This patch introduces an `OperatingSystem` base implementation in the
`lldb` python module to make it easier for lldb users to write their own
implementation.
The `OperatingSystem` base implementation is derived itself from the
`ScriptedThread` base implementation since they share some common grounds.
To achieve that, this patch makes changes to the `ScriptedThread`
initializer since it gets called by the `OperatingSystem` initializer.
I also took the opportunity to document the `OperatingSystem` base
class and methods.
Differential Revision: https://reviews.llvm.org/D159315
Signed-off-by: Med Ismail Bennani <ismail@bennani.ma>
This patch aims to consolidate the OperatingSystem scripting affordance
by introducing a stable interface that conforms to the
Scripted{,Python}Interface.
This unify the way we call into python methods from lldb while
also improving its capabilities by allowing us to pass lldb_private
objects are arguments.
Differential Revision: https://reviews.llvm.org/D159314
Signed-off-by: Med Ismail Bennani <ismail@bennani.ma>
- For instance, `collectPGOFuncNameStrings` is reused a lot in https://github.com/llvm/llvm-project/pull/66825 to get the compressed vtable names; and in some added callsites it's just confusing to see 'func' since context clearly shows it's not. This function currently just takes a list of strings as input so name it to `collectGlobalObjectNameStrings`
- Do the rename in a standalone patch since the method is used in non-llvm codebase. It's easier to rollback this NFC in case rename in that codebase takes longer.
The runtime needs to know about the acceptable launch bounds, especially
if the compiler (middle- or backend) assumed those bounds. While this
patch does not yet inform the runtime, it stores the bounds in a place
that can/will be accessed and is associated with the kernel.
We now provide the information about the min/max thread and team count
from to the OMPIRBuilder, no matter what the source was. That means we
unify `thread_limit`, `num_teams`, `num_threads` handling with the
target specific attriutes (`__launch_bounds__` and
`amdgpu_flat_work_group_size`). This is in preparation to pass the
values to the runtime, and to allow the middle-end (OpenMP-opt) to
tighten the values if it seems appropriate. There is no "real" change
after this commit.
The workaround code ensure we always call __kmpc_kernel_parallel, but it
did so in a racy manner as the initialization might not have been
completed yet. To avoid introducing a sync, we move the workaround into
the deinit function for now.
-mcmodel= is supported for a few architectures. Reject the option for
other architectures.
* -mcmodel= is unsupported on x86-32.
* -mcmodel=large is unsupported for PIC on AArch64.
* -mcmodel= is unsupported for aarch64_32 triples.
* https://reviews.llvm.org/D67066 (for RISC-V) made
-mcmodel=medany/-mcmodel=medlow aliases for all architectures. Restrict
this to RISC-V.
* llvm/lib/Target/Sparc has some small/medium/large support, but the
values listed on https://gcc.gnu.org/onlinedocs/gcc/SPARC-Options.html
had been supported before https://reviews.llvm.org/D67066. Consider
-mcmodel= unsupported for Sparc.
* https://reviews.llvm.org/D106371 translated -mcmodel=medium to
-mcmodel=large on AIX, even for 32-bit systems. Retain this behavior but
reject -mcmodel= for other PPC32 systems.
In general the accept/reject behavior is more similar to GCC.
err_drv_invalid_argument_to_option is less clear than
err_drv_unsupported_option_argument. As the supported values are
different for
different architectures, add a
err_drv_unsupported_option_argument_for_target
for better clarity.
It's a drop in the ocean considering the lack of documentation of our
diagnostics, but it's a start.
Co-authored-by: serge-sans-paille <sguelton@mozilla.com>
Improve diagnostic message to be more straight forward, fix handling of
casting to non-void and add new option AllowCastToVoid to control
casting
to void behavior.
Closes#66570
Fixed assertion failure
Basic Block in function 'main' does not have terminator!
label %land.end
caused by premature setting of CodeGenIP upon entry to
emitTargetDataCalls, where subsequent evaluation of logical expression
created new basic blocks, leaving CodeGenIP pointing to the wrong basic
block. CodeGenIP is now set near the end of the function, just prior to
generating a comparison of the logical expression result (from the if
clause) which uses CodeGenIP to insert new IR.
Adds an end-to-end test for `vector.contract` that targets SVE (i.e.
scalable vectors). Note that this requires lifting the restriction on
`vector.outerproduct` (to which `vector.contract` is lowered to) that
would deem the following as invalid by the Op verifier (*):
```
vector.outerproduct %27, %28, %26 {kind = #vector.kind<add>} : vector<3xf32>, vector<[2]xf32>
```
This is indeed valid as the end-to-end test demonstrates (at least when
compiling for SVE).
Add tosa-to-linalg-pipeline that calls the function
addTosaToLinalgPasses, so it gets tested in core
also added tests in tosa-to-linalg-pipeline.mlir
Signed-off-by: Tai Ly <tai.ly@arm.com>
This seems to be causing issues with using overloaded RVV intrinsics
that take scalar operands. If the scalar type passed in doesn't exactly
match the element type.
I blindly copied this feature from SVE. Since no one has asked for it
I'd prefer to remove it to make overloaded intrinsics work as expected.
By removing the lax conversions, types declared with __attribute__((riscv_rvv_vector_bits(__riscv_v_fixed_vlen)))
can only ever be used like their underlying RVV builtin type. No lax conversions
to other element sizes with the same LMUL.
Fixes#64404.
There are many tests that specify a target triple/CPU flags but no
DataLayout which can lead to IR being generated that has unusual
behaviour. This commit attempts to use the default DataLayout based
on the relevant flags if there is no explicit override on the command
line or in the IR file.
One thing that is not currently possible to differentiate from a missing
datalayout `target datalayout = ""` in the IR file since the current
APIs don't allow detecting this case. If it is considered useful to
support this case (instead of passing "-data-layout=" on the command
line), I can change IR parsers to track whether they have seen such a
directive and change the callback type.
Differential Revision: https://reviews.llvm.org/D141060
- FreeBSD removed big-endian arm with 12.0.
- OpenBSD never had big-endian arm support. I added it just in case, but it has
never been used.
- Remove sparcel bits. It was sprinkled in a few places but it will never be a
thing.
- Remove 32-bit sparc bits for FreeBSD. FreeBSD has never had 32-bit sparc
support.
- Remove sparc64 IAS test as support was enabled across the board awhile ago.
As discussed in [[Driver] Link Flang runtime on Solaris](https://github.com/llvm/llvm-project/pull/65644), `clang -r`
incorrectly passes both `-Bdynamic` and `-e _start` to `ld` which lets the linker choke.
This patch fixes this, omitting `-Bdynamic` completely which is the linker default.
Tested on `amd64-pc-solaris2.11` and `sparcv9-sun-solaris2.11`.
These are an artifact of how types are structured but serve little
purpose, merely showing that the type is sugared in some way. For
example, ElaboratedType's existence means struct S gets printed as
'struct S':'struct S' in the AST, which is unnecessary visual clutter.
Note that skipping the second print when the types have the same string
matches what we do for diagnostics, where the aka will be skipped.
https://github.com/llvm/llvm-project/pull/69824 added libc build, but
missed the folder in ninja command, is causing failures.
ninja: fatal: chdir to 'docs-libc-html' - No such file or directory
ninja: Entering directory `docs-libc-html'
The pre-index matcher just needs some small heuristics to make sure it
doesn't cause regressions. Apart from that it's a simple change, since
the only difference is an immediate operand of '1' vs '0' in the
instruction.
Some textual editing errors got through this pull request that was
merged a few weeks ago: https://github.com/llvm/llvm-project/pull/65876
This patch clears up the unintentional duplicated line, and white-space
at the end of the lines.
