[PATCH] drm/i915: Only free the oldest stale object before a fresh allocation

[PATCH] drm/i915: Only free the oldest stale object before a fresh allocation

[Chris Wilson]

Inspired by Tvrtko's critique of the reaping of the stale contexts
before allocating a new one, also limit the freed object reaping to the
oldest stale object before allocating a fresh object. Unlike contexts,
objects may have radically different sizes of backing storage, but
similar to contexts, whilst we want to prevent starvation due to
excessive freed lists, we also want do not want to delay fresh
allocations for too long. Only freeing the oldest on the freed object
list before each allocation is a reasonable compromise.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
---
 drivers/gpu/drm/i915/i915_gem.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 91145c39cd43..2048532c9cae 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -4473,9 +4473,13 @@ static void i915_gem_flush_free_objects(struct drm_i915_private *i915)
 {
 	struct llist_node *freed;
 
-	freed = llist_del_all(&i915->mm.free_list);
-	if (unlikely(freed))
+	freed = NULL;
+	if (!llist_empty(&i915->mm.free_list))
+		freed = llist_del_first(&i915->mm.free_list);
+	if (unlikely(freed)) {
+		freed->next = NULL;
 		__i915_gem_free_objects(i915, freed);
+	}
 }
 
 static void __i915_gem_free_work(struct work_struct *work)
-- 
2.13.2

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

✓ Fi.CI.BAT: success for drm/i915: Only free the oldest stale object before a fresh allocation

[Patchwork]

== Series Details ==

Series: drm/i915: Only free the oldest stale object before a fresh allocation
URL   : https://patchwork.freedesktop.org/series/26864/
State : success

== Summary ==

Series 26864v1 drm/i915: Only free the oldest stale object before a fresh allocation
https://patchwork.freedesktop.org/api/1.0/series/26864/revisions/1/mbox/

Test kms_busy:
        Subgroup basic-flip-default-b:
                dmesg-warn -> PASS       (fi-skl-6700hq) fdo#101144

fdo#101144 https://bugs.freedesktop.org/show_bug.cgi?id=101144

fi-bdw-5557u     total:279  pass:268  dwarn:0   dfail:0   fail:0   skip:11  time:437s
fi-blb-e6850     total:279  pass:224  dwarn:1   dfail:0   fail:0   skip:54  time:358s
fi-bsw-n3050     total:279  pass:243  dwarn:0   dfail:0   fail:0   skip:36  time:538s
fi-bxt-j4205     total:279  pass:260  dwarn:0   dfail:0   fail:0   skip:19  time:508s
fi-byt-j1900     total:279  pass:254  dwarn:1   dfail:0   fail:0   skip:24  time:485s
fi-byt-n2820     total:279  pass:250  dwarn:1   dfail:0   fail:0   skip:28  time:480s
fi-glk-2a        total:279  pass:260  dwarn:0   dfail:0   fail:0   skip:19  time:593s
fi-hsw-4770      total:279  pass:263  dwarn:0   dfail:0   fail:0   skip:16  time:435s
fi-hsw-4770r     total:279  pass:263  dwarn:0   dfail:0   fail:0   skip:16  time:415s
fi-ilk-650       total:279  pass:229  dwarn:0   dfail:0   fail:0   skip:50  time:427s
fi-ivb-3520m     total:279  pass:261  dwarn:0   dfail:0   fail:0   skip:18  time:496s
fi-ivb-3770      total:279  pass:261  dwarn:0   dfail:0   fail:0   skip:18  time:474s
fi-kbl-7500u     total:279  pass:261  dwarn:0   dfail:0   fail:0   skip:18  time:464s
fi-kbl-7560u     total:279  pass:268  dwarn:1   dfail:0   fail:0   skip:10  time:569s
fi-kbl-r         total:279  pass:260  dwarn:1   dfail:0   fail:0   skip:18  time:584s
fi-pnv-d510      total:279  pass:221  dwarn:3   dfail:0   fail:0   skip:55  time:556s
fi-skl-6260u     total:279  pass:269  dwarn:0   dfail:0   fail:0   skip:10  time:460s
fi-skl-6700hq    total:279  pass:262  dwarn:0   dfail:0   fail:0   skip:17  time:586s
fi-skl-6700k     total:279  pass:257  dwarn:4   dfail:0   fail:0   skip:18  time:463s
fi-skl-6770hq    total:279  pass:269  dwarn:0   dfail:0   fail:0   skip:10  time:480s
fi-skl-gvtdvm    total:279  pass:266  dwarn:0   dfail:0   fail:0   skip:13  time:437s
fi-snb-2520m     total:279  pass:251  dwarn:0   dfail:0   fail:0   skip:28  time:544s
fi-snb-2600      total:279  pass:250  dwarn:0   dfail:0   fail:0   skip:29  time:406s

24346e831017070c18f3c33b74a7b098682e20f7 drm-tip: 2017y-07m-04d-15h-39m-34s UTC integration manifest
51e95b2 drm/i915: Only free the oldest stale object before a fresh allocation

== Logs ==

For more details see: https://intel-gfx-ci.01.org/CI/Patchwork_5117/
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH] drm/i915: Only free the oldest stale object before a fresh allocation

[Tvrtko Ursulin]

On 05/07/2017 18:21, Chris Wilson wrote:
> Inspired by Tvrtko's critique of the reaping of the stale contexts
> before allocating a new one, also limit the freed object reaping to the
> oldest stale object before allocating a fresh object. Unlike contexts,
> objects may have radically different sizes of backing storage, but
> similar to contexts, whilst we want to prevent starvation due to
> excessive freed lists, we also want do not want to delay fresh
> allocations for too long. Only freeing the oldest on the freed object
> list before each allocation is a reasonable compromise.
> 
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
> ---
>   drivers/gpu/drm/i915/i915_gem.c | 8 ++++++--
>   1 file changed, 6 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
> index 91145c39cd43..2048532c9cae 100644
> --- a/drivers/gpu/drm/i915/i915_gem.c
> +++ b/drivers/gpu/drm/i915/i915_gem.c
> @@ -4473,9 +4473,13 @@ static void i915_gem_flush_free_objects(struct drm_i915_private *i915)
>   {
>   	struct llist_node *freed;
>   
> -	freed = llist_del_all(&i915->mm.free_list);
> -	if (unlikely(freed))
> +	freed = NULL;
> +	if (!llist_empty(&i915->mm.free_list))
> +		freed = llist_del_first(&i915->mm.free_list);

Looks like llist_del_first already handles the empty list case by 
returning NULL so you could simplify this.

> +	if (unlikely(freed)) { > +		freed->next = NULL;
>   		__i915_gem_free_objects(i915, freed);
> +	}
>   }
>   
>   static void __i915_gem_free_work(struct work_struct *work)
> 

With the simplification:

Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>

Regards,

Tvrtko
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH] drm/i915: Only free the oldest stale object before a fresh allocation

[Chris Wilson]

Quoting Tvrtko Ursulin (2017-07-06 12:40:56)
> 
> On 05/07/2017 18:21, Chris Wilson wrote:
> > Inspired by Tvrtko's critique of the reaping of the stale contexts
> > before allocating a new one, also limit the freed object reaping to the
> > oldest stale object before allocating a fresh object. Unlike contexts,
> > objects may have radically different sizes of backing storage, but
> > similar to contexts, whilst we want to prevent starvation due to
> > excessive freed lists, we also want do not want to delay fresh
> > allocations for too long. Only freeing the oldest on the freed object
> > list before each allocation is a reasonable compromise.
> > 
> > Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> > Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
> > ---
> >   drivers/gpu/drm/i915/i915_gem.c | 8 ++++++--
> >   1 file changed, 6 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
> > index 91145c39cd43..2048532c9cae 100644
> > --- a/drivers/gpu/drm/i915/i915_gem.c
> > +++ b/drivers/gpu/drm/i915/i915_gem.c
> > @@ -4473,9 +4473,13 @@ static void i915_gem_flush_free_objects(struct drm_i915_private *i915)
> >   {
> >       struct llist_node *freed;
> >   
> > -     freed = llist_del_all(&i915->mm.free_list);
> > -     if (unlikely(freed))
> > +     freed = NULL;
> > +     if (!llist_empty(&i915->mm.free_list))
> > +             freed = llist_del_first(&i915->mm.free_list);
> 
> Looks like llist_del_first already handles the empty list case by 
> returning NULL so you could simplify this.

I didn't want the extra call here (since object allocation is far more
frequent than context allocation), but that's just me.
-Chris
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH] drm/i915: Only free the oldest stale object before a fresh allocation

[Tvrtko Ursulin]

On 06/07/2017 12:44, Chris Wilson wrote:
> Quoting Tvrtko Ursulin (2017-07-06 12:40:56)
>>
>> On 05/07/2017 18:21, Chris Wilson wrote:
>>> Inspired by Tvrtko's critique of the reaping of the stale contexts
>>> before allocating a new one, also limit the freed object reaping to the
>>> oldest stale object before allocating a fresh object. Unlike contexts,
>>> objects may have radically different sizes of backing storage, but
>>> similar to contexts, whilst we want to prevent starvation due to
>>> excessive freed lists, we also want do not want to delay fresh
>>> allocations for too long. Only freeing the oldest on the freed object
>>> list before each allocation is a reasonable compromise.
>>>
>>> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
>>> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
>>> ---
>>>    drivers/gpu/drm/i915/i915_gem.c | 8 ++++++--
>>>    1 file changed, 6 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
>>> index 91145c39cd43..2048532c9cae 100644
>>> --- a/drivers/gpu/drm/i915/i915_gem.c
>>> +++ b/drivers/gpu/drm/i915/i915_gem.c
>>> @@ -4473,9 +4473,13 @@ static void i915_gem_flush_free_objects(struct drm_i915_private *i915)
>>>    {
>>>        struct llist_node *freed;
>>>    
>>> -     freed = llist_del_all(&i915->mm.free_list);
>>> -     if (unlikely(freed))
>>> +     freed = NULL;
>>> +     if (!llist_empty(&i915->mm.free_list))
>>> +             freed = llist_del_first(&i915->mm.free_list);
>>
>> Looks like llist_del_first already handles the empty list case by
>> returning NULL so you could simplify this.
> 
> I didn't want the extra call here (since object allocation is far more
> frequent than context allocation), but that's just me.

Ah OK then, I didn't spot that angle. R-b applies to v1 then. Could be 
useful to stick a comment in there so someone does not clean it up in 
the future. :)

Regards,

Tvrtko

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH v2] drm/i915: Only free the oldest stale object before a fresh allocation

[Chris Wilson]

Inspired by Tvrtko's critique of the reaping of the stale contexts
before allocating a new one, also limit the freed object reaping to the
oldest stale object before allocating a fresh object. Unlike contexts,
objects may have radically different sizes of backing storage, but
similar to contexts, whilst we want to prevent starvation due to
excessive freed lists, we also want do not want to delay fresh
allocations for too long. Only freeing the oldest on the freed object
list before each allocation is a reasonable compromise.

v2: Only a single consumer of llist_del_first() is allowed (although
multiple llist_add are still allowed in parallel). Unlike
i915_gem_context, i915_gem_flush_free_objects() is itself not serialized
and so we need to add our own spinlock. Otherwise KASAN eventually spots
a use-after-free for the race on *first->next.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> #v1
---
 drivers/gpu/drm/i915/i915_drv.h |  1 +
 drivers/gpu/drm/i915/i915_gem.c | 14 ++++++++++++--
 2 files changed, 13 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 667fb5c44483..e1c2bc66e0fd 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -1447,6 +1447,7 @@ struct i915_gem_mm {
 	 */
 	struct llist_head free_list;
 	struct work_struct free_work;
+	spinlock_t free_lock;
 
 	/** Usable portion of the GTT for GEM */
 	dma_addr_t stolen_base; /* limited to low memory (32-bit) */
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 2c2907fef612..d57603f64e96 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -4542,9 +4542,18 @@ static void i915_gem_flush_free_objects(struct drm_i915_private *i915)
 {
 	struct llist_node *freed;
 
-	freed = llist_del_all(&i915->mm.free_list);
-	if (unlikely(freed))
+	/* Free the oldest, most stale object to keep the free_list short */
+	freed = NULL;
+	if (!llist_empty(&i915->mm.free_list)) { /* quick test for hotpath */
+		/* Only one consumer of llist_del_first() allowed */
+		spin_lock(&i915->mm.free_lock);
+		freed = llist_del_first(&i915->mm.free_list);
+		spin_unlock(&i915->mm.free_lock);
+	}
+	if (unlikely(freed)) {
+		freed->next = NULL;
 		__i915_gem_free_objects(i915, freed);
+	}
 }
 
 static void __i915_gem_free_work(struct work_struct *work)
@@ -5026,6 +5035,7 @@ i915_gem_load_init(struct drm_i915_private *dev_priv)
 	INIT_WORK(&dev_priv->mm.free_work, __i915_gem_free_work);
 
 	spin_lock_init(&dev_priv->mm.obj_lock);
+	spin_lock_init(&dev_priv->mm.free_lock);
 	init_llist_head(&dev_priv->mm.free_list);
 	INIT_LIST_HEAD(&dev_priv->mm.unbound_list);
 	INIT_LIST_HEAD(&dev_priv->mm.bound_list);
-- 
2.13.3

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx