diff options
| author | buzbee <buzbee@google.com> | 2013-04-12 14:39:29 -0700 |
|---|---|---|
| committer | buzbee <buzbee@google.com> | 2013-05-13 12:34:28 -0700 |
| commit | a5abf7091711eed1e9f1d0e1538fe9963ebdf31c (patch) | |
| tree | e256df83ca632744d144854403a326d90cb683a7 /src/compiler/dex/mir_dataflow.cc | |
| parent | bf47e5f28b1aa39748dce8ac5abbabca1baee093 (diff) | |
| download | android_art-a5abf7091711eed1e9f1d0e1538fe9963ebdf31c.tar.gz android_art-a5abf7091711eed1e9f1d0e1538fe9963ebdf31c.tar.bz2 android_art-a5abf7091711eed1e9f1d0e1538fe9963ebdf31c.zip | |
Compiler: replace DOM traversal computation
Originally the old trace JIT used a few recursive graph walking
algorithms - which was perfectly reasonable given that the graph
size was capped at a few dozen nodes at most. These were replaced
with iterative walk order computations - or at least I thought
they all were. Missed one of them, which caused a stack overflow
on a pathologically large method compilation.
Renaming of some arena_allocator items for consistency and clarity.
More detailed memory usage logging. Reworked the allocator to waste
less space when an allocation doesn't fit and a new block must be
allocated.
Change-Id: I4d84dded3c47819eefa0de90ebb821dd12eb8be8
Diffstat (limited to 'src/compiler/dex/mir_dataflow.cc')
| -rw-r--r-- | src/compiler/dex/mir_dataflow.cc | 7 |
1 files changed, 0 insertions, 7 deletions
diff --git a/src/compiler/dex/mir_dataflow.cc b/src/compiler/dex/mir_dataflow.cc index 444874dce5..9f61d73d6b 100644 --- a/src/compiler/dex/mir_dataflow.cc +++ b/src/compiler/dex/mir_dataflow.cc @@ -1202,13 +1202,6 @@ void MIRGraph::CompilerInitializeSSAConversion() } } -/* Clear the visited flag for each BB */ -bool MIRGraph::ClearVisitedFlag(struct BasicBlock* bb) -{ - bb->visited = false; - return true; -} - /* * This function will make a best guess at whether the invoke will * end up using Method*. It isn't critical to get it exactly right, |