On u-boot v2026.04, we need to add PKG_CONFIG_SYSROOT_DIR flag to
fix in-tree dtc tools build without yaml[1-2]. Fixes:
/usr/bin/ld: yamltree.c:(.text+0xa7c): undefined reference to `yaml_sequence_start_event_initialize'
/usr/bin/ld: yamltree.c:(.text+0xa87): undefined reference to `yaml_emitter_emit'
/usr/bin/ld: yamltree.c:(.text+0xaa4): undefined reference to `yaml_sequence_end_event_initialize'
/usr/bin/ld: yamltree.c:(.text+0xaaf): undefined reference to `yaml_emitter_emit'
/usr/bin/ld: yamltree.c:(.text+0xac1): undefined reference to `yaml_document_end_event_initialize'
/usr/bin/ld: yamltree.c:(.text+0xacc): undefined reference to `yaml_emitter_emit'
/usr/bin/ld: yamltree.c:(.text+0xadc): undefined reference to `yaml_stream_end_event_initialize'
/usr/bin/ld: yamltree.c:(.text+0xae7): undefined reference to `yaml_emitter_emit'
/usr/bin/ld: yamltree.c:(.text+0xaf3): undefined reference to `yaml_emitter_delete'
[1] https://github.com/u-boot/u-boot/commit/0535e46d55d7423094d8b553d704a0f7f678de77
[2] https://github.com/u-boot/u-boot/commit/807bcd844afeadbfe47ac132524cc7e3b699bc46
Signed-off-by: Shiji Yang <yangshiji66@outlook.com>
The simpleImage contains a payload already compressed with lzma-based
xz (by default), so further compressing it with lzma will often make
the result larger. On the contrary, compressing these simpleImages
with gzip can make the result smaller, so replace lzma with
libdeflate-gzip to compress kernel for ws-ap3715i.
Signed-off-by: Edward Chow <equu@openmail.cc>
Link: https://github.com/openwrt/openwrt/pull/23121
Signed-off-by: Robert Marko <robimarko@gmail.com>
find_mtd_part() outputs /dev/mtdblockX, to which fw_setenv cannot
write, "/dev/mtd$(find_mtd_index '<vol name>')" could be used instead.
The envsize should also be changed to 0x1000 to make the CRC checksum
valid and the env block recognized by the uboot-envtools, but the
flash sector size remains 0x10000, otherwise the env block will be
readable but not writable.
The "read-only" mark within device tree is also removed.
Signed-off-by: Edward Chow <equu@openmail.cc>
Link: https://github.com/openwrt/openwrt/pull/23121
Signed-off-by: Robert Marko <robimarko@gmail.com>
Add the basic bits to allow 100base-FX SFP mode on the RTL8214FC.
While this looks good fom ethtool perspective, it does not really
change the phy registers to enforce the mode. The SFP is still
driven in 1000base-X.
While it might seem useless at the moment this at least opens
up a new phy control method. This comes handy with one known bug.
In rare cases a SFP that is plugged in during boot does not get
a link. One option to revive the dead port seems to be
root@OpenWrt:~# ethtool -s lan28 speed 100 duplex full autoneg off
rtl83xx-switch 1b000000.switchcore:ethernet-switch lan28: Link is Up - 100Mbps/Full - flow control off
switch: port 28(lan28) entered blocking state
switch: port 28(lan28) entered forwarding state
rtl83xx_fib_event: FIB_RULE ADD/DEL for IPv6 not supported
rtl83xx_fib_event: FIB_RULE ADD/DEL for IPv6 not supported
Signed-off-by: Markus Stockhausen <markus.stockhausen@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/23087
Signed-off-by: Robert Marko <robimarko@gmail.com>
The RTL839x allows to add an extended page operator during phy
access. This is not needed for the standard linux kernel C22
access. Give the hardcoded 0x1ff value a meaningful define.
Although it is not needed, add the corresponding register define.
This makes clear where the mask belongs to.
Signed-off-by: Markus Stockhausen <markus.stockhausen@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/23204
Signed-off-by: Robert Marko <robimarko@gmail.com>
The rtmdio_931x_setup_ctrl() function currently initializes the c22/c45
and the proprietary format bit of the controller. This works because of
the order these calls are arranged. Narrow down the update to the really
needed bits.
- c22/c45 (bit 1) is updated here
- standard/proprietary (bit 0) is updated in rtmdio_931x_setup_polling()
Adapt the confusing comment "Std. C45, non-standard is 0x3" it basically
explains the other function.
Signed-off-by: Markus Stockhausen <markus.stockhausen@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/23204
Signed-off-by: Robert Marko <robimarko@gmail.com>
Provide a separate function to setup the ability (SDS/MDIO) of a RTL930x
port. This simplifies rtmdio_930x_setup_polling(). With this commit the
driver does no longer unconditionally overwrite reserved register bits.
Add a return value for the new function to indicate failure/success. As
of now this will be silently ignored in the caller. A future commit will
take care about that.
Signed-off-by: Markus Stockhausen <markus.stockhausen@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/23204
Signed-off-by: Robert Marko <robimarko@gmail.com>
Provide a separate function to setup the ability (SDS/MDIO) of a RTL931x
port. This simplifies rtmdio_931x_setup_polling(). With this commit the
driver does no longer unconditionally overwrite reserved register bits.
Add a return value for the new function to indicate failure/success. As
of now this will be silently ignored in the caller. A future commit will
take care about that.
Signed-off-by: Markus Stockhausen <markus.stockhausen@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/23204
Signed-off-by: Robert Marko <robimarko@gmail.com>
Topology setup is used during probing. Improve error checking
to be 100% sure that hardware setup works as expected. While
we are here:
- use GENMASK()
- Be type consistent and add u32 cast for value calculation
- use __ffs(mask) for code deduplication and better readability
Signed-off-by: Markus Stockhausen <markus.stockhausen@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/23204
Signed-off-by: Robert Marko <robimarko@gmail.com>
With its current usage type RTMDIO_PHY_POLL_MMD() definition is ok.
But for the sake of consistency add brackets around the macro
parameters and use masks to avoid calculation inconsistencies.
Signed-off-by: Markus Stockhausen <markus.stockhausen@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/23204
Signed-off-by: Robert Marko <robimarko@gmail.com>
Try to describe the hardware capabilities with consistent defines
and configuration variables. As raw_page is always num_pages - 1
better use num_pages naming where needed and provide a macro that
converts this naming.
While we are here:
- Sort the configuration variables alphabetically
- Provide num_pages defines per architecture
- adapt RTMDIO_839X_C22_DATA() macro to use the new define
Signed-off-by: Markus Stockhausen <markus.stockhausen@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/23204
Signed-off-by: Robert Marko <robimarko@gmail.com>
In AP+STA mode, wifi-scripts emits start_disabled=1 in the per-BSS
section of the generated hostapd config so that hostapd defers
beaconing on every BSS until apsta_state up clears the flag for the
whole iface (uc_hostapd_iface_start clears start_disabled on every BSS
and calls ieee802_11_set_beacon).
When a new BSS is added later via iface.add_bss while the iface is
already in HAPD_IFACE_ENABLED state, the freshly parsed config still
carries start_disabled=1 for that BSS. hostapd_setup_bss is invoked
with start_beacon=true, but hostapd_start_beacon then skips
ieee802_11_set_beacon because conf->start_disabled is set. The kernel
netdev is created without ever starting beacons, the carrier never
comes up, and probe-response transmission attempts fail with
"handle_probe_req: send failed".
Mirror what iface.start does: when the iface is already enabled, the
apsta channel selection has happened, so clear start_disabled for the
incoming BSS before starting it.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Surface SAE confirm mismatches (wrong password) through the same
key-mismatch ubus notification that is already used for PSK failures, so
consumers can react uniformly regardless of the authentication method.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
mld_set_config() treated any call with empty prev_mld as a fresh
configuration and triggered a full Reload all interfaces, even when the
new config was also empty (the typical path on non-MLD devices).
Reloading every BSS on each netifd reconf disrupted associated stations
including PMF-protected backhaul STAs, which would self-deauth after the
SA Query timeout.
Only treat the call as a new configuration when the new config is
actually non-empty.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
This contains many fixes from upstream Linux.
The code block from this patch was moved a bit in the function:
subsys/110-mac80211_keep_keys_on_stop_ap.patch
This patch was applied upstream:
subsys/330-mac80211-fix-crash-in-ieee80211_chan_bw_change-for-A.patch
Link: https://github.com/openwrt/openwrt/pull/23167
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Harmonize the read/write functions for better readability. This
aligns with the naming convention of the upstream rtl9300 mdio
driver.
read_phy -> read_c22
read_phy_mmd -> read_c45
write_phy -> write_c22
write_phy_mmd -> write_c45
Swap order of c22/c45 functions to keep structures sorted
alphabetically.
Signed-off-by: Markus Stockhausen <markus.stockhausen@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/23197
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Add support for changing the LED mode via the device tree.
Currently it always defaults to SERIAL mode. With this change,
one can also use the SINGLE_COLOR_SCAN and BI_COLOR_SCAN modes.
Signed-off-by: Manuel Stocker <mensi@mensi.ch>
Link: https://github.com/openwrt/openwrt/pull/23160
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
