We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
This issue is to keep track of breaking changes we'd like to make in the next major version release.
global_workarounds.h
THRUST_HOST_BACKEND
Agent*
detail::<algorithm name>::Agent*
cub::DeviceSpmv
LEGACY_PTX_ARCH
::result_type
debug_synchronous
thrust::[unary|binary]_function
unary_traits
binary_traits
not1
not2
cub::min
thrust::remove_cvref
cub::max
logical_metafunctions.h
thrust::[make_][integer|index]_sequence
thrust/thrust/system/detail/error_*
thrust::system::cuda::errc
thrust::cuda_cub::core::AgentLauncher
thrust::cuda_cub::launcher::triple_chevron
THRUST_TUNING_ARCH
thrust::mr::validator
simple_defines.h
thrust/detail/preprocessor.h
THRUST_PP*
thrust::tuple_for_each
memory_algorithms.h
event_error.h
thrust::has_difference_type
advance.inl
DeviceSegmentedReduce::Arg{Min,Max}
cub::KeyValuePair
CUB_IS_INT128_ENABLED
cuda::std::numeric_limits
__half
__nv_bfloat16
__nv_fp8_e4m3
__nv_fp8_e5m2
cub::FpLimits
thrust::identity
cuda::std::identity
cub/[block|warp]/specializations
cub::min/max
cuda::std::min/max
util_ptx.cuh
GridBarrier
GridBarrierLifetime
We should backport those to the 2.8 branch.
thrust::discard_iterator
value_type
void
thrust::identity::operator()
cuda::std::identity::operator()
ceil_div
DispatchBatchMemcpy
DispatchReduce
DispatchScan
universal_host_pinned_memory_resource
universal_ptr
<complex>
result_of_adaptable_function
thrust/thrust/system/detail/sequential/*
<algorithm>
THRUST_UNUSED_VAR
cuda::(std::)
__half/bfloat16/fp8
thrust/thrust/system/cuda/config.h
The text was updated successfully, but these errors were encountered:
shfl_sync
bernhardmgruber
No branches or pull requests
This issue is to keep track of breaking changes we'd like to make in the next major version release.
Removals
global_workarounds.h
THRUST_HOST_BACKEND
Agent*
alias templates for the respectivedetail::<algorithm name>::Agent*
#3467cub::DeviceSpmv
in favor of cuSPARSE #3070LEGACY_PTX_ARCH
parameters::result_type
etc. from Thrust function objects (and alias them to the libcu++ ones) #3271debug_synchronous
parameter #3329thrust::[unary|binary]_function
#3273unary_traits
andbinary_traits
#3259not1
andnot2
#3263cub::min
andthrust::remove_cvref
#3450cub::max
Deprecations
logical_metafunctions.h
which can be replaced by fold expressionsthrust::[make_][integer|index]_sequence
in favor of libcu++thrust/thrust/system/detail/error_*
thrust::system::cuda::errc
thrust::cuda_cub::core::AgentLauncher
(move to detail)thrust::cuda_cub::launcher::triple_chevron
(move to detail)THRUST_TUNING_ARCH
thrust::mr::validator
(move to detail)simple_defines.h
thrust/detail/preprocessor.h
andTHRUST_PP*
macrosthrust::tuple_for_each
memory_algorithms.h
event_error.h
thrust::has_difference_type
inadvance.inl
(move to detail)DeviceSegmentedReduce::Arg{Min,Max}
interface that returns result ascub::KeyValuePair
#3482CUB_IS_INT128_ENABLED
by a CCCL macro and deprecate it #3403cuda::std::numeric_limits
for__half
,__nv_bfloat16
,__nv_fp8_e4m3
,__nv_fp8_e5m2
and deprecatecub::FpLimits
#3349thrust::identity
bycuda::std::identity
#2762cub/[block|warp]/specializations
into a detail namespace #3409cub::min/max
bycuda::std::min/max
#3404util_ptx.cuh
#3289cub::DeviceSpmv
#896GridBarrier
andGridBarrierLifetime
#1389We should backport those to the 2.8 branch.
API breaking changes
thrust::discard_iterator
'svalue_type
void
#3399thrust::identity::operator()
have the same return type ascuda::std::identity::operator()
ceil_div
should return the resulting type of its operation #2845DispatchBatchMemcpy
andDispatchReduce
so policy hub is lastDispatchScan
to be unsigneduniversal_host_pinned_memory_resource
's pointer type should probably be auniversal_ptr
#2659ABI breaking changes
Behavioral changes
<complex>
Enable users to disable normalization for division too. libcudacxx#474 (review)::result_type
detection from Thrustresult_of_adaptable_function
Cleanup
thrust/thrust/system/detail/sequential/*
by libcu++<algorithm>
THRUST_UNUSED_VAR
by some CCCL macrocuda::(std::)
types for__half/bfloat16/fp8
#525thrust/thrust/system/cuda/config.h
The text was updated successfully, but these errors were encountered: