Nadav Rotem 0b37f14371 LoopVectorizer: Fix a bug in the code that updates the loop exiting block. ...

LCSSA PHIs may have undef values. The vectorizer updates values that are used by outside users such as PHIs.
The bug happened because undefs are not loop values. This patch handles these PHIs.

PR14725

llvm-svn: 171251

2012-12-30 07:47:00 +00:00

..

X86

Fix typo "Makre" -> "Make".

2012-12-24 19:55:47 +00:00

12-12-11-if-conv.ll

PR14574. Fix a bug in the code that calculates the mask the converted PHIs in if-conversion.

2012-12-11 21:30:14 +00:00

2012-10-20-infloop.ll

LoopVectorizer: Fix an endless loop in the code that looks for reductions.

2012-12-24 01:22:06 +00:00

2012-10-22-isconsec.ll

Move the target-specific tests, which require specific backends, to dirs that only run if the target is present.

2012-10-26 18:52:01 +00:00

calloc.ll

If all of the write objects are identified then we can vectorize the loop even if the read objects are unidentified.

2012-12-26 23:30:53 +00:00

cast-induction.ll

Loop Vectorize: optimize the vectorization of trunc(induction_var). The truncation is now done on scalars.

2012-12-11 18:58:10 +00:00

cpp-new-array.ll

Loop Vectorize: optimize the vectorization of trunc(induction_var). The truncation is now done on scalars.

2012-12-11 18:58:10 +00:00

flags.ll

LoopVectorize: Preserve NSW, NUW and IsExact flags.

2012-10-31 21:40:39 +00:00

gcc-examples.ll

LoopVectorizer: Optimize the vectorization of consecutive memory access when the iteration step is -1

2012-12-26 19:08:17 +00:00

if-conv-crash.ll

Fix PR14565. Don't if-convert loops that have switch statements in them.

2012-12-11 04:55:10 +00:00

if-conversion-reduction.ll

Fix a bug in vectorization of if-converted reduction variables. If the

2012-12-04 22:40:22 +00:00

if-conversion.ll

Add support for reduction variables when IF-conversion is enabled.

2012-12-04 18:17:33 +00:00

increment.ll

LoopVectorize: Preserve NSW, NUW and IsExact flags.

2012-10-31 21:40:39 +00:00

induction_plus.ll

Loop Vectorize: optimize the vectorization of trunc(induction_var). The truncation is now done on scalars.

2012-12-11 18:58:10 +00:00

intrinsic.ll

LoopVectorize: Enable vectorization of the fmuladd intrinsic

2012-12-25 23:21:29 +00:00

lcssa-crash.ll

LoopVectorizer: Fix a bug in the code that updates the loop exiting block.

2012-12-30 07:47:00 +00:00

lit.local.cfg

…

no_int_induction.ll

Add support for pointer induction variables even when there is no integer induction variable.

2012-11-25 08:41:35 +00:00

non-const-n.ll

LoopVectorize: Preserve NSW, NUW and IsExact flags.

2012-10-31 21:40:39 +00:00

read-only.ll

LoopVectorizer: Add a basic cost model which uses the VTTI interface.

2012-10-24 20:36:32 +00:00

reduction.ll

LoopVectorize: Emit reductions as log2(vectorsize) shuffles + vector ops instead of scalar operations.

2012-12-18 18:40:20 +00:00

runtime-check.ll

Add support for memory runtime check. When we can, we calculate array bounds.

2012-11-09 07:09:44 +00:00

same-base-access.ll

Fix a bug in the code that checks if we can vectorize loops while using dynamic

2012-12-21 00:07:35 +00:00

scalar-select.ll

LoopVectorizer: Add a basic cost model which uses the VTTI interface.

2012-10-24 20:36:32 +00:00

small-loop.ll

LoopVectorize: Do not vectorize loops with tiny constant trip counts.

2012-10-31 03:31:07 +00:00

small-size.ll

LoopVectorizer: Use the "optsize" attribute to decide if we are allowed to increase the function size.

2012-12-12 19:29:45 +00:00

start-non-zero.ll

LoopVectorize: Preserve NSW, NUW and IsExact flags.

2012-10-31 21:40:39 +00:00

write-only.ll

LoopVectorize: Add support for write-only loops when the write destination is a single pointer.

2012-10-30 18:36:45 +00:00