USBMS problem on Forma

mzel · 02-19-2022, 11:47 AM

We've had a similar problem in Nickel, but now I stumbled upon it in KoReader. Originally Kobo could not recognize one of the two USB ports of my laptop. A couple of firmwares back Kobo fixed it, and now Nickel connects to both of them just fine.
However KoReader establishes the problem with the same port. When you connect it to the "bad" port it says it is "power only connection"

NiLuJe · 02-19-2022, 01:32 PM

That probably hints at a bad port and/or cable, making the USB negotiation wonky.

You can check the syslog (run logread in a shell; the matching excerpt from the kernel log might also be helpful, so launching klogd beforehand will take care of that) to see how badly it's misdetected.

This is mostly a hardware issue, the main difference may be that I'm overly cautious, in order to avoid the numerous "oh, hey, I just thrashed your filesystem" issues that could otherwise be caused by USBMS

.

Frenzie · 02-20-2022, 02:36 PM

Quote:

Originally Posted by NiLuJe

That probably hints at a bad port and/or cable, making the USB negotiation wonky.

Possibly beyond obvious but important note on that point: the bad port could be on your computer/USB hub. It needn't be the device.

mzel · 03-12-2022, 08:59 PM

I am doing something wrong
I am running telnet server to run logread on kobo. I am starting it from NickelMenu from Nickel. And as I see the log only has info after the last restart. I.e. it has no info whatsoever from my koreader session.
Please advice

NiLuJe · 03-13-2022, 08:37 PM

That's not how you debug USBMS issues: this is.

mzel · 03-29-2022, 11:41 PM

Is this what you were looking for:
-----------

Code:

Mar 29 23:35:05 usbms[1891]: Initializing USBMS v1.3.0 (2022-03-21 12:36:02 +0100)
Mar 29 23:35:05 usbms[1891]: [FBInk] Detected a Kobo Forma (377 => Frost @ Mark 7)
Mar 29 23:35:05 usbms[1891]: [FBInk] Enabled Kobo Mark 7 quirks
Mar 29 23:35:05 usbms[1891]: [FBInk] Clock tick frequency appears to be 100 Hz
Mar 29 23:35:05 usbms[1891]: [FBInk] Screen density set to 300 dpi
Mar 29 23:35:05 usbms[1891]: [FBInk] Variable fb info: 1440x1920, 8bpp @ rotation: 3 (Counter Clockwise, 270°)
Mar 29 23:35:05 usbms[1891]: [FBInk] Fixed fb info: ID is "mxc_epdc_fb", length of fb mem: 11796480 bytes & line length: 1440 bytes
Mar 29 23:35:05 usbms[1891]: [FBInk] Canonical rotation: 0 (Upright, 0°)
Mar 29 23:35:05 usbms[1891]: [FBInk] Fontsize set to 32x32 (IBM (Default) base glyph size: 8x8)
Mar 29 23:35:05 usbms[1891]: [FBInk] Line length: 45 cols, Page size: 60 rows
Mar 29 23:35:05 usbms[1891]: [FBInk] Horizontal fit is perfect!
Mar 29 23:35:05 usbms[1891]: [FBInk] Pen colors set to #000000 for the foreground and #FFFFFF for the background
Mar 29 23:35:05 usbms[1891]: Initialized FBInk v1.24.0-164-g3682fb7 for Kobo
Mar 29 23:35:05 usbms[1891]: [setup_usb_ids:138] USB product ID: 0x4229
Mar 29 23:35:05 usbms[1891]: Initialized libue v1.4.0
Mar 29 23:35:05 usbms[1891]: Initialized libevdev v1.12.0 for device `gpio-keys`
Mar 29 23:35:05 usbms[1891]: [FBInk] Font `/mnt/usbms/resources/fonts/CaskaydiaCove_NF.otf` loaded for style 'Regular'
Mar 29 23:35:05 usbms[1891]: Internal storage partition wasn't busy, it's been successfully marked as expired.
Mar 29 23:35:05 usbms[1891]: External storage device not available.
Mar 29 23:35:05 usbms[1891]: Starting USBMS shenanigans
Mar 29 23:35:05 usbms[1891]: Checking charger type
Mar 29 23:35:05 usbms[1891]: CDP (Charging Downstream Port) charger detected
Mar 29 23:35:05 usbms[1891]: Charger type is not SDP PC, aborting
Mar 29 23:35:10 usbms[1891]: Bye!

NiLuJe · 03-30-2022, 02:19 PM

Indeed! It seems that port is being negotiated as a potentially data-less charging port, which is understandably problematic...$

EDIT: Nope, data-less is DCP, not CDP

. (Thanks for the confusing acronyms, USB-IF ^^).

NiLuJe · 03-30-2022, 02:20 PM

I wouldn't mind the matching kernel logs, to see what it has to say about that

. (that's the klogd + logread stuff in the message I linked to).

mzel · 03-30-2022, 09:48 PM

Quote:

Originally Posted by NiLuJe

Indeed! It seems that port is being negotiated as a potentially data-less charging port, which is understandably problematic...

Except that Nickel can utilize both ports of my laptop after that bug fix. And before the fix it also saw it as the power-only port

NiLuJe · 04-02-2022, 12:35 PM

Which is why I asked for kernel logs ;o).

mzel · 04-02-2022, 07:24 PM

I do not see much of additional information with klogd on:

Code:

Apr  2 19:20:06 kernel: command exec successfully!
Apr  2 19:20:11 usbms[2633]: Initializing USBMS v1.3.0 (2022-03-21 12:36:02 +010
0)
Apr  2 19:20:11 usbms[2633]: [FBInk] Detected a Kobo Forma (377 => Frost @ Mark
7)
Apr  2 19:20:11 usbms[2633]: [FBInk] Enabled Kobo Mark 7 quirks
Apr  2 19:20:11 usbms[2633]: [FBInk] Clock tick frequency appears to be 100 Hz
Apr  2 19:20:11 usbms[2633]: [FBInk] Screen density set to 300 dpi
Apr  2 19:20:11 usbms[2633]: [FBInk] Variable fb info: 1440x1920, 8bpp @ rotatio
n: 3 (Counter Clockwise, 270°)
Apr  2 19:20:11 usbms[2633]: [FBInk] Fixed fb info: ID is "mxc_epdc_fb", length
of fb mem: 11796480 bytes & line length: 1440 bytes
Apr  2 19:20:11 usbms[2633]: [FBInk] Canonical rotation: 0 (Upright, 0°)
Apr  2 19:20:11 usbms[2633]: [FBInk] Fontsize set to 32x32 (IBM (Default) base g
lyph size: 8x8)
Apr  2 19:20:11 usbms[2633]: [FBInk] Line length: 45 cols, Page size: 60 rows
Apr  2 19:20:11 usbms[2633]: [FBInk] Horizontal fit is perfect!
Apr  2 19:20:11 usbms[2633]: [FBInk] Pen colors set to #000000 for the foregroun
d and #FFFFFF for the background
Apr  2 19:20:11 usbms[2633]: Initialized FBInk v1.24.0-164-g3682fb7 for Kobo
Apr  2 19:20:11 usbms[2633]: [setup_usb_ids:138] USB product ID: 0x4229
Apr  2 19:20:11 usbms[2633]: Initialized libue v1.4.0
Apr  2 19:20:11 kernel: ntx_power_key_hook():state=0
Apr  2 19:20:11 usbms[2633]: Initialized libevdev v1.12.0 for device `gpio-keys`

Apr  2 19:20:11 usbms[2633]: [FBInk] Font `/mnt/usbms/resources/fonts/CaskaydiaC
ove_NF.otf` loaded for style 'Regular'
Apr  2 19:20:11 usbms[2633]: Internal storage partition wasn't busy, it's been s
uccessfully marked as expired.
Apr  2 19:20:11 usbms[2633]: External storage device not available.
Apr  2 19:20:11 usbms[2633]: Starting USBMS shenanigans
Apr  2 19:20:11 usbms[2633]: Checking charger type
Apr  2 19:20:11 usbms[2633]: CDP (Charging Downstream Port) charger detected
Apr  2 19:20:11 usbms[2633]: Charger type is not SDP PC, aborting
Apr  2 19:20:16 usbms[2633]: Bye!
Apr  2 19:20:16 kernel: ntx_power_key_hook():state=0
Apr  2 19:20:16 kernel: lit_ctl
Apr  2 19:20:16 kernel: command exec successfully!

NiLuJe · 04-03-2022, 09:15 AM

Sorry, I had a specific procedure in mind that I forgot to describe

.

The kernel logs would be most helpful if you were to only plug the device while *already* running the USBMS tool (i.e., start it via Exit > Start USBMS).

Barring that, I would need more backlog, to actually *see* the (initial) plug event in the kernel logs

.

NiLuJe · 04-03-2022, 09:22 AM

Also, what kind of USB port is that on the computer, and do you have something like "USB Charging" or "USB PD" setup in your BIOS/UEFI FW?

NiLuJe · 04-03-2022, 10:25 AM

Hah. Turns out the kernels were updated on newer devices, and they are now slightly more sensible (as in, they do allow CDP, as it can rightfully enumerate (meaning it supports data)). I've updated the tool to match that (to the best of our ability, some funky > 500mA SDPs might not be able to get detected properly).

mzel · 04-03-2022, 12:44 PM

Quote:

Originally Posted by NiLuJe

Hah. Turns out the kernels were updated on newer devices, and they are now slightly more sensible (as in, they do allow CDP, as it can rightfully enumerate (meaning it supports data)). I've updated the tool to match that (to the best of our ability, some funky > 500mA SDPs might not be able to get detected properly).

I did not get this part. Is there something I need to/should do to pick this up?

02-19-2022, 11:47 AM	#1
mzel Zealot Posts: 107 Karma: 519930 Join Date: Apr 2016 Device: Kobo Forma	USBMS problem on Forma We've had a similar problem in Nickel, but now I stumbled upon it in KoReader. Originally Kobo could not recognize one of the two USB ports of my laptop. A couple of firmwares back Kobo fixed it, and now Nickel connects to both of them just fine. However KoReader establishes the problem with the same port. When you connect it to the "bad" port it says it is "power only connection"

03-30-2022, 02:19 PM	#7
NiLuJe BLAM! Posts: 13,506 Karma: 26047202 Join Date: Jun 2010 Location: Paris, France Device: Kindle 2i, 3g, 4, 5w, PW, PW2, PW5; Kobo H2O, Forma, Elipsa, Sage, C2E	Indeed! It seems that port is being negotiated as a potentially data-less charging port, which is understandably problematic...$ EDIT: Nope, data-less is DCP, not CDP . (Thanks for the confusing acronyms, USB-IF ^^). Last edited by NiLuJe; 04-03-2022 at 10:26 AM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
USBMS Shenanigans...	sherman	Kobo Developer's Corner	21	11-25-2023 11:02 AM
Problem with Kobo Forma and Overdrive	Teeg	Kobo Reader	2	12-22-2021 07:47 AM
problem with my forma reader	anacreon	Devices	10	11-29-2020 01:58 PM
Forma Connector Problem	pgilster	Kobo Reader	2	11-14-2020 03:56 PM
Problem with my Forma	Strether	Kobo Reader	33	11-06-2019 05:46 PM

02-19-2022, 01:32 PM	#2
NiLuJe BLAM! Posts: 13,506 Karma: 26047202 Join Date: Jun 2010 Location: Paris, France Device: Kindle 2i, 3g, 4, 5w, PW, PW2, PW5; Kobo H2O, Forma, Elipsa, Sage, C2E	That probably hints at a bad port and/or cable, making the USB negotiation wonky. You can check the syslog (run logread in a shell; the matching excerpt from the kernel log might also be helpful, so launching klogd beforehand will take care of that) to see how badly it's misdetected. This is mostly a hardware issue, the main difference may be that I'm overly cautious, in order to avoid the numerous "oh, hey, I just thrashed your filesystem" issues that could otherwise be caused by USBMS .

03-12-2022, 08:59 PM	#4
mzel Zealot Posts: 107 Karma: 519930 Join Date: Apr 2016 Device: Kobo Forma	I am doing something wrong I am running telnet server to run logread on kobo. I am starting it from NickelMenu from Nickel. And as I see the log only has info after the last restart. I.e. it has no info whatsoever from my koreader session. Please advice

03-13-2022, 08:37 PM	#5
NiLuJe BLAM! Posts: 13,506 Karma: 26047202 Join Date: Jun 2010 Location: Paris, France Device: Kindle 2i, 3g, 4, 5w, PW, PW2, PW5; Kobo H2O, Forma, Elipsa, Sage, C2E	That's not how you debug USBMS issues: this is.

03-30-2022, 02:20 PM	#8
NiLuJe BLAM! Posts: 13,506 Karma: 26047202 Join Date: Jun 2010 Location: Paris, France Device: Kindle 2i, 3g, 4, 5w, PW, PW2, PW5; Kobo H2O, Forma, Elipsa, Sage, C2E	I wouldn't mind the matching kernel logs, to see what it has to say about that . (that's the klogd + logread stuff in the message I linked to).

04-02-2022, 12:35 PM	#10
NiLuJe BLAM! Posts: 13,506 Karma: 26047202 Join Date: Jun 2010 Location: Paris, France Device: Kindle 2i, 3g, 4, 5w, PW, PW2, PW5; Kobo H2O, Forma, Elipsa, Sage, C2E	Which is why I asked for kernel logs ;o).

04-03-2022, 09:15 AM	#12
NiLuJe BLAM! Posts: 13,506 Karma: 26047202 Join Date: Jun 2010 Location: Paris, France Device: Kindle 2i, 3g, 4, 5w, PW, PW2, PW5; Kobo H2O, Forma, Elipsa, Sage, C2E	Sorry, I had a specific procedure in mind that I forgot to describe . The kernel logs would be most helpful if you were to only plug the device while already running the USBMS tool (i.e., start it via Exit > Start USBMS). Barring that, I would need more backlog, to actually see the (initial) plug event in the kernel logs .

04-03-2022, 09:22 AM	#13
NiLuJe BLAM! Posts: 13,506 Karma: 26047202 Join Date: Jun 2010 Location: Paris, France Device: Kindle 2i, 3g, 4, 5w, PW, PW2, PW5; Kobo H2O, Forma, Elipsa, Sage, C2E	Also, what kind of USB port is that on the computer, and do you have something like "USB Charging" or "USB PD" setup in your BIOS/UEFI FW?

04-03-2022, 10:25 AM	#14
NiLuJe BLAM! Posts: 13,506 Karma: 26047202 Join Date: Jun 2010 Location: Paris, France Device: Kindle 2i, 3g, 4, 5w, PW, PW2, PW5; Kobo H2O, Forma, Elipsa, Sage, C2E	Hah. Turns out the kernels were updated on newer devices, and they are now slightly more sensible (as in, they do allow CDP, as it can rightfully enumerate (meaning it supports data)). I've updated the tool to match that (to the best of our ability, some funky > 500mA SDPs might not be able to get detected properly).

Advert

Advert