previous descriptor set and update only the ones affected by the bound constant buffer.

Credits to ripinperiperi for the initial idea and making me aware of how SMO does these draws
2022-11-02 17:46:07 +00:00
Billy Laws
89edd9b303 Reset megabuffer binding for disabled vertex buffers 2022-11-02 17:46:07 +00:00
Billy Laws
6a1615a104 Expose color and depth attachments to Draw 2022-11-02 17:46:07 +00:00
Billy Laws
aae957819e Simplify BufferView locking by requiring buffer manager be locked
Avoids the need to repeat the lookup after locking since recreations are impossible if buffer manager is locked.
2022-11-02 17:46:07 +00:00
Billy Laws
9449b52f36 Reduce minimum megabuffer alignment to 128 bytes 2022-11-02 17:46:07 +00:00
Billy Laws
b3cf9c40ba Update megabuffer execution/sequence numbers after updating an allocation 2022-11-02 17:46:07 +00:00
Billy Laws
4b2b6fc6e9 Avoid calling SynchronizeGuest when attempting to megabuffer unless necessary 2022-11-02 17:46:07 +00:00
Billy Laws
e5919e84a1 Pipeline state if statment cleanups 2022-11-02 17:46:07 +00:00
Billy Laws
bf536aa168 Sync pipeline descriptors every draw 2022-11-02 17:46:07 +00:00
Billy Laws
9223d7f524 Fix descriptor initialisation order
They need to be setup before the pipeline is created to avoid passing in garbage data.
2022-11-02 17:46:07 +00:00
Billy Laws
4652cc5a0a Avoid parsing descriptors for disabled shader stages 2022-11-02 17:46:07 +00:00
Billy Laws
3456fb39fa Fix pipeline to shader stage conversion when filling in shader infos
The two vertex pipeline stages need to be both treated as a single stage, and all subsequent stages need to be offset by -1
2022-11-02 17:46:07 +00:00
Billy Laws
a9213debc7 Implement constant buffer reading 2022-11-02 17:46:07 +00:00
Billy Laws
afcfe8a7fa Don't update scissor state >0 unless multiview is supported 2022-11-02 17:46:07 +00:00
Billy Laws
55d77b7eb0 Update user code for new megabuffering 2022-11-02 17:46:07 +00:00
Billy Laws
cc776ae395 Keep track of an 'execution number' in CommandExecutor
Allows users to efficiently cache resources that are valid for only one execution without resorting to callbacks.
2022-11-02 17:46:07 +00:00
Billy Laws
99a34df4cc Avoid trapping frequently synced buffers by using megabuffer copies
When a buffer is trapped nearly every frame, the cost of trapping and synchronising its contents starts to quickly add up. By always using the megabuffer when this is the case, since megabuffer copies are done directly from the guest, we skip the need to synchronise/trap the backing.
2022-11-02 17:46:07 +00:00
Billy Laws
a24aec03a6 Rework per-view megabuffering to cache allocs in the buffer itself
The original intention was to cache on the user side, but especially with shader constant buffers that's difficult and costly. Instead we can cache on the buffer side, with a page-table like structure to hold variable sized allocations indexed by the aligned view base address. This avoids most redundant copies from repeated use of the same buffer without updates inbetween.
2022-11-02 17:46:07 +00:00
Billy Laws
b810470601 Invalidate HLE macro state on macro updates 2022-11-02 17:46:07 +00:00
Billy Laws
2360ca24da Implement constant buffer and storage buffer pipeline descriptor types 2022-11-02 17:46:07 +00:00
Billy Laws
25255b01c7 Keep track of more pipeline descriptor information
This is needed in order to allow allocating per-draw descriptor arrays without recounting their lengths each time
2022-11-02 17:46:07 +00:00
Billy Laws
ad0275dbef Expose active pipeline for access by Maxwell3D class 2022-11-02 17:46:07 +00:00
Billy Laws
6e22373b59 Add array support to AllocateUntracked 2022-11-02 17:46:07 +00:00
Billy Laws
388cff3353 Implement simple pipeline transition cache
Avoids the need to hash PipelineState when we can guess the pipeline that will be used next. This could very easily be optimised in the future with generational, usage-based caching if necessary.
2022-11-02 17:46:07 +00:00
Billy Laws
302b2fcc3f Force flush when dirty refresh returns true 2022-11-02 17:46:07 +00:00
Billy Laws
ec4ea5c5d7 Supply dispatcher manually for shader creation 2022-11-02 17:46:07 +00:00
Billy Laws
3404a3abdb Implement macro HLE for instanced draw macros
gm20b performs instanced draws by repeating draw methods for each instance, the code to detect this together with the cost of interpreting macros took up around 6% of GPFIFO time in Metro Kingdom. By detecting these specific macros and performing an instanced draw directly much of that cost can be avoided.
2022-11-02 17:46:07 +00:00
Billy Laws
cf0752f937 Use NCE memory tracking for guest shaders
Prevents needing to hash them for every single pipeline state update, without this just hashing shaders takes up a significant amount of time.
2022-11-02 17:46:07 +00:00
Billy Laws
19a75c3f65 Bind all pipeline states to main pipeline dirty state 2022-11-02 17:46:07 +00:00
Billy Laws
a04d8fb5cf Setup minimal viewport Vulkan pipeline state 2022-11-02 17:46:07 +00:00
Billy Laws
fe51db366b Mark all dirty resources as dirty initially 2022-11-02 17:46:07 +00:00
Billy Laws
abfa5929f1 Treat vertex buffers with base addr 0 as disabled 2022-11-02 17:46:07 +00:00
Billy Laws
e71ca05f19 Avoid bitfields for signed enum types in PackedPipelineState 2022-11-02 17:46:07 +00:00
Billy Laws
2f2b615780 Add dynamic state support to VK graphics pipeline cache 2022-11-02 17:46:07 +00:00
Billy Laws
ad045058ee Cleanup some redundant comments and includes in pipeline state 2022-11-02 17:46:07 +00:00
Billy Laws
9b05c9c0c3 Introduce a pipeline manager and partial pipeline object
gpu-new will use a monolithic pipeline object for each pipeline to store state, keyed by the PackedPipelineState contents. This allows for a greater level of per-pipeline optimisations and a reduction in the overall number of lookups in a draw compared to the previous system.
2022-11-02 17:46:07 +00:00
Billy Laws
2c1b40c9a8 Add some missed pipeline state members used by Hades 2022-11-02 17:46:07 +00:00
Billy Laws
0c6fa22c6b Introduce pipeline shader stage state
Simple state that generates a hash that can be used in the packed state and spans over the binary for pipeline creation.
2022-11-02 17:46:07 +00:00
Billy Laws
a6bb716123 Move packed pipeline state to a seperate file 2022-11-02 17:46:07 +00:00
Billy Laws
4dcbf5c3a0 Drop the caching aspect of shader manager entirely
Caching here was deemed unnecessary since it will be done implicitly by the pipeline cache and creates issues with the legacy attribute conversion pass. It now purely serves as a frontend for Hades.
2022-11-02 17:46:07 +00:00
Billy Laws
e77e4891dc Tidy up Maxwell 3D regs a bit 2022-11-02 17:46:07 +00:00
Billy Laws
405d26fc22 Introduce Maxwell 3D shader state
Simple state holder that hashes the stored shader and reads it into a buffer
2022-11-02 17:46:07 +00:00
Billy Laws
6865f0bdaf Transition color blend state to packed pipeline state 2022-11-02 17:46:07 +00:00
Billy Laws
7049a521d2 Use Vulkan types directly in PackedPipelineState where possible 2022-11-02 17:46:07 +00:00
Billy Laws
effeb074b6 Move pipeline cache Key to cpp file and rename to PackedPipelineState
The new name is more indicative of what the struct contains and how it will be used as more than just a key.
2022-11-02 17:46:07 +00:00
Billy Laws
e1512c91a0 Transition depth stencil state to pipeline cache key 2022-11-02 17:46:07 +00:00
Billy Laws
1f844e2c18 Transition rasterization state to pipeline cache key 2022-11-02 17:46:07 +00:00
Billy Laws
9a6efb091c Transition tessellation state to pipeline cache key
Also adds dirty tracking and removes it from direct state while we're at it. Since we no longer use Vulkan structs directly there's no benefit to it.
2022-11-02 17:46:07 +00:00
Billy Laws
ae5d419586 Transition input assembly state to pipeline cache key 2022-11-02 17:46:07 +00:00
Billy Laws
3f9161fb74 Transition vertex input state to pipeline cache key
Also adds dirty tracking and removes it from direct state while we're at it. Since we no longer use Vulkan structs directly there's no benefit to it.
2022-11-02 17:46:07 +00:00
Billy Laws
ffe24aa075 Introduce a base pipeline cache key starting with RTs
It was determined that a general purpose Vulkan pipeline cache isn't viable for the significant performance reqs of Draw(), by using a Maxwell 3D specific key we can shrink state significantly more than if we used Vulkan structs.
2022-11-02 17:46:07 +00:00