[DEV- MOD] [Under Development] Multiple USB drives MTCB Units - MTCB Software Development

Hey guys,
Currently i`m digging into this matter. So here's the concept:
[HU] - one of the back USB's
|
[USB Hub] - powered form an external power supply (external 5V)
| |
U1 U2 - (32GB USB Drives)
A.) Current behaviour
- only the first inserted USB drive get's mounted under /mnt/usb_storage2
- the second usb drive is recognised by the kernel and does not get mounted by the system (not assigned to any "drive letter" so to speak(i think)).
B.) Current investigations:
B.1 ) by unpacking the boot.img - fstab.rk30board.bootmode.*
Code:
/devices/platform/usb20_otg/usb /mnt/usb_storage vfat defaults voldmanaged=usb_storage:auto
/devices/platform/usb20_host/usb /mnt/usb_storage2 vfat defaults voldmanaged=usb_storage2:auto
B.2) by unpacking the boot.img - init.rk30board.rc
Code:
[13] export THIRD_VOLUME_STORAGE /mnt/usb_storage
[14] mkdir /mnt/usb_storage 0000 system system
[15] export THIRD_VOLUME_STORAGE2 /mnt/usb_storage2
[16] mkdir /mnt/usb_storage2 0000 system system
On the Application Side, for example MTCAPKInstall_source/com/hct/apkinstall/DevListAdapter.java
Code:
} else if (path.equals("/mnt/usb_storage")) {
holder.Image.setImageResource(R.drawable.usbhost);
holder.ImageName.setText(this.mContext.getString(R.string.usb1));
} else if (path.equals("/mnt/usb_storage2")) {
holder.Image.setImageResource(R.drawable.usbhost);
holder.ImageName.setText(this.mContext.getString(R.string.usb2));
This can be solved easily by creating a for, or by duplication each else if (the copy/paste way )
I need to do some more tests... i`m also interested for your opinions/experience with this (What have you tested, What worked, What did not work, Why is not working, etc..)
So cheers to everyone and keep you posted

---

So far i rulled out the usb driver, internal hub.. It has something to do with the block creation and vold...
Will dig more tomorrow(i hope)

---

okey, so baseline the issue is related to VOLD
VOLD does not generate the block device for the new devices... i`m not 100% sure that VOLD handles the block device generation
for the time being i managed to mount 3 separate partitions from the same USB storage -> by doing this i ruled out the kernel itself, the irq assignment, etc..
I need to dig deeper into VOLD and to see exactly what is the process for generating the block device eg (/dev/block/vold/8:X)
from this point it would be nice to exchange some experience with someone that really know how things work on Android, i`m an iOS user, and the HU is my first Android System

---

I can't help as haven't gone down the road of digging to deep into kernels and Roms due to other commitments..
But just want to offer words of encouragement as you are working towards a good course..
I don't use a mobile modem and so at the moment that spare usbport is doing nothing. Would be nice to load it up with a spare USB drive I have around here .
That then brings up the powering up thread which the boys have already sorted..
So once you crack this and get a kernel mode going we are good to go..
Sent from my SM-G900I using Tapatalk

---

I tried to have a look in the binary code of vold shipped with the firmware, but friday was short... I'll continue my work next week
Keep you guys posted

---

In the past i did some test with a 4 Port USB hub. I connected several USB sticks and a 350 gb harddisk. They were not recognised by rootexplorer or android itself, but ESF file Explorer did recognised all of the devices.
Another strange thing is that when i connect a USB RTL dab+ dongle on the First USB Port on my hui and the harddisk on the other usb port my GPS is not working propperly.

---

well i`m 100% sure that the hub is not the issue, because:
<<1'st usb drive>>
<6>[ 1757.992589] usb 2-1.3: new high speed USB device number 30 using usb20_host
<6>[ 1758.093203] usb 2-1.3: New USB device found, idVendor=0781, idProduct=5204
<6>[ 1758.093230] usb 2-1.3: New USB device strings: Mfr=1, Product=2, SerialNumber=3
<6>[ 1758.093282] usb 2-1.3: Product: Cruzer Crossfire
<6>[ 1758.093303] usb 2-1.3: Manufacturer: SanDisk Corporation
<6>[ 1758.093322] usb 2-1.3: SerialNumber: 000016A29874531B
<4>[ 1758.094142] --mtc usb storage 2-1.3 0781 5204
<6>[ 1758.094989] scsi11 : usb-storage 2-1.3:1.0
<5>[ 1759.092957] scsi 11:0:0:0: Direct-Access SanDisk Cruzer 4.05 PQ: 0 ANSI: 2
<5>[ 1759.094087] sd 11:0:0:0: [sda] 16055949 512-byte logical blocks: (8.22 GB/7.65 GiB)
<5>[ 1759.094577] sd 11:0:0:0: [sda] Write Protect is off
<7>[ 1759.094601] sd 11:0:0:0: [sda] Mode Sense: 03 00 00 00
<3>[ 1759.095072] sd 11:0:0:0: [sda] No Caching mode page present
<3>[ 1759.095095] sd 11:0:0:0: [sda] Assuming drive cache: write through
<3>[ 1759.098162] sd 11:0:0:0: [sda] No Caching mode page present
<3>[ 1759.098183] sd 11:0:0:0: [sda] Assuming drive cache: write through
<6>[ 1759.100316] sda: sda1
<3>[ 1759.104583] sd 11:0:0:0: [sda] No Caching mode page present
<3>[ 1759.104621] sd 11:0:0:0: [sda] Assuming drive cache: write through
<5>[ 1759.104663] sd 11:0:0:0: [sda] Attached SCSI removable disk
<7>[ 1759.846560] SELinux: initialized (dev sda1, type vfat), uses genfs_contexts
<<2'nd usb drive>>
<6>[ 1956.392593] usb 2-1.2: new high speed USB device number 31 using usb20_host
<6>[ 1956.493577] usb 2-1.2: New USB device found, idVendor=0951, idProduct=1666
<6>[ 1956.493616] usb 2-1.2: New USB device strings: Mfr=1, Product=2, SerialNumber=3
<6>[ 1956.493647] usb 2-1.2: Product: DataTraveler 3.0
<6>[ 1956.493666] usb 2-1.2: Manufacturer: Kingston
<6>[ 1956.493683] usb 2-1.2: SerialNumber: 60A44C425294BE10A9820D20
<4>[ 1956.494632] --mtc usb storage 2-1.2 0951 1666
<4>[ 1960.190620] ~ arm_rev_16bits err0 0
the GUI applications are dependent of Device Block Assignment, as you can clearly see, kernel recognises both USB Storage devices, but only the first is assigned with a block device.

---

okey, so i had a chance to check with IDA the vold binary from our units... seems that vold binary is custom than from other rom's
i will do some tests on the unit itself with a modified version of vold to see what happens
Later Edit,
vold has nothing to do with the block assignment ... on the front USB port also known by the unit : usb20_otg works like a charm... i added 4 USB Drivers each one named accordingly -> sda, sdb, etc.. i managed to mount them manually -> success
On the rear USB also known by the unit usb20_hub it still does not work, VOLD was stopped and removed during tests
clearly things do not add up.. and VOLD so far seems to be ok..
i ran out of ideas

---

i'm not a developer, but if you find a solution for that problem, you'll be a hero. hope you have finally the idea that will be the right one