[PATCH] Cell SPU task notification

[PATCH] Cell SPU task notification

[Maynard Johnson]

[-- Attachment #1: Type: text/plain, Size: 1 bytes --]



[-- Attachment #2: spu-notifier.patch --]
[-- Type: text/x-diff, Size: 4586 bytes --]

Subject: Enable SPU switch notification to detect currently active SPU tasks.

From: Maynard Johnson <maynardj@us.ibm.com>

This patch adds to the capability of spu_switch_event_register so that the
caller is also notified of currently active SPU tasks.  It also exports
spu_switch_event_register and spu_switch_event_unregister.

Signed-off-by: Maynard Johnson <mpjohn@us.ibm.com>


Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c
===================================================================
--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/sched.c	2006-12-04 10:56:04.730698720 -0600
+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c	2007-01-11 09:45:37.918333128 -0600
@@ -46,6 +46,8 @@
 
 #define SPU_MIN_TIMESLICE 	(100 * HZ / 1000)
 
+int notify_active[MAX_NUMNODES];
+
 #define SPU_BITMAP_SIZE (((MAX_PRIO+BITS_PER_LONG)/BITS_PER_LONG)+1)
 struct spu_prio_array {
 	unsigned long bitmap[SPU_BITMAP_SIZE];
@@ -81,18 +83,45 @@
 static void spu_switch_notify(struct spu *spu, struct spu_context *ctx)
 {
 	blocking_notifier_call_chain(&spu_switch_notifier,
-			    ctx ? ctx->object_id : 0, spu);
+				     ctx ? ctx->object_id : 0, spu);
+}
+
+static void notify_spus_active(void)
+{
+	int node;
+	/* Wake up the active spu_contexts. When the awakened processes 
+	 * sees their notify_active flag is set, they will call
+	 * spu_notify_already_active().
+	 */
+	for (node = 0; node < MAX_NUMNODES; node++) {
+		struct spu *spu;
+		mutex_lock(&spu_prio->active_mutex[node]);
+                list_for_each_entry(spu, &spu_prio->active_list[node], list) {
+			struct spu_context *ctx = spu->ctx;
+			wake_up_all(&ctx->stop_wq);
+			notify_active[ctx->spu->number] = 1;
+			smp_mb();
+		}
+                mutex_unlock(&spu_prio->active_mutex[node]);
+	}
+	yield();
 }
 
 int spu_switch_event_register(struct notifier_block * n)
 {
-	return blocking_notifier_chain_register(&spu_switch_notifier, n);
+	int ret;
+	ret = blocking_notifier_chain_register(&spu_switch_notifier, n);
+	if (!ret)
+		notify_spus_active();
+	return ret;
 }
+EXPORT_SYMBOL_GPL(spu_switch_event_register);
 
 int spu_switch_event_unregister(struct notifier_block * n)
 {
 	return blocking_notifier_chain_unregister(&spu_switch_notifier, n);
 }
+EXPORT_SYMBOL_GPL(spu_switch_event_unregister);
 
 
 static inline void bind_context(struct spu *spu, struct spu_context *ctx)
@@ -250,6 +279,14 @@
 	return spu_get_idle(ctx, flags);
 }
 
+void spu_notify_already_active(struct spu_context *ctx)
+{
+	struct spu *spu = ctx->spu;
+	if (!spu)
+		return;
+	spu_switch_notify(spu, ctx);
+}
+
 /* The three externally callable interfaces
  * for the scheduler begin here.
  *
Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/spufs.h
===================================================================
--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/spufs.h	2007-01-08 18:18:40.093354608 -0600
+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/spufs.h	2007-01-08 18:31:03.610345792 -0600
@@ -183,6 +183,7 @@
 void spu_yield(struct spu_context *ctx);
 int __init spu_sched_init(void);
 void __exit spu_sched_exit(void);
+void spu_notify_already_active(struct spu_context *ctx);
 
 extern char *isolated_loader;
 
Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/run.c
===================================================================
--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/run.c	2007-01-08 18:33:51.979311680 -0600
+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/run.c	2007-01-11 10:17:20.777344984 -0600
@@ -10,6 +10,8 @@
 
 #include "spufs.h"
 
+extern int notify_active[MAX_NUMNODES];
+
 /* interrupt-level stop callback function. */
 void spufs_stop_callback(struct spu *spu)
 {
@@ -45,7 +47,9 @@
 	u64 pte_fault;
 
 	*stat = ctx->ops->status_read(ctx);
-	if (ctx->state != SPU_STATE_RUNNABLE)
+	smp_mb();
+
+	if (ctx->state != SPU_STATE_RUNNABLE || notify_active[ctx->spu->number])
 		return 1;
 	spu = ctx->spu;
 	pte_fault = spu->dsisr &
@@ -319,6 +323,11 @@
 		ret = spufs_wait(ctx->stop_wq, spu_stopped(ctx, &status));
 		if (unlikely(ret))
 			break;
+		if (unlikely(notify_active[ctx->spu->number])) {
+			notify_active[ctx->spu->number] = 0;
+			if (!(status & SPU_STATUS_STOPPED_BY_STOP))
+				spu_notify_already_active(ctx);
+		}
 		if ((status & SPU_STATUS_STOPPED_BY_STOP) &&
 		    (status >> SPU_STOP_STATUS_SHIFT == 0x2104)) {
 			ret = spu_process_callback(ctx);

Re: [PATCH] Cell SPU task notification

[Michael Ellerman]

[-- Attachment #1: Type: text/plain, Size: 6175 bytes --]

> Subject: Enable SPU switch notification to detect currently active SPU tasks.
> 
> From: Maynard Johnson <maynardj@us.ibm.com>
> 
> This patch adds to the capability of spu_switch_event_register so that the
> caller is also notified of currently active SPU tasks.  It also exports
> spu_switch_event_register and spu_switch_event_unregister.

Hi Maynard,

It'd be really good if you could convince your mailer to send patches inline :)

> Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c
> ===================================================================
> --- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/sched.c	2006-12-04 10:56:04.730698720 -0600
> +++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c	2007-01-11 09:45:37.918333128 -0600
> @@ -46,6 +46,8 @@
>  
>  #define SPU_MIN_TIMESLICE 	(100 * HZ / 1000)
>  
> +int notify_active[MAX_NUMNODES];

You're basing the size of the array on MAX_NUMNODES
(1 << CONFIG_NODES_SHIFT), but then indexing it by spu->number.

It's quite possible we'll have a system with MAX_NUMNODES == 1, but > 1
spus, in which case this code is going to break. The PS3 is one such
system.

Instead I think you should have a flag in the spu struct.

>  #define SPU_BITMAP_SIZE (((MAX_PRIO+BITS_PER_LONG)/BITS_PER_LONG)+1)
>  struct spu_prio_array {
>  	unsigned long bitmap[SPU_BITMAP_SIZE];
> @@ -81,18 +83,45 @@
>  static void spu_switch_notify(struct spu *spu, struct spu_context *ctx)
>  {
>  	blocking_notifier_call_chain(&spu_switch_notifier,
> -			    ctx ? ctx->object_id : 0, spu);
> +				     ctx ? ctx->object_id : 0, spu);
> +}

Try not to make whitespace only changes in the same patch as actual code changes.

> +
> +static void notify_spus_active(void)
> +{
> +	int node;
> +	/* Wake up the active spu_contexts. When the awakened processes 
> +	 * sees their notify_active flag is set, they will call
> +	 * spu_notify_already_active().
> +	 */
> +	for (node = 0; node < MAX_NUMNODES; node++) {
> +		struct spu *spu;
> +		mutex_lock(&spu_prio->active_mutex[node]);
> +                list_for_each_entry(spu, &spu_prio->active_list[node], list) {
> +			struct spu_context *ctx = spu->ctx;
> +			wake_up_all(&ctx->stop_wq);
> +			notify_active[ctx->spu->number] = 1;
> +			smp_mb();
> +		}

I don't understand why you're setting the notify flag after you do the
wake_up_all() ?

You only need a smp_wmb() here.

Does the scheduler guarantee that ctxs won't swap nodes? Otherwise
between releasing the lock on one node and getting the lock on the next,
a ctx could migrate between them - which would cause either spurious
wake ups, or missing a ctx altogether. Although I'm not sure if it's
that important.

> +                mutex_unlock(&spu_prio->active_mutex[node]);
> +	}
> +	yield();
>  }
>  
>  int spu_switch_event_register(struct notifier_block * n)
>  {
> -	return blocking_notifier_chain_register(&spu_switch_notifier, n);
> +	int ret;
> +	ret = blocking_notifier_chain_register(&spu_switch_notifier, n);
> +	if (!ret)
> +		notify_spus_active();
> +	return ret;
>  }
> +EXPORT_SYMBOL_GPL(spu_switch_event_register);
>  
>  int spu_switch_event_unregister(struct notifier_block * n)
>  {
>  	return blocking_notifier_chain_unregister(&spu_switch_notifier, n);
>  }
> +EXPORT_SYMBOL_GPL(spu_switch_event_unregister);
>  
>  
>  static inline void bind_context(struct spu *spu, struct spu_context *ctx)
> @@ -250,6 +279,14 @@
>  	return spu_get_idle(ctx, flags);
>  }
>  
> +void spu_notify_already_active(struct spu_context *ctx)
> +{
> +	struct spu *spu = ctx->spu;
> +	if (!spu)
> +		return;
> +	spu_switch_notify(spu, ctx);
> +}
> +
>  /* The three externally callable interfaces
>   * for the scheduler begin here.
>   *
> Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/spufs.h
> ===================================================================
> --- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/spufs.h	2007-01-08 18:18:40.093354608 -0600
> +++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/spufs.h	2007-01-08 18:31:03.610345792 -0600
> @@ -183,6 +183,7 @@
>  void spu_yield(struct spu_context *ctx);
>  int __init spu_sched_init(void);
>  void __exit spu_sched_exit(void);
> +void spu_notify_already_active(struct spu_context *ctx);
>  
>  extern char *isolated_loader;
>  
> Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/run.c
> ===================================================================
> --- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/run.c	2007-01-08 18:33:51.979311680 -0600
> +++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/run.c	2007-01-11 10:17:20.777344984 -0600
> @@ -10,6 +10,8 @@
>  
>  #include "spufs.h"
>  
> +extern int notify_active[MAX_NUMNODES];
> +
>  /* interrupt-level stop callback function. */
>  void spufs_stop_callback(struct spu *spu)
>  {
> @@ -45,7 +47,9 @@
>  	u64 pte_fault;
>  
>  	*stat = ctx->ops->status_read(ctx);
> -	if (ctx->state != SPU_STATE_RUNNABLE)
> +	smp_mb();

And smp_rmb() should be sufficient here.

> +	if (ctx->state != SPU_STATE_RUNNABLE || notify_active[ctx->spu->number])
>  		return 1;
>  	spu = ctx->spu;
>  	pte_fault = spu->dsisr &
> @@ -319,6 +323,11 @@
>  		ret = spufs_wait(ctx->stop_wq, spu_stopped(ctx, &status));
>  		if (unlikely(ret))
>  			break;
> +		if (unlikely(notify_active[ctx->spu->number])) {
> +			notify_active[ctx->spu->number] = 0;
> +			if (!(status & SPU_STATUS_STOPPED_BY_STOP))
> +				spu_notify_already_active(ctx);
> +		}
>  		if ((status & SPU_STATUS_STOPPED_BY_STOP) &&
>  		    (status >> SPU_STOP_STATUS_SHIFT == 0x2104)) {
>  			ret = spu_process_callback(ctx);

cheers

-- 
Michael Ellerman
OzLabs, IBM Australia Development Lab

wwweb: http://michael.ellerman.id.au
phone: +61 2 6212 1183 (tie line 70 21183)

We do not inherit the earth from our ancestors,
we borrow it from our children. - S.M.A.R.T Person

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [PATCH] Cell SPU task notification

[Michael Ellerman]

[-- Attachment #1: Type: text/plain, Size: 6175 bytes --]

> Subject: Enable SPU switch notification to detect currently active SPU tasks.
> 
> From: Maynard Johnson <maynardj@us.ibm.com>
> 
> This patch adds to the capability of spu_switch_event_register so that the
> caller is also notified of currently active SPU tasks.  It also exports
> spu_switch_event_register and spu_switch_event_unregister.

Hi Maynard,

It'd be really good if you could convince your mailer to send patches inline :)

> Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c
> ===================================================================
> --- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/sched.c	2006-12-04 10:56:04.730698720 -0600
> +++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c	2007-01-11 09:45:37.918333128 -0600
> @@ -46,6 +46,8 @@
>  
>  #define SPU_MIN_TIMESLICE 	(100 * HZ / 1000)
>  
> +int notify_active[MAX_NUMNODES];

You're basing the size of the array on MAX_NUMNODES
(1 << CONFIG_NODES_SHIFT), but then indexing it by spu->number.

It's quite possible we'll have a system with MAX_NUMNODES == 1, but > 1
spus, in which case this code is going to break. The PS3 is one such
system.

Instead I think you should have a flag in the spu struct.

>  #define SPU_BITMAP_SIZE (((MAX_PRIO+BITS_PER_LONG)/BITS_PER_LONG)+1)
>  struct spu_prio_array {
>  	unsigned long bitmap[SPU_BITMAP_SIZE];
> @@ -81,18 +83,45 @@
>  static void spu_switch_notify(struct spu *spu, struct spu_context *ctx)
>  {
>  	blocking_notifier_call_chain(&spu_switch_notifier,
> -			    ctx ? ctx->object_id : 0, spu);
> +				     ctx ? ctx->object_id : 0, spu);
> +}

Try not to make whitespace only changes in the same patch as actual code changes.

> +
> +static void notify_spus_active(void)
> +{
> +	int node;
> +	/* Wake up the active spu_contexts. When the awakened processes 
> +	 * sees their notify_active flag is set, they will call
> +	 * spu_notify_already_active().
> +	 */
> +	for (node = 0; node < MAX_NUMNODES; node++) {
> +		struct spu *spu;
> +		mutex_lock(&spu_prio->active_mutex[node]);
> +                list_for_each_entry(spu, &spu_prio->active_list[node], list) {
> +			struct spu_context *ctx = spu->ctx;
> +			wake_up_all(&ctx->stop_wq);
> +			notify_active[ctx->spu->number] = 1;
> +			smp_mb();
> +		}

I don't understand why you're setting the notify flag after you do the
wake_up_all() ?

You only need a smp_wmb() here.

Does the scheduler guarantee that ctxs won't swap nodes? Otherwise
between releasing the lock on one node and getting the lock on the next,
a ctx could migrate between them - which would cause either spurious
wake ups, or missing a ctx altogether. Although I'm not sure if it's
that important.

> +                mutex_unlock(&spu_prio->active_mutex[node]);
> +	}
> +	yield();
>  }
>  
>  int spu_switch_event_register(struct notifier_block * n)
>  {
> -	return blocking_notifier_chain_register(&spu_switch_notifier, n);
> +	int ret;
> +	ret = blocking_notifier_chain_register(&spu_switch_notifier, n);
> +	if (!ret)
> +		notify_spus_active();
> +	return ret;
>  }
> +EXPORT_SYMBOL_GPL(spu_switch_event_register);
>  
>  int spu_switch_event_unregister(struct notifier_block * n)
>  {
>  	return blocking_notifier_chain_unregister(&spu_switch_notifier, n);
>  }
> +EXPORT_SYMBOL_GPL(spu_switch_event_unregister);
>  
>  
>  static inline void bind_context(struct spu *spu, struct spu_context *ctx)
> @@ -250,6 +279,14 @@
>  	return spu_get_idle(ctx, flags);
>  }
>  
> +void spu_notify_already_active(struct spu_context *ctx)
> +{
> +	struct spu *spu = ctx->spu;
> +	if (!spu)
> +		return;
> +	spu_switch_notify(spu, ctx);
> +}
> +
>  /* The three externally callable interfaces
>   * for the scheduler begin here.
>   *
> Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/spufs.h
> ===================================================================
> --- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/spufs.h	2007-01-08 18:18:40.093354608 -0600
> +++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/spufs.h	2007-01-08 18:31:03.610345792 -0600
> @@ -183,6 +183,7 @@
>  void spu_yield(struct spu_context *ctx);
>  int __init spu_sched_init(void);
>  void __exit spu_sched_exit(void);
> +void spu_notify_already_active(struct spu_context *ctx);
>  
>  extern char *isolated_loader;
>  
> Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/run.c
> ===================================================================
> --- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/run.c	2007-01-08 18:33:51.979311680 -0600
> +++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/run.c	2007-01-11 10:17:20.777344984 -0600
> @@ -10,6 +10,8 @@
>  
>  #include "spufs.h"
>  
> +extern int notify_active[MAX_NUMNODES];
> +
>  /* interrupt-level stop callback function. */
>  void spufs_stop_callback(struct spu *spu)
>  {
> @@ -45,7 +47,9 @@
>  	u64 pte_fault;
>  
>  	*stat = ctx->ops->status_read(ctx);
> -	if (ctx->state != SPU_STATE_RUNNABLE)
> +	smp_mb();

And smp_rmb() should be sufficient here.

> +	if (ctx->state != SPU_STATE_RUNNABLE || notify_active[ctx->spu->number])
>  		return 1;
>  	spu = ctx->spu;
>  	pte_fault = spu->dsisr &
> @@ -319,6 +323,11 @@
>  		ret = spufs_wait(ctx->stop_wq, spu_stopped(ctx, &status));
>  		if (unlikely(ret))
>  			break;
> +		if (unlikely(notify_active[ctx->spu->number])) {
> +			notify_active[ctx->spu->number] = 0;
> +			if (!(status & SPU_STATUS_STOPPED_BY_STOP))
> +				spu_notify_already_active(ctx);
> +		}
>  		if ((status & SPU_STATUS_STOPPED_BY_STOP) &&
>  		    (status >> SPU_STOP_STATUS_SHIFT == 0x2104)) {
>  			ret = spu_process_callback(ctx);

cheers

-- 
Michael Ellerman
OzLabs, IBM Australia Development Lab

wwweb: http://michael.ellerman.id.au
phone: +61 2 6212 1183 (tie line 70 21183)

We do not inherit the earth from our ancestors,
we borrow it from our children. - S.M.A.R.T Person

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [PATCH] Cell SPU task notification

[Maynard Johnson]

Michael,
Thanks for your comments!  My responses are below.

-Maynard

Michael Ellerman wrote:

>>Subject: Enable SPU switch notification to detect currently active SPU tasks.
>>
>>From: Maynard Johnson <maynardj@us.ibm.com>
>>
>>This patch adds to the capability of spu_switch_event_register so that the
>>caller is also notified of currently active SPU tasks.  It also exports
>>spu_switch_event_register and spu_switch_event_unregister.
> 
> 
> Hi Maynard,
> 
> It'd be really good if you could convince your mailer to send patches inline :)
Mozilla Mail is my client, and I don't see any option to force patches 
inline.  Of course, there is an option to display received attachments 
as inline, but that doesn't help you.
> 
> 
>>Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c
>>===================================================================
>>--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/sched.c	2006-12-04 10:56:04.730698720 -0600
>>+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c	2007-01-11 09:45:37.918333128 -0600
>>@@ -46,6 +46,8 @@
>> 
>> #define SPU_MIN_TIMESLICE 	(100 * HZ / 1000)
>> 
>>+int notify_active[MAX_NUMNODES];
DOH!  Right, this is patently wrong.  Somehow, I misunderstood what 
MAX_NUMNODES was.  I'll fix it.
> 
> 
> You're basing the size of the array on MAX_NUMNODES
> (1 << CONFIG_NODES_SHIFT), but then indexing it by spu->number.
> 
> It's quite possible we'll have a system with MAX_NUMNODES == 1, but > 1
> spus, in which case this code is going to break. The PS3 is one such
> system.
> 
> Instead I think you should have a flag in the spu struct.
Yes, that's certainly an option I'll look at.
> 
> 
>> #define SPU_BITMAP_SIZE (((MAX_PRIO+BITS_PER_LONG)/BITS_PER_LONG)+1)
>> struct spu_prio_array {
>> 	unsigned long bitmap[SPU_BITMAP_SIZE];
>>@@ -81,18 +83,45 @@
>> static void spu_switch_notify(struct spu *spu, struct spu_context *ctx)
>> {
>> 	blocking_notifier_call_chain(&spu_switch_notifier,
>>-			    ctx ? ctx->object_id : 0, spu);
>>+				     ctx ? ctx->object_id : 0, spu);
>>+}
> 
> 
> Try not to make whitespace only changes in the same patch as actual code changes.
> 
> 
>>+
>>+static void notify_spus_active(void)
>>+{
>>+	int node;
>>+	/* Wake up the active spu_contexts. When the awakened processes 
>>+	 * sees their notify_active flag is set, they will call
>>+	 * spu_notify_already_active().
>>+	 */
>>+	for (node = 0; node < MAX_NUMNODES; node++) {
>>+		struct spu *spu;
>>+		mutex_lock(&spu_prio->active_mutex[node]);
>>+                list_for_each_entry(spu, &spu_prio->active_list[node], list) {
>>+			struct spu_context *ctx = spu->ctx;
>>+			wake_up_all(&ctx->stop_wq);
>>+			notify_active[ctx->spu->number] = 1;
>>+			smp_mb();
>>+		}
> 
> 
> I don't understand why you're setting the notify flag after you do the
> wake_up_all() ?
Right, I'll move it.
> 
> You only need a smp_wmb() here.
OK.
> 
> Does the scheduler guarantee that ctxs won't swap nodes? Otherwise
Don't know.  Arnd would probably know off the top of his head.
> between releasing the lock on one node and getting the lock on the next,
> a ctx could migrate between them - which would cause either spurious
> wake ups, or missing a ctx altogether. Although I'm not sure if it's
> that important.
How important is it?  If the SPUs were executing different code or 
taking different paths through the same code, then seeing the samples 
for each SPU is important.  But if this were the case, the user would 
easily see the misssing data for the missing SPU(s).  Since this sounds 
like a very small window, re-running the profile would _probably_ yield 
a complete profile.  On the other hand, if we can avoid it, we should. 
When I repost the fixed-up patch, I'll add a comment/question for Arnd 
about this issue.
> 
> 
>>+                mutex_unlock(&spu_prio->active_mutex[node]);
>>+	}
>>+	yield();
>> }
>> 
>> int spu_switch_event_register(struct notifier_block * n)
>> {
>>-	return blocking_notifier_chain_register(&spu_switch_notifier, n);
>>+	int ret;
>>+	ret = blocking_notifier_chain_register(&spu_switch_notifier, n);
>>+	if (!ret)
>>+		notify_spus_active();
>>+	return ret;
>> }
>>+EXPORT_SYMBOL_GPL(spu_switch_event_register);
>> 
>> int spu_switch_event_unregister(struct notifier_block * n)
>> {
>> 	return blocking_notifier_chain_unregister(&spu_switch_notifier, n);
>> }
>>+EXPORT_SYMBOL_GPL(spu_switch_event_unregister);
>> 
>> 
>> static inline void bind_context(struct spu *spu, struct spu_context *ctx)
>>@@ -250,6 +279,14 @@
>> 	return spu_get_idle(ctx, flags);
>> }
>> 
>>+void spu_notify_already_active(struct spu_context *ctx)
>>+{
>>+	struct spu *spu = ctx->spu;
>>+	if (!spu)
>>+		return;
>>+	spu_switch_notify(spu, ctx);
>>+}
>>+
>> /* The three externally callable interfaces
>>  * for the scheduler begin here.
>>  *
>>Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/spufs.h
>>===================================================================
>>--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/spufs.h	2007-01-08 18:18:40.093354608 -0600
>>+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/spufs.h	2007-01-08 18:31:03.610345792 -0600
>>@@ -183,6 +183,7 @@
>> void spu_yield(struct spu_context *ctx);
>> int __init spu_sched_init(void);
>> void __exit spu_sched_exit(void);
>>+void spu_notify_already_active(struct spu_context *ctx);
>> 
>> extern char *isolated_loader;
>> 
>>Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/run.c
>>===================================================================
>>--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/run.c	2007-01-08 18:33:51.979311680 -0600
>>+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/run.c	2007-01-11 10:17:20.777344984 -0600
>>@@ -10,6 +10,8 @@
>> 
>> #include "spufs.h"
>> 
>>+extern int notify_active[MAX_NUMNODES];
>>+
>> /* interrupt-level stop callback function. */
>> void spufs_stop_callback(struct spu *spu)
>> {
>>@@ -45,7 +47,9 @@
>> 	u64 pte_fault;
>> 
>> 	*stat = ctx->ops->status_read(ctx);
>>-	if (ctx->state != SPU_STATE_RUNNABLE)
>>+	smp_mb();
> 
> 
> And smp_rmb() should be sufficient here.
OK
> 
> 
>>+	if (ctx->state != SPU_STATE_RUNNABLE || notify_active[ctx->spu->number])
>> 		return 1;
>> 	spu = ctx->spu;
>> 	pte_fault = spu->dsisr &
>>@@ -319,6 +323,11 @@
>> 		ret = spufs_wait(ctx->stop_wq, spu_stopped(ctx, &status));
>> 		if (unlikely(ret))
>> 			break;
>>+		if (unlikely(notify_active[ctx->spu->number])) {
>>+			notify_active[ctx->spu->number] = 0;
>>+			if (!(status & SPU_STATUS_STOPPED_BY_STOP))
>>+				spu_notify_already_active(ctx);
>>+		}
>> 		if ((status & SPU_STATUS_STOPPED_BY_STOP) &&
>> 		    (status >> SPU_STOP_STATUS_SHIFT == 0x2104)) {
>> 			ret = spu_process_callback(ctx);
> 
> 
> cheers
> 

Re: [PATCH] Cell SPU task notification

[Maynard Johnson]

Michael,
Thanks for your comments!  My responses are below.

-Maynard

Michael Ellerman wrote:

>>Subject: Enable SPU switch notification to detect currently active SPU tasks.
>>
>>From: Maynard Johnson <maynardj@us.ibm.com>
>>
>>This patch adds to the capability of spu_switch_event_register so that the
>>caller is also notified of currently active SPU tasks.  It also exports
>>spu_switch_event_register and spu_switch_event_unregister.
> 
> 
> Hi Maynard,
> 
> It'd be really good if you could convince your mailer to send patches inline :)
Mozilla Mail is my client, and I don't see any option to force patches 
inline.  Of course, there is an option to display received attachments 
as inline, but that doesn't help you.
> 
> 
>>Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c
>>===================================================================
>>--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/sched.c	2006-12-04 10:56:04.730698720 -0600
>>+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c	2007-01-11 09:45:37.918333128 -0600
>>@@ -46,6 +46,8 @@
>> 
>> #define SPU_MIN_TIMESLICE 	(100 * HZ / 1000)
>> 
>>+int notify_active[MAX_NUMNODES];
DOH!  Right, this is patently wrong.  Somehow, I misunderstood what 
MAX_NUMNODES was.  I'll fix it.
> 
> 
> You're basing the size of the array on MAX_NUMNODES
> (1 << CONFIG_NODES_SHIFT), but then indexing it by spu->number.
> 
> It's quite possible we'll have a system with MAX_NUMNODES == 1, but > 1
> spus, in which case this code is going to break. The PS3 is one such
> system.
> 
> Instead I think you should have a flag in the spu struct.
Yes, that's certainly an option I'll look at.
> 
> 
>> #define SPU_BITMAP_SIZE (((MAX_PRIO+BITS_PER_LONG)/BITS_PER_LONG)+1)
>> struct spu_prio_array {
>> 	unsigned long bitmap[SPU_BITMAP_SIZE];
>>@@ -81,18 +83,45 @@
>> static void spu_switch_notify(struct spu *spu, struct spu_context *ctx)
>> {
>> 	blocking_notifier_call_chain(&spu_switch_notifier,
>>-			    ctx ? ctx->object_id : 0, spu);
>>+				     ctx ? ctx->object_id : 0, spu);
>>+}
> 
> 
> Try not to make whitespace only changes in the same patch as actual code changes.
> 
> 
>>+
>>+static void notify_spus_active(void)
>>+{
>>+	int node;
>>+	/* Wake up the active spu_contexts. When the awakened processes 
>>+	 * sees their notify_active flag is set, they will call
>>+	 * spu_notify_already_active().
>>+	 */
>>+	for (node = 0; node < MAX_NUMNODES; node++) {
>>+		struct spu *spu;
>>+		mutex_lock(&spu_prio->active_mutex[node]);
>>+                list_for_each_entry(spu, &spu_prio->active_list[node], list) {
>>+			struct spu_context *ctx = spu->ctx;
>>+			wake_up_all(&ctx->stop_wq);
>>+			notify_active[ctx->spu->number] = 1;
>>+			smp_mb();
>>+		}
> 
> 
> I don't understand why you're setting the notify flag after you do the
> wake_up_all() ?
Right, I'll move it.
> 
> You only need a smp_wmb() here.
OK.
> 
> Does the scheduler guarantee that ctxs won't swap nodes? Otherwise
Don't know.  Arnd would probably know off the top of his head.
> between releasing the lock on one node and getting the lock on the next,
> a ctx could migrate between them - which would cause either spurious
> wake ups, or missing a ctx altogether. Although I'm not sure if it's
> that important.
How important is it?  If the SPUs were executing different code or 
taking different paths through the same code, then seeing the samples 
for each SPU is important.  But if this were the case, the user would 
easily see the misssing data for the missing SPU(s).  Since this sounds 
like a very small window, re-running the profile would _probably_ yield 
a complete profile.  On the other hand, if we can avoid it, we should. 
When I repost the fixed-up patch, I'll add a comment/question for Arnd 
about this issue.
> 
> 
>>+                mutex_unlock(&spu_prio->active_mutex[node]);
>>+	}
>>+	yield();
>> }
>> 
>> int spu_switch_event_register(struct notifier_block * n)
>> {
>>-	return blocking_notifier_chain_register(&spu_switch_notifier, n);
>>+	int ret;
>>+	ret = blocking_notifier_chain_register(&spu_switch_notifier, n);
>>+	if (!ret)
>>+		notify_spus_active();
>>+	return ret;
>> }
>>+EXPORT_SYMBOL_GPL(spu_switch_event_register);
>> 
>> int spu_switch_event_unregister(struct notifier_block * n)
>> {
>> 	return blocking_notifier_chain_unregister(&spu_switch_notifier, n);
>> }
>>+EXPORT_SYMBOL_GPL(spu_switch_event_unregister);
>> 
>> 
>> static inline void bind_context(struct spu *spu, struct spu_context *ctx)
>>@@ -250,6 +279,14 @@
>> 	return spu_get_idle(ctx, flags);
>> }
>> 
>>+void spu_notify_already_active(struct spu_context *ctx)
>>+{
>>+	struct spu *spu = ctx->spu;
>>+	if (!spu)
>>+		return;
>>+	spu_switch_notify(spu, ctx);
>>+}
>>+
>> /* The three externally callable interfaces
>>  * for the scheduler begin here.
>>  *
>>Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/spufs.h
>>===================================================================
>>--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/spufs.h	2007-01-08 18:18:40.093354608 -0600
>>+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/spufs.h	2007-01-08 18:31:03.610345792 -0600
>>@@ -183,6 +183,7 @@
>> void spu_yield(struct spu_context *ctx);
>> int __init spu_sched_init(void);
>> void __exit spu_sched_exit(void);
>>+void spu_notify_already_active(struct spu_context *ctx);
>> 
>> extern char *isolated_loader;
>> 
>>Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/run.c
>>===================================================================
>>--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/run.c	2007-01-08 18:33:51.979311680 -0600
>>+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/run.c	2007-01-11 10:17:20.777344984 -0600
>>@@ -10,6 +10,8 @@
>> 
>> #include "spufs.h"
>> 
>>+extern int notify_active[MAX_NUMNODES];
>>+
>> /* interrupt-level stop callback function. */
>> void spufs_stop_callback(struct spu *spu)
>> {
>>@@ -45,7 +47,9 @@
>> 	u64 pte_fault;
>> 
>> 	*stat = ctx->ops->status_read(ctx);
>>-	if (ctx->state != SPU_STATE_RUNNABLE)
>>+	smp_mb();
> 
> 
> And smp_rmb() should be sufficient here.
OK
> 
> 
>>+	if (ctx->state != SPU_STATE_RUNNABLE || notify_active[ctx->spu->number])
>> 		return 1;
>> 	spu = ctx->spu;
>> 	pte_fault = spu->dsisr &
>>@@ -319,6 +323,11 @@
>> 		ret = spufs_wait(ctx->stop_wq, spu_stopped(ctx, &status));
>> 		if (unlikely(ret))
>> 			break;
>>+		if (unlikely(notify_active[ctx->spu->number])) {
>>+			notify_active[ctx->spu->number] = 0;
>>+			if (!(status & SPU_STATUS_STOPPED_BY_STOP))
>>+				spu_notify_already_active(ctx);
>>+		}
>> 		if ((status & SPU_STATUS_STOPPED_BY_STOP) &&
>> 		    (status >> SPU_STOP_STATUS_SHIFT == 0x2104)) {
>> 			ret = spu_process_callback(ctx);
> 
> 
> cheers
> 

Re: [PATCH] Cell SPU task notification -- updated patch: #1

[Maynard Johnson]

[-- Attachment #1: Type: text/plain, Size: 565 bytes --]

Attached is an updated patch that addresses Michael Ellerman's comments.

One comment made by Michael has not yet been addressed:
The comment was in regard to the for-loop in 
spufs/sched.c:notify_spus_active().  He wondered if the scheduler can 
swap a context from one node to another.  If so, there's a small window 
in this loop (where we switch the lock from one node's active list to 
the next) where it may be possible we might miss waking up a context and 
send a spurious wakeup to another.
	Arnd . . . can you comment on this question?

Thanks.
-Maynard


[-- Attachment #2: spu-notifier.patch --]
[-- Type: text/x-diff, Size: 4285 bytes --]

Subject: Enable SPU switch notification to detect currently active SPU tasks.

From: Maynard Johnson <maynardj@us.ibm.com>

This patch adds to the capability of spu_switch_event_register so that the
caller is also notified of currently active SPU tasks.  It also exports
spu_switch_event_register and spu_switch_event_unregister.

Signed-off-by: Maynard Johnson <mpjohn@us.ibm.com>


Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c
===================================================================
--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/sched.c	2006-12-04 10:56:04.730698720 -0600
+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c	2007-01-15 16:22:31.808461448 -0600
@@ -84,15 +84,42 @@
 			    ctx ? ctx->object_id : 0, spu);
 }
 
+static void notify_spus_active(void)
+{
+       int node;
+	/* Wake up the active spu_contexts. When the awakened processes 
+	 * sees their notify_active flag is set, they will call
+	 * spu_notify_already_active().
+	 */
+	for (node = 0; node < MAX_NUMNODES; node++) {
+		struct spu *spu;
+		mutex_lock(&spu_prio->active_mutex[node]);
+                list_for_each_entry(spu, &spu_prio->active_list[node], list) {
+			struct spu_context *ctx = spu->ctx;
+			spu->notify_active = 1;
+			wake_up_all(&ctx->stop_wq);
+			smp_wmb();
+		}
+                mutex_unlock(&spu_prio->active_mutex[node]);
+	}
+	yield();
+}
+
 int spu_switch_event_register(struct notifier_block * n)
 {
-	return blocking_notifier_chain_register(&spu_switch_notifier, n);
+	int ret;
+	ret = blocking_notifier_chain_register(&spu_switch_notifier, n);
+	if (!ret)
+		notify_spus_active();
+	return ret;
 }
+EXPORT_SYMBOL_GPL(spu_switch_event_register);
 
 int spu_switch_event_unregister(struct notifier_block * n)
 {
 	return blocking_notifier_chain_unregister(&spu_switch_notifier, n);
 }
+EXPORT_SYMBOL_GPL(spu_switch_event_unregister);
 
 
 static inline void bind_context(struct spu *spu, struct spu_context *ctx)
@@ -250,6 +277,14 @@
 	return spu_get_idle(ctx, flags);
 }
 
+void spu_notify_already_active(struct spu_context *ctx)
+{
+	struct spu *spu = ctx->spu;
+	if (!spu)
+		return;
+	spu_switch_notify(spu, ctx);
+}
+
 /* The three externally callable interfaces
  * for the scheduler begin here.
  *
Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/spufs.h
===================================================================
--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/spufs.h	2007-01-08 18:18:40.093354608 -0600
+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/spufs.h	2007-01-08 18:31:03.610345792 -0600
@@ -183,6 +183,7 @@
 void spu_yield(struct spu_context *ctx);
 int __init spu_sched_init(void);
 void __exit spu_sched_exit(void);
+void spu_notify_already_active(struct spu_context *ctx);
 
 extern char *isolated_loader;
 
Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/run.c
===================================================================
--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/run.c	2007-01-08 18:33:51.979311680 -0600
+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/run.c	2007-01-15 16:31:30.104499992 -0600
@@ -45,9 +45,11 @@
 	u64 pte_fault;
 
 	*stat = ctx->ops->status_read(ctx);
-	if (ctx->state != SPU_STATE_RUNNABLE)
-		return 1;
+	smp_rmb();
+
 	spu = ctx->spu;
+	if (ctx->state != SPU_STATE_RUNNABLE || spu->notify_active)
+		return 1;
 	pte_fault = spu->dsisr &
 	    (MFC_DSISR_PTE_NOT_FOUND | MFC_DSISR_ACCESS_DENIED);
 	return (!(*stat & 0x1) || pte_fault || spu->class_0_pending) ? 1 : 0;
@@ -304,6 +306,7 @@
 		   u32 *npc, u32 *event)
 {
 	int ret;
+	struct * spu;
 	u32 status;
 
 	if (down_interruptible(&ctx->run_sema))
@@ -317,8 +320,16 @@
 
 	do {
 		ret = spufs_wait(ctx->stop_wq, spu_stopped(ctx, &status));
+		spu = ctx->spu;
 		if (unlikely(ret))
 			break;
+		if (unlikely(spu->notify_active)) {
+			spu->notify_active = 0;
+			if (!(status & SPU_STATUS_STOPPED_BY_STOP)) {
+				spu_notify_already_active(ctx);
+				continue;
+			}
+		}
 		if ((status & SPU_STATUS_STOPPED_BY_STOP) &&
 		    (status >> SPU_STOP_STATUS_SHIFT == 0x2104)) {
 			ret = spu_process_callback(ctx);

Re: [PATCH] Cell SPU task notification -- updated patch: #1

[Maynard Johnson]

[-- Attachment #1: Type: text/plain, Size: 565 bytes --]

Attached is an updated patch that addresses Michael Ellerman's comments.

One comment made by Michael has not yet been addressed:
The comment was in regard to the for-loop in 
spufs/sched.c:notify_spus_active().  He wondered if the scheduler can 
swap a context from one node to another.  If so, there's a small window 
in this loop (where we switch the lock from one node's active list to 
the next) where it may be possible we might miss waking up a context and 
send a spurious wakeup to another.
	Arnd . . . can you comment on this question?

Thanks.
-Maynard


[-- Attachment #2: spu-notifier.patch --]
[-- Type: text/x-diff, Size: 4285 bytes --]

Subject: Enable SPU switch notification to detect currently active SPU tasks.

From: Maynard Johnson <maynardj@us.ibm.com>

This patch adds to the capability of spu_switch_event_register so that the
caller is also notified of currently active SPU tasks.  It also exports
spu_switch_event_register and spu_switch_event_unregister.

Signed-off-by: Maynard Johnson <mpjohn@us.ibm.com>


Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c
===================================================================
--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/sched.c	2006-12-04 10:56:04.730698720 -0600
+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c	2007-01-15 16:22:31.808461448 -0600
@@ -84,15 +84,42 @@
 			    ctx ? ctx->object_id : 0, spu);
 }
 
+static void notify_spus_active(void)
+{
+       int node;
+	/* Wake up the active spu_contexts. When the awakened processes 
+	 * sees their notify_active flag is set, they will call
+	 * spu_notify_already_active().
+	 */
+	for (node = 0; node < MAX_NUMNODES; node++) {
+		struct spu *spu;
+		mutex_lock(&spu_prio->active_mutex[node]);
+                list_for_each_entry(spu, &spu_prio->active_list[node], list) {
+			struct spu_context *ctx = spu->ctx;
+			spu->notify_active = 1;
+			wake_up_all(&ctx->stop_wq);
+			smp_wmb();
+		}
+                mutex_unlock(&spu_prio->active_mutex[node]);
+	}
+	yield();
+}
+
 int spu_switch_event_register(struct notifier_block * n)
 {
-	return blocking_notifier_chain_register(&spu_switch_notifier, n);
+	int ret;
+	ret = blocking_notifier_chain_register(&spu_switch_notifier, n);
+	if (!ret)
+		notify_spus_active();
+	return ret;
 }
+EXPORT_SYMBOL_GPL(spu_switch_event_register);
 
 int spu_switch_event_unregister(struct notifier_block * n)
 {
 	return blocking_notifier_chain_unregister(&spu_switch_notifier, n);
 }
+EXPORT_SYMBOL_GPL(spu_switch_event_unregister);
 
 
 static inline void bind_context(struct spu *spu, struct spu_context *ctx)
@@ -250,6 +277,14 @@
 	return spu_get_idle(ctx, flags);
 }
 
+void spu_notify_already_active(struct spu_context *ctx)
+{
+	struct spu *spu = ctx->spu;
+	if (!spu)
+		return;
+	spu_switch_notify(spu, ctx);
+}
+
 /* The three externally callable interfaces
  * for the scheduler begin here.
  *
Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/spufs.h
===================================================================
--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/spufs.h	2007-01-08 18:18:40.093354608 -0600
+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/spufs.h	2007-01-08 18:31:03.610345792 -0600
@@ -183,6 +183,7 @@
 void spu_yield(struct spu_context *ctx);
 int __init spu_sched_init(void);
 void __exit spu_sched_exit(void);
+void spu_notify_already_active(struct spu_context *ctx);
 
 extern char *isolated_loader;
 
Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/run.c
===================================================================
--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/run.c	2007-01-08 18:33:51.979311680 -0600
+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/run.c	2007-01-15 16:31:30.104499992 -0600
@@ -45,9 +45,11 @@
 	u64 pte_fault;
 
 	*stat = ctx->ops->status_read(ctx);
-	if (ctx->state != SPU_STATE_RUNNABLE)
-		return 1;
+	smp_rmb();
+
 	spu = ctx->spu;
+	if (ctx->state != SPU_STATE_RUNNABLE || spu->notify_active)
+		return 1;
 	pte_fault = spu->dsisr &
 	    (MFC_DSISR_PTE_NOT_FOUND | MFC_DSISR_ACCESS_DENIED);
 	return (!(*stat & 0x1) || pte_fault || spu->class_0_pending) ? 1 : 0;
@@ -304,6 +306,7 @@
 		   u32 *npc, u32 *event)
 {
 	int ret;
+	struct * spu;
 	u32 status;
 
 	if (down_interruptible(&ctx->run_sema))
@@ -317,8 +320,16 @@
 
 	do {
 		ret = spufs_wait(ctx->stop_wq, spu_stopped(ctx, &status));
+		spu = ctx->spu;
 		if (unlikely(ret))
 			break;
+		if (unlikely(spu->notify_active)) {
+			spu->notify_active = 0;
+			if (!(status & SPU_STATUS_STOPPED_BY_STOP)) {
+				spu_notify_already_active(ctx);
+				continue;
+			}
+		}
 		if ((status & SPU_STATUS_STOPPED_BY_STOP) &&
 		    (status >> SPU_STOP_STATUS_SHIFT == 0x2104)) {
 			ret = spu_process_callback(ctx);

Re: [Cbe-oss-dev] [PATCH] Cell SPU task notification

[Christoph Hellwig]

Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c
===================================================================
--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/sched.c	2006-12-04 10:56:04.730698720 -0600
+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c	2007-01-15 16:22:31.808461448 -0600
@@ -84,15 +84,42 @@
 			    ctx ? ctx->object_id : 0, spu);
 }
 
+static void notify_spus_active(void)
+{
+       int node;
+	/* Wake up the active spu_contexts. When the awakened processes 
+	 * sees their notify_active flag is set, they will call
+	 * spu_notify_already_active().
+	 */
+	for (node = 0; node < MAX_NUMNODES; node++) {
+		struct spu *spu;
+		mutex_lock(&spu_prio->active_mutex[node]);
+                list_for_each_entry(spu, &spu_prio->active_list[node], list) {

	You seem to have some issues with tabs vs spaces for indentation
	here.

+			struct spu_context *ctx = spu->ctx;
+			spu->notify_active = 1;


	Please make this a bit in the sched_flags field that's added in
	the scheduler patch series I sent out.

+			wake_up_all(&ctx->stop_wq);
+			smp_wmb();
+		}
+                mutex_unlock(&spu_prio->active_mutex[node]);
+	}
+	yield();
+}

	Why do you add the yield() here?  yield is pretty much a sign
	for a bug

+void spu_notify_already_active(struct spu_context *ctx)
+{
+	struct spu *spu = ctx->spu;
+	if (!spu)
+		return;
+	spu_switch_notify(spu, ctx);
+}

	Please just call spu_switch_notify directly from the only
	caller.  Also the check for ctx->spu beeing there is not
	required if you look a the caller.


 	*stat = ctx->ops->status_read(ctx);
-	if (ctx->state != SPU_STATE_RUNNABLE)
-		return 1;
+	smp_rmb();


	What do you need the barrier for here?

Re: [Cbe-oss-dev] [PATCH] Cell SPU task notification

[Christoph Hellwig]

Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c
===================================================================
--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/sched.c	2006-12-04 10:56:04.730698720 -0600
+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c	2007-01-15 16:22:31.808461448 -0600
@@ -84,15 +84,42 @@
 			    ctx ? ctx->object_id : 0, spu);
 }
 
+static void notify_spus_active(void)
+{
+       int node;
+	/* Wake up the active spu_contexts. When the awakened processes 
+	 * sees their notify_active flag is set, they will call
+	 * spu_notify_already_active().
+	 */
+	for (node = 0; node < MAX_NUMNODES; node++) {
+		struct spu *spu;
+		mutex_lock(&spu_prio->active_mutex[node]);
+                list_for_each_entry(spu, &spu_prio->active_list[node], list) {

	You seem to have some issues with tabs vs spaces for indentation
	here.

+			struct spu_context *ctx = spu->ctx;
+			spu->notify_active = 1;


	Please make this a bit in the sched_flags field that's added in
	the scheduler patch series I sent out.

+			wake_up_all(&ctx->stop_wq);
+			smp_wmb();
+		}
+                mutex_unlock(&spu_prio->active_mutex[node]);
+	}
+	yield();
+}

	Why do you add the yield() here?  yield is pretty much a sign
	for a bug

+void spu_notify_already_active(struct spu_context *ctx)
+{
+	struct spu *spu = ctx->spu;
+	if (!spu)
+		return;
+	spu_switch_notify(spu, ctx);
+}

	Please just call spu_switch_notify directly from the only
	caller.  Also the check for ctx->spu beeing there is not
	required if you look a the caller.


 	*stat = ctx->ops->status_read(ctx);
-	if (ctx->state != SPU_STATE_RUNNABLE)
-		return 1;
+	smp_rmb();


	What do you need the barrier for here?

Re: [Cbe-oss-dev] [PATCH] Cell SPU task notification

[Maynard Johnson]

Christoph Hellwig wrote:

>Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c
>===================================================================
>--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/sched.c	2006-12-04 10:56:04.730698720 -0600
>+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c	2007-01-15 16:22:31.808461448 -0600
>@@ -84,15 +84,42 @@
> 			    ctx ? ctx->object_id : 0, spu);
> }
>
>+static void notify_spus_active(void)
>+{
>+       int node;
>+	/* Wake up the active spu_contexts. When the awakened processes
>+	 * sees their notify_active flag is set, they will call
>+	 * spu_notify_already_active().
>+	 */
>+	for (node = 0; node < MAX_NUMNODES; node++) {
>+		struct spu *spu;
>+		mutex_lock(&spu_prio->active_mutex[node]);
>+                list_for_each_entry(spu, &spu_prio->active_list[node], list) {
>
>	You seem to have some issues with tabs vs spaces for indentation
>	here.
>  
>
fixed

>+			struct spu_context *ctx = spu->ctx;
>+			spu->notify_active = 1;
>
>
>	Please make this a bit in the sched_flags field that's added in
>	the scheduler patch series I sent out.
>  
>
I haven't seen that the scheduler patch series got applied yet.  This 
Cell spu task notification patch is a pre-req for OProfile development 
to support profiling SPUs.   When the scheduler patch gets applied to a 
kernel version that fits our needs for our OProfile development, I don't 
see any problem in using the sched_flags field instead of notify_active.

>+			wake_up_all(&ctx->stop_wq);
>+			smp_wmb();
>+		}
>+                mutex_unlock(&spu_prio->active_mutex[node]);
>+	}
>+	yield();
>+}
>
>	Why do you add the yield() here?  yield is pretty much a sign
>	for a bug
>  
>
Yes, the yield() and the memory barriers were leftovers from an earlier 
ill-conceived attempt at solving this problem.  They should have been 
removed.  They're gone now.

>+void spu_notify_already_active(struct spu_context *ctx)
>+{
>+	struct spu *spu = ctx->spu;
>+	if (!spu)
>+		return;
>+	spu_switch_notify(spu, ctx);
>+}
>
>	Please just call spu_switch_notify directly from the only
>  
>
I hesitated doing this since it would entail changing spu_switch_notify 
from being static to non-static.  I'd like to get Arnd's opinion on this 
question before going ahead and making such a change.

>	caller.  Also the check for ctx->spu beeing there is not
>	required if you look a the caller.
>
>
> 	*stat = ctx->ops->status_read(ctx);
>-	if (ctx->state != SPU_STATE_RUNNABLE)
>-		return 1;
>+	smp_rmb();
>  
>
>
>	What do you need the barrier for here?
>  
>
Removed.

Re: [Cbe-oss-dev] [PATCH] Cell SPU task notification

[Maynard Johnson]

Christoph Hellwig wrote:

>Index: linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c
>===================================================================
>--- linux-2.6.19-rc6-arnd1+patches.orig/arch/powerpc/platforms/cell/spufs/sched.c	2006-12-04 10:56:04.730698720 -0600
>+++ linux-2.6.19-rc6-arnd1+patches/arch/powerpc/platforms/cell/spufs/sched.c	2007-01-15 16:22:31.808461448 -0600
>@@ -84,15 +84,42 @@
> 			    ctx ? ctx->object_id : 0, spu);
> }
>
>+static void notify_spus_active(void)
>+{
>+       int node;
>+	/* Wake up the active spu_contexts. When the awakened processes
>+	 * sees their notify_active flag is set, they will call
>+	 * spu_notify_already_active().
>+	 */
>+	for (node = 0; node < MAX_NUMNODES; node++) {
>+		struct spu *spu;
>+		mutex_lock(&spu_prio->active_mutex[node]);
>+                list_for_each_entry(spu, &spu_prio->active_list[node], list) {
>
>	You seem to have some issues with tabs vs spaces for indentation
>	here.
>  
>
fixed

>+			struct spu_context *ctx = spu->ctx;
>+			spu->notify_active = 1;
>
>
>	Please make this a bit in the sched_flags field that's added in
>	the scheduler patch series I sent out.
>  
>
I haven't seen that the scheduler patch series got applied yet.  This 
Cell spu task notification patch is a pre-req for OProfile development 
to support profiling SPUs.   When the scheduler patch gets applied to a 
kernel version that fits our needs for our OProfile development, I don't 
see any problem in using the sched_flags field instead of notify_active.

>+			wake_up_all(&ctx->stop_wq);
>+			smp_wmb();
>+		}
>+                mutex_unlock(&spu_prio->active_mutex[node]);
>+	}
>+	yield();
>+}
>
>	Why do you add the yield() here?  yield is pretty much a sign
>	for a bug
>  
>
Yes, the yield() and the memory barriers were leftovers from an earlier 
ill-conceived attempt at solving this problem.  They should have been 
removed.  They're gone now.

>+void spu_notify_already_active(struct spu_context *ctx)
>+{
>+	struct spu *spu = ctx->spu;
>+	if (!spu)
>+		return;
>+	spu_switch_notify(spu, ctx);
>+}
>
>	Please just call spu_switch_notify directly from the only
>  
>
I hesitated doing this since it would entail changing spu_switch_notify 
from being static to non-static.  I'd like to get Arnd's opinion on this 
question before going ahead and making such a change.

>	caller.  Also the check for ctx->spu beeing there is not
>	required if you look a the caller.
>
>
> 	*stat = ctx->ops->status_read(ctx);
>-	if (ctx->state != SPU_STATE_RUNNABLE)
>-		return 1;
>+	smp_rmb();
>  
>
>
>	What do you need the barrier for here?
>  
>
Removed.

Re: [Cbe-oss-dev] [PATCH] Cell SPU task notification

[Christoph Hellwig]

On Wed, Jan 17, 2007 at 09:56:12AM -0600, Maynard Johnson wrote:
> I haven't seen that the scheduler patch series got applied yet.  This 
> Cell spu task notification patch is a pre-req for OProfile development 
> to support profiling SPUs.   When the scheduler patch gets applied to a 
> kernel version that fits our needs for our OProfile development, I don't 
> see any problem in using the sched_flags field instead of notify_active.

I'll hopefull commit these patches this weekend, I'm at a conference
currently so not really able to do a lot of work.  If you need to make
more progress until than just apply the hunk that introduces sched_flags
before doing your patch.

> Yes, the yield() and the memory barriers were leftovers from an earlier 
> ill-conceived attempt at solving this problem.  They should have been 
> removed.  They're gone now.

Ok.

> I hesitated doing this since it would entail changing spu_switch_notify 
> from being static to non-static.  I'd like to get Arnd's opinion on this 
> question before going ahead and making such a change.

There is no difference in impact between marking a function non-static
and adding a trivial wrapper around it, only that the latter creates
more bloat.  So I don't think there's a good argument against this.

Re: [Cbe-oss-dev] [PATCH] Cell SPU task notification

[Christoph Hellwig]

On Wed, Jan 17, 2007 at 09:56:12AM -0600, Maynard Johnson wrote:
> I haven't seen that the scheduler patch series got applied yet.  This 
> Cell spu task notification patch is a pre-req for OProfile development 
> to support profiling SPUs.   When the scheduler patch gets applied to a 
> kernel version that fits our needs for our OProfile development, I don't 
> see any problem in using the sched_flags field instead of notify_active.

I'll hopefull commit these patches this weekend, I'm at a conference
currently so not really able to do a lot of work.  If you need to make
more progress until than just apply the hunk that introduces sched_flags
before doing your patch.

> Yes, the yield() and the memory barriers were leftovers from an earlier 
> ill-conceived attempt at solving this problem.  They should have been 
> removed.  They're gone now.

Ok.

> I hesitated doing this since it would entail changing spu_switch_notify 
> from being static to non-static.  I'd like to get Arnd's opinion on this 
> question before going ahead and making such a change.

There is no difference in impact between marking a function non-static
and adding a trivial wrapper around it, only that the latter creates
more bloat.  So I don't think there's a good argument against this.

Re: [Cbe-oss-dev] [PATCH] Cell SPU task notification - repost of patch with updates

[Maynard Johnson]

[-- Attachment #1: Type: text/plain, Size: 469 bytes --]

This patch makes all the changes suggested by Christopher, with the 
exception of the suggestion to use the sched_flags field versus a new 
member to the spu_context struct to signal the need to "notify already 
active".  I don't have the sched_flags change in my 2.6.20-rc1 tree.  I 
can send another patch later if/when the sched_flags changes appears in 
the kernel version we end up picking for final oprofile-spu development.

Comments welcome.  Thanks.

-Maynard

[-- Attachment #2: spu-notifier.patch --]
[-- Type: text/x-diff, Size: 4589 bytes --]

Subject: Enable SPU switch notification to detect currently active SPU tasks.

From: Maynard Johnson <maynardj@us.ibm.com>

This patch adds to the capability of spu_switch_event_register so that the
caller is also notified of currently active SPU tasks.  It also exports
spu_switch_event_register and spu_switch_event_unregister.

Signed-off-by: Maynard Johnson <mpjohn@us.ibm.com>


Index: linux-2.6.20-rc1/arch/powerpc/platforms/cell/spufs/sched.c
===================================================================
--- linux-2.6.20-rc1.orig/arch/powerpc/platforms/cell/spufs/sched.c	2007-01-18 16:43:14.324526032 -0600
+++ linux-2.6.20-rc1/arch/powerpc/platforms/cell/spufs/sched.c	2007-01-26 16:16:35.219668640 -0600
@@ -78,21 +78,46 @@
 
 static BLOCKING_NOTIFIER_HEAD(spu_switch_notifier);
 
-static void spu_switch_notify(struct spu *spu, struct spu_context *ctx)
+void spu_switch_notify(struct spu *spu, struct spu_context *ctx)
 {
 	blocking_notifier_call_chain(&spu_switch_notifier,
 			    ctx ? ctx->object_id : 0, spu);
 }
 
+static void notify_spus_active(void)
+{
+       int node;
+	/* Wake up the active spu_contexts. When the awakened processes 
+	 * sees their notify_active flag is set, they will call
+	 * spu_switch_notify();
+	 */
+	for (node = 0; node < MAX_NUMNODES; node++) {
+		struct spu *spu;
+		mutex_lock(&spu_prio->active_mutex[node]);
+                list_for_each_entry(spu, &spu_prio->active_list[node], list) {
+			struct spu_context *ctx = spu->ctx;
+			spu->notify_active = 1;
+			wake_up_all(&ctx->stop_wq);
+		}
+                mutex_unlock(&spu_prio->active_mutex[node]);
+	}
+}
+
 int spu_switch_event_register(struct notifier_block * n)
 {
-	return blocking_notifier_chain_register(&spu_switch_notifier, n);
+	int ret;
+	ret = blocking_notifier_chain_register(&spu_switch_notifier, n);
+	if (!ret)
+		notify_spus_active();
+	return ret;
 }
+EXPORT_SYMBOL_GPL(spu_switch_event_register);
 
 int spu_switch_event_unregister(struct notifier_block * n)
 {
 	return blocking_notifier_chain_unregister(&spu_switch_notifier, n);
 }
+EXPORT_SYMBOL_GPL(spu_switch_event_unregister);
 
 
 static inline void bind_context(struct spu *spu, struct spu_context *ctx)
Index: linux-2.6.20-rc1/arch/powerpc/platforms/cell/spufs/spufs.h
===================================================================
--- linux-2.6.20-rc1.orig/arch/powerpc/platforms/cell/spufs/spufs.h	2007-01-18 16:43:14.340523600 -0600
+++ linux-2.6.20-rc1/arch/powerpc/platforms/cell/spufs/spufs.h	2007-01-26 16:26:49.733703448 -0600
@@ -180,6 +180,7 @@
 int spu_activate(struct spu_context *ctx, u64 flags);
 void spu_deactivate(struct spu_context *ctx);
 void spu_yield(struct spu_context *ctx);
+void spu_switch_notify(struct spu *spu, struct spu_context *ctx);
 int __init spu_sched_init(void);
 void __exit spu_sched_exit(void);
 
Index: linux-2.6.20-rc1/arch/powerpc/platforms/cell/spufs/run.c
===================================================================
--- linux-2.6.20-rc1.orig/arch/powerpc/platforms/cell/spufs/run.c	2007-01-18 16:43:14.340523600 -0600
+++ linux-2.6.20-rc1/arch/powerpc/platforms/cell/spufs/run.c	2007-01-26 16:24:38.979744856 -0600
@@ -45,9 +45,10 @@
 	u64 pte_fault;
 
 	*stat = ctx->ops->status_read(ctx);
-	if (ctx->state != SPU_STATE_RUNNABLE)
-		return 1;
+
 	spu = ctx->spu;
+	if (ctx->state != SPU_STATE_RUNNABLE || spu->notify_active)
+		return 1;
 	pte_fault = spu->dsisr &
 	    (MFC_DSISR_PTE_NOT_FOUND | MFC_DSISR_ACCESS_DENIED);
 	return (!(*stat & 0x1) || pte_fault || spu->class_0_pending) ? 1 : 0;
@@ -305,6 +306,7 @@
 		   u32 *npc, u32 *event)
 {
 	int ret;
+	struct spu * spu;
 	u32 status;
 
 	if (down_interruptible(&ctx->run_sema))
@@ -318,8 +320,16 @@
 
 	do {
 		ret = spufs_wait(ctx->stop_wq, spu_stopped(ctx, &status));
+		spu = ctx->spu;
 		if (unlikely(ret))
 			break;
+		if (unlikely(spu->notify_active)) {
+			spu->notify_active = 0;
+			if (!(status & SPU_STATUS_STOPPED_BY_STOP)) {
+				spu_switch_notify(spu, ctx);
+				continue;
+			}
+		}
 		if ((status & SPU_STATUS_STOPPED_BY_STOP) &&
 		    (status >> SPU_STOP_STATUS_SHIFT == 0x2104)) {
 			ret = spu_process_callback(ctx);
Index: linux-2.6.20-rc1/include/asm-powerpc/spu.h
===================================================================
--- linux-2.6.20-rc1.orig/include/asm-powerpc/spu.h	2007-01-18 16:43:19.932605072 -0600
+++ linux-2.6.20-rc1/include/asm-powerpc/spu.h	2007-01-24 12:17:30.209676992 -0600
@@ -144,6 +144,7 @@
 
 	void* pdata; /* platform private data */
 	struct sys_device sysdev;
+	int notify_active;
 };
 
 struct spu *spu_alloc(void);

Re: [Cbe-oss-dev] [PATCH] Cell SPU task notification - repost of patch with updates

[Maynard Johnson]

[-- Attachment #1: Type: text/plain, Size: 469 bytes --]

This patch makes all the changes suggested by Christopher, with the 
exception of the suggestion to use the sched_flags field versus a new 
member to the spu_context struct to signal the need to "notify already 
active".  I don't have the sched_flags change in my 2.6.20-rc1 tree.  I 
can send another patch later if/when the sched_flags changes appears in 
the kernel version we end up picking for final oprofile-spu development.

Comments welcome.  Thanks.

-Maynard

[-- Attachment #2: spu-notifier.patch --]
[-- Type: text/x-diff, Size: 4589 bytes --]

Subject: Enable SPU switch notification to detect currently active SPU tasks.

From: Maynard Johnson <maynardj@us.ibm.com>

This patch adds to the capability of spu_switch_event_register so that the
caller is also notified of currently active SPU tasks.  It also exports
spu_switch_event_register and spu_switch_event_unregister.

Signed-off-by: Maynard Johnson <mpjohn@us.ibm.com>


Index: linux-2.6.20-rc1/arch/powerpc/platforms/cell/spufs/sched.c
===================================================================
--- linux-2.6.20-rc1.orig/arch/powerpc/platforms/cell/spufs/sched.c	2007-01-18 16:43:14.324526032 -0600
+++ linux-2.6.20-rc1/arch/powerpc/platforms/cell/spufs/sched.c	2007-01-26 16:16:35.219668640 -0600
@@ -78,21 +78,46 @@
 
 static BLOCKING_NOTIFIER_HEAD(spu_switch_notifier);
 
-static void spu_switch_notify(struct spu *spu, struct spu_context *ctx)
+void spu_switch_notify(struct spu *spu, struct spu_context *ctx)
 {
 	blocking_notifier_call_chain(&spu_switch_notifier,
 			    ctx ? ctx->object_id : 0, spu);
 }
 
+static void notify_spus_active(void)
+{
+       int node;
+	/* Wake up the active spu_contexts. When the awakened processes 
+	 * sees their notify_active flag is set, they will call
+	 * spu_switch_notify();
+	 */
+	for (node = 0; node < MAX_NUMNODES; node++) {
+		struct spu *spu;
+		mutex_lock(&spu_prio->active_mutex[node]);
+                list_for_each_entry(spu, &spu_prio->active_list[node], list) {
+			struct spu_context *ctx = spu->ctx;
+			spu->notify_active = 1;
+			wake_up_all(&ctx->stop_wq);
+		}
+                mutex_unlock(&spu_prio->active_mutex[node]);
+	}
+}
+
 int spu_switch_event_register(struct notifier_block * n)
 {
-	return blocking_notifier_chain_register(&spu_switch_notifier, n);
+	int ret;
+	ret = blocking_notifier_chain_register(&spu_switch_notifier, n);
+	if (!ret)
+		notify_spus_active();
+	return ret;
 }
+EXPORT_SYMBOL_GPL(spu_switch_event_register);
 
 int spu_switch_event_unregister(struct notifier_block * n)
 {
 	return blocking_notifier_chain_unregister(&spu_switch_notifier, n);
 }
+EXPORT_SYMBOL_GPL(spu_switch_event_unregister);
 
 
 static inline void bind_context(struct spu *spu, struct spu_context *ctx)
Index: linux-2.6.20-rc1/arch/powerpc/platforms/cell/spufs/spufs.h
===================================================================
--- linux-2.6.20-rc1.orig/arch/powerpc/platforms/cell/spufs/spufs.h	2007-01-18 16:43:14.340523600 -0600
+++ linux-2.6.20-rc1/arch/powerpc/platforms/cell/spufs/spufs.h	2007-01-26 16:26:49.733703448 -0600
@@ -180,6 +180,7 @@
 int spu_activate(struct spu_context *ctx, u64 flags);
 void spu_deactivate(struct spu_context *ctx);
 void spu_yield(struct spu_context *ctx);
+void spu_switch_notify(struct spu *spu, struct spu_context *ctx);
 int __init spu_sched_init(void);
 void __exit spu_sched_exit(void);
 
Index: linux-2.6.20-rc1/arch/powerpc/platforms/cell/spufs/run.c
===================================================================
--- linux-2.6.20-rc1.orig/arch/powerpc/platforms/cell/spufs/run.c	2007-01-18 16:43:14.340523600 -0600
+++ linux-2.6.20-rc1/arch/powerpc/platforms/cell/spufs/run.c	2007-01-26 16:24:38.979744856 -0600
@@ -45,9 +45,10 @@
 	u64 pte_fault;
 
 	*stat = ctx->ops->status_read(ctx);
-	if (ctx->state != SPU_STATE_RUNNABLE)
-		return 1;
+
 	spu = ctx->spu;
+	if (ctx->state != SPU_STATE_RUNNABLE || spu->notify_active)
+		return 1;
 	pte_fault = spu->dsisr &
 	    (MFC_DSISR_PTE_NOT_FOUND | MFC_DSISR_ACCESS_DENIED);
 	return (!(*stat & 0x1) || pte_fault || spu->class_0_pending) ? 1 : 0;
@@ -305,6 +306,7 @@
 		   u32 *npc, u32 *event)
 {
 	int ret;
+	struct spu * spu;
 	u32 status;
 
 	if (down_interruptible(&ctx->run_sema))
@@ -318,8 +320,16 @@
 
 	do {
 		ret = spufs_wait(ctx->stop_wq, spu_stopped(ctx, &status));
+		spu = ctx->spu;
 		if (unlikely(ret))
 			break;
+		if (unlikely(spu->notify_active)) {
+			spu->notify_active = 0;
+			if (!(status & SPU_STATUS_STOPPED_BY_STOP)) {
+				spu_switch_notify(spu, ctx);
+				continue;
+			}
+		}
 		if ((status & SPU_STATUS_STOPPED_BY_STOP) &&
 		    (status >> SPU_STOP_STATUS_SHIFT == 0x2104)) {
 			ret = spu_process_callback(ctx);
Index: linux-2.6.20-rc1/include/asm-powerpc/spu.h
===================================================================
--- linux-2.6.20-rc1.orig/include/asm-powerpc/spu.h	2007-01-18 16:43:19.932605072 -0600
+++ linux-2.6.20-rc1/include/asm-powerpc/spu.h	2007-01-24 12:17:30.209676992 -0600
@@ -144,6 +144,7 @@
 
 	void* pdata; /* platform private data */
 	struct sys_device sysdev;
+	int notify_active;
 };
 
 struct spu *spu_alloc(void);

[PATCH] CELL SPU task notification

[Carl Love]

This is the updates to the patch previously posted by Maynard Johnson as
"PATCH 3/4 Add support to OProfile for profiling CELL".  All concerns
have been addressed.  The only open issue is the movement of the
notify_active flag to sched_flag in a future patch by Christoph
Hellwig's for SPU scheduling.

Subject: Enable SPU switch notification to detect currently active SPU
tasks.

From: Maynard Johnson <maynardj@us.ibm.com>, Carl Love
<carll@us.ibm.com>

This patch adds to the capability of spu_switch_event_register so that
the
caller is also notified of currently active SPU tasks.  It also exports
spu_switch_event_register and spu_switch_event_unregister.

Signed-off-by: Maynard Johnson <mpjohn@us.ibm.com>


Index: linux-2.6.20-rc1/arch/powerpc/platforms/cell/spufs/sched.c
===================================================================
--- linux-2.6.20-rc1.orig/arch/powerpc/platforms/cell/spufs/sched.c
2007-01-18 16:43:14.324526032 -0600
+++ linux-2.6.20-rc1/arch/powerpc/platforms/cell/spufs/sched.c
2007-01-30 09:35:03.237837184 -0600
@@ -78,21 +78,46 @@
 
 static BLOCKING_NOTIFIER_HEAD(spu_switch_notifier);
 
-static void spu_switch_notify(struct spu *spu, struct spu_context *ctx)
+void spu_switch_notify(struct spu *spu, struct spu_context *ctx)
 {
 	blocking_notifier_call_chain(&spu_switch_notifier,
 			    ctx ? ctx->object_id : 0, spu);
 }
 
+static void notify_spus_active(void)
+{
+	int node;
+	/* Wake up the active spu_contexts. When the awakened processes 
+	 * sees their notify_active flag is set, they will call
+	 * spu_switch_notify();
+	 */
+	for (node = 0; node < MAX_NUMNODES; node++) {
+		struct spu *spu;
+		mutex_lock(&spu_prio->active_mutex[node]);
+		list_for_each_entry(spu, &spu_prio->active_list[node], list) {
+			struct spu_context *ctx = spu->ctx;
+			spu->notify_active = 1;
+			wake_up_all(&ctx->stop_wq);
+		}
+		mutex_unlock(&spu_prio->active_mutex[node]);
+	}
+}
+
 int spu_switch_event_register(struct notifier_block * n)
 {
-	return blocking_notifier_chain_register(&spu_switch_notifier, n);
+	int ret;
+	ret = blocking_notifier_chain_register(&spu_switch_notifier, n);
+	if (!ret)
+		notify_spus_active();
+	return ret;
 }
+EXPORT_SYMBOL_GPL(spu_switch_event_register);
 
 int spu_switch_event_unregister(struct notifier_block * n)
 {
 	return blocking_notifier_chain_unregister(&spu_switch_notifier, n);
 }
+EXPORT_SYMBOL_GPL(spu_switch_event_unregister);
 
 
 static inline void bind_context(struct spu *spu, struct spu_context
*ctx)
Index: linux-2.6.20-rc1/arch/powerpc/platforms/cell/spufs/spufs.h
===================================================================
--- linux-2.6.20-rc1.orig/arch/powerpc/platforms/cell/spufs/spufs.h
2007-01-18 16:43:14.340523600 -0600
+++ linux-2.6.20-rc1/arch/powerpc/platforms/cell/spufs/spufs.h
2007-01-26 16:26:49.733703448 -0600
@@ -180,6 +180,7 @@
 int spu_activate(struct spu_context *ctx, u64 flags);
 void spu_deactivate(struct spu_context *ctx);
 void spu_yield(struct spu_context *ctx);
+void spu_switch_notify(struct spu *spu, struct spu_context *ctx);
 int __init spu_sched_init(void);
 void __exit spu_sched_exit(void);
 
Index: linux-2.6.20-rc1/arch/powerpc/platforms/cell/spufs/run.c
===================================================================
--- linux-2.6.20-rc1.orig/arch/powerpc/platforms/cell/spufs/run.c
2007-01-18 16:43:14.340523600 -0600
+++ linux-2.6.20-rc1/arch/powerpc/platforms/cell/spufs/run.c	2007-01-26
16:24:38.979744856 -0600
@@ -45,9 +45,10 @@
 	u64 pte_fault;
 
 	*stat = ctx->ops->status_read(ctx);
-	if (ctx->state != SPU_STATE_RUNNABLE)
-		return 1;
+
 	spu = ctx->spu;
+	if (ctx->state != SPU_STATE_RUNNABLE || spu->notify_active)
+		return 1;
 	pte_fault = spu->dsisr &
 	    (MFC_DSISR_PTE_NOT_FOUND | MFC_DSISR_ACCESS_DENIED);
 	return (!(*stat & 0x1) || pte_fault || spu->class_0_pending) ? 1 : 0;
@@ -305,6 +306,7 @@
 		   u32 *npc, u32 *event)
 {
 	int ret;
+	struct spu * spu;
 	u32 status;
 
 	if (down_interruptible(&ctx->run_sema))
@@ -318,8 +320,16 @@
 
 	do {
 		ret = spufs_wait(ctx->stop_wq, spu_stopped(ctx, &status));
+		spu = ctx->spu;
 		if (unlikely(ret))
 			break;
+		if (unlikely(spu->notify_active)) {
+			spu->notify_active = 0;
+			if (!(status & SPU_STATUS_STOPPED_BY_STOP)) {
+				spu_switch_notify(spu, ctx);
+				continue;
+			}
+		}
 		if ((status & SPU_STATUS_STOPPED_BY_STOP) &&
 		    (status >> SPU_STOP_STATUS_SHIFT == 0x2104)) {
 			ret = spu_process_callback(ctx);
Index: linux-2.6.20-rc1/include/asm-powerpc/spu.h
===================================================================
--- linux-2.6.20-rc1.orig/include/asm-powerpc/spu.h	2007-01-18
16:43:19.932605072 -0600
+++ linux-2.6.20-rc1/include/asm-powerpc/spu.h	2007-01-24
12:17:30.209676992 -0600
@@ -144,6 +144,7 @@
 
 	void* pdata; /* platform private data */
 	struct sys_device sysdev;
+	int notify_active;
 };
 
 struct spu *spu_alloc(void);