Found a problem in indirect call promotion in sample loader pass. Currently
if an indirect call is promoted for a target, and if the parent function is
inlined into some other function, the indirect call can be promoted for the
same target again. That is redundent which can harm performance and can cause
excessive compile time in some extreme case.
The patch fixes the issue. If a target is promoted for an indirect call, the
patch will write ICP metadata with the target call count being set to 0.
In the later ICP in sample profile loader, if it sees a target has 0 count
for an indirect call, it knows the target has been promoted and won't do
indirect call promotion for the indirect call.
The fix brings 0.1~0.2% performance on our search benchmark.
Differential Revision: https://reviews.llvm.org/D96806
We can always look through single-argument (LCSSA) phi nodes when
performing alias analysis. getUnderlyingObject() already does this,
but stripPointerCastsAndInvariantGroups() does not. We still look
through these phi nodes with the usual aliasPhi() logic, but
sometimes get sub-optimal results due to the restrictions on value
equivalence when looking through arbitrary phi nodes. I think it's
generally beneficial to keep the underlying object logic and the
pointer cast stripping logic in sync, insofar as it is possible.
With this patch we get marginally better results:
aa.NumMayAlias | 5010069 | 5009861
aa.NumMustAlias | 347518 | 347674
aa.NumNoAlias | 27201336 | 27201528
...
licm.NumPromoted | 1293 | 1296
I've renamed the relevant strip method to stripPointerCastsForAliasAnalysis(),
as we're past the point where we can explicitly spell out everything
that's getting stripped.
Differential Revision: https://reviews.llvm.org/D96668
SROA does not correctly account for offsets in TBAA/TBAA struct metadata.
This patch creates functionality for generating new MD with the corresponding
offset and updates SROA to use this functionality.
Differential Revision: https://reviews.llvm.org/D95826
This patch enables scalable vectorization of loops with integer/fast reductions, e.g:
```
unsigned sum = 0;
for (int i = 0; i < n; ++i) {
sum += a[i];
}
```
A new TTI interface, isLegalToVectorizeReduction, has been added to prevent
reductions which are not supported for scalable types from vectorizing.
If the reduction is not supported for a given scalable VF,
computeFeasibleMaxVF will fall back to using fixed-width vectorization.
Reviewed By: david-arm, fhahn, dmgreen
Differential Revision: https://reviews.llvm.org/D95245
The GPUDivergenceAnalysis is now renamed to just "DivergenceAnalysis"
since there is no conflict with LegacyDivergenceAnalysis. In the
legacy PM, this analysis can only be used through the legacy DA
serving as a wrapper. It is now made available as a pass in the new
PM, and has no relation with the legacy DA.
The new DA currently cannot handle irreducible control flow; its
presence can cause the analysis to run indefinitely. The analysis is
now modified to detect this and report all instructions in the
function as divergent. This is super conservative, but allows the
analysis to be used without hanging the compiler.
Reviewed By: aeubanks
Differential Revision: https://reviews.llvm.org/D96615
In the motivating example from https://llvm.org/PR49171 and
reduced test here, we would unroll and clone assumes so much
that compile-time effectively became infinite while analyzing
all of those assumes.
Currently, setting the `no-nans-fp-math` attribute to true will allow
loops with fmin/fmax to vectorize, though we should be requiring that
`no-signed-zeros-fp-math` is also set.
This patch adds the check for no-signed-zeros at the function level and includes
tests to make sure we don't vectorize functions with only one of the attributes
associated.
Reviewed By: spatel
Differential Revision: https://reviews.llvm.org/D96604
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
This refactors shouldFavorPostInc() and shouldFavorBackedgeIndex() into
getPreferredAddressingMode() so that we have one interface to steer LSR in
generating the preferred addressing mode.
Differential Revision: https://reviews.llvm.org/D96600
At this point, we can treat the case of GEP/GEP aliasing and
GEP/non-GEP aliasing in essentially the same way. The only
differences are that we need to do an additional negative GEP base
check, and that we perform a bailout on unknown sizes for the
GEP/non-GEP case (the latter exists only to limit compile-time).
This change is not quite NFC due to the peculiar effect that
the DecomposedGEP for V2 can actually be non-trivial even if V2
is not a GEP. The reason for this is that getUnderlyingObject()
can look through LCSSA phi nodes, while stripPointerCasts() doesn't.
This can lead to slightly better results if single-entry phi nodes
occur inside a loop, where looking through the phi node via aliasPhi()
would subject it to phi cycle equivalence restrictions. It would
probably make sense to adjust pointer cast stripping (for AA) to
handle this case, and ensure consistent results.
For two GEPs with identical offsets, we currently first perform
a base address query without size information, and then if it is
MayAlias, perform another with size information. This is pointless,
as the latter query should produce strictly better results.
This was not quite true historically due to the way that NoAlias
assumptions were handled, but that issue has since been resolved.
We currently detect GEPs that have exactly the same indexes by
comparing the Offsets and VarIndices. However, the latter implicitly
performs equality comparisons between two values, which is not
generally legal inside BasicAA, due to the possibility of comparisons
across phi cycles.
I believe that in this particular instance this actually ends up being
unproblematic, at least I wasn't able to come up with any cases that
could result in an incorrect root query result.
In the interest of being defensive, compute GetIndexDifference earlier
(which knows how to handle phi cycles properly) and use the result of
that to determine whether the offsets are identical.
This is a follow-up of D95238's LangRef update.
This patch updates `programUndefinedIfUndefOrPoison(V)` to return true if
`V` is used by any memory-accessing instruction.
Interestingly, this affected many tests in Attributors, mainly about adding noundefs.
The tests are updated using llvm/utils/update_test_checks.py. I checked that the diffs
are about updating noundefs.
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D96642
Instcombine will convert the nonnull and alignment assumption that use the boolean condtion
to an assumption that uses the operand bundles when knowledge retention is enabled.
Differential Revision: https://reviews.llvm.org/D82703
Rather than storing the query depth in AAResults, store it in AAQI.
This makes more sense, as it is a property of the query. This
sidesteps the issue of D94363, fixing slightly inaccurate AA
statistics. Additionally, I plan to use the Depth from BasicAA in
the future, where fetching it from AAResults would be unreliable.
This change is not quite as straightforward as it seems, because
we need to preserve the depth when creating a new AAQI for recursive
queries across phis. I'm adding a new method for this, as we may
need to preserve additional information here in the future.
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
Changes `getScalarizationOverhead` to return an invalid cost for scalable VFs
and adds some simple tests for loops containing a function for which
there is a vectorized variant available.
Reviewed By: david-arm
Differential Revision: https://reviews.llvm.org/D96356
The vector reduction intrinsics started life as experimental ops, so backend support
was lacking. As part of promoting them to 1st-class intrinsics, however, codegen
support was added/improved:
D58015
D90247
So I think it is safe to now remove this complication from IR.
Note that we still have an IR-level codegen expansion pass for these as discussed
in D95690. Removing that is another step in simplifying the logic. Also note that
x86 was already unconditionally forming reductions in IR, so there should be no
difference for x86.
I spot checked a couple of the tests here by running them through opt+llc and did
not see any asm diffs.
If we do find functional differences for other targets, it should be possible
to (at least temporarily) restore the shuffle IR with the ExpandReductions IR
pass.
Differential Revision: https://reviews.llvm.org/D96552
This patch changes the VecDesc struct to use ElementCount
instead of an unsigned VF value, in preparation for
future work that adds support for vectorized versions of
math functions using scalable vectors. Since all I'm doing
in this patch is switching the type I believe it's a
non-functional change. I changed getWidestVF to now return
both the widest fixed-width and scalable VF values, but
currently the widest scalable value will be zero.
Differential Revision: https://reviews.llvm.org/D96011
The motivation for this is that I'm looking at an example that uses shifts as induction variables. There's lots of other omissions, but one of the first I noticed is that we can't compute tight known bits. (This indirectly causes SCEV's range analysis to produce very poor results as well.)
Differential Revision: https://reviews.llvm.org/D96440
This reverts commit b7d870eae7 and the
subsequent fix "[Polly] Fix build after AssumptionCache change (D96168)"
(commit e6810cab09).
It caused indeterminism in the output, such that e.g. the
polly-x86_64-linux buildbot failed accasionally.
This will be needed in the loop-vectorizer where the minimum VF
requested may be a scalable VF. getMinimumVF now takes an additional
operand 'IsScalableVF' that indicates whether a scalable VF is required.
Reviewed By: kparzysz, rampitec
Differential Revision: https://reviews.llvm.org/D96020
The AssumptionCache mechanism is used to feed assumes into known bits computations. Most places in SCEV passed it in, but one place appears to have been missed.
Spotted via inspection, don't have a test case which actually exercises this, but it seemed like an obvious fixit.
Instcombine will convert the nonnull and alignment assumption that use the boolean condtion
to an assumption that uses the operand bundles when knowledge retention is enabled.
Differential Revision: https://reviews.llvm.org/D82703
When doing some recent debugging of the IROutliner, and using the similarity pass for debugging, just having the basic block and function isn't really enough to get all the information. This adds the first and last instruction to the output of the IRSimilarityPrinting pass to give better information to a user.
Reviewer: paquette
Differential Revision: https://reviews.llvm.org/D94304
This is based on the example/comments in:
https://llvm.org/PR48984
I tried just lifting the restriction in computeKnownBitsFromShiftOperator()
as suggested in the bug report, but that doesn't catch all of the cases
shown here. I didn't step through to see exactly why that happened. But it
seems like a reasonable compromise to cheaply check the special-case of
shifting a constant.
There's a slight regression on a cmp transform as noted, but this is likely
the more important/common pattern, so we can fix that icmp pattern later if
needed.
Differential Revision: https://reviews.llvm.org/D95959
This reverts commit 502a67dd7f.
This expose a failure in test-suite build on PowerPC,
revert to unblock buildbot first,
Dave will re-commit in https://reviews.llvm.org/D96287.
Thanks Dave.
PR49043 exposed a problem when it comes to RAUW llvm.assumes. While
D96106 would fix it for GVNSink, it seems a more general concern. To
avoid future problems this patch moves away from the vector of weak
reference model used in the assumption cache. Instead, we track the
llvm.assume calls with a callback handle which will remove itself from
the cache if the call is deleted.
Fixes PR49043.
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D96168
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
getIntrinsicInstrCost takes a IntrinsicCostAttributes holding various
parameters of the intrinsic being costed. It can either be called with a
scalar intrinsic (RetTy==Scalar, VF==1), with a vector instruction
(RetTy==Vector, VF==1) or from the vectorizer with a scalar type and
vector width (RetTy==Scalar, VF>1). A RetTy==Vector, VF>1 is considered
an error. Both of the vector modes are expected to be treated the same,
but because this is confusing many backends end up getting it wrong.
Instead of trying work with those two values separately this removes the
VF parameter, widening the RetTy/ArgTys by VF used called from the
vectorizer. This keeps things simpler, but does require some other
modifications to keep things consistent.
Most backends look like this will be an improvement (or were not using
getIntrinsicInstrCost). AMDGPU needed the most changes to keep the code
from c230965ccf working. ARM removed the fix in
dfac521da1, webassembly happens to get a fixup for an SLP cost
issue and both X86 and AArch64 seem to now be using better costs from
the vectorizer.
Differential Revision: https://reviews.llvm.org/D95291
MemorySSA currently treats lifetime.end intrinsics as not aliasing
anything. This breaks MemorySSA-based MemCpyOpt, because we'll happily
move a read of a pointer below a lifetime.end intrinsic, as no clobber
is reported.
I think the MemorySSA modelling here isn't correct: lifetime.end(p)
has approximately the same effect as doing a memcpy(p, undef), and
should be treated as a clobber.
This patch removes the special handling of lifetime.end, leaving
alias analysis to handle it appropriately.
Differential Revision: https://reviews.llvm.org/D95763
Extend applyLoopGuards() to take into account conditions/assumes proving some
value %v to be divisible by D by rewriting %v to (%v / D) * D. This lets the
loop unroller and the loop vectorizer identify more loops as not requiring
remainder loops.
Differential Revision: https://reviews.llvm.org/D95521
This is another step (see D95452) towards correcting fast-math-flags
bugs in vector reductions.
There are multiple bugs visible in the test diffs, and this is still
not working as it should. We still use function attributes (rather
than FMF) to drive part of the logic, but we are not checking for
the correct FP function attributes.
Note that FMF may not be propagated optimally on selects (example
in https://llvm.org/PR35607 ). That's why I'm proposing to union the
FMF of a fcmp+select pair and avoid regressions on existing vectorizer
tests.
Differential Revision: https://reviews.llvm.org/D95690
This is a (rather delayed) follow up to commit 0129cd5. This commit is entirely NFC, the semantic change to leverage the new information will be submitted separate with a test case.
We use `EquivalenceClasses` to cache the notion that two SCEVs are equivalent,
so save time in situation when `A` is equivalent to `B` and `B` is equivalent to `C`,
making check "if `A` is equivalent to `C`?" cheaper.
We also return `0` in the comparator when we reach max analysis depth to save
compile time. After doing this, we also cache them as being equivalent.
Now, imagine the following situation:
- `A` is proved equivalent to `B`;
- `C` is proved equivalent to `D`;
- Comparison of `A` against `D` is proved non-zero;
- Comparison of `B` against `C` reaches max depth (and gets cached as equivalence).
Now, before the invocation of compare(`B`, `C`), `A` and `D` belonged
to different equivalence classes, and their comparison returned non-zero.
After the the invocation of compare(`B`, `C`), equivalence classes get merged
and `A`, `B`, `C` and `D` all fall into the same equivalence class. So the comparator
will change its behavior for couple `A` and `D`, with weird consequences following it.
This comparator is finally used in `std::stable_sort`, and this behavior change
makes it crash (looks like it's causing a memory corruption).
Solution: this patch changes `CompareSCEVComplexity` to return `None`
when the max depth is reached. So in this case, we do not cache these SCEVs
(and their parents in the tree) as being equivalent.
Differential Revision: https://reviews.llvm.org/D94654
Reviewed By: lebedev.ri
Imported functions and variable get the visibility from the module supplying the
definition. However, non-imported definitions do not get the visibility from
(ELF) the most constraining visibility among all modules (Mach-O) the visibility
of the prevailing definition.
This patch
* adds visibility bits to GlobalValueSummary::GVFlags
* computes the result visibility and propagates it to all definitions
Protected/hidden can imply dso_local which can enable some optimizations (this
is stronger than GVFlags::DSOLocal because the implied dso_local can be
leveraged for ELF -shared while default visibility dso_local has to be cleared
for ELF -shared).
Note: we don't have summaries for declarations, so for ELF if a declaration has
the most constraining visibility, the result visibility may not be that one.
Differential Revision: https://reviews.llvm.org/D92900
In computeLoadConstantCompareExitLimit, the addrec used to compute the
exit count should be from the loop which the exiting block belongs to.
Reviewed by: mkazantsev
Differential Revision: https://reviews.llvm.org/D92367
This change leverages the work done in D83743 to replay in the SampleProfile inliner to also be used in the CGSCC inliner. NOTE: currently restricted to non-ML advisors only.
The added switch `-cgscc-inline-replay=<remarks file>` will replay the inlining decisions in that file where the remarks file is generated via `-Rpass=inline`. The aim here is to make it easier to analyze changes that would modify inlining heuristics to be separated from this behavior. Doing so allows easier examination of assembly and runtime behavior compared to the baseline rather than trying to dig through the large churn caused by inlining.
In LTO compilation, since inlining is done twice you can separately specify replay by passing the flag to the FE (`-cgscc-inline-replay=`) and to the linker (`-Wl,cgscc-inline-replay=`) with the remarks generated from their respective places.
Testing on mysqld by comparing the inline decisions between base (generates remarks.txt) and diff (replay using identical input/tools with remarks.txt) and examining the inlining sites with `diff` shows 14,000 mismatches out of 247,341 for a ~94% replay accuracy. I believe this gap can be narrowed further though for the general case we may never achieve full accuracy. For my personal use, this is close enough to be representative: I set the baseline as the one generated by the replay on identical input/toolset and compare that to my modified input/toolset using the same replay.
Testing:
ninja check-llvm
newly added test correctly replays CGSCC inlining decisions
Reviewed By: mtrofin, wenlei
Differential Revision: https://reviews.llvm.org/D94334
This is similar to D94106, but for the
isGuaranteedToTransferExecutionToSuccessor() helper. We should not
assume that readonly functions will return, as this is only true for
mustprogress functions (in which case we already infer willreturn).
As with the DCE change, for now continue assuming that readonly
intrinsics will return, as not all target intrinsics have been
annotated yet.
Differential Revision: https://reviews.llvm.org/D95288
This is to support the memory routines vec_malloc, vec_calloc, vec_realloc, and vec_free. These routines manage memory that is 16-byte aligned. And they are only available on AIX.
Differential Revision: https://reviews.llvm.org/D94710
We tend to assume that the AA pipeline is by default the default AA
pipeline and it's confusing when it's empty instead.
PR48779
Initially reverted due to BasicAA running analyses in an unspecified
order (multiple function calls as parameters), fixed by fetching
analyses before the call to construct BasicAA.
Reviewed By: asbirlea
Differential Revision: https://reviews.llvm.org/D95117
Having a custom inliner doesn't really fit in with the new PM's
pipeline. It's also extra technical debt.
amdgpu-inline only does a couple of custom things compared to the normal
inliner:
1) It disables inlining if the number of BBs in a function would exceed
some limit
2) It increases the threshold if there are pointers to private arrays(?)
These can all be handled as TTI inliner hooks.
There already exists a hook for backends to multiply the inlining
threshold.
This way we can remove the custom amdgpu-inline pass.
This caused inline-hint.ll to fail, and after some investigation, it
looks like getInliningThresholdMultiplier() was previously getting
applied twice in amdgpu-inline (https://reviews.llvm.org/D62707 fixed it
not applying at all, so some later inliner change must have fixed
something), so I had to change the threshold in the test.
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D94153
This adds cost modelling for the inloop vectorization added in
745bf6cf44. Up until now they have been modelled as the original
underlying instruction, usually an add. This happens to works OK for MVE
with instructions that are reducing into the same type as they are
working on. But MVE's instructions can perform the equivalent of an
extended MLA as a single instruction:
%sa = sext <16 x i8> A to <16 x i32>
%sb = sext <16 x i8> B to <16 x i32>
%m = mul <16 x i32> %sa, %sb
%r = vecreduce.add(%m)
->
R = VMLADAV A, B
There are other instructions for performing add reductions of
v4i32/v8i16/v16i8 into i32 (VADDV), for doing the same with v4i32->i64
(VADDLV) and for performing a v4i32/v8i16 MLA into an i64 (VMLALDAV).
The i64 are particularly interesting as there are no native i64 add/mul
instructions, leading to the i64 add and mul naturally getting very
high costs.
Also worth mentioning, under NEON there is the concept of a sdot/udot
instruction which performs a partial reduction from a v16i8 to a v4i32.
They extend and mul/sum the first four elements from the inputs into the
first element of the output, repeating for each of the four output
lanes. They could possibly be represented in the same way as above in
llvm, so long as a vecreduce.add could perform a partial reduction. The
vectorizer would then produce a combination of in and outer loop
reductions to efficiently use the sdot and udot instructions. Although
this patch does not do that yet, it does suggest that separating the
input reduction type from the produced result type is a useful concept
to model. It also shows that a MLA reduction as a single instruction is
fairly common.
This patch attempt to improve the costmodelling of in-loop reductions
by:
- Adding some pattern matching in the loop vectorizer cost model to
match extended reduction patterns that are optionally extended and/or
MLA patterns. This marks the cost of the reduction instruction correctly
and the sext/zext/mul leading up to it as free, which is otherwise
difficult to tell and may get a very high cost. (In the long run this
can hopefully be replaced by vplan producing a single node and costing
it correctly, but that is not yet something that vplan can do).
- getExtendedAddReductionCost is added to query the cost of these
extended reduction patterns.
- Expanded the ARM costs to account for these expanded sizes, which is a
fairly simple change in itself.
- Some minor alterations to allow inloop reduction larger than the highest
vector width and i64 MVE reductions.
- An extra InLoopReductionImmediateChains map was added to the vectorizer
for it to efficiently detect which instructions are reductions in the
cost model.
- The tests have some updates to show what I believe is optimal
vectorization and where we are now.
Put together this can greatly improve performance for reduction loop
under MVE.
Differential Revision: https://reviews.llvm.org/D93476
This reverts commit d97f776be5.
The original problem was due to build failures in shared lib builds. D95079
moved ImportedFunctionsInliningStatistics under Analysis, unblocking
this.
This is related to D94982. We want to call these APIs from the Analysis
component, so we can't leave them under Transforms.
Differential Revision: https://reviews.llvm.org/D95079
When using 2 InlinePass instances in the same CGSCC - one for other
mandatory inlinings, the other for the heuristic-driven ones - the order
in which the ImportedFunctionStats would be output-ed would depend on
the destruction order of the inline passes, which is not deterministic.
This patch moves the ImportedFunctionStats responsibility to the
InlineAdvisor to address this problem.
Differential Revision: https://reviews.llvm.org/D94982
Currently LLVM is relying on ValueTracking's `isKnownNonZero` to attach `nonnull`, which can return true when the value is poison.
To make the semantics of `nonnull` consistent with the behavior of `isKnownNonZero`, this makes the semantics of `nonnull` to accept poison, and return poison if the input pointer isn't null.
This makes many transformations like below legal:
```
%p = gep inbounds %x, 1 ; % p is non-null pointer or poison
call void @f(%p) ; instcombine converts this to call void @f(nonnull %p)
```
Instead, this semantics makes propagation of `nonnull` to caller illegal.
The reason is that, passing poison to `nonnull` does not immediately raise UB anymore, so such program is still well defined, if the callee does not use the argument.
Having `noundef` attribute there re-allows this.
```
define void @f(i8* %p) { ; functionattr cannot mark %p nonnull here anymore
call void @g(i8* nonnull %p) ; .. because @g never raises UB if it never uses %p.
ret void
}
```
Another attribute that needs to be updated is `align`. This patch updates the semantics of align to accept poison as well.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D90529
Just like llvm.assume, there are a lot of cases where we can just ignore llvm.experimental.noalias.scope.decl.
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D93042
Split impliesPoison into two recursive walks, one over V, the
other over ValAssumedPoison. This allows us to reason about poison
implications in a number of additional cases that are important
in practice. This is a generalized form of D94859, which handles
the cmp to cmp implication in particular.
Differential Revision: https://reviews.llvm.org/D94866
83daa49758 made loop-rotate more conservative in the presence of
function calls in the prepare-for-lto stage. The code did not properly
account for calls that are no actual function calls, like calls to
intrinsics. This patch updates the code to ensure only calls that are
lowered to actual calls are considered inline candidates.
D84108 exposed a bad interaction between inlining and loop-rotation
during regular LTO, which is causing notable regressions in at least
CINT2006/473.astar.
The problem boils down to: we now rotate a loop just before the vectorizer
which requires duplicating a function call in the preheader when compiling
the individual files ('prepare for LTO'). But this then prevents further
inlining of the function during LTO.
This patch tries to resolve this issue by making LoopRotate more
conservative with respect to rotating loops that have inline-able calls
during the 'prepare for LTO' stage.
I think this change intuitively improves the current situation in
general. Loop-rotate tries hard to avoid creating headers that are 'too
big'. At the moment, it assumes all inlining already happened and the
cost of duplicating a call is equal to just doing the call. But with LTO,
inlining also happens during full LTO and it is possible that a previously
duplicated call is actually a huge function which gets inlined
during LTO.
From the perspective of LV, not much should change overall. Most loops
calling user-provided functions won't get vectorized to start with
(unless we can infer that the function does not touch memory, has no
other side effects). If we do not inline the 'inline-able' call during
the LTO stage, we merely delayed loop-rotation & vectorization. If we
inline during LTO, chances should be very high that the inlined code is
itself vectorizable or the user call was not vectorizable to start with.
There could of course be scenarios where we inline a sufficiently large
function with code not profitable to vectorize, which would have be
vectorized earlier (by scalarzing the call). But even in that case,
there probably is no big performance impact, because it should be mostly
down to the cost-model to reject vectorization in that case. And then
the version with scalarized calls should also not be beneficial. In a way,
LV should have strictly more information after inlining and make more
accurate decisions (barring cost-model issues).
There is of course plenty of room for things to go wrong unexpectedly,
so we need to keep a close look at actual performance and address any
follow-up issues.
I took a look at the impact on statistics for
MultiSource/SPEC2000/SPEC2006. There are a few benchmarks with fewer
loops rotated, but no change to the number of loops vectorized.
Reviewed By: sanwou01
Differential Revision: https://reviews.llvm.org/D94232
We can fold x*C1/C2 <= x to true if C1 <= C2. This is valid even
if the multiplication is not nuw: https://alive2.llvm.org/ce/z/vULors
The multiplication or division can be replaced by shifts. We don't
handle the case where both are shifts, as that should get folded
away by InstCombine.
There are no changes relative to the original commit. However, an issue
this exposed in BasicAA assumption tracking has been fixed in the
previous commit.
-----
An alias query currently works out roughly like this:
* Look up location pair in cache.
* Perform BasicAA logic (including cache lookup and insertion...)
* Perform a recursive query using BestAAResults.
* Look up location pair in cache (and thus do not recurse into BasicAA)
* Query all the other AA providers.
* Query all the other AA providers.
This is a lot of unnecessary work, all ultimately caused by the
BestAAResults query at the end of aliasCheck(). The reason we perform
it, is that aliasCheck() is getting called recursively, and we of
course want those recursive queries to also make use of other AA
providers, not just BasicAA. We can solve this by making the recursive
queries directly use BestAAResults (which will check both BasicAA
and other providers), rather than recursing into aliasCheck().
There are some tradeoffs:
* We can no longer pass through the precomputed underlying object
to aliasCheck(). This is not a major concern, because nowadays
getUnderlyingObject() is quite cheap.
* Results from other AA providers are no longer cached inside
BasicAA. The way this worked was already a bit iffy, in that a
result could be cached, but if it was MayAlias, we'd still end
up re-querying other providers anyway. If we want to cache
non-BasicAA results, we should do that in a more principled manner.
In any case, despite those tradeoffs, this works out to be a decent
compile-time improvment. I think it also simplifies the mental model
of how BasicAA works. It took me quite a while to fully understand
how these things interact.
Differential Revision: https://reviews.llvm.org/D90094
D91936 placed the tracking for the assumptions into BasicAA.
However, when recursing over phis, we may use fresh AAQI instances.
In this case AssumptionBasedResults from an inner AAQI can reesult
in a removal of an element from the outer AAQI.
To avoid this, move the tracking into AAQI. This generally makes
more sense, as the NoAlias assumptions themselves are also stored
in AAQI.
The test case only produces an assertion failure with D90094
reapplied. I think the issue exists independently of that change
as well, but I wasn't able to come up with a reproducer.
Expanding from D94808 - we ensure the same InlineAdvisor is used by both
InlinerPass instances. The notion of mandatory inlining is moved into
the core InlineAdvisor: advisors anyway have to handle that case, so
this change also factors out that a bit better.
Differential Revision: https://reviews.llvm.org/D94825
This reverts commit a3904cc77f.
It causes the compiler to crash while building Harfbuzz for ARM in
Chromium, reduced reproducer forthcoming:
https://crbug.com/1167305
An alias query currently works out roughly like this:
* Look up location pair in cache.
* Perform BasicAA logic (including cache lookup and insertion...)
* Perform a recursive query using BestAAResults.
* Look up location pair in cache (and thus do not recurse into BasicAA)
* Query all the other AA providers.
* Query all the other AA providers.
This is a lot of unnecessary work, all ultimately caused by the
BestAAResults query at the end of aliasCheck(). The reason we perform
it, is that aliasCheck() is getting called recursively, and we of
course want those recursive queries to also make use of other AA
providers, not just BasicAA. We can solve this by making the recursive
queries directly use BestAAResults (which will check both BasicAA
and other providers), rather than recursing into aliasCheck().
There are some tradeoffs:
* We can no longer pass through the precomputed underlying object
to aliasCheck(). This is not a major concern, because nowadays
getUnderlyingObject() is quite cheap.
* Results from other AA providers are no longer cached inside
BasicAA. The way this worked was already a bit iffy, in that a
result could be cached, but if it was MayAlias, we'd still end
up re-querying other providers anyway. If we want to cache
non-BasicAA results, we should do that in a more principled manner.
In any case, despite those tradeoffs, this works out to be a decent
compile-time improvment. I think it also simplifies the mental model
of how BasicAA works. It took me quite a while to fully understand
how these things interact.
Differential Revision: https://reviews.llvm.org/D90094
This change modifies the source location formatting from:
LineNumber.Discriminator
to:
LineNumber:ColumnNumber.Discriminator
The motivation here is to enhance location information for inline replay that currently exists for the SampleProfile inliner. This will be leveraged further in inline replay for the CGSCC inliner in the related diff.
The ReplayInlineAdvisor is also modified to read the new format and now takes into account the callee for greater accuracy.
Testing:
ninja check-llvm
Reviewed By: mtrofin
Differential Revision: https://reviews.llvm.org/D94333
This is a partial fix for https://bugs.llvm.org/show_bug.cgi?id=44403.
Folding gep p, q-p to q is only legal if p and q have the same
provenance. This fold should probably be guarded by something like
getUnderlyingObject(p) == getUnderlyingObject(q).
This patch is a partial fix that removes the special handling for
gep p, 0-p, which will fold to a null pointer, which would certainly
not pass an underlying object check (unless p is also null, in which
case this would fold trivially anyway). Folding to a null pointer
is particularly problematic due to the special handling it receives
in many places, making end-to-end miscompiles more likely.
Differential Revision: https://reviews.llvm.org/D93820
This patch fixes a bug that could result in miscompiles (at least
in an OOT target). The problem could be seen by adding checks that
the DominatorTree used in BasicAliasAnalysis and ValueTracking was
valid (e.g. by adding DT->verify() call before every DT dereference
and then running all tests in test/CodeGen).
Problem was that the LegacyPassManager calculated "last user"
incorrectly for passes such as the DominatorTree when not telling
the pass manager that there was a transitive dependency between
the different analyses. And then it could happen that an incorrect
dominator tree was used when doing alias analysis (which was a pretty
serious bug as the alias analysis result could be invalid).
Fixes: https://bugs.llvm.org/show_bug.cgi?id=48709
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D94138
Currently make_early_inc_range cannot be used with iterators with
operator* implementations that do not return a reference.
Most notably in the LLVM codebase, this means the User iterator ranges
cannot be used with make_early_inc_range, which slightly simplifies
iterating over ranges while elements are removed.
Instead of directly using BaseT::reference as return type of operator*,
this patch uses decltype to get the actual return type of the operator*
implementation in WrappedIteratorT.
This patch also updates a few places to use make use of
make_early_inc_range.
Reviewed By: dblaikie
Differential Revision: https://reviews.llvm.org/D93992
Similar to the Arm VCTP intrinsics, if the operands of an
active.lane.mask are both known, the constant lane mask can be
calculated. This can come up after unrolling the loops.
Differential Revision: https://reviews.llvm.org/D94103
When creating pi-blocks we try to avoid creating duplicate edges
between outside nodes and the pi-block when an edge is of the
same kind and direction as another one that has already been
created. We do this by keeping track of the edges in an
enumerated array called EdgeAlreadyCreated. The problem is that
this array is declared local to the loop that iterates over the
nodes in the pi-block, so the information gets lost every time a
new inside-node is iterated over. The fix is to move the
declaration to the outer loop.
Reviewed By: Meinersbur
Differential Revision: https://reviews.llvm.org/D94094
Add support for mixed pre/post CFG views.
Update usages of the MemorySSAUpdater to use the new DT API by
requesting the DT updates to be done by the MSSAUpdater.
Differential Revision: https://reviews.llvm.org/D93371
Change the way NoAlias assumptions in BasicAA are handled. Instead of
handling this inside the phi-phi code, always initially insert a
NoAlias result into the map and keep track whether it is used.
If it is used, then we require that we also get back NoAlias from
the recursive queries. Otherwise, the entry is changed to MayAlias.
Additionally, keep track of all location pairs we inserted that may
still be based on assumptions higher up. If it turns out one of those
assumptions is incorrect, we flush them from the cache.
The compile-time impact for the new implementation is significantly
higher than the previous iteration of this patch:
https://llvm-compile-time-tracker.com/compare.php?from=c0bb9859de6991cc233e2dedb978dd118da8c382&to=c07112373279143e37568b5bcd293daf81a35973&stat=instructions
However, it should avoid the exponential runtime cases we run into
if we don't cache assumption-based results entirely.
This also produces better results in some cases, because NoAlias
assumptions can now start at any root, rather than just phi-phi pairs.
This is not just relevant for analysis quality, but also for BatchAA
consistency: Otherwise, results would once again depend on query order,
though at least they wouldn't be wrong.
This ended up both more complicated and more expensive than I hoped,
but I wasn't able to come up with another solution that satisfies all
the constraints.
Differential Revision: https://reviews.llvm.org/D91936
I don't believe this has an observable effect, because the only
thing we care about here is replacing the operand with a constant
so following folds can apply. This change is just to make the
representation follow canonical unary shuffle form.
Calling null or undef results in immediate undefined behavior.
Return poison instead of undef in this case, similar to what
we do for immediate UB due to division by zero.
Make InstSimplify return poison rather than undef for out-of-bounds
shifts, as specified by LandRef:
> If op2 is (statically or dynamically) equal to or larger than the
> number of bits in op1, this instruction returns a poison value.
Differential Revision: https://reviews.llvm.org/D93998
Previously when trying to support CoroSplit's function splitting, we
added in a hack that simply added the new function's node into the
original function's SCC (https://reviews.llvm.org/D87798). This is
incorrect since it might be in its own SCC.
Now, more similar to the previous design, we have callers explicitly
notify the LazyCallGraph that a function has been split out from another
one.
In order to properly support CoroSplit, there are two ways functions can
be split out.
One is the normal expected "outlining" of one function into a new one.
The new function may only contain references to other functions that the
original did. The original function must reference the new function. The
new function may reference the original function, which can result in
the new function being in the same SCC as the original function. The
weird case is when the original function indirectly references the new
function, but the new function directly calls the original function,
resulting in the new SCC being a parent of the original function's SCC.
This form of function splitting works with CoroSplit's Switch ABI.
The second way of splitting is more specific to CoroSplit. CoroSplit's
Retcon and Async ABIs split the original function into multiple
functions that all reference each other and are referenced by the
original function. In order to keep the LazyCallGraph in a valid state,
all new functions must be processed together, else some nodes won't be
populated. To keep things simple, this only supports the case where all
new edges are ref edges, and every new function references every other
new function. There can be a reference back from any new function to the
original function, putting all functions in the same RefSCC.
This also adds asserts that all nodes in a (Ref)SCC can reach all other
nodes to prevent future incorrect hacks.
The original hacks in https://reviews.llvm.org/D87798 are no longer
necessary since all new functions should have been registered before
calling updateCGAndAnalysisManagerForPass.
This fixes all coroutine tests when opt's -enable-new-pm is true by
default. This also fixes PR48190, which was likely due to the previous
hack breaking SCC invariants.
Reviewed By: rnk
Differential Revision: https://reviews.llvm.org/D93828
This patch
- Adds containsPoisonElement that checks existence of poison in constant vector elements,
- Renames containsUndefElement to containsUndefOrPoisonElement to clarify its behavior & updates its uses properly
With this patch, isGuaranteedNotToBeUndefOrPoison's tests w.r.t constant vectors are added because its analysis is improved.
Thanks!
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D94053
This patch teaches the inliner to compute the full cost for a call
site where the newly introduced cost benefit analysis is enabled.
Note that the cost benefit analysis requires the full cost to be
computed. However, without this patch or the -inline-cost-full
option, the early termination logic would kick in when the cost
exceeds the threshold, so we don't get to perform the cost benefit
analysis. For this reason, we would need to specify four clang
options:
-mllvm -inline-cost-full
-mllvm -inline-enable-cost-benefit-analysis
This patch eliminates the need to specify -inline-cost-full.
Differential Revision: https://reviews.llvm.org/D93658
Allow loop nests with empty basic blocks without loops in different
levels as perfect.
Reviewers: Meinersbur
Differential Revision: https://reviews.llvm.org/D93665
This function no longer does anything useful. It probably did something originally but latter changes removed them and didn't clean up this function.
The checks are already done in the callers as well.
Differential Revision: https://reviews.llvm.org/D94055
While here, rename the inaccurate getRecurrenceBinOp()
because that was also used to get CmpInst opcodes.
The recurrence/reduction kind should always refer to the
expected opcode for a reduction. SLP appears to be the
only direct caller of createSimpleTargetReduction(), and
that calling code ideally should not be carrying around
both an opcode and a reduction kind.
This should allow us to generalize reduction matching to
use intrinsics instead of only binops.
Allow loop nests with empty basic blocks without loops in different
levels as perfect.
Reviewers: Meinersbur
Differential Revision: https://reviews.llvm.org/D93665
A new TTI interface has been added 'Optional <unsigned>getMaxVScale' that
returns the maximum vscale for a given target.
When known getMaxVScale is used to compute the cost of masked gather scatter
for scalable vector.
Depends on D92094
Differential Revision: https://reviews.llvm.org/D93030
As the comment already indicates, performing an operation with
nnan/ninf flags on a nan/inf or undef results in poison. Now that
we have a proper poison value, we no longer need to relax it to
undef.
Div/rem by zero is immediate undefined behavior and anything goes.
Currently we fold it to undef, this patch changes it to fold to
poison instead, which is slightly stronger.
Differential Revision: https://reviews.llvm.org/D93995
This is the same change as D93990, but for extractelement rather
than insertelement.
> If idx exceeds the length of val for a fixed-length vector, the
> result is a poison value. For a scalable vector, if the value of
> idx exceeds the runtime length of the vector, the result is a
> poison value.
This is a simple patch that updates InstSimplify to return poison if the index is/can be out-of-bounds
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D93990
Let getTruncateExpr() short-circuit to zero when the value being truncated is
known to have at least as many trailing zeros as the target type.
Differential Revision: https://reviews.llvm.org/D93973
This is almost all mechanical search-and-replace and
no-functional-change-intended (NFC). Having a single
enum makes it easier to match/reason about the
reduction cases.
The goal is to remove `Opcode` from reduction matching
code in the vectorizers because that makes it harder to
adapt the code to handle intrinsics.
The code in RecurrenceDescriptor::AddReductionVar() is
the only place that required closer inspection. It uses
a RecurrenceDescriptor and a second InstDesc to sometimes
overwrite part of the struct. It seem like we should be
able to simplify that logic, but it's not clear exactly
which cmp+sel patterns that we are trying to handle/avoid.
LVI previously handled "if (L && R)" conditions, but not
"if (L || R)" conditions. The latter case can still produce
useful information if L and R both constrain the same variable.
This adds support for handling the "if (L || R)" case as well.
The only difference is that we take the union instead of the
intersection of the lattice values.
Here we let non-intrinsic calls be considered legal and valid for
similarity only if the call is not indirect, and has a name.
For two calls to be considered similar, they must have the same name,
the same function types, and the same set of parameters, including tail
calls and calling conventions.
Tests are found in unittests/Analysis/IRSimilarityIdentifierTest.cpp.
Reviewers: jroelofs, paquette
Differential Revision: https://reviews.llvm.org/D87312
This might also make it easier to adapt if we want
to match min/max intrinsics rather than cmp+sel idioms.
The 'const' part is to potentially avoid confusion
in calling code. There's some surprising and possibly
wrong behavior related to matching min/max reductions
differently than other reductions.
GetElementPtr instructions require the extra check that all operands
after the first must only be constants and be exactly the same to be
considered similar.
Tests are found in unittests/Analysis/IRSimilarityIdentifierTest.cpp.
This patch makes SCEV recognize 'select A, B, false' and 'select A, true, B'.
This is a performance improvement that will be helpful after unsound select -> and/or transformation is removed, as discussed in D93065.
SCEV's answers for the select form should be a bit more conservative than the equivalent `and A, B` / `or A, B`.
Take this example: https://alive2.llvm.org/ce/z/NsP9ue .
To check whether it is valid for SCEV's computeExitLimit to return min(n, m) as ExactNotTaken value, I put llvm.assume at tgt.
It fails because the exit limit becomes poison if n is zero and m is poison. This is problematic if e.g. the exit value of i is replaced with min(n, m).
If either n or m is constant, we can revive the analysis again. I added relevant tests and put alive2 links there.
If and is used instead, this is okay: https://alive2.llvm.org/ce/z/K9rbJk . Hence the existing analysis is sound.
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D93882
The last use of the function, located in RemovePredecessorAndSimplify,
was removed on Dec 25, 2020 in commit
46bea9b297.
The last use of RemovePredecessorAndSimplify was removed on Sep 29,
2010 in commit 99c985c37d.
As mentioned in D93793, there are quite a few places where unary `IRBuilder::CreateShuffleVector(X, Mask)` can be used
instead of `IRBuilder::CreateShuffleVector(X, Undef, Mask)`.
Let's update them.
Actually, it would have been more natural if the patches were made in this order:
(1) let them use unary CreateShuffleVector first
(2) update IRBuilder::CreateShuffleVector to use poison as a placeholder value (D93793)
The order is swapped, but in terms of correctness it is still fine.
Reviewed By: spatel
Differential Revision: https://reviews.llvm.org/D93923
The x86_amx is used for AMX intrisics. <256 x i32> is bitcast to x86_amx when
it is used by AMX intrinsics, and x86_amx is bitcast to <256 x i32> when it
is used by load/store instruction. So amx intrinsics only operate on type x86_amx.
It can help to separate amx intrinsics from llvm IR instructions (+-*/).
Thank Craig for the idea. This patch depend on https://reviews.llvm.org/D87981.
Differential Revision: https://reviews.llvm.org/D91927
This PR adds impliesPoison(ValAssumedPoison, V) that returns true if V is
poison under the assumption that ValAssumedPoison is poison.
For example, impliesPoison('icmp X, 10', 'icmp X, Y') return true because
'icmp X, Y' is poison if 'icmp X, 10' is poison.
impliesPoison can be used for sound optimization of select, as discussed in
D77868.
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D78152
Confusingly, BinaryOperator is not an Operator,
OverflowingBinaryOperator is... We were implicitly assuming that
the multiply is an Instruction here.
This fixes the assertion failure reported in
https://reviews.llvm.org/D92726#2472827.
This patch updates isImpliedCondition/isKnownNonZero to look into select form of
and/or as well.
See llvm.org/pr48353 and D93065 for more context
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D93845
Following the discussion in D93065, this adds m_LogicalAnd() and
m_LogicalOr() matchers, that match A && B and A || B logical
operations, either as bitwise operations or select expressions.
As an example usage, LVI is adapted to use these matchers for its
condition reasoning.
The plan here is to switch other parts of LLVM that reason about
and/or of conditions to also support the select forms, and then
merge D93065 (or a variant thereof) to disable the poison-unsafe
select to and/or transform.
Differential Revision: https://reviews.llvm.org/D93827
In 35676a4f9a I've added handling for
non-trivial dominating conditions that imply non-zero on the true
branch. This adds the same support for the false branch.
The changes in pr45360.ll change block ordering and naming, but
don't change the control flow. The urem is still guaraded by a
non-zero check correctly.
The dominating condition handling in isKnownNonZero() currently
only takes into account conditions of the form "x != 0" or "x == 0".
However, there are plenty of other conditions that imply non-zero,
a common one being "x s> 0".
Peculiarly, the handling for assumes was already dealing with more
general non-zero-ness conditions, so this just reuses the same
logic for the dominating condition case.
D71264 started using a context instruction in a computeKnownBits()
call. However, if aliasing between two GEPs is checked, then the
choice of context instruction will be different for alias(GEP1, GEP2)
and alias(GEP2, GEP1), which is not supposed to happen.
Resolve this by remembering which GEP a certain VarIndex belongs to,
and use that as the context instruction. This makes the choice of
context instruction predictable and symmetric.
It should be noted that this choice of context instruction is
non-optimal (just like the previous choice): The AA query result is
only valid at points that are reachable from *both* instructions.
Using either one of them is conservatively correct, but a larger
context may also be valid to use.
Differential Revision: https://reviews.llvm.org/D93183
Some predicates, can be considered the same as long as the operands are
flipped. For example, a > b gives the same result as b > a. This maps
instructions in a greater than form, to their appropriate less than
form, swapping the operands in the IRInstructionData only, allowing for
more flexible matching.
Tests:
llvm/test/Transforms/IROutliner/outlining-isomorphic-predicates.ll
llvm/unittests/Analysis/IRSimilarityIdentifierTest.cpp
Reviewers: jroelofs, paquette
Recommit of commit 0503926602
Differential Revision: https://reviews.llvm.org/D87310
Some predicates, can be considered the same as long as the operands are
flipped. For example, a > b gives the same result as b > a. This maps
instructions in a greater than form, to their appropriate less than
form, swapping the operands in the IRInstructionData only, allowing for
more flexible matching.
Tests:
llvm/test/Transforms/IROutliner/outlining-isomorphic-predicates.ll
llvm/unittests/Analysis/IRSimilarityIdentifierTest.cpp
Reviewers: jroelofs, paquette
Differential Revision: https://reviews.llvm.org/D87310
Certain instructions, such as adds and multiplies can have the operands
flipped and still be considered the same. When we are analyzing
structure, this gives slightly more flexibility to create a mapping from
one region to another. We can add both operands in a corresponding
instruction to an operand rather than just the exact match. We then try
to eliminate items from the set, until there is only one valid mapping
between the regions of code.
We do this for adds, multiplies, and equality checking. However, this is
not done for floating point instructions, since the order can still
matter in some cases.
Tests:
llvm/test/Transforms/IROutliner/outlining-commutative-fp.ll
llvm/test/Transforms/IROutliner/outlining-commutative.ll
llvm/unittests/Analysis/IRSimilarityIdentifierTest.cpp
Reviewers: jroelofs, paquette
Differential Revision: https://reviews.llvm.org/D87311
Current approach doesn't work well in cases when multiple paths are predicted to be "cold". By "cold" paths I mean those containing "unreachable" instruction, call marked with 'cold' attribute and 'unwind' handler of 'invoke' instruction. The issue is that heuristics are applied one by one until the first match and essentially ignores relative hotness/coldness
of other paths.
New approach unifies processing of "cold" paths by assigning predefined absolute weight to each block estimated to be "cold". Then we propagate these weights up/down IR similarly to existing approach. And finally set up edge probabilities based on estimated block weights.
One important difference is how we propagate weight up. Existing approach propagates the same weight to all blocks that are post-dominated by a block with some "known" weight. This is useless at least because it always gives 50\50 distribution which is assumed by default anyway. Worse, it causes the algorithm to skip further heuristics and can miss setting more accurate probability. New algorithm propagates the weight up only to the blocks that dominates and post-dominated by a block with some "known" weight. In other words, those blocks that are either always executed or not executed together.
In addition new approach processes loops in an uniform way as well. Essentially loop exit edges are estimated as "cold" paths relative to back edges and should be considered uniformly with other coldness/hotness markers.
Reviewed By: yrouban
Differential Revision: https://reviews.llvm.org/D79485
When __builtin_dynamic_object_size returns a non-constant expression, it cannot
be -1 since that is an invalid return value for object size. However since
passes running after the substitution don't know this, they are unable to
optimize away the comparison and hence the comparison and branch stays in there.
This change generates an appropriate call to llvm.assume to help the optimizer
folding the test.
glibc is considering adopting __builtin_dynamic_object_size for additional
protection[1] and this change will help reduce branching overhead in fortified
implementations of all of the functions that don't have the __builtin___*_chk
type builtins, e.g. __ppoll_chk.
Also remove the test limit-max-iterations.ll because it was deemed unnecessary
during review.
[1] https://sourceware.org/pipermail/libc-alpha/2020-November/120191.html
Differential Revision: https://reviews.llvm.org/D93015
byval arguments should mostly get the same treatment as noalias
arguments in alias analysis. This was not the case for the
isIdentifiedFunctionLocal() function.
Marking byval arguments as identified function local means that
they cannot alias with other arguments, which I believe is correct.
Differential Revision: https://reviews.llvm.org/D93602
The transform wasn't checking that the LHS of the comparison
*is* the `X` in question...
This is the miscompile that was holding up D87188.
Thanks to Dave Green for producing an actionable reproducer!
Temporarily revert commit 8b1c4e310c.
After 8b1c4e310c the compile-time for `MultiSource/Benchmarks/MiBench/consumer-lame`
dramatically increases with -O3 & LTO, causing issues for builders with
that configuration.
I filed PR48553 with a smallish reproducer that shows a 10-100x compile
time increase.
This is split off from D91718 and adds a new target hook
supportsScalableVectors that can be queried to check if scalable vectors
are supported by the backend. For AArch64 this returns true if SVE is
enabled.
Reviewed By: david-arm
Differential Revision: https://reviews.llvm.org/D93060
This patch adds an alternative cost metric for the inliner to take
into account both the cost (i.e. size) and cycle count savings into
account.
Without this patch, we decide to inline a given call site if the size
of inlining the call site is below the threshold that is computed
according to the hotness of the call site.
This patch adds a new cost metric, turned off by default, to take over
the handling of hot call sites. Specifically, with the new cost
metric, we decide to inline a given call site if the ratio of cycle
savings to size exceeds a threshold. The cycle savings are computed
from call site costs, parameter propagation, folded conditional
branches, etc, all weighted by their respective profile counts. The
size is primarily the callee size, but we subtract call site costs and
the size of basic blocks that are never executed.
The new cost metric implicitly takes advantage of the machine function
splitter recently introduced by Snehasish Kumar, which dramatically
reduces the cost of duplicating (e.g. inlining) cold basic blocks by
placing cold basic blocks of hot functions in the .text.split
section.
We evaluated the new cost metric on clang bootstrap and SPECInt 2017.
For clang bootstrap, we observe 0.69% runtime improvement.
For SPECInt we report the change in IntRate the C/C++ benchmarks. All
benchmarks apart from perlbench and omnetpp improve, on average by
0.21% with the max for mcf at 1.96%.
Benchmark % Change
500.perlbench_r -0.45
502.gcc_r 0.13
505.mcf_r 1.96
520.omnetpp_r -0.28
523.xalancbmk_r 0.49
525.x264_r 0.00
531.deepsjeng_r 0.00
541.leela_r 0.35
557.xz_r 0.21
Differential Revision: https://reviews.llvm.org/D92780
The last use of the function was removed on Sep 18, 2016 in commit
5f8cc0c346.
The function was later moved to llvm/lib/Analysis/IVDescriptors.cpp on
Sep 12, 2018 in commit 7e98d69847.
This is being recommitted to try and address the MSVC complaint.
This patch implements a DDG printer pass that generates a graph in
the DOT description language, providing a more visually appealing
representation of the DDG. Similar to the CFG DOT printer, this
functionality is provided under an option called -dot-ddg and can
be generated in a less verbose mode under -dot-ddg-only option.
Reviewed By: Meinersbur
Differential Revision: https://reviews.llvm.org/D90159
Per http://llvm.org/OpenProjects.html#llvm_loopnest, the goal of this
patch (and other following patches) is to create facilities that allow
implementing loop nest passes that run on top-level loop nests for the
New Pass Manager.
This patch extends the functionality of LoopPassManager to handle
loop-nest passes by specializing the definition of LoopPassManager that
accepts both kinds of passes in addPass.
Only loop passes are executed if L is not a top-level one, and both
kinds of passes are executed if L is top-level. Currently, loop nest
passes should have the following run method:
PreservedAnalyses run(LoopNest &, LoopAnalysisManager &,
LoopStandardAnalysisResults &, LPMUpdater &);
Reviewed By: Whitney, ychen
Differential Revision: https://reviews.llvm.org/D87045
This patch implements a DDG printer pass that generates a graph in
the DOT description language, providing a more visually appealing
representation of the DDG. Similar to the CFG DOT printer, this
functionality is provided under an option called -dot-ddg and can
be generated in a less verbose mode under -dot-ddg-only option.
Differential Revision: https://reviews.llvm.org/D90159
his is a preparation patch for supporting multiple exits in the loop vectorizer, by itself it should be mostly NFC. This patch moves the loop structure checks from LAA to their respective consumers (where duplicates don't already exist). Moving the checks does end up changing some of the optimization warnings and debug output slightly, but nothing that appears to be a regression.
Why do this? Well, after auditing the code, I can't actually find anything in LAA itself which relies on having all instructions within a loop execute an equal number of times. This patch simply makes this explicit so that if one consumer - say LV in the near future (hopefully) - wants to handle a broader class of loops, it can do so.
Differential Revision: https://reviews.llvm.org/D92066
D82227 has added a proper check to limit PHI vectorization to the
maximum vector register size. That unfortunately resulted in at
least a couple of regressions on SystemZ and x86.
This change reverts PHI handling from D82227 and replaces it with
a more general check in SLPVectorizerPass::tryToVectorizeList().
Moved to tryToVectorizeList() it allows to restart vectorization
if initial chunk fails.
However, this function is more general and handles not only PHI
but everything which SLP handles. If vectorization factor would
be limited to maximum vector register size it would limit much
more vectorization than before leading to further regressions.
Therefore a new TTI callback getMaximumVF() is added with the
default 0 to preserve current behavior and limit nothing. Then
targets can decide what is better for them.
The callback gets ElementSize just like a similar getMinimumVF()
function and the main opcode of the chain. The latter is to avoid
regressions at least on the AMDGPU. We can have loads and stores
up to 128 bit wide, and <2 x 16> bit vector math on some
subtargets, where the rest shall not be vectorized. I.e. we need
to differentiate based on the element size and operation itself.
Differential Revision: https://reviews.llvm.org/D92059
InstCombine canonicalizes X>C && X<C' style comparisons into
(X+C1)<C2. This type of expression is recognized by some analyses
like LVI, but currently not when used inside assumptions, because
AssumptionCache does not track affected values for it.
BasicAA currently handles cases like Scale*V0 + (-Scale)*V1 where
V0 != V1, but does not handle the simpler case of Scale*V with
V != 0. Add it based on an isKnownNonZero() call.
I'm not passing a context instruction for now, because the existing
approach of always using GEP1 for context could result in symmetry
issues.
Differential Revision: https://reviews.llvm.org/D93162
The last use of isLoop was removed on Apr 29, 2002 in commit
09bbb5c015 as part of an effort to
remove "old induction varaible cannonicalization pass built on top of
interval analysis".
If we have two unknown sizes and one GEP operand and one non-GEP
operand, then we currently simply return MayAlias. The comment says
we can't do anything useful ... but we can! We can still check that
the underlying objects are different (and do so for the GEP-GEP case).
To reduce the compile-time impact, this a) checks this early, before
doing the relatively expensive GEP decomposition that will not be
used and b) doesn't do the check if the other operand is a phi or
select. In that case, the phi/select will already recurse, so this
would just do two slightly different recursive walks that arrive at
the same roots.
Compile-time is still a bit of a mixed bag: https://llvm-compile-time-tracker.com/compare.php?from=624af932a808b363a888139beca49f57313d9a3b&to=845356e14adbe651a553ed11318ddb5e79a24bcd&stat=instructions
On average this is a small improvement, but sqlite with ThinLTO has
a 0.5% regression (lencod has a 1% improvement).
The BasicAA test case checks this by using two memsets with unknown
size. However, the more interesting case where this is useful is
the LoopVectorize test case, as analysis of accesses in loops tends
to always us unknown sizes.
Differential Revision: https://reviews.llvm.org/D92401
Wei Mi