On Wed, Feb 26, 2025 at 07:28:17PM +0000, Juntong Deng wrote:
> This patch declare context-sensitive kfunc groups that can be used by
^
declares
> different SCX operations.
...
> +/* Each flag corresponds to a btf kfunc id set */
> +enum scx_ops_kf_flags {
> + SCX_OPS_KF_ANY = 0,
> + SCX_OPS_KF_UNLOCKED = 1 << 1,
nit: Any specific reason to skip bit 0?
> + SCX_OPS_KF_CPU_RELEASE = 1 << 2,
> + SCX_OPS_KF_DISPATCH = 1 << 3,
> + SCX_OPS_KF_ENQUEUE = 1 << 4,
> + SCX_OPS_KF_SELECT_CPU = 1 << 5,
> +};
> +
> +static const u32 scx_ops_context_flags[] = {
> + [SCX_OP_IDX(select_cpu)] = SCX_OPS_KF_SELECT_CPU | SCX_OPS_KF_ENQUEUE,
> + [SCX_OP_IDX(enqueue)] = SCX_OPS_KF_ENQUEUE,
> + [SCX_OP_IDX(dequeue)] = SCX_OPS_KF_ANY,
> + [SCX_OP_IDX(dispatch)] = SCX_OPS_KF_DISPATCH | SCX_OPS_KF_ENQUEUE,
> + [SCX_OP_IDX(tick)] = SCX_OPS_KF_ANY,
> + [SCX_OP_IDX(runnable)] = SCX_OPS_KF_ANY,
> + [SCX_OP_IDX(running)] = SCX_OPS_KF_ANY,
> + [SCX_OP_IDX(stopping)] = SCX_OPS_KF_ANY,
> + [SCX_OP_IDX(quiescent)] = SCX_OPS_KF_ANY,
> + [SCX_OP_IDX(yield)] = SCX_OPS_KF_ANY,
> + [SCX_OP_IDX(core_sched_before)] = SCX_OPS_KF_ANY,
> + [SCX_OP_IDX(set_weight)] = SCX_OPS_KF_ANY,
> + [SCX_OP_IDX(set_cpumask)] = SCX_OPS_KF_ANY,
> + [SCX_OP_IDX(update_idle)] = SCX_OPS_KF_ANY,
> + [SCX_OP_IDX(cpu_acquire)] = SCX_OPS_KF_ANY,
> + [SCX_OP_IDX(cpu_release)] = SCX_OPS_KF_CPU_RELEASE,
> + [SCX_OP_IDX(init_task)] = SCX_OPS_KF_UNLOCKED,
> + [SCX_OP_IDX(exit_task)] = SCX_OPS_KF_ANY,
> + [SCX_OP_IDX(enable)] = SCX_OPS_KF_ANY,
> + [SCX_OP_IDX(disable)] = SCX_OPS_KF_ANY,
> + [SCX_OP_IDX(dump)] = SCX_OPS_KF_DISPATCH,
Shouldn't this be SCX_OPS_KF_UNLOCKED?
Thanks.
--
tejun
