This fixes PR47297.
When ProcessBlock() was able to constant fold the terminator's
condition, but not do any more transformations, the function would
return false, which would lead to the JumpThreading pass returning an
incorrect modified status. This patch makes so that ProcessBlock()
returns true in such cases. This will trigger an unnecessary invocation
of ProcessBlock() in such cases, but this should be rare to occur.
This was caught using the check introduced by D80916.
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D87392
Turns out this was use-after-move of function_ref, which is trivially
copyable and movable, so the move did nothing and use after move was
safe.
But since this function_ref is being copied into a std::function, change
the function_ref to be std::function to avoid extra layers of type
erasure indirection - and then it's a real use after move, and fix that
by referring to the moved-to member variable rather than the moved-from
parameter.
gcov is an "Edge Profiling with Edge Counters" application according to
Optimally Profiling and Tracing Programs (1994).
The minimum number of counters necessary is |E|-(|V|-1). The unmeasured edges
form a spanning tree. Both GCC --coverage and clang -fprofile-generate leverage
this optimization. This patch implements the optimization for clang --coverage.
The produced .gcda files are much smaller now.
i.e. change the work flow from
* .gcno for function A
* .gcno for function B
* .gcno for function C
* .gcda for function A
* .gcda for function B
* .gcda for function C
to
* .gcno for function A
* .gcda for function A
* .gcno for function B
* .gcda for function B
* .gcno for function C
* .gcda for function C
Currently there is duplicate logic in .gcno & .gcda processing: how functions
are filtered, which edges are instrumented, etc. This refactor enables simplification.
Since we always process .gcno, in -fprofile-arcs -fno-test-coverage mode,
__llvm_internal_gcov_emit_function_args.0 will have non-zero checksums.
When deleting stores at the end of a function, we have to do PHI
translation, otherwise we might miss reads in different iterations of a
loop. See multiblock-loop-carried-dependence.ll for details.
This fixes a mis-compile and surprisingly also increases the number of
eliminated stores from 26047 to 26572 for MultiSource/SPEC2000/SPEC2006
on X86 with -O3 -flto. This is most likely because we save budget by not
exploring through MemoryPhis, which are less likely to result in valid
candidates for elimination.
The issue was reported post-commit for fb109c42d9.
This allows the backend to tell the vectorizer to produce inloop
reductions through a TTI hook.
For the moment on ARM under MVE this means allowing integer add
reductions of the correct size. In the future this can include integer
min/max too, under -Os.
Differential Revision: https://reviews.llvm.org/D75512
NOTE: There is a mailing list discussion on this: http://lists.llvm.org/pipermail/llvm-dev/2019-December/137632.html
Complemantary to the assumption outliner prototype in D71692, this patch
shows how we could simplify the code emitted for an alignemnt
assumption. The generated code is smaller, less fragile, and it makes it
easier to recognize the additional use as a "assumption use".
As mentioned in D71692 and on the mailing list, we could adopt this
scheme, and similar schemes for other patterns, without adopting the
assumption outlining.
This is a followup to D86834, which partially fixed this issue in
InstSimplify. However, InstCombine repeats the same transform while
dropping poison flags -- which does not cover cases where poison is
introduced in some other way.
The fix here is a bit more comprehensive, because things are quite
entangled, and it's hard to only partially address it without
regressing optimization. There are really two changes here:
* Export the SimplifyWithOpReplaced API from InstSimplify, with an
added AllowRefinement flag. For replacements inside the TrueVal
we don't actually care whether refinement occurs or not, the
replacement is always legal. This part of the transform is now
done in InstSimplify only. (It should be noted that the current
AllowRefinement check is not sufficient -- that's an issue we
need to address separately.)
* Change the InstCombine fold to work by temporarily dropping
poison generating flags, running the fold and then restoring the
flags if it didn't work out. This will ensure that the InstCombine
fold is correct as long as the InstSimplify fold is correct.
Differential Revision: https://reviews.llvm.org/D87445
The test example based on PR47450 shows that we can
match non-byte-sized shifts, but those won't ever be
bswap opportunities. This isn't a full fix (we'd still
match if the shifts were by 8-bits for example), but
this should be enough until there's evidence that we
need to do more (this is a borderline case for
vectorization in the first place).
When inlining functions containing allocas of scalable vectors we
cannot specify the size in the lifetime markers, since we don't
know this at compile time.
Added new test here:
test/Transforms/Inline/AArch64/sve-alloca-merge.ll
Differential Revision: https://reviews.llvm.org/D87139
In addition to calculate hash consistently by swapping SELECT's
operands, we also need to inverse the select pattern favor to match the
original logic.
[EarlyCSE] Equivalent SELECTs should hash equally
DenseMap<SimpleValue> assumes that, if its isEqual method returns true
for two elements, then its getHashValue method must return the same value
for them. This invariant is broken when one SELECT node is a min/max
operation, and the other can be transformed into an equivalent min/max by
inverting its predicate and swapping its operands. This patch fixes an
assertion failure that would occur intermittently while compiling the
following IR:
define i32 @t(i32 %i) {
%cmp = icmp sle i32 0, %i
%twin1 = select i1 %cmp, i32 %i, i32 0
%cmpinv = icmp sgt i32 0, %i
%twin2 = select i1 %cmpinv, i32 0, i32 %i
%sink = add i32 %twin1, %twin2
ret i32 %sink
}
Differential Revision: https://reviews.llvm.org/D86843
The tests have been updated and I plan to move them from the MSSA
directory up.
Some end-to-end tests needed small adjustments. One difference to the
legacy DSE is that legacy DSE also deletes trivially dead instructions
that are unrelated to memory operations. Because MemorySSA-backed DSE
just walks the MemorySSA, we only visit/check memory instructions. But
removing unrelated dead instructions is not really DSE's job and other
passes will clean up.
One noteworthy change is in llvm/test/Transforms/Coroutines/ArgAddr.ll,
but I think this comes down to legacy DSE not handling instructions that
may throw correctly in that case. To cover this with MemorySSA-backed
DSE, we need an update to llvm.coro.begin to treat it's return value to
belong to the same underlying object as the passed pointer.
There are some minor cases MemorySSA-backed DSE currently misses, e.g. related
to atomic operations, but I think those can be implemented after the switch.
This has been discussed on llvm-dev:
http://lists.llvm.org/pipermail/llvm-dev/2020-August/144417.html
For the MultiSource/SPEC2000/SPEC2006 the number of eliminated stores
goes from ~17500 (legayc DSE) to ~26300 (MemorySSA-backed). More numbers
and details in the thread on llvm-dev.
Impact on CTMark:
```
Legacy Pass Manager
exec instrs size-text
O3 + 0.60% - 0.27%
ReleaseThinLTO + 1.00% - 0.42%
ReleaseLTO-g. + 0.77% - 0.33%
RelThinLTO (link only) + 0.87% - 0.42%
RelLO-g (link only) + 0.78% - 0.33%
```
http://llvm-compile-time-tracker.com/compare.php?from=3f22e96d95c71ded906c67067d75278efb0a2525&to=ae8be4642533ff03803967ee9d7017c0d73b0ee0&stat=instructions
```
New Pass Manager
exec instrs. size-text
O3 + 0.95% - 0.25%
ReleaseThinLTO + 1.34% - 0.41%
ReleaseLTO-g. + 1.71% - 0.35%
RelThinLTO (link only) + 0.96% - 0.41%
RelLO-g (link only) + 2.21% - 0.35%
```
http://195.201.131.214:8000/compare.php?from=3f22e96d95c71ded906c67067d75278efb0a2525&to=ae8be4642533ff03803967ee9d7017c0d73b0ee0&stat=instructions
Reviewed By: asbirlea, xbolva00, nikic
Differential Revision: https://reviews.llvm.org/D87163
DenseMap<SimpleValue> assumes that, if its isEqual method returns true
for two elements, then its getHashValue method must return the same value
for them. This invariant is broken when one SELECT node is a min/max
operation, and the other can be transformed into an equivalent min/max by
inverting its predicate and swapping its operands. This patch fixes an
assertion failure that would occur intermittently while compiling the
following IR:
define i32 @t(i32 %i) {
%cmp = icmp sle i32 0, %i
%twin1 = select i1 %cmp, i32 %i, i32 0
%cmpinv = icmp sgt i32 0, %i
%twin2 = select i1 %cmpinv, i32 0, i32 %i
%sink = add i32 %twin1, %twin2
ret i32 %sink
}
Differential Revision: https://reviews.llvm.org/D86843
Bail from maskIsAllZeroOrUndef and maskIsAllOneOrUndef prior to iterating over the number of
elements for scalable vectors.
Assert that the mask type is not scalable in possiblyDemandedEltsInMask .
Assert that the types are correct in all three functions.
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D87424
Previously we could match fcmp+select to a reduction if the fcmp had
the nonans fast math flag. But if the select had the nonans fast
math flag, InstCombine would turn it into a fminnum/fmaxnum intrinsic
before SLP gets to it. Seems fairly likely that if one of the
fcmp+select pair have the fast math flag, they both would.
My plan is to start vectorizing the fmaxnum/fminnum version soon,
but I wanted to get this code out as it had some of the strangest
fast math flag behaviors.
PGOInstrumentation runs `SplitIndirectBrCriticalEdges` but some IndirectBrInst
critical edge cannot be split. `getInstrBB` will crash when calling `SplitCriticalEdge`, e.g.
int foo(char *p) {
void *targets[2];
targets[0] = &&indirect;
targets[1] = &&end;
for (;; p++)
if (*p == 7) {
indirect:
goto *targets[p[1]]; // the self loop is critical in -O
}
end:
return 0;
}
Skip such critical edges to prevent a crash.
Reviewed By: davidxl, lebedev.ri
Differential Revision: https://reviews.llvm.org/D87435
The argument promotion pass currently fails to copy function annotations
over to the modified function after promoting arguments.
This patch copies the original function annotation to the new function.
Reviewed By: fhann
Differential Revision: https://reviews.llvm.org/D86630
See discussion in D87149. Dropping volatile stores here is legal
per LLVM semantics, but causes issues for real code and may result
in a change to LLVM volatile semantics. Temporarily treat volatile
stores as "not guaranteed to transfer execution" in just this place,
until this issue has been resolved.
MemoryLocation has been taught about memcpy.inline, which means we can
get the memory locations read and written by it. This means DSE can
handle memcpy.inline
This patch enables inserting freeze when JumpThreading converts a select to
a conditional branch when it is run in LTO.
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D85534
As code size is the only thing we care about at minsize, query the
cost of materialising immediates when calculating the cost of a SCEV
expansion. We also modify the CostKind to TCK_CodeSize for minsize,
instead of RecipThroughput.
Differential Revision: https://reviews.llvm.org/D76434
This patch fixes pr45956 (https://bugs.llvm.org/show_bug.cgi?id=45956 ).
To minimize its impact to the quality of generated code, I suggest enabling
this only for LTO as a start (it has two JumpThreading passes registered).
This patch contains a flag that makes JumpThreading enable it.
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D84940
The test in PR47457 demonstrates a situation when candidate load's pointer's SCEV
is no loger a SCEVAddRec after loop versioning. The code there assumes that it is
always a SCEVAddRec and crashes otherwise.
This patch makes sure that we do not consider candidates for which this requirement
is broken after the versioning.
Differential Revision: https://reviews.llvm.org/D87355
Reviewed By: asbirlea
Atomic stores are modeled as MemoryDef to model the fact that they may
not be reordered, depending on the ordering constraints.
Atomic stores that are monotonic or weaker do not limit re-ordering, so
we do not have to treat them as potential read clobbers.
Note that llvm/test/Transforms/DeadStoreElimination/MSSA/atomic.ll
already contains a set of negative test cases.
Reviewed By: asbirlea
Differential Revision: https://reviews.llvm.org/D87386
The entry block is split at the first instruction where `shouldKeepInEntry`
returns false. The created basic block has a br jumping to the original entry
block. The new basic block causes the function label line and the other entry
block lines to be covered by different basic blocks, which can affect line
counts with special control flows (fork/exec in the entry block requires
heuristics in llvm-cov gcov to get consistent line counts).
int main() { // BB0
return 0; // BB2 (due to entry block splitting)
}
// BB1 is the exit block (since gcov 4.8)
This patch adds a synthetic entry block (like PGOInstrumentation and GCC) and
inserts an edge from the synthetic entry block to the original entry block. We
can thus remove the tricky `shouldKeepInEntry` and entry block splitting. The
number of basic blocks does not change, but the emitted .gcno files will be
smaller because we can save one GCOV_TAG_LINES tag.
// BB0 is the synthetic entry block with a single edge to BB2
int main() { // BB2
return 0; // BB2
}
// BB1 is the exit block (since gcov 4.8)
This is the initial part of the implementation of the C++20 likelihood
attributes. It handles the attributes in an if statement.
Differential Revision: https://reviews.llvm.org/D85091
If a function had at most one return block, the pass would return false
regardless if an unified unreachable block was created.
This patch fixes that by refactoring runOnFunction into two separate
helper functions for handling the unreachable blocks respectively the
return blocks, as suggested by @bjope in a review comment.
This was caught using the check introduced by D80916.
Reviewed By: serge-sans-paille
Differential Revision: https://reviews.llvm.org/D85818
This patch follows D85345 and adds more noundef attributes to return values/arguments of library functions
that are mostly about accessing the file system or processes.
A few functions like `chmod` or `times` use typedef `mode_t` and `clock_t`.
They are neither struct nor union, so they cannot contain undef even if they're lowered to iN in IR. So, it is fine to add noundef to them.
- clock_t's actual type is size_t (C17, 7.27.1.3), so it isn't struct or union.
- For mode_t, either int or long is used in practice because programmers use bit manipulation. So, I think it is okay that it's never aggregate in practice.
After this patch, the remaining library functions are those that eagerly participate in optimizations: they can be removed, reordered, or
introduced by a transformation from primitive IR operations.
For them, a few testings is needed, since it may not be valid to add noundef anymore even if C standard says it's okay.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D85894
This patch adds isGuaranteedNotToBePoison and programUndefinedIfUndefOrPoison.
isGuaranteedNotToBePoison will be used at D75808. The latter function is used at isGuaranteedNotToBePoison.
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D84242
The MemorySSAWrapperPass depends on AAResultsWrapperPass and if
MemorySSA is preserved but AAResultsWrapperPass is not, this could lead
to a crash when updating the last user of the MemorySSAWrapperPass.
Alternatively AAResultsWrapperPass could be marked preserved by GVN, but
I am not sure if that would be safe. I am not sure what is required in
order to preserve AAResultsWrapperPass. At the moment, it seems like a
couple of passes that do similar transforms to GVN are preserving it.
Reviewed By: asbirlea
Differential Revision: https://reviews.llvm.org/D87137
This commit cleans up the ::initialize method of various AAs in the
following ways:
- If an associated function is required, give up on declarations.
This was discovered as a real problem when lots of llvm.dbg.XXX
call sites were assumed `noreturn` until proven otherwise. That
does not make any sense and caused huge regressions and missed
deductions.
- Require more associated declarations for function interface AAs.
- Use the IRAttribute::initialize to determine if function interface
AAs can be used in IPO, don't replicate the checks (especially
isFunctionIPOAmendable) all over the place. Arguably the function
declaration check should be moved to some central place to.
If we have a callback, call site arguments were already associated with
the callback callee. Now we also associate the function with the
callback callee, thus we know ensure that the following holds true (if
all return nonnull):
`getAssociatedArgument()->getParent() == getAssociatedFunction()`
To test this an early exit from
`AAMemoryBehaviorCallSiteArgument::initialize``
is included as well. Without the change to getAssociatedFunction() this
kind of early exit for declarations would cause callback call site
arguments to miss out.
As we handle callback calls we need to disambiguate the call site
argument number from the callee argument number. While always equal in
non-callback calls, a callback comes with a partial parameter-argument
mapping so there is no implicit correspondence. Here we split
`IRPosition::getArgNo()` into two public functions, `getCallSiteArgNo()`
and `getCalleeArgNo()`. Usages are adjusted to pick the right one for
their purpose. This fixed some problems that would have been exposed as
we more aggressively optimize callbacks.
While operand bundles carry unpredictable semantics, we know some of
them and can therefore "ignore" them. In this case we allow to look at
the declaration of `llvm.assume` when asked for the attributes at a call
site. The assume operand bundles we have do not invalidate the
declaration attributes.
We cannot test this in isolation because the llvm.assume attributes are
determined by the parser. However, a follow up patch will provide test
coverage.
In `MultiSource/Benchmarks/tramp3d-v4/tramp3d-v4.cpp` we initialized
attributes until stack frame ~35k caused space to run out. The initial
size 1024 is pretty much random.
The get{Return,Unwind,Unreachable}Block functions in
UnifyFunctionExitNodes have not been used for many years,
so just remove them.
Reviewed By: bjope
Differential Revision: https://reviews.llvm.org/D87078
If we know that the abs operand is known negative, we can replace
it with a neg.
To avoid computing known bits twice, I've removed the fold for the
non-negative case from InstSimplify. Both the non-negative and the
negative case are handled by InstCombine now, with one known bits call.
Differential Revision: https://reviews.llvm.org/D87196
D66230 attempted to fix a problem where when there are allocas used before CoroBegin.
It keeps allocas and their uses stay in put if there are no escapse/changes to the data before CoroBegin.
Unfortunately that's incorrect.
Consider this code:
%var = alloca i32
%1 = getelementptr .. %var; stays put
%f = call i8* @llvm.coro.begin
store ... %1
After this fix, %1 will now stay put, however if a store happens after coro.begin and hence modifies the content, this change will not be reflected in the coroutine frame (and will eventually be DCEed).
To generalize the problem, if any alias ptr is created before coro.begin for an Alloca and that alias ptr is latter written into after coro.begin, it will lead to incorrect behavior.
There are also a few other minor issues, such as incorrect dominate condition check in the ptr visitor, unhandled memory intrinsics and etc.
Ths patch attempts to fix some of these issue, and make it more robust to deal with aliases.
While visiting through the alloca pointer, we also keep track of all aliases created that will be used after CoroBegin. We track the offset of each alias, and then reacreate these aliases after CoroBegin using these offset.
It's worth noting that this is not perfect and there will still be cases we cannot handle. I think it's impractical to handle all cases given the current design.
This patch makes it more robust and should be a pure win.
In the meantime, we need to think about what how to completely elimiante these issues, likely through the route as @rjmccall mentioned in D66230.
Differential Revision: https://reviews.llvm.org/D86859
In GenerateConstantOffsetsImpl, we may generate non canonical Formula
if BaseRegs of that Formula is updated and includes a recurrent expr reg
related with current loop while its ScaledReg is not.
Patched by: mdchen
Reviewed By: qcolombet
Differential Revision: https://reviews.llvm.org/D86939
The CloneFunctionInto has implicit requirements with regards to the
linkage and visibility of the function. We now update these after we did
the CloneFunctionInto on the copy with the same linkage and visibility
as the original.
Deleting or replacing anything is certainly a modification. This caused
a later assertion in IPSCCP when compiling 400.perlbench with the new PM.
I'm not sure how to test this.
This was reverted in 503deec218
because it caused gigantic increase (3x) in branch mispredictions
in certain benchmarks on certain CPU's,
see https://reviews.llvm.org/D84108#2227365.
It has since been investigated and here are the results:
https://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20200907/827578.html
> It's an amazingly severe regression, but it's also all due to branch
> mispredicts (about 3x without this). The code layout looks ok so there's
> probably something else to deal with. I'm not sure there's anything we can
> reasonably do so we'll just have to take the hit for now and wait for
> another code reorganization to make the branch predictor a bit more happy :)
>
> Thanks for giving us some time to investigate and feel free to recommit
> whenever you'd like.
>
> -eric
So let's just reland this.
Original commit message:
I've been looking at missed vectorizations in one codebase.
One particular thing that stands out is that some of the loops
reach vectorizer in a rather mangled form, with weird PHI's,
and some of the loops aren't even in a rotated form.
After taking a more detailed look, that happened because
the loop's headers were too big by then. It is evident that
SimplifyCFG's common code hoisting transform is at fault there,
because the pattern it handles is precisely the unrotated
loop basic block structure.
Surprizingly, `SimplifyCFGOpt::HoistThenElseCodeToIf()` is enabled
by default, and is always run, unlike it's friend, common code sinking
transform, `SinkCommonCodeFromPredecessors()`, which is not enabled
by default and is only run once very late in the pipeline.
I'm proposing to harmonize this, and disable common code hoisting
until //late// in pipeline. Definition of //late// may vary,
here currently i've picked the same one as for code sinking,
but i suppose we could enable it as soon as right after
loop rotation happens.
Experimentation shows that this does indeed unsurprizingly help,
more loops got rotated, although other issues remain elsewhere.
Now, this undoubtedly seriously shakes phase ordering.
This will undoubtedly be a mixed bag in terms of both compile- and
run- time performance, codesize. Since we no longer aggressively
hoist+deduplicate common code, we don't pay the price of said hoisting
(which wasn't big). That may allow more loops to be rotated,
so we pay that price. That, in turn, that may enable all the transforms
that require canonical (rotated) loop form, including but not limited to
vectorization, so we pay that too. And in general, no deduplication means
more [duplicate] instructions going through the optimizations. But there's still
late hoisting, some of them will be caught late.
As per benchmarks i've run {F12360204}, this is mostly within the noise,
there are some small improvements, some small regressions.
One big regression i saw i fixed in rG8d487668d09fb0e4e54f36207f07c1480ffabbfd, but i'm sure
this will expose many more pre-existing missed optimizations, as usual :S
llvm-compile-time-tracker.com thoughts on this:
http://llvm-compile-time-tracker.com/compare.php?from=e40315d2b4ed1e38962a8f33ff151693ed4ada63&to=c8289c0ecbf235da9fb0e3bc052e3c0d6bff5cf9&stat=instructions
* this does regress compile-time by +0.5% geomean (unsurprizingly)
* size impact varies; for ThinLTO it's actually an improvement
The largest fallout appears to be in GVN's load partial redundancy
elimination, it spends *much* more time in
`MemoryDependenceResults::getNonLocalPointerDependency()`.
Non-local `MemoryDependenceResults` is widely-known to be, uh, costly.
There does not appear to be a proper solution to this issue,
other than silencing the compile-time performance regression
by tuning cut-off thresholds in `MemoryDependenceResults`,
at the cost of potentially regressing run-time performance.
D84609 attempts to move in that direction, but the path is unclear
and is going to take some time.
If we look at stats before/after diffs, some excerpts:
* RawSpeed (the target) {F12360200}
* -14 (-73.68%) loops not rotated due to the header size (yay)
* -272 (-0.67%) `"Number of live out of a loop variables"` - good for vectorizer
* -3937 (-64.19%) common instructions hoisted
* +561 (+0.06%) x86 asm instructions
* -2 basic blocks
* +2418 (+0.11%) IR instructions
* vanilla test-suite + RawSpeed + darktable {F12360201}
* -36396 (-65.29%) common instructions hoisted
* +1676 (+0.02%) x86 asm instructions
* +662 (+0.06%) basic blocks
* +4395 (+0.04%) IR instructions
It is likely to be sub-optimal for when optimizing for code size,
so one might want to change tune pipeline by enabling sinking/hoisting
when optimizing for size.
Reviewed By: mkazantsev
Differential Revision: https://reviews.llvm.org/D84108
This reverts commit 503deec218.
For intrinsics supported by ConstantRange, compute the result range
based on the argument ranges. We do this independently of whether
some or all of the input ranges are full, as we can often still
constrain the result in some way.
Differential Revision: https://reviews.llvm.org/D87183
This was supposed to be an NFC cleanup, but there's
a real logic difference (did not drop 'nsw') visible
in some tests in addition to an efficiency improvement.
This is because in the case where we have 2 GEPs,
the code was *always* swapping the operands and
negating the result. But if we have 2 GEPs, we
should *never* need swapping/negation AFAICT.
This is part of improving flags propagation noticed
with PR47430.
Normal dead code elimination ignores assume intrinsics, so we fail to
delete assumes that are not meaningful (and potentially worse if they
cause conflicts with other assumptions).
The motivating example in https://llvm.org/PR47416 suggests that we
might have problems upstream from here (difference between C and C++),
but this should be a cheap way to make sure we remove more dead code.
Differential Revision: https://reviews.llvm.org/D87149
To enable the cost of constants, the helper function has been
reorganised:
- A struct has been introduced to hold SCEV operand information so
that we know the user of the operand, as well as the operand index.
The Worklist now uses instead instead of a bare SCEV.
- The costing of each SCEV, and collection of its operands, is now
performed in a helper function.
Differential Revision: https://reviews.llvm.org/D86050
Modify FoldBranchToCommonDest to consider the cost of inserting
instructions when attempting to combine predicates to fold blocks.
The threshold can be controlled via a new option:
-simplifycfg-branch-fold-threshold which defaults to '2' to allow
the insertion of a not and another logical operator.
Differential Revision: https://reviews.llvm.org/D86526
Similar to D87168, but for abs. If we have a dominating x >= 0
condition, then we know that abs(x) is x. This fold is in
InstCombine, because we need to create a sub instruction for
the x < 0 case.
Differential Revision: https://reviews.llvm.org/D87184
When a switch case is folded into default's case, that's an IR change that
should be reported, update ConstantFoldTerminator accordingly.
Differential Revision: https://reviews.llvm.org/D87142
When checking call sites, give special handling to indirect call, as the
callee may be unknown and can lead to nullptr dereference later. Assume
conservatively that the ICV always changes in such case.
Reviewed By: sstefan1
Differential Revision: https://reviews.llvm.org/D87104
This patch updates MemCpyOpt to preserve MemorySSA. It uses the
MemoryDef at the insertion point of the builder and inserts the new def
after that def.
In some cases, we just modify a memory instruction. In that case, get
the defining access, then remove the memory access and add a new one.
If the defining access is in a different block, insert a new def at the
beginning of the current block, otherwise after the defining access.
Reviewed By: asbirlea
Differential Revision: https://reviews.llvm.org/D86651
As far as I am aware, the placement of MergedLoadStoreMotion in the
pipeline is not heavily tuned currently. It seems to not matter much if
we do it after DSE in the LTO pipeline (no binary changes for -O3 -flto
on MultiSource/SPEC2000/SPEC2006). Moving it after DSE however has a
major benefit: MemorySSA is constructed by LICM and is consumed by DSE,
so if MergedLoadStoreMotion happens after DSE, we do not need to
preserve MemorySSA in it.
If there are any concerns with this move, I can also update
MergedLoadStoreMotion to preserve MemorySSA.
This patch together with D86651 (preserve MemSSA in MemCpyOpt) and
D86534 (preserve MemSSA in GVN) are the remaining patches to bring down
compile-time for DSE + MemorySSA to the levels outlined in
http://lists.llvm.org/pipermail/llvm-dev/2020-August/144417.html
Once they land, we should be able to start with flipping the switch on
enabling DSE + MmeorySSA.
Reviewed By: asbirlea
Differential Revision: https://reviews.llvm.org/D86967
Preserve MemorySSA if it is available before running GVN.
DSE with MemorySSA will run closely after GVN. If GVN and 2 other
passes preserve MemorySSA, DSE can re-use MemorySSA used by LICM
when doing LTO.
Reviewed By: asbirlea
Differential Revision: https://reviews.llvm.org/D86534
This is a followup to 1ccfb52a61, which made a number of changes
including the apparently innocuous reordering of required passes in
MemCpyOptimizer. This however altered the creation order of BasicAA vs
Phi Values analysis, meaning BasicAA did not pick up PhiValues as a
cached result. Instead if we require MemoryDependence first it will
require PhiValuesAnalysis allowing BasicAA to use it for better results.
I don't claim this is an excellent design, but it fixes a nasty little
regressions where a query later in JumpThreading was getting worse
results.
Differential Revision: https://reviews.llvm.org/D87027
Currently IPSCCP (and others like CVP/GVN) blindly propagate pointer
equalities. In certain cases, that leads to dereferenceable pointers
being replaced, as in the example test case.
I think this is not allowed, as it introduces an access of an
un-dereferenceable pointer. Note that the pointer is inbounds, but one
past the last element, so it is valid, but not dereferenceable.
This patch is mostly to highlight the issue and start a discussion.
Currently it only checks for specifically looking
one-past-the-last-element pointers with array typed bases.
This causes the mis-compile outlined in
https://stackoverflow.com/questions/55754313/is-this-gcc-clang-past-one-pointer-comparison-behavior-conforming-or-non-standar
In the test case, if we replace %p with the GEP for the store, we
subsequently determine that the store and the load cannot alias, because
they are to different underlying objects.
Note that Alive2 seems to think that the replacement is valid:
https://alive2.llvm.org/ce/z/2rorhk
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D85332
First, shuffle cost for scalable type is not known for scalable type;
Second, we cannot reason if the narrowed shuffle mask for scalable type
is a splat or not.
E.g., Bitcast splat vector from type <vscale x 4 x i32> to <vscale x 8 x i16>
will involve narrowing shuffle mask <vscale x 4 x i32> zeroinitializer to
<vscale x 8 x i32> with element sequence of <0, 1, 0, 1, ...>, which cannot be
reasoned if it's a valid splat or not.
Reviewed By: spatel
Differential Revision: https://reviews.llvm.org/D86995
In IPSCCP when a function is optimized to return undef, it should clear the returned attribute for all its input arguments
and its corresponding call sites.
The bug is exposed when the value of an input argument of the function is assigned to a physical register and
because of the argument having a returned attribute, the value of this physical register will continue to be used
as the function return value right after the call instruction returns, even if the value that this register holds may
be clobbered during the function call. This potentially results in incorrect values being used afterwards.
Reviewed By: jdoerfert, fhahn
Differential Revision: https://reviews.llvm.org/D84220
When marking a global variable constant, and simplifying users using
CleanupConstantGlobalUsers(), the pass could incorrectly return false if
there were still some uses left, and no further optimizations was done.
This was caught using the check introduced by D80916.
This fixes PR46749.
Reviewed By: fhahn
Differential Revision: https://reviews.llvm.org/D85837
These transforms will now be performed irrespective of the number of uses for the expression "1.0/sqrt(X)":
1.0/sqrt(X) * X => X/sqrt(X)
X * 1.0/sqrt(X) => X/sqrt(X)
We already handle more general cases, and we are intentionally not creating extra (and likely expensive)
fdiv ops in IR. This pattern is the exception to the rule because we always expect the Backend to reduce
X/sqrt(X) to sqrt(X), if it has the necessary (reassoc) fast-math-flags.
Ref: DagCombiner optimizes the X/sqrt(X) to sqrt(X).
Differential Revision: https://reviews.llvm.org/D86726
This is an enhancement to D81766 to allow loading the minimum target
vector type into an IR vector with a different number of elements.
In one of the motivating tests from PR16739, SLP creates <2 x float>
load ops mixed with <4 x float> insert ops, so we want to handle that
pattern in addition to potential oversized vectors created by the
vectorizers.
For now, we are assuming the insert/extract subvector with undef is
free because there is no exact corresponding TTI modeling for that.
Differential Revision: https://reviews.llvm.org/D86160
This patch makes it possible for AAUB to use information from AANoUndef.
This is the next patch of D86983
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D86984
When the associated value is undef, we immediately forced to indicate a pessimistic fixpoint so far.
This patch changes the initialization to check the attribute given in IR at first and to indicate an optimistic fixpoint when it is given.
This change will enable us to catch , for example, the following case in AAUB.
```
call void @foo(i32 noundef undef)
```
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D86983
Summary:
Analyses are preserved in MemCpyOptimizer.
Get analyses before running the pass and store the pointers, instead of
using lambdas and getting them every time on demand.
Reviewers: lenary, deadalnix, mehdi_amini, nikic, efriedma
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D74494
Interleave for small loops that have reductions inside,
which breaks dependencies and expose.
This gives very significant performance improvements for some benchmarks.
Because small loops could be in very hot functions in real applications.
Differential Revision: https://reviews.llvm.org/D81416
Loop Idiom Recognize Pass (LIRP) attempts to transform loops with subscripted arrays
into memcpy/memset function calls. In some particular situation, this transformation
introduces negative impacts. For example: https://bugs.llvm.org/show_bug.cgi?id=47300
This patch will enable users to disable a particular part of the transformation, while
he/she can still enjoy the benefit brought about by the rest of LIRP. The default
behavior stays unchanged: no part of LIRP is disabled by default.
Reviewed By: etiotto (Ettore Tiotto)
Differential Revision: https://reviews.llvm.org/D86262
getValuesInOffloadArrays goes through the offload arrays in __tgt_target_data_begin_mapper getting the values stored in them before the call is issued.
call void @__tgt_target_data_begin_mapper(arg0, arg1,
i8** %offload_baseptrs, i8** %offload_ptrs, i64* %offload_sizes,
...)
Diferential Revision: https://reviews.llvm.org/D86300
The 1st try was reverted because I missed an assert that
needed softening.
As discussed in D86798 / rG09652721 , we were potentially
returning a different result for whether an Instruction
is commutable depending on if we call the base class or
derived class method.
This requires relaxing asserts in GVN, but that pass
seems to be working otherwise.
NewGVN requires more work because it uses different
code paths for numbering binops and calls.
For an instruction in the basic block BB, SinkingPass enumerates basic blocks
dominated by BB and BB's successors. For each enumerated basic block,
SinkingPass uses `AllUsesDominatedByBlock` to check whether the basic
block dominates all of the instruction's users. This is inefficient.
Use the nearest common dominator of all users to avoid enumerating the
candidate. The nearest common dominator may be in a parent loop which is
not beneficial. In that case, find the ancestors in the dominator tree.
In the case that the instruction has no user, with this change we will
not perform unnecessary move. This causes some amdgpu test changes.
A stage-2 x86-64 clang is a byte identical with this change.
As discussed in D86798 / rG09652721 , we were potentially
returning a different result for whether an Instruction
is commutable depending on if we call the base class or
derived class method.
This requires relaxing an assert in GVN, but that pass
seems to be working otherwise.
NewGVN requires more work because it uses different
code paths for numbering binops and calls.
addRuntimeChecks uses SCEVExpander, which relies on the DT/LoopInfo to
be up-to-date. Changing the CFG afterwards may invalidate some inserted
instructions, especially LCSSA phis.
Reorder the code to first update the CFG and then create the runtime
checks. This should not have any impact on the generated code, as we
adjust the CFG and generate runtime checks together.
Fixes PR47343.
As discussed in D86798, it's not clear if the caller code
works with a more liberal definition of "commutative" that
includes intrinsics like min/max. This makes the binop
restriction (current functionality is unchanged) explicit
until the code is audited/tested.
The problem with module slice has been addressed in D86319
Introduce two new AAs. AAICVTrackerFunctionReturned which checks if a
function can have a unique ICV value after it is finished, and
AAICVCallSiteReturned which checks AAICVTrackerFunctionReturned for a
call site. This enables us to check the value of a call and if it
changes the ICV. This also changes the approach in
`getReplacementValues()` to a worklist-based approach so we can explore
all relevant BBs.
Differential Revision: https://reviews.llvm.org/D85544
Summary:
The module slice describes which functions we can analyze and transform
while working on an SCC as part of the Attributor-CGSCC pass. So far we
simply restricted it to the SCC.
Reviewers: jdoerfert
Differential Revision: https://reviews.llvm.org/D86319
This is the next patch of D86842
When we check `noundef` attribute violation at callsites, we do not have to require `nonnull` in the following two cases.
1. An argument is known to be simplified to undef
2. An argument is known to be dead
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D86845
Even though `noundef` IR attribute might be attached to non-void type values, AANoUndef is mistakenly identified for pointer type values only.
This patch fixes that.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D86737
Move bail out when optimizing for size before runtime check generation.
In that case, we do not use the result of the expansion, the expanded
instruction will be dead and cleaned up later.
By doing the check before expanding the runtime-checks, we can save a
bit of unnecessary work.
EarlyCSE has a mode to verify the invariant that hash equality equals
key equality, but EliminateDuplicatePHINodes() doesn't.
I've verified that this would have caught the stage2-stage3 mismatches
5ec2b757cc revert has fixed,
that were introduced last time in 3e69871ab5.
This patch fixes AANoUndef manifestation.
We should not manifest noundef for positions that will be changed to undef.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D86835
Handling the new min/max intrinsics is the motivation, but it
turns out that we have a bunch of other intrinsics with this
missing bit of analysis too.
The FP min/max tests show that we are intersecting FMF,
so that part should be safe too.
As noted in https://llvm.org/PR46897 , there is a commutative
property specifier for intrinsics, but no corresponding function
attribute, and so apparently no uses of that bit. We may want to
remove that next.
Follow-up patches should wire up the Instruction::isCommutative()
to this IntrinsicInst specialization. That requires updating
callers to be aware of the more general commutative property
(not just binops).
Differential Revision: https://reviews.llvm.org/D86798
The original take 1 was 6102310d81,
which taught InstSimplify to do that, which seemed better at time,
since we got EarlyCSE support for free.
However, it was proven that we can not do that there,
the simplified-to PHI would not be reachable from the original PHI,
and that is not something InstSimplify is allowed to do,
as noted in the commit ed90f15efb
that reverted it:
> It appears to cause compilation non-determinism and caused stage3 mismatches.
Then there was take 2 3e69871ab5,
which was InstCombine-specific, but it again showed stage2-stage3 differences,
and reverted in bdaa3f86a0.
This is quite alarming.
Here, let's try to change how we find existing PHI candidate:
due to the worklist order, and the way PHI nodes are inserted
(it may be inserted as the first one, or maybe not), let's look at *all*
PHI nodes in the block.
Effects on vanilla llvm test-suite + RawSpeed:
```
| statistic name | baseline | proposed | Δ | % | \|%\| |
|----------------------------------------------------|-----------|-----------|-------:|---------:|---------:|
| asm-printer.EmittedInsts | 7942329 | 7942457 | 128 | 0.00% | 0.00% |
| assembler.ObjectBytes | 254295632 | 254312480 | 16848 | 0.01% | 0.01% |
| correlated-value-propagation.NumPhis | 18412 | 18347 | -65 | -0.35% | 0.35% |
| early-cse.NumCSE | 2183283 | 2183267 | -16 | 0.00% | 0.00% |
| early-cse.NumSimplify | 550105 | 541842 | -8263 | -1.50% | 1.50% |
| instcombine.NumAggregateReconstructionsSimplified | 73 | 4506 | 4433 | 6072.60% | 6072.60% |
| instcombine.NumCombined | 3640311 | 3644419 | 4108 | 0.11% | 0.11% |
| instcombine.NumDeadInst | 1778204 | 1783205 | 5001 | 0.28% | 0.28% |
| instcombine.NumPHICSEs | 0 | 22490 | 22490 | 0.00% | 0.00% |
| instcombine.NumWorklistIterations | 2023272 | 2024400 | 1128 | 0.06% | 0.06% |
| instcount.NumCallInst | 1758395 | 1758802 | 407 | 0.02% | 0.02% |
| instcount.NumInvokeInst | 59478 | 59502 | 24 | 0.04% | 0.04% |
| instcount.NumPHIInst | 330557 | 330545 | -12 | 0.00% | 0.00% |
| instcount.TotalBlocks | 1077138 | 1077220 | 82 | 0.01% | 0.01% |
| instcount.TotalFuncs | 101442 | 101441 | -1 | 0.00% | 0.00% |
| instcount.TotalInsts | 8831946 | 8832606 | 660 | 0.01% | 0.01% |
| simplifycfg.NumHoistCommonCode | 24186 | 24187 | 1 | 0.00% | 0.00% |
| simplifycfg.NumInvokes | 4300 | 4410 | 110 | 2.56% | 2.56% |
| simplifycfg.NumSimpl | 1019813 | 999767 | -20046 | -1.97% | 1.97% |
```
So it fires 22490 times, which is less than ~24k the take 1 did,
but more than what take 2 did (22228 times)
.
It allows foldAggregateConstructionIntoAggregateReuse() to actually work
after PHI-of-extractvalue folds did their thing. Previously SimplifyCFG
would have done this PHI CSE, of all places. Additionally, allows some
more `invoke`->`call` folds to happen (+110, +2.56%).
All in all, expectedly, this catches less things overall,
but all the motivational cases are still caught, so all good.
While the original variant with doing this in InstSimplify (rightfully)
caused questions and ultimately was detected to be a culprit
of stage2-stage3 mismatch, it was expected that
InstCombine-based implementation would be fine.
But apparently it's not, as
http://lab.llvm.org:8011/builders/clang-with-thin-lto-ubuntu/builds/24095/steps/compare-compilers/logs/stdio
suggests.
Which suggests that somewhere in InstCombine there is a loop
over nondeterministically sorted container, which causes
different worklist ordering.
This reverts commit 3e69871ab5.
As we've established, if it takes more than two iterations
(one to perform folding and one to ensure that no folding opportunities
remain) per function, then there are worklist management issues.
So it may be interesting to keep track of it.
The original take was 6102310d81,
which taught InstSimplify to do that, which seemed better at time,
since we got EarlyCSE support for free.
However, it was proven that we can not do that there,
the simplified-to PHI would not be reachable from the original PHI,
and that is not something InstSimplify is allowed to do,
as noted in the commit ed90f15efb
that reverted it :
> It appears to cause compilation non-determinism and caused stage3 mismatches.
However InstCombine already does many different optimizations,
so it should be a safe place to do it here.
Note that we still can't just compare incoming values ranges,
because there is no guarantee that these PHI's we'd simplify to
were already re-visited and sorted.
However coming up with a test is problematic.
Effects on vanilla llvm test-suite + RawSpeed:
```
| statistic name | baseline | proposed | Δ | % | |%| |
|----------------------------------------------------|-----------|-----------|-------:|---------:|---------:|
| instcombine.NumPHICSEs | 0 | 22228 | 22228 | 0.00% | 0.00% |
| asm-printer.EmittedInsts | 7942329 | 7942456 | 127 | 0.00% | 0.00% |
| assembler.ObjectBytes | 254295632 | 254313792 | 18160 | 0.01% | 0.01% |
| early-cse.NumCSE | 2183283 | 2183272 | -11 | 0.00% | 0.00% |
| early-cse.NumSimplify | 550105 | 541842 | -8263 | -1.50% | 1.50% |
| instcombine.NumAggregateReconstructionsSimplified | 73 | 4506 | 4433 | 6072.60% | 6072.60% |
| instcombine.NumCombined | 3640311 | 3666911 | 26600 | 0.73% | 0.73% |
| instcombine.NumDeadInst | 1778204 | 1783318 | 5114 | 0.29% | 0.29% |
| instcount.NumCallInst | 1758395 | 1758804 | 409 | 0.02% | 0.02% |
| instcount.NumInvokeInst | 59478 | 59502 | 24 | 0.04% | 0.04% |
| instcount.NumPHIInst | 330557 | 330549 | -8 | 0.00% | 0.00% |
| instcount.TotalBlocks | 1077138 | 1077221 | 83 | 0.01% | 0.01% |
| instcount.TotalFuncs | 101442 | 101441 | -1 | 0.00% | 0.00% |
| instcount.TotalInsts | 8831946 | 8832611 | 665 | 0.01% | 0.01% |
| simplifycfg.NumInvokes | 4300 | 4410 | 110 | 2.56% | 2.56% |
| simplifycfg.NumSimpl | 1019813 | 999740 | -20073 | -1.97% | 1.97% |
```
So it fires ~22k times, which is less than ~24k the take 1 did.
It allows foldAggregateConstructionIntoAggregateReuse() to actually work
after PHI-of-extractvalue folds did their thing. Previously SimplifyCFG
would have done this PHI CSE, of all places. Additionally, allows some
more `invoke`->`call` folds to happen (+110, +2.56%).
All in all, expectedly, this catches less things overall,
but all the motivational cases are still caught, so all good.
instruction can decrement the reference count, not whether it can alter
it
This prevents the state transition from S_Use to S_CanRelease when doing
a bottom-up traversal and the transition from S_Retain to S_CanRelease
when doing a top-down traversal when the visited instruction can
increment the ref count but cannot decrement it. This allows the ARC
optimizer to remove retain/release pairs which were previously not
removed.
rdar://problem/21793154
There's a special case in hasAttribute for None when pImpl is null. If pImpl is not null we dispatch to pImpl->hasAttribute which will always return false for Attribute::None.
So if we just want to check for None its sufficient to just check that pImpl is null. Which can even be done inline.
This patch adds a helper for that case which I hope will speed up our getSubtargetImpl implementations.
Differential Revision: https://reviews.llvm.org/D86744
Since doInitialization() in the legacy pass modifies the module, the NPM
pass is a Module pass.
Reviewed By: ahatanak, ychen
Differential Revision: https://reviews.llvm.org/D86178
This patch fixes this crash https://gcc.godbolt.org/z/Ps8d1e
And gives SROA the ability to remove assumes if it allows promoting an alloca to register
Without removing assumes when it can't promote to register.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D86570
This patch changes ElementCount so that the Min and Scalable
members are now private and can only be accessed via the get
functions getKnownMinValue() and isScalable(). In addition I've
added some other member functions for more commonly used operations.
Hopefully this makes the class more useful and will reduce the
need for calling getKnownMinValue().
Differential Revision: https://reviews.llvm.org/D86065
This changes getDomMemoryDef to check if a Current is a valid
candidate for elimination before checking for reads. Before the change,
we were spending a lot of compile-time in checking for read accesses for
Current that might not even be removable.
This patch flips the logic, so we skip Current if they cannot be
removed before checking all their uses. This is much more efficient in
practice.
It also adds a more aggressive limit for checking partially overlapping
stores. The main problem with overlapping stores is that we do not know
if they will lead to elimination until seeing all of them. This patch
limits adds a new limit for overlapping store candidates, which keeps
the number of modified overlapping stores roughly the same.
This is another substantial compile-time improvement (while also
increasing the number of stores eliminated). Geomean -O3 -0.67%,
ReleaseThinLTO -0.97%.
http://llvm-compile-time-tracker.com/compare.php?from=0a929b6978a068af8ddb02d0d4714a2843dd8ba9&to=2e630629b43f64b60b282e90f0d96082fde2dacc&stat=instructions
Reviewed By: asbirlea
Differential Revision: https://reviews.llvm.org/D86487
strspn, strncmp, strcspn, strcasecmp, strncasecmp, memcmp, memchr,
memrchr, memcpy, memmove, memcpy, mempcpy, strchr, strrchr, bcmp
should all only access memory through their arguments.
I broke out strcoll, strcasecmp, strncasecmp because the result
depends on the locale, which might get accessed through memory.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D86724
David Stenberg