Chris Lattner 7d75f78b92 Instcombine should not promote whole computation trees to "strange" ... integer types, unless they are already strange. This prevents it from turning the code produced by SROA into crazy libcalls and stuff that the code generator can't handle. In the attached example, the result was an i96 multiply that caused the x86 backend to assert. Note that if TargetData had an idea of what the legal types are for a target that this could be used to stop instcombine from introducing i64 muls, as Scott wanted. llvm-svn: 68598		2009-04-08 05:41:03 +00:00
..
ADCE
ArgumentPromotion
BlockPlacement
BranchFolding
CodeExtractor
CodeGenPrepare	reenable test	2008-11-24 21:27:20 +00:00
CondProp	Ignore dbg intrinsics while propagating conditional expression info. Take 2.	2009-02-05 23:32:52 +00:00
ConstantMerge
ConstProp
DeadArgElim	aha, DAE does have to think about PHI nodes. Many thanks to "Dr Evil" (aka Duncan)	2009-03-18 16:48:45 +00:00
DeadStoreElimination	Reimplement the inner loop of DSE. It now uniformly uses getDependence(),	2008-12-06 00:53:22 +00:00
FunctionAttrs	Remove alloca tracking from nocapture analysis. Not only	2009-01-07 19:39:06 +00:00
GlobalDCE	If an alias is dead and so is its aliasee, then globaldce would	2009-02-17 23:05:26 +00:00
GlobalOpt	Ignore debug info while evaluating function.	2009-03-09 23:04:12 +00:00
GVN	Reapply r68211, with the miscompilations it caused fixed.	2009-04-01 23:53:49 +00:00
GVNPRE
IndMemRem	Actually run the test in this directory.	2009-01-25 08:05:07 +00:00
IndVarSimplify	Back out the change in 64918 that used sign-extensions when promoting	2009-02-23 23:20:35 +00:00
Inline	If a function is marked alwaysinline, it must be inlined (possibly for correctness). Do so even if the callee has dynamic alloca and the caller doesn't.	2009-03-10 07:57:50 +00:00
InstCombine	Instcombine should not promote whole computation trees to "strange"	2009-04-08 05:41:03 +00:00
Internalize	Teach the internalize pass to also internalize	2009-01-05 21:24:45 +00:00
IPConstantProp
JumpThreading	Fix PR3353, infinitely jump threading an infinite loop make from switches.	2009-01-19 21:20:34 +00:00
LCSSA
LICM	Before deleting a basic block, give other loop passes a chance cleanup analysis values, related to the instructions in the basic block.	2009-03-25 23:57:48 +00:00
LoopDeletion	don't spew tons of stuff to the output. This testcase is *not* for	2008-12-03 06:41:50 +00:00
LoopIndexSplit	Loop Index Split can eliminate a loop if it can determin if loop body is executed only once. There was a bug in determining IV based value of the iteration for which the loop body is executed. Fix it.	2009-03-30 22:24:10 +00:00
LoopRotate	Fix PR3408 by making a non-obvious assumption very obvious, and	2009-01-26 02:11:30 +00:00
LoopSimplify
LoopStrengthReduce	Implement "superhero" strength reduction, or full strength	2009-02-20 04:17:46 +00:00
LoopUnroll	remove a test that depends on -debug.	2009-03-13 20:31:48 +00:00
LoopUnswitch	Actually test something. Use PR3170 test case.	2008-12-08 23:44:46 +00:00
LowerInvoke
LowerSetJmp
LowerSwitch
Mem2Reg	no need to print output	2008-10-27 06:56:35 +00:00
MemCpyOpt
MergeFunc	Fix tests not to emit IR output	2008-11-04 23:02:39 +00:00
PredicateSimplifier
PruneEH	Do not inline functions with (dynamic) alloca into	2009-01-08 21:45:23 +00:00
RaiseAllocations
Reassociate
ScalarRepl	Fix a bug.	2009-03-18 07:56:13 +00:00
SCCP	adjust for asmprinter change.	2009-03-01 00:25:46 +00:00
SimplifyCFG	While hoisting instruction to speculatively execute simple bb, ignore dbg intrinsics.	2009-03-06 06:00:17 +00:00
SimplifyLibCalls	Let the strcat optimizer return the pointer to the start of the buffer,	2009-04-06 13:06:48 +00:00
SRETPromotion
StripSymbols
TailCallElim
TailDup