Commit Graph

19 Commits

Author SHA1 Message Date
eopXD fa488ea864 [LoopNest][LoopFlatten] Change LoopFlattenPass to LoopNest pass
This patch changes LoopFlattenPass from FunctionPass to LoopNestPass.

Utilize LoopNest and let function 'Flatten' generate information from it.

Reviewed By: Whitney

Differential Revision: https://reviews.llvm.org/D102904
2021-05-28 15:43:12 +00:00
eopXD e96d6f4821 Revert "[LoopNest][LoopFlatten] Change LoopFlattenPass to LoopNest pass"
This reverts commit 7952ddb21f.

Differential Revision: https://reviews.llvm.org/D103302
2021-05-28 07:58:06 +00:00
eopXD 7e06cf8f1b Revert "[LoopNest][LoopFlatten] Change LoopFlattenPass to LoopNest pass"
This reverts commit ffc4d3e068.
2021-05-28 07:48:04 +00:00
eopXD ffc4d3e068 [LoopNest][LoopFlatten] Change LoopFlattenPass to LoopNest pass
This patch changes LoopFlattenPass from FunctionPass to LoopNestPass.

Utilize LoopNest and let function 'Flatten' generate information from it.

Reviewed By: Whitney

Differential Revision: https://reviews.llvm.org/D102904
2021-05-28 07:25:53 +00:00
eopXD 7952ddb21f [LoopNest][LoopFlatten] Change LoopFlattenPass to LoopNest pass
This patch changes LoopFlattenPass from FunctionPass to LoopNestPass.

Utilize LoopNest and let function 'Flatten' generate information from it.

Reviewed By: Whitney

Differential Revision: https://reviews.llvm.org/D102904
2021-05-28 07:11:26 +00:00
Stelios Ioannou 1124ad2f5d [LoopFlatten] Simplify loops so that the pass can operate on unsimplified loops.
The loop flattening pass requires loops to be in simplified form. If the
loops are not in simplified form, the pass cannot operate. This patch
simplifies all loops before flattening. As a result, all loops will be
simplified regardless of whether anything ends up being flattened.

This change was inspired by observing a certain loop that was not flatten
because the loops were not in simplified form. This loop is added as a
test to verify that it is now flattened.

Differential Revision: https://reviews.llvm.org/D102249

Change-Id: I45bcabe70fb99b0d89f0effafc82eb9e0585ec30
2021-05-12 19:22:01 +01:00
Simon Pilgrim b8aba76a4e LoopFlatten - CanWidenIV - Fix uninitialized variable warnings and use for-range loop. NFCI.
Fix static analysis uninitialized variable warnings, and use for-range loop iteration across WideIVs array.
2021-04-06 12:24:20 +01:00
Yevgeny Rouban 1ed53d44d8 [LoopFlatten] Do not report CFG analyses as up-to-date
Removes CFGAnalyses from the preserved analyses set
returned by LoopFlattenPass::run().

Reviewed By: Dave Green, Ta-Wei Tu

Differential Revision: https://reviews.llvm.org/D99700
2021-04-01 15:52:36 +07:00
Ta-Wei Tu 4d9d736875 [NFC] Improve debug message and test description in 4c1f74a 2021-03-24 18:21:13 +08:00
Ta-Wei Tu 4c1f74a76c [LoopFlatten] Fix invalid assertion (PR49571)
The `InductionPHI` is not necessarily the increment instruction, as
demonstrated in pr49571.ll.
This patch removes the assertion and instead bails out from the
`LoopFlatten` pass if that happens.

This fixes https://bugs.llvm.org/show_bug.cgi?id=49571

Reviewed By: SjoerdMeijer

Differential Revision: https://reviews.llvm.org/D99252
2021-03-24 18:08:27 +08:00
Ta-Wei Tu 8fde25b3c3 [NFC] Remove redundant `struct` prefix
Reviewed By: SjoerdMeijer, fhahn

Differential Revision: https://reviews.llvm.org/D99251
2021-03-24 17:58:33 +08:00
Sander de Smalen ae27274b2f NFC: Migrate LoopFlatten to work on InstructionCost.
This patch migrates cost values and arithmetic to work on InstructionCost.
When the interfaces to TargetTransformInfo are changed, any InstructionCost
state will propagate naturally.

See this patch for the introduction of the type: https://reviews.llvm.org/D91174
See this thread for context: http://lists.llvm.org/pipermail/llvm-dev/2020-November/146408.html

Reviewed By: david-arm

Differential Revision: https://reviews.llvm.org/D96029
2021-02-06 11:47:04 +00:00
Sjoerd Meijer 33b2c88fa8 [LoopFlatten] Widen IV, support ZExt.
I disabled the widening in fa5cb4b because it run in an assert, which was
related to replacing values with different types. I forgot that an extend could
also be a zero-extend, which I have added now. This means that the approach now
is to create and insert a trunc value of the outerloop for each user, and use
that to replace IV values.

Differential Revision: https://reviews.llvm.org/D91690
2020-11-23 08:57:19 +00:00
Sjoerd Meijer fa5cb4b936 [LoopFlatten] Disable IV widening
Disable widening of the IV in LoopFlatten while I investigate an assertion
failures. Please note that the pass is also not yet enabled by default.
2020-11-16 22:30:52 +00:00
Benjamin Kramer 2e7455f00a [LoopFlatten] Fold variable into assert. NFC. 2020-11-16 11:51:39 +01:00
Sjoerd Meijer 9aa773381b [LoopFlatten] Widen the IV
Widen the IV to the widest available and legal integer type, which makes this
transformations always safe so that we can skip overflow checks.

Motivation is to let this pass trigger on 64-bit targets too, and this is the
last patch in a serie to achieve this: D90402 moves pass LoopFlatten to just
before IndVarSimplify so that IVs are not already widened, D90421 factors out
widening from IndVarSimplify into Utils/SimplifyIndVar so that we can also use
it in LoopFlatten.

Differential Revision: https://reviews.llvm.org/D90640
2020-11-16 10:20:13 +00:00
Sjoerd Meijer 706ead0e87 [LoopFlatten] Make it a FunctionPass
This converts LoopFlatten from a LoopPass to a FunctionPass so that we don't
run into problems of a loop pass deleting a (inner)loop.

Differential Revision: https://reviews.llvm.org/D90940
2020-11-10 20:03:31 +00:00
Sjoerd Meijer e2dcea4489 [LoopFlatten] FlattenInfo bookkeeping. NFC.
Introduce struct FlattenInfo to group some of the bookkeeping. Besides this
being a bit of a clean-up, it is a prep step for next additions (D90640). I
could take things a bit further, but thought this was a good first step also
not to make this change too large.

Differential Revision: https://reviews.llvm.org/D90408
2020-11-09 14:50:26 +00:00
Sjoerd Meijer d53b4bee0c [LoopFlatten] Add a loop-flattening pass
This is a simple pass that flattens nested loops.  The intention is to optimise
loop nests like this, which together access an array linearly:

  for (int i = 0; i < N; ++i)
    for (int j = 0; j < M; ++j)
      f(A[i*M+j]);

into one loop:

  for (int i = 0; i < (N*M); ++i)
    f(A[i]);

It can also flatten loops where the induction variables are not used in the
loop. This can help with codesize and runtime, especially on simple cpus
without advanced branch prediction.

This is only worth flattening if the induction variables are only used in an
expression like i*M+j. If they had any other uses, we would have to insert a
div/mod to reconstruct the original values, so this wouldn't be profitable.

This partially fixes PR40581 as this pass triggers on one of the two cases. I
will follow up on this to learn LoopFlatten a few more (small) tricks. Please
note that LoopFlatten is not yet enabled by default.

Patch by Oliver Stannard, with minor tweaks from Dave Green and myself.

Differential Revision: https://reviews.llvm.org/D42365
2020-10-01 13:54:45 +01:00