Repository navigation
docs(coverage): sync scenario catalog with sv2-spec main - #28
Conversation
| | GEN-13 | Ω1 | Θ1 | Server does not treat a non-printable byte in a client-supplied `STR0_255` field as a protocol error (accepts the message or silently ignores the bad field) | [§3.1](https://github.com/stratum-mining/sv2-spec/blob/main/03-Protocol-Overview.md#31-data-types-mapping) `STR0_255` "Every byte MUST be a printable ASCII character (0x20 to 0x7E). A receiver MUST treat any other byte as a protocol error." | Send a byte-level `SetupConnection` and `OpenStandardMiningChannel`/`OpenExtendedMiningChannel` with one `STR0_255` field at a time set to a control byte or a byte > 0x7E (all other fields valid; `endpoint_host`, `vendor`, `hardware_version`, `firmware`, `device_id`, `user_identity`); assert a framed error or disconnect | `SetupConnection`, `OpenStandardMiningChannel`, `OpenExtendedMiningChannel`, `SetupConnection.Error`, `OpenMiningChannel.Error` | | ||
| | GEN-14 | Ω1 | Θ1 | Server does not treat a non-finite or negative `nominal_hash_rate` as a protocol error (accepts the request or silently ignores the value) | [§3.1](https://github.com/stratum-mining/sv2-spec/blob/main/03-Protocol-Overview.md#31-data-types-mapping) `F32` "A receiver MUST treat NaN, ±Infinity, and negative values (including -0.0) as protocol errors." | Send `OpenStandardMiningChannel`, `OpenExtendedMiningChannel`, and `UpdateChannel` with each invalid `F32` bit pattern (`NaN`, `+Infinity`, `-Infinity`, `-0.0`, negative); assert a framed error or disconnect | `OpenStandardMiningChannel`, `OpenExtendedMiningChannel`, `UpdateChannel`, `OpenMiningChannel.Error`, `UpdateChannel.Error` | |
There was a problem hiding this comment.
clanker finding:
GEN-13 and GEN-14 each bundle active probes on several messages (
SetupConnection,OpenStandardMiningChannel,OpenExtendedMiningChannel, andUpdateChannelfor GEN-14) into one row. The "Symmetric split" rule at the top of the file says scenarios that apply to a standard/extended pair get separate IDs because they exercise different server code paths, and C-SC-9 already files aSTR0_255probe onSetupConnectionunderC-SC, notGEN. With one row, Suite Coverage can only flip to ✅ when every message is covered, and a server can fail one and pass the others.Since IDs are never renumbered, I'd rather split these before merge: one row per message, filed under that message's table (
C-SC,M-OSMC,M-OEMC,M-UC), each citing the §3.1 type rule. GEN-3 and GEN-15 are fine as they are, they are passive checks on everything the server sends.
There was a problem hiding this comment.
ACK, addressed
Update STR0_255 references, add message-specific scenarios for STR0_255 and F32 validation, B0_32 length checks, SetTarget on already-received future jobs, extension 0x0002, and tighten several detection rules. Checked against sv2-spec@main:efade57715f1f5146966a52d1c6e8921ab839728.
8d1468e to
9e80c0e
Compare
I've run an AI in the COVERAGE.md file and asked to check if the references are still up-to-date;
as a results some changes are needed;
checked againt sv2-spec@main:efade57715f1f5146966a52d1c6e8921ab839728