Changes in directory llvm/lib/Target/ARM:
ARMInstrInfo.td updated: 1.107 -> 1.108
---
Log message:
Make jumptable non-predicable for now.
---
Diffs of the changes: (+10 -11)
ARMInstrInfo.td | 21 ++---
1 files changed, 10 insertions(+), 11 deletions(-)
Index: llvm/lib/Ta
Changes in directory llvm/lib/CodeGen:
IfConversion.cpp updated: 1.18 -> 1.19
---
Log message:
Remove a bogus check. Even terminators in a ifcvt need to be predicated.
Unconditional branches can usually be converted to conditional ones.
---
Diffs of the changes: (+1 -3)
IfConversion.cpp |
Changes in directory llvm/lib/CodeGen:
IfConversion.cpp updated: 1.17 -> 1.18
---
Log message:
Allow multiple ifcvt candidates to share children blocks; add some debugging
code.
---
Diffs of the changes: (+20 -7)
IfConversion.cpp | 27 ---
1 files changed, 20 inse
Changes in directory llvm-test/SingleSource/Benchmarks/Misc:
flops.c updated: 1.3 -> 1.4
---
Log message:
Add SMALL_PROBLEM_SIZE support.
---
Diffs of the changes: (+4 -1)
flops.c |5 -
1 files changed, 4 insertions(+), 1 deletion(-)
Index: llvm-test/SingleSource/Benchmarks/Misc
Changes in directory llvm-test/SingleSource/Benchmarks/Misc-C++:
stepanov_v1p2.cpp updated: 1.2 -> 1.3
---
Log message:
Add SMALL_PROBLEM_SIZE support.
---
Diffs of the changes: (+4 -0)
stepanov_v1p2.cpp |4
1 files changed, 4 insertions(+)
Index: llvm-test/SingleSource/Benchma
On May 31, 2007, at 3:34 PM, Lauro Ramos Venancio wrote:
> I'm not disabling the use of RS when FP is used. About 30 lines
> before my
> change, there is this code:
>
> if (RS && RegInfo->hasFP(Fn)) {
>int SFI = RS->getScavengingFrameIndex();
>if (SFI >= 0) {
> ...
>
> The bug was: the
Changes in directory llvm-www/devmtg/2007-05:
06-Spencer-HLVM.pdf updated: 1.1 -> 1.2
---
Log message:
Remove "Confidential & Proprietary" notice from slides.`
---
Diffs of the changes: (+0 -0)
06-Spencer-HLVM.pdf |0
1 files changed
Index: llvm-www/devmtg/2007-05/06-Spencer-HLVM.pd
Changes in directory llvm/lib/Transforms/Scalar:
GVNPRE.cpp updated: 1.7 -> 1.8
---
Log message:
Add a topological sort function.
---
Diffs of the changes: (+56 -2)
GVNPRE.cpp | 58 --
1 files changed, 56 insertions(+), 2 deletions(
I'm not disabling the use of RS when FP is used. About 30 lines before my
change, there is this code:
if (RS && RegInfo->hasFP(Fn)) {
int SFI = RS->getScavengingFrameIndex();
if (SFI >= 0) {
...
The bug was: the PrologEpilogInserter has two chunks of code to define the
ScavengingFrameInde
Nice! Thanks Devang,
-Chris
On May 31, 2007, at 2:56 PM, [EMAIL PROTECTED] wrote:
> Revision: 127876
> Author: dpatel
> Date: 2007-05-31 14:56:52 -0700 (Thu, 31 May 2007)
>
> Log Message:
> ---
> Explicitly set alignment for Objective-C protocols.
> Set Global Value alignment if i
Changes in directory llvm-www/devmtg/2007-05:
index.html updated: 1.7 -> 1.8
---
Log message:
Add link to Reid's slides.
---
Diffs of the changes: (+5 -3)
index.html |8 +---
1 files changed, 5 insertions(+), 3 deletions(-)
Index: llvm-www/devmtg/2007-05/index.html
diff -u llvm-w
Changes in directory llvm-www/devmtg/2007-05:
06-Spencer-HLVM.pdf added (r1.1)
---
Log message:
Add Reid's Slides.
---
Diffs of the changes: (+0 -0)
06-Spencer-HLVM.pdf |0
1 files changed
Index: llvm-www/devmtg/2007-05/06-Spencer-HLVM.pdf
Revision: 127876
Author: dpatel
Date: 2007-05-31 14:56:52 -0700 (Thu, 31 May 2007)
Log Message:
---
Explicitly set alignment for Objective-C protocols.
Set Global Value alignment if it is specified by User.
Modified Paths:
--
apple-local/branches/llvm/gcc/llvm-backen
Changes in directory llvm/lib/CodeGen:
BranchFolding.cpp updated: 1.60 -> 1.61
---
Log message:
Arrange for only 1 of multiple branches to landing pad to be kept.
Do not remove empty landing pads (EH table needs to be updated)
---
Diffs of the changes: (+13 -4)
BranchFolding.cpp | 17 +++
Hi Lauro,
This is the wrong fix.
Even if disabling the use of RS when FP is used is correct, that's up
to the target to not instantiate RS when it detects that. PEI
shouldn't do this check. Please revert this fix. I'm going to reopen
the bug because we need a proper fix.
Evan
On May 31, 2
Changes in directory llvm/lib/CodeGen:
IfConversion.cpp updated: 1.16 -> 1.17
---
Log message:
Fix a typo.
---
Diffs of the changes: (+1 -1)
IfConversion.cpp |2 +-
1 files changed, 1 insertion(+), 1 deletion(-)
Index: llvm/lib/CodeGen/IfConversion.cpp
diff -u llvm/lib/CodeGen/IfConve
Changes in directory llvm/include/llvm/CodeGen:
AsmPrinter.h updated: 1.66 -> 1.67
---
Log message:
Fix the asmprinter so that a globalvalue can specify an explicit alignment
smaller than the preferred alignment, but so that the target can actually
specify a minimum alignment if needed. This f
Changes in directory llvm/lib/CodeGen:
AsmPrinter.cpp updated: 1.163 -> 1.164
---
Log message:
Fix the asmprinter so that a globalvalue can specify an explicit alignment
smaller than the preferred alignment, but so that the target can actually
specify a minimum alignment if needed. This fixes
Changes in directory llvm/lib/Target/ARM:
ARMAsmPrinter.cpp updated: 1.74 -> 1.75
---
Log message:
Fix the asmprinter so that a globalvalue can specify an explicit alignment
smaller than the preferred alignment, but so that the target can actually
specify a minimum alignment if needed. This fi
Changes in directory llvm/test/CodeGen/ARM:
2007-05-31-RegScavengerInfiniteLoop.ll added (r1.1)
---
Log message:
Add a test for PR1424: http://llvm.org/PR1424 .
---
Diffs of the changes: (+237 -0)
2007-05-31-RegScavengerInfiniteLoop.ll | 237 +
1 files chan
Changes in directory llvm/lib/CodeGen:
PrologEpilogInserter.cpp updated: 1.87 -> 1.88
---
Log message:
Fix PR1424: http://llvm.org/PR1424 .
When a function has FP, the register scavenging spill slot offset already
was calculated.
---
Diffs of the changes: (+1 -1)
PrologEpilogInserter.cpp
On May 22, 2007, at 10:41 PM, Reid Spencer wrote:
> On Tue, 22 May 2007 22:12:17 -0700
> Chris Lattner <[EMAIL PROTECTED]> wrote:
>>
>> On May 22, 2007, at 12:33 PM, Reid Spencer wrote:
>>
>>> The attached patch contains a small modification to get the index
>>> of the
>>> iAny types correct for
On May 30, 2007, at 2:51 PM, Duncan Sands wrote:
> TypeInfos are passed around as pointers (i8*), yet when
> a null value is needed it is passed as an i32 value of 0,
> rather than an i8* null. The attached patches change to
> using an i8* null in LLVM and llvm-gcc. At the same time
> I elimina
Revision: 127873
Author: clattner
Date: 2007-05-31 11:18:57 -0700 (Thu, 31 May 2007)
Log Message:
---
>From Duncan:
"TypeInfos are passed around as pointers (i8*), yet when
a null value is needed it is passed as an i32 value of 0,
rather than an i8* null. The attached patches chang
On May 31, 2007, at 11:12 AM, Chris Lattner wrote:
>
> Hi Devang,
>
>> +extern int flag_next_runtime;
>> +
>
> Is this around in non-objc compilers?
yes.
>
>
>> // Non-zero if bytecode from PCH is successfully read.
>> int flag_llvm_pch_read;
>>
>> @@ -733,6 +735,9 @@
>> } else if (const ch
Hi Devang,
> +extern int flag_next_runtime;
> +
Is this around in non-objc compilers?
> // Non-zero if bytecode from PCH is successfully read.
> int flag_llvm_pch_read;
>
> @@ -733,6 +735,9 @@
> } else if (const char *Section =
> LLVM_IMPLICIT_TARGET_GLOBAL_VAR_SECTION(d
Changes in directory llvm-test/MultiSource/Benchmarks/Prolangs-C++/life:
life.cpp updated: 1.3 -> 1.4
---
Log message:
Doh.
---
Diffs of the changes: (+1 -1)
life.cpp |2 +-
1 files changed, 1 insertion(+), 1 deletion(-)
Index: llvm-test/MultiSource/Benchmarks/Prolangs-C++/life/life.c
Revision: 127872
Author: dpatel
Date: 2007-05-31 10:39:08 -0700 (Thu, 31 May 2007)
Log Message:
---
Fix objective-c runtime data alignment.
Modified Paths:
--
apple-local/branches/llvm/gcc/llvm-backend.cpp
Modified: apple-local/branches/llvm/gcc/llvm-backend.cpp
===
On May 31, 2007, at 6:18 AM, Duncan Sands wrote:
> Hi Chris, this breaks the Ada build since the TYPE_SIZE may not be
> a constant:
>
>> + if (TYPE_SIZE(type) == 0 ||
>> + PadStartBits >= int(TREE_INT_CST_LOW(TYPE_SIZE(type))) ||
>
> I've attached a fix, which includes some other tweaks f
Hi Chris, this breaks the Ada build since the TYPE_SIZE may not be a constant:
> + if (TYPE_SIZE(type) == 0 ||
> + PadStartBits >= int(TREE_INT_CST_LOW(TYPE_SIZE(type))) ||
I've attached a fix, which includes some other tweaks for arrays
(not very important) plus whitespace cleanup.
Ciao,
Changes in directory llvm-test/SingleSource/UnitTests/Integer:
reduce_xor.c added (r1.1)
reduce_xor.reference_output added (r1.1)
---
Log message:
Add a new test case for reduce_xor
---
Diffs of the changes: (+121 -0)
reduce_xor.c| 41 ++
reduce_xor.re
31 matches
Mail list logo
