Commits · db8ec2d75a90ef7f0b8ab8b0e5bc78075c4dbe5c · Roger Ferrer / llvm-epi · GitLab

Dec 09, 2008
- Add sub/mul overflow intrinsics. This currently doesn't have a · db8ec2d7
  Bill Wendling authored Dec 09, 2008
  db8ec2d7
- Fix PR3117: not all nodes being legalized. The · 445071c4
  Duncan Sands authored Dec 09, 2008
  445071c4
- Fix getNode to allow a vector for the shift amount for shifts of vectors. · 4dd832d2
  Mon P Wang authored Dec 09, 2008
  4dd832d2
Dec 08, 2008
- Factor out the code for sign-extending/truncating gep indices · 4c31524b
  Dan Gohman authored Dec 08, 2008
  4c31524b
Dec 05, 2008
- Fix test to pass on Linux. · 0733759b
  Dale Johannesen authored Dec 05, 2008
  0733759b
- Make LoopStrengthReduce smarter about hoisting things out of · 9efd2ce5
  Dale Johannesen authored Dec 05, 2008
  9efd2ce5
- This test also requires -mattr=+sse41. · 7a15646d
  Evan Cheng authored Dec 05, 2008
  7a15646d
- Effectively undo 60461 in PIC mode which simply transform V_SET0 /... · fd8c4d59
  Evan Cheng authored Dec 05, 2008
  fd8c4d59
- Re-did 60519. It turns out Darwin's handling of hidden visibility symbols are... · 2a03c7e9
  Evan Cheng authored Dec 05, 2008
  2a03c7e9
Dec 04, 2008
- Temporarily revert r60519. It was causing a bootstrap failure: · 6949f613
  Bill Wendling authored Dec 04, 2008
  6949f613
- Visibility hidden GVs do not require extra load of symbol address from the GOT or non-lazy-ptr. · 011c4fa8
  Evan Cheng authored Dec 04, 2008
  011c4fa8
Dec 03, 2008
- Use mmx (punpckldq VR64, (mmx_v_set0)) to clear high 32-bits of a VR64 register. · 1339e72d
  Evan Cheng authored Dec 03, 2008
  1339e72d
- Fix test. · b5a97ff6
  Evan Cheng authored Dec 03, 2008
  b5a97ff6
- Mark x86's V_SET0 and V_SETALLONES with isSimpleLoad, and teach X86's · cc78cdf2
  Dan Gohman authored Dec 03, 2008
  cc78cdf2
- Change label to 'carry' for unsigned adds. · e3402692
  Bill Wendling authored Dec 03, 2008
  e3402692
- Fix byval arguments in the fastcc calling convention. The fastcc convention · 5d3d1f69
  Dan Gohman authored Dec 03, 2008
  5d3d1f69
- Add nounwind attributes to this test. · 971c88f3
  Dan Gohman authored Dec 03, 2008
  971c88f3
- testcases for recent dag combiner changes · b43a6895
  Dale Johannesen authored Dec 03, 2008
  b43a6895
Dec 02, 2008
- Fix PR3124: overly strict assert. · 1718fd43
  Evan Cheng authored Dec 02, 2008
  1718fd43
- Second stab at target-dependent lowering of everyone's favorite nodes: [SU]ADDO · 30e9dc81
  Bill Wendling authored Dec 02, 2008
  30e9dc81
Dec 01, 2008
- Add rdar reference, make this actually fail when the patch isn't applied. · b2f131a4
  Chris Lattner authored Dec 01, 2008
  b2f131a4
- Consider only references to an IV within the loop when · 069a4eee
  Dale Johannesen authored Dec 01, 2008
  069a4eee
Nov 30, 2008
- Followup to r60283: optimize arbitrary width signed divisions as well · c8228d26
  Eli Friedman authored Nov 30, 2008
  c8228d26
- Fix for PR2164: allow transforming arbitrary-width unsigned divides into · 1b7fc154
  Eli Friedman authored Nov 30, 2008
  1b7fc154
- APIntify a test which is potentially unsafe otherwise, and fix the · bd0f5782
  Eli Friedman authored Nov 30, 2008
  bd0f5782
Nov 27, 2008
- XFAil test due to reverting of patch. · 077eb6fc
  Bill Wendling authored Nov 27, 2008
  077eb6fc
- Avoid inserting noop's in the middle of a loop. · 37611437
  Evan Cheng authored Nov 27, 2008
  37611437
- On x86 favors folding short immediate into some arithmetic operations (e.g.... · 83bdb389
  Evan Cheng authored Nov 27, 2008
  83bdb389
- Add -march=x86. · d1dda533
  Evan Cheng authored Nov 27, 2008
  d1dda533
Nov 26, 2008
- Add x86-specific test for add-with-overflow intrinsics. · a69ced6b
  Bill Wendling authored Nov 26, 2008
  a69ced6b
- Turn on my codegen prepare heuristic by default. It doesn't affect · 397a11cc
  Chris Lattner authored Nov 26, 2008
  397a11cc
- This adds in some code (currently disabled unless you pass · eb3e4fb6
  Chris Lattner authored Nov 26, 2008
  eb3e4fb6
- Teach CodeGenPrepare to look through Bitcast instructions when attempting to · a9ab165b
  Chris Lattner authored Nov 26, 2008
  a9ab165b
- fix an over-reduced test. · f0e01def
  Chris Lattner authored Nov 26, 2008
  f0e01def
- this doesn't need EH · 0f98f74c
  Chris Lattner authored Nov 26, 2008
  0f98f74c
Nov 25, 2008
- Initial support for anti-dependence breaking. Currently this code does not · ad2134d4
  Dan Gohman authored Nov 25, 2008
  ad2134d4
Nov 19, 2008
- Add support for rematerialization in pre-alloc-splitting. · 1af37c2f
  Owen Anderson authored Nov 19, 2008
  1af37c2f
Nov 17, 2008
- Removed 2008-10-17-SpillerBug.ll as it does not provide an accurate test of PR2898. · 41df6394
  Lang Hames authored Nov 16, 2008
  41df6394
Nov 16, 2008
- 2008-10-17-SpillerBug.ll is currently failing, but this doesn't reflect an... · 782ec1a7
  Lang Hames authored Nov 16, 2008
  782ec1a7
- Improved shuffle normalization to avoid using extract/build when we · 7a824743
  Mon P Wang authored Nov 16, 2008
  7a824743