Hi Reiji,
On 2022-02-19 04:50, Reiji Watanabe wrote:
Hi Marc,
On Thu, Feb 17, 2022 at 1:12 AM Marc Zyngier <maz@xxxxxxxxxx> wrote:
On Thu, 17 Feb 2022 04:52:10 +0000,
Oliver Upton <oupton@xxxxxxxxxx> wrote:
> Would it make sense to test some opt-in capabilities that expose
> additional registers (PMU, SVE, etc.)?
I think this is important. System registers are usually saved/restored
Yes, I will fix the test to include registers for opt-in features
when supported.
in groups, and due to the way we walk the sysreg array, timings are
unlikely to be uniform. Getting a grip on that could help restructure
the walking if required (either per-group arrays, or maybe a tree
structure).
The biggest system register table that I know is sys_reg_descs[],
and KVM_SET_ONE_REG/KVM_GET_ONE_REG/emulation code already uses
binary search to find the target entry. So, the search itself
isn't that bad. The difference between the min and the max
latency of KVM_GET_ONE_REG for the registers is always around
200nsec on Ampere Altra machine as far as I checked.
Even if it is OK so far, it is bound to get worse over time, as
the architecture keeps adding all sort of things that we'll
eventually have to save/restore.
I see this test as a way to monitor this trend and work out when
we need to invest in something better.
Thanks,
M.
--
Jazz is not dead. It just smells funny...
