Skip to content

device/lynx: record AoC-first Wi-Fi association and HTTPS - #751

Merged
derekross merged 2 commits into
derekross:mainfrom
latentoperator:docs/lynx-aoc-wifi-results-20261113
Nov 15, 2026
Merged

derekross merged 2 commits into
derekross:mainfrom
latentoperator:docs/lynx-aoc-wifi-results-20261113

Conversation

@latentoperator

@latentoperator latentoperator commented Nov 14, 2026 •

Copy link
Copy Markdown
Contributor

What and why

The lynx Wi-Fi plan and matrix still stop at the ROM stall. Following the AoC-first report on #36, the original test phone has now passed one guarded firmware/scan, association, DHCPv4 and verified HTTPS run on a local experimental kernel.

Record the result in a new device/lynx/AOC-WIFI-PLAN.md, link it from the existing Wi-Fi investigation, add dated device notes, and change only lynx's Wi-Fi status in the matrix to WIP. Keep the earlier WMI/TX stall and USB loss alongside the successful MSI-candidate retest. One success does not identify the earlier failure's cause or qualify the MSI change as a fix.

The kernel changes and overlays remain unpublished. Drivers stay blacklisted at boot; after the owner's subsequent reboot, guarded Wi-Fi startup has not been repeated. Reconnect, sustained traffic, suspend/resume, Bluetooth coexistence and unattended startup remain unqualified. No release enablement is proposed.

Device and build

  • Phone: Pixel 7a (lynx), the original test unit from Pixel 7a (lynx, gs201 IPOP) port: stock facts and first light #36.
  • Successful connection kernel: 7.3.0-rc6-omamobile-lynx-msi-20261113+, published kernel base 2db2bc35beecbf2deccb7f2e9bc8c3843b1edc9c plus local changes described in the new plan. The base alone does not reproduce the build.
  • Earlier firmware/scan kernel: 7.3.0-rc6-omamobile-lynx-wifi-aoc-20261113+.
  • Packages changed: none; four Markdown files only.

Checks

  • tools/pr-check origin/main: pass (identifiers, blobs, attribution and scripts).
  • tests/all: pass.
  • tests/packages: 31/32 package suites pass; only omamobile-modem fails at "the old overlay still loads after a failure at object" with omamobile-modem-nv-overlay: error: [Errno 32] Broken pipe. The same failure reproduces with tests/packages omamobile-modem on a separate untouched origin/main checkout (6649b76). No package or test files differ in this PR.
  • New relative links and heading anchor, matrix column counts, saved-serial exclusion and SPDX headers checked; git diff --check passes.
  • Area-prefixed commits with explanatory bodies and Assisted-by: OpenAI Codex; no package version change applies.

Recorded hardware results

The November 13 connection run passed Trusty IPC, AoC (82 services, zero restarts), PCIe/MSI, one stock firmware load, passive scan (24 BSS entries), association, DHCPv4, three gateway pings and Wi-Fi-bound HTTPS (HTTP 200, 577 bytes, TLS verified). Fresh SSH confirmed USB access; six protected hashes matched, the storage guard held and all six required smoke checks passed. No new WMI/TX timeout, oops or panic appeared in the connection capture. Watchdogs and rollback timers were disarmed after fresh healthy status.

The full saved smoke report below also retains the optional failed CMA-reservation and missing GPU-render-node rows. Neither is qualified by the Wi-Fi result.

No phone operation was performed to prepare this documentation PR, and no fresh smoke run is needed for a docs-only change (CONTRIBUTING section 6.5). Firmware, raw captures, network identifiers, partition images and calibration hashes remain private. No protected-partition write or storage-guard change is included.

November 14 documentation follow-up

Address the earlier #439/#88 review follow-ups in WIFI-PLAN.md: record kernel Bluetooth #98 as merged, correct the PERIC0 pin-bank reference, remove an unattributed helper-review claim, and mark the host-stack/calibration-wait hypothesis as answered by the recorded negative fs_ready test with a direct source link. Independent source review, tools/pr-check, git diff --check, and repository tests/all pass locally. This follow-up changes one Markdown file and performs no phone operation. Fresh head CI passes both Policy and Tests (repository tests and every package) on 206a83815ca6cdc43c67c63758c46a31aa1e0f72.

Historical phone-smoke report from the successful connection boot

phone-smoke: lynx

Kernel 7.3.0-rc6-omamobile-lynx-msi-20261113+, tools at 12258d9, saved after the successful 2026-11-13 connection run (uptime 1384 s).

Check Result Detail
boot (required) ok kernel 7.3.0-rc6-omamobile-lynx-msi-20261113+, slot b, up 1384s
storage guard (required) ok 55 partitions; only userdata and slot B's boot images writable
root (required) ok ext4 on userdata, 99G free
kernel health (required) ok no oops or BUG; 0 WARN, 6 error lines in dmesg (5 known, 1 new)
new dmesg errors not ok cma: Failed to reserve N MiB x1
systemd units ok none failed
USB gadget (required) ok bound to 11210000.usb
display ok card0-Unknown-1 connected, 1080x2400
GPU not ok no render node
session (required) ok Hyprland and the Omarchy shell running
lock ok PAM file present, the shell can lock (passwordPam true, locked false)
touch ok Goodix Berlin Capacitive TouchScreen
buttons skip no key input device found by name
Wi-Fi ok wlp1s0 on ath11k_pci, link up
Bluetooth skip no controller
battery skip no battery power supply
sound cards skip none
sensors skip no sensor proxy
cameras skip no libcamera tools
thermal ok hottest zone 48 °C

Packages: hyprland 0.56.2-7, libcamera 0.7.2-4, mesa 1:26.2.5-1, omacam 0.3.3-1, omamobile-apps 0.2.0-1, omamobile-device-lynx 0.2.0-1, omamobile-firstboot 0.1.0-1, omamobile-keyring 20261013-1, omamobile-mirror 0.1.1-1, omamobile-omarchy 0.1.6-1, omamobile-owl-widgets 0.1.1-1, omamobile-shell 0.2.15-1, omarchy 4.0.6-1, quickshell 0.3.2.r7.g2b4501a-1.

The Wi-Fi plan and matrix still stop at the ROM stall. Record the AoC-first reproduction and one guarded association, DHCP and verified HTTPS test, alongside the earlier WMI/TX failure and recovery. Keep the local kernel provenance, boot-time blacklist and unqualified paths explicit.

Assisted-by: OpenAI Codex
Record the merged Bluetooth tree and the published PERIC0 pin bank,
and remove an unattributed helper-review claim. Mark the host-stack
plus calibration-wait hypothesis as answered by the recorded fs_ready
run, retaining its negative result and the later AoC qualification limits.

Assisted-by: OpenAI Codex
@derekross

Copy link
Copy Markdown
Owner

Thank you, Chris, and thanks to @alfikpl for the lead on #36. This is the first time a lynx has associated and carried traffic, and the record is written the way I'd want: the earlier stall and USB loss are kept beside the pass, and it says plainly that one run with the MSI change doesn't explain the stall. I checked the WIFI-PLAN follow-ups too. Kernel #98 is merged, PERIC0 at 0x10840000 with gpp16 is in the published gs201.dtsi and gs201-pinctrl.dtsi, and run 17 on #36 does answer the host-stack-and-wait row. CI is green and it merges cleanly, so I'm merging it.

A few things for before the next Wi-Fi run, not for this PR:

  1. Please push the local kernel tree as a branch on your fork and put its commit in AOC-WIFI-PLAN section 1 in place of the local worktree path. Then the pass can be tied to a source, and the Trusty and AoC parts, which touch the Pixel 7 Pro's paths, can start their way into kernel PRs from there.
  2. One line on Trusty, please: did any storageproxyd, or anything else that reaches Trusty storage, run on these boots? I'd expect not for an AoC start, but RPMB writes are one of the things I keep a close eye on.
  3. "The watchdogs were explicitly disabled": was that your stage feeder and rollback timers, or the hardware watchdog itself? A few words in the plan would make it clear.
  4. It would be nice to name alfikpl in the plan's opening paragraph, since the ordering and the four mainline fixes came from their report.

Thanks also for reporting the omamobile-modem test failure you hit on main. I'll look at it separately.

@derekross
derekross merged commit 3818f8f into derekross:main Nov 15, 2026
2 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants