summaryrefslogtreecommitdiff
path: root/tools/skl_compute_wrpll.c
AgeCommit message (Collapse)Author
2015-06-27stats: Spwan igt_init_with_size() from igt_init()Damien Lespiau
It's all about good looking APIs. Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
2015-06-27stats: Add a way to specify if the data set is a population or a sampleDamien Lespiau
This changes how we compute the variance. We want an unbiased variance when reasoning about a sample. Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
2015-06-27stats: Be more precise and talk about mean, not averageDamien Lespiau
There are several types of averages eg. mean, median and mode. Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
2015-06-25skl_compute_wrpll: Don't try other dividers if we find a 0 central freq ↵Damien Lespiau
deviation Paulo suggested that we could short-circuit the search for a good divider if we find a 0 deviation of the DCO frequency from the central frequency. Out of the 373 test frequencies, 34 hit that fast path. Suggested-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
2015-06-25skl_compute_wrpll: Sync a comment with from the kernel codeDamien Lespiau
Might as well try to keep the code in both this test and the kernel as close as possible. Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
2015-06-25skl_compute_wrpll: Fix the mininum deviation computationDamien Lespiau
Paulo noticed that, because we were only comparing positive deviations with positive deviations and negative deviations with negative deviations, we weren't actually always using the absolute minimal deviation at all. This improves the average deviation across all tested frequencies (373): before: average deviation: 215.13 after: average deviation: 194.47 Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
2015-06-25skl_compute_wrpll: Cycle through dividers, then central freqsDamien Lespiau
Follow Paulo's comment on the corresponding kernel patch. This means we also have to move the break when we have cycled through the even dividers as well. This improves the number of even dividers used across the tested frequencies (373) (at the expense of a slightly worse average deviation, but "even dividers take precedence over a lower deviation". before: even/odd dividers: 338/35 average deviation: 206.52 after: even/odd dividers: 363/10 average deviation: 215.13 Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
2015-06-25skl_compute_wrpll: Print the average deviationDamien Lespiau
It's interesting to watch the effect of some algorithm tweaks on the average deviation between the central freq and the dco freq. A metric we'd like to minimize. Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
2015-05-08skl_compute_wrpll: Prefer even dividersDamien Lespiau
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
2015-05-08skl_compute_wrpll: Count how many even/odd dividers we computeDamien Lespiau
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
2015-05-08skl_compute_wrpll: Make sure we respect the DCO frequency constraintsDamien Lespiau
We might as well verify that we have a semblance of all being in order by making sure the DCO frequency is within the expected bounds. Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
2015-05-08skl_compute_wrpll: Add a way to test the SKL WRPLL algorithmDamien Lespiau
I had various problems (infinite loops, unable to compute dividers for certain frequencies) after implementing a BSpec update. Much easier to debug that in userspace. Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
generated by cgit v1.2.3 (git 2.39.1) at 2024-05-15 23:07:15 +0000