Nick Lewycky d7c0e22d5f Improve 'tail' call marking in TRE. A bootstrap of clang goes from 375k calls marked tail in the IR to 470k, however this improvement does not carry into an improvement of the call/jmp ratio on x86. The most common pattern is a tail call + br to a block with nothing but a 'ret'. ... The number of tail call to loop conversions remains the same (1618 by my count). The new algorithm does a local scan over the use-def chains to identify local "alloca-derived" values, as well as points where the alloca could escape. Then, a visit over the CFG marks blocks as being before or after the allocas have escaped, and annotates the calls accordingly. llvm-svn: 208017		2014-05-05 23:59:03 +00:00
..
Hello	[Modules] Fix potential ODR violations by sinking the DEBUG_TYPE	2014-04-22 02:55:47 +00:00
InstCombine	Also handle ConstantAggregateZero when optimizing vpermilvar*.	2014-04-29 22:20:40 +00:00
Instrumentation	[ASan/Win] Fix issue 305 -- don't instrument .CRT initializer/terminator callbacks	2014-05-05 14:28:38 +00:00
IPO	Reapply: Add slp vectorization to LTO passes. The bug it exposed has been fixed by r207983. <radar://16641956>	2014-05-05 23:14:46 +00:00
ObjCARC	[C++] Use 'nullptr'.	2014-04-28 04:05:08 +00:00
Scalar	Improve 'tail' call marking in TRE. A bootstrap of clang goes from 375k calls marked tail in the IR to 470k, however this improvement does not carry into an improvement of the call/jmp ratio on x86. The most common pattern is a tail call + br to a block with nothing but a 'ret'.	2014-05-05 23:59:03 +00:00
Utils	Teach GlobalDCE how to remove empty global_ctor entries.	2014-05-02 18:35:25 +00:00
Vectorize	Always set alignment of vectorized LD/ST in SLP-Vectorizer. <rdar://problem/16812145>	2014-05-05 17:59:14 +00:00
CMakeLists.txt
LLVMBuild.txt
Makefile