[STF] Rename getStream into get stream#7946
Conversation
place_partition.cuh does not use exec_affinity; the include was a leftover. Made-with: Cursor
This reverts commit 7fd0545.
- Refactor data_place::to_string() to use cascaded ternary operators - Replace assert() calls with _CCCL_ASSERT for consistent diagnostics - Remove exec_place::impl::create_stream() and inline into stream_pool::next() - Simplify exec_place::operator->* to use exec_place_guard RAII pattern - Minor cleanup: remove redundant parentheses, reorder operator< logic
|
/ok to test |
@andralex, there was an error processing your request: See the following link for more information: https://docs.gha-runners.nvidia.com/cpr/e/1/ |
|
/ok to test 4e1b59b |
|
|
||
| void* allocate(::std::ptrdiff_t, cudaStream_t) const override | ||
| { | ||
| throw ::std::logic_error("Cannot allocate from device_auto data_place directly"); |
There was a problem hiding this comment.
note: some methods throw, some other return NOT IMPLEMENTED
| */ | ||
| virtual bool equals(const data_place_interface& other) const | ||
| { | ||
| return typeid(*this) == typeid(other) && get_device_ordinal() == other.get_device_ordinal(); |
There was a problem hiding this comment.
this looks like a remains of having the base data_place == data_place_device ? (this was in exec_place_device), maybe this should be a pure virtual = 0 ?
😬 CI Workflow Results🟥 Finished in 36m 56s: Pass: 93%/48 | Total: 15h 01m | Max: 28m 28s | Hits: 59%/21875See results here. |
miscco
left a comment
There was a problem hiding this comment.
Can you rebase?
there are a lot of unrelated changes
| bool allocation_is_stream_ordered() const override | ||
| { | ||
| return true; | ||
| } | ||
|
|
||
| // Provide affine exec place implementation for data_place::affine_exec_place() | ||
| ::std::shared_ptr<void> get_affine_exec_impl() const override; | ||
|
|
| // under the Apache License v2.0 with LLVM Exceptions. | ||
| // See https://llvm.org/LICENSE.txt for license information. | ||
| // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception | ||
| // SPDX-FileCopyrightText: Copyright (c) 2022-2024 NVIDIA CORPORATION & AFFILIATES. |
…VIDIA#7926) * prepare tests for var segments * adds tests for small variable-size segments * restores tests for types * reverts now-superfluous segment fixup test code * variable segment size tests * split up test tus, restore utils * adds tests for variable-size keys * fixes tu splitup * style fix
I thought I'd make your life easier so I put the refactoring in #7976 for easy review. Will ping you when ready for review. Thanks! |
4e1b59b to
b30ffa3
Compare
Description
Use a consistent naming scheme
closes
Checklist