Re: flowgraph crashes when I enable a block

[Jameson Collins]

I just tried with the latest commit, same issue.  Maybe its the python
version?  I'm using 3.9.18 in conda.  This is the message I see, but I feel
like these X errors aren't usually the actual problem:

(gnuradio-companion:1759026): Gdk-WARNING **: 06:23:29.883: The program
'gnuradio-companion' received an X Window System error.
This probably reflects a bug in the program.
The error was 'BadAlloc (insufficient resources for operation)'.
  (Details: serial 6258 error_code 11 request_code 131 (MIT-SHM) minor_code
5)
  (Note to programmers: normally, X errors are reported asynchronously;
   that is, you will receive the error a while after causing it.
   To debug your program, run it with the GDK_SYNCHRONIZE environment
   variable to change this behavior. You can then get a meaningful
   backtrace from your debugger if you break on the gdk_x_error() function.)

On Wed, Jul 17, 2024 at 10:26 AM Jameson Collins 
wrote:

> I'm using v3.10.2.0 from source, in conda.  I'll try with another version.
>
> On Wed, Jul 17, 2024 at 10:22 AM Volker Schroer  wrote:
>
>> Just tested with a freesh build from main
>> and it works. I can enable the block and run the generated flowgraph.
>>
>>
>> Am 17.07.24 um 13:19 schrieb Jameson Collins:
>> > This flowgraph is based on the example in the PDU to Stream
>> > documentation.  The difference is that now 'N' is now a formula instead
>> > of a constant ( int(samp_rate*tx_dur) )
>> >
>> > As soon as I enable the one disabled block then gnuradio will crash with
>> > only a X window system error.  And as far as I can tell this only
>> > happens when N is variable.
>> >
>> > VEC is defined as `pmt.init_f32vector( N, np.arange(N) )`
>> >
>> > image.png
>>
>>
>>

Re: flowgraph crashes when I enable a block

[Jameson Collins]

I also tried with python 3.8, also in conda.  Same problem.

On Thu, Jul 18, 2024 at 6:26 AM Jameson Collins 
wrote:

> I just tried with the latest commit, same issue.  Maybe its the python
> version?  I'm using 3.9.18 in conda.  This is the message I see, but I feel
> like these X errors aren't usually the actual problem:
>
> (gnuradio-companion:1759026): Gdk-WARNING **: 06:23:29.883: The program
> 'gnuradio-companion' received an X Window System error.
> This probably reflects a bug in the program.
> The error was 'BadAlloc (insufficient resources for operation)'.
>   (Details: serial 6258 error_code 11 request_code 131 (MIT-SHM)
> minor_code 5)
>   (Note to programmers: normally, X errors are reported asynchronously;
>that is, you will receive the error a while after causing it.
>To debug your program, run it with the GDK_SYNCHRONIZE environment
>variable to change this behavior. You can then get a meaningful
>backtrace from your debugger if you break on the gdk_x_error()
> function.)
>
> On Wed, Jul 17, 2024 at 10:26 AM Jameson Collins <
> jameson.coll...@gmail.com> wrote:
>
>> I'm using v3.10.2.0 from source, in conda.  I'll try with another version.
>>
>> On Wed, Jul 17, 2024 at 10:22 AM Volker Schroer  wrote:
>>
>>> Just tested with a freesh build from main
>>> and it works. I can enable the block and run the generated flowgraph.
>>>
>>>
>>> Am 17.07.24 um 13:19 schrieb Jameson Collins:
>>> > This flowgraph is based on the example in the PDU to Stream
>>> > documentation.  The difference is that now 'N' is now a formula instead
>>> > of a constant ( int(samp_rate*tx_dur) )
>>> >
>>> > As soon as I enable the one disabled block then gnuradio will crash
>>> with
>>> > only a X window system error.  And as far as I can tell this only
>>> > happens when N is variable.
>>> >
>>> > VEC is defined as `pmt.init_f32vector( N, np.arange(N) )`
>>> >
>>> > image.png
>>>
>>>
>>>

Re: flowgraph crashes when I enable a block

[Volker Schroer]

I'm using python 3.12.4 and Fedora 40

I also tried with python 3.8, also in conda.  Same problem.

Re: flowgraph crashes when I enable a block

[Jameson Collins]

Python 3.11 is the latest I can reasonably test.  And it still is broken.
I'm going to guess its not related to that.

Short of compiling with debug flags and running in gdb, any thoughts on how
I can isolate this?

On Thu, Jul 18, 2024 at 11:23 AM Volker Schroer  wrote:

> I'm using python 3.12.4 and Fedora 40
> > I also tried with python 3.8, also in conda.  Same problem.
> >
>
>

Re: flowgraph crashes when I enable a block

[Jameson Collins]

I was able to make a simpler flowgraph.  N is no longer a formula.  If
N=425, it works.  If N=426 is crashes.

[image: image.png]

On Thu, Jul 18, 2024 at 12:57 PM Jameson Collins 
wrote:

> Python 3.11 is the latest I can reasonably test.  And it still is broken.
> I'm going to guess its not related to that.
>
> Short of compiling with debug flags and running in gdb, any thoughts on
> how I can isolate this?
>
> On Thu, Jul 18, 2024 at 11:23 AM Volker Schroer  wrote:
>
>> I'm using python 3.12.4 and Fedora 40
>> > I also tried with python 3.8, also in conda.  Same problem.
>> >
>>
>>

Re: flowgraph crashes when I enable a block

[Volker Schroer]

Do you get a coredump it crashes?Am 18.07.2024 um 19:07 schrieb Jameson Collins :I was able to make a simpler flowgraph.  N is no longer a formula.  If N=425, it works.  If N=426 is crashes.On Thu, Jul 18, 2024 at 12:57 PM Jameson Collins  wrote:Python 3.11 is the latest I can reasonably test.  And it still is broken.  I'm going to guess its not related to that.Short of compiling with debug flags and running in gdb, any thoughts on how I can isolate this?On Thu, Jul 18, 2024 at 11:23 AM Volker Schroer  wrote:I'm using python 3.12.4 and Fedora 40
> I also tried with python 3.8, also in conda.  Same problem.
>

Re: flowgraph crashes when I enable a block

[Jameson Collins]

No, just that X windows error.

(gnuradio-companion:2635330): Gdk-WARNING **: 13:44:30.934: The program
'gnuradio-companion' received an X Window System error.
This probably reflects a bug in the program.
The error was 'BadAlloc (insufficient resources for operation)'.
  (Details: serial 16821 error_code 11 request_code 131 (MIT-SHM)
minor_code 5)
  (Note to programmers: normally, X errors are reported asynchronously;
   that is, you will receive the error a while after causing it.
   To debug your program, run it with the GDK_SYNCHRONIZE environment
   variable to change this behavior. You can then get a meaningful
   backtrace from your debugger if you break on the gdk_x_error() function.)

On Thu, Jul 18, 2024 at 1:28 PM Volker Schroer  wrote:

> Do you get a coredump it crashes?
>
> Am 18.07.2024 um 19:07 schrieb Jameson Collins  >:
>
> 
> I was able to make a simpler flowgraph.  N is no longer a formula.  If
> N=425, it works.  If N=426 is crashes.
>
> 
>
>
> On Thu, Jul 18, 2024 at 12:57 PM Jameson Collins <
> jameson.coll...@gmail.com> wrote:
>
>> Python 3.11 is the latest I can reasonably test.  And it still is
>> broken.  I'm going to guess its not related to that.
>>
>> Short of compiling with debug flags and running in gdb, any thoughts on
>> how I can isolate this?
>>
>> On Thu, Jul 18, 2024 at 11:23 AM Volker Schroer  wrote:
>>
>>> I'm using python 3.12.4 and Fedora 40
>>> > I also tried with python 3.8, also in conda.  Same problem.
>>> >
>>>
>>>