summaryrefslogtreecommitdiffstats
path: root/llvm/lib
Commit message (Collapse)AuthorAgeFilesLines
...
* use high-level functions in CCStateChris Lattner2007-02-281-15/+6
| | | | llvm-svn: 34739
* add methods for analysis of call results and return nodes.Chris Lattner2007-02-281-12/+42
| | | | llvm-svn: 34738
* make use of helper functions in CCState for analyzing formals and calls.Chris Lattner2007-02-281-49/+15
| | | | llvm-svn: 34737
* add methods to analyze calls and formals.Chris Lattner2007-02-281-0/+34
| | | | llvm-svn: 34736
* add a newline at end of fileChris Lattner2007-02-281-0/+1
| | | | llvm-svn: 34735
* switch LowerFastCCCallTo over to using the new fastcall description.Chris Lattner2007-02-281-97/+40
| | | | llvm-svn: 34734
* switch LowerFastCCArguments over to using the autogenerated Fastcall ↵Chris Lattner2007-02-281-174/+59
| | | | | | description. llvm-svn: 34733
* add new CC_X86_32_FastCall calling conv, which describes fastcall on win32.Chris Lattner2007-02-281-10/+29
| | | | | | | Factor out a CC_X86_32_Common convention, which is the part shared between ccc, stdcall and fastcall llvm-svn: 34732
* rearrange codeChris Lattner2007-02-281-300/+292
| | | | llvm-svn: 34731
* remove fastcc (not fastcall) supportChris Lattner2007-02-282-58/+26
| | | | llvm-svn: 34730
* switch LowerCCCArguments over to using autogenerated CC.Chris Lattner2007-02-281-94/+62
| | | | llvm-svn: 34729
* simplify sret handlingChris Lattner2007-02-281-13/+5
| | | | llvm-svn: 34728
* switch LowerCCCCallTo over to using an autogenerated callingconvChris Lattner2007-02-282-110/+71
| | | | llvm-svn: 34727
* rename stuffChris Lattner2007-02-281-24/+24
| | | | llvm-svn: 34726
* rename some CCActions, add CCIfInRegChris Lattner2007-02-281-7/+12
| | | | llvm-svn: 34725
* switch return value passing and the x86-64 calling convention informationChris Lattner2007-02-281-137/+9
| | | | | | over to being autogenerated from the X86CallingConv.td file. llvm-svn: 34722
* make subtarget references work.Chris Lattner2007-02-281-3/+7
| | | | llvm-svn: 34721
* Implement reading of arbitrary precision integers.Reid Spencer2007-02-281-4/+9
| | | | llvm-svn: 34718
* Implement writing of arbitrary precision integers.Reid Spencer2007-02-281-4/+14
| | | | llvm-svn: 34717
* Regenerate.Reid Spencer2007-02-285-3362/+4592
| | | | llvm-svn: 34716
* Implement arbitrary integer constants through the use of APInt values.Reid Spencer2007-02-281-13/+45
| | | | | | | Positive, negative, and hexadecimal integer constants will now return an APInt for values having > 64 bits of precision. llvm-svn: 34715
* Implement support for aribrary precision integers by creating two newReid Spencer2007-02-281-2/+41
| | | | | | | | tokens: ESAPINTVAL and EUAPINTVAL and adding an APInt* as a semantic value. This allows us to extend the definition of an integer constant to allow arbitrary precision integer constant values. llvm-svn: 34714
* Fix a bug in RoundDoubleToAPInt where it would force the size to 64 bitsReid Spencer2007-02-281-1/+1
| | | | | | instead of honoring the client's requested bit width. llvm-svn: 34712
* Make requiresRegisterScavenging determination on a per MachineFunction basis.Evan Cheng2007-02-284-12/+14
| | | | llvm-svn: 34711
* Temporary: make R12 available in ARM mode if RegScavenger is being used.Evan Cheng2007-02-281-12/+53
| | | | llvm-svn: 34709
* Start making use of RegScavenger.Evan Cheng2007-02-281-8/+12
| | | | llvm-svn: 34708
* PEI now passes a RegScavenger ptr to eliminateFrameIndex.Evan Cheng2007-02-2811-16/+24
| | | | llvm-svn: 34707
* MRegisterInfo disowns RegScavenger. It's immutable.Evan Cheng2007-02-281-2/+2
| | | | llvm-svn: 34706
* Join two lines that can be joined.Reid Spencer2007-02-271-2/+1
| | | | llvm-svn: 34705
* Pull out the stops.Reid Spencer2007-02-271-4/+2
| | | | llvm-svn: 34703
* Remove some redundancy.Reid Spencer2007-02-271-4/+1
| | | | llvm-svn: 34702
* Minor tweak. Allocate r0 to r3 in reverse order, r3 is least likely to be ↵Evan Cheng2007-02-271-5/+5
| | | | | | livein to a function. llvm-svn: 34701
* Some more code clean up.Evan Cheng2007-02-271-23/+29
| | | | llvm-svn: 34700
* reenable generation of CC infoChris Lattner2007-02-271-1/+1
| | | | llvm-svn: 34699
* Oops.Evan Cheng2007-02-271-1/+4
| | | | llvm-svn: 34698
* Implement countLeadingOnes() and getMinSignedBits(). This helps to minimizeReid Spencer2007-02-271-1/+34
| | | | | | | the bit width of negative numbers by computing the minimum bit width for a negative value. E.g. 0x1800000000000000 could be just 0x8000000000000000 llvm-svn: 34695
* Back out previous commit temporarily.Evan Cheng2007-02-271-1/+1
| | | | llvm-svn: 34694
* Let MRegisterInfo now owns RegScavenger; eliminateFrameIndex must preserve ↵Evan Cheng2007-02-272-30/+51
| | | | | | register kill info. llvm-svn: 34692
* Let MRegisterInfo now owns RegScavenger.Evan Cheng2007-02-271-3/+3
| | | | llvm-svn: 34691
* RegScavenger interface change to make it more flexible.Evan Cheng2007-02-271-16/+18
| | | | llvm-svn: 34690
* implement CCMatchIfCC in terms of CCMatchIfChris Lattner2007-02-271-4/+4
| | | | llvm-svn: 34685
* build cc infoChris Lattner2007-02-271-1/+1
| | | | llvm-svn: 34684
* Adjust to changes in the APInt interface.Reid Spencer2007-02-272-5/+5
| | | | llvm-svn: 34681
* Improve APInt interface:Reid Spencer2007-02-271-34/+1
| | | | | | | | | | 1. Add unsigned and signed versions of methods so a "bool" argument doesn't need to be passed in. 2. Make the various getMin/getMax functions all be inline since they are so simple. 3. Simplify sdiv and srem code. llvm-svn: 34680
* Conform to single person attribution in file comment.Reid Spencer2007-02-271-3/+2
| | | | llvm-svn: 34678
* For PR1205:Reid Spencer2007-02-271-14/+5
| | | | | | | | | | | Implement review feedback: 1. Use new APInt::RoundDoubleToAPInt interface to specify the bit width so that we don't have to truncate or extend in constant folding. 2. Fix a pasteo in SDiv that prevented a check for overflow. 3. Fix the shift operators: undef happens when the shift amount is equal to the bitwidth. llvm-svn: 34677
* 1. Fix three serious bugs in the comparison code.Reid Spencer2007-02-271-18/+22
| | | | | | | 2. Change RoundDoubleToAPInt to take a bit width parameter. Use that parameter to limit the bit width of the result. llvm-svn: 34673
* a noteChris Lattner2007-02-271-0/+42
| | | | llvm-svn: 34670
* Make getPassManagerType() const.Devang Patel2007-02-272-3/+5
| | | | llvm-svn: 34669
* Fix parenthesis for BCTRL_{ELF|Macho} test.Nicolas Geoffray2007-02-271-1/+1
| | | | llvm-svn: 34668
OpenPOWER on IntegriCloud