Changes in directory llvm/lib/System/Win32:
DynamicLibrary.inc updated: 1.18 -> 1.19
---
Log message:
Provide hook for alloca on VCPP. Patch by Scott Graham
---
Diffs of the changes: (+18 -4)
DynamicLibrary.inc | 22 ++
1 files changed, 18 insertions(+), 4 deletions(-
Changes in directory llvm/lib/Target/X86:
X86ATTAsmPrinter.cpp updated: 1.107 -> 1.108
X86ATTAsmPrinter.h updated: 1.18 -> 1.19
---
Log message:
Say AT&T instead of Intel in the comments for AT&T support.
---
Diffs of the changes: (+3 -3)
X86ATTAsmPrinter.cpp |4 ++--
X86ATTAsmPrinter.
Changes in directory llvm/lib/Target/X86:
X86InstrSSE.td updated: 1.184 -> 1.185
---
Log message:
Fix loadv2i32 to be loadv4i32, though it isn't actually used anywhere yet.
---
Diffs of the changes: (+1 -1)
X86InstrSSE.td |2 +-
1 files changed, 1 insertion(+), 1 deletion(-)
Index: l
Changes in directory llvm/lib/Target/X86:
X86InstrSSE.td updated: 1.185 -> 1.186
---
Log message:
Make minor adjustments to whitespace and comments to reduce differences
between SSE1 instructions and their respective SSE2 analogues.
---
Diffs of the changes: (+28 -25)
X86InstrSSE.td | 53
Hi All,
Here is the llvm-gcc patch for PR1373. It makes llvm-gcc to emit noalias
parameter attribute as well as noalias intrinsic. Note this patch is just
for review. Please give me some suggestion to improve it. Thanks.
Sheng.
Index: gcc/llvm-convert.cpp
===
Changes in directory llvm/include/llvm/Target:
TargetLowering.h updated: 1.129 -> 1.130
---
Log message:
Generalize MVT::ValueType and associated functions to be able to represent
extended vector types. Remove the special SDNode opcodes used for pre-legalize
vector operations, and the special M
Changes in directory llvm/lib/Target/X86:
X86ISelLowering.cpp updated: 1.414 -> 1.415
---
Log message:
Generalize MVT::ValueType and associated functions to be able to represent
extended vector types. Remove the special SDNode opcodes used for pre-legalize
vector operations, and the special MVT
Changes in directory llvm/lib/VMCore:
ValueTypes.cpp updated: 1.18 -> 1.19
---
Log message:
Generalize MVT::ValueType and associated functions to be able to represent
extended vector types. Remove the special SDNode opcodes used for pre-legalize
vector operations, and the special MVT::Vector ty
Changes in directory llvm/include/llvm/CodeGen:
CallingConvLower.h updated: 1.5 -> 1.6
SelectionDAG.h updated: 1.151 -> 1.152
SelectionDAGNodes.h updated: 1.192 -> 1.193
ValueTypes.h updated: 1.35 -> 1.36
---
Log message:
Generalize MVT::ValueType and associated functions to be able to represen
Changes in directory llvm/tools/llvm2cpp:
CppWriter.cpp updated: 1.52 -> 1.53
---
Log message:
Fix PR1525: http://llvm.org/PR1525 :
Use a better determinator for identifying constant array initializers that
are or are not zero terminated and generate code appropriately.
---
Diffs of the chang
Revision: 128805
Author: clattner
Date: 2007-06-25 10:36:34 -0700 (Mon, 25 Jun 2007)
Log Message:
---
remove a bogus change.
Modified Paths:
--
apple-local/branches/llvm/gcc/c-lex.c
Modified: apple-local/branches/llvm/gcc/c-lex.c
Changes in directory llvm/lib/Transforms/Scalar:
GVNPRE.cpp updated: 1.57 -> 1.58
---
Log message:
Use the built-in postorder iterators rather than computing a postorder walk by
hand.
---
Diffs of the changes: (+3 -35)
GVNPRE.cpp | 38 +++---
1 files chan
Changes in directory llvm/test/Transforms/Inline:
2007-06-25-WeakInline.ll added (r1.1)
---
Log message:
new testcase, the inliner shouldn't inline this.
---
Diffs of the changes: (+14 -0)
2007-06-25-WeakInline.ll | 14 ++
1 files changed, 14 insertions(+)
Index: llvm/test/
Changes in directory llvm/lib/Transforms/IPO:
InlineSimple.cpp updated: 1.86 -> 1.87
---
Log message:
fix Transforms/Inline/2007-06-25-WeakInline.ll by not inlining functions
with weak linkage.
---
Diffs of the changes: (+8 -4)
InlineSimple.cpp | 12
1 files changed, 8 inser
Changes in directory llvm/lib/CodeGen:
RegAllocBigBlock.cpp updated: 1.1 -> 1.2
---
Log message:
A bunch of fixes to the BigBlock allocator improve compile-time by ~20%
and code quality by ~2% on my tests.
A big thank you to Roman Levenstein for this patch! See
http://lists.cs.uiuc.edu/piperma
Changes in directory llvm/lib/CodeGen:
RegAllocBigBlock.cpp updated: 1.2 -> 1.3
---
Log message:
tidy this file up a bit
---
Diffs of the changes: (+82 -54)
RegAllocBigBlock.cpp | 136 ++-
1 files changed, 82 insertions(+), 54 deletions(-)
Changes in directory llvm/include/llvm:
Module.h updated: 1.89 -> 1.90
---
Log message:
Add a convenince member function for appending strings to a module's
module-level asm string data.
---
Diffs of the changes: (+7 -0)
Module.h |7 +++
1 files changed, 7 insertions(+)
Index: ll
Changes in directory llvm/utils/TableGen:
IntrinsicEmitter.cpp updated: 1.32 -> 1.33
---
Log message:
Fix a typo in a comment.
---
Diffs of the changes: (+1 -1)
IntrinsicEmitter.cpp |2 +-
1 files changed, 1 insertion(+), 1 deletion(-)
Index: llvm/utils/TableGen/IntrinsicEmitter.cpp
Changes in directory llvm/include/llvm/Target:
TargetInstrInfo.h updated: 1.132 -> 1.133
---
Log message:
Revert the earlier change that removed the M_REMATERIALIZABLE machine
instruction flag, and use the flag along with a virtual member function
hook for targets to override if there are instr
Changes in directory llvm/lib/Target/ARM:
ARMInstrInfo.cpp updated: 1.39 -> 1.40
ARMInstrInfo.h updated: 1.16 -> 1.17
ARMInstrInfo.td updated: 1.114 -> 1.115
ARMInstrThumb.td updated: 1.32 -> 1.33
---
Log message:
Revert the earlier change that removed the M_REMATERIALIZABLE machine
instruction
Changes in directory llvm/utils/TableGen:
CodeGenInstruction.h updated: 1.30 -> 1.31
CodeGenTarget.cpp updated: 1.94 -> 1.95
InstrInfoEmitter.cpp updated: 1.62 -> 1.63
---
Log message:
Revert the earlier change that removed the M_REMATERIALIZABLE machine
instruction flag, and use the flag along
Changes in directory llvm/lib/Target/X86:
X86InstrFPStack.td updated: 1.10 -> 1.11
X86InstrInfo.cpp updated: 1.92 -> 1.93
X86InstrInfo.h updated: 1.67 -> 1.68
X86InstrInfo.td updated: 1.308 -> 1.309
X86InstrMMX.td updated: 1.33 -> 1.34
X86InstrSSE.td updated: 1.186 -> 1.187
---
Log message:
Rev
Changes in directory llvm/lib/Target:
Target.td updated: 1.105 -> 1.106
---
Log message:
Revert the earlier change that removed the M_REMATERIALIZABLE machine
instruction flag, and use the flag along with a virtual member function
hook for targets to override if there are instructions that are
Changes in directory llvm-test:
Makefile.programs updated: 1.269 -> 1.270
---
Log message:
give the bigblock allocator a workout
---
Diffs of the changes: (+1 -1)
Makefile.programs |2 +-
1 files changed, 1 insertion(+), 1 deletion(-)
Index: llvm-test/Makefile.programs
diff -u llvm-
On Jun 25, 2007, at 8:57 AM, Zhou Sheng wrote:
Hi All,
Here is the llvm-gcc patch for PR1373. It makes llvm-gcc to emit
noalias parameter attribute as well as noalias intrinsic. Note this
patch is just for review. Please give me some suggestion to improve
it. Thanks.
The patch looks
> Generalize MVT::ValueType and associated functions to be able to
> represent
> extended vector types.
Whoa, I'm surprised to see this so soon!
> @@ -45,10 +46,6 @@
>
> isVoid = 12, // This has no value
>
> -Vector = 13, // This is an abstract vector type,
> wh
On Jun 25, 2007, at 11:02 PM, Chris Lattner wrote:
>> + /// MVT::isExtendedValueType - Test if the given ValueType is
>> extended
>> + /// (as opposed to being simple).
>> + static inline bool isExtendedValueType(ValueType VT) {
>> +return VT & ~SimpleTypeMask;
>>}
Oh, one other thi
> @@ -129,7 +130,9 @@
>/// specified value type. This means that it has a register
> that directly
>/// holds it without promotions or expansions.
>bool isTypeLegal(MVT::ValueType VT) const {
> -return RegClassForVT[VT] != 0;
> +return !MVT::isExtendedValueType(VT) ?
> +
> @@ -14,13 +14,21 @@
> #include "llvm/CodeGen/ValueTypes.h"
> #include "llvm/Type.h"
> #include "llvm/DerivedTypes.h"
> +#include
> using namespace llvm;
Please use utostr() in llvm/ADT/StringExtras.h instead of sstream.
-Chris
___
llvm-commits ma
Another great change.
> @@ -856,6 +844,10 @@
>ConstantSDNode *N0C = dyn_cast(N0);
>ConstantSDNode *N1C = dyn_cast(N1);
>MVT::ValueType VT = N0.getValueType();
> +
> + // fold vector ops
> + SDOperand FoldedVOp = SimplifyVBinOp(N);
> + if (FoldedVOp.Val) return FoldedVOp;
I'm co
30 matches
Mail list logo
