[llvm-commits] CVS: llvm/lib/Transforms/Scalar/LoopUnswitch.cpp

Changes in directory llvm/lib/Transforms/Scalar: LoopUnswitch.cpp updated: 1.6 - 1.7 --- Log message: Simplify the loop-unswitch pass, by not even trying to unswitch loops with uses of loop values outside the loop. We need loop-closed SSA form to do this right, or to use SSA rewriting if we

[llvm-commits] CVS: llvm/lib/Transforms/Scalar/LoopUnswitch.cpp

Changes in directory llvm/lib/Transforms/Scalar: LoopUnswitch.cpp updated: 1.7 - 1.8 --- Log message: Make the threshold a parameter --- Diffs of the changes: (+7 -3) LoopUnswitch.cpp | 10 +++--- 1 files changed, 7 insertions(+), 3 deletions(-) Index:

[llvm-commits] CVS: llvm/include/llvm/CodeGen/SelectionDAG.h

Changes in directory llvm/include/llvm/CodeGen: SelectionDAG.h updated: 1.96 - 1.97 --- Log message: Added SelectionDAG::InsertISelMapEntry(). This is used to workaround the gcc problem where it inline the map insertion call too aggressively. Before this change it was producing a frame size of

[llvm-commits] CVS: llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp

Changes in directory llvm/lib/CodeGen/SelectionDAG: SelectionDAG.cpp updated: 1.255 - 1.256 --- Log message: Added SelectionDAG::InsertISelMapEntry(). This is used to workaround the gcc problem where it inline the map insertion call too aggressively. Before this change it was producing a frame

[llvm-commits] CVS: llvm/utils/TableGen/DAGISelEmitter.cpp

Changes in directory llvm/utils/TableGen: DAGISelEmitter.cpp updated: 1.171 - 1.172 --- Log message: Call InsertISelMapEntry rather than map insertion operator to prevent overly aggrssive inlining. This reduces Select_store frame size from 24k to 10k. --- Diffs of the changes: (+64 -45)

[llvm-commits] CVS: llvm/lib/Target/X86/X86ISelDAGToDAG.cpp

Changes in directory llvm/lib/Target/X86: X86ISelDAGToDAG.cpp updated: 1.45 - 1.46 --- Log message: Match tblgen change. --- Diffs of the changes: (+2 -2) X86ISelDAGToDAG.cpp |4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) Index: llvm/lib/Target/X86/X86ISelDAGToDAG.cpp

[llvm-commits] CVS: llvm/lib/Transforms/Scalar/LoopUnswitch.cpp

Changes in directory llvm/lib/Transforms/Scalar: LoopUnswitch.cpp updated: 1.8 - 1.9 --- Log message: Simplify control flow a bit, note that unswitch preserves canonical loop form --- Diffs of the changes: (+12 -10) LoopUnswitch.cpp | 22 -- 1 files changed, 12

[llvm-commits] CVS: llvm-test/SingleSource/UnitTests/2005-07-17-INT-To-FP.c

Changes in directory llvm-test/SingleSource/UnitTests: 2005-07-17-INT-To-FP.c updated: 1.6 - 1.7 --- Log message: Fixing the test case: it was testing boundry cases where the result of conversion is different depending on whether x86 SSE or 387 instructions are used. --- Diffs of the

[llvm-commits] CVS: llvm-test/MultiSource/Benchmarks/Prolangs-C/dixie/Makefile add.c bind.c compare.c dixie.h error.c globals.c init.c list.c modify.c modrdn.c parse.c read.c remove.c search.c send.c

Changes in directory llvm-test/MultiSource/Benchmarks/Prolangs-C/dixie: Makefile (r1.2) removed add.c (r1.1) removed bind.c (r1.1) removed compare.c (r1.1) removed dixie.h (r1.1) removed error.c (r1.1) removed globals.c (r1.1) removed init.c (r1.1) removed list.c (r1.1) removed modify.c (r1.1)

[llvm-commits] CVS: llvm-test/MultiSource/Benchmarks/Prolangs-C/Makefile

Changes in directory llvm-test/MultiSource/Benchmarks/Prolangs-C: Makefile updated: 1.5 - 1.6 --- Log message: Removed dixie --- Diffs of the changes: (+1 -1) Makefile |2 +- 1 files changed, 1 insertion(+), 1 deletion(-) Index: llvm-test/MultiSource/Benchmarks/Prolangs-C/Makefile

[llvm-commits] CVS: llvm/lib/Transforms/Scalar/LoopUnswitch.cpp

Changes in directory llvm/lib/Transforms/Scalar: LoopUnswitch.cpp updated: 1.9 - 1.10 --- Log message: Implement unconditional unswitching of 'trivial' loops, those loops that contain branches in their entry block that control whether or not the loop is a noop or not. --- Diffs of the

[llvm-commits] CVS: llvm/lib/Transforms/Scalar/LoopUnswitch.cpp

Changes in directory llvm/lib/Transforms/Scalar: LoopUnswitch.cpp updated: 1.10 - 1.11 --- Log message: When unswitching a trivial loop, do admit we are doing it! :) --- Diffs of the changes: (+7 -0) LoopUnswitch.cpp |7 +++ 1 files changed, 7 insertions(+) Index:

[llvm-commits] CVS: llvm/lib/Transforms/Scalar/LoopUnswitch.cpp

Changes in directory llvm/lib/Transforms/Scalar: LoopUnswitch.cpp updated: 1.11 - 1.12 --- Log message: Move code around to be more logical, no functionality change. --- Diffs of the changes: (+32 -26) LoopUnswitch.cpp | 58 ++- 1

[llvm-commits] CVS: llvm/lib/Transforms/Scalar/LoopUnswitch.cpp

Changes in directory llvm/lib/Transforms/Scalar: LoopUnswitch.cpp updated: 1.12 - 1.13 --- Log message: add some notes, move some code around. Implement unswitching of loops with branches on partially invariant computations. --- Diffs of the changes: (+64 -19) LoopUnswitch.cpp | 83

[llvm-commits] CVS: llvm/lib/Target/Sparc/SparcAsmPrinter.cpp SparcISelDAGToDAG.cpp SparcInstrInfo.td

Changes in directory llvm/lib/Target/Sparc: SparcAsmPrinter.cpp updated: 1.53 - 1.54 SparcISelDAGToDAG.cpp updated: 1.82 - 1.83 SparcInstrInfo.td updated: 1.120 - 1.121 --- Log message: Use the auto-generated call matcher. Remove a broken impl of the frameaddr/returnaddr intrinsics. Autogen