itimers: remove the per-cpu-ish-ness

Either we bounce once cacheline per cpu per tick, yielding n^2 bounces
or we just bounce a single..

Also, using per-cpu allocations for the thread-groups complicates the
per-cpu allocator in that its currently aimed to be a fixed sized
allocator and the only possible extention to that would be vmap based,
which is seriously constrained on 32 bit archs.

So making the per-cpu memory requirement depend on the number of
processes is an issue.

Lastly, it didn't deal with cpu-hotplug, although admittedly that might
be fixable.

Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

1 files changed, 18 insertions, 11 deletions

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 4cae9b81a1f8..c20943eabb4c 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -450,6 +450,7 @@ struct task_cputime {
         cputime_t utime;
         cputime_t stime;
         unsigned long long sum_exec_runtime;
+        spinlock_t lock;
 };
 /* Alternate field names when used to cache expirations. */
 #define prof_exp        stime
@@ -465,7 +466,7 @@ struct task_cputime {
  * used for thread group CPU clock calculations.
  */
 struct thread_group_cputime {
-        struct task_cputime *totals;
+        struct task_cputime totals;
 };
 
 /*
@@ -2180,24 +2181,30 @@ static inline int spin_needbreak(spinlock_t *lock)
  * Thread group CPU time accounting.
  */
 
-extern int thread_group_cputime_alloc(struct task_struct *);
-extern void thread_group_cputime(struct task_struct *, struct task_cputime *);
-
-static inline void thread_group_cputime_init(struct signal_struct *sig)
+static inline
+void thread_group_cputime(struct task_struct *tsk, struct task_cputime *times)
 {
-        sig->cputime.totals = NULL;
+        struct task_cputime *totals = &tsk->signal->cputime.totals;
+        unsigned long flags;
+
+        spin_lock_irqsave(&totals->lock, flags);
+        *times = *totals;
+        spin_unlock_irqrestore(&totals->lock, flags);
 }
 
-static inline int thread_group_cputime_clone_thread(struct task_struct *curr)
+static inline void thread_group_cputime_init(struct signal_struct *sig)
 {
-        if (curr->signal->cputime.totals)
-                return 0;
-        return thread_group_cputime_alloc(curr);
+        sig->cputime.totals = (struct task_cputime){
+                .utime = cputime_zero,
+                .stime = cputime_zero,
+                .sum_exec_runtime = 0,
+        };
+
+        spin_lock_init(&sig->cputime.totals.lock);
 }
 
 static inline void thread_group_cputime_free(struct signal_struct *sig)
 {
-        free_percpu(sig->cputime.totals);
 }
 
 /*