Commits

Branch selector

 37b25ab

User selector

All users

Datepicker

All time

Commit History

Commits on Aug 30, 2010

• Add a new example to the LLVM distribution: a trace-based Brainfuck compiler that uses LLVM as its code generator.

  Show description for 37b25ab
  resistor

  committed
  37b25ab

  Copy full SHA for 37b25ab

• Attempt to remove the MSIL backend from CMake as well based on Chris's r112375.

  Show description for 83944b7
  chandlerc

  committed
  83944b7

  Copy full SHA for 83944b7

• Revert r112461. It was failing on PPC...

  Show description for c040719
  isanbard

  committed
  c040719

  Copy full SHA for c040719

• Create Thumb2sI_cpsr and T2sI_cpsr. These new classes indicate that CPSR is the

  Show description for 4822bce
  isanbard

  committed
  4822bce

  Copy full SHA for 4822bce

• When adding a register, we should mark it as "def" if it can optionally define

  Show description for 25b1419
  isanbard

  committed
  25b1419

  Copy full SHA for 25b1419

Commits on Aug 29, 2010

• revert 112457, it looks like it broke selfhost.

  Show description for a11ce95
  lattner

  committed
  a11ce95

  Copy full SHA for a11ce95

• rewrite DwarfEHPrepare to use SSAUpdater to promote its allocas

  Show description for 4455608
  lattner

  committed
  4455608

  Copy full SHA for 4455608

• inline function into its only caller.

  Show description for c34c220
  lattner

  committed
  c34c220

  Copy full SHA for c34c220

• two changes: 1) make AliasSet hold the list of call sites with an

  Show description for cb7f653
  lattner

  committed
  cb7f653

  Copy full SHA for cb7f653

• LICM does get dead instructions input to it. Instead of sinking them

  Show description for 0de5cad
  lattner

  committed
  0de5cad

  Copy full SHA for 0de5cad

• use moveBefore instead of remove+insert, it avoids some

  Show description for d9a5dae
  lattner

  committed
  d9a5dae

  Copy full SHA for d9a5dae

• revert 112448 for now.

  Show description for 1b9c848
  lattner

  committed
  1b9c848

  Copy full SHA for 1b9c848

• optimize LICM::hoist to use moveBefore. Correct its updating

  Show description for bb07ed3
  lattner

  committed
  bb07ed3

  Copy full SHA for bb07ed3

• fix some bugs (found by inspection) where LICM would not update

  Show description for 9891750
  lattner

  committed
  9891750

  Copy full SHA for 9891750

• rework the ownership of subloop alias information: instead of

  Show description for 4282e32
  lattner

  committed
  4282e32

  Copy full SHA for 4282e32

• apparently unswitch had the same "Feature". Stop its

  Show description for ea10923
  lattner

  committed
  ea10923

  Copy full SHA for ea10923

• now that loop passes don't use DomFrontier, there is no reason

  Show description for 9377901
  lattner

  committed
  9377901

  Copy full SHA for 9377901

• Make IVUsers iterative instead of recursive.

  Show description for eaa40ff
  

  Dan Gohman

  committed
  eaa40ff

  Copy full SHA for eaa40ff

• Optionally rerun dedicated-register filtering after applying

  Show description for 4f7e18d
  

  Dan Gohman

  committed
  4f7e18d

  Copy full SHA for 4f7e18d

• Fix several areas in LSR to do a better job keeping the main

  Show description for 25608f7
  

  Dan Gohman

  committed
  25608f7

  Copy full SHA for 25608f7

• Refactor the three main groups of code out of

  Show description for 4aa5c2e
  

  Dan Gohman

  committed
  4aa5c2e

  Copy full SHA for 4aa5c2e

• Delete a bogus check.

  Show description for e39a47c
  

  Dan Gohman

  committed
  e39a47c

  Copy full SHA for e39a47c

• Add some comments.

  Show description for 6a83271
  

  Dan Gohman

  committed
  6a83271

  Copy full SHA for 6a83271

• Move this debug output into GenerateAllReuseFormula, to declutter

  Show description for 3902f9f
  

  Dan Gohman

  committed
  3902f9f

  Copy full SHA for 3902f9f

• Delete an unused declaration.

  Show description for a7d0d64
  

  Dan Gohman

  committed
  a7d0d64

  Copy full SHA for a7d0d64

• Do one lookup instead of two.

  Show description for 46fd7a6
  

  Dan Gohman

  committed
  46fd7a6

  Copy full SHA for 46fd7a6

• Restructure the {A,+,B}<L> * {C,+,D}<L> folding so that it folds

  Show description for d578a40
  

  Dan Gohman

  committed
  d578a40

  Copy full SHA for d578a40

• Batch up subtracts along with adds, when analyzing long chains of

  Show description for b8fc62b
  

  Dan Gohman

  committed
  b8fc62b

  Copy full SHA for b8fc62b

• Micro-optimize GroupByComplexity.

  Show description for c6a8e99
  

  Dan Gohman

  committed
  c6a8e99

  Copy full SHA for c6a8e99

• Hold AddRec->getLoop() in a variable, to make the Mul code more consistent

  Show description for 0f32ae3
  

  Dan Gohman

  committed
  0f32ae3

  Copy full SHA for 0f32ae3

• Rename a variable, for consistency.

  Show description for 30cbc86
  

  Dan Gohman

  committed
  30cbc86

  Copy full SHA for 30cbc86

• Use iterators instead of indices.

  Show description for 7e1fee7
  

  Dan Gohman

  committed
  7e1fee7

  Copy full SHA for 7e1fee7

• Don't worry about union types.

  Show description for 8f49168
  

  Dan Gohman

  committed
  8f49168

  Copy full SHA for 8f49168

• Make this test less dependent on register allocation choices.

  Show description for 2b5258c
  

  Dan Gohman

  committed
  2b5258c

  Copy full SHA for 2b5258c

• Use exec.

  Show description for 0584163
  

  Dan Gohman

  committed
  0584163

  Copy full SHA for 0584163

Pagination

Previous

Next