TTG device tasks with coroutines #265
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Most functionality is an extension/copy of the CUDA support. Also ported the bspmm example to HIP. Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
avoids CMake errors related to lapack including OpenMP_C dependence that MADNESS does not know anything about
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Notable changes: - add ttg::device::current_device and ttg::device::current_stream to query current streams - Consistently start counting devices at 0 - The host has ID -2 - Fixes in the broadcast Signed-off-by: Joseph Schuchart <[email protected]>
This fixes a segfault in multi-gpu operations. We should not destroy the full data_t but only remove the host-side copy and release our reference on the data. The data will be destroyed once all gpu copies have been removed. Signed-off-by: Joseph Schuchart <[email protected]>
evaleev
force-pushed
the
ttg-device-support-master-coro-with-stream-tasks
branch
from
2166d5e
to
f05530f
Compare
evaleev
force-pushed
the
ttg-device-support-master-coro-with-stream-tasks
branch
from
753ccb9
to
1ffaf6a
Compare
Signed-off-by: Joseph Schuchart <[email protected]>
For small objects it is beneficial of packing the data into the activation message instead of going through an extra GET. Signed-off-by: Joseph Schuchart <[email protected]>
Work may be performed by the owner of A, B, or C, or the default honoring the R parameter. Signed-off-by: Joseph Schuchart <[email protected]>
For matrix A, we make sure that tiles are sent out in column-order (left to right). For matrix B, we want to make sure that tiles are sent out in row-order (top to bottom). Signed-off-by: Joseph Schuchart <[email protected]>
…thub.com:devreal/ttg into ttg-device-support-master-coro-with-stream-tasks
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
Signed-off-by: Joseph Schuchart <[email protected]>
… only produce if Catch's `verbosity` != `quiet`
evaleev
approved these changes
Feb 24, 2024
| @@ -17,7 +17,7 @@ if (NOT TARGET PaRSEC::parsec) | |||
|
|
|||
| FetchContent_Declare( | |||
| PARSEC | |||
| GIT_REPOSITORY https://github.com/ICLDisco/parsec.git | |||
| GIT_REPOSITORY https://github.com/devreal/parsec-1.git | |||
There was a problem hiding this comment.
is there a PR on https://github.com/ICLDisco/parsec.git that we can quote here?
evaleev
merged commit f89cf45
into
TESSEorg:ttg-device-support-master-coro-with-stream-tasks
4 checks passed
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Supersedes #253