| commit | 878194421e436322d3f5855135d339d177edf7a3 | [log] [tgz] |
|---|---|---|
| author | Craig Topper <craig.topper@intel.com> | Wed Aug 22 23:27:50 2018 +0000 |
| committer | Craig Topper <craig.topper@intel.com> | Wed Aug 22 23:27:50 2018 +0000 |
| tree | 7db8e92cce149b72a312edef3563268d27ff4a09 | |
| parent | f1d427341c3da5889eddae53c351c832223f35c7 [diff] |
[ValueTracking] Teach computeNumSignBits to understand min/max clamp patterns with constant/splat values If we have a min/max pair we can do a better job of counting sign bits if we look at them together. This is similar to what is done in the SelectionDAG version of computeNumSignBits for ISD::SMAX/SMIN. Differential Revision: https://reviews.llvm.org/D51112 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@340480 91177308-0d34-0410-b5e6-96231b3b80d8