Commits · fb6e02931c4c909b55bfa383a80e6917aa8eaa4a · Roger Ferrer / llvm-epi-0.8

Mar 23, 2006
- Add support for 8 bit immediates with 16/32 bit cmp instructions · fb6e0293
  Nate Begeman authored Mar 23, 2006
```
llvm-svn: 26966
```
  fb6e0293
Mar 22, 2006
- Added a ValueType operand to isShuffleMaskLegal(). For now, x86 will not do · 021bb7c9
  Evan Cheng authored Mar 22, 2006
```
64-bit vector shuffle.

llvm-svn: 26964
```
  021bb7c9
- SHUFP* are two address code. · ed794cd2
  Evan Cheng authored Mar 22, 2006
```
llvm-svn: 26959
```
  ed794cd2
- Some clean up. · bc047228
  Evan Cheng authored Mar 22, 2006
```
llvm-svn: 26957
```
  bc047228
- - Supposely movlhps is faster / better than unpcklpd. · d4e15579
  Evan Cheng authored Mar 22, 2006
```
- Don't forget pshufd is only available with sse2.

llvm-svn: 26956
```
  d4e15579
- - Implement X86ISelLowering::isShuffleMaskLegal(). We currently only support · 68ad48bd
  Evan Cheng authored Mar 22, 2006
```
  splat and PSHUFD cases.
- Clean up shuffle / splat matching code.

llvm-svn: 26954
```
  68ad48bd
- - VECTOR_SHUFFLE of v4i32 / v4f32 with undef second vector always matches · 8fdbdf20
  Evan Cheng authored Mar 22, 2006
```
  PSHUFD. We can make permutes entries which point to the undef pointing
  anything we want.
- Change some names to appease Chris.

llvm-svn: 26951
```
  8fdbdf20
- Fix PSHUF* and SHUF* jit code emission problems · 3617caf5
  Evan Cheng authored Mar 22, 2006
```
llvm-svn: 26949
```
  3617caf5
- fix a warning · f5e36c8b
  Chris Lattner authored Mar 22, 2006
```
llvm-svn: 26941
```
  f5e36c8b
- Some splat and shuffle support. · d097e675
  Evan Cheng authored Mar 22, 2006
```
llvm-svn: 26940
```
  d097e675
- Add a couple more pseudo instructions. · b1d3c64d
  Evan Cheng authored Mar 22, 2006
```
llvm-svn: 26939
```
  b1d3c64d
- Didn't mean to check this in. No MMX support yet. · baea59c6
  Evan Cheng authored Mar 21, 2006
```
llvm-svn: 26933
```
  baea59c6
- - Use movaps to store 128-bit vector integers. · d5e905d7
  Evan Cheng authored Mar 21, 2006
```
- Each scalar to vector v8i16 and v16i8 is a any_extend followed by a movd.

llvm-svn: 26932
```
  d5e905d7
Mar 21, 2006
- These targets don't support EXTRACT_VECTOR_ELT, though, in time, X86 will. · 00f4683b
  Chris Lattner authored Mar 21, 2006
```
llvm-svn: 26930
```
  00f4683b
- Combine 2 entries · 2d819f5f
  Evan Cheng authored Mar 21, 2006
```
llvm-svn: 26921
```
  2d819f5f
- Add a note about x86 register coallescing · aeebc960
  Evan Cheng authored Mar 21, 2006
```
llvm-svn: 26920
```
  aeebc960
- - Remove scalar to vector pseudo ops. They are just wrong. · 1208d917
  Evan Cheng authored Mar 21, 2006
```
- Handle FR32 to VR128:v4f32 and FR64 to VR128:v2f64 with aliases of MOVAPS
and MOVAPD. Mark them as move instructions and *hope* they will be deleted.

llvm-svn: 26919
```
  1208d917
- x86 ISD::SCALAR_TO_VECTOR support. · e4d14162
  Evan Cheng authored Mar 21, 2006
```
llvm-svn: 26911
```
  e4d14162
- Junk unused vector register classes. · fb872b41
  Evan Cheng authored Mar 21, 2006
```
llvm-svn: 26910
```
  fb872b41
Mar 20, 2006
- Add a build_vector node · 80b6bd27
  Chris Lattner authored Mar 20, 2006
```
llvm-svn: 26895
```
  80b6bd27
- Move a few things around. · e6448448
  Evan Cheng authored Mar 20, 2006
```
llvm-svn: 26893
```
  e6448448
Mar 19, 2006
- add a note with a testcase · d16f6fdd
  Chris Lattner authored Mar 19, 2006
```
llvm-svn: 26877
```
  d16f6fdd
- Vector undef's · f7c2e362
  Evan Cheng authored Mar 19, 2006
```
llvm-svn: 26870
```
  f7c2e362
- Turning on LSR by default · 5111c81a
  Evan Cheng authored Mar 19, 2006
```
llvm-svn: 26861
```
  5111c81a
- Remember which tests are hurt by LSR. · 66a9c0de
  Evan Cheng authored Mar 19, 2006
```
llvm-svn: 26860
```
  66a9c0de
- rename these nodes · f7b6e721
  Chris Lattner authored Mar 19, 2006
```
llvm-svn: 26848
```
  f7b6e721
Mar 18, 2006
- Use the generic vector register classes VR64 / VR128 rather than V4F32, · 9bf978dc
  Evan Cheng authored Mar 18, 2006
```
V8I16, etc.

llvm-svn: 26838
```
  9bf978dc
Mar 17, 2006
- Darwin should use _setjmp/_longjmp instead of setjmp/longjmp. · b09a56f3
  Evan Cheng authored Mar 17, 2006
```
llvm-svn: 26833
```
  b09a56f3
- Move some pattern fragments to the right files. · 4f674921
  Evan Cheng authored Mar 17, 2006
```
llvm-svn: 26831
```
  4f674921
- Disable x86 fastcc from passing args in registers · 388fc4d9
  Chris Lattner authored Mar 17, 2006
```
llvm-svn: 26824
```
  388fc4d9
- Parameterize the number of integer arguments to pass in registers · 43798850
  Chris Lattner authored Mar 17, 2006
```
llvm-svn: 26818
```
  43798850
- Also fold MOV8r0, MOV16r0, MOV32r0 + store to MOV8mi, MOV16mi, and MOV32mi. · bfc2e973
  Evan Cheng authored Mar 17, 2006
```
llvm-svn: 26817
```
  bfc2e973
- Add some missing entries to X86RegisterInfo::foldMemoryOperand(). e.g. · aca7915b
  Evan Cheng authored Mar 17, 2006
```
ADD32ri8.

llvm-svn: 26816
```
  aca7915b
- - Nuke 16-bit SBB instructions. We'll never use them. · 27750f32
  Evan Cheng authored Mar 17, 2006
```
- Nuke a bogus comment.

llvm-svn: 26815
```
  27750f32
- Remove BRTWOWAY* · bb01d4f2
  Nate Begeman authored Mar 17, 2006
```
Make the PPC backend not dependent on BRTWOWAY_CC and make the branch
selector smarter about the code it generates, fixing a case in the
readme.

llvm-svn: 26814
```
  bb01d4f2
Mar 16, 2006
- A new entry. · c11fccee
  Evan Cheng authored Mar 16, 2006
```
llvm-svn: 26810
```
  c11fccee
- Bug fix: condition inverted. · f75555fe
  Evan Cheng authored Mar 16, 2006
```
llvm-svn: 26804
```
  f75555fe
- Added a way for TargetLowering to specify what values can be used as the · 20931a79
  Evan Cheng authored Mar 16, 2006
```
scale component of the target addressing mode.

llvm-svn: 26802
```
  20931a79
Mar 14, 2006
- Added getTargetLowering() to TargetMachine. Refactored targets to support this. · 2dd2c652
  Evan Cheng authored Mar 13, 2006
```
llvm-svn: 26742
```
  2dd2c652
- Add LSR hooks. · af598d24
  Evan Cheng authored Mar 13, 2006
```
llvm-svn: 26740
```
  af598d24