wrapper returns a vector of integers when passed a vector of pointers) by having
getIntPtrType itself return a vector of integers in this case.  Outside of this
wrapper, I didn't find anywhere in the codebase that was relying on the old
behaviour for vectors of pointers, so give this a whirl through the buildbots.

llvm-svn: 166939

Change the PassManagerBuilder (used by -O3) loop vectorizer flag from -vectorize to -vectorize-loops because we dont want to share the same flag as the bb-vectorizer.

llvm-svn: 166937

split module can see each other. If it is keeping a symbol that already has
a non local linkage, it doesn't need to change it.

llvm-svn: 166908

output of both

llvm-extract foo.ll -func=bar
and
llvm-extract foo.ll -func=bar -delete

so the two new files could not be linked together anymore. With this change
alias are handled almost like functions and global variables. Almost because
with alias we cannot just clear the initializer/body, we have to create a new
declaration and replace the alias with it.

The net result is that now the output of the above commands can be linked
even if foo.ll has aliases.

llvm-svn: 166907

I don't think this is possible with the current implementation but that may change eventually.

llvm-svn: 166877

This turns loops like
  for (unsigned i = 0; i != n; ++i)
    p[i] = p[i+1];
into memmove, which has a highly optimized implementation in most libcs.

This was really easy with the new DependenceAnalysis :)

llvm-svn: 166875

Requires a lot less code and complexity on loop-idiom's side and the more
precise analysis can catch more cases, like the one I included as a test case.
This also fixes the edge-case miscompilation from PR9481.

Compile time performance seems to be slightly worse, but this is mostly due
to an extra LCSSA run scheduled by the PassManager and should be fixed there.

llvm-svn: 166874

The monolithic interface for instruction costs has been split into
several functions. This is the corresponding change. No functionality
change is intended.

llvm-svn: 166865

1. Fix a bug in getTypeConversion. When a *simple* type is split, we need to return the type of the split result.
2. Change the maximum vectorization width from 4 to 8.
3. A test for both.

llvm-svn: 166864

Refactor the VectorTargetTransformInfo interface.

Add getCostXXX calls for different families of opcodes, such as casts, arithmetic, cmp, etc.

Port the LoopVectorizer to the new API.

The LoopVectorizer now finds instructions which will remain uniform after vectorization. It uses this information when calculating the cost of these instructions.

llvm-svn: 166836

list of externals. This makes sense since a shared library with no symbols
can still be useful if it has static constructors.

llvm-svn: 166795

This is currently true, but may change when DA grows more aggressive caching.
Without this setting it's impossible to use DA from a LoopPass because DA is a
function pass and cannot be properly scheduled in between LoopPasses. The
LoopManager reacts to this with an infinite loop which made this really annoying
to debug.

llvm-svn: 166788

The LoopSimplify bug is pretty harmless because the loop goes from unanalyzable
to analyzable but the LCSSA bug is very nasty. It only comes into play with a
specific order of the LoopPassManager worklist and can cause actual
miscompilations, when a SCEV refers to a value that has been replaced with PHI
node. SCEVExpander may then insert code into the wrong place, either violating
domination or randomly miscompiling stuff.

Comes with an extensive test case reduced from the test-suite with
bugpoint+SCEVValidator.

llvm-svn: 166787

This change reflects VTTI refactoring; no functionality change intended.

llvm-svn: 166752

Once vector-of-pointer support works, then this can be reverted.

llvm-svn: 166741

This is needed so that perl's SHA can be compiled (otherwise
BBVectorize takes far too long to find its fixed point).

I'll try to come up with a reduced test case.

llvm-svn: 166738

This is the first of several steps to incorporate information from the new
TargetTransformInfo infrastructure into BBVectorize. Two things are done here:

 1. Target information is used to determine if it is profitable to fuse two
    instructions. This means that the cost of the vector operation must not
    be more expensive than the cost of the two original operations. Pairs that
    are not profitable are no longer considered (because current cost information
    is incomplete, for intrinsics for example, equal-cost pairs are still
    considered).

 2. The 'cost savings' computed for the profitability check are also used to
    rank the DAGs that represent the potential vectorization plans. Specifically,
    for nodes of non-trivial depth, the cost savings is used as the node
    weight.

The next step will be to incorporate the shuffle costs into the DAG weighting;
this will give the edges of the DAG weights as well. Once that is done, when
target information is available, we should be able to dispense with the
depth heuristic.

llvm-svn: 166716

llvm-svn: 166715

The isValueEqualityComparison() guard at the top of SimplifySwitch()
only applies to some of the possible transformations.

The newer transformations work just fine on large switches, and the
check on predecessor count is nonsensical.

llvm-svn: 166710

smaller integer loads and stores.

The high-level motivation is that the frontend sometimes generates
a single whole-alloca integer load or store during ABI lowering of
splittable allocas. We need to be able to break this apart in order to
see the underlying elements and properly promote them to SSA values. The
hope is that this fixes some performance regressions on x86-32 with the
new SROA pass.

Unfortunately, this causes quite a bit of churn in the test cases, and
bloats some IR that comes out. When we see an alloca that consists soley
of bits and bytes being extracted and re-inserted, we now do some
splitting first, before building widened integer "bucket of bits"
representations. These are always well folded by instcombine however, so
this shouldn't actually result in missed opportunities.

If this splitting of all-integer allocas does cause problems (perhaps
due to smaller SSA values going into the RA), we could potentially go to
some extreme measures to only do this integer splitting trick when there
are non-integer component accesses of an alloca, but discovering this is
quite expensive: it adds yet another complete walk of the recursive use
tree of the alloca.

Either way, I will be watching build bots and LNT bots to see what
fallout there is here. If anyone gets x86-32 numbers before & after this
change, I would be very interested.

llvm-svn: 166662

Patch by Paul Redmond <paul.redmond@intel.com>.

llvm-svn: 166649

llvm-svn: 166643

llvm-svn: 166642

llvm-svn: 166634

GVN will now generate ptrtoint instructions for vectors of pointers.
Fixes PR14166.

llvm-svn: 166624

llvm-svn: 166622

llvm-svn: 166620

llvm-svn: 166607

Back out r166591, not sure why this made it through since I cancelled the command. Bleh, sorry about this!

llvm-svn: 166596

llvm-svn: 166591

This checkin also adds in some tests that utilize these paths and updates some of the
clients.

llvm-svn: 166578

Use the AliasAnalysis isIdentifiedObj because it also understands mallocs and c++ news.

PR14158.

llvm-svn: 166491

llvm-svn: 166475

 %V = mul i64 %N, 4
 %t = getelementptr i8* bitcast (i32* %arr to i8*), i32 %V
into
 %t1 = getelementptr i32* %arr, i32 %N
 %t = bitcast i32* %t1 to i8*
incorporating the multiplication into the getelementptr.
This happens all the time in dragonegg, for example for
  int foo(int *A, int N) {
    return A[N];
  }
because gcc turns this into byte pointer arithmetic before it hits the plugin:
  D.1590_2 = (long unsigned int) N_1(D);
  D.1591_3 = D.1590_2 * 4;
  D.1592_5 = A_4(D) + D.1591_3;
  D.1589_6 = *D.1592_5;
  return D.1589_6;
The D.1592_5 line is a POINTER_PLUS_EXPR, which is turned into a getelementptr
on a bitcast of A_4 to i8*, so this becomes exactly the kind of IR that the
transform fires on.

An analogous transform (with no testcases!) already existed for bitcasts of
arrays, so I rewrote it to share code with this one.

llvm-svn: 166474

every TU where it's implicitly instantiated, even if there's an implicit
instantiation for the same types available in another TU.

llvm-svn: 166470

llvm-svn: 166456

llvm-svn: 166454

deterministic, replace it with a DenseMap<std::pair<unsigned, unsigned>,
PHINode*> (we already have a map from BasicBlock to unsigned).

<rdar://problem/12541389>

llvm-svn: 166435

Fix by Shivarama Rao <Shivarama.Rao@amd.com>

llvm-svn: 166427

llvm-svn: 166424