6c81ebc97d
Add comment
...
llvm-svn: 6415
2003-05-29 20:26:30 +00:00
3442844742
Fix bug: Instcombine/2003-05-27-ConstExprCrash.ll
...
llvm-svn: 6352
2003-05-27 16:40:51 +00:00
c16b210349
* Actually USE the statistic that we made
...
* Implement SRoA for arrays
llvm-svn: 6349
2003-05-27 16:09:27 +00:00
fb41a50d15
Implementation of the simple "scalar replacement of aggregates" transformation
...
llvm-svn: 6346
2003-05-27 15:45:27 +00:00
196897c424
Fix bug: InstCombine/2003-05-26-CastMiscompile.ll
...
llvm-svn: 6338
2003-05-26 23:41:32 +00:00
8d0a71aff8
Remove using declarations
...
llvm-svn: 6306
2003-05-22 22:00:07 +00:00
471bd76281
Minor cleanups.
...
This hunk:
- } else if (Src->getNumOperands() == 2 && Src->use_size() == 1) {
+ } else if (Src->getNumOperands() == 2) {
Allows GEP folding to be more aggressive, which reduces the number of instructions
and can dramatically speed up BasicAA in some cases.
llvm-svn: 6286
2003-05-22 19:07:21 +00:00
373086db75
Hopefully, the final fix for `[Pp]ropogate'.
...
llvm-svn: 6251
2003-05-20 21:01:22 +00:00
e5838c4d72
s/convertable/convertible/g
...
llvm-svn: 6248
2003-05-20 18:45:36 +00:00
652e7bf8a8
Fix long standing bug
...
llvm-svn: 6232
2003-05-15 18:25:13 +00:00
0895025b30
Fix bug: LoopPreheaders/2003-05-12-PreheaderExitOfChild.ll
...
llvm-svn: 6153
2003-05-12 22:04:34 +00:00
b02697210e
Handle va_arg instruction correctly
...
llvm-svn: 6030
2003-05-08 02:50:13 +00:00
3666378690
Fix spelling
...
llvm-svn: 5983
2003-05-02 19:26:34 +00:00
df7cc91159
Fix Bug: LowerSwitch/2003-05-01-PHIProblem.ll
...
llvm-svn: 5979
2003-05-01 22:21:21 +00:00
dd65d863a9
Fix another case where constexprs could cause a crash
...
llvm-svn: 5972
2003-04-30 22:34:06 +00:00
9244df60d3
Fix constant folding of constexprs
...
llvm-svn: 5971
2003-04-30 22:19:10 +00:00
b68a34eb3f
Fix bug: ADCE/2003-04-25-PHIPostDominateProblem.ll
...
llvm-svn: 5952
2003-04-25 22:53:27 +00:00
3c98276168
Fix major problem with PHI node traversal: rechecking PHIs should go into
...
the visit function. The worklist is for when their lattice value changes.
llvm-svn: 5927
2003-04-25 03:35:10 +00:00
cccc5c76c8
Minor non-functional changes:
...
* Spell propagate right
* Improve performance of phi node handling
* Delete using directive
* Other minor changes
llvm-svn: 5920
2003-04-25 02:50:03 +00:00
889f620841
Remove unnecesary &*'s
...
llvm-svn: 5872
2003-04-23 16:37:45 +00:00
1b094a0f2e
New pass to lower switch instructions to branch instructions
...
llvm-svn: 5865
2003-04-23 16:23:59 +00:00
1f45f4e4d8
Minor cleanups
...
llvm-svn: 5863
2003-04-23 16:18:14 +00:00
46b3d30bf6
Change the interface to constant expressions to allow automatic folding
...
llvm-svn: 5793
2003-04-16 22:40:51 +00:00
9023155621
Initial checkin of PRE on LLVM. This implementation is still lacking in
...
several ways:
* Load expressions are not PRE'd well. Alias Analysis should be used to
get accurate information when computing anticipatibility.
* The expression collection implementation does not handle PHI nodes properly,
thus the implementation misses many opportunities to PRE.
* This code could be sped up quite a bit
Despite these flaws, the code seems to work well, and handles PR's as one
would expect.
llvm-svn: 5759
2003-03-31 19:55:43 +00:00
c42046f9f1
Move BreakCriticalEdges pass to lib/Transforms/Utils
...
llvm-svn: 5754
2003-03-21 21:43:19 +00:00
7fb29e19f0
Add the following instcombine xforms:
...
- Implement simple reassociation: (A|c1)|(B|c2) == (A|B)|(c1|c2)
- (A & C1)+(B & C2) -> (A & C1)|(B & C2) iff C1&C2 == 0
- (A & C1)^(B & C2) -> (A & C1)|(B & C2) iff C1&C2 == 0
llvm-svn: 5743
2003-03-11 00:12:48 +00:00
934a64cfa4
Implement: -A*-B == A*B
...
llvm-svn: 5740
2003-03-10 23:23:04 +00:00
3e327a4334
Add new transformation: // (~A | ~B) == (~(A & B))
...
llvm-svn: 5738
2003-03-10 23:13:59 +00:00
bb74e2284a
Generalize not and neg comparison testers to allow constant to be considered not'able and neg'able. This
...
allows optimization of this:
int %test4(int %A, int %B) {
%a = xor int %A, -1
%c = and int %a, 5 ; 5 = ~c2
%d = xor int %c, -1
ret int %d
}
into this:
int %test4(int %A, int %B) { ; No predecessors!
%c.demorgan = or int %A, -6 ; <int> [#uses=1]
ret int %c.demorgan
}
llvm-svn: 5736
2003-03-10 23:06:50 +00:00
dcf240a649
Generalize (A+c1)+c2 optimization to work with all associative operators
...
llvm-svn: 5733
2003-03-10 21:43:22 +00:00
3695257724
Minor change, no functionality diff
...
llvm-svn: 5731
2003-03-10 19:20:30 +00:00
e8d6c604bb
Fix bug: (x << 100) wasn't folded to 0, but (x >> 100) was (when x is unsigned)
...
Implement new shift optimizations for shifting the result of a shift.
llvm-svn: 5729
2003-03-10 19:16:08 +00:00
1bbb7b6f19
Implement: (A|B)^B == A & (~B)
...
llvm-svn: 5728
2003-03-10 18:24:17 +00:00
235af5673b
Implement %test7 in InstCombine/getelementptr.ll
...
llvm-svn: 5704
2003-03-05 22:33:14 +00:00
0592bb7654
Convert LICM over to use AliasSetTracker. Besides being nicer, this automatically
...
allows LICM to use access sizes to help alias analysis be more precise.
llvm-svn: 5693
2003-03-03 23:32:45 +00:00
e27406eb59
Change the mem2reg interface to accept a TargetData argument
...
llvm-svn: 5685
2003-03-03 17:25:18 +00:00
e19a9f8ef7
Fix bug: LICM/2003-02-28-PromoteDifferentType.ll
...
llvm-svn: 5675
2003-02-28 19:21:40 +00:00
49eb0e3026
Fix bug I introduced yesterday :(
...
llvm-svn: 5669
2003-02-28 16:54:17 +00:00
32a39c26e0
Fix bug: LICM/2003-02-27-PreheaderExitNodeUpdate.ll
...
llvm-svn: 5667
2003-02-28 03:07:54 +00:00
f2d9f94a33
Fix bug: 2003-02-27-PreheaderExitNodeUpdate.ll by updating exit node info
...
llvm-svn: 5664
2003-02-27 22:48:57 +00:00
10b2b055e8
Add a new assertion to check that stuff is happening right
...
Ironically the exit block modification code wasn't updating the exit block
information itself. Fix this.
llvm-svn: 5659
2003-02-27 22:31:07 +00:00
1ad80e2511
Fix bug: 2003-02-27-StoreSinkPHIs.ll
...
llvm-svn: 5658
2003-02-27 21:59:36 +00:00
4e2fbfb4c3
* Don't forget to update Loop information!
...
* Remove bogus assertion: there may be a single outside predecessor and
still need a new loop-preheader if the predecessor has multiple
successors. See bug: LICM/2003-02-27-PreheaderProblem.ll
llvm-svn: 5655
2003-02-27 21:50:19 +00:00
650096a061
* Significant changes to the preheader insertion pass:
...
- Now we perform loop exit-block splitting to ensure exit blocks are
always dominated by the loop header.
- We now preserve dominance frontier information
- This fixes bug: LICM/2003-02-26-LoopExitNotDominated.ll
llvm-svn: 5652
2003-02-27 20:27:08 +00:00
71b82bb494
Convert to work with new AliasAnalysis interface by conservatively assuming all pointers are arbitrarily large accesses
...
llvm-svn: 5636
2003-02-26 19:28:57 +00:00
4869f3700e
Rename Instruction::hasSideEffects() -> mayWriteToMemory()
...
llvm-svn: 5620
2003-02-24 20:48:32 +00:00
45d67d6ea7
Initial implementation of Loop invariant memory->scalar promotion
...
llvm-svn: 5613
2003-02-24 03:52:32 +00:00
26f159046c
Split mem2reg promotion into two parts: a function which does the work, and
...
a pass which wraps the function. This allows other passes to use the functionality
llvm-svn: 5610
2003-02-22 23:57:48 +00:00
57c8d99bbd
4 new transformations:
...
* X*C + X --> X * (C+1)
* X + X*C --> X * (C+1)
* X - X*C --> X * (1-C)
* X*C - X --> X * (C-1)
llvm-svn: 5592
2003-02-18 19:57:07 +00:00
3082c5a070
Add a variety of new transformations:
...
* A & ~A == 0
* A / (2^c) == A >> c if unsigned
* 0 / A == 0
* 1.0 * A == A
* A * (2^c) == A << c
* A ^ ~A == -1
* A | ~A == -1
* 0 % X = 0
* A % (2^c) == A & (c-1) if unsigned
* A - (A & B) == A & ~B
* -1 - A == ~A
llvm-svn: 5587
2003-02-18 19:28:33 +00:00
f683599915
Fix typo
...
llvm-svn: 5458
2003-02-01 04:50:59 +00:00
6fa1629c47
Fix bug: ADCE/2003-01-22-PredecessorProblem.ll
...
llvm-svn: 5405
2003-01-23 02:12:18 +00:00
3aa7767352
Fix a huge performance problem in reassociate by introducing a
...
rank map cache for instruction ranks
llvm-svn: 5030
2002-12-15 03:56:00 +00:00
f96c8befe0
Minor changes:
...
* Reword comment to make more clear
* Don't print out BB's after modification made
* Don't delete and new an instruction when we need to move something, just move it.
llvm-svn: 5029
2002-12-15 03:49:50 +00:00
688ac5b723
Remove dead code
...
llvm-svn: 4952
2002-12-07 18:42:13 +00:00
c6509f487b
Fix bug: 2002-12-05-MissedConstProp.ll pointed out by Casey Carter
...
llvm-svn: 4935
2002-12-05 22:41:53 +00:00
98cf1f5d64
- Eliminated the deferred symbol table stuff in Module & Function, it really
...
wasn't an optimization and it was causing lots of bugs.
llvm-svn: 4779
2002-11-20 18:36:02 +00:00
a2620ac156
Fix warning
...
llvm-svn: 4649
2002-11-09 00:49:43 +00:00
0da31d6648
Fix spelling error
...
llvm-svn: 4645
2002-11-08 23:18:37 +00:00
5d606a09ff
Instcombine this away:
...
%A = getelementptr [4 x int]* %M, long 0, long 0
%B = getelementptr int* %A, long 2
llvm-svn: 4536
2002-11-04 16:43:32 +00:00
1085bdf49f
Add a transformation to turn:
...
malloc Ty, C
int
malloc [C x Ty], 1
llvm-svn: 4534
2002-11-04 16:18:53 +00:00
669c6cf459
Make sure to _delete_ memory allocated by worklists
...
llvm-svn: 4523
2002-11-04 02:54:22 +00:00
8fdf75cde5
Fixes to the reassociate pass to make it respect dominance properties
...
Huge thanks go to Casey Carter for writing this fix, reassociate is now
reoperational!
llvm-svn: 4471
2002-10-31 17:12:59 +00:00
12764c85e8
BreakCriticalEdges should update dominance frontier information as well as
...
other dominance stuff. Patch contributed by Casey Carter
llvm-svn: 4457
2002-10-31 02:44:36 +00:00
632df28deb
Fix spelling of `propagate'.
...
llvm-svn: 4423
2002-10-29 23:06:16 +00:00
820d971233
- Rename AnalysisUsage::preservesAll to getPreservesAll & preservesCFG to
...
setPreservesCFG to be less confusing.
llvm-svn: 4255
2002-10-21 20:00:28 +00:00
b7368500dc
- Fix bug: cee/2002-10-07-NoImmediateDominator.ll
...
llvm-svn: 4081
2002-10-08 21:53:51 +00:00
736709feda
- Checkin LARGE number of Changes to CEE pass that will make it much more
...
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
2002-10-08 21:34:15 +00:00
5ac72defc9
Expose isCriticalEdge & SplitCriticalEdge methods from crit-edges pass
...
llvm-svn: 4075
2002-10-08 21:06:27 +00:00
394617f38b
It is illegal for PHI nodes to have zero values, delete the code to handle them
...
llvm-svn: 4071
2002-10-08 17:07:39 +00:00
2e0fb39d87
Fold ashr -1, X into -1
...
llvm-svn: 4070
2002-10-08 16:16:40 +00:00
bf3a099a62
Updates to work with recent Statistic's changes:
...
* Renamed StatisticReporter.h/cpp to Statistic.h/cpp
* Broke constructor to take two const char * arguments instead of one, so
that indendation can be taken care of automatically.
* Sort the list by pass name when printing
* Make sure to print all statistics as a group, instead of randomly when
the statistics dtors are called.
* Updated ProgrammersManual with new semantics.
llvm-svn: 4002
2002-10-01 22:38:41 +00:00
05e86302f4
Fix bug in LICM that caused the previous big win. :(
...
llvm-svn: 3980
2002-09-29 22:26:07 +00:00
6443769b34
Hoist the contents of Loops in depth first order in the dominator tree,
...
rather than in random order. This causes LICM to be DRAMATICALLY more
effective. For example, on bzip2.c, it is able to hoist 302 loads and
2380 total instructions, as opposed to 44/338 before. This obviously
makes other transformations much more powerful as well!
llvm-svn: 3978
2002-09-29 21:46:09 +00:00
03a9e15a9f
Fix major bugs in dominator set & tree information updating
...
llvm-svn: 3975
2002-09-29 21:41:38 +00:00
d57f3f5073
- Further cleanups of LICM pass, remove extra work from previous implementation
...
- Do not clone instructions then insert clone outside of loop. Just move them.
llvm-svn: 3951
2002-09-26 19:40:25 +00:00
f64f2d3fb3
Improve comments, doxygenize more
...
llvm-svn: 3950
2002-09-26 16:52:07 +00:00
718b221ff8
Clean up LICM significantly now that it is guaranteed to have loop preheaders
...
llvm-svn: 3947
2002-09-26 16:38:03 +00:00
ca056914c6
Change pass name to something sane
...
llvm-svn: 3946
2002-09-26 16:37:37 +00:00
d771fdfd75
Loop invariant code motion now depends on the LoopPreheader pass. Dead code
...
has not yet been removed.
llvm-svn: 3945
2002-09-26 16:19:31 +00:00
bedbd6bc16
- Cleanup break-crit-edges pass by making SplitCriticalEdge a member method.
...
- break-crit-edges pass does not invalidate loop-preheader pass.
llvm-svn: 3944
2002-09-26 16:18:51 +00:00
61992f66d8
Checkin new loop-preheader insertion pass.
...
llvm-svn: 3942
2002-09-26 16:17:31 +00:00
3cab9f0396
Change LowerAllocations pass to 'require' TargetData instead of it being
...
passed in.
llvm-svn: 3929
2002-09-25 23:47:47 +00:00
4357e54760
Eliminate extraneous #include
...
llvm-svn: 3916
2002-09-24 21:18:40 +00:00
75f80bd0b6
- Do not expose Critical Edge breaking mechanics outside the BCE pass, thus
...
removing it from Transforms/Local.h and from Transforms/Utils/*
llvm-svn: 3910
2002-09-24 15:51:56 +00:00
b4e7e22be3
Correlated Exprs pass now requires BCE pass instead of doing it manually
...
llvm-svn: 3908
2002-09-24 15:43:56 +00:00
4bec665b09
- Expose passinfo from BreakCriticalEdges pass so that it may be "Required"
...
by other passes. Make BCE pass be in anonymous namespace now.
llvm-svn: 3907
2002-09-24 15:43:12 +00:00
b03832d5d0
Add new BreakCriticalEdges pass
...
llvm-svn: 3903
2002-09-24 00:08:39 +00:00
555518c70f
Optimize away cases like:
...
%cast109 = cast uint %cast212 to short ; <short> [#uses=1]
%cast214 = cast short %cast109 to uint ; <uint> [#uses=1]
%cast215 = cast uint %cast214 to short ; <short> [#uses=1]
llvm-svn: 3897
2002-09-23 23:39:43 +00:00
7ccb217f21
* Fix bug: CorrelatedExprs/2002-09-23-PHIUpdateBug.ll
...
* Make sure "Changed" is updated correctly
llvm-svn: 3891
2002-09-23 20:06:22 +00:00
a8339e3613
Fix bug: test/Regression/Transforms/InstCombine/2002-09-17-GetElementPtrCrash.ll
...
Thanks to Casey for finding it!
llvm-svn: 3783
2002-09-17 21:05:42 +00:00
dba59921d7
Extract most of the transformation into an externally accessible
...
function -- DecomposeArrayRef(GetElementPtrInst* GEP) -- that can
be invoked on a single instruction at a time.
llvm-svn: 3755
2002-09-16 16:40:07 +00:00
2e8086f046
Function.h is unnecessary when Module.h is included.
...
llvm-svn: 3716
2002-09-14 03:04:02 +00:00
136dab7d86
- Change getelementptr instruction to use long indexes instead of uint
...
indexes for sequential types.
llvm-svn: 3682
2002-09-11 01:21:33 +00:00
cbc532603a
Fix the last of the silly bugs I just introduced. :(
...
llvm-svn: 3674
2002-09-10 23:46:10 +00:00
55f3d94b40
Add cannonicalization of shl X, 1 -> add X, X
...
llvm-svn: 3671
2002-09-10 23:04:09 +00:00
a239e6879d
Clean up code due to auto-insert constructors
...
llvm-svn: 3665
2002-09-10 22:38:47 +00:00
28a8d2468a
Simplify code (somtimes dramatically), by using the new "auto-insert" feature
...
of instruction constructors.
llvm-svn: 3656
2002-09-10 17:04:02 +00:00
4184bcc701
* Clean up code a little bit
...
* Fix bug: test/Regression/Transforms/IndVarsSimplify/2002-09-09-PointerIndVar.ll
llvm-svn: 3644
2002-09-10 05:24:05 +00:00
e103d178c5
Hack unneccesary now that shifts of pointers are no longer legal!
...
llvm-svn: 3640
2002-09-10 03:50:54 +00:00
Chris Lattner