15 Commits

Author SHA1 Message Date
Jingyue Wu
62bdd72542 [SLSR] S's basis must have the same type as S
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@240910 91177308-0d34-0410-b5e6-96231b3b80d8
2015-06-28 17:45:05 +00:00
Jingyue Wu
40ccf16b92 [NFC] more comments in SLSR
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@239984 91177308-0d34-0410-b5e6-96231b3b80d8
2015-06-18 03:35:57 +00:00
Matt Arsenault
b1d220c21a SLSR: Pass address space to isLegalAddressingMode
This only updates one of the uses. The other is used in cases
that may never touch memory, so I'm not sure why this is even
calling it. Perhaps there should be a new, similar hook for such
cases or pass -1 for unknown address space.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@239540 91177308-0d34-0410-b5e6-96231b3b80d8
2015-06-11 16:13:39 +00:00
Jingyue Wu
2703007b7b [ScalarEvolution] refactor: extract interface getGEPExpr
Summary:
This allows other passes (such as SLSR) to compute the SCEV expression for an
imaginary GEP.

Test Plan: no regression

Reviewers: atrick, sanjoy

Reviewed By: sanjoy

Subscribers: llvm-commits

Differential Revision: http://reviews.llvm.org/D9786

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@237589 91177308-0d34-0410-b5e6-96231b3b80d8
2015-05-18 17:03:25 +00:00
Jingyue Wu
ef0b6c4b1e [SLSR] handle (B | i) * S
Summary:
Consider (B | i) * S as (B + i) * S if B and i have no bits set in
common.

Test Plan: @or in slsr-mul.ll

Reviewers: broune, meheff

Subscribers: llvm-commits

Differential Revision: http://reviews.llvm.org/D9788

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@237456 91177308-0d34-0410-b5e6-96231b3b80d8
2015-05-15 17:07:48 +00:00
Jingyue Wu
3062272092 [SLSR] handles non-canonicalized Mul candidates
such as (2 + B) * S.

Tested by @non_canonicalized in slsr-mul.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@237216 91177308-0d34-0410-b5e6-96231b3b80d8
2015-05-13 00:03:17 +00:00
Jingyue Wu
423476d899 [SLSR] garbage-collect unused instructions
Summary:
After we rewrite a candidate, the instructions used by the old form may
become unused. This patch cleans up these unused instructions so that we
needn't run DCE after SLSR.

Test Plan: removed -dce in all the SLSR tests

Reviewers: broune, meheff

Subscribers: llvm-commits

Differential Revision: http://reviews.llvm.org/D9101

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@235410 91177308-0d34-0410-b5e6-96231b3b80d8
2015-04-21 19:56:18 +00:00
Jingyue Wu
d4ceea3837 [SLSR] handle candidate form (B + i * S)
Summary:
With this patch, SLSR may rewrite

S1: X = B + i * S
S2: Y = B + i' * S

to

S2: Y = X + (i' - i) * S

A secondary improvement: if (i' - i) is a power of 2, emit Y as X + (S << log(i' - i)). (S << log(i' -i)) is in a canonical form and thus more likely GVN'ed than (i' - i) * S.

Test Plan: slsr-add.ll

Reviewers: hfinkel, sanjoy, meheff, broune, eliben

Reviewed By: eliben

Subscribers: llvm-commits

Differential Revision: http://reviews.llvm.org/D8983

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@235019 91177308-0d34-0410-b5e6-96231b3b80d8
2015-04-15 16:46:13 +00:00
Jingyue Wu
10483b93ea [SLSR] consider &B[S << i] as &B[(1 << i) * S]
Summary: This reduces handling &B[(1 << i) * s] to handling &B[i * S].

Test Plan: slsr-gep.ll

Reviewers: meheff

Subscribers: sanjoy, llvm-commits

Differential Revision: http://reviews.llvm.org/D8837

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@234180 91177308-0d34-0410-b5e6-96231b3b80d8
2015-04-06 17:15:48 +00:00
David Blaikie
4a86b381a3 [opaque pointer type] More GEP IRBuilder API migrations...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@234058 91177308-0d34-0410-b5e6-96231b3b80d8
2015-04-03 21:33:42 +00:00
David Blaikie
cf57d81b6e [opaque pointer type] More GEP API migrations in IRBuilder uses
The plan here is to push the API changes out from the common components
(like Constant::getGetElementPtr and IRBuilder::CreateGEP related
functions) and just update callers to either pass the type if it's
obvious, or pass null.

Do this with LoadInst as well and anything else that comes up, then to
start porting specific uses to not pass null anymore - this may require
some refactoring in each case.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@234042 91177308-0d34-0410-b5e6-96231b3b80d8
2015-04-03 19:41:44 +00:00
Jingyue Wu
6dedfc073a [SLSR] handles off bounds GEPs
Summary:
The old requirement on GEP candidates being in bounds is unnecessary.
For off-bound GEPs, we still have

  &B[i * S] = B + (i * S) * e = B + (i * e) * S

Test Plan: slsr_offbound_gep in slsr-gep.ll

Reviewers: meheff

Subscribers: llvm-commits

Differential Revision: http://reviews.llvm.org/D8809

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@233949 91177308-0d34-0410-b5e6-96231b3b80d8
2015-04-02 21:18:32 +00:00
Jingyue Wu
439d17ca9f [SLSR] handle candidate form &B[i * S]
Summary:
This patch enhances SLSR to handle another candidate form &B[i * S]. If
we found two candidates

S1: X = &B[i * S]
S2: Y = &B[i' * S]

and S1 dominates S2, we can replace S2 with

Y = &X[(i' - i) * S]

Test Plan:
slsr-gep.ll
X86/no-slsr.ll: verify that we do not run SLSR on GEPs that already fit into
an addressing mode

Reviewers: eliben, atrick, meheff, hfinkel

Reviewed By: hfinkel

Subscribers: sanjoy, llvm-commits

Differential Revision: http://reviews.llvm.org/D7459

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@233286 91177308-0d34-0410-b5e6-96231b3b80d8
2015-03-26 16:49:24 +00:00
Aaron Ballman
e818df583c Fixing a -Wsign-compare warning; NFC
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@228142 91177308-0d34-0410-b5e6-96231b3b80d8
2015-02-04 14:01:08 +00:00
Jingyue Wu
2918efd551 Add straight-line strength reduction to LLVM
Summary:
Straight-line strength reduction (SLSR) is implemented in GCC but not yet in
LLVM. It has proven to effectively simplify statements derived from an unrolled
loop, and can potentially benefit many other cases too. For example,

LLVM unrolls

  #pragma unroll
  foo (int i = 0; i < 3; ++i) {
    sum += foo((b + i) * s);
  }

into

  sum += foo(b * s);
  sum += foo((b + 1) * s);
  sum += foo((b + 2) * s);

However, no optimizations yet reduce the internal redundancy of the three
expressions:

  b * s
  (b + 1) * s
  (b + 2) * s

With SLSR, LLVM can optimize these three expressions into:

  t1 = b * s
  t2 = t1 + s
  t3 = t2 + s

This commit is only an initial step towards implementing a series of such
optimizations. I will implement more (see TODO in the file commentary) in the
near future. This optimization is enabled for the NVPTX backend for now.
However, I am more than happy to push it to the standard optimization pipeline
after more thorough performance tests.

Test Plan: test/StraightLineStrengthReduce/slsr.ll

Reviewers: eliben, HaoLiu, meheff, hfinkel, jholewinski, atrick

Reviewed By: jholewinski, atrick

Subscribers: karthikthecool, jholewinski, llvm-commits

Differential Revision: http://reviews.llvm.org/D7310

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@228016 91177308-0d34-0410-b5e6-96231b3b80d8
2015-02-03 19:37:06 +00:00