On Tue, Feb 7, 2012 at 8:22 PM, Yoichi Yuasa <yuasa@xxxxxxxxxxxxxx> wrote: > Hi, > > On Tue, 7 Feb 2012 18:56:46 -0800 > Bjorn Helgaas <bhelgaas@xxxxxxxxxx> wrote: > >> On Mon, Jan 30, 2012 at 7:54 PM, Bjorn Helgaas <bhelgaas@xxxxxxxxxx> wrote: >> > On Mon, Jan 30, 2012 at 6:42 PM, Yoichi Yuasa <yuasa@xxxxxxxxxxxxxx> wrote: >> >> Hi, >> >> >> >> On Mon, 30 Jan 2012 09:58:05 -0700 >> >> Bjorn Helgaas <bhelgaas@xxxxxxxxxx> wrote: >> >> >> >>> Tell the PCI core about host bridge address translation so it can take >> >>> care of bus-to-resource conversion for us. >> >>> >> >>> N.B. cobalt_legacy_ide_fixup() does scary things with resources, but I >> >>> *think* it will still work because we're still making the same adjustments, >> >>> just in the opposite order. Here are the legacy IDE resource.start >> >>> values I expect: >> >>> >> >>> current new >> >>> >> >>> 0x1f0 (pci_setup_device) 0x1f0 >> >>> (pci_bus_to_resource) + 0xf0000000 >> >>> - 0xf0000000 (cobalt_legacy_ide_fixup) - 0xf0000000 >> >>> + 0xf0000000 (pcibios_fixup_bus) >> >>> ------------ ------------ >> >>> 0x1f0 final value 0x1f0 >> >> >> >> It seems to be unable to handle the offset correctly. >> >> >> >> failed on Cobalt: >> >> >> >> pata_via 0000:00:09.1: BAR 0: can't reserve [io 0x100001f0-0x100001f7] >> >> pata_via 0000:00:09.1: failed to request/iomap BARs for port 0 (errno=-16) >> >> pata_via 0000:00:09.1: BAR 2: can't reserve [io 0x10000170-0x10000177] >> >> pata_via 0000:00:09.1: failed to request/iomap BARs for port 1 (errno=-16) >> >> pata_via 0000:00:09.1: no available native port >> > >> > Thank you very much for testing this. Could you collect the complete >> > dmesg and /proc/ioport contents, before and after my patches? >> >> Ping :) I'll try to figure this out, but it would be easier if I had >> the info mentioned above. Thanks! > > It cannot mount rootfs with your patches. > I cannot get /proc/ioports yet. > > I attach only boot logs. Ah, I see the trouble. The original MIPS pcibios_fixup_bus() applied the io_offset to all IORESOURE_IO resources, but the generic pcibios_bus_to_resource() only applies the offset to things that are in the host bridge's window. The legacy IDE addresses aren't in the window, so we didn't apply the offset, and it didn't work. I'll rework this and update my patches. Thanks! Bjorn -- To unsubscribe from this list: send the line "unsubscribe linux-pci" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
