- 13 Jul, 2012 8 commits
-
-
Christoph Mallon authored
-
Christoph Mallon authored
-
Christoph Mallon authored
-
Christoph Mallon authored
-
Matthias Braun authored
-
Matthias Braun authored
Also add a GRAPH_PROPERTY for the analysis state.
-
Matthias Braun authored
-
yb9976 authored
-
- 10 Jul, 2012 3 commits
-
-
Matthias Braun authored
-
Matthias Braun authored
- Setting dep edges to NULL is illegal now - add_irn_dep is simpler but does not check for duplicates anymore
-
Matthias Braun authored
-
- 28 Jun, 2012 1 commit
-
-
Matthias Braun authored
Get away from the strange set_default_XXX_op style and instead let the code modules register their node ops directly.
-
- 21 Jun, 2012 2 commits
-
-
Matthias Braun authored
This also makes the compiler alot more deterministic avoiding, lots of "random" jumping around in series of stores and loads.
-
Matthias Braun authored
I expect this to be a common pattern
-
- 19 Jun, 2012 1 commit
-
-
Christoph Mallon authored
-
- 15 Jun, 2012 4 commits
-
-
Matthias Braun authored
The opt_manage framework didn't really do much except assuring that some analysis info is computed before a pass starts. We can achieve this simpler by providing an assure_irg_properties function. This commit also fixes some places where analysis info was potentially not invalidated before.
-
Matthias Braun authored
Noone was really using the return values anyway. This change makes the API more consistent.
-
Matthias Braun authored
This makes it consistent with other firm analyse functions. Also removes the unnecessary return value.
-
Matthias Braun authored
graph_properties are things that a transformation may destroy but which can be easily recalculated. graph_constraints are set after lowering phases have been running and forbid certain constructs that no (later) transformation is allowed to construct
-
- 14 Jun, 2012 1 commit
-
-
Christoph Mallon authored
-
- 13 Jun, 2012 3 commits
-
-
Matthias Braun authored
-
Matthias Braun authored
First it is unclear what is a good way to deconstruct a program into extended basic blocks, also none of our optimizations works on extended basic blocks.
-
Matthias Braun authored
-
- 16 May, 2012 1 commit
-
-
Matthias Braun authored
-
- 14 May, 2012 2 commits
-
-
Christoph Mallon authored
It probably was true when the start block had itself as predecessor.
-
Christoph Mallon authored
-
- 13 May, 2012 1 commit
-
-
Christoph Mallon authored
-
- 11 May, 2012 1 commit
-
-
uqxyd authored
-
- 10 May, 2012 1 commit
-
-
Matthias Braun authored
This was used as a hack to mark unreachable code by some people, but it is prefered to simply use a new Block without predecessors, or set NULL if you have to.
-
- 04 May, 2012 1 commit
-
-
Matthias Braun authored
This commit removes the strange differentiation between ir_visibility_external and ir_visibility_default. We now only have ir_visibility_external for all symbols visible across compilation units. You may or may not attach graphs/initializers to them.
-
- 26 Apr, 2012 1 commit
-
-
Matthias Braun authored
(This happens when you break on a Sel creation and the firm debugger tries to tell you about the created node)
-
- 25 Mar, 2012 1 commit
-
-
Christoph Mallon authored
-
- 13 Mar, 2012 1 commit
-
-
Christoph Mallon authored
-
- 06 Mar, 2012 1 commit
-
-
Christoph Mallon authored
-
- 01 Mar, 2012 2 commits
-
-
Matthias Braun authored
- The direct call did not respect the get_opt_constant_folding() flag
-
Matthias Braun authored
-
- 27 Feb, 2012 2 commits
-
-
Christoph Mallon authored
-
Christoph Mallon authored
-
- 09 Feb, 2012 2 commits
-
-
Matthias Braun authored
(otherwise CSE merges temporary Phi nodes without inputs)
-
Matthias Braun authored
-