If the vectorized insertelements instructions form indentity subvector
(the subvector at the beginning of the long vector), it is just enough
to extend the vector itself, no need to generate inserting subvector
shuffle.
Differential Revision: https://reviews.llvm.org/D107494
If the vectorized insertelements instructions form indentity subvector
(the subvector at the beginning of the long vector), it is just enough
to extend the vector itself, no need to generate inserting subvector
shuffle.
Differential Revision: https://reviews.llvm.org/D107344
Replace insertelement instructions for splats with just single
insertelement + broadcast shuffle. Also, try to merge these instructions
if they come from the same/shuffled gather node.
Differential Revision: https://reviews.llvm.org/D107104
For the nodes with reused scalars the user may be not only of the size
of the final shuffle but also of the size of the scalars themselves,
need to check for this. It is safe to just modify the check here, since
the order of the scalars themselves is preserved, only indeces of the
reused scalars are changed. So, the users with the same size as the
number of scalars in the node, will not be affected, they still will get
the operands in the required order.
Reported by @mstorsjo in D105020.
Differential Revision: https://reviews.llvm.org/D107080
If the instruction was previously deleted, it should not be treated as
an external user. This fixes cost estimation and removes dead
extractelement instructions.
Differential Revision: https://reviews.llvm.org/D107106
Need to check that the minimum acceptable vector factor is at least 2,
not 0, to avoid compiler crash during gathered loads analysis.
Differential Revision: https://reviews.llvm.org/D107058
Reworked reordering algorithm. Originally, the compiler just tried to
detect the most common order in the reordarable nodes (loads, stores,
extractelements,extractvalues) and then fully rebuilding the graph in
the best order. This was not effecient, since it required an extra
memory and time for building/rebuilding tree, double the use of the
scheduling budget, which could lead to missing vectorization due to
exausted scheduling resources.
Patch provide 2-way approach for graph reodering problem. At first, all
reordering is done in-place, it doe not required tree
deleting/rebuilding, it just rotates the scalars/orders/reuses masks in
the graph node.
The first step (top-to bottom) rotates the whole graph, similarly to the previous
implementation. Compiler counts the number of the most used orders of
the graph nodes with the same vectorization factor and then rotates the
subgraph with the given vectorization factor to the most used order, if
it is not empty. Then repeats the same procedure for the subgraphs with
the smaller vectorization factor. We can do this because we still need
to reshuffle smaller subgraph when buildiong operands for the graph
nodes with lasrger vectorization factor, we can rotate just subgraph,
not the whole graph.
The second step (bottom-to-top) scans through the leaves and tries to
detect the users of the leaves which can be reordered. If the leaves can
be reorder in the best fashion, they are reordered and their user too.
It allows to remove double shuffles to the same ordering of the operands in
many cases and just reorder the user operations instead. Plus, it moves
the final shuffles closer to the top of the graph and in many cases
allows to remove extra shuffle because the same procedure is repeated
again and we can again merge some reordering masks and reorder user nodes
instead of the operands.
Also, patch improves cost model for gathering of loads, which improves
x264 benchmark in some cases.
Gives about +2% on AVX512 + LTO (more expected for AVX/AVX2) for {625,525}x264,
+3% for 508.namd, improves most of other benchmarks.
The compile and link time are almost the same, though in some cases it
should be better (we're not doing an extra instruction scheduling
anymore) + we may vectorize more code for the large basic blocks again
because of saving scheduling budget.
Differential Revision: https://reviews.llvm.org/D105020
Need to fix several cost-related problems. The final type may be defined
incorrectly because of to early definition (we may end up with the wider
type), the CommonCost should not be redefined in ExtractElements
cost related calculations and the shuffle of the final insertelements
vectors should be calculated as a cost of single vector permutations
+ costs of two vector permutations for other n-1 incoming vectors.
Differential Revision: https://reviews.llvm.org/D106578
Need to fix several cost-related problems. The final type may be defined
incorrectly because of to early definition (we may end up with the wider
type), the CommonCost should not be redefined in ExtractElements
cost related calculations and the shuffle of the final insertelements
vectors should be calculated as a cost of single vector permutations
+ costs of two vector permutations for other n-1 incoming vectors.
Differential Revision: https://reviews.llvm.org/D106578
I have added a new FastMathFlags parameter to getArithmeticReductionCost
to indicate what type of reduction we are performing:
1. Tree-wise. This is the typical fast-math reduction that involves
continually splitting a vector up into halves and adding each
half together until we get a scalar result. This is the default
behaviour for integers, whereas for floating point we only do this
if reassociation is allowed.
2. Ordered. This now allows us to estimate the cost of performing
a strict vector reduction by treating it as a series of scalar
operations in lane order. This is the case when FP reassociation
is not permitted. For scalable vectors this is more difficult
because at compile time we do not know how many lanes there are,
and so we use the worst case maximum vscale value.
I have also fixed getTypeBasedIntrinsicInstrCost to pass in the
FastMathFlags, which meant fixing up some X86 tests where we always
assumed the vector.reduce.fadd/mul intrinsics were 'fast'.
New tests have been added here:
Analysis/CostModel/AArch64/reduce-fadd.ll
Analysis/CostModel/AArch64/sve-intrinsics.ll
Transforms/LoopVectorize/AArch64/strict-fadd-cost.ll
Transforms/LoopVectorize/AArch64/sve-strict-fadd-cost.ll
Differential Revision: https://reviews.llvm.org/D105432
The incoming values for PHI nodes may come from unreachable BasicBlocks,
need to handle this case.
Differential Revision: https://reviews.llvm.org/D106264
Part of D105020. Also, fixed FIXMEs that need to use wider vector type
when trying to calculate the cost of reused scalars. This may cause
regressions unless D100486 is landed to improve the cost estimations
for long vectors shuffling.
Differential Revision: https://reviews.llvm.org/D106060
The cost of the InsertSubvector shuffle kind cost is not complete and
may end up with just extracts + inserts costs in many cases. Added
a workaround to represent it as a generic PermuteSingleSrc, which is
still pessimistic but better than InsertSubvector.
Differential Revision: https://reviews.llvm.org/D105827
This bug was introduced with D105730 / 25ee55c0ba .
If we are not converting all of the operations of a reduction
into a vector op, we need to preserve the existing select form
of the remaining ops. Otherwise, we are potentially leaking
poison where it did not in the original code.
Alive2 agrees that the version that freezes some inputs
and then falls back to scalar is correct:
https://alive2.llvm.org/ce/z/erF4K2
The cost of the InsertSubvector shuffle kind cost is not complete and
may end up with just extracts + inserts costs in many cases. Added
a workaround to represent it as a generic PermuteSingleSrc, which is
still pessimistic but better than InsertSubvector.
Differential Revision: https://reviews.llvm.org/D105827
This has been a work-in-progress for a long time...we finally have all of
the pieces in place to handle vectorization of compare code as shown in:
https://llvm.org/PR41312
To do this (see PhaseOrdering tests), we converted SimplifyCFG and
InstCombine to the poison-safe (select) forms of the logic ops, so now we
need to have SLP recognize those patterns and insert a freeze op to make
a safe reduction:
https://alive2.llvm.org/ce/z/NH54Ah
We get the minimal patterns with this patch, but the PhaseOrdering tests
show that we still need adjustments to get the ideal IR in some or all of
the motivating cases.
Differential Revision: https://reviews.llvm.org/D105730
This makes it clearer when we have encountered the extra arg.
Also, we may need to adjust the way the operand iteration
works when handling logical and/or.
This is NFC-intended currently (so no test diffs). The motivation
is to eventually allow matching for poison-safe logical-and and
logical-or (these are in the form of a select-of-bools).
( https://llvm.org/PR41312 )
Those patterns will not have all of the same constraints as min/max
in the form of cmp+sel. We may also end up removing the cmp+sel
min/max matching entirely (if we canonicalize to intrinsics), so
this will make that step easier.
This patch removes the IsPairwiseForm flag from the Reduction Cost TTI
hooks, along with some accompanying code for pattern matching reductions
from trees starting at extract elements. IsPairWise is now assumed to be
false, which was the predominant way that the value was used from both
the Loop and SLP vectorizers. Since the adjustments such as D93860, the
SLP vectorizer has not relied upon this distinction between paiwise and
non-pairwise reductions.
This also removes some code that was detecting reductions trees starting
from extract elements inside the costmodel. This case was
double-counting costs though, adding the individual costs on the
individual instruction _and_ the total cost of the reduction. Removing
it changes the costs in llvm/test/Analysis/CostModel/X86/reduction.ll to
not double count. The cost of reduction intrinsics is still tested
through the various tests in
llvm/test/Analysis/CostModel/X86/reduce-xyz.ll.
Differential Revision: https://reviews.llvm.org/D105484
Patch tries to improve the vectorization of stores. Originally, we just
check the type and the base pointer of the store.
Patch adds some extra checks to avoid non-profitable vectorization
cases. It includes analysis of the scalar values to be stored and
triggers the vectorization attempt only if the scalar values have
same/alt opcode and are from same basic block, i.e. we don't end up
immediately with the gather node, which is not profitable.
This also improves compile time by filtering out non-profitable cases.
Part of D57059.
Differential Revision: https://reviews.llvm.org/D104122
Revived D101297 in its original form + added some changes in X86
legalization cehcking for masked gathers.
This solution is the most stable and the most correct one. We have to
check the legality before trying to build the masked gather in SLP.
Without this check we have incorrect cost (for SLP) in case if the masked gather
is not legal/slower than the gather. And we're missing some
vectorization opportunities.
This can be fixed in the cost model, but in this case we need to add
special checks for the cost of GEPs for ScatterVectorize node, add
special check for small trees, etc., i.e. there are a lot of corner
cases here and there, which insrease code base and make it harder to
maintain the code.
> Can't we rely on cost model to deal with this? This can be profitable for futher vectorization, when we can start from such gather loads as seed.
The question from D101297. Actually, no, it can't. Actually, simple
gather may give us better result, especially after we started
vectorization of insertelements. Plus, like I said before, the cost for
non-legal masked gathers leads to missed vectorization opportunities.
Differential Revision: https://reviews.llvm.org/D105042
The reduction matching was probably only dealing with binops
when it was written, but we have now generalized it to handle
select and intrinsics too, so assert on that too.
Compare type IDs and DFS numbering for basic block instead of addresses
to fix non-determinism.
Differential Revision: https://reviews.llvm.org/D105031
The function vectorizeChainsInBlock does not support scalable vector,
because function like canReuseExtract and isCommutative in the code
path assert with scalable vectors.
This patch avoids vectorizing blocks that have extract instructions with scalable
vector..
Differential Revision: https://reviews.llvm.org/D104809
Same as other CreateLoad-style APIs, these need an explicit type
argument to support opaque pointers.
Differential Revision: https://reviews.llvm.org/D105395
The compiler should not ignore UndefValue when gathering the scalars,
otherwise the resulting code may be less defined than the original one.
Also, grouped scalars to insert them at first to reduce the analysis in
further passes.
Differential Revision: https://reviews.llvm.org/D105275
Make getPointersDiff() and sortPtrAccesses() compatible with opaque
pointers by explicitly passing in the element type instead of
determining it from the pointer element type.
The SLPVectorizer result is slightly non-optimal in that unnecessary
pointer bitcasts are added.
Differential Revision: https://reviews.llvm.org/D104784
Perform better analysis when trying to vectorize PHIs.
1. Do not try to vectorize vector PHIs.
2. Do deeper analysis for more profitable nodes for the vectorization.
Before we just tried to vectorize the PHIs of the same type. Patch
improves this and tries to vectorize PHIs with incoming values which
come from the same basic block, have the same and/or alternative
opcodes.
It allows to save the compile time and provides better vectorization
results in general.
Part of D57059.
Differential Revision: https://reviews.llvm.org/D103638
This can be seen as a follow up to commit 0ee439b705,
that changed the second argument of __powidf2, __powisf2 and
__powitf2 in compiler-rt from si_int to int. That was to align with
how those runtimes are defined in libgcc.
One thing that seem to have been missing in that patch was to make
sure that the rest of LLVM also handle that the argument now depends
on the size of int (not using the si_int machine mode for 32-bit).
When using __builtin_powi for a target with 16-bit int clang crashed.
And when emitting libcalls to those rtlib functions, typically when
lowering @llvm.powi), the backend would always prepare the exponent
argument as an i32 which caused miscompiles when the rtlib was
compiled with 16-bit int.
The solution used here is to use an overloaded type for the second
argument in @llvm.powi. This way clang can use the "correct" type
when lowering __builtin_powi, and then later when emitting the libcall
it is assumed that the type used in @llvm.powi matches the rtlib
function.
One thing that needed some extra attention was that when vectorizing
calls several passes did not support that several arguments could
be overloaded in the intrinsics. This patch allows overload of a
scalar operand by adding hasVectorInstrinsicOverloadedScalarOpd, with
an entry for powi.
Differential Revision: https://reviews.llvm.org/D99439
It was found by chance revealing discrepancy between comment (few lines above),
the condition and how re-ordering of instruction is done inside the if statement
it guards. The condition was always evaluated to true.
Differential Revision: https://reviews.llvm.org/D104064
There is no need to schedule insertelement instructions. The compiler
did not schedule them before it started support their vectorization and
it should not do it after. We pre-schedule them manually when finding
a build vector sequence.
Disabling scheduling of insertelement instructions improves compile
time and vectorization of the very large basic blocks by saving
scheduling budget for other instructions.
Differential Revision: https://reviews.llvm.org/D104026
1. Better sorting of scalars to be gathered. Trying to insert
constants/arguments/instructions-out-of-loop at first and only then
the instructions which are inside the loop. It improves hoisting of
invariant insertelements instructions.
2. Better detection of shuffle candidates in gathering function.
3. The cost of insertelement for constants is 0.
Part of D57059.
Differential Revision: https://reviews.llvm.org/D103458
No need to recalculate the cost of extractelements, just no need to
compensate the cost of all extractelements, need to check before if this
is actually going to be removed at the vectorization. Also, no need to
generate new extractelement instruction, we may just regenerate the
original one. It may improve the final vectorization.
Differential Revision: https://reviews.llvm.org/D102933
tryToVectorizeList function allows to reorder only 2 scalars. Patch
allows to reorder >2 scalars. Also, to avoid possible regressions, it
allows extra vectorization of the remaining parts of the scalars
elements if possible.
Part of D57059.
Differential Revision: https://reviews.llvm.org/D103247
As noticed by NAKAMURA Takumi back in 2017, we cannot use
properlyDominates for std::stable_sort as properlyDominates only
partially orders blocks. That is, for blocks A, B, C, D, where A
dominates B and C dominates D, we have A == C, B == C, but A < B. This
is not a valid comparison function for std::stable_sort and causes
different results between libstdc++ and libc++. This change uses DFS
numbering to give deterministic results for all reachable blocks.
Unreachable blocks are ignored already, so do not need special
consideration.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D103441
As the existing test unreachable.ll shows, we should be doing more
work to avoid entering unreachable blocks: we should not stop
vectorization just because a PHI incoming value from an unreachable
block cannot be vectorized. We know that particular value will never
be used so we can just replace it with poison.
Implemented better scheme for perfect/shuffled matches of the gather
nodes which allows to fix the performance regressions introduced by
earlier patches. Starting detecting matches for broadcast nodes and
extractelement gathering.
Differential Revision: https://reviews.llvm.org/D102920
Alexey Bataev