[PATCH i-g-t CI run 03/14] tests/intel/xe_oa: Rename oa_exp_1_millisec to oa_exponent_default
Dixit, Ashutosh
ashutosh.dixit at intel.com
Sat Feb 22 00:31:11 UTC 2025
On Fri, 21 Feb 2025 16:29:03 -0800, Umesh Nerlige Ramappa wrote:
>
> On Fri, Feb 21, 2025 at 04:24:30PM -0800, Dixit, Ashutosh wrote:
> > On Tue, 18 Feb 2025 12:28:01 -0800, Umesh Nerlige Ramappa wrote:
> >>
> >> Rename the global exponent to default so it can be used in different
> >> modes.
> >>
> >> Signed-off-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa at intel.com>
> >> ---
> >> tests/intel/xe_oa.c | 34 +++++++++++++++++-----------------
> >> 1 file changed, 17 insertions(+), 17 deletions(-)
> >>
> >> diff --git a/tests/intel/xe_oa.c b/tests/intel/xe_oa.c
> >> index 9c74315df..e7d6ac5af 100644
> >> --- a/tests/intel/xe_oa.c
> >> +++ b/tests/intel/xe_oa.c
> >> @@ -299,7 +299,7 @@ static uint32_t devid;
> >> static struct drm_xe_engine_class_instance default_hwe;
> >>
> >> static struct intel_xe_perf *intel_xe_perf;
> >> -static uint64_t oa_exp_1_millisec;
> >> +static uint64_t oa_exponent_default;
> >> static size_t default_oa_buffer_size;
> >> static struct intel_mmio_data mmio_data;
> >> static igt_render_copyfunc_t render_copy;
> >> @@ -514,7 +514,7 @@ static size_t get_default_oa_buffer_size(int fd)
> >> /* OA unit configuration */
> >> DRM_XE_OA_PROPERTY_OA_METRIC_SET, default_test_set->perf_oa_metrics_set,
> >> DRM_XE_OA_PROPERTY_OA_FORMAT, __ff(default_test_set->perf_oa_format),
> >> - DRM_XE_OA_PROPERTY_OA_PERIOD_EXPONENT, oa_exp_1_millisec,
> >> + DRM_XE_OA_PROPERTY_OA_PERIOD_EXPONENT, oa_exponent_default,
> >
> > My suggestion here would be to (a) either rename this to
> > oa_exponent_default_1ms, or alternatively, (b) to rename the period
> > argument of max_oa_exponent_for_period_lte() to period_ns. So it's clear
> > from the code, without having to guess, what the value of the default
> > exponent is.
>
> So, the idea is to set this default to a smaller value like 1000 ns for
> simulation, so that it gets applied to all the tests, hence not using any
> specific values in the name. Thoughts?
Then let's do (b).
More information about the igt-dev
mailing list