Evan Cheng f215e55d5f - Allow target to specify when is register pressure "too high". In most cases, ... it's too late to start backing off aggressive latency scheduling when most of the registers are in use so the threshold should be a bit tighter. - Correctly handle live out's and extract_subreg etc. - Enable register pressure aware scheduling by default for hybrid scheduler. For ARM, this is almost always a win on # of instructions. It's runtime neutral for most of the tests. But for some kernels with high register pressure it can be a huge win. e.g. 464.h264ref reduced number of spills by 54 and sped up by 20%. llvm-svn: 109279		2010-07-23 22:39:59 +00:00
..
Analysis	Eliminate getCanonicalInductionVariableIncrement's last user and	2010-07-23 21:34:51 +00:00
Archive	Eliminate some unnessary Path::exists() calls.	2010-05-27 20:51:54 +00:00
AsmParser	Make NamedMDNode not be a subclass of Value, and simplify the interface	2010-07-21 23:38:33 +00:00
Bitcode	Make NamedMDNode not be a subclass of Value, and simplify the interface	2010-07-21 23:38:33 +00:00
CodeGen	- Allow target to specify when is register pressure "too high". In most cases,	2010-07-23 22:39:59 +00:00
CompilerDriver	Formatting.	2010-07-23 04:19:34 +00:00
ExecutionEngine	remove the dwarf sizing stuff which is now dead, it was	2010-07-22 21:20:39 +00:00
Linker	Make NamedMDNode not be a subclass of Value, and simplify the interface	2010-07-21 23:38:33 +00:00
MC	Initial modifications to MCAssembler and TargetMachine for the MCJIT.	2010-07-22 05:58:53 +00:00
Support	Don't pass StringRef by reference.	2010-07-14 22:38:02 +00:00
System	Rather than using an ifdef on the target to zero out fields,	2010-07-14 14:32:33 +00:00
Target	- Allow target to specify when is register pressure "too high". In most cases,	2010-07-23 22:39:59 +00:00
Transforms	Speculatively revert 109117	2010-07-22 18:44:00 +00:00
VMCore	undo 80 column trespassing I caused	2010-07-22 10:37:47 +00:00
Makefile