sgl-project/sglang · error · ValueError

kv-canary: lut_len must be non-negative, got {lut_len}

Error message

kv-canary: lut_len must be non-negative, got {lut_len}

What it means

lut_len reports the length of the full-to-SWA LUT passed to the offsets kernel and must be non-negative; negative lengths indicate a shape extraction bug in the caller.

Source

Thrown at python/sglang/kernels/ops/kv_canary/plan/offsets_kernel.py:173

        )
    if write_offsets_len <= 0:
        raise ValueError(
            f"kv-canary: write_offsets_len must be positive, got {write_offsets_len}"
        )
    if write_req_capacity < 0:
        raise ValueError(
            f"kv-canary: write_req_capacity must be non-negative, got {write_req_capacity}"
        )
    if verify_capacity < 0:
        raise ValueError(
            f"kv-canary: verify_capacity must be non-negative, got {verify_capacity}"
        )
    if req_to_token_stride0 <= 0:
        raise ValueError(
            f"kv-canary: req_to_token_stride0 must be positive, got {req_to_token_stride0}"
        )
    if lut_len < 0:
        raise ValueError(f"kv-canary: lut_len must be non-negative, got {lut_len}")
    if not isinstance(has_swa_lut, bool):
        raise ValueError(
            f"kv-canary: has_swa_lut must be bool, got {type(has_swa_lut).__name__}"
        )
    if has_swa_lut and lut_len <= 0:
        raise ValueError("kv-canary: lut_len must be positive when has_swa_lut is True")
    if not has_swa_lut and lut_len != 0:
        raise ValueError("kv-canary: lut_len must be 0 when has_swa_lut is False")

    _require_len(req_pool_indices, "req_pool_indices", bs)
    _require_len(prefix_lens, "prefix_lens", bs)
    _require_len(extend_seq_lens, "extend_seq_lens", bs)
    _require_2d(req_to_token, "req_to_token")
    _require_min_len(lut_tensor, "lut_tensor", max(lut_len, 1))
    _require_min_len(
        out_verify_offsets_scratch,
        "out_verify_offsets_scratch",
        _PLAN_BS_BLOCK_SIZE + 1,

View on GitHub (pinned to 0132848349)

Solutions

1. Use the prepare helper that derives (lut, lut_len, has_swa_lut) from the tensor or None, guaranteeing non-negative lut_len
2. Pass lut_len=0 with has_swa_lut=False when there is no LUT

Example fix

// before
launch_plan_offsets_kernel(..., lut_len=-1, has_swa_lut=False)
// after
launch_plan_offsets_kernel(..., lut_len=0, has_swa_lut=False)

lut_len = max(0, int(lut_len))
assert (lut_len > 0) == has_swa_lut or lut_len == 0

When it happens

Trigger: Calling launch_plan_offsets_kernel with lut_len < 0 (usually derived from int(lut.shape[0]) of a malformed tensor or an arithmetic underflow).

Common situations: LUT length computed as (a - b) that underflows; passing -1 as a 'no LUT' sentinel instead of 0 with has_swa_lut=False.

Understand the failure class

Background: "Must be a positive integer", "Invalid value", "Unsupported": the invalid-argument-value error family, when a library rejects the value you pass — this error's family across 35 libraries.

Related errors

• kv-canary: launch_canary_plan_kernels requires full_to_swa_i
• kv-canary: launch_canary_plan_kernels verify_capacity={verif
• kv-canary: launch_plan_entries_kernel requires req_to_verify
• kv-canary: write_offsets_len must be positive, got {write_of
• kv-canary: write_req_capacity must be non-negative, got {wri

AI-assisted analysis of sgl-project/sglang@0132848349 (2026-08-28). Data as JSON: /api/errors/32ef25ca632fd395. Report an issue: GitHub.