x86: add op class for int and fp instructions
Review Request #1866 - Created May 11, 2013 and submitted - Latest diff uploaded
| Information | |
|---|---|
| Nilay Vaish | |
| gem5 | |
| default | |
| Reviewers | |
| Default | |
Changeset 9686:4f4352a352e2 --------------------------- x86: add op class for int and fp instructions Currently all the integer instructions are marked as IntAluOp and the floating point instructions are marked as FpAddOp. This patch adds support for marking different instructions differently. Now IntMultOp, IntDivOp, FloatDivOp, FloatMultOp, FloatCvtOp, FloatSqrtOp classes will be used as well. This will help in providing different latencies for different op class.