Commits · e34c835bdef14f08b512ec5669f47b921ddd309f · Roger Ferrer / llvm-epi · GitLab

Sep 06, 2010
- speed up -gvn 3.4% on the testcase in PR7023 · e34c835b
  Chris Lattner authored Sep 06, 2010
  e34c835b
- Teach loop rotate to hoist trivially invariant instructions · b01c24a9
  Chris Lattner authored Sep 06, 2010
  b01c24a9
- pull a simple method out of LICM into a new · da24b9a4
  Chris Lattner authored Sep 06, 2010
  da24b9a4
- remove some dead code. t2addrmode_imm8s4 is never used in a · f43cb302
  Chris Lattner authored Sep 05, 2010
  f43cb302
Sep 05, 2010
- cleanups. · e40007a7
  Chris Lattner authored Sep 05, 2010
  e40007a7
- add a comment about where this should eventually move. · 419d0aa0
  Chris Lattner authored Sep 05, 2010
  419d0aa0
- update this. · 5cac0f71
  Chris Lattner authored Sep 05, 2010
  5cac0f71
- more cleanups · 1edf7434
  Chris Lattner authored Sep 05, 2010
  1edf7434
- Change lower atomic pass to use IntrinsicInst to simplify it a bit. · e6214557
  Chris Lattner authored Sep 05, 2010
  e6214557
- eliminate some non-obvious casts. UndefValue isa Constant. · 05ef361b
  Chris Lattner authored Sep 05, 2010
  05ef361b
- fix PR8063, a crash in globalopt in the malloc analysis code. · 72d283c8
  Chris Lattner authored Sep 05, 2010
  72d283c8
- Added initialisers for reduction rule counters. · 64a4a136
  Lang Hames authored Sep 05, 2010
  64a4a136
- Fix warning reported by MSVC++ builder. · e3ac69ec
  Nick Lewycky authored Sep 05, 2010
  e3ac69ec
- Switch FnSet to containing the ComparableFunction instead of a pointer to one. · f3a07ec3
  Nick Lewycky authored Sep 05, 2010
  f3a07ec3
- Fix many bugs when merging weak-strong and weak-weak pairs. We now merge all · 0095937b
  Nick Lewycky authored Sep 05, 2010
  0095937b
- implement rdar://6653118 - fastisel should fold loads where possible. · eeba0c73
  Chris Lattner authored Sep 05, 2010
  eeba0c73
Sep 04, 2010
- Remove dead code. · 313358fe
  Jakob Stoklund Olesen authored Sep 04, 2010
  313358fe
- dead method. · 09630481
  Chris Lattner authored Sep 04, 2010
  09630481
- zap dead code. · 65b48b5d
  Chris Lattner authored Sep 04, 2010
  65b48b5d
- remove dead code, mblaze uses SelectAddrRegImm/SelectAddrRegReg, · 55118e72
  Chris Lattner authored Sep 04, 2010
  55118e72
- Remove the last bit of isShuffleMaskLegal checks and improve the comment regarding mmx shuffles · c6accda7
  Bruno Cardoso Lopes authored Sep 04, 2010
  c6accda7
- make explicit that we not handle several mmx shuffles · 731bcc1a
  Bruno Cardoso Lopes authored Sep 04, 2010
  731bcc1a
- Fix LoopSimplify to notify ScalarEvolution when splitting a loop backedge · 487e2501
  Dan Gohman authored Sep 04, 2010
  487e2501
- Emit target specific nodes to handle palignr. Do not touch it for MMX versions yet. · 20779ee1
  Bruno Cardoso Lopes authored Sep 04, 2010
  20779ee1
- Emit target specific nodes to handle splats starting at zero indicies · cff7cd18
  Bruno Cardoso Lopes authored Sep 04, 2010
  cff7cd18
- Emit target specific nodes for isPSHUFHWMask and isPSHUFLWMask · 95759917
  Bruno Cardoso Lopes authored Sep 04, 2010
  95759917
- Emit target specific nodes for isSHUFPMask · 2b57008c
  Bruno Cardoso Lopes authored Sep 04, 2010
  2b57008c
- Previous isMOVLMask matching already emits targets nodes, remove check · 2f7af361
  Bruno Cardoso Lopes authored Sep 04, 2010
  2f7af361
- One more check from the original isShuffleMaskLegal goes away · 9f8e7041
  Bruno Cardoso Lopes authored Sep 04, 2010
  9f8e7041
- Remove a duplicated but useless check that i've inserted in the previous commit. · 16959372
  Bruno Cardoso Lopes authored Sep 04, 2010
  16959372
- Refactor some code and remove the extra checks for unpckl_undef and unpckh_undef · 44578d38
  Bruno Cardoso Lopes authored Sep 04, 2010
  44578d38
- fix a bug in my licm rewrite when a load from the promoted memory · 50506787
  Chris Lattner authored Sep 04, 2010
  50506787
- Remove check for unpckh mask · 7829d0e7
  Bruno Cardoso Lopes authored Sep 03, 2010
  7829d0e7
- Remove check for unpckl mask · d1dacc57
  Bruno Cardoso Lopes authored Sep 03, 2010
  d1dacc57
- Inline isShuffleMaskLegal into LowerVECTOR_SHUFFLE, so we can start · 207b9d62
  Bruno Cardoso Lopes authored Sep 03, 2010
  207b9d62
- Propagate non-local comparisons. Fixes PR1757. · c91c1a20
  Owen Anderson authored Sep 03, 2010
  c91c1a20
- Disable the asserts that check that normalization is perfectly · 47bec3cb
  Dan Gohman authored Sep 03, 2010
  47bec3cb
- Reapply considered harmfull part of rr112934 and r112942. · 2bef20ed
  Bruno Cardoso Lopes authored Sep 03, 2010
  2bef20ed
Sep 03, 2010
- previous patch was a little too tricky for its own good. Don't try to · 005155e2
  Jim Grosbach authored Sep 03, 2010
  005155e2
- Remove the rest of the nonexistent 64-bit AVX instructions. · 367afb5a
  Dale Johannesen authored Sep 03, 2010
  367afb5a