Graham Yiu 1f93c854d8 The cost of splitting a large vector instruction is not being taken into account by the getUserCost function. This was leading to some loops being over unrolled. The cost of a vector instruction is now being multiplied by the cost of the type legalization. This will return a more accurate cost. ...

Committing on behalf on Brad Nemanich (brad.nemanich@ibm.com)

Differential Revision: https://reviews.llvm.org/D38961

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@316174 91177308-0d34-0410-b5e6-96231b3b80d8

2017-10-19 18:16:31 +00:00

..

a2-high-cost-trip-count-computation.ll

…

a2-unrolling.ll

[LoopUnroll] Fix a PowerPC test broken by r277524.

2016-08-02 21:43:25 +00:00

lit.local.cfg

…

p7-unrolling.ll

…

p8-unrolling-legalize-vectors.ll

The cost of splitting a large vector instruction is not being taken into account by the getUserCost function. This was leading to some loops being over unrolled. The cost of a vector instruction is now being multiplied by the cost of the type legalization. This will return a more accurate cost.

2017-10-19 18:16:31 +00:00