uscanner + HPS' usb stack + nvidia-driver

[Andrew Muhametshin]

Warner Losh пишет:
>> I've tested it in -CURRENT with x11/nvidia-driver and official USB stack 
>> - problem can't be reproduced, but scanning speed is much lower than 
>> with HPS' USB stack.
>>
>> Is there any hope that it will be solved in the near future?
>>     
>
> Define: Much slower.  With my scanner, I can't measure a difference in
> speed using xsane between the two, although I have done only limited
> testing with hps' stack.
>   
Sorry for some confusion on my part, I thought that HPS' USB stack is 
already in -CURRENT. But anyway, here are results (and question still 
stands - is there any way to speed up scanning process; difference 
becomes hours for hundreds of images):


$ uname -a #(+ HPS' USB stack revision 538)
FreeBSD host01 6.2-STABLE FreeBSD 6.2-STABLE #1: Wed Aug  1 05:45:10 MSD 
2007     root at host01:/usr/obj/usr/src/sys/HOST01  i386

$ uname -a
FreeBSD host01.localdomain 7.0-CURRENT FreeBSD 7.0-CURRENT #0: Thu Jul 
26 07:45:03 MSD 2007 
root at host01.localdomain:/usr/obj/usr/src/sys/GENERIC  i386


$ uname -r
6.2-STABLE
$ time scanimage -p --mode color --resolution 600 --quick-format max > 
/PUBLIC/TMP/000-DEL/6_2-600-max01.pnm
Progress: 100.0%
real    0m40.984s
user    0m3.414s
sys     0m0.821s

$ uname -r
7.0-CURRENT
$ time scanimage -p --resolution 600 --quick-format max --mode color > 
/PUBLIC/TMP/000-DEL/7_0-600-max01.pnm
Progress: 100.0%
real    0m56.550s
user    0m3.729s
sys     0m0.597s

$ uname -r
6.2-STABLE
$time scanimage -p --mode color --depth 16 --resolution 600 
--quick-format max > /PUBLIC/TMP/000-DEL/6_2-600-16-max01.pnm
Progress: 100.0%
real    1m4.007s
user    0m7.485s
sys     0m1.731s

$ uname -r
7.0-CURRENT
$ time scanimage -p --mode color --depth 16 --quick-format max 
--resolution 600  > /PUBLIC/TMP/000-DEL/7_0-600-16-max01.pnm
Progress: 100.0%
real    1m40.989s
user    0m7.606s
sys     0m1.238