llvm-project/clang/lib/CodeGen
Alp Toker 541d5070bf Avoid dubious IdentifierInfo::getNameStart() uses
These cases in particular were incurring an extra strlen() when we already knew
the length. They appear to be leftovers from when the interfaces worked with C
strings that have continued to compile due to the implicit StringRef ctor.

llvm-svn: 210403
2014-06-07 23:30:53 +00:00
..
ABIInfo.h Sort all the #include lines with LLVM's utils/sort_includes.py which 2014-01-07 11:51:46 +00:00
BackendUtil.cpp Remove unused variable 2014-06-02 19:45:14 +00:00
CGAtomic.cpp [C++11] Use 'nullptr'. CodeGen edition. 2014-05-21 05:09:00 +00:00
CGBlocks.cpp Implement -Wframe-larger-than backend diagnostic 2014-06-05 22:10:59 +00:00
CGBlocks.h Don't indent inside namespaces. 2014-05-09 00:26:20 +00:00
CGBuilder.h This patch adds a helper class (CGLoopInfo) for marking memory instructions with llvm.mem.parallel_loop_access metadata. 2014-05-22 08:54:05 +00:00
CGBuiltin.cpp Add __builtin_operator_new and __builtin_operator_delete, which act like calls 2014-06-03 23:27:44 +00:00
CGCUDANV.cpp [C++11] Add 'override' keyword to virtual methods that override their base class. 2014-03-12 06:41:41 +00:00
CGCUDARuntime.cpp [C++11] Use 'nullptr'. CodeGen edition. 2014-05-21 05:09:00 +00:00
CGCUDARuntime.h CUDA: IR generation support for device stubs 2011-10-06 18:51:56 +00:00
CGCXX.cpp Update for llvm API change. 2014-06-03 02:42:01 +00:00
CGCXXABI.cpp [C++11] Use 'nullptr'. CodeGen edition. 2014-05-21 05:09:00 +00:00
CGCXXABI.h MS ABI: Update the thunk linkage computation 2014-06-06 20:04:01 +00:00
CGCall.cpp Add 'nonnull' parameter or return attribute when producing an llvm pointer type in a function type where the C++ type is a reference. Update the tests. 2014-05-28 09:56:42 +00:00
CGCall.h [C++11] Use 'nullptr'. CodeGen edition. 2014-05-21 05:09:00 +00:00
CGClass.cpp Start adding support for dllimport/dllexport on classes (PR11170) 2014-05-30 16:59:42 +00:00
CGCleanup.cpp [C++11] Use 'nullptr'. CodeGen edition. 2014-05-21 05:09:00 +00:00
CGCleanup.h Update for llvm api change. 2014-06-04 18:51:46 +00:00
CGDebugInfo.cpp Delete apparently unused method 2014-06-02 21:05:54 +00:00
CGDebugInfo.h Delete apparently unused method 2014-06-02 21:05:54 +00:00
CGDecl.cpp Implement -Wframe-larger-than backend diagnostic 2014-06-05 22:10:59 +00:00
CGDeclCXX.cpp Use comdats to avoid double initialization of weak data 2014-05-23 21:13:45 +00:00
CGException.cpp Update for llvm api change. 2014-06-04 18:51:46 +00:00
CGExpr.cpp Mangle predefined string constants names to merge them at link-time 2014-06-05 20:53:34 +00:00
CGExprAgg.cpp Remove incorrect assertion. 2014-06-03 08:40:27 +00:00
CGExprCXX.cpp [MS-ABI] Implement dynamic_cast 2014-06-06 22:12:37 +00:00
CGExprComplex.cpp [C++11] Use 'nullptr'. CodeGen edition. 2014-05-21 05:09:00 +00:00
CGExprConstant.cpp [C++11] Use 'nullptr'. CodeGen edition. 2014-05-21 05:09:00 +00:00
CGExprScalar.cpp [CodeGen] Don't cast and use SizeTy instead of Int32Ty when constructing {extract,insert} vector element instructions. 2014-05-31 00:22:12 +00:00
CGLoopInfo.cpp This patch adds a helper class (CGLoopInfo) for marking memory instructions with llvm.mem.parallel_loop_access metadata. 2014-05-22 08:54:05 +00:00
CGLoopInfo.h This patch adds a helper class (CGLoopInfo) for marking memory instructions with llvm.mem.parallel_loop_access metadata. 2014-05-22 08:54:05 +00:00
CGObjC.cpp [C++11] Use 'nullptr'. CodeGen edition. 2014-05-21 05:09:00 +00:00
CGObjCGNU.cpp Remove the overload of GetAddrOfConstantString method 2014-06-04 20:25:57 +00:00
CGObjCMac.cpp Avoid dubious IdentifierInfo::getNameStart() uses 2014-06-07 23:30:53 +00:00
CGObjCRuntime.cpp Update for llvm api change. 2014-06-04 18:51:46 +00:00
CGObjCRuntime.h [C++11] Use 'nullptr'. CodeGen edition. 2014-05-21 05:09:00 +00:00
CGOpenCLRuntime.cpp [C++11] Use 'nullptr'. CodeGen edition. 2014-05-21 05:09:00 +00:00
CGOpenCLRuntime.h Rewrite #includes for llvm/Foo.h to llvm/IR/Foo.h as appropriate to 2013-01-02 11:45:17 +00:00
CGOpenMPRuntime.cpp [OPENMP] Improve debug location codegen for OpenMP runtime library. 2014-05-30 05:48:40 +00:00
CGOpenMPRuntime.h [OPENMP] Improve debug location codegen for OpenMP runtime library. 2014-05-30 05:48:40 +00:00
CGRTTI.cpp [MS-ABI] Implements MS-compatible RTTI 2014-05-23 16:07:43 +00:00
CGRecordLayout.h When a module completes the definition of a class template specialization imported from another module, emit an update record, rather than using the broken decl rewriting mechanism. If multiple modules do this, merge the definitions together, much as we would if they were separate declarations. 2014-04-19 03:48:30 +00:00
CGRecordLayoutBuilder.cpp [C++11] Use 'nullptr'. CodeGen edition. 2014-05-21 05:09:00 +00:00
CGStmt.cpp Adding a new #pragma for the vectorize and interleave optimization hints. 2014-06-06 12:40:24 +00:00
CGStmtOpenMP.cpp This patch adds a helper class (CGLoopInfo) for marking memory instructions with llvm.mem.parallel_loop_access metadata. 2014-05-22 08:54:05 +00:00
CGVTT.cpp Remove the -fhidden-weak-vtables -cc1 option. It was dead, 2014-02-08 00:41:16 +00:00
CGVTables.cpp MS ABI: Update the thunk linkage computation 2014-06-06 20:04:01 +00:00
CGVTables.h Fix comment vs function name mismatch 2014-06-02 18:50:54 +00:00
CGValue.h [C++11] Use 'nullptr'. CodeGen edition. 2014-05-21 05:09:00 +00:00
CMakeLists.txt [MS-ABI] Implements MS-compatible RTTI 2014-05-23 16:07:43 +00:00
CodeGenABITypes.cpp [ms-cxxabi] Use x86_cdeclmethodcc for __cdecl methods on win32 2014-01-31 22:54:50 +00:00
CodeGenAction.cpp Provide fallback locations for backend remarks 2014-06-05 22:11:12 +00:00
CodeGenFunction.cpp This patch adds a helper class (CGLoopInfo) for marking memory instructions with llvm.mem.parallel_loop_access metadata. 2014-05-22 08:54:05 +00:00
CodeGenFunction.h Adding a new #pragma for the vectorize and interleave optimization hints. 2014-06-06 12:40:24 +00:00
CodeGenModule.cpp Implement -Wframe-larger-than backend diagnostic 2014-06-05 22:10:59 +00:00
CodeGenModule.h Implement -Wframe-larger-than backend diagnostic 2014-06-05 22:10:59 +00:00
CodeGenPGO.cpp InstrProf: Set profile data to visibility hidden 2014-05-16 01:24:00 +00:00
CodeGenPGO.h [C++11] Use 'nullptr'. CodeGen edition. 2014-05-21 05:09:00 +00:00
CodeGenTBAA.cpp [C++11] Use 'nullptr'. CodeGen edition. 2014-05-21 05:09:00 +00:00
CodeGenTBAA.h Fixed comment typo. 2013-06-08 00:27:19 +00:00
CodeGenTypes.cpp [C++11] Use 'nullptr'. CodeGen edition. 2014-05-21 05:09:00 +00:00
CodeGenTypes.h [ms-cxxabi] Use inalloca on win32 when passing non-trivial C++ objects 2014-02-01 00:04:45 +00:00
EHScopeStack.h [C++11] Use 'nullptr'. CodeGen edition. 2014-05-21 05:09:00 +00:00
ItaniumCXXABI.cpp MS ABI: Update the thunk linkage computation 2014-06-06 20:04:01 +00:00
Makefile
MicrosoftCXXABI.cpp MS ABI: Update the thunk linkage computation 2014-06-06 20:04:01 +00:00
MicrosoftRTTI.cpp Fix typos 2014-06-06 12:02:07 +00:00
ModuleBuilder.cpp Defer codegen of inline method definitions to the end of current top level declaration 2014-06-06 17:36:17 +00:00
README.txt
TargetInfo.cpp StringRefize TargetInfo::getABI() 2014-06-07 23:30:42 +00:00
TargetInfo.h [C++11] Use 'nullptr'. CodeGen edition. 2014-05-21 05:09:00 +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!

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