The transform 'scalarizeLoadExtract' can be applied to scalable
vector types if the index is less than the minimum number of elements.
The check whether the index is less than the minimum number of elements
locates at line 1175~1180. 'scalarizeLoadExtract' will call
'canScalarizeAccess' and check the returned result if this transform is safe.
At the beginning of the function 'canScalarizeAccess', the index will be
checked
1. If it is less than the number of elements of a fixed vector type.
2. If it is less than the minimum number of elements of a scalable vector type.
Otherwise 'canScalarizeAccess' will return unsafe and this transform
will be prevented.
On the very first iteration for the reductions, when trying to build
reduction for boolean logic operations, no need to compare LHS/RHS with
the Reduction(VectorizedTree), need to compare with actual parameters of
the reduction operations.
Continuing the patch series to get rid of debug intrinsics [0], instruction
insertion needs to be done with iterators rather than instruction pointers,
so that we can communicate information in the iterator class. This patch
adds an iterator-taking insertBefore method and converts various call sites
to take iterators. These are all sites where such debug-info needs to be
preserved so that a stage2 clang can be built identically; it's likely that
many more will need to be changed in the future.
At this stage, this is just changing the spelling of a few operations,
which will eventually become signifiant once the debug-info bearing
iterator is used.
[0] https://discourse.llvm.org/t/rfc-instruction-api-changes-needed-to-eliminate-debug-intrinsics-from-ir/68939
Differential Revision: https://reviews.llvm.org/D152537
If the scalar does not need to be scheduled and it was vectorized
already in one of the vector nodes, we still can try to vectorize it in
another node. Just does not need account its cost in the scalar total
cost, as it will be handled in the main vectorized node.
Differential Revision: https://reviews.llvm.org/D159205
The return value of the function is only used to get the debug location.
Directly return the debug location, as this avoids an extra null
check in the caller.
This directly models the flags as part of the recipe, which allows
dropping them using the VPlan infrastructure when required.
It also allows removing the full reference to InductionDescriptor and
limit it to only the opcode.
We still can try to vectorize the bundle of the instructions, even if
the
repeated number of instruction is non-power-of-2. In this case need to
adjust the cost (calculate the cost only for unique scalar instructions)
and cost of the extracts. Also, when scheduling the bundle need to
schedule only unique scalars to avoid compiler crash because of the
multiple dependencies. Can be safely applied only if all scalars's users
are also vectorized and do not require memory accesses (this one is
a temporarily requirement, can be relaxed later).
---------
Co-authored-by: Alexey Bataev <a.bataev@outlook.com>
VPWidenRecipe only needs the opcode to widen, all other information
(flags, debug loc and operands) is already modeled directly via the
recipe.
This removes the remaining uses of the underlying instruction from
VPWidenRecipe::execute.
Add a dedicated debug location to VPRecipeBase to remove another
unneeded use of the underlying LLVM IR instruction and also consolidate
various DL fields in sub classes.
Each recipe can have debug location and it shouldn't rely on reference
to the underlying LLVM IR instructions to retain it. See various recipes
that had separate DL fields already.
This patch removes the member TTI from VPReductionRecipe, as the
generation of reduction operations no longer requires TTI.
Reviewed By: fhahn
Differential Revision: https://reviews.llvm.org/D158148
Update VPBlendRecipe::print() to print the result directly, instead of
relying on the stored Phi pointer. This brings the recipe in line with
how other recipes are printed.
Debug and pseudo instructions aren't modeled in VPlan. Turn a check into
an assertion. This will help removing the direct use of Inst here in the
future.
Extend VPRecipeWithIRFlags to also manage predicates for compares. This
allows removing the custom ICmpULE opcode from VPInstruction which was a
workaround for missing proper predicate handling.
This simplifies the code a bit while also allowing compares with any
predicates. It also fixes a case where the compare predixcate wasn't
printed properly for VPReplicateRecipes.
Discussed/split off from D150398.
Reviewed By: Ayal
Differential Revision: https://reviews.llvm.org/D158992
When a loop has multiple reductions, each with an intermediate invariant
store, the order in which those reductions are processed is not considered.
This can result in the invariant stores outside the loop not preserving the
original order.
This patch sorts VPReductionPHIRecipes by the order in which they have
stores in the original loop before running
`InnerLoopVectorizer::fixReduction` function, and it helps to maintain
the correct order of stores.
Fixes https://github.com/llvm/llvm-project/issues/64047
Differential Revision: https://reviews.llvm.org/D157631
Split off from D150398 to avoid builder-related diff changes there.
Using IRBuilder to create ICmps simplifies the result if both operands
are constants.
Reviewed By: Ayal
Differential Revision: https://reviews.llvm.org/D158332
ICmp codegen for VPInstructionD will be extended for other predicates,
and the operands could be any values (not just IV and TC as implied by
the names). Suggested cleanup from 150398.
Suppose we have a nested loop like this:
void foo(int32_t *dst, int32_t *src, int m, int n) {
for (int i = 0; i < m; i++) {
for (int j = 0; j < n; j++) {
dst[(i * n) + j] += src[(i * n) + j];
}
}
}
We currently generate runtime memory checks as a precondition for
entering the vectorised version of the inner loop. However, if the
runtime-determined trip count for the inner loop is quite small
then the cost of these checks becomes quite expensive. This patch
attempts to mitigate these costs by adding a new option to
expand the memory ranges being checked to include the outer loop
as well. This leads to runtime checks that can then be hoisted
above the outer loop. For example, rather than looking for a
conflict between the memory ranges:
1. &dst[(i * n)] -> &dst[(i * n) + n]
2. &src[(i * n)] -> &src[(i * n) + n]
we can instead look at the expanded ranges:
1. &dst[0] -> &dst[((m - 1) * n) + n]
2. &src[0] -> &src[((m - 1) * n) + n]
which are outer-loop-invariant. As with many optimisations there
is a trade-off here, because there is a danger that using the
expanded ranges we may never enter the vectorised inner loop,
whereas with the smaller ranges we might enter at least once.
I have added a HoistRuntimeChecks option that is turned off by
default, but can be enabled for workloads where we know this is
guaranteed to be of real benefit. In future, we can also use
PGO to determine if this is worthwhile by using the inner loop
trip count information.
When enabling this option for SPEC2017 on neoverse-v1 with the
flags "-Ofast -mcpu=native -flto" I see an overall geomean
improvement of ~0.5%:
SPEC2017 results (+ is an improvement, - is a regression):
520.omnetpp: +2%
525.x264: +2%
557.xz: +1.2%
...
GEOMEAN: +0.5%
I didn't investigate all the differences to see if they are
genuine or noise, but I know the x264 improvement is real because
it has some hot nested loops with low trip counts where I can
see this hoisting is beneficial.
Tests have been added here:
Transforms/LoopVectorize/runtime-checks-hoist.ll
Differential Revision: https://reviews.llvm.org/D152366
Split off mask creation for tail folding and proactively create the mask for
the header block.
This simplifies createBlockInMask.
Reviewed By: Ayal
Differential Revision: https://reviews.llvm.org/D157037
The transform 'foldSingleElementStore' can be applied to scalable
vector types if the index is less than the minimum number of elements.
Reviewed By: dmgreen, nikic
Differential Revision: https://reviews.llvm.org/D157676
If the masked gathers can be reordered, it may produce strided access
pattern and the reordering does not affect common reodering, better to
try to reorder masked gathers for better performance.
Differential Revision: https://reviews.llvm.org/D157009
Split off independent suggestion from D157037. This simplifies the
condition to decide if a recipe needs to be inserted to the header phi
section or simply appended.
The assertion has been updated to allow cases where the first non-phi
recipe is the end of the block, in which case inserting before this
point is equivalent to appending.
All dependencies on code from LoopVectorize.cpp have been
removed/refactored. Move the ::execute implementations to other recipe
definitions in VPlanRecipes.cpp
This commit refactors the implementation of VPReductionRecipe to use
reference instead of pointer for member RdxDesc. Because the member
RdxDesc in VPReductionRecipe should not be a nullptr, using a reference
will provide clearer semantics.
Reviewed By: fhahn
Differential Revision: https://reviews.llvm.org/D158058
Address post-commit simplification suggestion for 8a56179bcd8c:
Store operator only for floating point inductions (i.e. the binary op is
a FPMathOperator).
