IRgen: Move EmitStoreThroughBitfieldLValue to use new CGBitfieldInfo::AccessInfo decomposition, instead of computing the access policy itself.
- Sadly, this doesn't seem to give any .ll size win so

IRgen: Move EmitStoreThroughBitfieldLValue to use new CGBitfieldInfo::AccessInfo decomposition, instead of computing the access policy itself.
- Sadly, this doesn't seem to give any .ll size win so far. It is possible to make this routine significantly smarter & avoid various shifting, masking, and zext/sext, but I'm not really convinced it is worth it. It is tricky, and this is really instcombine's job.

- No intended functionality change; the test case is just to increase coverage & serves as a demo file, it worked before this commit.

llvm-svn: 101222

show more ...

IRgen: Move EmitLoadOfBitfieldLValue to use new CGBitfieldInfo::AccessInfo decomposition, instead of computing the access policy itself.
- This lets the method focus slightly more on emitting clean

IRgen: Move EmitLoadOfBitfieldLValue to use new CGBitfieldInfo::AccessInfo decomposition, instead of computing the access policy itself.
- This lets the method focus slightly more on emitting clean IR to honor the policy which has been selected. On 403.gcc's combine.c, x86_64, -O0, this reduces the number of lines in the .ll file (~= # of instructions) by 2.5%.

- No intended functionality change -- at -O3 this should produce equivalent if not identical output. On 403.gcc's combine.c, x86_64, -O3, this isn't quite true and some of the changes are regressions, but I'm not going to worry about that until we move to a new access policy.

- There is still some room for improvement in the generated IR, in particular we can usually fold the sign-extension of the bit-field into one of the component access. See the FIXME.

llvm-svn: 101192

show more ...

fix PR6805: llvm.objectsize changed to take an i1 instead of an i32.

llvm-svn: 100938

IRgen: Move the bit-field access type into CGBitFieldInfo, and change bit-field LValues to just store the base address of object containing the bit-field.

llvm-svn: 100745

IRgen: Move BitFieldIsSigned bit into CGBitFieldInfo.

llvm-svn: 100513

Simplify.

llvm-svn: 100511

IRgen: Move BitField LValues to just hold a reference to the CGBitFieldInfo.
- Unfortunately, this requires some horrible code in CGObjCMac which always
allocats a CGBitFieldInfo because we don't

IRgen: Move BitField LValues to just hold a reference to the CGBitFieldInfo.
- Unfortunately, this requires some horrible code in CGObjCMac which always
allocats a CGBitFieldInfo because we don't currently build a proper layout
for Objective-C classes. It needs to be cleaned up, but I don't want the
bit-field cleanups to be blocked on that.

llvm-svn: 100474

show more ...

IRgen: Lift BitFieldInfo to CGBitFieldInfo at namespace level.

llvm-svn: 100433

IRGen: Move the auxiliary data structures tracking AST -> LLVM mappings out of CodeGenTypes, to per-record CGRecordLayout structures.
- I did a cursory check that this was perf neutral, FWIW.

llvm-

IRGen: Move the auxiliary data structures tracking AST -> LLVM mappings out of CodeGenTypes, to per-record CGRecordLayout structures.
- I did a cursory check that this was perf neutral, FWIW.

llvm-svn: 99978

show more ...

Introduce a new kind of derived-to-base cast which bypasses the need for
null checks, and make sure we elide null checks when accessing base class
members.

llvm-svn: 99963

Minor formatting/FIXME cleanups.

llvm-svn: 99944

Fix PR6473.

Clang's support for weakref is now better than llvm-gcc's :-)

We don't introduce a new symbol and we correctly mark undefined references weak only if there is no
definition or regular u

Fix PR6473.

Clang's support for weakref is now better than llvm-gcc's :-)

We don't introduce a new symbol and we correctly mark undefined references weak only if there is no
definition or regular undefined references in the same file.

llvm-svn: 97733

show more ...

IRgen: Add CreateIRTemp, which creates a temporary alloca but with type converted "not-for-memory". Dunno a better name.

llvm-svn: 96374

IRgen: Switch EmitCompoundLiteralLValue to use CreateMemTemp.

llvm-svn: 96373

When emitting an aggregate into a temporary, make sure we set the alignment
on the alloca.

The fact that codegen makes this class of bug so wonderfully easy to make is
embarrassing.

llvm-svn: 96204

More vtable layout dumper improvements. Handle destructors, dump the complete function type of the member functions (using PredefinedExpr::ComputeName.

llvm-svn: 95887

IRgen: Add CreateMemTemp, for creating an temporary memory object for a particular type, and flood fill. - CreateMemTemp sets the alignment on the alloca correctly, which fixes a great many places in

IRgen: Add CreateMemTemp, for creating an temporary memory object for a particular type, and flood fill. - CreateMemTemp sets the alignment on the alloca correctly, which fixes a great many places in IRgen where we were doing the wrong thing.

- This fixes many many more places than the test case, but my feeling is we need to audit alignment systematically so I'm not inclined to try hard to test the individual fixes in this patch. If this bothers you, patches welcome!

PR6240.

llvm-svn: 95648

show more ...

Reapply r95393, without the change to CGExpr. I was wrong in assuming that the
element type always matched the converted LLVM type for ExprType.

llvm-svn: 95596

Standardize the parsing of function type attributes in a way that
follows (as conservatively as possible) gcc's current behavior: attributes
written on return types that don't apply there are applie

Standardize the parsing of function type attributes in a way that
follows (as conservatively as possible) gcc's current behavior: attributes
written on return types that don't apply there are applied to the function
instead, etc. Only parse CC attributes as type attributes, not as decl attributes;
don't accepet noreturn as a decl attribute on ValueDecls, either (it still
needs to apply to other decls, like blocks). Consistently consume CC/noreturn
information throughout codegen; enforce this by removing their default values
in CodeGenTypes::getFunctionInfo().

llvm-svn: 95436

show more ...

Revert r95393, which broke Clang's self-host.

llvm-svn: 95430

IRgen: A few more ConvertType cleanups.

llvm-svn: 95423

IRgen: Factor out EmitAggExprToLValue.

llvm-svn: 95416

IRgen: Fix some CreateTempAlloca calls to use ConvertTypeForMem when that is
conceptually correct. Review appreciated (Chris, Eli, Anders).

llvm-svn: 95401

Now that we store calling conventions in the types, use them instead of
getting the calling convention from the target function, which may or may not
exist. Fixes PR5280.

llvm-svn: 95399

IRgen: Use hasAggregateLLVMType instead of isSingleValueType() for cases that
need to deal with aggregates specially; this is consistent with the rest of IRgen.

Also, simplify EmitParmDecl and don't

IRgen: Use hasAggregateLLVMType instead of isSingleValueType() for cases that
need to deal with aggregates specially; this is consistent with the rest of IRgen.

Also, simplify EmitParmDecl and don't worry about using Decl::getNameAsString.

llvm-svn: 95393

show more ...