llvm-project/clang/lib/CodeGen
Rafael Espindola df88f6fe98 Replace MarkVarRequired with a more generic
HandleCXXStaticMemberVarInstantiation. Suggested by Argyrios.

llvm-svn: 152320
2012-03-08 15:51:03 +00:00
..
ABIInfo.h Add field PaddingType to ABIArgInfo which specifies the type of padding that 2012-01-07 00:25:33 +00:00
BackendUtil.cpp Add a flag -fthread-sanitizer. 2012-03-01 22:27:08 +00:00
CGBlocks.cpp Implement "optimization" for lambda-to-block conversion which inlines the generated block literal for lambdas which are immediately converted to block pointer type. This simplifies the AST, avoids an unnecessary copy of the lambda and makes it much easier to avoid copying the result onto the heap. 2012-03-01 04:01:32 +00:00
CGBlocks.h Whenever explicitly activating or deactivating a cleanup, we 2011-11-10 10:43:54 +00:00
CGBuilder.h Make CGBuilderTy a typedef again; its functionality has been rolled back 2010-07-06 18:43:48 +00:00
CGBuiltin.cpp add a testcase for PR12094 and fix a crash on pointer to incomplete type, 2012-03-04 00:52:12 +00:00
CGCUDANV.cpp CUDA: IR generation support for device stubs 2011-10-06 18:51:56 +00:00
CGCUDARuntime.cpp CUDA: IR generation support for kernel call expressions 2011-10-06 18:29:37 +00:00
CGCUDARuntime.h CUDA: IR generation support for device stubs 2011-10-06 18:51:56 +00:00
CGCXX.cpp Whether an argument is required (in contrast with being an 2012-02-17 03:33:10 +00:00
CGCXXABI.cpp Whether an argument is required (in contrast with being an 2012-02-17 03:33:10 +00:00
CGCXXABI.h Split reinterpret_casts of member pointers out from CK_BitCast; this 2012-02-15 01:22:51 +00:00
CGCall.cpp Whether an argument is required (in contrast with being an 2012-02-17 03:33:10 +00:00
CGCall.h Whether an argument is required (in contrast with being an 2012-02-17 03:33:10 +00:00
CGClass.cpp Implement "optimization" for lambda-to-block conversion which inlines the generated block literal for lambdas which are immediately converted to block pointer type. This simplifies the AST, avoids an unnecessary copy of the lambda and makes it much easier to avoid copying the result onto the heap. 2012-03-01 04:01:32 +00:00
CGCleanup.cpp Taken into account Duncan's comments for r149481 dated by 2nd Feb 2012: 2012-03-08 07:06:48 +00:00
CGCleanup.h Simplify EH control flow by observing that EH scopes form a simple 2011-08-11 02:22:43 +00:00
CGDebugInfo.cpp Reapply r151702 with a small fix for a failure to cut and paste 2012-03-01 21:36:52 +00:00
CGDebugInfo.h Add in a caching mechanism so that forward declarations are replaced 2012-02-18 00:50:17 +00:00
CGDecl.cpp Make heap-allocation of std::initializer_list 'work'. 2012-02-19 16:03:09 +00:00
CGDeclCXX.cpp Emit the exact size for the invariant intrinsics. 2012-02-21 00:26:58 +00:00
CGException.cpp Remove dead code. 2012-02-19 11:57:29 +00:00
CGExpr.cpp AST representation for user-defined literals, plus just enough of semantic 2012-03-07 08:35:16 +00:00
CGExprAgg.cpp Make sure list-initialization of arrays works correctly in explicit type conversions. PR12121. 2012-02-29 00:00:28 +00:00
CGExprCXX.cpp Formatting. 2012-02-29 03:25:18 +00:00
CGExprComplex.cpp Add missing code for compound literals of complex type. <rdar://problem/10938628> 2012-02-27 20:26:13 +00:00
CGExprConstant.cpp Reinstate r151879, r151880, reverted in r151922, along with a bugfix for 2012-03-02 23:27:11 +00:00
CGExprScalar.cpp Don't even try to directly emit the value of a DeclRefExpr if that declaration 2012-03-07 01:58:44 +00:00
CGObjC.cpp Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, 2012-03-06 20:05:56 +00:00
CGObjCGNU.cpp Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, 2012-03-06 20:05:56 +00:00
CGObjCMac.cpp Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, 2012-03-06 20:05:56 +00:00
CGObjCRuntime.cpp Whether an argument is required (in contrast with being an 2012-02-17 03:33:10 +00:00
CGObjCRuntime.h Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, 2012-03-06 20:05:56 +00:00
CGOpenCLRuntime.cpp OpenCL: introduce support for function scope __local variables 2011-09-19 21:14:35 +00:00
CGOpenCLRuntime.h OpenCL: introduce support for function scope __local variables 2011-09-19 21:14:35 +00:00
CGRTTI.cpp simplify a bunch of code to use the well-known LLVM IR types computed by CodeGenModule. 2012-02-07 00:39:47 +00:00
CGRecordLayout.h clean up forward declarations of raw_ostream to use the new LLVM.h 2011-07-23 10:35:09 +00:00
CGRecordLayoutBuilder.cpp Silence a GCC warning about a set-but-not-used variable in release builds. 2012-03-04 12:16:40 +00:00
CGStmt.cpp Fold debug scope emission into the cleanup scope. 2012-02-23 00:43:07 +00:00
CGVTT.cpp simplify a bunch of code to use the well-known LLVM IR types computed by CodeGenModule. 2012-02-07 00:39:47 +00:00
CGVTables.cpp Whether an argument is required (in contrast with being an 2012-02-17 03:33:10 +00:00
CGVTables.h Move VTable builder to AST 2011-09-26 01:57:12 +00:00
CGValue.h Reuse forAddr to create ignored AggValueSlots. 2011-12-11 16:34:24 +00:00
CMakeLists.txt Add missing dependency to unbreak the CMake build 2012-02-01 19:53:01 +00:00
CodeGenAction.cpp Replace MarkVarRequired with a more generic 2012-03-08 15:51:03 +00:00
CodeGenFunction.cpp Implement "optimization" for lambda-to-block conversion which inlines the generated block literal for lambdas which are immediately converted to block pointer type. This simplifies the AST, avoids an unnecessary copy of the lambda and makes it much easier to avoid copying the result onto the heap. 2012-03-01 04:01:32 +00:00
CodeGenFunction.h Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, 2012-03-06 20:05:56 +00:00
CodeGenModule.cpp Replace MarkVarRequired with a more generic 2012-03-08 15:51:03 +00:00
CodeGenModule.h Replace MarkVarRequired with a more generic 2012-03-08 15:51:03 +00:00
CodeGenTBAA.cpp Revert my patches which removed Diagnostic.h includes by moving some operator overloads out of line. 2012-02-07 22:29:24 +00:00
CodeGenTBAA.h now that we have a centralized place to do so, add some using declarations for 2011-07-20 06:58:45 +00:00
CodeGenTypes.cpp Trying to increase my Ohloh ranking with trivial tweaks 2012-02-25 00:16:17 +00:00
CodeGenTypes.h Whether an argument is required (in contrast with being an 2012-02-17 03:33:10 +00:00
ItaniumCXXABI.cpp Whether an argument is required (in contrast with being an 2012-02-17 03:33:10 +00:00
Makefile BUILD_ARCHIVE is the default for libraries, no need to set it. 2010-07-18 00:14:47 +00:00
MicrosoftCXXABI.cpp remove unneeded llvm:: namespace qualifiers on some core types now that LLVM.h imports 2011-07-23 10:55:15 +00:00
ModuleBuilder.cpp Replace MarkVarRequired with a more generic 2012-03-08 15:51:03 +00:00
README.txt
TargetInfo.cpp Adding support for Microsoft's thiscall calling convention. Clang side of the patch. 2012-02-22 03:04:13 +00:00
TargetInfo.h Whether an argument is required (in contrast with being an 2012-02-17 03:33:10 +00:00

README.txt

IRgen optimization opportunities.

//===---------------------------------------------------------------------===//

The common pattern of
--
short x; // or char, etc
(x == 10)
--
generates an zext/sext of x which can easily be avoided.

//===---------------------------------------------------------------------===//

Bitfields accesses can be shifted to simplify masking and sign
extension. For example, if the bitfield width is 8 and it is
appropriately aligned then is is a lot shorter to just load the char
directly.

//===---------------------------------------------------------------------===//

It may be worth avoiding creation of alloca's for formal arguments
for the common situation where the argument is never written to or has
its address taken. The idea would be to begin generating code by using
the argument directly and if its address is taken or it is stored to
then generate the alloca and patch up the existing code.

In theory, the same optimization could be a win for block local
variables as long as the declaration dominates all statements in the
block.

NOTE: The main case we care about this for is for -O0 -g compile time
performance, and in that scenario we will need to emit the alloca
anyway currently to emit proper debug info. So this is blocked by
being able to emit debug information which refers to an LLVM
temporary, not an alloca.

//===---------------------------------------------------------------------===//

We should try and avoid generating basic blocks which only contain
jumps. At -O0, this penalizes us all the way from IRgen (malloc &
instruction overhead), all the way down through code generation and
assembly time.

On 176.gcc:expr.ll, it looks like over 12% of basic blocks are just
direct branches!

//===---------------------------------------------------------------------===//