<feed xmlns='http://www.w3.org/2005/Atom'>
<title>staging/blocktrron/target, branch master</title>
<subtitle>Staging tree of David Bauer</subtitle>
<id>https://git-03.infra.openwrt.org/openwrt/staging/blocktrron/atom?h=master</id>
<link rel='self' href='https://git-03.infra.openwrt.org/openwrt/staging/blocktrron/atom?h=master'/>
<link rel='alternate' type='text/html' href='https://git-03.infra.openwrt.org/openwrt/staging/blocktrron/'/>
<updated>2026-11-16T19:56:19Z</updated>
<entry>
<title>ramips: fix switch setup for AXIMCom MR-102N</title>
<updated>2026-11-16T19:56:19Z</updated>
<author>
<name>Christoph Krapp</name>
</author>
<published>2026-11-10T23:59:19Z</published>
<link rel='alternate' type='text/html' href='https://git-03.infra.openwrt.org/openwrt/staging/blocktrron/commit/?id=396f223d3df0a443394b66201de05ec4cec6900c'/>
<id>urn:sha1:396f223d3df0a443394b66201de05ec4cec6900c</id>
<content type='text'>
Commit a43cbfe2e3a1 ("ramips: remove default switch setup in 02_network")
added aximcom,mr-102n to two cases in ramips_setup_interfaces: the generic
"0:lan" "1:lan" "2:lan" "3:lan" "4:wan" "6@eth0" switch setup, and a
dedicated "4:lan" "6@eth0" one. As the first matching case wins, the
dedicated setup was never used, and port 4 is set up as WAN.

The dedicated setup is the intended one: before that commit, the board
used eth0.2 as its only LAN interface, and the DTS sets the switch
portmap to 0x2f, which leaves port 4 as the only port outside of the
default LAN group, so eth0.2 maps to port 4. With the generic setup,
this LAN port ends up as WAN.

Remove the board from the generic case so the dedicated one is used.

Fixes: a43cbfe2e3a1 ("ramips: remove default switch setup in 02_network")
Signed-off-by: Christoph Krapp &lt;achterin@gmail.com&gt;
Co-Authored-By: Claude Opus 5.5 &lt;noreply@anthropic.com&gt;
Link: https://github.com/openwrt/openwrt/pull/26609
Signed-off-by: Robert Marko &lt;robimarko@gmail.com&gt;
</content>
</entry>
<entry>
<title>bcm53xx: R6400: add the USB LED trigger sources</title>
<updated>2026-11-16T19:48:59Z</updated>
<author>
<name>Erik Conijn</name>
</author>
<published>2026-11-11T17:03:44Z</published>
<link rel='alternate' type='text/html' href='https://git-03.infra.openwrt.org/openwrt/staging/blocktrron/commit/?id=b49e1008af640410abca65c7245919b098357d14'/>
<id>urn:sha1:b49e1008af640410abca65c7245919b098357d14</id>
<content type='text'>
The R6400's USB LEDs have no trigger-sources in the device tree, so the
kernel's usbport trigger logs at boot:

  leds white:usb-1: Failed to get trigger sources for /leds/led-usb1

Add a kernel patch that names the ports as the other BCM5301X boards do
(USB 1: OHCI, EHCI and XHCI port 1; USB 2: OHCI and EHCI port 2) and
makes usbport the default trigger. The R6400 v2 includes the v1 DTS, so
this covers both versions. Refresh the switch LED patch that follows it.

Tested on the R6400 v1 and v2: the warning is gone, and a USB stick
lights USB 1 in the front port (USB 3.0) and USB 2 in the back port.

Fixes: 21d9bd899b ("bcm53xx: add support for Netgear R6400 v1")
Assisted-by: Claude:claude-opus-5-5
Signed-off-by: Erik Conijn &lt;egc112@msn.com&gt;
Link: https://github.com/openwrt/openwrt/pull/26631
Signed-off-by: Robert Marko &lt;robimarko@gmail.com&gt;
</content>
</entry>
<entry>
<title>bmips: bcm6318: fix syntax error in 09_fix_crc</title>
<updated>2026-11-16T19:47:05Z</updated>
<author>
<name>Christoph Krapp</name>
</author>
<published>2026-11-11T20:57:31Z</published>
<link rel='alternate' type='text/html' href='https://git-03.infra.openwrt.org/openwrt/staging/blocktrron/commit/?id=780113c89b591092463515ae134d05814610e54a'/>
<id>urn:sha1:780113c89b591092463515ae134d05814610e54a</id>
<content type='text'>
Commit 1d976f83e40c ("bmips: add support for TP-Link TD-W8968 V3")
added the new board as a second label to the existing case, but left
the closing parenthesis of the first label in place:

  comtrend,ar-5315u) |\
  tp-link,td-w8968-v3)

This is a syntax error, so the script can not be parsed at all. As
uci-defaults scripts are only removed once they succeed, it is retried
on every boot, and "mtd fixtrx firmware" never runs, neither for the
TP-Link TD-W8968 V3 nor for the Comtrend AR-5315u, where it worked
before.

Drop the stray parenthesis.

Fixes: 1d976f83e40c ("bmips: add support for TP-Link TD-W8968 V3")
Signed-off-by: Christoph Krapp &lt;achterin@gmail.com&gt;
Co-Authored-By: Claude Opus 5.5 &lt;noreply@anthropic.com&gt;
Link: https://github.com/openwrt/openwrt/pull/26641
Signed-off-by: Robert Marko &lt;robimarko@gmail.com&gt;
</content>
</entry>
<entry>
<title>rockchip: add upstream patch for FriendlyELEC NanoPi R76S ADC button</title>
<updated>2026-11-16T19:44:59Z</updated>
<author>
<name>Ryan Leung</name>
</author>
<published>2026-11-13T07:11:04Z</published>
<link rel='alternate' type='text/html' href='https://git-03.infra.openwrt.org/openwrt/staging/blocktrron/commit/?id=986dd058ffbe83d5fee79413fa44ebd3c8c66cb8'/>
<id>urn:sha1:986dd058ffbe83d5fee79413fa44ebd3c8c66cb8</id>
<content type='text'>
Add upstream patch for FriendlyELEC NanoPi R76S ADC button that has been merged to the maintainer's
staging tree and is scheduled for Linux kernel v7.5

Add the kmod-button-hotplug and kmod-input-adc-keys to the R76S default packages accordingly.

Signed-off-by: Ryan Leung &lt;untilscour@protonmail.com&gt;
Link: https://github.com/openwrt/openwrt/pull/26671
Signed-off-by: Robert Marko &lt;robimarko@gmail.com&gt;
</content>
</entry>
<entry>
<title>rockchip: use upstream patch for FriendlyELEC NanoPi M5 ADC buttons</title>
<updated>2026-11-16T19:44:59Z</updated>
<author>
<name>Ryan Leung</name>
</author>
<published>2026-11-13T02:27:16Z</published>
<link rel='alternate' type='text/html' href='https://git-03.infra.openwrt.org/openwrt/staging/blocktrron/commit/?id=f57464f55f84843c90b7b76b9ea14fbd57bd04f3'/>
<id>urn:sha1:f57464f55f84843c90b7b76b9ea14fbd57bd04f3</id>
<content type='text'>
Replace downstream patch for FriendlyELEC NanoPi M5 ADC buttons with upstream patch that has been
merged to the maintainer's staging tree and is scheduled for Linux kernel v7.5

The input event codes have changed which is not an issue because the device hasn't been backported
to a stable release branch.

Signed-off-by: Ryan Leung &lt;untilscour@protonmail.com&gt;
Link: https://github.com/openwrt/openwrt/pull/26671
Signed-off-by: Robert Marko &lt;robimarko@gmail.com&gt;
</content>
</entry>
<entry>
<title>mpc85xx: fix USB probe on boards without vbus-supply</title>
<updated>2026-11-16T19:42:25Z</updated>
<author>
<name>Lefteris Koutrakis</name>
</author>
<published>2026-11-11T18:37:24Z</published>
<link rel='alternate' type='text/html' href='https://git-03.infra.openwrt.org/openwrt/staging/blocktrron/commit/?id=517319b413a15daf1661bd1bb8dfff343b20eb4e'/>
<id>urn:sha1:517319b413a15daf1661bd1bb8dfff343b20eb4e</id>
<content type='text'>
devm_regulator_get_enable_optional() returns -ENODEV when the device
tree does not describe a vbus-supply. Treating that as a fatal error
makes fsl-mph-dr-of fail to probe on every mpc85xx board without a
vbus regulator (all of them except the TP-Link TL-WDR4900 v1), leaving
USB unusable:

  fsl-usb2-mph-dr ffe22000.usb: error -ENODEV: failed to get vbus regulator

Use devm_regulator_get_enable() instead. Without a vbus-supply it
falls back to a dummy regulator, so the probe succeeds, while boards
that describe one keep controlling it as before.

Tested on a WatchGuard Firebox T30.

Fixes: 9e43544bc6ec ("mpc85xx: use regulator for USB power")
Suggested-by: Rosen Penev &lt;rosenp@gmail.com&gt;
Signed-off-by: Lefteris Koutrakis &lt;lefteris.koutrakis@gmail.com&gt;
Link: https://github.com/openwrt/openwrt/pull/26667
Signed-off-by: Robert Marko &lt;robimarko@gmail.com&gt;
</content>
</entry>
<entry>
<title>realtek: dsa: rtl960x: add port_max_mtu and port_change_mtu</title>
<updated>2026-11-16T19:19:31Z</updated>
<author>
<name>Rustam Adilov</name>
</author>
<published>2026-11-01T18:12:15Z</published>
<link rel='alternate' type='text/html' href='https://git-03.infra.openwrt.org/openwrt/staging/blocktrron/commit/?id=3b45236abc2b3ba289a750239a432faabb2d67fc'/>
<id>urn:sha1:3b45236abc2b3ba289a750239a432faabb2d67fc</id>
<content type='text'>
RTL9607C switch is capable of changing max packet frame sizes
per port with RTL960X_ACCEPT_MAX_LEN register. Add support for
port_change_mtu and port_max_mtu following similar logic to
RTL8365MB except allow for any port, not just cpu port.

Since the CPU NIC does not currently support jumbo frames, the
actual limit is around ~1576 and DSA would throw an error if
some would try going higher.

Signed-off-by: Rustam Adilov &lt;adilov@tutamail.com&gt;
Link: https://github.com/openwrt/openwrt/pull/26592
Signed-off-by: Markus Stockhausen &lt;markus.stockhausen@gmx.de&gt;
</content>
</entry>
<entry>
<title>generic: backport SDIO UHS fixes for cards misreporting S18A</title>
<updated>2026-11-16T17:19:23Z</updated>
<author>
<name>David Härdeman</name>
</author>
<published>2026-11-16T14:13:46Z</published>
<link rel='alternate' type='text/html' href='https://git-03.infra.openwrt.org/openwrt/staging/blocktrron/commit/?id=a637a644fb4d9c7b9aef12c8fb8ed8d911a69f66'/>
<id>urn:sha1:a637a644fb4d9c7b9aef12c8fb8ed8d911a69f66</id>
<content type='text'>
Backport two mmc core changes from mmc/next, on their way to v7.5:

- "mmc: core: sdio: skip CMD11 if signalling is already 1.8V"
  (67a97316a31a): the RTL8822CS answers CMD5 with S18A = 1 even when
  its bus is hardwired to 1.8V. The core then sends CMD11, which times
  out, and the card ends up in 50 MHz high-speed mode instead of
  SDR104. Skip CMD11 when the host already signals at 1.8V.

- "mmc: core: sdio: retry without UHS if UHS init fails"
  (a3838c21e28e): if UHS init fails, e.g. because tuning finds no
  working phase, fall back to high-speed mode instead of leaving the
  card unusable. The first patch depends on this fallback for
  non-compliant cards that do need CMD11.

Only hosts that start at a fixed 1.8V vqmmc are affected; hosts that
start at 3.3V and cards that report S18A correctly are not.

Tested on a FriendlyElec NanoPi R76S (RTL8822CS, fixed 1.8V vqmmc):
the "Unexpected CMD11 timeout" at boot is gone and the WiFi card runs
in SDR104 at 198 MHz instead of high speed at 50 MHz. TCP throughput
as an AP goes from ~37 to ~60 Mbit/s (AP transmitting) and from ~78
to ~150 Mbit/s (AP receiving) with the in-tree rtw88 driver.
Checked on first boot, 5 warm reboots, a host unbind/bind and
repeated wifi down/up: SDR104 every time, no mmc or rtw88 errors.

Upstream:
https://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc.git/commit/?h=next&amp;id=67a97316a31a
https://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc.git/commit/?h=next&amp;id=a3838c21e28e

Link: https://github.com/openwrt/openwrt/issues/21361
Signed-off-by: David Härdeman &lt;david@hardeman.nu&gt;
Link: https://github.com/openwrt/openwrt/pull/26760
Signed-off-by: Robert Marko &lt;robimarko@gmail.com&gt;
</content>
</entry>
<entry>
<title>realtek: dsa: rtl960x: add more bridge port flags</title>
<updated>2026-11-16T15:44:20Z</updated>
<author>
<name>Rustam Adilov</name>
</author>
<published>2026-11-02T16:14:04Z</published>
<link rel='alternate' type='text/html' href='https://git-03.infra.openwrt.org/openwrt/staging/blocktrron/commit/?id=80ef55b34d7cc11b0d6d482ca66decb3bff93612'/>
<id>urn:sha1:80ef55b34d7cc11b0d6d482ca66decb3bff93612</id>
<content type='text'>
Add support for multicast, broadcast and unicast bridge flooding
in same way as rtl8365mb does it. Everything apart from register
address is pretty much identical. However, no BR_MCAST_FLOOD due
to lack of MDB support yet.

Signed-off-by: Rustam Adilov &lt;adilov@tutamail.com&gt;
Link: https://github.com/openwrt/openwrt/pull/26594
Signed-off-by: Markus Stockhausen &lt;markus.stockhausen@gmx.de&gt;
</content>
</entry>
<entry>
<title>realtek: rtl9607c: use new pinmux</title>
<updated>2026-11-16T12:58:00Z</updated>
<author>
<name>Michael Zavertkin</name>
</author>
<published>2026-11-02T13:58:58Z</published>
<link rel='alternate' type='text/html' href='https://git-03.infra.openwrt.org/openwrt/staging/blocktrron/commit/?id=eb0f561cc4064116e073bfb56345725e039244cd'/>
<id>urn:sha1:eb0f561cc4064116e073bfb56345725e039244cd</id>
<content type='text'>
RTL9607C benefits from specific pinmux using:
- ugly pinctrl-single nodes for each GPIO could be removed
- I2C no longer pinmuxes itself
- it opens road for LED driver implementation

Signed-off-by: Michael Zavertkin &lt;misha.zavertkin@mail.ru&gt;
Link: https://github.com/openwrt/openwrt/pull/23961
Signed-off-by: Markus Stockhausen &lt;markus.stockhausen@gmx.de&gt;
</content>
</entry>
</feed>
