Re: segfault with sane-umax

Michael K. Johnson (johnsonm@redhat.com)
Thu, 09 Oct 1997 22:10:43 -0400

"Iain O'Cain" writes:
>write failed; CMD = 0xa, result = 0x0, errno = 12
>Out of memory
>Could not do scan

Oh, that's helpful. That was a failed write to the sg device due to
the device driver saying that it is out of memory. What has happened
is that you have no large, unfragmented blocks of memory available
in the system for the sg device. At least, that's my guess. I've
never had this happen with my S6. Anyone want to give me an S12? :-)

You could redefine SG_BIG_BUFF to be 15872 throughout and see if that
solves the problem.

There's work going on for the 2.1 kernel that may eventually make
this problem go away. For now, we've got to work around it. Start
with the SG_BIG_BUFF hack and see if that fixes it. If it does,
we either need to wait for the kernel to catch up with us, or
implement a workaround. Not sure how long that would take, particularly
because I'm incredibly busy at the moment; it's after 10pm and I'm
still working on stuff for my job...

michaelkjohnson

"Ever wonder why the SAME PEOPLE make up ALL the conspiracy theories?"

--
Source code, list archive, and docs: http://www.mostang.com/sane/
To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com