cups fails to detect duplexer on HP m479fdw printer; this is a regression

[jikamens]

Describe the bug
When I add my HP m479fdw network printer to CUPS, it fails to detect that it has a duplexer unit installed, even though the printer is saying that it does and this used to work.

When I added this printer on another computer, also running Debian (probably Bookworm rather than Trixie, then), the duplexer was detected and double-sided printing was available.

To Reproduce
Steps to reproduce the behavior:

1. lpadmin -p m479 -m everywhere -E -v 'ipps://HP%20Color%20LaserJet%20Pro%20M478f-9f%20%5B3279B8%5D._ipps._tcp.local/'

Expected behavior
The resulting printer has double-sided printing available as an option.

Actual Behavior
Double-sided printing is not available when printing, and the PPD file in /etc/cups/ppd doesn't mention duplex.

System Information:

• OS: Debian Trixie
• cups-filters 1.28.17

Additional context
Add any other context about the problem here.

$ ipptool -tv 'ipps://HP%20Color%20LaserJet%20Pro%20M478f-9f%20%5B3279B8%5D._ipps._tcp.local/' get-printer-attributes.test | grep -i duplex
        job-constraints-supported (1setOf collection) = {resolver-name=fullbleed-sizes media-top-margin=0 media-bottom-margin=0 media-left-margin=0 media-right-margin=0 media=na_letter_8.5x11in,na_legal_8.5x14in,na_executive_7.25x10.5in,na_foolscap_8.5x13in,na_index-4x6_4x6in,na_index-5x8_5x8in,iso_a4_210x297mm,iso_a5_148x210mm,iso_a6_105x148mm,jis_b5_182x257mm,jis_b6_128x182mm,om_small-photo_100x150mm,na_oficio_8.5x13.4in,om_16k_195x270mm,om_16k_184x260mm,roc_16k_7.75x10.75in,jpn_hagaki_100x148mm,jpn_oufuku_148x200mm,na_number-10_4.125x9.5in,na_monarch_3.875x7.5in,iso_b5_176x250mm,iso_c5_162x229mm,iso_dl_110x220mm},{resolver-name=duplex-types sides=two-sided-short-edge,two-sided-long-edge media=na_letter_8.5x11in,na_legal_8.5x14in,na_executive_7.25x10.5in,na_foolscap_8.5x13in,iso_a4_210x297mm,jis_b5_182x257mm,na_oficio_8.5x13.4in,om_16k_195x270mm,om_16k_184x260mm,roc_16k_7.75x10.75in media-type=com.hp.cover-matte,cardstock,labels,envelope,envelope-heavyweight,photographic-film},{resolver-name=duplex-sizes sides=two-sided-short-edge,two-sided-long-edge media=na_index-4x6_4x6in,na_index-5x8_5x8in,iso_a5_148x210mm,iso_a6_105x148mm,jis_b6_128x182mm,om_small-photo_100x150mm,jpn_hagaki_100x148mm,jpn_oufuku_148x200mm,na_number-10_4.125x9.5in,na_monarch_3.875x7.5in,iso_b5_176x250mm,iso_c5_162x229mm,iso_dl_110x220mm}
        job-resolvers-supported (1setOf collection) = {resolver-name=fullbleed-sizes media-top-margin=423 media-bottom-margin=423 media-left-margin=423 media-right-margin=423},{resolver-name=duplex-types sides=one-sided},{resolver-name=duplex-sizes sides=one-sided}
        hp-default-duplex-standard (boolean) = false
        manual-duplex-supported (boolean) = false

[jikamens]

The same problem exists on the dynamic version of this printer made automatically visible by cups-browsed, i.e., I can't print duplex with that dynamically generated printer entry either.

[tillkamppeter]

Could you attach the following files (rename each file to have a .txt extension, then attach each file, one by one, do not compress them and not package them together):

• PPD file (/etc/cups/ppd/*.ppd) from Bookworm, where duplex worked
• PPD file (/etc/cups/ppd/*.ppd) from Trixie, where duplex does not work
• Output file of

ipptool -tv 'ipps://HP%20Color%20LaserJet%20Pro%20M478f-9f%20%5B3279B8%5D._ipps._tcp.local/' get-printer-attributes.test > out.txt

once for Bookwarm and once for Trixie.

Also run this command, once on Bookworm and once on Trixie and attach the resulting PPD files:

driverless 'ipps://HP%20Color%20LaserJet%20Pro%20M478f-9f%20%5B3279B8%5D._ipps._tcp.local/' > file.ppd.txt

Test your printer with these PPD files. Do you get duplex?

[jikamens]

I am no longer convinced this problem is due to a change in CUPS. Rather I think it's due to a change in my printer's firmware.

I tested the driverless command shown above to generate a PPD in Bullseye, Bookworm, and Trixie, and in none of those Debian versions did it result in a PPD file with Duplex lines in it.

I can therefore only assume that some time after I added a printer in July whose PPD file ended up correct, HP changed my printer firmware and changed the format of the information it's sending in response to the IPP query, such that CUPS no longer knows how to interpret it.

I am attaching three files:

• working.ppd.txt is the file that I believe was generated in July that has Duplex lines in it.
• broken.ppd.txt is the file that was generated by the driverless command in Trixie. The file generated in Bookworm is identical. The file generated in Bullseye is slightly different do to the earlier cups-filters version, but none of the differences are related to Duplex so I'm not bothering to attach it.
• ipptool.out.txt, the ipptool command output, which is functionally identical (just timestamps differ) in all three Debian versions.

broken.ppd.txt
ipptool.out.txt
working.ppd.txt

[tillkamppeter]

Thank you for sending the files.

Unfortuntaely, after having checked your files, I must tell you that you are right, HP has messed it up with the firmware.

To check for duplex support in IPP attributes you need to search for "sides" not for "duplex". I get

$ grep sides ipptool.out.txt
        job-creation-attributes-supported (1setOf keyword) = copies,finishings,finishings-col,job-pages-per-set,sides,orientation-requested,media,print-quality,printer-resolution,output-bin,media-col,output-mode,print-content-optimize,pclm-source-resolution,print-color-mode,ipp-attribute-fidelity,job-name,page-ranges,multiple-document-handling,job-mandatory-attributes,overrides,print-rendering-intent,print-scaling
        job-constraints-supported (1setOf collection) = {resolver-name=fullbleed-sizes media-top-margin=0 media-bottom-margin=0 media-left-margin=0 media-right-margin=0 media=na_letter_8.5x11in,na_legal_8.5x14in,na_executive_7.25x10.5in,na_foolscap_8.5x13in,na_index-4x6_4x6in,na_index-5x8_5x8in,iso_a4_210x297mm,iso_a5_148x210mm,iso_a6_105x148mm,jis_b5_182x257mm,jis_b6_128x182mm,om_small-photo_100x150mm,na_oficio_8.5x13.4in,om_16k_195x270mm,om_16k_184x260mm,roc_16k_7.75x10.75in,jpn_hagaki_100x148mm,jpn_oufuku_148x200mm,na_number-10_4.125x9.5in,na_monarch_3.875x7.5in,iso_b5_176x250mm,iso_c5_162x229mm,iso_dl_110x220mm},{resolver-name=duplex-types sides=two-sided-short-edge,two-sided-long-edge media=na_letter_8.5x11in,na_legal_8.5x14in,na_executive_7.25x10.5in,na_foolscap_8.5x13in,iso_a4_210x297mm,jis_b5_182x257mm,na_oficio_8.5x13.4in,om_16k_195x270mm,om_16k_184x260mm,roc_16k_7.75x10.75in media-type=com.hp.cover-matte,cardstock,labels,envelope,envelope-heavyweight,photographic-film},{resolver-name=duplex-sizes sides=two-sided-short-edge,two-sided-long-edge media=na_index-4x6_4x6in,na_index-5x8_5x8in,iso_a5_148x210mm,iso_a6_105x148mm,jis_b6_128x182mm,om_small-photo_100x150mm,jpn_hagaki_100x148mm,jpn_oufuku_148x200mm,na_number-10_4.125x9.5in,na_monarch_3.875x7.5in,iso_b5_176x250mm,iso_c5_162x229mm,iso_dl_110x220mm}
        job-resolvers-supported (1setOf collection) = {resolver-name=fullbleed-sizes media-top-margin=423 media-bottom-margin=423 media-left-margin=423 media-right-margin=423},{resolver-name=duplex-types sides=one-sided},{resolver-name=duplex-sizes sides=one-sided}
        sides-default (keyword) = one-sided
        sides-supported (keyword) = one-sided
$

Important is the last line. It tells which choices are supported for the option "sides" (which is "Duplex" in PPDs), and here only "one-sided" is supported which is wrong.

The printer does Duplex though, and even in IPP when sending an attribute like "sides=two-sided-long-edge" along with a job. The printer just does not report that it supports Duplex via IPP attributes. This is clearly a firmware bug.

And as many months ago your PPD file "working.ppd.txt" was auto-generated, your printer still has reported correctly that time, so this firmware bug is a regression.

"broken.ppd.txt" is generated from the current IPP attributes as we see them in the attached "ipptool.out.txt" file.

Also the capability description for Apple AirPrint clients

urf-supported (1setOf keyword) = CP1,MT1-2-8-9-10-11,PQ3-4-5,RS600,SRGB24,OB10,W8,DEVW8,DEVRGB24,ADOBERGB24,IS19-1-2,V1.5,FN3

does not report the printer as supporting Duplex, as it does not contain a "DM" (Duplex Mode) entry.

Now, should we do some special tricks to work around the firmware bugs, by searching the whole file for "two-sided-long-edge" and assume Duplex support if we find it somewhere, here in the "job-constraints-supported" attribute?

@michaelrsweet What do you think?

[michaelrsweet]

@tillkamppeter Do not add workarounds for things like this. Better to get HP to fix the firmware.

@wifiprintguy Can you take a quick look?