aboutsummaryrefslogtreecommitdiff
path: root/lib/CodeGen/SelectionDAG
AgeCommit message (Expand)Author
2005-10-10Add ISD::ADD to MaskedValueIsZeroChris Lattner
2005-10-10This function is now deadChris Lattner
2005-10-10Enable Nate's excellent DAG combiner work by default. This allows theChris Lattner
2005-10-09add a todo for something I noticedChris Lattner
2005-10-09(X & Y) & C == 0 if either X&C or Y&C are zeroChris Lattner
2005-10-09When emiting a CopyFromReg and the source is already a vreg, do not botherChris Lattner
2005-10-08Lo and behold, the last bits of SelectionDAG.cpp have been moved over.Nate Begeman
2005-10-07remove debugging codeChris Lattner
2005-10-07implement CodeGen/PowerPC/div-2.ll:test2-4 by propagating zero bits throughChris Lattner
2005-10-07fix indentationChris Lattner
2005-10-07Turn sdivs into udivs when we can prove the sign bits are clear. ThisChris Lattner
2005-10-06silence a bogus GCC warningChris Lattner
2005-10-06Make the legalizer completely non-recursiveChris Lattner
2005-10-05Let the combiner handle more casesNate Begeman
2005-10-05Remove some bad code from LegalizeNate Begeman
2005-10-05Check in some more DAGCombiner piecesNate Begeman
2005-10-05implement visitBR_CC so that PowerPC/inverted-bool-compares.ll passesChris Lattner
2005-10-05fix some pastosChris Lattner
2005-10-05Add a new HandleNode class, which is used to handle (haha) cases in theChris Lattner
2005-10-05Implement the code for PowerPC/inverted-bool-compares.ll, even though itChris Lattner
2005-10-05make sure that -view-isel-dags is the input to the isel, not the input toChris Lattner
2005-10-05Fix a crash compiling Olden/tspChris Lattner
2005-10-04Reverting to version - until problem isolated.Jim Laskey
2005-10-04Fix some faulty logic in the libcall inserter.Nate Begeman
2005-10-04Add back a workaround that fixes some breakages from chris's last change.Nate Begeman
2005-10-03Refactor gathering node info and emission.Jim Laskey
2005-10-02Fix a problem where the legalizer would run out of stack space on extremelyChris Lattner
2005-10-02silence a bogus warningChris Lattner
2005-10-02Add assertions to the trivial scheduler to check that the value types matchChris Lattner
2005-10-02Codegen CopyFromReg using the regclass that matches the valuetype of theChris Lattner
2005-10-01Add some very paranoid checking for operand/result reg class matchupChris Lattner
2005-10-01Fix VC++ warnings.Jeff Cohen
2005-10-01add a methodChris Lattner
2005-10-01typoJim Laskey
2005-10-011. Simplify the gathering of node groups.Jim Laskey
2005-09-301. Made things node-centric (from operand).Jim Laskey
2005-09-29Fix two bugs in my patch earlier today that broke int->fp conversion on X86.Chris Lattner
2005-09-29Silence VC++ redeclaration warnings.Jeff Cohen
2005-09-28Add FP versions of the binary operators, keeping the int and fp worlds seperate.Chris Lattner
2005-09-27If the target prefers it, use _setjmp/_longjmp should be used instead of setj...Chris Lattner
2005-09-27initialize new flagChris Lattner
2005-09-27Remove some redundancies.Jim Laskey
2005-09-26Addition of a simple two pass scheduler. This version is currently hacked upJim Laskey
2005-09-23Turn (X^C1) == C2 into X == C1^C2 iff X&~C1 = 0 (and move a function)Chris Lattner
2005-09-19Stub out the rest of the DAG Combiner. Just need to fill in theNate Begeman
2005-09-16More DAG combining. Still need the branch instructions, and select_ccNate Begeman
2005-09-13If a function has liveins, and if the target requested that they be ploppedChris Lattner
2005-09-10Allow targets to say they don't support truncstore i1 (which includes a maskChris Lattner
2005-09-09Add a missing #include, patch courtesy of Baptiste Lepilleur.Chris Lattner
2005-09-09Fix a problem duraid encountered on itanium where this folding:Chris Lattner