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
- Convert some more Neon tests to FileCheck. · faebdee4
  Bob Wilson authored Aug 26, 2009
  faebdee4
- Alter 79292 to produce output that actually assembles. · f582ac7c
  Dale Johannesen authored Aug 26, 2009
  f582ac7c
- Expand scalar_to_vector - we don't have any isel logic for it now · 0f756b27
  Anton Korobeynikov authored Aug 26, 2009
  0f756b27
- Don't use INSERT_SUBREG to model anyext operations on x86-64, as it · 6c23fa24
  Dan Gohman authored Aug 26, 2009
  6c23fa24
- Add dummy inline asm handling for 'r' constraint. This fixes PR4778 · a0e01bec
  Anton Korobeynikov authored Aug 26, 2009
  a0e01bec
- Updated i128 sext support for CellSPU backend, contributed by Ken Werner (IBM) · c5dd8bd8
  Scott Michel authored Aug 25, 2009
  c5dd8bd8
Aug 25, 2009
- remove some dead lines. · 3e6c7946
  Chris Lattner authored Aug 25, 2009
  3e6c7946
- convert to filecheck style · 88093c75
  Chris Lattner authored Aug 25, 2009
  88093c75
- 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
- Fixup register kills after scheduling. · ae6bc821
  David Goodwin authored Aug 25, 2009
  ae6bc821
- Provide dynamic_stackalloc lowering for MSP430. · 271cdda8
  Anton Korobeynikov authored Aug 25, 2009
  271cdda8
- 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
- - Remove SelectSEXTi128 from SPUISelDAGToDAG.cpp, evidently, this is redundant · ec89f0c4
  Scott Michel authored Aug 24, 2009
  ec89f0c4
- Prefer 'FileCheck' over 'grep'. · e208c945
  Scott Michel authored Aug 24, 2009
  e208c945
- 128-bit sign extension and vector shift cleanups, contributed by Ken Werner · 8d1602af
  Scott Michel authored Aug 24, 2009
  8d1602af
Aug 24, 2009
- Fix a typo. Somehow I thought this had passed before, but I guess not. · 9054d258
  Bob Wilson authored Aug 24, 2009
  9054d258
- Convert slow test to use FileCheck. · 5fe1d386
  Bob Wilson authored Aug 24, 2009
  5fe1d386
- Convert two gratuitous abuses of poor helpless CPU cycles to FileCheck. · 6969df0f
  Daniel Dunbar authored Aug 24, 2009
  6969df0f
- Split test into 3. · fbc9a2e3
  Dale Johannesen authored Aug 24, 2009
  fbc9a2e3
- Make linkerprivate work for ARM and PPC. Testcase covers · 6bbeda41
  Dale Johannesen authored Aug 24, 2009
  6bbeda41
Aug 23, 2009
- Rerevert (r75663 and r76805), seems there is more non-determinism. · a9e1d3f0
  Daniel Dunbar authored Aug 23, 2009
  a9e1d3f0
- Fix PR4753. · 972c8fab
  Jakob Stoklund Olesen authored Aug 23, 2009
  972c8fab
- Speculatively revert r76823 (i.e., reapply r75663 and r76805) to see if the real · 2982196f
  Daniel Dunbar authored Aug 23, 2009
  2982196f
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
- Add missing RUN line · 48e4a6c7
  Anton Korobeynikov authored Aug 22, 2009
  48e4a6c7
- Reduce the test · 5f47ecb9
  Anton Korobeynikov authored Aug 22, 2009
  5f47ecb9
- Use CHECK-NEXT to make sure we're only getting one copy of each shuffle · 616335f6
  Bob Wilson authored Aug 22, 2009
  616335f6
Aug 21, 2009
- Match VTRN, VZIP, and VUZP shuffles. Restore the tests for these operations, · a7062310
  Bob Wilson authored Aug 21, 2009
  a7062310
- Add fcopysign instructions · f31a44ec
  Anton Korobeynikov authored Aug 21, 2009
  f31a44ec
- Handle 'r' inline asm constraint · a39f96c6
  Anton Korobeynikov authored Aug 21, 2009
  a39f96c6
- Add some tests for vext.16 and vext.32. · f73af72d
  Bob Wilson authored Aug 21, 2009
  f73af72d
- Remove Neon intrinsics for VZIP, VUZP, and VTRN. We will represent these as · 51c7aa04
  Bob Wilson authored Aug 21, 2009
  51c7aa04
- 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
- Fix an obvious copy-n-paste bug. · 01de985a
  Evan Cheng authored Aug 20, 2009
  01de985a
- Use FileCheck for the test run where it's appropriate. · d16abc09
  Dale Johannesen authored Aug 20, 2009
  d16abc09