| 2004-02-10 |
Chris Lattner | An initial implementation of an LLVM ProfileInfo class... |
tree | commitdiff |
| 2004-02-10 |
Chris Lattner | Add #include |
tree | commitdiff |
| 2004-02-10 |
Chris Lattner | Do not use MachineOperand::isVirtualRegister either! |
tree | commitdiff |
| 2004-02-10 |
Chris Lattner | Stop using this method |
tree | commitdiff |
| 2004-02-10 |
Chris Lattner | Remove uses of MachineOperand::isVirtualRegister |
tree | commitdiff |
| 2004-02-10 |
Chris Lattner | Remvoe use of MO.isVirtualRegister(), turn an assertion... |
tree | commitdiff |
| 2004-02-10 |
Chris Lattner | Eliminate users of MachineOperand::isPhysicalRegister |
tree | commitdiff |
| 2004-02-10 |
Chris Lattner | Remove use of isPhysicalRegister |
tree | commitdiff |
| 2004-02-10 |
Chris Lattner | Don't use MachineOperator::is(Phys|Virt)Register |
tree | commitdiff |
| 2004-02-10 |
Chris Lattner | Tighten up checks |
tree | commitdiff |
| 2004-02-10 |
Chris Lattner | initialization calls now return argc. If the program... |
tree | commitdiff |
| 2004-02-10 |
Chris Lattner | Fix PR228: [sparc] Boolean constants are emitted as... |
tree | commitdiff |
| 2004-02-09 |
Misha Brukman | Doxygenify comments. |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | Only add the global variable with the abort message... |
tree | commitdiff |
| 2004-02-09 |
John Criswell | Fix PR#226: When emitting padding, always emit it as... |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | It turns out that the two dimensional vectors were... |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | Remove the statistics |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | Speed up type resolution some more. On the testcase... |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | When resolving upreferences, if multiple uprefs will... |
tree | commitdiff |
| 2004-02-09 |
Brian Gaeke | Move InstrSchedule's iterator begin/end methods inline. |
tree | commitdiff |
| 2004-02-09 |
Brian Gaeke | Make SchedGraph::dump() use SchedGraphNodeCommon's... |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | Implement the hashing scheme in an attempt to speed... |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | This debugging hook is no longer needed. |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | Code cleanup in preparation for later changes. Now... |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | Now that all of the derived types have disciplined... |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | Don't depend on auto data conversion |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | Adjust to the changed StructType interface. In particu... |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | Start using the new and improve interface to FunctionTy... |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | This #include is not needed, it should have been remove... |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | Instead of searching the entire type graph for a type... |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | Upreferences are always OpaqueTypes, meaning that it... |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | Another nice speedup for the register allocator. This... |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | Add a new (hidden) option that is useful for profiling. |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | Ugh, perform an optimization that GCC should be able... |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | Only do stuff for the REAL number of physical registers... |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | Change the PhysRegsUsed map into a dense array. Seeing... |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | Two problems with these lines of code: |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | The ConstantExpr::getCast call can cause a CPR to be... |
tree | commitdiff |
| 2004-02-09 |
Chris Lattner | Fix PR215: [bcwriter] Problem compactifying ConstantPoi... |
tree | commitdiff |
| 2004-02-08 |
Misha Brukman | Fix grammar-o. |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | Improve compatibility with programs that already have... |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | vi failed me again. :) |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | Rename the invoke 'except' destination to the 'unwind... |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | Change the 'exception' destination to the 'unwind'... |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | rename the "exceptional" destination of an invoke instr... |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | Fix PR225: [pruneeh] -pruneeh pass removes invoke instr... |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | splitBasicBlock "does the right thing" now, no reason... |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | Implement proper invoke/unwind lowering. |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | Print out all globals as they are emitted, not just... |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | There is no reason to #define fd |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | Add a call to 'write' right before the call to abort... |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | Add one that I missed |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | Instead of callign removeTriviallyDeadNodes on the... |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | Remove another unneeded call. |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | This call is no longer needed now that merging does... |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | Substantially improve the DSA code by removing 'forward... |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | Bugfix for ilist conversion. The ilist wants to make... |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | Switch the Nodes list from being an std::vector<DSNode... |
tree | commitdiff |
| 2004-02-08 |
Chris Lattner | Change to use node_iterators instead of direct access... |
tree | commitdiff |
| 2004-02-07 |
Chris Lattner | getNodes() is gone, use node_begin/end instead |
tree | commitdiff |
| 2004-02-07 |
Chris Lattner | getNodes() is gone |
tree | commitdiff |
| 2004-02-07 |
Chris Lattner | There is no need to clone over nodes that are going... |
tree | commitdiff |
| 2004-02-06 |
Alkis Evlogimenos | Increase code clarity. |
tree | commitdiff |
| 2004-02-06 |
Alkis Evlogimenos | Eliminate uneeded lookups by passing a Virt2PhysMap... |
tree | commitdiff |
| 2004-02-05 |
Chris Lattner | Fix another dominator update bug. These bugs keep... |
tree | commitdiff |
| 2004-02-05 |
Alkis Evlogimenos | Change live interval representation. Machine instructio... |
tree | commitdiff |
| 2004-02-05 |
Chris Lattner | Fix bug updating dominators |
tree | commitdiff |
| 2004-02-05 |
Chris Lattner | Add debug output |
tree | commitdiff |
| 2004-02-05 |
Chris Lattner | Fix PR223: Loopsimplify incorrectly updates dominator... |
tree | commitdiff |
| 2004-02-05 |
Alkis Evlogimenos | We don't need to scan the blocks that we are live-in... |
tree | commitdiff |
| 2004-02-05 |
Chris Lattner | In a "seeing the forest through the trees" kinda situat... |
tree | commitdiff |
| 2004-02-05 |
Chris Lattner | This is a big diff with no functionality change. We... |
tree | commitdiff |
| 2004-02-05 |
Chris Lattner | finegrainify namespacification |
tree | commitdiff |
| 2004-02-05 |
Tanya Lattner | Added missing include. |
tree | commitdiff |
| 2004-02-05 |
Tanya Lattner | Fixed Chris' typo. |
tree | commitdiff |
| 2004-02-05 |
Chris Lattner | Implement optimizations for handling large basic blocks. |
tree | commitdiff |
| 2004-02-04 |
Alkis Evlogimenos | Modify the two address instruction pass to remove the... |
tree | commitdiff |
| 2004-02-04 |
Chris Lattner | Minor speedup, don't query ValueMap each time through... |
tree | commitdiff |
| 2004-02-04 |
Brian Gaeke | Take away the default iostream argument of createMachin... |
tree | commitdiff |
| 2004-02-04 |
Chris Lattner | Two changes: |
tree | commitdiff |
| 2004-02-04 |
Alkis Evlogimenos | IMULri* instructions do not require their first two... |
tree | commitdiff |
| 2004-02-04 |
Chris Lattner | Optimize the case where we are inlining a function... |
tree | commitdiff |
| 2004-02-04 |
Chris Lattner | Adjust to the new BasicBlock ctor, which requires a... |
tree | commitdiff |
| 2004-02-04 |
Chris Lattner | Adjust to the new BB ctor |
tree | commitdiff |
| 2004-02-04 |
Chris Lattner | Remove unneeded code now that splitBasicBlock does... |
tree | commitdiff |
| 2004-02-04 |
Chris Lattner | When splitting a basic block, insert the new half immed... |
tree | commitdiff |
| 2004-02-04 |
Chris Lattner | More refactoring. Move alloca instructions and handle... |
tree | commitdiff |
| 2004-02-04 |
Chris Lattner | Move the cloning of the function body much earlier... |
tree | commitdiff |
| 2004-02-04 |
Chris Lattner | Give CloneBasicBlock an optional function argument... |
tree | commitdiff |
| 2004-02-04 |
Chris Lattner | If changing a parent, don't add then remove the object... |
tree | commitdiff |
| 2004-02-03 |
Chris Lattner | In BasicBlock::splitBasicBlock, just use islist::splice... |
tree | commitdiff |
| 2004-02-03 |
Chris Lattner | Bunch up all locally used allocas by the block they... |
tree | commitdiff |
| 2004-02-03 |
Chris Lattner | Handle extremely trivial cases extremely efficiently... |
tree | commitdiff |
| 2004-02-03 |
Chris Lattner | Generate ftst instructions for comparison against zero |
tree | commitdiff |
| 2004-02-03 |
Chris Lattner | Add the ftst instruction |
tree | commitdiff |
| 2004-02-03 |
Chris Lattner | Add support for one argument OneArgFP instructions |
tree | commitdiff |
| 2004-02-03 |
Alkis Evlogimenos | When an instruction like: A += B had both A and B virtu... |
tree | commitdiff |
| 2004-02-02 |
Alkis Evlogimenos | Revert changes. Will implement this using a different... |
tree | commitdiff |
| 2004-02-02 |
Alkis Evlogimenos | Fix debugging output. |
tree | commitdiff |
| 2004-02-02 |
Alkis Evlogimenos | Correctly update def/use information for modified machi... |
tree | commitdiff |
| next |
projects / oota-llvm.git / history