Summary:
Previosly we simply always said that `SCEVMinMaxExpr` is too costly to expand.
But this isn't really true, it expands into just a comparison+swap pair.
And again much like with add/mul, there will be one less such pair
than the number of operands. And we need to count the cost of operands themselves.
This does change a number of testcases, and as far as i can tell,
all of these changes are improvements, in the sense that
we fixed up more latches to do the [in]equality comparison.
This concludes cost-modelling changes, no other SCEV expressions exist as of now.
This is a part of addressing [[ https://bugs.llvm.org/show_bug.cgi?id=44668 | PR44668 ]].
Reviewers: reames, mkazantsev, wmi, sanjoy
Reviewed By: mkazantsev
Subscribers: hiraditya, javed.absar, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73744
Summary:
So, i wouldn't call this *obviously* correct,
but i think i got it right this time :)
Roughly, we have
```
Op0*x^0 + Op1*x^1 + Op2*x^2 ...
```
where `Op_{n} * x^{n}` is called term, and `n` the degree of term.
Due to the way they are stored internally in `SCEVAddRecExpr`,
i believe we can have `Op_{n}` to be `0`, so we should not charge for those.
I think it is most straight-forward to count the cost in 4 steps:
1. First, count it the same way we counted `scAddExpr`, but be sure to skip terms with zero constants.
Much like with `add` expr we will have one less addition than number of terms.
2. Each non-constant term (term degree >= 1) requires a multiplication between the `Op_{n}` and `x^{n}`.
But again, only charge for it if it is required - `Op_{n}` must not be 0 (no term) or 1 (no multiplication needed),
and obviously don't charge constant terms (`x^0 == 1`).
3. We must charge for all the `x^0`..`x^{poly_degree}` themselves.
Since `x^{poly_degree}` is `x * x * ... * x`, i.e. `poly_degree` `x`'es multiplied,
for final `poly_degree` term we again require `poly_degree-1` multiplications.
Note that all the `x^{0}`..`x^{poly_degree-1}` will be computed for the free along the way there.
4. And finally, the operands themselves.
Here, much like with add/mul exprs, we really don't look for preexisting instructions..
Reviewers: reames, mkazantsev, wmi, sanjoy
Reviewed By: mkazantsev
Subscribers: hiraditya, javed.absar, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73741
Summary:
While this resolves the regression from D73722 in `llvm/test/Transforms/IndVarSimplify/exit_value_test2.ll`,
this now regresses `llvm/test/Transforms/IndVarSimplify/elim-extend.ll` `@nestedIV` test,
we no longer can perform that expansion within default budget of `4`, but require budget of `6`.
That regression is being addressed by D73777.
The basic idea here is simple.
```
Op0, Op1, Op2 ...
| | |
\--+--/ |
| |
\---+---/
```
I.e. given N operands, we will have N-1 operations,
so we have to add cost of an add (mul) for **every** Op processed,
**except** the first one, plus we need to recurse into *every* Op.
I'm guessing there's already canonicalization that ensures we won't
have `1` operand in `scMulExpr`, and no `0` in `scAddExpr`/`scMulExpr`.
Reviewers: reames, mkazantsev, wmi, sanjoy
Reviewed By: mkazantsev
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73728
Summary:
If we don't believe this UDiv is actually a LShr in disguise, things are much worse.
First, we try to see if this UDiv actually originates from user code,
by looking for `S + 1`, and if found considering this UDiv to be free.
But otherwise, we always considered this UDiv to be high-cost.
However that is no longer the case with TTI-driven cost model:
our default budget is 4, which matches the default cost of UDiv,
so now we allow a single UDiv to not be counted as high-cost.
While that is the case, it is evident this is actually a regression
due to the fact that cost-modelling is incomplete - we did not account
for the `add`, `mul` costs yet. That is being addressed in D73728.
Cost-modelling for UDiv also seems pretty straight-forward:
subtract cost of the UDiv itself, and recurse into both the LHS and RHS.
Reviewers: reames, mkazantsev, wmi, sanjoy
Reviewed By: mkazantsev
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73722
Summary:
Like with casts, we need to subtract the cost of `lshr` instruction
from budget, and recurse into LHS operand.
Seems "pretty obviously correct" to me?
To be noted, there is a number of other shortcuts we //could// cost-model:
* `... + (-1 * ...)` -> `... - ...` <- likely very frequent case
* `x - (rem x, power-of-2)`, which is currently `(x udiv power-of-2) * power-of-2` -> `x & -log2(power-of-2)`
* `rem x, power-of-2`, which is currently `x - ((x udiv power-of-2) * power-of-2)` -> `x & log2(power-of-2)-1`
* `... * power-of-2` -> `... << log2(power-of-2)` <- likely not very beneficial
Reviewers: reames, mkazantsev, wmi, sanjoy
Reviewed By: mkazantsev
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73718
Summary:
This is not a NFC, although it does not change any of the existing tests.
I'm not really sure if we should have specific tests for the cost modelling itself.
This is the first patch that actually makes `SCEVExpander::isHighCostExpansionHelper()`
account for the cost of the SCEV expression, and consider the budget available,
by modelling cast expressions.
I believe the logic itself is "pretty obviously correct" - from budget,
we need to subtract the cost of the cast expression from inner type `Op->getType()`
to the `S->getType()` type, and recurse into the expression we are casting.
Reviewers: reames, mkazantsev, wmi, sanjoy
Reviewed By: mkazantsev
Subscribers: xbolva00, hiraditya, javed.absar, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73716
Summary:
Currently, as per `check-llvm`, we never call `SCEVExpander::isHighCostExpansion()` with null TTI,
so this appears to be a safe restriction.
Reviewers: reames, mkazantsev, wmi, sanjoy
Reviewed By: mkazantsev
Subscribers: javed.absar, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73712
Summary:
As far as i can tell this is still NFC.
Initially in rL146438 it was added at the top of the function,
later rL238507 dethroned it, and rL244474 did it again.
I'm not sure if we have already checked the cost of this expansion, we should be doing that again.
Reviewers: reames, mkazantsev, wmi, sanjoy, atrick, igor-laevsky
Reviewed By: mkazantsev
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73706
Summary:
In future patches`SCEVExpander::isHighCostExpansionHelper()` will respect the budget allocated by performing TTI cost modelling.
This is a fully NFC patch to make things reviewable.
Reviewers: reames, mkazantsev, wmi, sanjoy
Reviewed By: mkazantsev
Subscribers: hiraditya, zzheng, javed.absar, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73705
Summary:
Future patches will make use of TTI to perform cost-model-driven `SCEVExpander::isHighCostExpansionHelper()`
This is a fully NFC patch to make things reviewable.
Reviewers: reames, mkazantsev, wmi, sanjoy
Reviewed By: mkazantsev
Subscribers: hiraditya, zzheng, javed.absar, dmgreen, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73704
Summary:
Implement the DWARF register mapping described in
llvm/docs/AMDGPUUsage.rst
This is currently limited to wave64 VGPRs/AGPRs.
This also includes some minor changes in AMDGPUInstPrinter,
AMDGPUMCTargetDesc, and AMDGPUAsmParser to make generating CFI assembly
text and ELF sections possible to ease testing, although complete CFI
support is not yet implemented.
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D74915
Add a dump method that recursively prints an instruction and all
the instructions defining its operands and so on.
This is helpful when looking at combiner issue.
NFC
Differential Revision: https://reviews.llvm.org/D75094
This reverts commit 8d22100f66.
There was a functional regression reported (https://bugs.llvm.org/show_bug.cgi?id=44996). I'm not actually sure the patch is wrong, but I don't have time to investigate currently, and this line of work isn't something I'm likely to get back to quickly.
Summary:
This patch if motivated by D74560, specifically the subthread about what
to print upon encountering reserved initial length values.
If the debug_line prologue has an unsupported version, we skip parsing
the rest of the data. If we encounter an reserved initial length field,
we don't even parse the version. However, we still print out all members
(with value 0) in the dump function.
This patch introduces early exits in the Prologue::dump function so that
we print only the fields that were parsed successfully. In case of an
unsupported version, we skip printing all subsequent prologue fields --
because we don't even know if this version has those fields. In case of a
reserved unit length, we don't print anything -- if the very first field
of the prologue is invalid, it's hard to say if we even have a prologue
to begin with.
Note that the user will still be able to see the invalid/reserved
initial length value in the error message. I've modified (reordered)
debug_line_invalid.test to show that the error message comes straight
after the debug_line offset. I've also added some flush() calls to the
dumping code to ensure this is the case in all situations (without that,
the warnings could get out of sync if the output was not a terminal -- I
guess this is why std::iostreams have the tie() function).
Reviewers: jhenderson, ikudrin, dblaikie
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75043
Much like with reassociateShiftAmtsOfTwoSameDirectionShifts(),
as input, we have the following pattern:
icmp eq/ne (and ((x shift Q), (y oppositeshift K))), 0
We want to rewrite that as:
icmp eq/ne (and (x shift (Q+K)), y), 0 iff (Q+K) u< bitwidth(x)
While we know that originally (Q+K) would not overflow
(because 2 * (N-1) u<= iN -1), we may have looked past extensions of
shift amounts. so it may now overflow in smaller bitwidth.
To ensure that does not happen, we need to ensure that the total maximal
shift amount is still representable in that smaller bitwidth.
If the overflow would happen, (Q+K) u< bitwidth(x) check would be bogus.
https://bugs.llvm.org/show_bug.cgi?id=44802
As input, we have the following pattern:
Sh0 (Sh1 X, Q), K
We want to rewrite that as:
Sh x, (Q+K) iff (Q+K) u< bitwidth(x)
While we know that originally (Q+K) would not overflow
(because 2 * (N-1) u<= iN -1), we may have looked past extensions of
shift amounts. so it may now overflow in smaller bitwidth.
To ensure that does not happen, we need to ensure that the total maximal
shift amount is still representable in that smaller bitwidth.
If the overflow would happen, (Q+K) u< bitwidth(x) check would be bogus.
https://bugs.llvm.org/show_bug.cgi?id=44802
Code duplication (subsequently removed by refactoring) allowed
a logic discrepancy to creep in here.
We were being conservative about creating a vector binop -- but
not a vector cmp -- in the case where a vector op has the same
estimated cost as the scalar op. We want to be more aggressive
here because that can allow other combines based on reduced
instruction count/uses.
We can reverse the transform in DAGCombiner (potentially with a
more accurate cost model) if this causes regressions.
AFAIK, this does not conflict with InstCombine. We have a
scalarize transform there, but it relies on finding a constant
operand or a matching insertelement, so that means it eliminates
an extractelement from the sequence (so we won't have 2 extracts
by the time we get here if InstCombine succeeds).
Differential Revision: https://reviews.llvm.org/D75062
Summary:
This patch renames functions and TableGen classes for SVE gathers and
scatters. The original names implied that the corresponding
methods/classes are only suited for regular gathers/scatters (i.e. LD1
and ST1), which is not the case. Indeed, we will be re-using them for
non-temporal and first-faulting gathers/scatters in the forthcoming
patches. The new names also highlight the split into Vector-Scalar (VS)
and Scalar-Vector (SV) cases.
List of changes:
* `performLD1GatherCombine` and `performST1ScatterCombine` are renamed
as `performGatherLoadCombine` and `performScatterStoreCombine`,
respectively.
* Selection DAG types for scatters and gathers from
AArch64SVEInstrInfo.td are renamed. For example, `SDT_AArch64_GLD1` is
renamed as `SDT_AArch64_GATHER_SV`. SV stands for Scalar-Vector, as
opposed to Vector-Scalar (VS).
* The intrinsic classes from IntrinsicsAArch64.td are renamed. For
example, `AdvSIMD_GatherLoad_64bitOffset_Intrinsic` is renamed as
`AdvSIMD_GatherLoad_SV_64b_Offsets_Intrinsic`.
* Updated comments in `performGatherLoadCombine` and
`performScatterStoreCombine`.
Reviewers: sdesmalen, rengolin, efriedma
Reviewed By: sdesmalen
Subscribers: tschuett, kristof.beyls, hiraditya, rkruppe, psnobl, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75035
Summary:
Implements the following intrinsics:
* llvm.aarch64.sve.convert.to.svbool
* llvm.aarch64.sve.convert.from.svbool
For converting the ACLE svbool_t type (<n x 16 x i1>) to and from the
other predicate types: <n x 8 x i1>, <n x 4 x i1> and <n x 2 x i1>.
Reviewers: sdesmalen, kmclaughlin, efriedma, dancgr, rengolin
Reviewed By: sdesmalen, efriedma
Subscribers: tschuett, kristof.beyls, hiraditya, rkruppe, psnobl, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D74471
While the value of the CIE pointer field in a DWARF FDE record is
an offset to the corresponding CIE record from the beginning of
the section, for EH FDE records it is relative to the current offset.
Previously, we did not make that distinction when dumped both kinds
of FDE records and just printed the same value for the CIE pointer
field and the CIE offset; that was acceptable for DWARF FDEs but was
wrong for EH FDEs.
This patch fixes the issue by explicitly printing the offset of the
linked CIE object.
Differential Revision: https://reviews.llvm.org/D74613
Instead add it when we make the machine nodes during instruction
selections.
This makes this ISD node closer to ISD::MGATHER. Trying to see
if we remove the X86 specific ones.
The current set of custom combines are only really useful after
legalization, so move them there. There is a lot of overlap in the
boilerplate here, but I think we do want a pretty different set of
combines before and after legalize. I think we will want a lot of
overlap between the post-legalize and a post-regbankselect combiner.
MachineVerifier still takes 45-50% of total compile time with
-verify-machineinstrs, with calcRegsPassed dataflow taking ~50-60% of
MachineVerifier.
The majority of that time is spent in BBInfo::addPassed, mostly within
DenseSet implementing the sets the dataflow is operating over.
In particular, 1/4 of that DenseSet time is spent just iterating over it
(operator++), 40-50% on insertions, and most of the rest in ::count.
Given that, we're implementing custom sets just for this analysis here,
focusing on cheap insertions and O(n) iteration time (as opposed to
O(U), where U is the universe).
As it's based _mostly_ on BitVector for sparse and SmallVector for
dense, it may remotely resemble SparseSet. The difference is, our
solution is a lot less clever, doesn't have constant time `clear` that
we won't use anyway as reusing these sets across analyses is cumbersome,
and thus more space efficient and safer (got a resizable Universe and a
fallback to DenseSet for sparse if it gets too big).
With this patch MachineVerifier gets ~15-20% faster, its contribution to
total compile time drops from 45-50% to ~35%, while contribution of
calcRegsPassed to MachineVerifier drops from 50-60% to ~35% as well.
calcRegsPassed itself gets another 2x faster here.
All measured on a large suite of shaders targeting a number of GPUs.
Reviewers: bogner, stoklund, rudkx, qcolombet
Reviewed By: rudkx
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75033
Summary:
Terminators in LLVM aren't prohibited from returning values. This means that
the "callbr" instruction, which is used for "asm goto", can support "asm goto
with outputs."
This patch removes all restrictions against "callbr" returning values. The
heavy lifting is done by the code generator. The "INLINEASM_BR" instruction's
a terminator, and the code generator doesn't allow non-terminator instructions
after a terminator. In order to correctly model the feature, we need to copy
outputs from "INLINEASM_BR" into virtual registers. Of course, those copies
aren't terminators.
To get around this issue, we split the block containing the "INLINEASM_BR"
right before the "COPY" instructions. This results in two cheats:
- Any physical registers defined by "INLINEASM_BR" need to be marked as
live-in into the block with the "COPY" instructions. This violates an
assumption that physical registers aren't marked as "live-in" until after
register allocation. But it seems as if the live-in information only
needs to be correct after register allocation. So we're able to get away
with this.
- The indirect branches from the "INLINEASM_BR" are moved to the "COPY"
block. This is to satisfy PHI nodes.
I've been told that MLIR can support this handily, but until we're able to
use it, we'll have to stick with the above.
Reviewers: jyknight, nickdesaulniers, hfinkel, MaskRay, lattner
Reviewed By: nickdesaulniers, MaskRay, lattner
Subscribers: rriddle, qcolombet, jdoerfert, MatzeB, echristo, MaskRay, xbolva00, aaron.ballman, cfe-commits, JonChesterfield, hiraditya, llvm-commits, rnk, craig.topper
Tags: #llvm, #clang
Differential Revision: https://reviews.llvm.org/D69868
macro section dumping.
Summary: Previously macinfo infrastructure was using functions
names that were ambiguous i.e `getMacro/getMacroDWO` in a sense
of conveying stated intentions. This patch refactored them into more
reasonable `getDebugMacinfo/getDebugMacinfoDWO` names thus making
room for macro implementation.
Reviewers: aprantl, probinson, jini.susan.george, dblaikie
Reviewed By: dblaikie
Differential Revision: https://reviews.llvm.org/D75037
Changes the handling of odd breakdowns, and avoids using
G_EXTRACT/G_INSERT. Pad with undef to a wider size, and unmerge. Also
avoid introducing instructions for the fully undef components.
This is explicitly guaranteed in ARMARM. And it makes reasoning about
vectors easier: we can assume that if a vector operation is legal, the
corresponding scalar operation is also legal.
Differential Revision: https://reviews.llvm.org/D74993
Summary:
This patch moves the getIndexExpressionsFromGEP function from polly
into ScalarEvolution so that both polly and DependenceAnalysis can
use it for the purpose of subscript delinearization when the array
sizes are not parametric.
Authored By: bmahjour
Reviewer: Meinersbur, sebpop, fhahn, dmgreen, grosser, etiotto, bollu
Reviewed By: Meinersbur
Subscribers: hiraditya, arphaman, Whitney, ppc-slack, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73995
Summary:
These modificaitons will be used in D74883.
Fixed length C strings can have trailing NULLs or sometimes spaces (BSD archive files), so the fixed length C string defaults to stripping trailing NULLs, but can have the arguments specify to remove one or more kinds of spaces if needed. This is used to extract fixed length C strings from ELF NOTEs in D74883.
Reviewers: labath, dblaikie, aprantl
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D74991
This reverts commit 9fe769a961, and re-lands commit c2e272f8cf.
Summary: Add support for ?, DUP, and string initializers, as well as MASM syntax for named data locations.
This version avoids the use of a C++17-only feature, if-statements with initializer.
Reviewers: rnk, thakis
Reviewed By: thakis
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73226
Depending on the target, test suite, pipeline config and perhaps other
factors machine verifier when forced on with -verify-machineinstrs can
increase compile time 2-2.5 times over (Release, Asserts On), taking up
~60% of the time. An invaluable tool, it significantly slows down
machine verifier-enabled testing.
Nearly 75% of its time MachineVerifier spends in the calcRegsPassed
method. It's a classic forward dataflow analysis executed over sets, but
visiting MBBs in arbitrary order. We switch that to RPO here.
This speeds up MachineVerifier by about 35%, decreasing the overall
compile time with -verify-machineinstrs by 20-25% or so.
calcRegsPassed itself gets 2x faster here.
All measured on a large suite of shaders targeting a number of GPUs.
Reviewers: bogner, stoklund, rudkx, qcolombet
Reviewed By: bogner
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75032
Summary: Add support for ?, DUP, and string initializers, as well as MASM syntax for named data locations.
Reviewers: rnk, thakis
Reviewed By: thakis
Subscribers: merge_guards_bot, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73226
Previously we emitted an fmadd and a fmadd+fneg and combined them with a shufflevector. But this doesn't follow the correct exception behavior for unselected elements so the backend can't merge them into the fmaddsub/fmsubadd instructions.
This patch restores the the fmaddsub intrinsics so we don't have two arithmetic operations. We lose out on optimization opportunity in the non-strict FP case, but I don't think this is a big loss. If someone gives us a test case we can look into adding instcombine/dagcombine improvements. I'd rather not have the frontend do completely different things for strict and non-strict.
This still has problems because target specific intrinsics don't support strict semantics yet. We also still have all of the problems with masking. But we at least generate the right instruction in constrained mode now.
Differential Revision: https://reviews.llvm.org/D74268
GCC 9.2 seems to incorrectly issue warning about out of bounds
access. This situation should not happen in any way.
Differential Revision: https://reviews.llvm.org/D75071
This patch adds bindings to C and Go for
addCoroutinePassesToExtensionPoints, which is used to add coroutine
passes to the correct locations in PassManagerBuilder.
Differential Revision: https://reviews.llvm.org/D51642
This is the second patch as part of https://bugs.llvm.org/show_bug.cgi?id=36544
Merging in the ConstantSDNode variant of FoldConstantArithmetic. After this, I will begin merging in FoldConstantVectorArithmetic
I've ensured this patch can build & pass all lit tests in Windows and Linux environments.
Patch by @justice_adams (Justice Adams)
Differential Revision: https://reviews.llvm.org/D74881
Simply by implementing a few functions I was able to correctly
disassemble a much larger amount of instructions.
Differential Revision: https://reviews.llvm.org/D74045
Not all operands are correctly disassembled at the moment. This means
that some machine instructions won't have all the necessary operands
set.
To avoid asserting, print an error instead until the necessary support
has been implemented.
Differential Revision: https://reviews.llvm.org/D73958
A number of multiplication instructions (muls, mulsu, fmul, fmuls,
fmulsu) had the wrong register class for an operand. This resulted in
the wrong register being used for the instruction.
Example:
target datalayout = "e-P1-p:16:8-i8:8-i16:8-i32:8-i64:8-f32:8-f64:8-n8-a:8"
target triple = "avr-atmel-none"
define i16 @sliceAppend(i16, i16, i16, i16, i16, i16) addrspace(1) {
%d = mul i16 %0, %5
ret i16 %d
}
The first instruction would be muls r24, r31 before this patch. The r31
should have been r15 if you look at the intermediate forms during
instruction selection / register allocation, but the generated
instruction uses r31. After this patch, an extra movw is inserted to get
%5 in range for muls.
To make sure this bug is fixed everywhere, I checked all instructions
and found that most multiplication instructions suffered from this bug,
which I have fixed with this patch. No other instructions appear to be
affected.
Differential Revision: https://reviews.llvm.org/D74281
Summary:
Function descriptor csect on AIX should be 4 byte align instead of 1 byte align.
Reviewer: daltenty
Differential Revision: https://reviews.llvm.org/D74974
I'm hoping to begin improving shuffle combining across different vector sizes, but before that we must ensure that all existing getTargetShuffleInputs calls must bail if the inputs aren't the same size.
Extends the existing support for spilling and restoring the condition
register to the linkage area for 32-bit targets, and enables for AIX.
Differential Revision: https://reviews.llvm.org/D74349
D74976 will handle larger vector types, but since SLM doesn't support AVX+ then we will always be extracting from 128-bit vectors so don't need to scale the cost.
This adds infrastructure to print and parse MIR MachineOperand comments.
The motivation for the ARM backend is to print condition code names instead of
magic constants that are difficult to read (for human beings). For example,
instead of this:
dead renamable $r2, $cpsr = tEOR killed renamable $r2, renamable $r1, 14, $noreg
t2Bcc %bb.4, 0, killed $cpsr
we now print this:
dead renamable $r2, $cpsr = tEOR killed renamable $r2, renamable $r1, 14 /* CC::always */, $noreg
t2Bcc %bb.4, 0 /* CC:eq */, killed $cpsr
This shows that MachineOperand comments are enclosed between /* and */. In this
example, the EOR instruction is not conditionally executed (i.e. it is "always
executed"), which is encoded by the 14 immediate machine operand. Thus, now
this machine operand has /* CC::always */ as a comment. The 0 on the next
conditional branch instruction represents the equal condition code, thus now
this operand has /* CC:eq */ as a comment.
As it is a comment, the MI lexer/parser completely ignores it. The benefit is
that this keeps the change in the lexer extremely minimal and no target
specific parsing needs to be done. The changes on the MIPrinter side are also
minimal, as there is only one target hooks that is used to create the machine
operand comments.
Differential Revision: https://reviews.llvm.org/D74306
Summary:
Implements the @llvm.aarch64.sve.dupq.lane intrinsic.
As specified in the ACLE, the behaviour of:
svdupq_lane_u64(data, index)
...is identical to:
svtbl(data, svadd_x(svptrue_b64(),
svand_x(svptrue_b64(), svindex_u64(0, 1), 1),
index * 2))
If the index is in the range [0,3], the operation is equivalent
to a single DUP (.q) instruction.
Reviewers: sdesmalen, c-rhodes, cameron.mcinally, efriedma, dancgr, rengolin
Reviewed By: sdesmalen
Subscribers: tschuett, kristof.beyls, hiraditya, rkruppe, psnobl, cfe-commits, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D74734
Change the way that we remove the redundant iteration count code in
the presence of IT blocks. collectLocalKilledOperands has been
introduced to scan an instructions operands, collecting the killed
instructions and then visiting them too. This is used to delete the
code in the preheader which calculates the iteration count. We also
track any IT blocks within the preheader and, if we remove all the
instructions from the IT block, we also remove the IT instruction.
isSafeToRemove is used to remove any redundant uses of the iteration
count within the loop body.
Differential Revision: https://reviews.llvm.org/D74975
Summary:
This patch adds intrinsics and ISelDAG nodes for signed
and unsigned fixed-point division:
```
llvm.sdiv.fix.sat.*
llvm.udiv.fix.sat.*
```
These intrinsics perform scaled, saturating division
on two integers or vectors of integers. They are
required for the implementation of the Embedded-C
fixed-point arithmetic in Clang.
Reviewers: bjope, leonardchan, craig.topper
Subscribers: hiraditya, jdoerfert, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D71550
Summary:
There is no need to write out gcdas when forking because we can just reset the counters in the parent process.
Let say a counter is N before the fork, then fork and this counter is set to 0 in the child process.
In the parent process, the counter is incremented by P and in the child process it's incremented by C.
When dump is ran at exit, parent process will dump N+P for the given counter and the child process will dump 0+C, so when the gcdas are merged the resulting counter will be N+P+C.
About exec** functions, since the current process is replaced by an another one there is no need to reset the counters but just write out the gcdas since the counters are definitely lost.
To avoid to have lists in a bad state, we just lock them during the fork and the flush (if called explicitely) and lock them when an element is added.
Reviewers: marco-c
Reviewed By: marco-c
Subscribers: hiraditya, cfe-commits, #sanitizers, llvm-commits, sylvestre.ledru
Tags: #clang, #sanitizers, #llvm
Differential Revision: https://reviews.llvm.org/D74953
Summary:
The type used to represent functional units in MC is
'unsigned', which is 32 bits wide. This is currently
not a problem in any upstream target as no one seems
to have hit the limit on this yet, but in our
downstream one, we need to define more than 32
functional units.
Increasing the size does not seem to cause a huge
size increase in the binary (an llc debug build went
from 1366497672 to 1366523984, a difference of 26k),
so perhaps it would be acceptable to have this patch
applied upstream as well.
Subscribers: hiraditya, jsji, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D71210
This optimization bypasses GOT loads and calls/branches through stubs when the
ultimate target of the access/branch is found to be within range of the
reference.
Extra debugging output is also added to the generic JITLink algorithm and
basic GOT and Stubs builder utility to aid debugging.
The gather intrinsics use a floating point mask when the result
type is FP. But we call DemandedBits on the mask assuming its an
integer type. We also use integer types when we create it from
generic IR. So add a bitcast to the intrinsic path to guarantee
the integer type.
The type profile we use for the isel patterns lied about how
many operands the gather/scatter node has to skip the index
and scale operands. This allowed us to expand the baseptr
operand into base, displacement, and segment and then merge
the index and scale with them in the final instruction during
isel. This is kind of a hack that relies on isel not checking the
number of operands at all.
This commit switches to custom isel where we can manage this
directly without relying on holes in the isel checking.
Summary:
The IR printing always prints out all functions in a module with the new pass manager, even with -filter-print-funcs specified. This is being fixed in this change. However, there are two exceptions, i.e, with user-specified wildcast switch -filter-print-funcs=* or -print-module-scope, under which IR of all functions should be printed.
Test Plan:
make check-clang
make check-llvm
Reviewers: wenlei
Reviewed By: wenlei
Subscribers: wenlei, hiraditya, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D74814
Leave the gather/scatter subclasses, but make them inherit from
MemIntrinsicSDNode and delete their constructor and destructor.
This way we can still have the getIndex, getMask, etc. convenience
functions.
In order to build the Linux kernel, the back chain must be supported with
packed-stack. The back chain is then stored topmost in the register save
area.
Review: Ulrich Weigand
Differential Revision: https://reviews.llvm.org/D74506
This version fixes a buildbot failure cause by picking the wrong insert
point for XORs. We cannot pick the XOR binary operator as insert point,
as it is not guaranteed that both input operands for the overflow
intrinsic are defined before it.
This reverts the revert commit
c7fc0e5da6.
A question about this behavior came up on llvm-dev:
http://lists.llvm.org/pipermail/llvm-dev/2020-February/139003.html
...and as part of backend improvements in D73978.
We decided not to implement a more general change that would have
folded any FP binop with nearly arbitrary constant + undef operand
to undef because that is not theoretically correct (even if it is
practically correct).
This is the SDAG-equivalent to the IR change in D74713.
Add a map from BasicBlocks to overlap intervals. For partial writes, we
can keep track of those in IOLs. We only add candidates that are valid
for eliminations.
Reviewers: dmgreen, bryant, asbirlea, Tyker
Reviewed By: asbirlea
Differential Revision: https://reviews.llvm.org/D73757
If a simplication occurs the operand will be added to the worklist.
But since the demanded mask was based on N, we need to make sure
we revisit N in case there are more simplifications to be done.
Returning SDValue(N, 0) as we do, only tells DAG combine that
something changed, but that won't make it add anything to the
worklist.
Found while playing around with using VEXTRACT_STORE in more cases.
But I guess this doesn't affect any of our existing tests.
We can use MOVLPS which will load 64 bits, but we need a v4f32
result type. We already have isel patterns for this.
The code here is a little hacky. We can probably improve it with
more isel patterns.
This is similar to using movd which we do for sse2 targets.
I've added a DAG combine for VEXTRACT_STORE to use SimplifyDemandedVectorElts
to clean up some artifacts from type legalization.
The GenericLLVMIRPlatformSupport class runs a transform on all LLVM IR added to
the LLJIT instance to replace instances of llvm.global_ctors with a specially
named function that runs the corresponing static initializers (See
(GlobalCtorDtorScraper from lib/ExecutionEngine/Orc/LLJIT.cpp). This patch
updates the GenericIRPlatform class to check for this specially named function
in other materialization units that are added to the JIT and, if found, add
the function to the initializer work queue. Doing this allows object files
that were compiled from IR and cached to be reloaded in subsequent JIT sessions
without their initializers being skipped.
To enable testing this patch also updates the lli tool's -jit-kind=orc-lazy mode
to respect the -enable-cache-manager and -object-cache-dir options, and modifies
the CompileOnDemandLayer to rename extracted submodules to include a hash of the
names of their symbol definitions. This allows a simple object caching scheme
based on module names (which was already implemented in lli) to work with the
lazy JIT.
This patch adds new errors and error checking to the ObjectLinkingLayer to
catch cases where a compiled or loaded object either:
(1) Contains definitions not covered by its responsibility set, or
(2) Is missing definitions that are covered by its responsibility set.
Proir to this patch providing the correct set of definitions was treated as
an API contract requirement, however this requires that the client be confident
in the correctness of the whole compiler / object-cache pipeline and results
in difficult-to-debug assertions upon failure. Treating this as a recoverable
error results in clearer diagnostics.
The performance overhead of this check is one comparison of densemap keys
(symbol string pointers) per linking object, which is minimal. If this overhead
ever becomes a problem we can add the check under a flag that can be turned off
if the client fully trusts the rest of the pipeline.
I've noticed that it is not convenient to create YAMLs from
binaries (using obj2yaml) that have to be test cases for obj2yaml
later (after applying yaml2obj).
The problem, for example is that obj2yaml emits "DynamicSymbols:"
key instead of .dynsym. It also does not create .dynstr.
And when a YAML document without explicitly defined .dynsym/.dynstr
is given to yaml2obj, we have issues:
1) These sections are placed after non-allocatable sections (I've fixed it in D74756).
2) They have VA == 0. User needs create descriptions for such sections explicitly manually
to set a VA.
This patch addresses (2). I suggest to let yaml2obj assign virtual addresses by itself.
It makes an output binary to be much closer to "normal" ELF.
(It is still possible to use "Address: 0x0" for a section to get the original behavior
if it is needed)
Differential revision: https://reviews.llvm.org/D74764
Similar to what do for other operations that use a subset of bits.
Allows us to remove a pattern that shrinks a load. Which was
incorrect if the load was volatile.
Added two flags to omit uncommon or dead paths in the CFG graphs:
-cfg-hide-unreachable-paths
-cfg-hide-deoptimize-paths
The main purpose is performance analysis when such block are not
"interesting" from perspective of common path performance.
Reviewed By: apilipenko, davidxl
Differential Revision: https://reviews.llvm.org/D74346
Summary:
We already sorted the blocks when fixing up a set of mutual
loop entries, however, there can be multiple sets of such
mutual loop entries, and the order we encounter them
should not be random, so sort them too.
Fixes https://bugs.llvm.org/show_bug.cgi?id=44982
Patch by Alon Zakai (kripken)
Reviewers: aheejin, sbc100, dschuff
Subscribers: mgrang, sunfish, hiraditya, jgravelle-google, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D74999
This reverts commit 977cd661cf.
It breaks OpenCL testing. OpenCL Runtime is using PT_LOAD information
to calculate memory for global variables. This commit should be relanded once
the OpenCL runtime stops relying on PT_LOAD information for calculating global
variable memory size.
Differential Revision: https://reviews.llvm.org/D74995
Heads-up message: https://lists.llvm.org/pipermail/llvm-dev/2020-February/139390.html
GNU as started to emit warnings for changed sh_type or sh_flags in 2000.
GNU as>=2.35 will emit errors for most sh_type/sh_flags change, and error for entsize change.
Some cases remain warnings for legacy reasons:
.section .init_array,"ax", @progbits
.section .init_array,"ax", @init_array
# And some obscure sh_flags changes (OS/Processor specific flags)
The rationale of a diagnostic (warning or error) is that sh_type,
sh_flags or sh_entsize changes usually indicate user errors. The values
are taken from the first .section directive. Successive directives are ignored.
We just try to be rigid and emit errors for all sh_type/sh_flags/sh_entsize change.
A possible improvement in the future is to reuse
llvm-readobj/ELFDumper.cpp:getSectionTypeString so that we can name the
type in the diagnostics.
Reviewed By: psmith
Differential Revision: https://reviews.llvm.org/D73999
This patch adds a cache that is valid only for the duration of a call
to getKnownBits. With such short lived cache we avoid all the problems
of cache invalidation while still getting the benefits of reusing
the information we already computed.
This cache is useful whenever an instruction occurs more than once
in a chain of computation.
E.g.,
v0 = G_ADD v1, v2
v3 = G_ADD v0, v1
Previously we would compute the known bits for:
v1, v2, v0, then v1 again and finally v3.
With the patch, now we won't have to recompute v1 again.
NFC
Summary:
Blocks in a loop can be in any order as long as the loop header is the
first block in Blocks.
With some order of Blocks, cloneLoopWithPreheader would trigger the
assertion in addBasicBlockToLoop.
Example:
define void @test(i64 %N) {
preheader.i:
br label %header.i
header.i:
%i = phi i64 [ 0, %preheader.i ], [ %inc.i, %latch.i ]
br label %header.j
header.j:
%j = phi i64 [ 0, %header.i ], [ %inc.j, %latch.j ]
br label %header.k
header.k:
%k = phi i64 [ 0, %header.j ], [ %inc.k, %latch.k ]
call void @baz(i64 %i, i64 %j, i64 %k)
br label %latch.k
latch.k:
%inc.k = add nsw i64 %k, 1
%cmp.k = icmp slt i64 %inc.k, %N
br i1 %cmp.k, label %header.k, label %latch.j
latch.j:
%inc.j = add nsw i64 %j, 1
%cmp.j = icmp slt i64 %inc.j, %N
br i1 %cmp.j, label %header.j, label %latch.i
latch.i:
%inc.i = add nsw i64 %i, 1
%cmp.i = icmp slt i64 %inc.i, %N
br i1 %cmp.i, label %header.i, label %exit.i
exit.i:
ret void
}
declare void @baz(i64, i64, i64)
If the blocks of loop-i is in the order: header.i, latch.k, header.k,
header.j, latch.j, latch.i,
then cloneLoopWithPreheader would trigger the assertion in
addBasicBlockToLoop
assert(contains(SameHeader) && getHeader() == SameHeader->getHeader() &&
"Incorrect LI specified for this loop!");
As latch.k is in both loop-j and loop-k, it would be set as the header
of both loops after adding latch.k.
If we update loop headers during cloning blocks, then after adding
header.k,
the header of loop-k would be updated with header.k,
while the header of loop-j stays as latch.k.
When adding header.j, SameHeader is loop-k, SameHeader->getHeader() is
header.k, but getHeader() is latch.k, which trigger the assertion.
Reviewer: jdoerfert, Meinersbur, fhahn, kbarton, hfinkel, bmahjour,
etiotto
Reviewed By: Meinersbur
Subscribers: hiraditya, llvm-commits
Tag: LLVM
Differential Revision: https://reviews.llvm.org/D74382
Add support for DestructiveBinaryComm DestructiveInstType, as well as the lowering code to expand the new Pseudos into the final movprfx+instruction pairs.
Differential Revision: https://reviews.llvm.org/D73711
This should be the last step in the current cleanup.
Follow-ups should resolve the TODO about cost calc
and enable the more general case where we extract
different elements.
This moves all the logic of converting LLVM Triples to
MachO::CPU_(SUB_)TYPE from the specific target (Target)AsmBackend to
more convenient functions in lib/BinaryFormat.
This also gets rid of the separate two X86AsmBackend classes.
The previous attempt was to add it to libObject, but that adds an
unnecessary dependency to libObject from all the targets.
Differential Revision: https://reviews.llvm.org/D74808
isPrefix was added to support the patches to align branches.
it relies on a switch over instruction names.
This moves those opcodes to a new format so the information is
tablegen and we can just check for a specific value in some bits
in TSFlags instead.
I've left the other function in place for now so that the
existing patches in phabricator will still work. I'll work with
the owner to get them migrated.
Summary:
Added register + immediate and register + register addressing modes for the following intrinsics:
1. Masked load and stores:
* Sign and zero extended load and truncated stores.
* No extension or truncation.
2. Masked non-temporal load and store.
Reviewers: andwar, efriedma
Subscribers: cameron.mcinally, sdesmalen, tschuett, kristof.beyls, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D74254
The legalizer helper functions are unusably awkward to perform the 3-5
part legalization. This needs to be widened, scalarized, lowered, and
we should avoid creating vector extends and truncates. Manually do all
of this and expand.
I tried to use some of the new tablegen features to avoid creating
different operand list permutations, but I still don't see a way to
programmatically build a source pattern dag.
Also add GlobalISel tests, which now all import successfully.
Some of the fneg fold tests are incorrect, which need to be fixed in a
future commit
G_SHUFFLE_VECTOR is legal since it theoretically may help match op_sel
for VOP3P instructions. Expand it in some other way in case it doesn't
fold into the use instructions.
Summary:
When we have a long name for the undefined symbol, we would hit this assertion:
Assertion failed: I != StringIndexMap.end() && "String is not in table!"
This patch addresses that.
Reviewed by: DiggerLin, daltenty
Differential Revision: https://reviews.llvm.org/D74924
This fixes a small mistake from D72944: The worklist add should
happen before assigning the new operand, not after.
In case an actual replacement happens, the old operand needs to
be added for DCE. If no actual replacement happens, then old/new
are the same, so it doesn't matter.
This drops one iteration from the annotated test case.
Summary:
This prevents BFI queries on new blocks (from
MachineSinking::GetAllSortedSuccessors) and fixes a bunch of assert failures
under -check-bfi-unknown-block-queries=true.
Reviewers: davidxl
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D74511
Followup to D73919 with another batch of replacements of
setOperand() -> replaceOperand(), to make sure the old
operand gets DCEd right away.
Differential Revision: https://reviews.llvm.org/D74932
ToVectorTy is defined and used in multiple places. Hoist it to
VectorUtils.h to avoid duplication and improve re-usability.
Reviewers: rengolin, hsaito, Ayal, gilr, fpetrogalli
Reviewed By: fpetrogalli
Differential Revision: https://reviews.llvm.org/D74959
This changes the SimplifyLibCalls utility to accept an IRBuilderBase,
which allows us to pass through the IRBuilder used by InstCombine.
This will ensure that new instructions get added to the worklist.
The annotated test-case drops from 4 to 2 InstCombine iterations thanks
to this.
To achieve this, I'm adding an IRBuilderBase::OperandBundlesGuard,
which is basically the same as the existing InsertPointGuard and
FastMathFlagsGuard, but for operand bundles. Also add a
setDefaultOperandBundles() method so these can be set outside the
constructor.
Differential Revision: https://reviews.llvm.org/D74792
A cost query for a vector instruction should return a cost even without
target vector support, and not trigger an assert.
VectorCombine does this with an input containing source code vectors.
Review: Ulrich Weigand
We don't use this, and matching from the def doesn't make much sense.
There are multiple tablegen bugs with default operand
handling. undef_tied_input should work to handle the vdst_in
correctly, but this breaks the operand register class constraint which
it should be able to infer.
Can be used like
-debug-counter=dse-memoryssa-skip=10,dse-memoryssa-counter-count=20
Reviewers: dmgreen, rnk, efriedma, bryant, asbirlea
Reviewed By: asbirlea
Differential Revision: https://reviews.llvm.org/D72147
We should try the generated matchers before the manual selection. This
means the patterns are now handling the common cases, but the manual
selection code is not yet dead. It's still handling the non-s32/s64
cases (like v2s16 and v2s32). Currently tablegen doesn't have a nice
way to have a single pattern that covers multiple types.
We have patterns for s_pack* selection, but they assume the inputs are
a build_vector with 16-bit inputs, not a truncating build
vector. Since there's still outstanding work for how to handle
mismatched result and source element vector operations, and since I'm
trying a different packed vector strategy than SelectionDAG, just
manually select this for now.
There are few differences from the DAG handling. First, the DAG
handling uses a primitive selection pattern instead of custom
legalizing it. Because of this, this makes use of source modifiers
while the DAG does not.
Also instead of promoting f16, try to use the f16 log/exp. There's no
f16 fmul_legacy, so widen just for the multiply, although I'm not sure
that's the best solution.
This looked through copies to find the source modifiers, which may
have been SGPR->VGPR copies added to avoid potential constant bus
violations. Re-insert a copy to a VGPR if this happens.
Remove some cumbersome Darwin specific logic for updating the frame
offsets of the condition-register spill slots. The containing function has an
early return if the subtarget is not ELF based which makes the Darwin logic
dead.
Use the SelectionDAG::getValidShiftAmountConstant helper to get const/constsplat shift amounts, which allows us to drop the out of range shift amount early-out.
First step towards better non-uniform shift amount support in SimplifyDemandedBits.
The (overloaded) intrinsic is llvm.hexagon.V6.pred.typecast[.128B]. The
types of the operand and the return value are HVX boolean vector types.
For each cast, there needs to be a corresponding intrinsic declared,
with different suffixes appended to the name, e.g.
; cast <128 x i1> to <32 x i1>
declare <32 x i1> @llvm.hexagon.V6.pred.typecast.128B.s1(<128 x i1>)
; cast <32 x i1> to <64 x i1>
declare <64 x i1> @llvm.hexagon.V6.pred.typecast.128B.s2(<32 x i1>)
etc.
Summary:
There is a flaw in memory dependence analysis caching mechanism when memory accesses with TBAA are involved. Assume we first analysed and cached results for access with TBAA. Later we request dependence for the same memory but without TBAA (or different TBAA). By design these two queries should share one entry in the internal cache which corresponds to a general access (without TBAA). Thus upon second request internal cached is cleared and we continue analysis for access as if there is no TBAA.
The problem is that even though internal cache is cleared the set of visited nodes is not. That means we won't traverse visited nodes again and populate internal cache with the corresponding dependence results. So we end up with internal cache in an incomplete state. Current implementation tries to signal that situation by resetting CacheInfo->Pair at line 1104. But that doesn't actually help since later code ignores this invalidation and relies on 'Cache->empty()' property to decide on cache completeness.
Reviewers: reames, hfinkel, chandlerc, fedor.sergeev, asbirlea, fhahn, john.brawn, Prazek, sunfish
Reviewed By: john.brawn
Subscribers: DaniilSuchkov, kosarev, jfb, dantrushin, hiraditya, bmahjour, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73032
A question about this behavior came up on llvm-dev:
http://lists.llvm.org/pipermail/llvm-dev/2020-February/139003.html
...and as part of backend improvements in D73978, but this is an IR
change first because we already have fairly thorough tests in place
here.
We decided not to implement a more general change that would have
folded any FP binop with nearly arbitrary constant + undef operand
to undef because that is not theoretically correct (even if it is
practically correct).
Differential Revision: https://reviews.llvm.org/D74713
I believe this was carried over from getELFKindForNamedSection since
the wasm backend originally used ELF object writing as a template.
Differential Revision: https://reviews.llvm.org/D74565
At this point in the code we know that Op1 or Op2 is
all ones. Y points to the other operand. In the case that
Op2 is zero, Op1 must be all ones and Y is Op2. The OR
ORs Y into Res. But if Y is 0 the OR will be folded away by
getNode so we don't need to check for it.
The combineSelect code was casting to i64 without any check that
i64 was legal. This can break after type legalization.
It also required splitting the mmx register on 32-bit targets.
It's not clear that this makes sense. Instead switch to using
a cmov pseudo like we do for XMM/YMM/ZMM.
VK1 was being used as the output of the copy to regclass, but it
should be VK2/VK4. Shouldn't matter in practice though since
VK1/VK2/VK4/VK8/VK16 are all identicaly and just have different VTs.
For tracked globals that are unknown after solving, we expect all
non-store uses to be replaced.
This is a follow-up to f8045b250d, which removed forcedconstant.
We should not mark unknown loads as overdefined, as they either load
from an unknown pointer or an undef global. Restore the original logic
for loads.
This includes both GEPs where the indexed type is a scalable vector, and
GEPs where the result type is a scalable vector.
Differential Revision: https://reviews.llvm.org/D73602
The code at https://reviews.llvm.org/D74808 has broken builds that are
configured with -DBUILD_SHARED_LIBS=On.
This patch adds the correct library dependencies.
When analyzing PHIs, we gather the known bits for every operand and
merge them together to get the known bits of the result of the PHI.
It is not unusual that merging the information leads to know nothing
on the result (e.g., phi a: i8 3, b: i8 unknown, ..., after looking at the
second argument we know we will know nothing on the result), thus, as
soon as we reach that state, stop analyzing the following operand (i.e.,
on the previous example, we won't process anything after looking at `b`).
This improves compile time in particular with PHIs with a large number
of operands.
NFC.
The motivating case is seen in "splat4_v8f32_load_store" and based on code in PR42024:
https://bugs.llvm.org/show_bug.cgi?id=42024
(I haven't stepped through the v8i32 sibling test yet to see why that diverged.)
There are other potential improvements visible like allowing scalarization or vector
narrowing.
Differential Revision: https://reviews.llvm.org/D74909
This moves all the logic of converting LLVM Triples to
MachO::CPU_(SUB_)TYPE from the specific target (Target)AsmBackend to
more convenient functions in libObject.
This also gets rid of the separate two X86AsmBackend classes.
Differential Revision: https://reviews.llvm.org/D74808
There's a lot of old leftover code in LowerBRCOND. Especially
the detecting or AND or OR of X86ISD::SETCC nodes. Those were
needed before LegalizeDAG was changed to visit nodes before
their operands.
It also relied on reversing the output of LowerSETCC to find the
flags producing node to use for the X86ISD::BRCOND node.
Rather than using LowerSETCC this patch uses emitFlagsForSetcc to
handle the integer ISD::SETCC case. This gives the flag producer
and the comparison code to use directly. I've removed the addTest
flag and just produce a X86ISD::BRCOND and return immediately.
Floating point ISD::SETCC case is just an X86ISD::FCMP with special
care for OEQ and UNE derived from the previous code. I've left
f128 out so it will emit a test. And LowerSETCC will be called
later to produce a libcall and X86ISD::SETCC. We have combines
that can merge the test and X86ISD::SETCC.
We need to handle two cases for overflow ops. Either they are used
directly or they have a seteq 0 or setne 1 to invert the overflow.
The old code did not handle the setne 1 case, but I think some
other combines were making up for it.
If we fail to find a condition, we'll wrap an AND with 1 on the
original condition and tell emitFlagsForSetcc to emit a compare
with 0. This will pickup the LowerAndToBT and or the EmitTest case.
I kept the isTruncWithZeroHighBitsInput call, but we might be able
to fold that in to emitFlagsForSetcc.
Differential Revision: https://reviews.llvm.org/D74750
Only handle power of 2 element count for simplicity. Not sure what to do with vXf64->vXf16 fp_round to avoid double rounding
Differential Revision: https://reviews.llvm.org/D74886
Similar to what we already do with SimplifyDemandedVectorElts, call SimplifyDemandedBits across all the extracted elements of the source vector, treating it as single use.
There's a minor regression in store-weird-sizes.ll which will be addressed in an upcoming SimplifyDemandedBits patch.
Marking a section as ALLOC tells the ELF loader to load the section into memory.
As we do not want to load the notes into VRAM, the flag should not be there.
Differential Revision: https://reviews.llvm.org/D74600
GetDemandedBits mostly just calls SimplifyMultipleUseDemandedBits now, but it does a very blunt constant simplification that SimplifyMultipleUseDemandedBits avoids.
If we need to demand bits from constants we should handle this through ShrinkDemandedConstant/targetShrinkDemandedConstant.
@arsenm confirmed that the sign extended immediates are better for code size.
Differential Revision: https://reviews.llvm.org/D74857
Summary:
After updating cost model in AMDGPU target (47a5c36b37) the pass started to
ignore some BBs since they got all instructions estimated as free.
Reviewers: arsenm, chandlerc, nhaehnle
Reviewed By: nhaehnle
Subscribers: jvesely, wdng, nhaehnle, tpr, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D74825
Summary:
This patch adds two families of ACLE intrinsics: vqdmullbq and
vqdmulltq (including vector-vector and vector-scalar variants) and the
corresponding LLVM IR intrinsics llvm.arm.mve.vqdmull and
llvm.arm.mve.vqdmull.predicated.
Reviewers: simon_tatham, MarkMurrayARM, dmgreen, ostannard
Reviewed By: MarkMurrayARM
Subscribers: kristof.beyls, hiraditya, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D74845
Roman Lebedev