USB Host Mode for Harmattan? - Page 3 - maemo.org

unoace | # 21 | 2013-06-23, 12:13 | Report

Sad this died

Win7Mac | # 22 | 2013-06-23, 13:29 | Report

Originally Posted by unoace View Post

Sad this died

Wish I was wrong, but AFAIK, it's simply not possible on N9. Tivoization is the catchword.
joerg_rw (father of H-E-N) once described it as "you can see the hardware parts, you just can't access them. It's like they're sealed in epoxy".

The Following 3 Users Say Thank You to Win7Mac For This Useful Post:
F2thaK, nokiabot, Wikiwide

nokiabot | # 23 | 2013-06-23, 17:24 | Report

Cunning nokia .

The Following User Says Thank You to nokiabot For This Useful Post:
Wikiwide

mikecomputing | # 24 | 2013-06-23, 17:52 | Report

Who is Nokia? Is that a new company?

The Following 3 Users Say Thank You to mikecomputing For This Useful Post:
Artyom, nokiabot, Wikiwide

javispedro | # 25 | 2013-06-25, 13:09 | Report

There is no "tivoization" problem as long as you are happy with the current open mode, which you really should be.

The lack of USB host is just because lack of time, because there's not much to investigate other than just merge the same changes that were done on the N900. We used to think that there would be an easier method on the N9, because gaia is used as PHY instead of using a separate chip like on the N900, but it seems that even correct triggering of the ID pin on the N9 driver is not enough to get the OTG state machine rolling. So the only choice left is doing the same that was done on the N900.

In forum.meego.com there was an excellent thread about this, I hope the information DocScrutinizer posted there is not lost forever...

The Following 8 Users Say Thank You to javispedro For This Useful Post:
HtheB, imaginaryenemy, joerg_rw, Jordi, juiceme, rcolistete, trx, Wikiwide

nieldk | # 26 | 2013-06-25, 14:14 | Report

Originally Posted by javispedro View Post

There is no "tivoization" problem as long as you are happy with the current open mode, which you really should be.

The lack of USB host is just because lack of time, because there's not much to investigate other than just merge the same changes that were done on the N900. We used to think that there would be an easier method on the N9, because gaia is used as PHY instead of using a separate chip like on the N900, but it seems that even correct triggering of the ID pin on the N9 driver is not enough to get the OTG state machine rolling. So the only choice left is doing the same that was done on the N900.

In forum.meego.com there was an excellent thread about this, I hope the information DocScrutinizer posted there is not lost forever...

At least, not all:
http://web.archive.org/web/201110080...ead.php?t=4610

The Following 2 Users Say Thank You to For This Useful Post:
nokiabot, Wikiwide

joerg_rw | # 27 | 2013-06-27, 02:02 | Report

Originally Posted by nieldk View Post

At least, not all:
http://web.archive.org/web/201110080...ead.php?t=4610

Hrm.
Wayback Machine doesn't have that page
archived.
Want to search for all archived pages under
http://forum.meego.com/ ?

The Following 2 Users Say Thank You to joerg_rw For This Useful Post:
nokiabot, Wikiwide

nokiabot | # 28 | 2013-06-27, 05:37 | Report

How wayback machine archives the internet?? I mean the internet is so huge so how does it keep track of every site??
Just want to know

The Following User Says Thank You to nokiabot For This Useful Post:
Wikiwide

nieldk | # 29 | 2013-06-27, 06:58 | Report

Originally Posted by joerg_rw View Post

Hrm.
Wayback Machine doesn't have that page
archived.
Want to search for all archived pages under
http://forum.meego.com/ ?

Luckily, I copied all from that to a word document

Here is the plain-text version of that (was, unfortunately, not the complete thread) (I can upload somewhere the word doc, if someone wants, or in mail)

(Actually, that page, is still there http://web.archive.org/web/201110080...ead.php?t=4610 - but it seems to sometimes not show, guess it might disappear at some point)

Code:

09-25-2011, 10:30 AM 
joergrw 
Member
 
Join Date: Jul 2010
Posts: 50 
Thanked: 211 times

H-E-N9 USB hostmode enabler N9 

I succeeded in a first step towards USB host mode: finding out about charger chip and operating it in boost mode -> stable 5V supply to USB receptacle

I'll copy the following from http://talk.maemo.org/showthread.php...68#post1094568 and continue the h-e-n topic here (if only for not completely hijacking that infodump thread ;-D )

It's well understood that there are still massive obstacles regarding loading of kernel modules and also regarding aegis introduced restrictions at large. 
Some related links:
http://talk.maemo.org/showthread.php...339#post649339
http://talk.maemo.org/showthread.php?t=65232
Whatever, that's been my major application project, so here we go:

Code:
RM680-22-6_PR_RM680:~# i2cdetect -r 2
WARNING! This program can confuse your I2C bus, cause data loss and worse!
I will probe file /dev/i2c-2 using read byte commands.
I will probe address range 0x03-0x77.
Continue? [Y/n]
     0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
00:          -- -- -- -- -- -- -- -- -- -- -- UU --
10: UU -- -- -- -- -- -- -- -- UU -- -- -- -- -- --
20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
30: UU -- UU -- -- -- -- UU UU -- -- -- -- -- -- --
40: -- -- -- -- -- -- -- -- -- -- -- UU -- -- -- --
50: -- -- -- -- -- 55 -- -- -- -- -- -- -- -- -- --
60: UU -- -- -- -- -- -- -- -- -- -- 6b -- -- -- --
70: -- -- -- -- -- -- -- --
RM680-22-6_PR_RM680:~# echo "        · "UU". Probing was skipped, because this address is currently in use by a driver.  This  strongly  suggests  that
>          there is a chip at this address.
> "
        · UU. Probing was skipped, because this address is currently in use by a driver.  This  strongly  suggests  that
         there is a chip at this address.

RM680-22-6_PR_RM680:~#

RM680-22-6_PR_RM680:~# i2cdump 2 0x6b
No size specified (using byte-data access)
WARNING! This program can confuse your I2C bus, cause data loss and worse!
I will probe file /dev/i2c-2, address 0x6b, mode byte
Continue? [Y/n]
     0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f    0123456789abcdef
00: 40 30 0a 51 09 2c 40 ff ff ff ff ff ff ff ff ff    @0?Q?,@.........
[2011-09-23 17:29:10] <DocScrutinizer> candidate: bq24153A rev1.1
[2011-09-23 17:30:13] <DocScrutinizer> B4, B3:For I2C Address 6BH: 00--bq24151, 01–bq24150; 10–bq24153A; 11–NA.
[2011-09-23 17:32:19] <DocScrutinizer> 0x51=0101 0001= 010(vendor ID) | 10(bq24153A) | 001(001: Revision 1.1
[2011-09-23 17:33:45] <DocScrutinizer> Table 3. Status/Control Register (Read/Write)
[2011-09-23 17:33:45] <DocScrutinizer> Memory Location: 00, Reset State: x1xx 0xxx
[2011-09-23 17:34:01] <DocScrutinizer> Table 4. Control Register (Read/Write)
[2011-09-23 17:34:01] <DocScrutinizer> Memory Location: 01, Reset State: 0011 0000
[2011-09-23 17:34:14] <DocScrutinizer> Table 5. Control/Battery Voltage Register (Read/Write)
[2011-09-23 17:34:14] <DocScrutinizer> Memory Location: 02, Reset State: 0000 1010
[2011-09-23 17:34:28] <DocScrutinizer> Table 6. Vender/Part/Revision Register (Read only)
[2011-09-23 17:34:28] <DocScrutinizer> Memory Location: 03, Reset State: 0101 000x
[2011-09-23 17:34:43] <DocScrutinizer> Table 7. Battery Termination/Fast Charge Current Register (Read/Write)
[2011-09-23 17:34:43] <DocScrutinizer> Memory Location: 04, Reset State: 0000 0001
[2011-09-23 17:35:04] <DocScrutinizer> Table 8. Special Charger Voltage/Enable Pin Status Register
[2011-09-23 17:35:04] <DocScrutinizer> Memory location: 05, Reset state: 001X X100
[2011-09-23 17:36:03] <DocScrutinizer> reg0:is:40,doc:x1xx0xxx - match
[2011-09-23 17:36:36] <DocScrutinizer> reg1:is:30,doc:0011 0000 - match
[2011-09-23 17:37:12] <DocScrutinizer> reg2:is:0a,doc:0000 1010 - check
[2011-09-23 17:37:24] <DocScrutinizer> reg3: already checked
[2011-09-23 17:38:04] <DocScrutinizer> reg4:is:09,doc:0000 0001 - may match
[2011-09-23 17:38:46] <DocScrutinizer> reg5:is:2c,doc:001X X100 - seems to match
[2011-09-23 17:39:24] <DocScrutinizer> reg6:is:40,docoops
[2011-09-23 17:39:40] <DocScrutinizer> :-DD
[2011-09-23 17:39:43] <DocScrutinizer> Table 9. Safety Limit Register (READ/WRITE, Write only once after reset!)
[2011-09-23 17:39:43] <DocScrutinizer> Memory location: 06, Reset state: 01000000
[2011-09-23 17:39:46] <DocScrutinizer> check
[2011-09-23 17:40:08] <DocScrutinizer> WE HAVE A *WINNER*

charger chip in N950: i2cbus2, addr 0x6B, type BQ24153A http://www.ti.com/product/bq24153a
capable of powering VBUS with 5V@200mA (vboost mode)
according to reg6=40 we are locked down to 4200mV and 950mA max charge current
Code:
RM680-22-6_PR_RM680:~# i2cdump 2 0x55
No size specified (using byte-data access)
WARNING! This program can confuse your I2C bus, cause data loss and worse!
I will probe file /dev/i2c-2, address 0x55, mode byte
Continue? [Y/n]
     0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f    0123456789abcdef
00: 00 00 01 01 3f 31 e0 00 06 00 90 0b 0e 10 da ff    ..???1?.?.?????.
10: ce ff 6b 00 e7 9f ff 7f 00 00 0a 00 0a 00 46 0a    ?.k.??.?..?.?.F?
20: ca 0d 64 00 64 00 9c ff 0a 00 ff ff 18 fc 64 00    ??d.d.?.?...??d.
30: 00 00 50 fb 20 21 06 00 00 00 07 00 XX XX XX XX    ..P? !?...?.XXXX
40: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX    XXXXXXXXXXXXXXXX
50: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX    XXXXXXXXXXXXXXXX
60: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX    XXXXXXXXXXXXXXXX
70: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX    XXXXXXXXXXXXXXXX
80: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX    XXXXXXXXXXXXXXXX
90: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX    XXXXXXXXXXXXXXXX
a0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX    XXXXXXXXXXXXXXXX
b0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX    XXXXXXXXXXXXXXXX
c0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX    XXXXXXXXXXXXXXXX
d0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX    XXXXXXXXXXXXXXXX
e0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX    XXXXXXXXXXXXXXXX
f0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX    XXXXXXXXXXXXXXXX
RM680-22-6_PR_RM680:~#
also doesn't exactly look like a bq27200 dump :-S
dpkg -L bme
*candidate: /usr/lib/hwi/ic/bq27521.so

closest hit: http://www.ti.com/product/bq27520-g2 - pinout matches with N9 schematics
Quote:
RM680-22-6_PR_RM680:~# stop bme
Sep 24 13:52:51 (none) bme: terminating with exit status 0
RM680-22-6_PR_RM680:~# start bme
Sep 24 13:53:52 (none) bme: bme (May 19 2011 14:59:58) started, pid 5161
Sep 24 13:53:52 (none) bme: [bmeipc.c:bmeipc_open():356] unable to connect to server (sd=4): No such file or directory (2)
Sep 24 13:53:54 (none) bme: [tabby_get_battery_data_cmdline()] Fixed battery type: BL-4D
Sep 24 13:53:54 (none) bme: [tabby.c:tabby_get_battery_data_cmdline():775] Unknown battery type: 'BL-4D': Invalid argument (22)
Sep 24 13:53:54 (none) bme: Battery level OK for boot (4094 mv)
Sep 24 13:53:54 (none) bme: Stop and wait for upstart
Sep 24 13:53:54 (none) bme: Continue
RM680-22-6_PR_RM680:~/DL# ps|grep bme
606 root /usr/lib/hal/hald-addon-bme
5161 bme /usr/sbin/bme_RX-71 -u -l syslog -v 5 -c /usr/lib/hwi/hw/rx71.so
5164 root grep bme
RM680-22-6_PR_RM680:~/DL# 

Achieved stable VBUS 5V at USB using:
Code:
RM680-22-6_PR_RM680:~# stop bme; sleep 3; i2cset -y -m 0x07 2 0x6b 0x01 0x05; while true; do sleep 5; i2cget -y 2 0x6b 0x00; i2cset -y -m 0x80 2 0x6b 0x00 0x80; done
funny syslog lines: 
Sep 24 16:00:37 (none) kernel: [1135370.167449] twl5031_bcc twl5031_bcc: dcd failed 
Sep 24 16:00:40 (none) kernel: [1135373.303527] g_file_storage gadget: File-backed Storage Gadget, version: 20 November 2008
Sep 24 16:00:40 (none) kernel: [1135373.303588] g_file_storage gadget: Number of LUNs=1 Number of buffers=16 
(memstick attached, probably unrelated)

to be continued 

Last edited by joergrw; 09-25-2011 at 10:48 AM. 


The Following 26 Users Say Thank You to joergrw For This Useful Post: 

acid, almehdi, cassio, claesbas, druid23, et3rnal, gordonshowers, hawaii, isaacdfp, jalyst, javispedro, kollin, martink, mece, mikecomputing, mohammad7410, ovk, pelago, ph5, qole, reffy, rlinfati, sandst1, timsamoff, vkvraju, wicket

  #2   
09-25-2011, 10:49 AM 

kollin 
Junior Member
 
Join Date: Feb 2011
Posts: 25 
Thanked: 23 times

Re: H-E-N9 USB hostmode enabler N9 

Quote:
Originally Posted by joergrw 
I succeeded in a first step towards USB host mode: finding out about charger chip and operating it in boost mode -> stable 5V supply to USB receptacle[/COLOR]
Thank you, that actually provides a partial answer to my questions in this thread : 

link 

Last edited by kollin; 09-25-2011 at 10:57 AM. 


The Following User Says Thank You to kollin For This Useful Post: 

jalyst


  #3   
09-25-2011, 11:11 AM 
joergrw 
Member
 
Join Date: Jul 2010
Posts: 50 
Thanked: 211 times

Re: H-E-N9 USB hostmode enabler N9 

WANTED:
proper datasheet / user manual for RENO TWL5031 aka TPS65951 (or even twl5030/4030 tps65950) with register descriptions

(to save your time, this http://www.ti.com/lit/ds/symlink/tps65950.pdf has no proper register descriptions, it's mainly just about electrical and mechanical properties of the chip, not about actual usage, programming, etc. I'm looking for a file >>170 pages/2MB, probably) 


The Following 2 Users Say Thank You to joergrw For This Useful Post: 

jalyst, mikecomputing

  #4   
09-25-2011, 11:41 AM 
gordonshowers 
Junior Member
 
Join Date: Oct 2010
Posts: 1 
Thanked: 4 times

Re: H-E-N9 USB hostmode enabler N9 

Quote:
Originally Posted by joergrw 
WANTED:
proper datasheet / user manual for RENO TWL5031 aka TPS65951 (or even twl5030/4030 tps65950) with register descriptions

(to save your time, this http://www.ti.com/lit/ds/symlink/tps65950.pdf has no proper register descriptions, it's mainly just about electrical and mechanical properties of the chip, not about actual usage, programming, etc. I'm looking for a file >>170 pages/2MB, probably)
Is this document any good? It mentions registers:

http://www.tij.co.jp/jp/lit/an/swca090/swca090.pdf

Or maybe this one?

http://www.ti.com/lit/ds/symlink/tps65951.pdf 


The Following 4 Users Say Thank You to gordonshowers For This Useful Post: 

jalyst, joergrw, mece, mikecomputing


  #5   
09-25-2011, 12:17 PM 
joergrw 
Member
 
Join Date: Jul 2010
Posts: 50 
Thanked: 211 times

Re: H-E-N9 USB hostmode enabler N9 

Quote:
Originally Posted by gordonshowers 
Is this document any good? It mentions registers:

http://www.tij.co.jp/jp/lit/an/swca090/swca090.pdf
this one mentions the register names at least, but obviously doesn't explain a bit about how the registers are "working"
Quote:
Originally Posted by gordonshowers 
Or maybe this one?

http://www.ti.com/lit/ds/symlink/tps65951.pdf
This one is the *51 variant of the useless tps65950.pdf 171p electrical and mech sheet I linked to as an example of what doesn't help.

here's a datasheet for another similar chip that shows what I'm looking for: http://www.ti.com/lit/ug/swcu067a/swcu067a.pdf (see "2.3 USB Register Descriptions")

many thanks for the help so far, and please don't hesitate to post new findings as well :-)

cheers
jOERG 


The Following 2 Users Say Thank You to joergrw For This Useful Post: 

jalyst, mece

  #6   
09-25-2011, 05:07 PM 

ph5 
Junior Member
 
Join Date: Feb 2010
Posts: 22 
Thanked: 43 times

Re: H-E-N9 USB hostmode enabler N9 

http://www.ti.com/lit/ug/swcu050g/swcu050g.pdf 


The Following 6 Users Say Thank You to ph5 For This Useful Post: 

jalyst, joergrw, martink, mece, mikecomputing, qole


  #7   
09-25-2011, 06:06 PM 
shmerl 
Senior Member
 
Join Date: Sep 2010
Posts: 645 
Thanked: 399 times

Re: H-E-N9 USB hostmode enabler N9 

Great to see some progress in this area! This is such a useful feature, I really wonder why couldn't Nokia implement it by default? Any why couldn't they put microUSB-A in there in the first place (like N8 has for example)? 

Last edited by shmerl; 09-25-2011 at 06:29 PM. 


The Following 2 Users Say Thank You to shmerl For This Useful Post: 

jalyst, mikecomputing

  #8   
09-25-2011, 07:53 PM 
joergrw 
Member
 
Join Date: Jul 2010
Posts: 50 
Thanked: 211 times

Re: H-E-N9 USB hostmode enabler N9 

Quote:
Originally Posted by ph5 
http://www.ti.com/lit/ug/swcu050g/swcu050g.pdf
ph5, you're incredible.
Many thanks :-D
/j 


  #9   
10-01-2011, 09:28 PM 
joergrw 
Member
 
Join Date: Jul 2010
Posts: 50 
Thanked: 211 times

Re: H-E-N9 USB hostmode enabler N9 

Detected mayor hardware differences ¹) between N9 (according to schematics and deduction) vs N950 (according to deduction). Those hardware differences have direct impact on the way USB hostmode would work and render extensive tests on N950 relatively futile, as the results are not applicable to N9.

Quim, how's about some N9 for testing? (no nice display or anything needed, so some sorts of "broken" devices were completely sufficient, as long as core APE and charging and USB works [edit] preferably also working WLAN, for access via ssh :-) [/edit])

[edit2]
STILL WANTED: TPS65951 Technical Reference Manual (analog to http://www.ti.com/lit/ug/swcu050g/swcu050g.pdf TPS65950 version)
[/edit2]

cheers
jOERG


¹) http://www.tij.co.jp/jp/lit/an/swca090/swca090.pdf 

Last edited by joergrw; 10-01-2011 at 09:46 PM. 


The Following 2 Users Say Thank You to joergrw For This Useful Post: 

isaacdfp, jalyst

  #10   
10-02-2011, 03:07 PM 
jalyst 
Senior Member
 
Join Date: Oct 2010
Posts: 130 
Thanked: 27 times

Re: H-E-N9 USB hostmode enabler N9 

Quote:
Originally Posted by shmerl 
This is such a useful feature, I really wonder why couldn't Nokia implement it by default? Any why couldn't they put microUSB-A in there in the first place (like N8 has for example)?
Who knows, so many of their choices in the past 12mths have been an utter mystery.
Luckily we have people w/the skill-sets & willingness to help, & luckily Nokia's open to facilitating that.
Otherwise we'd be "up **** creek..." 

Last edited by jalyst; 10-02-2011 at 03:12 PM.

Last edited by nieldk; 2013-06-27 at 07:02.

The Following 11 Users Say Thank You to For This Useful Post:
ceteigrek, HtheB, Iryus, javispedro, joerg_rw, Jordi, juiceme, MaddogG, mrsellout, qwazix, Wikiwide

juiceme | # 30 | 2013-06-27, 11:09 | Report

Fairly excellent.

I did follow part of that discussion but not to the end as forum.meego.com was scrapped.
Now it would be nice time to start this again, conducting the further tests with N9.

The Following User Says Thank You to juiceme For This Useful Post:
HtheB