Le Sun, 26 Aug 2007 10:28:44 +0530,
Balbir Singh <[EMAIL PROTECTED]> a écrit :
> From what I understand, task->start_time and task->real_start_time
> are taken from the realtime clock. The accounting in CSA seems
> to be very similar to the accounting done in do_acct_process()
> (kernel/acct.c).
>
> Jay/Jonathan any comments?
Thanks, I see it is used to have a wall time version of btime.
Here is my current version of the refactoring for information.
The testcase is coming.
diff -r 27877ed82082 include/linux/tsacct_kern.h
--- a/include/linux/tsacct_kern.h Sat Aug 25 22:17:19 2007 +0200
+++ b/include/linux/tsacct_kern.h Sat Aug 25 21:48:51 2007 +0200
@@ -10,17 +10,23 @@
#include <linux/taskstats.h>
#ifdef CONFIG_TASKSTATS
-extern void bacct_add_tsk(struct taskstats *stats, struct task_struct *tsk);
+void bacct_fill_threadgroup(struct taskstats *stats, struct task_struct *task);
+void bacct_add_tsk(struct taskstats *stats, struct task_struct *task);
#else
-static inline void bacct_add_tsk(struct taskstats *stats, struct task_struct
*tsk)
+static inline void bacct_fill_threadgroup(struct taskstats *stats, struct
task_struct *task)
+{}
+static inline void bacct_add_tsk(struct taskstats *stats, struct task_struct
*task)
{}
#endif /* CONFIG_TASKSTATS */
#ifdef CONFIG_TASK_XACCT
-extern void xacct_add_tsk(struct taskstats *stats, struct task_struct *p);
+void xacct_fill_threadgroup(struct taskstats *stats, struct task_struct *task);
+void xacct_add_tsk(struct taskstats *stats, struct task_struct *p);
extern void acct_update_integrals(struct task_struct *tsk);
extern void acct_clear_integrals(struct task_struct *tsk);
#else
+static inline void xacct_fill_threadgroup(struct taskstats *stats, struct
task_struct *task)
+{}
static inline void xacct_add_tsk(struct taskstats *stats, struct task_struct
*p)
{}
static inline void acct_update_integrals(struct task_struct *tsk)
diff -r 27877ed82082 kernel/taskstats.c
--- a/kernel/taskstats.c Sat Aug 25 22:17:19 2007 +0200
+++ b/kernel/taskstats.c Sat Aug 25 22:11:23 2007 +0200
@@ -168,6 +168,55 @@ static void send_cpu_listeners(struct sk
up_write(&listeners->sem);
}
+/*
+ * Common stats for each thread in the thread group, @task is not necessarily
+ * a thread group leader.
+ */
+static void fill_threadgroup(struct taskstats *stats, struct task_struct *task)
+{
+ /*
+ * Each accounting subsystem adds calls to its functions to initialize
+ * relevant parts of struct taskstsats for a single tgid as follows:
+ *
+ * per-task-foo-fill_threadgroup(stats, task);
+ */
+
+ stats->version = TASKSTATS_VERSION;
+
+ /* fill in basic acct fields */
+ bacct_fill_threadgroup(stats, task);
+
+ /* fill in extended acct fields */
+ xacct_fill_threadgroup(stats, task);
+}
+
+/*
+ * Stats specific to each thread in the thread group. add_tsk() works in
+ * conjunction with fill_threadgroup(), both may write to the same field, so
+ * the ordering of these two calls indicates if the caller wants stats for the
+ * whole thread group or a specific thread.
+ */
+static void add_tsk(struct taskstats *stats, struct task_struct *task)
+{
+ /*
+ * Each accounting subsystem adds calls to its functions to combine
+ * relevant parts of struct taskstsats for a single pid as follows:
+ *
+ * per-task-foo-add_tsk(stats, task);
+ */
+ stats->nvcsw += task->nvcsw;
+ stats->nivcsw += task->nivcsw;
+
+ /* fill in delay acct fields */
+ delayacct_add_tsk(stats, task);
+
+ /* fill in basic acct fields */
+ bacct_add_tsk(stats, task);
+
+ /* fill in extended acct fields */
+ xacct_add_tsk(stats, task);
+}
+
static int fill_pid(pid_t pid, struct task_struct *tsk,
struct taskstats *stats)
{
@@ -185,23 +234,8 @@ static int fill_pid(pid_t pid, struct ta
get_task_struct(tsk);
memset(stats, 0, sizeof(*stats));
- /*
- * Each accounting subsystem adds calls to its functions to
- * fill in relevant parts of struct taskstsats as follows
- *
- * per-task-foo(stats, tsk);
- */
-
- delayacct_add_tsk(stats, tsk);
-
- /* fill in basic acct fields */
- stats->version = TASKSTATS_VERSION;
- stats->nvcsw = tsk->nvcsw;
- stats->nivcsw = tsk->nivcsw;
- bacct_add_tsk(stats, tsk);
-
- /* fill in extended acct fields */
- xacct_add_tsk(stats, tsk);
+ fill_threadgroup(stats, tsk);
+ add_tsk(stats, tsk);
/* Define err: label here if needed */
put_task_struct(tsk);
@@ -233,31 +267,16 @@ static int fill_tgid(pid_t tgid, struct
memset(stats, 0, sizeof(*stats));
tsk = first;
- do {
- if (tsk->exit_state)
- continue;
- /*
- * Accounting subsystem can call its functions here to
- * fill in relevant parts of struct taskstsats as follows
- *
- * per-task-foo(stats, tsk);
- */
- delayacct_add_tsk(stats, tsk);
-
- stats->nvcsw += tsk->nvcsw;
- stats->nivcsw += tsk->nivcsw;
- } while_each_thread(first, tsk);
-
+ do
+ if (!tsk->exit_state)
+ add_tsk(stats, tsk);
+ while_each_thread(first, tsk);
+
+ fill_threadgroup(stats, first);
unlock_task_sighand(first, &flags);
rc = 0;
out:
rcu_read_unlock();
-
- stats->version = TASKSTATS_VERSION;
- /*
- * Accounting subsytems can also add calls here to modify
- * fields of taskstats.
- */
return rc;
}
@@ -265,21 +284,16 @@ static void fill_tgid_exit(struct task_s
static void fill_tgid_exit(struct task_struct *tsk)
{
unsigned long flags;
+ struct taskstats *tg_stats;
spin_lock_irqsave(&tsk->sighand->siglock, flags);
- if (!tsk->signal->stats)
- goto ret;
-
- /*
- * Each accounting subsystem calls its functions here to
- * accumalate its per-task stats for tsk, into the per-tgid structure
- *
- * per-task-foo(tsk->signal->stats, tsk);
- */
- delayacct_add_tsk(tsk->signal->stats, tsk);
-ret:
+ tg_stats = tsk->signal->stats;
+
+ /* fill_threadgroup() will be called by the stats requester */
+ if (tg_stats)
+ add_tsk(tg_stats, tsk);
+
spin_unlock_irqrestore(&tsk->sighand->siglock, flags);
- return;
}
static int add_del_listener(pid_t pid, cpumask_t *maskp, int isadd)
diff -r 27877ed82082 kernel/tsacct.c
--- a/kernel/tsacct.c Sat Aug 25 22:17:19 2007 +0200
+++ b/kernel/tsacct.c Sat Aug 25 21:48:51 2007 +0200
@@ -22,50 +22,71 @@
#include <linux/acct.h>
#include <linux/jiffies.h>
-/*
- * fill in basic accounting fields
- */
-void bacct_add_tsk(struct taskstats *stats, struct task_struct *tsk)
+static void fill_wall_time(struct taskstats *stats, struct task_struct *task)
{
struct timespec uptime, ts;
s64 ac_etime;
+ /* calculate task elapsed time in timespec */
+ do_posix_clock_monotonic_gettime(&uptime);
+ ts = timespec_sub(uptime, task->start_time);
+
+ stats->ac_btime = get_seconds() - ts.tv_sec;
+
+ ac_etime = timespec_to_ns(&ts);
+ do_div(ac_etime, NSEC_PER_USEC);
+ stats->ac_etime = ac_etime;
+}
+
+/*
+ * fill in basic accounting fields
+ */
+
+void bacct_fill_threadgroup(struct taskstats *stats, struct task_struct *task)
+{
+ struct task_struct *leader;
+
BUILD_BUG_ON(TS_COMM_LEN < TASK_COMM_LEN);
- /* calculate task elapsed time in timespec */
- do_posix_clock_monotonic_gettime(&uptime);
- ts = timespec_sub(uptime, tsk->start_time);
- /* rebase elapsed time to usec */
- ac_etime = timespec_to_ns(&ts);
- do_div(ac_etime, NSEC_PER_USEC);
- stats->ac_etime = ac_etime;
- stats->ac_btime = get_seconds() - ts.tv_sec;
- if (thread_group_leader(tsk)) {
- stats->ac_exitcode = tsk->exit_code;
- if (tsk->flags & PF_FORKNOEXEC)
+ rcu_read_lock();
+ stats->ac_ppid = pid_alive(task) ?
+ rcu_dereference(task->real_parent)->tgid : 0;
+ rcu_read_unlock();
+
+ leader = task->group_leader;
+ get_task_struct(leader);
+ fill_wall_time(stats, leader);
+ put_task_struct(leader);
+
+ stats->ac_nice = task_nice(task);
+ stats->ac_sched = task->policy;
+ stats->ac_uid = task->uid;
+ stats->ac_gid = task->gid;
+ stats->ac_pid = task->pid;
+
+ strncpy(stats->ac_comm, task->comm, sizeof(stats->ac_comm));
+}
+
+void bacct_add_tsk(struct taskstats *stats, struct task_struct *task)
+{
+ if (thread_group_leader(task)) {
+ stats->ac_exitcode = task->exit_code;
+ if (task->flags & PF_FORKNOEXEC)
stats->ac_flag |= AFORK;
}
- if (tsk->flags & PF_SUPERPRIV)
+ if (task->flags & PF_SUPERPRIV)
stats->ac_flag |= ASU;
- if (tsk->flags & PF_DUMPCORE)
+ if (task->flags & PF_DUMPCORE)
stats->ac_flag |= ACORE;
- if (tsk->flags & PF_SIGNALED)
+ if (task->flags & PF_SIGNALED)
stats->ac_flag |= AXSIG;
- stats->ac_nice = task_nice(tsk);
- stats->ac_sched = tsk->policy;
- stats->ac_uid = tsk->uid;
- stats->ac_gid = tsk->gid;
- stats->ac_pid = tsk->pid;
- rcu_read_lock();
- stats->ac_ppid = pid_alive(tsk) ?
- rcu_dereference(tsk->real_parent)->tgid : 0;
- rcu_read_unlock();
- stats->ac_utime = cputime_to_msecs(tsk->utime) * USEC_PER_MSEC;
- stats->ac_stime = cputime_to_msecs(tsk->stime) * USEC_PER_MSEC;
- stats->ac_minflt = tsk->min_flt;
- stats->ac_majflt = tsk->maj_flt;
- strncpy(stats->ac_comm, tsk->comm, sizeof(stats->ac_comm));
+ fill_wall_time(stats, task);
+
+ stats->ac_utime += cputime_to_msecs(task->utime) * USEC_PER_MSEC;
+ stats->ac_stime += cputime_to_msecs(task->stime) * USEC_PER_MSEC;
+ stats->ac_minflt += task->min_flt;
+ stats->ac_majflt += task->maj_flt;
}
@@ -76,32 +97,34 @@ void bacct_add_tsk(struct taskstats *sta
/*
* fill in extended accounting fields
*/
-void xacct_add_tsk(struct taskstats *stats, struct task_struct *p)
+void xacct_fill_threadgroup(struct taskstats *stats, struct task_struct *task)
{
struct mm_struct *mm;
- /* convert pages-jiffies to Mbyte-usec */
- stats->coremem = jiffies_to_usecs(p->acct_rss_mem1) * PAGE_SIZE / MB;
- stats->virtmem = jiffies_to_usecs(p->acct_vm_mem1) * PAGE_SIZE / MB;
- mm = get_task_mm(p);
+ mm = get_task_mm(task);
if (mm) {
/* adjust to KB unit */
stats->hiwater_rss = mm->hiwater_rss * PAGE_SIZE / KB;
- stats->hiwater_vm = mm->hiwater_vm * PAGE_SIZE / KB;
+ stats->hiwater_vm = mm->hiwater_vm * PAGE_SIZE / KB;
mmput(mm);
}
- stats->read_char = p->rchar;
- stats->write_char = p->wchar;
- stats->read_syscalls = p->syscr;
- stats->write_syscalls = p->syscw;
+}
+
+void xacct_add_tsk(struct taskstats *stats, struct task_struct *p)
+{
+ /* convert pages-jiffies to Mbyte-usec */
+ stats->coremem += jiffies_to_usecs(p->acct_rss_mem1) * PAGE_SIZE / MB;
+ stats->virtmem += jiffies_to_usecs(p->acct_vm_mem1) * PAGE_SIZE / MB;
+
+ stats->read_char += p->rchar;
+ stats->write_char += p->wchar;
+ stats->read_syscalls += p->syscr;
+ stats->write_syscalls += p->syscw;
+
#ifdef CONFIG_TASK_IO_ACCOUNTING
- stats->read_bytes = p->ioac.read_bytes;
- stats->write_bytes = p->ioac.write_bytes;
- stats->cancelled_write_bytes = p->ioac.cancelled_write_bytes;
-#else
- stats->read_bytes = 0;
- stats->write_bytes = 0;
- stats->cancelled_write_bytes = 0;
+ stats->read_bytes += p->ioac.read_bytes;
+ stats->write_bytes += p->ioac.write_bytes;
+ stats->cancelled_write_bytes += p->ioac.cancelled_write_bytes;
#endif
}
#undef KB
--
Guillaume
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
