Commits · 66cb5db76fad10bc879ccd51e5f30d6bd6551869 · Roger Ferrer / llvm-epi · GitLab

Mar 28, 2007
- new testcase · 66cb5db7
  Chris Lattner authored Mar 28, 2007
  66cb5db7
- Avoid unnecessary APInt construction. · 117477e2
  Zhou Sheng authored Mar 28, 2007
  117477e2
- Dan pointed out that this is done, remove it! · a6527d6a
  Chris Lattner authored Mar 28, 2007
  a6527d6a
- First test check-in. · 734077b1
  Scott Michel authored Mar 28, 2007
  734077b1
- 1. Make more use of getLowBitsSet/getHighBitsSet. · 23f7a1c9
  Zhou Sheng authored Mar 28, 2007
  23f7a1c9
- It is not possible to determine dominance between two PHI nodes · 0b013d57
  Devang Patel authored Mar 28, 2007
  0b013d57
- 1. Make more use of getLowBitsSet/getHighBitsSet. · 2777a318
  Zhou Sheng authored Mar 28, 2007
  2777a318
- Notes on re-materialization. · d771485d
  Evan Cheng authored Mar 28, 2007
  d771485d
- Move rematerialization out of beta. · 603e49c6
  Evan Cheng authored Mar 28, 2007
  603e49c6
- Use UnknownBIts[BitWidth-1] instead of UnknownBIts & SignBits. · c2d3309b
  Zhou Sheng authored Mar 28, 2007
  c2d3309b
- Fix problem with llvm-config that prevented JIT from being used on x86_64 · d578f1f8
  Jeff Cohen authored Mar 28, 2007
  d578f1f8
- Delete the remanant of an obsolete comment. · 8cd7e819
  Reid Spencer authored Mar 28, 2007
  8cd7e819
- Remove unused APInt variable. · 18570b1f
  Zhou Sheng authored Mar 28, 2007
  18570b1f
- Undocument the implementation keyword · 15316f95
  Reid Spencer authored Mar 28, 2007
  15316f95
- Remove use of implementation keyword. · 44259a29
  Reid Spencer authored Mar 28, 2007
  44259a29
- Regenerate. · b7e45e70
  Reid Spencer authored Mar 28, 2007
  b7e45e70
- Replace the "implementation" keyword with "checkpoint". This more accurately · 2e35a588
  Reid Spencer authored Mar 28, 2007
  2e35a588
- The implementation keyword was replaced with checkpoint. · bad2921d
  Reid Spencer authored Mar 28, 2007
  bad2921d
- Clean up codes in ComputeMaskedBits(): · 57e3f732
  Zhou Sheng authored Mar 28, 2007
  57e3f732
- Scale 1 is always ok. · 4388043b
  Evan Cheng authored Mar 28, 2007
  4388043b
- Remove isLegalAddressImmediate. · c2cba18f
  Evan Cheng authored Mar 28, 2007
  c2cba18f
- No need to generate the implementation keyword any more. Its frivolous. · 6d8e75dd
  Reid Spencer authored Mar 28, 2007
  6d8e75dd
- implementation keyword is going .. going .. gone. · de7e7122
  Reid Spencer authored Mar 28, 2007
  de7e7122
- GEP index sinking fixes: · 07c42d43
  Evan Cheng authored Mar 28, 2007
  07c42d43
- don't use 'not' when we can use a positive test · 4ba1036a
  Chris Lattner authored Mar 28, 2007
  4ba1036a
- new test case for PR1280 · 90bb12c2
  Reid Spencer authored Mar 28, 2007
  90bb12c2
- For PR1280: · a5c18bf7
  Reid Spencer authored Mar 28, 2007
  a5c18bf7
- For PR1280: · 94a8cb4b
  Reid Spencer authored Mar 28, 2007
  94a8cb4b
- Added a test case for PR1279. · 1f4acbfe
  Evan Cheng authored Mar 28, 2007
  1f4acbfe
- new testcase · 4776ebc1
  Chris Lattner authored Mar 28, 2007
  4776ebc1
- Fix for PR1279. Dead def has a live interval of length 1. Copy coalescing should · b41697c0
  Evan Cheng authored Mar 28, 2007
  b41697c0
- upgrade a random test · ae0f8dcd
  Chris Lattner authored Mar 28, 2007
  ae0f8dcd
- Remove cruft I put in there... · 2338e21b
  Bill Wendling authored Mar 28, 2007
  2338e21b
- Fix a problem building llvm-gcc on amd64-unknown-freebsd6.2, due to the · c56e4920
  Chris Lattner authored Mar 28, 2007
  c56e4920
- Unbreak mmx arithmetic. It was barfing trying to do v8i8 arithmetic. · ad2db4a4
  Bill Wendling authored Mar 28, 2007
  ad2db4a4
- new testcase · 51788696
  Chris Lattner authored Mar 28, 2007
  51788696
- Fixing doxygen.cfg target. I'm not a makefile expert, so hopefully this... · b453c429
  Tanya Lattner authored Mar 27, 2007
  b453c429
Mar 27, 2007
- Add the "unpack low packed data" instructions. This should be the last of · 999c77f8
  Bill Wendling authored Mar 27, 2007
  999c77f8
- Use SmallPtrSet. · 5a1bd404
  Devang Patel authored Mar 27, 2007
  5a1bd404
- Test that a VIEW_CONVERT_EXPR used as an lvalue has the right type. · 62574103
  Duncan Sands authored Mar 27, 2007
  62574103