Commits · df8a8a8c6f4d528084c55aa4c690098774a49100 · Roger Ferrer / llvm-epi · GitLab

Feb 24, 2010
- Change the scheduler from adding nodes in allnodes order · df8a8a8c
  Chris Lattner authored Feb 24, 2010
  df8a8a8c
- LowerCall() should always do getCopyFromReg() to reference the stack pointer. · 6ad4bcb0
  Jim Grosbach authored Feb 24, 2010
  6ad4bcb0
- Re-apply 96540 and 96556 with fixes. · 328a6074
  Evan Cheng authored Feb 24, 2010
  328a6074
- DIV8r must define %AX since X86DAGToDAGISel::Select() sometimes uses it · a2d8c97b
  Jakob Stoklund Olesen authored Feb 24, 2010
  a2d8c97b
Feb 23, 2010
- Remember to handle sub-registers when moving imp-defs to a rematted instruction. · fe0a8cd2
  Jakob Stoklund Olesen authored Feb 23, 2010
  fe0a8cd2
- Keep track of phi join registers explicitly in LiveVariables. · 38b76e27
  Jakob Stoklund Olesen authored Feb 23, 2010
  38b76e27
- Adding the MicroBlaze backend. · e4801e49
  Wesley Peck authored Feb 23, 2010
  e4801e49
- Lower BR_JT on the XCore to a jump into a series of jump instructions. · f5781969
  Richard Osborne authored Feb 23, 2010
  f5781969
- These should not have been committed. · 2a33390e
  Evan Cheng authored Feb 22, 2010
  2a33390e
- no need to run llvm-as here. · 72334622
  Chris Lattner authored Feb 22, 2010
  72334622
- Instcombine constant folding can normalize gep with negative index to index... · 3688b8fa
  Evan Cheng authored Feb 22, 2010
  3688b8fa
Feb 22, 2010
- Canonicalize ConstantInts to the right operand of commutative · 6c5ac6de
  Dan Gohman authored Feb 22, 2010
  6c5ac6de
- Actually enable the -enable-unsafe-fp-math tests. · be24c455
  Dan Gohman authored Feb 22, 2010
  be24c455
- Mark the return address stack slot as mutable when moving the return address · 30ece5b8
  Arnold Schwaighofer authored Feb 22, 2010
  30ece5b8
- Remove the logic for reasoning about NaNs from the code that forms · b87de8d3
  Dan Gohman authored Feb 22, 2010
  b87de8d3
- When emitting an instruction which depends on both a post-incremented · 4506fcb3
  Dan Gohman authored Feb 22, 2010
  4506fcb3
Feb 21, 2010
- add some no-unwinds, other minor cleanups. · 745219ea
  Chris Lattner authored Feb 21, 2010
  745219ea
- add a triple so that this doesn't fail due to linux/ppc register printing · c43c88eb
  Chris Lattner authored Feb 21, 2010
  c43c88eb
- filecheckize and add nouwinds. · 53485469
  Chris Lattner authored Feb 21, 2010
  53485469
- IT turns out that during jumpless setcc lowering eq and ne were swapped. · e96503fa
  Anton Korobeynikov authored Feb 21, 2010
  e96503fa
- fix and un-xfail X86/vec_ss_load_fold.ll · 3c29aff9
  Chris Lattner authored Feb 21, 2010
  3c29aff9
- temporarily disable this. · 7d5f4a4c
  Chris Lattner authored Feb 21, 2010
  7d5f4a4c
Feb 19, 2010
- Check for overflow when scaling up an add or an addrec for · 85af2567
  Dan Gohman authored Feb 19, 2010
  85af2567
- Add support for the 'alignstack' attribute to the x86 backend. Fixes PR5254. · 7e477677
  Charles Davis authored Feb 19, 2010
  7e477677
- Revert commits 96556 and 96640, because commit 96556 breaks the · d0bf6f64
  Duncan Sands authored Feb 19, 2010
  d0bf6f64
- Transform (xor (setcc), (setcc)) == / != 1 to · d2d9252f
  Evan Cheng authored Feb 19, 2010
  d2d9252f
- When determining the set of interesting reuse factors, consider · 2446f575
  Dan Gohman authored Feb 19, 2010
  2446f575
Feb 18, 2010
- getSplatIndex assumes that the first element of the mask contains the splat index · c9489251
  Mon P Wang authored Feb 18, 2010
  c9489251
- Always normalize spill weights, also for intervals created by spilling. · c953acbd
  Jakob Stoklund Olesen authored Feb 18, 2010
  c953acbd
- Make CodePlacementOpt detect special EH control flow by · 5ffef745
  Dan Gohman authored Feb 18, 2010
  5ffef745
- remove empty file · 6a9bdade
  Chris Lattner authored Feb 18, 2010
  6a9bdade
- Use NEON vmin/vmax instructions for floating-point selects. · c6c13a35
  Bob Wilson authored Feb 18, 2010
  c6c13a35
- Some dag combiner goodness: · 0ceb68a5
  Evan Cheng authored Feb 18, 2010
  0ceb68a5
Feb 17, 2010
- Don't check for comments, which vary between subtargets. · 104207b4
  Dan Gohman authored Feb 17, 2010
  104207b4
- Don't attempt to divide INT_MIN by -1; consider such cases to · 5f10d6c5
  Dan Gohman authored Feb 17, 2010
  5f10d6c5
- roundss is an sse 4 thing, fix the test on non-sse41 builders · 1fc2773a
  Chris Lattner authored Feb 17, 2010
  1fc2773a
- Make g5 target explicit; scheduling affects register choice. · cee88742
  Dale Johannesen authored Feb 16, 2010
  cee88742
Feb 16, 2010
- fix rdar://7653908 , a crash on a case where we would fold a load · afac7dad
  Chris Lattner authored Feb 16, 2010
  afac7dad
- Adjust register numbers in tests to compensate for the · 0062f7bf
  Dale Johannesen authored Feb 16, 2010
  0062f7bf
- filecheckize · c98beb56
  Chris Lattner authored Feb 16, 2010
  c98beb56