Commits · 5fc43eb186f018b154380db4f432c51fab43615c · Raul Torres / llvm-target-spread · GitLab

Jun 19, 2010
- Silence compiler warnings. · 5fc43eb1
  Dan Gohman authored Jun 19, 2010
  5fc43eb1
- more refactoring! yay! big win over the intrinsics · 502c4fe6
  Bruno Cardoso Lopes authored Jun 19, 2010
  502c4fe6
- Remove isTwoAddress from here too. · 6bdbdb55
  Eric Christopher authored Jun 18, 2010
  6bdbdb55
- Fix typo, SSE1 should be used by XS, not SSE2 · 66d2d57d
  Bruno Cardoso Lopes authored Jun 18, 2010
  66d2d57d
- Remove isTwoAddress from 64-bit files. · 3577c1b8
  Eric Christopher authored Jun 18, 2010
  3577c1b8
- Move ARM if-conversion before post-ra scheduling. · 119824ed
  Evan Cheng authored Jun 18, 2010
  119824ed
- Teach regular and fast isel to set dead flags on unused implicit defs · 86936504
  Dan Gohman authored Jun 18, 2010
  86936504
- Apply some refactor to packed instructions · 2bfad417
  Bruno Cardoso Lopes authored Jun 18, 2010
  2bfad417
- Update cmake list. · 4f0781c9
  Evan Cheng authored Jun 18, 2010
  4f0781c9
- Thumb2 hazard recognizer. · 28593593
  Evan Cheng authored Jun 18, 2010
  28593593
- Only run CoalesceExtSubRegs when we can expect LiveIntervalAnalysis to clean up · 678927e0
  Jakob Stoklund Olesen authored Jun 18, 2010
  678927e0
- Allow ARM if-converter to be run after post allocation scheduling. · 2d51c7c5
  Evan Cheng authored Jun 18, 2010
  2d51c7c5
- back-end libcall handling for ATOMIC_SWAP (__sync_lock_test_and_set) · a57c2885
  Jim Grosbach authored Jun 18, 2010
  a57c2885
- Enable Expand handling of atomics for subtargets that can't do them inline. · 6860bb77
  Jim Grosbach authored Jun 18, 2010
  6860bb77
- TwoAddressInstructionPass::CoalesceExtSubRegs can insert INSERT_SUBREG · 07f4fa81
  Jakob Stoklund Olesen authored Jun 18, 2010
  07f4fa81
- Fix an inverted condition. · cf9e8a98
  Evan Cheng authored Jun 18, 2010
  cf9e8a98
- Use the new 'defm' class inheritance in SSE · 871439ab
  Bruno Cardoso Lopes authored Jun 18, 2010
  871439ab
- Fix cross initialization compilation error. · f5d62535
  Evan Cheng authored Jun 18, 2010
  f5d62535
Jun 18, 2010
- Teach iff-converter to properly count # of dups. It was not skipping over... · c0e0d85b
  Evan Cheng authored Jun 18, 2010
  c0e0d85b
- Add Expand-to-libcall support for additional atomics. This covers the usual · d64dfc15
  Jim Grosbach authored Jun 18, 2010
  d64dfc15
- Rewrite chained if's as switches and replace assertions with llvm_unreachable · a92e41a5
  Bob Wilson authored Jun 18, 2010
  a92e41a5
- Fix ARM/Thumb reversal in previous attempt. · 589ffb49
  Dale Johannesen authored Jun 18, 2010
  589ffb49
- When using ADDri to get the address of a stack object, 255 is a conservative · 22a212f9
  Jakob Stoklund Olesen authored Jun 18, 2010
  22a212f9
- Make this comment less specific. · a46d6075
  Dan Gohman authored Jun 18, 2010
  a46d6075
- Fix X86FastISel's address-mode folding to stay within the · af4903d6
  Dan Gohman authored Jun 18, 2010
  af4903d6
- An attempt to fix the problem Anton reported with · a06c2f79
  Dale Johannesen authored Jun 18, 2010
  a06c2f79
- Revert r106304 (105548 and friends), which are the SCEVComplexityCompare · 24ceda8e
  Dan Gohman authored Jun 18, 2010
  24ceda8e
- Reapply 105540, 105542, and 105548, and revert r105732. · 4c807fca
  Dan Gohman authored Jun 18, 2010
  4c807fca
- Reapply 105546. · 45073042
  Dan Gohman authored Jun 18, 2010
  45073042
- Reapply 105544. · 9136d9fb
  Dan Gohman authored Jun 18, 2010
  9136d9fb
- Enable tail calls on ARM by default, with some · c1570dda
  Dale Johannesen authored Jun 18, 2010
  c1570dda
- Don't leak RegClass2VRegMap, which is now a new[] array instead of a · e5457c27
  Dan Gohman authored Jun 18, 2010
  e5457c27
- Start TargetRegisterClass indices at 0 instead of 1, so that · 882bb298
  Dan Gohman authored Jun 18, 2010
  882bb298
- Last round of changes for ARM tail calls. · 3ac52b3e
  Dale Johannesen authored Jun 18, 2010
  3ac52b3e
- Fix PR7372: Conditional branches (at least on ARM) are treated as predicated, · f82c8fcc
  Bob Wilson authored Jun 18, 2010
  f82c8fcc
- Treat the ARM inline asm {cc} constraint as a physreg (%CPSR), just like X86 · b9f91667
  Jakob Stoklund Olesen authored Jun 18, 2010
  b9f91667
- Don't bother calling releaseMemory before destroying the DominatorTreeBase. · 9f58b3e1
  Dan Gohman authored Jun 18, 2010
  9f58b3e1
- Minor code simplifications. · 7edb39cc
  Dan Gohman authored Jun 18, 2010
  7edb39cc
- Give NamedRegionTimer an Enabled flag, allowing all its clients to · 6e681a5f
  Dan Gohman authored Jun 18, 2010
  6e681a5f
- Don't replace the old Ordering object with a new one; just clear() · 96ca25eb
  Dan Gohman authored Jun 18, 2010
  96ca25eb