This patch series makes kfunc filters support the use of struct_ops context information to reduce SCX runtime overhead. After improving kfunc filters, SCX no longer needs the mask-based runtime kfuncs call restriction, so this patch removes the mask-based runtime restriction and updates the corresponding test case. I added *st_ops as part of the context information to avoid kfuncs being incorrectly blocked when used in non-SCX scenarios where the member offsets would have a different meaning. Signed-off-by: Juntong Deng <juntong.deng@xxxxxxxxxxx> --- v2 -> v3: * Change scx_kfunc_ids_ops_context to the more descriptive scx_kfunc_ids_ops_context_sensitive * Change the target to sched_ext/for-6.15 * Remove RFC tag * v2 link: https://lore.kernel.org/bpf/AM6PR03MB5080855B90C3FE9B6C4243B099FE2@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx/T/#u v1 -> v2: * Use completely new design in filtering * v1 link: https://lore.kernel.org/bpf/AM6PR03MB5080261D024B49D26F3FFF0099F72@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx/T/#u Juntong Deng (5): bpf: Add struct_ops context information to struct bpf_prog_aux sched_ext: Declare context-sensitive kfunc groups that can be used by different SCX operations sched_ext: Add scx_kfunc_ids_ops_context_sensitive for unified filtering of context-sensitive SCX kfuncs sched_ext: Removed mask-based runtime restrictions on calling kfuncs in different contexts selftests/sched_ext: Update enq_select_cpu_fails to adapt to struct_ops context filter include/linux/bpf.h | 2 + include/linux/sched/ext.h | 24 -- kernel/bpf/verifier.c | 8 +- kernel/sched/ext.c | 385 ++++++++---------- kernel/sched/ext_idle.c | 13 +- .../sched_ext/enq_select_cpu_fails.c | 37 +- 6 files changed, 194 insertions(+), 275 deletions(-) -- 2.39.5
