On Mon, May 10, 2021 at 04:47:23PM +0100, Dave Martin wrote:
On Mon, May 10, 2021 at 04:22:53PM +0100, Mark Brown wrote:
quoted
On Mon, May 10, 2021 at 03:57:11PM +0100, Dave P Martin wrote:
quoted
quoted
If not, can we keep these combined? We could either keep the name
sve_flush_p (since the FFR is a predicate register, if a weird kind of
one), or go with something more descriptive such as sve_flush_p_and_ffr.
quoted
Not right now but I do expect to build on this in the not too distant
future.
OK, but due to the fact that we need to dirty a P-register in order to
zero the FFR, these still don't feel like independent operations.
Indeed, like I said I got confused by the way the macros end up being
written when I originally wrote this.
Can we control the clearing of the FFR (or not) with an argument to a
combined macro?
It's an option, I'm poking around with how to do it.