Bug 28900

Summary:	CUPS timeout too short for older Samsung ML-XXXX (SpliX driven) USB printers
Product:	Mageia	Reporter:	John L. ten Wolde <johnltw>
Component:	RPM Packages	Assignee:	Thierry Vignaud <thierry.vignaud>
Status:	NEW ---	QA Contact:
Severity:	normal
Priority:	Normal
Version:	8
Target Milestone:	---
Hardware:	All
OS:	Linux
Whiteboard:
Source RPM:	cups-2.3.3op2-1.mga8.src.rpm	CVE:
Status comment:

Description John L. ten Wolde 2021-05-11 08:39:54 CEST

Hi all. Before I get into my actual issue, I just want to say congratulations and thank you to the team members who take care of MCC's printer discovery/setup component(s). I've never had a more effortless experience setting up my printers than I have with Mageia 8. The input fields literally filled themselves in with the correct information. Kudos! :)

Unfortunately, I'm also left frustrated that my old -- and once hassle-free -- Samsung ML-2510 won't behave itself any longer. In fact it's now so randomly dysfunctional it's become uselessly unreliable.

Fortunately, I don't need to elaborate further because I'm far from alone in my pain. Upstream is already aware of the problem and has a proposed fix coming down the pipe.

It seems that in CUPS 2.3.3op2 the USB bulk-transfer timeout was lowered to what the CUPS devs consider a more sane value. Unfortunately this new value is now too low for old Samsung ML (Monochrome Laser) printers to cope with, forcing the value to be bumped up again next release.

So I'm essentially typing up this report to pre-order the update to CUPS 2.3.3op3 whenever it happens to drop. Soon, hopefully.

...

For a more thorough explanation of what's up, Arch user altermetax and his ML-1640 share my problem exactly. See:

https://github.com/OpenPrinting/cups/issues/171

But by the time he reported (only days ago), upstream was already working on a fix:

https://github.com/OpenPrinting/cups/pull/160

And that's about it. Thanks in advance, guys (and girls).

Comment 1 Lewis Smith 2021-05-11 21:50:47 CEST

Thank you for the very helpful report.
It looks as if the correction is already in the pipeline:
> Fri Apr 16 2021
> old samsung USB devices malfunction with the current (250ms) timeout
> for usb bulk transaction (rhbz#1935318)
Assigning this bug to Thierry for CUPS, but to test the eventual update, it will be important for you John to do that: QA are unlikely to have a suitable printer.
Can tv please advise John when the update hits core/updates_testing, or that it is already there, for him to try.

Assignee: bugsquad => thierry.vignaud

Comment 2 John L. ten Wolde 2021-05-11 23:32:35 CEST

(In reply to Lewis Smith from comment #1)

> ... to test the eventual update, it will be important for you John to do
> that: QA are unlikely to have a suitable printer.

Thanks, Lewis.  I'm waiting eagerly.  No candidate in the Updates Db yet.

I'll also add a direct link to the Red Hat bug you quoted:

https://bugzilla.redhat.com/show_bug.cgi?id=1935318

The second link I provided in my report mentioned that two Red Hat users were actively testing the fix, and now I see that user zdohnal is RH's CUPS maintainer. Nice.

@Thierry :  OK, I'm ready when you are.  Just give the word.  Oh, and in case it expedites building the test RPMs, you only need to compile for i586 as the old printer in question is hooked up to an even more ancient 32-bit box.  Thanks. :)