summaryrefslogtreecommitdiffstats
path: root/llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp
Commit message (Expand)AuthorAgeFilesLines
* fix a bad typeoChris Lattner2005-05-121-2/+2
* rename the ADJCALLSTACKDOWN/ADJCALLSTACKUP nodes to be CALLSEQ_START/BEGIN.Chris Lattner2005-05-121-6/+6
* handle a common case generated by the uint64 -> FP code path betterChris Lattner2005-05-121-1/+17
* Make legalize a bit more efficient, and canonicalize sub X, C -> add X, -CChris Lattner2005-05-121-6/+7
* Do not memoize ADJCALLSTACKDOWN nodes, provide a method to hack on them.Chris Lattner2005-05-111-3/+24
* Print bit count nodes correctlyChris Lattner2005-05-111-0/+6
* Silence some VC++ warningsJeff Cohen2005-05-101-2/+2
* Add support for READPORT, WRITEPORT, READIO, WRITEIOChris Lattner2005-05-091-1/+10
* Fold shifts into subsequent SHL's. These shifts often arise due to addrsesChris Lattner2005-05-091-0/+28
* Don't use the load/store instruction as the source pointer, use the pointerChris Lattner2005-05-091-0/+2
* memoize all nodes, even null Value* nodes. Do not add two token chain outputsChris Lattner2005-05-091-7/+5
* Print SrcValue nodes correctlyChris Lattner2005-05-091-0/+6
* Implement count leading zeros (ctlz), count trailing zeros (cttz), and countAndrew Lenharth2005-05-031-3/+6
* Add FSQRT, FSIN, FCOS nodes, patch contributed by Morten OfstadChris Lattner2005-04-281-1/+4
* Implement Value* tracking for loads and stores in the selection DAG. This en...Andrew Lenharth2005-04-271-9/+33
* Fold (X > -1) | (Y > -1) --> (X&Y > -1)Chris Lattner2005-04-261-1/+3
* implement some more logical compares with constants, so that:Chris Lattner2005-04-251-7/+20
* Codegen x < 0 | y < 0 as (x|y) < 0. This allows us to compile this to:Chris Lattner2005-04-251-1/+4
* Convert tabs to spacesMisha Brukman2005-04-221-6/+4
* Remove trailing whitespaceMisha Brukman2005-04-211-31/+31
* Improve and elimination. On PPC, for:Chris Lattner2005-04-211-6/+26
* Fold (x & 8) != 0 and (x & 8) == 8 into (x & 8) >> 3.Chris Lattner2005-04-211-0/+22
* Fold setcc of MVT::i1 operands into logical operationsChris Lattner2005-04-181-0/+39
* Another minor simplification: handle setcc (zero_extend x), c -> setcc(x, c')Chris Lattner2005-04-181-0/+45
* Another simple xformChris Lattner2005-04-181-0/+8
* Fold:Chris Lattner2005-04-181-0/+11
* Make the AND elimination operation recursive and significantly more powerful,Chris Lattner2005-04-181-26/+57
* Add a couple missing transforms in getSetCC that were triggering assertionsNate Begeman2005-04-141-1/+8
* Disbale the broken fold of shift + sz[ext] for nowNate Begeman2005-04-131-7/+30
* fix an infinite loopChris Lattner2005-04-131-1/+1
* fix some serious miscompiles on ia64, alpha, and ppcChris Lattner2005-04-131-1/+1
* avoid work when possible, perhaps fix the problem nate and andrew are seeingChris Lattner2005-04-131-0/+1
* add back the optimization that Nate added for shl X, (zext_inreg y)Chris Lattner2005-04-131-2/+23
* Oops, remove these too.Chris Lattner2005-04-131-6/+2
* Remove all foldings of ZERO_EXTEND_INREG, moving them to work for AND nodesChris Lattner2005-04-131-41/+46
* Fold shift x, [sz]ext(y) -> shift x, yNate Begeman2005-04-121-0/+16
* Fold shift by size larger than type size to undefNate Begeman2005-04-121-11/+4
* Remove some redundant checks, add a couple of new ones. This allows us toChris Lattner2005-04-121-7/+5
* canonicalize x <u 1 -> x == 0. On this testcase:Chris Lattner2005-04-121-0/+9
* Don't bother sign/zext_inreg'ing the result of an and operation if we knowChris Lattner2005-04-101-0/+19
* Fold zext_inreg(zextload), likewise for sext'sChris Lattner2005-04-101-0/+6
* add a simple xformChris Lattner2005-04-101-0/+6
* Fix a thinko. If the operand is promoted, pass the promoted value intoChris Lattner2005-04-101-0/+4
* add a little peephole optimization. This allows us to codegen:Chris Lattner2005-04-091-0/+11
* recognize some patterns as fabs operations, so that fabs at the source levelChris Lattner2005-04-091-0/+21
* print and fold BRCONDTWOWAY correctlyChris Lattner2005-04-091-11/+25
* canonicalize a bunch of operations involving fnegChris Lattner2005-04-091-0/+21
* If a target zero or sign extends the result of its setcc, allow folding ofChris Lattner2005-04-071-1/+13
* Remove somethign I had for testingChris Lattner2005-04-071-1/+1
* This patch does two things. First, it canonicalizes 'X >= C' -> 'X > C-1'Chris Lattner2005-04-071-7/+49
OpenPOWER on IntegriCloud