openwrt/staging/dedeckeh.git
6 years agohostapd: backport build fix when OWE is activated
Hauke Mehrtens [Wed, 3 Oct 2018 18:56:41 +0000 (20:56 +0200)]
hostapd: backport build fix when OWE is activated

This backports a compile fix form the hostapd project.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
6 years agohostapd: sync config with default configuration
Hauke Mehrtens [Tue, 2 Oct 2018 22:05:11 +0000 (00:05 +0200)]
hostapd: sync config with default configuration

This replaces the configuration files with the versions from the hostapd
project and the adaptions done by OpenWrt.

The resulting binaries should be the same.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
6 years agonetifd: update to latest git HEAD
Hauke Mehrtens [Sun, 14 Oct 2018 11:51:58 +0000 (13:51 +0200)]
netifd: update to latest git HEAD

22476ff wireless: Add Simultaneous Authentication of Equals (SAE)
c6c3a0d wireless: Add Opportunistic Wireless Encryption (OWE)
a117e41 wireless: Add WPA-EAP-SUITE-B-192 (WPA3-Enterprise)

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
6 years agouqmi: fix variable initilization for timeout handling
Florian Eckert [Thu, 11 Oct 2018 14:30:55 +0000 (16:30 +0200)]
uqmi: fix variable initilization for timeout handling

Also add logging output for SIM initilization.

Signed-off-by: Florian Eckert <fe@dev.tdt.de>
6 years agomac80211: fix A-MSDU packet handling with TCP retransmission
Felix Fietkau [Thu, 11 Oct 2018 13:16:28 +0000 (15:16 +0200)]
mac80211: fix A-MSDU packet handling with TCP retransmission

Improves local TCP throughput and fixes use-after-free bugs that could lead
to crashes.

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agobcm53xx: add pending pinctrl driver
Rafał Miłecki [Thu, 11 Oct 2018 11:55:57 +0000 (13:55 +0200)]
bcm53xx: add pending pinctrl driver

It's required to support devices using adjustable SoC pins for some
specific purpose (e.g. I2C, PWM, UART1).

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
6 years agoath10k-ct: Update to 29-09-2018
Robert Marko [Tue, 9 Oct 2018 13:05:56 +0000 (15:05 +0200)]
ath10k-ct: Update to 29-09-2018

Update ath10k-ct to be able to drop 210-ath10k-fix-recent-bandwidth-conversion-bug.patch
as its upstream.

b9989fbd5d6e ath10k-ct: Add upstream patch to fix peer rate reporting.
ac9224344dbf ath10k-ct: Support sending custom frames with no-ack flag.
bc938bc2021e ath10k-ct: Support sending pkts with specific rate on 10.4 firmware.

Runtime tested on:
- GL-iNet GL-B1300
- Mikrotik RB912 + QCA9882

Signed-off-by: Robert Marko <robimarko@gmail.com>
[Added list of all changes from previous version + add own test device]
Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agouqmi: update PKG_RELEASE version
Florian Eckert [Tue, 9 Oct 2018 09:32:36 +0000 (11:32 +0200)]
uqmi: update PKG_RELEASE version

update PKG_RELEASE

Signed-off-by: Florian Eckert <fe@dev.tdt.de>
6 years agouqmi: stop proto handler if verify pin count is not 3
Florian Eckert [Thu, 12 Apr 2018 08:31:29 +0000 (10:31 +0200)]
uqmi: stop proto handler if verify pin count is not 3

Check pin count value from pin status and stop verification the pin if
the value is less then 3. This should prevent the proto-handler to
lock the SIM. If SIM is locked then the PUK is needed.

Signed-off-by: Florian Eckert <fe@dev.tdt.de>
6 years agouqmi: evaluate pin-status output in qmi_setup function
Florian Eckert [Tue, 10 Apr 2018 10:55:11 +0000 (12:55 +0200)]
uqmi: evaluate pin-status output in qmi_setup function

Load the json output from uqmi --get-pin-status command and evaluate the
"pin1_status" value.

The following uqmi "pin1_status" values are evaluated:

- disabled
  Do not verify PIN because SIM verification is disabled on this SIM

- blocked
  Stop qmi_setup because SIM is locked and a PUK is required

- not_verified
  SIM is not yet verified. Do a uqmi --verify-pin1 command if a SIM is
  specified

- verified:
  Do not verify the PIN because this was already done before

Signed-off-by: Florian Eckert <fe@dev.tdt.de>
6 years agouqmi: do not block proto handler if SIM is uninitialized
Florian Eckert [Tue, 10 Apr 2018 13:56:24 +0000 (15:56 +0200)]
uqmi: do not block proto handler if SIM is uninitialized

QMI proto setup-handler will wait forever if SIM does not get initialized.
To fix this stop polling pin status and notify netifd. Netifd will generate
then a "ifup-failed" ACTION.

Signed-off-by: Florian Eckert <fe@dev.tdt.de>
6 years agouqmi: do not block proto handler if modem is unable to registrate
Florian Eckert [Thu, 12 Apr 2018 11:20:49 +0000 (13:20 +0200)]
uqmi: do not block proto handler if modem is unable to registrate

QMI proto setup-handler will wait forever if it is unable to registrate to
the mobile network. To fix this stop polling network registration status
and notify netifd. Netifd will generate then a "ifup-failed" ACTION.

Signed-off-by: Florian Eckert <fe@dev.tdt.de>
6 years agouqmi: add timeout option value
Florian Eckert [Thu, 12 Apr 2018 11:29:34 +0000 (13:29 +0200)]
uqmi: add timeout option value

This value will be used for now during following situations:
* Ask the sim with the uqmi --get-pin-status command.
* Wait for network registration with the uqmi --get-serving-system command.

This two commands wait forever in a while loop. Add a timeout to stop
waiting and so inform netifd.

Signed-off-by: Florian Eckert <fe@dev.tdt.de>
6 years agouqmi: redirect uqmi commands output to /dev/null
Florian Eckert [Thu, 12 Apr 2018 11:10:33 +0000 (13:10 +0200)]
uqmi: redirect uqmi commands output to /dev/null

Move uqmi std and error output on commands without using them to /dev/null.
This will remove useless outputs in the syslog.

Signed-off-by: Florian Eckert <fe@dev.tdt.de>
6 years agouqmi: fix indenting
Florian Eckert [Tue, 10 Apr 2018 14:29:05 +0000 (16:29 +0200)]
uqmi: fix indenting

fix indenting

Signed-off-by: Florian Eckert <fe@dev.tdt.de>
6 years agobcm53xx: replace SPI revert with a fix sent upstream
Rafał Miłecki [Thu, 11 Oct 2018 08:01:45 +0000 (10:01 +0200)]
bcm53xx: replace SPI revert with a fix sent upstream

Instead of reverting whole commit it's enough to just revert a single
line change. It seems the real problem with the regressing commit was a
bump of read chunk size. Switching back to 256 B chunks is enough to fix
the problem/regression.

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
6 years agoprocd: update to latest git HEAD
John Crispin [Thu, 11 Oct 2018 06:42:24 +0000 (08:42 +0200)]
procd: update to latest git HEAD

94944ab procd: Add cpu string to board detection

Signed-off-by: John Crispin <john@phrozen.org>
6 years agopackage/: fix $(PROJECT_GIT) usage
John Crispin [Thu, 11 Oct 2018 06:37:59 +0000 (08:37 +0200)]
package/: fix $(PROJECT_GIT) usage

Signed-off-by: John Crispin <john@phrozen.org>
6 years agoprocd: Install hotplug files as 600
Rosen Penev [Wed, 19 Sep 2018 19:03:32 +0000 (12:03 -0700)]
procd: Install hotplug files as 600

procd runs as root, so it only makes sense that its files are restricted.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
6 years agolinux-atm: Install hotplug file as 600
Rosen Penev [Wed, 19 Sep 2018 19:03:31 +0000 (12:03 -0700)]
linux-atm: Install hotplug file as 600

The hotplug files is only used by procd, which runs as root.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
6 years agocomgt: Install hotplug and netifd files as 600
Rosen Penev [Wed, 19 Sep 2018 19:03:30 +0000 (12:03 -0700)]
comgt: Install hotplug and netifd files as 600

procd and netifd both run as root. These files are not used elsewhere.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
6 years agosamba36: Install several config files as 600
Rosen Penev [Wed, 19 Sep 2018 19:03:29 +0000 (12:03 -0700)]
samba36: Install several config files as 600

Hotplug is managed by procd, which runs as root. The other files are used
by root as well.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
6 years agosoloscli: Install hotplug file as 600
Rosen Penev [Wed, 19 Sep 2018 19:03:28 +0000 (12:03 -0700)]
soloscli: Install hotplug file as 600

Hotplug is managed by procd, which runs as root.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
6 years agofirewall: Install config files as 600
Rosen Penev [Wed, 19 Sep 2018 19:03:26 +0000 (12:03 -0700)]
firewall: Install config files as 600

None of the files in firewall are used by non-root.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
6 years agompc85xx: remove kernel 4.9 support
Magnus Kroken [Fri, 5 Oct 2018 23:46:24 +0000 (01:46 +0200)]
mpc85xx: remove kernel 4.9 support

Signed-off-by: Magnus Kroken <mkroken@gmail.com>
6 years agompc85xx: switch to kernel 4.14
Magnus Kroken [Fri, 5 Oct 2018 23:46:23 +0000 (01:46 +0200)]
mpc85xx: switch to kernel 4.14

Signed-off-by: Magnus Kroken <mkroken@gmail.com>
6 years agompc85xx: add kernel 4.14 support
Magnus Kroken [Fri, 5 Oct 2018 23:46:22 +0000 (01:46 +0200)]
mpc85xx: add kernel 4.14 support

Based on patches previously submitted by Achim Gottinger:
http://lists.infradead.org/pipermail/openwrt-devel/2018-June/012719.html

Tested on TP-Link TL-WDR4900 v1.

Signed-off-by: Magnus Kroken <mkroken@gmail.com>
6 years agonetfilter: add missing dependency for kernel 4.14
Koen Vandeputte [Wed, 10 Oct 2018 15:31:40 +0000 (17:31 +0200)]
netfilter: add missing dependency for kernel 4.14

Since kernel 4.14.75 commit ("netfilter: xt_cluster: add dependency on conntrack module")
a dependency is required on kmod-nf-conntrack.

It seems this was already present for kmod-ipt-clusterip
but not yet for kmod-ipt-cluster

Add it fixing a build error when including kmod-ipt-cluster:

Package kmod-ipt-cluster is missing dependencies for the following libraries:
nf_conntrack.ko
modules/netfilter.mk:665: recipe for target '/mnt/ramdisk/koen/firmware/builds/openwrt/bin/targets/cns3xxx/generic/packages/kmod-ipt-cluster_4.14.75-1_arm_mpcore_vfp.ipk' failed
make[3]: *** [/mnt/ramdisk/koen/firmware/builds/openwrt/bin/targets/cns3xxx/generic/packages/kmod-ipt-cluster_4.14.75-1_arm_mpcore_vfp.ipk] Error 1
make[3]: Leaving directory '/mnt/ramdisk/koen/firmware/builds/openwrt/package/kernel/linux'
Command exited with non-zero status 2
time: package/kernel/linux/compile#1.80#0.05#2.07
package/Makefile:107: recipe for target 'package/kernel/linux/compile' failed
make[2]: *** [package/kernel/linux/compile] Error 2
make[2]: Leaving directory '/mnt/ramdisk/koen/firmware/builds/openwrt'
package/Makefile:103: recipe for target '/mnt/ramdisk/koen/firmware/builds/openwrt/staging_dir/target-arm_mpcore+vfp_musl_eabi/stamp/.package_compile' failed
make[1]: *** [/mnt/ramdisk/koen/firmware/builds/openwrt/staging_dir/target-arm_mpcore+vfp_musl_eabi/stamp/.package_compile] Error 2
make[1]: Leaving directory '/mnt/ramdisk/koen/firmware/builds/openwrt'
/mnt/ramdisk/koen/firmware/builds/openwrt/include/toplevel.mk:216: recipe for target 'world' failed
make: *** [world] Error 2

Fixes: f983956a8b72 ("kernel: bump 4.14 to 4.14.75")
Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
[1] https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v4.14.75&id=b969656b46626a674232c0eadf92a394b89df07c

6 years agoar71xx: remove linux 4.9 support
Koen Vandeputte [Wed, 10 Oct 2018 09:42:42 +0000 (11:42 +0200)]
ar71xx: remove linux 4.9 support

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agoar71xx: switch to kernel 4.14
Koen Vandeputte [Thu, 30 Aug 2018 17:10:05 +0000 (19:10 +0200)]
ar71xx: switch to kernel 4.14

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agokernel: bump 4.14 to 4.14.75
Koen Vandeputte [Wed, 10 Oct 2018 10:41:32 +0000 (12:41 +0200)]
kernel: bump 4.14 to 4.14.75

Refreshed all patches.

Compile-tested on: ar71xx, cns3xxx, imx6
Runtime-tested on: ar71xx, cns3xxx, imx6

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agokernel: bump 4.9 to 4.9.132
Koen Vandeputte [Wed, 10 Oct 2018 09:36:49 +0000 (11:36 +0200)]
kernel: bump 4.9 to 4.9.132

Refreshed all patches.

Compile-tested on: ar71xx
Runtime-tested on: ar71xx

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agokernel: enable memory compaction
Felix Fietkau [Tue, 9 Oct 2018 11:22:46 +0000 (13:22 +0200)]
kernel: enable memory compaction

Compaction is the only memory management component to form high order (larger
physically contiguous) memory blocks reliably. The page allocator relies on
compaction heavily and the lack of the feature can lead to unexpected OOM
killer invocations for high order memory requests. You shouldn't disable this
option unless there really is a strong reason for it.

Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Michal Hrusecky <michal.hrusecky@nic.cz>
6 years agomt76: fix typo in kmod-mt76x0u package description
Felix Fietkau [Mon, 8 Oct 2018 11:48:32 +0000 (13:48 +0200)]
mt76: fix typo in kmod-mt76x0u package description

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agokernel: support gcc-optimized inlining on all architectures
Felix Fietkau [Thu, 25 Jan 2018 13:01:34 +0000 (14:01 +0100)]
kernel: support gcc-optimized inlining on all architectures

Optimized inlining was disabled by default when gcc 4 was still
relatively new. By now, all gcc versions handle this well and there
seems to be no real reason to keep it x86-only.

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agodnsmasq: add dhcp-ignore-names support - CERT VU#598349
Kevin Darbyshire-Bryant [Mon, 8 Oct 2018 19:40:36 +0000 (20:40 +0100)]
dnsmasq: add dhcp-ignore-names support - CERT VU#598349

dnsmasq v2.80test8 adds the ability to ignore dhcp client's requests for
specific hostnames.  Clients claiming certain hostnames and thus
claiming DNS namespace represent a potential security risk. e.g. a
malicious host could claim 'wpad' for itself and redirect other web
client requests to it for nefarious purpose. See CERT VU#598349 for more
details.

Some Samsung TVs are claiming the hostname 'localhost', it is believed
not (yet) for nefarious purposes.

/usr/share/dnsmasq/dhcpbogushostname.conf contains a list of hostnames
in correct syntax to be excluded. e.g.

dhcp-name-match=set:dhcp_bogus_hostname,localhost

Inclusion of this file is controlled by uci option dhcpbogushostname
which is enabled by default.

To be absolutely clear, DHCP leases to these requesting hosts are still
permitted, but they do NOT get to claim ownership of the hostname
itself and hence put into DNS for other hosts to be confused/manipulate by.

Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
6 years agowireguard: bump to 0.0.20181007
Kevin Darbyshire-Bryant [Tue, 9 Oct 2018 08:03:59 +0000 (09:03 +0100)]
wireguard: bump to 0.0.20181007

64750c1 version: bump snapshot
f11a2b8 global: style nits
4b34b6a crypto: clean up remaining .h->.c
06d9fc8 allowedips: document additional nobs
c32b5f9 makefile: do more generic wildcard so as to avoid rename issues
20f48d8 crypto: use BIT(i) & bitmap instead of (bitmap >> i) & 1
b6e09f6 crypto: disable broken implementations in selftests
fd50f77 compat: clang cannot handle __builtin_constant_p
bddaca7 compat: make asm/simd.h conditional on its existence
b4ba33e compat: account for ancient ARM assembler

Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
6 years agoe2fsprogs: fix glibc compile issue (FS#1749,FS#1796)
Hans Dedecker [Mon, 8 Oct 2018 14:57:01 +0000 (16:57 +0200)]
e2fsprogs: fix glibc compile issue (FS#1749,FS#1796)

Fixes the following build error:

.../toolchain-i386_pentium4_gcc-7.3.0_glibc/lib/gcc/i486-openwrt-linux-gnu/7.3.0/../../../../i486-openwrt-linux-gnu/bin/ld: ../lib/libcom_err.so: undefined reference to `sem_post’
.../toolchain-i386_pentium4_gcc-7.3.0_glibc/lib/gcc/i486-openwrt-linux-gnu/7.3.0/../../../../i486-openwrt-linux-gnu/bin/ld: ../lib/libcom_err.so: undefined reference to `sem_wait'
.../toolchain-i386_pentium4_gcc-7.3.0_glibc/lib/gcc/i486-openwrt-linux-gnu/7.3.0/../../../../i486-openwrt-linux-gnu/bin/ld: ../lib/libcom_err.so: undefined reference to `sem_init’
.../toolchain-i386_pentium4_gcc-7.3.0_glibc/lib/gcc/i486-openwrt-linux-gnu/7.3.0/../../../../i486-openwrt-linux-gnu/bin/ld: ../lib/libcom_err.so: undefined reference to `sem_destroy’

Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
6 years agoRevert "kernel: add a RPS balancer"
Stijn Tintel [Sun, 7 Oct 2018 10:25:09 +0000 (13:25 +0300)]
Revert "kernel: add a RPS balancer"

This reverts commit 7af1fb9faafbc842fc727c49108f5fc4edc08601.

With the RPS balancer patch, both my APU2s are crashing, sometimes just
after a few minutes of uptime.

[ 6241.170132] BUG: unable to handle kernel paging request at ffffffffa20a75a8
[ 6241.177248] IP: get_rps_cpu+0x41c/0x440
[ 6241.181140] PGD 2012067 P4D 2012067 PUD 2013063 PMD 0
[ 6241.186370] Oops: 0000 [#1] SMP NOPTI
[ 6241.190080] Modules linked in: pppoe ppp_async pptp pppox ppp_mppe ppp_generic nf_nat_pptp nf_flow_table_ipv6 nf_flow_table_ipv4 nf_flow_table_inet nf_conntrack_pptp lzo iptable_nat ipt_REJECT ipt_MASQUERADE ftdi_sio xt_time xt_tcpudp xt_tcpmss xt_string xt_statistic xt_state xt_recent xt_quota xt_policy xt_pkttype xt_owner xt_nat xt_multiport xt_mark xt_mac xt_limit xt_length xt_hl xt_helper xt_esp xt_ecn xt_dscp xt_conntrack xt_connmark xt_connlimit xt_connbytes xt_comment xt_bpf xt_addrtype xt_TCPMSS xt_REDIRECT xt_NFQUEUE xt_LOG xt_HL xt_FLOWOFFLOAD xt_DSCP xt_CT xt_CLASSIFY wireguard usbserial ts_fsm ts_bm spidev slhc rfcomm nft_set_rbtree nft_set_hash nft_reject_ipv6 nft_reject_ipv4 nft_reject_inet nft_reject nft_redir_ipv4 nft_redir nft_quota nft_numgen nft_nat nft_meta nft_masq_ipv4 nft_masq
[ 6241.261735]  nft_log nft_limit nft_flow_offload nft_exthdr nft_ct nft_counter nft_chain_route_ipv6 nft_chain_route_ipv4 nft_chain_nat_ipv4 nfnetlink_queue nf_tables_ipv6 nf_tables_ipv4 nf_tables_inet nf_tables nf_reject_ipv4 nf_nat_tftp nf_nat_snmp_basic nf_nat_sip nf_nat_redirect nf_nat_proto_gre nf_nat_masquerade_ipv6 nf_nat_masquerade_ipv4 nf_nat_irc nf_conntrack_ipv6 nf_nat_ipv6 nf_conntrack_ipv4 nf_nat_ipv4 nf_nat_h323 nf_nat_ftp nf_nat_amanda nf_nat nf_log_ipv4 nf_flow_table_hw nf_flow_table nf_defrag_ipv6 nf_defrag_ipv4 nf_conntrack_tftp nf_conntrack_snmp nf_conntrack_sip nf_conntrack_rtcache nf_conntrack_proto_gre nf_conntrack_netlink nf_conntrack_irc nf_conntrack_h323 nf_conntrack_ftp nf_conntrack_broadcast ts_kmp nf_conntrack_amanda macvlan lzo_decompress lzo_compress libcrc32c kvm irqbypass
[ 6241.333427]  iptable_raw iptable_mangle iptable_filter ipt_ah ipt_ECN ip_tables hidp hci_uart crc_ccitt cdc_acm btusb btintel br_netfilter bnep bluetooth sch_cake tcp_bbr sch_teql em_nbyte sch_codel sch_prio sch_pie act_ipt em_meta sch_gred sch_dsmark cls_basic em_cmp em_text act_police sch_sfq sch_fq sch_multiq sch_red act_connmark nf_conntrack act_skbedit act_mirred em_u32 cls_u32 cls_tcindex cls_flow cls_route cls_fw sch_tbf sch_htb sch_hfsc sch_ingress configs evdev i2c_piix4 kfifo_buf industrialio i2c_dev xt_set ip_set_list_set ip_set_hash_netiface ip_set_hash_netport ip_set_hash_netnet ip_set_hash_net ip_set_hash_netportnet ip_set_hash_mac ip_set_hash_ipportnet ip_set_hash_ipportip ip_set_hash_ipport ip_set_hash_ipmark ip_set_hash_ip ip_set_bitmap_port ip_set_bitmap_ipmac ip_set_bitmap_ip ip_set
[ 6241.405252]  nfnetlink ip6t_REJECT nf_reject_ipv6 nf_log_ipv6 nf_log_common ip6table_mangle ip6table_filter ip6_tables x_tables ip_gre gre igb i2c_algo_bit ifb ip6_vti ip_vti sit l2tp_netlink l2tp_core ipcomp6 xfrm6_tunnel xfrm6_mode_tunnel xfrm6_mode_transport xfrm6_mode_beet esp6 ah6 ipcomp xfrm4_tunnel xfrm4_mode_tunnel xfrm4_mode_transport xfrm4_mode_beet esp4 ah4 ip6_tunnel tunnel6 tunnel4 ip_tunnel veth snd_compress snd_pcm_oss snd_mixer_oss snd_pcm snd_timer snd_rawmidi snd_seq_device snd_hwdep snd soundcore mpls_gso mpls_iptunnel mpls_router af_key xfrm_user xfrm_ipcomp xfrm_algo br2684 atm regmap_mmio vxlan udp_tunnel ip6_udp_tunnel ecdh_generic sha512_ssse3 sha512_generic sha256_ssse3 sha1_ssse3 seqiv jitterentropy_rng drbg md5 kpp ccp_crypto rsa_generic mpi asn1_decoder akcipher ccp sha256_generic
[ 6241.477726]  sha1_generic hmac ghash_generic gcm echainiv des_generic deflate zlib_deflate ctr cmac ccm cbc authenc crypto_acompress sdhci_pltfm pf_ring sp5100_tco leds_apu2 gpio_nct5104d button_hotplug ptp pps_core
[ 6241.497122] CPU: 3 PID: 0 Comm: swapper/3 Not tainted 4.14.73 #0
[ 6241.503198] Hardware name: PC Engines apu2/apu2, BIOS v4.8.0.2 20180705
[ 6241.509858] task: ffff88012a0d8000 task.stack: ffffc90000070000
[ 6241.515841] RIP: 0010:get_rps_cpu+0x41c/0x440
[ 6241.520246] RSP: 0018:ffff88012ed83db0 EFLAGS: 00010286
[ 6241.525511] RAX: 00000000ffffffff RBX: 0000000000011ae8 RCX: 0000000000000001
[ 6241.532737] RDX: 00000000ffffffff RSI: ffff88012a0d8788 RDI: 0000000000000282
[ 6241.539956] RBP: ffff88012ed83e00 R08: 0000000000000001 R09: 0000000000000000
[ 6241.547183] R10: 0000000000000000 R11: 0000000000000000 R12: ffffffff82dae2e0
[ 6241.554403] R13: ffff880124de4480 R14: 0000000000000000 R15: ffff880128120000
[ 6241.561625] FS:  0000000000000000(0000) GS:ffff88012ed80000(0000) knlGS:0000000000000000
[ 6241.569820] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 6241.575651] CR2: ffffffffa20a75a8 CR3: 00000001251da000 CR4: 00000000000406e0
[ 6241.582830] Call Trace:
[ 6241.585322]  <IRQ>
[ 6241.587372]  ? lock_acquire+0x59/0x80
[ 6241.591102]  netif_receive_skb_internal+0x1e4/0x2d0
[ 6241.596037]  napi_gro_receive+0x48/0x90
[ 6241.599948]  igb_alloc_rx_buffers+0xc97/0x1b60 [igb]
[ 6241.604981]  ? note_gp_changes+0x76/0xc0
[ 6241.608963]  net_rx_action+0x10c/0x280
[ 6241.612752]  __do_softirq+0xf0/0x22d
[ 6241.616375]  irq_exit+0x5e/0xa0
[ 6241.619573]  do_IRQ+0xee/0x110
[ 6241.622682]  common_interrupt+0x8b/0x8b
[ 6241.626556]  </IRQ>
[ 6241.628699] RIP: 0010:native_safe_halt+0x6/0x10
[ 6241.633260] RSP: 0018:ffffc90000073e10 EFLAGS: 00000202 ORIG_RAX: ffffffffffffff3d
[ 6241.640910] RAX: ffff88012a0d8000 RBX: ffff88012a190c00 RCX: 0000000000000000
[ 6241.648110] RDX: ffff88012a0d8000 RSI: 0000000000000001 RDI: ffff88012a0d8000
[ 6241.655354] RBP: ffffc90000073e10 R08: 0000000000000006 R09: 0000000000000000
[ 6241.662554] R10: 0000000000000000 R11: 0000000000000000 R12: ffff88012a190c64
[ 6241.669791] R13: 0000000000000001 R14: 0000000000000001 R15: 0000000000000001
[ 6241.676986]  acpi_safe_halt.part.9+0xe/0x20
[ 6241.681226]  acpi_idle_do_entry+0x2c/0x40
[ 6241.685325]  acpi_idle_enter+0x1ee/0x2b0
[ 6241.689298]  ? sched_clock+0x9/0x10
[ 6241.692839]  cpuidle_enter_state+0x1f2/0x230
[ 6241.697162]  cpuidle_enter+0x12/0x20
[ 6241.700803]  call_cpuidle+0x38/0x40
[ 6241.704335]  do_idle+0xed/0x160
[ 6241.707557]  cpu_startup_entry+0x6e/0x70
[ 6241.711557]  start_secondary+0x1b4/0x1d0
[ 6241.715536]  secondary_startup_64+0xa5/0xb0
[ 6241.719777] Code: d4 48 8d 04 9b 48 8b 35 53 a7 ac 00 48 8d 1c 43 48 83 c6 64 48 c1 e3 03 48 8d bb e8 e2 da 82 e8 cb a9 ba ff 41 8b 14 1c 48 89 d0 <48> 0f a3 15 dc dc b6 00 0f 83 46 fc ff ff 48 83 c4 28 5b 41 5c
[ 6241.739100] RIP: get_rps_cpu+0x41c/0x440 RSP: ffff88012ed83db0
[ 6241.744993] CR2: ffffffffa20a75a8
[ 6241.748373] ---[ end trace 77367d9f9830d5bc ]---

Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
6 years agosunxi: add support for Orange Pi Zero
Julian Haupt [Sun, 30 Sep 2018 14:42:03 +0000 (14:42 +0000)]
sunxi: add support for Orange Pi Zero

Signed-off-by: Julian Haupt <julian.haupt@hauptmedia.de>
6 years agoimage: ignore usign build errors
Hauke Mehrtens [Tue, 2 Oct 2018 20:43:58 +0000 (22:43 +0200)]
image: ignore usign build errors

The tl-wa850re-v2 images from the ar71xx/tiny target are getting too big
with the default packages. The size check is done before the meta data
is added so there is no file to add meta data to or to sign. Originally
errors in Build/append-metadata were getting ignored, but if the signing
fails the error is not ignored.
This adds a check if the file to be signed is there and only does the
signing if it is there. This way it does not fail if the package
creation was already aborted earlier.

Fixes: 848b455d2e94 ("image: use ucert to append signature")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
6 years agofstools: filter unknown action in mount.hotplug script
Rosy Song [Sun, 23 Sep 2018 01:57:29 +0000 (09:57 +0800)]
fstools: filter unknown action in mount.hotplug script

Signed-off-by: Rosy Song <rosysong@rosinson.com>
6 years agobusybox: fix IPv6 dependency
Hans Dedecker [Sun, 7 Oct 2018 18:37:47 +0000 (20:37 +0200)]
busybox: fix IPv6 dependency

Commit 9f0cb135dd made BUSYBOX_CONFIG_FEATURE_IPV6 dependant on IPV6 but
did not make its default value BUSYBOX_DEFAULT_FEATURE_IPV6 dependant
on IPV6. BUSYBOX_DEFAULT_FEATURE_IPV6 will have as default value y if
IPV6 is enabled otherwise n.

Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
6 years agodnsmasq: bump to v2.80test8
Kevin Darbyshire-Bryant [Sun, 7 Oct 2018 07:25:08 +0000 (08:25 +0100)]
dnsmasq: bump to v2.80test8

e1791f3 Fix logging of DNSSEC queries in TCP mode. Destination server address was misleading.
0fdf3c1 Fix dhcp-match-name to match hostname, not complete FQDN.
ee1df06 Tweak strategy for confirming SLAAC addresses.
1e87eba Clarify manpage for --auth-sec-servers
0893347 Make interface spec optional in --auth-server.
7cbf497 Example config file fix for CERT Vulnerability VU#598349.

Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
6 years agonghttp2: bump to 1.34.0
Hans Dedecker [Sun, 7 Oct 2018 15:35:06 +0000 (17:35 +0200)]
nghttp2: bump to 1.34.0

2b085815 (tag: v1.34.0) Update manual pages
986fa302 Bump up version number to 1.34.0, LT revision to 31:1:17
7c8cb3a0 nghttpx: Improve CONNECT response status handling
334c439c Fix bug that regular CONNECT does not work
6700626c Rule out content-length in the successful response to CONNECT
15162add Update manual pages
93270777 Merge pull request #1235 from nghttp2/backend-conn-timeout
aeb92bbb nghttpx: Add read/write-timeout parameters to backend option
fc7489e0 nghttpx: Fix mruby parameter validation
87ac872f nghttpx: Update doc
c278adde nghttpx: Log error when mruby file cannot be opened
f94d7209 Merge pull request #1234 from nghttp2/nghttpx-rfc8441
9b9baa6b Update doc
02566ee3 nghttpx: Update doc
3002f31b src: Add debug output for SETTINGS_ENABLE_CONNECT_PROTOCOL
d2a594a7 nghttpx: Implement RFC 8441 Bootstrapping WebSocket with HTTP/2
651e1477 Allow client sending :protocol optimistically
a42faf1c nghttpx: Write TLS alert during handshake
4aac05e1 Merge pull request #1231 from nghttp2/ws-lib-only
b80dfaa8 Adjustment for RFC 8441
a19d8f5d Deal with :protocol pseudo header
33f6e90a Add NGHTTP2_TOKEN__PROTOCOL
ed7fabcb Add SETTINGS_ENABLE_CONNECT_PROTOCOL
8753b6da Update doc
f2de733b Update neverbleed to fix OpenSSL 1.1.1 issues
88ff8c69 Update mruby 1.4.1
a63558a1 nghttpx: Call OCSP_response_get1_basic only when OCSP status is successful
3575a132 nghttpx: Fix crash with plain text HTTP
e2de2fee Update bash_completion
9f415979 Update manual pages
4bfc0cd1 Merge pull request #1230 from nghttp2/nghttpx-faster-logging
9c824b87 nghttpx: Get rid of std::stringstream from Log
a1ea1696 Make VALID_HD_NAME_CHARS and VALID_HD_VALUE_CHARS const qualified
dfc0f248 Make static_table const qualified
ed7c9db2 nghttpx: Add mruby env.tls_handshake_finished
5b42815a nghttpx: Strip incoming Early-Data header field by default
cfe7fa9a nghttpx: Add --tls13-ciphers and --tls-client-ciphers options
cb8a9d58 src: Remove TLSv1.3 ciphers from DEFAULT_CIPHER_LIST
023b9448 Merge branch 'tls13-early-data'
9b03c64f nghttpx: Should postpone early data by default
b8eccec6 nghttpx: Disable OpenSSL anti-replay
9f212587 Specify SSL_CTX_set_max_early_data and add an option to change max value
47f60124 nghttpx: Add an option to postpone early data processing
770e44de Implement draft-ietf-httpbis-replay-02
2ab319c1 Don't hide error code from openssl
39923024 Remove SSL_ERROR_WANT_WRITE handling
b30f312a Honor SSL_read semantics
c5cdb78a nghttpx: Add TLSv1.3 0-RTT early data support
f79a5812 Bump up version number to 1.34.0

Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
6 years agoath79: add Fritz!Box 4020 switchconfig
David Bauer [Sun, 7 Oct 2018 15:09:48 +0000 (17:09 +0200)]
ath79: add Fritz!Box 4020 switchconfig

In contrast to ar71xx, uci switchconfig is missing in the ath79 target.

Signed-off-by: David Bauer <mail@david-bauer.net>
6 years agoiperf: fix --daemon option
Rafał Miłecki [Sun, 7 Oct 2018 12:17:50 +0000 (14:17 +0200)]
iperf: fix --daemon option

Support for -D got broken in the 2.0.11 release by the upstream commit
218d8c667944 ("first pass L2 mode w/UDP checks, v4 only"). After that
commit clients were still able to connect but no traffic was passed.
It was reported and is fixed now in the upstream git repository.

Backport two patches to fix this. The first one is just a requirement
for the later to apply. The second one is the real fix and it needed
only a small adjustment to apply without backporing the commit
10887b59c7e7 ("fix --txstart-time report messages").

Fixes: 457e6d5a27be ("iperf: bump to 2.0.12")
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
6 years agoath79: add support for TP-Link TL-WA901ND v2
David Bauer [Sun, 7 Oct 2018 13:13:31 +0000 (15:13 +0200)]
ath79: add support for TP-Link TL-WA901ND v2

This commit adds support for the TP-Link TL-WR901ND v2 access point.

CPU:   Atheros AR9132 400MHz
RAM:   32MB
FLASH: 4MiB
WiFi:  Atheros AR9103 3x3:2 bgn
LED:   Power (static on)
       LAN (controlled by PHY)
       SYS, WiFi, QSS toggleable
BTN:   Reset, QSS

Installation:
Upload the factory image via the vendor-GUI.

Signed-off-by: David Bauer <mail@david-bauer.net>
6 years agoath79: add support for Buffalo WZR-HP-G302H A1A0
INAGAKI Hiroshi [Tue, 2 Oct 2018 13:28:45 +0000 (22:28 +0900)]
ath79: add support for Buffalo WZR-HP-G302H A1A0

Buffalo WZR-HP-G302H is a 2T2R 2.4 GHz 11n router, based on Atheros
AR7242.

It is Japanese market model of WZR-HP-G300NH2, but there are some
diffrences. This commit is based on WZR-HP-G300NH2 in ar71xx.
And, G302H has several hardware versions and hardware is different
dependent on the versions. This commit adds support for "A1A0"
version.

Specification:

- Atheros AR7242
- 64 MB of RAM (DDR2)
- 32 MB of Flash
  - 2x 16 MB SPI-NOR flash
- 2.4 GHz 2T2R wifi
  - Atheros AR9283
- 5x 10/100/1000 Mbps Ethernet
  - Atheros AR8316
- 7x LEDs, 5x keys
  - LED: 1x gpio-leds, 6x ath9k-leds
  - key: 3x buttons, 2x slide switches
- UART header on PCB
  - Vcc, GND, TX, RX from ethernet port side
  - 115200n8

Flash instruction using factory image:

1. Boot WZR-HP-G302H normaly and connect the computer to its LAN port
2. Access to "http://192.168.11.1/" and move to firmware update page
("ファーム更新")
3. Select the OpenWrt factory image and click update ("更新実行")
button to perform firmware update
4. Wait ~200 seconds to complete flashing

Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>
6 years agoath79: add hwver parameter to buffalo-tag
INAGAKI Hiroshi [Sun, 7 Oct 2018 11:37:42 +0000 (20:37 +0900)]
ath79: add hwver parameter to buffalo-tag

In order to be able to set the value of "hardware version" other than
"3", I added the "hwver" parameter.

Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>
6 years agoath79: Add support for TP-Link TL-WR842N-v2
Robert Marko [Sun, 7 Oct 2018 10:21:45 +0000 (12:21 +0200)]
ath79: Add support for TP-Link TL-WR842N-v2

This PR adds support for TP-Link TL-WR842N-v2 router which is supported by ar71xx to ath79.

This is a low cost model with following specs:

CPU: Atheros AR9341 SoC
RAM: 32 MB DDR1
Flash: 8 MB NOR SPI
Switch: Internal AR9341 5 port 10/100 Mbit
Ports:  5x 10/100 Mbit(1x WAN, 4x LAN)
USB: 1x USB2.0
WLAN: 2.4 GHZ AR9341

Installation:

Simply flash the factory image through stock firmware WEB UI.

Signed-off-by: Robert Marko <robimarko@gmail.com>
6 years agoipq40xx: Create /var/lock directory in platform_do_upgrade_*
Sven Eckelmann [Mon, 1 Oct 2018 09:48:04 +0000 (11:48 +0200)]
ipq40xx: Create /var/lock directory in platform_do_upgrade_*

The sysupgrade_pre_upgrade hook was removed with 5e1b4c57ded7 ("base-files:
drop fwtool_pre_upgrade") while there were still scripts using it:

* target/linux/ar71xx/base-files/lib/upgrade/allnet.sh
* target/linux/ar71xx/base-files/lib/upgrade/openmesh.sh
* target/linux/ipq40xx/base-files/lib/upgrade/openmesh.sh

Not running the hooks can either prevent a successful upgrade or brick the
device because the fw_setenv program cannot be started correctly.

Instead of adding this hook again, the directory /var/lock for fw_setenv
can also just be created directly before fw_setenv is called.

Fixes: 5e1b4c57ded7 ("base-files: drop fwtool_pre_upgrade")
Signed-off-by: Sven Eckelmann <sven.eckelmann@openmesh.com>
6 years agoar71xx: Create /var/lock directory in platform_do_upgrade_*
Sven Eckelmann [Mon, 1 Oct 2018 09:48:04 +0000 (11:48 +0200)]
ar71xx: Create /var/lock directory in platform_do_upgrade_*

The sysupgrade_pre_upgrade hook was removed with 5e1b4c57ded7 ("base-files:
drop fwtool_pre_upgrade") while there were still scripts using it:

* target/linux/ar71xx/base-files/lib/upgrade/allnet.sh
* target/linux/ar71xx/base-files/lib/upgrade/openmesh.sh
* target/linux/ipq40xx/base-files/lib/upgrade/openmesh.sh

Not running the hooks can either prevent a successful upgrade or brick the
device because the fw_setenv program cannot be started correctly.

Instead of adding this hook again, the directory /var/lock for fw_setenv
can also just be created directly before fw_setenv is called.

Fixes: 5e1b4c57ded7 ("base-files: drop fwtool_pre_upgrade")
Signed-off-by: Sven Eckelmann <sven.eckelmann@openmesh.com>
6 years agoar71xx: Use sysupgrade's RAMFS_COPY_* for fw_(set|print)env
Sven Eckelmann [Mon, 1 Oct 2018 10:27:25 +0000 (12:27 +0200)]
ar71xx: Use sysupgrade's RAMFS_COPY_* for fw_(set|print)env

The install_bin from /lib/upgrade/common.sh is no longer creating the
symlinks when a secondary parameter is added. But the fw_setenv program was
always copied this way to the ramdisk for the upgrade.

Instead, this should be done using RAMFS_COPY_* like on all other
platforms.

Fixes: 438dcbfe74a6 ("base-files: automatically handle paths and symlinks for RAMFS_COPY_BIN")
Signed-off-by: Sven Eckelmann <sven.eckelmann@openmesh.com>
6 years agoath79: fix build warning
Mathias Kresin [Sun, 7 Oct 2018 13:33:35 +0000 (15:33 +0200)]
ath79: fix build warning

The image build code for the Ubiquiti Nanostation AC series adds the
factory image as to be build image. The same is already done by an
included recipe which results into an expanded IMAGE variable of:

  IMAGES = sysupgrade.bin factory.bin factory.bin

The build system doesn't like these duplicates and issues the following
warning:

  Makefile:82: warning: overriding recipe for target...

Get remove the duplicate factory image to get rid of the warning.

Fixes: 5736af8024be ("ath79: Add support for Ubiquiti NanoStation AC loco")
       fa3c2676ab0c ("ath79: Add support for Ubiquiti Nanostation AC")

Signed-off-by: Mathias Kresin <dev@kresin.me>
6 years agoodhcpd: update to latest git HEAD (FS#1853)
Hans Dedecker [Sun, 7 Oct 2018 13:04:49 +0000 (15:04 +0200)]
odhcpd: update to latest git HEAD (FS#1853)

57f639e (HEAD -> master, origin/master, origin/HEAD) odhcpd: make DHCPv6/RA/NDP support optional
402c274 dhcpv6: check return code of dhcpv6_ia_init()
ee7472a router: don't leak RA message in relay mode (FS#1853)

Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
6 years agoiw: strip a few more non-essential features from iw-tiny
Felix Fietkau [Sun, 7 Oct 2018 08:48:31 +0000 (10:48 +0200)]
iw: strip a few more non-essential features from iw-tiny

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agoiw: fix filtering linked object files for iw-tiny
Felix Fietkau [Sun, 7 Oct 2018 08:35:08 +0000 (10:35 +0200)]
iw: fix filtering linked object files for iw-tiny

It was broken by the recent commit that added iw-full

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agoramips: don't hijack the status led
Mathias Kresin [Wed, 29 Aug 2018 06:58:33 +0000 (08:58 +0200)]
ramips: don't hijack the status led

Don't hijack the status led to indicate the wireless state. If we don't
have a dedicated wireless led, it's as simply as the wireless status
can't be indicated.

Such a led misuse should be set by the user and not shipped by default.

Signed-off-by: Mathias Kresin <dev@kresin.me>
6 years agoramips: move status led unset to devicetree
Mathias Kresin [Tue, 28 Aug 2018 05:04:40 +0000 (07:04 +0200)]
ramips: move status led unset to devicetree

Release the led used for boot status indication via devicetree instead
of setting a default off trigger in userspace.

Signed-off-by: Mathias Kresin <dev@kresin.me>
6 years agoramips: add support for indicating the boot state using multiple leds
Mathias Kresin [Tue, 28 Aug 2018 04:54:27 +0000 (06:54 +0200)]
ramips: add support for indicating the boot state using multiple leds

Use diag.sh version used for other targets supporting different leds
for the different boot states.

The existing led sequences should be the same as before.

Signed-off-by: Mathias Kresin <dev@kresin.me>
6 years agoramips: set F5D8235 v1 usb led trigger via devicetree
Mathias Kresin [Fri, 17 Aug 2018 14:41:17 +0000 (16:41 +0200)]
ramips: set F5D8235 v1 usb led trigger via devicetree

Assign the usbdev trigger via devicetree and drop the userspace
handling of the usb leds.

Signed-off-by: Mathias Kresin <dev@kresin.me>
6 years agoramips: set rt2880 pci controller of_node
Mathias Kresin [Tue, 28 Aug 2018 21:35:33 +0000 (23:35 +0200)]
ramips: set rt2880 pci controller of_node

Set the PCI controller of_node such that PCI devices can be
instantiated via device tree.

Signed-off-by: Mathias Kresin <dev@kresin.me>
6 years agoramips: set usb led trigger via devicetree
Mathias Kresin [Mon, 13 Aug 2018 15:14:08 +0000 (17:14 +0200)]
ramips: set usb led trigger via devicetree

Assign the usbdev trigger via devicetree for all subtargets and drop
the userspace handling of the usb leds.

With the change all usb ports are triggering the usb led instead of
only usb 1.1 XOR usb 2.0 XOR usb 3.0 as it was before.

Signed-off-by: Mathias Kresin <dev@kresin.me>
6 years agoath79: Add support for Ubiquiti Nanostation AC
Tobias Schramm [Tue, 4 Sep 2018 00:08:01 +0000 (02:08 +0200)]
ath79: Add support for Ubiquiti Nanostation AC

5 GHz AC wireless outdoor PoE CPE with internal 2.4 GHz management radio

CPU:    Atheros AR9342 SoC
RAM:    64 MB DDR2
Flash:  16 MB NOR SPI
Switch: QCA8334
Ports:  2 GbE ports (1x PoE in, 1x PoE passthrough)
WLAN:   5 GHz QCA899X (PCI) and 2.4 GHZ AR9342

Successor to the old NanoStation M5 with AC wireless.

The integrated QCA899X is a Ubiquiti branded part with modified vendor and
product id (0777:11ac9).

Serial

Serial settings: 115200, 8N1

* = plated through hole
0 = nylon screw

      [Top of device]
+--------------------------+
|    [label]               |
|  0                       |
|                    0     |
|         [ubnt]           |
|         [logo]  3V3  *   |
|                  TX  *   |
|                  RX  *   |
|                 GND  *   |
|                          |
|                      *   |
|                      *   |
|                      *   |
|                      *   |
|    0                     |
|                 0        |
|                          |
|                          |

Installation

1. Connect to serial header on device
2. Power on device and enter uboot console
3. Set up tftp server serving an openwrt initramfs build
4. Load initramfs build using the command tftpboot in the uboot cli
5. Boot the loaded image using the command bootm
6. Copy squashfs openwrt sysupgrade build to the booted device
7. Use mtd to write sysupgrade to partition "firmware"
8. Reboot and enjoy

Signed-off-by: Tobias Schramm <tobleminer@gmail.com>
6 years agoath79: Add support for Ubiquiti NanoStation AC loco
Tobias Schramm [Tue, 23 Jan 2018 10:36:41 +0000 (11:36 +0100)]
ath79: Add support for Ubiquiti NanoStation AC loco

Atheros AR9342, 16 MB flash, 64 MB RAM

Successor to the old NanoStation M5 loco with AC wireless.

Includes a mac80211 patch for ath10k_pci because Ubiquiti uses a Ubiquiti
branded and customized QCA988X with vendor id 0777 and device id 11ac for
AC wireless.

Installation

1. Connect to serial header on device (8N1 115200)
2. Power on device and enter uboot console
3. Set up tftp server serving an openwrt initramfs build
4. Load initramfs build using the command tftpboot in the uboot cli
5. Boot the loaded image using the command bootm
6. Copy squashfs openwrt sysupgrade build to the booted device
7. Use mtd to write sysupgrade to partition "firmware"
8. Reboot and enjoy

Signed-off-by: Tobias Schramm <tobleminer@gmail.com>
6 years agoath79: Add generic device tree for Ubiquiti WA boards
Tobias Schramm [Tue, 4 Sep 2018 13:05:47 +0000 (15:05 +0200)]
ath79: Add generic device tree for Ubiquiti WA boards

Signed-off-by: Tobias Schramm <tobleminer@gmail.com>
6 years agomkfwimage: Add image type definition for WA images
Tobias Schramm [Wed, 24 Jan 2018 00:38:14 +0000 (01:38 +0100)]
mkfwimage: Add image type definition for WA images

This patch adds a new type of ubiquiti image, the WA image. First seen
on the NanoStation AC loco the generic name implies that we will see
this type of image on more ubiquiti devices thus it makes sense to
implement it in mkfwimage.

The main difference is that WA images are signed. The "END" header has
been replaced by a "ENDS" header followed by a 2048 bit RSA signature.
This signature is not being generated by mkfwimage and filled with 0x00.

Signed-off-by: Tobias Schramm <tobleminer@gmail.com>
6 years agoath79: fix MAC address for Buffalo BHR-4GRV
INAGAKI Hiroshi [Thu, 27 Sep 2018 03:47:43 +0000 (12:47 +0900)]
ath79: fix MAC address for Buffalo BHR-4GRV

I added mtd-mac-address for WZR-HP-G450H and BHR-4GRV in
1df1ea4d7e8b6ae3351780ed58800ccd9edd4c27, but that address in ART is
incorrect for BHR-4GRV.

WZR-HP-G450H has wlan eeprom and MAC address in ART, but BHR-4GRV
has only MAC address in ART.

- WZR-HP-G450H
  - eeprom: 0x1000
  - MAC:    0x1002

- BHR-4GRV
  - MAC:    0x0

Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>
6 years agointel-microcode: update to version 20180807a
Hauke Mehrtens [Sat, 6 Oct 2018 13:49:20 +0000 (15:49 +0200)]
intel-microcode: update to version 20180807a

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
6 years agouboot-lantiq: fix compile with modern host dtc
Thomas Nixon [Tue, 4 Sep 2018 20:54:58 +0000 (21:54 +0100)]
uboot-lantiq: fix compile with modern host dtc

In dtc version 1.4.6 the macro names in header include guards changed,
but the build relies on them matching in order to replace selected
headers. This is a horrible hack to work around this.

Signed-off-by: Thomas Nixon <tom@tomn.co.uk>
6 years agouboot-envtools: adds r7800 uboot env support
Ansuel Smith [Sun, 30 Sep 2018 17:21:39 +0000 (19:21 +0200)]
uboot-envtools: adds r7800 uboot env support

This fix and adds support for r7800 uboot env

Signed-off-by: Ansuel Smith <ansuelsmth@gmail.com>
6 years agobcm53xx: added basic dts for linksys ea6500v2
Rene Kjellerup [Sun, 22 Oct 2017 00:19:03 +0000 (17:19 -0700)]
bcm53xx: added basic dts for linksys ea6500v2

only thing not working is the b43 5GHz wifi band as upstream
kernel
doesn't supporthe 0x4360 chip so far

Signed-off-by: Rene Kjellerup <rk.katana.steel@gmail.com>
6 years agomac80211: Use @KERNEL alias instead of hardlink
Daniel Engberg [Fri, 5 Oct 2018 06:31:56 +0000 (08:31 +0200)]
mac80211: Use @KERNEL alias instead of hardlink

Use @KERNEL alias provided by toolchain instead of relying on a specific site.

Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
6 years agobinutils: Use 2.31.1 by default
Daniel Engberg [Fri, 17 Aug 2018 21:40:29 +0000 (23:40 +0200)]
binutils: Use 2.31.1 by default

Set binutils 2.31.1 as default

Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
6 years agotoolchain/binutils: Fix u-boot on some armhf platforms
Daniel Engberg [Sat, 6 Oct 2018 10:55:02 +0000 (12:55 +0200)]
toolchain/binutils: Fix u-boot on some armhf platforms

Backport commit 3d9c8f6b3f033a6092425b7344647fb51dbed5c6
Without this binutils doesn't properly link u-boot

Source:
https://sourceware.org/bugzilla/show_bug.cgi?id=23571

Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
6 years agotools/cmake: Update to 3.12.3
Daniel Engberg [Sat, 6 Oct 2018 10:37:22 +0000 (12:37 +0200)]
tools/cmake: Update to 3.12.3

Update CMake to 3.12.3

Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
6 years agotools/bison: Update to 3.1
Daniel Engberg [Sat, 6 Oct 2018 10:41:52 +0000 (12:41 +0200)]
tools/bison: Update to 3.1

Update bison to 3.1

Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
6 years agotoolchain/fortify-headers: Update to 1.0
Daniel Engberg [Sat, 6 Oct 2018 10:39:51 +0000 (12:39 +0200)]
toolchain/fortify-headers: Update to 1.0

Update fortify-headers to 1.0

Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
6 years agotools/libressl: Update to 2.8.1
Daniel Engberg [Sat, 6 Oct 2018 10:44:04 +0000 (12:44 +0200)]
tools/libressl: Update to 2.8.1

Update libressl to 2.8.1

Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
6 years agotools/mtools: Update to 4.0.19
Daniel Engberg [Sat, 6 Oct 2018 10:47:46 +0000 (12:47 +0200)]
tools/mtools: Update to 4.0.19

Update mtools to 4.0.19

Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
6 years agoubus: update to latest git HEAD
Hans Dedecker [Sat, 6 Oct 2018 19:23:53 +0000 (21:23 +0200)]
ubus: update to latest git HEAD

221ce7e ubusd_acl: event send access list support
da503db ubusd_acl: event listen access list support
c035bab ubusd_acl: rework wildcard support
73bd847 ubusd_event: move strmatch_len to ubus_common.h
0327a91 ubus/lua: add support for BLOBMSG_TYPE_DOUBLE

Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
6 years agowireguard: bump to 0.0.20181006
Jason A. Donenfeld [Sat, 6 Oct 2018 03:09:00 +0000 (05:09 +0200)]
wireguard: bump to 0.0.20181006

* Account for big-endian 2^26 conversion in Poly1305.
  * Account for big-endian NEON in Curve25519.
  * Fix macros in big-endian AArch64 code so that this will actually run there
    at all.
  * Prefer if (IS_ENABLED(...)) over ifdef mazes when possible.
  * Call simd_relax() within any preempt-disabling glue code every once in a
    while so as not to increase latency if folks pass in super long buffers.
  * Prefer compiler-defined architecture macros in assembly code, which puts us
    in closer alignment with upstream CRYPTOGAMS code, and is cleaner.
  * Non-static symbols are prefixed with wg_ to avoid polluting the global
    namespace.
  * Return a bool from simd_relax() indicating whether or not we were
    rescheduled.
  * Reflect the proper simd conditions on arm.
  * Do not reorder lines in Kbuild files for the simd asm-generic addition,
    since we don't want to cause merge conflicts.
  * WARN() if the selftests fail in Zinc, since if this is an initcall, it won't
    block module loading, so we want to be loud.
  * Document some interdependencies beside include statements.
  * Add missing static statement to fpu init functions.
  * Use union in chacha to access state words as a flat matrix, instead of
    casting a struct to a u8 and hoping all goes well. Then, by passing around
    that array as a struct for as long as possible, we can update counter[0]
    instead of state[12] in the generic blocks, which makes it clearer what's
    happening.
  * Remove __aligned(32) for chacha20_ctx since we no longer use vmovdqa on x86,
    and the other implementations do not require that kind of alignment either.
  * Submit patch to ARM tree for adjusting RiscPC's cflags to be -march=armv3 so
    that we can build code that uses umull.
  * Allow CONFIG_ARM[64] to imply [!]CONFIG_64BIT, and use zinc arch config
    variables consistently throughout.
  * Document rationale for the 2^26->2^64/32 conversion in code comments.
  * Convert all of remaining BUG_ON to WARN_ON.
  * Replace `bxeq lr` with `reteq lr` in ARM assembler to be compatible with old
    ISAs via the macro in <asm/assembler.h>.
  * Do not allow WireGuard to be a built-in if IPv6 is a module.
  * Writeback the base register and reorder multiplications in the NEON x25519
    implementation.
  * Try all combinations of different implementations in selftests, so that
    potential bugs are more immediately unearthed.
  * Self tests and SIMD glue code work with #include, which lets the compiler
    optimize these. Previously these files were .h, because they were included,
    but a simple grep of the kernel tree shows 259 other files that carry out
    this same pattern. Only they prefer to instead name the files with a .c
    instead of a .h, so we now follow the convention.
  * Support many more platforms in QEMU, especially big endian ones.
  * Kernels < 3.17 don't have read_cpuid_part, so fix building there.

Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
6 years agomt76: fix dependencies
Felix Fietkau [Sat, 6 Oct 2018 11:08:49 +0000 (13:08 +0200)]
mt76: fix dependencies

- make kmod-mt76-core hidden
- select kmod-mt76-core and kmod-usb-core from mt76-usb

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agomt76: update to the latest version
Felix Fietkau [Tue, 2 Oct 2018 11:00:39 +0000 (13:00 +0200)]
mt76: update to the latest version

053de5c mt76x0: pci: add mt76x0e_cleanup routine
c6b1926 mt76x2: move mt76x2 source files to mt76x2 folder
6afb1b8 mt76: fix header guard macro define names
05c7f24 mt76: fix frag length allocation for usb
6856ef6 mt76: usb: fix hw initialization sequence
3d43146 mt76x0: usb: stop cal/mac workqueues at hw stop
bef08cc mt76: move mt76x02_tx_get_max_txpwr_adj in mt76x02_util.c
cd42111 mt76: add get_tx_txpwr_adj function pointer to mt76_driver_ops
c49a3ae mt76: move mt76x02_mac_write_txwi in mt76x02-lib module
4bae980 mt76: usb: use mt76x02u_tx_prepare_skb to fill txwi
e0847b4 mt76x0: init: remove unnecessary configurations
e802b5c mt76: move mt76x02_phy_get_min_avg_rssi in mt76x02_phy.c
2742847 mt76: move mt76x02_rx_get_sta and mt76x02_rx_get_sta_wcid in mt76x02_util.h
c26f9fd mt76x0: mac: use sta ewma estimation for rssi tracking
330fbe3 mt76x0: remove unused variable in mt76x0_dev
a56efc5 mt76x0: remove hw_atomic_mutex mutex in mt76x0_dev
abd1daa mt76x2: move mt76x2_dev in mt76x02_util.h
d5d1e59 mt76x0: merge mt76x0_dev in mt76x02_dev
e236ba5 mt76: move mt76x02_mac_process_rx in mt76x02-lib module
724d547 mt76: unify rxwi parsing between mt76x2 and mt76x0 drivers
ec3d829 mt76: move mt76x02_tx in mt76x02-lib module
159f767 mt76: move txrx shared routines in mt76x02_txrx.c
9c874cd mt76: rename mt76x02_util.h in mt76x02.h
fd9174f mt76x2: remove leftover function declatarions
608708a mt76: move tpc routines in mt76x02-lib module
45b8f5e mt76: move mt76x02_tx_prepare_skb in mt76x02_txrx.c
9a39521 mt76: usb: move mt76x02u_tx_complete_skb in mt76x02_usb_core.c
3770c54 mt76: move mt76x02_mac_poll_tx_status in mt76x02-lib moudle
9730290 mt76: move mt76x02_tx_complete in mt76x02-lib module
99e0486 mt76: use mt76x02_dev instead of mt76_dev in mt76x02_mmio.c
3d2cfe3 mt76: move tx_tasklet management in mt76x02-lib moudle
e61671e mt76: move irq handler in mt76x02-lib moudle
d533f13 mt76x0: pci: report firmware version using ethtool

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agoiw: add iw-full package without size reduction hacks
Felix Fietkau [Wed, 3 Oct 2018 06:57:00 +0000 (08:57 +0200)]
iw: add iw-full package without size reduction hacks

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agomac80211: move marvell packaging code to marvell.mk
Felix Fietkau [Sat, 29 Sep 2018 18:13:55 +0000 (20:13 +0200)]
mac80211: move marvell packaging code to marvell.mk

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agomac80211: move intel packaging code to intel.mk
Felix Fietkau [Sat, 29 Sep 2018 18:10:23 +0000 (20:10 +0200)]
mac80211: move intel packaging code to intel.mk

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agomac80211: move realtek packaging code to realtek.mk
Felix Fietkau [Sat, 29 Sep 2018 18:07:06 +0000 (20:07 +0200)]
mac80211: move realtek packaging code to realtek.mk

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agomac80211: move broadcom packaging code to broadcom.mk
Felix Fietkau [Sat, 29 Sep 2018 18:02:28 +0000 (20:02 +0200)]
mac80211: move broadcom packaging code to broadcom.mk

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agomac80211: move atheros/qca packaging code to ath.mk
Felix Fietkau [Sat, 29 Sep 2018 17:56:26 +0000 (19:56 +0200)]
mac80211: move atheros/qca packaging code to ath.mk

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agomac80211: move rt2x00 packaging code to ralink.mk
Felix Fietkau [Sat, 29 Sep 2018 17:47:41 +0000 (19:47 +0200)]
mac80211: move rt2x00 packaging code to ralink.mk

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agomac80211: remove duplicated lines for building packages, use PKG_DRIVERS instead
Felix Fietkau [Sat, 29 Sep 2018 17:44:16 +0000 (19:44 +0200)]
mac80211: remove duplicated lines for building packages, use PKG_DRIVERS instead

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agobuild: insert blank line after KernelPackage template to allow chaining calls to it
Felix Fietkau [Mon, 1 Oct 2018 10:44:50 +0000 (12:44 +0200)]
build: insert blank line after KernelPackage template to allow chaining calls to it

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agopackage/lldp: don't link against libbsd on !USE_GLIBC builds
Alexander Couzens [Tue, 4 Sep 2018 01:10:24 +0000 (03:10 +0200)]
package/lldp: don't link against libbsd on !USE_GLIBC builds

Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
6 years agoar71xx: fix pci irq init on kernel 4.14
Koen Vandeputte [Thu, 4 Oct 2018 13:12:27 +0000 (15:12 +0200)]
ar71xx: fix pci irq init on kernel 4.14

The IRQ init structs are marked as __initconst which
means this memory can be free after init.

On this platform, the PCI IRQ init happens very late _after_ the
kernel already freed the memory allocated for these structs.

During IRQ allocation, the allocation function is passed
with invalid data at this point leading to following error:

[    0.000000] SoC: Qualcomm Atheros QCA9533 ver 2 rev 0
[    2.382828] Freeing unused kernel memory: 264K
[   34.414816] pci 0000:00:00.0: no irq found for pin 1

and

[    0.000000] SoC: Qualcomm Atheros QCA956X ver 1 rev 0
[    2.125401] Freeing unused kernel memory: 284K
[    9.526479] pci 0000:00:00.0: no irq found for pin 1

After this patch:

[   14.960814] pci 0000:00:00.0: using irq 40 for pin 1

Commit 318e19ba6755 ("ar71xx: add v4.14 support") fixed this for the
default targets already present in the source by default but forgot
to remove the __initconst attribute for targets QCA953x and QCA956x
which are only added later through platform patches.

Fixes: 318e19ba6755 ("ar71xx: add v4.14 support")
Reported-by: Sven Schönhoff <sven.schoenhoff@gmail.com>
Reported-by: Dirk Brenken <dev@brenken.org>
Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
Tested-by: Dirk Brenken <dev@brenken.org>
6 years agokernel: bump 4.14 to 4.14.74
Koen Vandeputte [Thu, 4 Oct 2018 09:26:55 +0000 (11:26 +0200)]
kernel: bump 4.14 to 4.14.74

Refreshed all patches.

Fixes CVE:

- CVE-2018-7755

Compile-tested on: cns3xxx, imx6
Runtime-tested on: cns3xxx, imx6

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agokernel: bump 4.9 to 4.9.131
Koen Vandeputte [Thu, 4 Oct 2018 09:26:19 +0000 (11:26 +0200)]
kernel: bump 4.9 to 4.9.131

Refreshed all patches.

Fixes CVE:

- CVE-2018-10880
- CVE-2018-7755

Compile-tested on: ar71xx
Runtime-tested on: ar71xx

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agoipset: update to 6.38
Syrone Wong [Mon, 1 Oct 2018 02:56:04 +0000 (10:56 +0800)]
ipset: update to 6.38

dropped already upstream patch

Signed-off-by: Syrone Wong <wong.syrone@gmail.com>
6 years agosunxi: Turn on CONFIG_SUN4I_A10_CCU for A20
Florian Fainelli [Wed, 3 Oct 2018 03:40:26 +0000 (20:40 -0700)]
sunxi: Turn on CONFIG_SUN4I_A10_CCU for A20

CONFIG_SUN4I_A10_CCU controls both the A10 and the A20 enabling of the
CCU (LCCF) driver, this will be necessary once we move beyond kernel
4.14 because 4.15 has commit f18698e1c66338b902de386e4ad97b8b1b9d999d
("ARM: dts: sun7i: Convert to CCU") which requires this driver.

Fixes: ad2b3bf310f7 ("sunxi: Add support for kernel 4.14")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>