Changes in directory llvm/lib/VMCore:
ConstantFold.cpp updated: 1.149 -> 1.150
---
Log message:
Add a comment to explain a folding transform.
---
Diffs of the changes: (+1 -0)
ConstantFold.cpp |1 +
1 files changed, 1 insertion(+)
Index: llvm/lib/VMCore/ConstantFold.cpp
diff -u llvm/
Changes in directory llvm/lib/VMCore:
ConstantFold.cpp updated: 1.148 -> 1.149
---
Log message:
fold constantexprs more aggressively, fixing PR1265: http://llvm.org/PR1265
---
Diffs of the changes: (+17 -2)
ConstantFold.cpp | 19 +--
1 files changed, 17 insertions(+), 2
Changes in directory llvm/lib/VMCore:
ConstantFold.cpp updated: 1.147 -> 1.148
---
Log message:
Fix constant fold of div by zero and rem by zero to match IEEE 754
requirements. We must return NaN in some cases and correctly signed
infinity in other cases. Passes CFP2006 (not that that says muc
Changes in directory llvm/lib/VMCore:
ConstantFold.cpp updated: 1.146 -> 1.147
---
Log message:
Fix constant folding of fp->int bitcast for vectors.
---
Diffs of the changes: (+2 -2)
ConstantFold.cpp |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
Index: llvm/lib/VMCore/Con
Changes in directory llvm/lib/VMCore:
ConstantFold.cpp updated: 1.145 -> 1.146
---
Log message:
Prefer non-virtual calls to ConstantInt::isZero over virtual calls to
Constant::isNullValue() in situations where it is possible.
---
Diffs of the changes: (+6 -6)
ConstantFold.cpp | 12 +
Changes in directory llvm/lib/VMCore:
ConstantFold.cpp updated: 1.144 -> 1.145
---
Log message:
Use the APInt versions of the bit-wise conversions of float/double to int
and back. While this is not strictly necessary, it does pave the way for
future changes in casting. It should now be possible
Changes in directory llvm/lib/VMCore:
ConstantFold.cpp updated: 1.143 -> 1.144
---
Log message:
Use a simpler constructor when constructing ConstantInst.
---
Diffs of the changes: (+19 -20)
ConstantFold.cpp | 39 +++
1 files changed, 19 insertions(+),
On Feb 27, 2007, at 3:35 PM, Reid Spencer wrote:
> On Tue, 2007-02-27 at 14:17 -0800, Chris Lattner wrote:
This seems like a poor choice of API. RoundDoubleToAPInt should
take
a result bitwidth. Without this, this appears like it could make a
"large" APInt, then truncate it
On Tue, 2007-02-27 at 14:17 -0800, Chris Lattner wrote:
> >> This seems like a poor choice of API. RoundDoubleToAPInt should take
> >> a result bitwidth. Without this, this appears like it could make a
> >> "large" APInt, then truncate it down to a small one when converting
> >> from double, whic
Changes in directory llvm/lib/VMCore:
ConstantFold.cpp updated: 1.142 -> 1.143
---
Log message:
Pull out the stops.
---
Diffs of the changes: (+2 -4)
ConstantFold.cpp |6 ++
1 files changed, 2 insertions(+), 4 deletions(-)
Index: llvm/lib/VMCore/ConstantFold.cpp
diff -u llvm/lib/
>> This seems like a poor choice of API. RoundDoubleToAPInt should take
>> a result bitwidth. Without this, this appears like it could make a
>> "large" APInt, then truncate it down to a small one when converting
>> from double, which is very inefficient.
>
> Done.
Thanks, but why the default ar
Changes in directory llvm/lib/VMCore:
ConstantFold.cpp updated: 1.140 -> 1.141
---
Log message:
For PR1205: http://llvm.org/PR1205 :
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 fold
On Tue, 2007-02-27 at 09:08 -0800, Chris Lattner wrote:
> On Feb 26, 2007, at 10:24 PM, Reid Spencer wrote:
>
> >case Instruction::FPToUI:
> > -if (const ConstantFP *FPC = dyn_cast(V))
> > - return ConstantInt::get(DestTy,(uint64_t) FPC->getValue());
> > +if (const ConstantFP *FPC
On Feb 26, 2007, at 10:24 PM, Reid Spencer wrote:
>case Instruction::FPToUI:
> -if (const ConstantFP *FPC = dyn_cast(V))
> - return ConstantInt::get(DestTy,(uint64_t) FPC->getValue());
> +if (const ConstantFP *FPC = dyn_cast(V)) {
> + APInt Val(APIntOps::RoundDoubleToAPInt(FP
Changes in directory llvm/lib/VMCore:
ConstantFold.cpp updated: 1.139 -> 1.140
---
Log message:
For PR1205: http://llvm.org/PR1205 :
Implement constant folding via APInt instead of uint64_t.
---
Diffs of the changes: (+135 -102)
ConstantFold.cpp | 237 +++--
Changes in directory llvm/lib/VMCore:
ConstantFold.cpp updated: 1.138 -> 1.139
Constants.cpp updated: 1.221 -> 1.222
ConstantFolding.h (r1.53) removed
---
Log message:
ConstantFolding.cpp/h -> ConstantFold.cpp/h to avoid confusion with
lib/Analysis/ConstantFolding.
---
Diffs of the changes:
16 matches
Mail list logo