This change introduces a new IR intrinsic named `llvm.pseudoprobe` for pseudo-probe block instrumentation. Please refer to https://reviews.llvm.org/D86193 for the whole story.
A pseudo probe is used to collect the execution count of the block where the probe is instrumented. This requires a pseudo probe to be persisting. The LLVM PGO instrumentation also instruments in similar places by placing a counter in the form of atomic read/write operations or runtime helper calls. While these operations are very persisting or optimization-resilient, in theory we can borrow the atomic read/write implementation from PGO counters and cut it off at the end of compilation with all the atomics converted into binary data. This was our initial design and we’ve seen promising sample correlation quality with it. However, the atomics approach has a couple issues:
1. IR Optimizations are blocked unexpectedly. Those atomic instructions are not going to be physically present in the binary code, but since they are on the IR till very end of compilation, they can still prevent certain IR optimizations and result in lower code quality.
2. The counter atomics may not be fully cleaned up from the code stream eventually.
3. Extra work is needed for re-targeting.
We choose to implement pseudo probes based on a special LLVM intrinsic, which is expected to have most of the semantics that comes with an atomic operation but does not block desired optimizations as much as possible. More specifically the semantics associated with the new intrinsic enforces a pseudo probe to be virtually executed exactly the same number of times before and after an IR optimization. The intrinsic also comes with certain flags that are carefully chosen so that the places they are probing are not going to be messed up by the optimizer while most of the IR optimizations still work. The core flags given to the special intrinsic is `IntrInaccessibleMemOnly`, which means the intrinsic accesses memory and does have a side effect so that it is not removable, but is does not access memory locations that are accessible by any original instructions. This way the intrinsic does not alias with any original instruction and thus it does not block optimizations as much as an atomic operation does. We also assign a function GUID and a block index to an intrinsic so that they are uniquely identified and not merged in order to achieve good correlation quality.
Let's now look at an example. Given the following LLVM IR:
```
define internal void @foo2(i32 %x, void (i32)* %f) !dbg !4 {
bb0:
%cmp = icmp eq i32 %x, 0
br i1 %cmp, label %bb1, label %bb2
bb1:
br label %bb3
bb2:
br label %bb3
bb3:
ret void
}
```
The instrumented IR will look like below. Note that each `llvm.pseudoprobe` intrinsic call represents a pseudo probe at a block, of which the first parameter is the GUID of the probe’s owner function and the second parameter is the probe’s ID.
```
define internal void @foo2(i32 %x, void (i32)* %f) !dbg !4 {
bb0:
%cmp = icmp eq i32 %x, 0
call void @llvm.pseudoprobe(i64 837061429793323041, i64 1)
br i1 %cmp, label %bb1, label %bb2
bb1:
call void @llvm.pseudoprobe(i64 837061429793323041, i64 2)
br label %bb3
bb2:
call void @llvm.pseudoprobe(i64 837061429793323041, i64 3)
br label %bb3
bb3:
call void @llvm.pseudoprobe(i64 837061429793323041, i64 4)
ret void
}
```
Reviewed By: wmi
Differential Revision: https://reviews.llvm.org/D86490
We do not need to use the implicit cast here. We can instead can rely on
a comparison between two TypeSize objects instead. This algorithm will
work fine with scalable vectors.
Reviewed By: DavidTruby
Differential Revision: https://reviews.llvm.org/D90146
All existing SCEV cast types operate on integers.
D89456 will add SCEVPtrToIntExpr cast expression type.
I believe this is best for consistency.
Reviewed By: mkazantsev
Differential Revision: https://reviews.llvm.org/D89455
Bail from maskIsAllZeroOrUndef and maskIsAllOneOrUndef prior to iterating over the number of
elements for scalable vectors.
Assert that the mask type is not scalable in possiblyDemandedEltsInMask .
Assert that the types are correct in all three functions.
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D87424
This helps SelectionDAGBuilder recognize the splat can be used as a uniform base.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D86371
There's a potential motivating case to increase this limit in PR47191:
http://bugs.llvm.org/PR47191
But first we should make it less hacky. The limit in InstCombine is directly tied
to this value because an increase there can cause asserts in the underlying value
tracking calls if not changed together. The usage in VectorUtils is independent,
but the comment suggests that we should use the same value unless there's a known
reason to diverge. There are similar limits in codegen analysis, but I think we
should leave those independent in case we intentionally want the optimization
power/cost to be different there.
Differential Revision: https://reviews.llvm.org/D86113
This intrinsic implements IEEE-754 operation roundToIntegralTiesToEven,
and performs rounding to the nearest integer value, rounding halfway
cases to even. The intrinsic represents the missed case of IEEE-754
rounding operations and now llvm provides full support of the rounding
operations defined by the standard.
Differential Revision: https://reviews.llvm.org/D75670
Now that load/store alignment is required, we no longer need most
of them. Also switch the getLoadStoreAlignment() helper to return
Align instead of MaybeAlign.
Summary:
This change exposes the vector name mangling with LLVM ISA (used as part
of vector-function-abi-variant) as a utility.
This can then be used by front-ends that add this attribute.
Note that all parameters passed in to the function will be mangled with
the "v" token to identify that they are of of vector type. So, it is the
responsibility of the caller to confirm that all parameters in the
vectorized variant is of vector type.
Added unit test to show vector name mangling.
Reviewed-By: fpetrogalli, simoll
Differential Revision: https://reviews.llvm.org/D79867
Summary:
This patch fix the following issues with visitExtractElementInst:
1. Restrict VectorUtils::findScalarElement to fixed-length vector.
For scalable type, the number of elements in shuffle mask is
unknown at compile-time.
2. Fix out-of-range calculation for fixed-length vector.
3. Skip scalable type when analysis rely on fixed number of elements.
4. Add unit tests to check functionality of extractelement for scalable type.
Reviewers: sdesmalen, efriedma, spatel, nikic
Reviewed By: efriedma
Subscribers: tschuett, hiraditya, rkruppe, psnobl, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D78267
DelSet was used to avoid invalidating the current iterator while
modifying the map we are iterating over.
By using an early_inc_range, (which increments to iterator 'early',
allowing us to remove the current element), we can get rid of DelSet.
Reviewers: gilr, rengolin, Ayal, hsaito
Reviewed By: Ayal
Differential Revision: https://reviews.llvm.org/D78420
Cost-modeling decisions are tied to the compute interleave groups
(widening decisions, scalar and uniform values). When invalidating the
interleave groups, those decisions also need to be invalidated.
Otherwise there is a mis-match during VPlan construction.
VPWidenMemoryRecipes created initially are left around w/o converting them
into VPInterleave recipes. Such a conversion indeed should not take place,
and these gather/scatter recipes may in fact be right. The crux is leaving around
obsolete CM_Interleave (and dependent) markings of instructions along with
their costs, instead of recalculating decisions, costs, and recipes.
Alternatively to forcing a complete recompute later on, we could try
to selectively invalidate the decisions connected to the interleave
groups. But we would likely need to run the uniform/scalar value
detection parts again anyways and the extra complexity is probably not
worth it.
Fixes PR45572.
Reviewers: gilr, rengolin, Ayal, hsaito
Reviewed By: Ayal
Differential Revision: https://reviews.llvm.org/D78298
This is similar to the recent move/addition of "scaleShuffleMask" (D76508),
but there are a couple of differences:
1. The existing x86 helper (canWidenShuffleElements) always tries to
divide-by-2, so it gets called iteratively and wouldn't handle the
general case of non-pow-2 length.
2. The existing x86 code handles "SM_SentinelZero" - we don't have
that in IR, but this code should be safe to use with that or other
special (negative) values.
The motivation is to enable shuffle folds in instcombine/vector-combine
that are similar to D76844 and D76727, but in the reverse-bitcast direction.
Those patterns are visible in the tests for D40633.
Differential Revision: https://reviews.llvm.org/D77881
As proposed in D77881, we'll have the related widening operation,
so this name becomes too vague.
While here, change the function signature to take an 'int' rather
than 'size_t' for the scaling factor, add an assert for overflow of
32-bits, and improve the documentation comments.
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: sunfish, sdesmalen, efriedma
Reviewed By: efriedma
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D77273
Summary: These were templated due to SelectionDAG using int masks for shuffles and IR using unsigned masks for shuffles. But now that D72467 has landed we have an int mask version of IRBuilder::CreateShuffleVector. So just use int instead of a template
Reviewers: spatel, efriedma, RKSimon
Reviewed By: efriedma
Subscribers: hiraditya, llvm-commits
Differential Revision: https://reviews.llvm.org/D77183
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
Summary:
For scalable vector, index out-of-bound can not be determined at compile-time.
The same apply for VectorUtil findScalarElement().
Add test cases to check the functionality of SimplifyInsert/ExtractElementInst for scalable vector.
Reviewers: sdesmalen, efriedma, spatel, apazos
Reviewed By: efriedma
Subscribers: cameron.mcinally, tschuett, hiraditya, rkruppe, psnobl, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75782
I was debug stepping through an x86 shuffle lowering and
noticed we were doing an N^2 search for splat index. I
didn't find the equivalent functionality anywhere else in
LLVM, so here's a helper that takes an array of int and
returns a splatted index while ignoring undefs (any
negative value).
This might also be used inside existing
ShuffleVectorInst/ShuffleVectorSDNode functions and/or
help with D72467.
Differential Revision: https://reviews.llvm.org/D74064
We want to allow splat value transforms to improve PR44588 and related bugs:
https://bugs.llvm.org/show_bug.cgi?id=44588
...but to do that, we need to know if values are splatted from the same,
specific index (lane) rather than splatted from an arbitrary index.
We can improve the undef handling with 1-liner follow-ups because the
Constant API optionally allow undefs now.
Differential Revision: https://reviews.llvm.org/D73549
Summary:
This patch makes sure that the field VFShape.VF is greater than zero
when demangling the vector function name of scalable vector functions
encoded in the "vector-function-abi-variant" attribute.
This change is required to be able to provide instances of VFShape
that can be used to query the VFDatabase for the vectorization passes,
as such passes always require a positive value for the Vectorization Factor (VF)
needed by the vectorization process.
It is not possible to extract the value of VFShape.VF from the mangled
name of scalable vector functions, because it is encoded as
`x`. Therefore, the VFABI demangling function has been modified to
extract such information from the IR declaration of the vector
function, under the assumption that _all_ vectors in the signature of
the vector function have the same number of lanes. Such assumption is
valid because it is also assumed by the Vector Function ABI
specifications supported by the demangling function (x86, AArch64, and
LLVM internal one).
The unit tests that demangle scalable names have been modified by
adding the IR module that carries the declaration of the vector
function name being demangled.
In particular, the demangling function fails in the following cases:
1. When the declaration of the scalable vector function is not
present in the module.
2. When the value of VFSHape.VF is not greater than 0.
Reviewers: jdoerfert, sdesmalen, andwar
Reviewed By: jdoerfert
Subscribers: mgorny, kristof.beyls, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73286
This is how it should've been and brings it more in line with
std::string_view. There should be no functional change here.
This is mostly mechanical from a custom clang-tidy check, with a lot of
manual fixups. It uncovers a lot of minor inefficiencies.
This doesn't actually modify StringRef yet, I'll do that in a follow-up.
Summary:
This commits is a rework of the patch in
https://reviews.llvm.org/D67572.
The rework was requested to prevent out-of-tree performance regression
when vectorizing out-of-tree IR intrinsics. The vectorization of such
intrinsics is enquired via the static function `isTLIScalarize`. For
detail see the discussion in https://reviews.llvm.org/D67572.
Reviewers: uabelho, fhahn, sdesmalen
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72734
This reverts commit 0be81968a2.
The VFDatabase needs some rework to be able to handle vectorization
and subsequent scalarization of intrinsics in out-of-tree versions of
the compiler. For more details, see the discussion in
https://reviews.llvm.org/D67572.
This patch introduced the VFDatabase, the framework proposed in
http://lists.llvm.org/pipermail/llvm-dev/2019-June/133484.html. [*]
In this patch the VFDatabase is used to bridge the TargetLibraryInfo
(TLI) calls that were previously used to query for the availability of
vector counterparts of scalar functions.
The VFISAKind field `ISA` of VFShape have been moved into into VFInfo,
under the assumption that different vector ISAs may provide the same
vector signature. At the moment, the vectorizer accepts any of the
available ISAs as long as the signature provided by the VFDatabase
matches the one expected in the vectorization process. For example,
when targeting AVX or AVX2, which both have 256-bit registers, the IR
signature of the two vector functions associated to the two ISAs is
the same. The `getVectorizedFunction` method at the moment returns the
first available match. We will need to add more heuristics to the
search system to decide which of the available version (TLI, AVX,
AVX2, ...) the system should prefer, when multiple versions with the
same VFShape are present.
Some of the code in this patch is based on the work done by Sumedh
Arani in https://reviews.llvm.org/D66025.
[*] Notice that in the proposal the VFDatabase was called SVFS. The
name VFDatabase is more in line with LLVM recommendations for
naming classes and variables.
Differential Revision: https://reviews.llvm.org/D67572
Summary:
This patch introduces an API to build and modify vector shapes.
The validity of a VFShape can be checked with the
`hasValidParameterList` method, which is also run in an assertion each
time a VFShape is modified.
The field VFISAKind has been moved to VFInfo under the assumption that
different ISAs can map to the same VFShape (as it can be in the case
of vector extensions with the same registers size, for example AVX and
AVX2).
Reviewers: sdesmalen, jdoerfert, simoll, hsaito
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D70513
This patch introduces a function pass to inject the scalar-to-vector
mappings stored in the TargetLIbraryInfo (TLI) into the Vector
Function ABI (VFABI) variants attribute.
The test is testing the injection for three vector libraries supported
by the TLI (Accelerate, SVML, MASSV).
The pass does not change any of the analysis associated to the
function.
Differential Revision: https://reviews.llvm.org/D70107
The attribute is stored at the `FunctionIndex` attribute set, with the
name "vector-function-abi-variant".
The get/set methods of the attribute have assertion to verify that:
1. Each name in the attribute is a valid VFABI mangled name.
2. Each name in the attribute correspond to a function declared in the
module.
Differential Revision: https://reviews.llvm.org/D69976
Summary:
Add an intrinsic that takes 2 unsigned integers with
the scale of them provided as the third argument and
performs fixed point multiplication on them. The
result is saturated and clamped between the largest and
smallest representable values of the first 2 operands.
This is a part of implementing fixed point arithmetic
in clang where some of the more complex operations
will be implemented as intrinsics.
Patch by: leonardchan, bjope
Reviewers: RKSimon, craig.topper, bevinh, leonardchan, lebedev.ri, spatel
Reviewed By: leonardchan
Subscribers: ychen, wuzish, nemanjai, MaskRay, jsji, jdoerfert, Ka-Ka, hiraditya, rjmccall, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D57836
llvm-svn: 371308
Summary:
Handle smul.fix.sat in the scalarizer. This is done by
adding smul.fix.sat to the set of "isTriviallyVectorizable"
intrinsics.
The addition of smul.fix.sat in isTriviallyVectorizable and
hasVectorInstrinsicScalarOpd can also be seen as a preparation
to be able to use hasVectorInstrinsicScalarOpd in ConstantFolding.
Reviewers: rengolin, RKSimon, dblaikie
Reviewed By: rengolin
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D63704
llvm-svn: 364177
We have the related getSplatValue() already in IR (see code just above the proposed addition).
But sometimes we only need to know that the value is a splat rather than capture the splatted
scalar value. Also, we have an isSplatValue() function already in SDAG.
Motivation - recent bugs that would potentially benefit from improved splat analysis in IR:
https://bugs.llvm.org/show_bug.cgi?id=37428https://bugs.llvm.org/show_bug.cgi?id=42174
Differential Revision: https://reviews.llvm.org/D63138
llvm-svn: 363106
This requires a couple of tweaks to existing vectorization functions as they were assuming that only the second call argument (ctlz/cttz/powi) could ever be the 'always scalar' argument, but for smul.fix + umul.fix its the third argument.
Differential Revision: https://reviews.llvm.org/D58616
llvm-svn: 354790
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
Hongtao Yu