summaryrefslogtreecommitdiffstats
path: root/llvm/lib/Support/APFloat.cpp
Commit message (Collapse)AuthorAgeFilesLines
...
* Fast-track obviously over-large and over-small exponents during decimal->Neil Booth2007-10-151-8/+44
| | | | | | | | | | integer conversion. In some such cases this makes us one or two orders of magnitude faster than NetBSD's libc. Glibc seems to have a similar fast path. Also, tighten up some upper bounds to save a bit of memory. llvm-svn: 42984
* Consolidate logic for creating NaNs. Silence compiler warning.Neil Booth2007-10-141-10/+15
| | | | llvm-svn: 42966
* Whether arithmetic is supported is a property of the semantics. Make itNeil Booth2007-10-141-41/+36
| | | | | | so, and clean up the checks by putting them in an inline function. llvm-svn: 42965
* Separate out parsing of decimal number. Use this to only allocateNeil Booth2007-10-141-97/+150
| | | | | | | memory for the significand once up-front. Also ignore insignificant trailing zeroes; this saves unnecessary multiplications later. llvm-svn: 42964
* If the power of 5 is exact, and the reciprocal exact, the error is zero not ↵Neil Booth2007-10-131-1/+1
| | | | | | one half-ulps. This prevents an infinite loop in rare cases. llvm-svn: 42950
* Remove duplicate comment.Neil Booth2007-10-121-11/+0
| | | | llvm-svn: 42913
* Implement correctly-rounded decimal->binary conversion, i.e. conversionNeil Booth2007-10-121-6/+349
| | | | | | | | | | | from user input strings. Such conversions are more intricate and subtle than they may appear; it is unlikely I have got it completely right first time. I would appreciate being informed of any bugs and incorrect roundings you might discover. llvm-svn: 42912
* Remove a field that was never used.Neil Booth2007-10-121-9/+6
| | | | llvm-svn: 42911
* If we're trying to be arbitrary precision, unsigned char clearly won't cut ↵Neil Booth2007-10-121-1/+1
| | | | | | it. Needed for dec->bin conversions. llvm-svn: 42910
* Next PPC long double bits. First cut at constants.Dale Johannesen2007-10-111-6/+157
| | | | | | | | | No compile-time support for constant operations yet, just format transformations. Make readers and writers work. Split constants into 2 doubles in Legalize. llvm-svn: 42865
* Use APInt::tcExtract. It's cleaner, and works :)Neil Booth2007-10-081-19/+14
| | | | llvm-svn: 42746
* Add back convertFromSignExtendedInteger.Neil Booth2007-10-071-0/+30
| | | | llvm-svn: 42735
* Now that convertFromUnsignedParts has a sane, constant interface,Neil Booth2007-10-071-5/+1
| | | | | | | convertFromZeroExtendedInteger can be simplified as it doesn't need to make a copy of the source bignum. llvm-svn: 42734
* Reimplement convertFromUnsignedInteger so it is passed a const bignum.Neil Booth2007-10-071-18/+29
| | | | | | | | | It used to modify its argument in-place. This interface is saner and the implementation more efficient. It will be needed for decimal->binary conversion. llvm-svn: 42733
* convertFromInteger, as originally written, expected sign-extendedNeil Booth2007-10-071-4/+6
| | | | | | | | input. APInt unfortunately zero-extends signed integers, so Dale modified the function to expect zero-extended input. Make this assumption explicit in the function name. llvm-svn: 42732
* combineLostFractions does not need to be a member functionNeil Booth2007-10-071-16/+15
| | | | llvm-svn: 42729
* Cleaner, more general exponent output.Neil Booth2007-10-061-21/+22
| | | | llvm-svn: 42690
* silence warnings in no-assert build.Chris Lattner2007-10-061-23/+20
| | | | llvm-svn: 42687
* Generalize tcFullMultiply so that the operands can be of differingNeil Booth2007-10-061-3/+4
| | | | | | | | part widths. Also, return the number of parts actually required to hold the result's value. Remove an over-cautious condition from rounding of float->hex conversion. llvm-svn: 42669
* Add APFloat -> hexadecimal string conversion, as per %a and %A in C99.Neil Booth2007-10-031-3/+240
| | | | | | Useful for diagnostics and debugging. llvm-svn: 42598
* Tweak RoundAwayFromZero the bit number below which is truncated, and makeNeil Booth2007-10-031-8/+13
| | | | | | | | it const. Preparation for APFloat -> hexadecimal string conversion. llvm-svn: 42576
* Simplify and fix signed int -> FP conversions.Dale Johannesen2007-09-301-11/+3
| | | | llvm-svn: 42483
* Whitespace and compiler warning cleanup.Neil Booth2007-09-261-76/+91
| | | | llvm-svn: 42373
* Make APFloat->int conversions deterministic even inDale Johannesen2007-09-251-5/+33
| | | | | | cases with undefined behavior. llvm-svn: 42328
* Fix long double<->shorter FP type conversionsDale Johannesen2007-09-251-12/+29
| | | | | | of zero, infinity, and NaNs. llvm-svn: 42298
* Handle storage complications of float->float conversions.Neil Booth2007-09-221-18/+23
| | | | llvm-svn: 42220
* Change APFloat::convertFromInteger to take the incomingDale Johannesen2007-09-211-16/+19
| | | | | | | | | | bit width instead of number of words allocated, which makes it actually work for int->APF conversions. Adjust callers. Add const to one of the APInt constructors to prevent surprising match when called with const argument. llvm-svn: 42210
* Fix PR 1688. See comments there.Dale Johannesen2007-09-201-13/+25
| | | | llvm-svn: 42181
* Remove spurious consts. This fixes warnings with compilers thatDan Gohman2007-09-141-2/+2
| | | | | | are strict about such things. llvm-svn: 41956
* Compensate for partCount change in Bogus definitionDale Johannesen2007-09-121-1/+1
| | | | | | (could break hash table in ConstantFP) llvm-svn: 41874
* Add APInt interfaces to APFloat (allows directlyDale Johannesen2007-09-111-15/+149
| | | | | | | | | access to bits). Use them in place of float and double interfaces where appropriate. First bits of x86 long double constants handling (untested, probably does not work). llvm-svn: 41858
* Fix mod so it actually works. Fix conversions toDale Johannesen2007-09-051-9/+33
| | | | | | native types to handle denormals correctly. llvm-svn: 41726
* Oops, should be part of 41664; won't work very well without this piece.Dale Johannesen2007-08-311-0/+38
| | | | llvm-svn: 41665
* Enhance APFloat to retain bits of NaNs (fixes oggenc).Dale Johannesen2007-08-311-77/+76
| | | | | | | Use APFloat interfaces for more references, mostly of ConstantFPSDNode. llvm-svn: 41632
* Change LegalFPImmediates to use APFloat.Dale Johannesen2007-08-301-1/+1
| | | | | | | | | Add APFloat interfaces to ConstantFP, SelectionDAG. Fix integer bit in double->APFloat conversion. Convert LegalizeDAG to use APFloat interface in ConstantFPSDNode uses. llvm-svn: 41587
* Poison APFloat::operator==. Replace existing uses with bitwiseIsEqual.Dale Johannesen2007-08-241-1/+1
| | | | | | This means backing out the preceding change to Constants.cpp, alas. llvm-svn: 41378
* Revised per review feedback from previous patch.Dale Johannesen2007-08-241-36/+15
| | | | llvm-svn: 41353
* silence some warnings in an optimized build.Chris Lattner2007-08-241-11/+8
| | | | llvm-svn: 41352
* Change internal representation of ConstantFP to use APFloat.Dale Johannesen2007-08-241-1/+187
| | | | | | Interface to rest of the compiler unchanged, as yet. llvm-svn: 41348
* initial checkin of Neil's APFloat work.Chris Lattner2007-08-201-0/+1488
llvm-svn: 41203
OpenPOWER on IntegriCloud