[sane-devel] Primax 400 (plustek_pp) only detected by kernel module

[Gerhard Jaeger]

On Monday 14 February 2005 20:03, john wrote:
> On Thu, 10 Feb 2005 16:02:22 +0100
> Gerhard Jaeger <gerhard@gjaeger.de> wrote:
> 
> > On Monday 07 February 2005 20:31, john wrote:
> > > On Mon, 7 Feb 2005 08:31:47 +0100
> > > Gerhard Jaeger <gerhard@gjaeger.de> wrote:
> > > 
> > > > On Friday 04 February 2005 22:57, john wrote:
> > > > > Hello,
> > > > > 
> > > > > I've got an antique "Primax 4800 Driect 30 bit" which is
> > > > > supported by the plustek_pp module.
> > > > > 
> > > > > If I load the kernel module, everything works fine. But using
> > > > > the[direct] user-mode driver, I get this from scanimage -L:
> > > > > 
> > > > > <snip>
> > > > > [plustek_pp] Compare data=0x7f and status=0xf, port=0x0
> > > > > [plustek_pp] p48xxReadWriteTest()
> > > > > [plustek_pp] Can't find your model, asic = 0x77
> > > > > [plustek_pp] detectScannerConnection() returns -9031.
> > > > > [plustek_pp] ************* ASIC9800x *************
> > > > > [plustek_pp] ResetPort()
> > > > > [plustek_pp] IOInitialize()
> > > > > [plustek_pp] * using readfunction >fnBiDirRead<
> > > > > [plustek_pp] ASIC = 0x00
> > > > > [plustek_pp] Unknown ASIC-ID
> > > > > [plustek_pp] *** DETECTION DONE, result: -9020 ***
> > > > > <snip>
> > > > > No scanners were identified.
> > > > > 
> > > > > I've tested on Debian unstable (2.4 kernel) and Slackware
> > > > > current(2.6 kernel) with the same results. Cn anyone point me to
> > > > > where I should start looking for the problem?
> > > > > 
> > > > Hi,
> > > > 
> > > > this could be because of different timing behaviour, when
> > > > accessing a device via kernel directly or via user-mode
> > > > application. As you didn't send the complete debug stuff, I simply
> > > > guess, that the parport didn't run in EPP mode (or is not detected
> > > > correctly), please make sure to have it in EPP mode (check BIOS
> > > > settings), also check the options for the parport_pc modules (no
> > > > IRQ should be assigned)...
> > > 
> > > Gerhard,
> > > 
> > > Thanks for the suggestions. Sadly, it's still not working. Here's
> > > the complete debugging output from "scanimage -L". My apologies to
> > > the list for posting such a long message.
> > > 
> > [SNIPSNAP]
> > > [plustek_pp] Assigning port handle 0
> > > [plustek_pp] ptdrvOpen(port=0x0)
> > > [plustek_pp] Try to claim the parport
> > > [plustek_pp] Setting SPP-mode
> > > [plustek_pp] Setting PS/2-mode
> > > [plustek_pp] We're using libIEEE1284 I/O
> > > [plustek_pp] Starting Scanner-Autodetection
> > 
> > Hmmm, did you really check your BIOS settings? Please make sure, that
> > NO ECP is enabled. You might try and set the port to SPP mode. Also
> > set the parport_pc modules parameter, i.e. 
> > options parport_pc io=0x378  irq=none
> > This seems to be necessary on some boards, esp. with VIA chipset.
> 
> Gerhard,
> 
> No, sorry, I've tried every combination, and nothing works.
> 
> I would be happy to try to debug this myself (and to send you
> a patch, if I could find one). Is there any way I can get hold of any
> documentation to understand the conversation between the PC and the
> scanner?
> 
Well, as the scanner fails in an early stage, there's no need for a lot of
documentation ;)
In your case, you should focus on the different behaviour between the code running 
as kernel module and as user-space program. Especially the detection and init stuff
could be interesting:
MiscInitPorts() in file plustek-pp_misc.c
detectScannerConnection() in file plustek-pp_detect.c

You might wanna dig somewhat deeper. As I do not have such a device, I could not
test...

The following links provide some useful information about the parport interface
and the ieee1284 lib
http://www.beyondlogic.org/spp/parallel.htm
http://cyberelk.net/tim/libieee1284/index.html 

Good luck,
Gerhard