On some occasions ReuseOrCreateCast may convert previously
expanded value to undefined. That value may be passed by
SCEVExpander as an argument to InsertBinop making IV chain
undefined.
Differential revision: https://reviews.llvm.org/D63928
llvm-svn: 365009
This reverts r364422 (git commit 1a3dc76186)
The inlining cost calculation is incorrect, leading to stack overflow due to large stack frames from heavy inlining.
llvm-svn: 365000
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 an example).
Reviewers: RKSimon, ABataev, dtemirbulatov, Ayal, hfinkel, rnk
Reviewed By: RKSimon, dtemirbulatov
Subscribers: hiraditya, phosek, rnk, rcorcs, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D60897
llvm-svn: 364964
I was actually wondering if there was some nicer way than m_Value()+cast,
but apparently what i was really "subconsciously" thinking about
was correctness issue.
hasNoUnsignedWrap()/hasNoUnsignedWrap() exist for Instruction,
not for BinaryOperator, so let's just use m_Instruction(),
thus both avoiding a cast, and a crash.
Fixes https://bugs.llvm.org/show_bug.cgi?id=42484,
https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=15587
llvm-svn: 364915
Use both one bit and signbit shifting to check for one bit merge.
Reviewers: lebedev.ri, spatel, efriedma, craig.topper
Reviewed By: lebedev.ri
Subscribers: llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D63903
llvm-svn: 364857
Extends the transform from:
rL364341
...to include another (more common?) pattern that tests whether a
value is a power-of-2 (including or excluding zero).
llvm-svn: 364856
Summary:
To be noted, this pattern is not unhandled by instcombine per-se,
it is somehow does end up being folded when one runs opt -O3,
but not if it's just -instcombine. Regardless, that fold is
indirect, depends on some other folds, and is thus blind
when there are extra uses.
This does address the regression being exposed in D63992.
https://godbolt.org/z/7DGltUhttps://rise4fun.com/Alive/EPO0
Fixes [[ https://bugs.llvm.org/show_bug.cgi?id=42459 | PR42459 ]]
Reviewers: spatel, nikic, huihuiz
Reviewed By: spatel
Subscribers: llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D63993
llvm-svn: 364792
Summary:
Given pattern:
`icmp eq/ne (and ((x shift Q), (y oppositeshift K))), 0`
we should move shifts to the same hand of 'and', i.e. rewrite as
`icmp eq/ne (and (x shift (Q+K)), y), 0` iff `(Q+K) u< bitwidth(x)`
It might be tempting to not restrict this to situations where we know
we'd fold two shifts together, but i'm not sure what rules should there be
to avoid endless combine loops.
We pick the same shift that was originally used to shift the variable we picked to shift:
https://rise4fun.com/Alive/6x1v
Should fix [[ https://bugs.llvm.org/show_bug.cgi?id=42399 | PR42399]].
Reviewers: spatel, nikic, RKSimon
Reviewed By: spatel
Subscribers: llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D63829
llvm-svn: 364791
As discussed in https://reviews.llvm.org/D63829
*if* *both* shifts are one-use, we'd most likely want to produce `lshr`,
and not rely on ordering.
Also, there should likely be a *separate* fold to do this reordering.
llvm-svn: 364772
To be noted, this pattern is not unhandled by instcombine per-se,
it is somehow does end up being folded when one runs opt -O3,
but not if it's just -instcombine. Regardless, that fold is
indirect, depends on some other folds, and is thus blind
when there are extra uses.
https://bugs.llvm.org/show_bug.cgi?id=42459https://rise4fun.com/Alive/EPO0
llvm-svn: 364749
Backend changes to enable WLS/LE low-overhead loops for armv8.1-m:
1) Use TTI to communicate to the HardwareLoop pass that we should try
to generate intrinsics that guard the loop entry, as well as setting
the loop trip count.
2) Lower the BRCOND that uses said intrinsic to an Arm specific node:
ARMWLS.
3) ISelDAGToDAG the node to a new pseudo instruction:
t2WhileLoopStart.
4) Add support in ArmLowOverheadLoops to handle the new pseudo
instruction.
Differential Revision: https://reviews.llvm.org/D63816
llvm-svn: 364733
This is the opposite direction of D62158 (we have to choose 1 form or the other).
Now that we have FMF on the select, this becomes more palatable. And the benefits
of having a single IR instruction for this operation (less chances of missing folds
based on extra uses, etc) overcome my previous comments about the potential advantage
of larger pattern matching/analysis.
Differential Revision: https://reviews.llvm.org/D62414
llvm-svn: 364721
This transform came up in D62414, but we should deal with it first.
We have LLVM intrinsics that correspond exactly to libm calls (unlike
most libm calls, these libm calls never set errno).
This holds without any fast-math-flags, so we should always canonicalize
to those intrinsics directly for better optimization.
Currently, we convert to fcmp+select only when we have FMF (nnan) because
fcmp+select does not preserve the semantics of the call in the general case.
Differential Revision: https://reviews.llvm.org/D63214
llvm-svn: 364714
Fixes https://bugs.llvm.org/show_bug.cgi?id=41998. Usually when we
have a truncated exit count we'll truncate the IV when comparing
against the limit, in which case exit count overflow in post-inc
form doesn't matter. However, for pointer IVs we don't do that, so
we have to be careful about incrementing the IV in the wide type.
I'm fixing this by removing the IVCount variable (which was
ExitCount or ExitCount+1) and replacing it with a UsePostInc flag,
and then moving the actual limit adjustment to the individual cases
(which are: pointer IV where we add to the wide type, integer IV
where we add to the narrow type, and constant integer IV where we
add to the wide type).
Differential Revision: https://reviews.llvm.org/D63686
llvm-svn: 364709
Introduce llvm.test.set.loop.iterations which sets the loop counter
and also produces an i1 after testing that the count is not zero.
Differential Revision: https://reviews.llvm.org/D63809
llvm-svn: 364628
This patch introduces a new function attribute, willreturn, to indicate
that a call of this function will either exhibit undefined behavior or
comes back and continues execution at a point in the existing call stack
that includes the current invocation.
This attribute guarantees that the function does not have any endless
loops, endless recursion, or terminating functions like abort or exit.
Patch by Hideto Ueno (@uenoku)
Reviewers: jdoerfert
Subscribers: mehdi_amini, hiraditya, steven_wu, dexonsmith, lebedev.ri, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D62801
llvm-svn: 364555
FunctionComparator attempts to produce a stable comparison of two Function
instances by looking at all available properties. Since ByVal attributes now
contain a Type pointer, they are not trivially ordered and FunctionComparator
should use its own Type comparison logic to sort them.
llvm-svn: 364523
For fold
(X & (signbit l>> Y)) ==/!= 0 -> (X << Y) >=/< 0
(X & (signbit << Y)) ==/!= 0 -> (X l>> Y) >=/< 0
Test cases of X being constant are positive tests not negative.
Prep work for D62818.
llvm-svn: 364497
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 an example).
Reviewers: RKSimon, ABataev, dtemirbulatov, Ayal, hfinkel, rnk
Reviewed By: RKSimon, dtemirbulatov
Subscribers: rnk, rcorcs, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D60897
llvm-svn: 364478
I don't think there was anything going wrong here,
but the auto-generating CHECK line script is known
to have problems with 'TMP' because it uses that
to match nameless values.
This is a retry of rL364452.
llvm-svn: 364477
I don't think there was anything going wrong here,
but the auto-generating CHECK line script is known
to have problems with 'TMP' because it uses that
to match nameless values.
llvm-svn: 364452
Summary:
Doing better separation of Cost and Threshold.
Cost counts the abstract complexity of live instructions, while Threshold is an upper bound of complexity that inlining is comfortable to pay.
There are two parts:
- huge 15K last-call-to-static bonus is no longer subtracted from Cost
but rather is now added to Threshold.
That makes much more sense, as the cost of inlining (Cost) is not changed by the fact
that internal function is called once. It only changes the likelyhood of this inlining
being profitable (Threshold).
- bonus for calls proved-to-be-inlinable into callee is no longer subtracted from Cost
but added to Threshold instead.
While calculations are somewhat different, overall InlineResult should stay the same since Cost >= Threshold compares the same.
Reviewers: eraman, greened, chandlerc, yrouban, apilipenko
Reviewed By: apilipenko
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D60740
llvm-svn: 364422
This allows later passes (in particular InstCombine) to optimize more
cases.
One that's important to us is `memcmp(p, q, constant) < 0` and memcmp(p, q, constant) > 0.
llvm-svn: 364412
This patch generalizes the UnrollLoop utility to support loops that exit
from the header instead of the latch. Usually, LoopRotate would take care
of must of those cases, but in some cases (e.g. -Oz), LoopRotate does
not kick in.
Codesize impact looks relatively neutral on ARM64 with -Oz + LTO.
Program master patch diff
External/S.../CFP2006/447.dealII/447.dealII 629060.00 627676.00 -0.2%
External/SPEC/CINT2000/176.gcc/176.gcc 1245916.00 1244932.00 -0.1%
MultiSourc...Prolangs-C/simulator/simulator 86100.00 86156.00 0.1%
MultiSourc...arks/Rodinia/backprop/backprop 66212.00 66252.00 0.1%
MultiSourc...chmarks/Prolangs-C++/life/life 67276.00 67312.00 0.1%
MultiSourc...s/Prolangs-C/compiler/compiler 69824.00 69788.00 -0.1%
MultiSourc...Prolangs-C/assembler/assembler 86672.00 86696.00 0.0%
Reviewers: efriedma, vsk, paquette
Reviewed By: paquette
Differential Revision: https://reviews.llvm.org/D61962
llvm-svn: 364398
This follows up the transform from rL363956 to use the ctpop intrinsic when checking for power-of-2-or-zero.
This is matching the isPowerOf2() patterns used in PR42314:
https://bugs.llvm.org/show_bug.cgi?id=42314
But there's at least 1 instcombine follow-up needed to match the alternate form:
(v & (v - 1)) == 0;
We should have all of the backend expansions handled with:
rL364319
(x86-specific changes still needed for optimal code based on subtarget)
And the larger patterns to exclude zero as a power-of-2 are joining with this change after:
rL364153 ( D63660 )
rL364246
Differential Revision: https://reviews.llvm.org/D63777
llvm-svn: 364341
This provides the low-level support to start using MVE vector types in
LLVM IR, loading and storing them, passing them to __asm__ statements
containing hand-written MVE vector instructions, and *if* you have the
hard-float ABI turned on, using them as function parameters.
(In the soft-float ABI, vector types are passed in integer registers,
and combining all those 32-bit integers into a q-reg requires support
for selection DAG nodes like insert_vector_elt and build_vector which
aren't implemented yet for MVE. In fact I've also had to add
`arm_aapcs_vfpcc` to a couple of existing tests to avoid that
problem.)
Specifically, this commit adds support for:
* spills, reloads and register moves for MVE vector registers
* ditto for the VPT predication mask that lives in VPR.P0
* make all the MVE vector types legal in ISel, and provide selection
DAG patterns for BITCAST, LOAD and STORE
* make loads and stores of scalar FP types conditional on
`hasFPRegs()` rather than `hasVFP2Base()`. As a result a few
existing tests needed their llc command lines updating to use
`-mattr=-fpregs` as their method of turning off all hardware FP
support.
Reviewers: dmgreen, samparker, SjoerdMeijer
Subscribers: javed.absar, kristof.beyls, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D60708
llvm-svn: 364329
The expensive buildbots highlighted the mir tests were broken, which
I've now updated and added --verify-machineinstrs to them. This also
uncovered a couple of bugs in the backend pass, so these have also
been fixed.
llvm-svn: 364323
This is a pre-commit of the tests introduced by the SuperNode SLP patch D63661.
Committed on behalf of @vporpo (Vasileios Porpodas)
Differential Revision: https://reviews.llvm.org/D63664
llvm-svn: 364320
Introduce three pseudo instructions to be used during DAG ISel to
represent v8.1-m low-overhead loops. One maps to set_loop_iterations
while loop_decrement_reg is lowered to two, so that we can separate
the decrement and branching operations. The pseudo instructions are
expanded pre-emission, where we can still decide whether we actually
want to generate a low-overhead loop, in a new pass:
ARMLowOverheadLoops. The pass currently bails, reverting to an sub,
icmp and br, in the cases where a call or stack spill/restore happens
between the decrement and branching instructions, or if the loop is
too large.
Differential Revision: https://reviews.llvm.org/D63476
llvm-svn: 364288
This is the Demorgan'd 'not' of the pattern handled in:
D63660 / rL364153
This is another intermediate IR step towards solving PR42314:
https://bugs.llvm.org/show_bug.cgi?id=42314
We can test if a value is not a power-of-2 using ctpop(X) > 1,
so combining that with an is-zero check of the input is the
same as testing if not exactly 1 bit is set:
(X == 0) || (ctpop(X) u> 1) --> ctpop(X) != 1
llvm-svn: 364246
Inference of nowrap flags in CVP has been disabled, because it
triggered a bug in LFTR (https://bugs.llvm.org/show_bug.cgi?id=31181).
This issue has been fixed in D60935, so we should be able to reenable
nowrap flag inference now.
Differential Revision: https://reviews.llvm.org/D62776
llvm-svn: 364228
Prefer the more exact intrinsic to remove a use of the input value
and possibly make further transforms easier (we will still need
to match patterns with funnel-shift of wider types as pieces of
bswap, especially if we want to canonicalize to funnel-shift with
constant shift amount). Discussed in D46760.
llvm-svn: 364187
Summary:
Handle smul.fix.sat in the scalarizer. This is done by
adding smul.fix.sat to the set of "isTriviallyVectorizable"
intrinsics.
The addition of smul.fix.sat in isTriviallyVectorizable and
hasVectorInstrinsicScalarOpd can also be seen as a preparation
to be able to use hasVectorInstrinsicScalarOpd in ConstantFolding.
Reviewers: rengolin, RKSimon, dblaikie
Reviewed By: rengolin
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D63704
llvm-svn: 364177
In rL364135, I taught IndVars to fold exiting branches in loops with a zero backedge taken count (i.e. loops that only run one iteration). This extends that to eliminate the dead comparison left around.
llvm-svn: 364155
This is another intermediate IR step towards solving PR42314:
https://bugs.llvm.org/show_bug.cgi?id=42314
We can test if a value is power-of-2-or-0 using ctpop(X) < 2,
so combining that with a non-zero check of the input is the
same as testing if exactly 1 bit is set:
(X != 0) && (ctpop(X) u< 2) --> ctpop(X) == 1
Differential Revision: https://reviews.llvm.org/D63660
llvm-svn: 364153
This turned out to be surprisingly effective. I was originally doing this just for completeness sake, but it seems like there are a lot of cases where SCEV's exit count reasoning is stronger than it's isKnownPredicate reasoning.
Once this is in, I'm thinking about trying to build on the same infrastructure to eliminate provably untaken checks. There may be something generally interesting here.
Differential Revision: https://reviews.llvm.org/D63618
llvm-svn: 364135
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 an example).
Committed on behalf of @vporpo (Vasileios Porpodas)
Differential Revision: https://reviews.llvm.org/D60897
llvm-svn: 364084
Summary:
The motivation for this was to propagate fast-math flags like nnan and
ninf on vector floating point operations to the corresponding scalar
operations to take advantage of follow-on optimizations. But I think
the same argument applies to all of our IR flags: if they apply to the
vector operation then they also apply to all the individual scalar
operations, and they might enable follow-on optimizations.
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D63593
llvm-svn: 364051
Summary:
LLVM Allows Targets to provide information that guides optimisations
made to LLVM IR. This is done with callbacks on a TargetTransformInfo object.
This patch adds a TargetTransformInfo class for RISC-V. This will allow us to
implement RISC-V specific callbacks as they become necessary.
This commit also adds the getIntImmCost callbacks, and tests them with a simple
constant hoisting test. Our immediate costs are on the conservative side, for
the moment, but we prevent hoisting in most circumstances anyway.
Previous review was on D63007
Reviewers: asb, luismarques
Reviewed By: asb
Subscribers: ributzka, MaskRay, llvm-commits, Jim, benna, psnobl, jocewei, PkmX, rkruppe, the_o, brucehoult, MartinMosbeck, rogfer01, edward-jones, zzheng, jrtc27, shiva0217, kito-cheng, niosHD, sabuasal, apazos, simoncook, johnrusso, rbar, hiraditya, mgorny
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D63433
llvm-svn: 364046
Also, add a FIXME for the unsafe transform on a unary FNeg. A unary FNeg can only be transformed to a FMul by -1.0 when the nnan flag is present. The unary FNeg project is a WIP, so the unsafe transformation is acceptable until that work is complete.
The bogus assert with introduced in D63445.
llvm-svn: 363998
As discussed in PR42314:
https://bugs.llvm.org/show_bug.cgi?id=42314
Improving the canonicalization for these patterns:
rL363956
...means we should adjust/enhance the related simplification.
https://rise4fun.com/Alive/w1cp
Name: isPow2 or zero
%x = and i32 %xx, 2048
%a = add i32 %x, -1
%r = and i32 %a, %x
=>
%r = i32 0
llvm-svn: 363997
Summary:
The getClobberingMemoryAccess API checks for clobbering accesses in a loop by walking the backedge. This may check if a memory access is being
clobbered by the loop in a previous iteration, depending how smart AA got over the course of the updates in MemorySSA (it does not occur when built from scratch).
If no clobbering access is found inside the loop, it will optimize to an access outside the loop. This however does not mean that access is safe to sink.
Given:
```
for i
load a[i]
store a[i]
```
The access corresponding to the load can be optimized to outside the loop, and the load can be hoisted. But it is incorrect to sink it.
In order to sink the load, we'd need to check no Def clobbers the Use in the same iteration. With this patch we currently restrict sinking to either
Defs not existing in the loop, or Defs preceding the load in the same block. An easy extension is to ensure the load (Use) post-dominates all Defs.
Caught by PR42294.
This issue also shed light on the converse problem: hoisting stores in this same scenario would be illegal. With this patch we restrict
hoisting of stores to the case when their corresponding Defs are dominating all Uses in the loop.
Reviewers: george.burgess.iv
Subscribers: jlebar, Prazek, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D63582
llvm-svn: 363982
I added a canonicalization to create this general pattern in:
rL363956
But as noted in PR42314:
https://bugs.llvm.org/show_bug.cgi?id=42314#c11
...we have a (potentially expensive) simplification for the version
of the code that we just canonicalized away from, so we should
add/adjust that code to match.
llvm-svn: 363981
The form that compares against 0 is better because:
1. It removes a use of the input value.
2. It's the more standard form for this pattern: https://graphics.stanford.edu/~seander/bithacks.html#DetermineIfPowerOf2
3. It results in equal or better codegen (tested with x86, AArch64, ARM, PowerPC, MIPS).
This is a root cause for PR42314, but probably doesn't completely answer the codegen request:
https://bugs.llvm.org/show_bug.cgi?id=42314
Alive proof:
https://rise4fun.com/Alive/9kG
Name: is power-of-2
%neg = sub i32 0, %x
%a = and i32 %neg, %x
%r = icmp eq i32 %a, %x
=>
%dec = add i32 %x, -1
%a2 = and i32 %dec, %x
%r = icmp eq i32 %a2, 0
Name: is not power-of-2
%neg = sub i32 0, %x
%a = and i32 %neg, %x
%r = icmp ne i32 %a, %x
=>
%dec = add i32 %x, -1
%a2 = and i32 %dec, %x
%r = icmp ne i32 %a2, 0
llvm-svn: 363956
Thought of this case while working on something else. We appear to get it right in all of the variations I tried, but that's by accident. So, add a test which would catch the potential bug.
llvm-svn: 363953
Teach IndVarSimply's LinearFunctionTestReplace transform to handle multiple exit loops. LFTR does two key things 1) it rewrites (all) exit tests in terms of a common IV potentially eliminating one in the process and 2) it moves any offset/indexing/f(i) style logic out of the loop.
This turns out to actually be pretty easy to implement. SCEV already has all the information we need to know what the backedge taken count is for each individual exit. (We use that when computing the BE taken count for the loop as a whole.) We basically just need to iterate through the exiting blocks and apply the existing logic with the exit specific BE taken count. (The previously landed NFC makes this super obvious.)
I chose to go ahead and apply this to all loop exits instead of only latch exits as originally proposed. After reviewing other passes, the only case I could find where LFTR form was harmful was LoopPredication. I've fixed the latch case, and guards aren't LFTRed anyways. We'll have some more work to do on the way towards widenable_conditions, but that's easily deferred.
I do want to note that I added one bit after the review. When running tests, I saw a new failure (no idea why didn't see previously) which pointed out LFTR can rewrite a constant condition back to a loop varying one. This was theoretically possible with a single exit, but the zero case covered it in practice. With multiple exits, we saw this happening in practice for the eliminate-comparison.ll test case because we'd compute a ExitCount for one of the exits which was guaranteed to never actually be reached. Since LFTR ran after simplifyAndExtend, we'd immediately turn around and undo the simplication work we'd just done. The solution seemed obvious, so I didn't bother with another round of review.
Differential Revision: https://reviews.llvm.org/D62625
llvm-svn: 363883
greater-than-sign is not a BRE special character.
POSIX.1-2017 XBD Section 9.3.2 indicates that the interpretation of `\>`
is undefined. This patch replaces the pattern.
llvm-svn: 363828
Summary:
Bug: https://bugs.llvm.org/show_bug.cgi?id=39024
The bug reports that a vectorized loop is stepped through 4 times and each step through the loop seemed to show a different path. I found two problems here:
A) An incorrect line number on a preheader block (for.body.preheader) instruction causes a step into the loop before it begins.
B) Instructions in the middle block have different line numbers which give the impression of another iteration.
In this patch I give all of the middle block instructions the line number of the scalar loop latch terminator branch. This seems to provide the smoothest debugging experience because the vectorized loops will always end on this line before dropping into the scalar loop. To solve problem A I have altered llvm::SplitBlockPredecessors to accommodate loop header blocks.
I have set up a separate review D61933 for a fix which is required for this patch.
Reviewers: samsonov, vsk, aprantl, probinson, anemet, hfinkel, jmorse
Reviewed By: hfinkel, jmorse
Subscribers: jmorse, javed.absar, eraman, kcc, bjope, jmellorcrummey, hfinkel, gbedwell, hiraditya, zzheng, llvm-commits
Tags: #llvm, #debug-info
Differential Revision: https://reviews.llvm.org/D60831
> llvm-svn: 363046
llvm-svn: 363786
Summary:
The test case does an (out of bounds) load from a global constant with
type <3 x float>. InstSimplify tried to turn this into an integer load
of the whole alloc size of the vector, which is 128 bits due to
alignment padding, and then bitcast this to <3 x vector> which failed
an assertion due to the type size mismatch.
The fix is to do an integer load of the normal size of the vector, with
no alignment padding.
Reviewers: tpr, arsenm, majnemer, dstuttard
Reviewed By: arsenm
Subscribers: hfinkel, wdng, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D63375
llvm-svn: 363784
[SROA] Enhance SROA to handle `addrspacecast`ed allocas
- Fix typo in original change
- Add additional handling to ensure all return pointers are properly
casted.
Summary:
- After `addrspacecast` is allowed to be eliminated in SROA, the
adjusting of storage pointer (from `alloca) needs to handle the
potential different address spaces between the storage pointer (from
alloca) and the pointer being used.
Reviewers: arsenm
Subscribers: wdng, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D63501
llvm-svn: 363743
Summary:
- After `addrspacecast` is allowed to be eliminated in SROA, the
adjusting of storage pointer (from `alloca) needs to handle the
potential different address spaces between the storage pointer (from
alloca) and the pointer being used.
Reviewers: arsenm
Subscribers: wdng, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D63501
llvm-svn: 363711
This patch really contains two pieces:
Teach SCEV how to fold a phi in the header of a loop to the value on the backedge when a) the backedge is known to execute at least once, and b) the value is safe to use globally within the scope dominated by the original phi.
Teach IndVarSimplify's rewriteLoopExitValues to allow loop invariant expressions which already exist (and thus don't need new computation inserted) even in loops where we can't optimize away other uses.
Differential Revision: https://reviews.llvm.org/D63224
llvm-svn: 363619
Recommit r363289 with a bug fix for crash identified in pr42279. Issue was that a loop exit test does not have to be an icmp, leading to a null dereference crash when new logic was exercised for that case. Test case previously committed in r363601.
Original commit comment follows:
This contains fixes for two cases where we might invalidate inbounds and leave it stale in the IR (a miscompile). Case 1 is when switching to an IV with no dynamically live uses, and case 2 is when doing pre-to-post conversion on the same pointer type IV.
The basic scheme used is to prove that using the given IV (pre or post increment forms) would have to already trigger UB on the path to the test we're modifying. As such, our potential UB triggering use does not change the semantics of the original program.
As was pointed out in the review thread by Nikita, this is defending against a separate issue from the hasConcreteDef case. This is about poison, that's about undef. Unfortunately, the two are different, see Nikita's comment for a fuller explanation, he explains it well.
(Note: I'm going to address Nikita's last style comment in a separate commit just to minimize chance of subtle bugs being introduced due to typos.)
Differential Revision: https://reviews.llvm.org/D62939
llvm-svn: 363613
Summary:
Update compare normalization in SimpleValue hashing to break ties (when
the same value is being compared to itself) by switching to the swapped
predicate if it has a lower numerical value. This brings the hashing in
line with isEqual, which already recognizes the self-compares with
swapped predicates as equal.
Fixes PR 42280.
Reviewers: spatel, efriedma, nikic, fhahn, uabelho
Reviewed By: nikic
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D63349
llvm-svn: 363598
When considering a loop containing nontemporal stores or loads for
vectorization, suppress the vectorization if the corresponding
vectorized store or load with the aligment of the original scaler
memory op is not supported with the nontemporal hint on the target.
This adds two new functions:
bool isLegalNTStore(Type *DataType, unsigned Alignment) const;
bool isLegalNTLoad(Type *DataType, unsigned Alignment) const;
to TTI, leaving the target independent default implementation as
returning true, but with overriding implementations for X86 that
check the legality based on available Subtarget features.
This fixes https://llvm.org/PR40759
Differential Revision: https://reviews.llvm.org/D61764
llvm-svn: 363581
If an addrspacecast needed to be inserted again, this was creating a
clone of the original cast for each user. Just use the original, which
also saves losing the value name.
llvm-svn: 363562
Should also be marked writeonly, but I think that would require
splitting the version with done set to a separate intrinsic
Test change is only from renumbering the attribute group numbers,
which for some reason the generated check lines consider.
llvm-svn: 363560
The HardwareLoops pass finds exit blocks with a scevable exit count.
If the target specifies to update the loop counter in a register,
through a phi, we need to ensure that the exit block is a latch so
that we can insert the phi with the correct value for the incoming
edge.
Differential Revision: https://reviews.llvm.org/D63336
llvm-svn: 363556
Summary:
Avoid that loop vectorizer creates loads/stores of vectors
with "irregular" types when interleaving. An example of
an irregular type is x86_fp80 that is 80 bits, but that
may have an allocation size that is 96 bits. So an array
of x86_fp80 is not bitcast compatible with a vector
of the same type.
Not sure if interleavedAccessCanBeWidened is the best
place for this check, but it solves the problem seen
in the added test case. And it is the same kind of check
that already exists in memoryInstructionCanBeWidened.
Reviewers: fhahn, Ayal, craig.topper
Reviewed By: fhahn
Subscribers: hiraditya, rkruppe, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D63386
llvm-svn: 363547
Second functional change following on from rL362687. Pass the
NoWrapFlags from the MulExpr to InsertBinop when we're generating a
shl or mul.
Differential Revision: https://reviews.llvm.org/D61934
llvm-svn: 363540
Third time's the charm.
This was reverted in r363220 due to being suspected of an internal benchmark
regression and a test failure, none of which turned out to be caused by this.
llvm-svn: 363529
SimplifyCFG has a bug that results in inconsistent prof branch_weights metadata
if unreachable switch cases are removed. This patch fixes this bug by making use
of the newly introduced SwitchInstProfUpdateWrapper class (see patch D62122).
A new test is created.
Differential Revision: https://reviews.llvm.org/D62186
llvm-svn: 363527
Fix folds of addo and subo with an undef operand to be:
`@llvm.{u,s}{add,sub}.with.overflow` all fold to `{ undef, false }`,
as per LLVM undef rules.
Same for commuted variants.
Based on the original version of the patch by @nikic.
Fixes [[ https://bugs.llvm.org/show_bug.cgi?id=42209 | PR42209 ]]
Differential Revision: https://reviews.llvm.org/D63065
llvm-svn: 363522
This is based on the example/discussion in PR37428:
https://bugs.llvm.org/show_bug.cgi?id=37428
Proper vector shift instructions don't appear until AVX2, so we may generate several
extra instructions within a loop trying to compensate for that. It's difficult to
recover from that shift expansion later than this, so use the existing TLI hook and
splat analysis to enable better codegen.
This extends CGP functionality introduced with:
rL201655
Differential Revision: https://reviews.llvm.org/D63233
llvm-svn: 363511
If we can detect that saturating math that depends on an IV cannot
overflow, replace it with simple math. This is similar to the CVP
optimization from D62703, just based on a different underlying
analysis (SCEV vs LVI) that catches different cases.
Differential Revision: https://reviews.llvm.org/D62792
llvm-svn: 363489
with 'objc_arc_inert'
Those calls are no-ops, so they can be safely deleted.
rdar://problem/49839633
Differential Revision: https://reviews.llvm.org/D62433
llvm-svn: 363468
There is a circular dependency between SROA and InferAddressSpaces
today that requires running both multiple times in order to be able to
eliminate all simple allocas and addrspacecasts. InferAddressSpaces
can't remove addrspacecasts when written to memory, and SROA helps
move pointers out of memory.
This should avoid inserting new commuting addrspacecasts with GEPs,
since there are unresolved questions about pointer wrapping between
different address spaces.
For now, don't replace volatile operations that don't match the alloca
addrspace, as it would change the address space of the access. It may
be still OK to insert an addrspacecast from the new alloca, but be
more conservative for now.
llvm-svn: 363462
I'm not 100% sure about this, since I'm worried about IR transforms
that might end up introducing divergence downstream once replaced with
a constant, but I haven't come up with an example yet.
llvm-svn: 363406
InsertBinop now accepts NoWrapFlags, so pass them through when
expanding a simple add expression.
This is the first re-commit of the functional changes from rL362687,
which was previously reverted.
Differential Revision: https://reviews.llvm.org/D61934
llvm-svn: 363364
Also add baseline tests to show effect of later patches.
There were a couple of regressions here that were never caught,
but my patch set that this is a preparation to will fix them.
This is the third attempt to land this patch.
Differential Revision: https://reviews.llvm.org/D61150
llvm-svn: 363319
This contains fixes for two cases where we might invalidate inbounds and leave it stale in the IR (a miscompile). Case 1 is when switching to an IV with no dynamically live uses, and case 2 is when doing pre-to-post conversion on the same pointer type IV.
The basic scheme used is to prove that using the given IV (pre or post increment forms) would have to already trigger UB on the path to the test we're modifying. As such, our potential UB triggering use does not change the semantics of the original program.
As was pointed out in the review thread by Nikita, this is defending against a separate issue from the hasConcreteDef case. This is about poison, that's about undef. Unfortunately, the two are different, see Nikita's comment for a fuller explanation, he explains it well.
(Note: I'm going to address Nikita's last style comment in a separate commit just to minimize chance of subtle bugs being introduced due to typos.)
Differential Revision: https://reviews.llvm.org/D62939
llvm-svn: 363289
Summary:
The logic in EarlyCSE that looks through 'not' operations in the
predicate recognizes e.g. that `select (not (cmp sgt X, Y)), X, Y` is
equivalent to `select (cmp sgt X, Y), Y, X`. Without this change,
however, only the latter is recognized as a form of `smin X, Y`, so the
two expressions receive different hash codes. This leads to missed
optimization opportunities when the quadratic probing for the two hashes
doesn't happen to collide, and assertion failures when probing doesn't
collide on insertion but does collide on a subsequent table grow
operation.
This change inverts the order of some of the pattern matching, checking
first for the optional `not` and then for the min/max/abs patterns, so
that e.g. both expressions above are recognized as a form of `smin X, Y`.
It also adds an assertion to isEqual verifying that it implies equal
hash codes; this fires when there's a collision during insertion, not
just grow, and so will make it easier to notice if these functions fall
out of sync again. A new flag --earlycse-debug-hash is added which can
be used when changing the hash function; it forces hash collisions so
that any pair of values inserted which compare as equal but hash
differently will be caught by the isEqual assertion.
Reviewers: spatel, nikic
Reviewed By: spatel, nikic
Subscribers: lebedev.ri, arsenm, craig.topper, efriedma, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D62644
llvm-svn: 363274
This patch uses the mechanism from D62995 to strengthen the
definitions of the reduction intrinsics by letting the scalar
result/accumulator type be overloaded from the vector element type.
For example:
; The LLVM LangRef specifies that the scalar result must equal the
; vector element type, but this is not checked/enforced by LLVM.
declare i32 @llvm.experimental.vector.reduce.or.i32.v4i32(<4 x i32> %a)
This patch changes that into:
declare i32 @llvm.experimental.vector.reduce.or.v4i32(<4 x i32> %a)
Which has the type-constraint more explicit and causes LLVM to check
the result type with the vector element type.
Reviewers: RKSimon, arsenm, rnk, greened, aemerson
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D62996
llvm-svn: 363240
TTI should report that it's not profitable to generate a hardware loop
if it, or one of its child loops, has already been converted.
Differential Revision: https://reviews.llvm.org/D63212
llvm-svn: 363234
This reverts 363226 and 363227, both NFC intended
I swear I fixed the test case that is failing, and ran
the tests, but I will look into it again.
llvm-svn: 363229
Also add baseline tests to show effect of later patches.
There were a couple of regressions here that were never caught,
but my patch set that this is a preparation to will fix them.
Differential Revision: https://reviews.llvm.org/D61150
llvm-svn: 363226
see if my changes change anything
Also add baseline tests to show effect of later patches.
Differential Revision: https://reviews.llvm.org/D61150
llvm-svn: 363222
We have observed some failures with internal builds with this revision.
- Performance regressions:
- llvm's SingleSource/Misc evalloop shows performance regressions (although these may be red herrings).
- Benchmarks for Abseil's SwissTable.
- Correctness:
- Failures for particular libicu tests when building the Google AppEngine SDK (for PHP).
hwennborg has already been notified, and is aware of reproducer failures.
llvm-svn: 363220
The issue addressed in r363180 is more broadly relevant. For the moment, we don't actually get any of these cases because we a) restrict SCEV formation due to SCEExpander needing to preserve LCSSA, and b) don't iterate between loops.
llvm-svn: 363192
SCEV does not propagate arguments through one-input Phis so as to make it easy for the SCEV expander (and related code) to preserve LCSSA. It's not entirely clear this restriction is neccessary, but for the moment it exists. For this reason, we don't analyze single-entry phi inputs. However it is possible that when an this input leaves the loop through LCSSA Phi, it is a provable constant. Missing that results in an order of optimization issue in loop exit value rewriting where we miss some oppurtunities based on order in which we visit sibling loops.
This patch teaches computeSCEVAtScope about this case. We can generalize it later, but so far we can only replace LCSSA Phis with their constant loop-exiting values. We should probably also add similiar logic directly in the SCEV construction path itself.
Patch by: mkazantsev (with revised commit message by me)
Differential Revision: https://reviews.llvm.org/D58113
llvm-svn: 363180
This was using its own, outdated list of possible captures. This was
at minimum not catching cmpxchg and addrspacecast captures.
One change is now any volatile access is treated as capturing. The
test coverage for this pass is quite inadequate, but this required
removing volatile in the lifetime capture test.
Also fixes some infrastructure issues to allow running just the IR
pass.
Fixes bug 42238.
llvm-svn: 363169
This changes the standalone pass only. Arguably the utility class
itself should assert there are no convergent calls. However, a target
pass with additional context may still be able to version a loop if
all of the dynamic conditions are sufficiently uniform.
llvm-svn: 363165
This case is slightly tricky, because loop distribution should be
allowed in some cases, and not others. As long as runtime dependency
checks don't need to be introduced, this should be OK. This is further
complicated by the fact that LoopDistribute partially ignores if LAA
says that vectorization is safe, and then does its own runtime pointer
legality checks.
Note this pass still does not handle noduplicate correctly, as this
should always be forbidden with it. I'm not going to bother trying to
fix it, as it would require more effort and I think noduplicate should
be removed.
https://reviews.llvm.org/D62607
llvm-svn: 363160
Implement the backend target hook to drive the HardwareLoops pass.
The low-overhead branch extension for Arm M-class cores is flexible
enough that we don't have to ensure correctness at this point, except
checking that the loop counter variable can be stored in LR - a
32-bit register. For it to be profitable, we want to avoid loops that
contain function calls, or any other instruction that alters the PC.
This implementation uses TargetLoweringInfo, to query type and
operation actions, looks at intrinsic calls and also performs some
manual checks for remainder/division and FP operations.
I think this should be a good base to start and extra details can be
filled out later.
Differential Revision: https://reviews.llvm.org/D62907
llvm-svn: 363149
We were only matching RHS being a loop invariant value, not the inverse. Since there's nothing which appears to canonicalize loop invariant values to RHS, this means we missed cases.
Differential Revision: https://reviews.llvm.org/D63112
llvm-svn: 363108
Summary:
Bug: https://bugs.llvm.org/show_bug.cgi?id=39024
The bug reports that a vectorized loop is stepped through 4 times and each step through the loop seemed to show a different path. I found two problems here:
A) An incorrect line number on a preheader block (for.body.preheader) instruction causes a step into the loop before it begins.
B) Instructions in the middle block have different line numbers which give the impression of another iteration.
In this patch I give all of the middle block instructions the line number of the scalar loop latch terminator branch. This seems to provide the smoothest debugging experience because the vectorized loops will always end on this line before dropping into the scalar loop. To solve problem A I have altered llvm::SplitBlockPredecessors to accommodate loop header blocks.
I have set up a separate review D61933 for a fix which is required for this patch.
Reviewers: samsonov, vsk, aprantl, probinson, anemet, hfinkel, jmorse
Reviewed By: hfinkel, jmorse
Subscribers: jmorse, javed.absar, eraman, kcc, bjope, jmellorcrummey, hfinkel, gbedwell, hiraditya, zzheng, llvm-commits
Tags: #llvm, #debug-info
Differential Revision: https://reviews.llvm.org/D60831
llvm-svn: 363046
As pointed out by Nikita in review, undef and poison need to be handled separately. Since we're no longer expecting any test improvements - just fixes for miscompiles - update the tests to bypass the existing undef check.
llvm-svn: 363002
As shown in PR41279, some basic blocks (such as catchswitch) cannot be
instrumented. This patch filters out these BBs in PGO instrumentation.
It also sets the profile count to the fail-to-instrument edge, so that we
can propagate the counts in the CFG.
Differential Revision: https://reviews.llvm.org/D62700
llvm-svn: 362995
There are two interesting sub-cases here. 1) Switching IVs is legal, but only in pre-increment form. and 2) Switching IVs is legal, and so is post-increment form.
llvm-svn: 362993
Flesh out a collection of tests for switching to a dead IV within LFTR, both for the current miscompile, and for some cases which we should be able to handle via simple reasoning.
llvm-svn: 362976
This was discussed as part of D62880. The basic thought is that computing BE taken count after widening should produce (on average) an equally good backedge taken count as the one before widening. Since there's only one test in the suite which is impacted by this change, and it's essentially equivelent codegen, that seems to be a reasonable assertion. This change was separated from r362971 so that if this turns out to be problematic, the triggering piece is obvious and easily revertable.
For the nestedIV example from elim-extend.ll, we end up with the following BE counts:
BEFORE: (-2 + (-1 * %innercount) + %limit)
AFTER: (-1 + (sext i32 (-1 + %limit) to i64) + (-1 * (sext i32 %innercount to i64))<nsw>)
Note that before is an i32 type, and the after is an i64. Truncating the i64 produces the i32.
llvm-svn: 362975
Similar to rL362909:
This isn't the ideal fix (use FMF on the select), but it's still an
improvement until we have better FMF propagation to selects and other
FP math operators.
I don't think there's much risk of regression from this change by
not including the FMF on the fcmp any more. The nsz/nnan FMF
should be the same on the fcmp and the fsub because they have the
same operand.
llvm-svn: 362943
Roman Lebedev