dma-buf: consolidate dma_fence subclass checking
Subsystems (3)
PATCH 1/4 dma-buf: consolidate dma_fence subclass checking
Consolidate the wrapper functions to check for dma_fence subclasses in the dma_fence header.
This makes it easier to document and also check the different requirements for fence containers in the subclasses.
--- include/linux/dma-fence-array.h | 15 +------------ include/linux/dma-fence-chain.h | 3 +-- include/linux/dma-fence.h | 38 +++++++++++++++++++++++++++++++++ 3 files changed, 40 insertions(+), 16 deletions(-)
include/linux/dma-fence-array.h +1 -14
@@ -45,19 +45,6 @@ struct dma_fence_array {struct irq_work work; }; -extern const struct dma_fence_ops dma_fence_array_ops; - -/** -* dma_fence_is_array - check if a fence is from the array subsclass -* @fence: fence to test -* -* Return true if it is a dma_fence_array and false otherwise. -*/ -static inline bool dma_fence_is_array (struct dma_fence *fence) -{ -return fence->ops == &dma_fence_array_ops; -} -/** * to_dma_fence_array - cast a fence to a dma_fence_array * @fence: fence to cast to a dma_fence_array
@@ -68,7 +55,7 @@ static inline bool dma_fence_is_array(struct dma_fence *fence)static inline struct dma_fence_array *to_dma_fence_array (struct dma_fence *fence) { -if (fence->ops != &dma_fence_array_ops) +if (!fence || !dma_fence_is_array (fence))return NULL ; return container_of(fence, struct dma_fence_array, base);
include/linux/dma-fence-chain.h +1 -2
@@ -49,7 +49,6 @@ struct dma_fence_chain {spinlock_t lock; }; -extern const struct dma_fence_ops dma_fence_chain_ops;/** * to_dma_fence_chain - cast a fence to a dma_fence_chain
@@ -61,7 +60,7 @@ extern const struct dma_fence_ops dma_fence_chain_ops;static inline struct dma_fence_chain *to_dma_fence_chain (struct dma_fence *fence) { -if (!fence || fence->ops != &dma_fence_chain_ops) +if (!fence || !dma_fence_is_chain (fence))return NULL ; return container_of(fence, struct dma_fence_chain, base);
include/linux/dma-fence.h +38 -0
@@ -587,4 +587,42 @@ struct dma_fence *dma_fence_get_stub(void);struct dma_fence *dma_fence_allocate_private_stub (void );u64 dma_fence_context_alloc (unsigned num); +extern const struct dma_fence_ops dma_fence_array_ops; +extern const struct dma_fence_ops dma_fence_chain_ops; + +/** +* dma_fence_is_array - check if a fence is from the array subclass +* @fence: the fence to test +* +* Return true if it is a dma_fence_array and false otherwise. +*/ +static inline bool dma_fence_is_array (struct dma_fence *fence) +{ +return fence->ops == &dma_fence_array_ops; +} + +/** +* dma_fence_is_chain - check if a fence is from the chain subclass +* @fence: the fence to test +* +* Return true if it is a dma_fence_chain and false otherwise. +*/ +static inline bool dma_fence_is_chain (struct dma_fence *fence) +{ +return fence->ops == &dma_fence_chain_ops; +} + +/** +* dma_fence_is_container - check if a fence is a container for other fences +* @fence: the fence to test +* +* Return true if this fence is a container for other fences, false otherwise. +* This is important since we can't build up large fence structure or otherwise +* we run into recursion during operation on those fences. +*/ +static inline bool dma_fence_is_container (struct dma_fence *fence) +{ +return dma_fence_is_array (fence) ||dma_fence_is_chain (fence); +} +#endif /* __LINUX_DMA_FENCE_H */
Signature · 1 line
2.25.1
PATCH 2/4 dma-buf: warn about dma_fence_array container rules
It's not allowed to nest another dma_fence container into a dma_fence_array or otherwise we can run into recursion.
Warn about that when we create a dma_fence_array.
--- drivers/dma-buf/dma-fence-array.c | 13 +++++++++++++ 1 file changed, 13 insertions(+)
drivers/dma-buf/dma-fence-array.c +13 -0
@@ -176,6 +176,19 @@ struct dma_fence_array *dma_fence_array_create(int num_fences, array->base.error =PENDING_ERROR ; +/* dma_fence_array objects should never contain any other fence +* containers or otherwise we run into recursion and potential kernel +* stack overflow on operations on the dma_fence_array. +* +* The correct way of handling this is to flatten out the array by the +* caller instead. +* +* Enforce this here by checking that we don't create a dma_fence_array +* with any container inside. +*/ +while (seqno--) +WARN_ON (dma_fence_is_container (fences[seqno])); +return array; }EXPORT_SYMBOL (dma_fence_array_create);
Signature · 1 line
2.25.1
Re: [PATCH 2/4] dma-buf: warn about dma_fence_array container rules
Quoted 20 lines · Christian König
It's not allowed to nest another
dma_fencecontainer into adma_fence_arrayor otherwise we can run into recursion.Warn about that when we create a
dma_fence_array.Signed-off-by Christian König <[email protected]>--- drivers/dma-buf/dma-fence-array.c | 13 +++++++++++++ 1 file changed, 13 insertions(+)drivers/dma-buf/dma-fence-array.c +2
-0@@ -176,6 +176,19 @@ struct dma_fence_array *dma_fence_array_create(int num_fences, array->base.error =PENDING_ERROR ; +/* dma_fence_array objects should never contain any other fence +* containers or otherwise we run into recursion and potential kernel
drivers/dma-buf/dma-fence-array.c struct dma_fence_array *dma_fence_array_create(int num_fences, · line 190
Quoted 10 lines
drivers/dma-buf/dma-fence-array.c +10
-0+* stack overflow on operations on the dma_fence_array. +* +* The correct way of handling this is to flatten out the array by the +* caller instead. +* +* Enforce this here by checking that we don't create a dma_fence_array +* with any container inside. +*/ +while (seqno--) +WARN_ON (dma_fence_is_container (fences[seqno]));
Quoted 7 lines
drivers/dma-buf/dma-fence-array.c +1
-0+return array; }EXPORT_SYMBOL (dma_fence_array_create);Signature · 1 line
2.25.1
Signature · 3 lines
Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch
PATCH 3/4 dma-buf: Warn about dma_fence_chain container rules
In reply to Christian KönigChaining of dma_fence_chain objects is only allowed through the prev fence and not through the contained fence.
Warn about that when we create a dma_fence_chain.
--- drivers/dma-buf/dma-fence-chain.c | 8 ++++++++ 1 file changed, 8 insertions(+)
drivers/dma-buf/dma-fence-chain.c +8 -0
@@ -254,5 +254,13 @@ void dma_fence_chain_init(struct dma_fence_chain *chain,dma_fence_init (&chain->base, &dma_fence_chain_ops, &chain->lock, context, seqno); + +/* Chaining dma_fence_chain container together is only allowed through +* the prev fence and not through the contained fence. +* +* The correct way of handling this is to flatten out the fence +* structure into a dma_fence_array by the caller instead. +*/ +WARN_ON (dma_fence_is_chain (fence)); }EXPORT_SYMBOL (dma_fence_chain_init);
Signature · 1 line
2.25.1
Re: [PATCH 3/4] dma-buf: Warn about dma_fence_chain container rules
Quoted 16 lines · Christian König
Chaining of
dma_fence_chainobjects is only allowed through the prev fence and not through the contained fence.Warn about that when we create a
dma_fence_chain.Signed-off-by Christian König <[email protected]>--- drivers/dma-buf/dma-fence-chain.c | 8 ++++++++ 1 file changed, 8 insertions(+)drivers/dma-buf/dma-fence-chain.c +0
-0@@ -254,5 +254,13 @@ void dma_fence_chain_init(struct dma_fence_chain *chain,
drivers/dma-buf/dma-fence-chain.c void dma_fence_chain_init(struct dma_fence_chain *chain, · line 264
Quoted 10 lines
drivers/dma-buf/dma-fence-chain.c +8
-0dma_fence_init (&chain->base, &dma_fence_chain_ops, &chain->lock, context, seqno); + +/* Chaining dma_fence_chain container together is only allowed through +* the prev fence and not through the contained fence. +* +* The correct way of handling this is to flatten out the fence +* structure into a dma_fence_array by the caller instead. +*/ +WARN_ON (dma_fence_is_chain (fence));
At first I was worried that you'd leave a chain fence in the array fence as an option, but we exclude that with the previous patch.
Quoted 5 lines
drivers/dma-buf/dma-fence-chain.c +0
-0}EXPORT_SYMBOL (dma_fence_chain_init);Signature · 1 line
2.25.1
Signature · 3 lines
Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch
PATCH 4/4 dma-buf: warn about containers in dma_resv object
In reply to Christian KönigDrivers should not add containers as shared fences to the dma_resv object, instead each fence should be added individually.
--- drivers/dma-buf/dma-resv.c | 5 +++++ 1 file changed, 5 insertions(+)
drivers/dma-buf/dma-resv.c +5 -0
@@ -256,6 +256,11 @@ void dma_resv_add_shared_fence(struct dma_resv *obj, struct dma_fence *fence)dma_resv_assert_held (obj); +/* Drivers should not add containers here, instead add each fence +* individually. +*/ +WARN_ON (dma_fence_is_container (fence)); + fobj =dma_resv_shared_list (obj); count = fobj->shared_count;
Signature · 1 line
2.25.1
Re: [PATCH 4/4] dma-buf: warn about containers in dma_resv object
Quoted 13 lines · Christian König
Drivers should not add containers as shared fences to the
dma_resvobject, instead each fence should be added individually.Signed-off-by Christian König <[email protected]>--- drivers/dma-buf/dma-resv.c | 5 +++++ 1 file changed, 5 insertions(+)drivers/dma-buf/dma-resv.c +0
-0
drivers/dma-buf/dma-resv.c void dma_resv_add_shared_fence(struct dma_resv *obj, struct dma_fence *fence) · line 262
Quoted 7 lines
drivers/dma-buf/dma-resv.c +4
-0@@ -256,6 +256,11 @@ void dma_resv_add_shared_fence(struct dma_resv *obj, struct dma_fence *fence)dma_resv_assert_held (obj); +/* Drivers should not add containers here, instead add each fence +* individually. +*/ +WARN_ON (dma_fence_is_container (fence));
I'm honestly not sure whether this could go boom, so good if we push this asap and let it soak in linux-next for the entire release cycle.
On the change itself, we'll see what it brings.
-Daniel
Quoted 7 lines
drivers/dma-buf/dma-resv.c +1
-0+ fobj =dma_resv_shared_list (obj); count = fobj->shared_count;Signature · 1 line
2.25.1
Signature · 3 lines
Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch
Quoted 99 lines · Christian König
Consolidate the wrapper functions to check for
dma_fencesubclasses in thedma_fenceheader.This makes it easier to document and also check the different requirements for fence containers in the subclasses.
Signed-off-by Christian König <[email protected]>--- include/linux/dma-fence-array.h | 15 +------------ include/linux/dma-fence-chain.h | 3 +-- include/linux/dma-fence.h | 38 +++++++++++++++++++++++++++++++++ 3 files changed, 40 insertions(+), 16 deletions(-)include/linux/dma-fence-array.h +1
-14@@ -45,19 +45,6 @@ struct dma_fence_array {struct irq_work work; }; -extern const struct dma_fence_ops dma_fence_array_ops; - -/** -* dma_fence_is_array - check if a fence is from the array subsclass -* @fence: fence to test -* -* Return true if it is a dma_fence_array and false otherwise. -*/ -static inline bool dma_fence_is_array (struct dma_fence *fence) -{ -return fence->ops == &dma_fence_array_ops; -} -/** * to_dma_fence_array - cast a fence to a dma_fence_array * @fence: fence to cast to a dma_fence_array @@ -68,7 +55,7 @@ static inline bool dma_fence_is_array(struct dma_fence *fence)static inline struct dma_fence_array *to_dma_fence_array (struct dma_fence *fence) { -if (fence->ops != &dma_fence_array_ops) +if (!fence || !dma_fence_is_array (fence))return NULL ; return container_of(fence, struct dma_fence_array, base);include/linux/dma-fence-chain.h +1
-2@@ -49,7 +49,6 @@ struct dma_fence_chain {spinlock_t lock; }; -extern const struct dma_fence_ops dma_fence_chain_ops;/** * to_dma_fence_chain - cast a fence to a dma_fence_chain @@ -61,7 +60,7 @@ extern const struct dma_fence_ops dma_fence_chain_ops;static inline struct dma_fence_chain *to_dma_fence_chain (struct dma_fence *fence) { -if (!fence || fence->ops != &dma_fence_chain_ops) +if (!fence || !dma_fence_is_chain (fence))return NULL ; return container_of(fence, struct dma_fence_chain, base);include/linux/dma-fence.h +24
-0@@ -587,4 +587,42 @@ struct dma_fence *dma_fence_get_stub(void);struct dma_fence *dma_fence_allocate_private_stub (void );u64 dma_fence_context_alloc (unsigned num); +extern const struct dma_fence_ops dma_fence_array_ops; +extern const struct dma_fence_ops dma_fence_chain_ops; + +/** +* dma_fence_is_array - check if a fence is from the array subclass +* @fence: the fence to test +* +* Return true if it is a dma_fence_array and false otherwise. +*/ +static inline bool dma_fence_is_array (struct dma_fence *fence) +{ +return fence->ops == &dma_fence_array_ops; +} + +/** +* dma_fence_is_chain - check if a fence is from the chain subclass +* @fence: the fence to test +* +* Return true if it is a dma_fence_chain and false otherwise. +*/ +static inline bool dma_fence_is_chain (struct dma_fence *fence) +{ +return fence->ops == &dma_fence_chain_ops; +}
include/linux/dma-fence.h struct dma_fence *dma_fence_get_stub(void); · line 623
Quoted 10 lines
include/linux/dma-fence.h +10
-0+ +/** +* dma_fence_is_container - check if a fence is a container for other fences +* @fence: the fence to test +* +* Return true if this fence is a container for other fences, false otherwise. +* This is important since we can't build up large fence structure or otherwise +* we run into recursion during operation on those fences. +*/ +static inline bool dma_fence_is_container (struct dma_fence *fence)
Code looks all good, but I'm not super enthusiastic about exporting the ops to drivers and letting them do random nonsense. At least i915 does pretty enormous amounts of stuff with that instead of having pushed priority boosting into dma-fence as a proper concept. And maybe a few other things.
Now i915-gem team having gone off the rails of good upstream conduct is another thing maybe, but I'd like to not encourage that.
So could we perhaps do this all in header which is entirely private to drivers/dma-buf, like dma-fence-internal or so? And maybe whack a big fixme onto the current abuse in drivers (of which __dma_fence_is_chain() gets a special price for "not how upstream should be done" *sigh*).
Cheers, Daniel
Quoted 8 lines
include/linux/dma-fence.h +4
-0+{ +return dma_fence_is_array (fence) ||dma_fence_is_chain (fence); +} +#endif /* __LINUX_DMA_FENCE_H */ Signature · 1 line
2.25.1
Signature · 3 lines
Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch
Am 19.01.22 um 18:16 schrieb Daniel Vetter:
Quoted 116 lines · Christian König
Quoted 99 lines
Consolidate the wrapper functions to check for
dma_fencesubclasses in thedma_fenceheader.This makes it easier to document and also check the different requirements for fence containers in the subclasses.
Signed-off-by Christian König <[email protected]>--- include/linux/dma-fence-array.h | 15 +------------ include/linux/dma-fence-chain.h | 3 +-- include/linux/dma-fence.h | 38 +++++++++++++++++++++++++++++++++ 3 files changed, 40 insertions(+), 16 deletions(-)include/linux/dma-fence-array.h +1
-14@@ -45,19 +45,6 @@ struct dma_fence_array {struct irq_work work; }; -extern const struct dma_fence_ops dma_fence_array_ops; - -/** -* dma_fence_is_array - check if a fence is from the array subsclass -* @fence: fence to test -* -* Return true if it is a dma_fence_array and false otherwise. -*/ -static inline bool dma_fence_is_array (struct dma_fence *fence) -{ -return fence->ops == &dma_fence_array_ops; -} -/** * to_dma_fence_array - cast a fence to a dma_fence_array * @fence: fence to cast to a dma_fence_array @@ -68,7 +55,7 @@ static inline bool dma_fence_is_array(struct dma_fence *fence)static inline struct dma_fence_array *to_dma_fence_array (struct dma_fence *fence) { -if (fence->ops != &dma_fence_array_ops) +if (!fence || !dma_fence_is_array (fence))return NULL ; return container_of(fence, struct dma_fence_array, base);include/linux/dma-fence-chain.h +1
-2@@ -49,7 +49,6 @@ struct dma_fence_chain {spinlock_t lock; }; -extern const struct dma_fence_ops dma_fence_chain_ops;/** * to_dma_fence_chain - cast a fence to a dma_fence_chain @@ -61,7 +60,7 @@ extern const struct dma_fence_ops dma_fence_chain_ops;static inline struct dma_fence_chain *to_dma_fence_chain (struct dma_fence *fence) { -if (!fence || fence->ops != &dma_fence_chain_ops) +if (!fence || !dma_fence_is_chain (fence))return NULL ; return container_of(fence, struct dma_fence_chain, base);include/linux/dma-fence.h +24
-0@@ -587,4 +587,42 @@ struct dma_fence *dma_fence_get_stub(void);struct dma_fence *dma_fence_allocate_private_stub (void );u64 dma_fence_context_alloc (unsigned num); +extern const struct dma_fence_ops dma_fence_array_ops; +extern const struct dma_fence_ops dma_fence_chain_ops; + +/** +* dma_fence_is_array - check if a fence is from the array subclass +* @fence: the fence to test +* +* Return true if it is a dma_fence_array and false otherwise. +*/ +static inline bool dma_fence_is_array (struct dma_fence *fence) +{ +return fence->ops == &dma_fence_array_ops; +} + +/** +* dma_fence_is_chain - check if a fence is from the chain subclass +* @fence: the fence to test +* +* Return true if it is a dma_fence_chain and false otherwise. +*/ +static inline bool dma_fence_is_chain (struct dma_fence *fence) +{ +return fence->ops == &dma_fence_chain_ops; +}include/linux/dma-fence.h struct dma_fence *dma_fence_get_stub(void); · line 623
Quoted 10 lines
include/linux/dma-fence.h +10
-0+ +/** +* dma_fence_is_container - check if a fence is a container for other fences +* @fence: the fence to test +* +* Return true if this fence is a container for other fences, false otherwise. +* This is important since we can't build up large fence structure or otherwise +* we run into recursion during operation on those fences. +*/ +static inline bool dma_fence_is_container (struct dma_fence *fence)Code looks all good, but I'm not super enthusiastic about exporting the ops to drivers and letting them do random nonsense. At least i915 does pretty enormous amounts of stuff with that instead of having pushed priority boosting into dma-fence as a proper concept. And maybe a few other things.
Quoted 7 lines
Now i915-gem team having gone off the rails of good upstream conduct is another thing maybe, but I'd like to not encourage that.
So could we perhaps do this all in header which is entirely private to
drivers/dma-buf, like dma-fence-internal or so? And maybe whack a big fixme onto the current abuse in drivers (of which__dma_fence_is_chain()gets a special price for "not how upstream should be done" *sigh*).
WTF is __dma_fence_is_chain? Seeing that for the first time now.
And yes even if you do priority boosting manually that code in i915 is just way to complicated.
I'm sure you don't have any objections that I clean up that mess now you pointed it out :)
Thanks,
Christian.
Quoted 11 lines
Cheers, Daniel
Quoted 8 lines
include/linux/dma-fence.h +4
-0+{ +return dma_fence_is_array (fence) ||dma_fence_is_chain (fence); +} +#endif /* __LINUX_DMA_FENCE_H */ Signature · 1 line
2.25.1
Quoted 121 lines · Christian König
Am 19.01.22 um 18:16 schrieb Daniel Vetter:
Quoted 119 lines · Christian König
Quoted 99 lines
Consolidate the wrapper functions to check for
dma_fencesubclasses in thedma_fenceheader.This makes it easier to document and also check the different requirements for fence containers in the subclasses.
Signed-off-by Christian König <[email protected]>--- include/linux/dma-fence-array.h | 15 +------------ include/linux/dma-fence-chain.h | 3 +-- include/linux/dma-fence.h | 38 +++++++++++++++++++++++++++++++++ 3 files changed, 40 insertions(+), 16 deletions(-)include/linux/dma-fence-array.h +1
-14@@ -45,19 +45,6 @@ struct dma_fence_array {struct irq_work work; }; -extern const struct dma_fence_ops dma_fence_array_ops; - -/** -* dma_fence_is_array - check if a fence is from the array subsclass -* @fence: fence to test -* -* Return true if it is a dma_fence_array and false otherwise. -*/ -static inline bool dma_fence_is_array (struct dma_fence *fence) -{ -return fence->ops == &dma_fence_array_ops; -} -/** * to_dma_fence_array - cast a fence to a dma_fence_array * @fence: fence to cast to a dma_fence_array @@ -68,7 +55,7 @@ static inline bool dma_fence_is_array(struct dma_fence *fence)static inline struct dma_fence_array *to_dma_fence_array (struct dma_fence *fence) { -if (fence->ops != &dma_fence_array_ops) +if (!fence || !dma_fence_is_array (fence))return NULL ; return container_of(fence, struct dma_fence_array, base);include/linux/dma-fence-chain.h +1
-2@@ -49,7 +49,6 @@ struct dma_fence_chain {spinlock_t lock; }; -extern const struct dma_fence_ops dma_fence_chain_ops;/** * to_dma_fence_chain - cast a fence to a dma_fence_chain @@ -61,7 +60,7 @@ extern const struct dma_fence_ops dma_fence_chain_ops;static inline struct dma_fence_chain *to_dma_fence_chain (struct dma_fence *fence) { -if (!fence || fence->ops != &dma_fence_chain_ops) +if (!fence || !dma_fence_is_chain (fence))return NULL ; return container_of(fence, struct dma_fence_chain, base);include/linux/dma-fence.h +24
-0@@ -587,4 +587,42 @@ struct dma_fence *dma_fence_get_stub(void);struct dma_fence *dma_fence_allocate_private_stub (void );u64 dma_fence_context_alloc (unsigned num); +extern const struct dma_fence_ops dma_fence_array_ops; +extern const struct dma_fence_ops dma_fence_chain_ops; + +/** +* dma_fence_is_array - check if a fence is from the array subclass +* @fence: the fence to test +* +* Return true if it is a dma_fence_array and false otherwise. +*/ +static inline bool dma_fence_is_array (struct dma_fence *fence) +{ +return fence->ops == &dma_fence_array_ops; +} + +/** +* dma_fence_is_chain - check if a fence is from the chain subclass +* @fence: the fence to test +* +* Return true if it is a dma_fence_chain and false otherwise. +*/ +static inline bool dma_fence_is_chain (struct dma_fence *fence) +{ +return fence->ops == &dma_fence_chain_ops; +}include/linux/dma-fence.h struct dma_fence *dma_fence_get_stub(void); · line 623
Quoted 10 lines
include/linux/dma-fence.h +10
-0+ +/** +* dma_fence_is_container - check if a fence is a container for other fences +* @fence: the fence to test +* +* Return true if this fence is a container for other fences, false otherwise. +* This is important since we can't build up large fence structure or otherwise +* we run into recursion during operation on those fences. +*/ +static inline bool dma_fence_is_container (struct dma_fence *fence)Code looks all good, but I'm not super enthusiastic about exporting the ops to drivers and letting them do random nonsense. At least i915 does pretty enormous amounts of stuff with that instead of having pushed priority boosting into dma-fence as a proper concept. And maybe a few other things.
Now i915-gem team having gone off the rails of good upstream conduct is another thing maybe, but I'd like to not encourage that.
Quoted 6 lines
Quoted 4 lines
So could we perhaps do this all in header which is entirely private to
drivers/dma-buf, like dma-fence-internal or so? And maybe whack a big fixme onto the current abuse in drivers (of which__dma_fence_is_chain()gets a special price for "not how upstream should be done" *sigh*).WTF is
__dma_fence_is_chain? Seeing that for the first time now.
Yes :-/
Quoted 5 lines
And yes even if you do priority boosting manually that code in i915 is just way to complicated.
I'm sure you don't have any objections that I clean up that mess now you pointed it out :)
Go for it. I think ideally we could get rid of the dma_fence_chain_ops export. Btw similar situation exists for dma_fence_is_array, so maybe check those out too.
I think if we do want an interface for drivers then really the only thing that should be accessible is a dma_fence_is_container and a dma_fence_container_for_each_fence. Really no one's business to dig into deeper details (at least once your dma_resv series has landed).
Thanks, Daniel
Quoted 15 lines
Thanks,
Christian.Quoted 11 lines
Cheers, Daniel
Quoted 8 lines
include/linux/dma-fence.h +4
-1+{ +return dma_fence_is_array (fence) ||dma_fence_is_chain (fence); +} +#endif /* __LINUX_DMA_FENCE_H */ --2.25.1
Signature · 3 lines
Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch