Re: 2.6.11-rc1 vs. PowerMac 8500/G3 (and VAIO laptop) [usb-storageoops]

[David Woodhouse]

On Thu, 2005-01-20 at 23:58 -0800, Greg KH wrote:
> Well, we should be byteswapping all of the fields that need to be
> swapped, right?  I'm guessing that userspace is expecting the fields
> to be in cpu endian, correct?

Userspace varies in that. Nobody expects _all_ the fields to be swapped;
the kernel only ever swapped those four. And in fact lsusb from the
stock usbutils expects it to be consistently little-endian. John's
version seems to be hacked to expect just those four fields to be host-
endian, while the rest remains little-endian.

We have a choice here -- we can preserve the ABI by continuing to be
stupidly inconsistent about endianness, or you can revert my patch and
stock usbutils can be correct.

> But if you want, I'll gladly revert your patch, as I don't have a ppc
> box to test this out on.

I'd revert it.

-- 
dwmw2


-
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/