Commits · 762c6b7346365eafc7a61e3a637cf32bd9e3ac93 · Roger Ferrer / llvm-epi-0.8 · GitLab

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

Feb 18, 2010
- Use NEON vmin/vmax instructions for floating-point selects. · c6c13a35
  Bob Wilson authored Feb 18, 2010
  c6c13a35
Feb 09, 2010
- tighten up eh.setjmp sequence a bit. · a570d052
  Jim Grosbach authored Feb 08, 2010
  a570d052
Feb 03, 2010
- Revert 95130. · 6f36a083
  Evan Cheng authored Feb 02, 2010
  6f36a083
Feb 02, 2010
- Pass callsite return type to TargetLowering::LowerCall and use that to check sibcall eligibility. · c1b0116f
  Evan Cheng authored Feb 02, 2010
  c1b0116f
Jan 27, 2010
- Eliminate target hook IsEligibleForTailCallOptimization. · 67a69dd2
  Evan Cheng authored Jan 27, 2010
  67a69dd2
Jan 18, 2010
- Patch by David Conrad: · 8546ec9c
  Jim Grosbach authored Jan 18, 2010
  8546ec9c
Dec 12, 2009
- Framework for atomic binary operations. The emitter for the pseudo instructions · 8f9a3ac1
  Jim Grosbach authored Dec 12, 2009
  8f9a3ac1
Dec 11, 2009
- Rough first pass at compare_and_swap atomic builtins for ARM mode. Work in progress. · 5c4e99fc
  Jim Grosbach authored Dec 11, 2009
  5c4e99fc
Dec 10, 2009
- Add memory barrier intrinsic support for ARM. Moving towards adding the atomic... · 53e88544
  Jim Grosbach authored Dec 10, 2009
  53e88544
Nov 12, 2009
- isLegalICmpImmediate should take a signed integer; code clean up. · 15b80e4a
  Evan Cheng authored Nov 12, 2009
  15b80e4a
Nov 11, 2009
- Add TargetLowering::isLegalICmpImmediate. It tells LSR what immediate can be... · 3d3c24a8
  Evan Cheng authored Nov 11, 2009
  3d3c24a8
Nov 09, 2009
- Use Unified Assembly Syntax for the ARM backend. · d7cf55cd
  Jim Grosbach authored Nov 09, 2009
  d7cf55cd
Oct 30, 2009
- Add ARM codegen for indirect branches. · 1cf0b030
  Bob Wilson authored Oct 30, 2009
  1cf0b030
Oct 28, 2009
- Use fconsts and fconstd to materialize small fp constants. · 4a609f3c
  Evan Cheng authored Oct 28, 2009
  4a609f3c
Sep 23, 2009
- ARM does not support offset folding (yet). Disable it for now. · 29a44df5
  Anton Korobeynikov authored Sep 23, 2009
  29a44df5
Sep 18, 2009
- Enhance EmitInstrWithCustomInserter() so target can specify CFG changes that... · 270d0f98
  Evan Cheng authored Sep 18, 2009
  270d0f98
Sep 02, 2009
- Retype from unsigned to CallingConv::ID accordingly. Approved by Bob Wilson. · 68c5f477
  Sandeep Patel authored Sep 02, 2009
  68c5f477
Aug 26, 2009
- Remove unneeded ARM-specific DAG nodes for VLD* and VST* Neon operations. · e0636a7a
  Bob Wilson authored Aug 26, 2009
  e0636a7a
Aug 21, 2009
- Match VTRN, VZIP, and VUZP shuffles. Restore the tests for these operations, · a7062310
  Bob Wilson authored Aug 21, 2009
  a7062310
- Fix some typos and use type-based isel for VZIP/VUZP/VTRN · 232b19c3
  Anton Korobeynikov authored Aug 21, 2009
  232b19c3
- Add lowering of ARM 4-element shuffles to multiple instructios via perfectshuffle-generated table. · 9a232f46
  Anton Korobeynikov authored Aug 21, 2009
  9a232f46
- Use masks not nodes for vector shuffle predicates. Provide set of 'legal'... · c32e99e3
  Anton Korobeynikov authored Aug 21, 2009
  c32e99e3
Aug 19, 2009
- Add support for Neon VEXT (vector extract) shuffles. · 32cd8550
  Bob Wilson authored Aug 19, 2009
  32cd8550
Aug 15, 2009
- Reapply r79127. It was fixed by d0k. · bae6b2cc
  Bill Wendling authored Aug 15, 2009
  bae6b2cc
- Revert r79127. It was causing compilation errors. · d3fade65
  Bill Wendling authored Aug 15, 2009
  d3fade65
- Change allowsUnalignedMemoryAccesses to take type argument since some targets · 52d4e647
  Evan Cheng authored Aug 15, 2009
  52d4e647
Aug 14, 2009
- Add Thumb2 lsr hooks. · dc49a8d3
  Evan Cheng authored Aug 14, 2009
  dc49a8d3
- Create a new ARM-specific DAG node, VDUP, to represent a splat from a · eb54d517
  Bob Wilson authored Aug 14, 2009
  eb54d517
- During legalization, change Neon vdup_lane operations from shuffles to · cce31f68
  Bob Wilson authored Aug 14, 2009
  cce31f68
Aug 13, 2009
- Revert r78852 for now. I want to do this differently, but I don't have time · ef6e602b
  Bob Wilson authored Aug 13, 2009
  ef6e602b
- Recognize Neon VDUP shuffles during legalization instead of selection. · ff2db102
  Bob Wilson authored Aug 12, 2009
  ff2db102
- Recognize Neon VREV shuffles during legalization instead of selection. · ea3a402a
  Bob Wilson authored Aug 12, 2009
  ea3a402a
Aug 11, 2009
- Rename MVT to EVT, in preparation for splitting SimpleValueType out into its own struct type. · 53aa7a96
  Owen Anderson authored Aug 10, 2009
  53aa7a96
Aug 07, 2009
- It turns out most of the thumb2 instructions are not allowed to touch SP. The... · b972e563
  Evan Cheng authored Aug 07, 2009
  b972e563
Aug 06, 2009
- Implement Neon VST[234] operations. · 0127031c
  Bob Wilson authored Aug 06, 2009
  0127031c
Aug 05, 2009
- Missed pieces for ARM HardFP ABI. · 22ef7515
  Anton Korobeynikov authored Aug 05, 2009
  22ef7515
- Major calling convention code refactoring. · f9bbcd1a
  Dan Gohman authored Aug 05, 2009
  f9bbcd1a
Aug 04, 2009
- Lower Neon VLD* intrinsics to custom DAG nodes, and manually allocate the · f45dee3a
  Bob Wilson authored Aug 04, 2009
  f45dee3a
Jul 29, 2009
- Optimize Thumb2 jumptable to use tbb / tbh when all the offsets fit in byte / halfword. · c6d70ae0
  Evan Cheng authored Jul 29, 2009
  c6d70ae0
Jul 28, 2009
- In thumb2 mode, add pc is unpredictable. Use add + mov pc instead (that is... · c8bed033
  Evan Cheng authored Jul 28, 2009
  c8bed033