I don't believe this was a problem in practice, as it's likely that the
boolean wasn't checked unless the backend condition was non-null.
llvm-svn: 318073
Summary:
If a compare instruction is same or inverse of the compare in the
branch of the loop latch, then return a constant evolution node.
This shall facilitate computations of loop exit counts in cases
where compare appears in the evolution chain of induction variables.
Will fix PR 34538
Reviewers: sanjoy, hfinkel, junryoungju
Reviewed By: sanjoy, junryoungju
Subscribers: javed.absar, llvm-commits
Differential Revision: https://reviews.llvm.org/D38494
llvm-svn: 318050
Max backedge taken count is always expected to be a constant; and this is
usually true by construction -- it is a SCEV expression with constant inputs.
However, if the max backedge expression ends up being computed to be a udiv with
a constant zero denominator[0], SCEV does not fold the result to a constant
since there is no constant it can fold it to (SCEV has no representation for
"infinity" or "undef").
However, in computeMaxBECountForLT we already know the denominator is positive,
and thus at least 1; and we can use this fact to avoid dividing by zero.
[0]: We can end up with a constant zero denominator if the signed range of the
stride is more precise than the unsigned range.
llvm-svn: 316615
Summary:
If a compare instruction is same or inverse of the compare in the
branch of the loop latch, then return a constant evolution node.
Currently scope of evaluation is limited to SCEV computation for
PHI nodes.
This shall facilitate computations of loop exit counts in cases
where compare appears in the evolution chain of induction variables.
Will fix PR 34538
Reviewers: sanjoy, hfinkel, junryoungju
Reviewed By: junryoungju
Subscribers: javed.absar, llvm-commits
Differential Revision: https://reviews.llvm.org/D38494
llvm-svn: 316054
This reverts commit r315713. It causes PR34968.
I think I know what the problem is, but I don't think I'll have time to fix it
this week.
llvm-svn: 315962
Summary:
This change uses the loop use list added in the previous change to remember the
loops that appear in the trip count expressions of other loops; and uses it in
forgetLoop. This lets us not scan every loop in the function on a forgetLoop
call.
With this change we no longer invalidate clear out backedge taken counts on
forgetValue. I think this is fine -- the contract is that SCEV users must call
forgetLoop(L) if their change to the IR could have changed the trip count of L;
solely calling forgetValue on a value feeding into the backedge condition of L
is not enough. Moreover, I don't think we can strengthen forgetValue to be
sufficient for invalidating trip counts without significantly re-architecting
SCEV. For instance, if we have the loop:
I = *Ptr;
E = I + 10;
do {
// ...
} while (++I != E);
then the backedge taken count of the loop is 9, and it has no reference to
either I or E, i.e. there is no way in SCEV today to re-discover the dependency
of the loop's trip count on E or I. So a SCEV client cannot change E to (say)
"I + 20", call forgetValue(E) and expect the loop's trip count to be updated.
Reviewers: atrick, sunfish, mkazantsev
Subscribers: mcrosier, llvm-commits
Differential Revision: https://reviews.llvm.org/D38435
llvm-svn: 315713
Summary:
This patch teaches SCEV to calculate the maxBECount when the end bound
of the loop can vary. Note that we cannot calculate the exactBECount.
This will only be done when both conditions are satisfied:
1. the loop termination condition is strictly LT.
2. the IV is proven to not overflow.
This provides more information to users of SCEV and can be used to
improve identification of finite loops.
Reviewers: sanjoy, mkazantsev, silviu.baranga, atrick
Reviewed by: mkazantsev
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D38825
llvm-svn: 315683
Summary:
Currently we do not correctly invalidate memoized results for add recurrences
that were created directly (i.e. they were not created from a `Value`). This
change fixes this by keeping loop use lists and using the loop use lists to
determine which SCEV expressions to invalidate.
Here are some statistics on the number of uses of in the use lists of all loops
on a clang bootstrap (config: release, no asserts):
Count: 731310
Min: 1
Mean: 8.555150
50th %time: 4
95th %tile: 25
99th %tile: 53
Max: 433
Reviewers: atrick, sunfish, mkazantsev
Subscribers: mcrosier, llvm-commits
Differential Revision: https://reviews.llvm.org/D38434
llvm-svn: 315672
Summary:
Add LLVM_FORCE_ENABLE_DUMP cmake option, and use it along with
LLVM_ENABLE_ASSERTIONS to set LLVM_ENABLE_DUMP.
Remove NDEBUG and only use LLVM_ENABLE_DUMP to enable dump methods.
Move definition of LLVM_ENABLE_DUMP from config.h to llvm-config.h so
it'll be picked up by public headers.
Differential Revision: https://reviews.llvm.org/D38406
llvm-svn: 315590
Summary:
This patch fixes an error in the patch to ScalarEvolution::createAddRecFromPHIWithCastsImpl
made in D37265. In that patch we handle the cases where the either the start or accum values can be
zero after truncation. But, we assume that the start value must be a constant if the accum is
zero. This is clearly an erroneous assumption. This change removes that assumption.
Reviewers: sanjoy, dorit, mkazantsev
Reviewed By: sanjoy
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D38814
llvm-svn: 315491
Summary:
A SCEV such as:
{%v2,+,((-1 * (trunc i64 (-1 * %v1) to i32)) + (-1 * (trunc i64 %v1 to i32)))}<%loop>
can be folded into, simply, {%v2,+,0}. However, the current code in ::getAddExpr()
will not try to apply the simplification m*trunc(x)+n*trunc(y) -> trunc(trunc(m)*x+trunc(n)*y)
because it only keys off having a non-multiplied trunc as the first term in the simplification.
This patch generalizes this code to try to do a more generic fold of these trunc
expressions.
Reviewers: sanjoy
Reviewed By: sanjoy
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D37888
llvm-svn: 313988
forgetLoop() has pretty bad performance because it goes over
the same instructions over and over again in particular when
nested loop are involved.
The refactoring changes the function to a not-recursive function
and reusing the allocation for data-structures and the Visited
set.
NFCI
Differential Revision: https://reviews.llvm.org/D37659
llvm-svn: 312920
Summary:
When constructing the predicate P1 in ScalarEvolution::createAddRecFromPHIWithCastsImpl() it is possible
for the PHISCEV from which the predicate is constructed to be a SCEVConstant instead of a SCEVAddRec. If
this happens, then the cast<SCEVAddRec>(PHISCEV) in the code will assert.
Such a PHISCEV is possible if either the start value or the accumulator value is a constant value
that not equal to its truncated value, and if the truncated value is zero.
This patch adds tests that demonstrate the cast<> assertion, and fixes this problem by checking
whether the PHISCEV is a constant before constructing the P1 predicate; if it is, then P1 is
equivalent to one of P2 or P3. Additionally, if we know that the start value or accumulator
value are constants then we check whether the P2 and/or P3 predicates are known false at compile
time; if either is, then we bail out of constructing the AddRec.
Reviewers: sanjoy, mkazantsev, silviu.baranga
Reviewed By: mkazantsev
Subscribers: mkazantsev, llvm-commits
Differential Revision: https://reviews.llvm.org/D37265
llvm-svn: 312568
In LLVM IR the following code:
%r = urem <ty> %t, %b
is equivalent to
%q = udiv <ty> %t, %b
%s = mul <ty> nuw %q, %b
%r = sub <ty> nuw %t, %q ; (t / b) * b + (t % b) = t
As UDiv, Mul and Sub are already supported by SCEV, URem can be implemented
with minimal effort using that relation:
%r --> (-%b * (%t /u %b)) + %t
We implement two special cases:
- if %b is 1, the result is always 0
- if %b is a power-of-two, we produce a zext/trunc based expression instead
That is, the following code:
%r = urem i32 %t, 65536
Produces:
%r --> (zext i16 (trunc i32 %a to i16) to i32)
Note that while this helps get a tighter bound on the range analysis and the
known-bits analysis, this exposes some normalization shortcoming of SCEVs:
%div = udim i32 %a, 65536
%mul = mul i32 %div, 65536
%rem = urem i32 %a, 65536
%add = add i32 %mul, %rem
Will usually not be reduced.
llvm-svn: 312329
Pushes the sext onto the operands of a Sub if NSW is present.
Also adds support for propagating the nowrap flags of the
llvm.ssub.with.overflow intrinsic during analysis.
Differential Revision: https://reviews.llvm.org/D35256
llvm-svn: 310117
The patch rL309080 was reverted because it did not clean up the cache on "forgetValue"
method call. This patch re-enables this change, adds the missing check and introduces
two new unit tests that make sure that the cache is cleaned properly.
Differential Revision: https://reviews.llvm.org/D36087
llvm-svn: 309925
If SCEV can prove that the backedge taken count for a loop is zero, it does not
need to "understand" a recursive PHI to compute its exiting value.
This should fix PR33885.
llvm-svn: 309758
This patch reworks the function that searches constants in Add and Mul SCEV expression
chains so that now it does not visit a node more than once, and also renames this function
for better correspondence between its implementation and semantics.
Differential Revision: https://reviews.llvm.org/D35931
llvm-svn: 309367
This reverts commit r309080. The patch needs to clear out the
ScalarEvolution::ExitLimits cache in forgetMemoizedResults.
I've replied on the commit thread for the patch with more details.
llvm-svn: 309357
This patch adds a cache for computeExitLimit to save compilation time. A lot of examples of
tests that take extensive time to compile are attached to the bug 33494.
Differential Revision: https://reviews.llvm.org/D35827
llvm-svn: 309080
`SCEVUnknown::allUsesReplacedWith` does not need to call `forgetMemoizedResults`
since RAUW does a value-equivalent replacement by assumption. If this
assumption was false then the later setValPtr(New) call would be incorrect too.
This is a non-trivial performance optimization for functions with a large number
of loops since `forgetMemoizedResults` walks all loop backedge taken counts to
see if any of them use the SCEVUnknown being RAUWed. However, this improvement
is difficult to demonstrate without checking in an excessively large IR file.
llvm-svn: 309072
When SCEV calculates product of two SCEVAddRecs from the same loop, it
tries to combine them into one big AddRecExpr. If the sizes of the initial
SCEVs were `S1` and `S2`, the size of their product is `S1 + S2 - 1`, and every
operand of the resulting SCEV is combined from operands of initial SCEV and
has much higher complexity than they have.
As result, if we try to calculate something like:
%x1 = {a,+,b}
%x2 = mul i32 %x1, %x1
%x3 = mul i32 %x2, %x1
%x4 = mul i32 %x3, %x2
...
The size of such SCEVs grows as `2^N`, and the arguments
become more and more complex as we go forth. This leads
to long compilation and huge memory consumption.
This patch sets a limit after which we don't try to combine two
`SCEVAddRecExpr`s into one. By default, max allowed size of the
resulting AddRecExpr is set to 16.
Differential Revision: https://reviews.llvm.org/D35664
llvm-svn: 308847
using runtime checks
Extend the SCEVPredicateRewriter to work a bit harder when it encounters an
UnknownSCEV for a Phi node; Try to build an AddRecurrence also for Phi nodes
whose update chain involves casts that can be ignored under the proper runtime
overflow test. This is one step towards addressing PR30654.
Differential revision: http://reviews.llvm.org/D30041
llvm-svn: 308299
Going through the Constant methods requires redetermining that the Constant is a ConstantInt and then calling isZero/isOne/isMinusOne.
llvm-svn: 307292
In rL300494 there was an attempt to deal with excessive compile time on
invocations of getSign/ZeroExtExpr using local caching. This approach only
helps if we request the same SCEV multiple times throughout recursion. But
in the bug PR33431 we see a case where we request different values all the time,
so caching does not help and the size of the cache grows enormously.
In this patch we remove the local cache for this methods and add the recursion
depth limit instead, as we do for arithmetics. This gives us a guarantee that the
invocation sequence is limited and reasonably short.
Differential Revision: https://reviews.llvm.org/D34273
llvm-svn: 306785
In LLVM IR the following code:
%r = urem <ty> %t, %b
is equivalent to:
%q = udiv <ty> %t, %b
%s = mul <ty> nuw %q, %b
%r = sub <ty> nuw %t, %q ; (t / b) * b + (t % b) = t
As UDiv, Mul and Sub are already supported by SCEV, URem can be
implemented with minimal effort this way.
Note: While SRem and SDiv are also related this way, SCEV does not
provides SDiv yet.
llvm-svn: 306695
Summary:
This patch changes getRange to getRangeRef and returns a reference to the ConstantRange object stored inside the DenseMap caches. We then take advantage of that to add new helper methods that can return min/max value of a signed or unsigned ConstantRange using that reference without first copying the ConstantRange.
getRangeRef calls itself recursively and I believe the reference return is fine for those calls.
I've left getSignedRange and getUnsignedRange returning a ConstantRange object so they will make a copy now. This is to ensure safety since the reference will be invalidated if the DenseMap changes.
I'm sure there are still more places that can take advantage of the reference and I'll submit future patches as I find them.
Reviewers: sanjoy, davide
Reviewed By: sanjoy
Subscribers: zzheng, llvm-commits, mzolotukhin
Differential Revision: https://reviews.llvm.org/D32978
llvm-svn: 306229
MulOpsInlineThreshold option of SCEV is defaulted to 1000, which is inadequately high.
When constructing SCEVs of expressions like:
x1 = a * a
x2 = x1 * x1
x3 = x2 * x2
...
We actually have huge SCEVs with max allowed amount of operands inlined.
Such expressions are easy to get from unrolling of loops looking like
x = a
for (i = 0; i < n; i++)
x = x * x
Or more tricky cases where big powers are involved. If some non-linear analysis
tries to work with a SCEV that has 1000 operands, it may lead to excessively long
compilation. The attached test does not pass within 1 minute with default threshold.
This patch decreases its default value to 32, which looks much more reasonable if we
use analyzes with complexity O(N^2) or O(N^3) working with SCEV.
Differential Revision: https://reviews.llvm.org/D34397
llvm-svn: 305882
The description of this option was copy-pasted from another one and does not
correspond to reality.
Differential Revision: https://reviews.llvm.org/D34390
llvm-svn: 305782
This is a fix for PR33292 that shows a case of extremely long compilation
of a single .c file with clang, with most time spent within SCEV.
We have a mechanism of limiting recursion depth for getAddExpr to avoid
long analysis in SCEV. However, there are calls from getAddExpr to getMulExpr
and back that do not propagate the info about depth. As result of this, a chain
getAddExpr -> ... .> getAddExpr -> getMulExpr -> getAddExpr -> ... -> getAddExpr
can be extremely long, with every segment of getAddExpr's being up to max depth long.
This leads either to long compilation or crash by stack overflow. We face this situation while
analyzing big SCEVs in the test of PR33292.
This patch applies the same limit on max expression depth for getAddExpr and getMulExpr.
Differential Revision: https://reviews.llvm.org/D33984
llvm-svn: 305463
I did this a long time ago with a janky python script, but now
clang-format has built-in support for this. I fed clang-format every
line with a #include and let it re-sort things according to the precise
LLVM rules for include ordering baked into clang-format these days.
I've reverted a number of files where the results of sorting includes
isn't healthy. Either places where we have legacy code relying on
particular include ordering (where possible, I'll fix these separately)
or where we have particular formatting around #include lines that
I didn't want to disturb in this patch.
This patch is *entirely* mechanical. If you get merge conflicts or
anything, just ignore the changes in this patch and run clang-format
over your #include lines in the files.
Sorry for any noise here, but it is important to keep these things
stable. I was seeing an increasing number of patches with irrelevant
re-ordering of #include lines because clang-format was used. This patch
at least isolates that churn, makes it easy to skip when resolving
conflicts, and gets us to a clean baseline (again).
llvm-svn: 304787
Params DT and LI are redundant, because these values are contained in fields anyways.
Differential Revision: https://reviews.llvm.org/D33668
llvm-svn: 304204
The optimistic delinearization implemented in LLVM detects array sizes by
looking for non-linear products between parameters and induction variables.
In OpenCL code, such products often look like:
A[get_global_id(0) * N + get_global_id(1)]
Hence, the IV is hidden in the get_global_id() call and consequently
delinearization would fail as no induction variable is available that helps
us to identify N as array size parameter.
We now use a very simple heuristic to change this. We assume that each parameter
that comes directly from a function call is a hidden induction variable. As
a result, we can delinearize the access above to:
A[get_global_id(0)][get_global_id(1]
llvm-svn: 304073
The patch rL303730 was reverted because test lsr-expand-quadratic.ll failed on
many non-X86 configs with this patch. The reason of this is that the patch
makes a correctless fix that changes optimizer's behavior for this test.
Without the change, LSR was making an overconfident simplification basing on a
wrong SCEV. Apparently it did not need the IV analysis to do this. With the
change, it chose a different way to simplify (that wasn't so confident), and
this way required the IV analysis. Now, following the right execution path,
LSR tries to make a transformation relying on IV Users analysis. This analysis
is target-dependent due to this code:
// LSR is not APInt clean, do not touch integers bigger than 64-bits.
// Also avoid creating IVs of non-native types. For example, we don't want a
// 64-bit IV in 32-bit code just because the loop has one 64-bit cast.
uint64_t Width = SE->getTypeSizeInBits(I->getType());
if (Width > 64 || !DL.isLegalInteger(Width))
return false;
To make a proper transformation in this test case, the type i32 needs to be
legal for the specified data layout. When the test runs on some non-X86
configuration (e.g. pure ARM 64), opt gets confused by the specified target
and does not use it, rejecting the specified data layout as well. Instead,
it uses some default layout that does not treat i32 as a legal type
(currently the layout that is used when it is not specified does not have
legal types at all). As result, the transformation we expect to happen does
not happen for this test.
This re-enabling patch does not have any source code changes compared to the
original patch rL303730. The only difference is that the failing test is
moved to X86 directory and now has requirement of running on x86 only to comply
with the specified target triple and data layout.
Differential Revision: https://reviews.llvm.org/D33543
llvm-svn: 303971
This continues the changes started when computeSignBit was replaced with this new version of computeKnowBits.
Differential Revision: https://reviews.llvm.org/D33431
llvm-svn: 303773
When folding arguments of AddExpr or MulExpr with recurrences, we rely on the fact that
the loop of our base recurrency is the bottom-lost in terms of domination. This assumption
may be broken by an expression which is treated as invariant, and which depends on a complex
Phi for which SCEVUnknown was created. If such Phi is a loop Phi, and this loop is lower than
the chosen AddRecExpr's loop, it is invalid to fold our expression with the recurrence.
Another reason why it might be invalid to fold SCEVUnknown into Phi start value is that unlike
other SCEVs, SCEVUnknown are sometimes position-bound. For example, here:
for (...) { // loop
phi = {A,+,B}
}
X = load ...
Folding phi + X into {A+X,+,B}<loop> actually makes no sense, because X does not exist and cannot
exist while we are iterating in loop (this memory can be even not allocated and not filled by this moment).
It is only valid to make such folding if X is defined before the loop. In this case the recurrence {A+X,+,B}<loop>
may be existant.
This patch prohibits folding of SCEVUnknown (and those who use them) into the start value of an AddRecExpr,
if this instruction is dominated by the loop. Merging the dominating unknown values is still valid. Some tests that
relied on the fact that some SCEVUnknown should be folded into AddRec's are changed so that they no longer
expect such behavior.
llvm-svn: 303730
This is a re-application of a r303497 that was reverted in r303498.
I thought it had broken a bot when it had not (the breakage did not
go away with the revert).
This change makes the split between the "exact" backedge taken count
and the "maximum" backedge taken count a bit more obvious. Both of
these are upper bounds on the number of times the loop header
executes (since SCEV does not account for most kinds of abnormal
control flow), but the latter is guaranteed to be a constant.
There were a few places where the max backedge taken count *was* a
non-constant; I've changed those to compute constants instead.
At this point, I'm not sure if the constant max backedge count can be
computed by calling `getUnsignedRange(Exact).getUnsignedMax()` without
losing precision. If it can, we can simplify even further by making
`getMaxBackedgeTakenCount` a thin wrapper around
`getBackedgeTakenCount` and `getUnsignedRange`.
llvm-svn: 303531
Javed Absar