RE: 2.6.11, USB: High latency?
From: Alan Stern (stern_at_rowland.harvard.edu)
Date: 03/31/05
- Previous message: Chris Friesen: "Re: [RFC/PATCH] network configs: disconnect network options from drivers"
- In reply to: kus Kusche Klaus: "RE: 2.6.11, USB: High latency?"
- Next in thread: Ingo Molnar: "Re: 2.6.11, USB: High latency?"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Thu, 31 Mar 2005 11:48:11 -0500 (EST) To: kus Kusche Klaus <kus@keba.com>
On Thu, 31 Mar 2005, kus Kusche Klaus wrote:
> > The latencies are almost certainly caused by the USB host controller
> > driver. I'm planning improvements to uhci-hcd which should
> > help reduce
> > the latency, but it will still be on the large side. And I
> > won't have
> > time to write the changes to the driver for several months.
>
> Any numbers about the expected "large side"?
> We would need <30 microseconds irq latency,
> and <<1 milliseconds rt application latency.
It's pretty hard to say, considering that: (1) None of the code has been
written yet; (2) It's impossible to say how big a difference new code will
make just by reading it -- you have to actually test it; and (3) I don't
have any hardware suitable for testing latencies.
The biggest advantage would come from using a bottom-half handler to do
most of the work. Right now the uhci-hcd driver does everything in its
interrupt handler. This would certainly help IRQ latency; it might not
affect application latency very much.
I'll try adding a bottom half in my next series of patches. Maybe it will
be ready in time to appear in the -mm kernels before 2.6.12-final is
released.
> > The best solution is to stop using uhci-hcd. Get a PCI card
> > with an OHCI
> > or EHCI (high-speed) controller. They do much more work in hardware,
> > reducing the amount of time the driver needs to spend with interrupts
> > disabled.
>
> The hardware is invariable. It is an embedded system with no PCI slots.
>
> And it seems to be possible with UHCI,
> because vxWorks allows USB stick transfers in operation without
> missing latency requirements.
>
> I do not require rt on the USB, it may block its own irq as long as
> it likes, but it should not affect other irqs.
We'll see what happens with the upcoming changes. Maybe you'll be able to
test them for me?
Alan Stern
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
- Previous message: Chris Friesen: "Re: [RFC/PATCH] network configs: disconnect network options from drivers"
- In reply to: kus Kusche Klaus: "RE: 2.6.11, USB: High latency?"
- Next in thread: Ingo Molnar: "Re: 2.6.11, USB: High latency?"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|
|
