Re: [PATCH] fix ppc ioremap prototype

[Paul Mackerras]

Christoph Hellwig writes:

> Yupp.  This makes me question again how the phys_addr_t thing is
> supposed to work at all given struct resource uses unsigned long
> for start and len, so the whole generic resource infrastructure
> doesn't know about the higher bits...

That's what fixup_bigphys_addr is for.  Basically, on the 440 the
first 4GB of physical address space is all RAM.  PCI memory space
occupies 2GB from 380000000 - 3ffffffff.  So if ioremap is given an
address between 2GB and 4GB, it is assumed to be from a PCI driver,
and fixup_bigphys_addr adds on the 0x300000000.

> Could someone point me a to a driver actually making use of the
> extented ioremap address on ppc 44x?

drivers/net/ibm_emac/ibm_ocp_enet.c in 2.4 does - it's an ocp driver,
not a pci driver.  Yes, we owe Jeff Garzik a 2.5 version of that
driver.

Paul.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/