maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/Transforms
History
Hao Liu 32c0539691 [LoopVectorize] Teach Loop Vectorizor about interleaved memory accesses. 
Interleaved memory accesses are grouped and vectorized into vector load/store and shufflevector.
E.g. for (i = 0; i < N; i+=2) {
       a = A[i];         // load of even element
       b = A[i+1];       // load of odd element
       ...               // operations on a, b, c, d
       A[i] = c;         // store of even element
       A[i+1] = d;       // store of odd element
     }

  The loads of even and odd elements are identified as an interleave load group, which will be transfered into vectorized IRs like:
     %wide.vec = load <8 x i32>, <8 x i32>* %ptr
     %vec.even = shufflevector <8 x i32> %wide.vec, <8 x i32> undef, <4 x i32> <i32 0, i32 2, i32 4, i32 6>
     %vec.odd = shufflevector <8 x i32> %wide.vec, <8 x i32> undef, <4 x i32> <i32 1, i32 3, i32 5, i32 7>

  The stores of even and odd elements are identified as an interleave store group, which will be transfered into vectorized IRs like:
     %interleaved.vec = shufflevector <4 x i32> %vec.even, %vec.odd, <8 x i32> <i32 0, i32 4, i32 1, i32 5, i32 2, i32 6, i32 3, i32 7> 
     store <8 x i32> %interleaved.vec, <8 x i32>* %ptr

This optimization is currently disabled by defaut. To try it by adding '-enable-interleaved-mem-accesses=true'. 

llvm-svn: 239291
 		2015-06-08 06:39:56 +00:00
..
ADCE [opaque pointer type] Add textual IR support for explicit type parameter to load instruction 2015-02-27 21:17:42 +00:00
AddDiscriminators IR: Give 'DI' prefix to debug info metadata 2015-04-29 16:38:44 +00:00
AlignmentFromAssumptions [opaque pointer type] Add textual IR support for explicit type parameter to load instruction 2015-02-27 21:17:42 +00:00
ArgumentPromotion IR: Give 'DI' prefix to debug info metadata 2015-04-29 16:38:44 +00:00
AtomicExpand/ARM Use target-dependent emitLeading/TrailingFence instead of the target-independent insertLeading/TrailingFence (in AtomicExpandPass) 2014-09-03 21:01:03 +00:00
BBVectorize Changed renaming of local symbols by inserting a dot vefore the numeric suffix. 2015-05-12 16:47:30 +00:00
BDCE [BDCE] Don't forget uses of root instructions seen before the instruction itself 2015-02-18 03:12:28 +00:00
BranchFolding Delete -std-compile-opts. 2014-10-16 20:00:02 +00:00
CodeExtractor [opaque pointer type] Add textual IR support for explicit type parameter to load instruction 2015-02-27 21:17:42 +00:00
CodeGenPrepare Forgot to add lit.local.cfg for new R600 directory 2015-05-26 17:01:16 +00:00
ConstProp [ConstantFolding] Fix wrong folding of intrinsic 'convert.from.fp16'. 2015-05-14 18:01:48 +00:00
ConstantHoisting [opaque pointer type] Add textual IR support for explicit type parameter to the call instruction 2015-04-16 23:24:18 +00:00
ConstantMerge [opaque pointer type] Add textual IR support for explicit type parameter to gep operator 2015-03-13 18:20:45 +00:00
CorrelatedValuePropagation [CVP] Don't assume Constants of type i1 can be known to be true or false 2015-06-06 04:56:51 +00:00
DeadArgElim IR: Give 'DI' prefix to debug info metadata 2015-04-29 16:38:44 +00:00
DeadStoreElimination IR: Give 'DI' prefix to debug info metadata 2015-04-29 16:38:44 +00:00
EarlyCSE Correct a mistaken comment from 238071 [NFC] 2015-05-23 00:05:43 +00:00
Float2Int Reapply r233175 and r233183: float2int. 2015-03-27 10:36:57 +00:00
FunctionAttrs Remove conflicting attributes before adding deduced readonly/readnone 2015-05-25 19:46:38 +00:00
GCOVProfiling IR: Give 'DI' prefix to debug info metadata 2015-04-29 16:38:44 +00:00
GVN [PHITransAddr] Don't translate unreachable values 2015-06-01 00:15:08 +00:00
GlobalDCE [opaque pointer type] Add textual IR support for explicit type parameter to the call instruction 2015-04-16 23:24:18 +00:00
GlobalOpt IR: Give 'DI' prefix to debug info metadata 2015-04-29 16:38:44 +00:00
IPConstantProp Skip extra LLVM IR assemble/disassemble steps in some tests 2015-04-24 18:06:09 +00:00
IRCE [IRCE] Fix a regression introduced in r232444. 2015-03-24 19:29:18 +00:00
IndVarSimplify Enable exitValue rewrite only when the cost of expansion is low. 2015-05-28 21:49:07 +00:00
Inline [inliner] Fix the early-exit of the inline cost analysis to correctly 2015-05-27 02:49:05 +00:00
InstCombine [InstCombine] Don't miscompile select to poison 2015-06-06 02:30:43 +00:00
InstMerge [opaque pointer type] Add textual IR support for explicit type parameter to gep operator 2015-03-13 18:20:45 +00:00
InstSimplify [InstSimplify] Handle some overflow intrinsics in InstSimplify 2015-05-22 03:56:46 +00:00
Internalize [opaque pointer type] Add textual IR support for explicit type parameter to load instruction 2015-02-27 21:17:42 +00:00
JumpThreading [JumpThreading] Simplify comparisons when simplifying branches 2015-05-07 00:19:14 +00:00
LCSSA [opaque pointer type] Add textual IR support for explicit type parameter to the invoke instruction 2015-04-24 19:32:54 +00:00
LICM [LICM] Sinking doesn't involve the preheader 2015-05-22 02:14:05 +00:00
LoadCombine [opaque pointer type] Add textual IR support for explicit type parameter to load instruction 2015-02-27 21:17:42 +00:00
LoopDeletion [opaque pointer type] Add textual IR support for explicit type parameter to load instruction 2015-02-27 21:17:42 +00:00
LoopDistribute [LoopAccesses] Rearrange printed lines in -analyze 2015-05-18 15:36:57 +00:00
LoopIdiom Changed renaming of local symbols by inserting a dot vefore the numeric suffix. 2015-05-12 16:47:30 +00:00
LoopInterchange Add support to interchange loops with reductions. 2015-04-23 04:51:44 +00:00
LoopReroll [opaque pointer type] Add textual IR support for explicit type parameter to load instruction 2015-02-27 21:17:42 +00:00
LoopRotate IR: Give 'DI' prefix to debug info metadata 2015-04-29 16:38:44 +00:00
LoopSimplify Verifier: Call verifyModule() from llc and opt 2015-03-27 22:04:28 +00:00
LoopStrengthReduce Make the test introduced in r239015 more targeted. 2015-06-04 07:21:42 +00:00
LoopUnroll [LoopUnroll] Fix truncation bug in canUnrollCompletely. 2015-06-06 05:24:10 +00:00
LoopUnswitch Changed renaming of local symbols by inserting a dot vefore the numeric suffix. 2015-05-12 16:47:30 +00:00
LoopVectorize [LoopVectorize] Teach Loop Vectorizor about interleaved memory accesses. 2015-06-08 06:39:56 +00:00
LowerAtomic [opaque pointer type] Add textual IR support for explicit type parameter to load instruction 2015-02-27 21:17:42 +00:00
LowerBitSets Changed renaming of local symbols by inserting a dot vefore the numeric suffix. 2015-05-12 16:47:30 +00:00
LowerExpectIntrinsic [opaque pointer type] Add textual IR support for explicit type parameter to the call instruction 2015-04-16 23:24:18 +00:00
LowerInvoke Verifier: Call verifyModule() from llc and opt 2015-03-27 22:04:28 +00:00
LowerSwitch Changed renaming of local symbols by inserting a dot vefore the numeric suffix. 2015-05-12 16:47:30 +00:00
Mem2Reg IR: Give 'DI' prefix to debug info metadata 2015-04-29 16:38:44 +00:00
MemCpyOpt [MemCpyOpt] Do move the memset, but look at its dest's dependencies. 2015-05-21 01:43:39 +00:00
MergeFunc MergeFunctions: Two different sized allocas are *not* the same 2015-05-12 21:42:22 +00:00
MetaRenamer [opaque pointer type] Add textual IR support for explicit type parameter to load instruction 2015-02-27 21:17:42 +00:00
NaryReassociate [NaryReassociate] Run EarlyCSE after NaryReassociate 2015-05-28 04:56:52 +00:00
ObjCARC IR: Give 'DI' prefix to debug info metadata 2015-04-29 16:38:44 +00:00
PartiallyInlineLibCalls PartiallyInlineLibCalls: Check sqrt result type before transforming it. 2014-08-01 23:21:21 +00:00
PhaseOrdering [opaque pointer type] Add textual IR support for explicit type parameter to the call instruction 2015-04-16 23:24:18 +00:00
PlaceSafepoints Add a GCStrategy for CoreCLR 2015-05-20 01:07:23 +00:00
PruneEH Fix invalid LLVM IR in PruneEH tests 2015-02-11 02:06:47 +00:00
Reassociate [Reassociate] Canonicalizing 'x [+-] (-Constant * y)' isn't always a win 2015-05-28 06:16:39 +00:00
Reg2Mem
…
RewriteStatepointsForGC [RewriteStatepointsForGC] Strip deref info after rewriting. 2015-06-02 22:33:37 +00:00
SCCP [opaque pointer type] Add textual IR support for explicit type parameter to the call instruction 2015-04-16 23:24:18 +00:00
SLPVectorizer Let llc and opt override "-target-cpu" and "-target-features" via command line 2015-05-06 23:54:14 +00:00
SROA [ConstantFold] Don't fold ppc_fp128 <-> int bitcasts 2015-03-28 16:44:57 +00:00
SampleProfile Add function entry counts from sample profiles. 2015-05-13 17:04:29 +00:00
ScalarRepl IR: Give 'DI' prefix to debug info metadata 2015-04-29 16:38:44 +00:00
Scalarizer IR: Give 'DI' prefix to debug info metadata 2015-04-29 16:38:44 +00:00
SeparateConstOffsetFromGEP SeparateConstOffsetFromGEP: Pass address space to isLegalAddressingMode 2015-06-07 20:17:44 +00:00
SimplifyCFG Changed renaming of local symbols by inserting a dot vefore the numeric suffix. 2015-05-12 16:47:30 +00:00
Sink Teach the IR Sink pass to (conservatively) respect convergent annotations. 2015-06-01 17:20:31 +00:00
SpeculativeExecution Add a speculative execution pass 2015-05-15 17:54:48 +00:00
StraightLineStrengthReduce [SLSR] handle (B | i) * S 2015-05-15 17:07:48 +00:00
StripSymbols IR: Give 'DI' prefix to debug info metadata 2015-04-29 16:38:44 +00:00
StructurizeCFG [opaque pointer type] Add textual IR support for explicit type parameter to load instruction 2015-02-27 21:17:42 +00:00
TailCallElim TRE: Just erase dead BBs and tweak the iteration loop not to increment the deleted BB iterator. 2015-02-28 16:47:27 +00:00
TailDup [opaque pointer type] Add textual IR support for explicit type parameter to the call instruction 2015-04-16 23:24:18 +00:00
Util Changed renaming of local symbols by inserting a dot vefore the numeric suffix. 2015-05-12 16:47:30 +00:00