This case is interesting because ScalarEvolutionExpander lowers min(a,
b) as ~max(~a,~b). I think the profitability heuristics can be made
more clever/aggressive, but this is a start.
Differential Revision: http://reviews.llvm.org/D7821
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230285 91177308-0d34-0410-b5e6-96231b3b80d8
When emitting the increment operation, SCEVExpander marks the
operation as nuw or nsw based on the flags on the preincrement SCEV.
This is incorrect because, for instance, it is possible that {-6,+,1}
is <nuw> while {-6,+,1}+1 = {-5,+,1} is not.
This change teaches SCEV to mark the increment as nuw/nsw only if it
can explicitly prove that the increment operation won't overflow.
Apart from the attached test case, another (more realistic) manifestation
of the bug can be seen in Transforms/IndVarSimplify/pr20680.ll.
NOTE: this change was landed with an incorrect commit message in
rL230275 and was reverted for that reason in rL230279. This commit
message is the correct one.
Differential Revision: http://reviews.llvm.org/D7778
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230280 91177308-0d34-0410-b5e6-96231b3b80d8
230275 got committed with an incorrect commit message due to a mixup
on my side. Will re-land in a few moments with the correct commit
message.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230279 91177308-0d34-0410-b5e6-96231b3b80d8
The bug was a result of getPreStartForExtend interpreting nsw/nuw
flags on an add recurrence more strongly than is legal. {S,+,X}<nsw>
implies S+X is nsw only if the backedge of the loop is taken at least
once.
Differential Revision: http://reviews.llvm.org/D7808
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230275 91177308-0d34-0410-b5e6-96231b3b80d8
This patch adds the isProfitableToHoist API. For AArch64, we want to prevent a
fmul from being hoisted in cases where it is more profitable to form a
fmsub/fmadd.
Phabricator Review: http://reviews.llvm.org/D7299
Patch by Lawrence Hu <lawrence@codeaurora.org>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230241 91177308-0d34-0410-b5e6-96231b3b80d8
calculations. Semantically non-functional change.
This gets rid of some of the SCEV -> Value -> SCEV round tripping and
the Construct(SMin|SMax)Of and MaybeSimplify helper routines.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230150 91177308-0d34-0410-b5e6-96231b3b80d8
Previously, this pass ran over every function in the Module if added to the pass order. With this change, it runs only over those with a GC attribute where the GC explicitly opts in. A GC can also choose which of entry safepoint polls, backedge safepoint polls, and call safepoints it wants. I hope to get these exposed as checks on the GCStrategy at some point, but for now, the checks are manual string comparisons.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230097 91177308-0d34-0410-b5e6-96231b3b80d8
Yet another chapter in the endless story. While this looks like we leave
the loop in a non-canonical state this replicates the logic in
LoopSimplify so it doesn't diverge from the canonical form in any way.
PR21968
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230058 91177308-0d34-0410-b5e6-96231b3b80d8
This patch introduces a new mechanism that allows IR modules to co-operatively
build pointer sets corresponding to addresses within a given set of
globals. One particular use case for this is to allow a C++ program to
efficiently verify (at each call site) that a vtable pointer is in the set
of valid vtable pointers for the class or its derived classes. One way of
doing this is for a toolchain component to build, for each class, a bit set
that maps to the memory region allocated for the vtables, such that each 1
bit in the bit set maps to a valid vtable for that class, and lay out the
vtables next to each other, to minimize the total size of the bit sets.
The patch introduces a metadata format for representing pointer sets, an
'@llvm.bitset.test' intrinsic and an LTO lowering pass that lays out the globals
and builds the bitsets, and documents the new feature.
Differential Revision: http://reviews.llvm.org/D7288
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230054 91177308-0d34-0410-b5e6-96231b3b80d8
Before calling Function::getGC to test for enablement, we need to make sure there's actually a GC at all via Function::hasGC. Otherwise, we'd crash on functions without a GC. Thankfully, this only mattered if you manually scheduled the pass, but still, oops. :(
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230040 91177308-0d34-0410-b5e6-96231b3b80d8
This change addresses a deficiency pointed out in PR22629. To copy from the bug
report:
[from the bug report]
Consider this code:
int f(int x) {
int a[] = {12};
return a[x];
}
GCC knows to optimize this to
movl $12, %eax
ret
The code generated by recent Clang at -O3 is:
movslq %edi, %rax
movl .L_ZZ1fiE1a(,%rax,4), %eax
retq
.L_ZZ1fiE1a:
.long 12 # 0xc
[end from the bug report]
This definitely seems worth fixing. I've also seen this kind of code before (as
the base case of generic vector wrapper templates with one element).
The general idea is to look at the GEP feeding a load or a store, which has
some variable as its first non-zero index, and determine if that index must be
zero (or else an out-of-bounds access would occur). We can do this for allocas
and globals with constant initializers where we know the maximum size of the
underlying object. When we find such a GEP, we create a new one for the memory
access with that first variable index replaced with a constant zero.
Even if we can't eliminate the memory access (and sometimes we can't), it is
still useful because it removes unnecessary indexing calculations.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229959 91177308-0d34-0410-b5e6-96231b3b80d8
When back merging the changes in 229945 I noticed that I forgot to mark the test cases with the appropriate GC. We want the rewriting to be off by default (even when manually added to the pass order), not on-by default. To keep the current test working, mark them as using the statepoint-example GC and whitelist that GC.
Longer term, we need a better selection mechanism here for both actual usage and testing. As I migrate more tests to the in tree version of this pass, I will probably need to update the enable/disable logic as well.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229954 91177308-0d34-0410-b5e6-96231b3b80d8
This patch consists of a single pass whose only purpose is to visit previous inserted gc.statepoints which do not have gc.relocates inserted yet, and insert them. This can be used either immediately after IR generation to perform 'early safepoint insertion' or late in the pass order to perform 'late insertion'.
This patch is setting the stage for work to continue in tree. In particular, there are known naming and style violations in the current patch. I'll try to get those resolved over the next week or so. As I touch each area to make style changes, I need to make sure we have adequate testing in place. As part of the cleanup, I will be cleaning up a collection of test cases we have out of tree and submitting them upstream. The tests included in this change are very basic and mostly to provide examples of usage.
The pass has several main subproblems it needs to address:
- First, it has identify any live pointers. In the current code, the use of address spaces to distinguish pointers to GC managed objects is hard coded, but this will become parametrizable in the near future. Note that the current change doesn't actually contain a useful liveness analysis. It was seperated into a followup change as the code wasn't ready to be shared. Instead, the current implementation just considers any dominating def of appropriate pointer type to be live.
- Second, it has to identify base pointers for each live pointer. This is a fairly straight forward data flow algorithm.
- Third, the information in the previous steps is used to actually introduce rewrites. Rather than trying to do this by hand, we simply re-purpose the code behind Mem2Reg to do this for us.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229945 91177308-0d34-0410-b5e6-96231b3b80d8
This re-applies r223862, r224198, r224203, and r224754, which were
reverted in r228129 because they exposed Clang misalignment problems
when self-hosting.
The combine caused the crashes because we turned ISD::LOAD/STORE nodes
to ARMISD::VLD1/VST1_UPD nodes. When selecting addressing modes, we
were very lax for the former, and only emitted the alignment operand
(as in "[r1:128]") when it was larger than the standard alignment of
the memory type.
However, for ARMISD nodes, we just used the MMO alignment, no matter
what. In our case, we turned ISD nodes to ARMISD nodes, and this
caused the alignment operands to start being emitted.
And that's how we exposed alignment problems that were ignored before
(but I believe would have been caught with SCTRL.A==1?).
To fix this, we can just mirror the hack done for ISD nodes: only
take into account the MMO alignment when the access is overaligned.
Original commit message:
We used to only combine intrinsics, and turn them into VLD1_UPD/VST1_UPD
when the base pointer is incremented after the load/store.
We can do the same thing for generic load/stores.
Note that we can only combine the first load/store+adds pair in
a sequence (as might be generated for a v16f32 load for instance),
because other combines turn the base pointer addition chain (each
computing the address of the next load, from the address of the last
load) into independent additions (common base pointer + this load's
offset).
rdar://19717869, rdar://14062261.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229932 91177308-0d34-0410-b5e6-96231b3b80d8
First, don't combine bit masking into vector shuffles (even ones the
target can handle) once operation legalization has taken place. Custom
legalization of vector shuffles may exist for these patterns (making the
predicate return true) but that custom legalization may in some cases
produce the exact bit math this matches. We only really want to handle
this prior to operation legalization.
However, the x86 backend, in a fit of awesome, relied on this. What it
would do is mark VSELECTs as expand, which would turn them into
arithmetic, which this would then match back into vector shuffles, which
we would then lower properly. Amazing.
Instead, the second change is to teach the x86 backend to directly form
vector shuffles from VSELECT nodes with constant conditions, and to mark
all of the vector types we support lowering blends as shuffles as custom
VSELECT lowering. We still mark the forms which actually support
variable blends as *legal* so that the custom lowering is bypassed, and
the legal lowering can even be used by the vector shuffle legalization
(yes, i know, this is confusing. but that's how the patterns are
written).
This makes the VSELECT lowering much more sensible, and in fact should
fix a bunch of bugs with it. However, as you'll see in the test cases,
right now what it does is point out the *hilarious* deficiency of the
new vector shuffle lowering when it comes to blends. Fortunately, my
very next patch fixes that. I can't submit it yet, because that patch,
somewhat obviously, forms the exact and/or pattern that the DAG combine
is matching here! Without this patch, teaching the vector shuffle
lowering to produce the right code infloops in the DAG combiner. With
this patch alone, we produce terrible code but at least lower through
the right paths. With both patches, all the regressions here should be
fixed, and a bunch of the improvements (like using 2 shufps with no
memory loads instead of 2 andps with memory loads and an orps) will
stay. Win!
There is one other change worth noting here. We had hilariously wrong
vectorization cost estimates for vselect because we fell through to the
code path that assumed all "expand" vector operations are scalarized.
However, the "expand" lowering of VSELECT is vector bit math, most
definitely not scalarized. So now we go back to the correct if horribly
naive cost of "1" for "not scalarized". If anyone wants to add actual
modeling of shuffle costs, that would be cool, but this seems an
improvement on its own. Note the removal of 16 and 32 "costs" for doing
a blend. Even in SSE2 we can blend in fewer than 16 instructions. ;] Of
course, we don't right now because of OMG bad code, but I'm going to fix
that. Next patch. I promise.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229835 91177308-0d34-0410-b5e6-96231b3b80d8
Don't spend the entire iteration space in the scalar loop prologue if
computing the trip count overflows. This change also gets rid of the
backedge check in the prologue loop and the extra check for
overflowing trip-count.
Differential Revision: http://reviews.llvm.org/D7715
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229731 91177308-0d34-0410-b5e6-96231b3b80d8
See the comment in the code.
This is part of the patchset that converts LoopAccessAnalysis into an
actual analysis pass.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229627 91177308-0d34-0410-b5e6-96231b3b80d8
InstCombiner::visitGetElementPtrInst was using getFirstNonPHI to compute the
insertion point, which caused the verifier to complain when a GEP was inserted
before a landingpad instruction. This commit fixes it to use getFirstInsertionPt
instead.
rdar://problem/19394964
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229619 91177308-0d34-0410-b5e6-96231b3b80d8
When visiting the initial list of "root" instructions (those which must always
be alive), for those that are integer-valued (such as invokes returning an
integer), we mark their bits as (initially) all dead (we might, obviously, find
uses of those bits later, but all bits are assumed dead until proven
otherwise). Don't do so, however, if we're already seen a use of those bits by
another root instruction (such as a store).
Fixes a miscompile of the sanitizer unit tests on x86_64.
Also, add a debug line for visiting the root instructions, and remove a debug
line which tried to print instructions being removed (printing dead
instructions is dangerous, and can sometimes crash).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229618 91177308-0d34-0410-b5e6-96231b3b80d8
BDCE is a bit-tracking dead code elimination pass. It is based on ADCE (the
"aggressive DCE" pass), with the added capability to track dead bits of integer
valued instructions and remove those instructions when all of the bits are
dead.
Currently, it does not actually do this all-bits-dead removal, but rather
replaces the instruction's uses with a constant zero, and lets instcombine (and
the later run of ADCE) do the rest. Because we essentially get a run of ADCE
"for free" while tracking the dead bits, we also do what ADCE does and removes
actually-dead instructions as well (this includes instructions newly trivially
dead because all bits were dead, but not all such instructions can be removed).
The motivation for this is a case like:
int __attribute__((const)) foo(int i);
int bar(int x) {
x |= (4 & foo(5));
x |= (8 & foo(3));
x |= (16 & foo(2));
x |= (32 & foo(1));
x |= (64 & foo(0));
x |= (128& foo(4));
return x >> 4;
}
As it turns out, if you order the bit-field insertions so that all of the dead
ones come last, then instcombine will remove them. However, if you pick some
other order (such as the one above), the fact that some of the calls to foo()
are useless is not locally obvious, and we don't remove them (without this
pass).
I did a quick compile-time overhead check using sqlite from the test suite
(Release+Asserts). BDCE took ~0.4% of the compilation time (making it about
twice as expensive as ADCE).
I've not looked at why yet, but we eliminate instructions due to having
all-dead bits in:
External/SPEC/CFP2006/447.dealII/447.dealII
External/SPEC/CINT2006/400.perlbench/400.perlbench
External/SPEC/CINT2006/403.gcc/403.gcc
MultiSource/Applications/ClamAV/clamscan
MultiSource/Benchmarks/7zip/7zip-benchmark
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229462 91177308-0d34-0410-b5e6-96231b3b80d8
We won't find a root with index zero in any loop that we are able to reroll.
However, we may find one in a non-rerollable loop, so bail gracefully instead
of failing hard.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229406 91177308-0d34-0410-b5e6-96231b3b80d8
If a PHI has no users, don't crash; bail gracefully. This shouldn't
happen often, but we can make no guarantees that previous passes didn't leave
dead code around.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229405 91177308-0d34-0410-b5e6-96231b3b80d8
We didn't properly handle the out-of-bounds case for
ConstantAggregateZero and UndefValue. This would manifest as a crash
when the constant folder was asked to fold a load of a constant global
whose struct type has no operands.
This fixes PR22595.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229352 91177308-0d34-0410-b5e6-96231b3b80d8
The "dereferenceable" attribute cannot be added via .addAttribute(),
since it also expects a size in bytes. AttrBuilder#addAttribute or
AttributeSet#addAttribute is wrapped by classes Function, InvokeInst,
and CallInst. Add corresponding wrappers to
AttrBuilder#addDereferenceableAttr.
Having done this, propagate the dereferenceable attribute via
gc.relocate, adding a test to exercise it. Note that -datalayout is
required during execution over and above -instcombine, because
InstCombine only optionally requires DataLayoutPass.
Differential Revision: http://reviews.llvm.org/D7510
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229265 91177308-0d34-0410-b5e6-96231b3b80d8
If we know that the sign bit of a value being sign extended is zero, we can use a zero extension instead. This is motivated by the fact that zero extensions are generally cheaper on x86 (and most other architectures?). We already apply a similar transform in DAGCombine, this just extends that to the IR level.
This comes up when we eagerly canonicalize gep indices to the width of a machine register (i64 on x86_64). To do so, we insert sign extensions (sext) to promote smaller types.
Differential Revision: http://reviews.llvm.org/D7255
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229189 91177308-0d34-0410-b5e6-96231b3b80d8
This patch fixes a problem I accidentally introduced in an instruction combine
on select instructions added at r227197. That revision taught the instruction
combiner how to fold a cttz/ctlz followed by a icmp plus select into a single
cttz/ctlz with flag 'is_zero_undef' cleared.
However, the new rule added at r227197 would have produced wrong results in the
case where a cttz/ctlz with flag 'is_zero_undef' cleared was follwed by a
zero-extend or truncate. In that case, the folded instruction would have
been inserted in a wrong location thus leaving the CFG in an inconsistent
state.
This patch fixes the problem and add two reproducible test cases to
existing test 'InstCombine/select-cmp-cttz-ctlz.ll'.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229124 91177308-0d34-0410-b5e6-96231b3b80d8
SimplifyCFG now knows how to speculate calls to intrinsic cttz/ctlz that are
'cheap' for the target. Therefore, some of the logic in CodeGenPrepare
that was originally added at revision 224899 can now be removed.
This patch is basically a no functional change. It removes the duplicated
logic in CodeGenPrepare and converts all the existing target specific tests
for cttz/ctlz into SimplifyCFG tests.
Differential Revision: http://reviews.llvm.org/D7608
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229105 91177308-0d34-0410-b5e6-96231b3b80d8
The issues with the new unroll analyzer are more fundamental than code
cleanup, algorithm, or data structure changes. I've sent an email to the
original commit thread with details and a proposal for how to redesign
things. I'm disabling this for now so that we don't spend time
debugging issues with it in its current state.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229064 91177308-0d34-0410-b5e6-96231b3b80d8
- First, there's a crash when we try to combine that pointers into `icmp`
directly by creating a `bitcast`, which is invalid if that two pointers are
from different address spaces.
- It's not always appropriate to cast one pointer to another if they are from
different address spaces as that is not no-op cast. Instead, we only combine
`icmp` from `ptrtoint` if that two pointers are of the same address space.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229063 91177308-0d34-0410-b5e6-96231b3b80d8
propagating of metadata.
We were propagating !nonnull metadata even when the newly formed load is
no longer of a pointer type. This is clearly broken and results in LLVM
failing the verifier and aborting. This patch just restricts the
propagation of !nonnull metadata to when we actually have a pointer
type.
This bug report and the initial version of this patch was provided by
Charles Davis! Many thanks for finding this!
We still need to add logic to round-trip the metadata correctly if we
combine from pointer types to integer types and then back by using range
metadata for the integer type loads. But this is the minimal and safe
version of the patch, which is important so we can backport it into 3.6.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229029 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
Instances of the AssumptionCache are per function, so we can't re-use
the same AssumptionCache instance when recursing in the CallAnalyzer to
analyze a different function. Instead we have to pass the
AssumptionCacheTracker to the CallAnalyzer so it can get the right
AssumptionCache on demand.
Reviewers: hfinkel
Subscribers: llvm-commits, hans
Differential Revision: http://reviews.llvm.org/D7533
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@228957 91177308-0d34-0410-b5e6-96231b3b80d8