bp@www.zefox.net wrote:

[apologies for the wide lines]

Theo <theom+news@chiark.greenend.org.uk> wrote:

Depending on what kind of device it appears as (serial, keyboard, mass storage, network, ...) see if you can work out what other app might have grabbed it. ('sudo lsof' is handy here, if it appears as a device node that
you can grep for)

Sudo lsof is loquatious, but not obviously informative. These tests were
on a Pi5 running up-to-date Bookworm, but the behavior is outwardly the
same as Trixie on Pi2.

What I'm looking for is something having grabbed the USB device. This is
hard to specify without knowing how the OWON does its USB, but for example
if it was serial I'd 'lsof | grep /dev/tty' and look for things using the /dev/ttyUSB0 or /devttyACM0 device the OWON probably uses.

Since the OWON app has failed to grab the USB you're looking for other apps that have grabbed it, ie not at the OWON app.

Theo

--- PyGate Linux v1.5.2
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

Theo <theom+news@chiark.greenend.org.uk> wrote:

bp@www.zefox.net wrote:

[apologies for the wide lines]

Theo <theom+news@chiark.greenend.org.uk> wrote:

Depending on what kind of device it appears as (serial, keyboard, mass storage, network, ...) see if you can work out what other app might have grabbed it. ('sudo lsof' is handy here, if it appears as a device node that
you can grep for)

Sudo lsof is loquatious, but not obviously informative. These tests were
on a Pi5 running up-to-date Bookworm, but the behavior is outwardly the same as Trixie on Pi2.

What I'm looking for is something having grabbed the USB device. This is
hard to specify without knowing how the OWON does its USB, but for example
if it was serial I'd 'lsof | grep /dev/tty' and look for things using the /dev/ttyUSB0 or /devttyACM0 device the OWON probably uses.

Since the OWON app has failed to grab the USB you're looking for other apps that have grabbed it, ie not at the OWON app.

Another debugging trick is to run your app under strace:

strace -ft -o strace.log <whatever command>

That outputs a huge logfile which takes a bit of picking through. What
you're looking for is when it tries to access the USB nodes and fails.
Maybe that gives some clues as to where the error is coming from.

eg
strace -ft -o strace.log picocom /dev/ttyUSB0

has a lot of stuff but it says:

15875 11:16:22 openat(AT_FDCWD, "/dev/ttyUSB0", O_RDWR|O_NOCTTY|O_NONBLOCK)
= -1 ENOENT (No such file or directory)
15875 11:16:22 write(2, "\r\nFATAL: ", 9) = 9
15875 11:16:22 write(2, "cannot open /dev/ttyUSB0: No suc"..., 51) = 51
15875 11:16:22 write(2, "\r\n", 2) = 2
15875 11:16:22 exit_group(1) = ?
15875 11:16:22 +++ exited with 1 +++

which tells me it couldn't open the /dev/ttyUSB0 device because it doesn't exist, and then puts a FATAL message on the console. This error is correct because I have no USB serial devices plugged into this machine.

Theo

--- PyGate Linux v1.5.2
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On 09/12/2025 10:46, Theo wrote:

bp@www.zefox.net wrote:

[apologies for the wide lines]

Theo <theom+news@chiark.greenend.org.uk> wrote:

Depending on what kind of device it appears as (serial, keyboard, mass
storage, network, ...) see if you can work out what other app might have >>> grabbed it. ('sudo lsof' is handy here, if it appears as a device node that
you can grep for)

Sudo lsof is loquatious, but not obviously informative. These tests were
on a Pi5 running up-to-date Bookworm, but the behavior is outwardly the
same as Trixie on Pi2.

What I'm looking for is something having grabbed the USB device. This is
hard to specify without knowing how the OWON does its USB, but for example
if it was serial I'd 'lsof | grep /dev/tty' and look for things using the /dev/ttyUSB0 or /devttyACM0 device the OWON probably uses.

Since the OWON app has failed to grab the USB you're looking for other apps that have grabbed it, ie not at the OWON app.

Theo

OOI I discovered a weird thing
IF I have a raspberry PI PICO plugged into my desktop as a power supply
and running code , then my PC does not detect when the samsung mobile
phone is plugged in...unless I unplug the PI first...

My friend also reports weird shit with his (PI's) USB. I think it's not
as rock solid as we sometimes assume.




--
There is something fascinating about science. One gets such wholesale
returns of conjecture out of such a trifling investment of fact.

Mark Twain


--- PyGate Linux v1.5.2
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

Theo <theom+news@chiark.greenend.org.uk> wrote:

bp@www.zefox.net wrote:

[apologies for the wide lines]

Theo <theom+news@chiark.greenend.org.uk> wrote:

Depending on what kind of device it appears as (serial, keyboard, mass
storage, network, ...) see if you can work out what other app might have >> > grabbed it. ('sudo lsof' is handy here, if it appears as a device node that
you can grep for)

Sudo lsof is loquatious, but not obviously informative. These tests were
on a Pi5 running up-to-date Bookworm, but the behavior is outwardly the
same as Trixie on Pi2.

What I'm looking for is something having grabbed the USB device. This is
hard to specify without knowing how the OWON does its USB, but for example
if it was serial I'd 'lsof | grep /dev/tty' and look for things using the /dev/ttyUSB0 or /devttyACM0 device the OWON probably uses.

Since the OWON app has failed to grab the USB you're looking for other apps that have grabbed it, ie not at the OWON app.

A chance dmesg output reported that owon was associated with tty1. The obvious
candidate for conflict was gtkterm, which occupied tty0. Alas, killing gtkterm and removing the ft232 usb-serial adapter didn't fix/change anything.

As an aside, I also tried connecting with the Owon tty using gtkterm.
Can't get a rise out of it by tinkering with the speed or sending break.
That probably isn't relevant, but it did surprise me a little. I thought
at least some garbage would appear if I sent a carriage return.

Thanks for reading,

bob prohaska



--- PyGate Linux v1.5.2
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

Daniel James <daniel@me.invalid> wrote:

On 08/12/2025 13:13, Andy Burns wrote:

Have you done that?

It only says it's a warning, not an error.

When something gives a warning that says 'this thing probably won't work',
it pays to pay heed, even if it doesn't immediately fail stop.

True, but if you $(SEARCH_ENGINE_OF_CHOICE) for it you will find that
there is a suggestion that some things may not work in the situations
that trigger the warning.

It's surprising how often clutching at straws yields results!

Indeed!

On the subject of trying random things, the one Java+USB weirdness I know
about (using the Altera JTAG stack, which is mostly C++ but some higher
level tools are in Java) is fixed via installing the libudev0 package.

Seemingly the JTAG tools try to use the bindings for libudev but an old
version and then try something less reliable if they aren't found. I've
found that symlinking /usr/lix/x86_64-linux-gnu/libudev.so.1 to libudev.so.0 fixes that - only just discovered there's a 'libudev0' package that does it properly.

I doubt this is the OP's problem though.

Theo

--- PyGate Linux v1.5.2
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

bp@www.zefox.net wrote:

bp@www.zefox.net wrote:

Daniel James <daniel@me.invalid> wrote:

On 07/12/2025 18:55, bp@www.zefox.net wrote:

WARNING: Use --enable-native-access=ALL-UNNAMED to avoid a warning for callers in this module

Have you done that?

After a fashion, yes. I tried appending
--enable-native-access=ALL-UNNAMED
on the command line, but it made no difference.

Clearly, the restriction is new(ish) and there's no obvious reason to
think that's the right place put the option. The references I found on-line >> to this change date from 2025, the application was last updated in 2022 IIRC.

Thanks for writing!

bob prohaska

It turns out there's a problem report that resembles mine at https://github.com/florentbr/OWON-VDS1022/issues/76
but it doesn't offer an explicit resolution.

At this point I have Trixie and Bookworm installs that don't connect
and a somewhat older Bookworm install that worked out of the box. So
far, the one that works is on a 33 bit Pi2.

I made a copy of the working microSD, booted that and verified it could communicate with the 'scope. Then I invited the system to upgrade and
it made about 350 changes to packages. After that, the 'scope couldn't
be connected to.

So, clearly, an incompatibility has been introduced by the upgrade.

Is there some way to selectively update and revert when trouble strikes?
It isn't something I _have_ to fix, but it might be useful to know now.

Thanks for reading,

bob prohaska


--- PyGate Linux v1.5.2
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)