Changeset 183718 in webkit
- Timestamp:
- May 2, 2015 10:33:36 AM (9 years ago)
- Location:
- trunk/Source/JavaScriptCore
- Files:
-
- 3 edited
Legend:
- Unmodified
- Added
- Removed
-
trunk/Source/JavaScriptCore/ChangeLog
r183715 r183718 1 2015-05-02 Filip Pizlo <fpizlo@apple.com> 2 3 Unreviewed, add FIXMEs referencing https://bugs.webkit.org/show_bug.cgi?id=144524 and 4 https://bugs.webkit.org/show_bug.cgi?id=144525. 5 6 * dfg/DFGLICMPhase.cpp: 7 (JSC::DFG::LICMPhase::attemptHoist): 8 * dfg/DFGPhantomInsertionPhase.cpp: 9 1 10 2015-05-02 Yusuke Suzuki <utatane.tea@gmail.com> 2 11 -
trunk/Source/JavaScriptCore/dfg/DFGLICMPhase.cpp
r183497 r183718 214 214 } 215 215 216 // FIXME: At this point if the hoisting of the full node fails but the node has type checks, 217 // we could still hoist just the checks. 218 // https://bugs.webkit.org/show_bug.cgi?id=144525 219 216 220 if (readsOverlap(m_graph, node, data.writes)) { 217 221 if (verbose) { -
trunk/Source/JavaScriptCore/dfg/DFGPhantomInsertionPhase.cpp
r183406 r183718 82 82 void handleBlock(BasicBlock* block) 83 83 { 84 // FIXME: For blocks that have low register pressure, it would make the most sense to 85 // simply insert Phantoms at the last point possible since that would obviate the need to 86 // query bytecode liveness: 87 // 88 // - If we MovHint @x into loc42 then put a Phantom on the last MovHinted value in loc42. 89 // - At the end of the block put Phantoms for each MovHinted value. 90 // 91 // This will definitely not work if there are any phantom allocations. For those blocks 92 // where this would be legal, it remains to be seen how profitable it would be even if there 93 // was high register pressure. After all, a Phantom would cause a spill but it wouldn't 94 // cause a fill. 95 // 96 // https://bugs.webkit.org/show_bug.cgi?id=144524 97 84 98 m_values.fill(nullptr); 85 99
Note: See TracChangeset
for help on using the changeset viewer.