When the STA is brought up, it is set to DISABLED before adding the bss to ucode,
so the first trigger to disable the AP is missed.
Reported-by: Michael-cy Lee (李峻宇) <Michael-cy.Lee@mediatek.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Commit e978072baaca ("Do prune_association only after the STA is
authorized") causes issues when an STA roams from one interface to
another interface on the same PHY. The mt7915 driver is not able to
handle this properly. While the commits fixes a DoS, there are other
devices and drivers with the same limitation, so revert to the orginal
behavior for now, until we have a better solution in place.
Fixes: #13156
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
It cannot be properly cloned, since it is attached to the resource type.
Use a separate registry for data. Fixes object confusion issues
Signed-off-by: Felix Fietkau <nbd@nbd.name>
An active client mode interface could prevent the AP from claiming its channel
and mess up the bringup sequence order
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Drop obsolete control interface patches.
This fixes some corner cases in the previous code where the segment 0 center
frequency was not adjusted properly, leading to logspam and non-working AP
interfaces.
Additionally, shutting down the AP was broken, because the next beacon update
would re-enable it, leading to a race condition on assoc.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Fixes this error: hostapd: nl80211: kernel reports: integer out of range
Reported-by: Hartmut Birr <e9hack@gmail.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Check the phy before removing unrelated netdevs on the same hw device
Reported-by: Hartmut Birr <e9hack@gmail.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Include AP ucode source file
Fixes: e56c5f7b27 ("hostapd: add ucode support, use ucode for the main ubus object")
Signed-off-by: Felix Fietkau <nbd@nbd.name>
This implements vastly improved dynamic configuration reload support.
It can handle configuration changes on individual wifi interfaces, as well
as adding/removing interfaces.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
This can be used to run a standalone EAP server that can be used from
other APs. It uses json as user database format and can automatically
handle reload.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Add an UCI option to enable Multiple BSSID Advertisement. Enabling this
will announce all BSSIDS on a phy in a single beacon frame. The
interface that is brought up first will be the transmitting profile, all
others are non-transmitting profiles and will be advertised in the
Multiple BSSID element in Beacon and Probe Response frames of the first
interface.
This depends on driver and client support. Enabling this will result in
all but the first interface not being visible at all for clients that do
not support it.
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
Based on Paul Fertser <fercerpav@gmail.com>'s guidance:
Change AUTORELEASE in rules.mk to:
```
AUTORELEASE = $(if $(DUMP),0,$(shell sed -i "s/\$$(AUTORELEASE)/$(call commitcount,1)/" $(CURDIR)/Makefile))
```
then update all affected packages by:
```
for i in $(git grep -l PKG_RELEASE:=.*AUTORELEASE | sed 's^.*/\([^/]*\)/Makefile^\1^';);
do
make package/$i/clean
done
```
Signed-off-by: Tianling Shen <cnsztl@immortalwrt.org>
This reduces open coding and allows to easily add a knob to enable
it treewide, where chosen packages can still opt-out via "no-lto".
Some packages used LTO, but not the linker plugin. This unifies 'em
all to attempt to produce better code.
Quoting man gcc(1):
"This improves the quality of optimization by exposing more code to the
link-time optimizer."
Also use -flto=auto instead of -flto=jobserver, as it's not guaranteed
that every buildsystem uses +$(MAKE) correctly.
Signed-off-by: Andre Heider <a.heider@gmail.com>
This reduces open coding and allows to easily add a knob to
enable it treewide, where chosen packages can still opt-out via
"no-gc-sections".
Note: libnl, mbedtls and opkg only used the CFLAGS part without the
LDFLAGS counterpart. That doesn't help at all if the goal is to produce
smaller binaries. I consider that an accident, and this fixes it.
Note: there are also packages using only the LDFLAGS part. I didn't
touch those, as gc might have been disabled via CFLAGS intentionally.
Signed-off-by: Andre Heider <a.heider@gmail.com>
This allows adding backup servers, in case the primary ones fail.
Assume that port and shared secret are going to be the same.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
This patch fixes a corner case when using passwords that are exactly 64
characters in length with mesh mode or passwords longer than 63 characters
with SAE because 'psk' is used instead of 'sae_password'.
SAE is obligatory for 802.11s (mesh point).
The 'psk' option for hostapd is suited for WPA2 and enforces length
restrictions on passwords. Values of 64 characters are treated as PMKs.
With SAE, PMKs are always generated during the handshake and there are no
length restrictions.
The 'sae_password' option is more suited for SAE and should be used
instead.
Before this patch, the 'sae_password' option is only used with mesh mode
passwords that are not 64 characters long.
As a consequence:
- mesh passwords can't be 64 characters in length
- SAE only works with passwords with lengths >8 and <=63 (due to psk
limitation).
Fix this by always using 'sae_password' with SAE/mesh and applying the PMK
differentiation only when PSK is used.
Fixes: #11324
Signed-off-by: Leon M. Busch-George <leon@georgemail.eu>
[ improve commit description ]
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
The ABI of the wolfssl library changed a bit between version 5.5.3 and
5.5.4. This release update will trigger a rebuild of all packages which
are using wolfssl to make sure they are adapted to the new ABI.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The libraries libpthread, libdl, libutil, libanl have been integrated
into the libc library in version 2.34. it is not needed to explicitly
link them any more.
Most of the functions have been moved from the librt.so into libc.so
some time ago already.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
hostapd's compile time option CONFIG_IEEE80211N was removed almost 3 years
ago, 80.211n/HT is always included since then.
Noticed because `hostapd -v11n` confusingly returned an error.
See hostapd's commit:
f3bcd69603 "Remove CONFIG_IEEE80211N build option"
Signed-off-by: Andre Heider <a.heider@gmail.com>
eee80211_frequency_to_channel() isn't used anymore, which is a leftover from:
2a31e9ca97 "hostapd: add op-class to get_status output"
Signed-off-by: Andre Heider <a.heider@gmail.com>
dtim_period is a bss property, not a device one.
It is already handled properly in mac80211.sh
Fixes: 30c64825c7 ("hostapd: add dtim_period, local_pwr_constraint, spectrum_mgmt_required")
Signed-off-by: Felix Fietkau <nbd@nbd.name>
