Hi Alan,
> > Hi,
> >
> > I'm currently coding a device driver for a high speed device.
> > Unfortunately,
> > the test have shown, that the transfer rate is real slow.
> >
> > The code is as follows:
> >
> > urb = usb_alloc_urb( 0, GFP_KERNEL );
> > .
> > usb_fill_bulk_u
Alan,
Your code below seems to be missing a curly bracket. A closing bracket
would be missing the block beginning with "if (pdev->vendor". If not the
first, then an opening bracket would be missing for the block beginning with
"if (pci_read_config".
Please correct and repost if necessary. I
Alan,
My USB card is based on a VIA chip, so this patch may help me out. Can you
provide instructions on incorporating a patch like this or a web link for
this?
Thanks,
-Clay
.
.
.
.
.
- Original Message -
From: "Alan Stern" <[EMAIL PROTECTED]>
To: "Igor Yakushin" <[EMAIL PROTECTED]>
On Sun, 22 Jan 2006, Igor Yakushin wrote:
> Alan,
>
>
> >
> > A few people with 64-bit Linux systems have reported problems with USB.
> > At this point it's impossible to tell whether it's caused by hardware
> > issues or problems with the driver.
>
>
> I just ran my usbburn.pl script under
Hi Igor,
I don't know really much about USB but to somehow disable the caching
(for your testing purposes) you could run the command "sync" after
each write operation in your script. "sync" works for all block
devices, so it will freeze your script until all cached data is
flushed.
Good luck,
Andr
Alan,
>
> A few people with 64-bit Linux systems have reported problems with USB.
> At this point it's impossible to tell whether it's caused by hardware
> issues or problems with the driver.
I just ran my usbburn.pl script under Windows for three hours. It worked fine:
3 GB file (Fedora
Co
On Sat, 21 Jan 2006 [EMAIL PROTECTED] wrote:
> I think i'm attached this files on 18.1.2006., but here it again.
Yes, sorry, I lost the other message.
> I'm sorry about wrong naming convention.
> There is few different situation. Hub, Usb Audio is on Port2, Camera on
> Port4,
> everything is c
On Sat, 21 Jan 2006, Igor Yakushin wrote:
> > Mainly you are measuring the computer's buffer
> > capacity, not the transfer speed to the drive.
>
> Which buffer is that? Is it configurable or is it built into hardware?
It is Linux's block buffer system. As far as I know it is not
configurabl
