When floating-point operations are legalized to operations of a higher
precision (e.g. f16 fadd being legalized to f32 fadd) then we get
narrowing then widening operations between each operation. With the
appropriate fast math flags (nnan ninf contract) we can eliminate these
casts.
Add lowering in tablegen for PARTIAL_REDUCE_U/SMLA ISD nodes. Only
happens when the combine has been performed on the ISD node. Also adds
in check to only do the DAG combine when the node can then eventually be
lowered, so changes neon tests too.
---------
Co-authored-by: James Chesterman <james.chesterman@arm.com>
Based off feedback for #129695 - we need to be able to determine the
load offset of smaller loads when trying to determine whether a multiple
use load should be split (in particular for AVX subvector extractions).
This patch adds a std::optional<unsigned> ByteOffset argument to
shouldReduceLoadWidth calls for where we know the constant offset to
allow targets to make use of it in future patches.
DAGCombiner::hoistLogicOpWithSameOpcodeHands will hoist
(or disjoint (ext a), (ext b)) -> (ext (or disjoint a, b))
So this adds patterns to match vwadd[u].v{v,x} in this case.
We have to teach the combine to preserve the disjoint flag.
Rename one signature of getAtomic to getAtomicLoad and pass LoadExtType.
Previously we had to set the extension type after the node was created,
but we don't usually modify SDNodes once they are created. It's possible
the node already existed and has been CSEd. If that happens, modifying
the node may affect the other users. It's therefore safer to add the
extension type at creation so that it is part of the CSE information.
I don't know of any failures related to the current implementation. I
only noticed that it doesn't match how we usually do things.
Fold an AND or OR of two NaN SETCC nodes into a single SETCC where
possible. This optimization already exists in InstCombine but adding in
here as well can allow for additional folding if more logical operations
are exposed.
Folds patterns such as:
unsigned foo(unsigned x, unsigned y) {
return x >= y ? x - y : x;
}
Before, on RISC-V:
sltu a2, a0, a1
addi a2, a2, -1
and a1, a1, a2
subw a0, a0, a1
Or, with Zicond:
sltu a2, a0, a1
czero.nez a1, a1, a2
subw a0, a0, a1
After, with Zbb:
subw a1, a0, a1
minu a0, a0, a1
Only applies to unsigned comparisons.
If `x >= y` then `x - y` is less than or equal `x`.
Otherwise, `x - y` wraps and is greater than `x`.
Almost all of the rotate idioms that are valid for an 'or' are also
valid when the halves are combined with an 'add'. Further, many of these
cases are not handled by common bits tracking meaning that the 'add' is
not converted to a 'disjoint or'.
Similar to what is done for visitEXTRACT_VECTOR_ELT - if all uses of a vector are EXTRACT_SUBVECTOR, then determine the accumulated demanded elts across all users and call SimplifyDemandedVectorElts in "AssumeSingleUse" use.
Second try after #133130 was reverted by #133331 due to it affecting reverted test files
Similar to what is done for visitEXTRACT_VECTOR_ELT - if all uses of a
vector are EXTRACT_SUBVECTOR, then determine the accumulated demanded
elts across all users and call SimplifyDemandedVectorElts in
"AssumeSingleUse" use.
This patch improves DAGCombiner's handling of potential store merges by
detecting function calls between loads and stores. When a function call
exists in the chain between a load and its corresponding store, we avoid
merging these stores if the spilling is unprofitable.
We had to implement a hook on TLI, since TTI is unavailable in
DAGCombine. Currently, it's only enabled for riscv.
This is the DAG equivalent of PR #129258
The pattern
```LLVM
%shl = shl i32 %x, 31
%ashr = ashr i32 %shl, 31
```
would be combined to `SIGN_EXTEND_INREG %x, ValueType:ch:i1` by
SelectionDAG.
However InstCombine normalizes this pattern to:
```LLVM
%and = and i32 %x, 1
%neg = sub i32 0, %and
```
This adds matching code to DAGCombiner to catch this variant as well.
This option was added to improve test coverage for SVE lowering code
that is impossible to reach otherwise. Given it is not possible to
trigger a bug without it and the generated code is universally worse
with it, I figure the option has no value and should be removed.
Add generic DAG combine for ISD::PARTIAL_REDUCE_U/SMLA nodes. Transforms
the DAG from:
PARTIAL_REDUCE_MLA(Acc, MUL(EXT(MulOpLHS), EXT(MulOpRHS)), Splat(1)) to
PARTIAL_REDUCE_MLA(Acc, MulOpLHS, MulOpRHS).
Since N2 will be reused in the fold, we cannot skip N2's undef elements
if the corresponding element in N1 is well-defined.
For example:
```
t2: v4i32 = BUILD_VECTOR Constant:i32<0>, Constant:i32<0>, Constant:i32<0>, Constant:i32<0>
t24: v4i32 = BUILD_VECTOR undef:i32, undef:i32, Constant:i32<1>, undef:i32
t11: v4i32 = vselect t8, t2, t10
```
Before this patch, we fold t11 into:
```
t26: v4i32 = sign_extend t8
t27: v4i32 = add t26, t24
```
The last element of t27 is incorrect.
Closes https://github.com/llvm/llvm-project/issues/129181.
Begin extending replaceShuffleOfInsert to handle other forms of scalar insertion into a vector.
I've limited this to targets that have Custom/Legal ISD::INSERT_VECTOR_ELT handling for now - although we can probably always fold this before LegalOperations.
In DAGCombiner.cpp preserve range metadata when load is narrowed to load
LSBs if original range metadata bounds can fit in the narrower type.
Utilize preserved range metadata to reduce 64-bit shl to 32-bit shl.
---------
Signed-off-by: John Lu <John.Lu@amd.com>
The Target hook convertSelectOfConstantsToMath() needs to be used within
SimplifySelectCC helper combine function in SelectionDAG Isel, where
generic select folding with constants is happening into simple maths op
using the condition as it is.
It necessarily fixes#121145.
`N` may get merged with existing nodes inside the loop. Early exit when
it is deleted to avoid the crash.
Alternative solution: use `DAGNodeDeletedListener` to refresh the value
of N.
Closes https://github.com/llvm/llvm-project/issues/128143.
Shift amounts in SelectionDAG don't have to match the result type
of the shift. SelectionDAGBuilder will aggressively truncate shift
amounts to the target's preferred type. This may result in a zero extend
that existed in IR being removed.
If we look through a truncate here, we can't guarantee the upper bits
of the truncate input are zero. There may have been a zext that was
removed. Unfortunately, this regresses tests where no truncate was
involved. The only way I can think to fix this is to add an assertzext
when SelectionDAGBuilder truncates a shift amount or remove the
early truncation of shift amounts from SelectionDAGBuilder all together.
Fixes#126889.
Fix the case where the vector element type of the loaded extractelement
input does not match the result type of the extract.
This fixes a regression reported after
c55a7659b38946350315ac4a18d9805deb1f0a54
Happened to notice some odd things related to chains in this code.
The code calls hasOneUse on LoadSDNode* which will check users
of the data and the chain. I think this was trying to check that
the data had one use so one of the loads would definitely be
removed by the transform. Load chains don't always have users so
our testing may not have noticed that the chains being used would
block the transform.
The code makes all users of ld1's chain use the new load's chain, but
we don't know that ld1 becomes dead. This can cause incorrect dependencies if
ld1's chain is used and it isn't deleted. I think the better thing to do
is use makeEquivalentMemoryOrdering to make all users of ld0 and ld1
depend on the new load and the original loads. If the olds loads become
dead, their chain will be cleaned up later.
I'm having trouble getting a test for any ordering issue with the current code.
areNonVolatileConsecutiveLoads requires the two loads to have the same
input chain. Given that, I don't know how to use one of the load chain
results without also using the other. If they are both used we don't
do the transform because SDNode::hasOneUse will return false for both.
Previously this combine would undo AMDGPU's new custom legalization of
wide vector shuffles into 2 element pieces. The comment also
states that this combine is only done before legalization,
but the case with a build_vector source was unconditional.
We probably don't want to do this if the multiple uses are full
scalarization of the vector, but this seems to work well enough.
Scalarizing extracts should have folded out pre-legalize.
Once we get to SelectionDAG the IR should not be changing anymore, so we
can use BatchAAResults rather than AAResults to cache AA queries.
This should be a NFC change for targets that enable AA during codegen
(such as AArch64), but also give a nice compile-time improvement in some
cases. See:
https://github.com/llvm/llvm-project/pull/123787#issuecomment-2606797041
Note: This follows Nikita's suggestion on #123787.
For shuffle vector splats with undef lanes in the mask,
this was introducing real values. Filter out build_vector
results based on the undef elements in the mask.
This avoids AMDGPU test regressions in a future change.
test/CodeGen/X86/urem-seteq-illegal-types.ll looks worse
but I didn't investigate.
