aboutsummaryrefslogtreecommitdiffstats
path: root/package/network
Commit message (Collapse)AuthorAgeFilesLines
* hostapd: fix IEEE 802.11r (fast roaming) defaultsGospod Nassa2018-05-241-21/+27
| | | | | | | | | | | | | | | | | | | | | | | | | Use ft_psk_generate_local=1 by default, as it makes everything else fairly trivial. All of the r0kh/r1kh and key management stuff goes away and hostapd fairly much does it all for us. We do need to provide nas_identifier, which can be derived from the BSSID, and we need to generate a mobility_domain, for which we default to the first four chars of the md5sum of the SSID. The complex manual setup should also still work, but the defaults also now work easily out of the box. Verified by manually running hostapd (with the autogenerated config) and watching the debug output: wlan2: STA ac:37:43:a0:a6:ae WPA: FT authentication already completed - do not start 4-way handshake This was previous submitted to LEDE in https://github.com/lede-project/source/pull/1382 [dwmw2: Rewrote commit message] Signed-off-by: Gospod Nassa <devianca@gmail.com> Signed-off-by: David Woodhouse <dwmw2@infradead.org> (cherry picked from commit 3cc56a5534b8b49a7e9ba57edf9878ec32bdd27a)
* wireguard: no longer need portability patchKevin Darbyshire-Bryant2018-05-242-19/+1
| | | | | | | | | | Drop package/network/services/wireguard/patches/100-portability.patch Instead pass 'PLATFORM=linux' to make since we are always building FOR linux. Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk> (cherry picked from commit f06def4221c25f4ee43d31c2b90a1371ba4d6d33)
* ltq-xdsl-app: start after led scriptMathias Kresin2018-05-242-3/+2
| | | | | | | | | | | | | During handshake we are highjack and reset a LED to the configured trigger afterwards. ltq-xdsl-app need to start after the LED init script, to ensure that the LED init script doesn't re-highjack the LED we are currently using for handshake indication. Drop the comment about the atm dependency. The dependency was fixed quite some time ago by using hotplug scripts for br2684ctl. Signed-off-by: Mathias Kresin <dev@kresin.me> (cherry picked from commit 0678cc850ccc068ba33eb87f5f1bf8294a87e611)
* Revert "dnsmasq: use "hostsdir" instead of "addn-hosts""Hans Dedecker2018-05-241-1/+1
| | | | | | | | | This reverts commit c97189e26dfb9b0eaf5b9cd7e33a99e2f780cb58 as it has different issues: -Host file is not written in a directory unique per dnsmasq instance -odhcpd writes host info into the same directory but still sends a SIGHUP to dnsmasq Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
* dnsmasq: use "hostsdir" instead of "addn-hosts"Christian Schoenebeck2018-05-231-1/+1
| | | | | | | 1.) "addn-hosts" per default point to a file (but it supports directory) 2.) "hostsdir" only support directory with the additional benefit: New or changed files are read automatically. Signed-off-by: Christian Schoenebeck <christian.schoenebeck@gmail.com>
* wireguard: bump to 20180519Jason A. Donenfeld2018-05-221-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | * chacha20poly1305: add mips32 implementation "The OpenWRT Commit" - this significantly speeds up performance on cheap plastic MIPS routers, and presumably the remaining MIPS32r2 super computers out there. * timers: reinitialize state on init * timers: round up instead of down in slack_time * timers: remove slack_time * timers: clear send_keepalive timer on sending handshake response * timers: no need to clear keepalive in persistent keepalive Andrew He and I have helped simplify the timers and remove some old warts, making the whole system a bit easier to analyze. * tools: fix errno propagation and messages Error messages are now more coherent. * device: remove allowedips before individual peers This avoids an O(n^2) traversal in favor of an O(n) one. Before systems with many peers would grind when deleting the interface. Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
* curl: bump to 7.60.0Hans Dedecker2018-05-173-15/+4
| | | | | | | | | Refresh patches; remove 320-mbedtls_dont_use_deprecated_sha256_function patch as upstream fixed For changes in version 2.60 see https://curl.haxx.se/changes.html#7_60_0 Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
* wireguard: bump to 20180514Kevin Darbyshire-Bryant2018-05-172-9/+10
| | | | | | | | | | | | | | | 52be69b version: bump snapshot 4884b45 ncat-client-server: add wg-quick variant a333551 wg-quick: add darwin implementation f5bf84d compat: backport for OpenSUSE 15 fe1ae1b wg-quick: add wg symlink ecc1c5f wg-quick: add android implementation 3e6bb79 tools: reorganize for multiplatform wg-quick b289d12 allowedips: Fix graphviz output after endianness patch Refresh cross compile compatibility patch Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
* wireguard: Add support for ip6prefix config optionToke Høiland-Jørgensen2018-05-151-0/+6
| | | | | | | | | | | | This makes it easier to distribute prefixes over a wireguard tunnel interface, by simply setting the ip6prefix option in uci (just like with other protocols). Obviously, routing etc needs to be setup properly for things to work; this just adds the config option so the prefix can be assigned to other interfaces. Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk>
* hostapd: fix VHT80 for encrypted mesh channel settingsSven Eckelmann2018-05-142-1/+46
| | | | | | | | | | | | | | | | | | | The max_oper_chwidth settings was parsed incorrectly for big endian system. This prevented the system to switch to VHT80 (or VHT160). Instead they were mapped to: * HT20: 20MHz * VHT20: 20MHz * HT40: 40MHz * VHT40: 40MHz * VHT80: 40MHz * VHT160: 40MHz This happened because each max_oper_chwidth setting in the config file was parsed as "0" instead of the actual value. Fixes: a4322eba2b12 ("hostapd: fix encrypted mesh channel settings") Signed-off-by: Sven Eckelmann <sven.eckelmann@openmesh.com>
* firewall3: update to latest git HEADJohn Crispin2018-05-141-4/+4
| | | | | | | | b45e162 helpers: fix the set_helper in the rule structure f742ba7 helpers.conf: support also tcp in the CT sip helper 08b2c61 helpers: make the proto field as a list rather than one option Signed-off-by: John Crispin <john@phrozen.org>
* hostapd: fix mesh+APDaniel Golle2018-05-1415-220/+171
| | | | | | | | Fix encrypted (or DFS) AP+MESH interface combination in a way similar to how it's done for AP+STA and fix netifd shell script. Refresh patches while at it. Signed-off-by: Daniel Golle <daniel@makrotopia.org>
* wireguard: bump to 0.0.20180513Kevin Darbyshire-Bryant2018-05-141-2/+2
| | | | | | | | | | | | | | | | | | | 6b4a340 version: bump snapshot faa2103 compat: don't clear header bits on RHEL 4014532 compat: handle RHEL 7.5's recent backports 66589bc queueing: preserve pfmemalloc header bit 37f114a chacha20poly1305: make gcc 8.1 happy 926caae socket: use skb_put_data 724d979 wg-quick: preliminary support for go implementation c454c26 allowedips: simplify arithmetic 71d44be allowedips: produce better assembly with unsigned arithmetic 5e3532e allowedips: use native endian on lookup 856f105 allowedips: add selftest for allowedips_walk_by_peer 41df6d2 embeddable-wg-library: zero attribute padding 9a1bea6 keygen-html: add zip file example f182b1a qemu: retry on 404 in wget for kernel.org race Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
* dnsmasq: bump to 2.80test2Kevin Darbyshire-Bryant2018-05-1212-8/+1172
| | | | | | | | | | | | | | | | | | | Refresh patches and backport upstream to current HEAD: 1f1873a Log warning on very large cachesize config, instead of truncating it. 0a496f0 Do unsolicited RAs for interfaces which appear after dnsmasq startup. e27825b Fix logging in previous. 1f60a18 Retry SERVFAIL DNSSEC queries to a different server, if possible. a0088e8 Handle query retry on REFUSED or SERVFAIL for DNSSEC-generated queries. 34e26e1 Retry query to other servers on receipt of SERVFAIL rcode. 6b17335 Add packet-dump debugging facility. 07ed585 Add logging for DNS error returns from upstream and local configuration. 0669ee7 Fix DHCP broken-ness when --no-ping AND --dhcp-sequential-ip are set. f84e674 Be persistent with broken-upstream-DNSSEC warnings. Compile & run tested: ar71xx Archer C7 v2 Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
* dnsmasq: add specific interface procd triggershux2018-05-112-2/+14
| | | | | | | | | | | | | | | | | | Right now interface.update events are sent out by netifd upon interface state, route, address (lifetime), prefix lifetime changes. Dnsmasq is only interested in interface state changes and currently adds an interface trigger for all the "interface.*" events. In combination with commit 23bba9cb330, which triggers a SIGHUP signal to dnsmasq, IPv6 address/prefix lifetime changes on the wan will trigger dnsmasq reloads which can become frequent in case of shorter lifetimes. To avoid frequent dnsmasq reload, this patch adds specific interface triggers. During dnsmasq init it loops dhcp uci section; if the value of the ignore option is set to 0, then the corresponding interface trigger is not installed. Otherwise, if the ignore option value is 1, then procd_add_interface_trigger is called which adds the interface trigger. Signed-off-by: hux <xinxing.huchn@gmail.com>
* igmpproxy: bump to 0.2.1Kevin Darbyshire-Bryant2018-05-0710-679/+4
| | | | | | | | | | | | | | The sender domain has a DMARC Reject/Quarantine policy which disallows sending mailing list messages using the original "From" header. To mitigate this problem, the original message has been wrapped automatically by the mailing list software. Point at github which is new, maintained location for igmpproxy. Remove all patches as all have been upstreamed. Closes FS#1456 Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
* igmpproxy: fix creation of firewall rulesJaap Buurman2018-05-073-9/+14
| | | | | | | | The init sccript for igmpproxy uses the option 'network' both as an interface name for fetching the l3_device name and for creating the firewall rules. This only works if the name of the network and firewall zone are identical. This commit introduces a new option 'zone' for configuring the upstream and downstream firewall zones in order for the init script to create the required firewall rules automatically. When no such options are given, the init script falls back to not creating the firewall rules and the user can opt to create these manually. Signed-off-by: Jaap Buurman <jaapbuurman@gmail.com>
* hostapd: add channel utilization as config optionNick Hainke2018-05-071-2/+7
| | | | | | Add the channel utilization as hostapd configuration option. Signed-off-by: Nick Hainke <vincent@systemli.org>
* iproute2: import latest cakeKevin Darbyshire-Bryant2018-05-072-446/+425
| | | | | | | | | | | | | | | | | | The sender domain has a DMARC Reject/Quarantine policy which disallows sending mailing list messages using the original "From" header. To mitigate this problem, the original message has been wrapped automatically by the mailing list software. Bearing fruits of the latest upstreaming efforts on cake. Changes: diffserv-llt dropped. The paper describing this DSCP allocation has gone stale and doesn't appear used. The userspace to kernel netlink messages for cake have been reworked in a backwards incompatible way, so tc & cake must be bumped together this once. Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
* iproute2: backport json_print-fix-hidden-64-bit-type-promotionKevin Darbyshire-Bryant2018-05-072-1/+289
| | | | | | | | | | | | | | | | | | | | | | | The sender domain has a DMARC Reject/Quarantine policy which disallows sending mailing list messages using the original "From" header. To mitigate this problem, the original message has been wrapped automatically by the mailing list software. print_uint() will silently promote its variable type to uint64_t, but there is nothing that ensures that the format string specifier passed along with it fits (and the function name suggest to pass "%u"). Fix this by changing print_uint() to use a native 'unsigned int' type, and introduce a separate print_u64() function for printing 64-bit values. All call sites that were actually printing 64-bit values using print_uint() are converted to use print_u64() instead. Since print_int() was already using native int types, just add a print_s64() to match, but don't convert any call sites. Fixes wonkyness in some stats from some qdiscs under tc Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
* hostapd: fix a mesh mode crash with CONFIG_TAXONOMY enabledFelix Fietkau2018-05-031-0/+23
| | | | Signed-off-by: Felix Fietkau <nbd@nbd.name>
* samba36: add hotplug supportRosy Song2018-05-023-0/+114
| | | | | | | | Add hotplug handle script for storage devices, this will add corresponding option in the /etc/config/samba file automatically. Signed-off-by: Rosy Song <rosysong@rosinson.com>
* iproute2: update to 4.16Hans Dedecker2018-05-024-17/+7
| | | | | | | | | Update to latest version of iproute2, refresh patches. See https://lkml.org/lkml/2018/4/2/349 for a full overview of the changes in 4.16. Build and tested on AR7xxx against musl Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
* iptables: fix per object LDFLAGS for aggragate object buildsJohn Crispin2018-05-011-5/+5
| | | | | | | Without this patch the extra LDFLAGS of objects were selected based on the name of the extension being built, which breaks for aggregate so builds. Signed-off-by: John Crispin <john@phrozen.org>
* odhcp6c: update to latest git HEADHans Dedecker2018-04-291-3/+3
| | | | | | | 5316d7f ra: always trigger update in case of RA parameter change 327f73d dhcpv6: fix strncpy bounds Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
* netifd: update to latest git HEAD (Coverity fixes)Hans Dedecker2018-04-271-3/+3
| | | | | | | 56ceced interface-ip: remove superfluous iface check in interface_ip_set_enabled() 4f4a8c0 system-linux: fix strncpy bounds Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
* iftop: bump to latestKevin Darbyshire-Bryant2018-04-271-3/+3
| | | | | | | | | | | | The sender domain has a DMARC Reject/Quarantine policy which disallows sending mailing list messages using the original "From" header. To mitigate this problem, the original message has been wrapped automatically by the mailing list software. Choose first running interface, rather than first "up" interface (Redhat #1403025) Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk> Tested-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
* odhcpd: update to latest git HEADHans Dedecker2018-04-231-4/+4
| | | | | | | 4136529 dhcpv6-ia: keep tentative assignments alive for a short time 200cc8f dhcpv6-ia: make assignment lookup more strict Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
* wireguard: bump to 20180420Kevin Darbyshire-Bryant2018-04-201-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | 7cc2668 version: bump snapshot 860c7c7 poly1305: do not place constants in different sections 5f1e4ca compat: remove unused dev_recursion_level backport 7e4b991 blake2s: remove unused helper 13225fc send: simplify skb_padding with nice macro a1525bf send: account for route-based MTU bbb2fde wg-quick: account for specified fwmark in auto routing mode c452105 qemu: bump default version dbe5223 version: bump snapshot 1d3ef31 chacha20poly1305: put magic constant behind macro cdc164c chacha20poly1305: add self tests from wycheproof 1060e54 curve25519: add self tests from wycheproof 0e1e127 wg-quick.8: fix typo 2b06b8e curve25519: precomp const correctness 8102664 curve25519: memzero in batches 1f54c43 curve25519: use cmov instead of xor for cswap fa5326f curve25519: use precomp implementation instead of sandy2x 9b19328 compat: support OpenSUSE 15 3102d28 compat: silence warning on frankenkernels 8f64c61 compat: stable kernels are now receiving b87b619 62127f9 wg-quick: hide errors on save Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
* hostapd: fix encrypted mesh channel settingsDaniel Golle2018-04-205-1/+226
| | | | | | | | | | | | | | | Import two patches from Peter Oh to allow setting channel bandwidth in the way it already works for managed interfaces. This fixes mesh interfaces on 802.11ac devices always coming up in VHT80 mode. Add a patch to allow HT40 also on 2.4GHz if noscan option is set, which also skips secondary channel scan just like noscan works in AP mode. This time also make sure to add all files to the patch before committing it... Signed-off-by: Daniel Golle <daniel@makrotopia.org>
* Revert "hostapd: fix encrypted mesh channel settings"Felix Fietkau2018-04-205-213/+1
| | | | | | | This reverts commit 7f52919a2f2894125b4dca611eb2d30181af7e0b, which is currently breaking the builds and needs to be reworked Signed-off-by: Felix Fietkau <nbd@nbd.name>
* hostapd: fix encrypted mesh channel settingsDaniel Golle2018-04-205-1/+213
| | | | | | | | | | | | Import two patches from Peter Oh to allow setting channel bandwidth in the way it already works for managed interfaces. This fixes mesh interfaces on 802.11ac devices always coming up in VHT80 mode. Add a patch to allow HT40 also on 2.4GHz if noscan option is set, which also skips secondary channel scan just like noscan works in AP mode. Signed-off-by: Daniel Golle <daniel@makrotopia.org>
* netifd: update to latest git HEADDaniel Golle2018-04-201-3/+3
| | | | | | b3dca7b wireless: include noscan option in common wdev vars Signed-off-by: Daniel Golle <daniel@makrotopia.org>
* odhcpd: update to latest git HEAD (FS#1457)Hans Dedecker2018-04-191-4/+4
| | | | | | dcfc06a router: fix managed address configuration setting Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
* network/uqmi: pipe the output off qmi_wds_stop to /dev/nullFlorian Eckert2018-04-191-3/+8
| | | | | | | | | | Pipe uqmi output from qmi_wds_stop function into /dev/null. This will supress the following output in proto teardown. netifd: wwan (x): "No effect" netifd: wwan (x): Command failed: Permission denied Signed-off-by: Florian Eckert <fe@dev.tdt.de>
* hostapd: mesh: make forwarding configurableDaniel Golle2018-04-183-2/+245
| | | | | | | | | | | For unencrypted mesh networks our scripts take care of setting the various mesh_param values. wpa_supplicant changes somes of them when being used for SAE encrypted mesh and previously didn't allow configuring any of them. Add support for setting mesh_fwding (which has to be set to 0 when using other routing protocols on top of 802.11s) and update our script to pass the value to wpa_supplicant. Signed-off-by: Daniel Golle <daniel@makrotopia.org>
* curl: Deprecate idn(1) support and switch to xz tarballDaniel Engberg2018-04-162-8/+9
| | | | | | | libidn(1) is deprecated, add libidn2 support Switch to xz tarball (smaller size) Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
* netifd: update to latest git HEAD (coverity fixes)Hans Dedecker2018-04-161-3/+3
| | | | | | | | | | | | 513eb27 system-linux: check ioctl return value in system_vlan() df1625d system-linux: check ioctl return value in system_if_flags() 209c508 system-linux: fix segfault on alloc failure in system_if_check() 4a8e20e system-linux: fix segfault on error in system_add_ip6_tunnel() 36e4700 handler: fix resource leak on error in netifd_init_script_handlers() 86a0e7c system-linux: remove unnecessary open call in system_if_dump_info() 1e2cf67 system-linux: fix memory leak on error in system_add_vxlan() Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
* lldpd: bump to 1.0.1Stijn Tintel2018-04-151-2/+2
| | | | Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
* hostapd: fix compile of -mini variantsDaniel Golle2018-04-131-0/+23
| | | | | | Fixes commit d88934aa5a (hostapd: update to git snapshot of 2018-04-09) Signed-off-by: Daniel Golle <daniel@makrotopia.org>
* hostapd: update to git snapshot of 2018-04-09Daniel Golle2018-04-1332-76/+1085
| | | | | | | | | And import patchset to allow 802.11s mesh on DFS channels, see also http://lists.infradead.org/pipermail/hostap/2018-April/038418.html Fix sae_password for encryption mesh (sent upstream as well). Also refreshed existing patches and fixed 463-add-mcast_rate-to-11s. Signed-off-by: Daniel Golle <daniel@makrotopia.org>
* ebtables: update to latest git 2018-04-11Matthias Schiffer2018-04-122-5/+5
| | | | | | | | | | | 2e783b227766 ebt_ip: add support for matching IGMP type b5fbb8d786c9 ebt_ip: add support for matching ICMP type and code c5e5b784fd1a Move ICMP type handling functions from ebt_ip6 to useful_functions.c 11da52177196 include: sync linux/netfilter_bridge/ebt_ip.h with kernel Note: the new features require at least kernel 4.17 or backported patches. Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
* iptables: split physdev match out as a separate packageMatthias Schiffer2018-04-091-1/+10
| | | | | | | Split physdev match out of ipt-extra to allow installing ipt-extra without pulling in br-netfilter. Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
* dnsmasq: remove example domains from rfc6761.conf (FS#1447)Hans Dedecker2018-04-092-5/+1
| | | | | | | | | | | | RFC6771 does not exclude the forwarding of the example domain as it states : "Caching DNS servers SHOULD NOT recognize example names as special and SHOULD resolve them normally." Example domains cannot be assigned to any user or person by DNS registrars as they're registered in perpetuity to IANA meaning they can be resolved; therefore let's remove the example domains from the rfc6761.conf file. Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
* odhcpd: update to latest git HEADHans Dedecker2018-04-091-4/+4
| | | | | | 31f217f router: improve RFC7084 compliancy Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
* odhcp6c: update to latest git HEADHans Dedecker2018-04-051-4/+4
| | | | | | | | 74b5a3 script: fix possible negative delay 473f248 dhcpv6: always trigger script update in case of IA updates ea18935 ra: rework route information option handling Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
* firewall: update to the latest version, adds hw flow offload supportFelix Fietkau2018-04-051-3/+3
| | | | | | 35b3e74 defaults: add support for setting --hw on the xt_FLOWOFFLOAD rule Signed-off-by: Felix Fietkau <nbd@nbd.name>
* kernel: add support for enabling hardware flow offload via iptablesFelix Fietkau2018-04-051-5/+82
| | | | Signed-off-by: Felix Fietkau <nbd@nbd.name>
* samba36: fix some security problemsHauke Mehrtens2018-04-038-3/+322
| | | | | | | | | | This Adds fixes for the following security problems based on debians patches: CVE-2016-2125: Unconditional privilege delegation to Kerberos servers in trusted realms CVE-2017-12163: Server memory information leak over SMB1 CVE-2017-12150: SMB1/2/3 connections may not require signing where they should CVE-2018-1050: Denial of Service Attack on external print server. Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
* netifd: update to latest git HEADHans Dedecker2018-04-031-3/+3
| | | | | | 3dc8c91 interface-ip: fix memory leak in interface_ip_add_target_route() Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
generated by cgit v1.2.3 (git 2.25.1) at 2024-07-10 20:49:36 +0000