[wasm][coreCLR] fix ASAN/ASSERTIONS issues

[pavelsavara]

I compiled coreCLR for WASM with maybe? ASAN? flags and started getting "memory out of bounds".
I see memset or memcpy validating target pointer even when size is zero.

For example here when prevCapacity is 0 then from->ppOutBBs is -1.

runtime/src/coreclr/interpreter/compiler.cpp

Line 451 in a2d130a

memcpy(newa, from->ppOutBBs, from->outCount * sizeof(InterpBasicBlock*));

In another place cFields is 0 and bmtMFDescs->ppFieldDescList is -1

runtime/src/coreclr/vm/methodtablebuilder.cpp

Lines 3680 to 3681 in a2d130a

	bmtMFDescs->ppFieldDescList = new (GetStackingAllocator()) FieldDesc*[bmtMetaData->cFields];
	ZeroMemory(bmtMFDescs->ppFieldDescList, bmtMetaData->cFields * sizeof(FieldDesc *));

[dotnet-policy-service]

Tagging subscribers to 'arch-wasm': @lewing, @pavelsavara
See info in area-owners.md if you want to be subscribed.

[dotnet-policy-service]

Tagging subscribers to this area: @mangod9
See info in area-owners.md if you want to be subscribed.

[AaronRobinsonMSFT]

@pavelsavara How can I repro this locally? What commands/flags did you use?

[pavelsavara]

I don't know :) And it took me full clean and restart of the shell to recover from that.
Maybe it was some corrupt state of the build. But the methods that triggered the trap were plausible enough.

Normally emscripten ASAN experience is somewhat difficult.
https://emscripten.org/docs/debugging/Sanitizers.html

I think this is not high priority and maybe red herring.

[pavelsavara]

It's quite possible I compiled with -sASSERTIONS=1

https://github.com/dotnet/runtime/blob/main/src/native/corehost/browserhost/CMakeLists.txt#L68