Commit Graph

9 Commits

Author SHA1 Message Date
Florian Hahn 38609fa9e4 Recommit "[SCCP] Use SimplifyBinOp for non-integer constant/expressions & overdef."
This includes a fix reported with simplifications in the presence of
NaN.

This reverts the revert commit 06408451bf.
2020-04-14 11:48:52 +01:00
Benjamin Kramer 06408451bf Revert "[SCCP] Use SimplifyBinOp for non-integer constant/expressions & overdef."
This reverts commit 1a02aaeaa4. Crashes on
the following test case:

$ cat crash.ll
source_filename = "__compute_module"
target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
target triple = "x86_64-grtev4-linux-gnu"

@0 = private unnamed_addr constant [24 x i8] c"\00\00\C0\7F\00\00\C0\7F\09\85\08?\ED\C94\FE~\EB/\F3\90\CF\BA\C1"
@1 = private unnamed_addr constant [24 x i8] c"\00\00\C0\7F\A3\A0\0FA\00\00\C0\7F\00\00\C0\7F\00\00\00\00\02\9AA\00"

define void @IgammaSpecialValues.448() {
entry:
  br label %fusion.26.loop_header.dim.0

fusion.26.loop_header.dim.0:                      ; preds = %fusion.26.loop_header.dim.0, %entry
  %fusion.26.invar_address.dim.0.0 = phi i64 [ 0, %entry ], [ %invar.inc17, %fusion.26.loop_header.dim.0 ]
  %0 = getelementptr inbounds [6 x float], [6 x float]* bitcast ([24 x i8]* @0 to [6 x float]*), i64 0, i64 %fusion.26.invar_address.dim.0.0
  %1 = load float, float* %0
  %2 = fmul float %1, 0.000000e+00
  %3 = getelementptr inbounds [6 x float], [6 x float]* bitcast ([24 x i8]* @1 to [6 x float]*), i64 0, i64 %fusion.26.invar_address.dim.0.0
  %4 = load float, float* %3
  %5 = fneg float %4
  %6 = fadd float %2, %5
  %invar.inc17 = add nuw nsw i64 %fusion.26.invar_address.dim.0.0, 1
  br label %fusion.26.loop_header.dim.0
}

$ opt -ipsccp -S < crash.ll
opt: llvm/include/llvm/Analysis/ValueLattice.h:251: bool llvm::ValueLatticeElement::markConstant(llvm::Constant *, bool): Assertion `getConstant() == V && "Marking constant with different value"' failed.
2020-04-13 11:23:26 +02:00
Florian Hahn 1a02aaeaa4 [SCCP] Use SimplifyBinOp for non-integer constant/expressions & overdef.
For non-integer constants/expressions and overdefined, I think we can
just use SimplifyBinOp to do common folds. By just passing a context
with the DL, SimplifyBinOp should not try to get additional information
from looking at definitions.

For overdefined values, it should be enough to just pass the original
operand.

Note: The comment before the `if (isconstant(V1State)...` was wrong
originally: isConstant() also matches integer ranges with a single
element. It is correct now.

Reviewers: efriedma, davide, mssimpso, aartbik

Reviewed By: efriedma

Differential Revision: https://reviews.llvm.org/D76459
2020-04-10 11:02:57 +01:00
Florian Hahn 8a36594a7e [SCCP] Use constant ranges for binary operators.
If one of the operands of a binary operator is a constant range, we can
use ConstantRange::binaryOp to approximate the result.

We still handle single element constant ranges as we did previously,
with ConstantExpr::get(), because ConstantRange::binaryOp still gives
worse results in a few cases for single element ranges.

Also note that we bail out early if any of the operands is still unknown.

Reviewers: davide, efriedma, mssimpso

Reviewed By: efriedma

Differential Revision: https://reviews.llvm.org/D71936
2020-03-19 09:35:48 +00:00
Eric Christopher cee313d288 Revert "Temporarily Revert "Add basic loop fusion pass.""
The reversion apparently deleted the test/Transforms directory.

Will be re-reverting again.

llvm-svn: 358552
2019-04-17 04:52:47 +00:00
Eric Christopher a863435128 Temporarily Revert "Add basic loop fusion pass."
As it's causing some bot failures (and per request from kbarton).

This reverts commit r358543/ab70da07286e618016e78247e4a24fcb84077fda.

llvm-svn: 358546
2019-04-17 02:12:23 +00:00
Dan Gohman 1ccecdb2fd Reapply r155682, making constant folding more consistent, with a fix to work
properly with how the code handles all-undef PHI nodes.

llvm-svn: 155721
2012-04-27 17:50:22 +00:00
NAKAMURA Takumi 6008dfdb70 Revert r155682, "Use ConstantExpr::getExtractElement when constant-folding vectors"
It broke stage2 build. stage1/clang sometimes crashed.

llvm-svn: 155699
2012-04-27 07:59:20 +00:00
Dan Gohman 90f3798f26 Use ConstantExpr::getExtractElement when constant-folding vectors
instead of getAggregateElement. This has the advantage of being
more consistent and allowing higher-level constant folding to
procede even if an inner extract element cannot be folded.

Make ConstantFoldInstruction call ConstantFoldConstantExpression
on the instruction's operands, making it more consistent with 
ConstantFoldConstantExpression itself. This makes sure that
ConstantExprs get TargetData-aware folding before being handed
off as operands for further folding.

This causes more expressions to be folded, but due to a known
shortcoming in constant folding, this currently has the side effect
of stripping a few more nuw and inbounds flags in the non-targetdata
side of constant-fold-gep.ll. This is mostly harmless.

This fixes rdar://11324230.

llvm-svn: 155682
2012-04-27 00:54:36 +00:00