Commits · d5f935921ab03291ec93431b098ef8722cc762c9 · Roger Ferrer / llvm-epi · GitLab

Nov 04, 2008
- Fix PR3011: LegalizeTypes support for scalarizing · d5f93592
  Duncan Sands authored Nov 04, 2008
  d5f93592
- The ANDMask node folds to a constant, and isn't the node that needs to · b9110e7f
  Dan Gohman authored Nov 03, 2008
  b9110e7f
Nov 03, 2008
- Change how extended types are represented in MVTs. Instead of fiddling · d7546abb
  Dan Gohman authored Nov 03, 2008
  d7546abb
- Make VAARG work with x86 long double (which is · 0207a3f8
  Duncan Sands authored Nov 03, 2008
  0207a3f8
Oct 31, 2008
- Add a bunch of libcalls for ppcf128 that were somehow · 87588519
  Duncan Sands authored Oct 31, 2008
  87588519
- Use MOVSSmr instead of EXTRACTPSmr in the case of extracting · 99cdf889
  Dan Gohman authored Oct 31, 2008
  99cdf889
Oct 30, 2008
- Testcase for PR2986. · 8311657b
  Duncan Sands authored Oct 30, 2008
  8311657b
- Resolve bug 2947: vararg-marked functions must spill registers R3-R79 to stack · 487c4341
  Scott Michel authored Oct 30, 2008
  487c4341
Oct 29, 2008
- add testcase for PR2964 · 7ee00127
  Chris Lattner authored Oct 29, 2008
  7ee00127
- Fix PR2977: LegalizeTypes support for expanding · fbb10bbe
  Duncan Sands authored Oct 29, 2008
  fbb10bbe
- - More pre-split fixes: spill slot live interval computation bug; restore point bug. · ce3ccc1e
  Evan Cheng authored Oct 29, 2008
  ce3ccc1e
Oct 28, 2008
- Fix darwin ppc llvm-gcc build breakage: intercept · 4068a7f3
  Duncan Sands authored Oct 28, 2008
  4068a7f3
- Turn off LegalizeTypes for this test for the · b68694f1
  Duncan Sands authored Oct 28, 2008
  b68694f1
- Fix a testcase provided by Bill in which the node · f3e5850f
  Duncan Sands authored Oct 28, 2008
  f3e5850f
- Fix a nasty miscompilation of 176.gcc on linux/x86 where we synthesized · 38461f6b
  Chris Lattner authored Oct 28, 2008
  38461f6b
- Avoid putting a split past the end of the live range; always shrink wrap live... · fab31680
  Evan Cheng authored Oct 28, 2008
  fab31680
- Remove val# defined by a remat'ed def that is now dead. · f46642ad
  Evan Cheng authored Oct 27, 2008
  f46642ad
Oct 27, 2008
- rename vec_spat -> vec_splat, pointed out by duncan · b5cf101f
  Chris Lattner authored Oct 27, 2008
  b5cf101f
- Turn on LegalizeTypes, the new type legalization · 8475d567
  Duncan Sands authored Oct 27, 2008
  8475d567
- For now, don't split live intervals around x87 stack register barriers.... · f7137229
  Evan Cheng authored Oct 27, 2008
  f7137229
Oct 26, 2008
- remove eh output from this test. · a43f2b21
  Chris Lattner authored Oct 26, 2008
  a43f2b21
- Do not shrink wrap live interval in a mbb if it's livein any of its successor... · ed033ede
  Evan Cheng authored Oct 26, 2008
  ed033ede
- Handle cases where there aren't uses in the barrier mbb. · f48367b8
  Evan Cheng authored Oct 25, 2008
  f48367b8
Oct 25, 2008
- Related to PR2911, reject as invalid non-pointer GC roots. · bf40eee9
  Gordon Henriksen authored Oct 25, 2008
  bf40eee9
- If val# def is ~0U, meaning it's defined by a PHI, and it's previously split,... · 85d71d45
  Evan Cheng authored Oct 25, 2008
  85d71d45
Oct 24, 2008
- Be kind to non-x86 hosts. · e45896fc
  Dale Johannesen authored Oct 24, 2008
  e45896fc
- Fix translateX86CC: if SetCCOpcode is SETULE and · 014f5bba
  Duncan Sands authored Oct 24, 2008
  014f5bba
- Avoid splitting an interval multiple times; avoid splitting re-materializable val# (for now). · 4bac4d0a
  Evan Cheng authored Oct 24, 2008
  4bac4d0a
Oct 22, 2008
- Fix PR2907 by digging through constant expressions to find FP constants that · 35b40f8c
  Chris Lattner authored Oct 22, 2008
  35b40f8c
Oct 21, 2008
- Fix SelectionDAGBuild lowering of Select instructions to · 8b44b88e
  Dan Gohman authored Oct 21, 2008
  8b44b88e
- really fix run line · 192f27cb
  Chris Lattner authored Oct 21, 2008
  192f27cb
- fix run line · b4ee2aeb
  Chris Lattner authored Oct 21, 2008
  b4ee2aeb
- remove some unneeded eh generation · 0b641e47
  Chris Lattner authored Oct 21, 2008
  0b641e47
- Don't create TargetGlobalAddress nodes with offsets that don't fit · 269246b0
  Dan Gohman authored Oct 21, 2008
  269246b0
- Optimized FCMP_OEQ and FCMP_UNE for x86. · 97d95d6d
  Dan Gohman authored Oct 21, 2008
  97d95d6d
- When the coalescer is doing rematerializing, have it remove · c835458d
  Dan Gohman authored Oct 21, 2008
  c835458d
- Fix gcc.c-torture/compile/920520-1.c by inserting bitconverts · 4396e0d2
  Chris Lattner authored Oct 21, 2008
  4396e0d2
Oct 18, 2008
- Reapply r57699 with a fix to not crash on asms with multiple results. Unlike · 160e8abd
  Chris Lattner authored Oct 18, 2008
  160e8abd
- Teach DAGCombine to fold constant offsets into GlobalAddress nodes, · 2fe6bee5
  Dan Gohman authored Oct 18, 2008
  2fe6bee5
- Revert r57699. It's causing regressions in · 6de25562
  Dan Gohman authored Oct 18, 2008
  6de25562