On 17/03/2020 12:34, Wilco Dijkstra wrote:
Hi Andrea,
I think the first part is fine when approved, but the 2nd part is problematic
like Szabolcs
already pointed out. We can't just change the ABI or semantics, and these
builtins are critical
for GLIBC performance. We would first need to change GLIBC back to using inline
assembler
so it will still write zeroes in the top 32 bits (since that is the current
ABI).
Cheers,
Wilco
No, you shouldn't ever arbitrarily write zeros to the upper bits (or
even worse, write garbage as we do now). That's the whole point of
these patches.
R.
