Fix skipping frames when deinterlace ring buffer is present with sync…

[caner-milko]

… multi out

This is caused by not submitting for the skipped frame in deinterlacing ring buffer's execute.

[Copilot]

Pull Request Overview

This PR fixes skipped-frame handling in the deinterlacing ring buffer by adding a SkipExecute path and wiring it through the ring node, and bumps the plugin version.

• Bump plugin version to 3.10.8
• Add SkipExecute to ResourceInterface and implement it in GPU/CPU resource classes
• Wire SkipExecuteRingNode into DeinterlacedBufferRingNode for non-deinterlaced fields

Reviewed Changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 1 comment.

File	Description
Plugins/nosUtilities/Utilities.noscfg	Bumped version from 3.10.7 to 3.10.8
Plugins/nosUtilities/Source/Ring.h	Added SkipExecute API, implemented in resources, and introduced SkipExecuteRingNode
Plugins/nosUtilities/Source/DeinterlacedBufferRing.cpp	Updated else branch to call SkipExecuteRingNode instead of only scheduling

Comments suppressed due to low confidence (3)

Plugins/nosUtilities/Source/DeinterlacedBufferRing.cpp:85

• [nitpick] Indentation for SendScheduleRequest(1) is inconsistent with the newly introduced braces. Consider aligning its indentation within the block for clarity.

SendScheduleRequest(1);

Plugins/nosUtilities/Source/Ring.h:58

• SkipExecute does not accept a ringExecuteName parameter, so resource implementations cannot use the actual ring name for logging or profiling. Consider updating its signature to pass ringExecuteName through.

virtual nosResult SkipExecute(nosNodeExecuteParams* params) { return NOS_RESULT_SUCCESS; }

Plugins/nosUtilities/Source/DeinterlacedBufferRing.cpp:84

• The new SkipExecute path and its behavior for deinterlacing should be covered by unit or integration tests to ensure skipped frames are handled correctly.

res = SkipExecuteRingNode(params, NOS_NAME_STATIC("DeinterlacedBufferRing"));

[Copilot]

[nitpick] The SkipExecute implementations in GPUBufferResource and GPUTextureResource duplicate the BeginCmd/EndCmd logic. Consider extracting this into a shared helper to reduce duplication.

Suggested change

	// Force submit to ensure passes run correctly
	vkss::EndCmd(vkss::BeginCmd(NOS_NAME("SkipExecute"), executeParams->NodeId), true, nullptr);
	return NOS_RESULT_SUCCESS;
	}
	// Use shared helper to perform SkipExecute logic
	ExecuteSkipCommand(NOS_NAME("SkipExecute"), executeParams->NodeId);
	return NOS_RESULT_SUCCESS;
	}
	static void ExecuteSkipCommand(const char* commandName, uint32_t nodeId)
	{
	// Force submit to ensure passes run correctly
	vkss::EndCmd(vkss::BeginCmd(commandName, nodeId), true, nullptr);
	}