low_latency_layer - vulkan layer for reducing latency

Age	Commit message (Expand)	Author	Files
10 hours	LowLatency2: Avoid fast path when monitor is workingHEAD main	Nicolas James	2
29 hours	LowLatency2: Check semaphore value before signalling	Nicolas James	6
2 days	Don't discard signal semaphore on missing swapchain monitor without signallin...	Nicolas James	3
5 days	Signal all semaphores on swapchain_monitor teardown	Nicolas James	1
5 days	Guard against bad vkQueueSubmit before notifying contexts	Nicolas James	1
5 days	Don't take exclusive lock when collecting work for LowLatency2	Nicolas James	1
5 days	Removed unused import	Nicolas James	1
5 days	Reduce thread contention by introducing an atomic time point class - greatly ...	Nicolas James	6
5 days	Fix clearing outstanding semaphores instead of outstanding work for LowLatency2	Nicolas James	1
5 days	Fix refactor latency regression for VK_NV_LowLatency2 by checking if work has...	Nicolas James	8
6 days	Refactor storing submissions into FrameSpan class, reduce AntiLag thread cont...	Nicolas James	16
6 days	Check for Vulkan 1.1 via not null GetPhysicalDeviceQueueFamilyProperties2KHR	Nicolas James	1
6 days	Notify swapchain deletion before Vulkan deletes it	Nicolas James	1
6 days	Guard against nullptr instance in GIPA	Nicolas James	1
6 days	Fix QueueContext build error	Nicolas James	1
6 days	Only create query pool when we support doing anything	Nicolas James	1
6 days	Add ONE_TIME_SUBMIT_BIT to timestamp flags	Nicolas James	1
6 days	Remove unnecessary assertions for unique shared pointers in destructors of In...	Nicolas James	2
6 days	Fix validation warning when calling GetPhysicalDeviceQueueFamilyProperties2KH...	Nicolas James	3
6 days	Remove dead code, upgrade env bools to const	Nicolas James	4
6 days	Fix potential race in DestroyInstance and DestroyDevice	Nicolas James	1
7 days	Fix rare race when timestamps are returned to the pool and reused before thei...	Nicolas James	6
7 days	Fix race during teardown, don't enable PresentID - it's not a requirement	Nicolas James	2
7 days	Don't fail hard when the swapchain is erased and we can't find it	Nicolas James	1
7 days	Add refactored VK_NV_low_latency2 impl, (fixes many threading issues)	Nicolas James	8
7 days	Add WIP refactored reflex impl	Nicolas James	16
8 days	Implement frame limit for AL2 strategy	Nicolas James	2
8 days	Fix race in VkQueueSubmit and VkQueueSubmit2	Nicolas James	1
8 days	Implement refactored AL2, todo frame limit	Nicolas James	14
8 days	Add boilerplate for separate implementations	Nicolas James	17
8 days	Nuke old implementation, silence warnings	Nicolas James	9
8 days	Use plain std::arrays in TimestampPool::Handle's getters - avoid wrapper class	Nicolas James	1
9 days	Signal oldest semaphore instead of newest for Reflex	Nicolas James	1
9 days	Fix device_clock::ticks_to_time for negative time deltas	Nicolas James	1
9 days	Fix duplicate mutex shadowing 'real' mutex in base class	Nicolas James	1
9 days	Fix swallowing of result in QueuePresentKHR	Nicolas James	1
9 days	Fix bad return type for GetPhysicalDeviceSurfaceCapabitilies2KHR	Nicolas James	1
9 days	Fix potential duplicate extension insertion in EnumerateDeviceExtensionProper...	Nicolas James	1
9 days	Ensure swapchain monitor signals outstanding semaphores on shutdown	Nicolas James	1
9 days	Fix nullptr pName in GetInstanceProcAddr	Nicolas James	2
9 days	Fix erroneous microsecond conversion to milliseconds	Nicolas James	5
9 days	Relax VkQueuePresentKHR error condition to not include VK_SUBOPTIMAL_KHR	Nicolas James	1
9 days	Fix instances not tracking physical device children	Nicolas James	1
9 days	Fix division by zero in AntiLagUpdateAMD	Nicolas James	1
9 days	Lock in swapchain_monitor::update_params, fix missing CMakeList.txt modification	Nicolas James	6
9 days	Fix CmdWriteTimpstamps bit forward, re-add 'no-unused*' compiler warnings	Nicolas James	3
11 days	Fix bullet point rendering in README.md	Nicolas James	1
11 days	Update README.md, add benchmarks	Nicolas James	1
12 days	Fix potential for early semaphore submission with Reflex	Nicolas James	3
12 days	Fix AntiLagUpdateAMD race, reduce thread contention in both paths	Nicolas James	2