Commits · b52220cd21c88ee8d8b3fc2ced4f3982294c37f8 · Roger Ferrer / llvm-epi-0.8

Oct 11, 2002
- Use PARALLEL_DIRS. · b52220cd
  Vikram S. Adve authored Oct 11, 2002
```
llvm-svn: 4114
```
  b52220cd
- Added capability to get execution count of a loop if it is a predictable · 33022f07
  Misha Brukman authored Oct 11, 2002
```
number of iterations.

llvm-svn: 4113
```
  33022f07
- Added helper functions in LoopInfo: isLoopExit and numBackEdges. · 3845be20
  Misha Brukman authored Oct 11, 2002
```
llvm-svn: 4112
```
  3845be20
Oct 10, 2002
- Stop using DataStructureGraph.h · 1a535e18
  Chris Lattner authored Oct 10, 2002
```
llvm-svn: 4106
```
  1a535e18
- - Dramatically simplify the ConstantMerge code now that · eac4dcd0
  Chris Lattner authored Oct 09, 2002
```
    Value::replaceAllUsesWith works with constants correctly.

llvm-svn: 4104
```
  eac4dcd0
- - Make Value::replaceAllUsesWith work with constants correctly. This fixes · 2c6abeac
  Chris Lattner authored Oct 09, 2002
```
    bug FuncResolve/2002-08-19-ResolveGlobalVars.ll and gzip looks better.

llvm-svn: 4103
```
  2c6abeac
- - Add new Constant::replaceUsesOfWithOnConstant which has an end result · b1dd9bbc
  Chris Lattner authored Oct 09, 2002
```
    similar to User::replaceUsesOfWith but actually does the right thing for
    constants.

llvm-svn: 4102
```
  b1dd9bbc
Oct 09, 2002
- Almost a complete rewrite of FunctionResolution to now resolve functions · 013eca00
  Chris Lattner authored Oct 09, 2002
```
and global variables.

This fixes bug: FuncResolve/2002-08-19-ResolveGlobalVarsEasier.ll
And bug: SingleSource/UnitTests/2002-10-09-ArrayResolution.c

Note that this does not fix bug:
FunctionResolve/2002-08-19-ResolveGlobalVars.ll because replaceAllUsesWith
breaks when a constantexpr is pointing to the thing being replaced.  This
is more of an infrastructure problem than anything.

llvm-svn: 4099
```
  013eca00
- Minor, non-functionality changing, formatting fix · fa01a521
  Chris Lattner authored Oct 09, 2002
```
llvm-svn: 4091
```
  fa01a521
- - Remove Value::use_remove · 6023f638
  Chris Lattner authored Oct 09, 2002
```
llvm-svn: 4090
```
  6023f638
- - Rename MTy to FTy (no methods exist anymore) · a4e2653c
  Chris Lattner authored Oct 09, 2002
```
  - Fix bug: LevelRaise/2002-10-08-VarArgCallInfLoop.ll

llvm-svn: 4088
```
  a4e2653c
- Fix NASTY N^2 behavior that was causing the gzip benchmark to take forever to · 3ea0c188
  Chris Lattner authored Oct 08, 2002
```
assemble.  Now we scan the use-list from the back when removing users instead
of from the front.

llvm-svn: 4086
```
  3ea0c188
- Fix bug: Assembler/2002-10-08-LargeArrayPerformance.ll by using · 0d779712
  Chris Lattner authored Oct 08, 2002
```
std::vector::reserve when possible

llvm-svn: 4085
```
  0d779712
- - Fix bug: LevelRaise/2002-10-08-VarArgCall.ll · a185db3d
  Chris Lattner authored Oct 08, 2002
```
llvm-svn: 4083
```
  a185db3d
Oct 08, 2002
- - Fix bug: cee/2002-10-07-NoImmediateDominator.ll · b7368500
  Chris Lattner authored Oct 08, 2002
```
llvm-svn: 4081
```
  b7368500
- Changes to support PHINode::removeIncoming changes · caf5b504
  Chris Lattner authored Oct 08, 2002
```
llvm-svn: 4080
```
  caf5b504
- · a704ac8a
  Chris Lattner authored Oct 08, 2002
```
Changes to support PHINode::removeIncoming changes

llvm-svn: 4079
```
  a704ac8a
- - Change PHINode::removeIncomingValue to delete the phi node if the last · fae98a2d
  Chris Lattner authored Oct 08, 2002
```
    incoming value is removed!

llvm-svn: 4078
```
  fae98a2d
- - Checkin LARGE number of Changes to CEE pass that will make it much more · 736709fe
  Chris Lattner authored Oct 08, 2002
```
    powerful, but that are largely disabled.  The basic idea here is that it
    is trying to forward branches across basic blocks that have PHI nodes in
    it, which are crucial to be able to handle cases like whet.ll.
    Unfortunately we are not updating SSA correctly, causing sim.c to die, and I
    don't have time to fix the regression now, so I must disable the
    functionality.

llvm-svn: 4077
```
  736709fe
- Expose isCriticalEdge & SplitCriticalEdge methods from crit-edges pass · 5ac72def
  Chris Lattner authored Oct 08, 2002
```
llvm-svn: 4075
```
  5ac72def
- Expose new "recalculate" method from dominatorset · acb038e5
  Chris Lattner authored Oct 08, 2002
```
llvm-svn: 4074
```
  acb038e5
- It is illegal for PHI nodes to have zero values, delete the code to handle them · 394617f3
  Chris Lattner authored Oct 08, 2002
```
llvm-svn: 4071
```
  394617f3
- Fold ashr -1, X into -1 · 2e0fb39d
  Chris Lattner authored Oct 08, 2002
```
llvm-svn: 4070
```
  2e0fb39d
Oct 07, 2002
- Non-functionality change just to make it more clear what is going on · 68f21bcb
  Chris Lattner authored Oct 07, 2002
```
llvm-svn: 4060
```
  68f21bcb
- Global variables are now external if they don't have initializers, not · d131986b
  Chris Lattner authored Oct 06, 2002
```
"uninitialized"

llvm-svn: 4052
```
  d131986b
- Check that we don't have external varaibles with internal linkage · 78bc0fa2
  Chris Lattner authored Oct 06, 2002
```
llvm-svn: 4051
```
  78bc0fa2
- The parser now accepts "external" global variables, in addition to the backwards · e2dd92be
  Chris Lattner authored Oct 06, 2002
```
compatible "uninitialized" global variables

llvm-svn: 4050
```
  e2dd92be
Oct 06, 2002
- PHI nodes are not allowed to exist with zero incoming values, check that · 4cd9df88
  Chris Lattner authored Oct 06, 2002
```
there aren't any like this.

llvm-svn: 4044
```
  4cd9df88
- Bug fix: In preventing static global variables from being printed twice, · 2cc6609e
  Vikram S. Adve authored Oct 05, 2002
```
I also prevented external globals from being printed twice, but they
should (extern declaration and definition).

llvm-svn: 4043
```
  2cc6609e
Oct 05, 2002
- Added #include<unistd.h> to compile with solaris gcc3.2 · b8aef8ef
  Anand Shukla authored Oct 04, 2002
```
llvm-svn: 4042
```
  b8aef8ef
- added cast to unsigned to compile with gcc3.2 (sparc) · 195d3fb3
  Anand Shukla authored Oct 04, 2002
```
llvm-svn: 4041
```
  195d3fb3
Oct 04, 2002

Handle post dominance correctly in the case where blocks do not have a path to · d7b3bb0f
Chris Lattner authored Oct 04, 2002
```
the exit node.

llvm-svn: 4038
```
d7b3bb0f

Fix a nasty problem with dominance calculation for unreachable blocks. · 2cedf1ce

Chris Lattner authored Oct 04, 2002

If we had a CFG that look like Entry -> B, Unreachable -> B, then we would
not correctly determine that Entry dominated B, because Entry did not
apparently dominate "unreachable".  This patch fixes this by making the entry
node dominate all blocks, including unreachable ones.

llvm-svn: 4037

2cedf1ce

Oct 03, 2002
- Prune function nodes that are no longer referenced due to inlining · 9df1cf30
  Chris Lattner authored Oct 03, 2002
```
llvm-svn: 4036
```
  9df1cf30
- sgefa uses truely huge data structures nodes. Only print part of them if they · 9dfb9f44
  Chris Lattner authored Oct 03, 2002
```
are so big

llvm-svn: 4035
```
  9dfb9f44
- Handle bug exposed by power benchmark · 4ee4381a
  Chris Lattner authored Oct 03, 2002
```
llvm-svn: 4033
```
  4ee4381a
- Add parens around constant calls to getelemptr to properly associate the reference · afff0a25
  Nick Hildenbrandt authored Oct 03, 2002
```
llvm-svn: 4032
```
  afff0a25
- Reimplement/port the Bottom Up Closure pass · c9c681e4
  Chris Lattner authored Oct 03, 2002
```
llvm-svn: 4031
```
  c9c681e4
- DataStructure.h doesn't include DSGraph.h · 3f69054d
  Chris Lattner authored Oct 02, 2002
```
llvm-svn: 4029
```
  3f69054d
Oct 02, 2002
- I was wrong on the removing of those references on the last commit. · 2d3eb5a3
  Nick Hildenbrandt authored Oct 02, 2002
```
llvm-svn: 4023
```
  2d3eb5a3