This patch adds `olGetEventElapsedTime` to the new LLVM Offload API, as
requested in
[#185728](https://github.com/llvm/llvm-project/issues/185728), and adds
the corresponding support in `plugins-nextgen`.
A main motivation for this change is to make it possible to measure the
elapsed time of work submitted to a queue, especially kernel launches.
This is relevant to the intended use of the new Offload API for
microbenchmarking GPU libc math functions.
### Summary
The new API returns the elapsed time, in milliseconds, between two
events on the same device.
To support the common pattern `create start event → enqueue kernel →
create end event → sync end event → get elapsed time`, `olCreateEvent`
now always creates and records a backend event through the device
interface. For backends that materialize real event state, this gives
the event concrete backend state that can be used for elapsed-time
measurement. For backends that do not materialize backend event state,
`EventInfo` may still remain null and existing event operations continue
to treat such events as trivially complete.
Previously, an event created on an empty queue could be represented only
as a logical event. That representation was sufficient for sync and
completion queries, but it was not suitable for elapsed-time measurement
because there was no backend event state to timestamp. The new behavior
preserves the meaning of completion of prior work while also allowing
backends with timing support to attach real event state.
### Changes in `plugins-nextgen`
#### Common interface
Add elapsed-time support to the common device and plugin interfaces:
* `GenericPluginTy::get_event_elapsed_time`
* `GenericDeviceTy::getEventElapsedTime`
* `GenericDeviceTy::getEventElapsedTimeImpl`
#### AMDGPU
* Add the required ROCr declarations and wrappers.
* Enable queue profiling at queue creation time.
* Record events by enqueuing a real barrier marker packet on the stream.
* Retain the timing signal needed to query the recorded marker later.
* Implement `getEventElapsedTimeImpl` using
`hsa_amd_profiling_get_dispatch_time`, converting the result to
milliseconds with `HSA_SYSTEM_INFO_TIMESTAMP_FREQUENCY`.
This follows the ROCm/HIP approach of enabling queue profiling at HSA
queue creation time, while keeping the AMDGPU queue path simpler than
the lazy-enable alternative discussed during review.
#### CUDA
* Add the required CUDA driver declarations and wrappers.
* Implement `getEventElapsedTimeImpl` with `cuEventElapsedTime`.
#### Host
* Add `getEventElapsedTimeImpl` that stores `0.0f` in the output
pointer, when present, and returns success.
Reason: the host plugin does not materialize backend event state and
already treats event operations as trivially successful. Returning
`0.0f` preserves that model without introducing a new failure mode.
#### Level Zero
* Add `getEventElapsedTimeImpl`, but leave it unimplemented.
Reason: the Level Zero plugin currently does not provide standalone
backend event support for this event model. For example, `waitEventImpl`
/ `syncEventImpl` are still unimplemented there.
---------
Signed-off-by: Leandro Augusto Lacerda Campos <leandrolcampos@yahoo.com.br>
Signed-off-by: Leandro A. Lacerda Campos <leandrolcampos@yahoo.com.br>
This commit removes the `LIBOMPTARGET_SHARED_MEMORY_SIZE` envar and
outputs a runtime warning if it is defined. Access to dynamic shared memory
should be obtained through the `dyn_groupprivate` clause (OpenMP 6.1) or
the launch arguments in liboffload kernel launch.
Add liboffload asynchronous queue query API for libomptarget migration
This PR adds liboffload asynchronous queue query API that needed to make
libomptarget to use liboffload
Summary:
This was a lot of code that was only used for upstream LLVM builds of
AMDGPU offloading. We have a generic and fast `malloc` in `libc` now so
just use that. Simplifies code, can be added back if we start providing
alternate forms but I don't think there's a single use-case that would
justify it yet.
Introduced in OpenMP 6.0, the device UID shall be a unique identifier of
a device on a given system. (Not necessarily a UUID.) Since it is not
guaranteed that the (U)UIDs defined by the device vendor libraries, such
as HSA, do not overlap with those of other vendors, the device UIDs in
offload are always combined with the offload plugin name. In case the
vendor library does not specify any device UID for a given device, we
fall back to the offload-internal device ID.
The device UID can be retrieved using the `llvm-offload-device-info`
tool.
This implements two pieces to restore the interop functionality (that I
broke) when the 6.0 interfaces were added:
* A set of wrappers that support the old interfaces on top of the new
ones
* The same level of interop support for the CUDA amd AMD plugins
Summary:
This was originally added in as a hack to work around CUDA's limitation
on allocation. The `libc` implementation now isn't even used for CUDA so
this code is never hit. Even if this case, this code never truly worked.
A true solution would be to use CUDA's virtual memory API instead to
allocate 2MiB slabs independenctly from the normal memory management
done in the stream.
Summary:
I made the GPU flags accept more of the default LLVM warnings, which
triggered some new cases. Clean those up and fix some other ones while
I'm at it.
Summary:
Turns out the new CUDA ABI now applies retroactively to all the other
SMs if you upgrade to CUDA 13.0. This patch changes the scheme, keeping
all the SM flags consistent but using an offset.
Fixes: https://github.com/llvm/llvm-project/issues/159088
Summary:
Currently we have this `__tgt_device_image` indirection which just takes
a reference to some pointers. This was all find and good when the only
usage of this was from a section of GPU code that came from an ELF
constant section. However, we have expanded beyond that and now need to
worry about managing lifetimes. We have code that references the image
even after it was loaded internally. This patch changes the
implementation to instaed copy the memory buffer and manage it locally.
This PR reworks the JIT and other image handling to directly manage its
own memory. We now don't need to duplicate this behavior externally at
the Offload API level. Also we actually free these if the user unloads
them.
Upside, less likely to crash and burn. Downside, more latency when
loading an image.
This is equivalent to `cuOccupancyMaxPotentialBlockSize`. It is
currently
only implemented on Cuda; AMDGPU and Host return unsupported.
---------
Co-authored-by: Callum Fare <callum@codeplay.com>
Add the following properties in Offload device info:
* VENDOR_ID
* NUM_COMPUTE_UNITS
* [SINGLE|DOUBLE|HALF]_FP_CONFIG
* NATIVE_VECTOR_WIDTH_[CHAR|SHORT|INT|LONG|FLOAT|DOUBLE|HALF]
* MAX_CLOCK_FREQUENCY
* MEMORY_CLOCK_RATE
* ADDRESS_BITS
* MAX_MEM_ALLOC_SIZE
* GLOBAL_MEM_SIZE
Add a bitfield option to enumerators, allowing the values to be
bit-shifted instead of incremented. Generate the per-type enums using
`foreach` to reduce code duplication.
Use macros in unit test definitions to reduce code duplication.
The purpose of this fence is to ensure that any `dataSubmit`s inserted
into a queue before a `dataFence` finish before finish before any
`dataSubmit`s
inserted after it begin.
This is a no-op for most queues, since they are in-order, and by design
any operations inserted into them occur in order.
But the interface is supposed to be functional for out-of-order queues.
The addition of the interface means that any operations that rely on
such ordering (like ATTACH map-type support in #149036) can invoke it,
without worrying about whether the underlying queue is in-order or
out-of-order.
Once a plugin supports out-of-order queues, the plugin can implement
this function, without requiring any change at the libomptarget level.
---------
Co-authored-by: Alex Duran <alejandro.duran@intel.com>
This sprinkles a few mutexes around the plugin interface so that the
olLaunchKernel CTS test now passes when ran on multiple threads.
Part of this also involved changing the interface for device synchronise
so that it can optionally not free the underlying queue (which
introduced a race condition in liboffload).
Add a device function to check if a device queue is empty. If liboffload
tries to create an event for an empty queue, we create an "empty" event
that is already complete.
This allows `olCreateEvent`, `olSyncEvent` and `olWaitEvent` to run
quickly for empty queues.
`MAX_WORK_GROUP_SIZE` now represents the maximum total number of work
groups the device can allocate, rather than the maximum per dimension.
`MAX_WORK_GROUP_SIZE_PER_DIMENSION` has been added, which has the old
behaviour.
Summary:
We rely on these flags to do things in the runtime and print the
contents of binaries correctly. CUDA updated their ABI encoding recently
and we didn't handle that. it's a new ABI entirely so we just select on
it when it shows up.
Fixes: https://github.com/llvm/llvm-project/issues/148703
The `GlobalTy` helper has been extended to make both the Size and Ptr be
optional. Now `getGlobalMetadataFromDevice`/`Image` is able to write the
size of the global to the struct, instead of just verifying it.
After #146345 the device info implementation requires a value for every
query, rather than silently returning an empty string. This broke the
test for `OL_DEVICE_INFO_VENDOR` on CUDA.
Add a value to the CUDA plugin. We can quite safely hard code this one.
Previously, the user was not able to use more than 48 KB of shared
memory on NVIDIA GPUs. In order to do so, setting the function attribute
`CU_FUNC_ATTRIBUTE_MAX_THREADS_PER_BLOCK` is required, which was not
present in the code base. With this commit, we add the ability toset
this attribute, allowing the user to utilize the full power of their
GPU.
In order to not have to reset the function attribute for each launch of
the same kernel, we keep track of the maximum memory limit (as the
variable `MaxDynCGroupMemLimit`) and only set the attribute if our
desired amount exceeds the limit. By default, this limit is set to 48
KB.
Feedback is greatly appreciated, especially around setting the new
variable as mutable. I did this becuase the `launchImpl` method is const
and I am not able to modify my variable otherwise.
---------
Co-authored-by: Giorgi Gvalia <ggvalia@login33.chn.perlmutter.nersc.gov>
Co-authored-by: Giorgi Gvalia <ggvalia@login07.chn.perlmutter.nersc.gov>
AMD treats this value as a string, so for consistency require this in
NVIDIA as well. This shouldn't change the output of the
`llvm-offload-device-info` tool, but does fix an issue in liboffload
when it tries to query the version.
This allows removal of a specific Image from a Device, rather than
requiring all image data to outlive the device they were created for.
This is required for `ol_program_handle_t`s, which now specify the
lifetime of the buffer used to create the program.
Previously, device info was returned as a queue with each element having
a "Level" field indicating its nesting level. This replaces this queue
with a more traditional tree-like structure.
This should not result in a change to the output of
`llvm-offload-device-info`.
[Offload] Use new error code handling mechanism
This removes the old ErrorCode-less error method and requires
every user to provide a concrete error code. All calls have been
updated.
In addition, for consistency with error messages elsewhere in LLVM, all
messages have been made to start lower case.
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.
