cups printer prints raw postscript

I have an OKI B430d printer, which is connected to an Odroid C2 (a single-board computer similar to the Raspberry Pi) via USB. The Odroid runs Arch Linux Arm with a cups server, so every device in the local network can use the printer. The printer is configured as RAW printer. My desktop computer runs Arch Linux with a cups instance, too. This cups instance connects to the Odroid via ipp and also has the ppd file for the printer. Now to the issue I started noticing ~8 months ago: Sometimes the printer would print a blank page containing only

--------------------------

or

.0 0
    %-------------------------------------------------

or

source 1.0 0
             %-------------------------------------------------

The issue also occured when printing from a different PC than mine (but also running cups on Arch Linux). Yesterday I printed some more PDFs and this time the printer printed a blank page containing the following in front of 3 jobs:

OKI_Driver_Procedure_Patch_Resource 1.0 0
                                         %------------------------------------

Finally something to search for! I found a PPD file which contained a similar line, so I searched the PPD file I use for the printer for the line and found it on line 108: https://www.oki.com/eu/printing/support/drivers-and-utilities/?id=46262901FZ01 Today, I printed a [test PDF](https://upload.wikimedia.org/wikipedia/commons/d/d3/Test.pdf) after enabling debug logging for both cups instances. I pasted the debug logs here: https://gist.github.com/jkhsjdhjs/aab53920040f62b43c878c74a61860ea One thing I did change for the printer locally was lpadmin -p -o pdftops-renderer-default=pdftops. It didn't work without it, but I don't think that's the cause. Let me know if there's more info I can provide. Thanks in advance. EDIT: So I connected the printer directly to my PC, same issue. Then I created a test printing queue using the same PPD but printing the postscript output to a file instead. I noticed that the output that I suspected to from the PPD file is also in the postscript document that is sent to the printer:

%%BeginResource: OKI_Driver_Procedure_Patch_Resource 1.0 0
%-------------------------------------------------
%% OKI Driver Procedure ver 1.0
%%    Date: Dec. 27, 2005  Edt: 03
%-------------------------------------------------

/OK@setpagedevice_sadj where {pop}{
 /OK@setpagedevice_sadj /setpagedevice load def
/OK@PageDevice_Install_proc currentpagedevice /Install get def
> setpagedevice
}ifelse

So for some reason the printer fails to correctly interpret the PostScript sometimes.