Re: [PATCH 3/6] sched: prctl() cookie manipulation for core scheduling.

[Peter Zijlstra <peterz@infradead.org>]

On Fri, Mar 19, 2021 at 04:32:50PM -0400, Joel Fernandes (Google) wrote:

>  include/linux/sched.h            |   7 ++
>  include/linux/sched/task.h       |   4 +-
>  include/uapi/linux/prctl.h       |   7 ++
>  kernel/sched/core.c              |  11 +-
>  kernel/sched/coretag.c           | 197 +++++++++++++++++++++++++++++--
>  kernel/sched/sched.h             |   2 +
>  kernel/sys.c                     |   7 ++
>  tools/include/uapi/linux/prctl.h |   7 ++
>  8 files changed, 230 insertions(+), 12 deletions(-)
> 

> diff --git a/kernel/sched/core.c b/kernel/sched/core.c
> index 2e3024a6f6e1..a62e8ad5ce58 100644
> --- a/kernel/sched/core.c
> +++ b/kernel/sched/core.c
> @@ -4654,11 +4654,17 @@ unsigned long nr_iowait(void)
>   * sched_exec - execve() is a valuable balancing opportunity, because at
>   * this point the task has the smallest effective memory and cache footprint.
>   */
> -void sched_exec(void)
> +int sched_exec(void)
>  {
>  	struct task_struct *p = current;
>  	unsigned long flags;
>  	int dest_cpu;
> +	int ret;
> +
> +	/* this may change what tasks current can share a core with */
> +	ret = sched_core_exec();
> +	if (ret)
> +		return ret;

Did the hunk for fs/exec.c go missing?