Commit Graph

899 Commits

Author SHA1 Message Date
Chris Lattner d7e7b8e411 first wave of fixes for @encode sema support. This is part of PR3648.
The big difference here is that (like string literal) @encode has 
array type, not pointer type.

llvm-svn: 65391
2009-02-24 22:18:39 +00:00
Fariborz Jahanian 6fe4306195 Set flag for -fobjc-gc in IMAGE_INFO variable.
llvm-svn: 65387
2009-02-24 21:08:09 +00:00
Daniel Dunbar d4ecca135a Fix IRgen of constant expressions referring to external/static
variables.
 - PR3657.

llvm-svn: 65381
2009-02-24 18:41:57 +00:00
Daniel Dunbar 0b0dcd987d Some initial Obj-C zero cost EH support.
- Only handles cases with @try with no @catch blocks, and there are a
   number of problems with the implementation. Nevertheless, this is
   good enough to handled @synchronized correctly, and some other
   basic uses.

llvm-svn: 65378
2009-02-24 07:47:38 +00:00
Daniel Dunbar 36ae309434 Add a note about an IRgen optimization opportunity.
llvm-svn: 65376
2009-02-24 06:34:04 +00:00
Anders Carlsson 729a8202d0 Prevent accidental copying of CodeGenFunction and CodeGenModule.
llvm-svn: 65372
2009-02-24 04:21:31 +00:00
Anders Carlsson b9c9e1d16f Pass the CodeGenModule object to GenerateBlockFunction, instead of *this (which will call the copy constructor).
Also, since we're creating a new CodeGenFunction object for each block function, we don't need to clear the BreakContinueStack.

llvm-svn: 65371
2009-02-24 04:19:41 +00:00
Mike Stump 18e91257e4 Ensure that we can't break or continue out of a block.
llvm-svn: 65370
2009-02-24 02:59:30 +00:00
Daniel Dunbar 94ceb61574 Fix two @synchronized bugs found by inspection: the expression to sychronize on should only be evaluated once, and it is evaluated outside the cleanup scope.
Also, lift SyncEnter and SyncExit up in nervous anticipation of x86-64
zero cost EH.

llvm-svn: 65362
2009-02-24 01:43:46 +00:00
Douglas Gregor e62c0a45dd Improve merging of function declarations. Specifically:
- When we are declaring a function in local scope, we can merge with
    a visible declaration from an outer scope if that declaration
    refers to an entity with linkage. This behavior now works in C++
    and properly ignores entities without linkage.
  - Diagnose the use of "static" on a function declaration in local
    scope.
  - Diagnose the declaration of a static function after a non-static
    declaration of the same function.
  - Propagate the storage specifier to a function declaration from a
    prior declaration (PR3425)
  - Don't name-mangle "main"

llvm-svn: 65360
2009-02-24 01:23:02 +00:00
Fariborz Jahanian 392124c78e We should not generate __weak write barrier on indirect reference
of a pointer to object; This patch does this odd behavior according to
gcc.

llvm-svn: 65334
2009-02-23 18:59:50 +00:00
Daniel Dunbar 1234749853 Add low level support for generating invoke instead of calls.
- No functionality change.

llvm-svn: 65325
2009-02-23 17:26:39 +00:00
Anders Carlsson e14282e274 Make sure to reset the DidCallStackSave variable before emitting a compound statement. Fixes PR3649.
llvm-svn: 65291
2009-02-22 18:44:21 +00:00
Fariborz Jahanian 0773533b27 More objc gc work. Match gcc's treatment of ivar access
true a local pointer to objective-c object in generating
write barriers.

llvm-svn: 65290
2009-02-22 18:40:18 +00:00
Mike Stump e8b2d066ea Copy some clenaups from Eli to code that I copied. :-)
llvm-svn: 65286
2009-02-22 13:55:31 +00:00
Mike Stump 624497c29a Cleanp code with some recent suggestions.
llvm-svn: 65285
2009-02-22 13:27:11 +00:00
Daniel Dunbar bb525c3c7f x86_64 ABI: Actually, we can always pass things we want to pass in
memory using Indirect; this was a holdover from when CGCall wasn't as
robust.

llvm-svn: 65278
2009-02-22 08:17:51 +00:00
Eli Friedman 825fe7565e Eliminate a bunch of code which should be dead.
llvm-svn: 65267
2009-02-22 07:29:04 +00:00
Daniel Dunbar 075d642e24 x86_64 ABI: Make sure to pass vectors that we want to pass in memory
as byval. Otherwise LLVM will have its own opinion about where to put
things.

We now pass all gcc dg.compat tests on x86_64.

llvm-svn: 65266
2009-02-22 07:22:25 +00:00
Daniel Dunbar c64c481d18 x86_64 ABI: Pass 32-bit vectors as Integer to match gcc. We don't care
about these much but <2 x i16> shows up in the gcc test suite.

llvm-svn: 65264
2009-02-22 04:48:22 +00:00
Daniel Dunbar f25afad9e5 x86_64 ABI: Classify <1 x i64> as INTEGER (match gcc not llvm-gcc).
Also, make sure to pass <1 x i64> as i64 (not <1 x i64>, which doesn't
quite work yet in the backend).

llvm-svn: 65262
2009-02-22 04:16:10 +00:00
Eli Friedman 34ef305759 Last part of PR3254: use the same alignment computation in Sema and
CodeGen.  I'm not sure whether this actually makes any visible 
difference, but it's better to be consistent anyway.

llvm-svn: 65259
2009-02-22 03:23:43 +00:00
Fariborz Jahanian a9fecf3f9e local array of objects are non-gc'able.
llvm-svn: 65251
2009-02-21 23:37:19 +00:00
Steve Naroff 8487e3e541 More work to integrate newly added ObjCQualifiedClassType into the type system.
This is necessary 'plumbing' to fix <rdar://problem/6497631> Message lookup is sometimes different than gcc's.

llvm-svn: 65248
2009-02-21 21:17:01 +00:00
Mike Stump 2ac40a9ff8 The blocks ABI is wrong, add a FIXME.
llvm-svn: 65243
2009-02-21 20:07:44 +00:00
Mike Stump cb2fbcb0c9 Add CodeGen support for the helper for BlockDeclRefExprs. The easier
stuff is mostly done.  Move BlockHasCopyDispose up.

llvm-svn: 65242
2009-02-21 20:00:35 +00:00
Fariborz Jahanian 10bec1099c Handle case of none gc'able objects regardless of their
type.

llvm-svn: 65205
2009-02-21 00:30:43 +00:00
Daniel Dunbar 3c81dabd76 Emit extern_weak when needed.
- PR3629.

llvm-svn: 65203
2009-02-21 00:24:10 +00:00
Daniel Dunbar 0c4eda5923 We must always mangle attribute overloadable functions; even if in a
system header.
 - Prevents a codegen crash when anything used anything in tgmath! :)

llvm-svn: 65200
2009-02-20 23:09:27 +00:00
Daniel Dunbar 2f3a3d9f7a Add some IRgen improvement notes.
llvm-svn: 65146
2009-02-20 19:34:45 +00:00
Daniel Dunbar 27032de34b Shorten; no functionality change.
llvm-svn: 65145
2009-02-20 19:34:33 +00:00
Daniel Dunbar 5006f4a5f0 Take advantage of noreturn attribute to add unreachable instruction &
clear insertion point. The rest of IRgen should theoretically take
advantage of this to avoid emitting dead code. Theory != Practice.

llvm-svn: 65141
2009-02-20 18:54:31 +00:00
Chris Lattner a499715c83 remove some more methods from objc decls, using the iterator
interfaces more consistently.

llvm-svn: 65138
2009-02-20 18:43:26 +00:00
Daniel Dunbar cdbb5e336d Set call attribute for direct calls (i.e. noreturn).
- Remove an unused variant of EmitCallExpr overload.

llvm-svn: 65130
2009-02-20 18:06:48 +00:00
Chris Lattner 90669d0500 switch ObjCMethodDecl's parameter list from being explicitly managed to an ObjCList.
llvm-svn: 65114
2009-02-20 06:23:21 +00:00
Eli Friedman 719ed1a9ab Initialize the Init variable to something reasonable when we emit an
error, so we don't crash.

llvm-svn: 65099
2009-02-20 01:18:21 +00:00
Fariborz Jahanian c86fb5ecb4 More objc gc's ir-gen fixes.
llvm-svn: 65097
2009-02-20 01:14:43 +00:00
Mike Stump 2a9baebae1 Deox and Capitolize.
llvm-svn: 65093
2009-02-20 00:45:51 +00:00
Mike Stump f5a5c4fa11 Fix spacing.
llvm-svn: 65089
2009-02-20 00:19:45 +00:00
Fariborz Jahanian 0f466c746f More objc's gc ir-gen stuff.
llvm-svn: 65077
2009-02-19 23:36:06 +00:00
Mike Stump a67033294a Add enough checking to ensure that non-constant block literals don't
appear to be constant.  I'll probably redo this and throw it all away
later once we have codegen for BlockDeclRefExprs.

llvm-svn: 65070
2009-02-19 22:01:56 +00:00
Daniel Dunbar 18e1444f82 Remove IRgen constant emission assumption that LValue APValue results
only occur for pointer types; they are also possible for integer types
now.
 - No intended functionality change, IntExprEvaluate doesn't return
   LValue results yet.

llvm-svn: 65066
2009-02-19 21:44:24 +00:00
Anders Carlsson f96954ca30 Add irgen support for the noinline attribute.
llvm-svn: 65056
2009-02-19 19:22:11 +00:00
Fariborz Jahanian 3114e94e10 Generate the conservative objc gc's API for now.
llvm-svn: 65051
2009-02-19 18:29:24 +00:00
Daniel Dunbar f0acf7bd4c Don't emit K&R unprototyped function definitions as varargs.
- <rdar://problem/6584606> clang/x86-64 - too many reg saves

llvm-svn: 65032
2009-02-19 07:15:39 +00:00
Daniel Dunbar b2f4cdb233 Emission of global variable initialializer was broken in rare
situation where a tentative decl was emitted *after* the actual
initialization. This occurs in some rare situations with static decls.
 - PR3613.

 - I'm not particularly happy with this fix, but I don't see a simpler
   or more elegant solution yet.

llvm-svn: 65018
2009-02-19 05:36:41 +00:00
Mike Stump 5d2534ada7 More codegen for blocks. The type of block literals should be better.
The size calculation is improved.

llvm-svn: 64994
2009-02-19 01:01:04 +00:00
Fariborz Jahanian 83e3eea5fc Some code simplification. ir gen for gc'able array
of objects in objc.

llvm-svn: 64992
2009-02-19 00:48:05 +00:00
Douglas Gregor 5f361c9f1e Address Chris's comments regarding C++ name mangling.
llvm-svn: 64984
2009-02-18 23:53:56 +00:00
Daniel Dunbar a228a67637 Add anti-FIXME.
llvm-svn: 64969
2009-02-18 22:52:09 +00:00