On Tue, 2006-05-23 at 08:26 +0200, Stefano Brivio wrote:
> -static int bcm43xx_geo_init(struct bcm43xx_private *bcm)
> +static void bcm43xx_geo_init(struct bcm43xx_private *bcm)
> {
> - struct ieee80211_geo *geo;
> + struct ieee80211_geo geo;
This doesn't look like PCI-E support to me, and I haven't seen it in
your earlier patchset.
> - if (bcm->chip_rev < 5) {
> + if (bcm->core_pci.rev <= 5 && bcm->core_pci.id !=
> BCM43xx_COREID_PCIE) {
> sbimconfiglow = bcm43xx_read32(bcm,
> BCM43xx_CIR_SBIMCONFIGLOW);
> sbimconfiglow &= ~ BCM43xx_SBIMCONFIGLOW_REQUEST_TOUT_MASK;
> sbimconfiglow &= ~ BCM43xx_SBIMCONFIGLOW_SERVICE_TOUT_MASK;
> - if (bcm->bustype == BCM43xx_BUSTYPE_PCI)
> - sbimconfiglow |= 0x32;
> - else if (bcm->bustype == BCM43xx_BUSTYPE_SB)
> - sbimconfiglow |= 0x53;
> - else
> - assert(0);
> + sbimconfiglow |= 0x32;
> bcm43xx_write32(bcm, BCM43xx_CIR_SBIMCONFIGLOW,
> sbimconfiglow);
> }
Are you sure we can remove those things for bus type SB? Because, afaik,
that's how some of the WRT... devices are done. Not sure if we handle
them that way though. I forgot who created the patch for those, please
speak up!
> + if (bcm->current_core->id == BCM43xx_COREID_PCI) {
[...]
> + } else {
Shouldn't that be an else if .. PCIE? Or is it guaranteed that only
those ever end up there? I'm not sure.
> /* Set the MAC address in the networking subsystem */
> - if (is_valid_ether_addr(bcm->sprom.et1macaddr))
> + if (bcm43xx_current_phy(bcm)->type == BCM43xx_PHYTYPE_A)
> memcpy(bcm->net_dev->dev_addr, bcm->sprom.et1macaddr, 6);
> else
> memcpy(bcm->net_dev->dev_addr, bcm->sprom.il0macaddr, 6);
I thought we wanted to simply try both mac addresses?
> + bcm43xx_geo_init(bcm);
> +
Again, unrelated change.
johannes
signature.asc
Description: This is a digitally signed message part
