Is there any syscall with more accurate than the ddi_get_lbolt() in the kernel?

Question

Hi, I used with the ddi_get_lbolt() to evaluate the performance
of my some section code of driver in the kernel, but the return
tick of ddi_get_lbolt() is the 10 milliseconds too long to
evaluate. It seemed no change when I used with the drv_usectohz()
or drv_hztousec() to reduce the time interval. Would you give me
methods to accurately evaluate.
Thanks in advance.

Answer 1

hi,
I tested this code with 10 thousands lines. The performance of this code is very bad. Actually, as per your question, I just think out this code which is the same with SUMIF function in excel. Having another way is using grouping and merge table as below
and this is faster then.
let
    Source = Excel.CurrentWorkbook(){[Name="Table1"]}[Content],
    Group = Table.Group(Source, {"Date"}, {{"Total", each List.Sum([Time]), type number}}),
    Merge = Table.NestedJoin(Source,"Date",Group,"Date","NewColumn"),
    Expand = Table.ExpandTableColumn(Merge, "NewColumn", {"Total"}, {"Total"})
in
    Expand
Regards,