Bump pyasn1 to 0.6.*

[Avasam]

Closes #11664

None of these change were detected by stubtest because:

• Names are omitted from __all__ (see comment below)
• RealEncoder.supportIndefLenMode: bool being a subclass of int, this technically passed.

[AlexWaygood]

Should we be adding symbols if they're not included in __all__? The reason why stubtest doesn't complain about those being missing is that it assumes stub authors won't want to add things to stubs that aren't public API. And if it's missing from __all__, it's been explicitly marked as not being public API.

(If there's a specific reason to add any of these symbols, or we think they were just accidentally omitted from __all__, then it's obviously fine to add them — I just don't want us doing it just for "completeness"!)

[Avasam]

Maybe? but as far as I can tell, nearly all symbols from pyasn1 are currently exposed, so I went with established precedent for this stub.
These were really not complex to add (a quick merge from stubgen), I don't mind reverting it and only keeping the int --> bool change.

As a sidenote, these stubs are really simple and quite incomplete as far as annotations go. 6.0 just dropped support for Python 2.7 and 3.7. If the pyasn1 maintainers are open to it, it shouldn't be too hard to merge these stubs upstream.

[AlexWaygood]

Thanks! I don't feel too strongly about the symbols that aren't included in __all__, but I do think it's good if we think about why we're adding symbols marked as private API whenever we do so (it's much harder to remove something from a stub than it is to add it, due to backwards compatibility!)

[github-actions]

According to mypy_primer, this change has no effect on the checked open source code. 🤖🎉