1. 21 Mar, 2020 13 commits
    • Paul Spooren's avatar
      x86: switch image generation to new code · cb007a7b
      Paul Spooren authored
      This commit introduces few related changes which need to be done in
      single commit to keep images buildable between git revisions. In result
      it retains all previous image creation possibilities with slight name
      change of generated images. Brief summary of the commit:
      * Split up image generation recipe to smaller chunks to make it more
        generic and reusable.
      * Make iso images x86 specific and drop their definition as root
      * Convert image creation process to generic code specified in image.mk.
      * Make geode subtarget inherit features from the main target instead of
        redefining them.
      * For subtargets create device definitions with basic packages set.
      Signed-off-by: default avatarTomasz Maciej Nowak <tomek_n@o2.pl>
      Signed-off-by: default avatarPaul Spooren <mail@aparcar.org>
    • Paul Spooren's avatar
      x86/grub2: move grub2 image creation to package · 5a5df62d
      Paul Spooren authored
      Let the grub2 package take care of creating installable grub2 images,
      this will allow creating grub2 images without first calling x86 image
      generation recipe. Also as side effect, since those images are now
      shared, it'll reduce the number of calling grub-mkimage.
      Signed-off-by: default avatarTomasz Maciej Nowak <tomek_n@o2.pl>
      [rebase, adjusted commit title]
      Signed-off-by: default avatarPaul Spooren <mail@aparcar.org>
    • Kevin Darbyshire-Bryant's avatar
      kernel: connmark set-dscpmark fix build on 5.4 · e6e1e12d
      Kevin Darbyshire-Bryant authored
      Fix header change that was done for kernel but 4.19 got missed for 5.4.
      Solves nasty errors like:
                      from connmark_listener.c:30:
      /builder/shared-workdir/build/sdk/staging_dir/toolchain-aarch64_cortex-a53_gcc-8.4.0_musl/include/linux/netfilter/xt_connmark.h:23:2: error: enumerator value for 'XT_CONNMARK_VALUE' is not an integer constant
      /builder/shared-workdir/build/sdk/staging_dir/toolchain-aarch64_cortex-a53_gcc-8.4.0_musl/include/linux/netfilter/xt_connmark.h:25:1: error: enumerator value for 'XT_CONNMARK_DSCP' is not an integer constant
      Signed-off-by: default avatarKevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
    • Jason A. Donenfeld's avatar
      wireguard: bump to 0.0.20200318 · 2bd56595
      Jason A. Donenfeld authored
      WireGuard had a brief professional security audit. The auditors didn't find
      any vulnerabilities, but they did suggest one defense-in-depth suggestion to
      protect against potential API misuse down the road, mentioned below. This
      compat snapshot corresponds with the patches I just pushed to Dave for
      * curve25519-x86_64: avoid use of r12
      This buys us 100 extra cycles, which isn't much, but it winds up being even
      faster on PaX kernels, which use r12 as a RAP register.
      * wireguard: queueing: account for skb->protocol==0
      This is the defense-in-depth change. We deal with skb->protocol==0 just fine,
      but the advice to deal explicitly with it seems like a good idea.
      * receive: remove dead code from default packet type case
      A default case of a particular switch statement should never be hit, so
      instead of printing a pretty debug message there, we full-on WARN(), so that
      we get bug reports.
      * noise: error out precomputed DH during handshake rather than config
      All peer keys will now be addable, even if they're low order. However, no
      handshake messages will be produced successfully. This is a more consistent
      behavior with other low order keys, where the handshake just won't complete if
      they're being used anywhere.
      * send: use normaler alignment formula from upstream
      We're trying to keep a minimal delta with upstream for the compat backport.
      Signed-off-by: default avatarJason A. Donenfeld <Jason@zx2c4.com>
    • Jason A. Donenfeld's avatar
      wireguard-tools: bump to 1.0.20200319 · 858c6b17
      Jason A. Donenfeld authored
      * netlink: initialize mostly unused field
      * curve25519: squelch warnings on clang
      Code quality improvements.
      * man: fix grammar in wg(8) and wg-quick(8)
      * man: backlink wg-quick(8) in wg(8)
      * man: add a warning to the SaveConfig description
      Man page improvements. We hope to rewrite our man pages in mdocml at some
      point soon.
      Signed-off-by: default avatarJason A. Donenfeld <Jason@zx2c4.com>
    • Chuanhong Guo's avatar
      mediatek: move mt76 and wpad-bacic to mt7622 subtarget · ed7751f9
      Chuanhong Guo authored
      mt76 is a target default package for mt7622-wmac only.
      mt7623 doesn't have integrated wireless support and wifi drivers for
      pcie cards should be added as device specific package.
      mt7629-wmac isn't supported by mt76 yet.
      Signed-off-by: default avatarChuanhong Guo <gch981213@gmail.com>
    • Chuanhong Guo's avatar
      mediatek: mt7629: fix firmware partition · b7fe92b9
      Chuanhong Guo authored
      rename "kernel" partition in upstream dts to "firmware" and add
      denx,fit as compatible string for mtdsplit.
      Signed-off-by: default avatarChuanhong Guo <gch981213@gmail.com>
    • Chuanhong Guo's avatar
      mediatek: backport spi-mem based mtk spinor driver · 2d2e9d29
      Chuanhong Guo authored
      This new driver has full quadspi and DMA support, providing way better
      reading performance.
      Signed-off-by: default avatarChuanhong Guo <gch981213@gmail.com>
    • Chuanhong Guo's avatar
    • Chuanhong Guo's avatar
      mediatek: mt7629: fix ethernet default config · 1bbb8807
      Chuanhong Guo authored
      1. fix typo: ucidef_set_interfaces_wan -> ucidef_set_interface_wan
      2. change board name to mt7629-rfb to match upstream dts
      Signed-off-by: default avatarChuanhong Guo <gch981213@gmail.com>
    • Luiz Angelo Daros de Luca's avatar
      kernel: load rtl8366_smi and rtl8367b on boot · 5a326688
      Luiz Angelo Daros de Luca authored
      This external switch driver should be loaded on boot for network support
      in failsafe mode.
      Signed-off-by: default avatarLuiz Angelo Daros de Luca <luizluca@gmail.com>
      [alter commit message]
      Signed-off-by: default avatarChuanhong Guo <gch981213@gmail.com>
    • Serge Vasilugin's avatar
      kernel: rtl8367b: add configuration for extif2 · 9801d61c
      Serge Vasilugin authored
      Both rtl8367b and rtl8367s have two extended interface
      rtl8367rb: 5 port + 2*RGMII/MII
      rtl8367s:  5 port + SGMII/HSGMI + RGMII/MII
      (?)rtl8367sb:  5 port + 2*RGMII/MII
      These interfaces correspond to EXT1 and EXT2 (ports 6 and 7 respectivly).
      This patch allow to configure EXT2 in dts-file:
      	rtl8367rb {
      		compatible = "realtek,rtl8367b";
      		cpu_port = <7>;
      		realtek,extif2 = <1 0 1 1 1 1 1 1 2>;
      		mii-bus = <&mdio0>;
      		phy_id = <29>;
      Signed-off-by: default avatarSerge Vasilugin <vasilugin@yandex.ru>
      [fix indent, replace magic value, alter commit message]
      Signed-off-by: default avatarChuanhong Guo <gch981213@gmail.com>
    • Chuanhong Guo's avatar
      generic: rtl8367b: add definition of debug reg · c0bc706c
      Chuanhong Guo authored
      This commit adds definition of DEBUG0 and DEBUG1 registers and replace
      magic values with proper register modifying.
      Signed-off-by: default avatarChuanhong Guo <gch981213@gmail.com>
  2. 20 Mar, 2020 3 commits
    • Oldřich Jedlička's avatar
      mac80211: do not try to setup hostapd-managed interfaces. · 23a885bf
      Oldřich Jedlička authored
      For virtual access points (when multiple SSIDs are used for one
      physical AP), there exist one physical network interface and
      multiple virtual interfaces, which are fully under control of
      hostapd. When networking is setup, the script
      `/lib/netifd/wireless/mac80211.sh` is called, which tries to bring
      the interface up by a call to `ip link set dev <iface> up`. This
      call might fail for virtual APs, because the virtual interface
      might not have been created by hostapd yet. There are some artifical
      delays in the script most probably to handle this, but when DFS
      channel availability check on 5GHz band is issued, hostapd can
      delay creating virtual interfaces by a minute.
      In order to fix this (or work around it), do not try to bring the
      interface up (this is responsibility of hostapd anyway) and
      do not try to set txpower on the virtual interface.
      Fixes FS#2698.
      Signed-off-by: default avatarOldřich Jedlička <oldium.pro@gmail.com>
    • Chen Minqiang's avatar
      x86: make crashdump works · ec5e8461
      Chen Minqiang authored
      1. KERNEL_CRASH_DUMP should depends on KERNEL_PROC_KCORE (kexec use it)
      2. select crashkernel mem size by totalmem
         mem <= 256M disable crashkernel by default
         mem >= 4G use 256M for crashkernel
         mem >= 8G use 512M for crashkernel
         default use 128M
      3. set BOOT_IMAGE in kdump.init
      4. resolve a "Unhandled rela relocation: R_X86_64_PLT32" error
      Tested on x86_64
      Signed-off-by: default avatarChen Minqiang <ptpt52@gmail.com>
    • Jan Kardell's avatar
      readline: needs host depend on ncurses to build · ecef29b2
      Jan Kardell authored
      We must ensure that host ncurses is build before host readline.
      Signed-off-by: default avatarJan Kardell <jan.kardell@telliq.com>
  3. 19 Mar, 2020 1 commit
  4. 18 Mar, 2020 9 commits
    • Hauke Mehrtens's avatar
      toolchain: Update GCC 9 to version 9.3.0 · cb44ab4f
      Hauke Mehrtens authored
      The removed patch is included in GCC 9.3.0.
      Signed-off-by: default avatarHauke Mehrtens <hauke@hauke-m.de>
    • Hauke Mehrtens's avatar
      toolchain: Update GCC 8 to version 8.4.0 · db700776
      Hauke Mehrtens authored
      The removed patch is included in GCC 8.4.0.
      Signed-off-by: default avatarHauke Mehrtens <hauke@hauke-m.de>
    • Jeffery To's avatar
      cryptodev-linux: Fix error when compiling with 5.4 kernel · d3320082
      Jeffery To authored
      Currently, cryptodev-linux fails to compile with a
      '"crypto_givcipher_type" undefined' error for targets on the 5.4 kernel,
      e.g. armvirt[1].
      This backports an upstream patch[2] that fixes this error.
      [1]: https://downloads.openwrt.org/snapshots/faillogs/aarch64_generic/base/cryptodev-linux/compile.txt
      [2]: https://github.com/cryptodev-linux/cryptodev-linux/commit/f971e0cd4a0ebe59fb2e8e17240399bf6901b09b
      Signed-off-by: default avatarJeffery To <jeffery.to@gmail.com>
    • Daniel Golle's avatar
      hostapd: fix segfault in wpa_supplicant ubus · 50a59b3a
      Daniel Golle authored
      When introducing ubus reload support, ubus initialization was moved
      to the service level instead of being carried out when adding a BSS
      configuration. While this works when using wpa_supplicant in that way,
      it breaks the ability to run wpa_supplicant on the command line, eg.
      for debugging purposes.
      Fix that by re-introducing ubus context intialization when adding
      Reported-by: @PolynomialDivision https://github.com/openwrt/openwrt/pull/2417
      Fixes: 60fb4c92
       ("hostapd: add ubus reload")
      Signed-off-by: default avatarDaniel Golle <daniel@makrotopia.org>
    • Davide Fioravanti's avatar
      ath79: add support for TP-Link Archer D7/D7b v1 · 213250b5
      Davide Fioravanti authored
      TP-Link Archer D7 v1 is a dual-band AC1750 router + modem.
      The router section is based on Qualcomm/Atheros QCA9558 + QCA9880.
      The "DSL" section is based on BCM6318 but it's currently not supported.
      The Archer D7b seems to differ from the Archer D7 only in the
      partition table.
      Router section - Specification:
      775/650/258 MHz (CPU/DDR/AHB)
      128 MB of RAM (DDR2)
      16 MB of FLASH (SPI NOR)
      3T3R 2.4 GHz
      3T3R 5 GHz
      4x 10/100/1000 Mbps Ethernet
      7x LED, 2x button
      UART header on PCB
      Known issues:
      - Broadband LED (missing GPIO - probably driven by the BCM6318)
      - Internet LED (missing GPIO - probably driven by the BCM6318)
      - WIFI LED (working only for one interface at a time, while in the
        OEM firmware works for both wifi interfaces; thus, this patch does
        not set a trigger by default)
      - DSL not working (eth0)
      UART connection
      J1 HEADER (Qualcomm CPU)
      . VCC
      . GND
      . RX
      O TX
      J41 HEADER (Broadcom CPU)
      . VCC
      . GND
      . RX
      O TX
      The following instructions require a connection to the J1 UART header
      and are tested for the Archer D7 v1.
      For the Archer D7b v1, names should be changed accordingly.
      Flash instructions under U-Boot, using UART
       1. Press "tpl" to stop autobooting and obtain U-Boot CLI access.
       2. Setup ip addresses for U-Boot and your tftp server.
       3. Issue below commands:
      	tftpboot 0x81000000 openwrt-ath79-generic-tplink_archer-d7-v1-squashfs-sysupgrade.bin
      	erase 0x9f020000 +f90000
      	cp.b 0x81000000 0x9f020000 0xf90000
      Initramfs instructions under U-Boot for testing, using UART
       1. Press "tpl" to stop autobooting and obtain U-Boot CLI access.
       2. Setup ip addresses for U-Boot and your tftp server.
       3. Issue below commands:
      	tftpboot 0x81000000 openwrt-ath79-generic-tplink_archer-d7-v1-initramfs-kernel.bin
      	bootm 0x81000000
       4. Here you can backup the original firmware and/or flash the sysupgrade openwrt if you want
      Restore the original firmware
       0. Backup every partition using the OpenWrt web interface
       1. Download the OEM firmware from the TP-Link website
       2. Extract the bin file in a folder (eg. Archer_D7v1_1.6.0_0.9.1_up_boot(160216)_2016-02-16_15.55.48.bin)
       3. Remove the U-Boot and the Broadcom image part from the file.
          Issue the following command:
      	dd if="Archer_D7v1_1.6.0_0.9.1_up_boot(160216)_2016-02-16_15.55.48.bin" of="Archer_D7v1_1.6.0_0.9.1_up_boot(160216)_2016-02-16_15.55.48.bin.mod" skip=257 bs=512 count=31872
       4. Double check the .mod file size. It must be 16318464 bytes.
       5. Flash it using the OpenWrt web interface. Force the update if needed.
          WARNING: Remember to NOT keep settings.
       5b. (Alternative to 5.) Flash it using the U-Boot and UART connection.
           Issue below commands in the U-Boot:
      	tftpboot 0x81000000 Archer_D7v1_1.6.0_0.9.1_up_boot(160216)_2016-02-16_15.55.48.bin.mod
      	erase 0x9f020000 +f90000
      	cp.b 0x81000000 0x9f020000 0xf90000
      Signed-off-by: default avatarDavide Fioravanti <pantanastyle@gmail.com>
      [cosmetic DTS changes, remove TPLINK_HWREVADD := 0, do not use two
      phyXtpt at once, add missing buttons, minor commit message adjustments]
      Signed-off-by: default avatarAdrian Schmutzler <freifunk@adrianschmutzler.de>
    • Adrian Schmutzler's avatar
      ath79: add support for TP-Link Archer C60 v3 · 646d95c3
      Adrian Schmutzler authored
      TP-Link Archer C60 v3 is a dual-band AC1350 router,
      based on Qualcomm/Atheros QCA9561 + QCA9886.
      It seems to be identical to the v2 revision, except that
      it lacks a WPS LED and has different GPIO for amber WAN LED.
      - 775/650/258 MHz (CPU/DDR/AHB)
      - 64 MB of RAM (DDR2)
      - 8 MB of FLASH (SPI NOR)
      - 3T3R 2.4 GHz
      - 2T2R 5 GHz
      - 5x 10/100 Mbps Ethernet
      - 6x LED, 2x button
      - UART header on PCB
      Flash instruction (WebUI):
      Download *-factory.bin image and upload it via the firmwary upgrade
      function of the stock firmware WebUI.
      Flash instruction (TFTP):
      1. Set PC to fixed IP address
      2. Download *-factory.bin image and rename it to tp_recovery.bin
      3. Start a tftp server with the file tp_recovery.bin in its root
      4. Turn off the router
      5. Press and hold reset button
      6. Turn on router with the reset button pressed and wait ~15 seconds
      7. Release the reset button and after a short time the firmware should
         be transferred from the tftp server
      8. Wait ~30 second to complete recovery
      While TFTP works for OpenWrt images, my device didn't accept the
      only available official firmware "Archer C60(EU)_V3.0_190115.bin".
      In contrast to earlier revisions (v2), the v3 contains the (same)
      MAC address twice, once in 0x1fa08 and again in 0x1fb08.
      While the partition-table on the device refers to the latter, the
      firmware image contains a different partition-table for that region:
      name           device            firmware
      factory-boot   0x00000-0x1fb00   0x00000-0x1fa00
      default-mac    0x1fb00-0x1fd00   0x1fa00-0x1fc00
      pin            0x1fd00-0x1fe00   0x1fc00-0x1fd00
      product-info   0x1fe00-0x1ff00   0x1fd00-0x1ff00
      device-id      0x1ff00-0x20000   0x1ff00-0x20000
      While the MAC address is present twice, other data like the PIN isn't,
      so with the partitioning from the firmware image the PIN on the device
      would actually be outside of its partition.
      Consequently, the patch uses the MAC location from the device (which
      is the same as for the v2).
      Signed-off-by: default avatarAdrian Schmutzler <freifunk@adrianschmutzler.de>
    • Russell Senior's avatar
      ath79: rename ubnt_bullet-m to ubnt_bullet-m-ar7241 · 8cacb840
      Russell Senior authored
      Since there exists another variant of the Bullet M with AR7240 SoC
      under the same name, this patch introduces the SoC into the device
      name to be able to distinguish these variants.
      Signed-off-by: default avatarRussell Senior <russell@personaltelco.net>
      [add commit message, adjust model in DTS, fix 02_network and
      Signed-off-by: default avatarAdrian Schmutzler <freifunk@adrianschmutzler.de>
    • Russell Senior's avatar
      ath79: generalize ubnt_xm dtsi for ar7240 and ar7241 · 572278d3
      Russell Senior authored
      * Prepare to support the AR7240 variant of ubiquiti bullet m, by
        reorganizing the related dtsi files.
      * Distribute SOC variable across ubnt-xm devices.
      Signed-off-by: default avatarRussell Senior <russell@personaltelco.net>
    • Bjørn Mork's avatar
      bcm27xx: enable THERMAL_WRITABLE_TRIPS on bcm2710 and bcm2711 · dabe8471
      Bjørn Mork authored
      The PoE hat supported on the RPi 3B+ and 4B comes with a rather
      loud fan, tripped in two steps by the brcmstb_thermal
      thermal_zone driver.  Enabling writable trip points allow
      users to adjust the fan trip points according to their
      preferred temp/noise ratio.
      This setting is enabled on most other targets with similar
      thermal_zone controlled cooling devices.
      Signed-off-by: default avatarBjørn Mork <bjorn@mork.no>
  5. 17 Mar, 2020 9 commits
  6. 16 Mar, 2020 5 commits