NVIDIA / NVIDIA/cutlass

[FEA] make_tma_copy doesnot support shared-to-shared copy

Open
#1,657 3 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

feature request inactive-30d inactive-90d
Dominant language
C++
Stars
10.5k
Forks
2.1k
Avg merge
3d 11h
Merged PRs (30d)
7

Description

Describe the bug
In PTX, I noticed that

cp.async.bulk.dst.src.completion_mechanism [dstMem], [srcMem], size, [mbar]

.dst =                  { .shared::cluster }
.src =                  { .shared::cta }
.completion_mechanism = { .mbarrier::complete_tx::bytes }

supports shared-to-shared copy. But in cute, make_tma_copy has to be global to shared. Why? Could you modify it?

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by locating make_tma_copy and reviewing how its source and destination memory spaces are represented. Compare the supported forms with the PTX shared-to-shared cp.async.bulk instruction shown in the issue; done means make_tma_copy supports that shared-to-shared case without regressing existing global-to-shared behavior.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
hpc
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.