aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* toolchain: remove references to $(LIBC)/utils subtargetMichael Pratt2023-02-251-3/+0
| | | | | | | | | | | The utils subtarget was used for uClibc and eglibc which are both obsolete and replaced in Openwrt by mainline musl and glibc, which do not have a utils Makefile target. Ref: 5d19a38d8 ("toolchain: sync eglibc headers/build split with uclibc changes") Fixes: 63fb17520 ("toolchain: remove uClibc-ng") Suggested-by: Andre Heider <a.heider@gmail.com> Signed-off-by: Michael Pratt <mcpratt@pm.me>
* toolchain: cleanup dependenciesMichael Pratt2023-02-251-4/+3
| | | | | | | Reorganize dependencies between toolchain targets to save a line. This also puts them in a linear order which is easier to read. Signed-off-by: Michael Pratt <mcpratt@pm.me>
* toolchain: fix stampfile dependencyMichael Pratt2023-02-251-2/+2
| | | | | | | | | | | | | | | | | | Use the gcc/final/install target stampfile as the prerequisite for the toolchain/compile target stampfile. The previous filename was also incorrect, the stampfile name format was changed (see reference) but was never updated here. The issue was never noticed because the usage of timestamp.pl was broken (fixed in previous commit) causing the extra prereq files passed to 'call stampfile' to never get checked at all. Fixes: 5149ed151 ("gcc: split up the build process into three distinct stages") Signed-off-by: Michael Pratt <mcpratt@pm.me>
* iproute2: update to 6.2Nick Hainke2023-02-2516-69/+26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Release Notes: https://lwn.net/Articles/923952/ Refresh patches: - 110-darwin_fixes.patch - 115-add-config-xtlibdir.patch - 140-allow_pfifo_fast.patch - 140-keep_libmnl_optional.patch - 145-keep_libelf_optional.patch - 150-keep_libcap_optional.patch - 155-keep_tirpc_optional.patch - 170-ip_tiny.patch - 175-reduce-dynamic-syms.patch - 180-drop_FAILED_POLICY.patch - 190-fix-nls-rpath-link.patch - 195-build_variant_ip_tc.patch - 200-drop_libbsd_dependency.patch - 300-selinux-configurable.patch Remove upstreamed: - 320-configure-Remove-include-sys-stat.h.patch While working on it remove AUTORELEASE. Signed-off-by: Nick Hainke <vincent@systemli.org>
* ethtool: update to 6.2Nick Hainke2023-02-254-555/+2
| | | | | | | | | | | | | | | | | | Release notes: - Feature: link down event statistics (no option) - Feature: JSON output for coalesce (-c) - Feature: new link modes (no option) - Feature: JSON output for ring (-g) - Feature: netlink handler for RSS get (-x) - Fix: fix boolean value output in JSON output - Fix: fix build errors and warnings Remove upstreamed patches: - 100-uapi-Bring-in-if-h.patch - 101-netlink-Fix-maybe-uninitialized-meters-variable.patch - 102-raw-marvell-c-Fix-build-with-musl-libc.patch Signed-off-by: Nick Hainke <vincent@systemli.org>
* libcap: update to 2.67Nick Hainke2023-02-251-3/+3
| | | | | | | | | | Release notes: https://sites.google.com/site/fullycapable/release-notes-for-libcap#h.o8papfkfh1x9 While working on it, remove $(AUTORELEASE). Tested-by: Linhui Liu liulinhui36@gmail.com # Xiaomi AX3600 Signed-off-by: Nick Hainke <vincent@systemli.org>
* base-files: add support for retrieving IPv6 assignmentsMark Baker2023-02-241-0/+14
| | | | | | | In DHCPv6-PD enabled environments, addresses are assigned to interfaces. These new functions retrieve the IPv6 assigned prefix(es). Signed-off-by: Mark Baker <mark@vpost.net>
* bcm47xx: switch default kernel to the 5.15Rafał Miłecki2023-02-241-2/+1
| | | | Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
* bcm47xx: relocate LZMA loader #2Rafał Miłecki2023-02-241-2/+2
| | | | | | | | | | | | | | | | | | | | Increased size of the 5.15 kernel requires bumping BZ_TEXT_START again. Without this CFE hangs at the: Starting program at 0x80001000 This fixes booting 5.15 based mips74k images on: 1. BCM4706 (Luxul XWR-1750) 2. BCM5357B0 (Linksys E1000 V2.1) 3. BCM47186B0 (Luxul XWR-600) 4. BCM5358 (Linksys E2500 V3) It isn't needed but also doesn't break: 1. BCM5354 (Asus WL-500gP V2) Ref: 4cd97e476089 ("bcm47xx: relocate LZMA loader") Cc: Hauke Mehrtens <hauke@hauke-m.de> Signed-off-by: Rafał Miłecki <rafal@milecki.pl> Acked-by: Hauke Mehrtens <hauke@hauke-m.de>
* kernel: Add CONFIG_PPC_QUEUED_SPINLOCKS configuration optionHauke Mehrtens2023-02-232-1/+1
| | | | | | | | | | The CONFIG_PPC_QUEUED_SPINLOCKS configuration option is not defined for kernel 5.15, it is defined for kernel 5.10. This fixes the compilation of mpc85xx/p2020 with kernel 5.15. Reviewed-by: Robert Marko <robimarko@gmail.com> Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
* mac80211: ath11k: sync with ath-nextRobert Marko2023-02-2220-4/+1616
| | | | | | | | | | | | Synchronize the ath11k backports with the current ath-next tree. This backports several memory leak issues, PCI IRQ fixup, peer add locking fix as well as IPQ5018 support, though IPQ5018 support is unused for now. This allows to easily backport further fixes as cherry picking them has started requiring manual conflict resolution. Signed-off-by: Robert Marko <robimarko@gmail.com>
* kernel: backport NVMEM late fixes from v6.2Rafał Miłecki2023-02-2310-16/+282
| | | | | | | They were backported to stable kernels but we backport more stuff on our own so we have to pick up few remaining. Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
* mpc85xx: switch default kernel to 5.15Nick Hainke2023-02-221-2/+1
| | | | | | | | | | The default kernel should be switched to 5.15 in order to enable testing by a broader audience. Tested on TP-Link TL-WDR4900 v1. Acked-by: Josef Schlehofer <pepe.schlehofer@gmail.com> Signed-off-by: Nick Hainke <vincent@systemli.org>
* openssl: fix powerpc & arc libatomic dependenciesEneas U de Queiroz2023-02-222-3/+3
| | | | | | | | | | | PowerPC CONFIG_ARCH is defined as powerpc, not ppc. Fix that in the DEPENDS condition. Arc needs to be built with libatomic. Change the OpenSSL configuration file, and add it to the libatomic DEPENDS condition. Fixes: 7e7e76afca "openssl: bump to 3.0.8" Signed-off-by: Eneas U de Queiroz <cotequeiroz@gmail.com>
* mvebu: mochabin: enlarge PCI memory windowRobert Marko2023-02-211-0/+1
| | | | | | | | | | | | | | | | Armada 7040 uses a rather small 15MB memory window for every PCI adapter, however this is not sufficient for Qualcomm QCA6390 802.11ax cards that are shipped along with the OpenWrt WLAN model of MOCHAbin as ath11k requires at least 16MB of memory. So, similar to what MACCHIATOBin has been doing for years, lets move to using the second PCIe 2 memory window and expand it to 128MB to make it future proof. This has been already sent upstream [1]. [1] https://lore.kernel.org/linux-arm-kernel/20230219121418.1395401-1-robert.marko@sartura.hr/ Signed-off-by: Robert Marko <robert.marko@sartura.hr>
* kernel: update NVMEM subsystem to the v6.3Rafał Miłecki2023-02-2124-0/+2664
| | | | Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
* mcp85xx: Switch TP-Link TL-WDR4900 v1 to DSAPawel Dembicki2023-02-208-25/+105
| | | | | | | | | This patch introduces DSA support for TP-Link TL-WDR4900 v1 switch. Swconfig driver for QCA8327 switch is removed because this router is only one device which use Qualcom swconfig switch. Signed-off-by: Pawel Dembicki <paweldembicki@gmail.com> Tested-by: Nick Hainke <vincent@systemli.org> # TP Link WDR4900 v1 (5.15)
* mpc85xx: refresh kernel configPawel Dembicki2023-02-202-27/+15
| | | | | | It was done by "make kernel_oldconfig" command for 5.10 and 5.15. Signed-off-by: Pawel Dembicki <paweldembicki@gmail.com>
* openssl: bump to 3.0.8Eneas U de Queiroz2023-02-2016-3835/+87
| | | | | | | | | | This is a major update to the current LTS version, supported until 2026-09-07. Changelog: https://github.com/openssl/openssl/blob/openssl-3.0.8/CHANGES.md Signed-off-by: Eneas U de Queiroz <cotequeiroz@gmail.com>
* tools/squashfskit4: drop unused toolChristian Marangi2023-02-205-154/+0
| | | | | | | Drop squashfskit4 tool as it got replaced by new version of squashfs4 tool. Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
* image: update LZMA_XZ_OPTIONS with new squashfs4 toolChristian Marangi2023-02-201-1/+1
| | | | | | | | | The -Xpreset option changed format and dropped the -Xe and just require the extreme string to be passed to the -Xpreset option. Update the LZMA_XZ_OPTIONS to reflect this change. Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
* tools/squashfs4: add new tool for squashfs4 imagesChristian Marangi2023-02-2011-3/+713
| | | | | | | | | | | | | | squashfs tool is finally reborn and correctly maintained. Introduce the new version as a replacement for squasfs4kit as it was a fork and also abandoned. Add additional patch to add the missing feature present in squashfskit4 but still missing on this new project. Backport each required patch that fix compilation error on macos. Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
* mac80211: fix mesh issues and improve performanceFelix Fietkau2023-02-206-1/+939
| | | | | | | fix forwarding received mesh a-msdu packets add fast xmit support for mesh to improve performance Signed-off-by: Felix Fietkau <nbd@nbd.name>
* toolchain: remove installing twice in the "initial" subdirAndre Heider2023-02-193-10/+2
| | | | | | | | | | | | | | | | This was apparently introduced to recreate the toolchain (wipe staging_dir/toolchain*, but keep build_dir/toolchain*, followed by a `make toolchain/compile`). But it leaves leftovers and causes re-links to happen at src_install phase, because of the changed paths, possibly adding yet another source of issues. With the prior commits removing various hacks related to the "initial" folder we can remove installing it twice altogether. The recreated toolchain is exactly the same as before. Signed-off-by: Andre Heider <a.heider@gmail.com>
* toolchain/musl: remove libgcc_initial hackAndre Heider2023-02-192-5/+1
| | | | | | | | | | | | This reverts r15599 [0]. This was added to fix an uclibc rebuild issue, but since uclibc isn't supported anymore [Fixes:] this can be removed. [0] https://git.openwrt.org/?p=openwrt/svn-archive/archive.git;a=commitdiff;h=016a052efeee6bc2da3f501a8c66b5aea20350b4 Fixes: 63fb1752 "toolchain: remove uClibc-ng" Signed-off-by: Andre Heider <a.heider@gmail.com>
* toolchain/glibc: fixup the libm.so linker script tooAndre Heider2023-02-191-1/+1
| | | | | | | | | | Add it to the existing list to fix it up. Found by mold: toolchain-x86_64_gcc-12.2.0_glibc/lib/libm.so:4: GROUP ( /lib/libm.so.6 AS_NEEDED ( /lib/libmvec.so.1 ) ) ^ library not found: /lib/libm.so.6 Signed-off-by: Andre Heider <a.heider@gmail.com>
* toolchain/gcc: remove glibc libgcc_eh hackAndre Heider2023-02-191-2/+0
| | | | | | | | | A modern glibc only links against libgcc_eh for its tests or when building it static, which doesn't happen here. Reverts a hunk of: a3edea1b "add support for alternative C libraries..." Signed-off-by: Andre Heider <a.heider@gmail.com>
* toolchain/binutils: get rid of outdated linaro version string hackAndre Heider2023-02-191-1/+0
| | | | | | | | | | This reverts commit 9d1b619cb1a7dd18cd12cab0d9f93452b10e2e11. No Linaro binutils are supported anymore, the leftover removal commit missed one line. Fixes: b648e1c0 "binutils: remove 2.25.1 support and leftovers of older versions" Signed-off-by: Andre Heider <a.heider@gmail.com>
* toolchain/binutils: get rid of outdated libiberty hackAndre Heider2023-02-191-1/+0
| | | | | | | | | | | | | | | This reverts commit 68fcb15f90edd92ef36990c5696b9e79df07a6ec. binutils syncs libiberty from gcc, which defaults to not installing itself since [0], see also [1]. That change is apparently part of binutils 2.24, and the oldest supported version here is 2.37, hence we can remove this hack. [0] https://gcc.gnu.org/git/?p=gcc.git;a=commitdiff;h=48d7db63ff581b026e057badf39f1f23fdf47928 [1] https://gcc.gnu.org/onlinedocs/libiberty/Using.html Signed-off-by: Andre Heider <a.heider@gmail.com>
* toolchain/binutils: get rid of ARC leftoversAndre Heider2023-02-191-6/+0
| | | | | | | | | | | | In [0] support for ARC specific binutils were added, which was later removed again [Fixes:]. Upstream binutils never had an extlib. [0] 8b5d644b "toolchain: add support of ARC architecture" Fixes: 8568dcd9 "toolchain: Remove binutils 2.29 for ARC targets" Signed-off-by: Andre Heider <a.heider@gmail.com>
* toolchain/binutils: fill in missing GRAPHITE_CONFIGUREAndre Heider2023-02-191-0/+6
| | | | | | | This var is used further down, but it's empty. Fix that with the very same values gcc uses. Signed-off-by: Andre Heider <a.heider@gmail.com>
* hostapd: always use sae_password for mesh/SAE authLeon M. Busch-George2023-02-191-4/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch fixes a corner case when using passwords that are exactly 64 characters in length with mesh mode or passwords longer than 63 characters with SAE because 'psk' is used instead of 'sae_password'. SAE is obligatory for 802.11s (mesh point). The 'psk' option for hostapd is suited for WPA2 and enforces length restrictions on passwords. Values of 64 characters are treated as PMKs. With SAE, PMKs are always generated during the handshake and there are no length restrictions. The 'sae_password' option is more suited for SAE and should be used instead. Before this patch, the 'sae_password' option is only used with mesh mode passwords that are not 64 characters long. As a consequence: - mesh passwords can't be 64 characters in length - SAE only works with passwords with lengths >8 and <=63 (due to psk limitation). Fix this by always using 'sae_password' with SAE/mesh and applying the PMK differentiation only when PSK is used. Fixes: #11324 Signed-off-by: Leon M. Busch-George <leon@georgemail.eu> [ improve commit description ] Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
* hostapd: add quotes in assignmentsLeon M. Busch-George2023-02-191-6/+6
| | | | | | It's generally advised to use quotes for variable assignments in bash. Signed-off-by: Leon M. Busch-George <leon@georgemail.eu>
* ramips: add alternative device name for Wiflyer WF3526-PShiji Yang2023-02-191-0/+2
| | | | | | | | | Wiflyer WF3526-P and Zbtlink ZBT-WE1326 have the same circuit design. Installing the misunderstading firmware of ZBT-WE3526 will cause Wi-Fi not work due to allocate the wrong pcie port. Add alternative name to help users easily build or download the correct firmware. Signed-off-by: Shiji Yang <yangshiji66@qq.com>
* ramips: correct the PCIe port number for Zbtlink ZBT-WE1326Shiji Yang2023-02-191-2/+2
| | | | | | | | | | | | | | | | | MT7621 gets a new PCIe driver in the 5.15+ kernel. Allocating wrong PCIe port will cause the PCIe NIC to not work properly. This commit fixes the wrong port numbers on Zbtlink ZBT-WE1326. According to the bootlog, MT7612E (5 GHz) is connected to pcie1, and MT7603E (2 GHz) is connected to pcie2: [4.197658] mt7621-pci 1e140000.pcie: pcie0 no card, disable it (RST & CLK) [4.204609] mt7621-pci 1e140000.pcie: PCIE1 enabled [4.209476] mt7621-pci 1e140000.pcie: PCIE2 enabled ... [4.307988] pci 0000:01:00.0: [14c3:7662] type 00 class 0x028000 [4.367206] pci 0000:02:00.0: [14c3:7603] type 00 class 0x028000 Signed-off-by: Shiji Yang <yangshiji66@qq.com>
* base-files: sysfixtime: Fix time on the fake RTCYuan Tao2023-02-191-6/+16
| | | | | | | | | | | | | | | On some devices the chip has RTC but no battery save time. This leads back to getting the wrong time and skipping the check of the last file modification date. This commit ensures that the file time is checked even if the RTC exists. which would ordinarily return an approbiate system time used for e.g. certificate generation. Tested-on: NanoPi R2S Signed-off-by: Yuan Tao <ty@wevs.org>
* tools/squashfs: rename to squashfs3-lzmaChristian Marangi2023-02-189-12/+17
| | | | | | | | | | | | | | | | The name of squashfs is confusing since in reality it's a really old version using an old lzma library. This tools is used for old ath79 netgear target and to produde a fake squasfs3 image needed for some specific bootloader from some OEM (AVM for example) Rename squashfs tool to squasfs3-lzma to better describe it. Rename the installed bin from mksquashfs-lzma to mksquashfs3-lzma. Use tar transform to migrate the root directory in tar to the new naming. Drop redundant PKG_CAT variable not needed anymore. Also update any user of this tool. Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
* wireguard-tools: remove unnecessary .mk includesAndre Heider2023-02-181-2/+0
| | | | | | | | | Including kernel.mk moves the package build folder in the linux one, which is confusing since this isn't building any kernel modules. package-defaults.mk is already included my package.mk. Signed-off-by: Andre Heider <a.heider@gmail.com>
* ipq40xx: chromium: Enable kmod-ramoops by defaultBrian Norris2023-02-181-1/+5
| | | | | | | | Chromium devices (like Google WiFi) have ramoops memory reserved by the bootloader. Let's enable the ramoops kernel module by default, so we get better crash logging. Signed-off-by: Brian Norris <computersforpeace@gmail.com>
* ipq806x: chromium: Enable kmod-ramoops by defaultBrian Norris2023-02-181-1/+5
| | | | | | | | Chromium devices (like OnHub) have ramoops memory reserved by the bootloader. Let's enable the ramoops kernel module by default, so we get better crash logging. Signed-off-by: Brian Norris <computersforpeace@gmail.com>
* kernel: kmod-ramoops: Include pstore console supportBrian Norris2023-02-181-1/+2
| | | | | | | | | | | | | | | | | Pstore ramoops support is useful even when there isn't an explicit panic/crash. We can log all kernel messages via a "console", and then retrieve them in the event of some non-kernel-panic reset (e.g., watchdog). Since the buffer memory is already reserved, there isn't much overhead to doing this. The new console files will show up as: /sys/fs/pstore/console-ramoops-N Cc: Hannu Nyman <hannu.nyman@iki.fi> Signed-off-by: Brian Norris <computersforpeace@gmail.com>
* ramips: add support for Huasifei WS1208V2Arınç ÜNAL2023-02-183-0/+200
| | | | | | | | | | | | | | | | | | | | | | | | | | | | The Huasifei WS1208V2 is an AC1200 router featuring 5 Ethernet ports with a Quectel RM520N-GL cellular modem which supports QMI and MBIM modes. Specifications: - MT7621AT, 256 MiB RAM, 16 MiB SPI Flash - MT7603EN 2.4 GHz & MT7612EN 5 GHz WLAN - Quectel RM520N-GL Cellular Modem - 2 WLAN & 4 Cellular Antennas - 5 Gigabit Ethernet Ports - 1 USB 2.0 port - 1 PCI-E Slot - 1 M.2 slot - 1 SIM card slot - 1 SD card slot Installation: - Install sysupgrade image via ROOter OS. TFTP Recovery: - Connect to serial console. - Boot initramfs image by choosing option 1 when U-Boot prompts. - Install sysupgrade image via OpenWrt. Link: https://www.huasifei.com/a/Products/5G%20CPE/240.html Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com>
* elfutils: fix build with GCC 11Andre Heider2023-02-181-1/+9
| | | | | | | | GCC 11 doesn't know about -Wno-error=use-after-free and aborts compilation. Fixes: 2748c45d "elfutils: Ignore wrong use-after-free error" Signed-off-by: Andre Heider <a.heider@gmail.com>
* kernel: can: fix MCP251x CAN controller module autoloadTim Harvey2023-02-181-1/+1
| | | | | | Fix autoload module name for can-mcp251x kmod. Signed-off-by: Tim Harvey <tharvey@gateworks.com>
* kernel: bump 5.10 to 5.10.168John Audia2023-02-1810-84/+16
| | | | | | | | | | | | | | | | | | Manually rebased: backport-5.10/804-v5.14-0001-nvmem-core-allow-specifying-of_node.patch Removed upstreamed: generic-backport/807-v5.17-0003-nvmem-core-Fix-a-conflict-between-MTD-and-NVMEM-on-w.patch[1] All other patches automatically rebased. 1. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v5.10.168&id=34ec4c7831c416ac56619477f1701986634a7efc Build system: x86_64 Build-tested: ramips/tplink_archer-a6-v3 Run-tested: ramips/tplink_archer-a6-v3 Signed-off-by: John Audia <therealgraysky@proton.me>
* tools/zstd: update to 1.5.4Rosen Penev2023-02-181-3/+3
| | | | | | Mostly performance improvements, as usual. Signed-off-by: Rosen Penev <rosenp@gmail.com>
* netfilter: add kmod-nf-conntrackAviana Cruz2023-02-182-1/+29
| | | | | | | | | There have been some demands for the `ct count` expression, like https://forum.openwrt.org/t/22-03-2-unable-to-use-ct-count-nft-rules/146680. This adds the required kernel modules for the expression to work. Signed-off-by: Aviana Cruz <gwencroft@proton.me>
* ramips: add label MAC address for EPG600 and ESR600Michael Pratt2023-02-181-0/+1
| | | | | | | set label mac for the current ESR600 and new EPG600 using the board.d scripts Signed-off-by: Michael Pratt <mcpratt@pm.me>
* ramips: add support for Senao Engenius EPG600Michael Pratt2023-02-184-0/+246
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | FCC ID: A8J-EPG600 Engenius EPG600 is an indoor wireless router with 1 Gb ethernet switch, dual-band wireless, internal antenna plates, USB, and phone lines (not supported) this board is a Senao device: the hardware is equivalent to EnGenius ESR600 (except for phone lines) the software is Senao SDK which is based on openwrt and uboot which uses the legacy Senao header with Vendor / Product IDs to verify the firmware upgrade image. **Specification:** - MT7620 SOC MIPS 24kec, 2.4 GHz WMAC, 2x2 - RT5592N WLAN PCI chip, 5 GHz, 2x2 - QCA8337N Gb SW RGMII GbE, SW P0 -- SOC P5, 5 LEDs - 40 MHz clock - 16 MB FLASH MX25L12845EMI-10G - 64 MB RAM NT5TU32M16 - UART console J2, populated - USB 2.0 port direct to SOC - 6 GPIO LEDs power, 2G, 5G, wps2g, wps5g, line - 3 buttons reset, wps, "reg" (registeration) - 4 antennas internal omni-directional plates NOT YET SUPPORTED: VoIP - Si3050-FT + Si3019-FT Voice DAA, SPI control, PCM data - Phone Ports "TEL", "LINE" RJ11, 4P2C (2 pins) **MAC addresses:** MAC address labeled as MAC ADDRESS MACs present in both wifi cal data and uboot environment eth0.1/phy1 ---- *:82 rf 0x4 phy0 ---- *:83 factory 0x4 eth0.2 MAC *:b8 "wanaddr" **Installation:** Method 1: Firmware upgrade page: (if you cannot access the APs webpage) factory reset with the reset button connect ethernet to a computer OEM webpage at 192.168.0.1 username and password 'admin' Navigate to gear icon, "Device Management", "Tools" select the factory.dlf image Upload and verify checksum Method 2: Serial to upload initramfs: Follow directions for TFTP recovery upload and boot initramfs and do a sysupgrade **TFTP recovery:** Requires UART serial console, reset button does nothing rename initramfs-kernel.bin to 'uImageEPG600' make available on TFTP server at 192.168.99.8 power board, interrupt boot with "4" execute `tftpboot` and `bootm` (with the load address) **Return to OEM:** Images from OEM are provided, but not compatible with openwrt sysupgrade. So it must be modified. Alternatively, back up all mtd partitions before flashing **Note on switch registers:** The necessary registers needed for the QCA8337 switch can be read from interrupted boot (tftpboot, bootm) by using the following lines in the switch driver ar8327.c in the function 'ar8327_hw_config_of' where 'qca,ar8327-initvals' is parsed from DTS before the new register values are written: pr_info("0x04 %08x\n", ar8xxx_read(priv, AR8327_REG_PAD0_MODE)); pr_info("0x08 %08x\n", ar8xxx_read(priv, AR8327_REG_PAD5_MODE)); pr_info("0x0c %08x\n", ar8xxx_read(priv, AR8327_REG_PAD6_MODE)); pr_info("0x10 %08x\n", ar8xxx_read(priv, AR8327_REG_POWER_ON_STRAP)); Signed-off-by: Michael Pratt <mcpratt@pm.me>
* ramips: mt7620: split gsw hw_init to mac_init, ephy_initMichael Pratt2023-02-181-16/+24
| | | | | | | | | | | | | | | | | | | in order for the option ephy-disable to work without also needing ephy-base option, we have to skip all the lines that write to mdio addresses that assume those addresses do not have an external switch. Otherwise, ephy ports will be disabled in hardware, but register writes still happen as if they are enabled. Split the functions so that other things are done first, and ephy port setup can be skipped with a simple "return". Tested on Engenius EPG600 (MT7620A ver:2 eco:3) with QCA8337 external switch Ref: cc6fd6fbb505 ("ramips: mt7620: add ephy-disable option to switch driver") Signed-off-by: Michael Pratt <mcpratt@pm.me>
generated by cgit v1.2.3 (git 2.25.1) at 2025-02-10 09:50:57 +0000