Hi, On Fri, Jan 17, 2003 at 07:13:46PM +0300, Sergey Vlasov wrote: > > > Hmm. Looking at the Linux USB scanner kernel driver, I suspect that > > > the device scan would be blocked in open() when it hits an used > > > device. > > > > I dond't think so. The locks are only active during the call to > > open_scanner(), not while the device is open. Scanning an already > > opened device will run into the "if (scn->isopen)" test and exit. > > At least that's the case in recent kernels. > > > > The following example with three scanners works here: > > > > $ scanimage -L > > device gt68xx:/dev/usb/scanner2' is a Plustek OpticPro 1248U flatbed > > scanner > > device gt68xx:/dev/usb/scanner1' is a Mustek BearPaw 2400 CU Plus flatbed > > scanner > > device gt68xx:/dev/usb/scanner0' is a Mustek ScanExpress 1200 UB Plus > > flatbed scanner > > > > $ scanimage -d gt68xx:/dev/usb/scanner0 >/tmp/image0.pnm & \ > > scanimage -d gt68xx:/dev/usb/scanner1 >/tmp/image1.pnm & \ > > scanimage -d gt68xx:/dev/usb/scanner2 >/tmp/image2.pnm > > But what if you do 'scanimage -L' while some scanning is in progress?
$ scanimage -d gt68xx:/dev/usb/scanner0 >/tmp/image0.pnm & scanimage -L device gt68xx:/dev/usb/scanner2' is a Plustek OpticPro 1248U flatbed scanner device gt68xx:/dev/usb/scanner1' is a Mustek BearPaw 2400 CU Plus flatbed scanner scanner0 isn't printed, but that's ok. The output is a bit slow because of all that "kernel: scanner.c: open_scanner(0): Scanner device is already open" messages printed to syslog. I guess I'll downgrade this error to a debug. But it's not blocked while the device is open. Bye, Henning
