@frank-w: your suggestions from july
ip link set eth0 up
ip addr add 192.168.1.1/24 dev br-lan
ip link set br-lan up
has no effect. Device remains offline at eth0 or br-lan. eth1 is working well (pppoe-device)
which kernel (5.4.58+)? have you added your port(s) to br-lan? any error in dmesg/while applying the commands? if you mean the trgmii-patch than it can be included already in your kernelversion so patch cannot be applied again
i wonder why eth1 should work…it is not connected in mainline kernel
maybe a “ip a” and dmesg output may help
as I see ./target/linux/mediatek/Makefile says kernel 4.14 root@BananaPi:/# dmesg | tail [ 17.282486] mt7530 3.switch lan2: Link is Down [ 19.358091] br-lan: port 4(lan3) entered blocking state [ 19.363284] br-lan: port 4(lan3) entered forwarding state [ 19.369002] mt7530 3.switch lan3: Link is Up - 1Gbps/Full - flow control off [ 19.376205] mt7530 3.switch wan: Link is Up - 1Gbps/Full - flow control rx/tx [ 19.376327] mt7530 3.switch lan2: Link is Up - 1Gbps/Full - flow control rx/tx [ 19.384122] IPv6: ADDRCONF(NETDEV_CHANGE): wan: link becomes ready [ 19.397004] br-lan: port 3(lan2) entered blocking state [ 19.402246] br-lan: port 3(lan2) entered forwarding state [ 19.408059] IPv6: ADDRCONF(NETDEV_CHANGE): wan.7: link becomes ready and root@BananaPi:/# ip a 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN qlen 1000 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo valid_lft forever preferred_lft forever inet6 ::1/128 scope host valid_lft forever preferred_lft forever 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP qlen 1000 link/ether 3a:e4:f5:29:9d:ff brd ff:ff:ff:ff:ff:ff inet6 fe80::38e4:f5ff:fe29:9dff/64 scope link valid_lft forever preferred_lft forever 3: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP qlen 1000 link/ether 9a:de:54:ec:04:9a brd ff:ff:ff:ff:ff:ff inet6 fe80::98de:54ff:feec:49a/64 scope link valid_lft forever preferred_lft forever 4: wan@eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP qlen 1000 link/ether 9a:de:54:ec:04:9a brd ff:ff:ff:ff:ff:ff inet6 fe80::98de:54ff:feec:49a/64 scope link valid_lft forever preferred_lft forever 5: lan0@eth0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue master br-lan state LOWERLAYERDOWN qlen 1000 link/ether 3a:e4:f5:29:9d:ff brd ff:ff:ff:ff:ff:ff 6: lan1@eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue master br-lan state UP qlen 1000 link/ether 3a:e4:f5:29:9d:ff brd ff:ff:ff:ff:ff:ff 7: lan2@eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue master br-lan state UP qlen 1000 link/ether 3a:e4:f5:29:9d:ff brd ff:ff:ff:ff:ff:ff 8: lan3@eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue master br-lan state UP qlen 1000 link/ether 3a:e4:f5:29:9d:ff brd ff:ff:ff:ff:ff:ff 9: br-lan: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP qlen 1000 link/ether 0e:12:76:f3:b0:dd brd ff:ff:ff:ff:ff:ff inet 192.168.6.101/24 brd 192.168.6.255 scope global br-lan valid_lft forever preferred_lft forever inet6 fd84:72d7:dedd::1/60 scope global valid_lft forever preferred_lft forever inet6 fe80::c12:76ff:fef3:b0dd/64 scope link valid_lft forever preferred_lft forever 15: wan.7@wan: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP qlen 1000 link/ether 9a:de:54:ec:04:9a brd ff:ff:ff:ff:ff:ff inet6 fe80::98de:54ff:feec:49a/64 scope link valid_lft forever preferred_lft forever 16: pppoe-wan: <POINTOPOINT,MULTICAST,NOARP,UP,LOWER_UP> mtu 1492 qdisc fq_codel state UNKNOWN qlen 3
Ip address and several other stuff are predefined in ./files/etc/config if you´re wondering about the cruel ip. But so I have also the credentials for v-dsl on sd-card ready for boot link/ppp inet 87.176.255.65 peer 62.155.247.131/32 scope global pppoe-wan valid_lft forever preferred_lft forever inet6 2003:c1:e7ff:2867:569:2cc4:79f9:22d0/64 scope global dynamic valid_lft 14343sec preferred_lft 1743sec inet6 fe80::569:2cc4:79f9:22d0/10 scope link valid_lft forever preferred_lft forever
Btw: how can I format this readable? Bootlog is here
You use not master? Then it may be the old problem where gmac0 and 1 needs to be configured to same mode (rgmii)…can you check it? And it explains,why you can use gmac1 because of old second-gmac patches.
You can format code/logs by select the text and press <> button
You use not master?
I thougt it 
git branch reports
* (HEAD losgelöst bei v19.07.3) master
How to check? I found ./build_dir/…/arm/boot/dts/mt7623n-bananapi-bpi-r2.dts with marks from this thread: bpi-r2-adding-second-gmac-to-4-14
gmac0 mode=trgmii gmac1 mode=rgmii
may be, I missed to compile switch module for mt7630?
Strange,that your makefile points to 4.14
By default 5.4 is used. But if you want to keep 4.14,change trgmii to rgmii in the bpi-r2.dts (i hope i do not get overridden). Normally trgmii on gmac0 and rgmii on gmac1 is right,but driver in 4.14 is buggy and does not setup trgmii right so change to rgmii
If you see dsa-ports like wan and lanX you have switch-driver included.
put my repo in trash and clone new…
May be I have done “git checkout” instead of “git clone” some months ago.
Now I have kernel 5.4 and old problem: generated image not bootable, hangs on
[BLDR] jump to 0x81E00000 [BLDR] <0x81E00000>=0xEA0000B8 [BLDR] <0x81E00004>=0xE59FF014
I used build_dir/target…/tmp/openwrt-…img.gz, unzipped it and wrote to /dev/sdb
I’m not deep in openwrts build process,but afair this comes up if uboot is missing/not bootable
Found it: Bug in ./target/linux/mediatek/image/gen_mtk_mmc_img.sh, line 123/124:
> echo -en "${SDMMC_BOOT}" | dd bs=1 of="${OUTPUT}" seek=0 conv=notrunc`
> echo -en "${BRLYT}" | dd bs=1 of="${OUTPUT}" seek=512 conv=notrunc`
writes the boot code as ASCII into the image file, including a leading “-en”. As I see, the option -en has no effect, because also the trailing 0x0a is written to $output.
I used okteta to convert it into binary code and voila, the image starts from sd.
BTW: Box is online, including eth0
1 Like
Thank you for the Update
@dwmw2 as author of the script can you please check/fix this? for me it looks right
echo -e should convert \xXX to the byte with hexvalue XX (-n for no newline at the end)
you see only eth0? you should see wan/lanX and need assign ip to these dsa-interfaces (not to eth0)
i tried with this simple script:
1 #!/bin/bash
2
3 OUTPUT=test.img
4
5 dd if=/dev/zero of=${OUTPUT} bs=1M count=2;
6
7 SDMMC_BOOT="SDMMC_BOOT\x00\x00\x01\x00\x00\x00\x00\x02\x00\x00"
8 EMMC_BOOT="EMMC_BOOT\x00\x00\x00\x01\x00\x00\x00\x00\x02\x00\x00"
9 BRLYT="BRLYT\x00\x00\x00\x01\x00\x00\x00\x00\x08\x00\x00\
10 \x00\x08\x00\x00\x42\x42\x42\x42\x08\x00\x01\x00\x00\x08\x00\x00\
11 \x00\x08\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00"
12
13
14 echo -en "${SDMMC_BOOT}" | dd bs=1 of="${OUTPUT}" seek=0 conv=notrunc
15 echo -en "${BRLYT}" | dd bs=1 of="${OUTPUT}" seek=512 conv=notrunc
16
which results as it should:
00000000 53 44 4d 4d 43 5f 42 4f 4f 54 00 00 01 00 00 00 |SDMMC_BOOT......|
00000010 00 02 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
00000020 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
*
00000200 42 52 4c 59 54 00 00 00 01 00 00 00 00 08 00 00 |BRLYT...........|
00000210 00 08 00 00 42 42 42 42 08 00 01 00 00 08 00 00 |....BBBB........|
00000220 00 08 00 00 01 00 00 00 00 00 00 00 00 00 00 00 |................|
00000230 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
*
00200000Box is fully working including all interfaces (lan, wan and wan6). Openvpn is stopped. No further investigation ATM. Probably thursday.
Because I have altered the script for debugging purposed, I copied the original from github. Same results, the string on 0:0 begins with “-en SDMMC_BOOT\x00\x00\x01” and so on, any null bytes. I´m running ubuntu 20.04.1 on my system, if that matters. I also created test.sh like frank-w and get the correct results on my build system.
I guess script is not executed in local bash…more in own buildchain where echo does not support flags -en
Btw. For everyone seeing retransmitts with kernel 5.4+: i posted patch for adding pause statement in bpi-r2 dts for switchport 6
may it help to set /bin/echo in script?
i don’t know…maybe it runs in chroot or similar…
Thanks for your work. For newbies: how to add this patch in openwrt buildroot? Brute force: patching mt7623n-bananapi-bpi-r2.dts?
If my wife let me hack today, I will report. Tolerance level is low after weekend 
Change the shebang at the top of the script to #!/bin/bash instead of the current #!/bin/sh.
I’ll work out a better way to do this and fix it upstream without requiring bash, if I can…