NVIDIA/cccl issues and pull requests

#2453 - Fix typo in CCCLBuildCompilerTargets.cmake

Pull Request - State: closed - Opened by alliepiper 4 days ago - 1 comment

#2452 - Ensure that we treat `cuda::mr::host_accessible` differently

Pull Request - State: open - Opened by miscco 4 days ago - 4 comments

#2451 - [BUG]: thrust::device_vector initialization with fancy iterators 4x slower than initialization + thrust::transform

Issue - State: open - Opened by rwarmstr 5 days ago - 10 comments
Labels: bug

#2450 - Drop superflous compile definition from thrust tests

Pull Request - State: open - Opened by miscco 6 days ago - 1 comment
Labels: thrust, infrastructure

#2449 - [BUG]: thrust::inclusive_scan not working correctly when -rdc=true

Issue - State: open - Opened by Olli1080 8 days ago - 1 comment
Labels: bug

#2448 - [FEA]: Refactor C++/C dispatch layers to avoid code duplication

Issue - State: open - Opened by gevtushenko 8 days ago
Labels: feature request

#2447 - Programmatically (`sed`) replace `CUDA C++ Core Libraries` with CUDA Core Compute Libraries`.

Pull Request - State: open - Opened by rwgk 8 days ago - 1 comment

#2446 - [CUDAX] Add placeholder green context type and logical device that can hold both a green ctx and a device

Pull Request - State: closed - Opened by pciolkosz 8 days ago - 7 comments

#2445 - [FEA]: make `bit_cast` compatible with vector version and arrays of extended precision types

Issue - State: open - Opened by fbusato 8 days ago
Labels: feature request

#2444 - Do not include our own string.h file

Pull Request - State: closed - Opened by miscco 8 days ago - 1 comment
Labels: libcu++, bug: functional

#2443 - Cleanup and modularize `<cuda/std/barrier>`

Pull Request - State: open - Opened by miscco 8 days ago - 1 comment
Labels: feature request, libcu++

#2442 - Add support for large `num_items` to `DevicePartition::ThreeWayPartition`

Issue - State: open - Opened by elstehle 8 days ago - 1 comment

#2441 - Cleanup the semaphore headers

Pull Request - State: open - Opened by miscco 9 days ago - 1 comment
Labels: feature request, libcu++

#2440 - [BUG]: Using cuda::atomic_ref fails when using nvrtc and --device-debug

Issue - State: open - Opened by shacklettbp 9 days ago
Labels: bug

#2439 - replace `_CCCL_ALWAYS_INLINE` with `_CCCL_FORCEINLINE`

Pull Request - State: open - Opened by ericniebler 9 days ago - 6 comments

#2438 - [BUG]: unify `_CCCL_FORCEINLINE` and `_CCCL_ALWAYS_INLINE`

Issue - State: open - Opened by ericniebler 9 days ago
Labels: bug

#2437 - Move nightly time

Pull Request - State: closed - Opened by bdice 9 days ago - 3 comments

#2436 - Reduce PR testing matrix.

Pull Request - State: closed - Opened by alliepiper 9 days ago - 2 comments

#2435 - Increase the libcu++ timeout

Pull Request - State: closed - Opened by miscco 9 days ago - 6 comments

#2434 - Make `bit_cast` play nice with extended floating point types

Pull Request - State: closed - Opened by miscco 10 days ago - 5 comments
Labels: feature request, libcu++

#2433 - Experiment with a common type to represent a device, green context and possibly more in the future

Issue - State: open - Opened by pciolkosz 10 days ago

#2432 - [FEA]: CMake `FetchContent_Populate` is deprecated

Issue - State: open - Opened by fbusato 10 days ago
Labels: feature request

#2431 - `uninitialized_buffer::get_resource` returns a ref to an `any_resource` that can be copied

Pull Request - State: closed - Opened by ericniebler 10 days ago - 3 comments
Labels: CUDA Next

#2430 - [BUG]: cudax `uninitialized_buffer::get_resource` returns a `resource_ref` that can dangle

Issue - State: closed - Opened by ericniebler 10 days ago
Labels: bug

#2429 - [WIP][FEA]: Validate cuda.parallel type matching in build and execution

Pull Request - State: open - Opened by rwgk 10 days ago - 1 comment

#2428 - Move c/include/cccl/.h files to c/include/cccl/c/.h

Pull Request - State: closed - Opened by rwgk 10 days ago - 3 comments

#2427 - Limit Linux CI parallel jobs for each PR.

Pull Request - State: closed - Opened by bdice 10 days ago - 1 comment

#2426 - [DO_NOT_MERGE] disable system header escape hatch

Pull Request - State: closed - Opened by miscco 10 days ago - 3 comments

#2425 - Make `any_resource` emplacable

Pull Request - State: closed - Opened by miscco 10 days ago - 5 comments
Labels: feature request, CUDA Next

#2424 - Replace `CUDA C++ Core Libraries` with `CUDA Core Compute Libraries` (only in README.md).

Pull Request - State: closed - Opened by rwgk 11 days ago - 3 comments

#2423 - [BUG]: Incorrect check for cudax flag

Issue - State: open - Opened by osayamenja 11 days ago - 1 comment
Labels: bug

#2422 - [FEA]: make `bit_cast` compatible with raw narrow precision types

Issue - State: closed - Opened by fbusato 11 days ago - 1 comment
Labels: feature request

#2421 - Fix thrust package to work with newer FindOpenMP.cmake.

Pull Request - State: closed - Opened by alliepiper 11 days ago - 3 comments

#2420 - Propagate compiler flags down to libcu++ LIT tests

Pull Request - State: open - Opened by Artem-B 11 days ago - 8 comments

#2419 - [BUG]: `is_common_type` trait is broken when mixing rvalue references

Issue - State: open - Opened by fbusato 11 days ago - 2 comments
Labels: bug

#2418 - Minor cleanup with `cuda/atomic`

Pull Request - State: closed - Opened by miscco 11 days ago - 3 comments

#2417 - Fix sccache errors in RAPIDS builds

Pull Request - State: closed - Opened by trxcllnt 11 days ago - 2 comments

#2416 - [FEA]: Validate cuda.parallel type matching in build and execution

Issue - State: open - Opened by gevtushenko 12 days ago
Labels: feature request

#2415 - [BUG]: libcu++ tests fail with clang builds because compiler flags are not propagated to lit tests.

Issue - State: open - Opened by Artem-B 12 days ago
Labels: bug

#2414 - [FEA]: Unify Half/Bfloat16 detection macros

Issue - State: open - Opened by fbusato 12 days ago
Labels: feature request

#2413 - Fix issues with `host` and `device` definitions

Pull Request - State: closed - Opened by miscco 12 days ago - 9 comments

#2412 - [FEA]: optimize the case of storing a `resource_ref` in an `any_resource`

Issue - State: open - Opened by ericniebler 16 days ago - 1 comment
Labels: feature request

#2411 - `mr::resource` concept should require `const` allocate and `noexcept` deallocate

Issue - State: open - Opened by ericniebler 16 days ago - 3 comments
Labels: bug, CUDA Next

#2410 - give `any_resource` shared ownership of any resource that is not "small"

Pull Request - State: closed - Opened by ericniebler 17 days ago - 7 comments

#2409 - [BUG]: Remove warning suppressions from Thrust build system.

Issue - State: open - Opened by alliepiper 17 days ago
Labels: bug

#2408 - [BUG]: Using cuda::std::complex <<, >> (ostream,istream) operators cause to compilation error

Issue - State: open - Opened by YoniSp 17 days ago
Labels: bug

#2407 - [BUG]: CCCL headers use `host`/`device` macros without defining them

Issue - State: closed - Opened by ericniebler 17 days ago - 1 comment
Labels: bug

#2406 - conditionally include `<crt/host_defines.h>` from `__cccl/execution_space.h` header

Pull Request - State: closed - Opened by ericniebler 17 days ago - 2 comments

#2405 - Implement `cub::DeviceFind::FindIf`

Pull Request - State: open - Opened by gonidelis 17 days ago - 8 comments

#2404 - Extend address stability detection for `thrust::transform`

Pull Request - State: open - Opened by bernhardmgruber 17 days ago
Labels: thrust

#2401 - nvrtc wants `integral_constant`'s functions host/device annotated

Pull Request - State: closed - Opened by ericniebler 18 days ago - 3 comments

#2400 - Adds support for large number of items in `DeviceSelect` and `DevicePartition`

Pull Request - State: open - Opened by elstehle 18 days ago - 11 comments

#2399 - Pull out `_LIBCUDACXX_UNREACHABLE` into its own file

Pull Request - State: closed - Opened by miscco 18 days ago - 3 comments

#2398 - Implement `cudax::shared_resource`

Pull Request - State: closed - Opened by miscco 18 days ago - 6 comments
Labels: feature request, CUDA Next

#2397 - [FEA] Performing device-to-device copies with thrust::copy is synchronous with `par_nosync`. Can this be improved?

Issue - State: closed - Opened by fkallen 18 days ago - 1 comment
Labels: feature request

#2396 - Add prefetching kernel as new fallback for `cub::DeviceTransform`

Pull Request - State: open - Opened by bernhardmgruber 18 days ago - 1 comment
Labels: cub

#2392 - Drop thrusts diagnostic suppression warnings

Pull Request - State: closed - Opened by miscco 20 days ago - 4 comments

#2390 - [DRAFT] Optimize, Cleanup, and Expose CUB Thread-Level Reduction

Pull Request - State: open - Opened by fbusato 22 days ago - 5 comments

#2388 - Introduce `cccl_configure_target` cmake function.

Pull Request - State: closed - Opened by alliepiper 22 days ago - 11 comments

#2387 - Add a `cccl_configure_target` method that handles common config steps (configuring binary output locations, c++ dialects, etc)

Issue - State: closed - Opened by alliepiper 22 days ago

#2386 - Share common compiler flags in new CCCL-level targets.

Pull Request - State: closed - Opened by alliepiper 22 days ago - 5 comments

#2385 - Create `cccl.compiler_interface_cppXX` targets with shared build flags.

Issue - State: closed - Opened by alliepiper 22 days ago

#2382 - Unify assert handling in cccl

Pull Request - State: open - Opened by miscco 22 days ago - 11 comments
Labels: feature request, libcu++

#2379 - Should `any_resource` be copyable?

Issue - State: closed - Opened by harrism 23 days ago - 23 comments
Labels: libcu++, question

#2378 - [Draft] Add ForEachN from CUB to cccl/c.

Pull Request - State: open - Opened by wmaxey 23 days ago - 1 comment

#2376 - Refactor `cuda::ceil_div` to take two different types

Pull Request - State: closed - Opened by miscco 23 days ago - 13 comments

#2375 - Replace cub::DivideAndRoundUp by cuda::ceil_div

Issue - State: closed - Opened by bernhardmgruber 23 days ago - 11 comments
Labels: cub, libcu++

#2370 - add some simple utilities for manipulating lists of types

Pull Request - State: closed - Opened by ericniebler 24 days ago - 19 comments

#2369 - [FEA]: unify CCCL's metaprogramming facilities

Issue - State: closed - Opened by ericniebler 24 days ago
Labels: feature request

#2366 - [FEA]: Consolidate shared CMake infrastructure.

Issue - State: open - Opened by alliepiper 24 days ago
Labels: feature request

#2356 - Remove a ton of lines in thrust tests

Pull Request - State: closed - Opened by gonidelis 25 days ago - 13 comments

#2354 - Unify shared infra

Pull Request - State: open - Opened by alliepiper 25 days ago - 1 comment

#2347 - [FEA]: Improve readability of thrust tests

Issue - State: closed - Opened by miscco 26 days ago - 2 comments
Labels: feature request, good first issue, thrust, help wanted

#2339 - [DOC]: Format function signatures

Issue - State: open - Opened by gevtushenko 29 days ago - 4 comments
Labels: doc

#2306 - [FEA]: Add a copy routine to support data copy between two `mdspan`s

Issue - State: open - Opened by leofang about 1 month ago - 8 comments
Labels: feature request

#2303 - [FEA]: Enable exceptions by default

Issue - State: open - Opened by PointKernel about 1 month ago - 8 comments
Labels: feature request

#2287 - [FEA]: Optimize, Cleanup, and Expose CUB Thread-Level Reduction

Issue - State: open - Opened by fbusato about 1 month ago
Labels: feature request

#2274 - [BUG]: Include path for C library is too generic

Issue - State: closed - Opened by gevtushenko about 1 month ago
Labels: bug

#2255 - Add `atomic_ref` support for 8 and 16b types.

Pull Request - State: open - Opened by wmaxey about 1 month ago - 7 comments

#2254 - Add CUB tests for segmented sort/radix sort with 64-bit num. items and segments

Pull Request - State: open - Opened by fbusato about 1 month ago - 23 comments

#2234 - Add simple kernel for deterministic reduction

Pull Request - State: open - Opened by SAtacker about 2 months ago - 3 comments

#2193 - FIFO fair semaphore implementation

Pull Request - State: open - Opened by ZelboK about 2 months ago - 1 comment

#2154 - Prune CUB's ChainedPolicy by __CUDA_ARCH_LIST__

Pull Request - State: open - Opened by bernhardmgruber about 2 months ago - 22 comments
Labels: cub

#2141 - [FEA]: A `cuda::uninitialized_async_buffer<T, Properties...>` that handles allocation / deallocation and other low level things for the user

Issue - State: closed - Opened by miscco about 2 months ago - 1 comment
Labels: feature request, CUDA Next, 2.7.0

GitHub / NVIDIA/cccl issues and pull requests