TI people, please review and test this patchset. Thanks, Mathieu
On Wed, Sep 23, 2026 at 12:44:51AM +0530, Siddharth Karanam wrote: > Instead of invoking ti_sci_handle's ti_sci_dev_ops member directly > in the driver code, this abstracts the property read of > "ti,sci-dev-id" and the function pointers supported by > ti_sci_dev_ops such as is_on, get_device and put_device. > > The goal of the patch is to maintain a single source > of origin for all ti_sci_dev_ops call sites. Any stray calls > to any dev_ops function pointer members in driver code should > eventually be abstracted through this header. > > The second patch abstracts certain device ops' firmware logic > even further (such as get_device/put_device) via pm_runtime. > > Siddharth Karanam (2): > remoteproc: Add ti_sci_dev_ops abstraction > remoteproc: k3: Abstract ti_sci_handle's dev_ops calls through > pm_runtime > > drivers/remoteproc/ti_k3_common.c | 16 ++-- > drivers/remoteproc/ti_k3_common.h | 4 +- > drivers/remoteproc/ti_k3_dsp_remoteproc.c | 14 ++-- > drivers/remoteproc/ti_k3_m4_remoteproc.c | 13 ++-- > drivers/remoteproc/ti_k3_r5_remoteproc.c | 40 ++++------ > drivers/remoteproc/ti_sci_dev.h | 89 +++++++++++++++++++++++ > 6 files changed, 127 insertions(+), 49 deletions(-) > create mode 100644 drivers/remoteproc/ti_sci_dev.h > > -- > 2.55.0 >
