Summary:
Currently we express umin as `~umax(~x, ~y)`. However, this becomes
a problem for operands in non-integral pointer spaces, because `~x`
is not something we can compute for `x` non-integral. However, since
comparisons are generally still allowed, we are actually able to
express `umin(x, y)` directly as long as we don't try to express is
as a umax. Support this by adding an explicit umin/smin representation
to SCEV. We do this by factoring the existing getUMax/getSMax functions
into a new function that does all four. The previous two functions were
largely identical.
Reviewed By: sanjoy
Differential Revision: https://reviews.llvm.org/D50167
llvm-svn: 360159
Summary:
Commit
rL331949: SCEV] Do not use induction in isKnownPredicate for simplification umax
changed the codepath for umax from isKnownPredicate to
isKnownViaNonRecursiveReasoning to avoid compile time blow up (and as
I found out also stack overflows). However, there is an exact copy of
the code for umax that was lacking this change. In D50167 I want to unify
these codepaths, but to avoid that being a behavior change for the smax
case, pull this independent bit out of it.
Reviewed By: sanjoy
Differential Revision: https://reviews.llvm.org/D61166
llvm-svn: 359693
This reverts commit 7bf4d7c07f2fac862ef34c82ad0fef6513452445.
After thinking about this more, this isn't right, the range is not exact
in the same sense as makeExactICmpRegion(). This needs a separate
function.
llvm-svn: 358876
Following D60632 makeGuaranteedNoWrapRegion() always returns an
exact nowrap region. Rename the function accordingly. This is in
line with the naming of makeExactICmpRegion().
llvm-svn: 358875
ConstantRanges have an annoying special case: If upper and lower are
the same, it can be either an empty or a full set. When constructing
constant ranges nearly always a full set is intended, but this still
requires an explicit check in many places.
This revision adds a getNonEmpty() constructor that disambiguates this
case: If upper and lower are the same, a full set is created.
Differential Revision: https://reviews.llvm.org/D60947
llvm-svn: 358854
This adds a WithOverflowInst class with a few helper methods to get
the underlying binop, signedness and nowrap type and makes use of it
where sensible. There will be two more uses in D60650/D60656.
The refactorings are all NFC, though I left some TODOs where things
could be improved. In particular we have two places where add/sub are
handled but mul isn't.
Differential Revision: https://reviews.llvm.org/D60668
llvm-svn: 358512
Summary:
Create a method to forget everything in SCEV.
Add a cl::opt and PassManagerBuilder option to use this in LoopUnroll.
Motivation: Certain Halide applications spend a very long time compiling in forgetLoop, and prefer to forget everything and rebuild SCEV from scratch.
Sample difference in compile time reduction: 21.04 to 14.78 using current ToT release build.
Testcase showcasing this cannot be opensourced and is fairly large.
The option disabled by default, but it may be desirable to enable by
default. Evidence in favor (two difference runs on different days/ToT state):
File Before (s) After (s)
clang-9.bc 7267.91 6639.14
llvm-as.bc 194.12 194.12
llvm-dis.bc 62.50 62.50
opt.bc 1855.85 1857.53
File Before (s) After (s)
clang-9.bc 8588.70 7812.83
llvm-as.bc 196.20 194.78
llvm-dis.bc 61.55 61.97
opt.bc 1739.78 1886.26
Reviewers: sanjoy
Subscribers: mehdi_amini, jlebar, zzheng, javed.absar, dmgreen, jdoerfert, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D60144
llvm-svn: 358304
Summary:
This lets us avoid e.g. checking if A >=s B in getSMaxExpr(A, B) if we've
already established that (A smax B) is the best we can do.
Fixes PR41225.
Reviewers: asbirlea
Subscribers: mcrosier, jlebar, bixia, jdoerfert, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D60010
llvm-svn: 357320
This adds ConstantRange::getFull(BitWidth) and
ConstantRange::getEmpty(BitWidth) named constructors as more readable
alternatives to the current ConstantRange(BitWidth, /* full */ false)
and similar. Additionally private getFull() and getEmpty() member
functions are added which return a full/empty range with the same bit
width -- these are commonly needed inside ConstantRange.cpp.
The IsFullSet argument in the ConstantRange(BitWidth, IsFullSet)
constructor is now mandatory for the few usages that still make use of it.
Differential Revision: https://reviews.llvm.org/D59716
llvm-svn: 356852
Summary:
This fixes an extremely long compile time caused by recursive analysis
of truncs, which were not previously subject to any depth limits unlike
some of the other ops. I decided to use the same control used for
sext/zext, since the routines analyzing these are sometimes mutually
recursive with the trunc analysis.
Reviewers: mkazantsev, sanjoy
Subscribers: sanjoy, jdoerfert, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D58994
llvm-svn: 355949
In some cases, MaxBECount can be less precise than ExactBECount for AND
and OR (the AND case was PR26207). In the OR test case, both ExactBECounts are
undef, but MaxBECount are different, so we hit the assertion below. This
patch uses the same solution the AND case already uses.
Assertion failed:
((isa<SCEVCouldNotCompute>(ExactNotTaken) || !isa<SCEVCouldNotCompute>(MaxNotTaken))
&& "Exact is not allowed to be less precise than Max"), function ExitLimit
This patch also consolidates test cases for both AND and OR in a single
test case.
Fixes https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=13245
Reviewers: sanjoy, efriedma, mkazantsev
Reviewed By: sanjoy
Differential Revision: https://reviews.llvm.org/D58853
llvm-svn: 355259
The value stored in SCEVConstant is of type ConstantInt*, which can
never be UndefValue. So we should never hit that code.
Reviewers: mkazantsev, sanjoy
Reviewed By: sanjoy
Differential Revision: https://reviews.llvm.org/D58851
llvm-svn: 355257
Currently, SCEV creates SCEVUnknown for every node of unreachable code. If we
have a huge amounts of such code, we will be littering SE with these nodes. We could
just state that they all are undef and save some memory.
Differential Revision: https://reviews.llvm.org/D57567
Reviewed By: sanjoy
llvm-svn: 353017
Currently SCEV attempts to limit transformations so that they do not work with
big SCEVs (that may take almost infinite compile time). But for this, it uses heuristics
such as recursion depth and number of operands, which do not give us a guarantee
that we don't actually have big SCEVs. This situation is still possible, though it is not
likely to happen. However, the bug PR33494 showed a bunch of simple corner case
tests where we still produce huge SCEVs, even not reaching big recursion depth etc.
This patch introduces a concept of 'huge' SCEVs. A SCEV is huge if its expression
size (intoduced in D35989) exceeds some threshold value. We prohibit optimizing
transformations if any of SCEVs we are dealing with is huge. This gives us a reliable
check that we don't spend too much time working with them.
As the next step, we can possibly get rid of old limiting mechanisms, such as recursion
depth thresholds.
Differential Revision: https://reviews.llvm.org/D35990
Reviewed By: reames
llvm-svn: 352728
The code of AddRec simplification is using wrong loop when it creates a new
AddRecExpr. It should be using AddRecLoop which we have saved and against which
all gate checks are made, and not calling AddRec->getLoop() over and over
again because AddRec may change and become an AddRecurrency from outer loop
during the transform iterations.
Considering this change trivial, commiting for postcommit review.
llvm-svn: 352451
This patch introduces the field `ExpressionSize` in SCEV. This field is
calculated only once on SCEV creation, and it represents the complexity of
this SCEV from arithmetical point of view (not from the point of the number
of actual different SCEV nodes that are used in the expression). Roughly
saying, it is the number of operands and operations symbols when we print this
SCEV.
A formal definition is following: if SCEV `X` has operands
`Op1`, `Op2`, ..., `OpN`,
then
Size(X) = 1 + Size(Op1) + Size(Op2) + ... + Size(OpN).
Size of SCEVConstant and SCEVUnknown is one.
Expression size may be used as a universal way to limit SCEV transformations
for huge SCEVs. Currently, we have a bunch of options that represents various
limits (such as recursion depth limit) that may not make any sense from the
point of view of a LLVM users who is not familiar with SCEV internals, and all
these different options pursue one goal. A more general rule that may
potentially allow us to get rid of this redundancy in options is "do not make
transformations with SCEVs of huge size". It can apply to all SCEV traversals
and transformations that may need to visit a SCEV node more than once, hence
they are prone to combinatorial explosions.
This patch only introduces SCEV sizes calculation as NFC, its utilization will
be introduced in follow-up patches.
Differential Revision: https://reviews.llvm.org/D35989
Reviewed By: reames
llvm-svn: 351725
to reflect the new license.
We understand that people may be surprised that we're moving the header
entirely to discuss the new license. We checked this carefully with the
Foundation's lawyer and we believe this is the correct approach.
Essentially, all code in the project is now made available by the LLVM
project under our new license, so you will see that the license headers
include that license only. Some of our contributors have contributed
code under our old license, and accordingly, we have retained a copy of
our old license notice in the top-level files in each project and
repository.
llvm-svn: 351636
We have a lot of various bugs that are caused by misuse of SCEV (in particular in LV),
all of them can simply be described as "we ask SCEV to prove some fact on invalid IR".
Some of examples of those are PR36311, PR37221, PR39160.
The problem is that these failues manifest differently (what we saw was failure of various
asserts across SCEV, but there can also be miscompiles). This patch adds an assert into two
SCEV methods that strongly rely on correctness of the IR and are involved in known failues.
This will at least allow us to have a clear indication of what was wrong in this case.
This patch also fixes a unit test with incorrect IR that fails this verification.
Differential Revision: https://reviews.llvm.org/D52930
Reviewed By: fhahn
llvm-svn: 346389
When we calculate a product of 2 AddRecs, we end up making quite massive
computations to deduce the operands of resulting AddRec. This process can
be optimized by computing all args of intermediate sum and then calling
`getAddExpr` once rather than calling `getAddExpr` with intermediate
result every time a new argument is computed.
Differential Revision: https://reviews.llvm.org/D53189
Reviewed By: rtereshin
llvm-svn: 345813
SCEV's transform that turns `{A1,+,A2,+,...,+,An}<L> * {B1,+,B2,+,...,+,Bn}<L>` into
a single AddRec of size `2n+1` with complex combinatorial coefficients can easily
trigger exponential growth of the SCEV (in case if nothing gets folded and simplified).
We tried to restrain this transform using the option `scalar-evolution-max-add-rec-size`,
but its default value seems to be insufficiently small: the test attached to this patch
with default value of this option `16` has a SCEV of >3M symbols (when printed out).
This patch reduces the simplification limit. It is not a cure to combinatorial
explosions, but at least it reduces this corner case to something more or less
reasonable.
Differential Revision: https://reviews.llvm.org/D53282
Reviewed By: sanjoy
llvm-svn: 344584
by `getTerminator()` calls instead be declared as `Instruction`.
This is the biggest remaining chunk of the usage of `getTerminator()`
that insists on the narrow type and so is an easy batch of updates.
Several files saw more extensive updates where this would cascade to
requiring API updates within the file to use `Instruction` instead of
`TerminatorInst`. All of these were trivial in nature (pervasively using
`Instruction` instead just worked).
llvm-svn: 344502
This reverts r319889.
Unfortunately, wrapping flags are not a part of SCEV's identity (they
do not participate in computing a hash value or in equality
comparisons) and in fact they could be assigned after the fact w/o
rebuilding a SCEV.
Grep for const_cast's to see quite a few of examples, apparently all
for AddRec's at the moment.
So, if 2 expressions get built in 2 slightly different ways: one with
flags set in the beginning, the other with the flags attached later
on, we may end up with 2 expressions which are exactly the same but
have their operands swapped in one of the commutative N-ary
expressions, and at least one of them will have "sorted by complexity"
invariant broken.
2 identical SCEV's won't compare equal by pointer comparison as they
are supposed to.
A real-world reproducer is added as a regression test: the issue
described causes 2 identical SCEV expressions to have different order
of operands and therefore compare not equal, which in its turn
prevents LoadStoreVectorizer from vectorizing a pair of consecutive
loads.
On a larger example (the source of the test attached, which is a
bugpoint) I have seen even weirder behavior: adding a constant to an
existing SCEV changes the order of the existing terms, for instance,
getAddExpr(1, ((A * B) + (C * D))) returns (1 + (C * D) + (A * B)).
Differential Revision: https://reviews.llvm.org/D40645
llvm-svn: 340777
as well as sext(C + x + ...) -> (D + sext(C-D + x + ...))<nuw><nsw>
similar to the equivalent transformation for zext's
if the top level addition in (D + (C-D + x * n)) could be proven to
not wrap, where the choice of D also maximizes the number of trailing
zeroes of (C-D + x * n), ensuring homogeneous behaviour of the
transformation and better canonicalization of such AddRec's
(indeed, there are 2^(2w) different expressions in `B1 + ext(B2 + Y)` form for
the same Y, but only 2^(2w - k) different expressions in the resulting `B3 +
ext((B4 * 2^k) + Y)` form, where w is the bit width of the integral type)
This patch generalizes sext(C1 + C2*X) --> sext(C1) + sext(C2*X) and
sext{C1,+,C2} --> sext(C1) + sext{0,+,C2} transformations added in
r209568 relaxing the requirements the following way:
1. C2 doesn't have to be a power of 2, it's enough if it's divisible by 2
a sufficient number of times;
2. C1 doesn't have to be less than C2, instead of extracting the entire
C1 we can split it into 2 terms: (00...0XXX + YY...Y000), keep the
second one that may cause wrapping within the extension operator, and
move the first one that doesn't affect wrapping out of the extension
operator, enabling further simplifications;
3. C1 and C2 don't have to be positive, splitting C1 like shown above
produces a sum that is guaranteed to not wrap, signed or unsigned;
4. in AddExpr case there could be more than 2 terms, and in case of
AddExpr the 2nd and following terms and in case of AddRecExpr the
Step component don't have to be in the C2*X form or constant
(respectively), they just need to have enough trailing zeros,
which in turn could be guaranteed by means other than arithmetics,
e.g. by a pointer alignment;
5. the extension operator doesn't have to be a sext, the same
transformation works and profitable for zext's as well.
Apparently, optimizations like SLPVectorizer currently fail to
vectorize even rather trivial cases like the following:
double bar(double *a, unsigned n) {
double x = 0.0;
double y = 0.0;
for (unsigned i = 0; i < n; i += 2) {
x += a[i];
y += a[i + 1];
}
return x * y;
}
If compiled with `clang -std=c11 -Wpedantic -Wall -O3 main.c -S -o - -emit-llvm`
(!{!"clang version 7.0.0 (trunk 337339) (llvm/trunk 337344)"})
it produces scalar code with the loop not unrolled with the unsigned `n` and
`i` (like shown above), but vectorized and unrolled loop with signed `n` and
`i`. With the changes made in this commit the unsigned version will be
vectorized (though not unrolled for unclear reasons).
How it all works:
Let say we have an AddExpr that looks like (C + x + y + ...), where C
is a constant and x, y, ... are arbitrary SCEVs. Let's compute the
minimum number of trailing zeroes guaranteed of that sum w/o the
constant term: (x + y + ...). If, for example, those terms look like
follows:
i
XXXX...X000
YYYY...YY00
...
ZZZZ...0000
then the rightmost non-guaranteed-zero bit (a potential one at i-th
position above) can change the bits of the sum to the left (and at
i-th position itself), but it can not possibly change the bits to the
right. So we can compute the number of trailing zeroes by taking a
minimum between the numbers of trailing zeroes of the terms.
Now let's say that our original sum with the constant is effectively
just C + X, where X = x + y + .... Let's also say that we've got 2
guaranteed trailing zeros for X:
j
CCCC...CCCC
XXXX...XX00 // this is X = (x + y + ...)
Any bit of C to the left of j may in the end cause the C + X sum to
wrap, but the rightmost 2 bits of C (at positions j and j - 1) do not
affect wrapping in any way. If the upper bits cause a wrap, it will be
a wrap regardless of the values of the 2 least significant bits of C.
If the upper bits do not cause a wrap, it won't be a wrap regardless
of the values of the 2 bits on the right (again).
So let's split C to 2 constants like follows:
0000...00CC = D
CCCC...CC00 = (C - D)
and represent the whole sum as D + (C - D + X). The second term of
this new sum looks like this:
CCCC...CC00
XXXX...XX00
----------- // let's add them up
YYYY...YY00
The sum above (let's call it Y)) may or may not wrap, we don't know,
so we need to keep it under a sext/zext. Adding D to that sum though
will never wrap, signed or unsigned, if performed on the original bit
width or the extended one, because all that that final add does is
setting the 2 least significant bits of Y to the bits of D:
YYYY...YY00 = Y
0000...00CC = D
----------- <nuw><nsw>
YYYY...YYCC
Which means we can safely move that D out of the sext or zext and
claim that the top-level sum neither sign wraps nor unsigned wraps.
Let's run an example, let's say we're working in i8's and the original
expression (zext's or sext's operand) is 21 + 12x + 8y. So it goes
like this:
0001 0101 // 21
XXXX XX00 // 12x
YYYY Y000 // 8y
0001 0101 // 21
ZZZZ ZZ00 // 12x + 8y
0000 0001 // D
0001 0100 // 21 - D = 20
ZZZZ ZZ00 // 12x + 8y
0000 0001 // D
WWWW WW00 // 21 - D + 12x + 8y = 20 + 12x + 8y
therefore zext(21 + 12x + 8y) = (1 + zext(20 + 12x + 8y))<nuw><nsw>
This approach could be improved if we move away from using trailing
zeroes and use KnownBits instead. For instance, with KnownBits we could
have the following picture:
i
10 1110...0011 // this is C
XX X1XX...XX00 // this is X = (x + y + ...)
Notice that some of the bits of X are known ones, also notice that
known bits of X are interspersed with unknown bits and not grouped on
the rigth or left.
We can see at the position i that C(i) and X(i) are both known ones,
therefore the (i + 1)th carry bit is guaranteed to be 1 regardless of
the bits of C to the right of i. For instance, the C(i - 1) bit only
affects the bits of the sum at positions i - 1 and i, and does not
influence if the sum is going to wrap or not. Therefore we could split
the constant C the following way:
i
00 0010...0011 = D
10 1100...0000 = (C - D)
Let's compute the KnownBits of (C - D) + X:
XX1 1 = carry bit, blanks stand for known zeroes
10 1100...0000 = (C - D)
XX X1XX...XX00 = X
--- -----------
XX X0XX...XX00
Will this add wrap or not essentially depends on bits of X. Adding D
to this sum, however, is guaranteed to not to wrap:
0 X
00 0010...0011 = D
sX X0XX...XX00 = (C - D) + X
--- -----------
sX XXXX XX11
As could be seen above, adding D preserves the sign bit of (C - D) +
X, if any, and has a guaranteed 0 carry out, as expected.
The more bits of (C - D) we constrain, the better the transformations
introduced here canonicalize expressions as it leaves less freedom to
what values the constant part of ((C - D) + x + y + ...) can take.
Reviewed By: mzolotukhin, efriedma
Differential Revision: https://reviews.llvm.org/D48853
llvm-svn: 337943
if the top level addition in (D + (C-D + x + ...)) could be proven to
not wrap, where the choice of D also maximizes the number of trailing
zeroes of (C-D + x + ...), ensuring homogeneous behaviour of the
transformation and better canonicalization of such expressions.
This enables better canonicalization of expressions like
1 + zext(5 + 20 * %x + 24 * %y) and
zext(6 + 20 * %x + 24 * %y)
which get both transformed to
2 + zext(4 + 20 * %x + 24 * %y)
This pattern is common in address arithmetics and the transformation
makes it easier for passes like LoadStoreVectorizer to prove that 2 or
more memory accesses are consecutive and optimize (vectorize) them.
Reviewed By: mzolotukhin
Differential Revision: https://reviews.llvm.org/D48853
llvm-svn: 337859
SCEV tries to constant-fold arguments of trunc operands in SCEVAddExpr, and when it does
that, it passes wrong flags into the recursion. It is only valid to pass flags that are proved for
narrow type into a computation in wider type if we can prove that trunc instruction doesn't
actually change the value. If it did lose some meaningful bits, we may end up proving wrong
no-wrap flags for sum of arguments of trunc.
In the provided test we end up with `nuw` where it shouldn't be because of this bug.
The solution is to conservatively pass `SCEV::FlagAnyWrap` which is always a valid thing to do.
Reviewed By: sanjoy
Differential Revision: https://reviews.llvm.org/D49471
llvm-svn: 337435
It's a bit neater to write T.isIntOrPtrTy() over `T.isIntegerTy() ||
T.isPointerTy()`.
I used Python's re.sub with this regex to update users:
r'([\w.\->()]+)isIntegerTy\(\)\s*\|\|\s*\1isPointerTy\(\)'
llvm-svn: 336462
Summary:
Comment on Transforms/LoopVersioning/incorrect-phi.ll: With the change
SCEV is able to prove that the loop doesn't wrap-self (due to zext i16
to i64), disabling the entire loop versioning pass. Removed the zext and
just use i64.
Reviewers: sanjoy
Subscribers: jlebar, hiraditya, javed.absar, bixia, llvm-commits
Differential Revision: https://reviews.llvm.org/D48409
llvm-svn: 336140
We can have AddRec with loops having many predecessors.
This changes an assert to an early return.
Differential Revision: https://reviews.llvm.org/D48766
llvm-svn: 335965
Summary:
This initiates a discussion on changing Polly accordingly while re-applying r335197 (D48338).
I have never worked on Polly. The proposed change to param_div_div_div_2.ll is not educated, but just patterns that match the output.
All LLVM files are already reviewed in D48338.
Reviewers: jdoerfert, bollu, efriedma
Subscribers: jlebar, sanjoy, hiraditya, llvm-commits, bixia
Differential Revision: https://reviews.llvm.org/D48453
llvm-svn: 335292
Summary:
Try to match udiv and urem patterns, and sink zext down to the leaves.
I'm not entirely sure why some unrelated tests change, but the added <nsw>s seem right.
Reviewers: sanjoy
Subscribers: jlebar, hiraditya, bixia, llvm-commits
Differential Revision: https://reviews.llvm.org/D48338
llvm-svn: 335197
This reverts r334428. It incorrectly marks some multiplications as nuw. Tim
Shen is working on a proper fix.
Original commit message:
[SCEV] Add nuw/nsw to mul ops in StrengthenNoWrapFlags where safe.
Summary:
Previously we would add them for adds, but not multiplies.
llvm-svn: 335016
Summary:
Sending for presubmit review out of an abundance of caution; it would be
bad to mess this up.
Reviewers: sanjoy
Subscribers: hiraditya, llvm-commits
Differential Revision: https://reviews.llvm.org/D48238
llvm-svn: 334875
Summary:
Obviates the need for mask/clear/setFlags helpers.
There are some expressions here which can be simplified, but to keep
this easy to review, I have not simplified them in this patch.
No functional change.
Reviewers: sanjoy
Subscribers: hiraditya, llvm-commits
Differential Revision: https://reviews.llvm.org/D48237
llvm-svn: 334874
Summary:
Specifically, we transform
zext(2^K * (trunc X to iN)) to iM ->
2^K * (zext(trunc X to i{N-K}) to iM)<nuw>
This is helpful because pulling the 2^K out of the zext allows further
optimizations.
Reviewers: sanjoy
Subscribers: hiraditya, llvm-commits, timshen
Differential Revision: https://reviews.llvm.org/D48158
llvm-svn: 334737
