On Sunday 19 December 2004 5:21 pm, li xinyi wrote:
> >
> > See how the 2.6 "ohci-hcd" driver handles this ... there
> > are several ARM based SOCs with OHCI support there. You're
> > working with old (2.4) code, so you'll have to kluge things
> > not unlike how the 2.4 SA-1100 code did.
> >
>
>
> See how the 2.6 "ohci-hcd" driver handles this ... there
> are several ARM based SOCs with OHCI support there. You're
> working with old (2.4) code, so you'll have to kluge things
> not unlike how the 2.4 SA-1100 code did.
>
Where can I find the sa1100 usb code. It seems not available in th
On Friday 17 December 2004 7:49 am, li xinyi wrote:
> I can understand DMA buffer initlisation for EDs and TDs. But, in my
> case I am working for OHCI connecting to ARM AHB bus with a seperate
> general purpose DMA controller connecting to AHB as well.
You'll be ignoring that completely, unless y
> That's what _enables_ PCI bus-mastering (DMA) by the device.
>
> OHCI controllers do memory accesses all the time, as normal
> part of operation. They have some registers that point to
> memory locations for the USB transaction schedule, and every
> millisecond they start scanning that schedule
Am Freitag, 17. Dezember 2004 14:33 schrieb li xinyi:
> > >
> > > which piece of code initiated the DMA?
> >
> > I am not a specialist on ohci, but have a look at ohci_hub_resume()
> >
>
> I can not find explicit code which initialises DMA. Could it be
> pci_set_master?
pci_set_master() indeed
> >
> > which piece of code initiated the DMA?
>
> I am not a specialist on ohci, but have a look at ohci_hub_resume()
>
I can not find explicit code which initialises DMA. Could it be pci_set_master?
> It dispatches among the drivers sharing a major number, but you are
> not required to use
Am Freitag, 17. Dezember 2004 12:16 schrieb li xinyi:
> >
> > The hci walks its lists on its own. DMA is initiated once and then
> > the hci works on its schedule which is described by descriptors in ram
> > which is read by DMA.
>
> which piece of code initiated the DMA?
I am not a specialist o
>
> The hci walks its lists on its own. DMA is initiated once and then
> the hci works on its schedule which is described by descriptors in ram
> which is read by DMA.
which piece of code initiated the DMA?
>
> PCI devices implement DMA on their own. Their is no equivalent to
> the ISA DMA con
On Fri, 17 Dec 2004, li xinyi wrote:
> Hi, all
>
> Here is two questions about USB drvier in kernel 2.4.20
>
> 1. In the usb-ohci.c, buffer for TD is allocated as below:
It isn't allocated. It is mapped into the hci's space.
> data = pci_map_single (ohci->ohci_dev,
>
Hi, all
Here is two questions about USB drvier in kernel 2.4.20
1. In the usb-ohci.c, buffer for TD is allocated as below:
data = pci_map_single (ohci->ohci_dev,
urb->transfer_buffer, data_len,
usb_pipeout (urb->pipe)
Does a usb-serial converter use rts/cts dtr/drs ?
If yes is this done in the hardware or by your serial-driver ?
--
Klaus
---
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See t
11 matches
Mail list logo
