kernel-hacking-2024-linux-s.../kernel/sched/core_sched.c
Linus Torvalds a41b74451b kernel.sys.v5.16
-----BEGIN PGP SIGNATURE-----
 
 iHUEABYKAB0WIQRAhzRXHqcMeLMyaSiRxhvAZXjcogUCYYvEbgAKCRCRxhvAZXjc
 og17AQDj+gsxk2lT4GsRo+WrI9qegGSvYHaxbOoqqSL6rHrrsQD+IU92dwVfuUXE
 oP+De6/TBmsdygnlECxITp8p4ByhGAM=
 =wi2X
 -----END PGP SIGNATURE-----

Merge tag 'kernel.sys.v5.16' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux

Pull prctl updates from Christian Brauner:
 "This contains the missing prctl uapi pieces for PR_SCHED_CORE.

  In order to activate core scheduling the caller is expected to specify
  the scope of the new core scheduling domain.

  For example, passing 2 in the 4th argument of

     prctl(PR_SCHED_CORE, PR_SCHED_CORE_CREATE, <pid>,  2, 0);

  would indicate that the new core scheduling domain encompasses all
  tasks in the process group of <pid>. Specifying 0 would only create a
  core scheduling domain for the thread identified by <pid> and 2 would
  encompass the whole thread-group of <pid>.

  Note, the values 0, 1, and 2 correspond to PIDTYPE_PID, PIDTYPE_TGID,
  and PIDTYPE_PGID. A first version tried to expose those values
  directly to which I objected because:

   - PIDTYPE_* is an enum that is kernel internal which we should not
     expose to userspace directly.

   - PIDTYPE_* indicates what a given struct pid is used for it doesn't
     express a scope.

  But what the 4th argument of PR_SCHED_CORE prctl() expresses is the
  scope of the operation, i.e. the scope of the core scheduling domain
  at creation time. So Eugene's patch now simply introduces three new
  defines PR_SCHED_CORE_SCOPE_THREAD, PR_SCHED_CORE_SCOPE_THREAD_GROUP,
  and PR_SCHED_CORE_SCOPE_PROCESS_GROUP. They simply express what
  happens.

  This has been on the mailing list for quite a while with all relevant
  scheduler folks Cced. I announced multiple times that I'd pick this up
  if I don't see or her anyone else doing it. None of this touches
  proper scheduler code but only concerns uapi so I think this is fine.

  With core scheduling being quite common now for vm managers (e.g.
  moving individual vcpu threads into their own core scheduling domain)
  and container managers (e.g. moving the init process into its own core
  scheduling domain and letting all created children inherit it) having
  to rely on raw numbers passed as the 4th argument in prctl() is a bit
  annoying and everyone is starting to come up with their own defines"

* tag 'kernel.sys.v5.16' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux:
  uapi/linux/prctl: provide macro definitions for the PR_SCHED_CORE type argument
2021-11-10 16:10:47 -08:00

234 lines
5 KiB
C

// SPDX-License-Identifier: GPL-2.0-only
#include <linux/prctl.h>
#include "sched.h"
/*
* A simple wrapper around refcount. An allocated sched_core_cookie's
* address is used to compute the cookie of the task.
*/
struct sched_core_cookie {
refcount_t refcnt;
};
static unsigned long sched_core_alloc_cookie(void)
{
struct sched_core_cookie *ck = kmalloc(sizeof(*ck), GFP_KERNEL);
if (!ck)
return 0;
refcount_set(&ck->refcnt, 1);
sched_core_get();
return (unsigned long)ck;
}
static void sched_core_put_cookie(unsigned long cookie)
{
struct sched_core_cookie *ptr = (void *)cookie;
if (ptr && refcount_dec_and_test(&ptr->refcnt)) {
kfree(ptr);
sched_core_put();
}
}
static unsigned long sched_core_get_cookie(unsigned long cookie)
{
struct sched_core_cookie *ptr = (void *)cookie;
if (ptr)
refcount_inc(&ptr->refcnt);
return cookie;
}
/*
* sched_core_update_cookie - replace the cookie on a task
* @p: the task to update
* @cookie: the new cookie
*
* Effectively exchange the task cookie; caller is responsible for lifetimes on
* both ends.
*
* Returns: the old cookie
*/
static unsigned long sched_core_update_cookie(struct task_struct *p,
unsigned long cookie)
{
unsigned long old_cookie;
struct rq_flags rf;
struct rq *rq;
bool enqueued;
rq = task_rq_lock(p, &rf);
/*
* Since creating a cookie implies sched_core_get(), and we cannot set
* a cookie until after we've created it, similarly, we cannot destroy
* a cookie until after we've removed it, we must have core scheduling
* enabled here.
*/
SCHED_WARN_ON((p->core_cookie || cookie) && !sched_core_enabled(rq));
enqueued = sched_core_enqueued(p);
if (enqueued)
sched_core_dequeue(rq, p);
old_cookie = p->core_cookie;
p->core_cookie = cookie;
if (enqueued)
sched_core_enqueue(rq, p);
/*
* If task is currently running, it may not be compatible anymore after
* the cookie change, so enter the scheduler on its CPU to schedule it
* away.
*/
if (task_running(rq, p))
resched_curr(rq);
task_rq_unlock(rq, p, &rf);
return old_cookie;
}
static unsigned long sched_core_clone_cookie(struct task_struct *p)
{
unsigned long cookie, flags;
raw_spin_lock_irqsave(&p->pi_lock, flags);
cookie = sched_core_get_cookie(p->core_cookie);
raw_spin_unlock_irqrestore(&p->pi_lock, flags);
return cookie;
}
void sched_core_fork(struct task_struct *p)
{
RB_CLEAR_NODE(&p->core_node);
p->core_cookie = sched_core_clone_cookie(current);
}
void sched_core_free(struct task_struct *p)
{
sched_core_put_cookie(p->core_cookie);
}
static void __sched_core_set(struct task_struct *p, unsigned long cookie)
{
cookie = sched_core_get_cookie(cookie);
cookie = sched_core_update_cookie(p, cookie);
sched_core_put_cookie(cookie);
}
/* Called from prctl interface: PR_SCHED_CORE */
int sched_core_share_pid(unsigned int cmd, pid_t pid, enum pid_type type,
unsigned long uaddr)
{
unsigned long cookie = 0, id = 0;
struct task_struct *task, *p;
struct pid *grp;
int err = 0;
if (!static_branch_likely(&sched_smt_present))
return -ENODEV;
BUILD_BUG_ON(PR_SCHED_CORE_SCOPE_THREAD != PIDTYPE_PID);
BUILD_BUG_ON(PR_SCHED_CORE_SCOPE_THREAD_GROUP != PIDTYPE_TGID);
BUILD_BUG_ON(PR_SCHED_CORE_SCOPE_PROCESS_GROUP != PIDTYPE_PGID);
if (type > PIDTYPE_PGID || cmd >= PR_SCHED_CORE_MAX || pid < 0 ||
(cmd != PR_SCHED_CORE_GET && uaddr))
return -EINVAL;
rcu_read_lock();
if (pid == 0) {
task = current;
} else {
task = find_task_by_vpid(pid);
if (!task) {
rcu_read_unlock();
return -ESRCH;
}
}
get_task_struct(task);
rcu_read_unlock();
/*
* Check if this process has the right to modify the specified
* process. Use the regular "ptrace_may_access()" checks.
*/
if (!ptrace_may_access(task, PTRACE_MODE_READ_REALCREDS)) {
err = -EPERM;
goto out;
}
switch (cmd) {
case PR_SCHED_CORE_GET:
if (type != PIDTYPE_PID || uaddr & 7) {
err = -EINVAL;
goto out;
}
cookie = sched_core_clone_cookie(task);
if (cookie) {
/* XXX improve ? */
ptr_to_hashval((void *)cookie, &id);
}
err = put_user(id, (u64 __user *)uaddr);
goto out;
case PR_SCHED_CORE_CREATE:
cookie = sched_core_alloc_cookie();
if (!cookie) {
err = -ENOMEM;
goto out;
}
break;
case PR_SCHED_CORE_SHARE_TO:
cookie = sched_core_clone_cookie(current);
break;
case PR_SCHED_CORE_SHARE_FROM:
if (type != PIDTYPE_PID) {
err = -EINVAL;
goto out;
}
cookie = sched_core_clone_cookie(task);
__sched_core_set(current, cookie);
goto out;
default:
err = -EINVAL;
goto out;
};
if (type == PIDTYPE_PID) {
__sched_core_set(task, cookie);
goto out;
}
read_lock(&tasklist_lock);
grp = task_pid_type(task, type);
do_each_pid_thread(grp, type, p) {
if (!ptrace_may_access(p, PTRACE_MODE_READ_REALCREDS)) {
err = -EPERM;
goto out_tasklist;
}
} while_each_pid_thread(grp, type, p);
do_each_pid_thread(grp, type, p) {
__sched_core_set(p, cookie);
} while_each_pid_thread(grp, type, p);
out_tasklist:
read_unlock(&tasklist_lock);
out:
sched_core_put_cookie(cookie);
put_task_struct(task);
return err;
}