Skip to content

wifi-xm530: what the image ships for XM711, without mac80211, and the ATBM6032 is another chip - #653

Merged
openipc-ai merged 2 commits into
masterfrom
wifi-xm530-current
Nov 14, 2026
Merged

openipc-ai merged 2 commits into
masterfrom
wifi-xm530-current

Conversation

@openipc-ai

Copy link
Copy Markdown
Collaborator

The page dated from 2022. It had users copy a cfg80211.ko from the factory firmware and load mac80211.ko with it, and it did not say it was only for the XM711 module. A reporter with an ATBM6032 board followed it in OpenIPC/firmware#2853 and got mac80211: Unknown symbol __cfg80211_send_disassoc / cfg80211_send_rx_auth ….

What changed, each point checked against the 2026-11-13 xm530_lite nightly:

  • XM711: the image already ships /usr/bin/wifi, xm711.ko, compat.ko, wifi_pdn.ko, dwc_otg.ko and the vendor's backported cfg80211.ko at the kernel path. So the page says wifi xm711 instead of hand-installing a script and a factory module. Its modules.dep has xiongmai/xm711.ko: kernel/net/wireless/cfg80211.ko xiongmai/compat.ko, with no mac80211.
  • The mac80211 line is gone. The firmware's own copy of this script has carried it commented out (# not used with grabbed cfg80211). Loading it against the backport fails with the unknown symbols above, which I reproduced on the qemu-bench XM550 machine.
  • How to tell the chips apart: power the module and read its USB ID.
  • ATBM6032 (007a:8888, XM713G) is another chip. The stock image has no driver or profile for it, and wlandev takes a profile name, not a module name. The factory xm713g.ko needs the factory cfg80211.ko: I compared the one attached to #2853 (md5 e8dcd9df…). It exports the same symbols as OpenIPC's kernel module plus 11 cfg80211_wext_*, meaning a stock 3.10 cfg80211 with CONFIG_CFG80211_WEXT. It is not the backport. The section points at xm530: bring up the AltoBeam ATBM6032 USB WiFi chip firmware#2316 and xm530: add anbiux A8B-3MP (IPC-RB-BLK530AI, ATBM6032 USB WiFi) builder#130 for the open driver, and at build: run the final depmod on kernels that see this checkout's git firmware#2854 for why earlier builds could not load it.

@qodo-free-for-open-source-projects

Copy link
Copy Markdown

PR Summary by Qodo

Clarify XM530 Wi-Fi setup for XM711 and ATBM6032

📝 Documentation 🕐 10-20 Minutes

Grey Divider

AI Description

• Scope the guide to first-generation XM530/XM550 images and identify Wi-Fi chips by USB ID.
• Use the shipped XM711 loader instead of copying factory modules or loading incompatible mac80211.
• Explain why ATBM6032 needs separate support and link to ongoing driver work.
Diagram

graph TD
  A["XM530 lite"] --> B["Power USB"] --> C{"USB chip?"} --> D["XM711 loader"] --> E["Network config"]
  C --> F["ATBM support pending"]
Loading
High-Level Assessment

Documenting chip identification before setup and using the loader already shipped with the image is appropriate. Recommending copied factory modules or a generic profile would obscure the distinct driver requirements and current ATBM6032 support gap.

Files changed (1) +61 / -18

Documentation (1) +61 / -18
wifi-xm530.mdSeparate XM711 setup from ATBM6032 guidance +61/-18

Separate XM711 setup from ATBM6032 guidance

• Scopes the guide to first-generation XM530/XM550 images and adds USB ID checks before selecting a Wi-Fi path. Replaces the hand-installed XM711 script with the shipped 'wifi xm711' command, warns against loading incompatible 'mac80211.ko', and explains ATBM6032's missing stock support and related work.

en/wifi-xm530.md

@qodo-free-for-open-source-projects

qodo-free-for-open-source-projects Bot commented Nov 14, 2026 •

Copy link
Copy Markdown

Code Review by Qodo

🐞 Bugs (0) 📘 Rule violations (0) 📎 Requirement gaps (0) 🎨 UX issues (0) 🔗 Cross-repo conflicts (0) 📜 Skill insights (0)

Grey Divider


Remediation recommended

1. USB chips appear under mismatched IDs ✓ Resolved
Description
The USB-ID command prints every idVendor value before any idProduct value, without showing which
device each value belongs to. When more than one USB device is listed, a reader cannot reliably
identify the 007a:8888 pair used to distinguish the ATBM6032.

2. Boards with another power pin stay unseen ✓ Resolved
Description
The chip-detection procedure loads wifi_pdn.ko with value=96 before telling readers that the
power GPIO varies by board. On a board using another GPIO, following the command as written does not
power its Wi-Fi module, so the subsequent USB-ID check cannot reveal that chip.

Grey Divider

Tip of the day
💡 Did you know, you can ask Qodo to dismiss a finding you disagree with, with your reason on record

More tips ↗ | Customize Qodo ↗ | Qodo docs ↗

Grey Divider

Qodo Logo

@openipc-ai

Copy link
Copy Markdown
Collaborator Author

Both Qodo findings are fixed in the latest commit:

  1. USB IDs printed unpaired: the probe now loops over the devices and prints vendor:product per device, and it says that 1d6b:0002 is the USB host itself. It is the same loop the qemu-bench test uses, which prints 007a:8888 and 1d6b:0002 for the ATBM dongle.
  2. The power GPIO was explained after the command that hardcodes it: the page now says first where to read the value (the factory firmware's wifi_pdn value= line), and the probe uses $PDN. A probe with no second device points back at it. The XM711 section also notes that /usr/bin/wifi hardcodes 96, and how to change it.

@openipc-ai
openipc-ai merged commit 0d99a62 into master Nov 14, 2026
@openipc-ai
openipc-ai deleted the wifi-xm530-current branch November 14, 2026 13:24
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.

1 participant