[Bugfix][Kernel] Check output alignment in vectorize_with_alignment (fixes misaligned-address crash for non-multiple-of-8 head sizes) (#45466)

Signed-off-by: HumphreySun98 <humphreysun98@gmail.com>
This commit is contained in:
Humphrey
2026-06-18 16:58:22 +00:00
committed by GitHub
parent 21da47dabe
commit 4583630b56
2 changed files with 60 additions and 5 deletions
+37
View File
@@ -428,6 +428,43 @@ def test_reshape_and_cache_flash(
torch.testing.assert_close(value_cache_compact, cloned_value_cache)
@pytest.mark.parametrize("dtype", DTYPES)
@pytest.mark.parametrize("kv_cache_dtype", KV_CACHE_DTYPE)
@pytest.mark.parametrize("kv_cache_layout", CACHE_LAYOUTS)
@pytest.mark.parametrize("implementation", RESHAPE_FLASH_IMPLEMENTATIONS)
@torch.inference_mode()
def test_reshape_and_cache_flash_unaligned_rows(
kv_cache_factory_flashinfer,
dtype: torch.dtype,
kv_cache_dtype: str,
kv_cache_layout: str,
implementation: str,
) -> None:
"""Regression test for https://github.com/vllm-project/vllm/issues/41257.
head_size=46 with num_heads=13 places KV-cache rows at byte offsets
that are not a multiple of the vector width (NHD row pitch
13*46*itemsize, HND head pitch 46*itemsize), unlike HEAD_SIZES above
which are all 16-byte multiples. The CUDA kernel used to issue
vectorized stores to those rows -> CUDA misaligned address.
"""
test_reshape_and_cache_flash(
kv_cache_factory_flashinfer,
num_tokens=42,
num_heads=13,
head_size=46,
block_size=16,
num_blocks=128,
dtype=dtype,
seed=0,
device=CUDA_DEVICES[0],
kv_cache_dtype=kv_cache_dtype,
kv_cache_layout=kv_cache_layout,
kv_scale_type="tensor",
implementation=implementation,
)
@pytest.mark.parametrize("direction", COPYING_DIRECTION)
@pytest.mark.parametrize("num_mappings", NUM_MAPPINGS)
@pytest.mark.parametrize("num_heads", NUM_HEADS)