Commits · 14d07db44d652b05187b51e00bdae6981ce24289 · Roger Ferrer / llvm-epi · GitLab

Dec 13, 2004
- More substantial simplifications and speedups. This makes ADCE about 20% faster · 14d07db4
  Chris Lattner authored Dec 12, 2004
  14d07db4
Dec 12, 2004
- More minor microoptimizations · 9115eb30
  Chris Lattner authored Dec 12, 2004
  9115eb30
- Remove some more set operations · d4298781
  Chris Lattner authored Dec 12, 2004
  d4298781
- Reduce number of set operations. · a538439b
  Chris Lattner authored Dec 12, 2004
  a538439b
- Optimize div/rem + select combinations more. · bf5b7cf6
  Chris Lattner authored Dec 12, 2004
  bf5b7cf6
- Add testcases for simple things we can handle that occur now in vortex. · 9803a0a5
  Chris Lattner authored Dec 12, 2004
  9803a0a5
- Dejagnu doesn't work if you do TESTSUITE=.../ · 360cda5f
  Chris Lattner authored Dec 12, 2004
  360cda5f
- Fix Regression/CodeGen/PowerPC/2004-12-12-ZeroSizeCommon.ll, and all programs · cbd47631
  Chris Lattner authored Dec 12, 2004
  cbd47631
- New testcase that the PPC backend miscompiles. It is emitting · a0dffecf
  Chris Lattner authored Dec 12, 2004
  a0dffecf
- CSE calls to getTypeSize. · 10b637ea
  Chris Lattner authored Dec 12, 2004
  10b637ea
- Properly implement copying of a global, fixing the 255.vortex & povray · 745196a5
  Chris Lattner authored Dec 12, 2004
  745196a5
- Simplify code and do not invalidate iterators. · 88deefa3
  Chris Lattner authored Dec 12, 2004
  88deefa3
- Use the target triple to pick this target. · 17550c45
  Chris Lattner authored Dec 12, 2004
  17550c45
- Get rid of subbullets for all of the "known problems" section. None of the · 1adea147
  Chris Lattner authored Dec 12, 2004
  1adea147
- Complete the list of MultiSource failures. · 23b2bc54
  Brian Gaeke authored Dec 12, 2004
  23b2bc54
- Do not internalize a module if -link-as-library is passed. · b92fd600
  Chris Lattner authored Dec 12, 2004
  b92fd600
- hbd should be working now. · ed934682
  Brian Gaeke authored Dec 12, 2004
  ed934682
- Finally enable the setcc-branch folding code. · 27358a44
  Brian Gaeke authored Dec 12, 2004
  27358a44
- Add (currently disabled) code for canFoldSetCC · e5351ad5
  Brian Gaeke authored Dec 12, 2004
  e5351ad5
- Though the previous xform applies to literally dozens (hundreds?) of variables · 1cbd5be7
  Chris Lattner authored Dec 12, 2004
  1cbd5be7
- Add stubs for setcc-branch folding support. · aa40b7d0
  Brian Gaeke authored Dec 12, 2004
  aa40b7d0
- If a variable can only hold two values, and is not already a bool, shrink it · 40e4cec9
  Chris Lattner authored Dec 12, 2004
  40e4cec9
- New testcase. Shrinking the variable to a bool allows instcombine to delete · 17fbc80f
  Chris Lattner authored Dec 12, 2004
  17fbc80f
- If one side of and/or is known to be 0/-1, it doesn't matter · cbc0161d
  Chris Lattner authored Dec 11, 2004
  cbc0161d
- New testcase. If one side of and/or is known to be 0/-1, it doesn't matter · 2c14b32e
  Chris Lattner authored Dec 11, 2004
  2c14b32e
Dec 11, 2004
- Print llvm code one function at a time. · 25ac8356
  Brian Gaeke authored Dec 11, 2004
  25ac8356
- Check in the file I forgot last night, to solve all of the crashes in every · 8d0b1b2a
  Chris Lattner authored Dec 11, 2004
  8d0b1b2a
- JIT should print LLVM each function before selecting instructions for it. · 72daaa46
  Brian Gaeke authored Dec 11, 2004
  72daaa46
- Rename Path::get -> Path::toString · 5625dec2
  Reid Spencer authored Dec 11, 2004
  5625dec2
- Update test · cc6037f8
  Chris Lattner authored Dec 11, 2004
  cc6037f8
- Only cound if we actually made a change. · 263b0a16
  Chris Lattner authored Dec 11, 2004
  263b0a16
- The split bb is really the exit of the old function · ffefea07
  Chris Lattner authored Dec 11, 2004
  ffefea07
- A little cleanup on this file. · 8f72ce0f
  Reid Spencer authored Dec 11, 2004
  8f72ce0f
- New feature · 7bf2e54d
  Chris Lattner authored Dec 11, 2004
  7bf2e54d
- Two bug fixes: · 2f687fd9
  Chris Lattner authored Dec 11, 2004
  2f687fd9
- Do not delete the entry block to a function. · 8525ebe4
  Chris Lattner authored Dec 11, 2004
  8525ebe4
- Bools are *also* not ints. Sigh. Furthermore, most of the TargetMachine · b3095dd8
  Brian Gaeke authored Dec 11, 2004
  b3095dd8
- Look for many more moves to fold (previously, we only · ba5e9f10
  Brian Gaeke authored Dec 11, 2004
  ba5e9f10
- Make GEPs not suck so much: · 52a9ed63
  Brian Gaeke authored Dec 11, 2004
  52a9ed63
- Update lists of failing benchmarks, including info on which · cba3135b
  Brian Gaeke authored Dec 11, 2004
  cba3135b