In my case, I found a workaround. By optimizing my BIOS setup (I only
tried what seems to be the best, mostly enabling all kinds of caches &
burst & buffers), I was getting a configuration where the
ohci->hcca->done_head ptr was valid except for the last two bytes.
Example : 0xaaef00 -> 0xaa0000.
This causes all the timeouts I've seen. To solve the problem, I modify the
call to pci_pool_create in usb-ohci.h and replace the value 16 (for
alignements) by 0x10000. And it works. This is only a WORK-AROUND.
Karl, would you be glad to:
[usb-ohci.c]
printk() the value of ohci->hcca->done_head in hc_interrupt()
[usb-ohci.h]
printk() the value of ed->dma in hash_add_td()
Maybe your pb is different from mine...
Benoit
On Tue, 15 May 2001, Karl Gutwin wrote:
> I can't test under Windows - I don't have it installed
> nor do I want to waste the time :)
>
> I too noticed the problem with the timeouts. I tried
> turning the timeouts off but it only caused kernel
> panics :( I think this may be a more serious problem.
> Perhaps someone on the usb-devel list would know
> enough to help.
>
> I was thinking too that it may be a problem with the
> PCI bus, but the only other PCI motherboard I've got
> is a sloow 486... all the same, I may try switching
> over to that briefly to eliminate it as a source of
> error.
>
> by the way, I just realized that this motherboard
> indeed has onboard USB support -- but for some reason
> the irq is routed to interrupt 0. Has anyone else
> experienced this problem before? is there an easy
> solution? I don't have one of those brackets with the
> usb connectors on them to plug into my motherboard, so
> it's not as big of a deal (I'd rather get the
> quadraport working) but all the same, it would be nice
> to get that working too.
>
> Thanks,
> -Karl
>
> --- Benoit PAPILLAULT <[EMAIL PROTECTED]>
> wrote:
> >
> > Maybe you have the same pb as mine. I don't think
> > sharing irq is the
> > problem. I looked briefly at the usb-ohci.c code. In
> > my case, I got the
> > same pb as you:
> >
> > USB device are not recognised as they are plug in.
> > It complains about URB
> > timeout and looking deeper, it times out when
> > sending the "usb set
> > address" request (ie the first usb request sent to
> > the device).
> >
> > I added a lot of traces in hc_interrupt() and just
> > discovered that the
> > address seen in ohci->hcca->done_head does not match
> > any of the URB sent
> > (TD ptr).
> >
> > I'm thinking about PCI bus pb (SiS hardware? mine is
> > UMC.). In your case,
> > does it work under windows? (in my case, I do not
> > have enought disk space
> > to install windows).
> >
> > Benoit
> >
> > On Sun, 13 May 2001, Karl Gutwin wrote:
> >
> > > Hello once again...
> > >
> > > I have been frustrated lately with trying to get
> > my
> > > Belkin Quadraport PCI USB card (OHCI) working. For
> > > those unfamiliar folks, it consists of four
> > separate
> > > USB buses, each with a single port, on a single
> > card.
> > > They all share a single interrupt.
> > >
> > > My problem is: the card is recognized fine, and
> > each
> > > of the four root hubs is recognized and shows up.
> > Each
> > > time I plug in a device, however, into any port, a
> > > message similar to the following is shown:
> > >
> > > usb_control/bulk_msg: timeout
> > > usb-ohci.c: unlink URB timeout
> > > usb.c: USB device not accepting new address = 8
> > (error
> > > = -110)
> > >
> > > The device never appears on the
> > /proc/bus/usb/devices
> > > list. I have tried every USB device I own,
> > including
> > > an unpowered hub. Everything seems to give the
> > same,
> > > or similar errors.
> > >
> > > Since I know debug info is key, but I don't like
> > > spamming the list, I have posted the following
> > files
> > > on my website:
> > >
> > http://www.brandeis.edu/~kgutwin/dmesg-5.13.2001.txt
> > >
> > http://www.brandeis.edu/~kgutwin/lspci-5.13.2001.txt
> > >
> >
> http://www.brandeis.edu/~kgutwin/devices-5.13.2001.txt
> > >
> > > The errors can be seen at the bottom of the dmesg.
> > At
> > > the time, I had plugged into one of the ports an
> > > unpowered hub and a serial port adapter into the
> > hub.
> > >
> > > I think my problem is rooted in the interrupts,
> > since
> > > I've seen the same type of error related to shared
> > > interrupts and/or SMP. Unfortunately, I seem to be
> > > forced to share my interrupt, because of the
> > > quadraport card. What is the status of the
> > interrupt
> > > code, related to irq sharing?
> > >
> > > Right now, I'm running kernel 2.4.3. I've also
> > tested
> > > 2.4.0, which had the same errors. I'm in the
> > process
> > > right now of compiling 2.2.19 and testing it.
> > However,
> > > I need to use the usb-storage driver, so the 2.2
> > > series, even if it works, won't be useful to me.
> > When
> > > I bought this card, I had the impression that it
> > was
> > > working - does anyone else use this card?
> > >
> > > Any help would be appreciated.
> > >
> > > Thanks,
> > > -Karl
> > >
> > > __________________________________________________
> > > Do You Yahoo!?
> > > Yahoo! Auctions - buy the things you want at great
> > prices
> > > http://auctions.yahoo.com/
> > >
> > > _______________________________________________
> > > [EMAIL PROTECTED]
> > > To unsubscribe, use the last form field at:
> > >
> >
> http://lists.sourceforge.net/lists/listinfo/linux-usb-users
> > >
> >
>
>
> __________________________________________________
> Do You Yahoo!?
> Yahoo! Auctions - buy the things you want at great prices
> http://auctions.yahoo.com/
>
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
http://lists.sourceforge.net/lists/listinfo/linux-usb-users
