-* Emitcontext
-
- Do we really need to instanciate this variable all the time?
-
- It could be static for all we care, and just use it for making
- sure that there are no recursive invocations on it.
+TODO:
-* Static-ization
+ 1. Create a "partial" emit context for each TypeContainer..
- Since AppDomain exists, maybe we can get rid of all the stuff
- that is part of the `compiler instance' and just use globals
- everywhere.
+ 2. EmitContext should be partially constructed. No IL Generator.
-* FindMembers
+ Optimize the internal type expressions in the parser (system_object, etc)
+ to be something other than DecomposeQI-based expressions.
- Move our utility FindMembers from TypeContainer to Decl, because interfaces
- are also scanned with it.
+ interface_type review.
-* Visibility
+ parameter_array, line 952: `note: must be a single dimension array type'. Validate this
- I am not reporting errors on visibility yet.
+Dead Code Elimination bugs:
+---------------------------
-* Enumerations
+ I should also resolve all the children expressions in Switch, Fixed, Using.
- Currently I am not resolving enumerations.
+Major tasks:
+------------
- Either I track them with `RecordEnum' as I do with classes,
- structs and interfaces or I rewrite the code to visit type
- containers and `walk' the enums with this process.
+ Pinned and volatile require type modifiers that can not be encoded
+ with Reflection.Emit.
-* Known problems:
+ Properties and 17.6.3: Finish it.
- Cast expressions
+ Implement base indexer access.
- They should should use:
+readonly variables and ref/out
+
+BUGS
+----
- OPEN_PARENS type CLOSE_PARENS
+* We suck at reporting what turns out to be error -6. Use the standard error message
+ instead.
- instead of the current production which is wrong, because it
- only handles a few cases.
+* Explicit indexer implementation is missing.
- Complex casts like:
-
- Array r = (string []) object
-
- Wont be parsed.
-
-* Interfaces
-
- For indexers, the output of ix2.cs is different from our
- compiler and theirs. They use a DefaultMemberAttribute, which
- I have yet to figure out:
-
- .class interface private abstract auto ansi INTERFACE
- {
- .custom instance void [mscorlib]System.Reflection.DefaultMemberAttribute::.ctor(string)
- = ( 01 00 04 49 74 65 6D 00 00 ) // ...Item..
- ...
- }
+* Check for Final when overriding, if the parent is Final, then we cant
+ allow an override.
* Interface indexers
Explanation: The reason for the `instance' attribute on
indexers is that indexers only apply to instances
-* Constructors
+* Indexer bugs:
- Currently it calls the parent constructor before initializing fields.
- It should do it the other way around.
+ the following wont work:
-* Use of EmitBranchable
+ x [0] = x [1] = N
- Currently I use brfalse/brtrue in the code for statements, instead of
- using the EmitBranchable function that lives in Binary
+ if x has indexers, the value of x [N] set is set to void. This needs to be
+ fixed.
-* Create an UnimplementedExpcetion
+* Array declarations
- And use that instead of plain Exceptions to flag compiler errors.
+ Multi-dim arrays are declared as [,] instead of [0..,0..]
-* ConvertImplicit
+* Break/Continue statements
- Currently ConvertImplicit will not catch things like:
+ A finally block should reset the InLoop/LoopBegin/LoopEnd, as
+ they are logically outside the scope of the loop.
- - IntLiteral in a float context to generate a -FloatLiteral.
- Instead it will perform an integer load followed by a conversion.
+* Break/continue part 2.
-* In class.cs: Method.Define
+ They should transfer control to the finally block if inside a try/catch
+ block.
- Need to use FindMembers to lookup the member for reporting
- whether a new is needed or not.
+* Method Registration and error CS111
-* virtual-method.cs breaks
+ The way we use the method registration to signal 111 is wrong.
+
+ Method registration should only be used to register methodbuilders,
+ we need an alternate method of checking for duplicates.
- It breaks on the call to: new B ();
+* We need to catch:
- Where B is a class defined in the source code, my guess is that
- the look for ".ctor" fails
+ extern string Property {
+ get { }
+ }
-* Foreach on structure returns does not work
+ The get there should only have a semicolon
+
+*
+> // CSC sets beforefieldinit
+> class X {
+> // .cctor will be generated by compiler
+> public static readonly object O = new System.Object ();
+> public static void Main () {}
+> }
+>
- I am generating invalid code instead of calling ldarga for the
- structure, I am calling ldarg:
+PENDING TASKS
+-------------
- struct X {
- public IEnumerator GetEnumerator ();
- }
+* IMprove error handling here:
- X x;
+ public static Process ()
- foreach (object a in x){
- ...
- }
+ THat assumes that it is a constructor, check if its the same name
+ as the class, if not report a different error than the one we use now.
- I need to get the address of that bad boy
+* Merge test 89 and test-34
-* Using Alias
+* Revisit
- Need to reset the aliases for each compilation unit, so an
- alias defined in a file does not have any effect on another one:
+ Primary-expression, as it has now been split into
+ non-array-creation-expression and array-creation-expression.
+
+* Static flow analysis
- File.cs
- =======
- namespace A {
- using X = Blah;
+ Required to warn about reachability of code and definite
+ assignemt as well as missing returns on functions.
- class Z : X { <-- This X is `Blah'
- }
+* Code cleanup
- File2.cs
- namespace {
- class Y : X { <-- This X Is not `Blah'
- }
- }
+ The information when registering a method in InternalParameters
+ is duplicated, you can always get the types from the InternalParameters
- I think we can implement Aliases by having an `Alias' context in all
- the toplevel TypeContainers of a compilation unit. The children typecontainers
- just chain to the parents to resolve the information.
+* Emit modreq for volatiles
- The driver advances the Alias for each file compiled, so that each file
- has its own alias set.
+ Handle modreq from public apis.
-* Tests
+* Emit `pinned' for pinned local variables.
- Write tests for the various reference conversions. We have
- test for all the numeric conversions.
+ Both `modreq' and pinned will require special hacks in the compiler.
-* Handle destructors specially
+* Make sure that we are pinning the right variable
- Turn ~X () { a () } into:
- void Finalize () { try { a (); } finally { base.Finalize (); } }
+* Maybe track event usage? Currently I am not tracking these, although they
+ are fields.
-* Handle volatile
+* Merge tree.cs, rootcontext.cs
-* Support Re-Throw exceptions:
+OPTIMIZATIONS
+-------------
- try {
- X ();
- } catch (SomeException e){
- LogIt ();
- throw;
- }
+* Implement loop inversion for `For' as well.
-* Remove the tree dumper
+* There is too much unshared code between MemberAccess.Resolve and SimpleName
+ resolve.
+
+* User Defined Conversions is doing way too many calls to do union sets that are not needed
- And make all the stuff which is `public readonly' be private unless
- required.
+* Add test case for destructors
-* Use of lexer.Location in the parser
+* Places that use `Ldelema' are basically places where I will be
+ initializing a value type. I could apply an optimization to
+ disable the implicit local temporary from being created (by using
+ the method in New).
- Currently we do:
+* Dropping TypeContainer as an argument to EmitContext
- TOKEN nt TERMINAL nt TERMINAL nt3 {
- $$ = new Blah ($2, $4, $6, lexer.Location);
- }
+ My theory is that I can get rid of the TypeBuilder completely from
+ the EmitContext, and have typecasts where it is used (from
+ DeclSpace to where it matters).
- This is bad, because the lexer.Location is for the last item in `nt3'
+ The only pending problem is that the code that implements Aliases
+ is on TypeContainer, and probably should go in DeclSpace.
- We need to change that to use this pattern:
+* Casts need to trigger a name resolution against types only.
- TOKEN { $$ = lexer.Location } nt TERMINAL nt TERMINAL nt3 {
- $$ = new Blah ($3, $5, $7, (Location) $2);
- }
+ currently we use a secret hand shake, probably we should use
+ a differen path, and only expressions (memberaccess, simplename)
+ would participate in this protocol.
+
+* Use of local temporary in UnaryMutator
+
+ We should get rid of the Localtemporary there for some cases
+
+* Emitcontext
- Notice how numbering of the arguments changes, as the { $$ =
- lexer.Location } takes up a number
+ Do we really need to instanciate this variable all the time?
-* Method Names
+ It could be static for all we care, and just use it for making
+ sure that there are no recursive invocations on it.
- Method names could be; `IFACE.NAME' in the method declaration,
- stating that they implement a specific interface method.
+* Use of EmitBranchable
- We currently fail to parse it.
+ Currently I use brfalse/brtrue in the code for statements, instead of
+ using the EmitBranchable function that lives in Binary
-* Namespaces
+* ConvertImplicit
- Apparently:
+ Currently ConvertImplicit will not catch things like:
- namespace X {
- }
+ - IntLiteral in a float context to generate a -FloatLiteral.
+ Instead it will perform an integer load followed by a conversion.
- namespace X {
- }
+* Tests
- Is failing to create a single namespace
+ Write tests for the various reference conversions. We have
+ test for all the numeric conversions.
-* Arrays
+* Optimizations
- We need to make sure at *compile time* that the arguments in
- the expression list of an array creation are always positive.
+ In Indexers and Properties, probably support an EmitWithDup
+ That emits the code to call Get and then leaves a this pointer
+ in the stack, so that later a Store can be emitted using that
+ this pointer (consider Property++ or Indexer++)
-* Fix the new parameter mess that we introduced to support ref/outo
+* Optimizations: variable allocation.
-* Reducer and -Literal
+ When local variables of a type are required, we should request
+ the variable and later release it when we are done, so that
+ the same local variable slot can be reused later on.
- Maybe we should never handle -Literal in Unary expressions and let
- the reducer take care of it always?
+* Add a cache for the various GetArrayMethod operations.
-* Implement dead code elimination in statement.cs
+* TypeManager.FindMembers:
- It is pretty simple to implement dead code elimination in
- if/do/while
+ Instead of having hundreds of builder_to_blah hash table, have
+ a single one that maps a TypeBuilder `t' to a set of classes
+ that implement an interface that supports FindMembers.
-* Implement short circuit evaluation.
+* MakeUnionSet Callers
- Currently our and/or operations do not implement short circuit
- evaluation.
+ If the types are the same, there is no need to compute the unionset,
+ we can just use the list from one of the types.
-* Foreach and arrays
+* Factor all the FindMembers in all the FindMembers providers.
- Support foreach (T t in Array)
+* Factor the lookup code for class declarations an interfaces
+ (interface.cs:GetInterfaceByName)
- And optimize to deal with the array rather than getting the enumerator
- out of it.
+RECOMMENDATIONS
+---------------
-* Indexer bugs:
+* Use of lexer.Location in the parser
- the following wont work:
+ Currently we do:
- x [0] = x [1] = N
+ TOKEN nt TERMINAL nt TERMINAL nt3 {
+ $$ = new Blah ($2, $4, $6, lexer.Location);
+ }
- if x has indexers, the value of x [N] set is set to void. This needs to be
- fixed.
+ This is bad, because the lexer.Location is for the last item in `nt3'
-* Preincrement/pre-decrement is not working on array elements.
+ We need to change that to use this pattern:
+
+ TOKEN { oob_stack.Push (lexer.Location) } nt TERMINAL nt TERMINAL nt3 {
+ $$ = new Blah ($3, $5, $7, (Location) oob_stack.Pop ());
+ }
- We are catching this now. But it does not work.
+ Notice how numbering of the arguments changes as the
+ { oob_stack.Push (lexer.Location) } takes a "slot" in the productions.
* local_variable_declaration
Not sure that this grammar is correct, we might have to
resolve this during semantic analysis.
-* Optimizations
+* Idea
- In Indexers and Properties, probably support an EmitWithDup
- That emits the code to call Get and then leaves a this pointer
- in the stack, so that later a Store can be emitted using that
- this pointer (consider Property++ or Indexer++)
+ MethodGroupExpr
+
+ These guys should only appear as part of an Invocation, so we
+ probably can afford to have a special callback:
+
+ Expression.ResolveAllowMemberGroups
+
+ This is only called by Invocation (or anyone that consumes
+ MethodGroupExprs)
+
+ And the regular DoResolve and DoResolveLValue do emit the error
+ 654 `Method referenced without argument list'.
+
+ Otherwise, a resolution will return a MethodGroupExpr which is
+ not guaranteed to have set its `Expression.Type' to a non-null
+ value.