On 02/24/2012 09:53 PM, Ján Veselý wrote: > please test/review I went through all your changes, but especially the ones related to enabling the intelpci driver on ppc32. As a prove, here is a negligible nit I found in ohci_batch.c:
+ assert(addr_to_phys(ohci_ep->td) == ed_head_td(ohci_batch->ed)); + assert(addr_to_phys(ohci_ep->td) == ed_tail_td(ohci_batch->ed)); +// const uint32_t pa = addr_to_phys(ohci_ep->td); +// assert(pa == (ohci_batch->ed->td_head & ED_TDHEAD_PTR_MASK)); +// assert(pa == (ohci_batch->ed->td_tail & ED_TDTAIL_PTR_MASK)); The commented lines look like they will stay forever. As for your approach for reusing the intelpci driver, can a similar split of registers be found e.g. in Linux? As we discussed on IRC, it still seems to me like at least something in the definition of hc registers should be volatile, ergo ioport[8,16,32]_t. Otherwise, the changes look good. > PS: I removed all optical separators :) Hopefully you used some automated tool, only reading through those changes was quite tedious. Cheers, Jakub _______________________________________________ HelenOS-devel mailing list [email protected] http://lists.modry.cz/cgi-bin/listinfo/helenos-devel
