Henning Meier-Geinitz wrote:
>
> Hi,
>
> On Thu, Jun 29, 2000 at 02:14:25PM +0200, abel deuring wrote:
> > > Same here. I looked at the scanning times when the backend does nothing but
> > > getting data from the scanner and ignoring it. There was no big change in
> > > scanning time (about 5 %). With the original SCSI adapter the Mustek
> > > scanners are about twice as slow as with Windows despite large (4 MB) SCSI
> > > buffers and tweaking the Linux SCSI driver.
> >
> > Which adapter is shipped with the Mustek? And do other adpaters work
> > better?
>
> DTC-3181 or 53C400A. No interrupt (at least with Linux). The driver
> (g_NCR5380) blocks the whole system while waiting for a request. With some
> tricks (inserting usleeps) the driver is usable but gets slower. With the
> standard 128 MB SCSI buffer size scanning in high resolution is about 10
> times slower than with Windows because of backtracking after each read. With
> larger buffersizes I can go up to 2 times slower than windows.
>
> With a different SCSI host adapter (Adaptec 2940 or NCR 810 clone) it's
> better and the scan performance is about the same as with Windows in most
> (but not all) situations.
Henning,
I think this simply shows that a SCSI adapter or driver, which don't
support interrupts or DMA, are not very suitable for a real multitasking
OS like Linux. I even don't expect that it is possible under Windows to
really run another program while such a cheap SCSI adapter is working.
Abel
-- Source code, list archive, and docs: http://www.mostang.com/sane/ To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com
This archive was generated by hypermail 2b29 : Mon Jul 03 2000 - 09:29:07 PDT