From owner-freebsd-hackers Fri Jan 13 04:10:39 1995 Return-Path: hackers-owner Received: (from root@localhost) by freefall.cdrom.com (8.6.9/8.6.6) id EAA06771 for hackers-outgoing; Fri, 13 Jan 1995 04:10:39 -0800 Received: from fub46.zedat.fu-berlin.de (fub46.zedat.fu-berlin.de [130.133.1.46]) by freefall.cdrom.com (8.6.9/8.6.6) with SMTP id EAA06759 for ; Fri, 13 Jan 1995 04:10:15 -0800 Received: by fub46.zedat.fu-berlin.de (5.61/1.34) id AA15877; Fri, 13 Jan 95 13:11:16 GMT Date: Fri, 13 Jan 95 13:11:16 GMT From: gusw@fub46.zedat.fu-berlin.de (Gunther Schadow) Message-Id: <9501131311.AA15877@fub46.zedat.fu-berlin.de> To: gusw@fub46.zedat.fu-berlin.de, jkh@time.cdrom.com Subject: Re: new device driver for scanner Cc: FreeBSD-hackers@freefall.cdrom.com Sender: hackers-owner@FreeBSD.org Precedence: bulk Hi Jordan, > Hey, that's really great, thanks! [Translation: "Boah! Ist das aber Geil!!"] > :-) [Oops, where have you learned that german slang?] > > for people with other hardware. Will the driver work for other handys? Or > > will it work after only few modifications? Please send me your experiences > > and I am also willing to help. > > Will do. Let me see how hard this is to integrate - not very, I hope! :) What kind of scanner do you have? As long as it's semantics is similar the major changes would be in gscreg.h and in the geometry array in gsc.h. As soon as there is an other peace of hardware that uses different ports, I will invent an array which maps base port addresses x hardware types to the registers. I am currently working on code that uses larger dma buffers (64k) and will read one buffer in advance. This is because there is always a nasty stripe at the buffer boundaries when the scanner is in halftone mode. I expect to minimize the time that the scanner proceeds without any dma transaction running which occurs when the buffer is moved, inverted and read until it's again empty. There will be a new version coming soon. regards -Gunther