it seems no problem with CPU-detection but there is a nullpointer dereference near mt2701_afe_pcm_driver_init
but this is new to me (only comes up in 4.9 and not from any patch)
do you use now 4.9-config from lede-repo?
i hoped that this kernel works better (but was only created for comparision and not tested), because there are not modifications by me…only the official patches from garys lede-repo
added pinmux for uart2, but still hang after loading kernel
after adding serial-drivers i get boot-output
[ 2.884457] VFS: Cannot open root device "mmcblk0p2" or unknown-block(179,2)9
[ 2.892415] Please append a correct "root=" boot option; here are the availa:
[ 2.900766] b300 7634944 mmcblk0 [ 2.904585] driver: mmcblk
[ 2.907366] b318 512 mmcblk0rpmb [ 2.911513] (driver?)
[ 2.913861] b310 4096 mmcblk0boot1 [ 2.918121] (driver?)
[ 2.920471] b308 4096 mmcblk0boot0 [ 2.924717] (driver?)
[ 2.927066] b320 30375936 mmcblk1 [ 2.930869] driver: mmcblk
[ 2.933647] b321 262144 mmcblk1p1 000ed6cd-01[ 2.938764]
[ 2.940251] b322 7270400 mmcblk1p2 000ed6cd-02[ 2.945357]
[ 2.946844] Kernel panic - not syncing: VFS: Unable to mount root fs on unkn)
so i have to swap mmc0/1
hanging now on
/sbin/init: /lib/arm-linux-gnueabihf/libc.so.6: version `LIBPAM_1.0' not found )
/sbin/init: /lib/arm-linux-gnueabihf/libc.so.6: version `LIBKMOD_5' not found ()
had some trouble because “make mt7623n_evb_fwu_defconfig” did not import arch/arm/configs/mt7623n_evb_fwu_defconfig…now making manual copy and this works…
Linux bpi-r2 4.9.44-bpi-r2+ #153 SMP Mon Mar 19 19:05:03 CET 2018 armv7l
[18:54] root@bpi-r2:~# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group defau1
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
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP g0
link/ether 12:5a:b1:db:d9:c5 brd ff:ff:ff:ff:ff:ff
3: eth1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qle0
link/ether 4a:b5:c9:5a:a5:7a brd ff:ff:ff:ff:ff:ff
i miss the lan/wan-Ports…maybe some additional driver is missing (had used my config from 4.14)…
added hnat and got compile error…fixed…now the section mismatch
WARNING: drivers/net/ethernet/mediatek/mtk_hnat/mtkhnat.o(.text+0x628): Section mismatch in reference from the function hnat_probe() to the function .init.text:hnat_init_debugfs()
The function hnat_probe() references
the function __init hnat_init_debugfs().
This is often because hnat_probe lacks a __init
annotation or the annotation of hnat_init_debugfs is wrong.
Great so you are able to compile kernel 4.9 and boot! We should have all lede kernel feature.
For port-separation, you are missing the DSA driver to see multiple lan0-3 nic. However is it need? All lan nic you should act as hardware swicth with gmac eth0, or i am miss-understanding?
Can you confirm, you have both hwnat, hwqos, vlan, enable?
that is great! you made it! DSA is fine on 4.9 lede.
I got a merge conflict while update your repo, have to re-download the all kernel.
Can you test the vlan?
I run some test on kernel 4.4.120 using your repo.
For clarification, without DSA driver you don’t see any of the LAN/WAN interface, however the system does act as a physical on eth0, so need to bridge. Or i am missing something?
I got the 2 gmac
[ 3.766296] mtk_eth_soc 1b100000.ethernet: generated random MAC address da:ce:d2:f9:b4:78
[ 3.774770] mtk_eth_soc 1b100000.ethernet: connected mac 0 to PHY at fixed-0:00 [uid=00000000, driver=Generic PHY]
[ 3.785654] mtk_eth_soc 1b100000.ethernet eth0: mediatek frame engine at 0xe32c0000, irq 71
[ 3.793998] mtk_eth_soc 1b100000.ethernet: generated random MAC address 3e:38:12:1d:2d:14
[ 3.802430] mtk_eth_soc 1b100000.ethernet: connected mac 1 to PHY at fixed-0:01 [uid=00000000, driver=Generic PHY]
[ 3.813229] mtk_eth_soc 1b100000.ethernet eth1: mediatek frame engine at 0xe32c0000, irq 71
also I found something not present in other kerne 49 and 4.14, MTK AHCI SATA support, AHCI_MTK
However my SATA drive are detected but no device…
[ 1.730721] ata1: SATA max UDMA/133 abar m512@0x60000000 port 0x60000100 irq 63
[ 1.737996] ata2: SATA max UDMA/133 abar m512@0x60000000 port 0x60000180 irq 63
...
[ 2.310571] ata1: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
[ 2.311326] ata1.00: ATA-10: INTEL SSDSC2KW480H6, LSF036C, max UDMA/133
[ 2.311331] ata1.00: 937703088 sectors, multi 16: LBA48 NCQ (depth 31/32), AA
[ 2.311859] ata1.00: configured for UDMA/133
[ 2.326174] scsi 0:0:0:0: Direct-Access ATA INTEL SSDSC2KW48 036C PQ: 0 ANSI: 5
...
[ 2.870576] ata2: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
[ 2.877458] ata2.00: ATA-10: INTEL SSDSC2KW480H6, LSF036C, max UDMA/133
[ 2.884135] ata2.00: 937703088 sectors, multi 16: LBA48 NCQ (depth 31/32), AA
[ 2.891747] ata2.00: configured for UDMA/133
[ 2.896486] scsi 1:0:0:0: Direct-Access ATA INTEL SSDSC2KW48 036C PQ: 0 ANSI: 5
Thanks for the info, when is the next release?
I don’t care much about the Kernel version, however i am looking to have all feature working properly. It can be 4.14, .4.9, 4.4 but need support for:
This time i try your config on kernel4.9, same compiling warning but it does boot
CC drivers/net/ethernet/mediatek/mtk_eth_soc.o
drivers/net/ethernet/mediatek/mtk_eth_soc.c: In function 'mtk_start_xmit':
drivers/net/ethernet/mediatek/mtk_eth_soc.c:677:9: warning: 'fport' may be used uninitialized in this function [-Wmaybe-uninitialized]
fport |= 0x4 << TX_DMA_FPORT_SHIFT;
drivers/net/ethernet/mediatek/mtk_eth_soc.c:668:26: note: 'fport' was declared here
u32 txd3 = 0, txd4 = 0, fport;
^~~~~
LD drivers/net/ethernet/mediatek/built-in.o
CC [M] drivers/net/ethernet/mediatek/mtk_hnat/hnat_nf_hook.o
LD [M] drivers/net/ethernet/mediatek/mtk_hnat/mtkhnat.o
WARNING: drivers/net/ethernet/mediatek/mtk_hnat/mtkhnat.o(.text+0x604): Section mismatch in reference from the function hnat_probe() to the function .init.text:hnat_init_debugfs()
The function hnat_probe() references
the function __init hnat_init_debugfs().
This is often because hnat_probe lacks a __init
annotation or the annotation of hnat_init_debugfs is wrong.
LD drivers/net/ethernet/built-in.o
The udev rule don’t assing a MAC by gmac as both as detected as KERNELS=="1b100000.ethernet"
I wonder how should the MAC be assign, same for lan0-3 and eth0 then a second MAC for eth1 and wan?
Should i bridge all lan0-3 or only eth0?
Also as i will vtag the wan traffic, should i use eth1 or WAN nic?
How to get the best network performance?
Manage to have kernel 4.9 with hwnat, second gmac, port-separation and vlan.
There is some compilation warning but it does work.
Will give feedback on performance.
OBJCOPY arch/arm/boot/zImage
Kernel: arch/arm/boot/zImage is ready
DTC arch/arm/boot/dts/mt2701-evb.dtb
DTC arch/arm/boot/dts/mt6580-evbp1.dtb
DTC arch/arm/boot/dts/mt6589-aquaris5.dtb
DTC arch/arm/boot/dts/mt6592-evb.dtb
DTC arch/arm/boot/dts/mt7623-evb.dtb
Warning (reg_format): "reg" property in /mt7530@0 has invalid length (4 bytes) (#address-cells == 2, #size-cells == 2)
DTC arch/arm/boot/dts/mt7623-eMMC.dtb
Warning (reg_format): "reg" property in /switch has invalid length (4 bytes) (#address-cells == 2, #size-cells == 2)
DTC arch/arm/boot/dts/mt7623-NAND.dtb
DTC arch/arm/boot/dts/mt7623-NAND-ePHY.dtb
DTC arch/arm/boot/dts/mt7623n-bananapi-bpi-r2.dtb
Warning (reg_format): "reg" property in /mt7530@0 has invalid length (4 bytes) (#address-cells == 2, #size-cells == 2)
Warning (reg_format): "reg" property in /switch has invalid length (4 bytes) (#address-cells == 2, #size-cells == 2)
DTC arch/arm/boot/dts/mt8127-moose.dtb
DTC arch/arm/boot/dts/mt8135-evbp1.dtb
drivers/misc/mediatek/connectivity/wlan/Makefile:6: include gen2
drivers/misc/mediatek/connectivity/wlan/gen2/Makefile:220: y
Building modules, stage 2.
MODPOST 110 modules
WARNING: drivers/net/ethernet/mediatek/mtk_hnat/mtkhnat.o(.text+0x604): Section mismatch in reference from the function hnat_probe() to the function .init.text:hnat_init_debugfs()
The function hnat_probe() references
the function __init hnat_init_debugfs().
This is often because hnat_probe lacks a __init
annotation or the annotation of hnat_init_debugfs is wrong.