internal/pkgbits, cmd/compile/internal/noder: document string section

To understand this change, we begin with a short description of the UIR
file format.

Every file is a header followed by a series of sections. Each section
has a kind, which determines the type of elements it contains. An
element is just a collection of one or more primitives, as defined by
package pkgbits.

Strings have their own section. Elements in the string section contain
only string primitives. To use a string, elements in other sections
encode a reference to the string section.

To illustrate, consider a simple file which exports nothing at all.

  package p

In the meta section, there is an element representing a package stub.
In that package stub, a string ("p") represents both the path and name
of the package. Again, these are encoded as references.

To manage references, every element begins with a reference table.
Instead of writing the bytes for "p" directly, the package stub encodes
an index in this reference table. At that index, a pair of numbers is
stored, indicating:

  1. which section
  2. which element index within the section

Effectively, elements always use *2* layers of indirection; first to the
reference table, then to the bytes themselves.

With some minor hand-waving, an encoding for the above package is given
below, with (S)ections, (E)lements and (P)rimitives denoted.

+ Header
| + Section Ends                           // each section has 1 element
| | + 1                                    // String is elements [0, 1)
| | + 2                                    // Meta   is elements [1, 2)
| + Element Ends
| | + 1                                    // "p"    is bytes    [0, 1)
| | + 6                                    // stub   is bytes    [1, 6)
+ Payload
| + (S) String
| | + (E) String
| | | + (P) String           { byte } 0x70 // "p"
| + (S) Meta
| | + (E) Package Stub
| | | + Reference Table
| | | | + (P) Entry Count    uvarint  1    // there is a single entry
| | | | + (P) 0th Section    uvarint  0    // to String, 0th section
| | | | + (P) 0th Index      uvarint  0    // to 0th element in String
| | | + Internals
| | | | + (P) Path           uvarint  0    // 0th entry in table
| | | | + (P) Name           uvarint  0    // 0th entry in table

Note that string elements do not have reference tables like other
elements. They behave more like a primitive.

As this is a bit complicated and getting into details of the UIR file
format, we omit some details in the documentation here. The structure
will become clearer as we continue documenting.

Change-Id: I12a5ce9a34251c5358a20f2f2c4d0f9bd497f4d0
Reviewed-on: https://go-review.googlesource.com/c/go/+/671997
Reviewed-by: Robert Griesemer <gri@google.com>
Auto-Submit: Mark Freeman <mark@golang.org>
TryBot-Bypass: Mark Freeman <mark@golang.org>

src/cmd/compile/internal/noder/doc.go

@@ -20,7 +20,7 @@ The payload is a series of sections. Each section has a kind which determines
 its index in the series.
 
 SectionKind = Uint64 .
-Payload     = SectionString  // TODO(markfreeman) Define.
+Payload     = SectionString
               SectionMeta
               SectionPosBase // TODO(markfreeman) Define.
               SectionPkg     // TODO(markfreeman) Define.
@@ -40,6 +40,12 @@ accessed using an index relative to the start of the section.
 // TODO(markfreeman): Rename to SectionIndex.
 RelIndex = Uint64 .
 
+## String Section
+String values are stored as elements in the string section. Elements outside
+the string section access string values by reference.
+
+SectionString = { String } .
+
 ## Meta Section
 The meta section provides fundamental information for a package. It contains
 exactly two elements — a public root and a private root.

src/internal/pkgbits/doc.go

@@ -16,51 +16,62 @@ zvarint = (* a zig-zag encoded signed variable-width integer *) .
 uvarint = (* an unsigned variable-width integer *) .
 
 # Strings
-Strings are not encoded directly. Rather, they are deduplicated during encoding
-and referenced where needed.
+A string is a series of bytes.
 
-String      = [ Sync ] StringRef .
-StringRef   = [ Sync ] Uint64 . // TODO(markfreeman): Document.
+// TODO(markfreeman): Does this need a marker?
+String    = { byte } .
 
-StringSlice = Uint64            // the number of strings in the slice
-              { String }
-              .
+Strings are typically not encoded directly. Rather, they are deduplicated
+during encoding and referenced where needed; this process is called interning.
 
-// TODO(markfreeman) It is awkward to discuss references (and by extension
-// strings and constants). We cannot explain how they resolve without mention
-// of foreign concepts. Ideally, references would be defined in familar terms —
-// perhaps using an index on the byte array.
+StringRef = [ Sync ] Ref[String] .
+
+Note that StringRef is *not* equivalent to Ref[String] due to the extra marker.
+
+# References
+References specify the location of a value. While the representation here is
+fixed, the interpretation of a reference is left to other packages.
+
+Ref[T] = [ Sync ] Uint64 . // points to a value of type T
+
+# Slices
+Slices are a convenience for encoding a series of values of the same type.
+
+// TODO(markfreeman): Does this need a marker?
+Slice[T] = Uint64 // the number of values in the slice
+           { T }  // the values
+           .
 
 # Constants
 Constants appear as defined via the package constant.
 
 Constant = [ Sync ]
-           Bool       // whether the constant is a complex number
-           Scalar     // the real part
-           [ Scalar ] // if complex, the imaginary part
+           Bool        // whether the constant is a complex number
+           Scalar      // the real part
+           [ Scalar ]  // if complex, the imaginary part
            .
 
 A scalar represents a value using one of several potential formats. The exact
 format and interpretation is distinguished by a code preceding the value.
 
 Scalar   = [ Sync ]
-           Uint64     // the code
+           Uint64      // the code indicating the type of Val
            Val
            .
 
 Val      = Bool
          | Int64
-         | String
-         | Term       // big integer
-         | Term Term  // big ratio, numerator / denominator
-         | BigBytes   // big float, precision 512
+         | StringRef
+         | Term        // big integer
+         | Term Term   // big ratio, numerator / denominator
+         | BigBytes    // big float, precision 512
            .
 
 Term     = BigBytes
-           Bool       // whether the term is negative
+           Bool        // whether the term is negative
            .
 
-BigBytes = String .   // bytes of a big value
+BigBytes = StringRef . // bytes of a big value
 
 # Markers
 Markers provide a mechanism for asserting that encoders and decoders are