Working on top of D69252, this adds canonicalisation patterns for ssub.with.overflow to ssub.sats.
Differential Revision: https://reviews.llvm.org/D69753
This adds to D69245, adding extra signed patterns for folding from a
sadd_with_overflow to a sadd_sat. These are more complex than the
unsigned patterns, as the overflow can occur in either direction.
For the add case, the positive overflow can only occur if both of the
values are positive (same for both the values being negative). So there
is an extra select on whether to use the positive or negative overflow
limit.
Differential Revision: https://reviews.llvm.org/D69252
It was added in 2014 in 732e0aa9fb with one use in Scalarizer.cpp.
That one use was then removed when porting to the new pass manager in
2018 in b6f76002d9.
While the RFC and the desire to get off of static initializers for
cl::opt all still stand, this code is now dead, and I think we should
delete this code until someone is ready to do the migration.
There were many clients of CommandLine.h that were it transitively
through LLVMContext.h, so I cleaned that up in 4c1a1d3cf9.
Reviewers: beanz
Differential Revision: https://reviews.llvm.org/D70280
This is another step towards having FMF apply only to FP values
rather than those + fcmp. See PR38086 for one of the original
discussions/motivations:
https://bugs.llvm.org/show_bug.cgi?id=38086
And the test here is derived from PR39535:
https://bugs.llvm.org/show_bug.cgi?id=39535
Currently, we lose FMF when converting any phi to select in
SimplifyCFG. There are a small number of similar changes needed
to correct within SimplifyCFG, so it should be quick to patch
this pass up.
FMF was extended to select and phi with:
D61917
D67564
Differential Revision: https://reviews.llvm.org/D70208
This is a patch to support D66328, which was reverted until this lands.
Enable a compiler-rt test that used to fail previously with D66328.
Differential Revision: https://reviews.llvm.org/D67283
This patch introduces a function pass to inject the scalar-to-vector
mappings stored in the TargetLIbraryInfo (TLI) into the Vector
Function ABI (VFABI) variants attribute.
The test is testing the injection for three vector libraries supported
by the TLI (Accelerate, SVML, MASSV).
The pass does not change any of the analysis associated to the
function.
Differential Revision: https://reviews.llvm.org/D70107
ValueInfo has user-defined 'operator bool' which allows incorrect implicit conversion
to GlobalValue::GUID (which is unsigned long). This causes bugs which are hard to
track and should be removed in future.
Summary:
When scalarizing PHI nodes we might try to examine/rewrite
InsertElement nodes in predecessors. If those predecessors
are unreachable from entry, then the IR in those blocks could
have unexpected properties resulting in infinite loops in
Scatterer::operator[].
By simply treating values originating from instructions in
unreachable blocks as undef we do not need to analyse them
further.
This fixes PR41723.
Reviewers: bjope
Reviewed By: bjope
Subscribers: bjope, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D70171
getFirstNonPHI iterates over all the instructions in a block until it
finds a non-PHI.
Then, the loop starts from the beginning of the block and goes through
all the instructions until it reaches the instruction found by
getFirstNonPHI.
Instead of doing that, just stop when a non-PHI is found.
This reduces the compile-time of a test case discussed in
https://reviews.llvm.org/D47023 by 13x.
Not entirely sure how to come up with a test case for this since it's a
compile time issue that would significantly slow down running the tests.
Differential Revision: https://reviews.llvm.org/D70016
This reverts commit e511c4b0dff1692c267addf17dce3cebe8f97faa:
Temporarily Revert:
"[SLP] Generalization of stores vectorization."
"[SLP] Fix -Wunused-variable. NFC"
"[SLP] Vectorize jumbled stores."
after fixing the problem with compile time.
The vectoriser queries TTI->preferPredicateOverEpilogue to determine if
tail-folding is preferred for a loop, but it was not respecting loop hint
'predicate' that can disable this, which has now been added. This showed that
we were incorrectly initialising loop hint 'vectorize.predicate.enable' with 0
(i.e. FK_Disabled) but this should have been FK_Undefined, which has been
fixed.
Differential Revision: https://reviews.llvm.org/D70125
This is a resubmission of bbb29738b5 that
was reverted due to clang tests failures. It includes the fix and
additional IR tests for the missed case.
Summary:
In case when all incoming values of a PHI are equal pointers, this
transformation inserts a definition of such a pointer right after
definition of the base pointer and replaces with this value both PHI and
all it's incoming pointers. Primary goal of this transformation is
canonicalization of this pattern in order to enable optimizations that
can't handle PHIs. Non-inbounds pointers aren't currently supported.
Reviewers: spatel, RKSimon, lebedev.ri, apilipenko
Reviewed By: apilipenko
Tags: #llvm
Subscribers: hiraditya, llvm-commits
Differential Revision: https://reviews.llvm.org/D68128
This patch adds an assertion check for exported read/write-only
variables to be also in import list for module. If they aren't
we may face linker errors, because read/write-only variables are
internalized in their source modules. The patch also changes
export lists to store ValueInfo instead of GUID for performance
considerations.
Differential revision: https://reviews.llvm.org/D70128
Summary:
This fixes PR43081, where the transformation of `strchr(p, 0) -> p +
strlen(p)` can cause a segfault, if `-fno-builtin-strlen` is used. In
that case, `emitStrLen` returns nullptr, which CreateGEP is not designed
to handle. Also add the minimized code from the PR as a test case.
Reviewers: xbolva00, spatel, jdoerfert, efriedma
Reviewed By: efriedma
Subscribers: lebedev.ri, hiraditya, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D70143
This file lists every pass in LLVM, and is included by Pass.h, which is
very popular. Every time we add, remove, or rename a pass in LLVM, it
caused lots of recompilation.
I found this fact by looking at this table, which is sorted by the
number of times a file was changed over the last 100,000 git commits
multiplied by the number of object files that depend on it in the
current checkout:
recompiles touches affected_files header
342380 95 3604 llvm/include/llvm/ADT/STLExtras.h
314730 234 1345 llvm/include/llvm/InitializePasses.h
307036 118 2602 llvm/include/llvm/ADT/APInt.h
213049 59 3611 llvm/include/llvm/Support/MathExtras.h
170422 47 3626 llvm/include/llvm/Support/Compiler.h
162225 45 3605 llvm/include/llvm/ADT/Optional.h
158319 63 2513 llvm/include/llvm/ADT/Triple.h
140322 39 3598 llvm/include/llvm/ADT/StringRef.h
137647 59 2333 llvm/include/llvm/Support/Error.h
131619 73 1803 llvm/include/llvm/Support/FileSystem.h
Before this change, touching InitializePasses.h would cause 1345 files
to recompile. After this change, touching it only causes 550 compiles in
an incremental rebuild.
Reviewers: bkramer, asbirlea, bollu, jdoerfert
Differential Revision: https://reviews.llvm.org/D70211
Summary:
This temporarily disables the large working set size behavior in profile guided
size optimization due to internal benchmark regressions.
Reviewers: davidxl
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D70207
The bug manifests as replacing a reduction operand with an undef
value.
The problem appears to be limited to cases where a min/max reduction
has extra uses of the compare operand to the select.
In the general case, we are tracking "ExternallyUsedValues" and
an "IgnoreList" of the reduction operations, but those may not apply
to the final compare+select in a min/max reduction.
For that, we use replaceAllUsesWith (RAUW) to ensure that the new
vectorized reduction values are transferred to all subsequent users.
Differential Revision: https://reviews.llvm.org/D70148
As noted by the FIXME comment, this is not correct based on our current FMF semantics.
We should be propagating FMF from the final value in a sequence (in this case the
'select'). So the behavior even without this patch is wrong, but we did not allow FMF
on 'select' until recently.
But if we do the correct thing right now in this patch, we'll inevitably introduce
regressions because we have not wired up FMF propagation for 'phi' and 'select' in
other passes (like SimplifyCFG) or other places in InstCombine. I'm not seeing a
better incremental way to make progress.
That said, the potential extra damage over the existing wrong behavior from this
patch is very limited. AFAIK, the only way to have different FMF on IR in the same
function is if we have LTO inlined IR from 2 modules that were compiled using
different fast-math settings.
As seen in the tests, we may actually see some improvements with this patch because
adding the FMF to the 'select' allows matching to min/max intrinsics that were
previously missed (in the common case, the 'fcmp' and 'select' should have identical
FMF to begin with).
Next steps in the transition:
Make similar changes in instcombine as needed.
Enable phi-to-select FMF propagation in SimplifyCFG.
Remove dependencies on fcmp with FMF.
Deprecate FMF on fcmp.
Differential Revision: https://reviews.llvm.org/D69720
I think we have to be a bit more careful when it comes to moving
ops across shuffles, if the op does restrict undef. For example, without
this patch, we would move 'and %v, <0, 0, -1, -1>' over a
'shufflevector %a, undef, <undef, undef, 1, 2>'. As a result, the first
2 lanes of the result are undef after the combine, but they really
should be 0, unless I am missing something.
For ops that do fold to undef on undef operands, the current behavior
should be fine. I've add conservative check OpDoesRestrictUndef, maybe
there's a better existing utility?
Reviewers: spatel, RKSimon, lebedev.ri
Reviewed By: spatel
Differential Revision: https://reviews.llvm.org/D70093
In case when all incoming values of a PHI are equal pointers, this
transformation inserts a definition of such a pointer right after
definition of the base pointer and replaces with this value both PHI and
all it's incoming pointers. Primary goal of this transformation is
canonicalization of this pattern in order to enable optimizations that
can't handle PHIs. Non-inbounds pointers aren't currently supported.
Reviewers: spatel, RKSimon, lebedev.ri, apilipenko
Reviewed By: apilipenko
Tags: #llvm
Subscribers: hiraditya, llvm-commits
Differential Revision: https://reviews.llvm.org/D68128
Don't try to canonicalize loads to scalable vector types to loads
of integers.
This removes one assertion when trying to use a TypeSize as a parameter
to DataLayout::isLegalInteger. It does not handle the second part of the
function (which looks at bitcasts).
This patch also contains a NFC fix for Load Analysis, where a variable
initialization that would cause the same assertion is moved closer to
its use. This allows us to run the new test for InstCombine without
having to teach LocationSize to play nicely with scalable vectors.
Differential Revision: https://reviews.llvm.org/D70075
Currently we have limited support for outer loops with multiple basic
blocks after the inner loop exit. But the current checks for creating
PHIs for loop exit values only assumes the header and latches of the
outer loop. It is better to just skip incoming values defined in the
original inner loops. Those are handled earlier.
Reviewers: efriedma, mcrosier
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D70059
Summary:
This patch introduces align attribute deduction for callsite argument, function argument, function returned and floating value based on must-be-executed-context.
Reviewers: jdoerfert, sstefan1
Reviewed By: jdoerfert
Subscribers: hiraditya, jfb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69797
Summary: This patch introduces a new heuristic for guiding operand reordering. The new "look-ahead" heuristic can look beyond the immediate predecessors. This helps break ties when the immediate predecessors have identical opcodes (see lit test for examples).
Reviewers: RKSimon, ABataev, dtemirbulatov, Ayal, hfinkel, rnk
Reviewed By: RKSimon, dtemirbulatov
Subscribers: xbolva00, Carrot, hiraditya, phosek, rnk, rcorcs, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D60897
The attribute is stored at the `FunctionIndex` attribute set, with the
name "vector-function-abi-variant".
The get/set methods of the attribute have assertion to verify that:
1. Each name in the attribute is a valid VFABI mangled name.
2. Each name in the attribute correspond to a function declared in the
module.
Differential Revision: https://reviews.llvm.org/D69976
Re-try because earlier attempts were reverted due to use-after-free.
Hopefully, diagnosed correctly this time - we replace/remove the
invariant.start first rather than the invariant.end to avoid angering
worklist-based iteration.
We gather a set of white-listed instructions in isAllocSiteRemovable() and then
replace/erase them. But we don't know in general if the instructions in the set
have uses amongst themselves, so order of deletion makes a difference.
There's already a special-case for the llvm.objectsize intrinsic, so add another
for llvm.invariant.start.
Should fix:
https://bugs.llvm.org/show_bug.cgi?id=43723
Differential Revision: https://reviews.llvm.org/D69977
Summary:
SimplifySelectsFeedingBinaryOp simplified binary ops when both operands
were selects with the same condition. This patch extends it to handle
these cases where only one operand is a select:
X op (C ? P : Q) -> C ? (X op P) : (X op Q)
// if X op P and X op Q both simplify
(C ? P : Q) op Y -> C ? (P op Y) : (Q op Y)
// if P op Y and Q op Y both simplify
For example: X *fast (C ? 1.0 : 0.0) -> C ? X : 0.0
Reviewers: mcberg2017, majnemer, craig.topper, qcolombet, mcrosier
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D64713
The _m64 type is represented in IR as <1 x i64>. The x86-64 ABI
on Linux passes <1 x i64> as a double. MMX intrinsics use x86_mmx
type in IR.These things result in a lot of bitcasts in mmx code.
There's another instcombine that tries to turn bitcast <1 x i64>
to double into extractelement and a bitcast.
The combine here tries to reverse this extractelement conversion
if we see an mmx type.
Re-try rGef02831f0a4e (reverted due to use-after-free), but bail out completely
if we encounter an unexpected llvm.invariant.start.
We gather a set of white-listed instructions in isAllocSiteRemovable() and then
replace/erase them. But we don't know in general if the instructions in the set
have uses amongst themselves, so order of deletion makes a difference.
There's already a special-case for the llvm.objectsize intrinsic, so add another
for llvm.invariant.end.
Should fix:
https://bugs.llvm.org/show_bug.cgi?id=43723
Differential Revision: https://reviews.llvm.org/D69977
Summary: A helper function to get argument number of a arg operand Use.
Reviewers: jdoerfert, uenoku
Subscribers: hiraditya, lebedev.ri, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D66844
We gather a set of white-listed instructions in isAllocSiteRemovable() and then
replace/erase them. But we don't know in general if the instructions in the set
have uses amongst themselves, so order of deletion makes a difference.
There's already a special-case for the llvm.objectsize intrinsic, so add another
for llvm.invariant.end.
Should fix:
https://bugs.llvm.org/show_bug.cgi?id=43723
Differential Revision: https://reviews.llvm.org/D69977
This recommits 11ed1c0239 (reverted in
9f08ce0d21 for failing an assert) with a fix:
tryToWidenMemory() now first checks if the widening decision is to interleave,
thus maintaining previous behavior where tryToInterleaveMemory() was called
first, giving priority to interleave decisions over widening/scalarization. This
commit adds the test case that exposed this bug as a LIT.
Summary: A user can force a function to be inlined by specifying the always_inline attribute. Currently, thinlto implementation is not aware of always_inline functions and does not guarantee import of such functions, which in turn can prevent inlining of such functions.
Patch by Bharathi Seshadri <bseshadr@cisco.com>
Reviewers: tejohnson
Reviewed By: tejohnson
Subscribers: mehdi_amini, inglorion, hiraditya, steven_wu, dexonsmith, arphaman, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D70014
Patch enables import of write-only variables with non-trivial initializers
to fix linker errors. Initializers of imported variables are converted to
'zeroinitializer' to avoid promotion of referenced objects.
Differential revision: https://reviews.llvm.org/D70006
This patch implements a correct, but not terribly useful, transform. In particular, if we have a dynamic alloca in a loop which is guaranteed to execute, and provably not captured, we hoist the alloca out of the loop. The capture tracking is needed so that we can prove that each previous stack region dies before the next one is allocated. The transform decreases the amount of stack allocation needed by a linear factor (e.g. the iteration count of the loop).
Now, I really hope no one is actually using dynamic allocas. As such, why this patch?
Well, the actual problem I'm hoping to make progress on is allocation hoisting. There's a large draft patch out for review (https://reviews.llvm.org/D60056), and this patch was the smallest chunk of testable functionality I could come up with which takes a step vaguely in that direction.
Once this is in, it makes motivating the changes to capture tracking mentioned in TODOs testable. After that, I hope to extend this to trivial malloc free regions (i.e. free dominating all loop exits) and allocation functions for GCed languages.
Differential Revision: https://reviews.llvm.org/D69227
The change itself is straight forward and obvious, but ... there's an existing test checking for exactly the opposite. Both I and Artur think this is simply conservatism in the initial implementation. If anyone bisects a problem to this, a counter example will be very interesting.
Differential Revision: https://reviews.llvm.org/D69907
This recommits 100e797adb (reverted in
009e032634 for failing an assert). While the
root cause was independently reverted in eaff300401,
this commit includes a LIT to make sure IVDescriptor's SinkAfter logic does not
try to sink branch instructions.
x86_mmx is conceptually a vector already. Don't introduce an extra conversion between it and scalar i64.
I'm using VectorType::isValidElementType which checks for floating point, integer, and pointers to hopefully make this more readable than just blacklisting x86_mmx.
Differential Revision: https://reviews.llvm.org/D69964
Summary:
I need to make use of this pass from a driver program that isn't opt.
Therefore this patch moves this pass into the LLVM library so that it is
available for use elsewhere.
There was one function I kept in tools/opt which is exportDebugifyStats()
this is because it's serializing the statistics into a human readable
format and this seemed more in keeping with opt than a library function
Reviewers: vsk, aprantl
Subscribers: mgorny, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69926
Instcombiner pass was erasing trivially dead instruction without updating dependent llvm.dbg.value.
which was not showing programmer current state of variables while debugging.
As a part of this fix I did following,
Iterate throught all the users (llvm.dbg) of a instruction which is trivially dead and set each if them undef, Before deleting the instruction.
Now user will see optimized out, when try to print those variables.
This fixes
https://bugs.llvm.org/show_bug.cgi?id=43893
This is my first fix to llvm.
Patch by kamlesh kumar!
Differential Revision: https://reviews.llvm.org/D69809
shift (logic (shift X, C0), Y), C1 --> logic (shift X, C0+C1), (shift Y, C1)
This is an IR translation of an existing SDAG transform added here:
rL370617
So we again have 9 possible patterns with a commuted IR variant of each pattern:
https://rise4fun.com/Alive/VlIhttps://rise4fun.com/Alive/n1mhttps://rise4fun.com/Alive/1Vn
Part of the motivation is to allow easier recognition and subsequent
canonicalization of bswap patterns as discussed in PR43146:
https://bugs.llvm.org/show_bug.cgi?id=43146
We had to delay this transform because it used to allow the SLP vectorizer
to create awful reductions out of simple load-combines.
That problem was fixed with:
rL375025
(we'll bring back load combining in IR someday...)
The backend is also better equipped to deal with these patterns now
using hooks like TLI.getShiftAmountThreshold().
The only remaining potential controversy is that the -reassociate pass
tends to reverse this kind of pattern (to help GVN?). But since -reassociate
doesn't do anything with these specific patterns, there is no conflict currently.
Finally, there's a new pass proposal at D67383 for general tree-height-reduction
reassociation, and it could use a cost model to decide how to optimally rearrange
these kinds of ops for a target. That patch appears to be stalled.
Differential Revision: https://reviews.llvm.org/D69842
Patch allows importing declarations of functions and variables, referenced
by the initializer of some other readonly variable.
Differential revision: https://reviews.llvm.org/D69561
We have a vector compare reduction problem seen in PR39665 comment 2:
https://bugs.llvm.org/show_bug.cgi?id=39665#c2
Or slightly reduced here:
define i1 @cmp2(<2 x double> %a0) {
%a = fcmp ogt <2 x double> %a0, <double 1.0, double 1.0>
%b = extractelement <2 x i1> %a, i32 0
%c = extractelement <2 x i1> %a, i32 1
%d = and i1 %b, %c
ret i1 %d
}
SLP would not attempt to turn this into a vector reduction because there is an
artificial lower limit on that transform. We can not completely remove that limit
without inducing regressions though, so this patch just hacks an extra attempt at
creating a 2-way reduction to the end of the analysis.
As shown in the test file, we are still not getting some of the motivating cases,
so follow-on patches will be needed to solve those cases.
Differential Revision: https://reviews.llvm.org/D59710
Summary:
When adjusting function entry counts after inlining, Funciton::setEntryCount is called without providing an import function list. The side effect of that is the previously set import function list will be dropped. The import function list is used by ThinLTO to help import hot cross module callee for LTO inlining, so dropping that during ThinLTO pre-link may adversely affect LTO inlining. The fix is to keep the list while updating entry counts for inlining.
Reviewers: wmi, davidxl, tejohnson
Subscribers: mehdi_amini, hiraditya, dexonsmith, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69736
"[SLP] Generalization of stores vectorization."
"[SLP] Fix -Wunused-variable. NFC"
"[SLP] Vectorize jumbled stores."
As they're causing significant (10-30x) compile time regressions on
vectorizable code.
The primary cause of the compile-time regression is f228b53716.
This reverts commits:
f228b537165503455ccb21d498c9c0
The basic idea of the transform is to convert variant loop exit conditions into invariant exit conditions by changing the iteration on which the exit is taken when we know that the trip count is unobservable. See the original patch which introduced the code for a more complete explanation.
The individual parts of this have been reviewed, the result has been fuzzed, and then further analyzed by hand, but despite all of that, I will not be suprised to see breakage here. If you see problems, please don't hesitate to revert - though please do provide a test case. The most likely class of issues are latent SCEV bugs and without a reduced test case, I'll be essentially stuck on reducing them.
(Note: A bunch of tests were opted out of the new transform to preserve coverage. That landed in a previous commit to simplify revert cycles if they turn out to be needed.)
Summary:
This patch factors out code to clone instructions -- partly for
readability and partly to facilitate an upcoming patch of my own.
Reviewers: wmi
Subscribers: hiraditya, jfb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69861
We had a subtle, but nasty bug in our definition of a widenable branch, and thus in the transforms which used that utility. Specifically, we returned true for any branch which included a widenable condition within it's condition, regardless of whether that widenable condition also had other uses.
The problem is that the result of the WC() call is defined to be one particular value. As such, all users must agree as to what that value is. If we widen a branch without also updating *all other users* of the WC in the same way, we have broken the required semantics.
Most of the textual diff is updating existing transforms not to leave dead uses hanging around. They're largely NFC as the dead instructions would be immediately deleted by other passes. The reason to make these changes is so that the transforms preserve the widenable branch form.
In practice, we don't get bitten by this only because it isn't profitable to CSE WC() calls and the lowering pass from guards uses distinct WC calls per branch.
Differential Revision: https://reviews.llvm.org/D69916
This patch fixes two issues noticed by inspection when going to enable the loop predication code in IndVarSimplify.
Issue 1 - Both the LoopPredication transform, and the already on by default optimizeLoopExits transform, modify the exit count of the exits they modify. (either to 0 or Infinity) Looking at the code more closely, this was not reflected into SCEV and we were instead running later transforms with incorrect SCEVs. Fixing this requires forgetting the loop, weakening a too strong assert, and updating SCEV to not pessimize results when a loop is provable untaken. I haven't been able to find a test case to demonstrate the miscompile.
Issue 2 - For modules without a data layout, we can end up with unsized pointer typed exit counts. Just bail out of this case.
I think these are the last two issues which need addressed before we enable this by default. The code has already survived a decent amount of fuzzing without revealing either of the above.
Differential Revision: https://reviews.llvm.org/D69695
Summary:
I believe this bisects to https://reviews.llvm.org/D44983
(`[LoopUnroll] Only peel if a predicate becomes known in the loop body.`)
While that revision did contain tests that showed arguably-subpar peeling
for [in]equality predicates that [not] happen in the middle of the loop,
it also disabled peeling for the *first* loop iteration,
because latch would be canonicalized to [in]equality comparison..
That was intentional as per https://reviews.llvm.org/D44983#1059583.
I'm not 100% sure that i'm using correct checks here,
but this fix appears to be going in the right direction..
Let me know if i'm missing some checks here..
Fixes [[ https://bugs.llvm.org/show_bug.cgi?id=43840 | PR43840 ]].
Reviewers: fhahn, mkazantsev, efriedma
Reviewed By: fhahn
Subscribers: xbolva00, hiraditya, zzheng, llvm-commits, fhahn
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69617
We have two ways to steer creating a predicated vector body over creating a
scalar epilogue. To force this, we have 1) a command line option and 2) a
pragma available. This adds a third: a target hook to TargetTransformInfo that
can be queried whether predication is preferred or not, which allows the
vectoriser to make the decision without forcing it.
While this change behaves as a non-functional change for now, it shows the
required TTI plumbing, usage of this new hook in the vectoriser, and the
beginning of an ARM MVE implementation. I will follow up on this with:
- a complete MVE implementation, see D69845.
- a patch to disable this, i.e. we should respect "vector_predicate(disable)"
and its corresponding loophint.
Differential Revision: https://reviews.llvm.org/D69040
Summary:
This patch factors out code to merge a basic block with its sole
successor -- partly for readability and partly to facilitate an
upcoming patch of my own.
Reviewers: wmi
Subscribers: hiraditya, jfb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69852
This recommits 2be17087f8 (reverted in
d3ec06d219 for heap-use-after-free) with a fix
in IAI's reset() which was not clearing the set of interleave groups after
deleting them.
When eliminating a pair of
`llvm.objc.autoreleaseReturnValue`
followed by
`llvm.objc.retainAutoreleasedReturnValue`
we need to make sure that the instructions in between are safe to
ignore.
Other than bitcasts and useless GEPs, it's also safe to ignore lifetime
markers for both static allocas (lifetime.start/lifetime.end) and dynamic
allocas (stacksave/stackrestore).
These get added by the inliner as part of the return sequence and can
prevent the transformation from happening in practice.
Differential Revision: https://reviews.llvm.org/D69833
Summary:
This patch factors out common code to update the SSA form in
JumpThreading.cpp -- partly for readability and partly to facilitate
an coming patch of my own.
Reviewers: wmi
Subscribers: hiraditya, jfb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69811
Summary:
That fold keeps growing and growing :(
I think this may be one of the last pieces for it.
Since D67677/D67725, the fold knowns the general form
of the pattern - where some masking is needed:
https://rise4fun.com/Alive/F5Rhttps://rise4fun.com/Alive/gslRa
But there is one more huge piece missing - if you are extracting some bits,
it is not impossible that the origin is wider than the extraction,
i.e. there may be a truncation. And we don't deal with that yet.
But we can, and the generalization remains fully identical:
https://rise4fun.com/Alive/Uarhttps://rise4fun.com/Alive/5SW
After a preparatory cleanup i think the diff looks rather clean.
One missing piece is that in some patterns (especially pat. b),
`-1` only needs to be `-1` in final type, but that is for later..
https://bugs.llvm.org/show_bug.cgi?id=42563
Reviewers: spatel, nikic
Reviewed By: spatel
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69125
This transformation is a variation on the GuardWidening transformation we have checked in as it's own pass. Instead of focusing on merge (i.e. hoisting and simplifying) two widenable branches, this transform makes the observation that simply removing a second slowpath block (by reusing an existing one) is often a very useful canonicalization. This may lead to later merging, or may not. This is a useful generalization when the intermediate block has loads whose dereferenceability is hard to establish.
As noted in the patch, this can be generalized further, and will be.
Differential Revision: https://reviews.llvm.org/D69689
This reverts commit 004ed2b0d1.
Original commit hash 6d03890384
Summary:
This adds a clang option to disable inline line tables. When it is used,
the inliner uses the call site as the location of the inlined function instead of
marking it as an inline location with the function location.
https://reviews.llvm.org/D67723
Summary:
If the GEP instructions are going to be vectorized, the indices in those
GEP instructions must be of the same type. Otherwise, the compiler may
crash when trying to build the vector constant.
Reviewers: RKSimon, spatel
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69627
This was an experiment made possible by a non-standard feature of the
Android dynamic loader.
It required introducing a flag to tell the compiler which ABI was being
targeted.
This flag is no longer needed, since the generated code now works for
both ABI's.
We leave that flag untouched for backwards compatibility. This also
means that if we need to distinguish between targeted ABI's again
we can do that without disturbing any existing workflows.
We leave a comment in the source code and mention in the help text to
explain this for any confused person reading the code in the future.
Patch by Matthew Malcomson
Differential Revision: https://reviews.llvm.org/D69574
The sink-after and interleave-group vectorization decisions were so far applied to
VPlan during initial VPlan construction, which complicates VPlan construction – also because of
their inter-dependence. This patch refactors buildVPlanWithRecipes() to construct a simpler
initial VPlan and later apply both these vectorization decisions, in order, as VPlan-to-VPlan
transformations.
Differential Revision: https://reviews.llvm.org/D68577
Dependences between two abstract attributes SRC and TRG come naturally in
two flavors:
Either (1) "some" information of SRC is *required* for TRG to derive
information, or (2) SRC is just an *optional* way for TRG to derive
information.
While it is not strictly necessary to distinguish these types
explicitly, it can help us to converge faster, in terms of iterations,
and also cut down the number of `AbstractAttribute::update` calls.
As far as I can tell, we only use optional dependences for liveness so
far but that might change in the future. With this change the Attributor
can be informed about the "dependence class" and it will perform
appropriate actions when an Attribute is set to an invalid state, thus
one that cannot be used by others to derive information from.
As discussed in https://bugs.llvm.org/show_bug.cgi?id=43870,
this transform is missing a crucial legality check:
the old (non-countable) loop would early-return upon first mismatch,
but there is no such guarantee for bcmp/memcmp.
We'd need to ensure that [PtrA, PtrA+NBytes) and [PtrB, PtrB+NBytes)
are fully dereferenceable memory regions. But that would limit
the transform to constant loop trip counts and would further
cripple it because dereferenceability analysis is *very* partial.
Furthermore, even if all that is done, every single test
would need to be rewritten from scratch.
So let's just give up.
BlockAddress users will not "call" the function so they do not qualify
as call sites in the first place. When we delete a function with
BlockAddress users we need to first remove the body so they are properly
discarded.
When we replace constant returns at the call site we did issue a cast in
the hopes it would be a no-op if the types are equal. Turns out that is
not the case and we have to check it ourselves first.
Reused an IPConstantProp test for coverage. No functional change to the
test wrt. IPConstantProp.
Even if the invoked function may-return, we can replace it with a call
and branch if it is nounwind. We had almost everything in place to do
this but did not which actually caused a crash when we removed the
landingpad from the actually dead unwind block.
Exposed by the IPConstantProp tests.
We did merge "known" and "assumed" liveness information into a single
set which caused various kinds of problems, especially because we did
not properly record when something was actually known. With this patch
we properly track the "known" bit and distinguish dead ends we know from
the ones we still need to explore in future updates.
We gave up on `noreturn` if `willreturn` was known for a while but we
now again try to always derive `noreturn`. This is useful because a
function that is `noreturn` + `willreturn` is basically dead as
executing it would lead to undefined behavior (UB).
This came up in the IPConstantProp cases where a function only contained
a unreachable but was not marked `noreturn` which caused missed
opportunities down the line.
In D69605 only the "cases" of a switch were handled but if none matched
we did not make the default case life. This is fixed now and properly
tested (with code from IPConstantProp/user-with-multiple-uses.ll).
We cannot simply replace arguments that carry attributes like `nest`,
`inalloca`, `sret`, and `byval`. Except for the last one, which we can
replace if it is not written, we bail for now.
Trying to deduce information for declarations and calls sites of
declarations is not useful in practice but only for testing. Add a flag
that disables this by default but also enable it in the tests.
The misc.ll test will verify the flag "works" as expected.
We cannot look at the subsuming positions and take their nocapture bit
as shown with the two tests for which we derived nocapture on the call
site argument and readonly on the argument of the second before.
This recommits cc0b9647b7 which was
reverted in d39d1a2f87.
I added a fix for an issue found when testing via distributed ThinLTO,
and added a test case for that failure.
Before we did not follow casts and geps when we looked at the users of a
pointer in the pointers must-be-executed-context. This caused us to fail
to determine if it was accessed for sure. With this change we follow
such users now.
The above extension exposed problems in getKnownNonNullAndDerefBytesForUse
which did not always check what the base pointer was. We also did not
handle negative offsets as conservative as we have to without explicit
loop handling. Finally, we should not derive a huge number if we access
a pointer that was traversed backwards first.
The problems exposed by this functional change are already tested in the
existing test cases as is the functional change.
Differential Revision: https://reviews.llvm.org/D69647
Summary:
Patch adds support for vectorization of the jumbled stores. The value
operands are vectorized and then shuffled in the right order before
store.
Reviewers: RKSimon, spatel, hfinkel, mkuper
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D43339
Summary:
In order to get context sensitivity from isKnownNonZero we need to
provide a context instruction *and* a dominator tree. The latter is
passed now to which actually allows to remove some initialization code.
Tests taken from PR43833.
Reviewers: uenoku, sstefan1
Subscribers: hiraditya, bollu, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69595
The replacement code only looks at the first index of the
extractvalue. If there are additional indices we'll end
up doing a bad replacement.
This only happens if the function returns a nested struct. Not
sure if clang ever generates such code. The original report came
from ispc.
Fixes PR43857
Differential Revision: https://reviews.llvm.org/D69656
This adds some patterns to transform uadd.with.overflow to uadd.sat
(with usub.with.overflow to usub.sat too). The patterns selects from
UINTMAX (or 0 for subs) depending on whether the operation overflowed.
Signed patterns are a little more involved (they can wrap in two
directions), but can be added here in a followup patch too.
Differential Revision: https://reviews.llvm.org/D69245
Since SCEV can cache information about location of an instruction, it should be invalidated when the instruction is moved.
There should be similar bug in code sinking part of LICM, it will be fixed in a follow-up change.
Patch Author: Daniil Suchkov
Reviewers: asbirlea, mkazantsev, reames
Reviewed By: asbirlea
Subscribers: hiraditya, javed.absar, llvm-commits
Differential Revision: https://reviews.llvm.org/D69370
Summary:
If control is transferred to a successor is the key question when it
comes to liveness. The new implementation puts that question in the
focus and thereby providing a clean way to assume certain CFG edges are
dead or instructions will not transfer control.
Reviewers: sstefan1, uenoku
Subscribers: hiraditya, bollu, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69605
Summary:
This patch introduces liveness (AAIsDead) for all positions, thus for
all kinds of values. For now, we say an instruction is dead if it would
be removed assuming all users are dead. A call site return is different
as we just look at the users. If all call site returns have been
eliminated, the return values can return undef instead of their original
value, eliminating uses.
We try to recursively delete dead instructions now and we introduce a
simple check interface for use-traversal.
This is the idea tried out in D68626 but implemented in the right way.
Reviewers: uenoku, sstefan1
Subscribers: hiraditya, bollu, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68925
Deleting blocks will require us to deal with dead edges, e.g.,
`br i1 false, label %live, label %dead`
explicitly. For now we just clear the blocks and move on.
This will be revisited once we actually fold branches.
Summary:
If there is a unique free of the allocated that has to be reached from
the malloc, we can apply the heap-2-stack transformation even if the
pointer escapes.
Reviewers: hfinkel, sstefan1, uenoku
Subscribers: hiraditya, bollu, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68958
If an attribute did not query any optimistic (=non-fixed) information to
justify its state, we know the attribute state will not change anymore.
Thus, we can indicate an optimistic fixpoint.
We pretended IRPosition came either as mutable or immutable objects
while they are basically always immutable, with a single (existing)
unfortunate exceptions. This patch cleans up the uses to deal with the
immutable version.
Summary:
Clang does not add type metadata to available_externally vtables. When
choosing a summary to look at for virtual function definitions, make
sure we skip summaries for any available externally vtables as they will
not describe any virtual function functions, which are only summarized
in the presence of type metadata on the vtable def. Simply look for the
corresponding strong def's summary.
Also add handling for same-named local vtables with the same GUID
because of same-named files without enough distinguishing path.
In that case we return a conservative result with no devirtualization.
Reviewers: pcc, davidxl, evgeny777
Subscribers: mehdi_amini, inglorion, hiraditya, steven_wu, dexonsmith, arphaman, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69452
Summary:
This adds a clang option to disable inline line tables. When it is used,
the inliner uses the call site as the location of the inlined function instead of
marking it as an inline location with the function location.
See https://bugs.llvm.org/show_bug.cgi?id=42344
Reviewers: rnk
Subscribers: hiraditya, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D67723
Summary:
in the following C code the branch is not removed by clang in O3.
```
int f1(char* p) {
int i1 = __builtin_strlen(p);
if (!p)
return -1;
return i1;
}
```
The issue is that the call to strlen is sunk to the following block by instcombine. In its new place the call to strlen doesn't dominate the use in the icmp anymore so value tracking can't see that p cannot be null.
This patch resolves the issue by inserting an assumption at the place of the call before sinking a call when that call can be used to prove an argument to be nonnull.
This resolves this issue at O3.
Reviewers: majnemer, xbolva00, fhahn, jdoerfert, spatel, efriedma
Reviewed By: jdoerfert
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69477
Summary:
Patch adds support for vectorization of the jumbled stores. The value
operands are vectorized and then shuffled in the right order before
store.
Reviewers: RKSimon, spatel, hfinkel, mkuper
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D43339
The address sanitizer ignore memory accesses from different address
spaces, however when instrumenting globals the check for different
address spaces is missing. This result in assertion failure. The fault
was found in an out of tree target.
The patch skip all globals of non default address space.
Reviewed By: leonardchan, vitalybuka
Differential Revision: https://reviews.llvm.org/D68790
Stores are vectorized with maximum vectorization factor of 16. Patch
tries to improve the situation and use maximal vectorization factor.
Reviewers: spatel, RKSimon, mkuper, hfinkel
Differential Revision: https://reviews.llvm.org/D43582
This is a fix for:
https://bugs.llvm.org/show_bug.cgi?id=43730
...and as shown there, we have existing test cases that show potential miscompiles.
We could just bail out for vector constants that contain any undef elements, or we can do as shown here:
allow the transform, but replace the undefs with a safe value.
For most of the tests shown, this results in a full splat constant (no undefs) which is probably a win
for further IR analysis because we conservatively don't match undefs in most cases. Codegen can probably
recover these kinds of undef lanes via demanded elements analysis if that's profitable.
Differential Revision: https://reviews.llvm.org/D69519
Summary:
Currently we only forget the loop we added LCSSA phis for. But SCEV
expressions in other loops could also depend on the instruction we added
a PHI for and currently we do not invalidate those expressions. This can
happen when we use ScalarEvolution before converting a function to LCSSA
form. The SCEV expressions will refer to the non-LCSSA value. If this
SCEV expression is then used with the expander, we do not preserve LCSSA
form.
This patch properly forgets the values we created PHIs for. Those need
to be recomputed again. This patch fixes PR43458.
Currently SCEV::verify does not catch this mismatch and any test would
need to run multiple passes to trigger the error (e.g. -loop-reduce
-loop-unroll). I will also look into catching this kind of mismatch in
the verifier. Also, we currently forget the whole loop in LCSSA and I'll
check if we can be more surgical.
Reviewers: efriedma, sanjoy.google, reames
Reviewed By: efriedma
Subscribers: zzheng, hiraditya, javed.absar, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68194
To make IntegerState more flexible but also less error prone we split it
up into (1) incrementing, (2) decrementing, and (3) bit-tracking states.
This adds functionality compared to before and disallows misuse, e.g.,
"incrementing" updates on a bit-tracking state.
Part of the change is a single operator in the base class which
simplifies helper functions that deal with states.
There are certain functional changes but all of which should actually be
corrections.
Summary:
MSan instrumentation adds stores and loads even to pure
readonly/writeonly functions. It is removing some of those attributes
from instrumented functions and call targets, but apparently not enough.
Remove writeonly, argmemonly and speculatable in addition to readonly /
readnone.
Reviewers: pcc, vitalybuka
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69541
Summary:
(Split of off D67120)
SizeOpts/MachineSizeOpts changes for profile guided size optimization.
(A second try after previously committed as r375254 and reverted as r375375.)
Subscribers: mgorny, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69409
Currently we may do iterleaving by more than estimated trip count
coming from the profile or computed maximum trip count. The solution is to
use "best known" trip count instead of exact one in interleaving analysis.
Patch by Evgeniy Brevnov.
Differential Revision: https://reviews.llvm.org/D67948
Summary:
Debug info affects output from "opt -inline", InlineFunction could
not handle the llvm.dbg.value when it exist between alloca
instructions.
Problem was that the first alloca in a sequence of allocas was
handled differently from the subsequence alloca instructions. Now
all static alloca instructions are treated the same (being removed
if the have no uses). So it does not matter if there are dbg
instructions (or any other instructions) in between.
Fix the issue: https://bugs.llvm.org/show_bug.cgi?id=43291k
Patch by: yechunliang (Chris Ye)
Reviewers: bjope, jmorse, vsk, probinson, jdoerfert, mtrofin, aprantl, fhahn
Reviewed By: bjope
Subscribers: uabelho, ormris, aprantl, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68633
This is an extra fold for a canonical form of uadd_sat, as shown in
D68651. It essentially selects uadd from an add and a select.
Differential Revision: https://reviews.llvm.org/D69244
Summary:
A new function pass (Transforms/CFGuard/CFGuard.cpp) inserts CFGuard checks on
indirect function calls, using either the check mechanism (X86, ARM, AArch64) or
or the dispatch mechanism (X86-64). The check mechanism requires a new calling
convention for the supported targets. The dispatch mechanism adds the target as
an operand bundle, which is processed by SelectionDAG. Another pass
(CodeGen/CFGuardLongjmp.cpp) identifies and emits valid longjmp targets, as
required by /guard:cf. This feature is enabled using the `cfguard` CC1 option.
Reviewers: thakis, rnk, theraven, pcc
Subscribers: ychen, hans, metalcanine, dmajor, tomrittervg, alex, mehdi_amini, mgorny, javed.absar, kristof.beyls, hiraditya, steven_wu, dexonsmith, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D65761
This phi simplification transform was added with:
D45448
However as shown in PR43802:
https://bugs.llvm.org/show_bug.cgi?id=43802
...we must be careful not to propagate poison when we do the substitution.
There might be some more complicated analysis possible to retain the overflow flag,
but it should always be safe and easy to drop flags (we have similar behavior in
instcombine and other passes).
Differential Revision: https://reviews.llvm.org/D69442
Summary:
If there are a GUID collision between two globals checking the
summarylist from the import index to make assumption can be dangerous.
Do not assume that a GlobalValue that has a GlobalVarSummary
actually is a GlobalVariable as it can be another GlobalValue with
the same GUID that the summary is connected to.
Patch by Joel Klinghed (the_jk@opera.com)
Reviewers: evgeny777, tejohnson
Reviewed By: tejohnson
Subscribers: tejohnson, dblaikie, MaskRay, mehdi_amini, inglorion, hiraditya, steven_wu, dexonsmith, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D67322
We were already going to all of the trouble of computing maximum constant exit counts for each loop exit, we might as well expose them through the API. The change in IndVars is mostly to demonstrate that the wired up code works, but it als very slightly strengthens the transform. The strengthened case is rather narrow though: it requires one exactly analyzeable exit, one imprecisely analyzeable exit (with the upper bound less than the precise one), and one unanalyzeable exit. I coudn't construct a reasonably stable test case.
This does increase the memory usage of the BackedgeTakenCount by a factor of 2 in the worst case.
I also noticed the loop in IndVars is O(#Exits ^ 2). This doesn't change with this patch. A future patch will cache this result inside of SCEV to avoid requering.
The MVE VADC instruction reads and writes the carry bit at bit 29 of
the FPSCR register. The corresponding ACLE intrinsic is specified to
work with an integer in which the carry bit is stored at bit 0. So if
a user writes a code sequence in C that passes the carry from one VADC
to the next, like this,
s0 = vadcq_u32(a0, b0, &carry);
s1 = vadcq_u32(a1, b1, &carry);
then clang will generate IR for each of those operations that shifts
the carry bit up into bit 29 before the VADC, and after it, shifts it
back down and masks off all but the low bit. But in this situation
what you really wanted was two consecutive VADC instructions, so that
the second one directly reads the value left in FPSCR by the first,
without wasting several instructions on pointlessly clearing the other
flag bits in between.
This commit explains to InstCombine that the other bits of the flags
operand don't matter, and adds a test that demonstrates that all the
code between the two VADC instructions can be optimized away as a
result.
Reviewers: dmgreen, miyuki, ostannard
Subscribers: kristof.beyls, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D67162
This adds an instcombine matcher for code that attempts to perform signed
saturating arithmetic by casting to a higher type. Unsigned cases are already
matched, this adds extra matches for the more complex signed cases, which
involves matching the min(max(add a b)) nodes with proper extends to ensure
legality.
Differential Revision: https://reviews.llvm.org/D68651
llvm-svn: 375505
Summary:
Reduce include dependencies by no longer including Pass.h from
DataLayout.h. That include seemed irrelevant to DataLayout, as
well as being irrelevant to several users of DataLayout.
Reviewers: rnk
Reviewed By: rnk
Subscribers: mehdi_amini, hiraditya, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D69261
llvm-svn: 375436
The static analyzer is warning about a potential null dereference, but we should be able to use cast<> directly and if not assert will fire for us.
llvm-svn: 375429
The static analyzer is warning about a potential null dereference, but we should be able to use cast<> directly and if not assert will fire for us.
llvm-svn: 375427
The static analyzer is warning about a potential null dereference, but we should be able to use cast<> directly and if not assert will fire for us.
llvm-svn: 375426
Summary:
When MemCpyOpt is handling aggregate type values, if an instruction (let's call it P) between the targeting load (L) and store (S) clobbers the source pointer of L, it will try to hoist S before P. This process will also hoist S's data dependency instructions.
However, the current implementation has a bug that if one of S's dependency instructions is //also// a user of P, MemCpyOpt will not prevent it from being hoisted above P and cause a use-before-define error. For example, in the newly added test file (i.e. `aggregate-type-crash.ll`), it will try to hoist both `store %my_struct %1, %my_struct* %3` and its dependent, `%3 = bitcast i8* %2 to %my_struct*`, above `%2 = call i8* @my_malloc(%my_struct* %0)`. Creating the following BB:
```
entry:
%1 = bitcast i8* %4 to %my_struct*
%2 = bitcast %my_struct* %1 to i8*
%3 = bitcast %my_struct* %0 to i8*
call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 %2, i8* align 4 %3, i64 8, i1 false)
%4 = call i8* @my_malloc(%my_struct* %0)
ret void
```
Where there is a use-before-define error between `%1` and `%4`.
Update: The compiler for the Pony Programming Language [also encounter the same bug](https://github.com/ponylang/ponyc/issues/3140)
Patch by Min-Yih Hsu (myhsu)
Reviewers: eugenis, pcc, dblaikie, dneilson, t.p.northover, lattner
Reviewed By: eugenis
Subscribers: lenary, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D66060
llvm-svn: 375403
Summary:
Allow for ignoring the check for a single use in SimplifyDemandedVectorElts
to be able to simplify operands if DemandedElts is known to contain
the union of elements used by all users.
It is a responsibility of a caller of SimplifyDemandedVectorElts to
supply correct DemandedElts.
Simplify a series of extractelement instructions if only a subset of
elements is used.
Reviewers: reames, arsenm, majnemer, nhaehnle
Reviewed By: nhaehnle
Subscribers: wdng, jvesely, nhaehnle, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D67345
llvm-svn: 375395
AAReturnedValues, AAMemoryBehavior, and AANoUnwind, can provide
information that helps during the tracking or even justifies no-capture.
We now use this information and enable no-capture in some test cases
designed a long while a ago for these cases.
llvm-svn: 375382
We can end up with two loop exits whose exit counts are equivalent, but whose textual representation is different and non-obvious. For the sub-case where we have a series of exits which dominate one another (common), eliminate any exits which would iterate *after* a previous exit on the exiting iteration.
As noted in the TODO being removed, I'd always thought this was a good idea, but I've now seen this in a real workload as well.
Interestingly, in review, Nikita pointed out there's let another oppurtunity to leverage SCEV's reasoning. If we kept track of the min of dominanting exits so far, we could discharge exits with EC >= MDE. This is less powerful than the existing transform (since later exits aren't considered), but potentially more powerful for any case where SCEV can prove a >= b, but neither a == b or a > b. I don't have an example to illustrate that oppurtunity, but won't be suprised if we find one and return to handle that case as well.
Differential Revision: https://reviews.llvm.org/D69009
llvm-svn: 375379
In this pattern, all the "magic" bits that we'd `add` are all
high sign bits, and in the value we'd be adding to they are all unset,
not unexpectedly, so we can have an `or` there:
https://rise4fun.com/Alive/ups
It is possible that `haveNoCommonBitsSet()` should be taught about this
pattern so that we never have an `add` variant, but the reasoning would
need to be recursive (because of that `select`), so i'm not really sure
that would be worth it just yet.
llvm-svn: 375378
This adds folds for comparing uadd.sat/usub.sat with zero:
* uadd.sat(a, b) == 0 => a == 0 && b == 0 => (a | b) == 0
* usub.sat(a, b) == 0 => a <= b
And inverted forms for !=.
Differential Revision: https://reviews.llvm.org/D69224
llvm-svn: 375374
Summary:
This problem consists of several parts:
* Basic sign bit extraction - `trunc? (?shr %x, (bitwidth(x)-1))`.
This is trivial, and easy to do, we have a fold for it.
* Shift amount reassociation - if we have two identical shifts,
and we can simplify-add their shift amounts together,
then we likely can just perform them as a single shift.
But this is finicky, has one-use restrictions,
and shift opcodes must be identical.
But there is a super-pattern where both of these work together.
to produce sign bit test from two shifts + comparison.
We do indeed already handle this in most cases.
But since we get that fold transitively, it has one-use restrictions.
And what's worse, in this case the right-shifts aren't required to be
identical, and we can't handle that transitively:
If the total shift amount is bitwidth-1, only a sign bit will remain
in the output value. But if we look at this from the perspective of
two shifts, we can't fold - we can't possibly know what bit pattern
we'd produce via two shifts, it will be *some* kind of a mask
produced from original sign bit, but we just can't tell it's shape:
https://rise4fun.com/Alive/cM0https://rise4fun.com/Alive/9IN
But it will *only* contain sign bit and zeros.
So from the perspective of sign bit test, we're good:
https://rise4fun.com/Alive/FRzhttps://rise4fun.com/Alive/qBU
Superb!
So the simplest solution is to extend `reassociateShiftAmtsOfTwoSameDirectionShifts()` to also have a
sudo-analysis mode that will ignore extra-uses, and will only check
whether a) those are two right shifts and b) they end up with bitwidth(x)-1
shift amount and return either the original value that we sign-checking,
or null.
This does not have any functionality change for
the existing `reassociateShiftAmtsOfTwoSameDirectionShifts()`.
All that being said, as disscussed in the review, this yet again
increases usage of instsimplify in instcombine as utility.
Some day that may need to be reevaluated.
https://bugs.llvm.org/show_bug.cgi?id=43595
Reviewers: spatel, efriedma, vsk
Reviewed By: spatel
Subscribers: xbolva00, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68930
llvm-svn: 375371
by ExtBinary format profile
Profile on-demand loading was added for ExtBinary format profile in rL374233,
but currently profile on-demand loading doesn't work well with profile
remapping. The patch adds the support.
Suppose a function in the current module has outline instance in the profile.
The function name in the module is different from the name of the outline
instance, but remapper knows the two names are equal. When loading profile
on-demand, the outline instance has to be loaded with remapper's help.
At the same time SampleProfileReaderItaniumRemapper is changed from a proxy
of SampleProfileReader to a helper member in SampleProfileReader.
Differential Revision: https://reviews.llvm.org/D68901
llvm-svn: 375295
Summary:
CVP, unlike InstCombine, does not run till exaustion.
It only does a single pass.
When dealing with those special binops, if we prove that they can
safely be demoted into their usual binop form,
we do set the no-wrap we deduced. But when dealing with usual binops,
we try to deduce both no-wraps.
So if we convert e.g. @llvm.uadd.with.overflow() to `add nuw`,
we won't attempt to check whether it can be `add nuw nsw`.
This patch proposes to call `processBinOp()` on newly-created binop,
which is identical to what we do for div/rem already.
Reviewers: nikic, spatel, reames
Reviewed By: nikic
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69183
llvm-svn: 375273
Summary:
It looks like this is the only missing statistic in the CVP pass.
Since we prove NSW and NUW separately i'd think we should count them separately too.
Reviewers: nikic, spatel, reames
Reviewed By: spatel
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68740
llvm-svn: 375230
In the process of writing D69009, I realized we have two distinct sets of invariants within this single function, and basically no shared logic. The optimize loop exit transforms (including the new one in D69009) only care about *analyzeable* exits. Loop predication, on the other hand, has to reason about *all* exits. At the moment, we have the property (due to the requirement for an exact btc) that all exits are analyzeable, but that will likely change in the future as we add widenable condition support.
llvm-svn: 375138
The static analyzer is warning about a potential null dereference, but we should be able to use cast<> directly and if not assert will fire for us.
llvm-svn: 375103
We can't normally stumble into that assertion because a tautological
*conditional* `br` in loop body is required, one that always
branches to loop latch. But that should have been always folded
to an unconditional branch before we get it.
But that is not guaranteed if the pass is run standalone.
So let's just promote the assertion into a proper check.
Fixes https://bugs.llvm.org/show_bug.cgi?id=43687
llvm-svn: 375100
Remove dead virtual functions from vtables with
replaceNonMetadataUsesWith, so that CGProfile metadata gets cleaned up
correctly.
Original commit message:
Currently, it is hard for the compiler to remove unused C++ virtual
functions, because they are all referenced from vtables, which are referenced
by constructors. This means that if the constructor is called from any live
code, then we keep every virtual function in the final link, even if there
are no call sites which can use it.
This patch allows unused virtual functions to be removed during LTO (and
regular compilation in limited circumstances) by using type metadata to match
virtual function call sites to the vtable slots they might load from. This
information can then be used in the global dead code elimination pass instead
of the references from vtables to virtual functions, to more accurately
determine which functions are reachable.
To make this transformation safe, I have changed clang's code-generation to
always load virtual function pointers using the llvm.type.checked.load
intrinsic, instead of regular load instructions. I originally tried writing
this using clang's existing code-generation, which uses the llvm.type.test
and llvm.assume intrinsics after doing a normal load. However, it is possible
for optimisations to obscure the relationship between the GEP, load and
llvm.type.test, causing GlobalDCE to fail to find virtual function call
sites.
The existing linkage and visibility types don't accurately describe the scope
in which a virtual call could be made which uses a given vtable. This is
wider than the visibility of the type itself, because a virtual function call
could be made using a more-visible base class. I've added a new
!vcall_visibility metadata type to represent this, described in
TypeMetadata.rst. The internalization pass and libLTO have been updated to
change this metadata when linking is performed.
This doesn't currently work with ThinLTO, because it needs to see every call
to llvm.type.checked.load in the linkage unit. It might be possible to
extend this optimisation to be able to use the ThinLTO summary, as was done
for devirtualization, but until then that combination is rejected in the
clang driver.
To test this, I've written a fuzzer which generates random C++ programs with
complex class inheritance graphs, and virtual functions called through object
and function pointers of different types. The programs are spread across
multiple translation units and DSOs to test the different visibility
restrictions.
I've also tried doing bootstrap builds of LLVM to test this. This isn't
ideal, because only classes in anonymous namespaces can be optimised with
-fvisibility=default, and some parts of LLVM (plugins and bugpoint) do not
work correctly with -fvisibility=hidden. However, there are only 12 test
failures when building with -fvisibility=hidden (and an unmodified compiler),
and this change does not cause any new failures for either value of
-fvisibility.
On the 7 C++ sub-benchmarks of SPEC2006, this gives a geomean code-size
reduction of ~6%, over a baseline compiled with "-O2 -flto
-fvisibility=hidden -fwhole-program-vtables". The best cases are reductions
of ~14% in 450.soplex and 483.xalancbmk, and there are no code size
increases.
I've also run this on a set of 8 mbed-os examples compiled for Armv7M, which
show a geomean size reduction of ~3%, again with no size increases.
I had hoped that this would have no effect on performance, which would allow
it to awlays be enabled (when using -fwhole-program-vtables). However, the
changes in clang to use the llvm.type.checked.load intrinsic are causing ~1%
performance regression in the C++ parts of SPEC2006. It should be possible to
recover some of this perf loss by teaching optimisations about the
llvm.type.checked.load intrinsic, which would make it worth turning this on
by default (though it's still dependent on -fwhole-program-vtables).
Differential revision: https://reviews.llvm.org/D63932
llvm-svn: 375094
Summary:
There are two cases where a block is merged into its predecessor and the
MergeBlockIntoPredecessor API is not used. Update the API so it can be
reused in the other cases, in order to avoid code duplication.
Cleanup motivated by D68659.
Reviewers: chandlerc, sanjoy.google, george.burgess.iv
Subscribers: llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68670
llvm-svn: 375050
The problem is that we can have two loop exits, 'a' and 'b', where 'a' and 'b' would exit at the same iteration, 'a' precedes 'b' along some path, and 'b' is predicated while 'a' is not. In this case (see the previously submitted test case), we causing the loop to exit through 'b' whereas it should have exited through 'a'.
This only applies to loop exits where the exit counts are not provably inequal, but that isn't as much of a restriction as it appears. If we could order the exit counts, we'd have already removed one of the two exits. In theory, we might be able to prove inequality w/o ordering, but I didn't really explore that piece. Instead, I went for the obvious restriction and ensured we didn't predicate exits following non-predicateable exits.
Credit goes to Evgeny Brevnov for figuring out the problematic case. Fuzzing probably also found it (failures seen), but due to some silly infrastructure problems I hadn't gotten to the results before Evgeny hand reduced it from a benchmark (he manually enabled the transform). Once this is fixed, I'll try to filter through the fuzzer failures to see if there's anything additional lurking.
Differential Revision https://reviews.llvm.org/D68956
llvm-svn: 375038
The 1st attempt at this modified the cost model in a bad way to avoid the vectorization,
but that caused problems for other users (the loop vectorizer) of the cost model.
I don't see an ideal solution to these 2 related, potentially large, perf regressions:
https://bugs.llvm.org/show_bug.cgi?id=42708https://bugs.llvm.org/show_bug.cgi?id=43146
We decided that load combining was unsuitable for IR because it could obscure other
optimizations in IR. So we removed the LoadCombiner pass and deferred to the backend.
Therefore, preventing SLP from destroying load combine opportunities requires that it
recognizes patterns that could be combined later, but not do the optimization itself (
it's not a vector combine anyway, so it's probably out-of-scope for SLP).
Here, we add a cost-independent bailout with a conservative pattern match for a
multi-instruction sequence that can probably be reduced later.
In the x86 tests shown (and discussed in more detail in the bug reports), SDAG combining
will produce a single instruction on these tests like:
movbe rax, qword ptr [rdi]
or:
mov rax, qword ptr [rdi]
Not some (half) vector monstrosity as we currently do using SLP:
vpmovzxbq ymm0, dword ptr [rdi + 1] # ymm0 = mem[0],zero,zero,..
vpsllvq ymm0, ymm0, ymmword ptr [rip + .LCPI0_0]
movzx eax, byte ptr [rdi]
movzx ecx, byte ptr [rdi + 5]
shl rcx, 40
movzx edx, byte ptr [rdi + 6]
shl rdx, 48
or rdx, rcx
movzx ecx, byte ptr [rdi + 7]
shl rcx, 56
or rcx, rdx
or rcx, rax
vextracti128 xmm1, ymm0, 1
vpor xmm0, xmm0, xmm1
vpshufd xmm1, xmm0, 78 # xmm1 = xmm0[2,3,0,1]
vpor xmm0, xmm0, xmm1
vmovq rax, xmm0
or rax, rcx
vzeroupper
ret
Differential Revision: https://reviews.llvm.org/D67841
llvm-svn: 375025
Summary:
This is something of a workaround to avoid a crash later on in type
legalizer (WidenVectorResult()).
Also added some f16 tests, including a non-working v3f16 case with
a FIXME.
Reviewers: arsenm, tpr, nhaehnle
Reviewed By: arsenm
Subscribers: kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68865
llvm-svn: 374993
Check that a call has an attached MemoryAccess before calling
getClobbering on the instruction.
If no access is attached, the instruction does not access memory.
Resolves PR43441.
llvm-svn: 374920
The 1st attempt at rL374828 inserted the code
at the wrong position (outside of the constant-shift-amount
block). Trying again with an additional test to verify
const-ness.
For a constant shift amount, add the following fold.
shl (zext (i1 X)), ShAmt --> select (X, 1 << ShAmt, 0)
https://rise4fun.com/Alive/IZ9
Fixes PR42257.
Based on original patch by @zvi (Zvi Rackover)
Differential Revision: https://reviews.llvm.org/D63382
llvm-svn: 374886
For a constant shift amount, add the following fold.
shl (zext (i1 X)), ShAmt --> select (X, 1 << ShAmt, 0)
https://rise4fun.com/Alive/IZ9
Fixes PR42257.
Based on original patch by @zvi (Zvi Rackover)
Differential Revision: https://reviews.llvm.org/D63382
llvm-svn: 374828
As reported by Joerg Sonnenberger in IRC, for 32-bit systems,
where pointer and size_t are 32-bit, if you use 64-bit-wide variable
in the loop, you could end up with loop exit count being of the type
wider than the size_t. Now, i'm not sure if we can produce `bcmp`
from that (just truncate?), but we certainly should not assert/miscompile.
llvm-svn: 374811
Add a pass to lower is.constant and objectsize intrinsics
This pass lowers is.constant and objectsize intrinsics not simplified by
earlier constant folding, i.e. if the object given is not constant or if
not using the optimized pass chain. The result is recursively simplified
and constant conditionals are pruned, so that dead blocks are removed
even for -O0. This allows inline asm blocks with operand constraints to
work all the time.
The new pass replaces the existing lowering in the codegen-prepare pass
and fallbacks in SDAG/GlobalISEL and FastISel. The latter now assert
on the intrinsics.
Differential Revision: https://reviews.llvm.org/D65280
llvm-svn: 374784
Add an extra parameter so the backend can take the alignment into
consideration.
Differential Revision: https://reviews.llvm.org/D68400
llvm-svn: 374763
This pass lowers is.constant and objectsize intrinsics not simplified by
earlier constant folding, i.e. if the object given is not constant or if
not using the optimized pass chain. The result is recursively simplified
and constant conditionals are pruned, so that dead blocks are removed
even for -O0. This allows inline asm blocks with operand constraints to
work all the time.
The new pass replaces the existing lowering in the codegen-prepare pass
and fallbacks in SDAG/GlobalISEL and FastISel. The latter now assert
on the intrinsics.
Differential Revision: https://reviews.llvm.org/D65280
llvm-svn: 374743
No-return and will-return are exclusive, assuming the latter is more
prominent we can avoid updates of the former unless will-return is not
known for sure.
llvm-svn: 374739
Even if an argument is captured, we cannot have an effect the function
does not have. This is fine except for the special case of `inalloca` as
it does not behave by the rules.
TODO: Maybe the special rule for `inalloca` is wrong after all.
llvm-svn: 374736
Summary:
This changes "CHECK" check lines to "ATTRIBUTOR" check lines where
necessary and also fixes the now exposed, mostly minor, problems.
Reviewers: sstefan1, uenoku
Subscribers: hiraditya, bollu, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68929
llvm-svn: 374735
Follow-up to D68244 to account for a corner case discussed in:
https://bugs.llvm.org/show_bug.cgi?id=43501
Add one more restriction: if the pointer is deref-or-null and in a non-default
(non-zero) address space, we can't assume inbounds.
Differential Revision: https://reviews.llvm.org/D68706
llvm-svn: 374728
Before, we eagerly split blocks even if it was not necessary, e.g., they
had a single unreachable instruction and only a single predecessor.
llvm-svn: 374703
We do not yet perform h2s because we know something is free'ed but we do
it because we know the pointer does not escape. Storing the pointer
allows it to escape so we have to prevent that.
llvm-svn: 374699
H2S did apply to mallocs of non-constant sizes if the uses were OK. This
is now forbidden through reording of the "good" and "bad" cases in the
conditional.
llvm-svn: 374698
The check for naked/optnone was insufficient for different reasons. We
now check before we initialize an abstract attribute and we do it for
all abstract attributes.
llvm-svn: 374694
Summary:
If the underlying alloca did not change, we do not necessarily need new
lifetime markers. This patch adds a check and reuses the old ones if
possible.
Reviewers: reames, ssarda, t.p.northover, hfinkel
Subscribers: hiraditya, bollu, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68900
llvm-svn: 374692
Summary:
This is a recommit, this originally landed in rL370454 but was
subsequently reverted in rL370788 due to
https://bugs.llvm.org/show_bug.cgi?id=43206
The reduced testcase was added to bcmp-negative-tests.ll
as @pr43206_different_loops - we must ensure that the SCEV's
we got are both for the same loop we are currently investigating.
Original commit message:
@mclow.lists brought up this issue up in IRC.
It is a reasonably common problem to compare some two values for equality.
Those may be just some integers, strings or arrays of integers.
In C, there is `memcmp()`, `bcmp()` functions.
In C++, there exists `std::equal()` algorithm.
One can also write that function manually.
libstdc++'s `std::equal()` is specialized to directly call `memcmp()` for
various types, but not `std::byte` from C++2a. https://godbolt.org/z/mx2ejJ
libc++ does not do anything like that, it simply relies on simple C++'s
`operator==()`. https://godbolt.org/z/er0Zwf (GOOD!)
So likely, there exists a certain performance opportunities.
Let's compare performance of naive `std::equal()` (no `memcmp()`) with one that
is using `memcmp()` (in this case, compiled with modified compiler). {F8768213}
```
#include <algorithm>
#include <cmath>
#include <cstdint>
#include <iterator>
#include <limits>
#include <random>
#include <type_traits>
#include <utility>
#include <vector>
#include "benchmark/benchmark.h"
template <class T>
bool equal(T* a, T* a_end, T* b) noexcept {
for (; a != a_end; ++a, ++b) {
if (*a != *b) return false;
}
return true;
}
template <typename T>
std::vector<T> getVectorOfRandomNumbers(size_t count) {
std::random_device rd;
std::mt19937 gen(rd());
std::uniform_int_distribution<T> dis(std::numeric_limits<T>::min(),
std::numeric_limits<T>::max());
std::vector<T> v;
v.reserve(count);
std::generate_n(std::back_inserter(v), count,
[&dis, &gen]() { return dis(gen); });
assert(v.size() == count);
return v;
}
struct Identical {
template <typename T>
static std::pair<std::vector<T>, std::vector<T>> Gen(size_t count) {
auto Tmp = getVectorOfRandomNumbers<T>(count);
return std::make_pair(Tmp, std::move(Tmp));
}
};
struct InequalHalfway {
template <typename T>
static std::pair<std::vector<T>, std::vector<T>> Gen(size_t count) {
auto V0 = getVectorOfRandomNumbers<T>(count);
auto V1 = V0;
V1[V1.size() / size_t(2)]++; // just change the value.
return std::make_pair(std::move(V0), std::move(V1));
}
};
template <class T, class Gen>
void BM_bcmp(benchmark::State& state) {
const size_t Length = state.range(0);
const std::pair<std::vector<T>, std::vector<T>> Data =
Gen::template Gen<T>(Length);
const std::vector<T>& a = Data.first;
const std::vector<T>& b = Data.second;
assert(a.size() == Length && b.size() == a.size());
benchmark::ClobberMemory();
benchmark::DoNotOptimize(a);
benchmark::DoNotOptimize(a.data());
benchmark::DoNotOptimize(b);
benchmark::DoNotOptimize(b.data());
for (auto _ : state) {
const bool is_equal = equal(a.data(), a.data() + a.size(), b.data());
benchmark::DoNotOptimize(is_equal);
}
state.SetComplexityN(Length);
state.counters["eltcnt"] =
benchmark::Counter(Length, benchmark::Counter::kIsIterationInvariant);
state.counters["eltcnt/sec"] =
benchmark::Counter(Length, benchmark::Counter::kIsIterationInvariantRate);
const size_t BytesRead = 2 * sizeof(T) * Length;
state.counters["bytes_read/iteration"] =
benchmark::Counter(BytesRead, benchmark::Counter::kDefaults,
benchmark::Counter::OneK::kIs1024);
state.counters["bytes_read/sec"] = benchmark::Counter(
BytesRead, benchmark::Counter::kIsIterationInvariantRate,
benchmark::Counter::OneK::kIs1024);
}
template <typename T>
static void CustomArguments(benchmark::internal::Benchmark* b) {
const size_t L2SizeBytes = []() {
for (const benchmark::CPUInfo::CacheInfo& I :
benchmark::CPUInfo::Get().caches) {
if (I.level == 2) return I.size;
}
return 0;
}();
// What is the largest range we can check to always fit within given L2 cache?
const size_t MaxLen = L2SizeBytes / /*total bufs*/ 2 /
/*maximal elt size*/ sizeof(T) / /*safety margin*/ 2;
b->RangeMultiplier(2)->Range(1, MaxLen)->Complexity(benchmark::oN);
}
BENCHMARK_TEMPLATE(BM_bcmp, uint8_t, Identical)
->Apply(CustomArguments<uint8_t>);
BENCHMARK_TEMPLATE(BM_bcmp, uint16_t, Identical)
->Apply(CustomArguments<uint16_t>);
BENCHMARK_TEMPLATE(BM_bcmp, uint32_t, Identical)
->Apply(CustomArguments<uint32_t>);
BENCHMARK_TEMPLATE(BM_bcmp, uint64_t, Identical)
->Apply(CustomArguments<uint64_t>);
BENCHMARK_TEMPLATE(BM_bcmp, uint8_t, InequalHalfway)
->Apply(CustomArguments<uint8_t>);
BENCHMARK_TEMPLATE(BM_bcmp, uint16_t, InequalHalfway)
->Apply(CustomArguments<uint16_t>);
BENCHMARK_TEMPLATE(BM_bcmp, uint32_t, InequalHalfway)
->Apply(CustomArguments<uint32_t>);
BENCHMARK_TEMPLATE(BM_bcmp, uint64_t, InequalHalfway)
->Apply(CustomArguments<uint64_t>);
```
{F8768210}
```
$ ~/src/googlebenchmark/tools/compare.py --no-utest benchmarks build-{old,new}/test/llvm-bcmp-bench
RUNNING: build-old/test/llvm-bcmp-bench --benchmark_out=/tmp/tmpb6PEUx
2019-04-25 21:17:11
Running build-old/test/llvm-bcmp-bench
Run on (8 X 4000 MHz CPU s)
CPU Caches:
L1 Data 16K (x8)
L1 Instruction 64K (x4)
L2 Unified 2048K (x4)
L3 Unified 8192K (x1)
Load Average: 0.65, 3.90, 4.14
---------------------------------------------------------------------------------------------------
Benchmark Time CPU Iterations UserCounters...
---------------------------------------------------------------------------------------------------
<...>
BM_bcmp<uint8_t, Identical>/512000 432131 ns 432101 ns 1613 bytes_read/iteration=1000k bytes_read/sec=2.20706G/s eltcnt=825.856M eltcnt/sec=1.18491G/s
BM_bcmp<uint8_t, Identical>_BigO 0.86 N 0.86 N
BM_bcmp<uint8_t, Identical>_RMS 8 % 8 %
<...>
BM_bcmp<uint16_t, Identical>/256000 161408 ns 161409 ns 4027 bytes_read/iteration=1000k bytes_read/sec=5.90843G/s eltcnt=1030.91M eltcnt/sec=1.58603G/s
BM_bcmp<uint16_t, Identical>_BigO 0.67 N 0.67 N
BM_bcmp<uint16_t, Identical>_RMS 25 % 25 %
<...>
BM_bcmp<uint32_t, Identical>/128000 81497 ns 81488 ns 8415 bytes_read/iteration=1000k bytes_read/sec=11.7032G/s eltcnt=1077.12M eltcnt/sec=1.57078G/s
BM_bcmp<uint32_t, Identical>_BigO 0.71 N 0.71 N
BM_bcmp<uint32_t, Identical>_RMS 42 % 42 %
<...>
BM_bcmp<uint64_t, Identical>/64000 50138 ns 50138 ns 10909 bytes_read/iteration=1000k bytes_read/sec=19.0209G/s eltcnt=698.176M eltcnt/sec=1.27647G/s
BM_bcmp<uint64_t, Identical>_BigO 0.84 N 0.84 N
BM_bcmp<uint64_t, Identical>_RMS 27 % 27 %
<...>
BM_bcmp<uint8_t, InequalHalfway>/512000 192405 ns 192392 ns 3638 bytes_read/iteration=1000k bytes_read/sec=4.95694G/s eltcnt=1.86266G eltcnt/sec=2.66124G/s
BM_bcmp<uint8_t, InequalHalfway>_BigO 0.38 N 0.38 N
BM_bcmp<uint8_t, InequalHalfway>_RMS 3 % 3 %
<...>
BM_bcmp<uint16_t, InequalHalfway>/256000 127858 ns 127860 ns 5477 bytes_read/iteration=1000k bytes_read/sec=7.45873G/s eltcnt=1.40211G eltcnt/sec=2.00219G/s
BM_bcmp<uint16_t, InequalHalfway>_BigO 0.50 N 0.50 N
BM_bcmp<uint16_t, InequalHalfway>_RMS 0 % 0 %
<...>
BM_bcmp<uint32_t, InequalHalfway>/128000 49140 ns 49140 ns 14281 bytes_read/iteration=1000k bytes_read/sec=19.4072G/s eltcnt=1.82797G eltcnt/sec=2.60478G/s
BM_bcmp<uint32_t, InequalHalfway>_BigO 0.40 N 0.40 N
BM_bcmp<uint32_t, InequalHalfway>_RMS 18 % 18 %
<...>
BM_bcmp<uint64_t, InequalHalfway>/64000 32101 ns 32099 ns 21786 bytes_read/iteration=1000k bytes_read/sec=29.7101G/s eltcnt=1.3943G eltcnt/sec=1.99381G/s
BM_bcmp<uint64_t, InequalHalfway>_BigO 0.50 N 0.50 N
BM_bcmp<uint64_t, InequalHalfway>_RMS 1 % 1 %
RUNNING: build-new/test/llvm-bcmp-bench --benchmark_out=/tmp/tmpQ46PP0
2019-04-25 21:19:29
Running build-new/test/llvm-bcmp-bench
Run on (8 X 4000 MHz CPU s)
CPU Caches:
L1 Data 16K (x8)
L1 Instruction 64K (x4)
L2 Unified 2048K (x4)
L3 Unified 8192K (x1)
Load Average: 1.01, 2.85, 3.71
---------------------------------------------------------------------------------------------------
Benchmark Time CPU Iterations UserCounters...
---------------------------------------------------------------------------------------------------
<...>
BM_bcmp<uint8_t, Identical>/512000 18593 ns 18590 ns 37565 bytes_read/iteration=1000k bytes_read/sec=51.2991G/s eltcnt=19.2333G eltcnt/sec=27.541G/s
BM_bcmp<uint8_t, Identical>_BigO 0.04 N 0.04 N
BM_bcmp<uint8_t, Identical>_RMS 37 % 37 %
<...>
BM_bcmp<uint16_t, Identical>/256000 18950 ns 18948 ns 37223 bytes_read/iteration=1000k bytes_read/sec=50.3324G/s eltcnt=9.52909G eltcnt/sec=13.511G/s
BM_bcmp<uint16_t, Identical>_BigO 0.08 N 0.08 N
BM_bcmp<uint16_t, Identical>_RMS 34 % 34 %
<...>
BM_bcmp<uint32_t, Identical>/128000 18627 ns 18627 ns 37895 bytes_read/iteration=1000k bytes_read/sec=51.198G/s eltcnt=4.85056G eltcnt/sec=6.87168G/s
BM_bcmp<uint32_t, Identical>_BigO 0.16 N 0.16 N
BM_bcmp<uint32_t, Identical>_RMS 35 % 35 %
<...>
BM_bcmp<uint64_t, Identical>/64000 18855 ns 18855 ns 37458 bytes_read/iteration=1000k bytes_read/sec=50.5791G/s eltcnt=2.39731G eltcnt/sec=3.3943G/s
BM_bcmp<uint64_t, Identical>_BigO 0.32 N 0.32 N
BM_bcmp<uint64_t, Identical>_RMS 33 % 33 %
<...>
BM_bcmp<uint8_t, InequalHalfway>/512000 9570 ns 9569 ns 73500 bytes_read/iteration=1000k bytes_read/sec=99.6601G/s eltcnt=37.632G eltcnt/sec=53.5046G/s
BM_bcmp<uint8_t, InequalHalfway>_BigO 0.02 N 0.02 N
BM_bcmp<uint8_t, InequalHalfway>_RMS 29 % 29 %
<...>
BM_bcmp<uint16_t, InequalHalfway>/256000 9547 ns 9547 ns 74343 bytes_read/iteration=1000k bytes_read/sec=99.8971G/s eltcnt=19.0318G eltcnt/sec=26.8159G/s
BM_bcmp<uint16_t, InequalHalfway>_BigO 0.04 N 0.04 N
BM_bcmp<uint16_t, InequalHalfway>_RMS 29 % 29 %
<...>
BM_bcmp<uint32_t, InequalHalfway>/128000 9396 ns 9394 ns 73521 bytes_read/iteration=1000k bytes_read/sec=101.518G/s eltcnt=9.41069G eltcnt/sec=13.6255G/s
BM_bcmp<uint32_t, InequalHalfway>_BigO 0.08 N 0.08 N
BM_bcmp<uint32_t, InequalHalfway>_RMS 30 % 30 %
<...>
BM_bcmp<uint64_t, InequalHalfway>/64000 9499 ns 9498 ns 73802 bytes_read/iteration=1000k bytes_read/sec=100.405G/s eltcnt=4.72333G eltcnt/sec=6.73808G/s
BM_bcmp<uint64_t, InequalHalfway>_BigO 0.16 N 0.16 N
BM_bcmp<uint64_t, InequalHalfway>_RMS 28 % 28 %
Comparing build-old/test/llvm-bcmp-bench to build-new/test/llvm-bcmp-bench
Benchmark Time CPU Time Old Time New CPU Old CPU New
---------------------------------------------------------------------------------------------------------------------------------------
<...>
BM_bcmp<uint8_t, Identical>/512000 -0.9570 -0.9570 432131 18593 432101 18590
<...>
BM_bcmp<uint16_t, Identical>/256000 -0.8826 -0.8826 161408 18950 161409 18948
<...>
BM_bcmp<uint32_t, Identical>/128000 -0.7714 -0.7714 81497 18627 81488 18627
<...>
BM_bcmp<uint64_t, Identical>/64000 -0.6239 -0.6239 50138 18855 50138 18855
<...>
BM_bcmp<uint8_t, InequalHalfway>/512000 -0.9503 -0.9503 192405 9570 192392 9569
<...>
BM_bcmp<uint16_t, InequalHalfway>/256000 -0.9253 -0.9253 127858 9547 127860 9547
<...>
BM_bcmp<uint32_t, InequalHalfway>/128000 -0.8088 -0.8088 49140 9396 49140 9394
<...>
BM_bcmp<uint64_t, InequalHalfway>/64000 -0.7041 -0.7041 32101 9499 32099 9498
```
What can we tell from the benchmark?
* Performance of naive equality check somewhat improves with element size,
maxing out at eltcnt/sec=1.58603G/s for uint16_t, or bytes_read/sec=19.0209G/s
for uint64_t. I think, that instability implies performance problems.
* Performance of `memcmp()`-aware benchmark always maxes out at around
bytes_read/sec=51.2991G/s for every type. That is 2.6x the throughput of the
naive variant!
* eltcnt/sec metric for the `memcmp()`-aware benchmark maxes out at
eltcnt/sec=27.541G/s for uint8_t (was: eltcnt/sec=1.18491G/s, so 24x) and
linearly decreases with element size.
For uint64_t, it's ~4x+ the elements/second.
* The call obvious is more pricey than the loop, with small element count.
As it can be seen from the full output {F8768210}, the `memcmp()` is almost
universally worse, independent of the element size (and thus buffer size) when
element count is less than 8.
So all in all, bcmp idiom does indeed pose untapped performance headroom.
This diff does implement said idiom recognition. I think a reasonable test
coverage is present, but do tell if there is anything obvious missing.
Now, quality. This does succeed to build and pass the test-suite, at least
without any non-bundled elements. {F8768216} {F8768217}
This transform fires 91 times:
```
$ /build/test-suite/utils/compare.py -m loop-idiom.NumBCmp result-new.json
Tests: 1149
Metric: loop-idiom.NumBCmp
Program result-new
MultiSourc...Benchmarks/7zip/7zip-benchmark 79.00
MultiSource/Applications/d/make_dparser 3.00
SingleSource/UnitTests/vla 2.00
MultiSource/Applications/Burg/burg 1.00
MultiSourc.../Applications/JM/lencod/lencod 1.00
MultiSource/Applications/lemon/lemon 1.00
MultiSource/Benchmarks/Bullet/bullet 1.00
MultiSourc...e/Benchmarks/MallocBench/gs/gs 1.00
MultiSourc...gs-C/TimberWolfMC/timberwolfmc 1.00
MultiSourc...Prolangs-C/simulator/simulator 1.00
```
The size changes are:
I'm not sure what's going on with SingleSource/UnitTests/vla.test yet, did not look.
```
$ /build/test-suite/utils/compare.py -m size..text result-{old,new}.json --filter-hash
Tests: 1149
Same hash: 907 (filtered out)
Remaining: 242
Metric: size..text
Program result-old result-new diff
test-suite...ingleSource/UnitTests/vla.test 753.00 833.00 10.6%
test-suite...marks/7zip/7zip-benchmark.test 1001697.00 966657.00 -3.5%
test-suite...ngs-C/simulator/simulator.test 32369.00 32321.00 -0.1%
test-suite...plications/d/make_dparser.test 89585.00 89505.00 -0.1%
test-suite...ce/Applications/Burg/burg.test 40817.00 40785.00 -0.1%
test-suite.../Applications/lemon/lemon.test 47281.00 47249.00 -0.1%
test-suite...TimberWolfMC/timberwolfmc.test 250065.00 250113.00 0.0%
test-suite...chmarks/MallocBench/gs/gs.test 149889.00 149873.00 -0.0%
test-suite...ications/JM/lencod/lencod.test 769585.00 769569.00 -0.0%
test-suite.../Benchmarks/Bullet/bullet.test 770049.00 770049.00 0.0%
test-suite...HMARK_ANISTROPIC_DIFFUSION/128 NaN NaN nan%
test-suite...HMARK_ANISTROPIC_DIFFUSION/256 NaN NaN nan%
test-suite...CHMARK_ANISTROPIC_DIFFUSION/64 NaN NaN nan%
test-suite...CHMARK_ANISTROPIC_DIFFUSION/32 NaN NaN nan%
test-suite...ENCHMARK_BILATERAL_FILTER/64/4 NaN NaN nan%
Geomean difference nan%
result-old result-new diff
count 1.000000e+01 10.00000 10.000000
mean 3.152090e+05 311695.40000 0.006749
std 3.790398e+05 372091.42232 0.036605
min 7.530000e+02 833.00000 -0.034981
25% 4.243300e+04 42401.00000 -0.000866
50% 1.197370e+05 119689.00000 -0.000392
75% 6.397050e+05 639705.00000 -0.000005
max 1.001697e+06 966657.00000 0.106242
```
I don't have timings though.
And now to the code. The basic idea is to completely replace the whole loop.
If we can't fully kill it, don't transform.
I have left one or two comments in the code, so hopefully it can be understood.
Also, there is a few TODO's that i have left for follow-ups:
* widening of `memcmp()`/`bcmp()`
* step smaller than the comparison size
* Metadata propagation
* more than two blocks as long as there is still a single backedge?
* ???
Reviewers: reames, fhahn, mkazantsev, chandlerc, craig.topper, courbet
Reviewed By: courbet
Subscribers: miyuki, hiraditya, xbolva00, nikic, jfb, gchatelet, courbet, llvm-commits, mclow.lists
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D61144
llvm-svn: 374662
In loop-vectorize, interleave count and vector factor depend on target register number. Currently, it does not
estimate different register pressure for different register class separately(especially for scalar type,
float type should not be on the same position with int type), so it's not accurate. Specifically,
it causes too many times interleaving/unrolling, result in too many register spills in loop body and hurting performance.
So we need classify the register classes in IR level, and importantly these are abstract register classes,
and are not the target register class of backend provided in td file. It's used to establish the mapping between
the types of IR values and the number of simultaneous live ranges to which we'd like to limit for some set of those types.
For example, POWER target, register num is special when VSX is enabled. When VSX is enabled, the number of int scalar register is 32(GPR),
float is 64(VSR), but for int and float vector register both are 64(VSR). So there should be 2 kinds of register class when vsx is enabled,
and 3 kinds of register class when VSX is NOT enabled.
It runs on POWER target, it makes big(+~30%) performance improvement in one specific bmk(503.bwaves_r) of spec2017 and no other obvious degressions.
Differential revision: https://reviews.llvm.org/D67148
llvm-svn: 374634
This patch adds a moveAfter method to VPRecipeBase, which can be used to
move elements after other elements, across VPBasicBlocks, if necessary.
Reviewers: dcaballe, hsaito, rengolin, hfinkel
Reviewed By: dcaballe
Differential Revision: https://reviews.llvm.org/D46825
llvm-svn: 374565
Currently, it is hard for the compiler to remove unused C++ virtual
functions, because they are all referenced from vtables, which are referenced
by constructors. This means that if the constructor is called from any live
code, then we keep every virtual function in the final link, even if there
are no call sites which can use it.
This patch allows unused virtual functions to be removed during LTO (and
regular compilation in limited circumstances) by using type metadata to match
virtual function call sites to the vtable slots they might load from. This
information can then be used in the global dead code elimination pass instead
of the references from vtables to virtual functions, to more accurately
determine which functions are reachable.
To make this transformation safe, I have changed clang's code-generation to
always load virtual function pointers using the llvm.type.checked.load
intrinsic, instead of regular load instructions. I originally tried writing
this using clang's existing code-generation, which uses the llvm.type.test
and llvm.assume intrinsics after doing a normal load. However, it is possible
for optimisations to obscure the relationship between the GEP, load and
llvm.type.test, causing GlobalDCE to fail to find virtual function call
sites.
The existing linkage and visibility types don't accurately describe the scope
in which a virtual call could be made which uses a given vtable. This is
wider than the visibility of the type itself, because a virtual function call
could be made using a more-visible base class. I've added a new
!vcall_visibility metadata type to represent this, described in
TypeMetadata.rst. The internalization pass and libLTO have been updated to
change this metadata when linking is performed.
This doesn't currently work with ThinLTO, because it needs to see every call
to llvm.type.checked.load in the linkage unit. It might be possible to
extend this optimisation to be able to use the ThinLTO summary, as was done
for devirtualization, but until then that combination is rejected in the
clang driver.
To test this, I've written a fuzzer which generates random C++ programs with
complex class inheritance graphs, and virtual functions called through object
and function pointers of different types. The programs are spread across
multiple translation units and DSOs to test the different visibility
restrictions.
I've also tried doing bootstrap builds of LLVM to test this. This isn't
ideal, because only classes in anonymous namespaces can be optimised with
-fvisibility=default, and some parts of LLVM (plugins and bugpoint) do not
work correctly with -fvisibility=hidden. However, there are only 12 test
failures when building with -fvisibility=hidden (and an unmodified compiler),
and this change does not cause any new failures for either value of
-fvisibility.
On the 7 C++ sub-benchmarks of SPEC2006, this gives a geomean code-size
reduction of ~6%, over a baseline compiled with "-O2 -flto
-fvisibility=hidden -fwhole-program-vtables". The best cases are reductions
of ~14% in 450.soplex and 483.xalancbmk, and there are no code size
increases.
I've also run this on a set of 8 mbed-os examples compiled for Armv7M, which
show a geomean size reduction of ~3%, again with no size increases.
I had hoped that this would have no effect on performance, which would allow
it to awlays be enabled (when using -fwhole-program-vtables). However, the
changes in clang to use the llvm.type.checked.load intrinsic are causing ~1%
performance regression in the C++ parts of SPEC2006. It should be possible to
recover some of this perf loss by teaching optimisations about the
llvm.type.checked.load intrinsic, which would make it worth turning this on
by default (though it's still dependent on -fwhole-program-vtables).
Differential revision: https://reviews.llvm.org/D63932
llvm-svn: 374539
Summary:
If we insert them from function pass some analysis may be missing or invalid.
Fixes PR42877.
Reviewers: eugenis, leonardchan
Reviewed By: leonardchan
Subscribers: hiraditya, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D68832
> llvm-svn: 374481
Signed-off-by: Vitaly Buka <vitalybuka@google.com>
llvm-svn: 374527
This is really a known bits style transformation, but known bits isn't context sensitive. The particular case which comes up happens to involve a range which allows range based reasoning to eliminate the mask pattern, so handle that case specifically in CVP.
InstCombine likes to generate the mask-by-low-bits pattern when widening an arithmetic expression which includes a zext in the middle.
Differential Revision: https://reviews.llvm.org/D68811
llvm-svn: 374506
Summary:
If we insert them from function pass some analysis may be missing or invalid.
Fixes PR42877.
Reviewers: eugenis, leonardchan
Reviewed By: leonardchan
Subscribers: hiraditya, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D68832
llvm-svn: 374481
This is just small refactoring to minimize changes in upcoming patch.
In the next path I'm going to introduce changes into heuristic for vectorization of "tiny trip count" loops.
Patch by Evgeniy Brevnov <evgueni.brevnov@gmail.com>
Reviewers: hsaito, Ayal, fhahn, reames
Reviewed By: hsaito
Differential Revision: https://reviews.llvm.org/D67690
llvm-svn: 374338
Summary:
`null` in the default address space (=AS 0) cannot be captured nor can
it alias anything. We make this clear now as it can be important for
callbacks and other cases later on. In addition, this patch improves the
debug output for noalias deduction.
Reviewers: sstefan1, uenoku
Subscribers: hiraditya, bollu, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68624
llvm-svn: 374280
in ExtBinary format
Currently for Text, Binary and ExtBinary format profiles, when we compile a
module with samplefdo, even if there is no function showing up in the profile,
we have to load all the function profiles from the profile input. That is a
waste of compile time.
CompactBinary format profile has already had the support of loading function
profiles on demand. In this patch, we add the support to load profile on
demand for ExtBinary format. It will work no matter the sections in ExtBinary
format profile are compressed or not. Experiment shows it reduces the time to
compile a server benchmark by 30%.
When profile remapping and loading function profiles on demand are both used,
extra work needs to be done so that the loading on demand process will take
the name remapping into consideration. It will be addressed in a follow-up
patch.
Differential Revision: https://reviews.llvm.org/D68601
llvm-svn: 374233
Add own version of the mathematical constants from the upcoming C++20 `std::numbers`.
Differential revision: https://reviews.llvm.org/D68257
llvm-svn: 374207
We failed to account for the target register width (max vector factor)
when vectorizing starting from GEPs. This causes vectorization to
proceed to obviously illegal widths as in:
https://bugs.llvm.org/show_bug.cgi?id=43578
For x86, this also means that SLP can produce rogue AVX or AVX512
code even when the user specifies a narrower vector width.
The AArch64 test in ext-trunc.ll appears to be better using the
narrower width. I'm not exactly sure what getelementptr.ll is trying
to do, but it's testing with "-slp-threshold=-18", so I'm not worried
about those diffs. The x86 test is an over-reduction from SPEC h264;
this patch appears to restore the perf loss caused by SLP when using
-march=haswell.
Differential Revision: https://reviews.llvm.org/D68667
llvm-svn: 374183
Summary:
The rule for the moveAllAfterMergeBlocks API si for all instructions
from `From` to have been moved to `To`, while keeping the CFG edges (and
block terminators) unchanged.
Update all the callsites for moveAllAfterMergeBlocks to follow this.
Pending follow-up: since the same behavior is needed everytime, merge
all callsites into one. The common denominator may be the call to
`MergeBlockIntoPredecessor`.
Resolves PR43569.
Reviewers: george.burgess.iv
Subscribers: Prazek, sanjoy.google, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68659
llvm-svn: 374177
When optimising for size and SCEV runtime checks need to be emitted to check
overflow behaviour, the loop vectorizer can run in this assert:
LoopVectorize.cpp:2699: void llvm::InnerLoopVectorizer::emitSCEVChecks(
llvm::Loop *, llvm::BasicBlock *): Assertion `!BB->getParent()->hasOptSize()
&& "Cannot SCEV check stride or overflow when opt
We should not generate predicates while optimising for size because
code will be generated for predicates such as these SCEV overflow runtime
checks.
This should fix PR43371.
Differential Revision: https://reviews.llvm.org/D68082
llvm-svn: 374166
David added the JamCRC implementation in r246590. More recently, Eugene
added a CRC-32 implementation in r357901, which falls back to zlib's
crc32 function if present.
These checksums are essentially the same, so having multiple
implementations seems unnecessary. This replaces the CRC-32
implementation with the simpler one from JamCRC, and implements the
JamCRC interface in terms of CRC-32 since this means it can use zlib's
implementation when available, saving a few bytes and potentially making
it faster.
JamCRC took an ArrayRef<char> argument, and CRC-32 took a StringRef.
This patch changes it to ArrayRef<uint8_t> which I think is the best
choice, and simplifies a few of the callers nicely.
Differential revision: https://reviews.llvm.org/D68570
llvm-svn: 374148
Also Revert "[LoopVectorize] Fix non-debug builds after rL374017"
This reverts commit 9f41deccc0.
This reverts commit 18b6fe07bc.
The patch is breaking PowerPC internal build, checked with author, reverting
on behalf of him for now due to timezone.
llvm-svn: 374091
Factor out CodeExtractor's analysis of allocas (for shrinkwrapping
purposes), and allow the analysis to be reused.
This resolves a quadratic compile-time bug observed when compiling
AMDGPUDisassembler.cpp.o.
Pre-patch (Release + LTO clang):
```
---User Time--- --System Time-- --User+System-- ---Wall Time--- --- Name ---
176.5278 ( 57.8%) 0.4915 ( 18.5%) 177.0192 ( 57.4%) 177.4112 ( 57.3%) Hot Cold Splitting
```
Post-patch (ReleaseAsserts clang):
```
---User Time--- --System Time-- --User+System-- ---Wall Time--- --- Name ---
1.4051 ( 3.3%) 0.0079 ( 0.3%) 1.4129 ( 3.2%) 1.4129 ( 3.2%) Hot Cold Splitting
```
Testing: check-llvm, and comparing the AMDGPUDisassembler.cpp.o binary
pre- vs. post-patch.
An alternate approach is to hide CodeExtractorAnalysisCache from clients
of CodeExtractor, and to recompute the analysis from scratch inside of
CodeExtractor::extractCodeRegion(). This eliminates some redundant work
in the shrinkwrapping legality check. However, some clients continue to
exhibit O(n^2) compile time behavior as computing the analysis is O(n).
rdar://55912966
Differential Revision: https://reviews.llvm.org/D68616
llvm-svn: 374089
Summary:
In D65186 and related patches, MustBeExecutedContextExplorer is introduced. This enables us to traverse instructions guaranteed to execute from function entry. If we can know the argument is used as `dereferenceable` or `nonnull` in these instructions, we can mark `dereferenceable` or `nonnull` in the argument definition:
1. Memory instruction (similar to D64258)
Trace memory instruction pointer operand. Currently, only inbounds GEPs are traced.
```
define i64* @f(i64* %a) {
entry:
%add.ptr = getelementptr inbounds i64, i64* %a, i64 1
; (because of inbounds GEP we can know that %a is at least dereferenceable(16))
store i64 1, i64* %add.ptr, align 8
ret i64* %add.ptr ; dereferenceable 8 (because above instruction stores into it)
}
```
2. Propagation from callsite (similar to D27855)
If `deref` or `nonnull` are known in call site parameter attributes we can also say that argument also that attribute.
```
declare void @use3(i8* %x, i8* %y, i8* %z);
declare void @use3nonnull(i8* nonnull %x, i8* nonnull %y, i8* nonnull %z);
define void @parent1(i8* %a, i8* %b, i8* %c) {
call void @use3nonnull(i8* %b, i8* %c, i8* %a)
; Above instruction is always executed so we can say that@parent1(i8* nonnnull %a, i8* nonnull %b, i8* nonnull %c)
call void @use3(i8* %c, i8* %a, i8* %b)
ret void
}
```
Reviewers: jdoerfert, sstefan1, spatel, reames
Reviewed By: jdoerfert
Subscribers: xbolva00, hiraditya, jfb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D65402
llvm-svn: 374063
Summary: This patch introduces a generic way to compose two structured deductions. This will be used for composing generic deduction with `MustBeExecutedExplorer` and other existing generic deduction.
Reviewers: jdoerfert, sstefan1
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D66645
llvm-svn: 374060
* Adds a TypeSize struct to represent the known minimum size of a type
along with a flag to indicate that the runtime size is a integer multiple
of that size
* Converts existing size query functions from Type.h and DataLayout.h to
return a TypeSize result
* Adds convenience methods (including a transparent conversion operator
to uint64_t) so that most existing code 'just works' as if the return
values were still scalars.
* Uses the new size queries along with ElementCount to ensure that all
supported instructions used with scalable vectors can be constructed
in IR.
Reviewers: hfinkel, lattner, rkruppe, greened, rovka, rengolin, sdesmalen
Reviewed By: rovka, sdesmalen
Differential Revision: https://reviews.llvm.org/D53137
llvm-svn: 374042
Summary: LoopRotate is a loop pass and SE should always be available.
Reviewers: anemet, asbirlea
Reviewed By: asbirlea
Differential Revision: https://reviews.llvm.org/D68573
llvm-svn: 374026
In loop-vectorize, interleave count and vector factor depend on target register number. Currently, it does not
estimate different register pressure for different register class separately(especially for scalar type,
float type should not be on the same position with int type), so it's not accurate. Specifically,
it causes too many times interleaving/unrolling, result in too many register spills in loop body and hurting performance.
So we need classify the register classes in IR level, and importantly these are abstract register classes,
and are not the target register class of backend provided in td file. It's used to establish the mapping between
the types of IR values and the number of simultaneous live ranges to which we'd like to limit for some set of those types.
For example, POWER target, register num is special when VSX is enabled. When VSX is enabled, the number of int scalar register is 32(GPR),
float is 64(VSR), but for int and float vector register both are 64(VSR). So there should be 2 kinds of register class when vsx is enabled,
and 3 kinds of register class when VSX is NOT enabled.
It runs on POWER target, it makes big(+~30%) performance improvement in one specific bmk(503.bwaves_r) of spec2017 and no other obvious degressions.
Differential revision: https://reviews.llvm.org/D67148
llvm-svn: 374017
The initialization logic has become part of the Attributor but the
patches that introduced these calls here were in development when the
transition happened.
We also now clean up (undefine) the macros used to create attributes.
llvm-svn: 373987
Local linkage is internal or private, and private is a specialization of
internal, so either is fine for all our "local linkage" queries.
llvm-svn: 373986
Summary:
When we iterate over uses of functions and expect them to be call sites,
we now use abstract call sites to allow callback calls.
Reviewers: sstefan1, uenoku
Subscribers: hiraditya, bollu, hfinkel, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D67871
llvm-svn: 373985
This can come up in Bit Stream abstractions.
The pattern looks big/scary, but it can't be simplified any further.
It only is so simple because a number of my preparatory folds had
happened already (shift amount reassociation / shift amount
reassociation in bit test, sign bit test detection).
Highlights:
* There are two main flavors: https://rise4fun.com/Alive/zWi
The difference is add vs. sub, and left-shift of -1 vs. 1
* Since we only change the shift opcode,
we can preserve the exact-ness: https://rise4fun.com/Alive/4u4
* There can be truncation after high-bit-extraction:
https://rise4fun.com/Alive/slHc1 (the main pattern i'm after!)
Which means that we need to ignore zext of shift amounts and of NBits.
* The sign-extending magic can be extended itself (in add pattern
via sext, in sub pattern via zext. not the other way around!)
https://rise4fun.com/Alive/NhG
(or those sext/zext can be sinked into `select`!)
Which again means we should pay attention when matching NBits.
* We can have both truncation of extraction and widening of magic:
https://rise4fun.com/Alive/XTw
In other words, i don't believe we need to have any checks on
bitwidths of any of these constructs.
This is worsened in general by the fact that we may have `sext` instead
of `zext` for shift amounts, and we don't yet canonicalize to `zext`,
although we should. I have not done anything about that here.
Also, we really should have something to weed out `sub` like these,
by folding them into `add` variant.
https://bugs.llvm.org/show_bug.cgi?id=42389
llvm-svn: 373964
True, no test coverage is being added here. But those non-canonical
predicates that are already handled here already have no test coverage
as far as i can tell. I tried to add tests for them, but all the patterns
already get handled elsewhere.
llvm-svn: 373962
Summary:
Currently, we pre-check whether we need to produce a mask or not.
This involves some rather magical constants.
I'd like to extend this fold to also handle the situation
when there's also a `trunc` before outer shift.
That will require another set of magical constants.
It's ugly.
Instead, we can just compute the mask, and check
whether mask is a pass-through (all-ones) or not.
This way we don't need to have any magical numbers.
This change is NFC other than the fact that we now compute
the mask and then check if we need (and can!) apply it.
Reviewers: spatel
Reviewed By: spatel
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68470
llvm-svn: 373961
Summary:
When we do `ConstantExpr::getZExt()`, that "extends" `undef` to `0`,
which means that for patterns a/b we'd assume that we must not produce
any bits for that channel, while in reality we simply didn't care
about that channel - i.e. we don't need to mask it.
Reviewers: spatel
Reviewed By: spatel
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68239
llvm-svn: 373960
Doing this makes MSVC complain that `empty(someRange)` could refer to
either C++17's std::empty or LLVM's llvm::empty, which previously we
avoided via SFINAE because std::empty is defined in terms of an empty
member rather than begin and end. So, switch callers over to the new
method as it is added.
https://reviews.llvm.org/D68439
llvm-svn: 373935
This reverts SVN r373833, as it caused a failed assert "Non-zero loop
cost expected" on building numerous projects, see PR43582 for details
and reproduction samples.
llvm-svn: 373882
I don't see an ideal solution to these 2 related, potentially large, perf regressions:
https://bugs.llvm.org/show_bug.cgi?id=42708https://bugs.llvm.org/show_bug.cgi?id=43146
We decided that load combining was unsuitable for IR because it could obscure other
optimizations in IR. So we removed the LoadCombiner pass and deferred to the backend.
Therefore, preventing SLP from destroying load combine opportunities requires that it
recognizes patterns that could be combined later, but not do the optimization itself (
it's not a vector combine anyway, so it's probably out-of-scope for SLP).
Here, we add a scalar cost model adjustment with a conservative pattern match and cost
summation for a multi-instruction sequence that can probably be reduced later.
This should prevent SLP from creating a vector reduction unless that sequence is
extremely cheap.
In the x86 tests shown (and discussed in more detail in the bug reports), SDAG combining
will produce a single instruction on these tests like:
movbe rax, qword ptr [rdi]
or:
mov rax, qword ptr [rdi]
Not some (half) vector monstrosity as we currently do using SLP:
vpmovzxbq ymm0, dword ptr [rdi + 1] # ymm0 = mem[0],zero,zero,..
vpsllvq ymm0, ymm0, ymmword ptr [rip + .LCPI0_0]
movzx eax, byte ptr [rdi]
movzx ecx, byte ptr [rdi + 5]
shl rcx, 40
movzx edx, byte ptr [rdi + 6]
shl rdx, 48
or rdx, rcx
movzx ecx, byte ptr [rdi + 7]
shl rcx, 56
or rcx, rdx
or rcx, rax
vextracti128 xmm1, ymm0, 1
vpor xmm0, xmm0, xmm1
vpshufd xmm1, xmm0, 78 # xmm1 = xmm0[2,3,0,1]
vpor xmm0, xmm0, xmm1
vmovq rax, xmm0
or rax, rcx
vzeroupper
ret
Differential Revision: https://reviews.llvm.org/D67841
llvm-svn: 373833
David Green