updating that something similar came up again in bug #39243 and bug #61591. 

running the following triggers similar warning behavior:

>> xdata = cosd (0:60:360);
>> ydata = sind (0:60:360);
>> h = patch (xdata, ydata, 'b', 'marker', 's'); %%draws hexagon
>> set (h, 'xdata', xdata(1:end-3))
>> warning: opengl_renderer: x/y/zdata must have the same dimensions.  Not rendering.
warning: opengl_renderer: x/y/zdata must have the same dimensions.  Not rendering.
warning: opengl_renderer: x/y/zdata must have the same dimensions.  Not rendering.
get(h,'xdata'), get(h, 'ydata')
ans =

   1.0000   0.5000  -0.5000  -1.0000

ans =

        0
   0.8660
   0.8660
        0
  -0.8660
  -0.8660
        0
>> set (h, 'ydata', ydata(1:end-3)  %%figure renders 4pt trapezoid, warnings cease

Note the warning appears after the >>, which doesn't reappear until the next typed command as every figure interaction triggers the warning repetition until the condition is fixed. (no more warnings after the ydata set command)

the warning is also triggered when creating a mismatched figure (which errors in matlab), but ...

>> xdata = cosd (0:60:360);
>> ydata = sind (0:60:360);
>> h = patch (xdata, ydata(1:end-1), 'b', 'marker', 's');
>> warning: opengl_renderer: x/y/zdata must have the same dimensions.  Not rendering.
warning: opengl_renderer: x/y/zdata must have the same dimensions.  Not rendering.
get(h,'xdata'), get(h, 'ydata')
ans =
   1.0000
   0.5000
  -0.5000
  -1.0000
  -0.5000
   0.5000
   1.0000

ans =
        0
   0.8660
   0.8660
        0
  -0.8660
  -0.8660

>> h = patch (xdata, ydata, 'b', 'marker', 's');
>> warning: opengl_renderer: x/y/zdata must have the same dimensions.  Not rendering.
warning: opengl_renderer: x/y/zdata must have the same dimensions.  Not rendering.
warning: opengl_renderer: x/y/zdata must have the same dimensions.  Not rendering.
>> get(h,'xdata'), get(h, 'ydata')
ans =
   1.0000
   0.5000
  -0.5000
  -1.0000
  -0.5000
   0.5000
   1.0000

ans =
        0
   0.8660
   0.8660
        0
  -0.8660
  -0.8660
        0

>> warning: opengl_renderer: x/y/zdata must have the same dimensions.  Not rendering.

Even after overwriting h, the 'every interaction' warning persists.  maybe this is just my misunderstanding of how figures work. I understand adding another plot to that figure wouldn't clear the problem creating the warning, but I thought overwriting h should. or does the invalid plot still exist there, just no longer named h?

retitling with a more accurate description and removing tie to specific opengl warning

ok. so on windows in Octave 6.3.0, Octave 6.3.1 - and 6.3.1 mxe default   (hg id: d3298a58049e Mon, Aug 30, 2021)

the behavior is all the same.  the plot never appears in the figure window. the message appears once after the plot command, and then every time the window is interacted with.  the CLI never shows a prompt again (making me think control hadn't been returned), but typing does elicit a response, after which it shows the prompt. see:

>> plot(x)
>> warning: opengl_renderer: data values greater than fl
oat capacity.  (1) Scale data, or (2) Use gnuplot
warning: opengl_renderer: data values greater than float
 capacity.  (1) Scale data, or (2) Use gnuplot
warning: opengl_renderer: data values greater than float
 capacity.  (1) Scale data, or (2) Use gnuplot
abc
error: 'abc' undefined near line 1, column 1
>>

so, something is different on windows that is both preventing it from acutally drawing the plot, and preventing it from showing the prompt again in the CLI.   haven't stepped through m files yet.

At least on Linux it isn't infinite.  The plot is made and a few warnings are issued.  New warnings are issued whenever the plot is interacted with such as moving the window, clicking on the plot with the mouse, or re-sizing the window.  Otherwise, however, no new warnings are printed and the user is free to work at the CLI.