Commits · c4e72c9a0a2cf9d4f15ca027c547f9792b4280b8 · Roger Ferrer / llvm-epi-0.8

Apr 14, 2004
- Remove the return type check for llvm.readio. This check is done for all · c4e72c9a
  John Criswell authored Apr 14, 2004
```
functions and is not needed here.
Simplify the pointer type check per Chris's suggestions.

llvm-svn: 12945
```
  c4e72c9a
- Added code to verify that llvm.readio's pointer argument returns something · 0c654c6a
  John Criswell authored Apr 14, 2004
```
that matches its return type.

llvm-svn: 12944
```
  0c654c6a
- Test for memory mapped I/O intrinsics. · 917b6416
  John Criswell authored Apr 14, 2004
```
llvm-svn: 12943
```
  917b6416
- Finish adding the llvm.readio and llvm.writeio intrinsics. · 23c48d63
  John Criswell authored Apr 14, 2004
```
Sorry these didn't get in yesterday.

llvm-svn: 12942
```
  23c48d63
- ADd a trivial instcombine: load null -> null · 6679e46b
  Chris Lattner authored Apr 14, 2004
```
llvm-svn: 12940
```
  6679e46b
- This is the real fix for Codegen/X86/2004-04-13-FPCMOV-Crash.llx which works · 0dc099c2
  Chris Lattner authored Apr 14, 2004
```
even when the "optimization" I added before is turned off.  It generates this
extremely pointless code:

test:
        fld QWORD PTR [%ESP + 4]
        mov %AL, 0
        test %AL, %AL
        fcmove %ST(0), %ST(0)
        ret

Good thing the optimizer will have removed this before code generation
anyway.  :)

llvm-svn: 12939
```
  0dc099c2
- Temporarily hack in the intrinsics that John added. I expect him to finish up... · 52acdaac
  Chris Lattner authored Apr 14, 2004
```
Temporarily hack in the intrinsics that John added.  I expect him to finish up their addition, but in the meantime, the build should not be broken.

llvm-svn: 12938
```
  52acdaac
- Added support for the llvm.readio and llvm.writeio intrinsics. · beded72a
  John Criswell authored Apr 13, 2004
```
On x86, memory operations occur in-order, so these are just lowered into
volatile loads and stores.

llvm-svn: 12936
```
  beded72a
Apr 13, 2004
- Implement a small optimization, which papers over the problem in · 9042e381
  Chris Lattner authored Apr 13, 2004
```
X86/2004-04-13-FPCMOV-Crash.llx

A more robust fix is to follow.

llvm-svn: 12935
```
  9042e381
- Fix bug, add support for +/- · ed94fef7
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12934
```
  ed94fef7
- new testcase that crashes the fp stackifier · 8662beee
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12932
```
  8662beee
- We actually don't have spiff anymore · ed7a7d20
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12929
```
  ed7a7d20
- Out with spiff, in with fpcmp · 9a3ac1a9
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12928
```
  9a3ac1a9
- Remove spiff. Though it looked good, it was not really as spiffy as it seemed · e542c1c4
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12927
```
  e542c1c4
- Okay, spiff is completely incapable of handling files of nontrivial size. · 5613bcc2
  Chris Lattner authored Apr 13, 2004
```
Here is a simple minimal program that does what we want.  Instead of taking
minutes to compare mesa's output, and crashing on binary files (like spiff
does), this take < .02s in the common case and doesn't crash.

llvm-svn: 12926
```
  5613bcc2
- Add a note · cb5e795f
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12922
```
  cb5e795f
- Add SCCP support for constant folding calls, implementing: · ff9362a8
  Chris Lattner authored Apr 13, 2004
```
test/Regression/Transforms/SCCP/calltest.ll

llvm-svn: 12921
```
  ff9362a8
- New testcase · 9d420840
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12920
```
  9d420840
- Add a simple call constant propagation interface. · ca52d046
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12919
```
  ca52d046
- Add new interfaces · 7dacdb6f
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12918
```
  7dacdb6f
- Constant propagation should remove the dead instructions · d0dc6d52
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12917
```
  d0dc6d52
- New testcase · e3debac3
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12916
```
  e3debac3
- Regenerated using autoconf-2.57. · 9a79ed7b
  Brian Gaeke authored Apr 13, 2004
```
llvm-svn: 12915
```
  9a79ed7b
- Rewrite POV-Ray check as per PR301 · fca47205
  Brian Gaeke authored Apr 13, 2004
```
llvm-svn: 12914
```
  fca47205
- Increase buffer size again. *sigh* · becf6365
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12912
```
  becf6365
- Don't index spiff. It should not count as lines of LLVM code · f37bd007
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12911
```
  f37bd007
- Fixed sized limits are bad... duh. · 763596e2
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12908
```
  763596e2
- Really spiff, don't chatter :) · 7b9733da
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12907
```
  7b9733da
- I don't think we have to have 4 extra allocated (but unused) bytes on the stack. · 6f6e4b9f
  Brian Gaeke authored Apr 13, 2004
```
llvm-svn: 12905
```
  6f6e4b9f
- I started working on casts, but I don't have anything compilable yet. · 969795b4
  Brian Gaeke authored Apr 13, 2004
```
llvm-svn: 12903
```
  969795b4
- Don't underline diffs · bacab88d
  Brian Gaeke authored Apr 13, 2004
```
llvm-svn: 12902
```
  bacab88d
- Emit the immediate form of in/out when possible. · c71b0966
  Chris Lattner authored Apr 13, 2004
```
Fix several bugs in the intrinsics:
  1. Make sure to copy the input registers before the instructions that use them
  2. Make sure to copy the value returned by 'in' out of EAX into the register
     it is supposed to be in.

This fixes assertions when using in/out and linear scan.

llvm-svn: 12896
```
  c71b0966
- Add immediate forms of in/out. Use let to shorten lines · a340febe
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12895
```
  a340febe
- Add support for new instruction type · eaa28a29
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12894
```
  eaa28a29
- Add support for the printImplicitDefsBefore flag · 27b2e2cd
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12893
```
  27b2e2cd
- Make the testcase more challenging · e342a8ef
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12892
```
  e342a8ef
- Reenable library now that the loopsimplify bug is fixed · 34fba8eb
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12891
```
  34fba8eb
- Fix LoopSimplify/2004-04-13-LoopSimplifyUpdateDomFrontier.ll · 89e959bb
  Chris Lattner authored Apr 13, 2004
```
LoopSimplify was not updating dominator frontiers correctly in some cases.

llvm-svn: 12890
```
  89e959bb
- New testcase where loop simplify is not updating domfrontiers correctly · a7aeb506
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12889
```
  a7aeb506
- Refactor code a bit to make it simpler and eliminate the goto · a6e22814
  Chris Lattner authored Apr 13, 2004
```
llvm-svn: 12888
```
  a6e22814