Changes in directory llvm/lib/System/Win32:
Path.inc updated: 1.41 -> 1.42
---
Log message:
Unlike Unix, Windows won't let a file be implicitly replaced via renaming
without explicit permission.
---
Diffs of the changes: (+1 -1)
Path.inc |2 +-
1 files changed, 1 insertion(+), 1 deleti
Changes in directory llvm/lib/Target/PowerPC:
PPCISelDAGToDAG.cpp updated: 1.180 -> 1.181
---
Log message:
New rlwimi implementation, which is superior to the old one. There are
still a couple missed optimizations, but we now generate all the possible
rlwimis for multiple inserts into the same
Changes in directory llvm/lib/Target:
TargetLowering.cpp updated: 1.59 -> 1.60
---
Log message:
Use ComputeMaskedBits to determine # sign bits as a fallback. This allows us
to handle all kinds of stuff, including silly things like:
sextinreg(setcc,i16) -> setcc.
---
Diffs of the changes: (
Changes in directory llvm/lib/Target:
TargetLowering.cpp updated: 1.58 -> 1.59
---
Log message:
Add some more sign propagation cases
---
Diffs of the changes: (+77 -10)
TargetLowering.cpp | 87 ++---
1 files changed, 77 insertions(+), 10 de
Changes in directory llvm/lib/Bytecode/Archive:
Archive.cpp updated: 1.10 -> 1.11
---
Log message:
Apply bug fix supplied by Greg Pettyjohn for a bug he found: '' is
not a legal path on Windows.
---
Diffs of the changes: (+1 -1)
Archive.cpp |2 +-
1 files changed, 1 insertion(+), 1 de
Changes in directory llvm/lib/CodeGen/SelectionDAG:
DAGCombiner.cpp updated: 1.161 -> 1.162
---
Log message:
Simplify some code, add a couple minor missed folds
---
Diffs of the changes: (+16 -21)
DAGCombiner.cpp | 37 -
1 files changed, 16 insertions(
Changes in directory llvm/lib/CodeGen/SelectionDAG:
SelectionDAG.cpp updated: 1.303 -> 1.304
---
Log message:
constant fold sign_extend_inreg
---
Diffs of the changes: (+9 -1)
SelectionDAG.cpp | 10 +-
1 files changed, 9 insertions(+), 1 deletion(-)
Index: llvm/lib/CodeGen/Sele
Changes in directory llvm/lib/CodeGen/SelectionDAG:
DAGCombiner.cpp updated: 1.160 -> 1.161
---
Log message:
remove cases handled elsewhere
---
Diffs of the changes: (+2 -16)
DAGCombiner.cpp | 18 ++
1 files changed, 2 insertions(+), 16 deletions(-)
Index: llvm/lib/Code
Changes in directory llvm/lib/Target:
TargetLowering.cpp updated: 1.57 -> 1.58
---
Log message:
Add some more simple sign bit propagation cases.
---
Diffs of the changes: (+67 -27)
TargetLowering.cpp | 94 +
1 files changed, 67 insertio
Changes in directory llvm/lib/CodeGen:
AsmPrinter.cpp updated: 1.67 -> 1.68
---
Log message:
Fix some loose ends in MASM support.
---
Diffs of the changes: (+5 -5)
AsmPrinter.cpp | 10 +-
1 files changed, 5 insertions(+), 5 deletions(-)
Index: llvm/lib/CodeGen/AsmPrinter.cpp
dif
Changes in directory llvm/lib/Target/X86:
X86AsmPrinter.cpp updated: 1.178 -> 1.179
X86IntelAsmPrinter.cpp updated: 1.42 -> 1.43
---
Log message:
Fix some loose ends in MASM support.
---
Diffs of the changes: (+72 -60)
X86AsmPrinter.cpp |3 +
X86IntelAsmPrinter.cpp | 129 +
Changes in directory llvm/test/Regression/CodeGen/PowerPC:
and_sext.ll updated: 1.1 -> 1.2
---
Log message:
new testcase we handle right now.
---
Diffs of the changes: (+20 -6)
and_sext.ll | 26 --
1 files changed, 20 insertions(+), 6 deletions(-)
Index: llvm/te
Changes in directory llvm/lib/CodeGen/SelectionDAG:
DAGCombiner.cpp updated: 1.159 -> 1.160
---
Log message:
Use the new TargetLowering::ComputeNumSignBits method to eliminate
sign_extend_inreg operations. Though ComputeNumSignBits is still rudimentary,
this is enough to compile this:
short
Changes in directory llvm/lib/Target:
TargetLowering.cpp updated: 1.56 -> 1.57
---
Log message:
Add some really really simple code for computing sign-bit propagation.
This will certainly be enhanced in the future.
---
Diffs of the changes: (+95 -0)
TargetLowering.cpp | 95 ++
Changes in directory llvm/include/llvm/Target:
TargetLowering.h updated: 1.70 -> 1.71
---
Log message:
Add some new methods for computing sign bit information.
---
Diffs of the changes: (+13 -0)
TargetLowering.h | 13 +
1 files changed, 13 insertions(+)
Index: llvm/include/
Changes in directory llvm/lib/CodeGen/SelectionDAG:
SelectionDAGISel.cpp updated: 1.235 -> 1.236
---
Log message:
When inserting casts, be careful of where we put them. We cannot insert
a cast immediately before a PHI node.
This fixes Regression/CodeGen/Generic/2006-05-06-GEP-Cast-Sink-Crash.
Changes in directory llvm/test/Regression/CodeGen/Generic:
2006-05-06-GEP-Cast-Sink-Crash.ll added (r1.1)
---
Log message:
new testcase
---
Diffs of the changes: (+33 -0)
2006-05-06-GEP-Cast-Sink-Crash.ll | 33 +
1 files changed, 33 insertions(+)
Index:
Changes in directory llvm/lib/Transforms/Scalar:
InstructionCombining.cpp updated: 1.476 -> 1.477
---
Log message:
Move some code around.
Make the "fold (and (cast A), (cast B)) -> (cast (and A, B))" transformation
only apply when both casts really will cause code to be generated. If one or
b
Changes in directory llvm/test/Regression/Transforms/InstCombine:
2006-05-06-Infloop.ll added (r1.1)
---
Log message:
new testcase from ghostscript that inf looped instcombine
---
Diffs of the changes: (+522 -0)
2006-05-06-Infloop.ll | 522 +
19 matches
Mail list logo
