This pull request is the second part of an ongoing effort to extends PGO
instrumentation to GPU device code and depends on #76587. This PR makes
the following changes:
- Introduces `__llvm_write_custom_profile` to PGO compiler-rt library.
This is an external function that can be used to write profiles with
custom data to target-specific files.
- Adds `__llvm_write_custom_profile` as weak symbol to libomptarget so
that it can write the collected data to a profraw file.
- Adds `PGODump` debug flag and only displays dump when the
aforementioned flag is set
Summary:
If the user deallocates an RPC device this can sometimes fail if the RPC
server is still running. This will happen if the modification happens
while the server is still checking it. This patch adds a mutex to guard
modifications to it.
Summary:
This patch just changes the interface to make starting the thread
multiple times permissable since it will only be done the first time.
Note that this does not refcount it or anything, so it's onto the user
to make sure that they don't shut down the thread before everyone is
done using it. That is the case today because the shutDown portion is
run by a single thread in the destructor phase.
Another question is if we should make this thread truly global state,
because currently it will be private to each plugin instance, so if you
have an AMD and NVIDIA image there will be two, similarly if you have
those inside of a shared library.
Summary:
Pretty dumb mistake of me, forgot that this is run per-device and
per-plugin, which fell through the cracks with my testing because I have
two GPUs that use different plugins.
Summary:
Handling the RPC server requires running through list of jobs that the
device has requested to be done. Currently this is handled by the thread
that does the waiting for the kernel to finish. However, this is not
sound on NVIDIA architectures and only works for async launches in the
OpenMP model that uses helper threads.
However, we also don't want to have this thread doing work
unnnecessarily. For this reason we track the execution of kernels and
cause the thread to sleep via a condition variable (usually backed by
some kind of futex or other intelligent sleeping mechanism) so that the
thread will be idle while no kernels are running.
Summary:
This patch is an NFC renaming to make using the offloading entry type
more portable between other targets. Right now this is just moving its
definition to LLVM so others can use it. Future work will rework the
struct layout.
Exposed by -Warray-bounds:
In file included from
../../../../../../../llvm/offload/plugins-nextgen/common/src/GlobalHandler.cpp:252:
../../../../../../../llvm/llvm/include/llvm/ProfileData/InstrProfData.inc:109:1:
error: array index 4 is past the end of the array (that has type 'const
std::remove_const<const uint16_t>::type[4]' (aka 'const unsigned
short[4]')) [-Werror,-Warray-bounds]
109 | INSTR_PROF_DATA(const uint16_t, Int16ArrayTy,
NumValueSites[IPVK_Last+1], \
| ^ ~~~~~~~~~~~
../../../../../../../llvm/offload/plugins-nextgen/common/src/GlobalHandler.cpp:250:15:
note: expanded from macro 'INSTR_PROF_DATA'
250 | outs() << ProfData.Name << " "; \
| ^ ~~~~
../../../../../../../llvm/llvm/include/llvm/ProfileData/InstrProfData.inc:109:1:
note: array 'NumValueSites' declared here
109 | INSTR_PROF_DATA(const uint16_t, Int16ArrayTy,
NumValueSites[IPVK_Last+1], \
| ^
../../../../../../../llvm/offload/plugins-nextgen/common/include/GlobalHandler.h:62:3:
note: expanded from macro 'INSTR_PROF_DATA'
62 | std::remove_const<Type>::type Name;
Avoid accessing out-of-bound data, but skip printing array data for now.
As there is no simple way to do this without hardcoding the
NumValueSites field.
---------
Co-authored-by: Ethan Luis McDonough <ethanluismcdonough@gmail.com>
This adds support for the loongarch64 architecture to the offload host
plugin.
Similar to #115773
To fix some test issues, I've had to add the LoongArch64 target to:
- CompilerInvocation::ParseLangArgs
- linkDevice in ClangLinuxWrapper.cpp
- OMPContext::OMPContext (to set the device_kind_cpu trait)
Reviewed By: jhuber6
Pull Request: https://github.com/llvm/llvm-project/pull/120173
Enables generic ISA, e.g., "--offload-arch=gfx11-generic" device code to
run on gfx11-generic ISA capable device.
Executable may contain one ELF that has specific target ISA and another
ELF that has compatible generic ISA.
Under that circumstance, this code should say both ELFs are compatible,
leaving the rest to PluginManager to handle.
Suggestions on how best to address that is welcome.
Reland #118503. Added a fix for builds with `-DBUILD_SHARED_LIBS=ON`
(see last commit). Otherwise the changes are identical.
---
### New API
Previous discussions at the LLVM/Offload meeting have brought up the
need for a new API for exposing the functionality of the plugins. This
change introduces a very small subset of a new API, which is primarily
for testing the offload tooling and demonstrating how a new API can fit
into the existing code base without being too disruptive. Exact designs
for these entry points and future additions can be worked out over time.
The new API does however introduce the bare minimum functionality to
implement device discovery for Unified Runtime and SYCL. This means that
the `urinfo` and `sycl-ls` tools can be used on top of Offload. A
(rough) implementation of a Unified Runtime adapter (aka plugin) for
Offload is available
[here](https://github.com/callumfare/unified-runtime/tree/offload_adapter).
Our intention is to maintain this and use it to implement and test
Offload API changes with SYCL.
### Demoing the new API
```sh
# From the runtime build directory
$ ninja LibomptUnitTests
$ OFFLOAD_TRACE=1 ./offload/unittests/OffloadAPI/offload.unittests
```
### Open questions and future work
* Only some of the available device info is exposed, and not all the
possible device queries needed for SYCL are implemented by the plugins.
A sensible next step would be to refactor and extend the existing device
info queries in the plugins. The existing info queries are all strings,
but the new API introduces the ability to return any arbitrary type.
* It may be sensible at some point for the plugins to implement the new
API directly, and the higher level code on top of it could be made
generic, but this is more of a long-term possibility.
A fair number of fixes to get standalone builds of offload working —
mostly copying missing bits from openmp. It's almost ready — I still
need to figure out why some of the tsts aren't linking to the right
libraries.
This is another attempt to reland the changes from #108413
The previous two attempts introduced regressions and were reverted. This
PR has been more thoroughly tested with various configurations so
shouldn't cause any problems this time. If anyone is aware of any likely
remaining problems then please let me know.
The changes are identical other than the fixes contained in the last 5
commits.
---
### New API
Previous discussions at the LLVM/Offload meeting have brought up the
need for a new API for exposing the functionality of the plugins. This
change introduces a very small subset of a new API, which is primarily
for testing the offload tooling and demonstrating how a new API can fit
into the existing code base without being too disruptive. Exact designs
for these entry points and future additions can be worked out over time.
The new API does however introduce the bare minimum functionality to
implement device discovery for Unified Runtime and SYCL. This means that
the `urinfo` and `sycl-ls` tools can be used on top of Offload. A
(rough) implementation of a Unified Runtime adapter (aka plugin) for
Offload is available
[here](https://github.com/callumfare/unified-runtime/tree/offload_adapter).
Our intention is to maintain this and use it to implement and test
Offload API changes with SYCL.
### Demoing the new API
```sh
# From the runtime build directory
$ ninja LibomptUnitTests
$ OFFLOAD_TRACE=1 ./offload/unittests/OffloadAPI/offload.unittests
```
### Open questions and future work
* Only some of the available device info is exposed, and not all the
possible device queries needed for SYCL are implemented by the plugins.
A sensible next step would be to refactor and extend the existing device
info queries in the plugins. The existing info queries are all strings,
but the new API introduces the ability to return any arbitrary type.
* It may be sensible at some point for the plugins to implement the new
API directly, and the higher level code on top of it could be made
generic, but this is more of a long-term possibility.
Summary:
This patch adds an RPC interface that lives directly in the OpenMP
device runtime. This allows OpenMP to implement custom opcodes.
Currently this is only providing the host call interface, which is the
raw version of reverse offloading. Previously this lived in `libc/` as
an extension which is not the correct place.
The interface here uses a weak symbol for the RPC client by the same
name that the `libc` interface uses. This means that it will defer to
the libc one if both are present so we don't need to set up multiple
instances.
The presense of this symbol is what controls whether or not we set up
the RPC server. Because this is an external symbol it normally won't be
optimized out, so there's a special pass in OpenMPOpt that deletes this
symbol if it is unused during linking. That means at `O0` the RPC server
will always be present now, but will be removed trivially if it's not
used at O1 and higher.
Relands #117704, which relanded changes from #108413 - this was reverted
due to build issues. The new offload library did not build with
`LIBOMPTARGET_OMPT_SUPPORT` enabled, which was not picked up by
pre-merge testing.
The last commit contains the fix; everything else is otherwise identical
to the approved PR.
___
### New API
Previous discussions at the LLVM/Offload meeting have brought up the
need for a new API for exposing the functionality of the plugins. This
change introduces a very small subset of a new API, which is primarily
for testing the offload tooling and demonstrating how a new API can fit
into the existing code base without being too disruptive. Exact designs
for these entry points and future additions can be worked out over time.
The new API does however introduce the bare minimum functionality to
implement device discovery for Unified Runtime and SYCL. This means that
the `urinfo` and `sycl-ls` tools can be used on top of Offload. A
(rough) implementation of a Unified Runtime adapter (aka plugin) for
Offload is available
[here](https://github.com/callumfare/unified-runtime/tree/offload_adapter).
Our intention is to maintain this and use it to implement and test
Offload API changes with SYCL.
### Demoing the new API
```sh
# From the runtime build directory
$ ninja LibomptUnitTests
$ OFFLOAD_TRACE=1 ./offload/unittests/OffloadAPI/offload.unittests
```
### Open questions and future work
* Only some of the available device info is exposed, and not all the
possible device queries needed for SYCL are implemented by the plugins.
A sensible next step would be to refactor and extend the existing device
info queries in the plugins. The existing info queries are all strings,
but the new API introduces the ability to return any arbitrary type.
* It may be sensible at some point for the plugins to implement the new
API directly, and the higher level code on top of it could be made
generic, but this is more of a long-term possibility.
Summary:
We should now use the official™ way to include the files from
`libc/shared`. This required some code to make sure that it's not
included twice if multiple people use it as well as a sanity check on
the directory.
Summary:
We can simply include this header from the shared directory now and do
not need to have this level of indirection. Simply stash it with the
other libc opcode handlers.
If we were able to move the printf handlers to the shared directory then
this could just be a header as well, which would HEAVILY simplify the
mess associated with building the RPC server first in the projects
build, then copying it to the runtimes build.
Summary:
We currently have an unnecessary level of indirection when initializing
the RPC client. This is a holdover from when the RPC client was not
trivially copyable and simply makes it more complicated. Here we use the
`asm` syntax to give the C++ variable a valid name so that we can just
copy to it directly.
Another advantage to this, is that if users want to piggy-back on the
same RPC interface they need only declare theirs as extern with the same
symbol name, or make it weak to optionally use it if LIBC isn't
avaialb.e
Relands changes from #108413 - this was reverted due to build issues.
The problem was just that the `offload-tblgen` tool was behind recent
changes to tablegen that ensure `const` records. This has been fixed and
the PR is otherwise identical.
___
### New API
Previous discussions at the LLVM/Offload meeting have brought up the
need for a new API for exposing the functionality of the plugins. This
change introduces a very small subset of a new API, which is primarily
for testing the offload tooling and demonstrating how a new API can fit
into the existing code base without being too disruptive. Exact designs
for these entry points and future additions can be worked out over time.
The new API does however introduce the bare minimum functionality to
implement device discovery for Unified Runtime and SYCL. This means that
the `urinfo` and `sycl-ls` tools can be used on top of Offload. A
(rough) implementation of a Unified Runtime adapter (aka plugin) for
Offload is available
[here](https://github.com/callumfare/unified-runtime/tree/offload_adapter).
Our intention is to maintain this and use it to implement and test
Offload API changes with SYCL.
### Demoing the new API
```sh
# From the runtime build directory
$ ninja LibomptUnitTests
$ OFFLOAD_TRACE=1 ./offload/unittests/OffloadAPI/offload.unittests
```
### Open questions and future work
* Only some of the available device info is exposed, and not all the
possible device queries needed for SYCL are implemented by the plugins.
A sensible next step would be to refactor and extend the existing device
info queries in the plugins. The existing info queries are all strings,
but the new API introduces the ability to return any arbitrary type.
* It may be sensible at some point for the plugins to implement the new
API directly, and the higher level code on top of it could be made
generic, but this is more of a long-term possibility.
Introduce `offload-tblgen` and an initial implementation of a subset of
the new API. The tablegen files are intended to be the single source of
truth for the new API, with the header files, documentation, and others
bits of source all automatically generated.
**TODO** (based on review feedback so far):
- [x] Check in the generated headers
- [x] Add an `offload-generate` target to trigger the generation rather
than building them every time
- [x] Decide how error handling should work
- [x] Finish up new error handling implementation
- [x] Decide naming convention
- [x] Add testing for the new API
- [x] Add tablegen specific testing
- [x] clang-tidy and use llvm:: types when possible
- [x] Add optional code location arguments
- [x] Avoid multiple returns from one function
### offload-tblgen
See the included
[README](d80db06491/offload/new-api/API/README.md)
for more information on how the API definition and generation works. I'm
happy to answer any questions about it and plan to walk through it in a
future LLVM Offload call.
It should be noted that struct definitions have not been fully
implemented/tested as they aren't used by the initial API definitions,
but finishing that off in the future shouldn't be too much work.
The tablegen tooling has been designed to be easily extended with new
backends, using the classes in `RecordTypes.hpp` to abstract over the
tablegen records.
### New API
Previous discussions at the LLVM/Offload meeting have brought up the
need for a new API for exposing the functionality of the plugins. This
change introduces a very small subset of a new API, which is primarily
for testing the offload tooling and demonstrating how a new API can fit
into the existing code base without being too disruptive. Exact designs
for these entry points and future additions can be worked out over time.
The new API does however introduce the bare minimum functionality to
implement device discovery for Unified Runtime and SYCL. This means that
the `urinfo` and `sycl-ls` tools can be used on top of Offload. A
(rough) implementation of a Unified Runtime adapter (aka plugin) for
Offload is available
[here](https://github.com/callumfare/unified-runtime/tree/offload_adapter).
Our intention is to maintain this and use it to implement and test
Offload API changes with SYCL.
### Demoing the new API
```sh
$ git clone -b offload_adapter https://github.com/callumfare/unified-runtime.git
$ cd unified-runtime
$ mkdir build
$ cd build
$ cmake .. -GNinja -DUR_BUILD_ADAPTER_OFFLOAD=ON \
-DUR_OFFLOAD_INSTALL_DIR=<offload build dir containing liboffload_new.so> \
-DUR_OFFLOAD_INCLUDE_DIR=<offload build dir containing 'offload' headers directory>
$ ninja urinfo
export LD_LIBRARY_PATH=<offload build dir containing offload plugin libraries>
$ UR_ADAPTERS_FORCE_LOAD=$PWD/lib/libur_adapter_offload.so ./bin/urinfo
[cuda:gpu][cuda:0] CUDA, NVIDIA GeForce GT 1030 [12030]
# Demo with tracing
$ OFFLOAD_TRACE=1 UR_ADAPTERS_FORCE_LOAD=$PWD/lib/libur_adapter_offload.so ./bin/urinfo
---> offloadPlatformGet(.NumEntries = 0, .phPlatforms = {}, .pNumPlatforms = 0x7ffd05e4d6e0 (2))-> OFFLOAD_RESULT_SUCCESS
---> offloadPlatformGet(.NumEntries = 2, .phPlatforms = {0x564bf4040220, 0x564bf4040240}, .pNumPlatforms = nullptr)-> OFFLOAD_RESULT_SUCCESS
...
```
### Open questions and future work
* The new API is implemented in a separate library
(`liboffload_new.so`). It could just as easily be part of the existing
`libomptarget` library - I have no strong feelings on which is better.
* Only some of the available device info is exposed, and not all the
possible device queries needed for SYCL are implemented by the plugins.
A sensible next step would be to refactor and extend the existing device
info queries in the plugins. The existing info queries are all strings,
but the new API introduces the ability to return any arbitrary type.
* It may be sensible at some point for the plugins to implement the new
API directly, and the higher level code on top of it could be made
generic, but this is more of a long-term possibility.
Summary:
This patch removes much of the `llvmlibc_rpc_server` interface. This
pretty much deletes all of this code and just replaces it with including
`rpc.h` directly. We still maintain the file to let `libc` handle the
opcodes, since those depend on the `printf` impelmentation.
This will need to be cleaned up more, but I don't want to put too much
into a single patch.
Following discussions in #110443, and the following earlier discussions
in https://lists.llvm.org/pipermail/llvm-dev/2017-October/117907.html,
https://reviews.llvm.org/D38482, https://reviews.llvm.org/D38489, this
PR attempts to overhaul the `TargetMachine` and `LLVMTargetMachine`
interface classes. More specifically:
1. Makes `TargetMachine` the only class implemented under
`TargetMachine.h` in the `Target` library.
2. `TargetMachine` contains target-specific interface functions that
relate to IR/CodeGen/MC constructs, whereas before (at least on paper)
it was supposed to have only IR/MC constructs. Any Target that doesn't
want to use the independent code generator simply does not implement
them, and returns either `false` or `nullptr`.
3. Renames `LLVMTargetMachine` to `CodeGenCommonTMImpl`. This renaming
aims to make the purpose of `LLVMTargetMachine` clearer. Its interface
was moved under the CodeGen library, to further emphasis its usage in
Targets that use CodeGen directly.
4. Makes `TargetMachine` the only interface used across LLVM and its
projects. With these changes, `CodeGenCommonTMImpl` is simply a set of
shared function implementations of `TargetMachine`, and CodeGen users
don't need to static cast to `LLVMTargetMachine` every time they need a
CodeGen-specific feature of the `TargetMachine`.
5. More importantly, does not change any requirements regarding library
linking.
cc @arsenm @aeubanks
This adds support for the riscv64 architecture to the offload host
plugin. The check to define FFI_DEFAULT_ABI is intentionally not guarded
by __riscv_xlen as the value is the same for riscv32 and riscv64
(support for OpenMP on riscv32 is still under review).
We had three `utils::` namespaces, all with different "meaning" (host,
device, hsa_utils). We should, when we can, keep "include/Shared"
accessible from host and device, thus RefCountTy has been moved to a
separate header. `hsa_utils` was introduced to make `utils::` less
overloaded. And common functionality was de-duplicated, e.g.,
`utils::advance` and `utils::advanceVoidPtr` -> `utils:advancePtr`. Type
punning now checks for the size of the result to make sure it matches
the source type.
No functional change was intended.
This pull request is a revised version of #76587. This pull request
fixes some build issues that were present in the previous version of
this change.
> This pull request is the first part of an ongoing effort to extends
PGO instrumentation to GPU device code. This PR makes the following
changes:
>
> - Adds blank registration functions to device RTL
> - Gives PGO globals protected visibility when targeting a supported
GPU
> - Handles any addrspace casts for PGO calls
> - Implements PGO global extraction in GPU plugins (currently only
dumps info)
>
> These changes can be tested by supplying `-fprofile-instrument=clang`
while targeting a GPU.
Since we can already track allocations, we can diagnose memory faults to
some degree. If the fault happens in a prior allocation (use after free)
or "close but outside" one, we can provide that information to the user.
Note that the fault address might be page aligned, and not all accesses
trigger a fault, especially for allocations that are backed by a
MemoryManager. Still, if people disable the MemoryManager or the
allocation is big enough, we can sometimes provide valueable feedback.
Error: CommandLine Error: Option 'attributor-manifest-internal'
registered more than once
During the standalone debug build of offload the above error is seen at
app runtime when using a prebuilt llvm with LLVM_LINK_LLVM_DYLIB=ON.
This is caused by linking both libLLVM.so and various archives that are
found via llvm_map_components_to_libnames for jit support.
Through the new `-foffload-via-llvm` flag, CUDA kernels can now be
lowered to the LLVM/Offload API. On the Clang side, this is simply done
by using the OpenMP offload toolchain and emitting calls to `llvm*`
functions to orchestrate the kernel launch rather than `cuda*`
functions. These `llvm*` functions are implemented on top of the
existing LLVM/Offload API.
As we are about to redefine the Offload API, this wil help us in the
design process as a second offload language.
We do not support any CUDA APIs yet, however, we could:
https://www.osti.gov/servlets/purl/1892137
For proper host execution we need to resurrect/rebase
https://tianshilei.me/wp-content/uploads/2021/12/llpp-2021.pdf
(which was designed for debugging).
```
❯❯❯ cat test.cu
extern "C" {
void *llvm_omp_target_alloc_shared(size_t Size, int DeviceNum);
void llvm_omp_target_free_shared(void *DevicePtr, int DeviceNum);
}
__global__ void square(int *A) { *A = 42; }
int main(int argc, char **argv) {
int DevNo = 0;
int *Ptr = reinterpret_cast<int *>(llvm_omp_target_alloc_shared(4, DevNo));
*Ptr = 7;
printf("Ptr %p, *Ptr %i\n", Ptr, *Ptr);
square<<<1, 1>>>(Ptr);
printf("Ptr %p, *Ptr %i\n", Ptr, *Ptr);
llvm_omp_target_free_shared(Ptr, DevNo);
}
❯❯❯ clang++ test.cu -O3 -o test123 -foffload-via-llvm --offload-arch=native
❯❯❯ llvm-objdump --offloading test123
test123: file format elf64-x86-64
OFFLOADING IMAGE [0]:
kind elf
arch gfx90a
triple amdgcn-amd-amdhsa
producer openmp
❯❯❯ LIBOMPTARGET_INFO=16 ./test123
Ptr 0x155448ac8000, *Ptr 7
Ptr 0x155448ac8000, *Ptr 42
```
The kernel names for OpenMP are manually mangled and not ideal when we
report something to the user. We demangle them now, providing the
function and line number of the target region, together with the actual
kernel name.
Similar to (de)allocation traces, we can record kernel launch stack
traces and display them in case of an error. However, the AMD GPU plugin
signal handler, which is invoked on memroy faults, cannot pinpoint the
offending kernel. Insteade print `<NUM>`, set via
`OFFLOAD_TRACK_NUM_KERNEL_LAUNCH_TRACES=<NUM>`, many traces. The
recoding/record uses a ring buffer of fixed size (for now 8).
For `trap` errors, we print the actual kernel name, and trace if
recorded.
As a first step towards a GPU sanitizer we now can track allocations and
deallocations in order to report double frees, and other problems during
deallocation.
