Date: Wed, 18 Mar 2009 20:46:42 +0900 From: Weongyo Jeong <weongyo.jeong@gmail.com> To: Hans Petter Selasky <hselasky@c2i.net> Cc: Sam Leffler <sam@freebsd.org>, freebsd-usb@freebsd.org, Andrew Thompson <thompsa@freebsd.org> Subject: Re: q: USB_SET_TIMEOUT in ugen. Message-ID: <20090318114642.GB12632@weongyo.cdnetworks.kr> In-Reply-To: <200903180852.48034.hselasky@c2i.net> References: <20090318073032.GA12632@weongyo.cdnetworks.kr> <200903180852.48034.hselasky@c2i.net>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Mar 18, 2009 at 08:52:47AM +0100, Hans Petter Selasky wrote:
> Hi,
>
> On Wednesday 18 March 2009, Weongyo Jeong wrote:
> > Hello,
> >
> > During porting uath(4) to usb2 I found the following code failed to run:
> >
> > data = open("/dev/usb/0.2.2", O_WRONLY, 0);
> > timeout = UATH_DATA_TIMEOUT;
> > if (ioctl(data, USB_SET_TIMEOUT, &timeout) < 0)
> > err(-1, "%s: USB_SET_TIMEOUT(%u)", datadev,
> > UATH_DATA_TIMEOUT);
> >
> > and the error was:
> >
> > uathload: /dev/usb/0.2.2: USB_SET_TIMEOUT(1000): Invalid
> > argument
>
> Use the following macro instead:
>
> #define USB_SET_TX_TIMEOUT _IOW ('U', 137, int)
>
> There also is a match for USB_SET_RX_TIMEOUT .
>
> USB_SET_TIMEOUT has been removed.
>
> Also don't forget to include "usb_ioctl.h" .
>
> Another idea:
>
> Port the program to libusb ....
>
> >
> > It looks it's ok with calling ioctl(msg, USB_SET_TIMEOUT, ..) to
> > /dev/usb/0.2.1
> >
> > And the behaviour of program isn't likely with usb1 that it worked well
> > in usb1. After writing some data into msg and data pipe there was no
> > any response using read(2). The problem source can be found at:.
> >
> > http://perforce.freebsd.org/depotTreeBrowser.cgi?FSPC=//depot/user/weongyo
> >/wireless/src/usr.sbin/uathload&HIDEDEL=NO
> >
> > Are there something I missed in here?
> >
>
> See hints above.
Thank you for advise. :-) I've solve `Invalid argument' error. However
the program looks it still doesn't work. All functions the problem used
are open(2), ioctl(2), read(2) and write(2). I think I can try to use
libusb but it'd be better if it works again using the above 4 system
calls.
Is it supported in CURRENT or is it a one of regression?
regards,
Weongyo Jeong
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20090318114642.GB12632>