Without blacklisting the kernel module I make the following observations:
1. The kernel module is not able to obtain the device ID from the printer. Trying to get it with "sudo usb_printerid /dev/usb/lp0" leads to an I/O error.
2. Independent of this I can print arbitrarily many print jobs through the CUPS USB backewnd of my current PPA, still in forced uni-directional mode. The error_log does not show any error but completely successful jobs and the printer does not crash. I do not need to power-cycle it. It is now turned on uninterruptedly since the beginning of comment #97.
3. Though the error_log looks fine and the printer does not crash, the jobs come out incomplete. On every job around the last third of the last page is missing. All pages but the last page come out correctly. It seems that the re-attaching of the kernel module after the job completion sends a reset to the printer and this makes the printer aborting the rest of the job which is already in the printer's memory but not yet printed. Another evidence for the attaching of the module sending a reset is that the print head makes some noise when loading the kernel module manually with "sudo modprobe usblp".
The fact that the kernel module sends some kind of reset to the printer when attached can perhaps cause other bugs like bug 997040.
Without blacklisting the kernel module I make the following observations:
1. The kernel module is not able to obtain the device ID from the printer. Trying to get it with "sudo usb_printerid /dev/usb/lp0" leads to an I/O error.
2. Independent of this I can print arbitrarily many print jobs through the CUPS USB backewnd of my current PPA, still in forced uni-directional mode. The error_log does not show any error but completely successful jobs and the printer does not crash. I do not need to power-cycle it. It is now turned on uninterruptedly since the beginning of comment #97.
3. Though the error_log looks fine and the printer does not crash, the jobs come out incomplete. On every job around the last third of the last page is missing. All pages but the last page come out correctly. It seems that the re-attaching of the kernel module after the job completion sends a reset to the printer and this makes the printer aborting the rest of the job which is already in the printer's memory but not yet printed. Another evidence for the attaching of the module sending a reset is that the print head makes some noise when loading the kernel module manually with "sudo modprobe usblp".
The fact that the kernel module sends some kind of reset to the printer when attached can perhaps cause other bugs like bug 997040.