Reduce the number of iteration in optimised address calibration
This commit is contained in:
parent
9b96280e78
commit
15907efb15
@ -510,6 +510,8 @@ pub unsafe fn calibrate_fixed_freq_2_thread<I: Iterator<Item = (usize, usize)>>(
|
||||
)
|
||||
}
|
||||
|
||||
const OPTIMISED_ADDR_ITER_FACTOR: u32 = 64;
|
||||
|
||||
// TODO : Add the optimised address support
|
||||
// TODO : Modularisation / factorisation of some of the common code with the single threaded no_std version ?
|
||||
|
||||
@ -614,7 +616,7 @@ fn calibrate_fixed_freq_2_thread_impl<I: Iterator<Item = (usize, usize)>>(
|
||||
None => None,
|
||||
};
|
||||
if image_antecedent.is_some() {
|
||||
options.hist_params.iterations *= 1024;
|
||||
options.hist_params.iterations *= OPTIMISED_ADDR_ITER_FACTOR;
|
||||
}
|
||||
|
||||
let old = sched_getaffinity(Pid::from_raw(0)).unwrap();
|
||||
|
Loading…
Reference in New Issue
Block a user