* Stack alloc parsed to spans

* PR feedback

* Document the breaking change

NOTE: we make a good effort to emit the attribute. In some scenarios attribute will not be emitted.
- attribute is missing
- the type is already made Obsolete by the user
- user uses some other kind of deprecation

* Call this.DeclaringCompilation.EnsureReadOnlyAttributeExists() when needed

* Promote EmbeddedAttribute to be used by both compilers

* Generate Embedded attributes on PE Module Builder

* Thread PEModuleBuilder through Symbol.AddSynthesizedAttributes()

* Fix failing tests

* Base case passing

* Refactor ReadOnlyAttribute to IsReadOnlyAttribute

* Fix build break

* More Tests

* Clean up

* PR Comments #1

PR Comments #3

PR Comments #4

PR Comments #5

PR Comments #6

PR Comments # 7

* Handle NoPIA

Added more tests

* Lambdas and Local functions no longer generate attributes in symbols

Added tests for constructors and operators

Signal need for synthesized attributes in local rewriter for lambdas an local functions

Clean up

* Clean up

* More PR Feedback

* Adding more tests

* More PR Comments

* More tests for explicit interface implementations

* Latest PR Comments

* Moved EnsureIsReadOnlyAttributeExists to AfterAddingTypeMembersChecks

* Fix failing tests

A single method could have spans in multiple source files. For example, synthesized constructors of partial types can have spans in multiple files if this type has field initializers in multiple files. Another example is methods with #line directives.

This change changes instrumentation for such methods - the instrumented IL for such methods will now include a call to a new overload of Microsoft.CodeAnalysis.Runtime.Instrumentation.CreatePayload() that accepts an array of file indices as opposed to single file index.

* Writing attribute to metadata

* Reading attribute in PE symbols

* Added more tests for lambdas, delegates, and other types

* Use ReadOnlyAttribute instead

* Fix build break

* Fix Failed Tests

* Enable back disabled tests

* Fix more tests

* Ban usage of ReadOnlyAttribute in source

* Rename System.Runtime.InteropServices.ReadOnlyAttribute to System.Runtime.CompilerServices.ReadOnlyAttribute

* Clean up

* Address CR comments

* More PR feedback

...for dynamic analysis instrumentation.  Guid.Parse is not available when
targeting .NET 2.0/3.5.

[Based on features/async-return with tests ported from that branch.]

For the most part, this change implements the same encoding
for tuples that we use for dynamic. This means that interface
implementations and generic constraints are not covered, but
base types and type parameters are.

This change also doesn't have full tests for "long" tuples (nested tuple
transformations) or tuple types that can only be constructed in
metadata.

Make work on both 4.5 portable and 2.0 desktop. Still need to create a meaningful test and respond to review feedback.

Made tuple ctors WellKnown members
Made ctors optional in the bound tuple creation nodes
ObsoleteAttributeData for tuple wrappers returns null (at least for now)

Rebased onto future branch
Fixed Tuple field signatures
Fixed tests affected by tuple parsing

Since compiler no longer needs to instrument code to call flush method. Remove the check for flush
method when instrumenting code is safe because it is provided as a binary reference.

Two things were preventing the EE from finding the type, which is needed
for XML literal operations in VB:

1. It is marked internal and the accessibility check happened in
AssemblySymbol, rather than the binder, so it was not hijacked.

2. It is marked with EmbeddedAttribute and is not from the SourceModule of
the compilation (it's from the module of the current frame).

We can resolve both of these problems by doing the lookups as though we
were in the module containing the current frame.  This required two
changes:

1. Remove InternalXmlHelper from WellKnownTypes and give it its own code
path.  Explicitly pass the appropriate module symbol to the new helper.

2. In Binder.CheckViability, look for a containing module before falling
back on the SourceModule of the compilation.

By preserving the accessibility checks during type lookup (but doing them
from the perspective of a different assembly), we retain the compiler's
protection against pulling in a copy from each VB reference (i.e.
ambiguity).

The main caveat is that My.InternalXmlHelper will not be found if it was
not embedded in the module containing the current frame - even if it is
unambiguously available in a referenced assembly.

Breaking change: unlike dev12, roslyn will only find My.InternalXmlHelper
in the source module.

Fixes #964

Two things were preventing the EE from finding the type, which is needed
for XML literal operations in VB:

1. It is marked internal and the accessibility check happened in
AssemblySymbol, rather than the binder, so it was not hijacked.

2. It is marked with EmbeddedAttribute and is not from the SourceModule of
the compilation (it's from the module of the current frame).

We can resolve both of these problems by doing the lookups as though we
were in the module containing the current frame.  This required two
changes:

1. Remove InternalXmlHelper from WellKnownTypes and give it its own code
path.  Explicitly pass the appropriate module symbol to the new helper.

2. In Binder.CheckViability, look for a containing module before falling
back on the SourceModule of the compilation.

By preserving the accessibility checks during type lookup (but doing them
from the perspective of a different assembly), we retain the compiler's
protection against pulling in a copy from each VB reference (i.e.
ambiguity).

The main caveat is that My.InternalXmlHelper will not be found if it was
not embedded in the module containing the current frame - even if it is
unambiguously available in a referenced assembly.

Breaking change: unlike dev12, roslyn will only find My.InternalXmlHelper
in the source module.

Fixes #964

These were accidentally changed to UTF-16LE during a recent scripted change.  This went unnoticed because the encoding change didn't register in CodeFlow (indeed if you view a shelve of this change in CodeFlow it will report no changes).  UTF-16 does not work well with Git as it is often interpreted as a binary file.  Moving back to UTF-8 so our GitHub experience works well for customers.

closes #66 (changeset 1403787)

Background:

As discussed in the dev team all hands all code under the GitHub dotnet foundation is using a single process for contribution, API review, infrastructure and coding style. The idea is to present a unified view to our customer and give them a single story for contributing to any project under the dotnet foundation.

https://github.com/dotnet/corefx/wiki/Contributing#c-coding-style

The coding style transition is automated using a Roslyn based rewrite tool:

https://github.com/dotnet/codeformatter

This will be applied in stages across our developer tree. Right now the focus is on the Open directory as this is what is being presented on github. Code owners will be contacted before the transition happens.

Note: this is a soft style requirement. There are no build errors that come from this change.
 (changeset 1403394)

Anonymous Type.ToString() implementation should use overload of String.Format that takes IFormatProvider. Also, C# implementation should explicitly call Object.ToString on each property value before passing it to the Format function, this ensures compatibility with code generated by native compiler.
***NO_CI***
 (changeset 1386476)

Apply AsyncStateMachineAttribute and IteratorStateMachineAttribute consistently on async/iterator methods in both C# and VB.
Remove DebuggerStepThroughAttribute from these methods - the methods don't contain or call any user code, so this attribute is not needed.

VB: Avoid generating async state machine when emitting metadata-only PE.
VB: Unify implementation of state machine type symbol creation and synthesis of attributes with C#.
 (changeset 1360205)

 (changeset 1343137)

Checking in for MattWar. (changeset 1340558)

Enable fast array literal initialization for enums if the target framework is 4.5 or above. (changeset 1339077)

This change updates the C# compiler to use “Array.Empty<T>()” instead of “new T[0]” when filling an empty params array argument (and when Empty<T> is available). (changeset 1336872)

Mimic native compiler behavior when applying ClassInterfaceAttribute for a "ComClass" (changeset 1321287)

Splits Microsoft.CodeAnalysis, Microsoft.CodeAnalysis.CSharp and Microsoft.CodeAnalysis.VisualBasic into portable and desktop assemblies.

Existing Core\Source, CSharp\Source and VisualBasic\Source directories are renamed to Core\Portable, CSharp\Portable and VisualBasic\Portable.
New sibling Desktop folders are added and non-portable source is moved there.
"System.Runtime" references has to be removed in order for the portable project system magic to automatically add facade references.
 (changeset 1281686)