Skip to content

[TESTING] Fix expected error in 2CTA consan test #10409

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

Merged
lezcano merged 1 commit into from
May 29, 2026
+7 −3
Merged

[TESTING] Fix expected error in 2CTA consan test #10409

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 7 additions & 3 deletions python/test/gluon/test_consan.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -933,10 +933,14 @@ def test_tcgen5_mma(FAILURE, MEM_ACCESS_KIND, TWO_CTAS, device, run_wrapper, mon
assert_expected_cuda_failure(result.exc)
if MEM_ACCESS_KIND == "tma_cp":
# shmem operands are being read by the tcgen05_mma
assert "Buffer being accessed has outstanding reads" in result.driver_stderr_output
assert (("Buffer being accessed has outstanding reads" in result.driver_stderr_output) or
(TWO_CTAS
and "Barrier used before initialization or after invalidation" in result.driver_stderr_output))
elif MEM_ACCESS_KIND in ["tmem_load", "tmem_store"]:
# tmem is being written by the tcgen05_mma
assert "Buffer being accessed has outstanding writes" in result.driver_stderr_output
assert (("Buffer being accessed has outstanding writes" in result.driver_stderr_output) or
(TWO_CTAS
and "Barrier used before initialization or after invalidation" in result.driver_stderr_output))
else:
assert result.exc is None
assert result.driver_stderr_output == ""
Expand Down Expand Up @@ -974,7 +978,7 @@ def kernel(input_desc, output_desc, FAILURE: ttgl.constexpr, MEM_ACCESS_KIND: tt
acc = blackwell.allocate_tensor_memory(ttgl.float32, [block_m, block_n], acc_layout)
mbarrier.init(mma_bar, count=1)
if MEM_ACCESS_KIND == "tma_cp":
tma_bar = mbarrier.allocate_mbarrier(two_ctas=TWO_CTAS)
tma_bar = mbarrier.allocate_mbarrier()
mbarrier.init(tma_bar, count=1)

blackwell.tcgen05_mma(smemA, smemB, acc)
Expand Down
Loading