2005-03-26 |
Chris Lattner | Cache mapping information for a call site after computi... |
tree | commitdiff |
2005-03-26 |
Chris Lattner | Remove more long dead code: dsa doesn't provide must... |
tree | commitdiff |
2005-03-26 |
Chris Lattner | remove some unsafe code that has long been dead |
tree | commitdiff |
2005-03-26 |
Chris Lattner | slightly improve mod/ref for DSAA by checking the globa... |
tree | commitdiff |
2005-03-26 |
Chris Lattner | Teach steens-aa two things about mod/ref information: |
tree | commitdiff |
2005-03-26 |
Chris Lattner | Interchange this loop so that we test all pointers... |
tree | commitdiff |
2005-03-25 |
Chris Lattner | no really, don't double count these nodes either! |
tree | commitdiff |
2005-03-25 |
Chris Lattner | Don't count all of the nodes in the SCC once for each... |
tree | commitdiff |
2005-03-25 |
Chris Lattner | Grow the EQ classes for globals at the end of the BU... |
tree | commitdiff |
2005-03-25 |
Chris Lattner | Treat free operations as volatile, since they cannot... |
tree | commitdiff |
2005-03-25 |
Chris Lattner | remove a debugging timer. |
tree | commitdiff |
2005-03-25 |
Chris Lattner | Two changes here: |
tree | commitdiff |
2005-03-25 |
Chris Lattner | Make the spliceFrom case where one graph is completely... |
tree | commitdiff |
2005-03-24 |
Chris Lattner | add a new DSGraph::spliceFrom method, which violently... |
tree | commitdiff |
2005-03-24 |
Chris Lattner | This replaces the correct but slow code with a more... |
tree | commitdiff |
2005-03-24 |
Chris Lattner | only look at successors of globals. This gets us down... |
tree | commitdiff |
2005-03-24 |
Chris Lattner | Unfortunately, a previous patch was not safe. Revert... |
tree | commitdiff |
2005-03-24 |
Chris Lattner | don't bother |'ing in 0's |
tree | commitdiff |
2005-03-24 |
Chris Lattner | be more aggressive about incompleteness marking |
tree | commitdiff |
2005-03-24 |
Chris Lattner | Fix a crash while promoting a value out of a loop from... |
tree | commitdiff |
2005-03-24 |
Chris Lattner | teach ds-aa about mod/ref for external function calls. |
tree | commitdiff |
2005-03-24 |
Chris Lattner | Simplify dead code into a fixme :) |
tree | commitdiff |
2005-03-24 |
Chris Lattner | fix a compiler crash in runtime/libprofile |
tree | commitdiff |
2005-03-23 |
Chris Lattner | wrap a long line |
tree | commitdiff |
2005-03-23 |
Chris Lattner | If we are calling an external function, chain to anothe... |
tree | commitdiff |
2005-03-23 |
Chris Lattner | Make this more efficient by only making one virtual... |
tree | commitdiff |
2005-03-23 |
Chris Lattner | Make this a bit more aggressive |
tree | commitdiff |
2005-03-23 |
Chris Lattner | a hack to allow count-aa to work with ds-aa :( |
tree | commitdiff |
2005-03-23 |
Chris Lattner | Add two options to allow -count-aa to print queries... |
tree | commitdiff |
2005-03-23 |
Chris Lattner | turn a dead conditional into an assert. |
tree | commitdiff |
2005-03-23 |
Chris Lattner | Totally gut mergeInGraph. There is absolutely no reaso... |
tree | commitdiff |
2005-03-23 |
Chris Lattner | wrap a long line |
tree | commitdiff |
2005-03-23 |
Chris Lattner | Make -steens-aa more conservative (aka correct) by... |
tree | commitdiff |
2005-03-23 |
Chris Lattner | implement Analysis/DSGraph/field-sensitive.ll |
tree | commitdiff |
2005-03-22 |
Chris Lattner | Several changes here: |
tree | commitdiff |
2005-03-22 |
Chris Lattner | Mark external globals incomplete in the BU Globals... |
tree | commitdiff |
2005-03-22 |
Chris Lattner | Directly count the number of memory instructions. |
tree | commitdiff |
2005-03-22 |
Chris Lattner | Remove an iteration pass over the entire scalarmap... |
tree | commitdiff |
2005-03-22 |
Chris Lattner | When making a clone of a DSGraph from the BU pass,... |
tree | commitdiff |
2005-03-22 |
Chris Lattner | Don't use operator[], use the new method instead, which... |
tree | commitdiff |
2005-03-22 |
Chris Lattner | Now that the dead ctor is gone, nothing uses the old... |
tree | commitdiff |
2005-03-22 |
Chris Lattner | remove a dead ctor |
tree | commitdiff |
2005-03-22 |
Chris Lattner | now that the second argument is always this->ReturnNode... |
tree | commitdiff |
2005-03-22 |
Chris Lattner | instead of using a local RetValMap, just use the graph... |
tree | commitdiff |
2005-03-22 |
Chris Lattner | now that the valuemapping is always the local scalar... |
tree | commitdiff |
2005-03-22 |
Chris Lattner | remove the second argument to cloneInto |
tree | commitdiff |
2005-03-22 |
Chris Lattner | add some timers, don't clone aux nodes |
tree | commitdiff |
2005-03-22 |
Chris Lattner | move this out of line |
tree | commitdiff |
2005-03-22 |
Chris Lattner | don't generate temporary scalarmaps. |
tree | commitdiff |
2005-03-21 |
Chris Lattner | allow passing clone flags down to cloneInto |
tree | commitdiff |
2005-03-21 |
Chris Lattner | Enhance the TD pass to build composite graphs when... |
tree | commitdiff |
2005-03-21 |
Chris Lattner | make this const correct |
tree | commitdiff |
2005-03-21 |
Chris Lattner | Fix a major problem where we didn't add call graph... |
tree | commitdiff |
2005-03-21 |
Chris Lattner | Ugh, for some reason, I can't call this unless the... |
tree | commitdiff |
2005-03-21 |
Chris Lattner | The reachability cloner should add arguments to merged... |
tree | commitdiff |
2005-03-21 |
Chris Lattner | 'note to self' |
tree | commitdiff |
2005-03-21 |
Chris Lattner | Change the way that the TD pass inlines graphs. Instea... |
tree | commitdiff |
2005-03-20 |
Chris Lattner | Don't strip modref bits when inlining down the call... |
tree | commitdiff |
2005-03-20 |
Chris Lattner | Remove the ability to keep track of inlined globals... |
tree | commitdiff |
2005-03-20 |
Chris Lattner | This call is always a noop, remove it. |
tree | commitdiff |
2005-03-20 |
Chris Lattner | remove some pointless asserts |
tree | commitdiff |
2005-03-20 |
Chris Lattner | -steens doesn't use the inlined globals facility. |
tree | commitdiff |
2005-03-20 |
Chris Lattner | Build EC's for globals twice. The first is after const... |
tree | commitdiff |
2005-03-20 |
Chris Lattner | Add a new DSNode::removeGlobal method. |
tree | commitdiff |
2005-03-20 |
Chris Lattner | Transform BU pass to not use the horrible DSCallSiteIte... |
tree | commitdiff |
2005-03-20 |
Chris Lattner | Implement new methods for dealing with DSNode global... |
tree | commitdiff |
2005-03-20 |
Chris Lattner | use simpler methods. |
tree | commitdiff |
2005-03-20 |
Chris Lattner | method was renamed. |
tree | commitdiff |
2005-03-20 |
Chris Lattner | print out equiv class info. |
tree | commitdiff |
2005-03-20 |
Chris Lattner | some methods got renamed |
tree | commitdiff |
2005-03-20 |
Chris Lattner | Make use of simpler DSNode methods. |
tree | commitdiff |
2005-03-20 |
Chris Lattner | #ifdef out a function only used by #ifdef'd code. |
tree | commitdiff |
2005-03-20 |
Chris Lattner | comment cleanup |
tree | commitdiff |
2005-03-19 |
Chris Lattner | Create an equivalence class of global variables that... |
tree | commitdiff |
2005-03-19 |
Chris Lattner | Switch to use the new interface for the EquivalenceClas... |
tree | commitdiff |
2005-03-18 |
Chris Lattner | do not bother inlining nullary functions without return... |
tree | commitdiff |
2005-03-18 |
Chris Lattner | fix a bogus assertion |
tree | commitdiff |
2005-03-18 |
Chris Lattner | another fastpath |
tree | commitdiff |
2005-03-18 |
Chris Lattner | expose this pass to both opt and analyze |
tree | commitdiff |
2005-03-18 |
Chris Lattner | remove a bogus optimization. This only works if there... |
tree | commitdiff |
2005-03-18 |
Chris Lattner | Rewrite DSAA::getModRefInfo to compute the mapping... |
tree | commitdiff |
2005-03-17 |
Chris Lattner | implement a new method. |
tree | commitdiff |
2005-03-17 |
Chris Lattner | add some possibly bogus assertions. |
tree | commitdiff |
2005-03-17 |
Chris Lattner | Do not include the Function* for direct call/invoke... |
tree | commitdiff |
2005-03-17 |
Chris Lattner | simplify this function a bit, allow DS-AA to build... |
tree | commitdiff |
2005-03-17 |
Chris Lattner | Clean up some code, handle null pointer specially to... |
tree | commitdiff |
2005-03-17 |
Chris Lattner | Two changes: |
tree | commitdiff |
2005-03-17 |
Chris Lattner | Fix the missing symbols problem Bill was hitting. ... |
tree | commitdiff |
2005-03-17 |
Chris Lattner | Do not create ridiculously huge DSNodes, as described... |
tree | commitdiff |
2005-03-16 |
Chris Lattner | remove use of compat_iterator. |
tree | commitdiff |
2005-03-15 |
Chris Lattner | make sure to mark nodes in the globals graph incomplete... |
tree | commitdiff |
2005-03-15 |
Chris Lattner | fix crashes when we only have a prototype for main. |
tree | commitdiff |
2005-03-15 |
Chris Lattner | Fix a crash that happens when mapping something like... |
tree | commitdiff |
2005-03-15 |
Chris Lattner | Make computeGGToGMapping compute an invnodemap |
tree | commitdiff |
2005-03-15 |
Chris Lattner | Finally fix (the right way) the problem where functions... |
tree | commitdiff |
2005-03-15 |
Chris Lattner | Start using retnodes_* for iteration. |
tree | commitdiff |
2005-03-15 |
Chris Lattner | avoid varialbe name collisions |
tree | commitdiff |
2005-03-15 |
Chris Lattner | This mega patch converts us from using Function::a... |
tree | commitdiff |
2005-03-15 |
Chris Lattner | Don't crash if computing a mapping to a node with zero... |
tree | commitdiff |
2005-03-15 |
Chris Lattner | rename method, add counterpart |
tree | commitdiff |
next |