On 27 Jun 2017, David Miller told this:
> From: Nagarathnam Muthusamy <nagarathnam.muthusamy@xxxxxxxxxx>
> Date: Mon, 26 Jun 2017 12:39:27 -0400
>
>> +static notrace inline u64
>> +vgetsns(struct vvar_data *vvar)
>> +{
>> + u64 v;
>> + u64 cycles;
>> +
>> + switch (vvar->vclock_mode) {
>> + case VCLOCK_TICK:
>> + cycles = vread_tick();
>> + break;
>> + case VCLOCK_STICK:
>> + cycles = vread_stick();
>> + break;
>> + default:
>> + return 0;
>> + }
>> + v = (cycles - vvar->clock.cycle_last) & vvar->clock.mask;
>> + return v * vvar->clock.mult;
>> +}
>
> If you're going to go through all of this trouble to make a vdso,
> it should be assembler code patched at boot time for the tick
> access etc.
Yeah, I thought of doing that and came to the conclusion that I was too
stupid to make it work, particularly given that all machines I know of
that don't support %stick are museum pieces so it would be very hard to
test the result (I spent some time hunting for one and couldn't find
any, and the last machine I owned that didn't support %stick went on the
landfill in about 2002). I mean this is userspace code so the
alternatives system won't do it: the patching would have to be manual...
and compared to *that* the vDSO was *easy*.
I certainly don't *object* to doing it that way: it's clearly preferable
if you don't mind the significantly increased maintenance complexity.
> It is wasteful to run this switch statement when on a given
> machine it will evaluate to the same value always.
Agreed!
--
NULL && (void)
--
To unsubscribe from this list: send the line "unsubscribe sparclinux" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
