Changeset d49bfa8


Ignore:
Timestamp:
Jul 20, 2017, 2:05:31 PM (4 years ago)
Author:
Andrew Beach <ajbeach@…>
Branches:
aaron-thesis, arm-eh, cleanup-dtors, deferred_resn, demangler, jacob/cs343-translation, jenkins-sandbox, master, new-ast, new-ast-unique-expr, new-env, no_list, persistent-indexer, resolv-new, with_gc
Children:
957453d
Parents:
ac10576
Message:

That might be everything from yesterday. Hungry now.

File:
1 edited

Legend:

Unmodified
Added
Removed
  • doc/proposals/virtual.txt

    rac10576 rd49bfa8  
    11Proposal for virtual functionality
     2
     3There are two types of virtual inheritance in this proposal, relaxed
     4(implicate) and strict (explicate). Relaxed is the simpler case that uses the
     5existing trait system with the addition of trait references and vtables.
     6Strict adds some constrants and requires some additional notation but allows
     7for down-casting.
     8
     9Relaxed Virtual Inheritance:
    210
    311Imagine the following code :
     
    2028void draw(line*);
    2129
    22 While all the members of this simple UI support drawing creating a UI that easily
    23 supports both these UI requires some tedious boiler-plate code :
     30While all the members of this simple UI support drawing creating a UI that
     31easily supports both these UI requires some tedious boiler-plate code :
    2432
    2533enum type_t { text, line };
     
    4149}
    4250
    43 While this code will work as indented, adding any new widgets or any new widget behaviors
    44 requires changing existing code to add the desired functionality. To ease this maintenance
    45 effort required CFA introduces the concept of dynamic types, in a manner similar to C++.
    46 
    47 A simple usage of dynamic type with the previous example would look like :
    48 
    49 drawable* objects[10];
     51While this code will work as indented, adding any new widgets or any new
     52widget behaviors requires changing existing code to add the desired
     53functionality. To ease this maintenance effort required CFA introduces the
     54concept of trait references.
     55
     56Using trait references to implement the above gives the following :
     57
     58trait drawable objects[10];
    5059fill_objects(objects);
    5160
    5261while(running) {
    53       for(drawable* object : objects) {
     62      for(drawable object : objects) {
    5463            draw(object);
    5564      }
    5665}
    5766
    58 However, this is not currently do-able in the current CFA and furthermore is not
    59 possible to implement statically. Therefore we need to add a new feature to handle
    60 having dynamic types like this (That is types that are found dynamically not types
    61 that change dynamically).
    62 
    63 C++ uses inheritance and virtual functions to find the
    64 desired type dynamically. CFA takes inspiration from this solution.
    65 
    66 What we really want to do is express the fact that calling draw() on a object
    67 should find the dynamic type of the parameter before calling the routine, much like the
    68 hand written example given above. We can express this by adding the virtual keyword on
    69 the parameter of the constraints on our trait:
     67The keyword trait is optional (by the same rules as the struct keyword). This
     68is not currently suported in CFA and the lookup is not possible to implement
     69statically. Therefore we need to add a new feature to handle having dynamic
     70lookups like this.
     71
     72What we really want to do is express the fact that calling draw() on a trait
     73reference should find the underlying type of the given parameter and find how
     74it implements the routine, as in the example with the enumeration and union.
     75
     76For instance specifying that the drawable trait reference looks up the type
     77of the first argument to find the implemenation would be :
    7078
    7179trait drawable(otype T) {
     
    7381};
    7482
    75 This expresses the idea that drawable is similar to an abstract base class in C++ and
    76 also gives meaning to trying to take a pointer of drawable. That is anything that can
    77 be cast to a drawable pointer has the necessary information to call the draw routine on
    78 that type. Before that drawable was only a abstract type while now it also points to a
    79 piece of storage which specify which behavior the object will have at run time.
    80 
    81 This storage needs to be allocate somewhere. C++ just adds an invisible pointer at
    82 the beginning of the struct but we can do something more explicit for users, actually
    83 have a visible special field :
    84 
    85 struct text {
    86       char* text;
    87       vtable drawable;
    88 };
    89 
    90 struct line{
    91       vtable drawable;
    92       vec2 start;
    93       vec2 end;
    94 };
    95 
    96 With these semantics, adding a "vtable drawable" means that text pointers and line pointers are now
    97 convertible to drawable pointers. This conversion will not necessarily be a type only change however, indeed,
    98 the drawable pointer will point to the field "vtable drawable" not the head of the struct. However, since all
    99 the types are known at compile time, converting pointers becomes a simple offset operations.
    100 
    101 The vtable field contains a pointer to a vtable which contains all the information needed for the caller
    102 to find the function pointer of the desired behavior.
    103 
    104 One of the limitations of this design is that it does not support double dispatching, which
    105 concretely means traits cannot have routines with more than one virtual parameter. This design
    106 would have many ambiguities if it did support multiple virtual parameter. A futher limitation is
    107 that traits over more than one type cannot have vtables meaningfully defined for them, as the
    108 particular vtable to use would be a function of the other type(s) the trait is defined over.
    109 
    110 It is worth noting that the function pointers in these vtables are bound at object construction, rather than
    111 function call-site, as in Cforall's existing polymorphic functions. As such, it is possible that two objects
    112 with the same static type would have a different vtable (consider what happens if draw(line*) is overridden
    113 between the definitions of two line objects). Given that the virtual drawable* erases static types though,
    114 this should not be confusing in practice. A more distressing possibility is that of creating an object that
    115 outlives the scope of one of the functions in its vtable. This is certainly a possible bug, but it is of a
    116 type that C programmers are familiar with, and should be able to avoid by the usual methods.
    117 
    118 Extensibility.
    119 
    120 One of the obvious critics of this implementation is that it lacks extensibility for classes
    121 that cannot be modified (ex: Linux C headers). However this solution can be extended to
    122 allow more extensibility by adding "Fat pointers".
    123 
    124 Indeed, users could already "solve" this issue by writing their own fat pointers as such:
    125 
    126 trait MyContext(otype T) {
    127       void* get_stack(virtual T*)
    128 };
    129 
    130 void* get_stack(ucontext_t *context);
    131 
    132 struct fat_ucontext_t {
    133       vtable MyContext;
    134       ucontext_t *context;
    135 }
    136 
    137 //Tedious forwarding routine
    138 void* get_stack(fat_ucontext_t *ptr) {
    139       return get_stack(ptr->context);
    140 }
    141 
    142 However, users would have to write all the virtual methods they want to override and make
    143 them all simply forward to the existing method that takes the corresponding POCO(Plain Old C Object).
    144 
    145 The alternative we propose is to use language level fat pointers :
    146 
    147 trait MyContext(otype T) {
    148       void* get_stack(virtual T*)
    149 };
    150 
    151 void* get_stack(ucontext_t *context);
    152 
    153 //The type vptr(ucontext_t) all
    154 vptr(ucontext_t) context;
    155 
    156 These behave exactly as the previous example but all the forwarding routines are automatically generated.
    157 
    158 Bikeshedding.
    159 
    160 It may be desirable to add fewer new keywords than discussed in this proposal; it is possible that "virtual"
    161 could replace both "vtable" and "vptr" above with unambiguous contextual meaning. However, for purposes of
    162 clarity in the design discussion it is beneficial to keep the keywords for separate concepts distinct.
    163 
     83This could be implied in simple cases like this one (single parameter on the
     84trait and single generic parameter on the function). In more complex cases it
     85would have to be explicately given, or a strong convention would have to be
     86enforced.
     87
     88Once a function in a trait has been marked as virtual it defines a new
     89function that takes in that trait's reference and then dynamically calls the
     90underlying type implemenation. Hence a trait reference becomes a kind of
     91abstract type, cannot be directly instantiated but can still be used.
     92
     93One of the limitations of this design is that it does not support double
     94dispatching, which concretely means traits cannot have routines with more than
     95one virtual parameter. The program must have a single table to look up the
     96function on. Using trait references with traits with more than one parameter
     97is also restricted, inistially forbidden, see extention.
     98
     99Extention: Multi-parameter Virtual Traits:
     100
     101This implementation can be extented to traits with multiple parameters if
     102one is called out as being the virtual trait. For example :
     103
     104trait iterator(otype T, dtype Item) {
     105        Maybe(Item) next(virtual T *);
     106}
     107
     108iterator(int) generators[10];
     109
     110Which creates a collection of iterators that produce integers, regardless of
     111how those iterators are implemented. This may require a note that this trait
     112is virtual on T and not Item, but noting it on the functions may be enough.
     113
     114
     115Strict Virtual Inheritance:
     116
     117One powerful feature relaxed virtual does not support is the idea of down
     118casting. Once something has been converted into a trait reference there is
     119very little we can do to recover and of the type information, only the trait's
     120required function implementations are kept.
     121
     122To allow down casting strict virtual requires that all traits and structures
     123involved be orginized into a tree. Each trait or struct must have a unique
     124position on this tree (no multiple inheritance).
     125
     126This is declared as follows :
     127
     128trait error(otype T) virtual {
     129        const char * msg(T *);
     130}
     131
     132trait io_error(otype T) virtual error {
     133        FILE * src(T *);
     134}
     135
     136struct eof_error virtual io_error {
     137        FILE * fd;
     138};
     139
     140So the trait error is the head of a new tree and io_error is a child of it.
     141
     142Also the parent trait is implicately part of the assertions of the children,
     143so all children implement the same operations as the parent. By the unique
     144path down the tree, we can also uniquely order them so that prefixs of a
     145child's vtable has the same format as the parent's.
     146
     147This gives us an important extra feature, runtime checking of the parent-child
     148relationship with a C++ dynamic_cast like operation. Allowing checked
     149convertions from trait references to more particular references, which works
     150if the underlying type is, or is a child of, the new trait type.
     151
     152Extention: Multiple Parents
     153
     154Although each trait/struct must have a unique position on each tree, it could
     155have positions on multiple trees. All this requires is the ability to give
     156multiple parents, as here :
     157
     158trait region(otype T) virtual drawable, collider;
     159
     160The restriction being, the parents must come from different trees. This
     161object (and all of its children) can be cast to either tree. This is handled
     162by generating a seperate vtable for each tree the structure is in.
     163
     164Extention: Multi-parameter Strict Virtual
     165
     166If a trait has multiple parameters than one must be called out to be the one
     167we generate seperate vtables for, as in :
     168
     169trait example(otype T, otype U) virtual(T) ...
     170
     171This can generate a seperate vtable for each U for which all the T+U
     172implementations are provided. These are then seperate nodes in the tree (or
     173the root of different trees) as if each was created individually.
     174
     175Example:
     176
     177trait argument(otype T) virtual {
     178        char short_name(virtual T *);
     179        bool is_set(virtual T *);
     180};
     181
     182trait value_argument(otype T, otype U) virtual(T) argument {
     183        U get_value(virtual T *);
     184};
     185
     186Extention: Structural Inheritance
     187
     188Currently traits must be the internal nodes and structs the leaf nodes.
     189Structs could be made internal nodes as well, and might specify structural
     190inheritance.
     191
     192
     193Storing the Virtual Lookup Table (vtable):
     194
     195We have so fare been silent on how the vtable is created, stored and accessed.
     196
     197Creation happens at compile time. Function pointers are found by using the
     198same best match rules as elsewhere (additional rules for defaults from the
     199parent may or may not be reqired). For strict virtual this must happen at the
     200global scope and forbiding static functions, to ensure that a single unique
     201vtable is created. Similarly, there may have to be stricter matching rules
     202for the functions that go into the vtable, possibly requiring an exact match.
     203Relaxed virtual could relax both restrictions, if we allow different vtable
     204at different conversion (struct to trait reference) sites. If it is allowed
     205local functions being bound to a vtable could cause issues when they go out
     206of scope, however this should follow the lifetime rules most C programs
     207already follow implicately.
     208
     209Most vtables should be stored statically, the only exception being some of
     210the relaxed vtables that could have local function pointers. These may be able
     211to be stack allocated. All vtables should be immutable and require no manual
     212cleanup.
     213
     214Access has two main options:
     215
     216The first is through the use of fat pointers, or a tuple of pointers. When the
     217object is converted to a trait reference, the pointers to its vtables are
     218stored along side it.
     219
     220This allows for compatability with existing structures (such as those imported
     221from C) and is the default storage method unless a different one is given.
     222
     223The other is by inlining the vtable pointer as "intrusive vtables". This adds
     224a field to the structure to the vtable. The trait reference than has a single
     225pointer to this field, the vtable includes an offset to find the beginning of
     226the structure again.
     227
     228This is used if you specify a vtable field in the structure. If given in the
     229trait the vtable pointer in the trait reference can then become a single
     230pointer to the vtable field and use that to recover the original object
     231pointer as well as retrive all operations.
     232
     233trait drawable(otype T) {
     234        vtable drawable;
     235};
     236
     237struct line {
     238        vtable drawable;
     239        vec2 start;
     240        vec2 end;
     241};
     242
     243This inline code allows trait references to be converted to plain pointers
     244(although they still must be called specially). The vtable field may just be
     245an opaque block of memory or it may allow user access to the vtable. If so
     246than there should be some way to retrive the type of the vtable, which will be
     247autogenerated and often unique.
     248
     249
     250Keyword Usage:
     251
     252It may be desirable to add fewer new keywords than discussed in this proposal.
     253It is possible that "virtual" could replace both "vtable" above with
     254unambiguous contextual meaning. However, for purposes of clarity in the design
     255discussion it is beneficial to keep the keywords for separate concepts distinct.
     256
     257
     258Trait References and Operations:
     259
     260sizeof(drawable) will return the size of the trait object itself. However :
     261
     262line a_line;
     263drawable widget = a_line;
     264sizeof(widget);
     265
     266Will instead return the sizeof the underlying object, although the trait must
     267require that its implementation is sized for there to be a meaningful value
     268to return. You may also get the size of the trait reference with
     269
     270sizeof(&widget);
     271
     272Calling free on a trait reference will free the memory for the object. It will
     273leave the vtables alone, as those are (always?) statically allocated.
Note: See TracChangeset for help on using the changeset viewer.