diff options
| author | Goran Jakovljevic <Goran.Jakovljevic@imgtec.com> | 2017-04-07 11:18:59 +0200 |
|---|---|---|
| committer | Goran Jakovljevic <Goran.Jakovljevic@imgtec.com> | 2017-04-07 11:18:59 +0200 |
| commit | 716d0738c0ed8f924a44e5c645fa89c1a6079c95 (patch) | |
| tree | eb16f7564328d18cf51b415d771a45e58616a8f2 /compiler/optimizing/graph_visualizer.cc | |
| parent | f040074c2b9ca348196c176e11491f603a151006 (diff) | |
MIPS64: Fix Math.ceil/floor intrinsics
The latest MIPS64R6 emulator supports NAN2008 standard. When
FCSR.NAN2008=1 (required from R6) ceil.l.d and floor.l.d instructions
will return -2^63 when the input value is smaller than -2^63. This
wasn't the case when FCSR.NAN2008=0. In that case, result from the
instructions was 2^63-1 when the input value was smaller than -2^63.
This fixes test 082-inline-execute.
Test: ./testrunner.py --target --optimizing --64 -j1 (in QEMU)
Change-Id: I3d2cdd1d55b3694549f88f4f5b903a0c4f1c0d09
Diffstat (limited to 'compiler/optimizing/graph_visualizer.cc')
0 files changed, 0 insertions, 0 deletions