Commits · f1abb5511b9dfaf79d18d975b6ccf91152f3dab8 · Roger Ferrer / llvm-epi-0.8 · GitLab

This project is archived. Its data is read-only.

Aug 27, 2009
- Expand i8 selects into control flow instead of 16-bit conditional · f1abb551
  Dan Gohman authored Aug 27, 2009
  f1abb551
Aug 26, 2009
- Don't use INSERT_SUBREG to model anyext operations on x86-64, as it · 6c23fa24
  Dan Gohman authored Aug 26, 2009
  6c23fa24
Aug 25, 2009
- convert to filecheck · 6d9d5a9c
  Chris Lattner authored Aug 25, 2009
  6d9d5a9c
- Switch abi-isel.ll to FileCheck; it's not much faster, but it now tests a lot · 9cc4970e
  Daniel Dunbar authored Aug 25, 2009
  9cc4970e
- Remove obsolete -f flags. · 0d4bbf2c
  Dan Gohman authored Aug 25, 2009
  0d4bbf2c
- Fix PR 4751, another difficulty with %a modifier on x86. · f8d37c6b
  Dale Johannesen authored Aug 25, 2009
  f8d37c6b
Aug 24, 2009
- Split test into 3. · fbc9a2e3
  Dale Johannesen authored Aug 24, 2009
  fbc9a2e3
Aug 23, 2009
- Fix PR4753. · 972c8fab
  Jakob Stoklund Olesen authored Aug 23, 2009
  972c8fab
Aug 22, 2009
- Make x86 test actually test x86 code generation. Fix the · 682d8c18
  Eli Friedman authored Aug 22, 2009
  682d8c18
- rename test, make more specific. · f09250f1
  Chris Lattner authored Aug 22, 2009
  f09250f1
Aug 21, 2009
- Use FileCheck even though this means testing for something · fa2b97e6
  Dale Johannesen authored Aug 20, 2009
  fa2b97e6
Aug 20, 2009
- Fix an x86 code size regression: prefer RIP-relative addressing · 05046085
  Dan Gohman authored Aug 20, 2009
  05046085
- Use FileCheck for the test run where it's appropriate. · d16abc09
  Dale Johannesen authored Aug 20, 2009
  d16abc09
- Handle 'a' modifier in X86 asms. PR 4742. · 1d764f61
  Dale Johannesen authored Aug 19, 2009
  1d764f61
Aug 19, 2009
- Make this test platform neutral. · 6c528bc7
  Bill Wendling authored Aug 19, 2009
  6c528bc7
- Add an x86 peep that narrows TEST instructions to forms that use · ac33a906
  Dan Gohman authored Aug 19, 2009
  ac33a906
- PR4737: Fix a nasty bug in load narrowing with non-power-of-two types. · 1e008c17
  Eli Friedman authored Aug 19, 2009
  1e008c17
- Legalize the shift amount operand of SRL_PARTS, SHL_PARTS, and · 4906f73a
  Dan Gohman authored Aug 18, 2009
  4906f73a
Aug 18, 2009
- Make this test less sensitive to assembler differences. · b85ad637
  Dan Gohman authored Aug 18, 2009
  b85ad637
- force a triple so this passes on darwin · 8b0e164a
  Chris Lattner authored Aug 18, 2009
  8b0e164a
- Make tail merging handle blocks with repeated predecessors correctly, and · a41fa359
  Dan Gohman authored Aug 18, 2009
  a41fa359
Aug 15, 2009
- Fix test on Linux. · b4d73122
  Eli Friedman authored Aug 15, 2009
  b4d73122
- implement support for CHECK-NEXT: in filecheck. · da108b4e
  Chris Lattner authored Aug 15, 2009
  da108b4e
- specify a target triple so global variable manglings are consistent etc. · c6a803be
  Chris Lattner authored Aug 15, 2009
  c6a803be
- convert to filecheck. · 3838c2da
  Chris Lattner authored Aug 15, 2009
  3838c2da
- rename this test to sse2.ll · bb193ece
  Chris Lattner authored Aug 15, 2009
  bb193ece
- merge a bunch more sse3 tests into sse3.ll · d3954e27
  Chris Lattner authored Aug 15, 2009
  d3954e27
- convert test to filecheck format. · 9bae01ec
  Chris Lattner authored Aug 15, 2009
  9bae01ec
- rename test · 912aa19c
  Chris Lattner authored Aug 15, 2009
  912aa19c
- this is a test for sse3, simplify it. · e5b9130e
  Chris Lattner authored Aug 15, 2009
  e5b9130e
- On x86-64, for a varargs function, don't store the xmm registers to · 0700a568
  Dan Gohman authored Aug 15, 2009
  0700a568
Aug 14, 2009
- Properly handle indirect win64 args when they're passed in memory · 75821f7c
  Anton Korobeynikov authored Aug 14, 2009
  75821f7c
Aug 13, 2009
- Change MCSectionELF to represent a section semantically instead of · 607cd3b6
  Bruno Cardoso Lopes authored Aug 13, 2009
  607cd3b6
- Various AsmWriter output cleanups. Use WriteAsOperand instead of · ef3d4571
  Dan Gohman authored Aug 13, 2009
  ef3d4571
Aug 12, 2009
- Use WriteAsOperand to print BasicBlock names. · 1c0e13fe
  Dan Gohman authored Aug 12, 2009
  1c0e13fe
- Test for 78821, sort of. While that bug is nondeterministic, · 58043874
  Dale Johannesen authored Aug 12, 2009
  58043874
- one last (?) bad x86 triple test. · 1235f207
  Chris Lattner authored Aug 12, 2009
  1235f207
- fix some pastos in triple lines. · 6628e173
  Chris Lattner authored Aug 12, 2009
  6628e173
- another bogus triple · 0ea6e4cc
  Chris Lattner authored Aug 12, 2009
  0ea6e4cc
- fix another broken target triple. · 8c2d846a
  Chris Lattner authored Aug 12, 2009
  8c2d846a