Commits · 282c2af4d843306ee6a9a9fc0e29a1d4807810c0 · Roger Ferrer / llvm-epi-0.8

Dec 05, 2005
- On some targets (e.g. X86), shift amounts are not the same as the value · 282c2af4
  Chris Lattner authored Dec 05, 2005
```
being shifted.  Don't assume they are.

llvm-svn: 24598
```
  282c2af4
- Add some explicit type casts so that tblgen knows the type of the shiftamount,... · c54cddd2
  Chris Lattner authored Dec 05, 2005
```
Add some explicit type casts so that tblgen knows the type of the shiftamount, which is not necessarily the same as the type being shifted.

llvm-svn: 24595
```
  c54cddd2
- Add some explicit type casts so that tblgen knows the type of the shift · f3322af5
  Chris Lattner authored Dec 05, 2005
```
amount, which is not necessarily the same as the type being shifted.

llvm-svn: 24594
```
  f3322af5
Dec 04, 2005
- The basic fneg cases are already autogen'd · efc86f5f
  Chris Lattner authored Dec 04, 2005
```
llvm-svn: 24592
```
  efc86f5f
- Autogen matching code for ADJCALLSTACK[UP|DOWN], thanks to Evan's tblgen · f9797947
  Chris Lattner authored Dec 04, 2005
```
improvements.

llvm-svn: 24591
```
  f9797947
- Finish moving uncond br over to .td file, remove from .cpp file. · fd857daa
  Chris Lattner authored Dec 04, 2005
```
llvm-svn: 24590
```
  fd857daa
- Define BR in the .td file now that Evan made tblgen smarter. · d9d18aff
  Chris Lattner authored Dec 04, 2005
```
llvm-svn: 24589
```
  d9d18aff
- Added isel patterns for RET, JMP, and WRITEPORT. · 95cb7638
  Evan Cheng authored Dec 04, 2005
```
llvm-svn: 24588
```
  95cb7638
- * Added instruction property hasCtrlDep for those which r/w control-flow · e8531381
  Evan Cheng authored Dec 04, 2005
```
  chains.
* Added DAG node property SDNPHasChain for nodes which r/w control-flow
  chains.
* Renamed SDTVT to SDTOther.
* Added several new SDTypeProfiles for BR, BRCOND, RET, and WRITEPORT.
* Added SDNode definitions for BR, BRCOND, RET, and WRITEPORT.

llvm-svn: 24586
```
  e8531381
- Fix PR672 another way which should be more robust · 7e79292f
  Chris Lattner authored Dec 04, 2005
```
llvm-svn: 24585
```
  7e79292f
Dec 03, 2005
- dbg.stoppoint returns a value, don't forget to init it · 8782b782
  Chris Lattner authored Dec 03, 2005
```
llvm-svn: 24583
```
  8782b782
- Fix SimplifyCFG/2005-12-03-IncorrectPHIFold.ll · 2820b8c8
  Chris Lattner authored Dec 03, 2005
```
llvm-svn: 24581
```
  2820b8c8
- Fix test/Regression/ExecutionEngine/2005-12-02-TailCallBug.ll and PR672. · ecfc7e56
  Chris Lattner authored Dec 03, 2005
```
This also fixes 177.mesa, the only program that fails with --enable-x86-fastcc
turned on.  Given a clean nightly tester run, we should be able to turn it
on by default!

llvm-svn: 24578
```
  ecfc7e56
Dec 02, 2005
- bah, must generate all results · f9b27d70
  Andrew Lenharth authored Dec 02, 2005
```
llvm-svn: 24574
```
  f9b27d70
- cycle counter fix · 73420b37
  Andrew Lenharth authored Dec 02, 2005
```
llvm-svn: 24573
```
  73420b37
- add a note · 986cb409
  Chris Lattner authored Dec 02, 2005
```
llvm-svn: 24572
```
  986cb409
- Don't remove two operand, two result nodes from the binary ops map. These · 0142afd6
  Chris Lattner authored Dec 01, 2005
```
should come from the arbitrary ops map.

This fixes Regression/CodeGen/PowerPC/2005-12-01-Crash.ll

llvm-svn: 24571
```
  0142afd6
Dec 01, 2005
- Promote line and column number information for our friendly 64-bit targets. · 05b0b457
  Chris Lattner authored Dec 01, 2005
```
llvm-svn: 24568
```
  05b0b457
- IA64 doesn't support the LOCATION node, and for some reason the ISelPattern · 93feffb4
  Chris Lattner authored Dec 01, 2005
```
stuff isn't using ISelLowering.cpp

llvm-svn: 24567
```
  93feffb4
- Make sure these get added into the codegenmap when appropriate · df928783
  Chris Lattner authored Dec 01, 2005
```
llvm-svn: 24566
```
  df928783
- This is a bugfix for SelectNodeTo. In certain situations, we could be · 9d0d715e
  Chris Lattner authored Dec 01, 2005
```
selecting a node and use a mix of getTargetNode() and SelectNodeTo.  Because
SelectNodeTo didn't check the CSE maps for a preexisting node and didn't insert
its result into the CSE maps, we would sometimes miss a CSE opportunity.

This is extremely rare, but worth fixing for completeness.

llvm-svn: 24565
```
  9d0d715e
- major think-o · 7bb09912
  Andrew Lenharth authored Dec 01, 2005
```
llvm-svn: 24564
```
  7bb09912
- Support multiple ValueTypes per RegisterClass, needed for upcoming vector · 006bb04f
  Nate Begeman authored Dec 01, 2005
```
work.  This change has no effect on generated code.

llvm-svn: 24563
```
  006bb04f
- Cosmetic change, better reflects actual values · aa5f8f2a
  Nate Begeman authored Dec 01, 2005
```
llvm-svn: 24562
```
  aa5f8f2a
- Fix a regression caused by a patch earlier today · bd099102
  Chris Lattner authored Dec 01, 2005
```
llvm-svn: 24561
```
  bd099102
- Flags where I think I need them, quick, before the nightly tester starts · ce68ef80
  Andrew Lenharth authored Dec 01, 2005
```
llvm-svn: 24560
```
  ce68ef80
- Proper support for shifts with register shift value. · 4b024261
  Evan Cheng authored Dec 01, 2005
```
llvm-svn: 24559
```
  4b024261
- Use a getCopyToReg() variant to generate a flaggy CopyToReg node. · d94aa71e
  Evan Cheng authored Dec 01, 2005
```
llvm-svn: 24558
```
  d94aa71e
- SelectNodeTo now returns its result, we must pay attention to it. · 4b25924d
  Chris Lattner authored Nov 30, 2005
```
llvm-svn: 24552
```
  4b25924d
- Pay attn to the node returned by SelectNodeTo · a75694aa
  Chris Lattner authored Nov 30, 2005
```
llvm-svn: 24551
```
  a75694aa
Nov 30, 2005
- SelectNodeTo now returns its result, we must pay attention to it. · af2e0373
  Chris Lattner authored Nov 30, 2005
```
llvm-svn: 24550
```
  af2e0373
- SelectNodeTo now returns N. Use it instead of return N directly. · e3189779
  Chris Lattner authored Nov 30, 2005
```
llvm-svn: 24549
```
  e3189779
- Make SelectNodeTo return N · be5dd5da
  Chris Lattner authored Nov 30, 2005
```
llvm-svn: 24548
```
  be5dd5da
- Fix Regression/CodeGen/PowerPC/2005-11-30-vastart-crash.ll · 3713e6b4
  Chris Lattner authored Nov 30, 2005
```
llvm-svn: 24547
```
  3713e6b4
- Fix a bug where we didn't realize that vaarg reads memory. This fixes · dc4ffef6
  Chris Lattner authored Nov 30, 2005
```
Transforms/DeadStoreElimination/2005-11-30-vaarg.ll

llvm-svn: 24545
```
  dc4ffef6
- Fix a typo in my latest change · 11695c05
  Nate Begeman authored Nov 30, 2005
```
llvm-svn: 24542
```
  11695c05
- No longer track value types for asm printer operands, and remove them as · 6f8c1ace
  Nate Begeman authored Nov 30, 2005
```
an argument to every operand printing function.  Requires some slight
tweaks to x86, the only user.

llvm-svn: 24541
```
  6f8c1ace
- CALLSEQ_START/END nodes don't get memoized, do not add them in when · c1740484
  Chris Lattner authored Nov 30, 2005
```
replaceAllUses'ing.

llvm-svn: 24539
```
  c1740484
- remove redundant code · 7ffe3aff
  Andrew Lenharth authored Nov 30, 2005
```
llvm-svn: 24538
```
  7ffe3aff
- At long last, you can say that f32 isn't supported for setcc · 6ee8566c
  Andrew Lenharth authored Nov 30, 2005
```
llvm-svn: 24537
```
  6ee8566c