'Error initializing the symbol reader' error when clrdbg is run from a directory with accents

[gregg-miskelly]

1. Create a new user account with accents in the name (example: Andrés)
2. Follow instructions on http://aka.ms/vscclrdogfood to install the C# extension
3. Try to debug

Result:
No symbols are loaded, and the following is written to the console -
ERROR: Unable to load symbols. Error initializing the symbol reader (Microsoft.DiaSymReader.PortablePdb.dll).

[gregg-miskelly]

@MattGertz FYI, we have a new RC2 bug

[gregg-miskelly]

The problem here is that the CoreCLR Unix hosting APIs are using CP_ACP, while the debugger expects them to use CP_UTF8. Need to talk with the CLR team to find out which one is correct.

https://github.com/dotnet/coreclr/blob/775003a4c72f0acc37eab84628fcef541533ba4e/src/dlls/mscoree/unixinterface.cpp#L127

[MattGertz]

There's a similar bug from another component touching on this issue as well (not sure of the id) -- I'll escalate.

[MattGertz]

See https://github.com/dotnet/cli/issues/2649. Does that address the point?

[jello-chen]

When I use vs code to debug,the console will print 'ERROR: Unable to load symbols. Error initializing the symbol reader (Microsoft.DiaSymReader.PortablePdb.dll).'.I don't know if it is for the reason that my user name contains chinese.

[gregg-miskelly]

Yes, this is the problem then. We hope to have a fix for this shortly. Sorry.

[gregg-miskelly]

@jello-chen what OS are you on?

[jello-chen]

Win 10 Professional

[gregg-miskelly]

Okay, good. Yes, my fix should address this then.