Commits · 9857d075b5628aabdb238842617839d539a2df18 · Roger Ferrer / llvm-epi-0.8

Jan 27, 2006
- Added notes about a x86 isel deficiency. · 9857d075
  Evan Cheng authored Jan 27, 2006
```
llvm-svn: 25706
```
  9857d075
- Remove out of date comment · 4c251f89
  Chris Lattner authored Jan 27, 2006
```
llvm-svn: 25705
```
  4c251f89
- Remove the only use of FAKE_SOURCES · a4575c73
  Chris Lattner authored Jan 27, 2006
```
llvm-svn: 25704
```
  a4575c73
- Remove FAKE_SOURCES · 35b707cf
  Chris Lattner authored Jan 27, 2006
```
llvm-svn: 25703
```
  35b707cf
- Added a temporary option -enable-x86-sse to enable sse support. It is used by · 1073ae07
  Evan Cheng authored Jan 27, 2006
```
llc-beta.

llvm-svn: 25701
```
  1073ae07
- Bye bye Pattern ISel, hello DAG ISel. · a814f0b3
  Evan Cheng authored Jan 27, 2006
```
llvm-svn: 25700
```
  a814f0b3
- X86 dag isel is now (soon) the default. · 822f6295
  Evan Cheng authored Jan 27, 2006
```
llvm-svn: 25699
```
  822f6295
- These are fp stack test cases. · 7ca86912
  Evan Cheng authored Jan 27, 2006
```
llvm-svn: 25698
```
  7ca86912
- XFAIL tailcall test cases until it's implemented. · a3b41703
  Evan Cheng authored Jan 27, 2006
```
llvm-svn: 25697
```
  a3b41703
- Remove TLI.LowerReturnTo, and just let targets custom lower ISD::RET for · 8c47c3a3
  Nate Begeman authored Jan 27, 2006
```
the same functionality.  This addresses another piece of bug 680.  Next,
on to fixing Alpha VAARG, which I broke last time.

llvm-svn: 25696
```
  8c47c3a3
- Using bit size of integers instead of ambiguous "long" et all. · 0cda006a
  Jim Laskey authored Jan 27, 2006
```
llvm-svn: 25694
```
  0cda006a
- A better workaround · afab7aa8
  Evan Cheng authored Jan 27, 2006
```
llvm-svn: 25692
```
  afab7aa8
- Sorry - really folowing convention. · 4a84e974
  Jim Laskey authored Jan 27, 2006
```
llvm-svn: 25691
```
  4a84e974
- force sse/3dnow off until they work. This fixes all the x86 failures last night · 4be147f4
  Chris Lattner authored Jan 27, 2006
```
llvm-svn: 25690
```
  4be147f4
- Following convention. · 116bb154
  Jim Laskey authored Jan 27, 2006
```
llvm-svn: 25689
```
  116bb154
- Unbreak the JIT with SSE · ed2bb856
  Chris Lattner authored Jan 27, 2006
```
llvm-svn: 25688
```
  ed2bb856
- fix build · fc3eca90
  Andrew Lenharth authored Jan 27, 2006
```
llvm-svn: 25687
```
  fc3eca90
- Fix build error that is apparently only a warning with some compilers. · ecd7e61a
  Chris Lattner authored Jan 27, 2006
```
llvm-svn: 25686
```
  ecd7e61a
- Forgot the version number. · 2b6efa9d
  Jim Laskey authored Jan 27, 2006
```
llvm-svn: 25685
```
  2b6efa9d
- Improve visibility/correctness of operand indices in "llvm.db" objects. · f98fc844
  Jim Laskey authored Jan 27, 2006
```
Handle 64 in DIEs.

llvm-svn: 25684
```
  f98fc844
- Catch up on some of the file addition/deletions. · 064716ab
  Jim Laskey authored Jan 27, 2006
```
llvm-svn: 25683
```
  064716ab
- Fix auto-upgrade of intrinsics to work properly with both assembly and · 76a8d45e
  Reid Spencer authored Jan 27, 2006
```
bytecode reading. This code is crufty, the result of much hacking to get things
working correctly. Cleanup patches will follow.

llvm-svn: 25682
```
  76a8d45e
- A test case for ensuring that conversion of old-format intrinsics with · 6e2135d5
  Reid Spencer authored Jan 27, 2006
```
signed arguments get auto-upgraded correctly.

llvm-svn: 25681
```
  6e2135d5
- This now passes. · b3cd77f9
  Reid Spencer authored Jan 27, 2006
```
llvm-svn: 25680
```
  b3cd77f9
- x86 CPU detection and proper subtarget support · cde9e30b
  Evan Cheng authored Jan 27, 2006
```
llvm-svn: 25679
```
  cde9e30b
- Subtarget feature can now set any variable to any value · d98701c6
  Evan Cheng authored Jan 27, 2006
```
llvm-svn: 25678
```
  d98701c6
- Use printInlineAsm to, well, print inline asm's. · 4d5a93ed
  Chris Lattner authored Jan 27, 2006
```
llvm-svn: 25677
```
  4d5a93ed
- Stub out a method · 061d9e2c
  Chris Lattner authored Jan 27, 2006
```
llvm-svn: 25676
```
  061d9e2c
- add a method · 6f9cc46a
  Chris Lattner authored Jan 27, 2006
```
llvm-svn: 25675
```
  6f9cc46a
- PHI and INLINEASM are now built-in instructions provided by Target.td · 12405746
  Chris Lattner authored Jan 27, 2006
```
llvm-svn: 25674
```
  12405746
- PHI and INLINEASM are now builtin instructions provided by Target.td · 543fe4b6
  Chris Lattner authored Jan 27, 2006
```
llvm-svn: 25673
```
  543fe4b6
- If we want to emit things in enum order, use getInstructionsByEnumValue to · db2a5f09
  Chris Lattner authored Jan 27, 2006
```
get the order, don't compute it ourselves.

Don't emit stuff like (14<<0), emit 14 instead.

Don't attempt to get target properties for builtin instructions.

llvm-svn: 25672
```
  db2a5f09
- There is at least a 'noitinerary' itinerary now · ac8b13b3
  Chris Lattner authored Jan 27, 2006
```
llvm-svn: 25671
```
  ac8b13b3
- Add a default NoItinerary class for targets to use. · 1c341ac1
  Chris Lattner authored Jan 27, 2006
```
llvm-svn: 25670
```
  1c341ac1
- Don't emit JIT code for these instructions · ec249a3d
  Chris Lattner authored Jan 27, 2006
```
llvm-svn: 25669
```
  ec249a3d
- Teach the scheduler to emit the appropriate INLINEASM MachineInstr for an · 4df279cf
  Chris Lattner authored Jan 26, 2006
```
ISD::INLINEASM node.

llvm-svn: 25668
```
  4df279cf
- Add a common INLINEASM opcode · f32b77c8
  Chris Lattner authored Jan 26, 2006
```
llvm-svn: 25667
```
  f32b77c8
- Teach the dag selectors to select InlineAsm nodes. · 58156692
  Chris Lattner authored Jan 26, 2006
```
Aren't we happy the pattern selectors are almost all gone?

llvm-svn: 25666
```
  58156692
Jan 26, 2006
- Dropped DwarfWriter:: · 0f36e7b7
  Jim Laskey authored Jan 26, 2006
```
llvm-svn: 25665
```
  0f36e7b7
- initial selectiondag support for new INLINEASM node. Note that inline asms · 476e67be
  Chris Lattner authored Jan 26, 2006
```
with outputs or inputs are not supported yet. :)

llvm-svn: 25664
```
  476e67be