Commit Graph

5 Commits

Author SHA1 Message Date
Sean Silva 284b0324e2 [PM] Avoid getResult on a higher level in LoopAccessAnalysis
Note that require<domtree> and require<loops> aren't needed because they
come in implicitly via the loop pass manager.

llvm-svn: 274712
2016-07-07 01:01:53 +00:00
Xinliang David Li 8a021317a2 [PM] Port LoopAccessInfo analysis to new PM
It is implemented as a LoopAnalysis pass as 
discussed and agreed upon.

llvm-svn: 274452
2016-07-02 21:18:40 +00:00
Adam Nemet ce48250f11 [LoopAccesses] Allow analysis to complete in the presence of uniform stores
(Re-apply r234361 with a fix and a testcase for PR23157)

Both run-time pointer checking and the dependence analysis are capable
of dealing with uniform addresses. I.e. it's really just an orthogonal
property of the loop that the analysis computes.

Run-time pointer checking will only try to reason about SCEVAddRec
pointers or else gives up. If the uniform pointer turns out the be a
SCEVAddRec in an outer loop, the run-time checks generated will be
correct (start and end bounds would be equal).

In case of the dependence analysis, we work again with SCEVs. When
compared against a loop-dependent address of the same underlying object,
the difference of the two SCEVs won't be constant. This will result in
returning an Unknown dependence for the pair.

When compared against another uniform access, the difference would be
constant and we should return the right type of dependence
(forward/backward/etc).

The changes also adds support to query this property of the loop and
modify the vectorizer to use this.

Patch by Ashutosh Nema!

llvm-svn: 234424
2015-04-08 17:48:40 +00:00
Adam Nemet e09a928c80 Revert "[LoopAccesses] Allow analysis to complete in the presence of uniform stores"
This reverts commit r234361.

It caused PR23157.

llvm-svn: 234387
2015-04-08 04:16:55 +00:00
Adam Nemet 0515c33b70 [LoopAccesses] Allow analysis to complete in the presence of uniform stores
Both run-time pointer checking and the dependence analysis are capable
of dealing with uniform addresses. I.e. it's really just an orthogonal
property of the loop that the analysis computes.

Run-time pointer checking will only try to reason about SCEVAddRec
pointers or else gives up. If the uniform pointer turns out the be a
SCEVAddRec in an outer loop, the run-time checks generated will be
correct (start and end bounds would be equal).

In case of the dependence analysis, we work again with SCEVs. When
compared against a loop-dependent address of the same underlying object,
the difference of the two SCEVs won't be constant. This will result in
returning an Unknown dependence for the pair.

When compared against another uniform access, the difference would be
constant and we should return the right type of dependence
(forward/backward/etc).

The changes also adds support to query this property of the loop and
modify the vectorizer to use this.

Patch by Ashutosh Nema!

llvm-svn: 234361
2015-04-07 21:46:16 +00:00