Support overlay spaces

[tclamb]

Trying to run the Deepcut analyzer on a project using overlay spaces fails with the following stacktrace:

Address are in different spaces demo_sub.ovl != demo_task.ovl
java.lang.IllegalArgumentException: Address are in different spaces demo_sub.ovl != demo_task.ovl
	at ghidra.program.model.address.OverlayAddressSpace.subtract(OverlayAddressSpace.java:80)
	at ghidra.program.model.address.GenericAddress.subtract(GenericAddress.java:117)
	at deepcut.EdgeInfo.<init>(EdgeInfo.java:50)
	at deepcut.FunctionCallGraph.createListOfEdges(FunctionCallGraph.java:117)
	at deepcut.FunctionCallGraph.<init>(FunctionCallGraph.java:77)
	at deepcut.DeepCutAnalyzer.added(DeepCutAnalyzer.java:101)
	at ghidra.app.plugin.core.analysis.OneShotAnalysisCommand.applyTo(OneShotAnalysisCommand.java:46)
	at ghidra.app.plugin.core.analysis.OneShotAnalysisCommand.applyTo(OneShotAnalysisCommand.java:30)
	at ghidra.app.plugin.core.analysis.AutoAnalysisManager$AnalysisTaskWrapper.run(AutoAnalysisManager.java:660)
	at ghidra.app.plugin.core.analysis.AutoAnalysisManager.startAnalysis(AutoAnalysisManager.java:760)
	at ghidra.app.plugin.core.analysis.AutoAnalysisManager.startAnalysis(AutoAnalysisManager.java:639)
	at ghidra.app.plugin.core.analysis.AutoAnalysisManager.startAnalysis(AutoAnalysisManager.java:604)
	at ghidra.app.plugin.core.analysis.AnalysisBackgroundCommand.applyTo(AnalysisBackgroundCommand.java:55)
	at ghidra.app.plugin.core.analysis.AnalysisBackgroundCommand.applyTo(AnalysisBackgroundCommand.java:33)
	at ghidra.framework.plugintool.mgr.BackgroundCommandTask.run(BackgroundCommandTask.java:103)
	at ghidra.framework.plugintool.mgr.ToolTaskManager.run(ToolTaskManager.java:351)
	at java.base/java.lang.Thread.run(Thread.java:1474)

To check if addresses are in the same space, there's Address#hasSameAddressSpace(Address). Physical addresses are available using Address#getPhysicalAddress(), so checks could be avoided like a.getPhysicalAddress().subtract(b.getPhysicalAddress()). That might work here, assuming edges wouldn't span overlapping spaces.

To start, I think assuming translation units won't cross overlay boundaries is fine. For this project, the only way translation units split across address spaces is for .vtables, which all occur in section order at the end of .data.

[tclamb]

I just tried out swapping line 50 of EdgeInfo to use #getPhysicalAddress(), and the analysis completed successfully. A few of the object boundaries are clearly wrong where they're spanning multiple overlapping overlays, but most of the analysis looks correct and will be helpful. Thanks!