Three firmware builds in one hour, and a published root password
Supports running OpenWrt on Amlogic, Allwinner, and Rockchip devices. Support a311d, s922x, s905x3, s905x2, s912, s905d, s905x, s905w, s905, s905l, rk3588, rk3568, rk3399, rk3328, h6, etc.
At a glance
- What is it?
- amlogic-s9xxx-openwrt compiles a router operating system for TV boxes and other boards from three chip vendors, and most of its readme is a device compatibility table. Its default information table publishes the administrator account, its password and a fixed address, and its releases come in threes cut from three different upstream trees.
- Who is it for?
- This project suits someone who owns a specific box from the listed models and wants a router system on it with current packages rather than whatever the vendor shipped, and the device table plus the per-model kernel column make it possible to check compatibility before flashing anything. Three things to do first.
- Can I use it commercially?
- Yes, with conditions. GPL-2.0 is a copyleft licence: if you distribute software that includes it, you must release that software's source code under the same licence. Running it internally without distributing it does not trigger that obligation.
- Is it still maintained?
- Yes. The repository received new commits within the last day.
- What is it written in?
- Mainly Shell, according to GitHub's language statistics.
Answers come from the project's GitHub data, last synced on November 15, 2026, and from our analysis. They are not legal advice.
Star history
- ophub/amlogic-s9xxx-openwrt3,095 ★ · +68
| Repository | 2026-09-24 | 2026-11-15 |
|---|---|---|
| ophub/amlogic-s9xxx-openwrt | 3k | 3.1k |
68 Stars in the last 52 days, measured from the daily GitHub snapshot. Compare with other projects
Editorial analysis
The default credentials are printed in a table
Near the top of the readme there is a table headed as the default information for the firmware, and it has five columns: system name, default user name, default password, port and address.
The values are the same for both rows. The account is the superuser account that this operating system family uses. The password is the word password. The port is the standard SSH port. The address is the conventional first host of a private range.
Publishing that is defensible as a first-login convenience, and it is how a surprising number of router projects behave, because the alternative is a device that nobody can get into. The problem is the combination rather than any one value. A board flashed from one of these images and plugged into a home network is reachable over SSH, on a well known port, with a credential printed on the project's front page.
The readme does mention the ability to install to eMMC and to update the kernel, so the expectation is that people do flash these images onto storage. Change the password before anything else.
Three release variants, cut within an hour from three different trees
The three most recent release tags are the clearest thing this repository says about itself.
They are not version numbers. Each is a compound name: a prefix for this project, then the name of the upstream tree it was built from, then the branch within that tree, then a marker, then a year and month. So one tag names the current upstream tree's main branch, one names a master branch on a fork that shares the project's lineage, and one names a master branch on a third fork that does the same.
All three were published on the same morning, within about an hour of each other.
So a release here is not a version, it is a monthly snapshot of three different systems. Anyone comparing two firmwares needs to compare the tags, because the number in the tag is a date and the rest of it is the source. It also means the project is not a port of one upstream: it maintains three build targets, and a change in any one of those trees can produce a different router from the same month.
Most of the readme is a device table, and the models live in another repository
The bulk of this readme is a compatibility table, and its shape tells you where the real engineering work is.
The rows are chip families, and the table is explicitly ordered by processor performance from high to low. Three vendors are covered: one supplies most of the boxes, one supplies the higher-numbered parts in its range, and a third supplies the rest. One row alone lists more than twenty models.
The second column is the interesting one. Almost every model name is a link into an issue on a *different* repository in the same organisation, one named for the board-support distribution rather than for this firmware. So the device database is not a file here; it is a history of issue discussions maintained somewhere else, and this table is an index into it.
The third column links to a third repository, the one that publishes the kernels. Every entry visible in that column points at the same stable kernel tag rather than at a per-device build, which tells you that by default the firmware expects one kernel to serve the whole table.
There are data quality wrinkles too. Three models in the largest row all resolve to the same issue, one model name appears twice in a row, and every one of those entries is a link that may or may not still resolve.
There is no source directory, because this is a build recipe
Look at the tree and the shape of this project becomes clear.
The repository has no source directory, no tests and no container file. What it has is a build directory, a configuration directory, a documentation directory, a workflow definition, and two files at the root that are not directories and have no extension.
The language of the repository is shell, which fits: this is a set of scripts that fetch upstream sources, apply configuration, build firmware images and publish them. The workflow definition at the root means that build runs in continuous integration, and the releases are its output rather than something produced on a maintainer's machine.
The file at the root with no extension is not explained anywhere in the readme. Neither is what the two scripts in the root-level directory do individually. So the entry point for anyone wanting to build this themselves is the user guide in the documentation directory, not the repository listing, which is a normal arrangement for a firmware build project and worth knowing before you start reading shell scripts.
The intended device is a television box on a home network
The readme opens with a description of the upstream operating system, and then makes a specific proposal in one sentence: you can replace the Android TV system on your TV box with this one, turning it into a router.
That sentence sets the whole context. The target is not an industrial board or a development kit, it is a consumer set-top box that shipped with an Android-based interface, sitting in a living room, on a home network, plugged into a television.
Two consequences follow that the readme does not draw. The default credentials table describes a device that is trivially reachable from anything else on the same network. And replacing a vendor's Android build means the box stops being the device its software was written and tested for, which is why the compatibility table is as long and as specific as it is.
The readme also credits the contributors by linking to the board-support repository, which is the same repository that holds the device discussions. Three repositories from one author, then: firmware, board support, and kernels.
Editorial conclusion
This project suits someone who owns a specific box from the listed models and wants a router system on it with current packages rather than whatever the vendor shipped, and the device table plus the per-model kernel column make it possible to check compatibility before flashing anything. Three things to do first. Change the default credentials immediately after flashing, because the account name, its password, the port and the address are all published in the readme, and those devices sit on home networks. Pick the upstream tree deliberately, because the three release variants are different systems rather than flavours. And read the user guide in the documents directory before you write to storage, since installation to eMMC is destructive and the readme does not spell out the recovery path.
Frequently asked questions
What is OpenWrt?
A Linux operating system for embedded devices that uses a fully writable filesystem with package management rather than a single fixed firmware image, so the software on the device can be changed after it ships. The readme this project carries describes it as a framework for developers as well as a router system for end users, with thousands of packages available.
What is an OpenWrt router?
That is the upstream project's term rather than this repository's. This repository builds that system for set-top boxes and other boards from three chip vendors, and its default information table lists the superuser account, the standard SSH port and a fixed address in the first private range.
Is OpenWrt still relevant?
This repository does not answer that. What it demonstrates is a monthly release cadence in which three variants are cut from three different upstream trees within a single hour, and a device table maintained model by model against a kernel published in a separate repository.
What is OpenWrt on my WiFi?
This repository does not explain that. It documents how to build firmware for the boards it lists, what credentials the flashed system starts with, and where the user guide lives. Anything about a router's appearance on a network is a question for the firmware's own documentation or for whoever configured it.
Official sources
Add this badge to your README
If you maintain this project, the badge below links readers to this analysis and shows its maintenance status from the daily GitHub snapshot. Paste the markdown into your README; add ?metric=license or ?metric=stars to the image URL for a different field.
[](https://hysenlabs.com/projects/ophub-amlogic-s9xxx-openwrt)Star history chart for your README
[](https://hysenlabs.com/projects/ophub-amlogic-s9xxx-openwrt)Was this useful?