Re: aha152x driver hangs for AVA-1502AP card

From: Kees Bakker (rnews@tasking.nl)
Date: Wed Apr 12 2000 - 10:05:58 EST


>>>>> "Matthias" == Matthias Juchem <juchem@uni-mannheim.de> writes:

Matthias> I used to use the aha152x driver for my AVA-1502AP card (came with my
Matthias> scanner) to work with my SCSI ZIP drive for about two years.

Is this a CANON scanner?

Matthias> Now, it doesn't work any more.

Matthias> Below, I've include some information about my system, i.e.
[...]

This problem was recently nailed down by Juergen E Fischer
<fischer@linux-buechse.de> (he is the maintainer of aha152x). Apparently
the problem is with SANE, but it wasn't discovered using the previous
version of the aha152x driver.

A quick fix is to disable the 'if' in aha152x_queue(). Here is what Juergen
told me:
>Hmmm, looks like scanimage is using the REQUEST SENSE command.

>AFAIK that command should only be used when a CHECK CONDITION is
>returned by a SCSI command. The 2.0 driver now does that implicitly
>when it happens (otherwise it would lock up using the new error
>handling code). For that reason REQUEST SENSE commands get ignored
>(borought from the aha1542 driver).

>That might be the problem. Either the driver doesn't request
>sense after the CHECK CONDITION or the sg driver or scanimage
>ignore it. scanimage then tries to request the sense
>information again, but that command get's ignored.

>Maybe it helps if you comment out the if statement in
>aha152x_queue. Then the REQUEST SENSE command should be
>processed normally and scanimage should be able to request
>sense.

Better would be to update your SANE (1.0.2) which would not use REQUEST
SENSE anymore. Here is what Juergen said:
>yet another piece of information: The CANON backend in SANE 1.0.2
>doesn't use REQUEST SENSE anymore. Upgrading should solve you
>problem without touching the driver.

I haven't tried the new SANE, because 2.3.99-pre4-[3-5] wasn't very stable
yet. But now we have 2.3.99-pre5! Well, a few minutes ago, that is.

--

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Sat Apr 15 2000 - 21:00:18 EST