LinuxLists.cc - Galbraith patch

[permalink] [raw]

Subject: Re: Galbraith patch

Here is!!

Em Qua, 2010-11-17 às 20:00 +0100, Dhaval Giani escreveu:
> On Wed, Nov 17, 2010 at 7:58 PM, Dhaval Giani <[email protected]> wrote:
> > [CC'ing Mike]
> >
> > On Wed, Nov 17, 2010 at 7:35 PM, Gilberto Nunes
> > <[email protected]> wrote:
> >> Hi folks...
> >>
> >> That is my first post...
> >> I tried compile a vanilla kernel with Galbraith patch, but without
> >> sucess...
> >> When I turn on "Automatic process group scheduling" and try to compile
> >> the kernel, I get this error:
> >>
> >> kernel/sched.c: In function ‘sched_init’:
> >> kernel/sched.c:7756: error: implicit declaration of function
> >> ‘autogroup_init’
> >> make[1]: ** [kernel/sched.o] Erro 1
> >> make: ** [kernel] Erro 2
> >>
> >>
> >> Someone know what is wrong here??
> >>
>
> Could you send your .config as well?
>
> Thanks,
> Dhaval

--
--
Gilberto Nunes

<><

Attachments:

.config (125.29 kB)

2010-11-17 23:21:57

[permalink] [raw]

Subject: Re: Galbraith patch

On Wed, 2010-11-17 at 17:18 -0200, Gilberto Nunes wrote:
> Here is!!

Hm. Looks like you must have a miss-applied patch. Builds fine here
with your config.

-Mike

2010-11-17 23:36:21

[permalink] [raw]

Subject: Re: Galbraith patch

Hi...

On fact, when I try apply the patch, I took some errors. Here's:

patching file include/linux/sched.h
Hunk #1 succeeded at 506 (offset -3 lines).
Hunk #2 succeeded at 575 (offset -3 lines).
Hunk #3 succeeded at 1905 (offset -31 lines).
patching file kernel/sched.c
Hunk #2 succeeded at 613 (offset 7 lines).
Hunk #3 FAILED at 2010.
Hunk #4 succeeded at 7753 (offset -230 lines).
Hunk #5 succeeded at 8283 (offset -230 lines).
Hunk #6 succeeded at 8308 (offset -230 lines).
1 out of 6 hunks FAILED -- saving rejects to file kernel/sched.c.rej
patching file kernel/fork.c
Hunk #1 succeeded at 173 (offset -1 lines).
Hunk #2 succeeded at 902 (offset -4 lines).
patching file drivers/tty/tty_io.c
Hunk #1 FAILED at 3160.
1 out of 1 hunk FAILED -- saving rejects to file
drivers/tty/tty_io.c.rej
patching file kernel/sched_autogroup.h
patching file kernel/sched_autogroup.c
patching file kernel/sysctl.c
Hunk #1 succeeded at 384 (offset 2 lines).
patching file init/Kconfig
Hunk #1 succeeded at 652 with fuzz 2 (offset -76 lines).
patching file Documentation/kernel-parameters.txt
Hunk #1 succeeded at 1610 (offset -12 lines).

Attached, you have the patch.

How can I fix this!

Note: I'm using Ubuntu Linux 10.10... But, as I said before, I'm using
vanilla kernel...

Thanks again

Em Qua, 2010-11-17 às 16:21 -0700, Mike Galbraith escreveu:
> On Wed, 2010-11-17 at 17:18 -0200, Gilberto Nunes wrote:
> > Here is!!
>
> Hm. Looks like you must have a miss-applied patch. Builds fine here
> with your config.
>
> -Mike
>

--
--
Gilberto Nunes

<><

Attachments:

patch (10.73 kB)

2010-11-18 15:43:39

[permalink] [raw]

Subject: Re: Galbraith patch

Hi...

Someone can help with this????

Thanks a lot

Em Qua, 2010-11-17 às 16:21 -0700, Mike Galbraith escreveu:
> On Wed, 2010-11-17 at 17:18 -0200, Gilberto Nunes wrote:
> > Here is!!
>
> Hm. Looks like you must have a miss-applied patch. Builds fine here
> with your config.
>
> -Mike
>

--
--
Gilberto Nunes

<><

Attachments:

patch (10.73 kB)

2010-11-18 16:06:29

[permalink] [raw]

Subject: Re: Galbraith patch

On Thu, 2010-11-18 at 13:43 -0200, Gilberto Nunes wrote:
> Hi...
>
> Someone can help with this????

Hey, patience please, I'm on vacation :)

You can try the below if you like. It's what I'm currently tinkering
with, and has a patch it depends on appended for ease of application.
Should apply cleanly to virgin 2.6.36.

-Mike

A recurring complaint from CFS users is that parallel kbuild has a negative
impact on desktop interactivity. This patch implements an idea from Linus,
to automatically create task groups. This patch implements only per session
autogroups, but leaves the way open for enhancement.

Implementation: each task's signal struct contains an inherited pointer to a
refcounted autogroup struct containing a task group pointer, the default for
all tasks pointing to the init_task_group. When a task calls setsid(), the
process wide reference to the default group is dropped, a new task group is
created, and the process is moved into the new task group. Children thereafter
inherit this task group, and increase it's refcount. On exit, a reference to the
current task group is dropped when the last reference to each signal struct is
dropped. The task group is destroyed when the last signal struct referencing
it is freed. At runqueue selection time, IFF a task has no cgroup assignment,
it's current autogroup is used.

The feature is enabled from boot by default if CONFIG_SCHED_AUTOGROUP is
selected, but can be disabled via the boot option noautogroup, and can be
also be turned on/off on the fly via..
echo [01] > /proc/sys/kernel/sched_autogroup_enabled.
..which will automatically move tasks to/from the root task group.

Some numbers.

A 100% hog overhead measurement proggy pinned to the same CPU as a make -j10

About measurement proggy:
pert/sec = perturbations/sec
min/max/avg = scheduler service latencies in usecs
sum/s = time accrued by the competition per sample period (1 sec here)
overhead = %CPU received by the competition per sample period

pert/s: 31 >40475.37us: 3 min: 0.37 max:48103.60 avg:29573.74 sum/s:916786us overhead:90.24%
pert/s: 23 >41237.70us: 12 min: 0.36 max:56010.39 avg:40187.01 sum/s:924301us overhead:91.99%
pert/s: 24 >42150.22us: 12 min: 8.86 max:61265.91 avg:39459.91 sum/s:947038us overhead:92.20%
pert/s: 26 >42344.91us: 11 min: 3.83 max:52029.60 avg:36164.70 sum/s:940282us overhead:91.12%
pert/s: 24 >44262.90us: 14 min: 5.05 max:82735.15 avg:40314.33 sum/s:967544us overhead:92.22%

Same load with this patch applied.

pert/s: 229 >5484.43us: 41 min: 0.15 max:12069.42 avg:2193.81 sum/s:502382us overhead:50.24%
pert/s: 222 >5652.28us: 43 min: 0.46 max:12077.31 avg:2248.56 sum/s:499181us overhead:49.92%
pert/s: 211 >5809.38us: 43 min: 0.16 max:12064.78 avg:2381.70 sum/s:502538us overhead:50.25%
pert/s: 223 >6147.92us: 43 min: 0.15 max:16107.46 avg:2282.17 sum/s:508925us overhead:50.49%
pert/s: 218 >6252.64us: 43 min: 0.16 max:12066.13 avg:2324.11 sum/s:506656us overhead:50.27%

Average service latency is an order of magnitude better with autogroup.
(Imagine that pert were Xorg or whatnot instead)

Using Mathieu Desnoyers' wakeup-latency testcase:

With taskset -c 3 make -j 10 running..

taskset -c 3 ./wakeup-latency& sleep 30;killall wakeup-latency

without:
maximum latency: 42963.2 µs
average latency: 9077.0 µs
missed timer events: 0

with:
maximum latency: 4160.7 µs
average latency: 149.4 µs
missed timer events: 0

Signed-off-by: Mike Galbraith <[email protected]>

---
Documentation/kernel-parameters.txt | 2
include/linux/sched.h | 19 ++++
init/Kconfig | 12 ++
kernel/fork.c | 5 -
kernel/sched.c | 13 ++
kernel/sched_autogroup.c | 170 ++++++++++++++++++++++++++++++++++++
kernel/sched_autogroup.h | 23 ++++
kernel/sched_debug.c | 29 +++---
kernel/sys.c | 4
kernel/sysctl.c | 11 ++
10 files changed, 270 insertions(+), 18 deletions(-)

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 1e2a6db..a111fac 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -506,6 +506,8 @@ struct thread_group_cputimer {
spinlock_t lock;
};

+struct autogroup;
+
/*
* NOTE! "signal_struct" does not have it's own
* locking, because a shared signal_struct always
@@ -573,6 +575,9 @@ struct signal_struct {

struct tty_struct *tty; /* NULL if no tty */

+#ifdef CONFIG_SCHED_AUTOGROUP
+ struct autogroup *autogroup;
+#endif
/*
* Cumulative resource counters for dead threads in the group,
* and for reaped dead child processes forked by this group.
@@ -1072,7 +1077,7 @@ struct sched_class {
struct task_struct *task);

#ifdef CONFIG_FAIR_GROUP_SCHED
- void (*moved_group) (struct task_struct *p, int on_rq);
+ void (*task_move_group) (struct task_struct *p, int on_rq);
#endif
};

@@ -1900,6 +1905,20 @@ int sched_rt_handler(struct ctl_table *table, int write,

extern unsigned int sysctl_sched_compat_yield;

+#ifdef CONFIG_SCHED_AUTOGROUP
+extern unsigned int sysctl_sched_autogroup_enabled;
+
+extern void sched_autogroup_create_attach(struct task_struct *p);
+extern void sched_autogroup_detach(struct task_struct *p);
+extern void sched_autogroup_fork(struct signal_struct *sig);
+extern void sched_autogroup_exit(struct signal_struct *sig);
+#else
+static inline void sched_autogroup_create_attach(struct task_struct *p) { }
+static inline void sched_autogroup_detach(struct task_struct *p) { }
+static inline void sched_autogroup_fork(struct signal_struct *sig) { }
+static inline void sched_autogroup_exit(struct signal_struct *sig) { }
+#endif
+
#ifdef CONFIG_RT_MUTEXES
extern int rt_mutex_getprio(struct task_struct *p);
extern void rt_mutex_setprio(struct task_struct *p, int prio);
diff --git a/kernel/sched.c b/kernel/sched.c
index dc85ceb..8d1f066 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -78,6 +78,7 @@

#include "sched_cpupri.h"
#include "workqueue_sched.h"
+#include "sched_autogroup.h"

#define CREATE_TRACE_POINTS
#include <trace/events/sched.h>
@@ -268,6 +269,10 @@ struct task_group {
struct task_group *parent;
struct list_head siblings;
struct list_head children;
+
+#if (defined(CONFIG_SCHED_AUTOGROUP) && defined(CONFIG_SCHED_DEBUG))
+ struct autogroup *autogroup;
+#endif
};

#define root_task_group init_task_group
@@ -612,11 +617,14 @@ static inline int cpu_of(struct rq *rq)
*/
static inline struct task_group *task_group(struct task_struct *p)
{
+ struct task_group *tg;
struct cgroup_subsys_state *css;

css = task_subsys_state_check(p, cpu_cgroup_subsys_id,
lockdep_is_held(&task_rq(p)->lock));
- return container_of(css, struct task_group, css);
+ tg = container_of(css, struct task_group, css);
+
+ return autogroup_task_group(p, tg);
}

/* Change a task's cfs_rq and parent entity if it moves across CPUs/groups */
@@ -1920,6 +1928,7 @@ static void deactivate_task(struct rq *rq, struct task_struct *p, int flags)
#include "sched_idletask.c"
#include "sched_fair.c"
#include "sched_rt.c"
+#include "sched_autogroup.c"
#ifdef CONFIG_SCHED_DEBUG
# include "sched_debug.c"
#endif
@@ -7749,7 +7758,7 @@ void __init sched_init(void)
#ifdef CONFIG_CGROUP_SCHED
list_add(&init_task_group.list, &task_groups);
INIT_LIST_HEAD(&init_task_group.children);
-
+ autogroup_init(&init_task);
#endif /* CONFIG_CGROUP_SCHED */

#if defined CONFIG_FAIR_GROUP_SCHED && defined CONFIG_SMP
@@ -8297,12 +8306,12 @@ void sched_move_task(struct task_struct *tsk)
if (unlikely(running))
tsk->sched_class->put_prev_task(rq, tsk);

- set_task_rq(tsk, task_cpu(tsk));
-
#ifdef CONFIG_FAIR_GROUP_SCHED
- if (tsk->sched_class->moved_group)
- tsk->sched_class->moved_group(tsk, on_rq);
+ if (tsk->sched_class->task_move_group)
+ tsk->sched_class->task_move_group(tsk, on_rq);
+ else
#endif
+ set_task_rq(tsk, task_cpu(tsk));

if (unlikely(running))
tsk->sched_class->set_curr_task(rq);
diff --git a/kernel/fork.c b/kernel/fork.c
index c445f8c..61f2802 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -173,8 +173,10 @@ static inline void free_signal_struct(struct signal_struct *sig)

static inline void put_signal_struct(struct signal_struct *sig)
{
- if (atomic_dec_and_test(&sig->sigcnt))
+ if (atomic_dec_and_test(&sig->sigcnt)) {
+ sched_autogroup_exit(sig);
free_signal_struct(sig);
+ }
}

void __put_task_struct(struct task_struct *tsk)
@@ -900,6 +902,7 @@ static int copy_signal(unsigned long clone_flags, struct task_struct *tsk)
posix_cpu_timers_init_group(sig);

tty_audit_fork(sig);
+ sched_autogroup_fork(sig);

sig->oom_adj = current->signal->oom_adj;
sig->oom_score_adj = current->signal->oom_score_adj;
diff --git a/kernel/sys.c b/kernel/sys.c
index 7f5a0cd..2745dcd 100644
--- a/kernel/sys.c
+++ b/kernel/sys.c
@@ -1080,8 +1080,10 @@ SYSCALL_DEFINE0(setsid)
err = session;
out:
write_unlock_irq(&tasklist_lock);
- if (err > 0)
+ if (err > 0) {
proc_sid_connector(group_leader);
+ sched_autogroup_create_attach(group_leader);
+ }
return err;
}

diff --git a/kernel/sched_debug.c b/kernel/sched_debug.c
index 2e1b0d1..44a41d5 100644
--- a/kernel/sched_debug.c
+++ b/kernel/sched_debug.c
@@ -87,6 +87,20 @@ static void print_cfs_group_stats(struct seq_file *m, int cpu,
}
#endif

+#if defined(CONFIG_CGROUP_SCHED) && \
+ (defined(CONFIG_FAIR_GROUP_SCHED) || defined(CONFIG_RT_GROUP_SCHED))
+static void task_group_path(struct task_group *tg, char *buf, int buflen)
+{
+ /* may be NULL if the underlying cgroup isn't fully-created yet */
+ if (!tg->css.cgroup) {
+ buf[0] = '\0';
+ autogroup_path(tg, buf, buflen);
+ return;
+ }
+ cgroup_path(tg->css.cgroup, buf, buflen);
+}
+#endif
+
static void
print_task(struct seq_file *m, struct rq *rq, struct task_struct *p)
{
@@ -115,7 +129,7 @@ print_task(struct seq_file *m, struct rq *rq, struct task_struct *p)
char path[64];

rcu_read_lock();
- cgroup_path(task_group(p)->css.cgroup, path, sizeof(path));
+ task_group_path(task_group(p), path, sizeof(path));
rcu_read_unlock();
SEQ_printf(m, " %s", path);
}
@@ -147,19 +161,6 @@ static void print_rq(struct seq_file *m, struct rq *rq, int rq_cpu)
read_unlock_irqrestore(&tasklist_lock, flags);
}

-#if defined(CONFIG_CGROUP_SCHED) && \
- (defined(CONFIG_FAIR_GROUP_SCHED) || defined(CONFIG_RT_GROUP_SCHED))
-static void task_group_path(struct task_group *tg, char *buf, int buflen)
-{
- /* may be NULL if the underlying cgroup isn't fully-created yet */
- if (!tg->css.cgroup) {
- buf[0] = '\0';
- return;
- }
- cgroup_path(tg->css.cgroup, buf, buflen);
-}
-#endif
-
void print_cfs_rq(struct seq_file *m, int cpu, struct cfs_rq *cfs_rq)
{
s64 MIN_vruntime = -1, min_vruntime, max_vruntime = -1,
diff --git a/kernel/sysctl.c b/kernel/sysctl.c
index 3a45c22..165eb9b 100644
--- a/kernel/sysctl.c
+++ b/kernel/sysctl.c
@@ -384,6 +384,17 @@ static struct ctl_table kern_table[] = {
.mode = 0644,
.proc_handler = proc_dointvec,
},
+#ifdef CONFIG_SCHED_AUTOGROUP
+ {
+ .procname = "sched_autogroup_enabled",
+ .data = &sysctl_sched_autogroup_enabled,
+ .maxlen = sizeof(unsigned int),
+ .mode = 0644,
+ .proc_handler = proc_dointvec,
+ .extra1 = &zero,
+ .extra2 = &one,
+ },
+#endif
#ifdef CONFIG_PROVE_LOCKING
{
.procname = "prove_locking",
diff --git a/init/Kconfig b/init/Kconfig
index 2de5b1c..666fc7e 100644
--- a/init/Kconfig
+++ b/init/Kconfig
@@ -652,6 +652,18 @@ config DEBUG_BLK_CGROUP

endif # CGROUPS

+config SCHED_AUTOGROUP
+ bool "Automatic process group scheduling"
+ select CGROUPS
+ select CGROUP_SCHED
+ select FAIR_GROUP_SCHED
+ help
+ This option optimizes the scheduler for common desktop workloads by
+ automatically creating and populating task groups. This separation
+ of workloads isolates aggressive CPU burners (like build jobs) from
+ desktop applications. Task group autogeneration is currently based
+ upon task session.
+
config MM_OWNER
bool

diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
index 8dd7248..1e02f1f 100644
--- a/Documentation/kernel-parameters.txt
+++ b/Documentation/kernel-parameters.txt
@@ -1610,6 +1610,8 @@ and is between 256 and 4096 characters. It is defined in the file
noapic [SMP,APIC] Tells the kernel to not make use of any
IOAPICs that may be present in the system.

+ noautogroup Disable scheduler automatic task group creation.
+
nobats [PPC] Do not use BATs for mapping kernel lowmem
on "Classic" PPC cores.

diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c
index db3f674..4c44b90 100644
--- a/kernel/sched_fair.c
+++ b/kernel/sched_fair.c
@@ -3824,13 +3824,26 @@ static void set_curr_task_fair(struct rq *rq)
}

#ifdef CONFIG_FAIR_GROUP_SCHED
-static void moved_group_fair(struct task_struct *p, int on_rq)
+static void task_move_group_fair(struct task_struct *p, int on_rq)
{
- struct cfs_rq *cfs_rq = task_cfs_rq(p);
-
- update_curr(cfs_rq);
+ /*
+ * If the task was not on the rq at the time of this cgroup movement
+ * it must have been asleep, sleeping tasks keep their ->vruntime
+ * absolute on their old rq until wakeup (needed for the fair sleeper
+ * bonus in place_entity()).
+ *
+ * If it was on the rq, we've just 'preempted' it, which does convert
+ * ->vruntime to a relative base.
+ *
+ * Make sure both cases convert their relative position when migrating
+ * to another cgroup's rq. This does somewhat interfere with the
+ * fair sleeper stuff for the first placement, but who cares.
+ */
+ if (!on_rq)
+ p->se.vruntime -= cfs_rq_of(&p->se)->min_vruntime;
+ set_task_rq(p, task_cpu(p));
if (!on_rq)
- place_entity(cfs_rq, &p->se, 1);
+ p->se.vruntime += cfs_rq_of(&p->se)->min_vruntime;
}
#endif

@@ -3882,7 +3895,7 @@ static const struct sched_class fair_sched_class = {
.get_rr_interval = get_rr_interval_fair,

#ifdef CONFIG_FAIR_GROUP_SCHED
- .moved_group = moved_group_fair,
+ .task_move_group = task_move_group_fair,
#endif
};

2010-11-18 16:40:56

[permalink] [raw]

Subject: Re: Galbraith patch

OK... excuse me...

Sorry for disturb!

I'll try

Thanks

Em Qui, 2010-11-18 às 09:06 -0700, Mike Galbraith escreveu:
> On Thu, 2010-11-18 at 13:43 -0200, Gilberto Nunes wrote:
> > Hi...
> >
> > Someone can help with this????
>
> Hey, patience please, I'm on vacation :)
>
> You can try the below if you like. It's what I'm currently tinkering
> with, and has a patch it depends on appended for ease of application.
> Should apply cleanly to virgin 2.6.36.
>
> -Mike
>
> A recurring complaint from CFS users is that parallel kbuild has a negative
> impact on desktop interactivity. This patch implements an idea from Linus,
> to automatically create task groups. This patch implements only per session
> autogroups, but leaves the way open for enhancement.
>
> Implementation: each task's signal struct contains an inherited pointer to a
> refcounted autogroup struct containing a task group pointer, the default for
> all tasks pointing to the init_task_group. When a task calls setsid(), the
> process wide reference to the default group is dropped, a new task group is
> created, and the process is moved into the new task group. Children thereafter
> inherit this task group, and increase it's refcount. On exit, a reference to the
> current task group is dropped when the last reference to each signal struct is
> dropped. The task group is destroyed when the last signal struct referencing
> it is freed. At runqueue selection time, IFF a task has no cgroup assignment,
> it's current autogroup is used.
>
> The feature is enabled from boot by default if CONFIG_SCHED_AUTOGROUP is
> selected, but can be disabled via the boot option noautogroup, and can be
> also be turned on/off on the fly via..
> echo [01] > /proc/sys/kernel/sched_autogroup_enabled.
> ..which will automatically move tasks to/from the root task group.
>
> Some numbers.
>
> A 100% hog overhead measurement proggy pinned to the same CPU as a make -j10
>
> About measurement proggy:
> pert/sec = perturbations/sec
> min/max/avg = scheduler service latencies in usecs
> sum/s = time accrued by the competition per sample period (1 sec here)
> overhead = %CPU received by the competition per sample period
>
> pert/s: 31 >40475.37us: 3 min: 0.37 max:48103.60 avg:29573.74 sum/s:916786us overhead:90.24%
> pert/s: 23 >41237.70us: 12 min: 0.36 max:56010.39 avg:40187.01 sum/s:924301us overhead:91.99%
> pert/s: 24 >42150.22us: 12 min: 8.86 max:61265.91 avg:39459.91 sum/s:947038us overhead:92.20%
> pert/s: 26 >42344.91us: 11 min: 3.83 max:52029.60 avg:36164.70 sum/s:940282us overhead:91.12%
> pert/s: 24 >44262.90us: 14 min: 5.05 max:82735.15 avg:40314.33 sum/s:967544us overhead:92.22%
>
> Same load with this patch applied.
>
> pert/s: 229 >5484.43us: 41 min: 0.15 max:12069.42 avg:2193.81 sum/s:502382us overhead:50.24%
> pert/s: 222 >5652.28us: 43 min: 0.46 max:12077.31 avg:2248.56 sum/s:499181us overhead:49.92%
> pert/s: 211 >5809.38us: 43 min: 0.16 max:12064.78 avg:2381.70 sum/s:502538us overhead:50.25%
> pert/s: 223 >6147.92us: 43 min: 0.15 max:16107.46 avg:2282.17 sum/s:508925us overhead:50.49%
> pert/s: 218 >6252.64us: 43 min: 0.16 max:12066.13 avg:2324.11 sum/s:506656us overhead:50.27%
>
> Average service latency is an order of magnitude better with autogroup.
> (Imagine that pert were Xorg or whatnot instead)
>
> Using Mathieu Desnoyers' wakeup-latency testcase:
>
> With taskset -c 3 make -j 10 running..
>
> taskset -c 3 ./wakeup-latency& sleep 30;killall wakeup-latency
>
> without:
> maximum latency: 42963.2 µs
> average latency: 9077.0 µs
> missed timer events: 0
>
> with:
> maximum latency: 4160.7 µs
> average latency: 149.4 µs
> missed timer events: 0
>
> Signed-off-by: Mike Galbraith <[email protected]>
>
> ---
> Documentation/kernel-parameters.txt | 2
> include/linux/sched.h | 19 ++++
> init/Kconfig | 12 ++
> kernel/fork.c | 5 -
> kernel/sched.c | 13 ++
> kernel/sched_autogroup.c | 170 ++++++++++++++++++++++++++++++++++++
> kernel/sched_autogroup.h | 23 ++++
> kernel/sched_debug.c | 29 +++---
> kernel/sys.c | 4
> kernel/sysctl.c | 11 ++
> 10 files changed, 270 insertions(+), 18 deletions(-)
>
> diff --git a/include/linux/sched.h b/include/linux/sched.h
> index 1e2a6db..a111fac 100644
> --- a/include/linux/sched.h
> +++ b/include/linux/sched.h
> @@ -506,6 +506,8 @@ struct thread_group_cputimer {
> spinlock_t lock;
> };
>
> +struct autogroup;
> +
> /*
> * NOTE! "signal_struct" does not have it's own
> * locking, because a shared signal_struct always
> @@ -573,6 +575,9 @@ struct signal_struct {
>
> struct tty_struct *tty; /* NULL if no tty */
>
> +#ifdef CONFIG_SCHED_AUTOGROUP
> + struct autogroup *autogroup;
> +#endif
> /*
> * Cumulative resource counters for dead threads in the group,
> * and for reaped dead child processes forked by this group.
> @@ -1072,7 +1077,7 @@ struct sched_class {
> struct task_struct *task);
>
> #ifdef CONFIG_FAIR_GROUP_SCHED
> - void (*moved_group) (struct task_struct *p, int on_rq);
> + void (*task_move_group) (struct task_struct *p, int on_rq);
> #endif
> };
>
> @@ -1900,6 +1905,20 @@ int sched_rt_handler(struct ctl_table *table, int write,
>
> extern unsigned int sysctl_sched_compat_yield;
>
> +#ifdef CONFIG_SCHED_AUTOGROUP
> +extern unsigned int sysctl_sched_autogroup_enabled;
> +
> +extern void sched_autogroup_create_attach(struct task_struct *p);
> +extern void sched_autogroup_detach(struct task_struct *p);
> +extern void sched_autogroup_fork(struct signal_struct *sig);
> +extern void sched_autogroup_exit(struct signal_struct *sig);
> +#else
> +static inline void sched_autogroup_create_attach(struct task_struct *p) { }
> +static inline void sched_autogroup_detach(struct task_struct *p) { }
> +static inline void sched_autogroup_fork(struct signal_struct *sig) { }
> +static inline void sched_autogroup_exit(struct signal_struct *sig) { }
> +#endif
> +
> #ifdef CONFIG_RT_MUTEXES
> extern int rt_mutex_getprio(struct task_struct *p);
> extern void rt_mutex_setprio(struct task_struct *p, int prio);
> diff --git a/kernel/sched.c b/kernel/sched.c
> index dc85ceb..8d1f066 100644
> --- a/kernel/sched.c
> +++ b/kernel/sched.c
> @@ -78,6 +78,7 @@
>
> #include "sched_cpupri.h"
> #include "workqueue_sched.h"
> +#include "sched_autogroup.h"
>
> #define CREATE_TRACE_POINTS
> #include <trace/events/sched.h>
> @@ -268,6 +269,10 @@ struct task_group {
> struct task_group *parent;
> struct list_head siblings;
> struct list_head children;
> +
> +#if (defined(CONFIG_SCHED_AUTOGROUP) && defined(CONFIG_SCHED_DEBUG))
> + struct autogroup *autogroup;
> +#endif
> };
>
> #define root_task_group init_task_group
> @@ -612,11 +617,14 @@ static inline int cpu_of(struct rq *rq)
> */
> static inline struct task_group *task_group(struct task_struct *p)
> {
> + struct task_group *tg;
> struct cgroup_subsys_state *css;
>
> css = task_subsys_state_check(p, cpu_cgroup_subsys_id,
> lockdep_is_held(&task_rq(p)->lock));
> - return container_of(css, struct task_group, css);
> + tg = container_of(css, struct task_group, css);
> +
> + return autogroup_task_group(p, tg);
> }
>
> /* Change a task's cfs_rq and parent entity if it moves across CPUs/groups */
> @@ -1920,6 +1928,7 @@ static void deactivate_task(struct rq *rq, struct task_struct *p, int flags)
> #include "sched_idletask.c"
> #include "sched_fair.c"
> #include "sched_rt.c"
> +#include "sched_autogroup.c"
> #ifdef CONFIG_SCHED_DEBUG
> # include "sched_debug.c"
> #endif
> @@ -7749,7 +7758,7 @@ void __init sched_init(void)
> #ifdef CONFIG_CGROUP_SCHED
> list_add(&init_task_group.list, &task_groups);
> INIT_LIST_HEAD(&init_task_group.children);
> -
> + autogroup_init(&init_task);
> #endif /* CONFIG_CGROUP_SCHED */
>
> #if defined CONFIG_FAIR_GROUP_SCHED && defined CONFIG_SMP
> @@ -8297,12 +8306,12 @@ void sched_move_task(struct task_struct *tsk)
> if (unlikely(running))
> tsk->sched_class->put_prev_task(rq, tsk);
>
> - set_task_rq(tsk, task_cpu(tsk));
> -
> #ifdef CONFIG_FAIR_GROUP_SCHED
> - if (tsk->sched_class->moved_group)
> - tsk->sched_class->moved_group(tsk, on_rq);
> + if (tsk->sched_class->task_move_group)
> + tsk->sched_class->task_move_group(tsk, on_rq);
> + else
> #endif
> + set_task_rq(tsk, task_cpu(tsk));
>
> if (unlikely(running))
> tsk->sched_class->set_curr_task(rq);
> diff --git a/kernel/fork.c b/kernel/fork.c
> index c445f8c..61f2802 100644
> --- a/kernel/fork.c
> +++ b/kernel/fork.c
> @@ -173,8 +173,10 @@ static inline void free_signal_struct(struct signal_struct *sig)
>
> static inline void put_signal_struct(struct signal_struct *sig)
> {
> - if (atomic_dec_and_test(&sig->sigcnt))
> + if (atomic_dec_and_test(&sig->sigcnt)) {
> + sched_autogroup_exit(sig);
> free_signal_struct(sig);
> + }
> }
>
> void __put_task_struct(struct task_struct *tsk)
> @@ -900,6 +902,7 @@ static int copy_signal(unsigned long clone_flags, struct task_struct *tsk)
> posix_cpu_timers_init_group(sig);
>
> tty_audit_fork(sig);
> + sched_autogroup_fork(sig);
>
> sig->oom_adj = current->signal->oom_adj;
> sig->oom_score_adj = current->signal->oom_score_adj;
> diff --git a/kernel/sys.c b/kernel/sys.c
> index 7f5a0cd..2745dcd 100644
> --- a/kernel/sys.c
> +++ b/kernel/sys.c
> @@ -1080,8 +1080,10 @@ SYSCALL_DEFINE0(setsid)
> err = session;
> out:
> write_unlock_irq(&tasklist_lock);
> - if (err > 0)
> + if (err > 0) {
> proc_sid_connector(group_leader);
> + sched_autogroup_create_attach(group_leader);
> + }
> return err;
> }
>
> diff --git a/kernel/sched_debug.c b/kernel/sched_debug.c
> index 2e1b0d1..44a41d5 100644
> --- a/kernel/sched_debug.c
> +++ b/kernel/sched_debug.c
> @@ -87,6 +87,20 @@ static void print_cfs_group_stats(struct seq_file *m, int cpu,
> }
> #endif
>
> +#if defined(CONFIG_CGROUP_SCHED) && \
> + (defined(CONFIG_FAIR_GROUP_SCHED) || defined(CONFIG_RT_GROUP_SCHED))
> +static void task_group_path(struct task_group *tg, char *buf, int buflen)
> +{
> + /* may be NULL if the underlying cgroup isn't fully-created yet */
> + if (!tg->css.cgroup) {
> + buf[0] = '\0';
> + autogroup_path(tg, buf, buflen);
> + return;
> + }
> + cgroup_path(tg->css.cgroup, buf, buflen);
> +}
> +#endif
> +
> static void
> print_task(struct seq_file *m, struct rq *rq, struct task_struct *p)
> {
> @@ -115,7 +129,7 @@ print_task(struct seq_file *m, struct rq *rq, struct task_struct *p)
> char path[64];
>
> rcu_read_lock();
> - cgroup_path(task_group(p)->css.cgroup, path, sizeof(path));
> + task_group_path(task_group(p), path, sizeof(path));
> rcu_read_unlock();
> SEQ_printf(m, " %s", path);
> }
> @@ -147,19 +161,6 @@ static void print_rq(struct seq_file *m, struct rq *rq, int rq_cpu)
> read_unlock_irqrestore(&tasklist_lock, flags);
> }
>
> -#if defined(CONFIG_CGROUP_SCHED) && \
> - (defined(CONFIG_FAIR_GROUP_SCHED) || defined(CONFIG_RT_GROUP_SCHED))
> -static void task_group_path(struct task_group *tg, char *buf, int buflen)
> -{
> - /* may be NULL if the underlying cgroup isn't fully-created yet */
> - if (!tg->css.cgroup) {
> - buf[0] = '\0';
> - return;
> - }
> - cgroup_path(tg->css.cgroup, buf, buflen);
> -}
> -#endif
> -
> void print_cfs_rq(struct seq_file *m, int cpu, struct cfs_rq *cfs_rq)
> {
> s64 MIN_vruntime = -1, min_vruntime, max_vruntime = -1,
> diff --git a/kernel/sysctl.c b/kernel/sysctl.c
> index 3a45c22..165eb9b 100644
> --- a/kernel/sysctl.c
> +++ b/kernel/sysctl.c
> @@ -384,6 +384,17 @@ static struct ctl_table kern_table[] = {
> .mode = 0644,
> .proc_handler = proc_dointvec,
> },
> +#ifdef CONFIG_SCHED_AUTOGROUP
> + {
> + .procname = "sched_autogroup_enabled",
> + .data = &sysctl_sched_autogroup_enabled,
> + .maxlen = sizeof(unsigned int),
> + .mode = 0644,
> + .proc_handler = proc_dointvec,
> + .extra1 = &zero,
> + .extra2 = &one,
> + },
> +#endif
> #ifdef CONFIG_PROVE_LOCKING
> {
> .procname = "prove_locking",
> diff --git a/init/Kconfig b/init/Kconfig
> index 2de5b1c..666fc7e 100644
> --- a/init/Kconfig
> +++ b/init/Kconfig
> @@ -652,6 +652,18 @@ config DEBUG_BLK_CGROUP
>
> endif # CGROUPS
>
> +config SCHED_AUTOGROUP
> + bool "Automatic process group scheduling"
> + select CGROUPS
> + select CGROUP_SCHED
> + select FAIR_GROUP_SCHED
> + help
> + This option optimizes the scheduler for common desktop workloads by
> + automatically creating and populating task groups. This separation
> + of workloads isolates aggressive CPU burners (like build jobs) from
> + desktop applications. Task group autogeneration is currently based
> + upon task session.
> +
> config MM_OWNER
> bool
>
> diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
> index 8dd7248..1e02f1f 100644
> --- a/Documentation/kernel-parameters.txt
> +++ b/Documentation/kernel-parameters.txt
> @@ -1610,6 +1610,8 @@ and is between 256 and 4096 characters. It is defined in the file
> noapic [SMP,APIC] Tells the kernel to not make use of any
> IOAPICs that may be present in the system.
>
> + noautogroup Disable scheduler automatic task group creation.
> +
> nobats [PPC] Do not use BATs for mapping kernel lowmem
> on "Classic" PPC cores.
>
> diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c
> index db3f674..4c44b90 100644
> --- a/kernel/sched_fair.c
> +++ b/kernel/sched_fair.c
> @@ -3824,13 +3824,26 @@ static void set_curr_task_fair(struct rq *rq)
> }
>
> #ifdef CONFIG_FAIR_GROUP_SCHED
> -static void moved_group_fair(struct task_struct *p, int on_rq)
> +static void task_move_group_fair(struct task_struct *p, int on_rq)
> {
> - struct cfs_rq *cfs_rq = task_cfs_rq(p);
> -
> - update_curr(cfs_rq);
> + /*
> + * If the task was not on the rq at the time of this cgroup movement
> + * it must have been asleep, sleeping tasks keep their ->vruntime
> + * absolute on their old rq until wakeup (needed for the fair sleeper
> + * bonus in place_entity()).
> + *
> + * If it was on the rq, we've just 'preempted' it, which does convert
> + * ->vruntime to a relative base.
> + *
> + * Make sure both cases convert their relative position when migrating
> + * to another cgroup's rq. This does somewhat interfere with the
> + * fair sleeper stuff for the first placement, but who cares.
> + */
> + if (!on_rq)
> + p->se.vruntime -= cfs_rq_of(&p->se)->min_vruntime;
> + set_task_rq(p, task_cpu(p));
> if (!on_rq)
> - place_entity(cfs_rq, &p->se, 1);
> + p->se.vruntime += cfs_rq_of(&p->se)->min_vruntime;
> }
> #endif
>
> @@ -3882,7 +3895,7 @@ static const struct sched_class fair_sched_class = {
> .get_rr_interval = get_rr_interval_fair,
>
> #ifdef CONFIG_FAIR_GROUP_SCHED
> - .moved_group = moved_group_fair,
> + .task_move_group = task_move_group_fair,
> #endif
> };
>
>
>
>
>

--
--
Gilberto Nunes
Departamento de TI
Selbetti Gestão de Documentos
Fone: (47) 3441-6004
Celular (47) 8861-6672

<><

2010-11-18 18:40:36

by Thomas Backlund

[permalink] [raw]

Subject: Re: Galbraith patch

Mike Galbraith skrev 18.11.2010 18:06:
> On Thu, 2010-11-18 at 13:43 -0200, Gilberto Nunes wrote:
>> Hi...
>>
>> Someone can help with this????
>
> Hey, patience please, I'm on vacation :)
>
> You can try the below if you like. It's what I'm currently tinkering
> with, and has a patch it depends on appended for ease of application.
> Should apply cleanly to virgin 2.6.36.
>

[...]

This diffstat does not match the patch:
---
> Documentation/kernel-parameters.txt | 2
> include/linux/sched.h | 19 ++++
> init/Kconfig | 12 ++
> kernel/fork.c | 5 -
> kernel/sched.c | 13 ++
> kernel/sched_autogroup.c | 170 ++++++++++++++++++++++++++++++++++++
> kernel/sched_autogroup.h | 23 ++++
> kernel/sched_debug.c | 29 +++---
> kernel/sys.c | 4
> kernel/sysctl.c | 11 ++

It's really:
> Documentation/kernel-parameters.txt | 2 ++
> include/linux/sched.h | 21 ++++++++++++++++++++-
> init/Kconfig | 12 ++++++++++++
> kernel/fork.c | 5 ++++-
> kernel/sched.c | 21 +++++++++++++++------
> kernel/sched_debug.c | 29 +++++++++++++++--------------
> kernel/sched_fair.c | 25 +++++++++++++++++++------
> kernel/sys.c | 4 +++-
> kernel/sysctl.c | 11 +++++++++++
> 9 files changed, 101 insertions(+), 29 deletions(-)

and more importantly it misses theese:
>> kernel/sched_autogroup.c | 170 ++++++++++++++++++++++++++++++++++++
>> kernel/sched_autogroup.h | 23 ++++

Wich means it wont build...

--
Thomas

2010-11-18 19:32:18

[permalink] [raw]

Subject: Re: Galbraith patch

On Thursday, November 18, 2010, Mike Galbraith wrote:
>On Thu, 2010-11-18 at 13:43 -0200, Gilberto Nunes wrote:
>> Hi...
>>
>> Someone can help with this????
>
>Hey, patience please, I'm on vacation :)
>
>You can try the below if you like. It's what I'm currently tinkering
>with, and has a patch it depends on appended for ease of application.
>Should apply cleanly to virgin 2.6.36.
>
> -Mike
[...]
Unforch is did not apply to a freshly unpacked 2.6.36, Mike.
I didn't see an error in the patch go by: (from my buildit26 script)
====
Applying patch Galbraith-4-latency.patch

patching file include/linux/sched.h
patching file kernel/sched.c
patching file kernel/fork.c
patching file kernel/sys.c
patching file kernel/sched_debug.c
patching file kernel/sysctl.c
patching file init/Kconfig
patching file Documentation/kernel-parameters.txt
patching file kernel/sched_fair.c

patch Galbraith-4-latency.patch done
======

Then I turned on some of the other cgroup stuff too, and make spits this
out:
======
kernel/sched.c:81:29: error: sched_autogroup.h: No such file or directory
kernel/sched.c: In function ‘task_group’:
kernel/sched.c:627: error: implicit declaration of function
‘autogroup_task_group’
kernel/sched.c:627: warning: return makes pointer from integer without a
cast
kernel/sched.c:1931:29: error: sched_autogroup.c: No such file or directory
CC arch/x86/mm/pat.o
kernel/sched.c: In function ‘sched_init’:
kernel/sched.c:7761: error: implicit declaration of function
‘autogroup_init’
make[1]: *** [kernel/sched.o] Error 1
make: *** [kernel] Error 2
make: *** Waiting for unfinished jobs....
========

I repeated from scratch but did a 'make headers_install" before running my
build/install script, but that didn't help. Looks like the patch should
have a new .h file creation too?

Many thanks for your efforts on this Mike. In my case, its kmail rebuilding
its indices that lags this machine several times a day and it looked like
this might help. .config is attached in case I turned on something I
shouldn't have.

It can also wait till you have the sand out of your shoes. :-)

--
Cheers, Gene
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If you're right 90% of the time, why quibble about the remaining 3%?

Attachments:

.config (83.26 kB)

2010-11-18 19:35:22

[permalink] [raw]

Subject: Re: Galbraith patch

On Thursday, November 18, 2010, Thomas Backlund wrote:
>Mike Galbraith skrev 18.11.2010 18:06:
>> On Thu, 2010-11-18 at 13:43 -0200, Gilberto Nunes wrote:
>>> Hi...
>>>
>>> Someone can help with this????
>>
>> Hey, patience please, I'm on vacation :)
>>
>> You can try the below if you like. It's what I'm currently tinkering
>> with, and has a patch it depends on appended for ease of application.
>> Should apply cleanly to virgin 2.6.36.
>
>[...]
>
>This diffstat does not match the patch:
>---
>
>> Documentation/kernel-parameters.txt | 2
>> include/linux/sched.h | 19 ++++
>> init/Kconfig | 12 ++
>> kernel/fork.c | 5 -
>> kernel/sched.c | 13 ++
>> kernel/sched_autogroup.c | 170
>> ++++++++++++++++++++++++++++++++++++ kernel/sched_autogroup.h
>> | 23 ++++
>> kernel/sched_debug.c | 29 +++---
>> kernel/sys.c | 4
>> kernel/sysctl.c | 11 ++
>
>It's really:
>> Documentation/kernel-parameters.txt | 2 ++
>> include/linux/sched.h | 21 ++++++++++++++++++++-
>> init/Kconfig | 12 ++++++++++++
>> kernel/fork.c | 5 ++++-
>> kernel/sched.c | 21 +++++++++++++++------
>> kernel/sched_debug.c | 29
>> +++++++++++++++-------------- kernel/sched_fair.c |
>> 25 +++++++++++++++++++------ kernel/sys.c |
>> 4 +++-
>> kernel/sysctl.c | 11 +++++++++++
>> 9 files changed, 101 insertions(+), 29 deletions(-)
>
>and more importantly it misses theese:
>>> kernel/sched_autogroup.c | 170
>>> ++++++++++++++++++++++++++++++++++++ kernel/sched_autogroup.h
>>> | 23 ++++
>
>Wich means it wont build...

As I just found also.
>
>--
>Thomas

--
Cheers, Gene
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Surprise your boss. Get to work on time.

2010-11-18 20:01:16

[permalink] [raw]

Subject: Re: Galbraith patch

On Thu, 2010-11-18 at 20:40 +0200, Thomas Backlund wrote:

> and more importantly it misses theese:

Heh. Apply the attached, autogroup last.

Attachments:

cgroup_fixup_broken_cgroup_movement.diff (3.01 kB)
sched_autogroup.diff (16.11 kB)
Download all attachments

2010-11-18 20:10:34

[permalink] [raw]

Subject: Re: Galbraith patch

On Thu, 2010-11-18 at 14:35 -0500, Gene Heskett wrote:

> >and more importantly it misses theese:
> >>> kernel/sched_autogroup.c | 170
> >>> ++++++++++++++++++++++++++++++++++++ kernel/sched_autogroup.h
> >>> | 23 ++++
> >
> >Wich means it wont build...
>
> As I just found also.

Creating patch while heading for the door doesn't always work:)

2010-11-18 21:16:43

by Thomas Backlund

[permalink] [raw]

Subject: Re: Galbraith patch

Mike Galbraith skrev 18.11.2010 22:00:
> On Thu, 2010-11-18 at 20:40 +0200, Thomas Backlund wrote:
>
>> and more importantly it misses theese:
>
> Heh. Apply the attached, autogroup last.
>
>

Thanks!

Seems to build ok...

--
Thomas

2010-11-18 21:28:19

[permalink] [raw]

Subject: Re: Galbraith patch

On Thursday, November 18, 2010, Mike Galbraith wrote:
>On Thu, 2010-11-18 at 20:40 +0200, Thomas Backlund wrote:
>> and more importantly it misses theese:
>Heh. Apply the attached, autogroup last.

Still incomplete I'd guess. From the make bzimage:
In file included from kernel/sched.c:1931:
kernel/sched_autogroup.c: In function ‘autogroup_create’:
kernel/sched_autogroup.c:56: error: ‘struct task_group’ has no member named
‘autogroup’

--
Cheers, Gene
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
We gave you an atomic bomb, what do you want, mermaids?
-- I. I. Rabi to the Atomic Energy Commission

2010-11-18 21:29:22

[permalink] [raw]

Subject: Re: Galbraith patch

On Thursday, November 18, 2010, Mike Galbraith wrote:
>On Thu, 2010-11-18 at 14:35 -0500, Gene Heskett wrote:
>> >and more importantly it misses theese:
>> >>> kernel/sched_autogroup.c | 170
>> >>> ++++++++++++++++++++++++++++++++++++ kernel/sched_autogroup.h
>> >>>
>> >>> | 23 ++++
>> >
>> >Wich means it wont build...
>>
>> As I just found also.
>
>Creating patch while heading for the door doesn't always work:)

Chuckle.

--
Cheers, Gene
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
We gave you an atomic bomb, what do you want, mermaids?
-- I. I. Rabi to the Atomic Energy Commission

2010-11-18 21:36:42

[permalink] [raw]

Subject: Re: Galbraith patch

On Thursday, November 18, 2010, Thomas Backlund wrote:
>Mike Galbraith skrev 18.11.2010 22:00:
>> On Thu, 2010-11-18 at 20:40 +0200, Thomas Backlund wrote:
>>> and more importantly it misses theese:
>> Heh. Apply the attached, autogroup last.
>
>Thanks!
>
>Seems to build ok...
>
But does not build here until I turn the autogroup option off. So how do we
use it then?

See Mike, you should have stayed the hell away from a comm port of any
kind... Generally speaking, we would probably be just as dumb and happy
next week, except its thanksgiving week and I won't have time due to a
wedding 5 states south of me. rc3 maybe? Or .37?

>--
>Thomas
>--
>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/

--
Cheers, Gene
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Swahili, n.:
The language used by the National Enquirer to print their retractions.
-- Johnny Hart

2010-11-18 22:29:43

by Thomas Backlund

[permalink] [raw]

Subject: Re: Galbraith patch

Gene Heskett skrev 18.11.2010 23:28:
> On Thursday, November 18, 2010, Mike Galbraith wrote:
>> On Thu, 2010-11-18 at 20:40 +0200, Thomas Backlund wrote:
>>> and more importantly it misses theese:
>> Heh. Apply the attached, autogroup last.
>
> Still incomplete I'd guess. From the make bzimage:
> In file included from kernel/sched.c:1931:
> kernel/sched_autogroup.c: In function ‘autogroup_create’:
> kernel/sched_autogroup.c:56: error: ‘struct task_group’ has no member named
> ‘autogroup’
>

Please recheck that you are using only the 2 patches Mike sent.
With them applied to Mandriva 2.6.36 kernel it built fine with
CONFIG_SCHED_AUTOGROUP=y

--
Thomas

2010-11-19 14:09:58

by Jordi Pujol

[permalink] [raw]

Subject: Re: Galbraith patch

Hello,

it requires a little modification,

CONFIG_SCHED_DEBUG should not be a condition to create autogroup in structure
task_group

remove
&& defined(CONFIG_SCHED_DEBUG)
in the following conditional:

+++ linux-2.6.36/kernel/sched.c
@@ -78,6 +78,7 @@

#include "sched_cpupri.h"
#include "workqueue_sched.h"
+#include "sched_autogroup.h"

#define CREATE_TRACE_POINTS
#include <trace/events/sched.h>
@@ -268,6 +269,10 @@ struct task_group {
struct task_group *parent;
struct list_head siblings;
struct list_head children;
+
+#if (defined(CONFIG_SCHED_AUTOGROUP) && defined(CONFIG_SCHED_DEBUG))
+ struct autogroup *autogroup;
+#endif
};

Kind,

Jordi Pujol

Live never ending Tale
GNU/Linux Live forever!
http://livenet.selfip.com

2010-11-19 15:05:18

[permalink] [raw]

Subject: Re: Galbraith patch

On Friday, November 19, 2010, Jordi Pujol wrote:
>Hello,
>
>it requires a little modification,
>
>CONFIG_SCHED_DEBUG should not be a condition to create autogroup in
>structure task_group
>
>remove
> && defined(CONFIG_SCHED_DEBUG)
>in the following conditional:
>
>+++ linux-2.6.36/kernel/sched.c
>@@ -78,6 +78,7 @@
>
> #include "sched_cpupri.h"
> #include "workqueue_sched.h"
>+#include "sched_autogroup.h"
>
> #define CREATE_TRACE_POINTS
> #include <trace/events/sched.h>
>@@ -268,6 +269,10 @@ struct task_group {
> struct task_group *parent;
> struct list_head siblings;
> struct list_head children;
>+
>+#if (defined(CONFIG_SCHED_AUTOGROUP) && defined(CONFIG_SCHED_DEBUG))
>+ struct autogroup *autogroup;
>+#endif
> };
>
>Kind,
>
>Jordi Pujol

Thanks Jordi, but my last attempted build, using only the last 2 patches
that Mike posted on a fresh unpack of 2.6.36, couldn't get past stage 2 of
the modules build, it kept squawking about all the IR control stuff that is
not even configured to be built.

I wonder how long it will take to get work this into mainline?

Along the same line, how does this compare to CK's bfs patch?

>Live never ending Tale
>GNU/Linux Live forever!
>http://livenet.selfip.com

--
Cheers, Gene
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
"What's the use of a good quotation if you can't change it?"
-- The Doctor

2010-11-19 16:10:17

[permalink] [raw]

Subject: Re: Galbraith patch

On Fri, 2010-11-19 at 15:09 +0100, Jordi Pujol wrote:
> Hello,
>
> it requires a little modification,
>
> CONFIG_SCHED_DEBUG should not be a condition to create autogroup in structure
> task_group

Yeah, thanks. It was initially a debug only thing, but evolution
changed that.

-Mike

2010-11-19 16:21:27

[permalink] [raw]

Subject: Re: Galbraith patch

On Fri, 2010-11-19 at 10:05 -0500, Gene Heskett wrote:
> On Friday, November 19, 2010, Jordi Pujol wrote:
> >Hello,
> >
> >it requires a little modification,
> >
> >CONFIG_SCHED_DEBUG should not be a condition to create autogroup in
> >structure task_group
> >
> >remove
> > && defined(CONFIG_SCHED_DEBUG)
> >in the following conditional:
> >
> >+++ linux-2.6.36/kernel/sched.c
> >@@ -78,6 +78,7 @@
> >
> > #include "sched_cpupri.h"
> > #include "workqueue_sched.h"
> >+#include "sched_autogroup.h"
> >
> > #define CREATE_TRACE_POINTS
> > #include <trace/events/sched.h>
> >@@ -268,6 +269,10 @@ struct task_group {
> > struct task_group *parent;
> > struct list_head siblings;
> > struct list_head children;
> >+
> >+#if (defined(CONFIG_SCHED_AUTOGROUP) && defined(CONFIG_SCHED_DEBUG))
> >+ struct autogroup *autogroup;
> >+#endif
> > };
> >
> >Kind,
> >
> >Jordi Pujol
>
> Thanks Jordi, but my last attempted build, using only the last 2 patches
> that Mike posted on a fresh unpack of 2.6.36, couldn't get past stage 2 of
> the modules build, it kept squawking about all the IR control stuff that is
> not even configured to be built.

No idea what's going on with that. Maybe you should wait until I get
the /proc display/tweak interface cobbled together and post a checked
over patch instead of hasty WIP snapshot.

> I wonder how long it will take to get work this into mainline?

If it goes in at all, it won't be before .38.

> Along the same line, how does this compare to CK's bfs patch?

Like apple vs orange.

-Mike

2010-11-19 16:49:21

by Jordi Pujol

[permalink] [raw]

Subject: Re: Galbraith patch

A Divendres 19 Novembre 2010 16:05:12, Gene Heskett va escriure:
> Thanks Jordi, but my last attempted build, using only the last 2 patches
> that Mike posted on a fresh unpack of 2.6.36, couldn't get past stage 2 of
> the modules build, it kept squawking about all the IR control stuff that
> is not even configured to be built.
>
> I wonder how long it will take to get work this into mainline?
>
> Along the same line, how does this compare to CK's bfs patch?

the autogroup-sched patch works well, right now writing this message from a
computer that runs that kernel.

this patch is much simpler than CK2, so I suppose that the implementation
should be more easy,
I am trying both and observe, from my user's point of view, that they give a
similar result.

Gene, if that may help to do your own kernel implementation, the Debian kernel
sources I've used are available at:
http://livenet.selfip.com/ftp/debian/linux-2.6/

2.6.36 + CK2, in the subdirectory
linux-2.6.36-0.jpp.10-lnet/
2.6.36 + autogroup_sched, in the subdirectory
linux-2.6.36-0.jpp.11-lnet/

Jordi Pujol

Live never ending Tale
GNU/Linux Live forever!
http://livenet.selfip.com

2010-11-19 17:11:35

[permalink] [raw]

Subject: Re: Galbraith patch

HI...

I have compile a kernel 2.6.36 with autogroup but I don't see
any /proc/sys/kernel/sched_autogroup_enabled

Also, I tried with noautogroup and autogroup on kernel line boot, but it
doesn't work...

Maybe I do samething wrong...

Em Sex, 2010-11-19 às 17:49 +0100, Jordi Pujol escreveu:
> A Divendres 19 Novembre 2010 16:05:12, Gene Heskett va escriure:
> > Thanks Jordi, but my last attempted build, using only the last 2 patches
> > that Mike posted on a fresh unpack of 2.6.36, couldn't get past stage 2 of
> > the modules build, it kept squawking about all the IR control stuff that
> > is not even configured to be built.
> >
> > I wonder how long it will take to get work this into mainline?
> >
> > Along the same line, how does this compare to CK's bfs patch?
>
> the autogroup-sched patch works well, right now writing this message from a
> computer that runs that kernel.
>
> this patch is much simpler than CK2, so I suppose that the implementation
> should be more easy,
> I am trying both and observe, from my user's point of view, that they give a
> similar result.
>
> Gene, if that may help to do your own kernel implementation, the Debian kernel
> sources I've used are available at:
> http://livenet.selfip.com/ftp/debian/linux-2.6/
>
> 2.6.36 + CK2, in the subdirectory
> linux-2.6.36-0.jpp.10-lnet/
> 2.6.36 + autogroup_sched, in the subdirectory
> linux-2.6.36-0.jpp.11-lnet/
>
> Jordi Pujol
>
> Live never ending Tale
> GNU/Linux Live forever!
> http://livenet.selfip.com
> --
> 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/

--
--
Gilberto Nunes
Departamento de TI
Selbetti Gestão de Documentos
Fone: (47) 3441-6004
Celular (47) 8861-6672

<><

2010-11-20 00:05:10