Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | On Sandybridge split unaligned 256bit stores into two xmm-sized stores. | Nadav Rotem | 2013-01-19 | 1 | -2/+2 |
| | | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@172894 91177308-0d34-0410-b5e6-96231b3b80d8 | ||||
* | Specify CPU model to avoid breaking ATOM builds | Michael Liao | 2012-10-10 | 1 | -2/+2 |
| | | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165638 91177308-0d34-0410-b5e6-96231b3b80d8 | ||||
* | Add support for FP_ROUND from v2f64 to v2f32 | Michael Liao | 2012-10-10 | 1 | -0/+61 |
- Due to the current matching vector elements constraints in ISD::FP_ROUND, rounding from v2f64 to v4f32 (after legalization from v2f32) is scalarized. Add a customized v2f32 widening to convert it into a target-specific X86ISD::VFPROUND to work around this constraints. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165631 91177308-0d34-0410-b5e6-96231b3b80d8 |