Bring back 802.11s mesh features to the level previously available
before the recent hostapd version bump. This is mostly to support use
of 802.11s on DFS channels, but also making mesh forwarding
configurable which is crucial for use of 802.11s MAC with other routing
protocols, such as batman-adv, on top.
While at it, fix new compiler warning by adapting 700-wifi-reload.patch
to upstream changes, now building without any warnings again.
Fixes: 0a3ec87a66 ("hostapd: update to latest Git hostap_2_9-1238-gdd2daf0848ed")
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
This version has the various slew of bug fixes and compat fixes and
such, but the most interesting thing from an OpenWRT perspective is that
WireGuard now plays nicely with cake and fq_codel. I'll be very
interested to hear from OpenWRT users whether this makes a measurable
difference. Usual set of full changes follows.
This release aligns with the changes I sent to DaveM for 5.7-rc7 and were
pushed to net.git about 45 minutes ago.
* qemu: use newer iproute2 for gcc-10
* qemu: add -fcommon for compiling ping with gcc-10
These enable the test suite to compile with gcc-10.
* noise: read preshared key while taking lock
Matt noticed a benign data race when porting the Linux code to OpenBSD.
* queueing: preserve flow hash across packet scrubbing
* noise: separate receive counter from send counter
WireGuard now works with fq_codel, cake, and other qdiscs that make use of
skb->hash. This should significantly improve latency spikes related to
buffer bloat. Here's a before and after graph from some data Toke measured:
https://data.zx2c4.com/removal-of-buffer-bloat-in-wireguard.png
* compat: support RHEL 8 as 8.2, drop 8.1 support
* compat: support CentOS 8 explicitly
* compat: RHEL7 backported the skb hash renamings
The usual RHEL churn.
* compat: backport renamed/missing skb hash members
The new support for fq_codel and friends meant more backporting work.
* compat: ip6_dst_lookup_flow was backported to 4.14, 4.9, and 4.4
The main motivation for releasing this now: three stable kernels were released
at the same time, with a patch that necessitated updating in our compat layer.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Upstream in commit 200c7693c9a1 ("Make WEP functionality an optional
build parameter") has made WEP functionality an optional build parameter
disabled as default, because WEP should not be used for anything
anymore. As a step towards removing it completely, they moved all WEP
related functionality behind CONFIG_WEP blocks and disabled it by
default.
This functionality is subject to be completely removed in a future
release.
So follow this good security advice, deprecation notice and disable WEP
by default, but still allow custom builds with WEP support via
CONFIG_WPA_ENABLE_WEP config option till upstream removes support for
WEP completely.
Signed-off-by: Petr Štetiar <ynezz@true.cz>
Bump package to latest upstream Git HEAD which is commit dd2daf0848ed
("HE: Process HE 6 GHz band capab from associating HE STA"). Since last
update there was 1238 commits done in the upstream tree with 618 files
changed, 53399 insertions, 24928 deletions.
I didn't bothered to rebase mesh patches as the changes seems not
trivial and I don't have enough knowledge of those parts to do/test that
properly, so someone else has to forward port them, ideally upstream
them so we don't need to bother anymore. I've just deleted them for now:
004-mesh-use-setup-completion-callback-to-complete-mesh-.patch
005-mesh-update-ssid-frequency-as-pri-sec-channel-switch.patch
006-mesh-inform-kernel-driver-DFS-handler-in-userspace.patch
007-mesh-apply-channel-attributes-before-running-Mesh.patch
011-mesh-Allow-DFS-channels-to-be-selected-if-dfs-is-ena.patch
013-mesh-do-not-allow-pri-sec-channel-switch.patch
015-mesh-do-not-use-offchan-mgmt-tx-on-DFS.patch
016-mesh-fix-channel-switch-error-during-CAC.patch
018-mesh-make-forwarding-configurable.patch
Refreshed all other patches, removed upstreamed patches:
051-wpa_supplicant-fix-race-condition-in-mesh-mpm-new-pe.patch
067-0001-AP-Silently-ignore-management-frame-from-unexpected-.patch
070-driver_nl80211-fix-WMM-queue-mapping-for-regulatory-.patch
071-driver_nl80211-fix-regulatory-limits-for-wmm-cwmin-c.patch
090-wolfssl-fix-crypto_bignum_sum.patch
091-0001-wolfssl-Fix-compiler-warnings-on-size_t-printf-forma.patch
091-0002-wolfssl-Fix-crypto_bignum_rand-implementation.patch
091-0003-wolfssl-Do-not-hardcode-include-directory-in-wpa_sup.patch
800-usleep.patch
Tested-by: Stefan Lippers-Hollmann <s.l-h@gmx.de> [ipq8065/NBG6817; ipq40xx/MAP-AC2200]
Signed-off-by: Petr Štetiar <ynezz@true.cz>
* ipc: add support for openbsd kernel implementation
* ipc: cleanup openbsd support
* wg-quick: add support for openbsd kernel implementation
* wg-quick: cleanup openbsd support
Very exciting! wg(8) and wg-quick(8) now support the kernel implementation for
OpenBSD. OpenBSD is the second kernel, after Linux, to receive full fledged
and supported WireGuard kernel support. We'll probably send our patch set up
to the list during this next week. `ifconfig wg0 create` to make an interface,
and `wg ...` like usual to configure WireGuard aspects of it, like usual.
* wg-quick: support dns search domains
If DNS= has a non-IP in it, it is now treated as a search domain in
resolv.conf. This new feature will be rolling out across our various GUI
clients in the next week or so.
* Makefile: simplify silent cleaning
* ipc: remove extra space
* git: add gitattributes so tarball doesn't have gitignore files
* terminal: specialize color_mode to stdout only
Small cleanups.
* highlighter: insist on 256-bit keys, not 257-bit or 258-bit
The highlighter's key checker is now stricter with base64 validation.
* wg-quick: android: support application whitelist
Android users can now have an application whitelist instead of application
blacklist.
* systemd: add wg-quick.target
This enables all wg-quick at .services to be restarted or managed as a unit via
wg-quick.target.
* Makefile: remember to install all systemd units
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
crypto_bignum_rand() use needless time-consuming filtering
which resulted in SAE no longer connecting within time limits.
Import fixes from hostap upstream to fix that.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Hotplug scripts are sourced so the #!/bin/sh is superfluous/deceptive.
Re-arrange script to only source 'procd' if we get to the stage of
needing to signal the process, reduce hotplug processing load a little.
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
add option to set management IP pattern
also add missing 'unconfigure system hostname'
for example pattern '!192.168.1.1' makes it possible that
WAN IP is selected instead of LAN IP
Signed-off-by: Daniel A. Maierhofer <git@damadmai.at>
[grammar and spelling fixes in commit message]
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
Samba 3.6 is completely unsupported, in addition to having tons of patches
It also causes kernel panics on some platforms when sendfile is enabled.
Example:
https://github.com/gnubee-git/GnuBee_Docs/issues/45
I have reproduced on ramips as well as mvebu in the past.
Samba 4 is an alternative available in the packages repo.
cifsd is a lightweight alternative available in the packages repo. It is
also a faster alternative to both Samba versions (lower CPU usage). It
was renamed to ksmbd.
To summarize, here are the alternatives:
- ksmbd + luci-app-cifsd
- samba4 + luci-app-samba4
Signed-off-by: Rosen Penev <rosenp@gmail.com>
[drop samba36-server from GEMINI_NAS_PACKAGES, ksmbd rename + summary]
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
* compat: timeconst.h is a generated artifact
Before we were trying to check for timeconst.h by looking in the kernel
source directory. This isn't quite correct on configurations in which
the object directory is separate from the kernel source directory, for
example when using O="elsewhere" as a make option when building the
kernel. The correct fix is to use $(CURDIR), which should point to
where we want.
* compat: use bash instead of bc for HZ-->USEC calculation
This should make packaging somewhat easier, as bash is generally already
available (at least for dkms), whereas bc isn't provided by distros by
default in their build meta packages.
* socket: remove errant restriction on looping to self
It's already possible to create two different interfaces and loop
packets between them. This has always been possible with tunnels in the
kernel, and isn't specific to wireguard. Therefore, the networking stack
already needs to deal with that. At the very least, the packet winds up
exceeding the MTU and is discarded at that point. So, since this is
already something that happens, there's no need to forbid the not very
exceptional case of routing a packet back to the same interface; this
loop is no different than others, and we shouldn't special case it, but
rather rely on generic handling of loops in general. This also makes it
easier to do interesting things with wireguard such as onion routing.
At the same time, we add a selftest for this, ensuring that both onion
routing works and infinite routing loops do not crash the kernel. We
also add a test case for wireguard interfaces nesting packets and
sending traffic between each other, as well as the loop in this case
too. We make sure to send some throughput-heavy traffic for this use
case, to stress out any possible recursion issues with the locks around
workqueues.
* send: cond_resched() when processing tx ringbuffers
Users with pathological hardware reported CPU stalls on CONFIG_
PREEMPT_VOLUNTARY=y, because the ringbuffers would stay full, meaning
these workers would never terminate. That turned out not to be okay on
systems without forced preemption. This commit adds a cond_resched() to
the bottom of each loop iteration, so that these workers don't hog the
core. We don't do this on encryption/decryption because the compat
module here uses simd_relax, which already includes a call to schedule
in preempt_enable.
* selftests: initalize ipv6 members to NULL to squelch clang warning
This fixes a worthless warning from clang.
* send/receive: use explicit unlikely branch instead of implicit coalescing
Some code readibility cleanups.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
* compat: support latest suse 15.1 and 15.2
* compat: support RHEL 7.8's faulty siphash backport
* compat: error out if bc is missing
* compat: backport hsiphash_1u32 for tests
We now have improved support for RHEL 7.8, SUSE 15.[12], and Ubuntu 16.04.
* compat: include sch_generic.h header for skb_reset_tc
A fix for a compiler error on kernels with weird configs.
* compat: import latest fixes for ptr_ring
* compat: don't assume READ_ONCE barriers on old kernels
* compat: kvmalloc_array is not required anyway
ptr_ring.h from upstream was imported, with compat modifications, to our
compat layer, to receive the latest fixes.
* compat: prefix icmp[v6]_ndo_send with __compat
Some distros that backported icmp[v6]_ndo_send still try to build the compat
module in some corner case circumstances, resulting in errors. Work around
this with the usual __compat games.
* compat: ip6_dst_lookup_flow was backported to 3.16.83
* compat: ip6_dst_lookup_flow was backported to 4.19.119
Greg and Ben backported the ip6_dst_lookup_flow patches to stable kernels,
causing breaking in our compat module, which these changes fix.
* git: add gitattributes so tarball doesn't have gitignore files
Distros won't need to clean this up manually now.
* crypto: do not export symbols
These don't do anything and only increased file size.
* queueing: cleanup ptr_ring in error path of packet_queue_init
Sultan Alsawaf reported a memory leak on an error path.
* main: mark as in-tree
Now that we're upstream, there's no need to set the taint flag.
* receive: use tunnel helpers for decapsulating ECN markings
ECN markings are now decapsulated using RFC6040 instead of the old RFC3168.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Init script checks for an already active DHCP server on the interface
and if such DHCP server is found, then it logs "refusing to start DHCP"
message, starts dnsmasq without DHCP service unless `option force 1` is
set and caches the DHCP server check result.
Each consecutive service start then uses this cached DHCP server check
result, but doesn't provide log feedback about disabled DHCP service
anymore.
So this patch ensures, that the log message about disabled DHCP service
on particular interface is always provided.
Acked-by: Hans Dedecker <dedeckeh@gmail.com>
Signed-off-by: Petr Štetiar <ynezz@true.cz>
Backport patch from hostapd.git master that fixes copy/paste error in
crypto_bignum_sub() in crypto_wolfssl.c.
This missing fix was discovered while testing SAE over a mesh interface.
With this fix applied and wolfssl >3.14.4 mesh+SAE works fine with
wpad-mesh-wolfssl.
Cc: Sean Parkinson <sean@wolfssl.com>
Signed-off-by: Antonio Quartulli <a@unstable.cc>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
f4d759b dhcp.c: further improve validation
Further improve input validation for CVE-2020-11752
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
cdac046 dns.c: fix input validation fix
Due to a slight foobar typo, failing to de-reference a pointer, previous
fix not quite as complete as it should have been.
Improve CVE-2020-11750 fix
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
When compiled with glibc the config_scan.c wants to use the
cpupolicy2numeric() function which is only available when
HAVE_SCHED_SETSCHEDULER is set. It looks like the wrong define was used here.
This fixes a build problem with glibc in combination with the force
ac_cv_func_sched_setscheduler=no in the OpenWrt CONFIGURE_VARS.
This fixes the following compile error with glibc:
----------------------------------------------------------------------
/bin/ld: config_scan.o: in function `socks_yylex':
dante-1.4.1/sockd/config_scan.l:461: undefined reference to `cpupolicy2numeric'
collect2: error: ld returned 1 exit status
make[5]: *** [Makefile:522: sockd] Error 1
Fixes: aaf46a8fe2 ("dante: disable sched_getscheduler() - not implemented in musl")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
This is primarily a maintenance release with bugfixes and improvements.
This release also fixes a security issue (CVE-2020-11810) which allows
disrupting service of a freshly connected client that has not yet
negotiated session keys. The vulnerability cannot be used to
inject or steal VPN traffic.
Release announcement:
https://openvpn.net/community-downloads/#heading-13812
Full list of changes:
https://community.openvpn.net/openvpn/wiki/ChangesInOpenvpn24#OpenVPN2.4.9
Signed-off-by: Magnus Kroken <mkroken@gmail.com>
The previous commit introduced a regression for netns jails without
jail_ifname set. Fix that.
Fixes: 4e4f7c6d2d ("netifd: network namespace jail improvements")
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
aaaca2e interface: allocate and free memory for jail name
d93126d interface: allow renaming interface when moving to jail netns
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
CONFIG_WRITE functionality is not used and could be removed.
Looks helpful for devices with small flash because wpad is also affected.
Little testing shows that about 6 KB could be saved.
Signed-off-by: Kirill Lukonin <klukonin@gmail.com>
Since linux kernel commit 0768e17073dc527ccd18ed5f96ce85f9985e9115
(2019-04-19) the asm-generic/sockios.h header no longer defines
SIOCGSTAMP. Instead it provides only SIOCGSTAMP_OLD.
The linux/sockios.h header now defines SIOCGSTAMP using either
SIOCGSTAMP_OLD or SIOCGSTAMP_NEW as appropriate. This linux only
header file is not included so we get a build failure.
Signed-off-by: Norbert van Bolhuis <nvbolhuis@aimvalley.nl>
Allows proper usage of the ss tool. Otherwise, several errors and bad
data gets thrown:
Cannot open netlink socket: Protocol not supported
Cannot open netlink socket: Protocol not supported
Cannot open netlink socket: Protocol not supported
Cannot open netlink socket: Protocol not supported
Cannot open netlink socket: Protocol not supported
Cannot open netlink socket: Protocol not supported
Cannot open netlink socket: Protocol not supported
Originally reported here: https://github.com/openwrt/packages/issues/8232
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Use upstream latest git HEAD as it allows to remove the patches
700-radius-Prevent-buffer-overflow-in-rc_mksid,
701-pppd-Fix-bounds-check-in-EAP-code and
702-pppd-Ignore-received-EAP-messages-when-not-doing-EAP and
take in other fixes.
41a7323 pppd: Fixed spelling 'unkown' => 'unknown' (#141)
6b014be pppd: Print version information to stdout instead of stderr (#133)
cba2736 pppd: Add RFC1990 (Multilink) to the See Also section of the man page
f2f9554 pppd: Add mppe.h to the list of headers to install if MPPE is defined
ae54fcf pppd: Obfuscate password argument string
8d45443 pppd: Ignore received EAP messages when not doing EAP
8d7970b pppd: Fix bounds check in EAP code
858976b radius: Prevent buffer overflow in rc_mksid()
Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
Bump to 2.81rc5 and re-work ipset-remove-old-kernel-support.
More runtime kernel version checking is done in 2.81rc5 in various parts
of the code, so expand the ipset patch' scope to inlude those new areas
and rename to something a bit more generic.:wq
Upstream changes from rc4
532246f Tweak to DNSSEC logging.
8caf3d7 Fix rare problem allocating frec for DNSSEC.
d162bee Allow overriding of ubus service name.
b43585c Fix nameserver list in auth mode.
3f60ecd Fixed resource leak on ubus_init failure.
0506a5e Handle old kernels that don't do NETLINK_NO_ENOBUFS.
e7ee1aa Extend stop-dns-rebind to reject IPv6 LL and ULA addresses. We also reject the loopback address if rebind-localhost-ok is NOT set.
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
201e359 Handle early exit when addrstring isn't set
fa4c464 Improve address logging on early exit messages (#83)
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
It seemed very confusing when trying to select the different variants of
hostapd which are somewhat scattered about under the menu 'Network'.
Moving all hostapd variants under a common submenu helps avoid
confusion.
Inspired-by: Kevin Mahoney <kevin.mahoney@zenotec.net>
[Fixup badly formatted patch, change menu name]
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
ab7a39a umdns: fix unused error
45c4953 dns: explicitly endian-convert all fields in header and question
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
gcc 8 & 9 appear to be more picky with regards access alignment to
packed structures, leading to this warning in dns.c:
dns.c:261:2: error: converting a packed ‘struct dns_question’ pointer
(alignment 1) to a ‘uint16_t’ {aka ‘short unsigned int’} pointer
(alignment 2) may result in an unaligned pointer value
[-Werror=address-of-packed-member]
261 | uint16_t *swap = (uint16_t *) q;
Work around what I think is a false positive by turning the warning off.
Not ideal, but not quite as not ideal as build failure.
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
Recent backports to 5.5 and 5.4 broke our compat layer. This release is
to keep things running with the latest upstream stable kernels.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
* queueing: backport skb_reset_redirect change from 5.6
* version: bump
This release has only one slight change, to put it closer to the 5.6
codebase, but its main purpose is to bump us to a 1.0.y version number.
Now that WireGuard 1.0.0 has been released for Linux 5.6 [1], we can put
the same number on the backport compat codebase.
When OpenWRT bumps to Linux 5.6, we'll be able to drop this package
entirely, which I look forward to seeing.
[1] https://lists.zx2c4.com/pipermail/wireguard/2020-March/005206.html
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
If the abridged flag is set to 1 the APs that are listed in the BSS
Transition Candidate List are prioritized. If the bit is not set, the
APs have the same prioritization as the APs that are not in the list.
If you want to steer a client, you should set the flag!
The flag can be set by adding {...,'abridged': true,...} to the normal
ubus call.
Signed-off-by: Nick Hainke <vincent@systemli.org>
Subscribe to beacon reports through ubus.
Can be used for hearing map and client steering purposes.
First enable rrm:
ubus call hostapd.wlan0 bss_mgmt_enable '{"beacon_report":True}'
Subscribe to the hostapd notifications via ubus.
Request beacon report:
ubus call hostapd.wlan0 rrm_beacon_req
'{"addr":"00:xx:xx:xx:xx:xx", "op_class":0, "channel":1,
"duration":1,"mode":2,"bssid":"ff:ff:ff:ff:ff:ff", "ssid":""}'
Signed-off-by: Nick Hainke <vincent@systemli.org>
[rework identation]
Signed-off-by: David Bauer <mail@david-bauer.net>
Daniel Golle
Jason A. Donenfeld
Petr Štetiar
Adrian Schmutzler
Kevin Darbyshire-Bryant
Daniel A. Maierhofer
Rosen Penev
Yangbo Lu
Hauke Mehrtens
Hans Dedecker
Josef Schlehofer
Antonio Quartulli
Magnus Kroken
Kirill Lukonin
Norbert van Bolhuis
Peter Stadler
Nick Hainke
Jesus Fernandez Manzano
