cups-pdf crash status -139

[Per olof Ljungmark]

On 2020-06-16 12:15, Michael Gmelin wrote:
> 
> 
>> On 16. Jun 2020, at 11:47, Per olof Ljungmark <peo at nethead.se> wrote:
>>
>> On 2020-06-15 16:14, Michael Gmelin wrote:
>>>> On Mon, 15 Jun 2020 12:03:17 +0200
>>>> Per olof Ljungmark <peo at nethead.se> wrote:
>>>> On 2020-06-15 09:50, Michael Gmelin wrote:
>>>>>
>>>>>    
>>>>>> On 15. Jun 2020, at 09:45, Per olof Ljungmark <peo at nethead.se>
>>>>>> wrote:
>>>>>
>>>>> What happens if you run the gs command on the pdf you’re printing
>>>>> directly:
>>>>>
>>>>>    
>>>>>> gs -q -dNOPAUSE -dBATCH -dSAFER -dNOMEDIAATTRS -sstdout=?
>>>>>> -sDEVICE=ps2write -dShowAcroForm -sOUTPUTFILE=? -dLanguageLevel=2
>>>>>> -r300 -dCompressFonts=false -dNoT3CCITT -dNOINTERPOLATE ? ? -f ?
>>>>
>>>> A PDF is created.
>>>>
>>>> And, according to logs gs is OK:
>>>>
>>>> D [15/Jun/2020:09:16:23 +0200] [Job 1070] Wrote 1 pages...
>>>> D [15/Jun/2020:09:16:23 +0200] [Job 1070] PID 37126 (pstops) exited
>>>> with no errors.
>>>> D [15/Jun/2020:09:16:23 +0200] [Job 1070] PID 37125 (gs) exited with
>>>> no errors.
>>>> D [15/Jun/2020:09:16:23 +0200] [Job 1070] PID 37123
>>>> (/usr/local/libexec/cups/filter/pdftops) exited with no errors.
>>>> I [15/Jun/2020:09:16:23 +0200] [Job 1070] Backend returned status
>>>> -139 (crashed)
>>> Are you printer(s) shown as okay in the cups web UI? (like, not
>>> temporarily halted, stopped, etc.?). I sometimes had some issues after
>>> upgrading, especially with hplip.
>>> This upstream issue reported earlier this year sounds similar to what
>>> you're seeing, maybe it helps: https://github.com/apple/cups/issues/5765
>>
>> The cups-pdf printer is paused with "Backend failed" but all other printers are OK. HPLIP is not installed.
>>
> 
> Can you enable the cups-pdf printer using the “cupsenable” command? And if so, will it stay enabled or go back to paused/failed at the next attempt?
> 

It reverts to "Backend failed".

Just tried to add the cups-pdf printer to the 12-STABLE workstation I'm 
writing this on with identical result so by now I see this on three 
different boxes, all running 12-STABLE. Also, I am confident that the 
problem surfaced after the updates to cups in late March.

If I have the time at some point I could try to revert to an older 
version but it would be nicer to sort the problem.

Not sure how to proceed from here...

Per