detach_pid(): restore optimization
PATCH 1/2 detach_pid(): restore optimization
Hello.
Kirill's kernel/pid.c rework broke optimization logic in detach_pid(). Non zero return from __detach_pid() was used to indicate, that this pid can probably be freed. Current version always (modulo idle threads) return non zero value, thus resulting in unneccesary pid_hash scanning.
Also, uninlining __detach_pid() reduces pid.o text size from 2492 to 1600 bytes.
Oleg.
kernel/pid.c renamed from kernel/pid.c~ +7 -4
@@ -178,15 +178,18 @@ int fastcall attach_pid(task_t *task, enreturn 0 ; } -static inline int __detach_pid (task_t *task,enum pid_type type) +static fastcall int __detach_pid(task_t *task, enum pid_type type) {struct pid *pid, *pid_next; -int nr; +int nr =0 ; pid = &task->pids[type];if (!hlist_unhashed (&pid->pid_chain)) {hlist_del (&pid->pid_chain); -if (!list_empty (&pid->pid_list)) { + +if (list_empty (&pid->pid_list)) + nr = pid->nr; +else { pid_next =list_entry (pid->pid_list.next, struct pid, pid_list);/* insert next pid from pid_list to hash */
@@ -194,8 +197,8 @@ static inline int __detach_pid(task_t *t &pid_hash[type][pid_hashfn(pid_next->nr)]); } } +list_del (&pid->pid_list); - nr = pid->nr; pid->nr =0 ;return nr;
I can sign off both patches. They look ok.
Quoted 9 lines
Kirill's
kernel/pid.crework broke optimization logic indetach_pid(). Non zero return from__detach_pid()was used to indicate, that this pid can probably be freed. Current version always (modulo idle threads) return non zero value, thus resulting in unneccesarypid_hashscanning.Also, uninlining
__detach_pid()reduces pid.o text size from 2492 to 1600 bytes.Oleg.
[...]
Kirill