Implements:
- P2291R3 Add Constexpr Modifiers to Functions to_chars and from_chars for
Integral Types in <charconv> Header
Reviewed By: #libc, ldionne
Differential Revision: https://reviews.llvm.org/D131317
This partly reverts D133535 and enables CTAD for more parts in format.
Reviewed By: ldionne, #libc
Differential Revision: https://reviews.llvm.org/D135292
(X << Z) / (Y << Z) --> X / Y
https://alive2.llvm.org/ce/z/CLKzqT
This requires a surprising "nuw" constraint because we have
to guard against immediate UB via signed-div overflow with
-1 divisor.
This extends 008a89037a and is another transform
derived from issue #58137.
This version is more composable and also simpler at the expense of being more explicit and more verbose.
This patch is not meant to be submitted but gives an idea of the change.
Codegen can be checked in https://godbolt.org/z/6z1dEoWbs by removing the "static inline" before individual functions.
Unittests are coming.
Suggested review order:
- utils
- op_base
- op_builtin
- op_generic
- op_x86 / op_aarch64
- *_implementations.h
Differential Revision: https://reviews.llvm.org/D135134
Need to set the insertpoint for extractelement to point to the first
instruction in the node to avoid possible crash during external uses
combine process. Without it we may endup with the incorrect
transformation.
Differential Revision: https://reviews.llvm.org/D135591
(X << Z) / (Y << Z) --> X / Y
https://alive2.llvm.org/ce/z/E5eaxU
This fixes the motivating example from issue #58137,
but it is not the most general transform. We should
probably also convert left-shift in the divisor to
right-shift in the dividend for that, but that exposes
another missed canonicalization for shifts and adds.
Freeze instruction in some cases makes codegen worse, so need to be very
careful when emitting it. Instead improve analysis in isUndefVector
function to generate mask of unused elements and use it in the analysis.
Differential Revision: https://reviews.llvm.org/D135382
Linalg ops are generally lowered to loops using a pass and the individual application
of patterns has no concrete uses.
If/when necessary, this can later be reintroduced via a proper transform dialect op.
Differential Revision: https://reviews.llvm.org/D135766
Root cause for the regression is that value %6598 is
defined as a label, and later as an i32 value.
Reviewed By: uweigand
Differential Revision: https://reviews.llvm.org/D135778
Polymorphic entities with allocatable or pointer attribute cannot be passed
directly as passed-object when the type-bound procedure is expecting a
simply dummy polymorphic entity. Use fir.rebox to adapt the fir.class box
to the tbp type.
Depends on D135649
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D135654
One constant generated in MemRefToLLVM had a hardcoded bitwidth of
64 bits. The fix uses the typeConverter to create a constant that
matches the bitwidth of the provided by the data layout. The issue was
detected in an attempt to add a verifier to the LLVM ICmp operation that
checks that the types of the compared arguments match.
Reviewed By: ftynse
Differential Revision: https://reviews.llvm.org/D135775
However, mark them as EXCLUDE_FROM_ALL when we don't want to build them.
Simply declaring the targets should be of no harm, and it allows other
projects to mention these targets regardless of whether they end up
being built or not.
While the diff may not make that obvious, this patch basically
moves the definition of e.g. `cxx_shared` out of the `if (LIBCXX_ENABLE_SHARED)`
and instead marks it as `EXCLUDE_FROM_ALL` conditionally on whether
LIBCXX_ENABLE_SHARED is passed. It then does the same for libunwind
and libc++abi targets.
Differential Revision: https://reviews.llvm.org/D134221
Lower call with polymorphic entities to fir.dispatch operation. This patch only
focus one lowering with simple scalar polymorphic entities. A follow-up patch
will deal with allocatble, pointer and array of polymorphic entities as they
require box manipulation for the passed-object.
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D135649
If the AM* atomic memory access instruction has the same register number as
rd and rj, the execution will trigger an Instruction Non-defined Exception.
If the AM* atomic memory access instruction has the same register number as
rd and rk, the execution result is uncertain.
Reference: https://github.com/loongson/LoongArch-Documentation
Differential Revision: https://reviews.llvm.org/D135641
Before this change, the `NoReturnFunctionChecker` was missing function pointers
with a `[[noreturn]]` attribute, while `CFG` was constructed taking that into
account, which leads CSA to take impossible paths. The reason was that the
`NoReturnFunctionChecker` was looking for the attribute in the type of the
entire call expression rather than the type of the function being called.
This change makes the `[[noreturn]]` attribute of a function pointer visible
to `NoReturnFunctionChecker`. This leads to a more coherent behavior of the
CSA on the AST involving.
Reviewed By: xazax.hun
Differential Revision: https://reviews.llvm.org/D135682
Fix two typos and remove the incorrectly added `extraClassDeclaration` in
`FIRTypes.td`.
Reviewed By: clementval
Differential Revision: https://reviews.llvm.org/D135665
This change fixes two issues in ValueObject::GetExpressionPath method:
1. Accessing members of struct references used to produce expression
paths such as "str.&str.member" (instead of the expected
"str.member"). This is fixed by assigning the flag tha the child
value is a dereference when calling Dereference() on references
and adjusting logic in expression path creation.
2. If the parent of member access is dereference, the produced
expression path was "*(ptr).member". This is incorrect, since it
dereferences the member instead of the pointer. This is fixed by
wrapping dereference expression into parenthesis, resulting with
"(*ptr).member".
Reviewed By: werat, clayborg
Differential Revision: https://reviews.llvm.org/D132734
This will allow to model elementwise ops with this special op instead of using
GenericOp.
Also allow MapOp and ReduceOp to have no result if the output type is not a tensor.
This is needed for buffer semantics.
Differential Revision: https://reviews.llvm.org/D135754
optimizeInductions may leave dead recipes which can prevent sinking.
Sinking on the other hand should not introduce new dead recipes, so
clean up dead recipes before sinking.
Reviewed By: Ayal
Differential Revision: https://reviews.llvm.org/D133762
The revision uses tablegen generated builders to convert the most common
LLVM IR instructions to MLIR LLVM dialect operations. All instructions
with special handlers, except for alloca and fence, still use manual
handlers. The revision also introduces an additional "instructions.ll"
test file to test the import of instructions that have tablegen builders
(except for the resume instruction whose test remains untouched). A part
of the test cases are new, for example the integer instruction test,
while others are migrated from the "basic.ll" test file.
Reviewed By: ftynse
Differential Revision: https://reviews.llvm.org/D135709
They have been scattered over the code. For better structuring, perform
them in one place. Potential CT drop is possible because we collect exit
blocks twice, but it's small price to pay for much better code structure.
Fix crash issue of D129537 and reopen it.
Currently the X86 shuffle lowering would widen the element type for
shuffle if the mask element value is adjacent. For below example
%t2 = add nsw <16 x i32> %t0, %t1
%t3 = sub nsw <16 x i32> %t0, %t1
%t4 = shufflevector <16 x i32> %t2, <16 x i32> %t3,
<16 x i32> <i32 16, i32 17, i32 2, i32 3, i32 4,
i32 5, i32 6, i32 7, i32 8, i32 9, i32 10,
i32 11, i32 12, i32 13, i32 14, i32 15>
ret <16 x i32> %t4
Compiler would transform the shuffle to
%t4 = shufflevector <8 x i64> %t2, <8 x i64> %t3,
<8 x i64> <i32 8, i32 1, i32 2, i32 3, i32 4,
i32 5, i32 6, i32 7>
This may lose the oppotunity to let ISel select mask instruction when
avx512 is enabled.
This patch is to prevent the tranform when avx512 feature is enabled.
Thank Simon for the idea.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D130830
The test dexter/feature_tests/commands/perfect/expect_step_kind/direction.cpp
was failing on a machine because __libc_start_call_main was not identified as a
"frame below main" (a frame we don't want to gather information from), causing
dexter to count one more step than expected in the test.
Add __libc_start_call_main to the list of "frames below main". There may
be a more robust way of doing this but this is a pragmatic solution we can
use for now.
The previous calculations seem to have assumed that the section address would be zero.
This is true for relocatable object files, but certainly not for linked files like shared libraries.
Fixed the calculations to make them identical to the "real" `getInstruction` call below & added a regression test.
Reviewed By: scott.linder, simon_tatham
Differential Revision: https://reviews.llvm.org/D135430
Mark de Wever