Commits · ec3b77e00dee940476e676d3e936050c0acbde18 · Roger Ferrer / llvm-epi · GitLab

Dec 03, 2011
- [arm-fast-isel] Unaligned stores of floats require special care. · ec3b77e0
  Chad Rosier authored Dec 03, 2011
  ec3b77e0
Dec 02, 2011
- [arm-fast-isel] After promoting a function parameter be sure to update the · 9fd0e55e
  Chad Rosier authored Dec 02, 2011
  9fd0e55e
Nov 30, 2011
- Make GlobalMerge honor the preferred alignment on globals without an... · 6cff9df2
  Eli Friedman authored Nov 30, 2011
  6cff9df2
- FileCheckize. · f50d2eaf
  Jakob Stoklund Olesen authored Nov 29, 2011
  f50d2eaf
Nov 29, 2011
- If fast-isel fails, remove dead instructions generated during the failed · 46addb9e
  Chad Rosier authored Nov 29, 2011
  46addb9e
Nov 28, 2011
- DAG combine should not increase alignment of loads / stores with alignment less · a4b6404c
  Evan Cheng authored Nov 28, 2011
  a4b6404c
Nov 27, 2011
- Upgrade syntax of tests using volatile instructions to use 'load volatile'... · 6a144a22
  Chris Lattner authored Nov 27, 2011
  6a144a22
Nov 17, 2011
- When fast iseling a GEP, accumulate the offset rather than emitting a series of · f83ab704
  Chad Rosier authored Nov 17, 2011
  f83ab704
Nov 15, 2011
- Add vmov.f32 to materialize f32 immediate splats which cannot be handled by · 7ca4b6eb
  Evan Cheng authored Nov 15, 2011
  7ca4b6eb
- ARM VLDR/VSTR instructions don't need a size suffix. · 3e2c6f38
  Jim Grosbach authored Nov 14, 2011
  3e2c6f38
Nov 14, 2011
- Add newline to end of file. Thanks, Eli. · 4e88fbeb
  Chad Rosier authored Nov 14, 2011
  4e88fbeb
- Add support for inlining small memcpys. · ab7223e9
  Chad Rosier authored Nov 14, 2011
  ab7223e9
- Fix a performance regression from r144565. Positive offsets were being lowered · 45110fdf
  Chad Rosier authored Nov 14, 2011
  45110fdf
- Add support for Thumb load/stores with negative offsets. · adfd200b
  Chad Rosier authored Nov 14, 2011
  adfd200b
- Fix early-clobber handling in shrinkToUses. · 7e6004a3
  Jakob Stoklund Olesen authored Nov 14, 2011
  7e6004a3
- Delete stale comment. · 7e07b388
  Jakob Stoklund Olesen authored Nov 14, 2011
  7e07b388
- Add support for ARM halfword load/stores and signed byte loads with negative · 2a1df883
  Chad Rosier authored Nov 14, 2011
  2a1df883
Nov 13, 2011
- The order in which the predicate is added differs between Thumb and ARM mode. ... · 1198d894
  Chad Rosier authored Nov 13, 2011
  1198d894
- Temporarily disable SelectIntrinsicCall when in ARM mode. This is causing failures. · a476e391
  Chad Rosier authored Nov 13, 2011
  a476e391
- Add support for emitting both signed- and zero-extend loads. Fix · c8cfd3a8
  Chad Rosier authored Nov 13, 2011
  c8cfd3a8
- Delete the 'standard' spiller with used the old spilling framework. · 7ef502f6
  Jakob Stoklund Olesen authored Nov 12, 2011
  7ef502f6
Nov 12, 2011
- RAGreedy is better about hinting now. · 0eac531b
  Jakob Stoklund Olesen authored Nov 12, 2011
  0eac531b
- Linear scan is going away. · 8ec1a92a
  Jakob Stoklund Olesen authored Nov 12, 2011
  8ec1a92a
- XFAIL test that depends on linear scan to remove dead code. · 654d6088
  Jakob Stoklund Olesen authored Nov 12, 2011
  654d6088
- Switch a few tests off linearscan. · 4deff7bc
  Jakob Stoklund Olesen authored Nov 12, 2011
  4deff7bc
- Don't try to form pre/post-indexed loads/stores until after LegalizeDAG runs. Fixes PR11029. · 9d448e4a
  Eli Friedman authored Nov 12, 2011
  9d448e4a
- Add support in fast-isel for selecting memset/memcpy/memmove intrinsics. · a7ebc561
  Chad Rosier authored Nov 11, 2011
  a7ebc561
- Loosen test by using REs. Approved by Devang. · ab1a4a23
  Chad Rosier authored Nov 11, 2011
  ab1a4a23
Nov 11, 2011
- Preserve MachineMemOperands in ARMLoadStoreOptimizer. · 28c1d184
  Andrew Trick authored Nov 11, 2011
  28c1d184
- Add support for using immediates with select instructions. · 7ddd63ce
  Chad Rosier authored Nov 11, 2011
  7ddd63ce
- Make sure to expand SIGN_EXTEND_INREG for NEON vectors. PR11319, round 3. · c4a00147
  Eli Friedman authored Nov 11, 2011
  c4a00147
- Add support for using MVN to materialize negative constants. · 2a3503e0
  Chad Rosier authored Nov 11, 2011
  2a3503e0
Nov 10, 2011
- When in ARM mode, LDRH/STRH require special handling of negative offsets. · d1762e00
  Chad Rosier authored Nov 10, 2011
  d1762e00
- For immediate encodings of icmp, zero or sign extend first. Then · 3fbd094a
  Chad Rosier authored Nov 10, 2011
  3fbd094a
- Make sure we correctly unroll conversions between v2f64 and v2i32 on ARM. · 2d4055b6
  Eli Friedman authored Nov 09, 2011
  2d4055b6
Nov 09, 2011
- Add check so we don't try to perform an impossible transformation. Fixes issue from PR11319. · 53218b6f
  Eli Friedman authored Nov 09, 2011
  53218b6f
- Use REs to remove dependencies on the register allocation order. · c22f6518
  Chad Rosier authored Nov 09, 2011
  c22f6518
- Add support for encoding immediates in icmp and fcmp. Hopefully, this will · 595d4194
  Chad Rosier authored Nov 09, 2011
  595d4194
Nov 08, 2011
- Add workaround for Cortex-M3 errata 602117 by replacing ldrd x, y, [x] with ldm or ldr pairs. · c3770ac6
  Evan Cheng authored Nov 08, 2011
  c3770ac6
- Lower mem-ops to unaligned i32/i16 load/stores on ARM where supported. · b85fcd07
  Lang Hames authored Nov 08, 2011
  b85fcd07