2 * Recheck gfxinput2 memory bleeding.
4 * TODO: Give each global variable its own RefStor for locking
7 * Assignments of the same variable to itself are a big problem. The
8 underlying storge is dropped and then copied, but atm the drop code
9 cleans out the storage (e.g. any pointers or references).
11 * &array[n] is causing extra hanging locks in the interpreter.
13 * &lvalue --- do not set SRS flags!!! The address is not the local stack
16 * ptr.method() calls (other than new()) just don't understand whether
17 the method object is locked or unlocked. This isn't a problem for
18 ptr->method() calls through classes since the contents of a class is
19 always naturally unlocked (but still ref'd of course).
21 * XXX *pptr or *ptr-to-compound-type does not know whether the indirected
22 object resides in naturally locked or naturally unlocked storage.
24 --> intention - disallow &locked-storage-ptr or &locked-storage-comptype,
25 require the use of 'unlocked' on the storage to be able to take its
26 address. Relax requirements if TF_HASLVPTR is false.
28 * Do not allow the address of any return argument be taken during a
29 call. Messes up content-locked return data.
31 * Global compound types need to be fixed to not try to maintain locks.
33 * Lock global and (class) object storage in GenGetDeclarationData()
35 * on LValueStor, do not flow-through LVRSGet's. Instead require that
36 LValueStor's passed as an argument or return value have their
37 underlying object locked (and other LValueStor's do not, even if on
40 This makes .new() work as the lock will be borrowed if relaying an
41 lvs passed in, and if not relaying the caller must lock the underlying.
43 When an object is forwarded by constructing an LValueStor for it,
44 the lock must be borrowed if the object is already locked. Access
45 semantics are such that this should always be the case.
47 * insnx86.c line 149 ish.. remove rea->regused ???
49 * gen_refstor.c ~720, make DOP_ARGS_STORAGE cacheable. But to
50 work properly RAS needs to do a dummy load of each argument
51 from EA[non-cache] to EA.CACHE to ensure that the register
52 is properly initialized. And for pointers, to sync back
55 * Cacheability of compound types in RAS ?
57 * see classes strlen() code. CMP/BRA opt causes RSPUT to be improperly
58 handled (fix by optimizing out the PRSGET/RSPUT sequence for the
59 comparison? what if both args need it?)
61 * optimize const global access to immediate value XXX doesn't currently
62 happen (strlen()'s test against NIL).
64 * optimized stack storage in llvm is not cleared
66 * return/result sequencing must be validated by the resolver. break/continue
67 may not cross a result; point and result; cannot be conditionalized in one
68 path unless it is also present in the other.
70 * Add 'finalize' to class/subclass decl indicating that no further subclassing
73 * libgen initializers for global declarations w/ assignments or type defaults.
75 * resolver propagate CTF_RESOLVING flag in run-time context.
77 * resolver and runtime and generator - validate when casting pointers
78 that an integral number of objects of the target fits in the source.
79 Otherwise the cast is illegal.
81 * tests/refine.d - method call through non-replicated (i.e. not part of
82 subclass) @ref currently executes the superclass method
83 instead of looking up the actual method via the type.
85 * generally speaking must change how interfaces and classes work when making
86 field accesses. Disallow in interfaces, allow otherwise but then restrict
87 how refine operates so field offsets remain consistent across all
88 sub-classes. Disallow alias access via @ref?
90 * resolver - Disallow binding of incompatible subclasses to a superclass.
91 Do this by flagging the incompatibility.
93 * resolver - Enforce that non-interface classes cannot contain interfaces.
95 * resolver - Enforce that it is not allowed to take the sizeof() an interface
98 * resolver - Enforce that one cannot instantiate/declare an interface type.
100 * resolver - Enforce that object accessors can only be to non-generic elements.
102 * resolver - Reorder generic elements in an interface to the end for offset
105 * resolver - Do not allow pointer arithmatic on references.
107 * libgen - chained temporary data returned along one side of an expressin
108 can get corrupted by the other side. On the top-level temporary data is
111 * libgen - RunRelsRefStor in operators a mess.
113 * libgen - exp.c RunCastExpPtrToRef() is basically messing with the
114 type of the passed-in lvalue, which is really a bad idea.
116 * read, write, sizeof, arysize, and bcopy currently taking Int32 !!!'s. FIX!
119 * When realizing persistent store we have to clean out any previously
122 * map files containing structures with pointers can get corrupted. Maybe
123 we should just disallow pointers entirely for any persistent store.
125 * Persistent store mmap() file path is currently hacked up. Needs command
126 line arg to set the home directory for the program, and the default
127 should be "$HOME/.<program>/rune_%s.map". libgen/heap.c
129 * infinintely recursive aliases are not currently detected.
134 * get rid of need for 'auto' prefix in casts to distinguish between
135 variable names and types.
137 * share pens and also share font information
139 * Handle simple exposure aggregation and use clip to limit update area
140 (see tests/gfxslider.d)
142 * Wide char for SSTRING ? i.e. 'x'
144 isg for resolveType's assigned expression
146 DONE<<constructors/destructors for global elements and statement blocks>> ?
148 reference type resolver and libi/exp.c TOK_STRIND and TOK_CALL. METHOD CALLS
149 via ref types (don't screw around with the exp, dup it instead?)
151 refine: check validity of procedural refinements (different return types,
152 argument types, etc...)
154 is d_Search now obsolete?
156 Add support for 'super'. Either 'structural->super' or 'super.<blah>'
157 (for method calls), to access a superclass element that has been overloaded
160 We need to be able to copy a statement tree (procedure body) on the fly
161 in the resolver, when a subclass is passed to a procedure expecting its
162 superclass. We have to keep track of compound types (proc args) in order
163 to cache the various versions of the procedure. We already have d_Super
164 to locate the procedure in a TOK_CALL if the procedure body is NULL.
165 The modified procedure must be able to pass a different sg_Level.. i.e.
166 the one for the superclass, even though all of its types will be the actual
167 class (the subclass), in order to locate the proper declaration (a declaration
168 may be invisible to the subclass but visible to the superclass).
173 int ary[exp] ... the exp is shared across qualifications of the type,
174 which is bad. Also, what about compound types sharing the same semgroup?
177 double check 'isglobal' TF/SGF_NOINIT cache
179 cast procedures generally work but we really need to formalize
180 how ex_Type is set verses what the exp node really wants (e.g. for
181 procedures... look at ex_Type or look at the procdecl's return type?).
183 really fix cast... in libi/exp.c
185 Check global scope when indirecting through a type or class rather
186 then storage. only allow global scope for storage in that case.
188 non-global procedure called through object -- automatic "this" reference.
189 procedures embedded in class accessing class elements????? global vs
192 Implement STF_FORWARD checking (do not allow forward reference stack
193 declarations to other decls at the same semantic level.)
195 Make procedural refinement work
199 Implement heap variables (implement as initialized pointers / use '@' ?)
201 Partial procedure args resolution
203 procedure pointers (point to the declaration)