Commits · dd937843d3237ae9f09a7ffe77e81433bbb7788f · Roger Ferrer / llvm-epi · GitLab

Apr 29, 2011
- Fix run-line, again. :( · dd937843
  Eli Friedman authored Apr 29, 2011
  dd937843
- Re-committing r130454, which does not in fact break anything. · 86caced3
  Eli Friedman authored Apr 29, 2011
  86caced3
- Add trunc->branch support, this won't help with clang's i8->i1 truncations · 8d46b477
  Eric Christopher authored Apr 29, 2011
  8d46b477
- SimplifyCFG: Expose phi node folding cost threshold as command line parameter · 616044ac
  Peter Collingbourne authored Apr 29, 2011
  616044ac
- SimplifyCFG: Add CostRemaining parameter to DominatesMergePoint · e3511e15
  Peter Collingbourne authored Apr 29, 2011
  e3511e15
- SimplifyCFG: Add Trunc, ZExt and SExt to the list of cheap instructions for phi node folding · 61f6602a
  Peter Collingbourne authored Apr 29, 2011
  61f6602a
- MCAsmLayout: Add support for computing the symbol offset of variables. Not · 72032861
  Daniel Dunbar authored Apr 29, 2011
  72032861
- MC: Change variable symbols to be recognized as defined, by assigning their... · bea7b93c
  Daniel Dunbar authored Apr 29, 2011
  bea7b93c
- Change DwarfCFIException's member variables to track what it actually · 697edc89
  Rafael Espindola authored Apr 29, 2011
  697edc89
- Teach Thumb2 isel to fold and->rotr ==> ROR. · e794e175
  Andrew Trick authored Apr 29, 2011
  e794e175
- Combine thumb2-ror tests. · 65266ed4
  Andrew Trick authored Apr 29, 2011
  65266ed4
- InstCombine: turn (C1 << A) << C2) into (C1 << C2) << A) · 16f18ed7
  Benjamin Kramer authored Apr 29, 2011
  16f18ed7
- Revert r130454; apparently this doesn't actually work. · 517728b1
  Eli Friedman authored Apr 28, 2011
  517728b1
- Fix runline. · 37b9ede9
  Eli Friedman authored Apr 28, 2011
  37b9ede9
- Fix a rather obscure crash caused by ARM fast-isel generating code which redefines a register. · e4ecd429
  Eli Friedman authored Apr 28, 2011
  e4ecd429
Apr 28, 2011
- fast-isel sret calls, try 2. We actually do need to do something on x86-32.... · 7cd5101a
  Eli Friedman authored Apr 28, 2011
  7cd5101a
- Actually revert r130348 correctly. · 3cf6d403
  Eli Friedman authored Apr 28, 2011
  3cf6d403
- final step needed to resolve PR6627, which allows us to flatten the code down to · 1777601a
  Chris Lattner authored Apr 28, 2011
  1777601a
- Revert r130348; causing buildbot issues on x86-32. · d5a80ca3
  Eli Friedman authored Apr 28, 2011
  d5a80ca3
- InstCombine: Merge "(trunc x) == C1 & (and x, CA) == C2" into a single and+icmp. · 4145c0d3
  Benjamin Kramer authored Apr 28, 2011
  4145c0d3
- teach GVN to widen integer loads when they are overaligned, when doing an · 827a270a
  Chris Lattner authored Apr 28, 2011
  827a270a
- Teach dwarf writer to handle complex address expression for .debug_loc entries. · 3e021533
  Devang Patel authored Apr 28, 2011
  3e021533
- Fix a silly mistake in r130338. · 33c13391
  Eli Friedman authored Apr 28, 2011
  33c13391
- PTX: support for bitwise operations on predicates · 18e6ac83
  Justin Holewinski authored Apr 28, 2011
  18e6ac83
- fast-isel sret. We actually don't need to do anything special on x86. :) rdar://problem/9303592 . · 8bd572fc
  Eli Friedman authored Apr 27, 2011
  8bd572fc
- Add tests for A8.6.110 NOP. · c3c70018
  Johnny Chen authored Apr 27, 2011
  c3c70018
- Fixes PR9730: indvars: An asserting value handle still pointed to this value · 29ac7b88
  Andrew Trick authored Apr 27, 2011
  29ac7b88
- Make the fast-isel code for literal 0.0 a bit shorter/faster, since 0.0 is... · 406c471b
  Eli Friedman authored Apr 27, 2011
  406c471b
Apr 27, 2011
- Teach BasicAA about arm.neon.vld1 and vst1. · 5394c70d
  Dan Gohman authored Apr 27, 2011
  5394c70d
- If converter was being too cute. It look for root BBs (which don't have · 9808d31b
  Evan Cheng authored Apr 27, 2011
  9808d31b
- When analyzing functions known to only access argument pointees, · 39b3a1ef
  Dan Gohman authored Apr 27, 2011
  39b3a1ef
- Simplify cfg inserts a call to trap when unreachable code is detected. Assign... · 12bf0ab4
  Devang Patel authored Apr 27, 2011
  12bf0ab4
- Also add <imp-def> operands for defined and dead super-registers when rewriting. · 71d3b895
  Jakob Stoklund Olesen authored Apr 27, 2011
  71d3b895
- Test case and comment for PR9633. · 01eff820
  Andrew Trick authored Apr 27, 2011
  01eff820
- Fix an edge case involving branches in fast-isel on x86. · 0eea0293
  Eli Friedman authored Apr 27, 2011
  0eea0293
Apr 26, 2011
- remove support for llvm.invariant.end from memdep. It is a · 6b96621a
  Chris Lattner authored Apr 26, 2011
  6b96621a
- Be careful about scheduling nodes above previous calls. It increase usages of · 1355bbdd
  Evan Cheng authored Apr 26, 2011
  1355bbdd
- This test should be in MC. It breaks with changes to scheduling / register... · dbb86b81
  Evan Cheng authored Apr 26, 2011
  dbb86b81
- make a couple of changes to the standard pass pipeline: · 029afe47
  Chris Lattner authored Apr 26, 2011
  029afe47
- Transform: "icmp eq (trunc (lshr(X, cst1)), cst" to "icmp (and X, mask), cst" · 1b06c716
  Chris Lattner authored Apr 26, 2011
  1b06c716