| Commit message (Expand) | Author | Age | Files | Lines |
* | Add some comments. | Chris Lattner | 2006-03-25 | 1 | -0/+4 |
* | add a vnot helper node for matching 'not' on vectors | Chris Lattner | 2006-03-25 | 1 | -2/+6 |
* | Fix a bug in ISD::isBuildVectorAllOnesInteger that caused it to always return | Chris Lattner | 2006-03-25 | 1 | -2/+3 |
* | Implement the ISD::isBuildVectorAllOnesInteger predicate | Chris Lattner | 2006-03-25 | 1 | -1/+32 |
* | Don't call SimplifyDemandedBits on vectors | Chris Lattner | 2006-03-25 | 1 | -1/+2 |
* | Add some logical operations | Chris Lattner | 2006-03-25 | 1 | -3/+19 |
* | Don't crash on packed logical ops | Chris Lattner | 2006-03-25 | 1 | -3/+6 |
* | Teach BinaryOperator::createNot to work with packed integer types | Chris Lattner | 2006-03-25 | 1 | -2/+9 |
* | Cast instruction not inserted into basic block. | Jim Laskey | 2006-03-25 | 1 | -41/+5 |
* | Added missing (any_extend (load ...)) patterns. | Evan Cheng | 2006-03-25 | 1 | -0/+3 |
* | Build arbitrary vector with more than 2 distinct scalar elements with a | Evan Cheng | 2006-03-25 | 3 | -5/+104 |
* | implement a bunch of intrinsics | Chris Lattner | 2006-03-25 | 1 | -3/+34 |
* | Move all Altivec stuff out into a new PPCInstrAltivec.td file. | Chris Lattner | 2006-03-25 | 2 | -239/+298 |
* | Add some basic patterns for other datatypes | Chris Lattner | 2006-03-25 | 2 | -10/+12 |
* | add all supported formats to the vector register file | Chris Lattner | 2006-03-25 | 1 | -1/+1 |
* | Add support for __builtin_altivec_vnmsubfp /vmaddfp | Chris Lattner | 2006-03-25 | 1 | -0/+5 |
* | #include Intrinsics.h into all dag isels | Chris Lattner | 2006-03-25 | 6 | -0/+12 |
* | Implement Intrinsic::getName | Chris Lattner | 2006-03-25 | 1 | -0/+11 |
* | Codegen things like: | Chris Lattner | 2006-03-25 | 4 | -2/+123 |
* | Added SSE cachebility ops | Evan Cheng | 2006-03-25 | 1 | -0/+30 |
* | Instruction encoding bug | Evan Cheng | 2006-03-25 | 1 | -1/+1 |
* | Add new intrinsic node definitions for tblgen use | Chris Lattner | 2006-03-25 | 1 | -0/+12 |
* | Added 128-bit packed integer subtraction. | Evan Cheng | 2006-03-25 | 2 | -0/+26 |
* | Added CVTTPS2PI. | Evan Cheng | 2006-03-25 | 1 | -0/+8 |
* | Added CVTSS2SI. | Evan Cheng | 2006-03-25 | 1 | -0/+5 |
* | Support for scalar to vector with zero extension. | Evan Cheng | 2006-03-24 | 4 | -51/+118 |
* | fix inverted conditional | Chris Lattner | 2006-03-24 | 1 | -2/+2 |
* | D'oh - should be even numbered. | Jim Laskey | 2006-03-24 | 1 | -15/+15 |
* | Added LDMXCSR | Evan Cheng | 2006-03-24 | 1 | -0/+7 |
* | plug the intrinsics into the patterns for movmsk* | Chris Lattner | 2006-03-24 | 1 | -2/+4 |
* | Add dwarf register numbering to register data. | Jim Laskey | 2006-03-24 | 5 | -332/+671 |
* | Add support for dwarf register numbering. | Jim Laskey | 2006-03-24 | 1 | -0/+21 |
* | Hack no more. | Jim Laskey | 2006-03-24 | 1 | -2/+0 |
* | add another note | Chris Lattner | 2006-03-24 | 1 | -0/+15 |
* | add a note | Chris Lattner | 2006-03-24 | 1 | -0/+12 |
* | Shuffle some includes around | Chris Lattner | 2006-03-24 | 2 | -4/+2 |
* | Only to vector shuffle for {x,x,y,y} cases when SCALAR_TO_VECTOR is free. | Evan Cheng | 2006-03-24 | 1 | -1/+2 |
* | expose intrinsic info to the targets. | Chris Lattner | 2006-03-24 | 1 | -0/+3 |
* | Fix a bad JIT encoding of VPERM. Why is VPERM D,A,B,C but vfmadd is D,A,C,B ?? | Chris Lattner | 2006-03-24 | 1 | -2/+2 |
* | Like the comment says, prefer to use the implicit add done by [r+r] addressing | Chris Lattner | 2006-03-24 | 1 | -4/+14 |
* | Fix indent. | Jim Laskey | 2006-03-24 | 1 | -1/+1 |
* | Clean up some commentary. | Jim Laskey | 2006-03-24 | 2 | -2/+20 |
* | Rename for truth in advertising. | Jim Laskey | 2006-03-24 | 1 | -2/+2 |
* | Disable the i32->float G5 optimization. It is unsafe, as documented in the | Chris Lattner | 2006-03-24 | 1 | -1/+7 |
* | add support for using vxor to build zero vectors. This implements | Chris Lattner | 2006-03-24 | 3 | -4/+44 |
* | Handle BUILD_VECTOR with all zero elements. | Evan Cheng | 2006-03-24 | 4 | -11/+73 |
* | prefer to generate constant pool loads over splats. This prevents us from | Chris Lattner | 2006-03-24 | 1 | -34/+33 |
* | fix spello | Chris Lattner | 2006-03-24 | 1 | -1/+1 |
* | Fix spello | Chris Lattner | 2006-03-24 | 1 | -2/+2 |
* | add the actual cost to the debug info | Chris Lattner | 2006-03-24 | 1 | -2/+4 |