Bring the Spanning Tree page in line with a typical managed switch's
per-port panel. Configuration gains the full-range path cost (raw
0..200000000, 0 = auto, replacing the old 1000x-scaled byte), a
point-to-point admin control (auto/on/off), and the priority is now a
0..240 step-16 dropdown. A new status table shows, per port, the Port
State, Role, Designated Bridge / Port ID / Cost (learned from received
BPDUs, kept per port and aged via the BPDU age), Operational Edge and
Operational Point-to-Point.
The designated fields fall back to presenting this switch as the
segment's designated bridge when no fresh BPDU has been heard (so a
quiet port shows our own bridge-id, as the vendor UIs do). /stp.json
carries the packed hex fields plus our own MAC for that fallback.
Space: reclaim BANK2 for the above by moving rtl837x_pins to HOME and
compacting leds_dump into a register-address table (~800B); bandwidth
returns to BANK1. No BANK3 - hardware-verified that PSBANK > 2 crashes
this SoC at boot (a bricked unit and an SPI-programmer recovery earlier
today); a warning to that effect is now in rtl837x_lldp.c.
Hardware-verified: cost 200000000 and p2p off round-trip through the CLI
and JSON, the status table populates correctly with STP enabled (all
ports Forwarding/Designated, oper-edge and oper-p2p True), LACP 3f/3f
and the LAN unaffected.
(cherry picked from commit 2ec62072f061dc9e78bc821ba1c297cb6819e206)
Enabling STP on a bridge whose management rides an in-band VLAN can cut
off that very management - and not only by our own blocking: on this
network the upstream TP-Link Easy Smart switch's "loop prevention"
reacted to our BPDU hellos by blocking ITS port towards us while our
ASIC was all-forwarding, isolating the whole segment until a power
cycle. Recoverable only by going quiet.
Add a commit-confirm watchdog: while STP is enabled, any HTTP request
re-arms a countdown ("stp failsafe <seconds>", default 180, 0 disables);
if management stays silent for the whole window, STP disables itself,
which also stops BPDU TX so a neighbour's loop protection can release
its block. The web UI polls /stp.json every 2 s, so an open browser
naturally keeps the watchdog re-armed. The trip is reported via
/stp.json (fs, fsT) and as a warning on the Spanning Tree page.
Deliberately not conditioned on our own MSTP port states - the incident
above proves the uplink can be dead while every local port forwards.
Also bound the NIC DMA busy-waits (nic_tx_packet, nic_rx_header,
nic_rx_packet): an unbounded spin on SFR_NIC_CTRL freezes the entire
main loop (timers, HTTP, ARP) if the ASIC ever fails to consume a
transfer; give up after ~65k polls and drop the frame instead.
Hardware-verified end to end: with priority 15 against a live RSTP
bridge the uplink died 6 s after "stp on" and the network recovered BY
ITSELF 66 s later (trip at 45 s + neighbour release), fsT=1, LACP and
LAN intact. Telemetry via syslog-to-edge-port host confirmed the full
chain: countdown 44->4, trip, hello TX stopping at the trip.
(cherry picked from commit 1fa9775156fd6d7ebfdda2382f73430b86601230)
Implements the standard 802.1D-2004/802.1w configuration surface:
Bridge: priority (0-15 x4096), hello time, max age, forward delay,
force-version (RSTP v2 / STP-compatible v0 Config BPDUs), tx hold
count (per-port per-second BPDU budget).
Per port: enable, admin edge (forwarding immediately - no listen gap),
auto edge (forwarding after 3 s of BPDU silence; DEFAULT, so
host-facing ports no longer take the full forward delay),
path cost (0=auto/20000), port priority, BPDU guard (port disabled
on BPDU receipt), root guard (never accept a better root on the
port), BPDU filter (no BPDUs in or out).
Engine additions: root max-age expiry (reclaim the tree when the root goes
silent), root path cost accounting (rx cost + root-port cost, advertised in
our BPDUs), loop detection (our own BPDU coming back blocks the port for a
listen period), topology-change counter, approximated per-port roles
(Root/Designated/Alternate) for diagnostics.
CLI: "stp prio|hello|maxage|fwd|txhold|version ..." and
"stp port <n> on|off|edge|cost|prio|guard|filter ..." (stp_parse, delegated
from cmd_parser); all forms accepted by the startup-config validator so the
whole configuration persists. /stp.json now reports config + status; the
Spanning Tree page exposes everything with immediate-apply controls and live
state/role columns (edit-in-flight guard against the 2 s refresh).
8051 memory: the module moves to code BANK2; internal-RAM pressure from
cross-bank calls resolved by xdata loop iterators/scratch, __reentrant on
the small helpers, and moving httpd's header-pointer globals to xdata.
Verified on hardware (SWTGW218AS): defaults land per standard; priority and
hello change live; admin-edge ports (the LACP bond uplinks) keep the LAN at
0% loss THROUGH "stp on"; auto-edge ports forward after 3 s; a port that
heard real BPDUs (a VM bridge behind physical port 6) correctly declined
auto-edge, sat out the full listen period and became Designated; tc counts
promotions; we win the root election at priority 16384 vs 32768.
(cherry picked from commit 09a34dc6acdc81ab9cab0727d2f4a59c68131a3e)
Add a Spanning Tree page: an on/off toggle driving the existing "stp"
command over /cmd, and a live status section fed by a new /stp.json
endpoint - the elected root bridge (priority + MAC), our path cost,
whether we are the root, and the per-port STP state read live from the
ASIC's MSTP register (same 2-bit encoding stp_setup() writes). Ports are
reported by their physical numbers.
Recovered-from: 3132319, 9365c86
Two hygiene fixes for the web UI, prompted by a login that appeared to fail
under privacy shields (Brave Shields / NoScript-family extensions):
- httpd: replace the partial "style-src 'self' 'unsafe-inline'" CSP with a
complete, first-party policy (default-src 'self'; script-src 'self'
'unsafe-inline'; style-src 'self' 'unsafe-inline'; img-src 'self' data:;
connect-src 'self'; form-action 'self'). Everything the UI needs is
same-origin; the explicit policy stops shields injecting their own
restrictive report-only probes (the noisy script-src-elem 'none' console
spam) and passes a strict-CSP audit. Verified: no CSP violations in-browser.
- login.html: add autocomplete="current-password" so password managers
recognise the field (they showed "unknown password" without it).
NOTE: these do NOT bypass a browser's LAN-device protection (NoScript "lan" /
Brave Shields), which strips the POST body of requests to a LAN address and is
why the login can fail in-browser while the same credentials work over curl.
That is a deliberate browser security feature; the user must allow the site in
their shields to log in. The backend password (default 1234) is unchanged and
correct.
Add a language selector to the System Settings page supporting English,
Japanese, and Chinese. The i18n dictionary already contained zh
translations but lacked the UI to switch languages.
Changes:
- html/i18n.js: add sys_language key to en/ja/zh
- html/system.html: add language selector dropdown
- html/system.js: add changeLang() and selector initialization
- tools/httpd_sim.c: fix cookie parsing for multi-cookie headers;
fix Set-Cookie response (missing Path=/, missing \r\n\r\n);
add SO_REUSEADDR for faster port reuse
A bare "vlan N" line never matches conf_cmds (a VLAN entry always carries
a member port or the mgmt/name keyword), so it is never stored. The
c !== "vlan N" comparison could therefore never match an entry; the
startsWith(prefix) check alone covers all stored vlan lines.
The firmware has a single management VLAN (one management_vlan variable),
so a stored config should never carry more than one "vlan N mgmt" line.
When parsing a mgmt command, drop any previously stored mgmt entry so
repeatedly changing the management VLAN no longer accumulates stale lines.
parseConf() only recognized "port N name" in conf_cmds, so "port N <speed>"
lines were treated as unknown commands and dropped on save - configured port
speeds never persisted.
Add the speed command (incl. optional half/full duplex suffix) to conf_cmds,
add a "port N" entry to conf_overwrite so re-saving a speed replaces the
previous value, and guard the per-port name entry so changing a port's speed
no longer wipes its configured name.
The vlan/mgmt lookahead split (commit dcc60c3) correctly separated the
patterns so that `vlan N mgmt` and `vlan N <members>` match different
conf_overwrite entries. But the filter that applies the overwrite still
used `item.startsWith(matchStr + " ")` with matchStr = "vlan N", which
also matches "vlan N mgmt" — so the management entry was removed whenever
the VLAN's membership definition was re-saved.
Example: with both `vlan 44 management 2t 4t 5 6t` and `vlan 44 mgmt`
in the config, changing the membership dropped `vlan 44 mgmt` entirely.
Added a `!item.endsWith(" mgmt")` guard so management entries survive the
filter. No reordering is needed: `vlan N mgmt` only sets the global
management_vlan variable (cmd_parser.c) and does not depend on the VLAN
table entry existing, so its position in the config is irrelevant.
The pattern /^vlan\s+\d{1,4}\b/ matched both VLAN membership entries
(vlan N <ports>) and management entries (vlan N mgmt), causing them
to overwrite each other in configuration[]. When saving, whichever
form came last in the cmd_log would dedupe the other out of the
final config — resulting in either lost membership or a stale mgmt
setting.
Split into two patterns using negative lookahead:
- /^vlan\s+\d{1,4}\s+mgmt$/ matches only mgmt entries
- /^vlan\s+\d{1,4}(?!\s+mgmt\b)/ matches everything else
Both dedup independently. Discovered via hardware test on 6XH-X
where 'vlan 44 mgmt' silently dropped 'vlan 44 management 2t 5u'
from saved config, locking out web UI on next boot.
Multiple related bugs in the Save-to-Flash path:
1. Multipart upload was missing the required filename argument,
causing the backend parser to fail. Added 'config.txt' to the
form.append call. This was likely the primary reason Save-to-Flash
was unreliable.
2. Web UI was not pausing its polling interval during flash write,
causing CPU contention and intermittent crashes. Added isSaving
lock and clearInterval before sendConfig.
3. conf_cmds whitelist was incomplete. Added: syslog, passwd, pvid,
ingress, port name, lag, laghash, isolate, stp, igmp, mtu, bw,
vlan N mgmt, vlan N d.
4. VLAN regex blocked named VLANs. New pattern allows optional name
(starts with letter, matching CLI parser semantics).
5. vlan N d (delete) was not persisted. parseConf now removes the
matching vlan N ... entry from configuration[] when seeing a
delete command, without storing the delete itself. Result: the
saved config describes the end state.
6. configuration[] was not cleared between flashSave invocations,
leading to stale entries from prior interactions.
7. conf_overwrite boundary fix: 'pvid 1' no longer matches 'pvid 10'
etc. Added trailing space in startsWith check.
8. All conf_cmds patterns now anchored with ^...$ for full-line
match. parseConf normalizes whitespace before testing.
9. Port range widened to \d{1,2} so ports 11+ are accepted.
Structural fixes (1, 2, 6, 7, 8) ported from mcaptur's closed PR #219;
remaining fixes (3, 4, 5, 9) and overall regex strategy are new.
The hardware bitmask format encodes both "untagged members" and
"non-members" in the upper 10 bits (per doc/vlan.md). The existing
fetchVLAN() correctly masks this with the membership bitmask before
display, but loadVlanTable() did not, causing non-member ports to
appear in the Untagged Ports column.
Tested on KeepLiNK KP-9000-6XH-X.
Below the configuration form, a new table lists all configured VLANs
with their port memberships: Member, Tagged, Untagged, and PVID
columns. Port ranges are formatted compactly (e.g. "1-2,5").
Port-to-bit mapping uses the existing physToLogPort[] array populated
by /status.json, so the table is consistent with the icon view in
fetchVLAN() and works on any supported machine.
Each row has a delete button that sends "vlan N d" via /cmd, with
a confirmation dialog. VLAN 1 cannot be deleted (no button shown).
Table and dropdown both refresh automatically after Update/Create
and after delete via a new refreshVlanViews() helper.
The N+1 request pattern (1x /vlanlist + Nx /vlan.json) keeps the
backend simple and avoids buffer overflow risks for systems with
many VLANs. For typical configurations (<50 VLANs), page load
remains under one second.
A new <select> element above the VLAN ID input lets users pick an
existing VLAN by name instead of typing the ID. Options are loaded
from /vlanlist on page load and re-loaded after successful
Update/Create operations.
Selection triggers the existing fetchVLAN() flow.
Falls back gracefully if /vlanlist returns an empty list or fails:
the dropdown is hidden and the existing ID input remains functional.
When device has no LOS pin and module has no extended status, the
RX LOS value will not be shown. When both are present, the equality
check is performed. When only one is present, the present value
will be shown.
json from the switch, will still contain the field, but can be
null for when pin is not available.
The state of the SFP module is being already send in status.json,
but was not parsed via Web UI. When debuging SFP LOS/Signal detection
it is usefull to see what module is reporting back.
Extended the SFP mouse-over information with:
- RX LOS as reported in 0x02 bit of register 238
- External TX Disabled from 0x80 bit of register 238
- TX fault from 0x04 of the same register
- The last state of RX LOS pin (available also when there is no 0x40
option on the module)
This should help identify missing/incorrect setup of TX disabled pin.
I noticed that entering some vlan config, clicking "Update/Create", then
clicking "Get Configuration" resulted in Tagged/Untagged/PVID displaying
completely wrong data. It looks like the parisng in fetchVLAN() in
vlan.js was just completely disconnected from the layout of the
registers read by vlan_get() in rtl837x_port.c. Who knows how that
happened.
1. Fix fetchVLAN() member/untag parsing: the old code read bits [9:0]
as untagged and bits [10:19] as tagged, but the VLAN table register
layout is members in [9:0] and untag in [19:10]. Now correctly
derives tagged (member && !untag) and untagged (member && untag).
2. Add PVID to vlan.json response: PVID is stored per-port in separate
PVID registers (RTL837x_PVID_BASE_REG), not in the VLAN table entry.
The old code nonsensically tried to read it from bits [20:29] of the
VLAN table. Add port_pvid_get() and build a pvid bitmask in
send_vlan() so the JS can parse it correctly.
3. Remove auto-PVID logic from setC(): setC() is called by fetchVLAN()
while it's loading existing config, and so this logic mangled the
display of the existing config. Also, it doesn't make sense to
auto-set the PVID for tagged members (PVID relates to untagged
ingress.)
d00f
Your Name
logicog
Lynn-Becky
Erdnusschokolade
UAb5eSMn
René van Dorst
logicog
feelfree69
feelfree69
Jan Hoffmann
diijkstra
Mark H. Spatz
chriz
TylerDurden-23