aboutsummaryrefslogtreecommitdiff
path: root/lib/Transforms/IPO/GlobalOpt.cpp
AgeCommit message (Expand)Author
2004-10-14Fix a bug John tracked down in libstdc++ where we were incorrectly deletingChris Lattner
2004-10-11This patch implements two things (sorry).Chris Lattner
2004-10-10Just because we cannot completely eliminate all uses of a global, we canChris Lattner
2004-10-10Fix 2004-10-10-CastStoreOnce.llx, by adjusting types back if we strip off a castChris Lattner
2004-10-10Implement GlobalOpt/deadglobal-2.llx, deletion of globals that are onlyChris Lattner
2004-10-10Avoid calling use_size() which could (in theory) be expensive if the globalChris Lattner
2004-10-09Eliminate global pointers that are only stored a single value and null ifChris Lattner
2004-10-09Fix infinite loop due to iterationChris Lattner
2004-10-08If we found a dead global, we should at least delete it...Chris Lattner
2004-10-08* Pull out the meat of runOnModule into another function for clarity.Chris Lattner
2004-10-08We might as well delete the known-dead global sooner rather than later sinceChris Lattner
2004-10-08Implement SRA for global variables. This allows the other global variableChris Lattner
2004-10-07Improve comments, no functionality changesChris Lattner
2004-10-07Fix a bug in the safety analysis routineChris Lattner
2004-10-07* Rename pass to globalopt, since we do more than just constifyChris Lattner
2004-10-06Implement GlobalConstifier/trivialstore.llx, and also do someChris Lattner
2004-09-20'Pass' should now not be derived from by clients. Instead, they should deriveChris Lattner
2004-09-01Changes For Bug 352Reid Spencer
2004-08-14Implement test/Regression/Transforms/GlobalConstifier/phi-select.llxChris Lattner
2004-07-20Ignore instructions that are in trivially dead functions. This allows usChris Lattner
2004-07-18Strip out and simplify some code. This also fixes the regression lastChris Lattner
2004-07-18bug 122:Reid Spencer
2004-02-27setcond instructions don't have aliasing implications.Chris Lattner
2004-02-25My faith in programmers has been found to be totally misplaced. One wouldChris Lattner
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2026-02-01 15:43:11 +0000