The bootloader of a number of recent TP-Link devices does not properly
initialise the MT7621's internal switch when booting from flash. To
enable the mt7530 driver to clear the reset on the switch, the ramips
reset controller must be allowed to toggle these.
Backport upstream commit 3f9ef7785a9c from mips-next to allow control of
the "mcm" reset line.
Signed-off-by: Sander Vanheule <sander@svanheule.net>
Hardware
--------
MediaTek MT7621AT
256M DDR3
32M SPI-NOR
MediaTek MT7603 2T2R 802.11n 2.4GHz
MediaTek MT7915 2T2R 802.11ax 5GHz
Not Working
-----------
- Bluetooth (connected to UART3)
UART
----
UART is located in the lower left corner of the board. Pinout is
0 - 3V3 (don't connect)
1 - RX
2 - TX
3 - GND
Console is 115200 8N1.
Boot
----
1. Connect to the serial console and connect power.
2. Double-press ESC when prompted
3. Set the fdt address
$ fdt addr $(fdtcontroladdr)
4. Remove the signature node from the control FDT
$ fdt rm /signature
5. Transfer and boot the OpenWrt initramfs image to the device.
Make sure to name the file C0A80114.img and have it reachable at
192.168.1.1/24
$ tftpboot; bootm
Installation
------------
1. Connect to the booted device at 192.168.1.20 using username/password
"ubnt".
2. Update the bootloader environment.
$ fw_setenv devmode TRUE
$ fw_setenv boot_openwrt "fdt addr \$(fdtcontroladdr);
fdt rm /signature; bootubnt"
$ fw_setenv bootcmd "run boot_openwrt"
3. Transfer the OpenWrt sysupgrade image to the device using SCP.
4. Check the mtd partition number for bs / kernel0 / kernel1
$ cat /proc/mtd
5. Set the bootselect flag to boot from kernel0
$ dd if=/dev/zero bs=1 count=1 of=/dev/mtdblock4
6. Write the OpenWrt sysupgrade image to both kernel0 as well as kernel1
$ dd if=openwrt.bin of=/dev/mtdblock6
$ dd if=openwrt.bin of=/dev/mtdblock7
7. Reboot the device. It should boot into OpenWrt.
Below are the original installation instructions prior to the discovery
of "devmode=TRUE". They are not required for installation and are
documentation only.
The bootloader employs signature verification on the FIT image
configurations. This way, booting unauthorized image without patching
the bootloader is not possible. Manually configuring the bootcmd in the
U-Boot envronment won't work, as this is restored to the default value
if modified.
The bootloader is made up of three different parts.
1. The SPL performing early board initialization and providing a XModem
recovery in case the PBL is missing
2. The PBL being the primary U-Boot application and containing the
control FDT. It is LZMA packed with a uImage header.
3. A Ubiquiti standalone U-Boot application providing the main boot
routine as well as their recovery mechanism.
In a perfect world, we would only replace the PBL, as the SPL does not
perform checks on the PBLs integrity. However, as the PBL is in the same
eraseblock as the SPL, we need to at least rewrite both.
The bootloader will only verify integrity in case it has a "signature"
node in it's control device-tree. Renaming the signature node to
something else will prevent this from happening.
Warning: These instructions are based on the firmware intially
shipped with the device and potentially brick your device in a way it
can only be recovered using a SPI flasher.
Only (!) proceed if you understand this!
1. Extract the bootloader from the U-Boot partition using the OpenWrt
initramfs image.
2. Split the bootloader into it's 3 components:
$ dd if=bootloader.bin of=spl.bin bs=1 skip=0 count=45056
$ dd if=bootloader.bin of=pbl.uimage bs=1 skip=45056 count=143360
$ dd if=bootloader.bin of=ubnt.uimage bs=1 skip=188416
3. Strip the uImage header from the PBL
$ dd if=pbl.uimage of=pbl.lzma bs=64 skip=1
4. Decompress the PBL
$ lzma -d pbl.lzma --single-stream
The decompressed PBL sha256sum should be
d8b406c65240d260cf15be5f97f40c1d6d1b6e61ec3abed37bb841c90fcc1235
5. Open the decompressed PBL using your favorite hexeditor. Locate the
control FDT at offset 0x4CED0 (0xD00DFEED). At offset 0x4D5BC, the
label for the signature node is located. Rename the "signature"
string at this offset to "signaturr".
The patched PBL sha256sum should be
d028e374cdb40ba44b6e3cef2e4e8a8c16a3b85eb15d9544d24fdd10eed64c97
6. Compress the patched PBL
$ lzma -z pbl --lzma1=dict=67108864
The resulting pbl.lzma file should have the sha256sum
7ae6118928fa0d0b3fe4ff81abd80ecfd9ba2944cb0f0a462b6ae65913088b42
7. Create the PBL uimage
$ SOURCE_DATE_EPOCH=1607909492 mkimage -A mips -O u-boot -C lzma
-n "U-Boot 2018.03 [UniFi,v1.1.40.71]" -a 84000000 -e 84000000
-T firmware -d pbl.lzma patched_pbl.uimage
The resulting patched_pbl.uimage should have the sha256sum
b90d7fa2dcc6814180d3943530d8d6b0d6a03636113c94e99af34f196d3cf2ce
8. Reassemble the complete bootloader
$ dd if=patched_pbl.uimage of=aligned_pbl.uimage bs=143360 count=1
conv=sync
$ cat spl.bin > patched_uboot.bin
$ cat aligned_pbl.uimage >> patched_uboot.bin
$ cat ubnt.uimage >> patched_uboot.bin
The resulting patched_uboot.bin should have the sha256sum
3e1186f33b88a525687285c2a8b22e8786787b31d4648b8eee66c672222aa76b
9. Transfer your patched bootloader to the device. Also install the
kmod-mtd-rw package using opkg and load it.
$ insmod mtd-rw.ko i_want_a_brick=1
Write the patched bootloader to mtd0
$ mtd write patched_uboot.bin u-boot
10. Erase the kernel1 partition, as the bootloader might otherwise
decide to boot from there.
$ mtd erase kernel1
11. Transfer the OpenWrt sysupgrade image to the device and install
using sysupgrade.
FIT configurations
------------------
In the future, the MT7621 UniFi6 family can be supported by a single
OpenWrt image.
config@1: U6 Lite
config@2: U6 IW
config@3: U6 Mesh
config@4: U6 Extender
config@5: U6 LR-EA (Early Access - GA is MT7622)
Signed-off-by: David Bauer <mail@david-bauer.net>
All modifications made by update_kernel.sh run in a fresh clone
without any existing toolchains.
Build system: x86_64
Build-tested: ipq806x/R7800, ath79/generic, bcm27xx/bcm2711
Run-tested: ipq806x/R7800
No dmesg regressions, everything functional
Signed-off-by: John Audia <graysky@archlinux.us>
Manually rebased patches:
ath79/patches-5.4/910-unaligned_access_hacks.patch
bcm27xx/patches-5.4/950-0135-spi-spi-bcm2835-Disable-forced-software-CS.patch
bcm27xx/patches-5.4/950-0414-SQUASH-Fix-spi-driver-compiler-warnings.patch
ipq806x/patches-5.4/093-4-v5.8-ipq806x-PCI-qcom-Use-bulk-clk-api-and-assert-on-error.patch
Removed since could be reverse-applied by quilt and found to be included upstream:
ipq806x/patches-5.4/096-PCI-qcom-Make-sure-PCIe-is-reset-before-init-for-rev.patch
All modifications made by update_kernel.sh
Build system: x86_64
Build-tested: ipq806x/R7800, ath79/generic, bcm27xx/bcm2711
Run-tested: ipq806x/R7800
No dmesg regressions, everything functional
Signed-off-by: John Audia <graysky@archlinux.us>
[refresh altered targets after rebase]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Currently sfp_select_interface() return the fastest interface that
the sfp modules supports even if the phy don't support that mode.
For example an GPON module that support both 2500basex and 1000basex.
Currently sfp_select_interface() picks 2500basex instead of 1000basex.
So limit the interfaces which both sides supports before calling
sfp_select_interface() or return an error if we don't have match.
Reviewed-by: John Thomson <git@johnthomson.fastmail.com.au>
Tested-by: Braihan Cantera <bcanterac@gmail.com> [MikroTik RB760iGS + Nokia G-010S-A 3FE46541AA SFP]
Tested-by: John Thomson <git@johnthomson.fastmail.com.au> [Mikrotik rb760igs + SFP SM/LC, SFP base1000T, SFP+ passive DAC]
Signed-off-by: René van Dorst <opensource@vdorst.com>
`mt7621_nfc_write_page_hwecc` may be called with `buf=NULL`, but
`mt7621_nfc_check_empty_page` always tries to read it.
That caused Oops:
`Unable to handle kernel paging request at virtual address 00000000`
Fixes: FS#3416
Signed-off-by: Anton Ryzhov <anton@ryzhov.me>
linux-mips has zboot code which can create a self-extracting kernel
image.
This allows enabling kernel zboot support for ramips targets.
Signed-off-by: Chuanhong Guo <gch981213@gmail.com>
The code is now much cleaner and works better than the old code.
Preparation for submitting it upstream (though with a different API)
Also add back MT7621 support and fix flow table coherence issues on
MT7622
Signed-off-by: Felix Fietkau <nbd@nbd.name>
All modifications made by update_kernel.sh/no manual intervention needed
Run-tested: ipq806x (R7800), ath79 (Archer C7v5), x86/64
No dmesg regressions, everything appears functional
Signed-off-by: John Audia <graysky@archlinux.us>
[add run test from PR]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Manually merged:
hack-5.4
230-openwrt_lzma_options.patch
bcm27xx
950-0283-hid-usb-Add-device-quirks-for-Freeway-Airmouse-T3-an.patch
x86
011-tune_lzma_options.patch
Remove upstreamed patches in collaboration with Ansuel Smith:
ipq806x
093-1-v5.8-ipq806x-PCI-qcom-Add-missing-ipq806x-clocks-in-PCIe-driver.patch
093-2-v5.8-ipq806x-PCI-qcom-Change-duplicate-PCI-reset-to-phy-reset.patch
093-3-v5.8-ipq806x-PCI-qcom-Add-missing-reset-for-ipq806x.patch
All other modifications made by update_kernel.sh
Build-tested: bcm27xx/bcm2708, ipq806x, x86/64
Run-tested: ipq806x (R7800), x86/64
No dmesg regressions, everything functional
Signed-off-by: John Audia <graysky@archlinux.us>
[update commit message/tested]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
This PR is a blend of several kernel bumps authored by ldir taken from his
staging tree w/ some further adjustments made by me and update_kernel.sh
Summary:
Deleted upstreamed patches:
generic:
742-v5.5-net-sfp-add-support-for-module-quirks.patch
743-v5.5-net-sfp-add-some-quirks-for-GPON-modules.patch
bcm63xx:
022-v5.8-mtd-rawnand-brcmnand-correctly-verify-erased-pages.patch
024-v5.8-mtd-rawnand-brcmnand-fix-CS0-layout.patch
mediatek:
0402-net-ethernet-mtk_eth_soc-Always-call-mtk_gmac0_rgmii.patch
Deleted patches applied differently upstream:
generic:
641-sch_cake-fix-IP-protocol-handling-in-the-presence-of.patch
Manually merged patches:
generic:
395-v5.8-net-sch_cake-Take-advantage-of-skb-hash-where-appropriate.patch
bcm27xx:
950-0132-lan78xx-Debounce-link-events-to-minimize-poll-storm.patch
layerscape:
701-net-0231-enetc-Use-DT-protocol-information-to-set-up-the-port.patch
Build system: x86_64
Build-tested: ath79/generic, bcm27xx/bcm2708, bcm27xx/bcm2711,
imx6, mvebu/cortexa9, sunxi/a53
Run-tested: Netgear R7800 (ipq806x)
No dmesg regressions, everything functional
Signed-off-by: John Audia <graysky@archlinux.us>
Tested-By: Lucian Cristian <Lucian.cristian@gmail.com> [mvebu]
Tested-By: Curtis Deptuck <curtdept@me.com> [x86/64]
[do not remove 395-v5.8-net-sch_cake-Take-advantage-... patch,
adjust and refresh patches, adjust commit message]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Tested-By: John Audia <graysky@archlinux.us> [ipq806x]
update_kernel.sh refreshed all patches, no human interaction was needed
Build system: x86_64
Run-tested: Netgear R7800 (ipq806x)
Signed-off-by: John Audia <graysky@archlinux.us>
At this point in v5.4 kernel we cannot use dwc2_readl() and
dwc2_writel() since they rely on the value hsotg->needs_byte_swap
which cannot be obtained before the controller wakes up.
We should use readl() and writel() to wake up the controller before
calling dwc2_check_core_endianness().
Fixes: 6be0da90a1 ("ramips: refresh patches")
Signed-off-by: Alexey Dobrovolsky <dobrovolskiy.alexey@gmail.com>
[fixed Fixes: tag]
Signed-off-by: Petr Štetiar <ynezz@true.cz>
In FS#2738 we can see that patch first introduced in
e8ebcff ("ramips: add a explicit reset to dwc2")
breaks USB functionality since 18.06. Thus, this patch should be removed.
Removed:
- 0032-USB-dwc2-add-device_reset.patch
Fixes: FS#2738
Fixes: FS#2964
Signed-off-by: Alexey Dobrovolsky <dobrovolskiy.alexey@gmail.com>
Upstream pinctrl driver in drivers/staging uses
groups/function/ralink,num-gpios instead of
ralink,group/ralink,function/ralink,nr-gpio
Replace these properties in dts as well as the pinctrl driver in
patches-4.14.
This commit is created using:
sed -i 's/ralink,group/groups/g'
sed -i 's/ralink,function/function/g'
sed -i 's/ralink,nr-gpio/ralink,num-gpios/g'
Signed-off-by: Chuanhong Guo <gch981213@gmail.com>
This hack is needed for old ethernet driver:
On mt7620, we have two exposed RGMII ports that connects to builtin
switch. However, swconfig has no way to interact with phy subsystem.
As a result, we have to register both PHYs to ethernet mac instead
and this patch prevents main ethernet interface from going down due
to phy link changes.
Also rename the patch for its actual purpose.
Signed-off-by: Chuanhong Guo <gch981213@gmail.com>
Instead of resetting flash to 3B address on remove hook, this
implementation only enters 4B mode when needed, which prevents more
unexpected reboot stuck. This implementation makes it only break when
a kernel panic happens during flash operation on 16M+ areas.
Also silent broken-flash-reset warning. We are not dealing with vendors
and it's unpleasant for users to see that unnecessary and long WARN_ON
print.
Signed-off-by: Chuanhong Guo <gch981213@gmail.com>
mt7621 SoC has its own 'ralink_soc_info' structure with some
information about the soc itself. Pcie controller and pcie phy
drivers for this soc which are still in staging git tree make uses
of 'soc_device_attribute' looking for revision 'E2' in order to
know if reset lines are or not inverted. This way of doing things
seems to be necessary in order to make things clean and properly.
Hence, introduce this 'soc_device' to be able to properly use those
attributes in drivers. Also set 'data' pointer points to the struct
'ralink_soc_info' to be able to export also current soc information
using this mechanism.
Signed-off-by: DENG Qingfang <dengqf6@mail2.sysu.edu.cn>
