Eli Friedman
617ba48fd5
Always check that the definition of a function has the correct type.
...
This fixes a crash on the included testcase (found in NetHack).
llvm-svn: 51767
2008-05-30 11:13:18 +00:00
Eli Friedman
43a6d51494
More complete/correct implementation of the comparison operators for
...
constant expressions.
llvm-svn: 51682
2008-05-29 15:17:49 +00:00
Eli Friedman
3c28524632
Fix an extremely subtle bug with pointer comparisons: they have to be
...
unsigned because it's possible (at least in theory) to have
have both positive and negative pointers pointing to the same object.
llvm-svn: 51681
2008-05-29 15:09:15 +00:00
Eli Friedman
133e8040ca
Rearrange EmitLValueForField a bit to work properly for _Bool bitfields
...
in unions (we don't want to do the union-specific bitcast for
bit-fields).
llvm-svn: 51678
2008-05-29 11:33:25 +00:00
Eli Friedman
045bf4ff82
Add codegen support for a few more kinds of initializer constant
...
expressions.
llvm-svn: 51677
2008-05-29 11:22:45 +00:00
Eli Friedman
174d9c26f1
Add CodeGen support for alignment on globals, both for unusual natural
...
alignment and alignment attributes.
llvm-svn: 51676
2008-05-29 11:10:27 +00:00
Eli Friedman
d72a9624a1
Tentative declarations are supposed to have common linkage, not weak.
...
This change makes clang generate the same thing as llvm-gcc.
llvm-svn: 51674
2008-05-29 11:03:17 +00:00
Eli Friedman
5a6d0ede91
Always use packed structs. This isn't really very nice, but there's
...
nothing fundamentally wrong with it. Emitting unpacked structs where
possible is more work for almost no practical benefit. We'll probably
want to fix it at some point anyway, but it's low priority.
The issue with long double in particular is that LLVM thinks an X86 long
double is 10 bytes, while clang considers it for all purposes to be
either 12 or 16 bytes, depending on the platform, even in a packed
struct.
llvm-svn: 51673
2008-05-29 10:58:49 +00:00
Devang Patel
1c56ace4df
Check first member alignment and uses packed struct if required.
...
llvm-svn: 51619
2008-05-27 22:45:40 +00:00
Devang Patel
d608074e09
Fix test case
...
llvm-svn: 51618
2008-05-27 22:44:22 +00:00
Eli Friedman
21911e89d5
A few more cases for aggregate values.
...
llvm-svn: 51596
2008-05-27 15:51:49 +00:00
Eli Friedman
5e2281ede8
Implementation of __builtin_ctlz.
...
llvm-svn: 51595
2008-05-27 15:32:46 +00:00
Eli Friedman
aee9e54dca
Don't swap function decls, and add them to the scope as they are
...
encountered. Mixing up the decls is unintuitive, and confuses the AST
destruction code. Fixes PR2360.
Note that there is a need to look up the characteristics and
declarations of a function associated with a particular name or decl,
but the original swapping code doesn't solve it properly.
http://lists.cs.uiuc.edu/pipermail/cfe-dev/2008-May/001644.html is one
suggestion for how to fix that.
llvm-svn: 51584
2008-05-27 05:07:37 +00:00
Eli Friedman
078c963a31
Add a more reliable check for whether a static declaration has already
...
been used. In preparation for the fix to PR2360, but also a minor bug
in its own right.
llvm-svn: 51583
2008-05-27 04:58:01 +00:00
Eli Friedman
bdef5df44e
Generalize the float type generation code, and specifically fix the
...
codegen of X86 long double.
llvm-svn: 51578
2008-05-27 04:20:05 +00:00
Eli Friedman
43f6b67135
A small testcase; no patch, since it was a bug in a patch I never
...
committed.
llvm-svn: 51370
2008-05-21 13:21:01 +00:00
Eli Friedman
f77daf5e58
Fix test (it was incorrectly succeeding).
...
llvm-svn: 51310
2008-05-20 09:04:19 +00:00
Eli Friedman
53e38bd6b6
Add codegen support for stack address intrinsics.
...
llvm-svn: 51309
2008-05-20 08:59:34 +00:00
Eli Friedman
4b0e2a30e0
Implement codegen for comma operator for structs.
...
llvm-svn: 51304
2008-05-20 07:56:31 +00:00
Eli Friedman
a3a4068949
Implement CodeGen for __builtin_memcpy.
...
llvm-svn: 51282
2008-05-19 23:27:48 +00:00
Eli Friedman
541268086a
Remove illegal test.
...
llvm-svn: 51267
2008-05-19 20:17:34 +00:00
Eli Friedman
f23b6fa41f
Fix the emission of expressions like char a[10] = "asdf"; previously,
...
they were causing bad code to be emitted. There are two fixes here: one
makes sure we emit a string that is long enough, and one makes sure we
properly handle string initialization in init lists.
llvm-svn: 51259
2008-05-19 17:51:16 +00:00
Eli Friedman
f2442dcc8d
Fix support for _Bool bitfields. The issue is that the bitfield width
...
used for _Bool is not the same as the primitive width (which for _Bool
is 1 bit). The load and store changes add some casts to make the
types consistent. The EmitLValue changes make sure that the pointer is
of an appropriate type for loading the bitfield.
This isn't perfect, but it's an improvement, and getting everything
right depends on actually laying out structs in an ABI-compliant way.
llvm-svn: 51224
2008-05-17 20:03:47 +00:00
Eli Friedman
cb92a3f03b
Patch for PR2334, and a similar ObjC bug.
...
llvm-svn: 51193
2008-05-16 17:37:11 +00:00
Eli Friedman
a1b4ed8003
Implementation of __builtin_shufflevector, a portable builtin capable of
...
expressing the full flexibility of the LLVM shufflevector instruction.
The expected immediate usage is in *mmintrin.h, so that they don't
depend on the mess of gcc-inherited (and not completely implemented)
shuffle builtins.
llvm-svn: 51113
2008-05-14 19:38:39 +00:00
Eli Friedman
9fd8b68568
Add codegen support for block-level compound literals.
...
llvm-svn: 51081
2008-05-13 23:18:27 +00:00
Eli Friedman
6a7087e455
Emit basic block for switch body; fixes PR2307.
...
llvm-svn: 50968
2008-05-12 16:08:04 +00:00
Eli Friedman
6e31321043
Don't try to take the address of a bitfield; fixes PR2310.
...
llvm-svn: 50966
2008-05-12 15:06:05 +00:00
Mon P Wang
b84407d6ec
Added support to generate some atomic operators (add, sub, and, or etc..)
...
llvm-svn: 50919
2008-05-09 22:40:52 +00:00
Devang Patel
ac25fc6f2e
Begin handling union bitfields.
...
Note, this is just beginning.
llvm-svn: 50835
2008-05-07 22:28:29 +00:00
Devang Patel
243d5cd283
Fix PR2101 - Codegen crash during bitfield initialization.
...
llvm-svn: 50769
2008-05-06 19:57:59 +00:00
Chris Lattner
033352b036
add a testcase
...
llvm-svn: 50608
2008-05-04 00:56:25 +00:00
Chris Lattner
73669defed
remove the 'unused static' diagnostic from codegen. This should be
...
reimplemented in Sema someday. This is tracked in PR2236. Until then
disabling it removes some bogus diagnostics (see 2236).
llvm-svn: 50263
2008-04-25 15:47:13 +00:00
Nate Begeman
1c906206b8
Fix PR2236
...
Add test
Implement feedback from Chris re: PR1998
llvm-svn: 50001
2008-04-20 20:38:08 +00:00
Nate Begeman
ce4d7fce6b
OCUVector -> ExtVector, shorthand for extended vector, per feedback from Chris.
...
llvm-svn: 49942
2008-04-18 23:10:10 +00:00
Chris Lattner
d71f83a8cb
this started failing because a GEP was constant folded,
...
improve the robustness of the test.
llvm-svn: 49618
2008-04-13 19:54:28 +00:00
Chris Lattner
aa9c7aed0f
Add support for C++ default arguments, and rework Parse-Sema
...
interaction for function parameters, fixing PR2046.
Patch by Doug Gregor!
llvm-svn: 49369
2008-04-08 04:40:51 +00:00
Chris Lattner
579a05d767
Fix PR2049, updating Eli's patch that fixes to mainline. This produces incorrect
...
code, but the codegen doesn't crash. I'll file a bugzilla for the AST being wrong.
llvm-svn: 49226
2008-04-04 18:42:16 +00:00
Chris Lattner
f3bc75afcf
Since isComplexType() no longer returns true for _Complex integers, the code
...
generator needs to call isAnyComplexType(). This fixes PR1960.
llvm-svn: 49220
2008-04-04 16:54:41 +00:00
Chris Lattner
e799eb5daa
Fix PR2081 (problems codegen'ing some recursive structures) patch
...
by Lauro Venancio!
llvm-svn: 49144
2008-04-03 05:50:42 +00:00
Chris Lattner
a21ad8058a
Fix several bugs in array -> pointer decomposition.
...
First, we got several CVR propagation cases wrong, which Eli pointed
out in PR2039.
Second, we didn't propagate address space qualifiers correctly, leading
to incorrect lowering of code in CodeGen/address-space.c.
Third, we didn't uniformly propagate the specifier in the array to the
pointer ("int[restrict 4]" -> "int *restrict").
This adds an ASTContext::getArrayDecayedType member that handles the
non-trivial logic for this seemingly simple operation.
llvm-svn: 49078
2008-04-02 05:18:44 +00:00
Chris Lattner
0fdc379475
The operand to the visibility attribute is required to be a quoted string, not a bare identifier.
...
llvm-svn: 47893
2008-03-04 18:08:48 +00:00
Gabor Greif
e2e5c3f394
The attribute nothrow maps to llvm attribute nounwind. Fix testcase.
...
llvm-svn: 47841
2008-03-03 14:53:09 +00:00
Chris Lattner
65079bd9be
fix codegen support for functions that are nothrow and noreturn.
...
llvm-svn: 47838
2008-03-03 03:45:26 +00:00
Chris Lattner
8496639787
Add a bunch of attributes, patch by Nuno Lopes.
...
llvm-svn: 47837
2008-03-03 03:28:21 +00:00
Lauro Ramos Venancio
01a72ff5d3
Implement codegen for the following static var init.
...
void g() {
static char a[10];
static char *b = a;
}
Now we can compile wget!
llvm-svn: 47627
2008-02-26 21:41:45 +00:00
Lauro Ramos Venancio
8af648ac55
Fix a bug in ConvertAsmString function.
...
llvm-svn: 47618
2008-02-26 19:19:58 +00:00
Lauro Ramos Venancio
c9fbd7301a
Convert x86 target specific inline asm constraints to LLVM.
...
llvm-svn: 47609
2008-02-26 18:33:46 +00:00
Eli Friedman
c62dd9d345
Some const initializer tests for stuff I just committed.
...
llvm-svn: 47442
2008-02-21 18:07:36 +00:00
Chris Lattner
3b418d8446
implement codegen support for sizeof(void), fixing PR2080.
...
llvm-svn: 47429
2008-02-21 05:45:29 +00:00