When TTI was updated to use an explicit cost, TCK_CodeSize was used
although the default implicit cost would have been the hand-wavey
cost of size and latency. So, revert back to this behaviour. This is
not expected to have (much) impact on targets since most (all?) of
them return the same value for SizeAndLatency and CodeSize.
When optimising for size, the logic has been changed to query
CodeSize costs instead of SizeAndLatency.
This patch also adds a testing option in the unroller so that
OptSize thresholds can be specified.
Differential Revision: https://reviews.llvm.org/D85723
As with other targets, set the throughput cost of control-flow
instructions to free so that we don't miss out of vectorization
opportunities.
Differential Revision: https://reviews.llvm.org/D85283
This patch uses the feature added in D79162 to fix the cost of a
sext/zext of a masked load, or a trunc for a masked store.
Previously, those were considered cheap or even free, but it's
not the case as we cannot split the load in the same way we would for
normal loads.
This updates the costs to better reflect reality, and adds a test for it
in test/Analysis/CostModel/ARM/cast.ll.
It also adds a vectorizer test that showcases the improvement: in some
cases, the vectorizer will now choose a smaller VF when
tail-predication is enabled, which results in better codegen. (Because
if it were to use a higher VF in those cases, the code we see above
would be generated, and the vmovs would block tail-predication later in
the process, resulting in very poor codegen overall)
Original Patch by Pierre van Houtryve
Differential Revision: https://reviews.llvm.org/D79163
Currently, getCastInstrCost has limited information about the cast it's
rating, often just the opcode and types. Sometimes there is a context
instruction as well, but it isn't trustworthy: for instance, when the
vectorizer is rating a plan, it calls getCastInstrCost with the old
instructions when, in fact, it's trying to evaluate the cost of the
instruction post-vectorization. Thus, the current system can get the
cost of certain casts incorrect as the correct cost can vary greatly
based on the context in which it's used.
For example, if the vectorizer queries getCastInstrCost to evaluate the
cost of a sext(load) with tail predication enabled, getCastInstrCost
will think it's free most of the time, but it's not always free. On ARM
MVE, a VLD2 group cannot be extended like a normal VLDR can. Similar
situations can come up with how masked loads can be extended when being
split.
To fix that, this path adds a new parameter to getCastInstrCost to give
it a hint about the context of the cast. It adds a CastContextHint enum
which contains the type of the load/store being created by the
vectorizer - one for each of the types it can produce.
Original patch by Pierre van Houtryve
Differential Revision: https://reviews.llvm.org/D79162
For a long time, the InstCombine pass handled target specific
intrinsics. Having target specific code in general passes was noted as
an area for improvement for a long time.
D81728 moves most target specific code out of the InstCombine pass.
Applying the target specific combinations in an extra pass would
probably result in inferior optimizations compared to the current
fixed-point iteration, therefore the InstCombine pass resorts to newly
introduced functions in the TargetTransformInfo when it encounters
unknown intrinsics.
The patch should not have any effect on generated code (under the
assumption that code never uses intrinsics from a foreign target).
This introduces three new functions:
TargetTransformInfo::instCombineIntrinsic
TargetTransformInfo::simplifyDemandedUseBitsIntrinsic
TargetTransformInfo::simplifyDemandedVectorEltsIntrinsic
A few target specific parts are left in the InstCombine folder, where
it makes sense to share code. The largest left-over part in
InstCombineCalls.cpp is the code shared between arm and aarch64.
This allows to move about 3000 lines out from InstCombine to the targets.
Differential Revision: https://reviews.llvm.org/D81728
If a vector body has live-out values, it is probably a reduction, which needs a
final reduction step after the loop. MVE has a VADDV instruction to reduce
integer vectors, but doesn't have an equivalent one for float vectors. A
live-out value that is not recognised as reduction later in the optimisation
pipeline will result in the tail-predicated loop to be reverted to a
non-predicated loop and this is very expensive, i.e. it has a significant
performance impact, which is what we hope to avoid with fine tuning the ARM TTI
hook preferPredicateOverEpilogue implementation.
Differential Revision: https://reviews.llvm.org/D82953
This refactors option -disable-mve-tail-predication to take different arguments
so that we have 1 option to control tail-predication rather than several
different ones.
This is also a prep step for D82953, in which we want to reject reductions
unless that is requested with this option.
Differential Revision: https://reviews.llvm.org/D83133
Summary:
This patch separates the peeling specific parameters from the UnrollingPreferences,
and creates a new struct called PeelingPreferences. Functions which used the
UnrollingPreferences struct for peeling have been updated to use the PeelingPreferences struct.
Author: sidbav (Sidharth Baveja)
Reviewers: Whitney (Whitney Tsang), Meinersbur (Michael Kruse), skatkov (Serguei Katkov), ashlykov (Arkady Shlykov), bogner (Justin Bogner), hfinkel (Hal Finkel), anhtuyen (Anh Tuyen Tran), nikic (Nikita Popov)
Reviewed By: Meinersbur (Michael Kruse)
Subscribers: fhahn (Florian Hahn), hiraditya (Aditya Kumar), llvm-commits, LLVM
Tag: LLVM
Differential Revision: https://reviews.llvm.org/D80580
This adjusts the MVE fp16 cost model, similar to how we already do for
integer casts. It uses the base cost of 1 per cvt for most fp extend /
truncates, but adjusts it for loads and stores where we know that a
extending load has been used to get the load into the correct lane, and
only an MVE VCVTB is then needed.
Differential Revision: https://reviews.llvm.org/D81813
This adds some default costs for fp extends and truncates, generally
costing them as 1 per lane. If the type is not legal then the cost will
include a call to an __aeabi_ function.
Some NEON code is also adjusted to make sure it applies to the expected
types, now that fp16 is a more common thing.
Differential Revision: https://reviews.llvm.org/D82458
This expands the existing extend costs with a few extras for larger
types than legal, which will usually be split under MVE. It also adds
trunk support for the same thing. These should not have a large effect
on many things, but makes the costs explicit and keeps a certain balance
between the trunks and extends.
Differential Revision: https://reviews.llvm.org/D82457
This alters getMemoryOpCost to use the Base TargetTransformInfo version
that includes some additional checks for whether extending loads are
legal. This will generally have the effect of making <2 x ..> and some
<4 x ..> loads/stores more expensive, which in turn should help favour
larger vector factors.
Notably it alters the cost of a <4 x half>, which with the current
codegen will be expensive if it is not extended.
Differential Revision: https://reviews.llvm.org/D82456
The main interface has been migrated to Align already but a few backends where broadening the type from Align to MaybeAlign.
This patch makes sure all implementations conform to the public API.
Differential Revision: https://reviews.llvm.org/D82465
Summary:
Get back `const` partially lost in one of recent changes.
Additionally specify explicit qualifiers in few places.
Reviewers: samparker
Reviewed By: samparker
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D82383
This emits new IR intrinsic @llvm.get.active.mask for tail-folded vectorised
loops if the intrinsic is supported by the backend, which is checked by
querying TargetTransform hook emitGetActiveLaneMask.
This intrinsic creates a mask representing active and inactive vector lanes,
which is used by the masked load/store instructions that are created for
tail-folded loops. The semantics of @llvm.get.active.mask are described here in
LangRef:
https://llvm.org/docs/LangRef.html#llvm-get-active-lane-mask-intrinsics
This intrinsic is also used to provide a hint to the backend. That is, the
second argument of the intrinsic represents the back-edge taken count of the
loop. For MVE, for example, we use that to set up tail-predication, which is a
new form of predication in MVE for vector loops that implicitely predicates the
last vector loop iteration by implicitely setting active/inactive lanes, i.e.
the tail loop is predicated. In order to set up a tail-predicated vector loop,
we need to know the number of data elements processed by the vector loop, which
corresponds the the tripcount of the scalar loop, which we can now reconstruct
using @llvm.get.active.mask.
Differential Revision: https://reviews.llvm.org/D79100
Similar to a recent change to the X86 backend, this changes things so
that we always produce a reduction intrinsics for all reduction types,
not just the legal ones. This gives a better chance in the backend to
custom lower them to something more suitable for MVE. Especially for
something like fadd the in-order reduction produced during DAG lowering
is already better than the shuffles produced in the midend, and we can
do even better with a bit of custom lowering.
Differential Revision: https://reviews.llvm.org/D81398
Add the remaining arithmetic opcodes into the generic implementation
of getUserCost and then call this from getInstructionThroughput. Most
of the backends have been modified to return the base implementation
for cost kinds other RecipThroughput. The outlier here is AMDGPU
which already uses getArithmeticInstrCost for all the cost kinds.
This change means that most of the opcodes can be removed from that
backends implementation of getUserCost.
Differential Revision: https://reviews.llvm.org/D80992
Add cases for icmp, fcmp and select into the switch statement of the
generic getUserCost implementation with getInstructionThroughput then
calling into it. The BasicTTI and backend implementations have be set
to return a default value (1) when a cost other than throughput is
being queried.
Differential Revision: https://reviews.llvm.org/D80550
Use getMemoryOpCost from the generic implementation of getUserCost
and have getInstructionThroughput return the result of that for loads
and stores.
This also means that the X86 implementation of getUserCost can be
removed with the functionality folded into its getMemoryOpCost.
Differential Revision: https://reviews.llvm.org/D80984
This is split off from D79100 and adds a new target hook emitGetActiveLaneMask
that can be queried to check if the intrinsic @llvm.get.active.lane.mask() is
supported by the backend and if it should be emitted for a given loop.
See also commit rG7fb8a40e5220 and its commit message for more details/context
on this new intrinsic.
Differential Revision: https://reviews.llvm.org/D80597
Add the remaining cast instruction opcodes to the base implementation
of getUserCost and directly return the result. This allows
getInstructionThroughput to return getUserCost for the casts. This
has required changes to PPC and SystemZ because they implement
getUserCost and/or getCastInstrCost with adjustments for vector
operations. Adjusts have also been made in the remaining backends
that implement the method so that they still produce a cost of zero
or one for cost kinds other than throughput.
Differential Revision: https://reviews.llvm.org/D79848
If the caller needs to reponsible for making sure the MaybeAlign
has a value, then we should just make the caller convert it to an Align
with operator*.
I explicitly deleted the relational comparison operators that
were being inherited from Optional. It's unclear what the meaning
of two MaybeAligns were one is defined and the other isn't
should be. So make the caller reponsible for defining the behavior.
I left the ==/!= operators from Optional. But now that exposed a
weird quirk that ==/!= between Align and MaybeAlign required the
MaybeAlign to be defined. But now we use the operator== from
Optional that takes an Optional and the Value.
Differential Revision: https://reviews.llvm.org/D80455
This patch adds a new TTI hook to allow targets to tell LSR that
a chain including some instruction is already profitable and
should not be optimized. This patch also adds an implementation
of this TTI hook for ARM so LSR doesn't optimize chains that include
the VCTP intrinsic.
Differential Revision: https://reviews.llvm.org/D79418
- Specifically check for sext/zext users which have 'long' form NEON
instructions.
- Add more entries to the table for sext/zexts so that we can report
more accurately the number of vmovls required for NEON.
- Pass the instruction to the pass implementation.
Differential Revision: https://reviews.llvm.org/D79561
getScalarizationOverhead is only ever called with vectors (and we already had a load of cast<VectorType> calls immediately inside the functions).
Followup to D78357
Reviewed By: @samparker
Differential Revision: https://reviews.llvm.org/D79341
Make the kind of cost explicit throughout the cost model which,
apart from making the cost clear, will allow the generic parts to
calculate better costs. It will also allow some backends to
approximate and correlate the different costs if they wish. Another
benefit is that it will also help simplify the cost model around
immediate and intrinsic costs, where we currently have multiple APIs.
RFC thread:
http://lists.llvm.org/pipermail/llvm-dev/2020-April/141263.html
Differential Revision: https://reviews.llvm.org/D79002
There are several different types of cost that TTI tries to provide
explicit information for: throughput, latency, code size along with
a vague 'intersection of code-size cost and execution cost'.
The vectorizer is a keen user of RecipThroughput and there's at least
'getInstructionThroughput' and 'getArithmeticInstrCost' designed to
help with this cost. The latency cost has a single use and a single
implementation. The intersection cost appears to cover most of the
rest of the API.
getUserCost is explicitly called from within TTI when the user has
been explicit in wanting the code size (also only one use) as well
as a few passes which are concerned with a mixture of size and/or
a relative cost. In many cases these costs are closely related, such
as when multiple instructions are required, but one evident diverging
cost in this function is for div/rem.
This patch adds an argument so that the cost required is explicit,
so that we can make the important distinction when necessary.
Differential Revision: https://reviews.llvm.org/D78635
The API for shuffles and reductions uses generic Type parameters,
instead of VectorType, and so assertions and casts are used a lot.
This patch makes those types explicit, which means that the clients
can't be lazy, but results in less ambiguity, and that can only be a
good thing.
Bugzilla: https://bugs.llvm.org/show_bug.cgi?id=45562
Differential Revision: https://reviews.llvm.org/D78357
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: grosbach, efriedma, sdesmalen
Reviewed By: efriedma
Subscribers: hiraditya, dmgreen, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D77271
Refines the gather/scatter cost model, but also changes the TTI
function getIntrinsicInstrCost to accept an additional parameter
which is needed for the gather/scatter cost evaluation.
This did require trivial changes in some non-ARM backends to
adopt the new parameter.
Extending gathers and truncating scatters are now priced cheaper.
Differential Revision: https://reviews.llvm.org/D75525
Summary: This is a first step before changing the types to llvm::Align and introduce functions to ease client code.
Reviewers: courbet
Subscribers: arsenm, sdardis, nemanjai, jvesely, nhaehnle, hiraditya, kbarton, jrtc27, atanasyan, jsji, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73785
The codegen for splitting a llvm.vector.reduction intrinsic into parts
will be better than the codegen for the generic reductions. This will
only directly effect when vectorization factors are specified by the
user.
Also added tests to make sure the codegen for larger reductions is OK.
Differential Revision: https://reviews.llvm.org/D72257
This is a very basic MVE gather/scatter cost model, based roughly on the
code that we will currently produce. It does not handle truncating
scatters or extending gathers correctly yet, as it is difficult to tell
that they are going to be correctly extended/truncated from the limited
information in the cost function.
This can be improved as we extend support for these in the future.
Based on code originally written by David Sherwood.
Differential Revision: https://reviews.llvm.org/D73021
We were previously not necessarily favouring postinc for the MVE loads
and stores, leading to extra code prior to the loop to set up the
preinc. MVE in general can benefit from postinc (as we don't have
unrolled loops), and certain instructions like the VLD2's only post-inc
versions are available.
Differential Revision: https://reviews.llvm.org/D70790
We don't unroll vector loops for MVE targets, but we miss the case
when loops only contain intrinsic calls. So just move the logic a
bit to catch this case.
Differential Revision: https://reviews.llvm.org/D72440
Adds a pass to the ARM backend that takes a v4i32
gather and transforms it into a call to MVE's
masked gather intrinsics.
Differential Revision: https://reviews.llvm.org/D71743
Soon Intrinsic::ID will be a plain integer, so this overload will not be
possible.
Rename both overloads to ensure that downstream targets observe this as
a build failure instead of a runtime failure.
Split off from D71320
Reviewers: efriedma
Differential Revision: https://reviews.llvm.org/D71381
With the extra optimisations we have done, these should now be fine to
enable by default. Which is what this patch does.
Differential Revision: https://reviews.llvm.org/D70968
This attempts to teach the cost model in Arm that code such as:
%s = shl i32 %a, 3
%a = and i32 %s, %b
Can under Arm or Thumb2 become:
and r0, r1, r2, lsl #3
So the cost of the shift can essentially be free. To do this without
trying to artificially adjust the cost of the "and" instruction, it
needs to get the users of the shl and check if they are a type of
instruction that the shift can be folded into. And so it needs to have
access to the actual instruction in getArithmeticInstrCost, which if
available is added as an extra parameter much like getCastInstrCost.
We otherwise limit it to shifts with a single user, which should
hopefully handle most of the cases. The list of instruction that the
shift can be folded into include ADC, ADD, AND, BIC, CMP, EOR, MVN, ORR,
ORN, RSB, SBC and SUB. This translates to Add, Sub, And, Or, Xor and
ICmp.
Differential Revision: https://reviews.llvm.org/D70966
This adds some extra cost model tests for shifts, and does some minor
adjustments to some Neon code to make it clear as to what it applies to.
Both NFC.
Now that we have the intrinsics, we can add VLD2/4 and VST2/4 lowering
for MVE. This works the same way as Neon, recognising the load/shuffles
combination and converting them into intrinsics in a pre-isel pass,
which just calls getMaxSupportedInterleaveFactor, lowerInterleavedLoad
and lowerInterleavedStore.
The main difference to Neon is that we do not have a VLD3 instruction.
Otherwise most of the code works very similarly, with just some minor
differences in the form of the intrinsics to work around. VLD3 is
disabled by making isLegalInterleavedAccessType return false for those
cases.
We may need some other future adjustments, such as VLD4 take up half the
available registers so should maybe cost more. This patch should get the
basics in though.
Differential Revision: https://reviews.llvm.org/D69392
This is a follow up of d90804d, to also flag fmcp instructions as instructions
that we do not support in tail-predicated vector loops.
Differential Revision: https://reviews.llvm.org/D70295
This implements TTI hook 'preferPredicateOverEpilogue' for MVE. This is a
first version and it operates on single block loops only. With this change, the
vectoriser will now determine if tail-folding scalar remainder loops is
possible/desired, which is the first step to generate MVE tail-predicated
vector loops.
This is disabled by default for now. I.e,, this is depends on option
-disable-mve-tail-predication, which is off by default.
I will follow up on this soon with a patch for the vectoriser to respect loop
hint 'vectorize.predicate.enable'. I.e., with this loop hint set to Disabled,
we don't want to tail-fold and we shouldn't query this TTI hook, which is
done in D70125.
Differential Revision: https://reviews.llvm.org/D69845
We have two ways to steer creating a predicated vector body over creating a
scalar epilogue. To force this, we have 1) a command line option and 2) a
pragma available. This adds a third: a target hook to TargetTransformInfo that
can be queried whether predication is preferred or not, which allows the
vectoriser to make the decision without forcing it.
While this change behaves as a non-functional change for now, it shows the
required TTI plumbing, usage of this new hook in the vectoriser, and the
beginning of an ARM MVE implementation. I will follow up on this with:
- a complete MVE implementation, see D69845.
- a patch to disable this, i.e. we should respect "vector_predicate(disable)"
and its corresponding loophint.
Differential Revision: https://reviews.llvm.org/D69040
Add generic DAG combine for extending masked loads.
Allow us to generate sext/zext masked loads which can access v4i8,
v8i8 and v4i16 memory to produce v4i32, v8i16 and v4i32 respectively.
Differential Revision: https://reviews.llvm.org/D68337
llvm-svn: 375085
Add an extra parameter so the backend can take the alignment into
consideration.
Differential Revision: https://reviews.llvm.org/D68400
llvm-svn: 374763
Masked loads and store fit naturally with MVE, the instructions being easily
predicated. This adds lowering for the simple cases of masked loads and stores.
It does not yet deal with widening/narrowing or pre/post inc, and so is
currently behind an option.
The llvm masked load intrinsic will accept a "passthru" value, dictating the
values used for the zero masked lanes. In MVE the instructions write 0 to the
zero predicated lanes, so we need to match a passthru that isn't 0 (or undef)
with a select instruction to pull in the correct data after the load.
Differential Revision: https://reviews.llvm.org/D67186
llvm-svn: 371932
This patch adds vecreduce_smax, vecredude_umax, vecreduce_smin, vecreduce_umin and selection for vmaxv and minv.
Differential Revision: https://reviews.llvm.org/D66413
llvm-svn: 371827
This patch adds vecreduce_add and the relevant instruction selection for
vaddv.
Differential revision: https://reviews.llvm.org/D66085
llvm-svn: 369245
This adds some sext costs for MVE, taken from the length of assembly sequences
that we currently generate.
Differential Revision: https://reviews.llvm.org/D66010
llvm-svn: 369244
MVE also has some sext of loads, which will be free just as scalar
instructions are.
Differential Revision: https://reviews.llvm.org/D66008
llvm-svn: 369118
The MVE architecture has the idea of "beats", where a vector instruction can be
executed over several ticks of the architecture. This adds a similar system
into the Arm backend cost model, multiplying the cost of all vector
instructions by a factor.
This factor essentially becomes the expected difference between scalar code
and vector code, on average. MVE Vector instructions can also overlap so the a
true cost of them is often lower. But equally scalar instructions can in some
situations be dual issued, or have other optimisations such as unrolling or
make use of dsp instructions. The default is chosen as 2. This should not
prevent vectorisation is a most cases (as the vector instructions will still be
doing at least 4 times the work), but it will help prevent over vectorising in
cases where the benefits are less likely.
This adds things so far to the obvious places in ARMTargetTransformInfo, and
updates a few related costs like not treating float instructions as cost 2 just
because they are floats.
Differential Revision: https://reviews.llvm.org/D66005
llvm-svn: 368733
This teaches the cost model that the sext or zext of a load is going to be
free.
Differential Revision: https://reviews.llvm.org/D66006
llvm-svn: 368593
A VDUP will perform a vector broadcast in a single instruction. Update the cost
model for MVE accordingly.
Code originally by David Sherwood.
Differential Revision: https://reviews.llvm.org/D63448
llvm-svn: 368589
This puts some of the calls in ARMTargetTransformInfo.cpp behind hasNeon()
checks, now that we have MVE, and updates all the tests accordingly.
Differential Revision: https://reviews.llvm.org/D63447
llvm-svn: 368587
Due to the nature of the beat system in the MVE architecture, along with tail
predication and low-overhead loops, unrolling has less benefit compared to
normal loops. You can not, for example, hide the latency of a load with other
instructions as you can for scalar code. Preventing unrolling also makes the
code easier to read and reason about.
So if a loop contains vector code, don't enable the runtime unrolling. At least
for the time being.
Differential Revision: https://reviews.llvm.org/D65803
llvm-svn: 368530
The code is now in a good enough state to pass the bunch of tests that
I have run (after fixing the bugs), so let's enable it by default.
Differential Revision: https://reviews.llvm.org/D65277
llvm-svn: 367297
Backend changes to enable WLS/LE low-overhead loops for armv8.1-m:
1) Use TTI to communicate to the HardwareLoop pass that we should try
to generate intrinsics that guard the loop entry, as well as setting
the loop trip count.
2) Lower the BRCOND that uses said intrinsic to an Arm specific node:
ARMWLS.
3) ISelDAGToDAG the node to a new pseudo instruction:
t2WhileLoopStart.
4) Add support in ArmLowOverheadLoops to handle the new pseudo
instruction.
Differential Revision: https://reviews.llvm.org/D63816
llvm-svn: 364733
The HardwareLoops pass finds exit blocks with a scevable exit count.
If the target specifies to update the loop counter in a register,
through a phi, we need to ensure that the exit block is a latch so
that we can insert the phi with the correct value for the incoming
edge.
Differential Revision: https://reviews.llvm.org/D63336
llvm-svn: 363556
TTI should report that it's not profitable to generate a hardware loop
if it, or one of its child loops, has already been converted.
Differential Revision: https://reviews.llvm.org/D63212
llvm-svn: 363234
Implement the backend target hook to drive the HardwareLoops pass.
The low-overhead branch extension for Arm M-class cores is flexible
enough that we don't have to ensure correctness at this point, except
checking that the loop counter variable can be stored in LR - a
32-bit register. For it to be profitable, we want to avoid loops that
contain function calls, or any other instruction that alters the PC.
This implementation uses TargetLoweringInfo, to query type and
operation actions, looks at intrinsic calls and also performs some
manual checks for remainder/division and FP operations.
I think this should be a good base to start and extra details can be
filled out later.
Differential Revision: https://reviews.llvm.org/D62907
llvm-svn: 363149
This option allows loops with small max trip counts to be fully unrolled. This
can help with code like the remainder loops from manually unrolled loops like
those that appear in the cmsis dsp library. We would apparently previously
runtime unroll them with the default unroll count (4).
Differential Revision: https://reviews.llvm.org/D63064
llvm-svn: 362928
This implements TargetTransformInfo method getMemcpyCost, which estimates the
number of instructions to which a memcpy instruction expands to.
Differential Revision: https://reviews.llvm.org/D59787
llvm-svn: 359547
This prevents Constant Hoisting from pulling the constant out of the block,
allowing us to still produce LDRH/UXTH nodes. LDRB/UXTB (255) is already cheap
by the default getIntImmCost, but I've added it for clarity.
Differential Revision: https://reviews.llvm.org/D57671
llvm-svn: 353040
to reflect the new license.
We understand that people may be surprised that we're moving the header
entirely to discuss the new license. We checked this carefully with the
Foundation's lawyer and we believe this is the correct approach.
Essentially, all code in the project is now made available by the LLVM
project under our new license, so you will see that the license headers
include that license only. Some of our contributors have contributed
code under our old license, and accordingly, we have retained a copy of
our old license notice in the top-level files in each project and
repository.
llvm-svn: 351636
optsize using masked wide loads
Under Opt for Size, the vectorizer does not vectorize interleave-groups that
have gaps at the end of the group (such as a loop that reads only the even
elements: a[2*i]) because that implies that we'll require a scalar epilogue
(which is not allowed under Opt for Size). This patch extends the support for
masked-interleave-groups (introduced by D53011 for conditional accesses) to
also cover the case of gaps in a group of loads; Targets that enable the
masked-interleave-group feature don't have to invalidate interleave-groups of
loads with gaps; they could now use masked wide-loads and shuffles (if that's
what the cost model selects).
Reviewers: Ayal, hsaito, dcaballe, fhahn
Reviewed By: Ayal
Differential Revision: https://reviews.llvm.org/D53668
llvm-svn: 345705
I noticed while fixing PR39368 that we don't have generic shuffle costs for broadcast style shuffles.
This patch adds SK_BROADCAST handling, but exposes ARM/AARCH64 lack of handling of this type, which I've added a fix for at the same time.
Differential Revision: https://reviews.llvm.org/D53570
llvm-svn: 345253
interleave-group
The vectorizer currently does not attempt to create interleave-groups that
contain predicated loads/stores; predicated strided accesses can currently be
vectorized only using masked gather/scatter or scalarization. This patch makes
predicated loads/stores candidates for forming interleave-groups during the
Loop-Vectorizer's analysis, and adds the proper support for masked-interleave-
groups to the Loop-Vectorizer's planning and transformation stages. The patch
also extends the TTI API to allow querying the cost of masked interleave groups
(which each target can control); Targets that support masked vector loads/
stores may choose to enable this feature and allow vectorizing predicated
strided loads/stores using masked wide loads/stores and shuffles.
Reviewers: Ayal, hsaito, dcaballe, fhahn, javed.absar
Reviewed By: Ayal
Differential Revision: https://reviews.llvm.org/D53011
llvm-svn: 344472
A simple MOVS rd, imm8 can materialize [-128, 127] in signed i8 type or
[0, 255] in unsigned i8 type on Thumb1.
Differential Revision: https://reviews.llvm.org/D52257
llvm-svn: 342898
This is a simple implementation of the unroll-and-jam classical loop
optimisation.
The basic idea is that we take an outer loop of the form:
for i..
ForeBlocks(i)
for j..
SubLoopBlocks(i, j)
AftBlocks(i)
Instead of doing normal inner or outer unrolling, we unroll as follows:
for i... i+=2
ForeBlocks(i)
ForeBlocks(i+1)
for j..
SubLoopBlocks(i, j)
SubLoopBlocks(i+1, j)
AftBlocks(i)
AftBlocks(i+1)
Remainder Loop
So we have unrolled the outer loop, then jammed the two inner loops into
one. This can lead to a simpler inner loop if memory accesses can be shared
between the now jammed loops.
To do this we have to prove that this is all safe, both for the memory
accesses (using dependence analysis) and that ForeBlocks(i+1) can move before
AftBlocks(i) and SubLoopBlocks(i, j).
Differential Revision: https://reviews.llvm.org/D41953
llvm-svn: 336062
As discussed on PR33744, this patch relaxes ShuffleKind::SK_Alternate which requires shuffle masks to only match an alternating pattern from its 2 sources:
e.g. v4f32: <0,5,2,7> or <4,1,6,3>
This seems far too restrictive as most SIMD hardware which will implement it using a general blend/bit-select instruction, so replaces it with SK_Select, permitting elements from either source as long as they are inline:
e.g. v4f32: <0,5,2,7>, <4,1,6,3>, <0,1,6,7>, <4,1,2,3> etc.
This initial patch just updates the name and cost model shuffle mask analysis, later patch reviews will update SLP to better utilise this - it still limits itself to SK_Alternate style patterns.
Differential Revision: https://reviews.llvm.org/D47985
llvm-svn: 334513
This is a simple implementation of the unroll-and-jam classical loop
optimisation.
The basic idea is that we take an outer loop of the form:
for i..
ForeBlocks(i)
for j..
SubLoopBlocks(i, j)
AftBlocks(i)
Instead of doing normal inner or outer unrolling, we unroll as follows:
for i... i+=2
ForeBlocks(i)
ForeBlocks(i+1)
for j..
SubLoopBlocks(i, j)
SubLoopBlocks(i+1, j)
AftBlocks(i)
AftBlocks(i+1)
Remainder
So we have unrolled the outer loop, then jammed the two inner loops into
one. This can lead to a simpler inner loop if memory accesses can be shared
between the now-jammed loops.
To do this we have to prove that this is all safe, both for the memory
accesses (using dependence analysis) and that ForeBlocks(i+1) can move before
AftBlocks(i) and SubLoopBlocks(i, j).
Differential Revision: https://reviews.llvm.org/D41953
llvm-svn: 333358
The DEBUG() macro is very generic so it might clash with other projects.
The renaming was done as follows:
- git grep -l 'DEBUG' | xargs sed -i 's/\bDEBUG\s\?(/LLVM_DEBUG(/g'
- git diff -U0 master | ../clang/tools/clang-format/clang-format-diff.py -i -p1 -style LLVM
- Manual change to APInt
- Manually chage DOCS as regex doesn't match it.
In the transition period the DEBUG() macro is still present and aliased
to the LLVM_DEBUG() one.
Differential Revision: https://reviews.llvm.org/D43624
llvm-svn: 332240
Currently EVT is in the IR layer only because of Function.cpp needing a very small piece of the functionality of EVT::getEVTString(). The rest of EVT is used in codegen making CodeGen a better place for it.
The previous code converted a Type* to EVT and then called getEVTString. This was only expected to handle the primitive types from Type*. Since there only a few primitive types, we can just print them as strings directly.
Differential Revision: https://reviews.llvm.org/D45017
llvm-svn: 328806
This is used by llvm tblgen as well as by LLVM Targets, so the only
common place is Support for now. (maybe we need another target for these
sorts of things - but for now I'm at least making them correct & we can
make them better if/when people have strong feelings)
llvm-svn: 328395
Fixup to rL325573 for large xor constants.
Thanks to Eli Friedman for the catch.
Differential revision: https://reviews.llvm.org/D43549
llvm-svn: 325761
We can always convert xor %a, -1 into MVN, even in thumb 1 where the -1
would not otherwise be considered a cheap constant. This prevents the
-1's from being pulled out into constants and potentially hoisted.
Differential Revision: https://reviews.llvm.org/D43451
llvm-svn: 325573
The penalty is currently getting applied in a bunch of places where it
doesn't make sense, like bitcasts (which are free) and calls (which
were getting the call penalty applied twice). Instead, just apply the
penalty to binary operators and floating-point casts.
While I'm here, also fix getFPOpCost() to do the right thing in more
cases, so we don't have to dig into function attributes.
Differential Revision: https://reviews.llvm.org/D41522
llvm-svn: 321332
All these headers already depend on CodeGen headers so moving them into
CodeGen fixes the layering (since CodeGen depends on Target, not the
other way around).
llvm-svn: 318490
Before, loop unrolling was only enabled for loops with a single
block. This restriction has been removed and replaced by:
- allow a maximum of two exiting blocks,
- a four basic block limit for cores with a branch predictor.
Differential Revision: https://reviews.llvm.org/D38952
llvm-svn: 316313
- Set the default runtime unroll count to 4 and use the newly added
UnrollRemainder option.
- Create loop cost and force unroll for a cost less than 12.
- Disable unrolling on Thumb1 only targets.
Differential Revision: https://reviews.llvm.org/D36134
llvm-svn: 310997
Enable runtime and partial loop unrolling of simple loops without
calls on M-class cores. The thresholds are calculated based on
whether the target is Thumb or Thumb-2.
Differential Revision: https://reviews.llvm.org/D34619
llvm-svn: 308956
Summary:
Similar to X86, it should be safe to inline callees if their
target-features are a subset of the caller. As some subtarget features
provide different instructions depending on whether they are set or
unset (e.g. ThumbMode and ModeSoftFloat), we use a whitelist of
target-features describing hardware capabilities only.
Reviewers: kristof.beyls, rengolin, t.p.northover, SjoerdMeijer, peter.smith, silviu.baranga, efriedma
Reviewed By: SjoerdMeijer, efriedma
Subscribers: dschuff, efriedma, aemerson, sdardis, javed.absar, arichardson, eraman, llvm-commits
Differential Revision: https://reviews.llvm.org/D34697
llvm-svn: 307889
getArithmeticInstrCost(), getShuffleCost(), getCastInstrCost(),
getCmpSelInstrCost(), getVectorInstrCost(), getMemoryOpCost(),
getInterleavedMemoryOpCost() implemented.
Interleaved access vectorization enabled.
BasicTTIImpl::getCastInstrCost() improved to check for legal extending loads,
in which case the cost of the z/sext instruction becomes 0.
Review: Ulrich Weigand, Renato Golin.
https://reviews.llvm.org/D29631
llvm-svn: 300052
This patch refactors and strengthens the type checks performed for interleaved
accesses. The primary functional change is to ensure that the interleaved
accesses have valid element types. The added test cases previously failed
because the element type is f128.
Differential Revision: https://reviews.llvm.org/D31817
llvm-svn: 299864
After r296750, we're able to match interleaved accesses having types wider than
128 bits. This patch updates the associated TTI costs.
Differential Revision: https://reviews.llvm.org/D29675
llvm-svn: 296751
There are no vldN/vstN f16 variants, even with +fullfp16.
We could use the i16 variants, but, in practice, even with +fullfp16,
the f16 sequence leading to the i16 shuffle usually gets scalarized.
We'd need to improve our support for f16 codegen before getting there.
Teach the cost model to consider f16 interleaved operations as
expensive. Otherwise, we are all but guaranteed to end up with
a large block of scalarized vector code.
llvm-svn: 294819
updated instructions:
pmulld, pmullw, pmulhw, mulsd, mulps, mulpd, divss, divps, divsd, divpd, addpd and subpd.
special optimization case which replaces pmulld with pmullw\pmulhw\pshuf seq.
In case if the real operands bitwidth <= 16.
Differential Revision: https://reviews.llvm.org/D28104
llvm-svn: 291657
This code seems to be target dependent which may not be the same for all targets.
Passed the decision whether the given stride is complex or not to the target by sending stride information via SCEV to getAddressComputationCost instead of 'IsComplex'.
Specifically at X86 targets we dont see any significant address computation cost in case of the strided access in general.
Differential Revision: https://reviews.llvm.org/D27518
llvm-svn: 291106
Materializing something like "-3" can be done as 2 instructions:
MOV r0, #3
MVN r0, r0
This has a cost of 2, not 3. It looks like we were already trying to detect this pattern in TII::getIntImmCost(), but were taking the complement of the zero-extended value instead of the sign-extended value which is unlikely to ever produce a number < 256.
There were no tests failing after changing this... :/
llvm-svn: 280928
constant hoisting. It not only takes into account the number of uses and the
cost of expressions in which constants appear, but now also the resulting
integer range of the offsets. Thus, the algorithm maximizes the number of uses
within an integer range that will enable more efficient code generation. On
ARM, for example, this will enable code size optimisations because less
negative offsets will be created. Negative offsets/immediates are not supported
by Thumb1 thus preventing more compact instruction encoding.
Differential Revision: http://reviews.llvm.org/D21183
llvm-svn: 275382
This is a follow-up for r273544 and r273853.
The end goal is to get rid of the isSwift / isCortexXY / isWhatever methods.
This commit also marks them as obsolete.
Differential Revision: http://reviews.llvm.org/D21796
llvm-svn: 274616
Summary:
This fixes bug: https://llvm.org/bugs/show_bug.cgi?id=28282
Currently the cost model of constant hoisting checks the bit width of the data type of the constants.
However, the actual immediate value is small enough and not need to be hoisted.
This patch checks for the actual bit width needed for the constant.
Reviewers: t.p.northover, rengolin
Subscribers: aemerson, rengolin, llvm-commits
Differential Revision: http://reviews.llvm.org/D21668
llvm-svn: 274073
Divisions by a constant can be converted into multiplies which are usually
cheaper, but this isn't possible if the constant gets separated (particularly
in loops). Fix this by telling ConstantHoisting that the immediate in a DIV is
cheap.
I considered making the check generic, but neither AArch64 (strangely) nor x86
showed any benefit on the tests I had.
llvm-svn: 266464
At some point, ARM stopped getting any benefit from ConstantHoisting because
the pass called a different variant of getIntImmCost. Reimplementing the
correct variant revealed some problems, however:
+ ConstantHoisting was modifying switch statements. This is simply invalid,
the cases must remain integer constants no matter the notional cost.
+ ConstantHoisting was mangling alloca instructions in the entry block. These
should be handled by FrameLowering, so constants actually have a cost of 0.
Worse, the resulting bitcasts meant they became dynamic allocas.
rdar://25707382
llvm-svn: 266260
Otherwise, we think that most types that look like they'd fit in a
legal vector type are legal (so, basically, *any* vector type with a
size between 33 and 128 bits, I think, since we use pow2 alignment;
e.g., v2i25, v3f32, ...).
DataLayout::getTypeAllocSize rounds up based on alignment.
When checking for target intrinsic legality, that's not what we want:
if rounding makes a difference, the type isn't legal, and the
target intrinsics shouldn't be used, as they are always assumed legal.
One could make the argument that alloc size is ultimately the most
relevant here, since we're dealing with LD/ST intrinsics. That's only
true if we did legalize them though; that's a problem for another day.
Use DataLayout::getTypeSizeInBits instead of getTypeAllocSizeInBits.
Type::getSizeInBits can't be used because that'd gratuitously break
pointer vector support.
Some of these uses are currently fine, because we only hit them when
the type is already known legal (e.g., r114454). Update them for
consistency. It's faster to avoid the rounding anyway!
llvm-svn: 255089
The underlying issues surrounding codegen for 32-bit vselects have been resolved. The pessimistic costs for 64-bit vselects remain due to the bad
scalarization that is still happening there.
I tested this on A57 in T32, A32 and A64 modes. I saw no regressions, and some improvements.
From my benchmarks, I saw these improvements in A57 (T32)
spec.cpu2000.ref.177_mesa 5.95%
lnt.SingleSource/Benchmarks/Shootout/strcat 12.93%
lnt.MultiSource/Benchmarks/MiBench/telecomm-CRC32/telecomm-CRC32 11.89%
I also measured A57 A32, A53 T32 and A9 T32 and found no performance regressions. I see much bigger wins in third-party benchmarks with this change
Differential Revision: http://reviews.llvm.org/D14743
llvm-svn: 253349
This also lets us remove the versions of the functions that took a statically sized array as we can rely on ArrayRef implicit conversion now.
llvm-svn: 251490
This avoid mentioning the table name an extra time and allows the lookup to be done directly in the ifs by relying on the bool conversion of the pointer.
While there make use of ArrayRef and std::find_if.
llvm-svn: 251382
Summary:
This change limits the minimum cost of an insert/extract
element operation to 2 in cases where this would result
in mixing of NEON and VFP code.
Reviewers: rengolin
Subscribers: mssimpso, aemerson, llvm-commits, rengolin
Differential Revision: http://reviews.llvm.org/D12030
llvm-svn: 245225
rather than 'unsigned' for their costs.
For something like costs in particular there is a natural "negative"
value, that of savings or saved cost. As a consequence, there is a lot
of code that subtracts or creates negative values based on cost, all of
which is prone to awkwardness or bugs when dealing with an unsigned
type. Similarly, we *never* want these values to wrap, as that would
cause Very Bad code generation (likely percieved as an infinite loop as
we try to emit over 2^32 instructions or some such insanity).
All around 'int' seems a much better fit for these basic metrics. I've
added asserts to ensure that at least the TTI interface never returns
negative numbers here. If we ever have a use case for negative numbers,
we can remove this, but this way a bug where someone used '-1' to
produce a 'very large' cost will be caught by the assert.
This passes all tests, and is also UBSan clean.
No functional change intended.
Differential Revision: http://reviews.llvm.org/D11741
llvm-svn: 244080
Summary:
Fix the cost of interleaved accesses for ARM/AArch64.
We were calling getTypeAllocSize and using it to check
the number of bits, when we should have called
getTypeAllocSizeInBits instead.
This would pottentially cause the vectorizer to
generate loads/stores and shuffles which cannot
be matched with an interleaved access instruction.
No performance changes are expected for now since
matching/generating interleaved accesses is still
disabled by default.
Reviewers: rengolin
Subscribers: aemerson, llvm-commits, rengolin
Differential Revision: http://reviews.llvm.org/D11524
llvm-svn: 243270
Summary:
This change is part of a series of commits dedicated to have a single
DataLayout during compilation by using always the one owned by the
module.
Reviewers: echristo
Subscribers: jholewinski, ted, yaron.keren, rafael, llvm-commits
Differential Revision: http://reviews.llvm.org/D11028
From: Mehdi Amini <mehdi.amini@apple.com>
llvm-svn: 241775
DataLayout is no longer optional. It was initialized with or without
a DataLayout, and the DataLayout when supplied could have been the
one from the TargetMachine.
Summary:
This change is part of a series of commits dedicated to have a single
DataLayout during compilation by using always the one owned by the
module.
Reviewers: echristo
Subscribers: jholewinski, llvm-commits, rafael, yaron.keren
Differential Revision: http://reviews.llvm.org/D11021
From: Mehdi Amini <mehdi.amini@apple.com>
llvm-svn: 241774
Summary: When evaluating floating point instructions in the inliner, ask the TTI whether it is an expensive operation. By default, it's not an expensive operation. This keeps the default behavior the same as before. The ARM TTI has been updated to return back TCC_Expensive for targets which don't have hardware floating point.
Reviewers: chandlerc, echristo
Reviewed By: echristo
Subscribers: t.p.northover, aemerson, llvm-commits
Differential Revision: http://reviews.llvm.org/D6936
llvm-svn: 228263
base which it adds a single analysis pass to, to instead return the type
erased TargetTransformInfo object constructed for that TargetMachine.
This removes all of the pass variants for TTI. There is now a single TTI
*pass* in the Analysis layer. All of the Analysis <-> Target
communication is through the TTI's type erased interface itself. While
the diff is large here, it is nothing more that code motion to make
types available in a header file for use in a different source file
within each target.
I've tried to keep all the doxygen comments and file boilerplate in line
with this move, but let me know if I missed anything.
With this in place, the next step to making TTI work with the new pass
manager is to introduce a really simple new-style analysis that produces
a TTI object via a callback into this routine on the target machine.
Once we have that, we'll have the building blocks necessary to accept
a function argument as well.
llvm-svn: 227685
type erased interface and a single analysis pass rather than an
extremely complex analysis group.
The end result is that the TTI analysis can contain a type erased
implementation that supports the polymorphic TTI interface. We can build
one from a target-specific implementation or from a dummy one in the IR.
I've also factored all of the code into "mix-in"-able base classes,
including CRTP base classes to facilitate calling back up to the most
specialized form when delegating horizontally across the surface. These
aren't as clean as I would like and I'm planning to work on cleaning
some of this up, but I wanted to start by putting into the right form.
There are a number of reasons for this change, and this particular
design. The first and foremost reason is that an analysis group is
complete overkill, and the chaining delegation strategy was so opaque,
confusing, and high overhead that TTI was suffering greatly for it.
Several of the TTI functions had failed to be implemented in all places
because of the chaining-based delegation making there be no checking of
this. A few other functions were implemented with incorrect delegation.
The message to me was very clear working on this -- the delegation and
analysis group structure was too confusing to be useful here.
The other reason of course is that this is *much* more natural fit for
the new pass manager. This will lay the ground work for a type-erased
per-function info object that can look up the correct subtarget and even
cache it.
Yet another benefit is that this will significantly simplify the
interaction of the pass managers and the TargetMachine. See the future
work below.
The downside of this change is that it is very, very verbose. I'm going
to work to improve that, but it is somewhat an implementation necessity
in C++ to do type erasure. =/ I discussed this design really extensively
with Eric and Hal prior to going down this path, and afterward showed
them the result. No one was really thrilled with it, but there doesn't
seem to be a substantially better alternative. Using a base class and
virtual method dispatch would make the code much shorter, but as
discussed in the update to the programmer's manual and elsewhere,
a polymorphic interface feels like the more principled approach even if
this is perhaps the least compelling example of it. ;]
Ultimately, there is still a lot more to be done here, but this was the
huge chunk that I couldn't really split things out of because this was
the interface change to TTI. I've tried to minimize all the other parts
of this. The follow up work should include at least:
1) Improving the TargetMachine interface by having it directly return
a TTI object. Because we have a non-pass object with value semantics
and an internal type erasure mechanism, we can narrow the interface
of the TargetMachine to *just* do what we need: build and return
a TTI object that we can then insert into the pass pipeline.
2) Make the TTI object be fully specialized for a particular function.
This will include splitting off a minimal form of it which is
sufficient for the inliner and the old pass manager.
3) Add a new pass manager analysis which produces TTI objects from the
target machine for each function. This may actually be done as part
of #2 in order to use the new analysis to implement #2.
4) Work on narrowing the API between TTI and the targets so that it is
easier to understand and less verbose to type erase.
5) Work on narrowing the API between TTI and its clients so that it is
easier to understand and less verbose to forward.
6) Try to improve the CRTP-based delegation. I feel like this code is
just a bit messy and exacerbating the complexity of implementing
the TTI in each target.
Many thanks to Eric and Hal for their help here. I ended up blocked on
this somewhat more abruptly than I expected, and so I appreciate getting
it sorted out very quickly.
Differential Revision: http://reviews.llvm.org/D7293
llvm-svn: 227669
Cross-class copies being expensive is actually a trait of the microarchitecture, but as I haven't yet seen an example of a microarchitecture where they're cheap it seems best to just enable this by default, covering the non-mcpu build case.
llvm-svn: 217674
"Unroll" is not the appropriate name for this variable. Clang already uses
the term "interleave" in pragmas and metadata for this.
Differential Revision: http://reviews.llvm.org/D5066
llvm-svn: 217528
This patch adds support to recognize division by uniform power of 2 and modifies the cost table to vectorize division by uniform power of 2 whenever possible.
Updates Cost model for Loop and SLP Vectorizer.The cost table is currently only updated for X86 backend.
Thanks to Hal, Andrea, Sanjay for the review. (http://reviews.llvm.org/D4971)
llvm-svn: 216371
This patch adds support to recognize patterns such as fadd,fsub,fadd,fsub.../add,sub,add,sub... and
vectorizes them as vector shuffles if they are profitable.
These patterns of vector shuffle can later be converted to instructions such as addsubpd etc on X86.
Thanks to Arnold and Hal for the reviews. http://reviews.llvm.org/D4015
llvm-svn: 211339
the stack of the analysis group because they are all immutable passes.
This is made clear by Craig's recent work to use override
systematically -- we weren't overriding anything for 'finalizePass'
because there is no such thing.
This is kind of a lame restriction on the API -- we can no longer push
and pop things, we just set up the stack and run. However, I'm not
invested in building some better solution on top of the existing
(terrifying) immutable pass and legacy pass manager.
llvm-svn: 203437
Sam Parker