Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

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

Jump to bottom

TA::retile support for DistArray with tensor-of-tensors tiles #474

Merged
merged 50 commits into from
Oct 29, 2024
Merged

TA::retile support for DistArray with tensor-of-tensors tiles #474

merged 50 commits into from
Oct 29, 2024

Conversation

bimalgaudel
Copy link
Member

No description provided.

@bimalgaudel bimalgaudel marked this pull request as ready for review September 18, 2024 19:36
@evaleev @bimalgaudel
TiledRange1: can construct using a range of tile hashmarks
1f186cc
@evaleev @bimalgaudel
<-> Eigen::{Vector,Matrix,Tensor} conversions can handle DistArrays w…
91b7f26 
…ith non-zero base Ranges
@evaleev @bimalgaudel
[skip ci] typo
d4be91c
@evaleev @bimalgaudel
remove duplicate vlock copy in concat
f2b319b
@evaleev @bimalgaudel
introduced tile_ranges_match_trange(DistArray) for validating tile ra…
bd5f35b 
…nges against trange
@evaleev @bimalgaudel
SizeArray is a viewable range
59b0a37
@evaleev @bimalgaudel
fixed assignment to block expression from an expression with nonzero …
aed7128 
…base
@evaleev @bimalgaudel
[skip ci] to_container.hpp -> range/conversion.hpp
1662f8b 
to_container.hpp is deprecated
@evaleev @bimalgaudel
[skip ci] svd dox fixup
5b26108
@evaleev @bimalgaudel
TiledRange1{int x} constructs an empty element range at [x,x)
9c84513
@evaleev @bimalgaudel
make Range1 printable and shiftable
64ba684
@evaleev @bimalgaudel
make TiledRange1 shiftable
6a8f75f
@evaleev @bimalgaudel
TiledRange1 printer reimplemented in terms of Range1 printer
c0ebdd3
@evaleev @bimalgaudel
[skip ci] dox++
b080325
@evaleev @bimalgaudel
Range::shift is const
c22bf70
@evaleev @bimalgaudel
Range::shift is nodiscard
78931bc
@evaleev @bimalgaudel
TiledRange is shiftable
a847f70
@evaleev @bimalgaudel
introduced BlkTsrExpr::{{set_,}trange_lobound,preserve_lobound}() tha…
9ef3a8a 
…t allow to use block tensor expressions even with DistArrays that have non-zero lobound
@evaleev @bimalgaudel
dox fixup [skip ci]
5391346
@evaleev @bimalgaudel
introduced TiledRange::is_congruent
0793929
@evaleev @bimalgaudel
like contraction, reduction expression and binary expression can igno…
d49c7bc 
…re absolute positions if ignore_tile_position() is on
@evaleev @bimalgaudel
introduced TiledRange1::{lo,up}bound which feel to have unambiguous m…
9e85c7d 
…eaning
@evaleev @bimalgaudel
more TsrExpr::block variants tagged by preserve_lobound_t
f9b3f25
@evaleev @bimalgaudel
TA::expressions::preserve_lobound_t -> TA::preserve_lobound_t
ea9c834
@evaleev @bimalgaudel
btas <-> ta tensor conversions work for non-0-lobound
8e58039
@evaleev @bimalgaudel
introduced member versions of TiledRange1::make_uniform
548d5ea
evaleev and others added 22 commits September 25, 2024 08:45
@evaleev @bimalgaudel
TiledRange1::make_uniform(rng) with empty range preserves its lobound
4957361
@evaleev @bimalgaudel
heig: work around the n=0 corner case
bb1e83a
@evaleev @bimalgaudel
bump MAD tag to pull in m-a-d-n-e-s-s/madness#547
6429585
@evaleev @bimalgaudel
BinaryExpr: account for ignore_tile_position when checking preconditions
28b1966
@evaleev @bimalgaudel
[python] simplify make_trange by using TiledRange1::make_uniform
9a3594a
@evaleev @bimalgaudel
[ci] greatly reduce the gitlab matrix, replace rel/deb builds with re…
6e7e250 
…lwithdebinfo
@evaleev @bimalgaudel
TA::host_allocator is serializable, so that btas::Tensor can be used …
18bc8f4 
…as a tile again
@evaleev @bimalgaudel
introduced TA::Tile::at_ordinal + strengthen disambiguation checks fo…
5d35fb4 
…r potential at_ordinal uses
@evaleev @bimalgaudel
hostEnv -> host::Env + fixup to make f294db3 build
793b7c4
@evaleev @bimalgaudel
[ci] do not use gcc toolchain on macos, instead try linux + enable cc…
28323a0 
…ache
@evaleev @bimalgaudel
[ci] try symlinking libscalapack-openmpi.so to libscalapack.so to hel…
91130e1 
…p out FindReferenceSCALAPACK
@evaleev @bimalgaudel
[unit] [cuda] another disambiguation via at_ordinal
cd2f48f
@evaleev @bimalgaudel
[ci] disable Gitlab jobs except CUDA
2cc0629
@evaleev @bimalgaudel
introduce {Tensor,Tile}::c{begin,end}
8bea078
@evaleev @bimalgaudel
TensorInterface is a range
d5d6b6f
@evaleev @bimalgaudel
dox fixup
469ae8a
@evaleev @bimalgaudel
btas::Tensor can be copied into from Tensor and TensorInterface ... t…
93b96a3 
…his allows TA::retile on DistArrays of btas::Tensors
@evaleev @bimalgaudel
[unit] btas_suite/tensor_ctor: initialize ta_tensor properly
3fdbaa6
@evaleev @bimalgaudel
fixup TensorInterface::Iterator::deference()
ac83135
@evaleev @bimalgaudel
pull in ValeevGroup/BTAS#179
280983e
@bimalgaudel
Bug fixes in iterator logic in UserPmap.
c12bc45
@bimalgaudel
Merge branch 'refs/heads/master' into gaudel/feature/retile_tot_arrays
4bf23ad 
# Conflicts:
#	.github/workflows/ci.yml
#	INSTALL.md
#	external/versions.cmake
@evaleev evaleev merged commit 486ae16 into master Oct 29, 2024
9 checks passed
@evaleev evaleev deleted the gaudel/feature/retile_tot_arrays branch October 29, 2024 22:13
@bimalgaudel bimalgaudel restored the gaudel/feature/retile_tot_arrays branch October 30, 2024 14:18
@evaleev evaleev deleted the gaudel/feature/retile_tot_arrays branch November 14, 2024 03:55
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@evaleev evaleev evaleev approved these changes

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@bimalgaudel @evaleev