2006-03-20 |
Chris Lattner | add vsplat instructions, fix sched description for...
|
commit | commitdiff | tree |
2006-03-20 |
Chris Lattner | Custom lower arbitrary VECTOR_SHUFFLE's to VPERM.
|
commit | commitdiff | tree |
2006-03-20 |
Chris Lattner | Claim to have v16i8 for perm masks
|
commit | commitdiff | tree |
2006-03-20 |
Chris Lattner | If a target supports splatting with SHUFFLE_VECTOR...
|
commit | commitdiff | tree |
2006-03-20 |
Chris Lattner | TargetData doesn't know the alignment of vectors :(
|
commit | commitdiff | tree |
2006-03-20 |
Chris Lattner | add the vperm instruction
|
commit | commitdiff | tree |
2006-03-20 |
Chris Lattner | Add some helper methods
|
commit | commitdiff | tree |
2006-03-20 |
Chris Lattner | Add another helper
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | Add very basic support for VECTOR_SHUFFLE
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | add a node
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | Add a helper method
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | add a note with a testcase
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | Add a note about the MUL -> FMADD vector bug.
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | Teach cee to propagate through switch statements. ...
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | new testcase, contributed by Eric Kidd.
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | Custom lower SCALAR_TO_VECTOR into lve*x.
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | Allow SCALAR_TO_VECTOR to be custom lowered.
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | Add SCALAR_TO_VECTOR support
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | PPC doesn't have SCALAR_TO_VECTOR
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | add support for vector undef
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | add a new testcase. This insertelement should be a...
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | Don't bother storing undef elements of BUILD_VECTOR's
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | minor fixes
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | notes
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | improve comments, add a new MVT::getVectorBaseType...
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | add another testcase, explicitly check stuff works...
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | we don't use lmw/stmw. When we want them they are...
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | Implement expand of BUILD_VECTOR containing variable...
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | fold insertelement(buildvector) -> buildvector if the...
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | add two new insert_element tests
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | implement basic support for INSERT_VECTOR_ELT.
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | rename these nodes
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | Rename ConstantVec -> BUILD_VECTOR and VConstant -...
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | Rename ConstantVec -> BUILD_VECTOR and VConstant -...
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | implement vector.ll:test_undef
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | Add three new testcases
|
commit | commitdiff | tree |
2006-03-19 |
Chris Lattner | Fix the remaining bugs in the vector expansion rework...
|
commit | commitdiff | tree |
2006-03-18 |
Chris Lattner | update testcases for x86 fastcc changes.
|
commit | commitdiff | tree |
2006-03-18 |
Chris Lattner | Change the structure of lowering vector stuff. Note...
|
commit | commitdiff | tree |
2006-03-18 |
Chris Lattner | Update comments.
|
commit | commitdiff | tree |
2006-03-18 |
Chris Lattner | Fix miscodegen of V_SET0 in PPC.
|
commit | commitdiff | tree |
2006-03-17 |
Chris Lattner | new testcase
|
commit | commitdiff | tree |
2006-03-17 |
Chris Lattner | add a couple enum values
|
commit | commitdiff | tree |
2006-03-17 |
Chris Lattner | add a couple of enum values
|
commit | commitdiff | tree |
2006-03-17 |
Chris Lattner | new testcase that broke the new f.e.
|
commit | commitdiff | tree |
2006-03-17 |
Chris Lattner | Disable x86 fastcc from passing args in registers
|
commit | commitdiff | tree |
2006-03-17 |
Chris Lattner | Parameterize the number of integer arguments to pass...
|
commit | commitdiff | tree |
2006-03-16 |
Chris Lattner | remove dead variable
|
commit | commitdiff | tree |
2006-03-16 |
Chris Lattner | add an assert to get a slightly better msg about this...
|
commit | commitdiff | tree |
2006-03-16 |
Chris Lattner | Fix a problem fully scalarizing values.
|
commit | commitdiff | tree |
2006-03-16 |
Chris Lattner | Strangely, calls clobber call-clobbered vector regs...
|
commit | commitdiff | tree |
2006-03-16 |
Chris Lattner | add a note
|
commit | commitdiff | tree |
2006-03-16 |
Chris Lattner | teach the ppc backend how to spill/reload vector regs
|
commit | commitdiff | tree |
2006-03-16 |
Chris Lattner | add callee saved vector regs
|
commit | commitdiff | tree |
2006-03-16 |
Chris Lattner | in functions that use a lot of callee saved regs, this...
|
commit | commitdiff | tree |
2006-03-16 |
Chris Lattner | Add support for copying registers. still needed: spilling...
|
commit | commitdiff | tree |
2006-03-16 |
Chris Lattner | Add support for CopyFromReg from vector values. Note...
|
commit | commitdiff | tree |
2006-03-16 |
Chris Lattner | Teach CreateRegForValue how to handle vector types.
|
commit | commitdiff | tree |
2006-03-16 |
Chris Lattner | set TransformToType correctly for vector types.
|
commit | commitdiff | tree |
2006-03-16 |
Chris Lattner | Add markers for the first and last vector value type.
|
commit | commitdiff | tree |
2006-03-16 |
Chris Lattner | New testcase, the new CFE compiles this into insertelement...
|
commit | commitdiff | tree |
2006-03-16 |
Chris Lattner | Save/restore VRSAVE once per function, not once per...
|
commit | commitdiff | tree |
2006-03-16 |
Chris Lattner | Fix a typo Eric van Riet Paap noticed
|
commit | commitdiff | tree |
2006-03-16 |
Chris Lattner | add support for the bitconvert node
|
commit | commitdiff | tree |
2006-03-15 |
Chris Lattner | add support for vector->vector casts
|
commit | commitdiff | tree |
2006-03-15 |
Chris Lattner | Add a note, this code should be moved to the dag combiner.
|
commit | commitdiff | tree |
2006-03-15 |
Chris Lattner | minor changes so that GCC builtin can be specified...
|
commit | commitdiff | tree |
2006-03-15 |
Chris Lattner | Teach the strip pass to strip type names in addition...
|
commit | commitdiff | tree |
2006-03-15 |
Chris Lattner | allow the GCCBuiltinName field to be optional
|
commit | commitdiff | tree |
2006-03-15 |
Chris Lattner | remove typo
|
commit | commitdiff | tree |
2006-03-15 |
Chris Lattner | Autogenerate a table of intrinsic names, so we can...
|
commit | commitdiff | tree |
2006-03-15 |
Chris Lattner | Autogenerate code to map from GCC builtin to LLVM intrinsic.
|
commit | commitdiff | tree |
2006-03-15 |
Chris Lattner | Set TargetPrefix on target-specific intrinsics. That...
|
commit | commitdiff | tree |
2006-03-14 |
Chris Lattner | Update this to match the documentation in the GC doc...
|
commit | commitdiff | tree |
2006-03-14 |
Chris Lattner | Fix the gcread/gcwrite intrinsic specifications, fixing
|
commit | commitdiff | tree |
2006-03-14 |
Chris Lattner | Bugfix, unbreaking CodeGen/PowerPC/cttz.ll
|
commit | commitdiff | tree |
2006-03-14 |
Chris Lattner | Fix an et-forest memory leak. Patch by Daniel Berlin.
|
commit | commitdiff | tree |
2006-03-14 |
Chris Lattner | add a note
|
commit | commitdiff | tree |
2006-03-14 |
Chris Lattner | Fix an off by one error that caused PPC LLC failures...
|
commit | commitdiff | tree |
2006-03-14 |
Chris Lattner | Implement a FIXME, recusively reassociating
|
commit | commitdiff | tree |
2006-03-14 |
Chris Lattner | new testcase from a FIXME in the code
|
commit | commitdiff | tree |
2006-03-14 |
Chris Lattner | extract some code into a method, no functionality change
|
commit | commitdiff | tree |
2006-03-14 |
Chris Lattner | transformation implemented
|
commit | commitdiff | tree |
2006-03-14 |
Chris Lattner | Promote shifts by a constant to multiplies so that...
|
commit | commitdiff | tree |
2006-03-14 |
Chris Lattner | Test that we can reassociate (x<<1)+(y<<1) -> (X+Y...
|
commit | commitdiff | tree |
2006-03-13 |
Chris Lattner | Handle builtins that directly correspond to GCC builtins.
|
commit | commitdiff | tree |
2006-03-13 |
Chris Lattner | emit a mapping from LLVM intrinsic -> GCC builtins.
|
commit | commitdiff | tree |
2006-03-13 |
Chris Lattner | Verify that packed type operands have the right size...
|
commit | commitdiff | tree |
2006-03-13 |
Chris Lattner | Add a first ppc altivec intrinsic. Add packed type...
|
commit | commitdiff | tree |
2006-03-13 |
Chris Lattner | For functions that use vector registers, save VRSAVE...
|
commit | commitdiff | tree |
2006-03-13 |
Chris Lattner | make sure dead token factor nodes are removed by the...
|
commit | commitdiff | tree |
2006-03-13 |
Chris Lattner | remove two implemented items
|
commit | commitdiff | tree |
2006-03-13 |
Chris Lattner | Fold X+Y -> X|Y when safe. This implements:
|
commit | commitdiff | tree |
2006-03-13 |
Chris Lattner | new testcase
|
commit | commitdiff | tree |
2006-03-13 |
Chris Lattner | I can't convince myself that this is safe, remove the...
|
commit | commitdiff | tree |
2006-03-13 |
Chris Lattner | add a couple of missing folds
|
commit | commitdiff | tree |
2006-03-13 |
Chris Lattner | For targets with FABS/FNEG support, lower copysign...
|
commit | commitdiff | tree |
2006-03-13 |
Chris Lattner | Fix a couple of bugs that broke the alpha tester build
|
commit | commitdiff | tree |
2006-03-13 |
Chris Lattner | Handle cracked instructions in dispatch group formation.
|
commit | commitdiff | tree |
2006-03-13 |
Chris Lattner | Mark instructions that are cracked by the PPC970 decoder...
|
commit | commitdiff | tree |
next |