Pete Zaitcev wrote:
>
> > > > > timeout is in jiffies, and is the time waited from URB submission to
> > > > > getting a -ETIMEOUT if things go wrong. Normally HZ is a good thing to
> > > > > put in here.
>
> Wolfgang is trying his best to be difficult, so I am not
> replying to his trolling. Although I have a word or two.
>
> For interfaces such as URB, I question the wisdom of timeouts
> counted in jiffies. First, arithmetics in jiffies are often
> confusing, and prone to underflow/overflow difficulties.
> Second, there is a movement afoot to make HZ changeable,
> to accomodate some mobile devices that want to hybernate.
>
> -- Pete
And I have another one or two:
a. timeout is (currently) in jiffies because it ends up calling
schedule_timeout(), which works in jiffies.
b. Some USB drivers use a "raw" number for timeout -- no units implied.
(see dabusb.c, dsbr100.c, net1080.c [says timeout is in milliseconds])
However, some (most) of them use n * HZ.
They should be consistent, whatever the timeout unit is.
~Randy
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
http://lists.sourceforge.net/lists/listinfo/linux-usb-devel
