When the source has a series of assignments, users reasonably want to
have the debugger step through each one individually. Turn off the combine
for adjacent stores so we get this behavior at -O0.
Similar to D7181.
Reviewed By: spatel, xgupta
Differential Revision: https://reviews.llvm.org/D115808
When the source has a series of assignments, users reasonably want to
have the debugger step through each one individually. Turn off the combine
for adjacent stores so we get this behavior at -O0.
Similar to D7181.
Differential Revision: https://reviews.llvm.org/D115808
Merge the node combines into a common DAGCombiner::visitFMinMax (like we do for IMINMAX).
Move the constant folding into SelectionDAG::foldConstantFPMath.
This allows us to fold the vecreduce-propagate-sd-flags.ll test as it reduces constants - so I've refactored it to take variables instead.
Differential Revision: https://reviews.llvm.org/D115952
We were using a function attribute to indicate a non-standard FP mode,
but now we can use intrinsics for that job as shown in the new tests.
Presumably the x86 asm could be improved for that IR with intrinsics,
but I have not worked out exactly how to do that. Note that the
transform to FTRUNC still requires a hacky check for "nsz" (because
FMF are not applied to FP casts).
This is a cleanup based on the clang change in D115804 / 8c7f2a4f87 .
This is effectively a revert of 5a90285bd9 + D46237 .
Differential Revision: https://reviews.llvm.org/D115885
Replace custom constant scalar/splat folding with FoldConstantArithmetic call and canonicalize commutative constant ops to the RHS before the SimplifyVBinOp call
This extends the custom lowering for truncating stores on
fixed length vectors in SVE to support masked truncating stores.
It also adds a DAG combine for truncates followed by masked
stores.
Reviewed By: peterwaller-arm, paulwalker-arm
Differential Revision: https://reviews.llvm.org/D108115
SimplifyVBinOp still has a FoldConstantArithmetic call, which now it isn't vector specific we should be able to remove (once fp binops are tidied up); but we can at least clean up the integer opcodes to perform the basic constant/undef handling in common code first.
After D113888 / 32b6c17b29 the MMO size of a masked loads/store is
unknown. When we are converting back to a standard load/store because
the mask is known all ones, we can refine that to the correct size from
the size of the vector being loaded/stored.
Differential Revision: https://reviews.llvm.org/D114582
As an extension to D111976, this converts clamp fptosi, clamped between
0 and (2^n)-1 to a fptoui.sat. This can greatly help on targets with
conversions that naturally saturate, such as Arm.
X86 disables the transform as some of the test cases increases in size.
A fptoui.sat necessitates a fp clamp without native support, so there is
little use in converting if the instruction is just going to be
expanded.
Differential Revision: https://reviews.llvm.org/D112428
combinePMULH currently only truncates vXi32/vXi64 multiplies to PMULHW/PMULUW if the source operands are SEXT/ZEXT instructions for a 'free' truncation.
But we can generalize this to any source operand with sufficient leading sign/zero bits that would allow PACKS/PACKUS to be used as a 'cheap' truncation.
This helps us avoid the wider multiplies, in exchange for truncation on both source operands instead of the result.
Differential Revision: https://reviews.llvm.org/D113371
The REM DAG combine uses the visitDivLike functions to try and get an
optimized DIV node to provide better codegen, however in some cases this
visitDivLike call ends up in the BuildSDIVPow2 target hook, which in
turn sometimes will return the same node passed in to indicate not to
change it. The REM DAG combine does not anticipate this and creates a
cycle in the DAG because of it.
Fix this by ensuring any such optimized div node returned is distinct
from the node being combined.
Differential Revision: https://reviews.llvm.org/D114716
This adds a fold in DAGCombine to create fptosi_sat from sequences for
smin(smax(fptosi(x))) nodes, where the min/max saturate the output of
the fp convert to a specific bitwidth (say INT_MIN and INT_MAX). Because
it is dealing with smin(/smax) in DAG they may currently be ISD::SMIN,
ISD::SETCC/ISD::SELECT, ISD::VSELECT or ISD::SELECT_CC nodes which need
to be handled similarly.
A shouldConvertFpToSat method was added to control when converting may
be profitable. The original fptosi will have a less strict semantics
than the fptosisat, with less values that need to produce defined
behaviour.
This especially helps on ARM/AArch64 where the vcvt instructions
naturally saturate the result.
Differential Revision: https://reviews.llvm.org/D111976
It causes builds to fail with this assert:
llvm/include/llvm/ADT/APInt.h:990:
bool llvm::APInt::operator==(const llvm::APInt &) const:
Assertion `BitWidth == RHS.BitWidth && "Comparison requires equal bit widths"' failed.
See comment on the code review.
> This adds a fold in DAGCombine to create fptosi_sat from sequences for
> smin(smax(fptosi(x))) nodes, where the min/max saturate the output of
> the fp convert to a specific bitwidth (say INT_MIN and INT_MAX). Because
> it is dealing with smin(/smax) in DAG they may currently be ISD::SMIN,
> ISD::SETCC/ISD::SELECT, ISD::VSELECT or ISD::SELECT_CC nodes which need
> to be handled similarly.
>
> A shouldConvertFpToSat method was added to control when converting may
> be profitable. The original fptosi will have a less strict semantics
> than the fptosisat, with less values that need to produce defined
> behaviour.
>
> This especially helps on ARM/AArch64 where the vcvt instructions
> naturally saturate the result.
>
> Differential Revision: https://reviews.llvm.org/D111976
This reverts commit 52ff3b0093.
This adds a fold in DAGCombine to create fptosi_sat from sequences for
smin(smax(fptosi(x))) nodes, where the min/max saturate the output of
the fp convert to a specific bitwidth (say INT_MIN and INT_MAX). Because
it is dealing with smin(/smax) in DAG they may currently be ISD::SMIN,
ISD::SETCC/ISD::SELECT, ISD::VSELECT or ISD::SELECT_CC nodes which need
to be handled similarly.
A shouldConvertFpToSat method was added to control when converting may
be profitable. The original fptosi will have a less strict semantics
than the fptosisat, with less values that need to produce defined
behaviour.
This especially helps on ARM/AArch64 where the vcvt instructions
naturally saturate the result.
Differential Revision: https://reviews.llvm.org/D111976
This allows the generic DAG combine to fold fp_extend/fp_trunc into
loads/stores which we can then lower into a integer extending
load/truncating store plus an FP_EXTEND/FP_ROUND.
The nuance here is that fixed-type FP_EXTEND/FP_ROUND require unpacked
types hence lowering them introduces an unpack/zip. By allowing these
nodes to be combined with loads/store we make it much easier to have
this unpack/zip combined into the load/store by our custom lowering.
Differential Revision: https://reviews.llvm.org/D114580
Patch to fix some of the regressions in D77804.
By folding to rotate/funnel-shift by constant amounts for illegal types, we prevent SimplifyDemandedBits from destroying the patterns prematurely, allowing us to use the rotate/funnel-shift legalization that was added in D112443.
Differential Revision: https://reviews.llvm.org/D113192
It's possible that the mask is already a NOT. At least if InstCombine
hasn't canonicalized the input. In that case we will form an ANDN with
X instead of with Y. So we don't need to worry about Y being a constant.
We might need to check that X isn't a constant instead, but we don't
have a test case for that yet.
This fixes a size regression found when trying to enable this combine
for RISCV in D113937.
Differential Revision: https://reviews.llvm.org/D113948
This was noted as a follow-up to D113212 / D113426:
4fc1fc40057e30404c3b11522cfcadhttps://alive2.llvm.org/ce/z/e4o96b
The canonicalization rules for these IR patterns are complicated,
and we were not matching the expected forms in 2 out of the 3
cases. We can make codegen more robust by matching the swapped
forms (and that will also work if these patterns are created late).
(Cond0 s> -1) ? N1 : 0 --> ~(Cond0 s>> BW-1) & N1
https://alive2.llvm.org/ce/z/mGCBrd
This was suggested as a potential enhancement in D113212 (also 7e30404c3b ).
There's an improvement for AArch that could be generalized ( X > -1 --> X >= 0 ).
For x86, we have a counter-acting fold for most cases that turns the shift+not
back into a setcc, so that needs a work-around to get more cases to use "pandn":
D113603
Note that this pattern (and a previous one) are not currently canonical forms
in IR:
https://alive2.llvm.org/ce/z/e4o96b
Adding swapped variants is left as a TODO item here, but is planned as
a near-term follow-up patch.
Differential Revision: https://reviews.llvm.org/D113426
As suggested on D113371, this adds a wrapper to SelectionDAG::ComputeNumSignBits, similar to the llvm::ComputeMinSignedBits wrapper.
I've included some usage, its not exhaustive, just the more obvious cases where the intention is obvious.
Differential Revision: https://reviews.llvm.org/D113396
We have several places where we need to extract the raw bits data from a BUILD_VECTOR node, so consolidate this to a single helper function that handles Undefs and Integer/FP constants, including implicit truncation.
This should make it easier to extend D113202 to handle more constant folding of bitcasted constant data.
Differential Revision: https://reviews.llvm.org/D113351
Currently FoldConstantArithmetic only handles binops, so replacing other uses of FoldConstantVectorArithmetic (in particular for SETCC nodes), still require more work.
This is the 'or' sibling for the fold added with:
D113212
https://alive2.llvm.org/ce/z/tgnp7K
Note that neither of these transforms is poison-safe,
but it does not seem to matter at this level. We have
had the scalar version of D113212 for a long time, so
this is just making optimizer behavior consistent.
We do not have the scalar version of *this* fold,
however, so that is another follow-up.
Another minor step towards merging FoldConstantVectorArithmetic into FoldConstantArithmetic.
We don't use SDNodeFlags in any constant folding inside DAG, so passing the Flags argument is a waste of time - an alternative would be to wire up FoldConstantArithmetic to take SDNodeFlags just-in-case we someday start using it, but we don't have any way to test it and I'd prefer to avoid dead code.
Differential Revision: https://reviews.llvm.org/D113276
(X s< 0) ? Y : 0 --> (X s>> BW-1) & Y
We canonicalize to the icmp+select form in IR, and we already have this fold
for scalar select in SDAG, so I think it's an oversight that we don't have
the fold for vectors. It seems neutral for AArch64 and saves some instructions
on x86.
Whether we should also have the sibling folds for the inverse condition or
all-ones true value may depend on target-specific factors such as whether
there's an "and-not" instruction.
Differential Revision: https://reviews.llvm.org/D113212
Fold (srl (mul (zext i32:$a to i64), i64:c), 32) -> (mulhu $a, $b),
if c can truncate to i32 without loss.
Reviewed By: frasercrmck, craig.topper, RKSimon
Differential Revision: https://reviews.llvm.org/D108129
Rely on the hasOperation() instead - as commented on D77804, the mid-term intention is to recognise rotate/funnel-by-constant pre-legalization to help avoid SimplifyDemandedBits regressions.
(i8 X ^ 128) & (i8 X s>> 7) --> usubsat X, 128
As suggested in D112085, we can substitute 'xor' with 'add'
in this pattern, and it is logically equivalent:
https://alive2.llvm.org/ce/z/eJtWWC
We canonicalize to 'xor' in IR, but SDAG does not do that
(and it probably should not - https://llvm.org/PR52267 ), so
it is possible to see either pattern in codegen. Note that
'sub' is a another potential pattern, but that is
canonicalized to 'add' in DAGCombiner, so we don't need to
worry about that variation.
Differential Revision: https://reviews.llvm.org/D112377
EXTRACT_SUBVECTOR indices are always constant, we don't need to check for ConstantSDNode, we should just use getConstantOperandVal which will assert for the constant.
Instead of returning a bool to indicate success and a separate
SDValue, return the SDValue and have the callers check if it is
null.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D112331
(i8 X ^ 128) & (i8 X s>> 7) --> usubsat X, 128
I haven't found a generalization of this identity:
https://alive2.llvm.org/ce/z/_sriEQ
Note: I was actually looking at the first form of the pattern in that link,
but that's part of a long chain of potential missed transforms in codegen
and IR....that I hope ends here!
The predicates for when this is profitable are a bit tricky. This version of
the patch excludes multi-use but includes custom lowering (as opposed to
legal only).
On x86 for example, we have custom lowering for some vector types, and that
uses umax and sub. So to enable that fold, we need add use checks to avoid
regressions. Even with legal-only lowering, we could see code with extra
reg move instructions for extra uses, so that constraint would have to be
eased very carefully to avoid penalties.
Differential Revision: https://reviews.llvm.org/D112085
With unoptimized code, we may see lots of stores and spend too much time in mergeTruncStores.
Fixes PR51827.
Reviewed By: spatel
Differential Revision: https://reviews.llvm.org/D111596
Inspired by D111968, provide a isNegatedPowerOf2() wrapper instead of obfuscating code with (-Value).isPowerOf2() patterns, which I'm sure are likely avenues for typos.....
Differential Revision: https://reviews.llvm.org/D111998
This is NFC-intended for the callers. Posting in case there are
other potential users that I missed.
I would also use this from VectorCombine in a patch for:
https://llvm.org/PR52178 ( D111901 )
Differential Revision: https://reviews.llvm.org/D111891
We were previously silently generating incorrect code when extracting a
fixed-width vector from a scalable vector. This is worse than crashing,
since the user will have no indication that this is currently unsupported
behaviour. I have fixed the code to only perform DAG combines when safe
to do so, i.e. the input and output vectors are both fixed-width or
both scalable.
Test added here:
CodeGen/AArch64/sve-extract-scalable-vector.ll
Differential revision: https://reviews.llvm.org/D110624
Stop using APInt constructors and methods that were soft-deprecated in
D109483. This fixes all the uses I found in llvm, except for the APInt
unit tests which should still test the deprecated methods.
Differential Revision: https://reviews.llvm.org/D110807
One of the cases identified in PR45116 - we don't need to limit extracted loads to ABI alignment, we can use allowsMemoryAccess - which tests using getABITypeAlign, but also checks if a target permits (fast) misaligned memory loads by checking allowsMisalignedMemoryAccesses as a fallback.
I've also cleaned up the alignment calculation code - if we have a constant extraction index then the alignment can be based on an offset from the original vector load alignment, but for non-constant indices we should assume the worst (single element alignment only).
Differential Revision: https://reviews.llvm.org/D110486
This patch adds a generic DAGCombine for vector-predicated (VP) nodes.
Those for which we can determine that no vector element is active can be
replaced by either undef or, for reductions, the start value.
This is tested rather trivially at the IR level, where it's possible
that we want to teach instcombine to perform this optimization.
However, we can also see the zero-evl case arise during SelectionDAG
legalization, when wide VP operations can be split into two and the
upper operation emerges as trivially false.
It's possible that we could perform this optimization "proactively"
(both on legal vectors and before splitting) and reduce the width of an
operation and insert it into a larger undef vector:
```
v8i32 vp_add x, y, mask, 4
->
v8i32 insert_subvector (v8i32 undef), (v4i32 vp_add xsub, ysub, mask, 4), i32 0
```
This is somewhat analogous to similar vector narrow/widening
optimizations, but it's unclear at this point whether that's beneficial
to do this for VP ops for any/all targets.
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D109148
One of the cases identified in PR45116 - we don't need to limit store narrowing to ABI alignment, we can use allowsMemoryAccess - which tests using getABITypeAlign, but also checks if a target permits (fast) misaligned memory access by checking allowsMisalignedMemoryAccesses as a fallback.
The fmul is a canonicalizing operation, and fneg is not so this would
break denormals that need flushing and also would not quiet signaling
nans. Fold to fsub instead, which is also canonicalizing.
This extends the custom lowering for extending loads on
fixed length vectors in SVE to support masked extending loads.
The existing tests for correct behaviour of masked extending loads
exhibit bad code generation due to the legalistaion of i1 vectors.
They have been left as-is and new tests have been added that do not
exhibit this behaviour.
Differential Revision: https://reviews.llvm.org/D108200
Soft deprecrate isNullValue/isAllOnesValue and update in tree
callers. This matches the changes to the APInt interface from
D109483.
Reviewed By: lattner
Differential Revision: https://reviews.llvm.org/D109535
This renames the primary methods for creating a zero value to `getZero`
instead of `getNullValue` and renames predicates like `isAllOnesValue`
to simply `isAllOnes`. This achieves two things:
1) This starts standardizing predicates across the LLVM codebase,
following (in this case) ConstantInt. The word "Value" doesn't
convey anything of merit, and is missing in some of the other things.
2) Calling an integer "null" doesn't make any sense. The original sin
here is mine and I've regretted it for years. This moves us to calling
it "zero" instead, which is correct!
APInt is widely used and I don't think anyone is keen to take massive source
breakage on anything so core, at least not all in one go. As such, this
doesn't actually delete any entrypoints, it "soft deprecates" them with a
comment.
Included in this patch are changes to a bunch of the codebase, but there are
more. We should normalize SelectionDAG and other APIs as well, which would
make the API change more mechanical.
Differential Revision: https://reviews.llvm.org/D109483
Given a select_cc producing a constant and a invertion of the constant
for a comparison more than zero, we can produce an xor with ashr
instead, which produces smaller code. The ashr either sets all bits or
clear all bits depending on if the value is negative. This is then xor'd
with the constant to optionally negate the value.
https://alive2.llvm.org/ce/z/DTFaBZ
This includes a OneUseCheck on the Cmp, which seems to make thinks a
little worse and will be removed in a followup.
Differential Revision: https://reviews.llvm.org/D109149
The check for whether a rotate is possible occurs before the
memory legality checks for the integer type. So it's possible we
decide we can use a rotate, but then fail the legality checks. If
that happens we should not fall back to a vector type. This triggers
an assertion in the rotate handling when it finds a vector type
instead of an integer type.
In theory we could use a shufflevector in place of the rotate, but
right now I'd just like to fix the crash.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D108839
Without this change only the preferred fusion opcode is tested
when attempting to combine FMA operations.
If both FMA and FMAD are available then FMA ops formed prior to
legalization will not be merged post legalization as FMAD becomes
the preferred fusion opcode.
Reviewed By: foad
Differential Revision: https://reviews.llvm.org/D108619
This is another bug exposed by https://llvm.org/PR51612
(and the one that triggered the initial assertion) in the report.
That example was suppressed with:
985b48f183
...but these would still crash because we created nodes
like UADDO without the expected 2 output values.
There are 2 bugs here:
1. We were not checking uses of operand 2 (the false value of the select).
2. We were not checking for multiple uses of nodes that produce >1 result.
Correcting those is enough to avoid the crash in the reduced test based on:
https://llvm.org/PR51612
The additional use check on operand 0 (the condition value of the select)
should not strictly be necessary because we are only replacing one use
with another (whether it makes performance sense to do the transform with
that pattern is not clear). But as noted in the TODO, changing that
uncovers another bug.
Note: there's at least one more bug here - we aren't propagating EVTs
correctly, but I plan to fix that in another patch.
For ISD::EXTRACT_SUBVECTOR, its second operand must be a constant
multiple of the known-minimum vector length of the result type.
Reviewed By: dmgreen
Differential Revision: https://reviews.llvm.org/D107795
One of the cases identified in PR45116 - we don't need to limit load combines to ABI alignment, we can use allowsMemoryAccess - which tests using getABITypeAlign, but also checks if a target permits (fast) misaligned memory loads by checking allowsMisalignedMemoryAccesses as a fallback.
One of the cases identified in PR45116 - we don't need to limit load combines (in this case for fp->int load/store copies) to ABI alignment, we can use allowsMemoryAccess - which tests using getABITypeAlign, but also checks if a target permits (fast) misaligned memory loads by checking allowsMisalignedMemoryAccesses as a fallback.
Differential Revision: https://reviews.llvm.org/D108318
One of the cases identified in PR45116 - we don't need to limit load combines (in this case for ISD::BUILD_PAIR) to ABI alignment, we can use allowsMemoryAccess - which tests using getABITypeAlign, but also checks if a target permits (fast) misaligned memory loads by checking allowsMisalignedMemoryAccesses as a fallback.
This helps in particular for 32-bit X86 cases loading 64-bit size data, reducing codegen diffs vs x86_64.
Differential Revision: https://reviews.llvm.org/D108307
Follow-up to D107068, attempt to fold nested concat_vectors/undefs, as long as both the vector and inner subvector types are legal.
This exposed the same issue in ARM's MVE LowerCONCAT_VECTORS_i1 (raised as PR51365) and AArch64's performConcatVectorsCombine which both assumed concat_vectors only took 2 subvector operands.
Differential Revision: https://reviews.llvm.org/D107597
visitEXTRACT_SUBVECTOR can sometimes create illegal BITCASTs when
removing "redundant" INSERT_SUBVECTOR operations. This patch adds
an extra check to ensure such combines only occur after operation
legalisation if any resulting BITBAST is itself legal.
Differential Revision: https://reviews.llvm.org/D108086
We don't have real demanded bits support for MULHU, but we can
still use the known bits based constant folding support at the end
of SimplifyDemandedBits to simplify a MULHU. This helps with cases
where we know the LHS and RHS have enough leading zeros so that
the high multiply result is always 0.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D106471
IR typically creates INSERT_SUBVECTOR patterns as a widening of the subvector with undefs to pad to the destination size, followed by a shuffle for the actual insertion - SelectionDAGBuilder has to do something similar for shuffles when source/destination vectors are different sizes.
This combine attempts to recognize these patterns by looking for a shuffle of a subvector (from a CONCAT_VECTORS) that starts at a modulo of its size into an otherwise identity shuffle of the base vector.
This uncovered a couple of target-specific issues as we haven't often created INSERT_SUBVECTOR nodes in generic code - aarch64 could only handle insertions into the bottom of undefs (i.e. a vector widening), and x86-avx512 vXi1 insertion wasn't keeping track of undef elements in the base vector.
Fixes PR50053
Differential Revision: https://reviews.llvm.org/D107068
This allows special constants like to 0 to be recognized. It's also
expected by isel patterns if a target had a mulh with immediate instructions.
The commuting done by tablegen won't commute patterns with immediates since it
expects DAGCombine to have done it.
Reviewed By: foad
Differential Revision: https://reviews.llvm.org/D107486
We had some similar hasOneUse/isNON_EXTLoad early-outs spread out over different parts of the method - we should pull them all together.
Noticed while triaging PR45116
This transform was added with D58874, but there were no tests for overflow ops.
We need to change this one way or another because it can crash as shown in:
https://llvm.org/PR51238
Note that if there are no uses of an overflow op's bool overflow result, we
reduce it to a regular math op, so we continue to fold that case either way.
If we have uses of both the math and the overflow bool, then we are likely
not saving anything by creating an independent sub instruction as seen in
the test diffs here.
This patch makes the behavior in SDAG consistent with what we do in
instcombine AFAICT.
Differential Revision: https://reviews.llvm.org/D106983
This patch adds a peephole optimization `SETCC(FREEZE(x),const)` => `FREEZE(SETCC(x,const))`
if the SETCC is only used by BRCOND.
Combined with `BRCOND(FREEZE(X)) => BRCOND(X)`, this leads to a nice improvement in the generated assembly when x is a masked loaded value.
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D105344
This patch builds on top of D106575 in which scalable-vector splats were
supported in `ISD::matchBinaryPredicate`. It teaches the DAGCombiner how
to perform a variety of the pre-existing saturating add/sub combines on
scalable-vector types.
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D106652
This patch extends support for (scalable-vector) splats in the
DAGCombiner via the `ISD::matchBinaryPredicate` function, which enable a
variety of simple combines of constants.
Users of this function may now have to distinguish between
`BUILD_VECTOR` and `SPLAT_VECTOR` vector operands. The way of dealing
with this in-tree follows the approach added for
`ISD::matchUnaryPredicate` implemented in D94501.
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D106575
I've setup the basic framework for the isGuaranteedNotToBeUndefOrPoison call and updated DAGCombiner::visitFREEZE to use it, further Opcodes can be handled when we have test coverage.
I'm not aware of any vector test freeze coverage so the DemandedElts (and the Depth) args are not being used yet - but they are in place.
SelectionDAG::isGuaranteedNotToBePoison wrappers have also been added.
Differential Revision: https://reviews.llvm.org/D106668
This adds custom lowering for truncating stores when operating on
fixed length vectors in SVE. It also includes a DAG combine to
fold extends followed by truncating stores into non-truncating
stores in order to prevent this pattern appearing once truncating
stores are supported.
Currently truncating stores are not used in certain cases where
the size of the vector is larger than the target vector width.
Differential Revision: https://reviews.llvm.org/D104471
Reland of 31859f896.
This change implements new DAG notes GLOBAL_GET/GLOBAL_SET, and
lowering methods for load and stores of reference types from IR
globals. Once the lowering creates the new nodes, tablegen pattern
matches those and converts them to Wasm global.get/set.
Reviewed By: tlively
Differential Revision: https://reviews.llvm.org/D104797
The existing rule about the operand type is strange. Instead, just say
the operand is a TargetConstant with the right width. (Legalization
ignores TargetConstants, so it doesn't matter if that width is legal.)
Highlights:
1. I had to substantially rewrite the AArch64 isel patterns to expect a
TargetConstant. Nothing too exotic, but maybe a little hairy. Maybe
worth considering a target-specific node with some dagcombines instead
of this complicated nest of isel patterns.
2. Our behavior on RV32 for vectors of i64 has changed slightly. In
particular, we correctly preserve the width of the arithmetic through
legalization. This changes the DAG a bit. Maybe room for
improvement here.
3. I explicitly defined the behavior around overflow. This is necessary
to make the DAGCombine transforms legal, and I don't think it causes any
practical issues.
Differential Revision: https://reviews.llvm.org/D105673
I'm going to extend the functionality started in D106058 so move the folds into their own method to reduce the amount of code in DAGCombiner::visitSELECT
Similar to the folds performed in InstCombinerImpl::foldSelectOpOp, this attempts to push a select further up to help merge a pair of binops.
I'm primarily interested in select(cond,add(x,y),add(x,z)) folds to help expose pointer math (see https://bugs.llvm.org/show_bug.cgi?id=51069 etc.) but I've tried to use the more generic isBinOp().
Differential Revision: https://reviews.llvm.org/D106058
This adds custom lowering for truncating stores when operating on
fixed length vectors in SVE. It also includes a DAG combine to
fold extends followed by truncating stores into non-truncating
stores in order to prevent this pattern appearing once truncating
stores are supported.
Currently truncating stores are not used in certain cases where
the size of the vector is larger than the target vector width.
Differential Revision: https://reviews.llvm.org/D104471
We already have reassociation code for Adds and Ors separately in DAG
combiner, this adds it for the combination of the two where Ors act like
Adds. It reassociates (add (or (x, c), y) -> (add (add (x, y), c)) where
we know that the Ors operands have no common bits set, and the Or has
one use.
Differential Revision: https://reviews.llvm.org/D104765
Reland of 31859f896.
This change implements new DAG notes GLOBAL_GET/GLOBAL_SET, and
lowering methods for load and stores of reference types from IR
globals. Once the lowering creates the new nodes, tablegen pattern
matches those and converts them to Wasm global.get/set.
Differential Revision: https://reviews.llvm.org/D104797
This ports the AArch64 SABD and USBD over to DAG Combine, where they can be
used by more backends (notably MVE in a follow-up patch). The matching code
has changed very little, just to handle legal operations and types
differently. It selects from (ABS (SUB (EXTEND a), (EXTEND b))), producing
a ubds/abdu which is zexted to the original type.
Differential Revision: https://reviews.llvm.org/D91937
This add as a fold of sub(0, splat(sub(0, x))) -> splat(x). This can
come up in the lowering of right shifts under AArch64, where we generate
a shift left of a negated number.
Differential Revision: https://reviews.llvm.org/D103755
The is from discussion in https://reviews.llvm.org/D104247#inline-993387
The contract and reassoc flags shouldn't imply each other .
All the aggressive fsub fusion reassociate operations,
we should guard them with reassoc flag check.
Reviewed By: mcberg2017
Differential Revision: https://reviews.llvm.org/D104723
According to IR LangRef, the FMF flag:
contract
Allow floating-point contraction (e.g. fusing a multiply followed by an
addition into a fused multiply-and-add).
reassoc
Allow reassociation transformations for floating-point instructions.
This may dramatically change results in floating-point.
My understanding is that these two flags shouldn't imply each other,
as we might have a SDNode that can be reassociated with others, but
not contractble.
eg: We may want following fmul/fad/fsub to freely reassoc, but don't
want fma being generated here.
%F = fmul reassoc double %A, %B ; <double> [#uses=1]
%G = fmul reassoc double %C, %D ; <double> [#uses=1]
%H = fadd reassoc double %F, %G ; <double> [#uses=1]
%I = fsub reassoc double %H, %E ; <double> [#uses=1]
Before https://reviews.llvm.org/D45710, `reassoc` flag actually
did not imply isContratable either.
The current implementation also only check the flag in fadd node,
ignoring fmul node, this patch update that as well.
Reviewed By: spatel, qiucf
Differential Revision: https://reviews.llvm.org/D104247
6e5628354e regressed the Windows build as
the return type no longer matched in both branches for the return value
type deduction. This uses a bit more compiler magic to deal with that.
The sorting, obviously, must be stable, else we will have random assembly fluctuations.
Apparently there was no test coverage that would benefit from that,
so i've added one test.
The sorting consists of two parts - just sort the input vectors,
and recompute the shuffle mask -> input vector mapping.
I don't believe we need to do anything else.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D104187
When reducing vector builds to shuffles it possible that
the DAG combiner may try to extract invalid subvectors.
This happens as the existing code assumes vectors will be power
of 2 sizes, which is already untrue, but becomes more noticable
with v6 and v7 types.
Specifically the existing code assumes that half PowerOf2Ceil of
a given vector index will fit twice into a given vector.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D103880
This change implements new DAG notes GLOBAL_GET/GLOBAL_SET, and
lowering methods for load and stores of reference types from IR
globals. Once the lowering creates the new nodes, tablegen pattern
matches those and converts them to Wasm global.get/set.
Reviewed By: tlively
Differential Revision: https://reviews.llvm.org/D95425
As shown in:
https://llvm.org/PR50623
...and the similar tests here, we were not accounting for
store merging of different sizes that do not cover the
entire range of the wide value to be stored.
This is the easy fix: just make sure that all of the
original stores are the same size, so when we calculate
the wide width, it's a simple N * M check.
This still allows all of the motivating optimizations from:
D86420 / 54a5dd485c
D87112 / 7a06b166b1
We could enhance this code to track individual bytes and
allow merging multiple sizes.
shuffle(concat(x,undef),concat(y,undef)) -> concat(shuffle(x,y),shuffle(x,y))
If the original shuffle references any of the upper (undef) subvector elements, ensure the split shuffle masks uses undef instead of an out-of-bounds value.
Fixes PR50609
This extends 434c8e013a and ede3982792 to handle signed
predicates by sign-extending the setcc operands.
This is not shown directly in https://llvm.org/PR50055 ,
but the pattern is visible by changing the unsigned convert
to signed in the source code.
This is a follow-up to D103280 that eases the use restrictions,
so we can handle the motivating case from:
https://llvm.org/PR50055
The loop code is adapted from similar use checks in
ExtendUsesToFormExtLoad() and SliceUpLoad(). I did not see an
easier way to filter out non-chain uses of load values.
Differential Revision: https://reviews.llvm.org/D103462
I accidentaly pushed a draft of D103280 that was discussed
during the review, but it was not supposed to be the final
version.
Rather than revert and recommit, I'm updating the existing
code. This way we have a record of the codegen diff that
would result if we decide to remove this predicate in the
future.
sext (vsetcc X, Y) --> vsetcc (zext X), (zext Y) --
(when the zexts are free and a bunch of other conditions)
We have a couple of similar folds to this already for vector selects,
but this pattern slips through because it is only a setcc.
The tests are based on the motivating case from:
https://llvm.org/PR50055
...but we need extra logic to get that example, so I've left that as
a TODO for now.
Differential Revision: https://reviews.llvm.org/D103280
extractelement is poison if the index is out-of-bounds, so just
scalarizing the load may introduce an out-of-bounds load, which is UB.
To avoid introducing new UB, we can mask the index so it only contains
valid indices.
Fixes PR50382.
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D103077
DAGCombine's `mergeStoresOfConstantsOrVecElts` optimization is told
whether it's to use vector types and also whether it's to issue a
truncating store. However, the truncating store code path assumes a
scalar integer `ConstantSDNode`, and when using vector types it creates
either a `BUILD_VECTOR` or `CONCAT_VECTORS` to store: neither of which
is a constant.
The `riscv64` target is able to expose a crash here because it switches
on both code paths at the same time. The `f32` is stored as `i32` which
must be promoted to `i64`, necessitating a truncating store.
It also decides later that it prefers a vector store of `v2f32`.
While vector truncating stores are legal, this combine is not able to
emit them. We also don't have a test case. This patch adds an assert to
catch this case more gracefully, and updates one of the caller functions
to the function to turn off the use of truncating stores when preferring
vectors.
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D103173
The select-of-constants transform was asserting that its constant vector
inputs did not implicitly truncate their input without that as an
explicit precondition to the function. This patch relaxes that assertion
into an early return to skip the optimization.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D102393
Fixes a bug in the DAG combiner that eliminates the stores because it missed
to inspect the address space of the pointers.
%v = load %ptr_as1
// no chain side effect
store %v, %ptr_as2
As well as
store %v, %ptr_as1
store %v, %ptr_as2
Fixes a test for above in X86.
Differential Revision: https://reviews.llvm.org/D102096
Expanding a fixed length operation involves wrapping the operation in an
insert/extract subvector pair, as such, when this is done to bitcast we
end up with an extract_subvector of a bitcast. DAGCombine tries to
convert this into a bitcast of an extract_subvector which restores the
initial fixed length bitcast, causing an infinite loop of legalization.
As part of this patch, we must make sure the above DAGCombine does not
trigger after legalization if the created bitcast would not be legal.
Differential Revision: https://reviews.llvm.org/D101990
This replaces D98479.
This allows type legalization to form SPLAT_VECTOR_PARTS so we don't
lose the splattedness when the scalar type is split.
I'm handling SPLAT_VECTOR_PARTS for fixed vectors separately so
we can continue using non-VL nodes for scalable vectors.
I limited to RV32+vXi64 because DAGCombiner::visitBUILD_VECTOR likes
to form SPLAT_VECTOR before seeing if it can replace the BUILD_VECTOR
with other operations. Especially interesting is a splat BUILD_VECTOR of
the extract_vector_elt which can become a splat shuffle, but won't if
we form SPLAT_VECTOR first. We either need to reorder visitBUILD_VECTOR
or add visitSPLAT_VECTOR.
Reviewed By: frasercrmck
Differential Revision: https://reviews.llvm.org/D100803
It is proper to relax non-negative limitation of step_vector.
Also this patch adds more combines for step_vector:
(sub X, step_vector(C)) -> (add X, step_vector(-C))
Differential Revision: https://reviews.llvm.org/D100812
This patch adds incrementally-better support for SPLAT_VECTOR in a
handful of vector combines by changing a few more
isBuildVectorAllOnes/isBuildVectorAllZeros to the equivalent
isConstantSplatVectorAllOnes/Zeros calls.
Reviewed By: paulwalker-arm
Differential Revision: https://reviews.llvm.org/D100851
This patch changes ISD::isBuildVectorAllZeros to
ISD::isConstantSplatVectorAllZeros which handles zero sclar vector.
TestPlan: check-llvm
Differential Revision: https://reviews.llvm.org/D100813
When trying to clamp a constant index into a scalable vector we can
test if the index is less than the minimum number of elements in the
vector. If so, we can simply return the index because we know it is
guaranteed to fit inside the vector.
Differential Revision: https://reviews.llvm.org/D100639
Main reason is preparation to transform AliasResult to class that contains
offset for PartialAlias case.
Reviewed By: asbirlea
Differential Revision: https://reviews.llvm.org/D98027
If the inner shuffle already contains undef elements, then accept them in the merged shuffle as well.
This helps some X86 HADD/SUB patterns where slow targets were ending up with HADD/SUB because the (un)merged shuffles were stuck either side of the ADD/SUB - meaning we ended up with a total cost much higher than the "2*shuffle+add" that a slow target usually expands a HADD/SUB to.
This patch adds a new isIntOrFPConstant helper function to check if a
SDValue is a integer of FP constant. This pattern is used in various
places.
There also are places that incorrectly just check for integer constants,
e.g. D99384, so hopefully this helper will help people avoid that issue.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D99428
Don't bother calling ComputeNumSignBits if N00Bits < ExtVTBits. No
matter what answer we get back this will be true:
(N00Bits - DAG.ComputeNumSignBits(N00, DemandedSrcElts)) < ExtVTBits)
So we might as well save the computation. This makes the code more
consistent with the similar (sext_in_reg (sext x)) handling above.
As commented by @craig.topper on rG1ba5c550d418, we can't guarantee that we'll be extending zero bits, just sign bit. So, revert to the old code for zero_extend_vector_inreg cases.
Followup to D96345, handle unary shuffles of binops (as well as binary shuffles) if we can merge the shuffle with inner operand shuffles.
Differential Revision: https://reviews.llvm.org/D98646
Extend this to support ComputeNumSignBits of the (used) source vector elements so that we can handle more than just the case where we're sext_in_reg from the source element signbit.
Noticed while investigating the poor codegen in D98587.
A 1-bit smulo overflows is both inputs are -1 since the result
should be +1 which can't be represented in a signed 1 bit value.
We can detect this with an AND and a setcc. The multiply result
can also use the same AND.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D97634
This prepares codegen for a change that will remove the identical
folds from IR because they are not poison-safe. See
D93065 / D97360
for details.
We already generically support scalar types, and there are various
target-specific transforms that overlap the vector folds. For example,
x86 recognizes the and patterns, but not or. We can end up with 1
extra instruction there, but I think that is still preferred over the
blendv alternative that loads a constant vector.
If this is not optimal, then it should be fixed with a later transform
(this change is not expected to result in any regressions because
InstCombine currently does the same thing).
Removing custom code and supporting undefs in constant-pattern-matching
can be follow-up changes.
Differential Revision: https://reviews.llvm.org/D97730
Peeking through AND is only valid if the input to both shifts is
the same. If the inputs are different, then the original pattern
ORs the two values when the masked shift amount is 0. This is ok
if the values are the same since the OR would be a NOP which is
why its ok for rotate.
Fixes PR49365 and reverts PR34641
Differential Revision: https://reviews.llvm.org/D97637
Even if the first computeKnownBits call doesn't have any zero
bits it is possible the other operand has bitwidth-1 leading zero.
In that case overflow is still impossible. So always call computeKnownBits
for both operands.
Using ComputeNumSignBits or computeKnownBits we might be able
to determine that overflow is impossible.
This especially helps after type legalization if the type was
promoted from a type with half the bits or more. Type legalization
conservatively creates a promoted smulo/umulo and an overflow
check for the promoted bits. The overflow from the promoted
smulo/umulo is ORed with the result of the promoted bits
overflow check. Proving that the promoted smulo/umulo can never
overflow will leave us with just the promoted bits overflow check.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D97160
This patch handles usubsat patterns hidden through zext/trunc and uses the getTruncatedUSUBSAT helper to determine if the USUBSAT can be correctly performed in the truncated form:
zext(x) >= y ? x - trunc(y) : 0 --> usubsat(x,trunc(umin(y,SatLimit)))
zext(x) > y ? x - trunc(y) : 0 --> usubsat(x,trunc(umin(y,SatLimit)))
Based on original examples:
void foo(unsigned short *p, int max, int n) {
int i;
unsigned m;
for (i = 0; i < n; i++) {
m = *--p;
*p = (unsigned short)(m >= max ? m-max : 0);
}
}
Differential Revision: https://reviews.llvm.org/D25987
If extload is legal, following transform
(zext (select c, load1, load2)) -> (select c, zextload1, zextload2)
can save one ext instruction.
Differential Revision: https://reviews.llvm.org/D95086
Adjust generateFMAsInMachineCombiner to return false if SVE is present
in order to combine fmul+fadd into fma. Also add new pseudo instructions
so as to select the most appropriate of FMLA/FMAD depending on register
allocation.
Depends on D96599
Differential Revision: https://reviews.llvm.org/D96424
Fold shuffle(bop(shuffle(x,y),shuffle(z,w)),bop(shuffle(a,b),shuffle(c,d))) -> bop(shuffle(x,y),shuffle(z,w)),bop(shuffle(a,b),shuffle(c,d))
Attempt to fold from a shuffle of a pair of binops to a binop of shuffles, as long as one/both of the binop sources are also shuffles that can be merged with the outer shuffle. This should guarantee that we remove one binop without introducing any additional shuffles.
Technically there's potential for a merged shuffle's lowering to be poorer than the original shuffle, but it could also be better, and I'm not seeing any regressions as long as we keep the 'don't merge splats' rule already present in MergeInnerShuffle.
This expands and generalizes an existing X86 combine and attempts to merge either of each binop's sources (with an on-the-fly commutation of the shuffle mask) - we couldn't do that in the x86 version as it had to stay in a form that DAGCombine's MergeInnerShuffle would still recognise.
Fixes issue raised by @saugustine in rG5aa8f4c0843a where we were failing to replace null shuffle operands from MergeInnerShuffle to UNDEFs.
Differential Revision: https://reviews.llvm.org/D96345
This reverts commit 5dfba562dd.
That commit causes an assertion failure with the following repro:
typedef long b __attribute__((__vector_size__(16)));
b *d;
b e;
b __attribute__((__always_inline__)) c(b h, b i) {
return (__attribute__((__vector_size__(8 * sizeof(short)))) short)h + i;
}
j() {
b k, l, m, n, o[6], p, q;
m = d[5];
b r = m;
b s = f(r, 8);
q = s;
l = d[1];
p = l;
t(q);
n = c(m, l);
o[1] = c(s, f(p, 8));
k = __builtin_shufflevector(n, o[1], 0, 2);
e = __builtin_ia32_psrlwi128(k, j);
}
./bin/clang -cc1 -triple x86_64-grtev4-linux-gnu -emit-obj -O1 -std=c99 test.c
Fold shuffle(bop(shuffle(x,y),shuffle(z,w)),bop(shuffle(a,b),shuffle(c,d))) -> bop(shuffle(x,y),shuffle(z,w)),bop(shuffle(a,b),shuffle(c,d))
Attempt to fold from a shuffle of a pair of binops to a binop of shuffles, as long as one/both of the binop sources are also shuffles that can be merged with the outer shuffle. This should guarantee that we remove one binop without introducing any additional shuffles.
Technically there's potential for a merged shuffle's lowering to be poorer than the original shuffle, but it could also be better, and I'm not seeing any regressions as long as we keep the 'don't merge splats' rule already present in MergeInnerShuffle.
This expands and generalizes an existing X86 combine and attempts to merge either of each binop's sources (with an on-the-fly commutation of the shuffle mask) - we couldn't do that in the x86 version as it had to stay in a form that DAGCombine's MergeInnerShuffle would still recognise.
Differential Revision: https://reviews.llvm.org/D96345
Begin transitioning the X86 vector code to recognise sub(umax(a,b) ,b) or sub(a,umin(a,b)) USUBSAT patterns to make it more generic and available to all targets.
This initial patch just moves the basic umin/umax patterns to DAG, removing some vector-only checks on the way - these are some of the patterns that the legalizer will try to expand back to so we can be reasonably relaxed about matching these pre-legalization.
We can handle the trunc(sub(..))) variants as well, which helps with patterns where we were promoting to a wider type to detect overflow/saturation.
The remaining x86 code requires some cleanup first - some of it isn't actually tested etc. I also need to resurrect D25987.
Differential Revision: https://reviews.llvm.org/D96413
Avoid doing the following combine for vector types:
```
copysign(x, fp_extend(y)) -> copysign(x, y)
copysign(x, fp_round(y)) -> copysign(x, y)
```
That combine seemed to impede the selection of vector instruction and cause
a mess in some circumstances.
Differential Revision: https://reviews.llvm.org/D96037
As of commit 284f2bffc9, the DAG Combiner gets rid of the masking of the
input to this node if the mask only keeps the bottom 16 bits. This is because
the underlying library function does not use the high order bits. However, on
PowerPC's ELFv2 ABI, it is the caller that is responsible for clearing the bits
from the register. Therefore, the library implementation of __gnu_h2f_ieee will
return an incorrect result if the bits aren't cleared.
This combine is desired for ARM (and possibly other targets) so this patch adds
a query to Target Lowering to check if this zeroing needs to be kept.
Fixes: https://bugs.llvm.org/show_bug.cgi?id=49092
Differential revision: https://reviews.llvm.org/D96283
Make sure scalable property is preserved by using getVectorElementCount().
Reviewed By: paulwalker-arm
Differential Revision: https://reviews.llvm.org/D95967
If sext_inreg is supported, we will turn this into sext_inreg. That
will then remove it if there are enough sign bits. But if sext_inreg
isn't supported, we can still remove the shift pair based on sign
bits.
Split from D95890.
This patch adds support for scalable-vector splats in DAGCombiner's
`isConstantOrConstantVector` and `ISD::matchUnaryPredicate` functions,
which enable the SelectionDAG div/rem-by-constant optimizations for
scalable vector types.
It also fixes up one case where the UDIV optimization was generating a
SETCC without first consulting the target for its preferred SETCC result
type.
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D94501
For now, we correct the result for sqrt if iteration > 0. This doesn't make
sense as they are not strict relative.
Reviewed By: dmgreen, spatel, RKSimon
Differential Revision: https://reviews.llvm.org/D94480
Add DemandedElts support inside the TRUNCATE analysis.
REAPPLIED - this was reverted by @hans at rGa51226057fc3 due to an issue with vector shift amount types, which was fixed in rG935bacd3a724 and an additional test case added at rG0ca81b90d19d
Differential Revision: https://reviews.llvm.org/D56387
It caused "Vector shift amounts must be in the same as their first arg"
asserts in Chromium builds. See the code review for repro instructions.
> Add DemandedElts support inside the TRUNCATE analysis.
>
> Differential Revision: https://reviews.llvm.org/D56387
This reverts commit cad4275d69.
Use the KnownBits icmp comparisons to determine when a ISD::UMIN/UMAX op is unnecessary should either op be known to be ULT/ULE or UGT/UGE than the other.
Differential Revision: https://reviews.llvm.org/D94532
MergeInnerShuffle currently attempts to merge shuffle(shuffle(x,y),z) patterns into a single shuffle, using 1 or 2 of the x,y,z ops.
However if we already match 2 ops we might be able to handle the third op if its also a shuffle that references one of the previous ops, allowing us to handle some cases like:
shuffle(shuffle(x,y),shuffle(x,y))
shuffle(shuffle(shuffle(x,z),y),z)
shuffle(shuffle(x,shuffle(x,y)),z)
etc.
This isn't an exhaustive match and is dependent on the order the candidate ops are encountered - if one of the matched ops was a shuffle that was peek-able we don't go back and try to split that, I haven't found much need for that amount of analysis yet.
This is a preliminary patch that will allow us to later improve x86 HADD/HSUB matching - but needs to be reviewed separately as its in generic code and affects existing Thumb2 tests.
Differential Revision: https://reviews.llvm.org/D94671
I'm hoping to reuse MergeInnerShuffle in some other folds - so ensure the candidate ops/mask are reset at the start of each run.
Also, move the second op matching before bailing to make it simpler to try to match other things afterward.
This patch resolves the suboptimal codegen described in http://llvm.org/pr47873 .
When CodeGenPrepare lowers select into a conditional branch, a freeze instruction is inserted.
It is then translated to `BRCOND(FREEZE(SETCC))` in SelDag.
The `FREEZE` in the middle of `SETCC` and `BRCOND` was causing a suboptimal code generation however.
This patch adds `BRCOND(FREEZE(cond))` -> `BRCOND(cond)` fold to DAGCombiner to remove the `FREEZE`.
To make this optimization sound, `BRCOND(UNDEF)` simply should nondeterministically jump to the branch or not, rather than raising UB.
It wasn't clear what happens when the condition was undef according to the comments in ISDOpcodes.h, however.
I updated the comments of `BRCOND` to make it explicit (as well as `BR_CC`, which is also a conditional branch instruction).
Note that it diverges from the semantics of `br` instruction in IR, which is explicitly UB.
Since the UB semantics was necessary to explain optimizations that use branching conditions, and SelDag doesn't seem to have such optimization, I think this divergence is okay.
Reviewed By: spatel
Differential Revision: https://reviews.llvm.org/D92015
We are checking the unsafe-fp-math for sqrt but not for fpow, which behaves inconsistent.
As the direction is to remove this global option, we need to remove the unsafe-fp-math
check for sqrt and update the test with afn fast-math flags.
Reviewed By: Spatel
Differential Revision: https://reviews.llvm.org/D93891
This looks to have been done to save some duplicated code under
two different if statements, but it ends up being harmful to D94073.
This speculative constant can be called on a scalable vector type
with i64 element size when i64 scalars aren't legal. The code tries
and fails to find a vector type with i32 elements that it can use.
So only create the node when we know it will be used.
This patch disables the FSUB(-0,X)->FNEG(X) DAG combine when we're flushing subnormals. It requires updating the existing AMDGPU tests to use the fneg IR instruction, in place of the old fsub(-0,X) canonical form, since AMDGPU is the only backend currently checking the DenormalMode flags.
Note that this will require follow-up optimizations to make sure the FSUB(-0,X) form is handled appropriately
Differential Revision: https://reviews.llvm.org/D93243
Fixes a bug introduced by D91589.
When folding `(sext (not i1 x)) -> (add (zext i1 x), -1)`, we try to replace the not first when possible. If we replace the not in-visit, then the now invalidated node will be returned, and subsequently we will return an invalid sext. In cases where the not is replaced in-visit we can simply return SDValue, as the not in the current sext should have already been replaced.
Thanks @jgorbe, for finding the below reproducer.
The following reduced test case crashes clang when built with `clang -O1 -frounding-math`:
```
template <class> class a {
int b() { return c == 0.0 ? 0 : -1; }
int c;
};
template class a<long>;
```
A debug build of clang produces this "assertion failed" error:
```
clang: /home/jgorbe/code/llvm/llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp:264: void {anonymous}::DAGCombiner::AddToWorklist(llvm::
SDNode*): Assertion `N->getOpcode() != ISD::DELETED_NODE && "Deleted Node added to Worklist"' failed.
```
Reviewed By: spatel
Differential Revision: https://reviews.llvm.org/D93274
Clean up a TODO, to support folding a shift of a constant by a
select of constants, on targets with different shift operand sizes.
Reviewed By: RKSimon, lebedev.ri
Differential Revision: https://reviews.llvm.org/D90349
This patch adds the following DAGCombines, which apply if isVectorLoadExtDesirable() returns true:
- fold (and (masked_gather x)) -> (zext_masked_gather x)
- fold (sext_inreg (masked_gather x)) -> (sext_masked_gather x)
LowerMGATHER has also been updated to fetch the LoadExtType associated with the
gather and also use this value to determine the correct masked gather opcode to use.
Reviewed By: sdesmalen
Differential Revision: https://reviews.llvm.org/D92230
Adds the ExtensionType flag, which reflects the LoadExtType of a MaskedGatherSDNode.
Also updated SelectionDAGDumper::print_details so that details of the gather
load (is signed, is scaled & extension type) are printed.
Reviewed By: sdesmalen
Differential Revision: https://reviews.llvm.org/D91084
LLVM intrinsic llvm.maxnum|minnum is overloaded intrinsic, can be used on any
floating-point or vector of floating-point type.
This patch extends current infrastructure to support scalable vector type.
This patch also fix a warning message of incorrect use of EVT::getVectorNumElements()
for scalable type, when DAGCombiner trying to split scalable vector.
Reviewed By: sdesmalen
Differential Revision: https://reviews.llvm.org/D92607
The refineIndexType & refineUniformBase functions added by D90942 can also be used to
improve CodeGen of masked gathers.
These changes were split out from D91092
Reviewed By: sdesmalen
Differential Revision: https://reviews.llvm.org/D92319
In previous code, when refineIndexType(...) is called and Index is undef, Index.getOperand(0) will raise a assertion fail.
Reviewed By: pengfei
Differential Revision: https://reviews.llvm.org/D92548
Move fold of (sext (not i1 x)) -> (add (zext i1 x), -1) from X86 to DAGCombiner to improve codegen on other targets.
Differential Revision: https://reviews.llvm.org/D91589
1. Removed #include "...AliasAnalysis.h" in other headers and modules.
2. Cleaned up includes in AliasAnalysis.h.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D92489
Move the X86 VSELECT->UADDSAT fold to DAGCombiner - there's nothing target specific about these folds.
The SSE42 test diffs are relatively benign - its avoiding an extra constant load in exchange for an extra xor operation - there are extra register moves, which is annoying as all those operations should commute them away.
Differential Revision: https://reviews.llvm.org/D91876
For now, we will hardcode the result as 0.0 if the input is denormal or 0. That will
have the impact the precision. As the fsqrt added belong to the cold path of the
cmp+branch, it won't impact the performance for normal inputs for PowerPC, but improve
the precision if the input is denormal.
Reviewed By: Spatel
Differential Revision: https://reviews.llvm.org/D80974
PowerPC has instruction ftsqrt/xstsqrtdp etc to do the input test for software square root.
LLVM now tests it with smallest normalized value using abs + setcc. We should add hook to
target that has test instructions.
Reviewed By: Spatel, Chen Zheng, Qiu Chao Fang
Differential Revision: https://reviews.llvm.org/D80706
If the scatter store is able to perform the sign/zero extend of
its index, this is folded into the instruction with refineIndexType().
Additionally, refineUniformBase() will return the base pointer and index
from an add + splat_vector.
Reviewed By: sdesmalen
Differential Revision: https://reviews.llvm.org/D90942
Fold
VT = (and (sign_extend NarrowVT to VT) #bitmask)
into
VT = (zero_extend NarrowVT)
With this combine, the test replaces a sign extended load + an
unsigned extention with a zero extended load to render one of the
operands of the last multiplication.
BEFORE | AFTER
f_i16_i32: | f_i16_i32:
.fnstart | .fnstart
ldrsh r0, [r0] | ldrh r1, [r1]
ldrsh r1, [r1] | ldrsh r0, [r0]
smulbb r0, r1, r0 | smulbb r0, r0, r1
uxth r1, r1 | mul r0, r0, r1
mul r0, r0, r1 | bx lr
bx lr |
Reviewed By: resistor
Differential Revision: https://reviews.llvm.org/D90605
Summary:
For vector element types which are not byte-sized, we would generate
incorrect scalar offsets and produce incorrect codegen.
This optimization could potentially be supported in the future, e.g. by
loading in bytes, then shifting and masking out the remaining bits of
the vector element. However, without an upstream target to test against
it's best to avoid the bad codegen in the simplest possible way.
Related to this bug:
https://bugs.llvm.org/show_bug.cgi?id=27600
Reviewed by: foad
Differential Revision: https://reviews.llvm.org/D78568
This patch uses the existing LowerFixedLengthReductionToSVE function to also lower
scalable vector reductions. A separate function has been added to lower VECREDUCE_AND
& VECREDUCE_OR operations with predicate types using ptest.
Lowering scalable floating-point reductions will be addressed in a follow up patch,
for now these will hit the assertion added to expandVecReduce() in TargetLowering.
Reviewed By: paulwalker-arm
Differential Revision: https://reviews.llvm.org/D89382
The modified code in visitSTORE was missing a scalable vector check, and still
using the now deprecated implicit cast of TypeSize to uint64_t through the
overloaded operator. This patch fixes these issues.
This brings the logic in line with the comment on the context line immediately
above the added precondition.
Add a test in sve-redundant-store.ll that the warning is not triggered.
Differential Revision: https://reviews.llvm.org/D89701
The modified code in visitSTORE was missing a scalable vector check, and still
using the now deprecated implicit cast of TypeSize to uint64_t through the
overloaded operator. This patch fixes these issues.
This brings the logic in line with the comment on the context line immediately
above the added precondition.
Add a test in Redundantstores.ll that the warning is not triggered.
From LangRef, FMF contract should not enable reassociating to form
arbitrary contractions. So it should not help rearrange nodes like
(fma (fmul x, c1), c2, y) into (fma x, c1*c2, y).
Reviewed By: spatel
Differential Revision: https://reviews.llvm.org/D89527
MULH is often expanded on targets.
This patch removes the isMulhCheaperThanMulShift hook and uses
isOperationLegalOrCustom instead.
Differential Revision: https://reviews.llvm.org/D80485
In certain places in llvm/lib/CodeGen we were relying upon the TypeSize
comparison operators when in fact the code was only ever expecting
either scalar values or fixed width vectors. This patch changes a few
functions that were always expecting to work on scalar or fixed width
types:
1. DAGCombiner::mergeTruncStores - deals with scalar integers only.
2. DAGCombiner::ReduceLoadWidth - not valid for vectors.
3. DAGCombiner::createBuildVecShuffle - should only be used for
fixed width vectors.
4. SelectionDAGLegalize::ExpandFCOPYSIGN and
SelectionDAGLegalize::getSignAsIntValue - only work on scalars.
Differential Revision: https://reviews.llvm.org/D88562
In certain places in llvm/lib/CodeGen we were relying upon the TypeSize
comparison operators when in fact the code was only ever expecting
either scalar values or fixed width vectors. I've changed some of these
places to use the equivalent scalar operator.
Differential Revision: https://reviews.llvm.org/D88482
In certain places in the code we can never end up in a situation where
we're mixing fixed width and scalable vector types. For example,
we can't have truncations and extends that change the lane count. Also,
in other places such as GenWidenVectorStores and GenWidenVectorLoads we
know from the behaviour of FindMemType that we can never choose a vector
type with a different scalable property.
In various places I have used EVT::bitsXY functions instead of
TypeSize::isKnownXY, where it probably makes sense to keep an assert
that scalable properties match.
Differential Revision: https://reviews.llvm.org/D88654
This passes existing X86 test but I'm not sure if it handles all type
legalization cases it needs to.
Alternative to D89200
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D89222
I have introduced a new template PolySize class, where the template
parameter determines the type of quantity, i.e. for an element
count this is just an unsigned value. The ElementCount class is
now just a simple derivation of PolySize<unsigned>, whereas TypeSize
is more complicated because it still needs to contain the uint64_t
cast operator, since there are still many places in the code that
rely upon this implicit cast. As such the class also still needs
some of it's own operators.
I've tried to minimise the amount of code in the base PolySize
class, which led to a couple of changes:
1. In some places we were relying on '==' operator comparisons
between ElementCounts and the scalar value 1. I didn't put this
operator in the new PolySize class, and thought it was actually
clearer to use the isScalar() function instead.
2. I removed the isByteSized function and replaced it with calls
to isKnownMultipleOf(8).
I've also renamed NextPowerOf2 to be coefficientNextPowerOf2 so
that it's more consistent with coefficientDivideBy.
Differential Revision: https://reviews.llvm.org/D88409
Summary: This patch is derived from D87384.
In this patch we expand the existing decomposition of mul-by-constant to be more general by implementing 2 patterns:
```
mul x, (2^N + 2^M) --> (add (shl x, N), (shl x, M))
mul x, (2^N - 2^M) --> (sub (shl x, N), (shl x, M))
```
The conversion will be trigged if the multiplier is a big constant that the target can't use a single multiplication instruction to handle. This is controlled by the hook `decomposeMulByConstant`.
More over, the conversion benefits from an ILP improvement since the instructions are independent. A case with the sequence like following also gets benefit since a shift instruction is saved.
```
*res1 = a * 0x8800;
*res2 = a * 0x8080;
```
Reviewed By: spatel
Differential Revision: https://reviews.llvm.org/D88201
In DAGCombiner::ForwardStoreValueToDirectLoad I have fixed up some
implicit casts from TypeSize -> uint64_t and replaced calls to
getVectorNumElements() with getVectorElementCount(). There are some
simple cases of forwarding that we can definitely support for
scalable vectors, i.e. when the store and load are both scalable
vectors and have the same size. I have added tests for the new
code paths here:
CodeGen/AArch64/sve-forward-st-to-ld.ll
Differential Revision: https://reviews.llvm.org/D87098
getNode handling for ISD:SETCC calls FoldSETCC which can canonicalize
FP constants to the RHS. When this happens we should create the node
with the FMF that was requested. By using FlagInserter when can ensure
any calls to getNode/getSetcc during canonicalization will also get the flags.
Differential Revision: https://reviews.llvm.org/D88063
In the motivating case from https://llvm.org/PR47517
we create a node that does not get constant folded
before getNegatedExpression is attempted from some
other node, and we crash.
By moving the fold into SelectionDAG::simplifyFPBinop(),
we get the constant fold sooner and avoid the problem.
After some recent upstream discussion we decided that it was best
to avoid having the / operator for both ElementCount and TypeSize,
since this could give the impression that these classes can be used
in the same way as basic integer integer types. However, division
for scalable types is a bit odd because we are only dividing the
minimum quantity by a value, as opposed to something like:
(MinSize * Vscale) / SomeValue
This is why when performing division it's important the caller
first establishes whether the operation makes sense, perhaps by
calling isKnownMultipleOf() prior to division. The caller must now
explictly call divideCoefficientBy() on the class to perform the
operation.
Differential Revision: https://reviews.llvm.org/D87700
If we're multiplying all elements of a vector by '0' or '1' then we can more efficiently perform this as a clearing mask (that is likely to further simplify to a shuffle blend).
This was noticed when reviewing D87502 but seems to help idiv/irem by constant cases even more as '0'/'1' values are often used for 'passthrough' cases.
Differential Revision: https://reviews.llvm.org/D88225
This is like FastMathFlagGuard in IR. Since we use SDAG instance to get
values, it's with SelectionDAG. By creating a FlagInserter in current
scope, all values created by getNode will get the flags if no Flags
argument provided.
In this patch, I applied it to floating point operations folding part in
DAG combiner, and removed Flags passing to getNode to show its effect.
Other places in DAG combiner and other helper methods similar to getNode
also need this. They can be done in follow-up patches.
Reviewed By: spatel
Differential Revision: https://reviews.llvm.org/D87361
An existing function Type::getScalarSizeInBits returns a uint64_t
instead of a TypeSize class because the caller is requesting a
scalar size, which cannot be scalable. This patch makes other
similar functions requesting a scalar size consistent with that,
thereby eliminating more than 1000 implicit TypeSize -> uint64_t
casts.
Differential revision: https://reviews.llvm.org/D87889
If we have an all ones mask, we can just a regular masked load. InstCombine already gets this in IR. But the all ones mask can appear after type legalization.
Only avx512 test cases are affected because X86 backend already looks for element 0 and the last element being 1. It replaces this with an unmasked load and blend. The all ones mask is a special case of that where the blend will be removed. That transform is only enabled on avx2 targets. I believe that's because a non-zero passthru on avx2 already requires a separate blend so its more profitable to handle mixed constant masks.
This patch adds a dedicated all ones handling to the target independent DAG combiner. I've skipped extending, expanding, and index loads for now. X86 doesn't use index so I don't know much about it. Extending made me nervous because I wasn't sure I could trust the memory VT had the right element count due to some weirdness in vector splitting. For expanding I wasn't sure if we needed different undef handling.
Differential Revision: https://reviews.llvm.org/D87788
The versions that take 'unsigned' will be removed in the future.
I tried to use getOriginalAlign instead of getAlign in some
places. getAlign factors in the minimum alignment implied by
the offset in the pointer info. Since we're also passing the
pointer info we can use the original alignment.
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D87592
Similar to D87415, this folds the various float min/max opcodes
with a constant INF or -INF operand, or FLT_MAX / -FLT_MAX operand
if the ninf flag is set. Some of the folds are only possible under
nnan.
The fminnum(X, INF) with nnan and fmaxnum(X, -INF) with nnan cases
are needed to improve the VECREDUCE_FMIN/FMAX lowerings on X86,
the rest is here for the sake of completeness.
Differential Revision: https://reviews.llvm.org/D87571
Previously, we formed ISD::PARITY by looking for (and (ctpop X), 1)
but the AND might be separated from the ctpop. For example if the
parity result is multiplied by 2, we'll pull the AND through the
shift.
So to handle more cases, move to SimplifyDemandedBits where we
can handle more cases that result in only the LSB of the CTPOP
being used.
DAG combiner folds (fma a 1.0 b) into (fadd a b) but the flag isn't
propagated into new fadd. This patch fixes that.
Some code in visitFMA is redundant and such support for vector constants
is missing. Need follow-up patch to clean.
Reviewed By: spatel
Differential Revision: https://reviews.llvm.org/D87037
Clang emits (and (ctpop X), 1) for __builtin_parity. If ctpop
isn't natively supported by the target, this leads to poor codegen
due to the expansion of ctpop being more complex than what is needed
for parity.
This adds a DAG combine to convert the pattern to ISD::PARITY
before operation legalization. Type legalization is updated
to handled Expanding and Promoting this operation. If after type
legalization, CTPOP is supported for this type, LegalizeDAG will
turn it back into CTPOP+AND. Otherwise LegalizeDAG will emit a
series of shifts and xors followed by an AND with 1.
I've avoided vectors in this patch to avoid more legalization
complexity for this patch.
X86 previously had a custom DAG combiner for this. This is now
moved to Custom lowering for the new opcode. There is a minor
regression in vector-reduce-xor-bool.ll, but a follow up patch
can easily fix that.
Fixes PR47433
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D87209
Shivam Gupta