[BRLTTY] Papenmeier halts keyboard input, was: New release soon.

[Sebastian Humenda]

Hi

Nicolas Pitre schrieb am 05.10.2024, 13:26 -0400:
>On Sat, 5 Oct 2024, Sebastian Humenda wrote:
>> Dave Mielke schrieb am 04.10.2024, 14:02 -0400:
>> >[quoted lines by Sebastian Humenda on 2024/10/04 at 16:50 +0200]
>> >
>> >>I've attached two logs, one with your log categories set, another with a few
>> >>more (though I forgot debug). As I can only access this device once a week, it
>> >>might be useful.
>> >
>> >It must be an interaction between the braille device and the kernel because the brltty log isn't showing anything. All brltty is seeing, isofar as USB is concerned, seems to be a couple of virtual hubs (which is normal).
>> 
>> Ok, thanks for your time. Do you have a suggestion on how to proceed debugging
>> this?
>
>Is your keyboard connected on the same USB hub as the braille display?

Actually, it's a laptop with an internal keyboard. However, this is a very
specific situation, i.e. a VM with USB-passthrough of the Papenmeier display.
I could confirm that it doesn't occur on a "physical" machine, so I guess it
is likely a Linux uinput bug, as Dave suggested.

>What if you connect the braille display to different USB ports?

I'll try that, though guess due to the virtualisation in between, it might not
matter.

@Sébastien Hinderer
> In addition to Nicolas' suggestion, you may try running dmesg as root and
> see whether the most recent messages (so the one displayed towards

I should have mentioned this, yes. There are some messages regarding "modem"
and I was suspecting that the display would be treated as a modem. Given that
neither apt nor lsmod revealed anything (loaded | installed) named "modem", I
didn't further investigate it. I'll post the exact output.

Thanks!
Sebastian
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <http://brltty.app/pipermail/brltty/attachments/20241007/3ed5dfdc/attachment.sig>