Floating point conversions inside vectorized loops have performance
implications but are very subtle. The user could specify a floating
point constant, or call a function without realizing that it will
force a change in the vector width. An example of this behaviour is
seen in https://godbolt.org/z/M3nT6c . The vectorizer should indicate
when this happens becuase it is most likely unintended behaviour.
This patch adds a simple check for this behaviour by following floating
point stores in the original loop and checking if a floating point
conversion operation occurs.
Reviewed By: fhahn
Differential Revision: https://reviews.llvm.org/D95539
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 is the preliminary patch of converting `LoopInterchange` pass to a loop-nest pass and has no intended functional change.
Changes that are not loop-nest related are split to D96650.
Reviewed By: Whitney
Differential Revision: https://reviews.llvm.org/D96644
As discussed in:
https://llvm.org/PR49179
...this pattern shows up in library code.
There are several potential generalizations as noted,
but we need to be careful that we get FP special-values
right, and it's not clear how much variation we should
expect to see from this exact idiom.
Also don't call function to update the call graph if there are no
clones. The function will fail.
rdar://74277860
Differential Revision: https://reviews.llvm.org/D96620
This is a split patch of D96644.
Explicitly pass both `InnerLoop` and `OuterLoop` to function `processLoop` to remove the need to swap elements in loop list and allow making loop list an `ArrayRef`.
Also, fix inconsistent spellings of `OuterLoopId` and `Inner Loop Id` in debug log.
Reviewed By: fhahn
Differential Revision: https://reviews.llvm.org/D96650
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
This patch updates codegen to use VPValues to manage the generated
scalarized instructions.
Reviewed By: gilr
Differential Revision: https://reviews.llvm.org/D92285
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
Loop canonicalization may end up deleting blocks from CFG. And
Scalar Evolution may still keep cached referenced to those blocks
unless updated properly.
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
This is unusual in the general (non-reciprocal) case because we need
an extra instruction, but that should be better for general FP
reassociation and codegen. We conservatively check for "arcp" FMF
here as we do with existing fdiv folds, but it is not strictly
necessary to have that.
This is part of solving:
https://llvm.org/PR49147
(The powi variant potentially has a different constraint.)
Differential Revision: https://reviews.llvm.org/D96648
This patch fixes pr48832 by correctly generating the mask when a poison value is involved.
Consider this CFG (which is a part of the input):
```
for.body: ; preds = %for.cond
br i1 true, label %cond.false, label %land.rhs
land.rhs: ; preds = %for.body
br i1 poison, label %cond.end, label %cond.false
cond.false: ; preds = %for.body, %land.rhs
br label %cond.end
cond.end: ; preds = %land.rhs, %cond.false
%cond = phi i32 [ 0, %cond.false ], [ 1, %land.rhs ]
```
The path for.body -> land.rhs -> cond.end should be taken when 'select i1 false, i1 poison, i1 false' holds (which means it's never taken); but VPRecipeBuilder::createEdgeMask was emitting 'and i1 false, poison' instead.
The former one successfully blocks poison propagation whereas the latter one doesn't, making the condition poison and thus causing the miscompilation.
SimplifyCFG has a similar bug (which didn't expose a real-world bug yet), and a patch for this is also ongoing (see https://reviews.llvm.org/D95026).
Reviewed By: bjope
Differential Revision: https://reviews.llvm.org/D95217
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
This allows for suspend point specific resume function types.
Return values from a suspend point can therefore be modelled as
arguments to the resume function. Allowing for directly passed return
types.
Differential Revision: https://reviews.llvm.org/D96136
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
Summary:
The negative test (with the feature being added disabled) caused MSAN failure and that's the added feature is supposed to fix. Therefore the negative test code is being removed.
Functions are currently processed by the sample profiler loader in a top-down order defined by the static call graph. The order is being adjusted to be a top-down order based on the input context-sensitive profile. One benefit is that the processing order of caller and callee in one SCC would follow the context order in the profile to favor more inlining. Another benefit is that the processing order of caller and callee through an indirect call (which is not on the static call graph) can be honored which in turn allows for more inlining.
The profile top-down order for SCC is also extended to support non-CS profiles.
Two switches `-mllvm -use-profile-indirect-call-edges` and `-mllvm -use-profile-top-down-order` are being introduced.
Reviewed By: wmi
Differential Revision: https://reviews.llvm.org/D95988
The IR/MIR pseudo probe intrinsics don't get materialized into real machine instructions and therefore they don't incur runtime cost directly. However, they come with indirect cost by blocking certain optimizations. Some of the blocking are intentional (such as blocking code merge) for better counts quality while the others are accidental. This change unblocks perf-critical optimizations that do not affect counts quality. They include:
1. IR InstCombine, sinking load operation to shorten lifetimes.
2. MIR LiveRangeShrink, similar to #1
3. MIR TwoAddressInstructionPass, i.e, opeq transform
4. MIR function argument copy elision
5. IR stack protection. (though not perf-critical but nice to have).
Reviewed By: wmi
Differential Revision: https://reviews.llvm.org/D95982
This is a special-case multiply that replicates bits of
the source operand. We need this fold to avoid regression
if we make canonicalization to `mul` more aggressive for
shl+or patterns.
I did not see a way to make Alive generalize the bit width
condition for even-number-of-bits only, but an example of
the proof is:
Name: i32
Pre: isPowerOf2(C1 - 1) && log2(C1) == C2 && (C2 * 2 == width(C2))
%m = mul nuw i32 %x, C1
%t = lshr i32 %m, C2
=>
%t = and i32 %x, C1 - 2
Name: i14
%m = mul nuw i14 %x, 129
%t = lshr i14 %m, 7
=>
%t = and i14 %x, 127
https://rise4fun.com/Alive/e52
Enable partial and runtime unrolling with a threshold of 30, which
was derived from a large number of kernels running on node and
wasmtime for amd64 and aarch64.
Unrolling is enabled by default at -O2 and -O3 and is disabled at
-Oz and -Os. Compiling with -Os is recommended if the wasm binary
size is the most important factor.
Differential Revision: https://reviews.llvm.org/D95125
This was reported as PR49104. The reproducer uses varargs but the issue
is the same, we know an argument is dead but can't change the signature
for some reason. The PR49104 situation was: We are in an CG-SCC
traversal and we remove all the uses of an argument and proof it thereby
dead. However, if we do not remove the argument, via signature rewrite,
we need to ensure that the `undef` we introduce at the call site doesn't
clash with a `noundef` attribute.
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
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.
Define an option -riscv-vector-bits-max to specify the maximum vector
bits for vectorizer. Loop vectorizer will use the value to check if it
is safe to use the whole vector registers to vectorize the loop.
It is not the optimum solution for loop vectorizing for scalable vector.
It assumed the whole vector registers will be used to vectorize the code.
If it is possible, we should configure vl to do vectorize instead of
using whole vector registers.
We only consider LMUL = 1 in this patch.
This patch just an initial work for loop vectorizer for RISC-V Vector.
Differential Revision: https://reviews.llvm.org/D95659
Link: https://lists.llvm.org/pipermail/llvm-dev/2020-October/146162.html "[RFC] FileCheck: (dis)allowing unused prefixes"
If a downstream project using lit needs time for transition,
add the following to `lit.local.cfg`:
```
from lit.llvm.subst import ToolSubst
fc = ToolSubst('FileCheck', unresolved='fatal')
config.substitutions.insert(0, (fc.regex, 'FileCheck --allow-unused-prefixes'))
```
Differential Revision: https://reviews.llvm.org/D95849
This patch improves the index management during constraint building.
Previously, the code rejected constraints which used values that were not
part of Value2Index, but after combining the coefficients of the new
indices were 0 (if ShouldAdd was 0).
In those cases, no new indices need to be added. Instead of adding to
Value2Index directly, add new indices to the NewIndices map. The caller
can then check if it needs to add any new indices.
This enables checking constraints like `a + x <= a + n` to `x <= n`,
even if there is no constraint for `a` directly.
As discussed in:
https://llvm.org/PR49055
We invert instcombine's add->or transform here
because it makes it easier to identify factorization
transforms like the mul in the motivating test.
This extends the logic added with:
https://reviews.llvm.org/rG70472f3https://reviews.llvm.org/rG93f3d7f
(I intentionally kept the formatting fix in this patch
to provide more context about the calling logic.)
This fixes PR49043 by invalidating the handle on RAUW. This will work
fine assuming all existing RAUW users add the new assumption to the
cache. That means, if a new llvm.assume call replaces an old one, you
need to add the new one now as a RAUW is not enough anymore.
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D96208
Type tests used only by assumes were original for devirtualization, but
are meant to be kept through the first invocation of LTT so that they
can be used for additional optimization. In the regular LTO case where
the IR is analyzed we may find a resolution for the type test and end up
rewriting the associated vtable global, which can have implications on
section splitting. Simply ignore these type tests.
Fixes PR48245.
Differential Revision: https://reviews.llvm.org/D96083
Context-sensitive profile effectively split a function profile into many copies each representing the CFG profile of a particular calling context. That makes the count distribution looks more flat as we now have more function profiles each with lower counts, which in turn leads to lower hot thresholds. Now we tells threshold computation to merge context profile first before calculating percentile based cutoffs to compensate for seemingly flat context profile. This can be controlled by swtich `sample-profile-contextless-threshold`.
Earlier measurement showed ~0.4% perf boost with this tuning on spec2k6 for CSSPGO (with pseudo-probe and new inliner).
Differential Revision: https://reviews.llvm.org/D95980
Summary:
This resolves an issue posted on Bugzilla. https://bugs.llvm.org/show_bug.cgi?id=48764
In this issue, the loop had multiple exit blocks, which resulted in the
function getExitBlock to return a nullptr, which resulted in hitting the assert.
This patch ensures that loops which only have one exit block as allowed to be
unrolled and jammed.
Reviewed By: Whitney, Meinersbur, dmgreen
Differential Revision: https://reviews.llvm.org/D95806
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
Dynamic allocas that still exist have been verified to be only used
'locally' not accross a suspend point.
rdar://73903220
Differential Revision: https://reviews.llvm.org/D96071
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
These tests demonstrate that LSR does not insert IV increment
into the latch block (as it supposes to) when it can use an
existing Phi as IV rather than creating a new LSR IV.
If we know that the scalar epilogue is required to run, modify the CFG to end the middle block with an unconditional branch to scalar preheader. This is instead of a conditional branch to either the preheader or the exit block.
The motivation to do this is to support multiple exit blocks. Specifically, the current structure forces us to identify immediate dominators and *which* exit block to branch from in the middle terminator. For the multiple exit case - where we know require scalar will hold - these questions are ill formed.
This is the last change needed to support multiple exit loops, but since the diffs are already large enough, I'm going to land this, and then enable separately. You can think of this as being NFCI-ish prep work, but the changes are a bit too involved for me to feel comfortable tagging the change that way.
Differential Revision: https://reviews.llvm.org/D94892
These attributes were all incorrect or inappropriate for LLVM to infer:
- inaccessiblememonly is generally wrong; user replacement operator new
can access memory that's visible to the caller, as can a new_handler
function.
- willreturn is generally wrong; a custom new_handler is not guaranteed
to terminate.
- noalias is inappropriate: Clang has a flag to determine whether this
attribute should be present and adds it itself when appropriate.
- noundef and nonnull on the return value should be specified by the
frontend on all 'operator new' functions if we want them, not here.
In any case, inferring attributes on functions declared 'nobuiltin' (as
these are when Clang emits them) seems questionable.
Several of the new attributes here were incorrect, and even the ones
that are generally correct were being added even to nobuiltin calls.
This reverts commit bb3f169b59.
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
The upstream callers (the vectorizers) were fixed with:
bbed5f2f8a ( D95690 )
77adbe6a8c
We should remove this pass entirely now that reduction
legalization/lowering is expected to work just as well,
but we need to confirm that the shuffle ops do not
regress (for x86 in particular).
This should be the last step needed to close:
https://llvm.org/PR23116
This patch extends the condition collection logic to allow adding
conditions from pre-headers to loop headers, by allowing cases where the
target block dominates some of its predecessors.
If amdgpu-unsafe-fp-atomics is specified, allow {flat|global}_atomic_add_f32 even if atomic modes don't match.
Differential Revision: https://reviews.llvm.org/D95391
C identifier name input sections such as __llvm_prf_* are GC roots so
they cannot be discarded. In LLD, the SHF_LINK_ORDER flag overrides the
C identifier name semantics.
The !associated metadata may be attached to a global object declaration
with a single argument that references another global object, and it
gets lowered to SHF_LINK_ORDER flag. When a function symbol is discarded
by the linker, setting up !associated metadata allows linker to discard
counters, data and values associated with that function symbol.
Note that !associated metadata is only supported by ELF, it does not have
any effect on non-ELF targets.
Differential Revision: https://reviews.llvm.org/D76802
Sample re-annotation is required in LTO time to achieve a reasonable post-inline profile quality. However, we have seen that such LTO-time re-annotation degrades profile quality. This is mainly caused by preLTO code duplication that is done by passes such as loop unrolling, jump threading, indirect call promotion etc, where samples corresponding to a source location are aggregated multiple times due to the duplicates. In this change we are introducing a concept of distribution factor for pseudo probes so that samples can be distributed for duplicated probes scaled by a factor. We hope that optimizations duplicating code well-maintain the branch frequency information (BFI) based on which probe distribution factors are calculated. Distribution factors are updated at the end of preLTO pipeline to reflect an estimated portion of the real execution count.
This change also introduces a pseudo probe verifier that can be run after each IR passes to detect duplicated pseudo probes.
A saturated distribution factor stands for 1.0. A pesudo probe will carry a factor with the value ranged from 0.0 to 1.0. A 64-bit integral distribution factor field that represents [0.0, 1.0] is associated to each block probe. Unfortunately this cannot be done for callsite probes due to the size limitation of a 32-bit Dwarf discriminator. A 7-bit distribution factor is used instead.
Changes are also needed to the sample profile inliner to deal with prorated callsite counts. Call sites duplicated by PreLTO passes, when later on inlined in LTO time, should have the callees’s probe prorated based on the Prelink-computed distribution factors. The distribution factors should also be taken into account when computing hotness for inline candidates. Also, Indirect call promotion results in multiple callisites. The original samples should be distributed across them. This is fixed by adjusting the callisites' distribution factors.
Reviewed By: wmi
Differential Revision: https://reviews.llvm.org/D93264
Inlining sometimes maps different instructions to be inlined onto the same instruction.
We must ensure to only remap the noalias scopes once. Otherwise the scope might disappear (at best).
This patch ensures that we only replace scopes for which the mapping is known.
This approach is preferred over tracking which instructions we already handled in a SmallPtrSet,
as that one will need more memory.
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D95862
Refactoring SampleProfileLoader::inlineHotFunctions to use helpers from CSSPGO inlining and reduce similar code in the inlining loop, plus minor cleanup for AFDO path.
This is resubmit of D95024, with build break and overtighten assertion fixed.
Test Plan:
This is a yet another hint that we will eventually need InstCombineInverter,
which would consistently sink inversions, but but for that we'll need
to consistently hoist inversions where possible, so let's do that here.
Example of a proof: https://alive2.llvm.org/ce/z/78SbDq
See https://bugs.llvm.org/show_bug.cgi?id=48995
A DLS lr, lr instruction only moves lr to itself. It need not be emitted
on it's own to save a instruction in the loop preheader.
Differential Revision: https://reviews.llvm.org/D78916
This patch updates IRBuilder::CreateMaskedGather/Scatter to work
with ScalableVectorType and adds isLegalMaskedGather/Scatter functions
to AArch64TargetTransformInfo. In addition I've fixed up
isLegalMaskedLoad/Store to return true for supported scalar types,
since this is what the vectorizer asks for.
In LoopVectorize.cpp I've changed
LoopVectorizationCostModel::getInterleaveGroupCost to return an invalid
cost for scalable vectors, since currently this relies upon using shuffle
vector for reversing vectors. In addition, in
LoopVectorizationCostModel::setCostBasedWideningDecision I have assumed
that the cost of scalarising memory ops is infinitely expensive.
I have added some simple masked load/store and gather/scatter tests,
including cases where we use gathers and scatters for conditional invariant
loads and stores.
Differential Revision: https://reviews.llvm.org/D95350
Refactoring SampleProfileLoader::inlineHotFunctions to use helpers from CSSPGO inlining and reduce similar code in the inlining loop, plus minor cleanup for AFDO path.
Test Plan:
Differential Revision: https://reviews.llvm.org/D95024
This change implemented call site prioritized BFS profile guided inlining for sample profile loader. The new inlining strategy maximize the benefit of context-sensitive profile as mentioned in the follow up discussion of CSSPGO RFC. The change will not affect today's AutoFDO as it's opt-in. CSSPGO now defaults to the new FDO inliner, but can fall back to today's replay inliner using a switch (`-sample-profile-prioritized-inline=0`).
Motivation
With baseline AutoFDO, the inliner in sample profile loader only replays previous inlining, and the use of profile is only for pruning previous inlining that turned out to be cold. Due to the nature of replay, the FDO inliner is simple with hotness being the only decision factor. It has the following limitations that we're improving now for CSSPGO.
- It doesn't take inline candidate size into account. Since it's doing replay, the size growth is bounded by previous CGSCC inlining. With context-sensitive profile, FDO inliner is no longer limited by previous inlining, so we need to take size into account to avoid significant size bloat.
- The way it looks at hotness is not accurate. It uses total samples in an inlinee as proxy for hotness, while what really matters for an inline decision is the call site count. This is an unfortunate fall back because call site count and callee entry count are not reliable due to dwarf based correlation, especially for inlinees. Now paired with pseudo-probe, we have accurate call site count and callee's entry count, so we can use that to gauge hotness more accurately.
- It treats all call sites from a block as hot as long as there's one call site considered hot. This is normally true, but since total samples is used as hotness proxy, this transitiveness within block magnifies the inacurate hotness heuristic. With pseduo-probe and the change above, this is no longer an issue for CSSPGO.
New FDO Inliner
Putting all the requirement for CSSPGO together, we need a top-down call site prioritized BFS inliner. Here're reasons why each component is needed.
- Top-down: We need a top-down inliner to better leverage context-sensitive profile, so inlining is driven by accurate context profile, and post-inline is also accurate. This is already implemented in https://reviews.llvm.org/D70655.
- Size Cap: For top-down inliner, taking function size into account for inline decision alone isn't sufficient to control size growth. We also need to explicitly cap size growth because with top-down inlining, we can grow inliner size significantly with large number of smaller inlinees even if each individually passes the cost/size check.
- Prioritize call sites: With size cap, inlining order also becomes important, because if we stop inlining due to size budget limit, we'd want to use budget towards the most beneficial call sites.
- BFS inline: Same as call site prioritization, if we stop inlining due to size budget limit, we want a balanced inline tree, rather than going deep on one call path.
Note that the new inliner avoids repeatedly evaluating same set of call site, so it should help with compile time too. For this reason, we could transition today's FDO inliner to use a queue with equal priority to avoid wasted reevaluation of same call site (TODO).
Speculative indirect call promotion and inlining is also supported now with CSSPGO just like baseline AutoFDO.
Tunings and knobs
I created tuning knobs for size growth/cap control, and for hot threshold separate from CGSCC inliner. The default values are selected based on initial tuning with CSSPGO.
Results
Evaluated with an internal LLVM fork couple months ago, plus another change to adjust hot-threshold cutoff for context profile (will send up after this one), the new inliner show ~1% geomean perf win on spec2006 with CSSPGO, while reducing code size too. The measurement was done using train-train setup, MonoLTO w/ new pass manager and pseudo-probe. Note that this is just a starting point - we hope that the new inliner will open up more opportunity with CSSPGO, but it will certainly take more time and effort to make it fully calibrated and ready for bigger workloads (we're working on it).
Differential Revision: https://reviews.llvm.org/D94001
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
C identifier name input sections such as __llvm_prf_* are GC roots so
they cannot be discarded. In LLD, the SHF_LINK_ORDER flag overrides the
C identifier name semantics.
The !associated metadata may be attached to a global object declaration
with a single argument that references another global object, and it
gets lowered to SHF_LINK_ORDER flag. When a function symbol is discarded
by the linker, setting up !associated metadata allows linker to discard
counters, data and values associated with that function symbol.
Note that !associated metadata is only supported by ELF, it does not have
any effect on non-ELF targets.
Differential Revision: https://reviews.llvm.org/D76802
Fixing up a couple places where `getCallSiteIdentifier` is needed to support pseudo-probe-based callsites.
Also fixing an issue in the extbinary profile reader where the metadata section is not fully scanned based on the number of profiles loaded only for the current module.
Reviewed By: wmi, wenlei
Differential Revision: https://reviews.llvm.org/D95791
The main point of committing this early is to have a negative test in tree. Nothing fails in the current tests if we implement this (currently unsound) optimization.
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
A == B map to A >= B && A <= B
(https://alive2.llvm.org/ce/z/_dwxKn).
This extends the constraint construction to return a list of
constraints, which can be used to properly de-compose nested AND & OR.
If the incoming block to a phi node is an EH pad, then we will
materialize into an EH pad, which is not supposed to happen. To fix
this, I added a check to see if incoming block of a phi node is an EH
pad before using it as the insertion point.
Differential Revision: https://reviews.llvm.org/D95019
Some benchmarks regress with non-trivial unswitching, so add an option
to opt-out of performing non-trivial unswitching while investigating.
Reviewed By: asbirlea
Differential Revision: https://reviews.llvm.org/D95796
The constant trunc/ext may not be the optimal pre-condition,
but I think that handles the common cases.
Example of Alive2 proof:
https://alive2.llvm.org/ce/z/sREeLC
This is another step towards canonicalizing to the intrinsics.
Narrowing was identified as source of potential regression for
abs(), so we need to handle this for min/max - see:
https://llvm.org/PR48816
If this is not enough, we could process intrinsics in
the trunc-driven matching in canEvaluateTruncated().
D90687 introduced a crash:
llvm::LoopVectorizationCostModel::computeMaxVF(llvm::ElementCount, unsigned int):
Assertion `WideningDecisions.empty() && Uniforms.empty() && Scalars.empty() &&
"No decisions should have been taken at this point"' failed.
when compiling the following C code:
typedef struct {
char a;
} b;
b *c;
int d, e;
int f() {
int g = 0;
for (; d; d++) {
e = 0;
for (; e < c[d].a; e++)
g++;
}
return g;
}
with:
clang -Os -target hexagon -mhvx -fvectorize -mv67 testcase.c -S -o -
This occurred since prior to D90687 computeFeasibleMaxVF would only be
called in computeMaxVF when a scalar epilogue was allowed, but now it's
always called. This causes the assert above since computeFeasibleMaxVF
collects all viable VFs larger than the default MaxVF, and for each VF
calculates the register usage which results in analysis being done the
assert above guards against. This can occur in computeFeasibleMaxVF if
TTI.shouldMaximizeVectorBandwidth and this target hook is implemented in
the hexagon backend to always return true.
Reported by @iajbar.
Reviewed By: fhahn
Differential Revision: https://reviews.llvm.org/D94869
If we determine that the invariant path through the loop has no effects,
we can directly branch to the exit block, instead to unswitching first.
Besides avoiding some extra work (unswitching first, then deleting the
loop again) this allows to be more aggressive than regular unswitching
with respect to cost-modeling. This approach should always be be
desirable.
This is similar in spirit to D93734, just that it uses the previously
added checks for loop-unswitching.
I tried to add the required no-op checks from scratch, as we only check
a subset of the loop. There is potential to unify the checks with
LoopDeletion, at the cost of adding a predicate whether a block should
be considered.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D95468
The reduction of a sanitizer build failure when enabling the dominance check (D95335) showed that loop peeling also needs to take care of scope duplication, just like loop unrolling (D92887).
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D95544
This patch fixes updating MemorySSA if the header contains memory
defs that do not clobber a duplicated instruction. We need to find the
first defining access outside the loop body and use that as defining
access of the duplicated instruction.
This fixes a crash caused by bee486851c.
This patch emits "instr_prof_hash_mismatch" function annotation metadata if
there is a hash mismatch while applying instrumented profiles.
During the PGO optimized build using instrumented profiles, if the CFG of
the function has changed since generating the profile, a hash mismatch is
encountered. This patch emits this information as annotation metadata. We
plan to use this with Propeller which is done at the machine IR level.
Propeller is usually applied on top of PGO and a hash mismatch during
PGO could be used to detect source drift.
Differential Revision: https://reviews.llvm.org/D95495
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
and fix a few edge cases that show up in the Swift compiler but
weren't caught by the existing tests. Most notably the old code wasn't
salvaging load operations correctly. The patch also gets rid of the
LoadFromFramePtr argument and replaces it with a more generalized
mechanism.
This change brings up support of context-sensitive profiles in the format of extended binary. Existing sample profile reader/writer/merger code is being tweaked to reflect the fact of bracketed input contexts, like (`[...]`). The paired brackets are also needed in extbinary profiles because we don't yet have an otherwise good way to tell calling contexts apart from regular function names since the context delimiter `@` can somehow serve as a part of the C++ mangled names.
Reviewed By: wmi, wenlei
Differential Revision: https://reviews.llvm.org/D95547
I am trying to untangle the fast-math-flags propagation logic
in the vectorizers (see a6f022127 for SLP).
The loop vectorizer has a mix of checking FP function attributes,
IR-level FMF, and just wrong assumptions.
I am trying to avoid regressions while fixing this, and I think
the IR-level logic is good enough for that, but it's hard to say
for sure. This would be the 1st step in the clean-up.
The existing test that I changed to include 'fast' actually shows
a miscompile: the function only had the equivalent of nnan, but we
created new instructions that had fast (all FMF set). This is
similar to the example in https://llvm.org/PR35538
Differential Revision: https://reviews.llvm.org/D95452
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
This change implements support for applying profile instrumentation
only to selected files or functions. The implementation uses the
sanitizer special case list format to select which files and functions
to instrument, and relies on the new noprofile IR attribute to exclude
functions from instrumentation.
Differential Revision: https://reviews.llvm.org/D94820
This patch improves the availability for variables stored in the
coroutine frame by emitting an alloca to hold the pointer to the frame
object and rewriting dbg.declare intrinsics to point inside the frame
object using salvaged DIExpressions. Finally, a new alloca is created
in the funclet to hold the FramePtr pointer to ensure that it is
available throughout the entire function at -O0.
This path also effectively reverts D90772. The testcase updates
highlight nicely how every removed CHECK for a dbg.value is preceded
by a new CHECK for a dbg.declare.
Thanks to JunMa, Yifeng, and Bruno for their thoughtful reviews!
Differential Revision: https://reviews.llvm.org/D93497
rdar://71866936
This change implements support for applying profile instrumentation
only to selected files or functions. The implementation uses the
sanitizer special case list format to select which files and functions
to instrument, and relies on the new noprofile IR attribute to exclude
functions from instrumentation.
Differential Revision: https://reviews.llvm.org/D94820
The existing test has less FMF than we might expect if
our FMF was fixed (on all FP values), so this additional
test is intended to check propagation in a more "normal"
example.
This patch adds additional checks to avoid partial unswitching
in cases where it won't be profitable, e.g. because the path directly
exits the loop anyways.
Loop peeling removes conditions from loop bodies that become invariant
after a small number of iterations. When triggered, this leads to fewer
compares and possibly PHIs in loop bodies, enabling further
optimizations. The current cost-model of loop peeling should be quite
conservative/safe, i.e. only peel if a condition in the loop becomes
known after peeling.
For example, see PR47671, where loop peeling enables vectorization by
removing a PHI the vectorizer does not understand. Granted, the
loop-vectorizer could also be taught about constant PHIs, but loop
peeling is likely to enable other optimizations as well.
This has an impact on quite a few benchmarks from
MultiSource/SPEC2000/SPEC2006 on X86 with -O3 -flto, for example
Same hash: 186 (filtered out)
Remaining: 51
Metric: loop-vectorize.LoopsVectorized
Program base patch diff
test-suite...ve-susan/automotive-susan.test 8.00 9.00 12.5%
test-suite...nal/skidmarks10/skidmarks.test 35.00 31.00 -11.4%
test-suite...lications/sqlite3/sqlite3.test 41.00 43.00 4.9%
test-suite...s/ASC_Sequoia/AMGmk/AMGmk.test 25.00 26.00 4.0%
test-suite...006/450.soplex/450.soplex.test 88.00 89.00 1.1%
test-suite...TimberWolfMC/timberwolfmc.test 120.00 119.00 -0.8%
test-suite.../CINT2006/403.gcc/403.gcc.test 215.00 216.00 0.5%
test-suite...006/447.dealII/447.dealII.test 957.00 958.00 0.1%
test-suite...ternal/HMMER/hmmcalibrate.test 75.00 75.00 0.0%
Same hash: 186 (filtered out)
Remaining: 51
Metric: loop-vectorize.LoopsAnalyzed
Program base patch diff
test-suite...ks/Prolangs-C/agrep/agrep.test 440.00 434.00 -1.4%
test-suite...nal/skidmarks10/skidmarks.test 312.00 308.00 -1.3%
test-suite...marks/7zip/7zip-benchmark.test 6399.00 6323.00 -1.2%
test-suite...lications/minisat/minisat.test 134.00 135.00 0.7%
test-suite...rks/FreeBench/pifft/pifft.test 295.00 297.00 0.7%
test-suite...TimberWolfMC/timberwolfmc.test 1879.00 1869.00 -0.5%
test-suite...pplications/treecc/treecc.test 689.00 691.00 0.3%
test-suite...T2000/300.twolf/300.twolf.test 1593.00 1597.00 0.3%
test-suite.../Benchmarks/Bullet/bullet.test 1394.00 1392.00 -0.1%
test-suite...ications/JM/ldecod/ldecod.test 1431.00 1429.00 -0.1%
test-suite...6/464.h264ref/464.h264ref.test 2229.00 2230.00 0.0%
test-suite...lications/sqlite3/sqlite3.test 2590.00 2589.00 -0.0%
test-suite...ications/JM/lencod/lencod.test 2732.00 2733.00 0.0%
test-suite...006/453.povray/453.povray.test 3395.00 3394.00 -0.0%
Note the -11% regression in number of loops vectorized for skidmarks. I
suspect this corresponds to the fact that those loops are gone now (see
the reduction in number of loops analyzed by LV).
Reviewed By: lebedev.ri
Differential Revision: https://reviews.llvm.org/D88471
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
When LSR converts a branch on the pre-inc IV into a branch on the
post-inc IV, the nowrap flags on the addition may no longer be valid.
Previously, a poison result of the addition might have been ignored,
in which case the program was well defined. After branching on the
post-inc IV, we might be branching on poison, which is undefined behavior.
Fix this by discarding nowrap flags which are not present on the SCEV
expression. Nowrap flags on the SCEV expression are proven by SCEV
to always hold, independently of how the expression will be used.
This is essentially the same fix we applied to IndVars LFTR, which
also performs this kind of pre-inc to post-inc conversion.
I believe a similar problem can also exist for getelementptr inbounds,
but I was not able to come up with a problematic test case. The
inbounds case would have to be addressed in a differently anyway
(as SCEV does not track this property).
Fixes https://bugs.llvm.org/show_bug.cgi?id=46943.
Differential Revision: https://reviews.llvm.org/D95286
or claimRV calls in the IR
Background:
This patch makes changes to the front-end and middle-end that are
needed to fix a longstanding problem where llvm breaks ARC's autorelease
optimization (see the link below) by separating calls from the marker
instructions or retainRV/claimRV calls. The backend changes are in
https://reviews.llvm.org/D92569.
https://clang.llvm.org/docs/AutomaticReferenceCounting.html#arc-runtime-objc-autoreleasereturnvalue
What this patch does to fix the problem:
- The front-end annotates calls with attribute "clang.arc.rv"="retain"
or "clang.arc.rv"="claim", which indicates the call is implicitly
followed by a marker instruction and a retainRV/claimRV call that
consumes the call result. This is currently done only when the target
is arm64 and the optimization level is higher than -O0.
- ARC optimizer temporarily emits retainRV/claimRV calls after the
annotated calls in the IR and removes the inserted calls after
processing the function.
- ARC contract pass emits retainRV/claimRV calls after the annotated
calls. It doesn't remove the attribute on the call since the backend
needs it to emit the marker instruction. The retainRV/claimRV calls
are emitted late in the pipeline to prevent optimization passes from
transforming the IR in a way that makes it harder for the ARC
middle-end passes to figure out the def-use relationship between the
call and the retainRV/claimRV calls (which is the cause of PR31925).
- The function inliner removes the autoreleaseRV call in the callee that
returns the result if nothing in the callee prevents it from being
paired up with the calls annotated with "clang.arc.rv"="retain/claim"
in the caller. If the call is annotated with "claim", a release call
is inserted since autoreleaseRV+claimRV is equivalent to a release. If
it cannot find an autoreleaseRV call, it tries to transfer the
attributes to a function call in the callee. This is important since
ARC optimizer can remove the autoreleaseRV call returning the callee
result, which makes it impossible to pair it up with the retainRV or
claimRV call in the caller. If that fails, it simply emits a retain
call in the IR if the call is annotated with "retain" and does nothing
if it's annotated with "claim".
- This patch teaches dead argument elimination pass not to change the
return type of a function if any of the calls to the function are
annotated with attribute "clang.arc.rv". This is necessary since the
pass can incorrectly determine nothing in the IR uses the function
return, which can happen since the front-end no longer explicitly
emits retainRV/claimRV calls in the IR, and change its return type to
'void'.
Future work:
- Use the attribute on x86-64.
- Fix the auto upgrader to convert call+retainRV/claimRV pairs into
calls annotated with the attributes.
rdar://71443534
Differential Revision: https://reviews.llvm.org/D92808
Fixes an infinite loop encountered in GVN.
GVN will delay PRE if it encounters critical edges, attempt to split
them later via calls to SplitCriticalEdge(), then restart.
The caller of GVN::splitCriticalEdges() assumed a return value of true
meant that critical edges were split, that the IR had changed, and that
PRE should be re-attempted, upon which we loop infinitely.
This was exposed after D88438, by compiling the Linux kernel for s390,
but the test case is reproducible on x86.
Fixes: https://github.com/ClangBuiltLinux/linux/issues/1261
Reviewed By: void
Differential Revision: https://reviews.llvm.org/D94996
turning off SampleFDO silently.
Currently sample loader pass turns off SampleFDO optimization silently when
it sees error in reading the profile. This behavior will defeat the tests
which could have caught those bad/incompatible profile problems. This patch
change the behavior to report error.
Differential Revision: https://reviews.llvm.org/D95269
We can sink extends after min/max if they match and would
not change the sign-interpreted compare. The only combo
that doesn't work is zext+smin/smax because the zexts
could change a negative number into positive:
https://alive2.llvm.org/ce/z/D6sz6J
Sext+umax/umin works:
define i32 @src(i8 %x, i8 %y) {
%0:
%sx = sext i8 %x to i32
%sy = sext i8 %y to i32
%m = umax i32 %sx, %sy
ret i32 %m
}
=>
define i32 @tgt(i8 %x, i8 %y) {
%0:
%m = umax i8 %x, %y
%r = sext i8 %m to i32
ret i32 %r
}
Transformation seems to be correct!
a6f0221276 enabled intersection of FMF on reduction instructions,
so it is safe to ease the check here.
There is still some room to improve here - it looks like we
have nearly duplicate flags propagation logic inside of the
LoopUtils helper but it is limited targets that do not form
reduction intrinsics (they form the shuffle expansion).
A @llvm.experimental.noalias.scope.decl is only useful if there is !alias.scope and !noalias metadata that uses the declared scope.
When that is not the case for at least one of the two, the intrinsic call can as well be removed.
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D95141
Similar to D92887, LoopRotation also needs duplicate the noalias scopes when rotating a `@llvm.experimental.noalias.scope.decl` across a block boundary.
This is based on the version from the Full Restrict paches (D68511).
The problem it fixes also showed up in Transforms/Coroutines/ex5.ll after D93040 (when enabling strict checking with -verify-noalias-scope-decl-dom).
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D94306
This is a fix for https://bugs.llvm.org/show_bug.cgi?id=39282. Compared to D90104, this version is based on part of the full restrict patched (D68484) and uses the `@llvm.experimental.noalias.scope.decl` intrinsic to track the location where !noalias and !alias.scope scopes have been introduced. This allows us to only duplicate the scopes that are really needed.
Notes:
- it also includes changes and tests from D90104
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D92887
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
In both cases, optimization is prevented because
"br X == C || X == C2" is converted into a switch. In one case
loop rotation is blocked, in the other vectorization.
As shown in the test diffs, we could miscompile by
propagating flags that did not exist in the original
code.
The flags required for fmin/fmax reductions will be
fixed in a follow-up patch.
With the addition of the `willreturn` attribute, functions that may
not return (e.g. due to an infinite loop) are well defined, if they are
not marked as `willreturn`.
This patch updates `wouldInstructionBeTriviallyDead` to not consider
calls that may not return as dead.
This patch still provides an escape hatch for intrinsics, which are
still assumed as willreturn unconditionally. It will be removed once
all intrinsics definitions have been reviewed and updated.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D94106
Insert a llvm.experimental.noalias.scope.decl intrinsic that identifies where a noalias argument was inlined.
This patch includes some refactorings from D90104.
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D93040
The test wasn't sensitive to alias analysis. As you can seen from D95117 when AA is added by default this is affected.
Updating the test so that it coveres both cases for AA analysis.
Note that this patch depends on D95117 to land first.
Differential Revision: https://reviews.llvm.org/D95247
This builds on the restricted after initial revert form of D93906, and adds back support for breaking backedges of inner loops. It turns out the original invalidation logic wasn't quite right, specifically around the handling of LCSSA.
When breaking the backedge of an inner loop, we can cause blocks which were in the outer loop only because they were also included in a sub-loop to be removed from both loops. This results in the exit block set for our original parent loop changing, and thus a need for new LCSSA phi nodes.
This case happens when the inner loop has an exit block which is also an exit block of the parent, and there's a block in the child which reaches an exit to said block without also reaching an exit to the parent loop.
(I'm describing this in terms of the immediate parent, but the problem is general for any transitive parent in the nest.)
The approach implemented here involves a potentially expensive LCSSA rebuild. Perf testing during review didn't show anything concerning, but we may end up needing to revert this if anyone encounters a practical compile time issue.
Differential Revision: https://reviews.llvm.org/D94378
Similar to binary operators like fadd/fmul/fsub, propagate shape info
through unary operators (fneg is the only one?).
Differential Revision: https://reviews.llvm.org/D95252
I have previously tried doing that in
b33fbbaa34 / d38205144f,
but eventually it was pointed out that the approach taken there
was just broken wrt how the uses of bonus instructions are updated
to account for the fact that they should now use either bonus instruction
or the cloned bonus instruction. In particluar, all that manual handling
of PHI nodes in successors was just wrong.
But, the fix is actually much much simpler than my initial approach:
just tell SSAUpdate about both instances of bonus instruction,
and let it deal with all the PHI handling.
Alive2 confirms that the reproducers from the original bugs (@pr48450*)
are now handled correctly.
This effectively reverts commit 59560e8589,
effectively relanding b33fbbaa34.
NewBonusInst just took name from BonusInst, so BonusInst has no name,
so BonusInst.getName() makes no sense.
So we need to ask NewBonusInst for the name.
In the motivating cases from https://llvm.org/PR48816 ,
we have a trailing trunc. But that is not required to
reduce the abs width:
https://alive2.llvm.org/ce/z/ECaz-p
...as long as we clear the int-min-is-poison bit (nsw).
We have some existing tests that are affected, and I'm
not sure what the overall implications are, but in general
we favor narrowing operations over preserving nsw/nuw.
If that causes problems, we could restrict this transform
based on type (shouldChangeType() and/or vector vs. scalar).
Differential Revision: https://reviews.llvm.org/D95235
The existing code did not deal with atomic loads correctly. Such loads
are represented as MemoryDefs. Bail out on any MemoryAccess that is not
a MemoryUse.
Because we were not looking for the llvm.coro.id.async intrinsic in the
early coro pass which triggers follow-up passes we relied on the
llvm.coro.end intrinsic being present. This might not be the case in
functions that end in unreachable code.
Differential Revision: https://reviews.llvm.org/D95144
Iff we know we can get rid of the inversions in the new pattern,
we can thus get rid of the inversion in the old pattern,
this decreasing instruction count.
Note that we could position this transformation as just hoisting
of the `not` (still, iff y is freely negatible), but the test changes
show a number of regressions, so let's not do that.
Iff we know we can get rid of the inversions in the new pattern,
we can thus get rid of the inversion in the old pattern,
this decreasing instruction count.
The case where BB ends with an unconditional branch,
and has a single predecessor w/ conditional branch
to BB and a single successor of BB is exactly the pattern
SpeculativelyExecuteBB() transform deals with.
(and in this case they both allow speculating only a single instruction)
Well, or FoldTwoEntryPHINode(), if the final block
has only those two predecessors.
Here, in FoldBranchToCommonDest(), only a weird subset of that
transform is supported, and it's glued on the side in a weird way.
In particular, it took me a bit to understand that the Cond
isn't actually a branch condition in that case, but just the value
we allow to speculate (otherwise it reads as a miscompile to me).
Additionally, this only supports for the speculated instruction
to be an ICmp.
So let's just unclutter FoldBranchToCommonDest(), and leave
this transform up to SpeculativelyExecuteBB(). As far as i can tell,
this shouldn't really impact optimization potential, but if it does,
improving SpeculativelyExecuteBB() will be more beneficial anyways.
Notably, this only affects a single test,
but EarlyCSE should have run beforehand in the pipeline,
and then FoldTwoEntryPHINode() would have caught it.
This reverts commit rL158392 / commit d33f4efbfd.
I have removed an unnecessary assert in LoopVectorizationCostModel::getInstructionCost
that prevented a cost being calculated for select instructions when using
scalable vectors. In addition, I have changed AArch64TTIImpl::getCmpSelInstrCost
to only do special cost calculations for fixed width vectors and fall
back to the base version for scalable vectors.
I have added a simple cost model test for cmps and selects:
test/Analysis/CostModel/sve-cmpsel.ll
and some simple tests that show we vectorize loops with cmp and select:
test/Transforms/LoopVectorize/AArch64/sve-basic-vec.ll
Differential Revision: https://reviews.llvm.org/D95039
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
This patch sets the default for llvm tests, with the exception of tests
under Reduce, because quite a few of them use 'FileCheck' as parameter
to a tool, and including a flag as that parameter would complicate
matters.
The rest of the patch undo-es the lit.local.cfg changes we progressively
introduced as temporary measure to avoid regressions under various
directories.
Differential Revision: https://reviews.llvm.org/D95111
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
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
Reviewed By: asbirlea
Differential Revision: https://reviews.llvm.org/D95117
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
In LoopInterchange, `findInnerReductionPhi()` looks for reduction
variables, which cannot be constants. Update it to return early in that
case.
This also addresses a blocker for removing use-lists from ConstantData,
whose users could be spread across arbitrary modules in the same
LLVMContext.
Differential Revision: https://reviews.llvm.org/D94712
If a function doesn't contain loops and does not call non-willreturn
functions, then it is willreturn. Loops are detected by checking
for backedges in the function. We don't attempt to handle finite
loops at this point.
Differential Revision: https://reviews.llvm.org/D94633
It turns out the vectorizer calls the getIntrinsicInstrCost functions
with a scalar return type and vector VF. This updates the costmodel to
handle that, still producing the correct vector costs.
A vectorizer test is added to show it vectorizing at the correct factor
again.
Summary:
The custom mapper API did not previously support the mapping names added previously. This means they were not present if a user requested debugging information while using the mapper functions. This adds basic support for passing the mapped names to the runtime library.
Reviewers: jdoerfert
Differential Revision: https://reviews.llvm.org/D94806
In https://llvm.org/PR48810 , we are crashing while trying to
propagate attributes from mempcpy (returns void*) to memcpy
(returns nothing - void).
We can avoid the crash by removing known incompatible
attributes for the void return type.
I'm not sure if this goes far enough (should we just drop all
attributes since this isn't the same function?). We also need
to audit other transforms in LibCallSimplifier to make sure
there are no other cases that have the same problem.
Differential Revision: https://reviews.llvm.org/D95088
This patch applies the idea from D93734 to LoopUnswitch.
It adds support for unswitching on conditions that are only
invariant along certain paths through a loop.
In particular, it targets conditions in the loop header that
depend on values loaded from memory. If either path from
the true or false successor through the loop does not modify
memory, perform partial loop unswitching.
That is, duplicate the instructions feeding the condition in the pre-header.
Then unswitch on the duplicated condition. The condition is now known
in the unswitched version for the 'invariant' path through the original loop.
On caveat of this approach is that one of the loops created can be partially
unswitched again. To avoid this behavior, `llvm.loop.unswitch.partial.disable`
metadata is added to the unswitched loops, to avoid subsequent partial
unswitching.
If that's the approach to go, I can move the code handling the metadata kind
into separate functions.
This increases the cases we unswitch quite a bit in SPEC2006/SPEC2000 &
MultiSource. It also allows us to eliminate a dead loop in SPEC2017's omnetpp
```
Tests: 236
Same hash: 170 (filtered out)
Remaining: 66
Metric: loop-unswitch.NumBranches
Program base patch diff
test-suite...000/255.vortex/255.vortex.test 2.00 23.00 1050.0%
test-suite...T2006/401.bzip2/401.bzip2.test 7.00 55.00 685.7%
test-suite :: External/Nurbs/nurbs.test 5.00 26.00 420.0%
test-suite...s-C/unix-smail/unix-smail.test 1.00 3.00 200.0%
test-suite.../Prolangs-C++/ocean/ocean.test 1.00 3.00 200.0%
test-suite...tions/lambda-0.1.3/lambda.test 1.00 3.00 200.0%
test-suite...yApps-C++/PENNANT/PENNANT.test 2.00 5.00 150.0%
test-suite...marks/Ptrdist/yacr2/yacr2.test 1.00 2.00 100.0%
test-suite...lications/viterbi/viterbi.test 1.00 2.00 100.0%
test-suite...plications/d/make_dparser.test 12.00 24.00 100.0%
test-suite...CFP2006/433.milc/433.milc.test 14.00 27.00 92.9%
test-suite.../Applications/lemon/lemon.test 7.00 12.00 71.4%
test-suite...ce/Applications/Burg/burg.test 6.00 10.00 66.7%
test-suite...T2006/473.astar/473.astar.test 16.00 26.00 62.5%
test-suite...marks/7zip/7zip-benchmark.test 78.00 121.00 55.1%
```
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D93764
This reverts commit d97f776be5.
The original problem was due to build failures in shared lib builds. D95079
moved ImportedFunctionsInliningStatistics under Analysis, unblocking
this.
Branch/assume conditions in PredicateInfo are currently handled in
a rather ad-hoc manner, with some arbitrary limitations. For example,
an `and` of two `icmp`s will be handled, but an `and` of an `icmp`
and some other condition will not. That also includes the case where
more than two conditions and and'ed together.
This patch makes the handling more general by looking through and/ors
up to a limit and considering all kinds of conditions (though operands
will only be taken for cmps of course).
Differential Revision: https://reviews.llvm.org/D94447
David Green