Basically two parts to this fix:
1. Stop using AtomicExpand to expand cmpxchg i128
2. Fix AArch64ExpandPseudoInsts to use a correct expansion.
From ARM architecture reference:
To atomically load two 64-bit quantities, perform a Load-Exclusive
pair/Store-Exclusive pair sequence of reading and writing the same value
for which the Store-Exclusive pair succeeds, and use the read values
from the Load-Exclusive pair.
Fixes https://bugs.llvm.org/show_bug.cgi?id=51102
Differential Revision: https://reviews.llvm.org/D106039
This removes the promotion of NEON AND, OR and XOR nodes to v2i32/v4i32,
treating them the same as the AArch64 and MVE backends where we just add
the relevant patterns for each legal type. This prevents a lot of
bitcasts from being added to the DAG, which have the potential to make
optimizations more difficult. It does mean adding extra patterns, and
some codegen can change due to the types now being legal, not promoted.
Differential Revision: https://reviews.llvm.org/D105588
The case for nxv2f32/nxv2i32 was already covered by D104573.
This patch builds on top of that by making the mechanism work for
nxv2[b]f16/nxv2i16, nxv4[b]f16/nxv4i16 as well.
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D106138
This is mostly a minor convenience, but the pattern seems frequent
enough to be worthwhile (and we'll probably add more uses in the
future).
Differential Revision: https://reviews.llvm.org/D105850
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
Additionally, lower the floating point compare SVE intrinsics to
SETCC_MERGE_ZERO ISD nodes to avoid duplicating ISel patterns.
Differential Revision: https://reviews.llvm.org/D105486
This patch adds a new ShuffleKind SK_Splice and then handle the cost in
getShuffleCost, as in experimental.vector.reverse.
Differential Revision: https://reviews.llvm.org/D104630
Improve codegen when lowering the common vector shuffle case from the
vectorizer (op1[last]:op2[0:last-1]). This patch only handles this
common case as it is difficult to handle this more generally when using
fixed length vectors, due to being unable to use the SVE ext instruction.
Differential Revision: https://reviews.llvm.org/D105289
Target-independent code only knows how to spill to the stack; instead,
use AArch64ISD::REINTERPRET_CAST.
Differential Revision: https://reviews.llvm.org/D104573
Inserting into a smaller-than-legal scalable vector would result in an
internal compiler error. For example, inserting a <vscale x 4 x i8> into
a <vscale x 8 x i8> (both illegal vector types for SVE) would cause a
crash.
This crash was happening because there was no code to promote (legalise)
the result of an INSERT_SUBVECTOR node.
This patch implements PromoteIntRes_INSERT_SUBVECTOR, which legalises
the ISD node. This is currently done by going through memory. This is
necessary because of the requirement that the SubVec parameter of the
INSERT_SUBVECTOR node must be smaller than the Vec parameter, which
means that INSERT_SUBVECTOR cannot always have a legal result/operand
types.
Co-Authored-by: Joe Ellis <joe.ellis@arm.com>
Differential Revision: https://reviews.llvm.org/D102766
Since gather lowering can now lower to nodes that may need expansion via
the vector legalizer, do MGATHER lowering via vector legalizer.
Additionally, as part of adding passthru support for fixed typed
gathers, fix passthru support for scalable types.
Depends on D104910
Differential Revision: https://reviews.llvm.org/D104217
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 custom lowers <4 x i8> vector loads using a 32-bit load, followed by 2
SSHLL instructions to extend it to e.g. a <4 x i32> vector. Before, it was
really inefficient and expensive to construct a <4 x i32> for this as 4 byte
loads and 4 moves were used. With this improvement SLP vectorisation might for
example become profitable, see D103629.
Differential Revision: https://reviews.llvm.org/D104782
This is a mechanical change. This actually also renames the
similarly named methods in the SmallString class, however these
methods don't seem to be used outside of the llvm subproject, so
this doesn't break building of the rest of the monorepo.
This also adds new interfaces for the fixed- and scalable case:
* LLT::fixed_vector
* LLT::scalable_vector
The strategy for migrating to the new interfaces was as follows:
* If the new LLT is a (modified) clone of another LLT, taking the
same number of elements, then use LLT::vector(OtherTy.getElementCount())
or if the number of elements is halfed/doubled, it uses .divideCoefficientBy(2)
or operator*. That is because there is no reason to specifically restrict
the types to 'fixed_vector'.
* If the algorithm works on the number of elements (as unsigned), then
just use fixed_vector. This will need to be fixed up in the future when
modifying the algorithm to also work for scalable vectors, and will need
then need additional tests to confirm the behaviour works the same for
scalable vectors.
* If the test used the '/*Scalable=*/true` flag of LLT::vector, then
this is replaced by LLT::scalable_vector.
Reviewed By: aemerson
Differential Revision: https://reviews.llvm.org/D104451
This only applies to FastIsel. GlobalIsel seems to sidestep
the issue.
This fixes https://bugs.llvm.org/show_bug.cgi?id=46996
One of the things we do in llvm is decide if a type needs
consecutive registers. Previously, we just checked if it
was an array or not.
(plus an SVE specific check that is not changing here)
This causes some confusion when you arbitrary IR like:
```
%T1 = type { double, i1 };
define [ 1 x %T1 ] @foo() {
entry:
ret [ 1 x %T1 ] zeroinitializer
}
```
We see it is an array so we call CC_AArch64_Custom_Block
which bails out when it sees the i1, a type we don't want
to put into a block.
This leaves the location of the double in some kind of
intermediate state and leads to odd codegen. Which then crashes
the backend because it doesn't know how to implement
what it's been asked for.
You get this:
```
renamable $d0 = FMOVD0
$w0 = COPY killed renamable $d0
```
Rather than this:
```
$d0 = FMOVD0
$w0 = COPY $wzr
```
The backend knows how to copy 64 bit to 64 bit registers,
but not 64 to 32. It can certainly be taught how but the real
issue seems to be us even trying to assign a register block
in the first place.
This change makes the logic of
AArch64TargetLowering::functionArgumentNeedsConsecutiveRegisters
a bit more in depth. If we find an array, also check that all the
nested aggregates in that array have a single member type.
Then CC_AArch64_Custom_Block's assumption of a type that looks
like [ N x type ] will be valid and we get the expected codegen.
New tests have been added to exercise these situations. Note that
some of the output is not ABI compliant. The aim of this change is
to simply handle these situations and not to make our processing
of arbitrary IR ABI compliant.
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D104123
Currently, Loop strengh reduce is not handling loops with scalable stride very well.
Take loop vectorized with scalable vector type <vscale x 8 x i16> for instance,
(refer to test/CodeGen/AArch64/sve-lsr-scaled-index-addressing-mode.ll added).
Memory accesses are incremented by "16*vscale", while induction variable is incremented
by "8*vscale". The scaling factor "2" needs to be extracted to build candidate formula
i.e., "reg(%in) + 2*reg({0,+,(8 * %vscale)}". So that addrec register reg({0,+,(8*vscale)})
can be reused among Address and ICmpZero LSRUses to enable optimal solution selection.
This patch allow LSR getExactSDiv to recognize special cases like "C1*X*Y /s C2*X*Y",
and pull out "C1 /s C2" as scaling factor whenever possible. Without this change, LSR
is missing candidate formula with proper scaled factor to leverage target scaled-index
addressing mode.
Note: This patch doesn't fully fix AArch64 isLegalAddressingMode for scalable
vector. But allow simple valid scale to pass through.
Reviewed By: sdesmalen
Differential Revision: https://reviews.llvm.org/D103939
Given a vecreduce_add node, detect the below pattern and convert it to the node
sequence with UABDL, [S|U]ADB and UADDLP.
i32 vecreduce_add(
v16i32 abs(
v16i32 sub(
v16i32 [sign|zero]_extend(v16i8 a), v16i32 [sign|zero]_extend(v16i8 b))))
=================>
i32 vecreduce_add(
v4i32 UADDLP(
v8i16 add(
v8i16 zext(
v8i8 [S|U]ABD low8:v16i8 a, low8:v16i8 b
v8i16 zext(
v8i8 [S|U]ABD high8:v16i8 a, high8:v16i8 b
Differential Revision: https://reviews.llvm.org/D104042
Don't require a specific kind of IRBuilder for TargetLowering hooks.
This allows us to drop the IRBuilder.h include from TargetLowering.h.
Differential Revision: https://reviews.llvm.org/D103759
This NEG node is just a vector negation, easily represented as a SUB
zero. Removing it from the one place it is generated is essentially an
NFC, but can allow some extra folding. The updated tests are now loading
different constant literals, which have already been negated.
Differential Revision: https://reviews.llvm.org/D103703
setcc (csel 0, 1, cond, X), 1, ne ==> csel 0, 1, !cond, X
Where X is a condition code setting instruction.
Co-authored-by: Paul Walker <paul.walker@arm.com>
Differential Revision: https://reviews.llvm.org/D103256
If a cmpxchg specifies acquire or seq_cst on failure, make sure we
generate code consistent with that ordering even if the success ordering
is not acquire/seq_cst.
At one point, it was ambiguous whether this sort of construct was valid,
but the C++ standad and LLVM now accept arbitrary combinations of
success/failure orderings.
This doesn't address the corresponding issue in AtomicExpand. (This was
reported as https://bugs.llvm.org/show_bug.cgi?id=33332 .)
Fixes https://bugs.llvm.org/show_bug.cgi?id=50512.
Differential Revision: https://reviews.llvm.org/D103284
SwiftTailCC has a different set of requirements than the C calling convention
for a tail call. The exact argument sequence doesn't have to match, but fewer
ABI-affecting attributes are allowed.
Also make sure the musttail diagnostic triggers if a musttail call isn't
actually a tail call.
This adds custom lowering for the MLOAD and MSTORE ISD nodes when
passed fixed length vectors in SVE. This is done by converting the
vectors to VLA vectors and using the VLA code generation.
Fixed length extending loads and truncating stores currently produce
correct code, but do not use the built in extend/truncate in the
load and store instructions. This will be fixed in a future patch.
Differential Revision: https://reviews.llvm.org/D101834
bswap.v2i16 + sitofp in LLVM IR generate a sequence of:
- REV32 + USHR for bswap.v2i16
- SHL + SSHR + SCVTF for sext to v2i32 and scvt
The shift instructions are excessive as noted in PR24820, and they can
be optimized to just SSHR.
Differential Revision: https://reviews.llvm.org/D102333
The implementation just extends the vector to a larger element type, and
extracts from that. Not fancy, but generates reasonable code.
There was discussion in the review of doing the promotion in
target-independent code, but I'm sticking with this to avoid making
LegalizeDAG infrastructure more complicated.
Differential Revision: https://reviews.llvm.org/D87651
Adding lowering support for bitreverse.
Previously, lowering bitreverse would expand it into a series of other instructions. This patch makes it so this produces a single rbit instruction instead.
Reviewed By: dmgreen
Differential Revision: https://reviews.llvm.org/D102397
Swift's new concurrency features are going to require guaranteed tail calls so
that they don't consume excessive amounts of stack space. This would normally
mean "tailcc", but there are also Swift-specific ABI desires that don't
naturally go along with "tailcc" so this adds another calling convention that's
the combination of "swiftcc" and "tailcc".
Support is added for AArch64 and X86 for now.
AArch64's fctv* instructions implement the saturating behaviour that the
fpto*i.sat intrinsics require, in cases where the destination width
matches the saturation width. Lowering them removes a lot of unnecessary
generated code.
Only scalar lowerings are supported for now.
Differential Revision: https://reviews.llvm.org/D102353
Addition of this node allows us to better utilize the different forms of
the SVE BIC instructions, including using the alias to an AND (immediate).
Differential Revision: https://reviews.llvm.org/D101831
This extends any frame record created in the function to include that
parameter, passed in X22.
The new record looks like [X22, FP, LR] in memory, and FP is stored with 0b0001
in bits 63:60 (CodeGen assumes they are 0b0000 in normal operation). The effect
of this is that tools walking the stack should expect to see one of three
values there:
* 0b0000 => a normal, non-extended record with just [FP, LR]
* 0b0001 => the extended record [X22, FP, LR]
* 0b1111 => kernel space, and a non-extended record.
All other values are currently reserved.
If compiling for arm64e this context pointer is address-discriminated with the
discriminator 0xc31a and the DB (process-specific) key.
There is also an "i8** @llvm.swift.async.context.addr()" intrinsic providing
front-ends access to this slot (and forcing its creation initialized to nullptr
if necessary).
The sve.convert.to.svbool lowering has the effect of widening a logical
<M x i1> vector representing lanes into a physical <16 x i1> vector
representing bits in a predicate register.
In general, if converting to svbool, the contents of lanes in the
physical register might not be known. For sve.convert.to.svbool the new
lanes are specified to be zeroed, requiring 'and' instructions to mask
off the new lanes. For lanes coming from a ptrue or a comparison,
however, they are known to be zero.
CodeGen Before:
ptrue p0.s, vl16
ptrue p1.s
ptrue p2.b
and p0.b, p2/z, p0.b, p1.b
ret
After:
ptrue p0.s, vl16
ret
Differential Revision: https://reviews.llvm.org/D101544
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
When using predicated intrinsics, if the predicate used is all lanes active,
use an unpredicated form of the instruction, additionally this allows for
better use of immediate forms.
This only includes instructions where the unpredicated/predicated forms
matched in such a way that instruction selection would not introduce extra
ptrue instructions. This allows us to convert the intrinsics directly to
architecture independent ISD nodes.
Depends on D101062
Differential Revision: https://reviews.llvm.org/D101828
When using predicated arithmetic intrinsics, if the predicate used is all
lanes active, use an unpredicated form of the instruction, additionally
this allows for better use of immediate forms.
This also includes a new complex isel pattern which allows matching an
all active predicate when the types are different but the predicate is a
superset of the type being used. For example, to allow a b8 ptrue for a
b32 predicate operand.
This only includes instructions where the unpredicated/predicated forms
are mismatched between variants, meaning that the removal of the
predicate is done during instruction selection in order to prevent
spurious re-introductions of ptrue instructions.
Co-authored-by: Paul Walker <paul.walker@arm.com>
Differential Revision: https://reviews.llvm.org/D101062
DAGCombiner tries to combine a (fpext (load)) to (fround (extload))
but SVE has no FP-extending loads. By marking these as expand,
the combine no longer happens.
This also fixes a similar issue for fptrunc, where the source type
is not a legal type.
Reviewed By: bsmith, kmclaughlin
Differential Revision: https://reviews.llvm.org/D102053
Since index_vector is lowered into step_vector in D100816, we can just remove
index_vector, use step_vector for codegen directly.
Differential Revision: https://reviews.llvm.org/D101593
Based off a discussion on D89281 - where the AARCH64 implementations were being replaced to use funnel shifts.
Any target that has efficient funnel shift lowering can handle the shift parts expansion using the same expansion, avoiding a lot of duplication.
I've generalized the X86 implementation and moved it to TargetLowering - so far I've found that AARCH64 and AMDGPU benefit, but many other targets (ARM, PowerPC + RISCV in particular) could easily use this with a few minor improvements to their funnel shift lowering (or the folding of their target ops that funnel shifts lower to).
NOTE: I'm trying to avoid adding full SHIFT_PARTS legalizer handling as I think it might actually be possible to remove these opcodes in the medium-term and use funnel shift / libcall expansion directly.
Differential Revision: https://reviews.llvm.org/D101987
Specifically, this allow us to rely on the lane zero'ing behaviour of
SVE reduce instructions.
Co-authored-by: Paul Walker <paul.walker@arm.com>
Differential Revision: https://reviews.llvm.org/D101369
This can come up in rare situations, where a csel is created with
identical operands. These can be folded simply to the original value,
allowing the csel to be removed and further simplification to happen.
This patch also removes FCSEL as it is unused, not being produced
anywhere or lowered to anything.
Differential Revision: https://reviews.llvm.org/D101687
Apply the same logic used to check if CMPXCHG nodes should be expanded
at -O0: the register allocator may end up spilling some register in
between the atomic load/store pairs, breaking the atomicity and possibly
stalling the execution.
Fixes PR48017
Reviewed By: efriedman
Differential Revision: https://reviews.llvm.org/D101163
These operations don't exist natively, so just let the
target-independent code expand to plain shifts.
The generated sequences could probably be optimized a bit more, but
they seem good enough for now.
Differential Revision: https://reviews.llvm.org/D101574
As discussed in D100107, this patch first convert index_vector to
step_vector, and convert step_vector back to index_vector after LegalizeDAG.
Differential Revision: https://reviews.llvm.org/D100816
The function AArch64TargetLowering::LowerFixedLengthVectorIntDivideToSVE
previously assumed the operands were full vectors, but this is not
always true. This function would produce bogus if the division operands
are not full vectors, resulting in miscompiles when dividing 8-bit or
16-bit vectors.
The fix is to perform an extend + div + truncate for non-full vectors,
instead of the usual unpacking and unzipping logic. This is an additive
change which reduces the non-full integer vector divisions to a pattern
recognised by the existing lowering logic.
For future reference, an example of code that would miscompile before
this patch is below:
1 int8_t foo(unsigned N, int8_t *a, int8_t *b, int8_t *c) {
2 int8_t result = 0;
3 for (int i = 0; i < N; ++i) {
4 result += (a[i] / b[i]) / c[i];
5 }
6 return result;
7 }
Differential Revision: https://reviews.llvm.org/D100370
This improves the lowering of v8i16 and v16i8 vector reverse shuffles.
Instead of going via a generic tbl it uses a rev64; ext pair, as already
happens for v4i32.
Differential Revision: https://reviews.llvm.org/D100882
There are no patterns for the AArch64ISD::BSP ISD node for anything
other than NEON vectors at the moment. As a result, if we hit these
combines for vectors wider than a NEON vector (such as what we might get
with fixed length SVE) we will fail to lower.
This patch simply prevents us from attempting the combines if the input
vector type is too wide.
Reviewed By: peterwaller-arm
Differential Revision: https://reviews.llvm.org/D100961
When inspecting the calling convention, for calling windows functions
from a non-windows function, inspect the calling convention of
the called function, not the caller.
Also remove an unnecessary parameter to AArch64CallLowering
OutgoingArgHandler.
Differential Revision: https://reviews.llvm.org/D100890
This patch changes the lowering of SELECT_CC from Legal to Expand for scalable
vector and adds support for scalable vectors in performSelectCombine.
When selecting the nodes to lower in visitSELECT it checks if it is possible to
use SELECT_CC in cases where SETCC is followed by SELECT. visistSELECT checks
if SELECT_CC is legal or custom to replace SELECT by SELECT_CC.
SELECT_CC used to be legal for scalable vector, so the node changes to
SELECT_CC. This used to crash the compiler as there is no support for SELECT_CC
with scalable vectors. So now the compiler lowers to VSELECT instead of
SELECT_CC.
Differential Revision: https://reviews.llvm.org/D100485
Mark MULHS/MULHU nodes as legal for both scalable and fixed SVE types,
and lower them to the appropriate SVE instructions.
Additionally now that the MULH nodes are legal, integer divides can be
expanded into a more performant code sequence.
Differential Revision: https://reviews.llvm.org/D100487
With this patch vbslq_f32(vnegq_s32(a), b, c) lowers to a BIT instruction.
Co-authored-by: Paul Walker <paul.walker@arm.com>
Differential Revision: https://reviews.llvm.org/D100304
On Windows, float arguments are normally passed in float registers
in the calling convention for regular functions. For variable
argument functions, floats are passed in integer registers. This
already was done correctly since many years.
However, the surprising bit was that floats among the fixed arguments
also are supposed to be passed in integer registers, contrary to regular
functions. (This also seems to be the behaviour on ARM though, both
on Windows, but also on e.g. hardfloat linux.)
In the calling convention, don't promote shorter floats to f64, but
convert them to integers of the same length. (Floats passed as part of
the actual variable arguments are promoted to double already on the
C/Clang level; the LLVM vararg calling convention doesn't do any
extra promotion of f32 to f64 - this matches how it works on X86 too.)
Technically, this is an ABI break compared to older LLVM versions,
but it fixes compatibility with the official platform ABI. (In practice,
floats among the fixed arguments in variable argument functions is
a pretty rare construct.)
Differential Revision: https://reviews.llvm.org/D100365
When attempting to truncate a FP vector and store the result out
to memory we crashed because we had no pattern for truncating FP
stores. In fact, we don't support these types of stores and the
correct fix is to stop marking these truncating stores as legal.
Tests have been added here:
CodeGen/AArch64/sve-fptrunc-store.ll
Differential Revision: https://reviews.llvm.org/D100025
When an SVE function calls another SVE function using the C calling
convention we use the more efficient SVE VectorCall PCS. However,
for the Fast calling convention we're incorrectly falling back to
the generic AArch64 PCS.
This patch adds the same "can use SVE vector calling convention"
detection used by CallingConv::C to CallingConv::Fast.
Co-authored-by: Paul Walker <paul.walker@arm.com>
Differential Revision: https://reviews.llvm.org/D99657
This marks FSIN and other operations to EXPAND for scalable
vectors, so that they are not assumed to be legal by the cost-model.
Depends on D97470
Reviewed By: dmgreen, paulwalker-arm
Differential Revision: https://reviews.llvm.org/D97471
Currently the code only checks for integer constants (ConstantSDNode)
and triggers an infinite cycle for single-element floating point
vector constants. We need to check for both FP and integer constants.
Reviewed By: t.p.northover
Differential Revision: https://reviews.llvm.org/D99384
Currently performExtendCombine assumes that the src-element bitwidth * 2
is a valid MVT. But this is not the case for i1 and it causes a crash on
the v64i1 test cases added in this patch.
It turns out that this code appears to not be needed; the same patterns are
handled by other code and we end up with the same results, even without the
custom lowering. I also added additional test cases in a50037aaa6.
Let's just remove the unneeded code.
Reviewed By: dmgreen
Differential Revision: https://reviews.llvm.org/D99437
This is currently performed in SelectionDAGLegalize, here we make it also
happen in LegalizeVectorOps, allowing a target to lower the SETCC condition
codes first in LegalizeVectorOps and then lower to a custom node afterwards,
without having to duplicate all of the SETCC condition legalization in the
target specific lowering.
As a result of this, fixed length floating point SETCC nodes can now be
properly lowered for SVE.
Differential Revision: https://reviews.llvm.org/D98939
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
The VSelectCombine handler within AArch64ISelLowering,
uses an interface call which only expects fixed vectors.
This generates a warning when the call is made on a
scalable vector. This warning has been suppressed with this change,
by using the ElementCount interface, which supports both fixed and scalable vectors.
I have also added a regression test which recreates the warning.
Differential Revision: https://reviews.llvm.org/D98249
This patch adds a new llvm.experimental.stepvector intrinsic,
which takes no arguments and returns a linear integer sequence of
values of the form <0, 1, ...>. It is primarily intended for
scalable vectors, although it will work for fixed width vectors
too. It is intended that later patches will make use of this
new intrinsic when vectorising induction variables, currently only
supported for fixed width. I've added a new CreateStepVector
method to the IRBuilder, which will generate a call to this
intrinsic for scalable vectors and fall back on creating a
ConstantVector for fixed width.
For scalable vectors this intrinsic is lowered to a new ISD node
called STEP_VECTOR, which takes a single constant integer argument
as the step. During lowering this argument is set to a value of 1.
The reason for this additional argument at the codegen level is
because in future patches we will introduce various generic DAG
combines such as
mul step_vector(1), 2 -> step_vector(2)
add step_vector(1), step_vector(1) -> step_vector(2)
shl step_vector(1), 1 -> step_vector(2)
etc.
that encourage a canonical format for all targets. This hopefully
means all other targets supporting scalable vectors can benefit
from this too.
I've added cost model tests for both fixed width and scalable
vectors:
llvm/test/Analysis/CostModel/AArch64/neon-stepvector.ll
llvm/test/Analysis/CostModel/AArch64/sve-stepvector.ll
as well as codegen lowering tests for fixed width and scalable
vectors:
llvm/test/CodeGen/AArch64/neon-stepvector.ll
llvm/test/CodeGen/AArch64/sve-stepvector.ll
See this thread for discussion of the intrinsic:
https://lists.llvm.org/pipermail/llvm-dev/2021-January/147943.html
Found by adding asserts to LegalizeDAG to make sure custom legalized
results had the right types.
Reviewed By: kmclaughlin
Differential Revision: https://reviews.llvm.org/D98968
Don't rewrite an add instruction with 2 SET_CC operands into a csel
instruction. The total instruction sequence uses an extra instruction and
register. Preventing this allows us to match a `(add, csel)` pattern and
rewrite this into a `cinc`.
Differential Revision: https://reviews.llvm.org/D98704
Previously NEON used a target specific intrinsic for frintn, given that
the FROUNDEVEN ISD node now exists, move over to that instead and add
codegen support for that node for both NEON and fixed length SVE.
Differential Revision: https://reviews.llvm.org/D98487
This commit folds sxtw'd or uxtw'd offsets into gather loads where
possible with a DAGCombine optimization.
As an example, the following code:
1 #include <arm_sve.h>
2
3 svuint64_t func(svbool_t pred, const int32_t *base, svint64_t offsets) {
4 return svld1sw_gather_s64offset_u64(
5 pred, base, svextw_s64_x(pred, offsets)
6 );
7 }
would previously lower to the following assembly:
sxtw z0.d, p0/m, z0.d
ld1sw { z0.d }, p0/z, [x0, z0.d]
ret
but now lowers to:
ld1sw { z0.d }, p0/z, [x0, z0.d, sxtw]
ret
Differential Revision: https://reviews.llvm.org/D97858
This patch implements the __rndr and __rndrrs intrinsics to provide access to the random
number instructions introduced in Armv8.5-A. They are only defined for the AArch64
execution state and are available when __ARM_FEATURE_RNG is defined.
These intrinsics store the random number in their pointer argument and return a status
code if the generation succeeded. The difference between __rndr __rndrrs, is that the latter
intrinsic reseeds the random number generator.
The instructions write the NZCV flags indicating the success of the operation that we can
then read with a CSET.
[1] https://developer.arm.com/docs/101028/latest/data-processing-intrinsics
[2] https://bugs.llvm.org/show_bug.cgi?id=47838
Differential Revision: https://reviews.llvm.org/D98264
Change-Id: I8f92e7bf5b450e5da3e59943b53482edf0df6efc
We previously have lowering for:
vecreduce.add(zext(X)) to vecreduce.add(UDOT(zero, X, one))
This extends that to also handle:
vecreduce.add(mul(zext(X), zext(Y)) to vecreduce.add(UDOT(zero, X, Y))
It extends the existing code to optionally handle a mul with equal
extends.
Differential Revision: https://reviews.llvm.org/D97280
This patch introduces a new intrinsic @llvm.experimental.vector.splice
that constructs a vector of the same type as the two input vectors,
based on a immediate where the sign of the immediate distinguishes two
variants. A positive immediate specifies an index into the first vector
and a negative immediate specifies the number of trailing elements to
extract from the first vector.
For example:
@llvm.experimental.vector.splice(<A,B,C,D>, <E,F,G,H>, 1) ==> <B, C, D, E> ; index
@llvm.experimental.vector.splice(<A,B,C,D>, <E,F,G,H>, -3) ==> <B, C, D, E> ; trailing element count
These intrinsics support both fixed and scalable vectors, where the
former is lowered to a shufflevector to maintain existing behaviour,
although while marked as experimental the recommended way to express
this operation for fixed-width vectors is to use shufflevector. For
scalable vectors where it is not possible to express a shufflevector
mask for this operation, a new ISD node has been implemented.
This is one of the named shufflevector intrinsics proposed on the
mailing-list in the RFC at [1].
Patch by Paul Walker and Cullen Rhodes.
[1] https://lists.llvm.org/pipermail/llvm-dev/2020-November/146864.html
Reviewed By: sdesmalen
Differential Revision: https://reviews.llvm.org/D94708
This is included from IR files, and IR doesn't/can't depend on Analysis
(because Analysis depends on IR).
Also fix the implementation - don't use non-member static in headers, as
it leads to ODR violations, inaccurate "unused function" warnings, etc.
And fix the header protection macro name (we don't generally include
"LIB" in the names, so far as I can tell).
explicitly emitting retainRV or claimRV calls in the IR
This reapplies ed4718eccb, which was reverted
because it was causing a miscompile. The bug that was causing the miscompile
has been fixed in 75805dce5f.
Original commit message:
Background:
This fixes a longstanding problem where llvm breaks ARC's autorelease
optimization (see the link below) by separating calls from the marker
instructions or retainRV/claimRV calls. The backend changes are in
https://reviews.llvm.org/D92569.
https://clang.llvm.org/docs/AutomaticReferenceCounting.html#arc-runtime-objc-autoreleasereturnvalue
What this patch does to fix the problem:
- The front-end adds operand bundle "clang.arc.attachedcall" to calls,
which indicates the call is implicitly followed by a marker
instruction and an implicit retainRV/claimRV call that consumes the
call result. In addition, it emits a call to
@llvm.objc.clang.arc.noop.use, which consumes the call result, to
prevent the middle-end passes from changing the return type of the
called function. This is currently done only when the target is arm64
and the optimization level is higher than -O0.
- ARC optimizer temporarily emits retainRV/claimRV calls after the calls
with the operand bundle in the IR and removes the inserted calls after
processing the function.
- ARC contract pass emits retainRV/claimRV calls after the call with the
operand bundle. It doesn't remove the operand bundle on the call since
the backend needs it to emit the marker instruction. The retainRV and
claimRV calls are emitted late in the pipeline to prevent optimization
passes from transforming the IR in a way that makes it harder for the
ARC middle-end passes to figure out the def-use relationship between
the call and the retainRV/claimRV calls (which is the cause of
PR31925).
- The function inliner removes an autoreleaseRV call in the callee if
nothing in the callee prevents it from being paired up with the
retainRV/claimRV call in the caller. It then inserts a release call if
claimRV is attached to the call since autoreleaseRV+claimRV is
equivalent to a release. If it cannot find an autoreleaseRV call, it
tries to transfer the operand bundle to a function call in the callee.
This is important since the ARC optimizer can remove the autoreleaseRV
returning the callee result, which makes it impossible to pair it up
with the retainRV/claimRV call in the caller. If that fails, it simply
emits a retain call in the IR if retainRV is attached to the call and
does nothing if claimRV is attached to it.
- SCCP refrains from replacing the return value of a call with a
constant value if the call has the operand bundle. This ensures the
call always has at least one user (the call to
@llvm.objc.clang.arc.noop.use).
- This patch also fixes a bug in replaceUsesOfNonProtoConstant where
multiple operand bundles of the same kind were being added to a call.
Future work:
- Use the operand bundle on x86-64.
- Fix the auto upgrader to convert call+retainRV/claimRV pairs into
calls with the operand bundles.
rdar://71443534
Differential Revision: https://reviews.llvm.org/D92808
This caused miscompiles of Chromium tests for iOS due clobbering of live
registers. See discussion on the code review for details.
> Background:
>
> This fixes a longstanding problem where llvm breaks ARC's autorelease
> optimization (see the link below) by separating calls from the marker
> instructions or retainRV/claimRV calls. The backend changes are in
> https://reviews.llvm.org/D92569.
>
> https://clang.llvm.org/docs/AutomaticReferenceCounting.html#arc-runtime-objc-autoreleasereturnvalue
>
> What this patch does to fix the problem:
>
> - The front-end adds operand bundle "clang.arc.attachedcall" to calls,
> which indicates the call is implicitly followed by a marker
> instruction and an implicit retainRV/claimRV call that consumes the
> call result. In addition, it emits a call to
> @llvm.objc.clang.arc.noop.use, which consumes the call result, to
> prevent the middle-end passes from changing the return type of the
> called function. This is currently done only when the target is arm64
> and the optimization level is higher than -O0.
>
> - ARC optimizer temporarily emits retainRV/claimRV calls after the calls
> with the operand bundle in the IR and removes the inserted calls after
> processing the function.
>
> - ARC contract pass emits retainRV/claimRV calls after the call with the
> operand bundle. It doesn't remove the operand bundle on the call since
> the backend needs it to emit the marker instruction. The retainRV and
> claimRV calls are emitted late in the pipeline to prevent optimization
> passes from transforming the IR in a way that makes it harder for the
> ARC middle-end passes to figure out the def-use relationship between
> the call and the retainRV/claimRV calls (which is the cause of
> PR31925).
>
> - The function inliner removes an autoreleaseRV call in the callee if
> nothing in the callee prevents it from being paired up with the
> retainRV/claimRV call in the caller. It then inserts a release call if
> claimRV is attached to the call since autoreleaseRV+claimRV is
> equivalent to a release. If it cannot find an autoreleaseRV call, it
> tries to transfer the operand bundle to a function call in the callee.
> This is important since the ARC optimizer can remove the autoreleaseRV
> returning the callee result, which makes it impossible to pair it up
> with the retainRV/claimRV call in the caller. If that fails, it simply
> emits a retain call in the IR if retainRV is attached to the call and
> does nothing if claimRV is attached to it.
>
> - SCCP refrains from replacing the return value of a call with a
> constant value if the call has the operand bundle. This ensures the
> call always has at least one user (the call to
> @llvm.objc.clang.arc.noop.use).
>
> - This patch also fixes a bug in replaceUsesOfNonProtoConstant where
> multiple operand bundles of the same kind were being added to a call.
>
> Future work:
>
> - Use the operand bundle on x86-64.
>
> - Fix the auto upgrader to convert call+retainRV/claimRV pairs into
> calls with the operand bundles.
>
> rdar://71443534
>
> Differential Revision: https://reviews.llvm.org/D92808
This reverts commit ed4718eccb.
Given a zero input for a udot, an add can be folded in to take the place
of the input, using thte addition that the instruction naturally
performs.
Differential Revision: https://reviews.llvm.org/D97188
This patch addresses issues arising from the fact that the index type
used for subvector insertion/extraction is inconsistent between the
intrinsics and SDNodes. The intrinsic forms require i64 whereas the
SDNodes use the type returned by SelectionDAG::getVectorIdxTy.
Rather than update the intrinsic definitions to use an overloaded index
type, this patch fixes the issue by transforming the index to the
correct type as required. Any loss of index bits going from i64 to a
smaller type is unexpected, and will be caught by an assertion in
SelectionDAG::getVectorIdxConstant.
The patch also updates the documentation for INSERT_SUBVECTOR and adds
an assertion to its creation to bring it in line with EXTRACT_SUBVECTOR.
This necessitated changes to AArch64 which was using i64 for
EXTRACT_SUBVECTOR but i32 for INSERT_SUBVECTOR. Only one test changed
its codegen after updating the backend accordingly.
Reviewed By: sdesmalen
Differential Revision: https://reviews.llvm.org/D97459
This is used to lower UDOT/SDOT instructions, as opposed to relying on
the intrinsic. Subsequent optimizations will be able to optimize them
more cleanly based on these nodes.
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
isFMAFasterThanFMulAndFAdd should return true for FP16 types when
HasFullFP16 is present, since we have the instructions to handle it for
both SVE and NEON. (SVE patterns and tests will follow).
Differential Revision: https://reviews.llvm.org/D96599
This patch fixes a codegen crash introduced in fde2466171, where the
DAGCombiner started generating optimized MULH[SU] or [SU]MUL_LOHI nodes
unless the target opted out. The AArch64 backend cannot currently select
any of these nodes, so ensure that they are not generated in the first
place.
This issue was raised by @huihuiz in D94501.
Reviewed By: paulwalker-arm
Differential Revision: https://reviews.llvm.org/D96849
ICMP & SELECT patterns extracting the sign of a value can be simplified
to OR & ASR (see https://alive2.llvm.org/ce/z/Xx4iZ0).
This does not save any instructions in IR, but it is profitable on
AArch64, because we need at least 2 extra instructions to materialize 1
and -1 for the SELECT.
The improvements result in ~5% speedups on loops of the form
static int sign_of(int x) {
if (x < 0) return -1;
return 1;
}
void foo(const int *x, int *res, int cnt) {
for (int i=0;i<cnt;i++)
res[i] = sign_of(x[i]);
}
Reviewed By: dmgreen
Differential Revision: https://reviews.llvm.org/D96596
This patch adds a new intrinsic experimental.vector.reduce that takes a single
vector and returns a vector of matching type but with the original lane order
reversed. For example:
```
vector.reverse(<A,B,C,D>) ==> <D,C,B,A>
```
The new intrinsic supports fixed and scalable vectors types.
The fixed-width vector relies on shufflevector to maintain existing behaviour.
Scalable vector uses the new ISD node - VECTOR_REVERSE.
This new intrinsic is one of the named shufflevector intrinsics proposed on the
mailing-list in the RFC at [1].
Patch by Paul Walker (@paulwalker-arm).
[1] https://lists.llvm.org/pipermail/llvm-dev/2020-November/146864.html
Differential Revision: https://reviews.llvm.org/D94883
explicitly emitting retainRV or claimRV calls in the IR
Background:
This fixes a longstanding problem where llvm breaks ARC's autorelease
optimization (see the link below) by separating calls from the marker
instructions or retainRV/claimRV calls. The backend changes are in
https://reviews.llvm.org/D92569.
https://clang.llvm.org/docs/AutomaticReferenceCounting.html#arc-runtime-objc-autoreleasereturnvalue
What this patch does to fix the problem:
- The front-end adds operand bundle "clang.arc.attachedcall" to calls,
which indicates the call is implicitly followed by a marker
instruction and an implicit retainRV/claimRV call that consumes the
call result. In addition, it emits a call to
@llvm.objc.clang.arc.noop.use, which consumes the call result, to
prevent the middle-end passes from changing the return type of the
called function. This is currently done only when the target is arm64
and the optimization level is higher than -O0.
- ARC optimizer temporarily emits retainRV/claimRV calls after the calls
with the operand bundle in the IR and removes the inserted calls after
processing the function.
- ARC contract pass emits retainRV/claimRV calls after the call with the
operand bundle. It doesn't remove the operand bundle on the call since
the backend needs it to emit the marker instruction. The retainRV and
claimRV calls are emitted late in the pipeline to prevent optimization
passes from transforming the IR in a way that makes it harder for the
ARC middle-end passes to figure out the def-use relationship between
the call and the retainRV/claimRV calls (which is the cause of
PR31925).
- The function inliner removes an autoreleaseRV call in the callee if
nothing in the callee prevents it from being paired up with the
retainRV/claimRV call in the caller. It then inserts a release call if
claimRV is attached to the call since autoreleaseRV+claimRV is
equivalent to a release. If it cannot find an autoreleaseRV call, it
tries to transfer the operand bundle to a function call in the callee.
This is important since the ARC optimizer can remove the autoreleaseRV
returning the callee result, which makes it impossible to pair it up
with the retainRV/claimRV call in the caller. If that fails, it simply
emits a retain call in the IR if retainRV is attached to the call and
does nothing if claimRV is attached to it.
- SCCP refrains from replacing the return value of a call with a
constant value if the call has the operand bundle. This ensures the
call always has at least one user (the call to
@llvm.objc.clang.arc.noop.use).
- This patch also fixes a bug in replaceUsesOfNonProtoConstant where
multiple operand bundles of the same kind were being added to a call.
Future work:
- Use the operand bundle on x86-64.
- Fix the auto upgrader to convert call+retainRV/claimRV pairs into
calls with the operand bundles.
rdar://71443534
Differential Revision: https://reviews.llvm.org/D92808
emitting retainRV or claimRV calls in the IR
This reapplies 3fe3946d9a without the
changes made to lib/IR/AutoUpgrade.cpp, which was violating layering.
Original commit message:
Background:
This patch makes changes to the front-end and middle-end that are
needed to fix a longstanding problem where llvm breaks ARC's autorelease
optimization (see the link below) by separating calls from the marker
instructions or retainRV/claimRV calls. The backend changes are in
https://reviews.llvm.org/D92569.
https://clang.llvm.org/docs/AutomaticReferenceCounting.html#arc-runtime-objc-autoreleasereturnvalue
What this patch does to fix the problem:
- The front-end adds operand bundle "clang.arc.rv" to calls, which
indicates the call is implicitly followed by a marker instruction and
an implicit retainRV/claimRV call that consumes the call result. In
addition, it emits a call to @llvm.objc.clang.arc.noop.use, which
consumes the call result, to prevent the middle-end passes from changing
the return type of the called function. This is currently done only when
the target is arm64 and the optimization level is higher than -O0.
- ARC optimizer temporarily emits retainRV/claimRV calls after the calls
with the operand bundle in the IR and removes the inserted calls after
processing the function.
- ARC contract pass emits retainRV/claimRV calls after the call with the
operand bundle. It doesn't remove the operand bundle on the call since
the backend needs it to emit the marker instruction. The retainRV and
claimRV calls are emitted late in the pipeline to prevent optimization
passes from transforming the IR in a way that makes it harder for the
ARC middle-end passes to figure out the def-use relationship between
the call and the retainRV/claimRV calls (which is the cause of
PR31925).
- The function inliner removes an autoreleaseRV call in the callee if
nothing in the callee prevents it from being paired up with the
retainRV/claimRV call in the caller. It then inserts a release call if
the call is annotated with claimRV since autoreleaseRV+claimRV is
equivalent to a release. If it cannot find an autoreleaseRV call, it
tries to transfer the operand bundle to a function call in the callee.
This is important since ARC optimizer can remove the autoreleaseRV
returning the callee result, which makes it impossible to pair it up
with the retainRV/claimRV call in the caller. If that fails, it simply
emits a retain call in the IR if the implicit call is a call to
retainRV and does nothing if it's a call to claimRV.
Future work:
- Use the operand bundle on x86-64.
- Fix the auto upgrader to convert call+retainRV/claimRV pairs into
calls annotated with the operand bundles.
rdar://71443534
Differential Revision: https://reviews.llvm.org/D92808
emitting retainRV or claimRV calls in the IR
Background:
This patch makes changes to the front-end and middle-end that are
needed to fix a longstanding problem where llvm breaks ARC's autorelease
optimization (see the link below) by separating calls from the marker
instructions or retainRV/claimRV calls. The backend changes are in
https://reviews.llvm.org/D92569.
https://clang.llvm.org/docs/AutomaticReferenceCounting.html#arc-runtime-objc-autoreleasereturnvalue
What this patch does to fix the problem:
- The front-end adds operand bundle "clang.arc.rv" to calls, which
indicates the call is implicitly followed by a marker instruction and
an implicit retainRV/claimRV call that consumes the call result. In
addition, it emits a call to @llvm.objc.clang.arc.noop.use, which
consumes the call result, to prevent the middle-end passes from changing
the return type of the called function. This is currently done only when
the target is arm64 and the optimization level is higher than -O0.
- ARC optimizer temporarily emits retainRV/claimRV calls after the calls
with the operand bundle in the IR and removes the inserted calls after
processing the function.
- ARC contract pass emits retainRV/claimRV calls after the call with the
operand bundle. It doesn't remove the operand bundle on the call since
the backend needs it to emit the marker instruction. The retainRV and
claimRV calls are emitted late in the pipeline to prevent optimization
passes from transforming the IR in a way that makes it harder for the
ARC middle-end passes to figure out the def-use relationship between
the call and the retainRV/claimRV calls (which is the cause of
PR31925).
- The function inliner removes an autoreleaseRV call in the callee if
nothing in the callee prevents it from being paired up with the
retainRV/claimRV call in the caller. It then inserts a release call if
the call is annotated with claimRV since autoreleaseRV+claimRV is
equivalent to a release. If it cannot find an autoreleaseRV call, it
tries to transfer the operand bundle to a function call in the callee.
This is important since ARC optimizer can remove the autoreleaseRV
returning the callee result, which makes it impossible to pair it up
with the retainRV/claimRV call in the caller. If that fails, it simply
emits a retain call in the IR if the implicit call is a call to
retainRV and does nothing if it's a call to claimRV.
Future work:
- Use the operand bundle on x86-64.
- Fix the auto upgrader to convert call+retainRV/claimRV pairs into
calls annotated with the operand bundles.
rdar://71443534
Differential Revision: https://reviews.llvm.org/D92808
The AArch64 DAG combine added by D90945 & D91433 extends the index
of a scalable masked gather or scatter to i32 if necessary.
This patch removes the combine and instead adds shouldExtendGSIndex, which
is used by visitMaskedGather/Scatter in SelectionDAGBuilder to query whether
the index should be extended before calling getMaskedGather/Scatter.
Reviewed By: david-arm
Differential Revision: https://reviews.llvm.org/D94525
or claimRV calls in the IR
Background:
This patch makes changes to the front-end and middle-end that are
needed to fix a longstanding problem where llvm breaks ARC's autorelease
optimization (see the link below) by separating calls from the marker
instructions or retainRV/claimRV calls. The backend changes are in
https://reviews.llvm.org/D92569.
https://clang.llvm.org/docs/AutomaticReferenceCounting.html#arc-runtime-objc-autoreleasereturnvalue
What this patch does to fix the problem:
- The front-end annotates calls with attribute "clang.arc.rv"="retain"
or "clang.arc.rv"="claim", which indicates the call is implicitly
followed by a marker instruction and a retainRV/claimRV call that
consumes the call result. This is currently done only when the target
is arm64 and the optimization level is higher than -O0.
- ARC optimizer temporarily emits retainRV/claimRV calls after the
annotated calls in the IR and removes the inserted calls after
processing the function.
- ARC contract pass emits retainRV/claimRV calls after the annotated
calls. It doesn't remove the attribute on the call since the backend
needs it to emit the marker instruction. The retainRV/claimRV calls
are emitted late in the pipeline to prevent optimization passes from
transforming the IR in a way that makes it harder for the ARC
middle-end passes to figure out the def-use relationship between the
call and the retainRV/claimRV calls (which is the cause of PR31925).
- The function inliner removes the autoreleaseRV call in the callee that
returns the result if nothing in the callee prevents it from being
paired up with the calls annotated with "clang.arc.rv"="retain/claim"
in the caller. If the call is annotated with "claim", a release call
is inserted since autoreleaseRV+claimRV is equivalent to a release. If
it cannot find an autoreleaseRV call, it tries to transfer the
attributes to a function call in the callee. This is important since
ARC optimizer can remove the autoreleaseRV call returning the callee
result, which makes it impossible to pair it up with the retainRV or
claimRV call in the caller. If that fails, it simply emits a retain
call in the IR if the call is annotated with "retain" and does nothing
if it's annotated with "claim".
- This patch teaches dead argument elimination pass not to change the
return type of a function if any of the calls to the function are
annotated with attribute "clang.arc.rv". This is necessary since the
pass can incorrectly determine nothing in the IR uses the function
return, which can happen since the front-end no longer explicitly
emits retainRV/claimRV calls in the IR, and change its return type to
'void'.
Future work:
- Use the attribute on x86-64.
- Fix the auto upgrader to convert call+retainRV/claimRV pairs into
calls annotated with the attributes.
rdar://71443534
Differential Revision: https://reviews.llvm.org/D92808
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.
Add the aarch64[_be]-*-gnu_ilp32 targets to support the GNU ILP32 ABI for AArch64.
The needed codegen changes were mostly already implemented in D61259, which added support for the watchOS ILP32 ABI. The main changes are:
- Wiring up the new target to enable ILP32 codegen and MC.
- ILP32 va_list support.
- ILP32 TLSDESC relocation support.
There was existing MC support for ELF ILP32 relocations from D25159 which could be enabled by passing "-target-abi ilp32" to llvm-mc. This was changed to check for "gnu_ilp32" in the target triple instead. This shouldn't cause any issues since the existing support was slightly broken: it was generating ELF64 objects instead of the ELF32 object files expected by the GNU ILP32 toolchain.
This target has been tested by running the full rustc testsuite on a big-endian ILP32 system based on the GCC ILP32 toolchain.
Reviewed By: kristof.beyls
Differential Revision: https://reviews.llvm.org/D94143
In most cases, the dup(*ext) pattern can be rearranged to perform
the extension on the vector side, allowing for further vector-specific
optimisations to be made. However the initial checks for this conversion
were insufficient, allowing invalid encodings to be attempted (causing
compilation to fail).
Differential Revision: https://reviews.llvm.org/D94778
In order to limit the number of combinations of REINTERPRET_CAST,
whilst at the same time prevent overlap with BITCAST, this patch
establishes the following rules:
1. The operand and result element types must be the same.
2. The operand and/or result type must be an unpacked type.
Differential Revision: https://reviews.llvm.org/D94593
This reverts commit dda60035e9.
This commit caused failures to compile some sources, erroring out
with "error in backend: Cannot select: t85: v2i32 = AArch64ISD::DUP t15",
see https://reviews.llvm.org/D91271 for the full reproduction case.
Following on from D91255, this patch is responsible for sinking relevant mul
operands to the same block so that umull/smull instructions can be correctly
generated by the mul combine implemented in the aforementioned patch.
Differential revision: https://reviews.llvm.org/D91271
Changes in this patch:
- When lowering floating-point masked gathers, cast the result of the
gather back to the original type with reinterpret_cast before returning.
- Added patterns for reinterpret_casts from integer to floating point, and
concat_vector patterns for bfloat16.
- Tests for various legalisation scenarios with floating point types.
Reviewed By: sdesmalen, david-arm
Differential Revision: https://reviews.llvm.org/D94171
Fixes a crash caused by D91255, when LLVMTy is null when
calling changeExtendedVectorElementType.
Differential Revision: https://reviews.llvm.org/D94234
Performing this rearrangement allows for existing patterns
to match cases where the vector may be built after an extend,
instead of before.
Differential Revision: https://reviews.llvm.org/D91255
Demanded bits may turn a sext or zext into an anyext if the top bits are
not needed. This currently prevents the lowering to instructions like
mull, addl and addw. This patch fixes the mull generation by keeping it
simple and treating them like zextends.
Differential Revision: https://reviews.llvm.org/D93832
CTLZ and CTPOP are lowered to CLZ and CNT instructions respectively.
CTTZ is not a native SVE operation but is instead lowered to:
CTTZ(V) => CTLZ(BITREVERSE(V))
In the case of fixed-length support using SVE we also lower CTTZ
operating on NEON sized vectors because of its reliance on
BITREVERSE which is also lowered to SVE intructions at these lengths.
Differential Revision: https://reviews.llvm.org/D93607
If neon is disabled, LowerCTPOP will return SDValue() to indicate
that normal legalization should be used. However, ReplaceNodeResults
does not check for this and pushes the empty SDValue() onto the
result vector, which will subsequently result in a crash.
Differential Revision: https://reviews.llvm.org/D93825
These operations are lowered to RBIT and REVB instructions
respectively. In the case of fixed-length support using SVE we
also lower BITREVERSE operating on NEON sized vectors as this
results in fewer instructions.
Differential Revision: https://reviews.llvm.org/D93606
This patch extends LowerMGATHER/MSCATTER to make use of the vector + reg/immediate
addressing modes for scalable masked gathers & scatters.
selectGatherScatterAddrMode checks if the base pointer is null, in which case
we can swap the base pointer and the index, e.g.
getelementptr nullptr, <vscale x N x T> (splat(%offset)) + %indices)
-> getelementptr %offset, <vscale x N x T> %indices
Reviewed By: david-arm
Differential Revision: https://reviews.llvm.org/D93132
X86 and AArch64 expand it as libcall inside the target. And PowerPC also
want to expand them as libcall for P8. So, propose an implement in the
legalizer to common the logic and remove the code for X86/AArch64 to
avoid the duplicate code.
Reviewed By: Craig Topper
Differential Revision: https://reviews.llvm.org/D91331
AddPromotedToType is being used to legalise INT_TO_FP operations
when the source is a predicate. The point where this introduces
vector extends might cause problems in the future so this patch
falls back to manual promotion within custom lowering.
Differential Revision: https://reviews.llvm.org/D90093
Changes in this patch:
- Minor changes to the LowerVECREDUCE_SEQ_FADD function added by @cameron.mcinally
to also work for scalable types
- Added TableGen patterns for FP reductions with unpacked types (nxv2f16, nxv4f16 & nxv2f32)
- Asserts added to expandFMINNUM_FMAXNUM & expandVecReduceSeq for scalable types
Reviewed By: cameron.mcinally
Differential Revision: https://reviews.llvm.org/D93050
This recommits a87fccb3ff with a fix to mark the destination operand
of the marker instruction as def, to fix a machine verifier failure.
This reverts the revert commit c0f2cea7c0.
This patch adds support for lowering function calls with the
rv_marker attribute. The goal is to expand such calls to the
following sequence of instructions:
BL @fn
mov x29, x29
This sequence of instructions triggers Objective-C runtime optimizations,
hence we want to ensure no instructions get moved in between them.
This patch achieves that by adding a new CALL_RVMARKER ISD node,
which gets turned into the BLR_RVMARKER pseudo, which eventually gets
expanded into the sequence mentioned above. The sequence is then marked
as instruction bundle, to avoid anything being moved in between.
@ahatanak is working on using this attribute in the front- & middle-end.
Together with the front- & middle-end changes, this should address
PR31925 for AArch64.
Reviewed By: t.p.northover
Differential Revision: https://reviews.llvm.org/D92569
This patch changes performMSCATTERCombine to also promote the indices of
masked gathers where the element type is i8 or i16, and adds various tests
for gathers with illegal types.
Reviewed By: sdesmalen
Differential Revision: https://reviews.llvm.org/D91433
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
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
All the crashes found compiling inline assembly are fixed in this
patch by changing AArch64TargetLowering::getRegForInlineAsmConstraint
to be more resilient to mismatched value and register types. For
example, it makes no sense to request a predicate register for
a nxv2i64 type and so on.
Tests have been added here:
test/CodeGen/AArch64/inline-asm-constraints-bad-sve.ll
Differential Revision: https://reviews.llvm.org/D92554
Sometimes people get minimal crash reports after a UBSAN incident. This change
tags each trap with an integer representing the kind of failure encountered,
which can aid in tracking down the root cause of the problem.
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
Lowers the llvm.masked.gather intrinsics (scalar plus vector addressing mode only)
Changes in this patch:
- Add custom lowering for MGATHER, using getGatherVecOpcode() to choose the appropriate
gather load opcode to use.
- Improve codegen with refineIndexType/refineUniformBase, added in D90942
- Tests added for gather loads with 32 & 64-bit scaled & unscaled offsets.
Reviewed By: sdesmalen
Differential Revision: https://reviews.llvm.org/D91092
Instead of trying to pattern match the code produced by ISD::ABS expansion, just custom legalize ISD::ABS to the desired sequence.
The one test change is because a DAG combine for (neg (abs)) is no longer firing because ISD::ABS is now Custom instead of Expand.
Differential Revision: https://reviews.llvm.org/D92154
This patch adds a target-specific DAG combine for mscatter to promote indices
with element types i8 or i16 before legalisation, plus various tests with illegal types.
Reviewed By: sdesmalen
Differential Revision: https://reviews.llvm.org/D90945
This patch implements out of line atomics for LSE deployment
mechanism. Details how it works can be found in llvm/docs/Atomics.rst
Options -moutline-atomics and -mno-outline-atomics to enable and disable it
were added to clang driver. This is clang and llvm part of out-of-line atomics
interface, library part is already supported by libgcc. Compiler-rt
support is provided in separate patch.
Differential Revision: https://reviews.llvm.org/D91157
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
When passing SVE types as arguments to function calls we can run
out of hardware SVE registers. This is normally fine, since we
switch to an indirect mode where we pass a pointer to a SVE stack
object in a GPR. However, if we switch over part-way through
processing a SVE tuple then part of it will be in registers and
the other part will be on the stack.
I've fixed this by ensuring that:
1. When we don't have enough registers to allocate the whole block
we mark any remaining SVE registers temporarily as allocated.
2. We temporarily remove the InConsecutiveRegs flags from the last
tuple part argument and reinvoke the autogenerated calling
convention handler. Doing this prevents the code from entering
an infinite recursion and, in combination with 1), ensures we
switch over to the Indirect mode.
3. After allocating a GPR register for the pointer to the tuple we
then deallocate any SVE registers we marked as allocated in 1).
We also set the InConsecutiveRegs flags back how they were before.
4. I've changed the AArch64ISelLowering LowerCALL and
LowerFormalArguments functions to detect the start of a tuple,
which involves allocating a single stack object and doing the
correct numbers of legal loads and stores.
Differential Revision: https://reviews.llvm.org/D90219
We have a frequent pattern where we're merging two KnownBits to get the common/shared bits, and I just fell for the gotcha where I tried to use the & operator to merge them........
Lowers the llvm.masked.scatter intrinsics (scalar plus vector addressing mode only)
Changes included in this patch:
- Custom lowering for MSCATTER, which chooses the appropriate scatter store opcode to use.
Floating-point scatters are cast to integer, with patterns added to match FP reinterpret_casts.
- Added the getCanonicalIndexType function to convert redundant addressing
modes (e.g. scaling is redundant when accessing bytes)
- Tests with 32 & 64-bit scaled & unscaled offsets
Reviewed By: sdesmalen
Differential Revision: https://reviews.llvm.org/D90941
For example if the sign extension is only used in for TBZ, and the value is used elsewhere with a zero extension, this can eliminate a sign extension.
Reviewed By: samparker
Differential Revision: https://reviews.llvm.org/D90606
Silence warning Undefined Behavior Sanitzer warning:
runtime error: negation of -9223372036854775808 cannot be represented in type 'int64_t' (aka 'long'); cast to an unsigned type to negate this value to itself
Reviewed By: paulwalker-arm
Differential Revision: https://reviews.llvm.org/D90710
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
We don't currently support passing unnamed variadic SVE arguments
so I've added a fatal error if we hit such cases to prevent any
silent ABI issues in future.
Differential Revision: https://reviews.llvm.org/D90230
If most elements of BUILD_VECTOR are the same, with a few different
elements, it is better to use DUP for the common elements and
INSERT_VECTOR_ELT for the different elements.
Currently this transform is guarded quite restrictively to only trigger
in clearly beneficial cases.
With D90176, the lowering for patterns originating from code like
` float32x4_t y = {a,a,a,0};` (common in 3D apps) are lowered even
better (unnecessary fmov is removed).
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D90233
vnot (xor -1) should be equivalent to the AArch64 specific AArch64ISD::NOT
node, but allow more folding thanks to all the target independent
optimizations. Specifically this allows select(icmp ne, x, y) to
become "cmeq; bsl y, x" as opposed to needing to convert the predicate
with "cmeq; mvn; bsl x, y"
Unfortunately there is a regression in a cmtst test, but the code it
selected from was already non-canonical, with instcombine preferring to
use an eq predicate instead. Plus the more common case of icmp ne is
improved.
Differential Revision: https://reviews.llvm.org/D90126
In many places in the AArch64 backend we are comparing TypeSize objects,
but in fact we are only ever expecting fixed width types. I've changed
all such comparisons to use their integer equivalents by replacing
calls to getSizeInBits() with getFixedSizeInBits(), etc.
Differential Revision: https://reviews.llvm.org/D89116
When passing SVE types as arguments to function calls we can run
out of hardware SVE registers. This is normally fine, since we
switch to an indirect mode where we pass a pointer to a SVE stack
object in a GPR. However, if we switch over part-way through
processing a SVE tuple then part of it will be in registers and
the other part will be on the stack. This is wrong and we'd like
to avoid any silent ABI compatibility issues in future. For now,
I've added a fatal error when this happens until we can get a
proper fix.
Differential Revision: https://reviews.llvm.org/D89326
This patch adds FP_EXTEND_MERGE_PASSTHRU & FP_ROUND_MERGE_PASSTHRU
ISD nodes, used to lower scalable vector fp_extend/fp_round operations.
fp_round has an additional argument, the 'trunc' flag, which is an integer of zero or one.
This also fixes a warning introduced by the new tests added to sve-split-fcvt.ll,
resulting from an implicit TypeSize -> uint64_t cast in SplitVecOp_FP_ROUND.
Reviewed By: sdesmalen, paulwalker-arm
Differential Revision: https://reviews.llvm.org/D88321
Splitting the operand of a scalable [S|U]INT_TO_FP results in a
concat_vectors operation where the operands are unpacked FP
scalable vectors (e.g. nxv2f32).
This patch adds custom lowering of concat_vectors which
checks that the number of operands is 2, and isel patterns
to match concat_vectors of scalable FP types with uzp1.
Reviewed By: efriedma, paulwalker-arm
Differential Revision: https://reviews.llvm.org/D88033
Iterating across all of integer_scalable_vector_valuetypes seems
wasteful when there's only a handful we care about.
Also removes some rouge whitespace.
Differential Revision: https://reviews.llvm.org/D88552
Essentially the same as the signed variants from D88259. Also includes a clean up of the lowering function.
Differential Revision: https://reviews.llvm.org/D88317
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
This patch is pretty similar to the VECREDUCE_ADD patch, with some minor tweaks.
Results from the AArch64ISD::[SMAX|SMIN]V_PRED return element sized results. This requires an ANY_EXTEND for results < 32-bits, since Legalization promotes those results.
There is no NEON i64 vector support for SMAXV|SMINV, so use SVE for those.
Differential Revision: https://reviews.llvm.org/D88259
This change adds the support for __builtin_return_address
for ARMv8.3A Pointer Authentication.
Location of the authentication code in the pointer depends on
the system configuration, therefore a dedicated instruction is used for
effectively removing the authentication code without
authenticating the pointer.
Reviewed By: chill
Differential Revision: https://reviews.llvm.org/D75044
With the exception of VECREDUCE_ADD, there are no NEON instructions to support vector of i64 reductions. This patch removes the Custom lowerings for those and adds some test coverage to confirm.
Differential Revision: https://reviews.llvm.org/D88161
This patch adds new ISD nodes, SCVTZ_MERGE_PASSTHRU &
UCVTZ_MERGE_PASSTHRU, which are used to lower both legal
scalable vector [S|U]INT_TO_FP operations and the following intrinsics:
- llvm.aarch64.sve.scvtf
- llvm.aarch64.sve.ucvtf
Reviewed By: sdesmalen, efriedma
Differential Revision: https://reviews.llvm.org/D87913
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
The current nodes, AArch64::SMAXV_PRED for example, are defined to
return a NEON vector result. This is incorrect because they modify
the complete SVE register and are thus changed to represent such.
This patch also adds nodes for UADDV_PRED and SADDV_PRED, which
unifies the handling of all SVE reductions.
NOTE: Floating-point reductions are already implemented correctly,
so this patch is essentially making everything consistent with those.
Differential Revision: https://reviews.llvm.org/D87843
This turns all jump table entries into deltas within the target
function because in the small memory model all code & static data must
be in a 4GB block somewhere in memory.
When the entries were a delta between the table location and a basic
block, the 32-bit signed entries are not enough to guarantee
reachability.
https://reviews.llvm.org/D87286
D75689 turns the faddp pattern into a shuffle with vector add.
Match this new pattern in target-specific DAG combine, rather than ISel,
because legalization (for v2f32) turns it into a bit of a mess.
- extended to cover f16, f32, f64 and i64
This patch adds new ISD nodes, FCVTZS_MERGE_PASSTHRU &
FCVTZU_MERGE_PASSTHRU, which are used to lower scalable vector
FP_TO_SINT/FP_TO_UINT operations and the following intrinsics:
- llvm.aarch64.sve.fcvtzu
- llvm.aarch64.sve.fcvtzs
Reviewed By: efriedma, paulwalker-arm
Differential Revision: https://reviews.llvm.org/D87232
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
As discussed on llvm-dev:
http://lists.llvm.org/pipermail/llvm-dev/2020-April/140729.html
This is hopefully the final remaining showstopper before we can remove
the 'experimental' from the reduction intrinsics.
No behavior was specified for the FP min/max reductions, so we have a
mess of different interpretations.
There are a few potential options for the semantics of these max/min ops.
I think this is the simplest based on current behavior/implementation:
make the reductions inherit from the existing llvm.maxnum/minnum intrinsics.
These correspond to libm fmax/fmin, and those are similar to the (now
deprecated?) IEEE-754 maxNum/minNum functions (NaNs are treated as missing
data). So the default expansion creates calls to libm functions.
Another option would be to inherit from llvm.maximum/minimum (NaNs propagate),
but most targets just crash in codegen when given those nodes because no
default expansion was ever implemented AFAICT.
We could also just assume 'nnan' semantics by default (we are already
assuming 'nsz' semantics in the maxnum/minnum intrinsics), but some targets
(AArch64, PowerPC) support the more defined behavior, so it doesn't make much
sense to not allow a tighter spec. Fast-math-flags (nnan) can be used to
loosen the semantics.
(Note that D67507 was proposed to update the LangRef to acknowledge the more
recent IEEE-754 2019 standard, but that patch seems to have stalled. If we do
update based on the new standard, the reduction instructions can seamlessly
inherit from whatever updates are made to the max/min intrinsics.)
x86 sees a regression here on 'nnan' tests because we have underlying,
longstanding bugs in FMF creation/propagation. Those need to be fixed apart
from this change (for example: https://llvm.org/PR35538). The expansion
sequence before this patch may not have been correct.
Differential Revision: https://reviews.llvm.org/D87391
Truncating from an illegal SVE type to a legal type, e.g.
`trunc <vscale x 4 x i64> %in to <vscale x 4 x i32>`
fails after PromoteIntOp_CONCAT_VECTORS attempts to
create a BUILD_VECTOR.
This patch changes the promote function to create a sequence of
INSERT_SUBVECTORs if the return type is scalable, and replaces
these with UNPK+UZP1 for AArch64.
Reviewed By: paulwalker-arm
Differential Revision: https://reviews.llvm.org/D86548
Add the functionality to lower SVE rounding operations for passthru variant.
Created a new test case file for all rounding operations.
Reviewed By: paulwalker-arm
Differential Revision: https://reviews.llvm.org/D86793
This patch changes ElementCount so that the Min and Scalable
members are now private and can only be accessed via the get
functions getKnownMinValue() and isScalable(). In addition I've
added some other member functions for more commonly used operations.
Hopefully this makes the class more useful and will reduce the
need for calling getKnownMinValue().
Differential Revision: https://reviews.llvm.org/D86065
Previously in addTypeForNeon, we would set the operations for bfloat vectors
like other generic types. But as bfloat is a storage-only type a number of
operations shouldn't be set. This patch fixes that.
Reviewed By: dmgreen
Differential Revision: https://reviews.llvm.org/D85101
This patch adds code to recognize vector shuffles which can be
represented as VDUP (splat) of a vector lane with of a different
(wider) type than the original vector lane type.
For example:
shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 0, i32 1, i32 0, i32 1>
is essentially:
shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 0, i32 0>
Such patterns are generated by the SelectionDAG machinery in some cases
(see DAGCombiner::visitBITCAST in DAGCombiner.cpp, the "Remove double
bitcasts from shuffles" part).
Reviewed By: dmgreen
Differential Revision: https://reviews.llvm.org/D86225
Also updates isConstOrConstSplatFP to allow the mul(A,-1) -> neg(A)
transformation when -1 is expressed as an ISD::SPLAT_VECTOR.
Differential Revision: https://reviews.llvm.org/D86415
There are no nxv16i8/nxv8i16 SDIV instructions, so these fixed width operations must be promoted to nxv4i32.
Differential Revision: https://reviews.llvm.org/D86114
This isn't necessaary for ACLE, but could be useful in other situations.
And the change is simple.
Differential Revision: https://reviews.llvm.org/D85251
Testing is performed when targeting 128, 256 and 512-bit wide vectors.
For 128-bit vectors, the original behavior of using NEON instructions is
preserved.
Differential Revision: https://reviews.llvm.org/D85479
In this patch I have fixed two issues:
1. Our SVE tuple get/set intrinsics were using the wrong constant type
for the index passed to EXTRACT_SUBVECTOR. I have fixed this by using the
function SelectionDAG::getVectorIdxConstant to create the value. Also, I
have updated the documentation for EXTRACT_SUBVECTOR describing what type
the constant index should be and we now enforce this when creating the
node.
2. The AArch64 backend was missing the appropriate patterns for
extracting certain subvectors (nxv4f16 and nxv2f32) from legal SVE types.
I have added them as part of this patch.
The only way that I could find to test the new patterns was to use the
SVE tuple get intrinsics, although I realise it looks a bit unusual.
Tests added here:
test/CodeGen/AArch64/sve-extract-subvector.ll
Differential Revision: https://reviews.llvm.org/D85516
These are useful instructions when lowering fixed length vector
extends, so I've broken this patch out as kind of NFC like work.
Differential Revision: https://reviews.llvm.org/D85546
This allows us to remove extra patterns from AArch64SVEInstrInfo.td
because we can reuse those required for fixed length vectors.
Differential Revision: https://reviews.llvm.org/D85328
NOTE: Also uses SVE code generation for NEON size vectors, instead
of expanding i64 based vector multiplications.
Differential Revision: https://reviews.llvm.org/D85327
Since there are no ill effects when performing these operations
with undefined elements, they are lowered to the already supported
unpredicated scalable vector equivalents.
Differential Revision: https://reviews.llvm.org/D85117
This fixes an issue triggered by the following code, where emitEpilogue
got confused when trying to restore the SVE registers after the call,
whereas the call to bar() is implemented as a TCReturn:
int non_sve();
int sve(svint32_t x) { return non_sve(); }
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D84869
The SVE instruction set only supports sdiv/udiv for 32-bit and 64-bit
integers. If we see an 8-bit or 16-bit divide, widen the operands to 32
bits, and narrow the result.
Differential Revision: https://reviews.llvm.org/D85170
This is the final bit of work to relax the register allocation
requirements when code generating normal LLVM IR, which rarely
care about the result of inactive lanes. By using _PRED nodes
we can make better use of SVE's reversed instructions.
Also removes a redundant parameter from the min/max tests.
Differential Revision: https://reviews.llvm.org/D85142
When building code at -O0 We weren't falling back to DAG ISel correctly
when encountering alloca instructions with scalable vector types. This
is because the alloca has no operands that are scalable. I've fixed this by
adding a check in AArch64ISelLowering::fallBackToDAGISel for alloca
instructions with scalable types.
Differential Revision: https://reviews.llvm.org/D84746
dacf8d3 added support for most fcmp operations, but there are some extra
variations I hadn't considered: SelectionDAG supports float comparisons
that are neither ordered nor unordered. Add support for the missing
operations.
Differential Revision: https://reviews.llvm.org/D84460
Summary:
Teach LLVM to recognize the above pattern, where the operands are
either signed or unsigned types.
Subscribers: kristof.beyls, hiraditya, danielkiss, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D83777
This isn't a natively supported operation, so convert it to a
mask+compare.
In addition to the operation itself, fix up some surrounding stuff to
make the testcase work: we need concat_vectors on i1 vectors, we need
legalization of i1 vector truncates, and we need to fix up all the
relevant uses of getVectorNumElements().
Differential Revision: https://reviews.llvm.org/D83811
It's useful for a debugger to be able to distinguish an @llvm.debugtrap
from a (noreturn) @llvm.trap, so this extends the existing Windows
behaviour to other platforms.
Lower the operations to predicated variants. This is prep work
required for fixed length code generation but also fixes a bug
whereby these operations fail selection when "unpacked" vector
types (e.g. MVT::nxv2f32) are used.
This patch also adds the missing "unpacked" patterns for FMA.
Differential Revision: https://reviews.llvm.org/D83765
This is currently bare-bones; we aren't taking advantage of any of the
FMA variant instructions. But it's enough to at least generate
code.
Differential Revision: https://reviews.llvm.org/D83444
Fixed length vector code generation for SVE does not yet custom
lower BUILD_VECTOR and instead relies on expansion. At the same
time custom lowering for VECTOR_SHUFFLE is also not available so
this patch updates isShuffleMaskLegal to reject vector types that
require SVE.
Related to this it also prevents the merging of stores after
legalisation because this only works when BUILD_VECTOR is either
legal or can be elminated. When this is not the case the code
generator enters an infinite legalisation loop.
Differential Revision: https://reviews.llvm.org/D83408
We use extact_subvector and insert_subvector to "cast" between
fixed length and scalable vectors. This patch adds custom c++
based ISel for the following cases:
fixed_vector = ISD::EXTRACT_SUBVECTOR scalable_vector, 0
scalable_vector = ISD::INSERT_SUBVECTOR undef(scalable_vector), fixed_vector, 0
Which result in either EXTRACT_SUBREG/INSERT_SUBREG for NEON sized
vectors or COPY_TO_REGCLASS otherwise.
Differential Revision: https://reviews.llvm.org/D82871
Summary:
Teach LLVM to recognize the above pattern, which is usually a
transformation of (a + b + 1) >> 1, where the operands are either
signed or unsigned types.
Subscribers: kristof.beyls, hiraditya, danielkiss, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D82669
There was a rogue 'assert' in AArch64ISelLowering for the tuple.get intrinsics,
that shouldn't really have been there (I suspect this was a remnant from when
we expected the wider vector always to have come from a vector CONCAT).
When I tried to create a more minimal reproducer, I found a bug in
DAGCombiner where it drops the scalable flag when trying to fold:
extract_subv (bitcast X), Index --> bitcast (extract_subv X, Index')
This patch fixes both issues.
Reviewers: david-arm, efriedma, spatel
Reviewed By: efriedma
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D82910
As per documentation of `hasPairLoad`:
"`RequiredAlignment` gives the minimal alignment constraints that must be met to be able to select this paired load."
In this sense, `0` is strictly equivalent to `1`. We make this obvious by using `Align` instead of unsigned.
There is only one implementor of this interface.
Differential Revision: https://reviews.llvm.org/D82958
We currently lower SDIV to SDIV_MERGE_OP1. This forces the value
for inactive lanes in a way that can hamper register allocation,
however, the lowering has no requirement for inactive lanes.
Instead this patch replaces SDIV_MERGE_OP1 with SDIV_PRED thus
freeing the register allocator. Once done the only user of
SDIV_MERGE_OP1 is intrinsic lowering so I've removed the node
and perform ISel on the intrinsic directly. This also allows
us to implement MOVPRFX based zeroing in the same manner as SUB.
This patch also renames UDIV_MERGE_OP1 and [F]ADD_MERGE_OP1 for
the same reason but in the ADD cases the ISel code is already
as required.
Differential Revision: https://reviews.llvm.org/D82783
This patch proposes a naming convention for operations that take
a general predicate (and are thus predicated) that specifies
what happens to the false lanes.
Currently the _PRED suffix is used, which doesn't really say much other
than that it takes a predicate. In some instances this means it has
merging predication and in other cases it means zeroing-predication.
This patch also changes the order of operands to
AArch64ISD::DUP_MERGE_PASSTHRU, to pass the predicate as the first
operand, which is in line with all other predicates nodes. It takes the
passthru value as an explicit passthru value, which is always passed as
the last operand.
Reviewers: paulwalker-arm, cameron.mcinally, eli.friedman, dancgr, efriedma
Reviewed By: paulwalker-arm
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D81850
Summary:
performPostLD1Combine will introduce either a LD1LANEpost
or LD1DUPpost node, which will cause selection failure if the
return type is a scalable vector.
Reviewers: sdesmalen, c-rhodes, efriedma
Reviewed By: efriedma
Subscribers: tschuett, kristof.beyls, hiraditya, rkruppe, psnobl, danielkiss, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D82670
Pre-commit for D82257, this adds a DemandedElts arg to ShrinkDemandedConstant/targetShrinkDemandedConstant which will allow future patches to (optionally) add vector support.
Remove the asserts in performLDNT1Combine & performST[NT]1Combine
to ensure we get a failure where the type is a bfloat16 and
hasBF16() is false, regardless of whether asserts are enabled.
Implement them on top of sdiv/udiv, similar to what we do for integer
types.
Potential future work: implementing i8/i16 srem/urem, optimizations for
constant divisors, optimizing the mul+sub to mls.
Differential Revision: https://reviews.llvm.org/D81511
Summary:
This patch adds base support for code generating fixed length
vector operations targeting a known SVE vector length. To achieve
this we lower fixed length vector operations to equivalent scalable
vector operations, whereby SVE predication is used to limit the
elements processed to those present within the fixed length vector.
Specifically this patch implements load and store operations, which
get lowered to their masked counterparts thusly:
V = load(Addr) =>
V = extract_fixed_vector(masked_load(make_pred(V.NumElts), Addr))
store(V, (Addr)) =>
masked_store(insert_fixed_vector(V), make_pred(V.NumElts), Addr))
Reviewers: rengolin, efriedma
Subscribers: tschuett, hiraditya, rkruppe, psnobl, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D80385
At the moment we use Global ISel by default at -O0, however it is
currently not capable of dealing with scalable vectors for two
reasons:
1. The register banks know nothing about SVE registers.
2. The LLT (Low Level Type) class knows nothing about scalable
vectors.
For now, the easiest way to avoid users hitting issues when using
the SVE ACLE is to fall back on normal DAG ISel when encountering
instructions that operate on scalable vector types.
I've added a couple of RUN lines to existing SVE tests to ensure
we can compile at -O0. I've also added some new tests to
CodeGen/AArch64/GlobalISel/arm64-fallback.ll
that demonstrate we correctly fallback to DAG ISel at -O0 when
lowering formal arguments or translating instructions that involve
scalable vector types.
Differential Revision: https://reviews.llvm.org/D81557
Adds aarch64-sve-vector-bits-{min,max} to allow the size of SVE
data registers (in bits) to be specified. This allows the code
generator to make assumptions it normally couldn't. As a starting
point this information is used to mark fixed length vector types
that can fit within the specified size as legal.
Reviewers: rengolin, efriedma
Subscribers: tschuett, kristof.beyls, hiraditya, rkruppe, psnobl, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D80384
When checking for an enum function attribute, use hasFnAttribute()
rather than hasAttribute() at FunctionIndex, because it is
significantly faster (and more concise to boot).
Summary:
Note to downstream target maintainers: this might silently change the semantics of your code if you override `TargetLowering::allowsMisalignedMemoryAccesses` without marking it override.
This patch is part of a series to introduce an Alignment type.
See this thread for context: http://lists.llvm.org/pipermail/llvm-dev/2019-July/133851.html
See this patch for the introduction of the type: https://reviews.llvm.org/D64790
Reviewers: courbet
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D81374
Summary:
This patch adds initial support for the following instrinsics:
* llvm.aarch64.sve.ld2
* llvm.aarch64.sve.ld3
* llvm.aarch64.sve.ld4
For loading two, three and four vectors worth of data. Basic codegen is
implemented with reg+reg and reg+imm addressing modes being addressed
in a later patch.
The types returned by these intrinsics have a number of elements that is a
multiple of the elements in a 128-bit vector for a given type and N, where N is
the number of vectors being loaded, i.e. 2, 3 or 4. Thus, for 32-bit elements
the types are:
LD2 : <vscale x 8 x i32>
LD3 : <vscale x 12 x i32>
LD4 : <vscale x 16 x i32>
This is implemented with target-specific intrinsics for each variant that take
the same operands as the IR intrinsic but return N values, where the type of
each value is a full vector, i.e. <vscale x 4 x i32> in the above example.
These values are then concatenated using the standard concat_vector intrinsic
to maintain type legality with the IR.
These intrinsics are intended for use in the Arm C Language
Extension (ACLE).
Reviewed By: sdesmalen
Differential Revision: https://reviews.llvm.org/D75751
The code for trying to split up stores is designed for NEON vectors,
where we support arbitrary alignments. It's an optimisation designed
to improve performance by using smaller, aligned stores. However,
we currently only support 16 byte alignments for SVE vectors anyway
so we may as well bail out early.
This change fixes up remaining warnings in a couple of tests:
CodeGen/AArch64/sve-callbyref-notailcall.ll
CodeGen/AArch64/sve-calling-convention-byref.ll
Differential Revision: https://reviews.llvm.org/D80720
When the input to a wide compare instruction is a DUP or SPLAT_VECTOR
node we should deal with cases where the DUP/SPLAT_VECTOR input
operand is not an immediate value. I've fixed the code to return
SDValue() in such cases and added a couple of tests - one each to
represent the signed and unsigned cases.
Differential Revision: https://reviews.llvm.org/D81167
Summary:
This patch adds the following intrinsics for creating two-tuple,
three-tuple and four-tuple scalable vectors:
* llvm.aarch64.sve.tuple.create2
* llvm.aarch64.sve.tuple.create3
* llvm.aarch64.sve.tuple.create4
As well as:
* llvm.aarch64.sve.tuple.get
* llvm.aarch64.sve.tuple.set
For extracting and inserting scalable vectors from vector tuples. These
intrinsics are intended to be used by the ACLE functions svcreate<n>,
svget and svset.
This patch also includes calling convention support for passing and
returning tuples of scalable vectors to/from functions.
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D75674
Summary:
This patch adds legalisation of extensions where the operand
of the extend is a legal scalable type but the result is not.
EXTRACT_SUBVECTOR is used to split the result, before
being replaced by target-specific [S|U]UNPK[HI|LO] operations.
For example:
```
zext <vscale x 16 x i8> %a to <vscale x 16 x i16>
```
should emit:
```
uunpklo z2.h, z0.b
uunpkhi z1.h, z0.b
```
Reviewers: sdesmalen, efriedma, david-arm
Reviewed By: efriedma
Subscribers: tschuett, hiraditya, rkruppe, psnobl, huihuiz, cfe-commits, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D79587
The concept of G_GLOBAL_VALUE is nice and simple, but always using it as the
representation for global var addressing until selection time creates some
problems in optimizing accesses in certain code/relocation models.
The problem comes from trying to optimize adrp -> add -> load/store sequences
in the most common "small" code model. These accesses can be optimized into an
adrp -> load with the add offset being folded into the load's immediate field.
If we try to keep all global var references as a single generic instruction
then by the time we get to the complex operand trying to match these, we end up
generating an adrp at the point of use. The real issue here is that we don't
have any form of CSE during selection, so the code size will bloat from many
redundant adrp's.
This patch custom legalizes small code mode non-GOT G_GLOBALs into target ADRP
and a new "target specific generic opcode" G_ADD_LOW. We also teach the
localizer to localize these instructions via the custom hook that was added
recently. Finally, the complex pattern for indexed loads/stores is extended to
try to fold these G_ADD_LOW instructions into the load immediate.
On -O0 CTMark, we see a 0.8% geomean code size improvement. We should also see
some minor performance improvements too.
Differential Revision: https://reviews.llvm.org/D78465
Treat it as callee-saved, and always back it up. When windows code calls
entry points in unix code, marked with the windows calling convention,
that unix code can call other functions that isn't compiled with
-ffixed-x18 which may clobber x18 freely. By backing it up and restoring
it on return, we preserve the register across the function call,
fulfilling this part of the windows calling convention on another OS.
This isn't enough for making sure that x18 is preseved when non-windows
code does a callback to windows code, but is a clear improvement over
the current status quo. Additionally, wine is nowadays building many
modules as PE DLLs, which avoids the callback issue altogether for those
DLLs.
Differential Revision: https://reviews.llvm.org/D61892
Let the codegen recognized the nomerge attribute and disable branch folding when the attribute is given
Differential Revision: https://reviews.llvm.org/D79537
When creating a new vector type based on another vector type we
should pass in the element count instead of the number of elements
and scalable flag separately.
I encountered this warning whilst compiling this test:
CodeGen/AArch64/sve-intrinsics-int-compares.ll
Differential revision: https://reviews.llvm.org/D80621
Summary:
This patch fixes a problem when pmull2 instruction is not
generated for vmull_high_p64 intrinsic.
ISel has a pattern for int_aarch64_neon_pmull64 intrinsic to generate
PMULL2 instruction. That pattern assumes that extraction operations
are located in the same basic block. We need to sink them
if they are not. Handle operands of int_aarch64_neon_pmull64
into AArch64TargetLowering::shouldSinkOperands.
Reviewed by: efriedma
Differential Revision: https://reviews.llvm.org/D80320
Summary:
The optimization has been refactored to fix certain bugs and
limitations. The condition for lowering to S[LR]I has been changed
to reflect the manual pseudocode description of SLI and SRI operation.
The optimization can now handle more cases of operand type and order.
Subscribers: kristof.beyls, hiraditya, danielkiss, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D79233
This patch stores the alignment for ConstantPoolSDNode as an
Align and updates the getConstantPool interface to take a MaybeAlign.
Removing getAlignment() will be done as a follow up.
Differential Revision: https://reviews.llvm.org/D79436
Now using patterns, since there's a single-instruction lowering. (We
could convert to VSELECT and pattern-match that, but there doesn't seem
to be much point.)
I think this might be the first instruction to use nested multiclasses
this way? It seems like a good way to reduce duplication between
different integer widths. Let me know if it seems like an improvement.
Also, while I'm here, fix the return type of SETCC so we don't try to
merge a sign-extend with a SETCC.
Differential Revision: https://reviews.llvm.org/D79193
Summary:
This patch adds AArch64ISD nodes for [S|U]MIN_PRED
and [S|U]MAX_PRED, and lowers both SVE intrinsics and
IR operations for min and max to these nodes.
There are two forms of these instructions for SVE: a predicated
form and an immediate (unpredicated) form. The patterns
which existed for the latter have been updated to match a
predicated node with an immediate and map this
to the immediate instruction.
Reviewers: sdesmalen, efriedma, dancgr, rengolin
Reviewed By: efriedma
Subscribers: huihuiz, tschuett, kristof.beyls, hiraditya, rkruppe, psnobl, cfe-commits, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D79087
Summary:
This patch implements custom floating-point reduction ISD nodes that
have vector results, which are used to lower the following intrinsics:
* llvm.aarch64.sve.fadda
* llvm.aarch64.sve.faddv
* llvm.aarch64.sve.fmaxv
* llvm.aarch64.sve.fmaxnmv
* llvm.aarch64.sve.fminv
* llvm.aarch64.sve.fminnmv
SVE reduction instructions keep their result within a vector register,
with all other bits set to zero.
Changes in this patch were implemented by Paul Walker and Sander de
Smalen.
Reviewers: sdesmalen, efriedma, rengolin
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D78723
Summary:
This patch maps IR operations for sdiv & udiv to the
@llvm.aarch64.sve.[s|u]div intrinsics.
A ptrue must be created during lowering as the div instructions
have only a predicated form.
Patch contains changes by Andrzej Warzynski.
Reviewers: sdesmalen, c-rhodes, efriedma, cameron.mcinally, rengolin
Reviewed By: efriedma
Subscribers: tschuett, kristof.beyls, hiraditya, rkruppe, psnobl, andwar, cfe-commits, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D78569
Summary:
The SVE masked load and store intrinsics introduced in D76688 rely on
common llvm.masked.load/store nodes. This patch creates new ISD nodes
for LD1(S) & ST1 to remove this dependency.
Additionally, this adds support for sign & zero extending
loads and truncating stores.
Reviewers: sdesmalen, efriedma, cameron.mcinally, c-rhodes, rengolin
Reviewed By: efriedma
Subscribers: tschuett, kristof.beyls, hiraditya, rkruppe, psnobl, danielkiss, andwar, cfe-commits, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D78204
This reverts commit 17b1869b72.
It is an attempt to fix the failure reported at
The patch differs from the original one reviwed at
https://reviews.llvm.org/D77435 only for the use of the std::make_tuple
in building the return value of `findAddrModeSVELoadStore`:
- return {IsRegReg ? Opc_rr : Opc_ri, NewBase, NewOffset};
+ return std::make_tuple(IsRegReg ? Opc_rr : Opc_ri, NewBase,
the original patch submitted at
fc4e954ed5
was failing the following build:
http://lab.llvm.org:8011/builders/clang-armv7-linux-build-cache/builds/29420/
with error:
/home/buildslave/buildslave/clang-armv7-linux-build-cache/llvm/llvm/lib/Target/AArch64/AArch64ISelDAGToDAG.cpp
/home/buildslave/buildslave/clang-armv7-linux-build-cache/llvm/llvm/lib/Target/AArch64/AArch64ISelDAGToDAG.cpp:1439:10:
error: chosen constructor is explicit in copy-initialization
return {IsRegReg ? Opc_rr : Opc_ri, NewBase, NewOffset};
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/usr/bin/../lib/gcc/arm-linux-gnueabihf/5.4.0/../../../../include/c++/5.4.0/tuple:479:19:
note: explicit constructor declared here
constexpr tuple(_UElements&&... __elements)
^
1 error generated.
Summary:
This change is fixing an issue where the dagcombine incorrectly used an addressing mode with scaled offsets (indices), instead of unscaled offsets.
Those addressing modes do not exist for `prfh` , `prfw` and `prfd`, hence we can reuse `prfb` because that has unscaled offsets, and because the pseudo-code in the XML spec suggests that the element size is not used for the amount of data that is prefetched by the instruction.
FWIW, GCC also emits a `prfb` for these cases.
Reviewers: sdesmalen, andwar, rengolin
Reviewed By: sdesmalen
Subscribers: tschuett, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D78069
Summary:
Fixed wrong conditions for generating (S[LR]I X, Y, C2) from
(or (and X, BvecC1), (lsl Y, C2)) and added ISel nodes to lower to S[LR]I. The
optimisation is also enabled by default now.
Subscribers: kristof.beyls, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D77387
Summary:
The renaming is necessary to make the naming scheme uniform with other
gather/scatter load/stores SVE intrinsics.
The naming of variables and functions have been adapted to make it
explicit whether we are dealing with a scalar offset (which is
unscaled) or an index (which is scaled according to the data type of
the lanes of the vector).
Reviewers: andwar, sdesmalen, rengolin
Reviewed By: andwar
Subscribers: tschuett, hiraditya, arphaman, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D77839
Summary:
Remove usages of asserting vector getters in Type in preparation for the
VectorType refactor. The existence of these functions complicates the
refactor while adding little value.
Reviewers: mcrosier, efriedma, sdesmalen
Reviewed By: efriedma
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D77269
Instead, represent the mask as out-of-line data in the instruction. This
should be more efficient in the places that currently use
getShuffleVector(), and paves the way for further changes to add new
shuffles for scalable vectors.
This doesn't change the syntax in textual IR. And I don't currently plan
to change the bitcode encoding in this patch, although we'll probably
need to do something once we extend shufflevector for scalable types.
I expect that once this is finished, we can then replace the raw "mask"
with something more appropriate for scalable vectors. Not sure exactly
what this looks like at the moment, but there are a few different ways
we could handle it. Maybe we could try to describe specific shuffles.
Or maybe we could define it in terms of a function to convert a fixed-length
array into an appropriate scalable vector, using a "step", or something
like that.
Differential Revision: https://reviews.llvm.org/D72467
On Darwin these need to be selected into a function call for the TLS
address lookup. As a result, they can't be moved below a physreg write,
which happens in call sequences. In the long term, we should have some
mechanism in the localizer to prevent localizing into target-specific
atomic instruction sequences.
rdar://60056248
Differential Revision: https://reviews.llvm.org/D76652
Summary:
In order to keep the names consistent with other SVE gather loads, the
intrinsics for gather prefetch are renamed as follows:
* @llvm.aarch64.sve.gather.prfb -> @llvm.aarch64.sve.prfb.gather
Reviewed by: fpetrogalli
Differential Revision: https://reviews.llvm.org/D76421
Summary:
This fixes a discrepancy between the non-temporal loads/store
intrinsics and other SVE load intrinsics (such as nf/ff), so
that Clang can use the same code to generate these intrinsics.
Reviewers: andwar, kmclaughlin, rengolin, efriedma
Reviewed By: efriedma
Subscribers: tschuett, kristof.beyls, hiraditya, rkruppe, psnobl, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D76237
Summary:
This intrinsic implements the unpredicated duplication of scalar values
and is mapped to (through ISD::SPLAT_VECTOR):
* DUP <Zd>.<T>, #<imm>
* DUP <Zd>.<T>, <R><n|SP>
Reviewed by: sdesmalen
Differential Revision: https://reviews.llvm.org/D75900
Summary:
This patch adds the following intrinsics for non-temporal gather loads
and scatter stores:
* aarch64_sve_ldnt1_gather_index
* aarch64_sve_stnt1_scatter_index
These intrinsics implement the "scalar + vector of indices" addressing
mode.
As opposed to regular and first-faulting gathers/scatters, there's no
instruction that would take indices and then scale them. Instead, the
indices for non-temporal gathers/scatters are scaled before the
intrinsics are lowered to `ldnt1` instructions.
The new ISD nodes, GLDNT1_INDEX and SSTNT1_INDEX, are only used as
placeholders so that we can easily identify the cases implemented in
this patch in performGatherLoadCombine and performScatterStoreCombined.
Once encountered, they are replaced with:
* GLDNT1_INDEX -> SPLAT_VECTOR + SHL + GLDNT1
* SSTNT1_INDEX -> SPLAT_VECTOR + SHL + SSTNT1
The patterns for lowering ISD::SHL for scalable vectors (required by
this patch) were missing, so these are added too.
Reviewed By: sdesmalen
Differential Revision: https://reviews.llvm.org/D75601
Summary:
This patch adds the following LLVM IR intrinsics for SVE:
1. non-temporal gather loads
* @llvm.aarch64.sve.ldnt1.gather
* @llvm.aarch64.sve.ldnt1.gather.uxtw
* @llvm.aarch64.sve.ldnt1.gather.scalar.offset
2. non-temporal scatter stores
* @llvm.aarch64.sve.stnt1.scatter
* @llvm.aarch64.sve.ldnt1.gather.uxtw
* @llvm.aarch64.sve.ldnt1.gather.scalar.offset
These intrinsic are mapped to the corresponding SVE instructions
(example for half-words, zero-extending):
* ldnt1h { z0.s }, p0/z, [z0.s, x0]
* stnt1h { z0.s }, p0/z, [z0.s, x0]
Note that for non-temporal gathers/scatters, the SVE spec defines only
one instruction type: "vector + scalar". For this reason, we swap the
arguments when processing intrinsics that implement the "scalar +
vector" addressing mode:
* @llvm.aarch64.sve.ldnt1.gather
* @llvm.aarch64.sve.ldnt1.gather.uxtw
* @llvm.aarch64.sve.stnt1.scatter
* @llvm.aarch64.sve.ldnt1.gather.uxtw
In other words, all intrinsics for gather-loads and scatter-stores
implemented in this patch are mapped to the same load and store
instruction, respectively.
The sve2_mem_gldnt_vs multiclass (and it's counterpart for scatter
stores) from SVEInstrFormats.td was split into:
* sve2_mem_gldnt_vec_vs_32_ptrs (32bit wide base addresses)
* sve2_mem_gldnt_vec_vs_62_ptrs (64bit wide base addresses)
This is consistent with what we did for
@llvm.aarch64.sve.ld1.scalar.offset and highlights the actual split in
the spec and the implementation.
Reviewed by: sdesmalen
Differential Revision: https://reviews.llvm.org/D74858
Summary:
The following intrinsics are added:
* @llvm.aarch64.sve.ldff1.gather
* @llvm.aarch64.sve.ldff1.gather.index
* @llvm.aarch64.sve.ldff1.gather_sxtw
* @llvm.aarch64.sve.ldff1.gather.uxtw
* @llvm.aarch64.sve.ldff1.gather_sxtw.index
* @llvm.aarch64.sve.ldff1.gather.uxtw.index
* @llvm.aarch64.sve.ldff1.gather.scalar.offset
Although this patch is quite substantial, the vast majority of the
implementation is just a 'copy & paste' of the implementation of regular
gather loads, including tests. There's only a handful of new
definitions:
* AArch64ISD nodes defined in AArch64ISelLowering.h (e.g. GLDFF1)
* Seleciton DAG Types in AArch64SVEInstrInfo.td (e.g.
AArch64ldff1_gather)
* intrinsics in IntrinsicsAArch64.td (e.g. aarch64_sve_ldff1_gather)
* Pseudo instructions in SVEInstrFormats.td to workaround the issue of
use-before-def for the FFR register.
Reviewed By: sdesmalen
Differential Revision: https://reviews.llvm.org/D75128
In some cases Clang does not perform merging of instructions AND and TST (aka
ANDS xzr).
Example:
tst x2, x1
and x3, x2, x1
to:
ands x3, x2, x1
This patch add such merging during instruction selection: when AND is replaced
with ANDS instruction in LowerSELECT_CC, all users of AND also should be
changed for using this ANDS instruction
Short discussion on mailing list:
http://llvm.1065342.n5.nabble.com/llvm-dev-ARM-Peephole-optimization-instructions-tst-add-tp133109.html
Patch by Pavel Kosov.
Differential Revision: https://reviews.llvm.org/D71701
This node reads the rounding control which means it needs to be ordered properly with operations that change the rounding control. So it needs to be chained to maintain order.
This patch adds a chain input and output to the node and connects it to the chain in SelectionDAGBuilder. I've update all in-tree targets to connect their chain through their lowering code.
Differential Revision: https://reviews.llvm.org/D75132
Summary:
This patch renames functions and TableGen classes for SVE gathers and
scatters. The original names implied that the corresponding
methods/classes are only suited for regular gathers/scatters (i.e. LD1
and ST1), which is not the case. Indeed, we will be re-using them for
non-temporal and first-faulting gathers/scatters in the forthcoming
patches. The new names also highlight the split into Vector-Scalar (VS)
and Scalar-Vector (SV) cases.
List of changes:
* `performLD1GatherCombine` and `performST1ScatterCombine` are renamed
as `performGatherLoadCombine` and `performScatterStoreCombine`,
respectively.
* Selection DAG types for scatters and gathers from
AArch64SVEInstrInfo.td are renamed. For example, `SDT_AArch64_GLD1` is
renamed as `SDT_AArch64_GATHER_SV`. SV stands for Scalar-Vector, as
opposed to Vector-Scalar (VS).
* The intrinsic classes from IntrinsicsAArch64.td are renamed. For
example, `AdvSIMD_GatherLoad_64bitOffset_Intrinsic` is renamed as
`AdvSIMD_GatherLoad_SV_64b_Offsets_Intrinsic`.
* Updated comments in `performGatherLoadCombine` and
`performScatterStoreCombine`.
Reviewers: sdesmalen, rengolin, efriedma
Reviewed By: sdesmalen
Subscribers: tschuett, kristof.beyls, hiraditya, rkruppe, psnobl, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75035
Summary:
Implements the following intrinsics:
* llvm.aarch64.sve.convert.to.svbool
* llvm.aarch64.sve.convert.from.svbool
For converting the ACLE svbool_t type (<n x 16 x i1>) to and from the
other predicate types: <n x 8 x i1>, <n x 4 x i1> and <n x 2 x i1>.
Reviewers: sdesmalen, kmclaughlin, efriedma, dancgr, rengolin
Reviewed By: sdesmalen, efriedma
Subscribers: tschuett, kristof.beyls, hiraditya, rkruppe, psnobl, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D74471
Summary:
Implements the @llvm.aarch64.sve.dupq.lane intrinsic.
As specified in the ACLE, the behaviour of:
svdupq_lane_u64(data, index)
...is identical to:
svtbl(data, svadd_x(svptrue_b64(),
svand_x(svptrue_b64(), svindex_u64(0, 1), 1),
index * 2))
If the index is in the range [0,3], the operation is equivalent
to a single DUP (.q) instruction.
Reviewers: sdesmalen, c-rhodes, cameron.mcinally, efriedma, dancgr, rengolin
Reviewed By: sdesmalen
Subscribers: tschuett, kristof.beyls, hiraditya, rkruppe, psnobl, cfe-commits, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D74734
Eli Friedman