- 30 Nov, 2014 2 commits
-
-
Christoph Mallon authored
-
Christoph Mallon authored
ir: Use mode_is_float()/mode_is_int() instead of get_mode_sort()/->sort == irms_float_number/irms_int_number.
-
- 29 Nov, 2014 5 commits
-
-
Matthias Braun authored
-
Matthias Braun authored
-
Matthias Braun authored
There are still similar problems left.
-
Matthias Braun authored
-
yb9976 authored
-
- 28 Nov, 2014 7 commits
-
-
Christoph Mallon authored
-
Christoph Mallon authored
-
Christoph Mallon authored
-
Christoph Mallon authored
-
Christoph Mallon authored
-
Christoph Mallon authored
-
Christoph Mallon authored
-
- 27 Nov, 2014 13 commits
-
-
Christoph Mallon authored
* This corrects the requirements of fStore.
-
Christoph Mallon authored
-
Christoph Mallon authored
-
Christoph Mallon authored
This properly models the data dependencies and avoids keep edges.
-
Christoph Mallon authored
-
Christoph Mallon authored
-
Christoph Mallon authored
Determine the pos from the node to push.
-
Christoph Mallon authored
-
Christoph Mallon authored
-
Christoph Mallon authored
-
Christoph Mallon authored
This replaces several 1/0 and true/false by yes/no.
-
Christoph Mallon authored
Many x87 nodes do not need the x87-specific attributes.
-
yb9976 authored
-
- 26 Nov, 2014 5 commits
-
-
yb9976 authored
-
yb9976 authored
-
yb9976 authored
-
Matthias Braun authored
Some instructions liks shifts or out/in can only handle some forms of immediates. Make matchers respect that. This fixes backend/ia32_shl.c, although the real question here is probably why that program wasn't folded to a point where the middleend would reduce the shift amount anyway because of the modulo shift mode.
-
Matthias Braun authored
-
- 25 Nov, 2014 5 commits
-
-
Matthias Braun authored
-
Matthias Braun authored
-
Matthias Braun authored
-
Matthias Braun authored
-
Matthias Braun authored
Add callbacks on functions that really need them instead.
-
- 23 Nov, 2014 3 commits
-
-
Christoph Mallon authored
-
Christoph Mallon authored
-
Christoph Mallon authored
When creating an entity for a floatingpoint constant, first try to reduce the size, then check, whether it already exists. This way only a single entity is created for e.g. a double and a float 1.0.
-