Suppress placeholder BMS faults during startup grace

[mjc]

Prevent the startup BMS message-age sentinel from falling through to telemetry threshold checks during the existing grace period.

While BMS data is stale, bms_update() now returns without reporting telemetry faults until the grace period expires. Afterward it reports the existing connection fault; fresh telemetry continues through the existing threshold checks unchanged.

[Copilot]

Pull request overview

This PR refines BMS startup behavior so placeholder telemetry values aren’t interpreted as real BMS data, by explicitly tracking when BMS status has been requested and when the first real sample has been received.

Changes:

• Add status_requested / status_received state to BMS telemetry handling, and gate fault evaluation until a request is made and a sample arrives (or a request timeout occurs).
• Route Lisp bridge BMS writes through bms_request_status / bms_store_status helpers instead of directly mutating the BMS struct.
• Update the Lisp BMS loop to call ext-bms early to register a status request.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 2 comments.

File	Description
src/main.c	Updates ext_bms to either record a status request (no args) or store a received sample (with args).
src/bms.h	Extends BMS state to track request/receipt and adds new helper APIs.
src/bms.c	Implements request/receive tracking and gates connection-fault logic until after a request.
lisp/bms.lisp	Adjusts the BMS loop to call ext-bms up front to mark a request before reading and forwarding values.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[mjc]

code review made me unsure about this and then I thought about it. minimal fix posted instead. sorry about the wild ride lmao.

[lukash]

Which case is your code fixing? Seems to me they result in identical behavior because at startup bms->fault_mask = BMSF_NONE;.

[mjc]

before:

bms->fault_mask = BMSF_NONE; // during init

// First bms_update():
fault_mask = BMSF_NONE;
if (42 > 5 && false) { ... } // does not return
if (0.0f < 2.7f) {
    set_fault(&fault_mask, BMSF_CELL_UNDER_VOLTAGE);
}
bms->fault_mask = fault_mask; // now contains false undervoltage

17d3c61 was supposed to fix this but didn't exactly do it right.

To clarify the scope: with a healthy BMS this is likely only present until the BMS Lisp thread makes its first ext-bms call, which normally happens almost immediately. It persists through the startup grace period only when no BMS snapshot is forwarded. So this is primarily a correctness fix for the intended grace period.