Change from original commit: move test (that uses an X86 triple) into the X86
subdirectory.
Original description:
Gating vectorizing reductions on *all* fastmath flags seems unnecessary;
`reassoc` should be sufficient.
Reviewers: tvvikram, mkuper, kristof.beyls, sdesmalen, Ayal
Reviewed By: sdesmalen
Subscribers: dcaballe, huntergr, jmolloy, mcrosier, jlebar, bixia, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D57728
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355889 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
Swift now generates PDBs for debugging on Windows. llvm and lldb
need a language enumerator value too properly handle the output
emitted by swiftc.
Subscribers: jdoerfert, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D59231
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355882 91177308-0d34-0410-b5e6-96231b3b80d8
Fixes https://bugs.llvm.org/show_bug.cgi?id=36796.
Implement basic legalizations (PromoteIntRes, PromoteIntOp,
ExpandIntRes, ScalarizeVecOp, WidenVecOp) for VECREDUCE opcodes.
There are more legalizations missing (esp float legalizations),
but there's no way to test them right now, so I'm not adding them.
This also includes a few more changes to make this work somewhat
reasonably:
* Add support for expanding VECREDUCE in SDAG. Usually
experimental.vector.reduce is expanded prior to codegen, but if the
target does have native vector reduce, it may of course still be
necessary to expand due to legalization issues. This uses a shuffle
reduction if possible, followed by a naive scalar reduction.
* Allow the result type of integer VECREDUCE to be larger than the
vector element type. For example we need to be able to reduce a v8i8
into an (nominally) i32 result type on AArch64.
* Use the vector operand type rather than the scalar result type to
determine the action, so we can control exactly which vector types are
supported. Also change the legalize vector op code to handle
operations that only have vector operands, but no vector results, as
is the case for VECREDUCE.
* Default VECREDUCE to Expand. On AArch64 (only target using VECREDUCE),
explicitly specify for which vector types the reductions are supported.
This does not handle anything related to VECREDUCE_STRICT_*.
Differential Revision: https://reviews.llvm.org/D58015
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355860 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
Extract the functionality of eliminating unreachable basic blocks
within a function, previously encapsulated within the
-unreachableblockelim pass, and make it available as a function within
BlockUtils.h. No functional change intended other than making the logic
reusable.
Exposing this logic makes it easier to implement
https://reviews.llvm.org/D59068, which fixes coroutines bug
https://bugs.llvm.org/show_bug.cgi?id=40979.
Reviewers: mkazantsev, wmi, davidxl, silvas, davide
Reviewed By: davide
Subscribers: llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D59069
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355846 91177308-0d34-0410-b5e6-96231b3b80d8
AMDGPU target run out of Subtarget feature flags hitting the limit of 64.
AssemblerPredicates uses at most uint64_t for their representation.
At the same time CodeGen has exhausted this a long time ago and switched
to a FeatureBitset with the current limit of 192 bits.
This patch completes transition to the bitset for feature bits extending
it to asm matcher and MC code emitter.
Differential Revision: https://reviews.llvm.org/D59002
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355839 91177308-0d34-0410-b5e6-96231b3b80d8
Narrow Scalar G_MUL for MIPS32.
Revisit NarrowScalar implementation in LegalizerHelper.
Introduce new helper function multiplyRegisters.
It performs generic multiplication of values held in multiple registers.
Generated instructions use only types NarrowTy and i1.
Destination can be same or two times size of the source.
Differential Revision: https://reviews.llvm.org/D58824
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355814 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
Binary formats often include various enumerations or bitsets, but using
endian-specific types for accessing them is tricky because they
currently only support integral types. This is particularly true for
scoped enums (enum class), as these are not implicitly convertible to
integral types, and so one has to perform two casts just to read the
enum value.
This fixes that support by adding first-class support for enumeration
types to endian-specific types. The support for them was already almost
working -- all I needed to do was overload getSwappedBytes for
enumeration types (which casts the enum to its underlying type and performs the
conversion there). I also add some convenience template aliases to simplify
declaring endian-specific enums.
Reviewers: Bigcheese, zturner
Subscribers: kristina, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D59141
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355812 91177308-0d34-0410-b5e6-96231b3b80d8
Includes a fix to emit a CheckOpcode for build_vector when immAllZerosV/immAllOnesV is used as a pattern root. This means it can't be used to look through bitcasts when used as a root, but that's probably ok. This extra CheckOpcode will ensure that the first match in the isel table will be a SwitchOpcode which is needed by the caching optimization in the ISel Matcher.
Original commit message:
Previously we had build_vector PatFrags that called ISD::isBuildVectorAllZeros/Ones. Internally the ISD::isBuildVectorAllZeros/Ones look through bitcasts, but we aren't able to take advantage of that in isel. Instead of we have to canonicalize the types of the all zeros/ones build_vectors and insert bitcasts. Then we have to pattern match those exact bitcasts.
By emitting specific matchers for these 2 nodes, we can make isel look through any bitcasts without needing to explicitly match them. We should also be able to remove the canonicalization to vXi32 from lowering, but I've left that for a follow up.
This removes something like 40,000 bytes from the X86 isel table.
Differential Revision: https://reviews.llvm.org/D58595
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355784 91177308-0d34-0410-b5e6-96231b3b80d8
InstructionSimplify currently has some code to determine the constant
range of integer instructions for some simple cases. It is used to
simplify icmps.
This change moves the relevant code into ValueTracking as
llvm::computeConstantRange(), so it can also be reused for other
purposes.
In particular this is with the optimization of overflow checks in
mind (ref D59071), where constant ranges cover some cases that
known bits don't.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355781 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
Right now, when we encounter a string equality check,
e.g. `if (memcmp(a, b, s) == 0)`, we try to expand to a comparison if `s` is a
small compile-time constant, and fall back on calling `memcmp()` else.
This is sub-optimal because memcmp has to compute much more than
equality.
This patch replaces `memcmp(a, b, s) == 0` by `bcmp(a, b, s) == 0` on platforms
that support `bcmp`.
`bcmp` can be made much more efficient than `memcmp` because equality
compare is trivially parallel while lexicographic ordering has a chain
dependency.
Subscribers: fedor.sergeev, jyknight, ckennelly, gchatelet, llvm-commits
Differential Revision: https://reviews.llvm.org/D56593
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355672 91177308-0d34-0410-b5e6-96231b3b80d8
to get the modules bots running again.
The LLVM_DEBUG macro only plays well with a modular build of LLVM when
the header is marked as textual, but doing so causes redefinition
errors.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355653 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
This adds support for 64 bit buffer atomic arithmetic instructions but does not include
cmpswap as that depends on a fix to the way the register pairs are handled
Change-Id: Ib207ea65fb69487ccad5066ea647ae8ddfe2ce61
Subscribers: arsenm, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, jfb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D58918
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355520 91177308-0d34-0410-b5e6-96231b3b80d8
This allows us to store more info about where we're emitting the remarks
without cluttering LLVMContext. This is needed for future support for
the remark section.
Differential Revision: https://reviews.llvm.org/D58996
Original llvm-svn: 355507
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355514 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
This allows generic code to query these parameters, and is a common
practice in a lot of other template classes.
Reviewers: zturner, Bigcheese
Subscribers: kristina, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D58969
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355504 91177308-0d34-0410-b5e6-96231b3b80d8
Some target might try to subclass the PostGenericScheduler to custom the scheduling strategy.
We need to declare the member data of PostGenericScheduler as "protected", which acts the same as "GenericScheduler".
Differential Revision: https://reviews.llvm.org/D58949
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355475 91177308-0d34-0410-b5e6-96231b3b80d8
Getting rid of the name "optimization remarks" for anything that
involves handling remarks on the client side.
It's safer to do this now, before we get stuck with that name in all the
APIs and public interfaces we decide to export to users in the future.
This renames llvm/tools/opt-remarks to llvm/tools/remarks-shlib, and now
generates `libRemarks.dylib` instead of `libOptRemarks.dylib`.
Differential Revision: https://reviews.llvm.org/D58535
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355439 91177308-0d34-0410-b5e6-96231b3b80d8
This caused the first matcher in the isel table for many targets to Opc_Scope instead of Opc_SwitchOpcode. This leads to a significant increase in isel match failures.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355433 91177308-0d34-0410-b5e6-96231b3b80d8
These arrays are both keyed by CPU name and go into the same tablegenerated file. Merge them so we only need to store keys once.
This also removes a weird space saving quirk where we used the ProcDesc.size() to create to build an ArrayRef for ProcSched.
Differential Revision: https://reviews.llvm.org/D58939
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355431 91177308-0d34-0410-b5e6-96231b3b80d8
The description for CPUs was just the CPU name wrapped with "Select the " and " processor". We can just do that directly in the help printer instead of making a separate version in the binary for each CPU.
Also remove the Value field that isn't needed and was always 0.
Differential Revision: https://reviews.llvm.org/D58938
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355429 91177308-0d34-0410-b5e6-96231b3b80d8
The SubtargetFeature class managed a list of features as strings. And it also had functions for setting bits in a FeatureBitset.
The methods that operated on the Feature list as strings are used in other parts of the backend. But the parts that operate on FeatureBitset are very tightly coupled to MCSubtargetInfo and requires passing in the arrays that MCSubtargetInfo owns. And the same struct type is used for ProcFeatures and ProcDesc.
This has led to MCSubtargetInfo having 2 arrays keyed by CPU name. One containing a mapping from a CPU name to its features. And one containing a mapping from CPU name to its scheduler model.
I would like to make a single CPU array containing all CPU information and remove some unneeded fields the ProcDesc array currently has. But I don't want to make SubtargetFeatures.h have to know about the scheduler model type and have to forward declare or pull in the header file.
Differential Revision: https://reviews.llvm.org/D58937
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355428 91177308-0d34-0410-b5e6-96231b3b80d8
Currently one can concatenate strings using hash(#),
but not lists, although that would be a natural thing to do.
This patch allows one to write something like:
def : A<!listconcat([1,2], [3,4])>;
simply as :
def : A<[1,2] # [3,4]>;
This was missing feature was highlighted by Nicolai
at FOSDEM talk.
Reviewed by: nhaehnle, hfinkel
Differential Revision: https://reviews.llvm.org/D58895
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355414 91177308-0d34-0410-b5e6-96231b3b80d8
This patch adds a new flag named -bottleneck-analysis to print out information
about throughput bottlenecks.
MCA knows how to identify and classify dynamic dispatch stalls. However, it
doesn't know how to analyze and highlight kernel bottlenecks. The goal of this
patch is to teach MCA how to correlate increases in backend pressure to backend
stalls (and therefore, the loss of throughput).
From a Scheduler point of view, backend pressure is a function of the scheduler
buffer usage (i.e. how the number of uOps in the scheduler buffers changes over
time). Backend pressure increases (or decreases) when there is a mismatch
between the number of opcodes dispatched, and the number of opcodes issued in
the same cycle. Since buffer resources are limited, continuous increases in
backend pressure would eventually leads to dispatch stalls. So, there is a
strong correlation between dispatch stalls, and how backpressure changed over
time.
This patch teaches how to identify situations where backend pressure increases
due to:
- unavailable pipeline resources.
- data dependencies.
Data dependencies may delay execution of instructions and therefore increase the
time that uOps have to spend in the scheduler buffers. That often translates to
an increase in backend pressure which may eventually lead to a bottleneck.
Contention on pipeline resources may also delay execution of instructions, and
lead to a temporary increase in backend pressure.
Internally, the Scheduler classifies instructions based on whether register /
memory operands are available or not.
An instruction is marked as "ready to execute" only if data dependencies are
fully resolved.
Every cycle, the Scheduler attempts to execute all instructions that are ready
to execute. If an instruction cannot execute because of unavailable pipeline
resources, then the Scheduler internally updates a BusyResourceUnits mask with
the ID of each unavailable resource.
ExecuteStage is responsible for tracking changes in backend pressure. If backend
pressure increases during a cycle because of contention on pipeline resources,
then ExecuteStage sends a "backend pressure" event to the listeners.
That event would contain information about instructions delayed by resource
pressure, as well as the BusyResourceUnits mask.
Note that ExecuteStage also knows how to identify situations where backpressure
increased because of delays introduced by data dependencies.
The SummaryView observes "backend pressure" events and prints out a "bottleneck
report".
Example of bottleneck report:
```
Cycles with backend pressure increase [ 99.89% ]
Throughput Bottlenecks:
Resource Pressure [ 0.00% ]
Data Dependencies: [ 99.89% ]
- Register Dependencies [ 0.00% ]
- Memory Dependencies [ 99.89% ]
```
A bottleneck report is printed out only if increases in backend pressure
eventually caused backend stalls.
About the time complexity:
Time complexity is linear in the number of instructions in the
Scheduler::PendingSet.
The average slowdown tends to be in the range of ~5-6%.
For memory intensive kernels, the slowdown can be significant if flag
-noalias=false is specified. In the worst case scenario I have observed a
slowdown of ~30% when flag -noalias=false was specified.
We can definitely recover part of that slowdown if we optimize class LSUnit (by
doing extra bookkeeping to speedup queries). For now, this new analysis is
disabled by default, and it can be enabled via flag -bottleneck-analysis. Users
of MCA as a library can enable the generation of pressure events through the
constructor of ExecuteStage.
This patch partially addresses https://bugs.llvm.org/show_bug.cgi?id=37494
Differential Revision: https://reviews.llvm.org/D58728
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355308 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
Before when we implemented the first EH proposal, 'catch <tag>'
instruction may not catch an exception so there were multiple EH pads an
exception can unwind to. That means a BB could have multiple EH pad
successors.
Now after we switched to the new proposal, every 'catch' instruction
catches an exception, and there is only one catchpad per catchswitch, so
we at most have one EH pad successor, making `ThrowUnwindDest` map in
`WasmEHInfo` unnecessary.
Keeping `ThrowUnwindDest` map in `WasmEHInfo` has its own problems,
because other optimization passes can split a BB that contains possibly
throwing calls (previously invokes), and we have to update the map every
time that happens, which is not easy for common CodeGen passes.
This also correctly updates successor info in LateEHPrepare when we add
a rethrow instruction.
Reviewers: dschuff
Subscribers: sbc100, jgravelle-google, sunfish, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D58486
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@355296 91177308-0d34-0410-b5e6-96231b3b80d8
Sanjoy Das