On Thu, Jun 11, 2026 at 10:15:30AM +0530, Gaurav Kohli wrote: > > > On 6/9/2026 5:33 PM, Konrad Dybcio wrote: > > On 6/9/26 12:22 PM, Gaurav Kohli wrote: > > > Add support for Thermal Mitigation Devices (TMDs) to enable > > > thermal throttling of remote processors through QMI. > > > > > > This enables the thermal framework to request mitigation when remote > > > subsystems (modem, CDSP) contribute to thermal issues. > > > > > > Signed-off-by: Gaurav Kohli <[email protected]> > > > Signed-off-by: Daniel Lezcano <[email protected]> > > > > There's no other signs of Daniel in this patch, please fix the > > tags chain > > > > [...] > > > > > > + return 0; > > > + > > > + /* Get the TMD names array */ > > > + num_tmds = of_property_count_strings(np, "tmd-names"); > > > > Is this something we can stuff into platform_data for a given rproc on a > > given SoC (for which we already store *some* data in the PAS driver)? > > > > Without tmd-names in DT, it is unclear which index corresponds to which > binding. With #cooling-cells, tmd-names in DT provides an explicit > name-to-index mapping. Please suggest.
Could you please explain, what are those thermal mitigations responsible for? The series applies the same limitations to both mechanisms used on kodiak for the modem. Is it always the case? If so, do we need to handled them separately at all? Can't we apply the same limits to all TMD resources provided by the DSP? -- With best wishes Dmitry
