bus-independent backend, with PCI and CardBus attachment code.
The committed code has two serious bugs:
1. The driver makes no attempt to recover resources when a (Cardbus)
instance is removed; bus resources are leaked.
2. In testing with a NetGear GA-511, the Cardbus card never responded
to a reset/wakeup if the card is powered down after attachment.
So for now, leave cardbus instances powered up at attachment
(insertion, or at boot if a card is already present).
That aside, it acutally works on my GA-511. Committed as-is despite
the bugs, after repeated requests to make the code available for
further testing. Also requires sys/dev/mii/miidevs rev 1.54 -> 1.55,
and consequent regen of miidevs{,_data}.h.
in the wanted bus_dma space. If an allocation fails during operation
(i.e. when hotplugging an umass device), fall back to using the reserve.
The amount can be configured as USB_MEM_RESERVE. The default value is
256k.
Ideally, there would be a way to steal pages in the desired area from
UVM, but that's far more complicated, and this is not intrusive, plus
it works.
attach time the way Linux does it. Reported and tested by Jonathan
Schleifer, I checked it didn't break my own 7012 device which doesn't need
that manipulation.
reviewed by christos@, cube@ and martin@.
While here, remove a printf() in mpu_attach() (ic/mpu.c) to remove
an empty line in some frontends (mpu_isa.c, etc).
attempt to map the I/O BAR if it is 0. This will have the side-effect
of causing the wm_reset() routine to fall back to memory BAR accesses
for those chips on which it would attempt I/O BAR accesses for the reset.
Also update the comments about why we attempt the I/O BAR accesses for
the reset in favor of the memory BAR accesses.
make sure that the pullup was done into trailing space at the end of
the existing first-mbuf-in-chain. If not, log an error and drop the
packet, because the mbuf chain no longer corresponds to the DMA map,
and we are already committed to transmitting the packet.
