Up until Power9, the performance profile for rlwinm., rldicl. and andi. looked
more or less equivalent. However with Power9, the rotates are still 2-way
cracked whereas the and-immediate is not.
This patch just ensures that we don't emit record-form rotates when an andi.
is adequate.
As first pointed out by Carrot in https://bugs.llvm.org/show_bug.cgi?id=30833
(this patch is a fix for that PR).
Differential Revision: https://reviews.llvm.org/D43977
llvm-svn: 326736
Summary:
Presently, InstCombiner::foldICmpWithCastAndCast() implicitly assumes that it is
only invoked with icmp instructions of integer type. If that assumption is broken,
and it is called with an icmp of vector type, then it fails (asserts/crashes).
This patch addresses the deficiency. It allows it to simplify
icmp (ptrtoint x), (ptrtoint/c) of vector type into a compare of the inputs,
much as is done when the type is integer.
Reviewers: apilipenko, fedor.sergeev, mkazantsev, anna
Reviewed By: anna
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D44063
llvm-svn: 326730
This patch teaches getMinimumFPType to support shrinking a vector of ConstantFPs. This should improve our ability to combine vector fptrunc with fp binops.
Differential Revision: https://reviews.llvm.org/D43774
llvm-svn: 326729
Despite building cleanly on my machine in three separate configs, it's failing on pretty much all bots due to missing includes among other things. Investigating.
llvm-svn: 326726
Summary:
It can be useful for tools to be able to retrieve the values of variables
declared via STATISTIC() directly without having to emit them and parse
them back. Use cases include:
* Needing to report specific statistics to a test harness
* Wanting to post-process statistics. For example, to produce a percentage of
functions that were fully selected by GlobalISel
Make this possible by adding llvm::GetStatistics() which returns an
iterator_range that can be used to inspect the statistics that have been
touched during execution. When statistics are disabled (NDEBUG and not
LLVM_ENABLE_STATISTICS) this method will return an empty range.
This patch doesn't address the effect of multiple compilations within the same
process. In such situations, the statistics will be cumulative for all
compilations up to the GetStatistics() call.
Reviewers: qcolombet, rtereshin, aditya_nandakumar, bogner
Reviewed By: rtereshin, bogner
Subscribers: llvm-commits, mgorny
Differential Revision: https://reviews.llvm.org/D43901
llvm-svn: 326723
The error occurs when reading i16 elements (as in the testcase) from a v8i8
with a pattern of <0,2,4,6>. As all the data in the vector is accessed, the
operation is not a VUZP. The patch stops the pattern recognition of VUZP when
EXTRACT_VECTOR_ELT has a different element type than BUILD_VECTOR.
llvm-svn: 326722
getCompare returns true, false or undef constants if the comparison can
be evaluated, or nullptr if it cannot. This is in line with what
ConstantExpr::getCompare returns. It also allows us to use
ConstantExpr::getCompare for comparing constants.
Reviewers: davide, mssimpso, dberlin, anna
Reviewed By: davide
Differential Revision: https://reviews.llvm.org/D43761
llvm-svn: 326720
Use the whole gammut of constant immediates available to set up a vector.
Instead of using, for example, `mov w0, #0xffff; dup v0.4s, w0`, which
transfers between register files, use the more efficient `movi v0.4s, #-1`
instead. Not limited to just a few values, but any immediate value that can
be encoded by all the variants of `FMOV`, `MOVI`, `MVNI`, thus eliminating
the need to there be patterns to optimize special cases.
Differential revision: https://reviews.llvm.org/D42133
llvm-svn: 326718
This is a NFC simple patch that changes the DEBUG dumping in the
MachineScheduler so that the dumping of the built SUnits is done before the
SchedImpl->initialize() is called.
This is better on SystemZ, since it has a strategy that does some dumping at
the start of the region, and it is not possible to easily read it if it is
output above a long list of SU.
Review: Javed Absar
https://reviews.llvm.org/D44089
llvm-svn: 326716
Summary:
Use the new resolver interface more explicitly, and avoid traversing
all the initializers multiple times.
Change-Id: I679e86988b309d19f25e6cca8b0b14ea150198a6
Reviewers: arsenm, craig.topper, tra, MartinO
Subscribers: wdng, llvm-commits
Differential Revision: https://reviews.llvm.org/D43654
llvm-svn: 326708
Summary:
Use the new resolver interface more explicitly, and avoid traversing
all the initializers multiple times.
Change-Id: Ia4dcc6d42dd8b65e6079d318c6a202f36f320fee
Reviewers: arsenm, craig.topper, tra, MartinO
Subscribers: wdng, llvm-commits
Differential Revision: https://reviews.llvm.org/D43653
llvm-svn: 326707
Summary:
Use the new resolver interface more explicitly, and avoid traversing
all the initializers multiple times.
Add a test case for a pattern that was broken by an earlier version
of this change.
An additional change is that we now remove *all* template arguments
after resolving them.
Change-Id: I86c828c8cc84c18b052dfe0f64c0d5cbf3c4e13c
Reviewers: arsenm, craig.topper, tra, MartinO
Subscribers: wdng, llvm-commits
Differential Revision: https://reviews.llvm.org/D43652
llvm-svn: 326706
Summary:
This changes the syntax of !foreach so that the first "parameter" is
a new syntactic variable: !foreach(x, lst, expr) will define the
variable x within the scope of expr, and evaluation of the !foreach
will substitute elements of the given list (or dag) for x in expr.
Aside from leading to a nicer syntax, this allows more complex
expressions where x is deeply nested, or even constant expressions
in which x does not occur at all.
!foreach is currently not actually used anywhere in trunk, but I
plan to use it in the AMDGPU backend. If out-of-tree targets are
using it, they can adjust to the new syntax very easily.
Change-Id: Ib966694d8ab6542279d6bc358b6f4d767945a805
Reviewers: arsenm, craig.topper, tra, MartinO
Subscribers: wdng, llvm-commits, tpr
Differential Revision: https://reviews.llvm.org/D43651
llvm-svn: 326705
Summary:
The intention is to allow us to more easily restructure how resolving is
done, e.g. resolving multiple variables simultaneously, or using the
resolving mechanism to implement !foreach.
Change-Id: I4b976b54a32e240ad4f562f7eb86a4d663a20ea8
Reviewers: arsenm, craig.topper, tra, MartinO
Subscribers: wdng, llvm-commits
Differential Revision: https://reviews.llvm.org/D43564
llvm-svn: 326704
Summary:
NAME has already worked for def in a multiclass, since the (protoype)
record including its NAME variable is created before parsing the
superclasses. Since defm's do not have an associated single record,
support for NAME has to be implemented differently here.
Original test cases provided by Artem Belevich (tra)
Change-Id: I933b74f328c0ff202e7dc23a35b78f3505760cc9
Reviewers: arsenm, craig.topper, tra, MartinO
Subscribers: wdng, llvm-commits
Differential Revision: https://reviews.llvm.org/D43656
llvm-svn: 326700
Summary:
We can discard initial blocks that do other work
We do not need to limit ourselves to just the first block in the chain.
Reviewers: courbet, davide
Reviewed By: courbet
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D44029
llvm-svn: 326698
We now check relocations offsets are within range, and the relocation
index is valid.
Also updated tests which contained invalid Wasm files that were
previously not checked.
Differential Revision: https://reviews.llvm.org/D43684
llvm-svn: 326697
These instructions require that the two S registers are adjacent (but not the R
registers), because only the first register is included in the encoding, but we
were not checking this in the assembler.
Differential revision: https://reviews.llvm.org/D44084
llvm-svn: 326696
This is required in order to enable relocs to be validated
as they are read in.
Also update tests with new section ordering.
Differential Revision: https://reviews.llvm.org/D43940
llvm-svn: 326694
This allows LLD to print the name for an InputGlobal when encountering
an error.
Differential Revision: https://reviews.llvm.org/D44033
llvm-svn: 326691
Previously, we passed "#" to --autocomplete to indicate to enable cc1
flags. For example, when -cc1 or -Xclang was passed to bash, bash
executed `clang --autocomplete=#-<flag they want to complete>`.
However, this was not a good implementation because it depends -Xclang
and -cc1 parsing to shell. So I changed this to pass all flags shell
has, so that Clang can handle them internally.
I had to change many testcases because API spec changed quite a lot.
Reviewers: teemperor, v.g.vassilev
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D39342
llvm-svn: 326684
Almost none of these usages were FP specific. And we had no clear guideliness on when to use hasAVX vs hasFP256.
I might also remove hasInt256 too since its an alias for hasAVX2.
llvm-svn: 326682
Loading a constant into a k-register in AVX512 requires a bitcast from a scalar constant. In the test case here we have a k-register store that gets split into multiple parts of KNL. MergeConsecutiveStores sees each of these pieces as a consecutive store and looks through the bitcast to find the underly scalar constant. But when we went to create the combined store we didn't look through the same bitcast.
llvm-svn: 326677
rL322525 - mmx zero constant support
rL322553 - mmx i32 zero extended value
rL326497 - mmx i64 general constant handling
Not all constants are folded, we generate some on the GPRs (similar to SSE build vector) where appropriate
llvm-svn: 326673
We were previously doing this with isel patterns. Moving it to op legalization gives us chance to see the required bitcast earlier. And it lets us remove some isel patterns.
llvm-svn: 326669
Iterating through predecessors of `TailBB` while removing their
terminators leads to use after-free, because the predecessor list is
changing on each removal.
llvm-svn: 326668
Summary:
`musttail` calls can't be naively splitted. The split blocks must
include not only the call instruction itself, but also (optional)
`bitcast` and `return` instructions that follow it.
Clone `bitcast` and `ret`, place them into the split blocks, and
remove the tail block when done.
Reviewers: junbuml, mcrosier, davidxl, davide, fhahn
Reviewed By: fhahn
Subscribers: JDevlieghere, llvm-commits
Differential Revision: https://reviews.llvm.org/D43729
llvm-svn: 326666
This caused some links to fail with ThinLTO due to missing symbols as
well as causing some binaries to have failures at runtime. We're working
with the author to get a test case, but want to get the tree green
again.
Further, it appears to introduce a data race. While the test usage of
threads was disabled in r325361 & r325362, that isn't an acceptable fix.
I've reverted both of these as well. This code needs to be thread safe.
Test cases for this are already on the original commit thread.
llvm-svn: 326638
X86 considers v1i1 a legal type under AVX512 and as such a truncate from a v1iX type to v1i1 can be turned into a scalar truncate plus a conversion to v1i1. We would much prefer a v1i1 SCALAR_TO_VECTOR over a one element BUILD_VECTOR.
During lowering we were detecting the v1i1 BUILD_VECTOR as a splat BUILD_VECTOR like we try to do for v2i1/v4i1/etc. In this case we create (select i1 splat_elt, v1i1 all-ones, v1i1 all-zeroes). That goes through some more legalization and we end up with a CMOV choosing between 0 and 1 in scalar and a scalar_to_vector.
Arguably we could detect the v1i1 BUILD_VECTOR and do this better in X86 target code. But just using a SCALAR_TO_VECTOR in legalization is much easier.
llvm-svn: 326637
The fast/linear DAG scheduler doesn't lower DBG_VALUEs except for
function entry nodes.
Patch by Joshua Cranmer!
Differential Revision: https://reviews.llvm.org/D43028
llvm-svn: 326631
Summary:
The symbolizer was checking for .debug as a subdirectory of the
binary file itself, not of the directory containing the binary. This led to
a failure to find split debug info when it was contained in a .debug directory.
Reviewers: rnk, glider, zturner
Subscribers: llvm-commits, aprantl
Differential Revision: https://reviews.llvm.org/D44025
llvm-svn: 326630
In stage2 -O3 builds of llc, this results in small but measurable
increases in the number of variables with locations, and in the number
of unique source variables overall.
(According to llvm-dwarfdump --statistics, there are 123 additional
variables with locations, which is just a 0.006% improvement).
The size of the .debug_loc section of the llc dsym increases by 0.004%.
llvm-svn: 326629
Nemanja Ivanovic