An upcoming commit will add a full system image for U7623 which will
contain the MBR partition table and U-Boot too.
That contrasts with the current image which only owns the eMMC from
sector 0xa00 onwards, and must start with a legacy uImage.
Prepare for sysupgrade to the new images, and cope with the fact that
the recovery partition will be /dev/mmcblk0p2 instead of /dev/mmcblk0p1
after the upgrade.
This commit could potentially be backported to 19.07 to allow for direct
sysupgrade to the new image layout.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
I'm about to change the layout of the images for UniElec U7623 so make it
find the recovery partition based on which the root partition is too.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
This board ships with an ancient 14.07-based OpenWrt using block2mtd, and
the MBR partition table contains nonsense.
It is possible to sysupgrade to an upstream OpenWrt image, but the
legacy layout of the OpenWrt images start at 0xA00 in the eMMC, with
a raw uImage. The legacy OpenWrt image doesn't "own" the beginning
of the device, including the MBR and U-Boot.
This means that when a user upgrades to upstream OpenWrt, it doesn't
boot because it can't find the right partitions. So hard-code them on
the kernel's command line using CONFIG_CMDLINE_PARTITION (for block).
Additionally, the vendor firmware doesn't cope with images larger than
about 36MiB, because it only overwrites the contents of its "firmware"
MTD partition. The current layout of the legacy image wastes a lot of
space, allowing over 32MiB for the kernel and another 10MiB for the FAT
recovery file system which is only created as 3MiB. So pull those in
to allow 4¾ MiB for the kernel, 3MiB for recovery, and then we have over
20MiB for the root file system.
This doesn't affect the new images which ship with a full eMMC image
including a different MBR layout and a partition for U-Boot, because
our modern U-Boot can actually pass the command line to the kernel, and
the built-in one doesn't get used anyway.
Tested by upgrading from vendor OpenWrt to the current legacy image,
from legacy to itself, to the previous legacy layout, and then to
finally the full-system image.
This commit probably wants backporting to 19.07, which also doesn't
install over the vendor OpenWrt and doesn't even have a full-system
installation option.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Commit e53ec043ba ("kirkwood: move usb support to modules") has moved
this config symbol into generic configs, so it could be removed from
other configs.
Suggested-by: Aleksander Jan Bajkowski <A.Bajkowski@stud.elka.pw.edu.pl>
Signed-off-by: Petr Štetiar <ynezz@true.cz>
Like many boards, the Banana Pi R2 doesn't have permanant storage of
its MAC address, and we store the first random one that the kernel
generates in order to use it later and at least be consistent.
Store it in the FAT boot partition, just as the U7623 board (and others)
do.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
The supported MT7623 boards are mostly identical (what with being a
System-on-Chip and all), so unify the DEVICE_PACKAGES for them, and add
ext4 and usb support for them.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Some options were not explicitly specified, causing the kernel build to
drop to interactive mode. Set the missing options.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
This drops the shebang from all target files for /lib and
/etc/uci-defaults folders, as these are sourced and the shebang
is useless.
While at it, fix the executable flag on a few of these files.
This does not touch ar71xx, as this target is just used for
backporting now and applying cosmetic changes would just complicate
things.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
kernel config and patch files do not need to be executable. 644 is enough.
Fixes: 01c8f2e97c ("mediatek: bump to v4.19")
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
The image creation for the mt7623a-rfb-emmc has been removed during
a patch refresh without specific comment. The corresponding base-files
entries and DTS patches for 4.14 are still there.
Since mt7623 is pretty dead and nobody has missed this device, let's
just remove the rest.
Fixes: 050da2107a ("mediatek: backport upstream mediatek patches")
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Acked-by: John Crispin <john@phrozen.org>
This splits some base-files across subtargets, as done previously
on ath79 and ramips and also introduced for mt7629 subtarget here
already. Most of the existing base-files content is specific to
mt7623.
While at it, apply the following fixes:
- Remove lots of trailing whitespaces
- Remove wildcard on unielec,u7623-02-emmc-512m
- Remove inconsistent quotation marks in cases
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Acked-by: John Crispin <john@phrozen.org>
Refreshed all patches.
Altered patches:
- 666-Add-support-for-MAP-E-FMRs-mesh-mode.patch
New symbol for arm targets:
- HARDEN_BRANCH_PREDICTOR
Compile-tested on: ar71xx, cns3xxx, imx6
Runtime-tested on: ar71xx, cns3xxx, imx6
Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
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>
This commit adds support for the MT7623A-based UniElec U7623-02 router,
with eMMC storage and 512MB RAM. The router can be delivered with NAND
Flash and more memory, but I only have access to the one configuration.
The DTS is structured in such a way that adding support for
more/different storage/memory should be straight forward.
The device has the following specifications:
* MT7623A (quad-core, 1.3 GHz)
* 512MB RAM (DDR3)
* 8GB storage (eMMC 4.5)
* 2x normal miniPCIe slots
* 1x miniPCIe slot that is connected via an internal USB OTG port
* 5x 1Gbps Ethernet (MT7530 switch)
* 1x UART header
* 1x USB 3.0 port
* 1x SATA 3.0
* 1x 40P*0.5mm FPC for MIPI LCD
* 1x SIM slot
* 12x LEDs (2 GPIO controlled)
* 1x reset button
* 1x DC jack for main power (12V)
The following has been tested and is working:
* Ethernet switch
* miniPCIe slots (tested with Wi-Fi cards)
* USB 3.0 port
* sysupgrade
* reset button
Not working:
* The miniPCIe connected via USB OTG. For the port to work, some MUSB
glue must be added. I am currently in the process of porting the glue
from the vendor SDK.
Not tested:
* SATA 3.0
* MIPI LCD
Installation:
The board ships with u-boot, and the first installation needs to be done
via the bootloader using tftp. Step number one is to update the MBR of
the eMMC, as the one that ships with the device is broken. Since the
device can ship with different storage sizes, I will not provide the
exact steps for creating a valid MBR. However, I have made some
assumptions about the disk layout - there must be one 8MB recovery
partition (FAT32) and a partition for the rootfs (Linux).
The board loads the kernel from block 0xA00 (2560) and I have reserved
32MB for the kernel (65536 blocks). I have aligned the partitions on the
erase block size (4096 byte), so the recovery partition must start on
block 69632 and end on 86016 (16385 sectors). The rootfs is assumed to
start on sector 90112.
In order to install the mbr, you run the following commands from the
u-boot command line:
* tftpboot ${loadaddr} <name of mbr file>
* mmc device 0
* mmc write ${loadaddr} 0x00 1
Run the following commands to install + boot OpenWRT:
* tftpboot ${loadaddr} openwrt-mediatek-mt7623-7623a-unielec-u7623-02-emmc-512m-squashfs-sysupgrade-emmc.bin.gz
* run boot_wr_img
* run boot_rd_img
* bootm
Recovery:
In order to recover the router, you need to follow the installation
steps above (no need to replace MBR).
Notes:
* F2FS is used as the overlay filesystem.
* The device does not ship with any valid MAC address, so a random
address has to be generated. As a work-around, I write the initial
random MAC to a file on the recovery partition. The MAC of the WAN
interface is set to the MAC-address contained in this file on each boot,
and the address of the LAN-interfaces are WAN + 1. The MAC file is kept
across sysupgrade/firstboot.
My approach is slightly different than what the stock image does. The
first fives bytes of the MAC addresses in the stock image are static,
and then the last byte is random. I believe it is better to create fully
random MAC addresses.
* In order to support the miniPCIe-slots, I needed to add missing
pcie-nodes to mt7623.dtsi. The nodes are just c&p from the upstream
dtsi.
* One of the USB3.0 phys (u3phy2) on the board can be used as either USB
or PCI, and one of the wifi-cards is connected to this phy. In order to
support switching the phy from USB to PCI, I needed to patch the
phy-driver. The patch is based on a rejected (at least last time I
checked) PCI-driver submitted to the linux-mediatek mailing list.
* The eMMC is configured to boot from the user area, and according to
the data sheet of the eMMC this value can't be changed.
* I tried to structure the MBR more nicely and use for example a
FAT32-parition for the kernel, so that we don't need to write/read from
some offset. The bootloader does not support reading from
FAT32-paritions. While the command (fatload) is there, it just throws an
error when I try to use it.
* I will submit and hope to get the DTS for the device accepted
upstream. If and when that happens, I will update the patches
accordingly.
Signed-off-by: Kristian Evensen <kristian.evensen@gmail.com>
