luci-app-attendedsysupgrade: add uci-defaults script support by firlin123 · Pull Request #8703 · openwrt/luci

firlin123 · 2026-06-14T02:59:56Z

Pull request details

Description

Added the ability to specify uci-defaults script (defaults field in request to ASU) to be sent to ASU server during firmware generation.

Note: This is my first PR ever and I'm an ESL speaker. Let me know if changes are needed!

Screenshot or video of changes (if applicable)

Configuration tab screenshot:

Upgrade modal screenshot:

Warning in upgrade modal screenshot:

Tested on

Board: Raspberry Pi Compute Module 4 (bcm27xx/bcm2711)
OpenWrt version: OpenWrt 25.12.2 (r32802-f505120278) and 25.12.4 (r32933-4ccb782af7)
LuCI version: LuCI openwrt-25.12 branch (26.158.67103~~e99e132) and LuCI openwrt-25.12 branch (26.163.74024~~5476212)
Web browser(s): Brave 1.90.121 (Official Build) (64-bit), Chromium: 148.0.7778.96

Checklist

This PR is not from my main or master branch 💩, but a separate branch. ✅
Each commit has a valid ✒️ Signed-off-by: <my@email.address> row (via git commit --signoff).
Each commit and PR title has a valid 📝 <package name>: title first line subject for packages.
Incremented 🆙 any PKG_VERSION in the Makefile.
(Optional) Includes what Issue it closes (e.g. openwrt/luci#issue-number).
(Optional) Includes what it depends on (e.g. openwrt/packages#pr-number in sister repo).

openwrt-ai

Reviewed the single new commit. The change is well-scoped: the new 40_luci-attendedsysupgrade uci-defaults script matches the established cache-clear/rpcd-reload convention used by other luci apps, the rpcd ucode get_defaults method (fs.stat/fs.readfile, size cap) is correct API usage, the ACL entry is wired to the new method, and the defaults field is correctly omitted from the build request when unset. Two minor inline notes below (trailing whitespace; silent rejection of oversized/invalid files). The commit message line-length warning is already reported by the formality bot.

Generated by Claude Code

Allow users to specify a custom first-boot script (uci-defaults) to be included in the firmware build request to the ASU server. Signed-off-by: Vladislav Pereyaslov <firlin123@gmail.com>

Introduce dedicated error codes for uci-defaults script handling to display explicit frontend warnings if the file is missing, empty, or exceeds the 40 KiB ASU server limit. Signed-off-by: Vladislav Pereyaslov <firlin123@gmail.com>

efahl · 2026-06-14T18:44:25Z

+		};
+	}
+	// Enforce the 40 KiB official ASU server limit
+	if (st.size > 40960) {


You can't hard code this number as it varies between servers and can change arbitrarily even on the "official" upstream. You need to fetch it from the server limits: curl https://sysupgrade.openwrt.org/json/v1/overview.json | jsonfilter -e '$.server.max_defaults_length'

Removed the hard-coded number and now using server's max_defaults_length in the latest commit.

efahl · 2026-06-14T18:51:04Z

+const ERR_FILE_TOO_BIG = 3;
+
+function get_defaults() {
+	const path = uci.cursor().get("attendedsysupgrade", "client", "defaults_path");


This value is already in the config in attendedsysupgrade.owut.init_script. See https://openwrt.org/docs/guide-user/installation/sysupgrade.owut#using_a_uci-defaults_script

I was not aware that a config for this already existed. Switched to using attendedsysupgrade.owut.init_script in the latest commit.

efahl · 2026-06-14T18:54:46Z

 						profile,
 						version: candidates[0][0],
 						packages: Object.keys(packages).sort(),
+						defaults_checked: '1'


Shouldn't this be a bool?

I tried making it bool but the form.Flag only accepts "1" and 1 (which it type-coerces to "1" anyway).

efahl · 2026-06-14T18:59:37Z

+			form.Value,
+			'defaults_path',
+			_('Defaults file path'),
+			_('Specify the path to your first-boot script (e.g., <code>/rom/etc/uci-defaults/99-asu-defaults</code> inserted by the Firmware Selector.)')


If no config value exists, but the /rom/.../99-asu-defaults file exists, this field should auto-populate with the existing file path.

Because I switched to attendedsysupgrade.owut.init_script as you suggested, I don't control the defaults of this config anymore, owut does.

openwrt-ai

Reviewed 1 new commit (d4e5e16 "improve uci-defaults validation"); no new issues found.

The new commit cleanly addresses both notes from my prior review: oversized/invalid/empty files now surface an explicit alert-message warning instead of silently dropping the option, and the trailing whitespace is gone. The ucode error-code handling (fs.stat type/size checks, length(trim(contents)) emptiness check) is correct, the expect/resolveDefault defaults were updated consistently, and the commit message matches the diff. Wrapping data.defaults_path in E('code', {}, ...).outerHTML also correctly escapes the path before it reaches innerHTML.

The substantive open points (hardcoded 40 KiB limit vs. the server's max_defaults_length, reuse of attendedsysupgrade.owut.init_script, the bool question, and path auto-population) are already covered in @efahl's inline comments on this head, so I won't duplicate them.

Generated by Claude Code

openwrt-ai

Reviewed 1 new commit (1327ff7 "switch to owut config and server-driven limits"). The switch to attendedsysupgrade.owut.init_script and the server-supplied max_defaults_length correctly addresses @efahl's earlier notes, and the new ERR_EMPTY_FILE/ERR_SERVER_FAILED/ERR_INIT_DISALLOWED error paths are wired consistently between the rpcd ucode and the frontend. One inline note on a TypeError reachable when the overview request rejects for a SNAPSHOT device.

Generated by Claude Code

openwrt-ai

Reviewed 1 new commit (c5482d5, replacing the previously-reviewed 1327ff7); no new issues found.

The only delta since my last review is the overview.js refactor that hoists init_script_path out of the branches and guards results[0] with optional chaining (results[0]?.ok). This correctly addresses the TypeError I flagged previously: when the overview request rejects (e.g. for a SNAPSHOT device), results[0] is the resolveDefault fallback rather than a response object, and ?.ok now yields null → server_info = null → ERR_SERVER_FAILED. The added typeof server_info.max_defaults_length !== 'number' check also avoids passing NaN into callGetInitScript(Math.floor(...)) when the server advertises allow_defaults without a numeric limit. The commit subject (75 chars) and message match the diff, and Test Formalities now passes on this head.

Generated by Claude Code

firlin123 · 2026-06-23T17:58:59Z

Fixed eslint no-case-declarations error in the latest force push

openwrt-ai

Reviewed the force-pushed head (4311997e, replacing the previously-reviewed c5482d5); no new issues found.

The only delta since my last review is wrapping the default: case body in overview.js (handleCheck's switch (init_script_error)) in a block { }. This correctly resolves the eslint no-case-declarations error caused by the lexical declarations (let msg, let alertBox, const kib) in that case, with no behavioral change. The reworded third commit subject is now 75 chars and matches its diff.

Generated by Claude Code

efahl · 2026-06-24T16:15:42Z

@@ -61,6 +66,22 @@ return view.extend({
 		o.default = '0';
 		o.rmempty = false;

+		// For now, only display if the owut section exists in uci (owut installed or section is manually added).
+		// We should probably move init script path from attendedsysupgrade.owut.init_script to some common place


Yeah, too late now. That would require a significant rewrite of how owut gather's uci default values, and corresponding upgrade scripting, so not going to happen. LuCI already grabs another owut value (rootfs_size), and will always be a second class citizen compared to the far-more-capable owut.

So should we not display that config at all and only make it configurable via cli (like the owut.rootfs_size)? Or just remove the "for now" phrasing from the comment?

efahl · 2026-06-24T16:29:15Z

+#!/bin/sh
+rm -rf /var/luci-modulecache/; rm -f /var/luci-indexcache;
+[ -x /etc/init.d/rpcd ] && /etc/init.d/rpcd reload
+exit 0


Why is this whole file here? The tmp directories are deleted on reboot, and the rpcd reloading isn't anything that attended sysupgrade should care about.

I honestly don't know. I just looked at other luci apps with ucode rpcd service and they all seem to have this file so I copied it here too. Is it not needed in this case?

efahl · 2026-06-24T16:36:28Z

+		return {
+			contents: "",
+			path: path,
+			error: ERR_EMPTY_FILE


If people want to include an empty placeholder script, I don't see any reason to disallow that.

Allowed empty files in latest revision.

…limits Reuse exiting 'owut.init_script' instead of separate 'client.defaults_path'. Use ASU server's 'max_defaults_length' instead of hardcoded 40 KiB limit. Signed-off-by: Vladislav Pereyaslov <firlin123@gmail.com>

This comment has been minimized.

Sign in to view

openwrt-ai reviewed Jun 14, 2026

View reviewed changes

Comment thread ...luci-app-attendedsysupgrade/htdocs/luci-static/resources/view/attendedsysupgrade/overview.js Outdated

Comment thread applications/luci-app-attendedsysupgrade/root/usr/share/rpcd/ucode/luci.attendedsysupgrade Outdated

firlin123 force-pushed the luci-app-attendedsysupgrade-defaults branch from 1888178 to d5eb355 Compare June 14, 2026 11:42

luci-app-attendedsysupgrade: add uci-defaults script support

79a264f

Allow users to specify a custom first-boot script (uci-defaults) to be included in the firmware build request to the ASU server. Signed-off-by: Vladislav Pereyaslov <firlin123@gmail.com>

This comment has been minimized.

Sign in to view

firlin123 force-pushed the luci-app-attendedsysupgrade-defaults branch from d5eb355 to d4e5e16 Compare June 14, 2026 15:53

efahl reviewed Jun 14, 2026

View reviewed changes

openwrt-ai reviewed Jun 14, 2026

View reviewed changes

This comment has been minimized.

Sign in to view

firlin123 force-pushed the luci-app-attendedsysupgrade-defaults branch from 33b9e93 to 1327ff7 Compare June 15, 2026 03:00

openwrt-ai reviewed Jun 15, 2026

View reviewed changes

Comment thread ...luci-app-attendedsysupgrade/htdocs/luci-static/resources/view/attendedsysupgrade/overview.js Outdated

firlin123 force-pushed the luci-app-attendedsysupgrade-defaults branch from 1327ff7 to c5482d5 Compare June 15, 2026 08:46

openwrt-ai reviewed Jun 15, 2026

View reviewed changes

firlin123 requested a review from efahl June 16, 2026 05:42

firlin123 force-pushed the luci-app-attendedsysupgrade-defaults branch from c5482d5 to 4311997 Compare June 23, 2026 17:56

openwrt-ai reviewed Jun 24, 2026

View reviewed changes

efahl reviewed Jun 24, 2026

View reviewed changes

firlin123 force-pushed the luci-app-attendedsysupgrade-defaults branch from 4311997 to 83ae949 Compare June 24, 2026 18:10

firlin123 force-pushed the luci-app-attendedsysupgrade-defaults branch from 83ae949 to ae849bc Compare June 24, 2026 18:10

Uh oh!

Conversation

firlin123 commented Jun 14, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Pull request details

Description

Screenshot or video of changes (if applicable)

Tested on

Checklist

Uh oh!

This comment has been minimized.

openwrt-ai left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

This comment has been minimized.

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

firlin123 Jun 15, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

firlin123 Jun 15, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

openwrt-ai left a comment

Choose a reason for hiding this comment

Uh oh!

This comment has been minimized.

openwrt-ai left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

openwrt-ai left a comment

Choose a reason for hiding this comment

Uh oh!

firlin123 commented Jun 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

openwrt-ai left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

firlin123 commented Jun 14, 2026 •

edited

Loading

firlin123 Jun 15, 2026 •

edited

Loading

firlin123 Jun 15, 2026 •

edited

Loading

firlin123 commented Jun 23, 2026 •

edited

Loading