When multiplying several AddRecs, we do the following simplification:
{A1,+,A2,+,...,+,An}<L> * {B1,+,B2,+,...,+,Bn}<L> = {x=1 in [ sum y=x..2x
[ sum z=max(y-x, y-n)..min(x,n) [
choose(x, 2x)*choose(2x-y, x-z)*A_{y-z}*B_z]]
],+,...up to x=2n}
This is done iteratively, pair by pair. So if we try to multiply three AddRecs
A1, A2, A3, then we'd try to simplify A1 * A2 to A1' and then try to
simplify A1' * A3 if A1' is also an AddRec.
The transform is only legal if the loops of the two AddRecs are the same.
It is checked in the code, but the loop of one of the AddRecs is stored
in a local variable and doesn't get updated when we simplify a pair to a new AddRec.
In the motivating test the new AddRec A1' was created for a different loop and,
as the loop variable didn't get updated, the check for different loops passed and
the transform worked for two AddRecs from different loops.
So it created a wrong SCEV. And it caused LSR to replace an instruction with another
one that had the same SCEV as the incorrectly computed one.
Differential Revision: https://reviews.llvm.org/D153254
The old code used to materialize constants as ops, immediately folded them into the resulting affine map and then deleted the constant ops again. Instead, directly fold the attributes into the affine map. Furthermore, all helpers accept `OpFoldResult` instead of `Value` now. This makes the code at call sites more efficient, because it is no longer necessary to materialize a `Value`, just to be able to use these helper functions.
Note: The API has changed (accepts OpFoldResult instead of Value), otherwise this change is NFC.
Differential Revision: https://reviews.llvm.org/D153324
We only check a subset of the constraints in the verifier:
* that we only call the intrinsic from functions with a restricted set of
calling conventions
* that the 'flags' argument is an immediate
Other checks are (probably) more appropriate for codegen.
Differential Revision: https://reviews.llvm.org/D151995
Add the amdgpu_cs_chain and amdgpu_cs_chain_preserve keywords to
LLVM IR and make sure we can parse and print them. Also make sure we
perform some basic checks in the IR verifier - similar to what we check
for many of the other AMDGPU calling conventions, plus the additional
restriction that we can't have direct calls to functions with these
calling conventions.
Differential Revision: https://reviews.llvm.org/D151994
Currently, all relocations that point inside a function are registered
as external references. If these relocations cannot be resolved as jump
tables or computed gotos, the containing function gets marked as
not-simple and excluded from optimizations.
RISC-V uses relocations for branches and jumps (to support linker
relaxation) and as such, almost no functions get marked as simple. This
patch fixes this by only registering relocations that originate outside
of the referenced function as external references.
Reviewed By: rafauler
Differential Revision: https://reviews.llvm.org/D153345
Thispatch implements the R_RISCV_ADD32 and R_RISCV_SUB32 relocations for
RISC-V.
Reviewed By: rafauler
Differential Revision: https://reviews.llvm.org/D146554
Previously if no matching multilib was found then the user would
typically see an error like "fatal error: 'stdio.h' file not found"
which gives no indication as to the underlying problem.
With this change the user will instead see an error like
clang: error: no multilib found matching flags: --target=thumbv7em-none-unknown-eabi -march=...
clang: note: available multilibs are:
--target=armv4t-none-unknown-eabi
--target=thumbv6m-none-unknown-eabi -mfpu=none
...
Differential Revision: https://reviews.llvm.org/D153292
This revision ensures SwitchOps with case and condition
bitwidths other than 32-bit are imported properly. It adds an
APInt based builder to the SwitchOp and implements
a verifier that checks that the condition and the case
value types match.
Reviewed By: Dinistro
Differential Revision: https://reviews.llvm.org/D153438
When LLVM_ENABLE_PER_TARGET_RUNTIME_DIR is enabled, place headers
in `include/<target>` directory, otherwise use `include/`.
Differential Revision: https://reviews.llvm.org/D152592
This change also prepares for 9119325 to land again.
Adds `mlir_c_runner_utils_hdrs` and `mlir_runner_utils_hdrs` targets which do not depend on `//llvm::Support`.
These can be used by other 'runner.so' targets if they are loaded along with the 'runner_utils.so' without calling `__mlir_execution_engine_init()` twice.
Remove the `DenseMapInfo<std::variant<Ts...>>` variant out from
`llvm/ADT/DenseMapInfo.h` into a separate header
`llvm/ADT/DenseMapInfoVariant.h`
This allows us to remove the `<variant>` include, which is being
transitively and unncessary included in all translation units that
include `llvm/ADT/DenseMap.h`.
There have been similar changes to move out specializations for
* `APInt.h` fd7e309e02 and
* `StringRef.h`/`ArrayRef.h`
983565a6fe
to reduce the compilation time. As we are unable to move the
specialization into `<variant>`, we create a separate
`DenseMapInfoVariant.h` header that can be used by anyone who needs this
specialization.
This reduces the total number of preprocessing tokens across the LLVM
source files in lib from (roughly) 1,964,876,961 to 1,936,551,496 - a
reduction of ~1.44%. This should result in a small improvement in
compilation time.
Differential Revision: https://reviews.llvm.org/D150997
Add ability to be able to group ops together in documentation generation. This
results in section of ops grouped together under one summary & description.
This doesn't do anything special for the groups beyond nesting yet.
Differential Revision: https://reviews.llvm.org/D153306
vfredmax/min may have different scheduling behavior on different
microarchitectures compared to other reductions. This different
behavior can be described by giving vfredmax/min separate
scheduling resources.
Differential Revision: https://reviews.llvm.org/D153450
Old pattern was missing some cases (e.g. swapping the arguments)
but it also allowed too many cases (e.g. non-empty "absent" or
different arguments for add/mul). This fixes the issues.
Reviewed By: K-Wu
Differential Revision: https://reviews.llvm.org/D153487
The word "attribute" has a specific meaning in LLVM. Avoid using it
here to mean something different.
This addresses feedback from D153180.
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D153444
It hasn't been written to since https://reviews.llvm.org/D46652, so it
was always empty. I don't have enough context on that diff to know if
the removal of the write to ShowIncludesPretendHeader in that diff was
intentional, but no one's complained about it for five years, so I
assume we're okay to just get rid of it entirely.
Reviewed By: hans
Differential Revision: https://reviews.llvm.org/D153176
`-H` displays a tree of included header files, but that tree is supposed
to omit two categories of header files:
1. Any header files pulled in via `-include`, which the code refers to
as the "predefines".
2. Any header files whose inclusion was skipped because they'd already
been included (assuming header guards or `#pragma once`).
`-fshow-skipped-includes` was intended to make `-H` display the second
category of files. It wasn't checking for the first category, however,
so you could end up with only the middle of the `-include` hierarchy
displayed, e.g. the added test would previously output:
```
... /data/users/smeenai/llvm-project/clang/test/Frontend/Inputs/test2.h
. /data/users/smeenai/llvm-project/clang/test/Frontend/Inputs/test.h
```
This diff adds a check to prevent that and correctly omit headers from
`-include` even when `-fshow-skipped-includes` is passed. While I'm
here, add tests for the interaction between `-fshow-skipped-includes`
and `-sys-header-deps` as well.
Reviewed By: hans
Differential Revision: https://reviews.llvm.org/D153175
Add support for couple of FIR types such as fir.ptr, fir.heap,
fir.box, fir.class
Reviewed By: razvanlupusoru
Differential Revision: https://reviews.llvm.org/D153461
The ScopedString class has two functions named append. One takes
a va_list, but on some platforms va_list is typedef'd to char*.
That means that this call:
std::string value;
Str.append("print this string %s", value.c_str());
The compiler can incorrectly think this is the va_list function,
leading to crashes when calling this. To fix this, change the name
of the va_list function to be vappend to avoid this.
Fix https://github.com/llvm/llvm-project/issues/62893
Reviewed By: Chia-hungDuan
Differential Revision: https://reviews.llvm.org/D153389
Previously, it wouldn't take into account files in ignore_format.txt
(at least not on OSX) because the `find` command would return file names
like `libcxx/src//new_handler.cpp`, which never matched the file names
in `ignore_format.txt`.
Differential Revision: https://reviews.llvm.org/D153416
Two new entitlements, com.apple.private.thread-set-state and
com.apple.private.set-exception-port should be included in the
debugserver entitlement plist when built internally at Apple.
The desktop builds of debugserver don't need these entitlements;
don't add them to debugserver-macosx-entitlements.plist.
rdar://108912676
rdar://103032208
This doesn't change the selection, but it expands the conditions to
add comments and make it clearer what's happening. It also removes a
-Wundef instance when we checked __ARM_ARCH_7K__ >= 2 without checking
that it is defined in the first place.
Differential Revision: https://reviews.llvm.org/D153413
Commit ec77747fbd regenerated the check lines
without being very careful about which lines were updated. This attempts to fix
them to make sure the V7 and V8 lines are emitted as needed.
For all other compute and data constructs, the data operands list
is named `dataClauseOperands`. Update `acc.host_data` to be
consistent with this naming.
Reviewed By: clementval
Differential Revision: https://reviews.llvm.org/D153425
Lower private and firstprivate operands through their corresponding
data entry operation to support array section.
Depends on D152972
Reviewed By: razvanlupusoru
Differential Revision: https://reviews.llvm.org/D152974
acc.firstprivate operation will be used as data entry operation
for the firstprivate operands.
Depends on D152970
Reviewed By: razvanlupusoru
Differential Revision: https://reviews.llvm.org/D152972
acc.private operation will be used as data entry operation
for the private operands.
Reviewed By: razvanlupusoru
Differential Revision: https://reviews.llvm.org/D152970
