> I just wanted to try the debian and obtain some
> performance figures.

Can you tell me more?

> I also read the WOL message from the serial
> terminal, but was not able to wake it up, even
> with the original firmware.

Didn't try that

> Have you documented somwhere how you prepared
> kernel, ramdisk and root-filesystem and what
> changes you had to do?

Not fully, but I'll will write it down and upload it to git hub.

https://github.com/cschil

Well, I'm trying to get nfs running, but it seems you left it out from the kernel.

root@WDebian:/srv# systemctl status nfs-kernel-server.service -l
● nfs-kernel-server.service - LSB: Kernel NFS server support
Loaded: loaded (/etc/init.d/nfs-kernel-server)
Active: active (exited) since Di 2016-11-29 09:33:33 UTC; 11min ago
Process: 8763 ExecStop=/etc/init.d/nfs-kernel-server stop (code=exited, status=0/SUCCESS)
Process: 8772 ExecStart=/etc/init.d/nfs-kernel-server start (code=exited, status=0/SUCCESS)

Nov 29 09:33:33 WDebian nfs-kernel-server[8772]: Not starting NFS kernel daemon: no support in current kernel. ... (warning).

I'll upload later this week the .config and the DTS for the mainline kernel to my github.
But I can recompile the Kernel with nfs as well.

Cheers

I made some test with rsync and samba on an encrypted raid0 ext4-filesystem.

I get arround 50MB/s read/write performance with the cpu beeing arround 18% idle and having arround 45% iowait. So I guess this is the maximum I can reach with the old laptop disks I did put into the device. The device seems be able to go faster if the discs would. I think thi is a good result.

Power consupmtion varies between 7W and 12W.

I also installed owncloud with nginx and php-fpm, which also performs well.

For vpn-test also the tun device is missing from the kernel.

Looks good so far.

Thank you a lot. I'd be glad to try to compile a kernel myself, but this is not urgent. I have still not decided if I will use debian on the device or try to compile lede/openwrt for it. At least I guess I could fit lede/openwrt into the flash and have only the data on disks.

So I attached dts and .config for the mainline-kernel to this post.
Couldn't test the kernel today or provide a patch.

Thanks, but how do you compile it? Are you using the WD provided toolchain and the xbuild script?

Hey, no as its for the mainline / stabile kernel from kernel.org. I just used the gcc-linux-gunaebihf compiler with the debian toolchain.
So nothing from WD or Marvell.

Hi Peacemaker,

I managed to compile the kernel, but it doesn't boot. Do you maybe have an additional hint on how you integrate the dts-file and generate the uImage?

Thanks!
Martin

Hey Martin,
I updated the kernel and it boot's like charm.

So you can do this your own

export ARCH=arm make CROSS_COMPILE=arm-linux-gnuabihf-
cp armada-385-wd.dts arch/arm/boot/dts/
make menuconfig 
make zImage
make armada-385-wd.dtb
cat arch/arm/boot/zImage arch/arm/boot/dts/armada-385-wd.dtb > zImage+dts
mkimage -A arm -O linux -T kernel -C none -a 0x00008000 -e 0x00008000 -n 'WD-385' -d zImage+dts uImage

Hi peacemaker,

no I have no luck, I built the kernel like you said above, but still nothing happens after loading the kernel:

boot_get_kernel : Reset button have been pushed
## Booting image at 00a00000 ...
## Booting kernel from Legacy Image at 00a00000 ...
   Image Name:   WD-385
   Created:      2016-12-02  13:58:37 UTC
   Image Type:   ARM Linux Kernel Image (uncompressed)
   Data Size:    3745492 Bytes = 3.6 MiB
   Load Address: 00008000
   Entry Point:  00008000
   Verifying Checksum ... OK
## Loading init Ramdisk from Legacy Image at 00f00000 ...
   Image Name:   Ramdisk
   Created:      2016-11-22  19:37:10 UTC
   Image Type:   ARM Linux RAMDisk Image (lzma compressed)
   Data Size:    2565382 Bytes = 2.4 MiB
   Load Address: 00000000
   Entry Point:  00000000
   Verifying Checksum ... OK
   Loading Kernel Image ... OK
OK

Starting kernel ...

After that nothing happens anymore. I'm using your ramdisk unmodified and I'm using the debian jessie crosstoolchain installed via crossbuild-essential-armhf.

Do you have some idea?

Thanks!
Martin

Hey are you sure, that the armada-wd-385.dtb is build correct?
It looks that there might be a problem, cheers

Yes, I extracted the dtb from your uImage and compared it to the one generated during my kernel build:

mm@ovid:/mnt/src$ cmp dtbdump_2.dtb linux-4.8.6/arch/arm/boot/dts/armada-385-wd.dtb
mm@ovid:/mnt/src$

Hi peacemaker,

I tried your uImage-4.8.11_wd-385 in your dropbox and it doesn't boot either.

bye
Martin

hm,.. strange.

Bytes transferred = 2806891 (2ad46b hex)
## Booting image at 01200000 ...
## Booting kernel from Legacy Image at 01200000 ...
   Image Name:   385-WD
   Created:      2016-12-01  20:04:18 UTC
   Image Type:   ARM Linux Kernel Image (uncompressed)
   Data Size:    3734188 Bytes = 3.6 MiB
   Load Address: 00008000
   Entry Point:  00008000
   Verifying Checksum ... OK
## Loading init Ramdisk from Legacy Image at 02000000 ...
   Image Name:   Ramdisk
   Created:      2016-11-28  20:57:41 UTC
   Image Type:   ARM Linux RAMDisk Image (lzma compressed)
   Data Size:    2806827 Bytes = 2.7 MiB
   Load Address: 00000000
   Entry Point:  00000000
   Verifying Checksum ... OK
   Loading Kernel Image ... OK
OK

Starting kernel ...

** 2 printk messages dropped ** [    0.000000] CPU: ARMv7 Processor [414fc091] revision 1 (ARMv7), cr=10c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT aliasing instruction cache
[    0.000000] OF: fdt:Machine model: Marvell Armada 385 Development Board
[    0.000000] Memory policy: Data cache writealloc
[    0.000000] percpu: Embedded 13 pages/cpu @ef7cf000 s22860 r8192 d22196 u53248
[    0.000000] Built 1 zonelists in Zone order, mobility grouping on.  Total pages: 260608
[    0.000000] Kernel command line: console=ttyS0,115200 root=LABEL=rootfs earlyprintk root=/dev/ram console=ttyS0,115200
[    0.000000] PID hash table entries: 4096 (order: 2, 16384 bytes)
[    0.000000] Dentry cache hash table entries: 131072 (order: 7, 524288 bytes)
[    0.000000] Inode-cache hash table entries: 65536 (order: 6, 262144 bytes)
[    0.000000] Memory: 1024880K/1048576K available (8192K kernel code, 252K rwdata, 1572K rodata, 1024K init, 230K bss, 23696K reserved, 0K cma-reserved, 262144K highmem)
[    0.000000] Virtual kernel memory layout:
[    0.000000]     vector  : 0xffff0000 - 0xffff1000   (   4 kB)
[    0.000000]     fixmap  : 0xffc00000 - 0xfff00000   (3072 kB)
[    0.000000]     vmalloc : 0xf0800000 - 0xff800000   ( 240 MB)
[    0.000000]     lowmem  : 0xc0000000 - 0xf0000000   ( 768 MB)
[    0.000000]     pkmap   : 0xbfe00000 - 0xc0000000   (   2 MB)
[    0.000000]     modules : 0xbf000000 - 0xbfe00000   (  14 MB)
[    0.000000]       .text : 0xc0008000 - 0xc0900000   (9184 kB)
[    0.000000]       .init : 0xc0b00000 - 0xc0c00000   (1024 kB)
[    0.000000]       .data : 0xc0c00000 - 0xc0c3f300   ( 253 kB)
[    0.000000]        .bss : 0xc0c3f300 - 0xc0c78bd8   ( 231 kB)
[    0.000000] Hierarchical RCU implementation.
[    0.000000]  Build-time adjustment of leaf fanout to 32.
[    0.000000]  RCU restricting CPUs from NR_CPUS=4 to nr_cpu_ids=2.
[    0.000000] RCU: Adjusting geometry for rcu_fanout_leaf=32, nr_cpu_ids=2
[    0.000000] NR_IRQS:16 nr_irqs:16 16
[    0.000000] mvebu_mbus: [Firmware Warn]: deprecated mbus-mvebu Device Tree, suspend/resume will not work
[    0.000000] L2C-310 enabling early BRESP for Cortex-A9
[    0.000000] L2C-310 full line of zeros enabled for Cortex-A9
[    0.000000] L2C-310 D prefetch enabled, offset 1 lines
[    0.000000] L2C-310 dynamic clock gating enabled, standby mode enabled
[    0.000000] L2C-310 Coherent cache controller enabled, 16 ways, 1024 kB
[    0.000000] L2C-310 Coherent: CACHE_ID 0x410054c9, AUX_CTRL 0x56070001
[    0.000000] Switching to timer-based delay loop, resolution 40ns
[    0.000003] sched_clock: 32 bits at 25MHz, resolution 40ns, wraps every 85899345900ns
[    0.000011] clocksource: armada_370_xp_clocksource: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 76450417870 ns
[    0.000203] Console: colour dummy device 80x30
[    0.000216] Calibrating delay loop (skipped), value calculated using timer frequency.. 50.00 BogoMIPS (lpj=250000)
[    0.000224] pid_max: default: 32768 minimum: 301
[    0.000286] Mount-cache hash table entries: 2048 (order: 1, 8192 bytes)
[    0.000291] Mountpoint-cache hash table entries: 2048 (order: 1, 8192 bytes)
[    0.000569] CPU: Testing write buffer coherency: ok
[    0.000678] CPU0: thread -1, cpu 0, socket 0, mpidr 80000000
[    0.000692] Setting up static identity map for 0x100000 - 0x100058
[    0.000854] mvebu-soc-id: MVEBU SoC ID=0x6820, Rev=0x4
[    0.000928] mvebu-pmsu: Initializing Power Management Service Unit
[    0.001478] Booting CPU 1
[    0.001651] CPU1: thread -1, cpu 1, socket 0, mpidr 80000001
[    0.001691] Brought up 2 CPUs
[    0.001697] SMP: Total of 2 processors activated (100.00 BogoMIPS).
[    0.001700] CPU: All CPU(s) started in SVC mode.
[    0.002145] devtmpfs: initialized
[    0.003913] VFP support v0.3: implementor 41 architecture 3 part 30 variant 9 rev 4
[    0.004010] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.004037] xor: measuring software checksum speed
[    0.099815]    arm4regs  :  2100.800 MB/sec
[    0.199713]    8regs     :  1617.200 MB/sec
[    0.299614]    32regs    :  1679.600 MB/sec
[    0.399514]    neon      :  1530.000 MB/sec
[    0.399517] xor: using function: arm4regs (2100.800 MB/sec)
[    0.399523] pinctrl core: initialized pinctrl subsystem
[    0.399915] NET: Registered protocol family 16
[    0.400513] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.421248] cpuidle: using governor ladder
[    0.421311] hw-breakpoint: found 5 (+1 reserved) breakpoint and 1 watchpoint registers.
[    0.421315] hw-breakpoint: maximum watchpoint size is 4 bytes.
[    0.629502] raid6: int32x1  gen()   175 MB/s
[    0.799248] raid6: int32x1  xor()   256 MB/s
[    0.969092] raid6: int32x2  gen()   278 MB/s
[    1.138795] raid6: int32x2  xor()   283 MB/s
[    1.308731] raid6: int32x4  gen()   328 MB/s
[    1.478462] raid6: int32x4  xor()   279 MB/s
[    1.648402] raid6: int32x8  gen()   351 MB/s
[    1.818212] raid6: int32x8  xor()   258 MB/s
[    1.987960] raid6: neonx1   gen()  1010 MB/s
[    2.157781] raid6: neonx1   xor()   759 MB/s
[    2.327603] raid6: neonx2   gen()  1075 MB/s
[    2.497441] raid6: neonx2   xor()   950 MB/s
[    2.667300] raid6: neonx4   gen()   969 MB/s
[    2.837101] raid6: neonx4   xor()   858 MB/s
[    3.006911] raid6: neonx8   gen()   860 MB/s
[    3.176763] raid6: neonx8   xor()   740 MB/s
[    3.176767] raid6: using algorithm neonx2 gen() 1075 MB/s
[    3.176770] raid6: .... xor() 950 MB/s, rmw enabled
[    3.176774] raid6: using intx1 recovery algorithm
[    3.177029] vgaarb: loaded
[    3.177163] SCSI subsystem initialized
[    3.177371] usbcore: registered new interface driver usbfs
[    3.177410] usbcore: registered new interface driver hub
[    3.177442] usbcore: registered new device driver usb
[    3.178165] clocksource: Switched to clocksource armada_370_xp_clocksource
[    3.178226] VFS: Disk quotas dquot_6.6.0
[    3.178256] VFS: Dquot-cache hash table entries: 1024 (order 0, 4096 bytes)
[    3.182482] NET: Registered protocol family 2
[    3.182772] TCP established hash table entries: 8192 (order: 3, 32768 bytes)
[    3.182818] TCP bind hash table entries: 8192 (order: 4, 65536 bytes)
[    3.182891] TCP: Hash tables configured (established 8192 bind 8192)
[    3.182930] UDP hash table entries: 512 (order: 2, 16384 bytes)
[    3.182956] UDP-Lite hash table entries: 512 (order: 2, 16384 bytes)
[    3.183035] NET: Registered protocol family 1
[    3.183202] RPC: Registered named UNIX socket transport module.
[    3.183207] RPC: Registered udp transport module.
[    3.183211] RPC: Registered tcp transport module.
[    3.183214] RPC: Registered tcp NFSv4.1 backchannel transport module.
[    3.183324] Trying to unpack rootfs image as initramfs...
[    4.021619] Freeing initrd memory: 2744K (c2000000 - c22ae000)
[    4.021777] hw perfevents: enabled with armv7_cortex_a9 PMU driver, 7 counters available
[    4.022369] futex hash table entries: 512 (order: 3, 32768 bytes)
[    4.022720] workingset: timestamp_bits=30 max_order=18 bucket_order=0
[    4.022885] squashfs: version 4.0 (2009/01/31) Phillip Lougher
[    4.023107] SGI XFS with security attributes, no debug enabled
[    4.023952] async_tx: api initialized (async)
[    4.023980] bounce: pool size: 64 pages
[    4.024022] Block layer SCSI generic (bsg) driver version 0.4 loaded (major 251)
[    4.024028] io scheduler noop registered
[    4.024032] io scheduler deadline registered
[    4.024040] io scheduler cfq registered (default)
[    4.024632] armada-38x-pinctrl f1018000.pinctrl: registered pinctrl driver
[    4.025464] mvebu-pcie soc:pcie-controller: PCI host bridge to bus 0000:00
[    4.025472] pci_bus 0000:00: root bus resource [io  0x1000-0xfffff]
[    4.025479] pci_bus 0000:00: root bus resource [mem 0xe0000000-0xe7ffffff]
[    4.025484] pci_bus 0000:00: root bus resource [bus 00-ff]
[    4.025741] PCI: bus0: Fast back to back transfers disabled
[    4.025748] pci 0000:00:01.0: bridge configuration invalid ([bus 00-00]), reconfiguring
[    4.025756] pci 0000:00:02.0: bridge configuration invalid ([bus 00-00]), reconfiguring
[    4.025818] PCI: bus1: Fast back to back transfers enabled
[    4.025884] PCI: bus2: Fast back to back transfers enabled
[    4.025907] pci 0000:00:01.0: PCI bridge to [bus 01]
[    4.025915] pci 0000:00:02.0: PCI bridge to [bus 02]
[    4.026013] mv_xor f1060800.xor: Marvell shared XOR driver
[    4.078545] mv_xor f1060800.xor: Marvell XOR (Descriptor Mode): ( xor cpy intr )
[    4.079403] mv_xor f1060900.xor: Marvell shared XOR driver
[    4.138514] mv_xor f1060900.xor: Marvell XOR (Descriptor Mode): ( xor cpy intr )
[    4.165496] Serial: 8250/16550 driver, 4 ports, IRQ sharing disabled
[    4.166483] console [ttyS0] disabled
[    4.186566] f1012000.serial: ttyS0 at MMIO 0xf1012000 (irq = 23, base_baud = 12500000) is a 16550A
[    4.979776] console [ttyS0] enabled
[    5.003825] f1012100.serial: ttyS1 at MMIO 0xf1012100 (irq = 24, base_baud = 12500000) is a 16550A
[    5.017650] brd: module loaded
[    5.023360] loop: module loaded
[    5.026853] ahci-mvebu f10a8000.sata: AHCI 0001.0000 32 slots 2 ports 6 Gbps 0x3 impl platform mode
[    5.035960] ahci-mvebu f10a8000.sata: flags: 64bit ncq sntf led only pmp fbs pio slum part sxs
[    5.045350] scsi host0: ahci-mvebu
[    5.048988] scsi host1: ahci-mvebu
[    5.052544] ata1: SATA max UDMA/133 mmio [mem 0xf10a8000-0xf10a9fff] port 0x100 irq 42
[    5.060509] ata2: SATA max UDMA/133 mmio [mem 0xf10a8000-0xf10a9fff] port 0x180 irq 42
[    5.068871] pxa3xx-nand f10d0000.flash: This platform can't do DMA on this device
[    5.076697] nand: device found, Manufacturer ID: 0xad, Chip ID: 0xda
[    5.083091] nand: Hynix H27U2G8F2CTR-BC
[    5.086943] nand: 256 MiB, SLC, erase size: 128 KiB, page size: 2048, OOB size: 64
[    5.094552] pxa3xx-nand f10d0000.flash: ECC strength 16, ECC step size 2048
[    5.101731] Bad block table found at page 131008, version 0x01
[    5.107909] Bad block table found at page 130944, version 0x01
[    5.113993] 8 ofpart partitions found on MTD device pxa3xx_nand-0
[    5.120118] Creating 8 MTD partitions on "pxa3xx_nand-0":
[    5.125541] 0x000000000000-0x000000500000 : "U-Boot"
[    5.131046] 0x000000500000-0x000000a00000 : "uImage"
[    5.136492] 0x000000a00000-0x000000f00000 : "uRamdisk"
[    5.142117] 0x000000f00000-0x00000c800000 : "image.cfs"
[    5.148259] 0x00000c800000-0x00000d700000 : "rescue fw"
[    5.154008] 0x00000d700000-0x00000eb00000 : "config"
[    5.159534] 0x00000eb00000-0x00000f500000 : "reserve1"
[    5.165175] 0x00000f500000-0x00000ff00000 : "reserve2"
[    5.171178] libphy: Fixed MDIO Bus: probed
[    5.175461] libphy: orion_mdio_bus: probed
[    5.181389] mvneta f1034000.ethernet eth0: Using hardware mac address 00:50:43:33:36:16
[    5.189531] usbcore: registered new interface driver rtl8150
[    5.195245] usbcore: registered new interface driver r8152
[    5.200792] usbcore: registered new interface driver ipheth
[    5.206388] ehci_hcd: USB 2.0 'Enhanced' Host Controller (EHCI) Driver
[    5.212946] ehci-pci: EHCI PCI platform driver
[    5.217435] ehci-orion: EHCI orion driver
[    5.221718] xhci-hcd f10f0000.usb3: xHCI Host Controller
[    5.227064] xhci-hcd f10f0000.usb3: new USB bus registered, assigned bus number 1
[    5.234651] xhci-hcd f10f0000.usb3: hcc params 0x0a000990 hci version 0x100 quirks 0x00010010
[    5.243239] xhci-hcd f10f0000.usb3: irq 45, io mem 0xf10f0000
[    5.249417] hub 1-0:1.0: USB hub found
[    5.253200] hub 1-0:1.0: 1 port detected
[    5.257311] xhci-hcd f10f0000.usb3: xHCI Host Controller
[    5.262713] xhci-hcd f10f0000.usb3: new USB bus registered, assigned bus number 2
[    5.270284] usb usb2: We don't know the algorithms for LPM for this host, disabling LPM.
[    5.278763] hub 2-0:1.0: USB hub found
[    5.282543] hub 2-0:1.0: 1 port detected
[    5.286702] xhci-hcd f10f8000.usb3: xHCI Host Controller
[    5.292062] xhci-hcd f10f8000.usb3: new USB bus registered, assigned bus number 3
[    5.299643] xhci-hcd f10f8000.usb3: hcc params 0x0a000990 hci version 0x100 quirks 0x00010010
[    5.308228] xhci-hcd f10f8000.usb3: irq 46, io mem 0xf10f8000
[    5.314355] hub 3-0:1.0: USB hub found
[    5.318143] hub 3-0:1.0: 1 port detected
[    5.322264] xhci-hcd f10f8000.usb3: xHCI Host Controller
[    5.327612] xhci-hcd f10f8000.usb3: new USB bus registered, assigned bus number 4
[    5.335182] usb usb4: We don't know the algorithms for LPM for this host, disabling LPM.
[    5.343647] hub 4-0:1.0: USB hub found
[    5.347427] hub 4-0:1.0: 1 port detected
[    5.351651] usbcore: registered new interface driver usb-storage
[    5.357831] mousedev: PS/2 mouse device common for all mice
[    5.363843] armada38x-rtc f10a3800.rtc: rtc core: registered f10a3800.rtc as rtc0
[    5.371514] i2c /dev entries driver
[    5.386011] orion_wdt: Initial timeout 171 sec
[    5.390554] md: linear personality registered for level -1
[    5.396063] md: raid0 personality registered for level 0
[    5.401402] md: raid1 personality registered for level 1
[    5.406734] md: raid10 personality registered for level 10
[    5.412332] md: raid6 personality registered for level 6
[    5.417665] md: raid5 personality registered for level 5
[    5.423012] md: raid4 personality registered for level 4
[    5.428461] device-mapper: ioctl: 4.35.0-ioctl (2016-06-23) initialised: dm-devel@redhat.com
[    5.436996] device-mapper: raid: Loading target version 1.9.1
[    5.442932] ledtrig-cpu: registered to indicate activity on CPUs
[    5.449664] marvell-cesa f1090000.crypto: CESA device successfully registered
[    5.456989] usbcore: registered new interface driver usbhid
[    5.462601] usbhid: USB HID core driver
[    5.466493] usbcore: registered new interface driver r8712u
[    5.472244] Netfilter messages via NETLINK v0.30.
[    5.476971] nfnl_acct: registering with nfnetlink.
[    5.481861] nf_conntrack version 0.5.0 (16384 buckets, 65536 max)
[    5.488148] NET: Registered protocol family 10
[    5.493048] sit: IPv6, IPv4 and MPLS over IPv4 tunneling driver
[    5.499342] NET: Registered protocol family 17
[    5.503818] bridge: automatic filtering via arp/ip/ip6tables has been deprecated. Update your scripts to load br_netfilter if you need this.
[    5.516496] Bridge firewalling registered
[    5.520526] 8021q: 802.1Q VLAN Support v1.8
[    5.524795] ThumbEE CPU extension supported.
[    5.529090] Registering SWP/SWPB emulation handler
[    5.534683] Btrfs loaded, crc32c=crc32c-generic, assert=on
[    5.540479] armada38x-rtc f10a3800.rtc: setting system clock to 2016-12-04 17:18:49 UTC (1480871929)
[    5.578180] ata2: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[    5.584404] ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[    5.592119] ata1.00: ATA-8: ST9500530NS,     DA03, max UDMA/133
[    5.598064] ata1.00: 976773168 sectors, multi 0: LBA48 NCQ (depth 31/32)
[    5.607116] ata1.00: configured for UDMA/133
[    5.611639] scsi 0:0:0:0: Direct-Access     ATA      ST9500530NS      DA03 PQ: 0 ANSI: 5
[    5.612164] ata2.00: ATA-8: WDC WD5001ABYS-18YNA0, 59.01D01, max UDMA/133
[    5.612168] ata2.00: 976773168 sectors, multi 0: LBA48 NCQ (depth 31/32)
[    5.613020] ata2.00: configured for UDMA/133
[    5.708519] sd 0:0:0:0: [sda] 976773168 512-byte logical blocks: (500 GB/466 GiB)
[    5.708554] sd 0:0:0:0: Attached scsi generic sg0 type 0
[    5.721465] sd 0:0:0:0: [sda] Write Protect is off
[    5.721531] scsi 1:0:0:0: Direct-Access     ATA      WDC WD5001ABYS-1 1D01 PQ: 0 ANSI: 5
[    5.734451] sd 0:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
[    5.767961]  sda: sda1
[    5.770790] sd 0:0:0:0: [sda] Attached SCSI removable disk
[    5.808460] sd 1:0:0:0: [sdb] 976773168 512-byte logical blocks: (500 GB/466 GiB)
[    5.808569] sd 1:0:0:0: Attached scsi generic sg1 type 0
[    5.821410] sd 1:0:0:0: [sdb] Write Protect is off
[    5.826264] sd 1:0:0:0: [sdb] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
[    5.840822]  sdb: sdb1
[    5.843610] sd 1:0:0:0: [sdb] Attached SCSI removable disk
[    5.849696] Freeing unused kernel memory: 1024K (c0b00000 - c0c00000)
mdadm: /dev/md/0 has been started with 2 drives.
### Press any key to stop and run shell... (2)Something went wrong. Dropping to a shell.
MAC set to 00:90:A9:ED:8A:A9
udhcpc: started, v1.25.1
Setting IP address 0.0.0.0 on eth0
udhcpc: sending discover
udhcpc: sending discover
udhcpc: sending discover
udhcpc: sending select for 192.168.0.101
udhcpc: lease of 192.168.0.101 obtained, lease time 864000
Setting IP address 192.168.0.101 on eth0
Deleting routers
route: SIOCDELRT: No such process
Adding router 192.168.0.100
Recreating /etc/resolv.conf
 Adding DNS server 192.168.0.100

Ok, I finally figured the problem out. When booting via the reset button from usb, the load adresses are different from booting via tftp and the size of the kernel seems to decide if it works or not. Booting the kernels via tftp works with your new one and also with the ones I compiled myself.

Thanks to this I managed to expose the red and blue hdd leds with the following change to your dts file.

mm@ovid:/mnt/src$ diff -u armada-385-wd.dts.org armada-385-wd.dts
--- armada-385-wd.dts.org       2016-12-05 14:56:03.192605998 +0100
+++ armada-385-wd.dts   2016-12-05 16:28:03.334899889 +0100
@@ -219,5 +219,34 @@

        };

+       gpio-leds {
+               compatible = "gpio-leds";
+               pinctrl-0 = <&sata_leds>;
+               pinctrl-names = "default";
+
+               red-sata1-led {
+                       label = "wdmcex2u:red:hdd1";
+                       gpios = <&gpio1 11 GPIO_ACTIVE_HIGH>;
+               };
+               red-sata2-led {
+                       label = "wdmcex2u:red:hdd2";
+                       gpios = <&gpio1 20 GPIO_ACTIVE_HIGH>;
+               };
+               blue-sata1-led {
+                       label = "wdmcex2u:blue:hdd1";
+                       gpios = <&gpio1 21 GPIO_ACTIVE_HIGH>;
+               };
+               blue-sata2-led {
+                       label = "wdmcex2u:blue:hdd2";
+                       gpios = <&gpio1 22 GPIO_ACTIVE_HIGH>;
+               };
+       };
+
 };

+&pinctrl {
+       sata_leds: sata-leds {
+               marvell,pins = "mpp43", "mpp52", "mpp53", "mpp54";
+               marvell,function = "gpio";
+       };
+};

good to know.
And thanks, I'll add your patch to my kernel and update my github.
Now the reset and WoL is missing :-)

Cheers Carl



Edited 1 time(s). Last edit at 12/05/2016 02:12PM by Peacemaker.

I soldered two usb-serial adapters to the Welltrend micro to capture the reboot and hibernate sequence, since my tries with strace in the original firmware failed, because strace was killed too early.

Lines starting with > are sent to the micro, lines starting wit < are the answers of the micro.
The numbers in the begining are epoch % 1000 and milliseconds.

Here is what gets send to the micro controller in case of a reboot via web-interface:

>  694.131  0xfa 0x02 0x00 0x5a 0x00 0x00 0xfb
<  694.134  0xfa 0x30 0x00 0x00 0x00 0x00 0xfb

>  701.099  0xfa 0x26 0x00 0x04 0x00 0x01 0xfb
<  701.102  0xfa 0x30 0x00 0x00 0x00 0x00 0xfb

>  701.403  0xfa 0x26 0x00 0x21 0x00 0x01 0xfb
<  701.406  0xfa 0x30 0x00 0x00 0x00 0x00 0xfb

>  702.138  0xfa 0x02 0x00 0x00 0x00 0x00 0xfb
<  702.140  0xfa 0x30 0x00 0x00 0x00 0x00 0xfb

>  704.158  0xfa 0x02 0x00 0x5a 0x00 0x00 0xfb
<  704.160  0xfa 0x30 0x00 0x00 0x00 0x00 0xfb

>  705.142  0xfa 0x03 0x03 0x02 0x01 0x0a 0xfb
<  705.145  0xfa 0x30 0x00 0x00 0x00 0x00 0xfb

and here is what gets sent to the micro in case of using "hibernate" in the web-interface:

>  915.070  0xfa 0x02 0x00 0x5a 0x00 0x00 0xfb 
<  915.072  0xfa 0x30 0x00 0x00 0x00 0x00 0xfb 

>  925.096  0xfa 0x02 0x00 0x5a 0x00 0x00 0xfb 
<  925.098  0xfa 0x30 0x00 0x00 0x00 0x00 0xfb 

>  935.121  0xfa 0x03 0x08 0x00 0x00 0x00 0xfb 
<  935.124  0xfa 0x03 0x08 0x00 0x00 0x6f 0xfb 
<  935.128  0xfa 0x30 0x00 0x00 0x00 0x00 0xfb 

>  936.127  0xfa 0x02 0x00 0x5a 0x00 0x00 0xfb 
<  936.130  0xfa 0x30 0x00 0x00 0x00 0x00 0xfb 

>  938.314  0xfa 0x26 0x00 0x01 0x00 0x01 0xfb 
<  938.316  0xfa 0x30 0x00 0x00 0x00 0x00 0xfb 

>  938.618  0xfa 0x26 0x00 0x04 0x00 0x01 0xfb 
<  938.620  0xfa 0x30 0x00 0x00 0x00 0x00 0xfb 

>  938.922  0xfa 0x26 0x00 0x04 0x00 0x01 0xfb 
<  938.925  0xfa 0x30 0x00 0x00 0x00 0x00 0xfb 

>  939.227  0xfa 0x26 0x00 0x01 0x00 0x01 0xfb 
<  939.229  0xfa 0x30 0x00 0x00 0x00 0x00 0xfb 

>  939.531  0xfa 0x03 0x03 0x01 0x01 0x01 0xfb 
<  939.533  0xfa 0x30 0x00 0x00 0x00 0x00 0xfb

I can wake the ex2 up with etherwake <MAC of ex2> after this.

The original firmware contains an extra binary for wakeonlan. After executing this manually in the original firmware, network connectivity is lost until I reboot the device.

Maybe this helps. Do you have your modified sources of the mcm-daemon somwhere?

I also observed, that when executing "service mcm-daemon stop" in Debian, the device immediatelly shuts off.

Woow, nice work with the soldering :-),
I'll upload the mcm-daemon to my github (you'll be welcome to contribute there:-))

Quote
I also observed, that when executing "service mcm-daemon stop" in Debian, the device immediatelly shuts off.

I know, thats why it should be killed as the last process. Somehow there might be a delay in the code, but this delay is not working :-/

But you can have a look at the sources of A Boehler https://www.aboehler.at/hg/dns320l-daemon, I have only changed a few things, yet.
But over all he uses inipraser, what I don't like as this is no standard debian / ubuntu / arch package and somethings I would change as well, as the control over telnet.



Edited 1 time(s). Last edit at 12/07/2016 01:47PM by Peacemaker.

Well, I aquired a third USB-serial adapter so I'm back in business with the serial console and booting the debian image.

I was not able to get wakeonlan working with debian by using the sniffed commands.

echo -e "\xfa\x26\x00\x01\x00\x01\xfb" > /dev/ttyS1
echo -e "\xfa\x26\x00\x04\x00\x01\xfb" > /dev/ttyS1
echo -e "\xfa\x26\x00\x04\x00\x01\xfb" > /dev/ttyS1
echo -e "\xfa\x26\x00\x01\x00\x01\xfb" > /dev/ttyS1
echo -e "\xfa\x03\x03\x01\x01\x01\xfb" > /dev/ttyS1

The device powers down and there seems no way to wake it, except by powercycling.

When looking at the original firmware, there are three ethernet interfaces available on the system. The main interface is egiga0 and there are two other interfaces named egiga1 and egiga2. I'm wondering if they might have to do something with the wakeonlan stuff.

I did an strace of the set_wol command in the original firmware:

execve("./usrsbin/set_wol", ["./usrsbin/set_wol"], [/* 8 vars */]) = 0
brk(0)                                  = 0x28000
uname({sys="Linux", node="MyCloudEX2Uegiga0: ltra", ...}) = 0link down

access("/etc/ld.so.preload", R_OK)      = -1 ENOENT (No such file or directoryegiga0: )
open("/etc/ldlink up.so.cache", O_RD, full duplexONLY|O_CLOEXEC) , speed = 3
fstat64(3, 1 Gbps
{st_mode=S_IFREG|0644, st_size=15905, ...}) = 0
mmap2(NULL, 15905, PROT_READ, MAP_PRIVATE, 3, 0) = 0xb6870000
close(3)                                = 0
open("/lib/libshare.so", O_RDONLY|O_CLOEXEC) = 3
read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0(\0\1\0\0\0 \"\0\0004\0\0\0"..., 512) = 512
fstat64(3, {st_mode=S_IFREG|0755, st_size=43967, ...}) = 0
mmap2(NULL, 107988, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE, 3, 0) = 0xb6850000
mprotect(0xb6860000, 32768, PROT_NONE)  = 0
mmap2(0xb6868000, 32768, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x40000) = 0xb6868000
close(3)                                = 0
open("/lib/libxmldbc.so", O_RDONLY|O_CLOEXEC) = 3
read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0(\0\1\0\0\0$\17\0\0004\0\0\0"..., 512) = 512
fstat64(3, {st_mode=S_IFREG|0755, st_size=11928, ...}) = 0
mmap2(NULL, 76052, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE, 3, 0) = 0xb6838000
mprotect(0xb6840000, 32768, PROT_NONE)  = 0
mmap2(0xb6848000, 32768, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0) = 0xb6848000
close(3)                                = 0
open("/lib/libgcc_s.so.1", O_RDONLY|O_CLOEXEC) = 3
read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0(\0\1\0\0\0H&\0\0004\0\0\0"..., 512) = 512
fstat64(3, {st_mode=S_IFREG|0644, st_size=40716, ...}) = 0
mmap2(NULL, 104980, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE, 3, 0) = 0xb6818000
mprotect(0xb6828000, 32768, PROT_NONE)  = 0
mmap2(0xb6830000, 32768, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x40000) = 0xb6830000
close(3)                                = 0
open("/lib/libc.so.6", O_RDONLY|O_CLOEXEC) = 3
read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0(\0\1\0\0\0\350|\1\0004\0\0\0"..., 512) = 512
fstat64(3, {st_mode=S_IFREG|0755, st_size=1221784, ...}) = 0
mmap2(NULL, 1295744, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE, 3, 0) = 0xb66d8000
mprotect(0xb6800000, 32768, PROT_NONE)  = 0
mmap2(0xb6808000, 65536, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x900000) = 0xb6808000
close(3)                                = 0
set_tls(0xb68aa880, 0xb68aaf58, 0xb68a8050, 0xb68aa880, 0xb68a8050) = 0
mprotect(0xb6808000, 32768, PROT_READ)  = 0
mprotect(0xb68a0000, 32768, PROT_READ)  = 0
munmap(0xb6870000, 15905)               = 0
socket(PF_LOCAL, SOCK_STREAM, 0)        = 3
fcntl64(3, F_SETFD, FD_CLOEXEC)         = 0
connect(3, {sa_family=AF_LOCAL, sun_path="/var/run/xmldb_sock"}, 110) = 0
send(3, "\1\0\34\0\0\0\0\0\0\0\1\0\340w", 14, MSG_NOSIGNAL) = 14
send(3, "/network_mgr/bonding/enable\0", 28, MSG_NOSIGNAL) = 28
select(4, [3], NULL, NULL, NULL)        = 1 (in [3])
read(3, "0\0", 1024)                    = 2
close(3)                                = 0
brk(0)                                  = 0x28000
brk(0x50000)                            = 0x50000
open("/tmp/mac_addr", O_RDONLY|O_LARGEFILE) = 3
fstat64(3, {st_mode=S_IFREG|0644, st_size=18, ...}) = 0
mmap2(NULL, 32768, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0xb6870000
read(3, "00:90:A9:EA:45:B2\n", 32768)   = 18
close(3)                                = 0
munmap(0xb6870000, 32768)               = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 3
ioctl(3, SIOCDEVPRIVATE, 0xbeb77c60)    = 0
close(3)                                = 0
nanosleep({0, 1000000}, NULL)           = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 3
ioctl(3, SIOCDEVPRIVATE, 0xbeb77c60)    = 0
close(3)                                = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 3
ioctl(3, SIOCDEVPRIVATE, 0xbeb77c60)    = 0
close(3)                                = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 3
ioctl(3, SIOCDEVPRIVATE, 0xbeb77c60)    = 0
close(3)                                = 0
nanosleep({0, 10000000}, NULL)          = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 3
ioctl(3, SIOCDEVPRIVATE, 0xbeb77c60)    = 0
close(3)                                = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 3
ioctl(3, SIOCDEVPRIVATE, 0xbeb77c60)    = 0
close(3)                                = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 3
ioctl(3, SIOCDEVPRIVATE, 0xbeb77c60)    = 0
close(3)                                = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 3
ioctl(3, SIOCDEVPRIVATE, 0xbeb77c60)    = 0
close(3)                                = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 3
ioctl(3, SIOCDEVPRIVATE, 0xbeb77c60)    = 0
close(3)                                = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 3
ioctl(3, SIOCDEVPRIVATE, 0xbeb77c60)    = 0
close(3)                                = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 3
ioctl(3, SIOCDEVPRIVATE, 0xbeb77c60)    = 0
close(3)                                = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 3
ioctl(3, SIOCDEVPRIVATE, 0xbeb77c60)    = 0
close(3)                                = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 3
ioctl(3, SIOCDEVPRIVATE, 0xbeb77c60)    = 0
close(3)                                = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 3
ioctl(3, SIOCDEVPRIVATE, 0xbeb77c60)    = 0
close(3)                                = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 3
ioctl(3, SIOCDEVPRIVATE, 0xbeb77c60)    = 0
close(3)                                = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 3
ioctl(3, SIOCDEVPRIVATE, 0xbeb77c60)    = 0
close(3)                                = 0
exit_group(0)                           = ?
+++ exited with 0 +++

Unfortunately I can make no sense out of it. It connects to the xmldb where it probably just changes the network status and does a series of private ioctl. I have no idea on how to find out what they are doing.

When executing the command manually in the original firmware, I get the following message written on the console:

egiga0: link down
egiga0: link up, full duplex, speed 1 Gbps

I can then echo the above sequence to the micro, the device shuts down and I can wake it up with etherwake.

If I don't use the command and send the sequence to the micro, the device reboots, when the ethernet cable is connected and shuts down when the cable is not connected, but powers on when I plug the cable back in. So this is quite strange and I have no more ideas on how to continue.

However rebooting the device with:

echo -e "\xfa\x03\x03\x02\x01\x0a\xfb" > /dev/ttyS1

works reliable in debian (and the original firmware). It shuts the power of for about half a second and the device reboots.

I guess the direction to continue is to make the device shutdown or reboot reliable in debian and forget about the wol stuff for now.

Quote
MM
I guess the direction to continue is to make the device shutdown or reboot reliable in debian and forget about the wol stuff for now.

Yes, I fully agree to you. I see 2 - 3 ways to do so.

1. quick and dirty: 2 bash-scripts add them to the systemd shutdown and reboot command.
Maybe I'll do it for first to get shit down :-) + combine this with a new and better fan script and get rid of the command-daemon

2. rewrite the whole command script,.. lots of work and I'll realy poor in coding C, and not the best way.

3. Build Kernelmodul for the weltrend within Fan- ,RCT- , Powercontrol.
In my oppinon it would be the best way but I have realy no idea how to do that.

So if you have an other idea, or like to help with it go for it :-D

Otherwise: I'm now trying getting raid booting and an installscript working from the initrd as well as unlock an encrypted bootdevice via ssh

-------------------------------------------------------------------------
GitHub https://github.com/cschil



Edited 1 time(s). Last edit at 12/08/2016 03:38PM by Peacemaker.

Hi Carl,

That sounds good, especially the encrypted boot-stuff via ssh (if you are bored, please include lvm :-).

I'd say the shortest option to get it to a at least usable state is 1) if it is possible. I'm not a particular fan of systemd which is now default in jessie and later and don't know much about it. I guess I'll have to learn it.

I'll have a look into 3) next week if I can find the hooks in the kernel to tap into for reboot and shutdown. I won't have time until Tuesday and it will probably take quite some time to figure things out. I wrote a very small kernel-module controlling an MCU about 12 years ago but have forgotten most of it.

As we have a sort of working fan solution, I'd try to implement reboot and shutdown first and later extend it to the fan, if at all. Userspace control of the fan isn't too bad after all. Look at all the fan-control stuff, replacing the bios on laptops. It's easier to adapt to the preferences of the user. I have no clear opinion on that yet.

I ordered a second ex2u since I want to keep the one with all the soldered wires for future research with the WD firmware. It should arrive on Monday and I'll start flashing kernels / ramdisks on that one.

Have a nice weekend.
Martin

Hi Carl,

ok, I'm back and the new ex2u arrived at my neighbor today. I did some research over the weekend online concerning kernel modules and according to what I found out this is probably no task for a kernel module. The reasons mentioned:

* What can be done in userspace should be done in userspace.
* Performance is not an issue.
* We have a working driver which already claims the device and controls it correctly (aka serial_io).
* serial access from within the kernel over the existing driver is possible but cumbersome as one would need to synchronize kernel access to the device with possible userspace access to the device.

The new module would be conflicting with the existing serial driver, which can't unloaded since it is need for the serial console. So after this research it's probably better to stick to the existing approach.

Did you find out something new regarding systemd and the possibility to execute commands that trigger shutdown/reboot or do you have strong reasons, why the hassle to claim the device from the existing driver would be worth it?

bye
Martin

Hey Martin,
I didn't do much for the NAS this weekend as well :-) (fixing my crappy pogoplug was enough).

Quote
MM
* What can be done in userspace should be done in userspace.
* We have a working driver which already claims the device and controls it correctly (aka serial_io).
* serial access from within the kernel over the existing driver is possible but cumbersome as one would need to synchronize kernel access to the device with possible userspace access to the device.

hm,.. I see the point, but still but let's say I'm not very happy with this.
To be clear I didn't thought about putting more control from userspace to kernel. I thought providing (virtual) devices by the kernel, that they can be access and controled from userspace: e.g. fan + temp as hwmon dev.
I think the great deal there would be: every distribution would work regardless systemd or not.

I thought we could pin the second serial to the "Weltrend" device via the *.dts and avoid so the conflicts beween serial_io / userspace <-> kernelmodule

Quote
MM
Did you find out something new regarding systemd and the possibility to execute commands that trigger shutdown/reboot or do you have strong reasons, why the hassle to claim the device from the existing driver would be worth it?

If I get it right: in the orignal control daemon (its for a dlink dns327 and WD MCM gen1), there the Shutdown Signal should be delayed for 10 sec after the daemon is killed.
I have no idea how this delay work and if or how it "desides" between reboot and shutdown, main problem is maybe I don't know how the inipraser used there is working.
To workaround the "missing" delay I switched in the systemd file ( /etc/systemd/system/mcm-daemon.service) the dependency so that the process is killed as the last process and so poweroff the device -> but reboot does not work.

Get systemd shutdown / reboot better working the easy way would be to add a new programm (or 2) to /etc/systemd/system/reboot.target.wants and /etc/systemd/system/shutdown.target.wants to send the right signal at the right time.

But as I can't patch the mcm-daemon to this there is need for a new and better one without telnet^^. (for fancontrol we could use a python script for dns327 whats working as well)

So my hope with kernel-module was: we do not need to frickle arround with all of this as the kernel do the reboot, halt, that the rtc and so on,...

-------------------------------------------------------------------------
GitHub https://github.com/cschil

Hi Carl,

Quote
Peacemaker
-------------------------------------------------------
> So my hope with kernel-module was: we do not need
> to frickle arround with all of this as the kernel
> do the reboot, halt, that the rtc and so on,...

Ok, can you explain what the rtc has to do with that? As far as I understood the hardware, there is a battery for the marvel rtc on the board, which is supported by the kernel and seems to get used (at least with the plain debian stretch testinstall I did with my kernel). So what does the rtc do, that supposedly exists in the MCU?

For normal use the rtc in the mcu does not matter, so you a right.
But to get the alarmclock / rtc-wakeup working the mcu-rtc has to work, and be set correct.

-------------------------------------------------------------------------
GitHub https://github.com/cschil

Hi Carl,

I got my second ex2u modified for serial console and started to try booting flashed kernels.

With that I experienced the same problem as when booting from usb, that the kernel size matters if it succeeds or not.

You can check if the kernel would work from flash by using the following adresses:

tftp 0x00a00000 uImage ; tftp 0x00f00000 uRamdisk.jessie-sd; bootm 0x00a00000 0x00f00000

I have now a device that is booting from flash and will start now looking into the mcm-daemon.

bye
Martin

MM Wrote:
-------------------------------------------------------
> Hi Carl,
>
> I got my second ex2u modified for serial console
> and started to try booting flashed kernels.
>
> With that I experienced the same problem as when
> booting from usb, that the kernel size matters if
> it succeeds or not.
>
> You can check if the kernel would work from flash
> by using the following adresses:
>
>

> tftp 0x00a00000 uImage ; tftp 0x00f00000
> uRamdisk.jessie-sd; bootm 0x00a00000 0x00f00000
>

>

Okay, so your right but, do you have a clou why this is a problem?

As with the different adress kernel boots and with wd-adresses it's loading but boot fails?

-------------------------------------------------------------------------
GitHub https://github.com/cschil



Edited 2 time(s). Last edit at 12/17/2016 07:41AM by Peacemaker.

Hi Carl,

yes I have an idea, but didn't verify it exactly. The kernel get's loaded at 0x00a00000 and is uncompressing itself to 0x00008000. Witch means, that the kernel uncompressing itself is overwriting it's compressed image somewhere after 0x00a00000 - 0x00008000 = 10452992 bytes. Since the start of the compressed kernel image is already decompressed, it will happen somwhere between 10452992 and 10452992+imagesize bytes.

BTW. I'm trying to handle power off and restart in the kernel at the moment, but it keeps OOPSing. However my kernel function is called correctly, it just don't work yet since it crashes when I try to write to the serial port registers.

bye
Martin

Okay, this makes sens and seems to be right:
With kernelsetting: optimized for size it's booting :-)

Cheers Carl

-------------------------------------------------------------------------
GitHub https://github.com/cschil

Hi Carl,

ok here is my restart/poweroff driver. It works for me at least. you need to disable or kill (with -9) the mcm-daemon first.

diff -r -u -N linux-4.8.12.org/drivers/power/reset/Kconfig linux-4.8.12/drivers/power/reset/Kconfig
--- linux-4.8.12.org/drivers/power/reset/Kconfig        2016-12-02 09:11:45.000000000 +0100
+++ linux-4.8.12/drivers/power/reset/Kconfig    2016-12-16 19:37:39.787988518 +0100
@@ -120,6 +120,15 @@

          Say Y if you have a QNAP NAS.

+config POWER_RESET_WDEX2U
+       bool "WDMC EX2 ultra power-off driver"
+       depends on OF_GPIO && PLAT_ORION
+       help
+         This driver supports turning off WD MyCloud EX2 ultra NAS devices by
+          sending commands to the microcontroller which controls the main power.
+
+         Say Y if you have a WD MyCloud EX2 ultra
+
 config POWER_RESET_RESTART
        bool "Restart power-off driver"
        help
diff -r -u -N linux-4.8.12.org/drivers/power/reset/Makefile linux-4.8.12/drivers/power/reset/Makefile
--- linux-4.8.12.org/drivers/power/reset/Makefile       2016-12-02 09:11:45.000000000 +0100
+++ linux-4.8.12/drivers/power/reset/Makefile   2016-12-16 19:38:06.395973742 +0100
@@ -12,6 +12,7 @@
 obj-$(CONFIG_POWER_RESET_MSM) += msm-poweroff.o
 obj-$(CONFIG_POWER_RESET_LTC2952) += ltc2952-poweroff.o
 obj-$(CONFIG_POWER_RESET_QNAP) += qnap-poweroff.o
+obj-$(CONFIG_POWER_RESET_WDEX2U) += wt-poweroff.o
 obj-$(CONFIG_POWER_RESET_RESTART) += restart-poweroff.o
 obj-$(CONFIG_POWER_RESET_ST) += st-poweroff.o
 obj-$(CONFIG_POWER_RESET_VERSATILE) += arm-versatile-reboot.o
diff -r -u -N linux-4.8.12.org/drivers/power/reset/wt-poweroff.c linux-4.8.12/drivers/power/reset/wt-poweroff.c
--- linux-4.8.12.org/drivers/power/reset/wt-poweroff.c  1970-01-01 01:00:00.000000000 +0100
+++ linux-4.8.12/drivers/power/reset/wt-poweroff.c      2016-12-17 17:21:42.423378737 +0100
@@ -0,0 +1,184 @@
+/*
+ * Poweroff and restart driver for the Western Digital 
+ * MyCloud EX2 Ultra NAS. It could be easily reused for
+ * other devices with a serial connected MCU.
+ *
+ * Copyright (C) 2016 Martin Mueller <wdmc@sig21.net>
+ *
+ * Based on the code from:
+ *
+ * Copyright (C) 2012  Andrew Lunn <andrew@lunn.ch>
+ * Copyright (C) 2009  Martin Michlmayr <tbm@cyrius.com>
+ * Copyright (C) 2008  Byron Bradley <byron.bbradley@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version
+ * 2 of the License, or (at your option) any later version.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/serial_reg.h>
+#include <linux/kallsyms.h>
+#include <linux/of.h>
+#include <linux/io.h>
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <asm/system_misc.h>
+
+#define UART_REG(x)    (base + ((UART_##x) << 2))
+
+struct power_mcu_cmds {
+       u32 baud;
+       char *poweroff;
+       char *restart;
+};
+
+static const struct power_mcu_cmds wdmcex2u_power_mcu_cmds = {
+       .baud = 19200,
+       .poweroff = "\xfa\x03\x03\x01\x01\x01\xfb",
+       .restart  = "\xfa\x03\x03\x02\x01\x0a\xfb",
+};
+
+static const struct of_device_id wt_power_off_of_match_table[] = {
+       { .compatible = "wdmcex2u,power-off",
+         .data = &wdmcex2u_power_mcu_cmds,
+       },
+       {}
+};
+MODULE_DEVICE_TABLE(of, wt_power_off_of_match_table);
+
+static void __iomem *base;
+static unsigned long tclk;
+static const struct power_mcu_cmds *cfg;
+
+static void *pm_power_off_org;
+static void *arm_pm_restart_org;
+
+static void wt_power_off(void)
+{
+       const unsigned divisor = ((tclk + (8 * cfg->baud)) / (16 * cfg->baud));
+       uint8_t i;
+
+       pr_err("%s: triggering power-off...\n", __func__);
+
+       /* hijack UART and reset into sane state */
+       writel(0x83, UART_REG(LCR));
+       writel(divisor & 0xff, UART_REG(DLL));
+       writel((divisor >> 8) & 0xff, UART_REG(DLM));
+       writel(0x03, UART_REG(LCR));
+       writel(0x00, UART_REG(IER));
+       writel(0x00, UART_REG(FCR));
+       writel(0x00, UART_REG(MCR));
+
+       /* send the power-off command to PIC */
+       for (i=0; i<strlen(cfg->poweroff); i++) {
+         writel(cfg->poweroff, UART_REG(TX));
+         mdelay(2);
+       }
+       mdelay(1000);
+       pr_err("%s: triggering power-off failed!\n", __func__);
+}
+
+static void wt_restart(enum reboot_mode reboot_mode, const char *cmd)
+{
+       const unsigned divisor = ((tclk + (8 * cfg->baud)) / (16 * cfg->baud));
+       uint8_t i;
+
+       pr_err("%s: triggering restart...\n", __func__);
+
+       /* hijack UART and reset into sane state */
+       writel(0x83, UART_REG(LCR));
+       writel(divisor & 0xff, UART_REG(DLL));
+       writel((divisor >> 8) & 0xff, UART_REG(DLM));
+       writel(0x03, UART_REG(LCR));
+       writel(0x00, UART_REG(IER));
+       writel(0x00, UART_REG(FCR));
+       writel(0x00, UART_REG(MCR));
+
+       /* send the restart command to PIC */
+       for (i=0; i<strlen(cfg->restart); i++) {
+         writel(cfg->restart, UART_REG(TX));
+         mdelay(2);
+       }
+       mdelay(1000);
+       pr_err("%s: triggering restart failed!\n", __func__);
+}
+
+static int wt_power_off_probe(struct platform_device *pdev)
+{
+       struct device_node *np = pdev->dev.of_node;
+       struct resource *res;
+       struct clk *clk;
+       char symname[KSYM_NAME_LEN];
+
+       const struct of_device_id *match =
+               of_match_node(wt_power_off_of_match_table, np);
+       cfg = match->data;
+
+       res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+       if (!res) {
+               dev_err(&pdev->dev, "Missing resource");
+               return -EINVAL;
+       }
+
+       base = devm_ioremap(&pdev->dev, res->start, resource_size(res));
+       if (!base) {
+               dev_err(&pdev->dev, "Unable to map resource");
+               return -EINVAL;
+       }
+
+       /* We need to know tclk in order to calculate the UART divisor */
+       clk = devm_clk_get(&pdev->dev, NULL);
+       if (IS_ERR(clk)) {
+               dev_err(&pdev->dev, "Clk missing");
+               return PTR_ERR(clk);
+       }
+
+       tclk = clk_get_rate(clk);
+
+       pm_power_off_org = pm_power_off;
+       if ( strlen(cfg->poweroff) ) {
+               if (pm_power_off) {
+                       lookup_symbol_name((ulong)pm_power_off, symname);
+                       dev_warn(&pdev->dev, "replacing pm_power_off (%s)", symname);
+               } else
+                       dev_notice(&pdev->dev, "installing pm_power_off handler");
+               pm_power_off = wt_power_off;
+       }
+
+       arm_pm_restart_org = arm_pm_restart;
+       if ( strlen(cfg->restart) ) {
+               if (arm_pm_restart) {
+                       lookup_symbol_name((ulong)arm_pm_restart, symname);
+                       dev_warn(&pdev->dev, "replacing arm_pm_restart (%s)", symname);
+               } else
+                       dev_notice(&pdev->dev, "installing arm_pm_restart handler");
+               arm_pm_restart = wt_restart;
+       }
+
+       return 0;
+}
+
+static int wt_power_off_remove(struct platform_device *pdev)
+{
+       pm_power_off = pm_power_off_org;
+       arm_pm_restart = arm_pm_restart_org;
+       return 0;
+}
+
+static struct platform_driver wt_power_off_driver = {
+       .probe  = wt_power_off_probe,
+       .remove = wt_power_off_remove,
+       .driver = {
+               .name   = "wt_power_off",
+               .of_match_table = of_match_ptr(wt_power_off_of_match_table),
+       },
+};
+module_platform_driver(wt_power_off_driver);
+
+MODULE_AUTHOR("Martin Mueller <wdmc@sig21.net>");
+MODULE_DESCRIPTION("Welltrend MCU power off driver");
+MODULE_LICENSE("GPL v2");
--- linux-4.8.12.org/.config    2016-12-16 18:13:15.719627882 +0100
+++ linux-4.8.12/.config        2016-12-17 17:21:52.707374317 +0100
@@ -105,8 +105,8 @@
 # CONFIG_RCU_EXPEDITE_BOOT is not set
 # CONFIG_BUILD_BIN2C is not set
 # CONFIG_IKCONFIG is not set
-CONFIG_LOG_BUF_SHIFT=14
-CONFIG_LOG_CPU_MAX_BUF_SHIFT=12
+CONFIG_LOG_BUF_SHIFT=17
+CONFIG_LOG_CPU_MAX_BUF_SHIFT=14
 CONFIG_NMI_LOG_BUF_SHIFT=13
 CONFIG_GENERIC_SCHED_CLOCK=y
 CONFIG_CGROUPS=y
@@ -2447,10 +2447,11 @@
 CONFIG_POWER_RESET=y
 # CONFIG_POWER_RESET_BRCMKONA is not set
 # CONFIG_POWER_RESET_BRCMSTB is not set
-CONFIG_POWER_RESET_GPIO=y
-CONFIG_POWER_RESET_GPIO_RESTART=y
+# CONFIG_POWER_RESET_GPIO is not set
+# CONFIG_POWER_RESET_GPIO_RESTART is not set
 # CONFIG_POWER_RESET_LTC2952 is not set
 # CONFIG_POWER_RESET_QNAP is not set
+CONFIG_POWER_RESET_WDEX2U=y
 # CONFIG_POWER_RESET_RESTART is not set
 # CONFIG_POWER_RESET_VERSATILE is not set
 # CONFIG_POWER_RESET_SYSCON is not set
--- linux-4.8.12.org/arch/arm/boot/dts/armada-385-wd.dts        2016-12-14 17:48:42.099921892 +0100
+++ linux-4.8.12/arch/arm/boot/dts/armada-385-wd.dts    2016-12-17 17:21:52.447374429 +0100
@@ -17,7 +17,7 @@
 #include <dt-bindings/gpio/gpio.h>

 / {
-       model = "Marvell Armada 385 Development Board";
+       model = "Western Digital MyCloud EX2 Ultra";
        compatible = "marvell,a385-db", "marvell,armada385", "marvell,armada38x";

        chosen {
@@ -103,14 +103,23 @@
                                no-1-8-v;
                        };

-                       serial@12100 {
+                       serial@12000 {
                                status = "okay";
                        };
-      
-                       serial@12000 {
+
+                       /* ttyS1 is connected to w Welltrend 6703F-OG240WT Micocontroller
+                        * which controls power fan and other things */
+                       serial@12100 {
                                status = "okay";
                        };

+                       poweroff@12100 {
+                               compatible = "wdmcex2u,power-off";
+                               reg = <0x12100 0x100>;
+                               clocks = <&coreclk 0>;
+                               status = "okay";
+                       };
+      
                        spi0: spi@10600 {
                                status = "okay";