If decided to resize the instruction, need to drop wrapping flags from
the resulting vector instructions to avoid incorrect
optimizations/assumptions later.
Fixes PR75995.
Issue is covered by existing test
llvm/test/Transforms/SLPVectorizer/RISCV/phi-const.ll
See issue #75632 for ideas for how we could catch these more easily in
the future.
As suggested post-commit for a00227197, replace unnecessary getNumUsers
calls by boolean variable to indicate if users changed. Note that this
also requires an early exit to detect the case where a value is replaced
by itself.
SLP Vectorizer can discard vector entries at unknown positions. This
example shows the behaviour:
https://godbolt.org/z/or43EM594
The following instruction inserts an element at an unknown position:
```
%2 = insertelement <3 x i64> poison, i64 %value, i64 %position
```
The position depends on an argument that is unknown at compile time.
After running SLP, one can see there is no more instruction present
referencing `%position`.
This happens as SLP parallelizes the two adds in the example. It then
needs to merge the original vector with the new vector.
Within `isUndefVector`, the SLP vectorizer constructs a bitmap
indicating which elements of the original vector are poison values. It
does this by walking the insertElement instructions.
If it encounters an insert with a non-constant position, it is ignored.
This will result in poison values to be used for all entries, where
there are no inserts with constant positions.
However, as the position is unknown, the element could be anywhere.
Therefore, I think it is only safe to assume none of the entries are
poison values and to simply take them all over when constructing the
shuffleVector instruction.
This fixes#75437
This reverts commit 19918ac34dc5d304ec6ad413ceae1d4394abe28f.
Fixes#75298. There is still a case where we miss the correct users
outside the main vector loop for reductions, and that is tail-folded
loops with reductions where the final value is stored after the loop.
This should be handled explicitly in #70253
This patch starts initial modeling of VF * UF in VPlan.
Initially, introduce a dedicated VFxUF VPValue, which is then
populated during VPlan::prepareToExecute. Initially, the VF * UF
applies only to the main vector loop region. Once we extend the
scope of VPlan in the future, we may want to associate different VFxUFs
with different vector loop regions (e.g. the epilogue vector loop)
This allows explicitly parameterizing recipes that rely on the
VF * UF, like the canonical induction increment. At the moment, this
mainly helps to avoid generating some duplicated calls to vscale with
scalable vectors. It should also allow using EVL as induction increments
explicitly in D99750. Referring to VF * UF is also needed in other
places that we plan to migrate to VPlan, like the minimum trip count
check during skeleton creation.
The first version creates the value for VF * UF directly in
prepareToExecute to limit the scope of the patch. A follow-on patch will
model VF * UF computation explicitly in VPlan using recipes.
Moved from Phabricator (https://reviews.llvm.org/D157322)
nodes, having same last instruction.
If the user nodes has the same last-instruction, used as insert points
for the buildvector nodes, finding the proper dependency is crucial.
Before, it depended on the indices of the buildvectors themselves but
looks like it should depend on indices of the user nodes, because it
identifies the vectorization order and, thus, properly aligns
buildvector nodes in terms of def-use chain.
Minor simplification applied to VFShape::getScalarShape,
VFShape::get, and VFABI::tryDemangleForVFABI methods.
Also, remove unnecessary `static_cast` in `SLPVectorizer.cpp`
A new disjoint flag was added for OR instructions in #72583.
Update VPRecipeWithIRFlags to also support the new flag. This
allows printing and preserving the disjoint flag in vectorized code.
Compiler crashes when the assertion triggered for zext nneg instruction,
that checks that the instruction cannot produce poison. Changed the base
class for widencast recipe to handle dropping nneg flag to avoid
compiler crash.
MinBWs contains entries that specify the minimum required bitwidth. In
some cases, the old and new bitwidths can be equal (see test case) and
in those cases no truncations are needed, so skip those cases.
Fixes#74307.
In some cases MinBWs may contain entries for live-ins that are not used
by VPWidenRecipe or VPWidenSelectRecipes. In those cases, the live-ins
won't get processed, so make sure we include them in the count when used
as operands in VPWidenCast and VPWidenSelectRecipe.
Fixes https://github.com/llvm/llvm-project/issues/74231
This patch replaces the IR based truncateToMinimalBitwidths with a VPlan
version. This has 3 benefits:
1) the VPlan-based version is simpler; we don't need to implement
special codegen for each supported instruction type like the IR based
one.
2) Removes a dependency on the cost-model after VPlan execution and
3) Removes a use of getVPValue that uses underlying values after VPlan
execution (See removed FIXME).
Depends on D149081.
Depends on D149079.
Reviewed By: Ayal
Differential Revision: https://reviews.llvm.org/D149903
After inserting a select for the final value, update the VPlan def-use
chains. At the moment, the incorrect live-out doesn't cause a
mis-compile, as computing the final reduction value is not yet modeled
in VPlan.
Currently we emit gathers for scalars being vectorized in the tree as
a pair of extractelement/insertelement instructions. Instead we can try
to find all required vectors and emit shuffle vector instructions
directly, improving the code and reducing compile time.
Part of non-power-of-2 vectorization.
Differential Revision: https://reviews.llvm.org/D110978
Currently, MinBWs map uses Value* as a key and stores mapping for each
value to be demoted. It make is it hard to get the actual MinBWs value
for the buildvector scalars(constants), since same constant might be
used in different nodes with the different MinBWs values/decisions.
Also, it consumes extra memory for the vectorized values/instructions
from the same nodes.
Better to map actual nodes. It fixes the bitwidth data fetching for
buildvector scalars and improves memory consumption/analysis time for
other instructions.
Part of the "RemoveDIs" project to remove debug intrinsics requires
passing block-positions around in iterators rather than as instruction
pointers, allowing some debug-info to reside in BasicBlock::iterator.
This means getInsertionPointAfterDef has to return an iterator, and as
it can return no-instruction that means returning an optional iterator.
This patch changes the signature for getInsertionPtAfterDef and then
patches up the various places that use it to handle the different type.
This would overall be an NFC patch, however in
InstCombinerImpl::freezeOtherUses I've started skipping any debug
intrinsics at the returned insert-position. This should not have any
_meaningful_ effect on the compiler output: at worst it means variable
assignments that are skipped will now cover the freeze instruction and
anything inserted before it, which should be inconsequential.
Sadly: this makes the function signature ugly. This is probably the
ugliest piece of fallout for the "RemoveDIs" work, but it serves the
overall purpose of improving compile times and not allowing `-g` to
affect compiler output, so should be worthwhile in the end.
No need to store optional boolean in the map, enough to store boolean
directly. Also, we can do preliminary check for instruction and if they
are not simple, mark as aliased without storing this result in the map.
If the gather node includes ordered loads only partially (not the whole
node consists of loads) and the other gathered scalar are not loads, and
no other dependency from other nodes is found, we still can improve the
cost of gather, if take into account the fact that these loads still can
be vectorized.
