Re: Linux/IA-64 byte order

Richard Gooch (rgooch@atnf.csiro.au)
Tue, 9 Mar 1999 13:55:56 +1100


David Mosberger writes:
> >>>>> On Tue, 9 Mar 1999 13:11:55 +1100, Richard Gooch <rgooch@atnf.csiro.au> said:
>
> >> There are often endianness selection bits in the page table
> >> entries, add a new flag to mmap, and voila.
>
> Richard> Ah, now that's an attractive option.
> Richard> Linus: would this be acceptable?
> Richard> David Mo: is this possible?
>
> As I said before: the question was regarding _default_ byteorder.
> Linus has said that he wants the byteorder to be little-endian and
> that is it.

Hold on. We seem to be talking at cross purposes here. What Dave
Miller is suggesting is to have LE the default, but allowed specific
mmap()ed regions to be accessed with BE. It would appear to me that
that wouldn't cause the ugliness that Linus is concerned about. All
the x86 compatibility isn't affected. Normal operation wouldn't be
affected either. Only applications specifically requiring this would
use this facility, and it shouldn't impact the user<->kernel interface
either (no mucking with ioctl() data).

That's why I'm asking Linus the question whether or not such a
solution (if supported by the hardware) is acceptable.

> It's of no use to press me for architectural information on IA-64.
> You'll just have to wait until the architecture is fully disclosed
> (or live with what has been disclosed already). Sorry.

OK. Didn't realise that this was something sensitive.

Regards,

Richard....

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