On Mon, Jan 04, 2016 at 08:28:52PM +0100, Ard Biesheuvel wrote:

On 4 January 2016 at 20:21, H. Peter Anvin [off-list ref] wrote:

quoted

I suspect that means we will also need to go back to arch-specific
sorting for x86.

AFAICT, Tony's patches are not incompatible with mine. The fixup
address is offset with a large constant, but this does not affect the
sort order (since that is based on the other member), and the swap
operation that adds/subtracts the delta should not care about the
class bits. (I don't see any changes to sort_extable() in Tony's
patch)

Correct. Sorting is by the "insn" field (which I did not change).
The "fixup" field is just modified by an offset value, so survives
the math when moved to a new slot by the sort.

@Tony: any comments? And do you have any objections to the ia64 patch
in this series?

The ia64 bits look OK. I haven't tested, but add my Acked-by: anyway.

-Tony