Ayman Musa c5490e5a29 [X86] Relocate code of replacement of subtarget unsupported masked memory intrinsics to run also on -O0 option. ... Currently, when masked load, store, gather or scatter intrinsics are used, we check in CodeGenPrepare pass if the subtarget support these intrinsics, if not we replace them with scalar code - this is a functional transformation not an optimization (not optional). CodeGenPrepare pass does not run when the optimization level is set to CodeGenOpt::None (-O0). Functional transformation should run with all optimization levels, so here I created a new pass which runs on all optimization levels and does no more than this transformation. Differential Revision: https://reviews.llvm.org/D32487 llvm-svn: 303050		2017-05-15 11:30:54 +00:00
..
AnalysisWrappers.cpp	[CG] Actually hoist up the generic CallGraphPrinter pass from a weird	2016-03-10 11:08:44 +00:00
BreakpointPrinter.cpp	Turn DISubprogram into a variable-length node.	2017-04-26 23:59:52 +00:00
BreakpointPrinter.h	…
CMakeLists.txt	[CMake] opt depends on intrinsics_gen	2016-11-19 03:18:50 +00:00
GraphPrinters.cpp	Cleanup dump() functions.	2017-01-28 02:02:38 +00:00
LLVMBuild.txt	Wrap some long lines in LLVMBuild files. NFC	2015-06-12 18:44:57 +00:00
NewPMDriver.cpp	[PM] Separate the LoopAnalysisManager from the LoopPassManager and move	2017-01-11 09:43:56 +00:00
NewPMDriver.h	[opt] Remove an unused argument to runPassPipeline().	2016-09-07 00:48:47 +00:00
PassPrinters.cpp	Fix Clang-tidy readability-redundant-string-cstr warnings	2016-11-02 16:43:50 +00:00
PassPrinters.h	…
PrintSCC.cpp	…
opt.cpp	[X86] Relocate code of replacement of subtarget unsupported masked memory intrinsics to run also on -O0 option.	2017-05-15 11:30:54 +00:00