Alan,
I found it is too complicated to make those changes...
Seems current best patch is still make ehci_qh all item 32-bit aligned...
Thanks,
Alek
On Wed, 17 Jun 2009 12:24:00 +0800
Alek Du <alek.du@xxxxxxxxx> wrote:
> On Tue, 16 Jun 2009 23:11:06 +0800
> Alan Stern <stern@xxxxxxxxxxxxxxxxxxx> wrote:
>
> > On Tue, 16 Jun 2009, Alek Du wrote:
> >
> > > Have to send v2 version again, fixed a qh_destroy issue.
> > >
>
> >
> > I'm sorry to put you through all this work, but the patch you wrote is
> > backwards!
> >
> It is not simple just let ehci_qh contain the soft part, since the ehci_shadow requires
> the hardware part to be the first part...
>
> /*
> * Entries in periodic shadow table are pointers to one of four kinds
> * of data structure. That's dictated by the hardware; a type tag is
> * encoded in the low bits of the hardware's periodic schedule. Use
> * Q_NEXT_TYPE to get the tag.
> *
> * For entries in the async schedule, the type tag always says "qh".
> */
> union ehci_shadow {
> struct ehci_qh *qh; /* Q_TYPE_QH */
> struct ehci_itd *itd; /* Q_TYPE_ITD */
> struct ehci_sitd *sitd; /* Q_TYPE_SITD */
> struct ehci_fstn *fstn; /* Q_TYPE_FSTN */
> __hc32 *hw_next; /* (all types) */
> void *ptr;
> };
>
>
> Unless if I split all the ehci_qh, ehci_itd, ehci_sitd, ehci_fstn part into two parts...
>
> union ehci_shadow {
> struct ehci_qh_hw *qh; /* Q_TYPE_QH */
> struct ehci_itd_hw *itd; /* Q_TYPE_ITD */
> struct ehci_sitd_hw *sitd; /* Q_TYPE_SITD */
> struct ehci_fstn_hw *fstn; /* Q_TYPE_FSTN */
> __hc32 *hw_next; /* (all types) */
> void *ptr;
> };
>
> Thanks,
> Alek
> --
> To unsubscribe from this list: send the line "unsubscribe linux-usb" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
