Commit Graph

95845 Commits

Author SHA1 Message Date
Bob Wilson 77ad2c4164 Update TypePrinter::PrintVector to handle new Neon vector types.
llvm-svn: 119302
2010-11-16 00:32:26 +00:00
Bob Wilson 118baf7611 Add support for "neon_vector_type" and "neon_polyvector_type" attributes
to create the special Neon vector types.  These are intended to be used in
Clang's version of <arm_neon.h> to define special Neon vector types that will
be mangled according to ARM's ABI.

llvm-svn: 119301
2010-11-16 00:32:24 +00:00
Bob Wilson 7795480992 ASTContext::getVectorType should preserve the vector kind for canonical types.
This is needed for Neon types when it is most natural to define them in terms
of a typedef.  For example, Neon poly8_t is a typedef for "signed char", and
we want to define polynomial vectors as vectors of that typedef.  Without this
change, the result will be a generic GCC-style vector.  I think this is safe
for other vector types as well, but I would appreciate a review of this.

llvm-svn: 119300
2010-11-16 00:32:20 +00:00
Bob Wilson e4652b1f22 Change CXXNameMangler::mangleNeonVectorType to require the vector type to be
one of the special Neon types.  We'll check for invalid Neon vectors when
they are created, so there's no point in handling them when mangling.

llvm-svn: 119299
2010-11-16 00:32:18 +00:00
Bob Wilson 4348a800ca Fix a comment typo.
llvm-svn: 119298
2010-11-16 00:32:16 +00:00
Bob Wilson 59f9dec8cd Use getTypeSize() method.
llvm-svn: 119297
2010-11-16 00:32:12 +00:00
Owen Anderson 05a8daee21 Add Thumb2 encodings for mov and friends.
llvm-svn: 119295
2010-11-16 00:29:56 +00:00
John McCall 36459fdbad Add another case to the whitelist of cast kinds that can convert to bool.
Fixes PR8608.

llvm-svn: 119293
2010-11-16 00:12:50 +00:00
Ted Kremenek a4c27ec73b CursorVisitor: cache worklists created for data-recursion to reduce malloc() traffic.
llvm-svn: 119290
2010-11-15 23:31:32 +00:00
John McCall 8d69a2160e Add a new expression kind, OpaqueValueExpr, which is useful for
certain internal type-checking procedures as well as for representing
certain implicitly-generated operations.  Uses to follow.

llvm-svn: 119289
2010-11-15 23:31:06 +00:00
Zhanyong Wan 24f6742b34 Adds a README for the Clang static analyzer to document its design and work flow. The contents are taken from email notes by kremenek and xuzhongxing.
llvm-svn: 119288
2010-11-15 23:19:25 +00:00
Ted Kremenek acd03f6fa3 Use LLVM_ATTRIBUTE_NOINLINE instead of attribute((noinline)).
llvm-svn: 119287
2010-11-15 23:11:54 +00:00
Rafael Espindola d1993eb2a4 Change the 11 byte nop to be a single instruction.
llvm-svn: 119286
2010-11-15 23:10:30 +00:00
Douglas Gregor c65929299d Tweak libclang's heuristics for building precompiled preambles and
caching global code-completion results. In particular, don't perform
either operation the first time we parse, but do both after the first
reparse. 

llvm-svn: 119285
2010-11-15 23:00:34 +00:00
Ted Kremenek 18fb1665bf Remove invalid assertion from CFG builder. When building the CFG pieces for a ternary '?' expression,
it is possible for the confluence block to only have a single predecessor due to calls to 'noreturn'
functions.  Fixes assertion failure reported in PR 8619.

llvm-svn: 119284
2010-11-15 22:59:22 +00:00
Chris Lattner bf9f2f2c29 fix a pasto that massively broke the ppc jit while the buildbots happened
to be broken for other reasons

llvm-svn: 119283
2010-11-15 22:50:50 +00:00
Ted Kremenek e6f030467c Cast pointers instead of returning a new value within RunVisitorWorkList().
llvm-svn: 119282
2010-11-15 22:23:26 +00:00
Ted Kremenek 51df4d5596 Annotate CursorVisitor::VisitDataRecursive() with attribute 'noinline'.
Clang currently uses a ridiculous amount of stack space when inlining
this function, which can lead to premature stack overflows.

llvm-svn: 119281
2010-11-15 22:23:24 +00:00
Jakob Stoklund Olesen 97154f67d9 Revert "Fix PR8612 in the standard spiller as well."
This reverts r119183 which borke the buildbots.

llvm-svn: 119270
2010-11-15 21:51:51 +00:00
Owen Anderson ea96321781 Attempt to provide encodings for some miscellaneous Thumb2 encodings.
llvm-svn: 119187
2010-11-15 21:30:39 +00:00
Evan Cheng 2ce016c7f8 Code clean up. The peephole pass should be the one updating the instruction
iterator, not TII->OptimizeCompareInstr.

llvm-svn: 119186
2010-11-15 21:20:45 +00:00
Owen Anderson 7d97a99f4c Provide Thumb2 encodings for sxtb and friends.
llvm-svn: 119185
2010-11-15 21:12:05 +00:00
Eric Christopher 964943780b Recommit this change and remove the failing part of the test - it didn't
pass in the first place and was masked by earlier failures not warning
and aborting the block.

llvm-svn: 119184
2010-11-15 21:11:06 +00:00
Jakob Stoklund Olesen 97825bcbfd Fix PR8612 in the standard spiller as well.
The live range of a register defined by an early clobber starts at the use slot,
not the def slot.

llvm-svn: 119183
2010-11-15 20:55:53 +00:00
Jakob Stoklund Olesen ddf25c341c When spilling a register defined by an early clobber, make sure that the new
live ranges for the spill register are also defined at the use slot instead of
the normal def slot.

This fixes PR8612 for the inline spiller. A use was being allocated to the same
register as a spilled early clobber def.

This problem exists in all the spillers. A fix for the standard spiller is
forthcoming.

llvm-svn: 119182
2010-11-15 20:55:49 +00:00
Ted Kremenek 43f50695a8 Add method ParentMap::addStmt().
llvm-svn: 119181
2010-11-15 20:54:24 +00:00
Jim Grosbach 38b469effd ARM LDR_PRE/LDR_POST/STR_PRE/STR_POST (and the *B counterparts) binary encoding.
llvm-svn: 119180
2010-11-15 20:47:07 +00:00
Ted Kremenek db73d599b7 Relax assertion in SValuator so that we don't crash when analyzing a call via a function pointer that
casts the return value to something completely different.  While we need better reasoning here,
we should definately not crash.

llvm-svn: 119177
2010-11-15 20:09:42 +00:00
Owen Anderson 2a3c22efba Add Thumb2 encodings for comparison and shift operators.
llvm-svn: 119176
2010-11-15 19:58:36 +00:00
Benjamin Kramer 202d91e433 Fix compiler warnigns.
llvm-svn: 119175
2010-11-15 19:20:50 +00:00
Fariborz Jahanian 9db7b8b1ef Some cleanup of block API code.
llvm-svn: 119174
2010-11-15 19:19:38 +00:00
Johnny Chen 13e64a7411 Check in a test case for rdar://problem/8668740:
'frame variable' output for namespace variables look wrong

llvm-svn: 119172
2010-11-15 19:10:53 +00:00
Johnny Chen 8ed80e6360 Add radar info for failed expression command test sequences which are now commented out.
llvm-svn: 119171
2010-11-15 18:49:03 +00:00
Owen Anderson 0e7d728327 Add correct Thumb2 encodings for mvn and friends.
llvm-svn: 119170
2010-11-15 18:45:17 +00:00
Dan Gohman 56f3a4c761 Update examples and documentation to explicitly add basicaa, now that it's
no longer included by default.

llvm-svn: 119169
2010-11-15 18:41:10 +00:00
Johnny Chen 913af510c7 Also add a test for "frame variable '(anonymous namespace)::i'",
plus expression command using fully qualified names.

llvm-svn: 119168
2010-11-15 18:40:06 +00:00
Jim Grosbach 5cf10ea1d1 Add FIXMEs.
llvm-svn: 119167
2010-11-15 18:36:48 +00:00
Rafael Espindola d1002f6fce Fixes PR8611.
Silence warning about -g not being used during linking. I couldn't find any
change in behavior in gcc liking when given -g. Please open another bug if
I missed something.

llvm-svn: 119166
2010-11-15 18:28:16 +00:00
Johnny Chen 6ba9d1f944 Add test sequences for 'frame variable' address-of operator and 'frame variable' fully-qualified name variable.
Also add some comments.

llvm-svn: 119165
2010-11-15 18:27:57 +00:00
Jim Grosbach 40a7f57d0d Nuke redundant encoding bit set.
llvm-svn: 119164
2010-11-15 18:17:24 +00:00
Dan Gohman b83aed95c9 -basicaa is no longer the default.
llvm-svn: 119163
2010-11-15 18:07:16 +00:00
Johnny Chen 92dac7ae1b Add test sequence for 'frame variable this' and 'expression this' when stopped
inside a ctor.

llvm-svn: 119162
2010-11-15 18:04:54 +00:00
Duncan Sands f12ba1dfe1 Teach InstructionSimplify the trick of skipping incoming phi
values that are equal to the phi itself.

llvm-svn: 119161
2010-11-15 17:52:45 +00:00
Johnny Chen deac523f7c Remove the @expectedFailure decorator and uncomment 'expression -- i + j" test,
as both rdar://problem/8659840 and rdar://problem/8660275 have been fixed.

llvm-svn: 119160
2010-11-15 17:42:22 +00:00
Fariborz Jahanian c961119dcd Restore these tests. I think I fixed the problem.
We shall see.

llvm-svn: 119158
2010-11-15 17:06:17 +00:00
Duncan Sands 01157b027b Move PHI tests to phi.ll, out of select.ll.
llvm-svn: 119153
2010-11-15 16:43:28 +00:00
Rafael Espindola 8c3039b67b Change MCExpr::EvaluateAsRelocatableImpl of variables to return the original
variable if recursing fails to simplify it.

Factor AliasedSymbol to be a method of MCSymbol.

Update MCAssembler::EvaluateFixup to match the change in
EvaluateAsRelocatableImpl.

Remove the WeakRefExpr hack, as the object writer now sees the weakref with
no extra effort needed.

Nothing else is using MCTargetExpr, but keep it for now.

Now that the ELF writer sees relocations with aliases, handle

    .weak    foo2
foo2:
    .weak    bar2
    .set    bar2,foo2
    .quad    bar2

the same way gas does and produce a relocation with bar2.

llvm-svn: 119152
2010-11-15 16:33:49 +00:00
Jason W Kim 96f4c01cb3 Dovetail with Dan Dunbar's rework of ELFObjectWriter.
Added 2 new subclasses - X86ELFObjectWriter and ARMELFObectWriter.
ARM and X86 require different code for RecordRelocation(), possibly others.

llvm-svn: 119149
2010-11-15 16:18:39 +00:00
Douglas Gregor ee430a39c3 Implement C++0x [temp.func.order]p3 (aka DR532) properly. In
particular, we only add the implement object parameter type if only
one of the function templates is a non-static member function
template.

Moreover, since this DR differs from existing practice in C++98/03,
this commit implements the existing practice (which ignores the
first parameter of the function template that is not the non-static
member function template) in C++98/03 mode.

llvm-svn: 119145
2010-11-15 15:41:16 +00:00
Rafael Espindola 46c79ef113 Fix PR8565.
This moves most of the isUsed logic to the MCSymbol itself. With this we
get a bit more relaxed about allowing definitions after uses: uses that
don't evaluate their argument immediately (jmp foo) are accepted.

ddunbar, this was the smallest compromise I could think of that lets us
accept gcc (and clang!) assembly.

llvm-svn: 119144
2010-11-15 14:40:36 +00:00