* [RFC 0/3] Improve lock annotations
@ 2024-12-02 12:53 David Marchand
2024-12-02 12:53 ` [RFC 1/3] eal: add enhanced " David Marchand
` (3 more replies)
0 siblings, 4 replies; 10+ messages in thread
From: David Marchand @ 2024-12-02 12:53 UTC (permalink / raw)
To: dev; +Cc: thomas
A recent bug (see 22aa9a9c7099 ("vhost: fix deadlock in Rx async path"))
made more visible a gap in the clang thread safety annotations that
DPDK uses: no distinction is made between releasing a read lock and
releasing a write lock.
Clang 3.6 and later offers improved thread safety checks.
Marking objects as "lockable" has evolved into flagging some named
"capability". clang reports the capability name when an error is
reported (making this report a bit easier to understand).
For example, a spinlock is now flagged as:
typedef struct __rte_capability("spinlock") {
volatile RTE_ATOMIC(int) locked;
} rte_spinlock_t;
For "exclusive" locking (spinlocks / write locks), the conversion is:
- exclusive_lock_function -> acquire_capability
- exclusive_trylock_function -> try_acquire_capability
- unlock_function -> release_capability
...
For "shared" locking (read locks):
- shared_lock_function -> acquire_shared_capability
- shared_trylock_function -> try_acquire_shared_capability
- unlock_function -> release_shared_capability
...
This RFC proposes to use those annotations (sticking to the
convention of simply prefixing the compiler attributes with __rte_).
The existing "old" annotations macros are left in place in case users
started to rely on them.
Note: DPDK requirements state that clang version must be >= 3.6
(following use of C11 standard).
No check about availability of the new attributes has been added to
this series, let's see what the CI thinks of this assumption...
Comments welcome.
--
David Marchand
David Marchand (3):
eal: add enhanced lock annotations
eal: enhance lock annotations for spinlock and seqlock
eal: enhance lock annotations for rwlock
doc/api/doxy-api.conf.in | 12 ++
drivers/bus/dpaa/base/qbman/qman.c | 4 +-
drivers/net/fm10k/fm10k_ethdev.c | 4 +-
lib/eal/common/eal_memalloc.h | 2 +-
lib/eal/common/eal_private.h | 2 +-
lib/eal/include/generic/rte_rwlock.h | 23 ++--
lib/eal/include/generic/rte_spinlock.h | 14 +-
lib/eal/include/rte_eal_memconfig.h | 28 ++--
lib/eal/include/rte_lock_annotations.h | 48 +++++++
lib/eal/include/rte_seqlock.h | 4 +-
lib/ethdev/ethdev_driver.c | 4 +-
lib/ethdev/ethdev_private.h | 4 +-
lib/ethdev/rte_ethdev.c | 4 +-
lib/graph/graph_private.h | 4 +-
lib/hash/rte_cuckoo_hash.c | 8 +-
lib/vhost/iotlb.h | 8 +-
lib/vhost/vdpa.c | 2 +-
lib/vhost/vhost.c | 10 +-
lib/vhost/vhost.h | 24 ++--
lib/vhost/vhost_crypto.c | 14 +-
lib/vhost/virtio_net.c | 170 ++++++++++++-------------
lib/vhost/virtio_net_ctrl.c | 2 +-
22 files changed, 228 insertions(+), 167 deletions(-)
--
2.47.0
^ permalink raw reply [flat|nested] 10+ messages in thread* [RFC 1/3] eal: add enhanced lock annotations 2024-12-02 12:53 [RFC 0/3] Improve lock annotations David Marchand @ 2024-12-02 12:53 ` David Marchand 2024-12-02 16:13 ` Stephen Hemminger 2024-12-02 12:53 ` [RFC 2/3] eal: enhance lock annotations for spinlock and seqlock David Marchand ` (2 subsequent siblings) 3 siblings, 1 reply; 10+ messages in thread From: David Marchand @ 2024-12-02 12:53 UTC (permalink / raw) To: dev; +Cc: thomas, Tyler Retzlaff Clang 3.6+ offers enhanced lock annotations when it comes to shared vs exclusive capability/lock release. Introduce macros for those new function attributes. Signed-off-by: David Marchand <david.marchand@redhat.com> --- doc/api/doxy-api.conf.in | 12 +++++++ lib/eal/include/rte_lock_annotations.h | 48 ++++++++++++++++++++++++++ 2 files changed, 60 insertions(+) diff --git a/doc/api/doxy-api.conf.in b/doc/api/doxy-api.conf.in index d23352d300..ac490e7631 100644 --- a/doc/api/doxy-api.conf.in +++ b/doc/api/doxy-api.conf.in @@ -103,6 +103,18 @@ PREDEFINED = __DOXYGEN__ \ __rte_shared_trylock_function(x)= \ __rte_assert_shared_lock(x)= \ __rte_unlock_function(x)= \ + __rte_capability(x)= \ + __rte_requires_capability(x)= \ + __rte_acquire_capability(x)= \ + __rte_try_acquire_capability(x)= \ + __rte_release_capability(x)= \ + __rte_assert_capability(x)= \ + __rte_requires_shared_capability(x)= \ + __rte_acquire_shared_capability(x)= \ + __rte_try_acquire_shared_capability(x)= \ + __rte_release_shared_capability(x)= \ + __rte_assert_shared_capability(x)= \ + __rte_exclude_capability(x)= \ __attribute__(x)= OPTIMIZE_OUTPUT_FOR_C = YES diff --git a/lib/eal/include/rte_lock_annotations.h b/lib/eal/include/rte_lock_annotations.h index 2456a69352..4240458c53 100644 --- a/lib/eal/include/rte_lock_annotations.h +++ b/lib/eal/include/rte_lock_annotations.h @@ -43,6 +43,36 @@ extern "C" { #define __rte_locks_excluded(...) \ __attribute__((locks_excluded(__VA_ARGS__))) + +#define __rte_capability(...) \ + __attribute__((capability(__VA_ARGS__))) + +#define __rte_requires_capability(...) \ + __attribute__((requires_capability(__VA_ARGS__))) +#define __rte_acquire_capability(...) \ + __attribute__((acquire_capability(__VA_ARGS__))) +#define __rte_try_acquire_capability(ret, ...) \ + __attribute__((try_acquire_capability(ret, __VA_ARGS__))) +#define __rte_release_capability(...) \ + __attribute__((release_capability(__VA_ARGS__))) +#define __rte_assert_capability(...) \ + __attribute__((assert_capability(__VA_ARGS__))) + +#define __rte_requires_shared_capability(...) \ + __attribute__((requires_shared_capability(__VA_ARGS__))) +#define __rte_acquire_shared_capability(...) \ + __attribute__((acquire_shared_capability(__VA_ARGS__))) +#define __rte_try_acquire_shared_capability(ret, ...) \ + __attribute__((try_acquire_shared_capability(ret, __VA_ARGS__))) +#define __rte_release_shared_capability(...) \ + __attribute__((release_shared_capability(__VA_ARGS__))) +#define __rte_assert_shared_capability(...) \ + __attribute__((assert_shared_capability(__VA_ARGS__))) + +#define __rte_exclude_capability(...) \ + __attribute__((exclude_capability(__VA_ARGS__))) + + #define __rte_no_thread_safety_analysis \ __attribute__((no_thread_safety_analysis)) @@ -67,6 +97,24 @@ extern "C" { #define __rte_locks_excluded(...) + +#define __rte_capability(...) + +#define __rte_requires_capability(...) +#define __rte_acquire_capability(...) +#define __rte_try_acquire_capability(...) +#define __rte_release_capability(...) +#define __rte_assert_capability(...) + +#define __rte_requires_shared_capability(...) +#define __rte_acquire_shared_capability(...) +#define __rte_try_acquire_shared_capability(...) +#define __rte_release_shared_capability(...) +#define __rte_assert_shared_capability(...) + +#define __rte_exclude_capability(...) + + #define __rte_no_thread_safety_analysis #endif /* RTE_ANNOTATE_LOCKS */ -- 2.47.0 ^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [RFC 1/3] eal: add enhanced lock annotations 2024-12-02 12:53 ` [RFC 1/3] eal: add enhanced " David Marchand @ 2024-12-02 16:13 ` Stephen Hemminger 0 siblings, 0 replies; 10+ messages in thread From: Stephen Hemminger @ 2024-12-02 16:13 UTC (permalink / raw) To: David Marchand; +Cc: dev, thomas, Tyler Retzlaff On Mon, 2 Dec 2024 13:53:14 +0100 David Marchand <david.marchand@redhat.com> wrote: > diff --git a/doc/api/doxy-api.conf.in b/doc/api/doxy-api.conf.in > index d23352d300..ac490e7631 100644 > --- a/doc/api/doxy-api.conf.in > +++ b/doc/api/doxy-api.conf.in > @@ -103,6 +103,18 @@ PREDEFINED = __DOXYGEN__ \ > __rte_shared_trylock_function(x)= \ > __rte_assert_shared_lock(x)= \ > __rte_unlock_function(x)= \ > + __rte_capability(x)= \ > + __rte_requires_capability(x)= \ > + __rte_acquire_capability(x)= \ > + __rte_try_acquire_capability(x)= \ > + __rte_release_capability(x)= \ > + __rte_assert_capability(x)= \ > + __rte_requires_shared_capability(x)= \ > + __rte_acquire_shared_capability(x)= \ > + __rte_try_acquire_shared_capability(x)= \ > + __rte_release_shared_capability(x)= \ > + __rte_assert_shared_capability(x)= \ > + __rte_exclude_capability(x)= \ > __attribute__(x)= I would suggest shortened names: __rte_acquires(x) __rte_releases(x) __rte_must_hold(x) Based on the original source of all these lock annotations which is sparse. ^ permalink raw reply [flat|nested] 10+ messages in thread
* [RFC 2/3] eal: enhance lock annotations for spinlock and seqlock 2024-12-02 12:53 [RFC 0/3] Improve lock annotations David Marchand 2024-12-02 12:53 ` [RFC 1/3] eal: add enhanced " David Marchand @ 2024-12-02 12:53 ` David Marchand 2024-12-05 6:18 ` Mattias Rönnblom 2024-12-02 12:53 ` [RFC 3/3] eal: enhance lock annotations for rwlock David Marchand 2024-12-12 16:00 ` [PATCH v2 0/3] Improve lock annotations David Marchand 3 siblings, 1 reply; 10+ messages in thread From: David Marchand @ 2024-12-02 12:53 UTC (permalink / raw) To: dev Cc: thomas, Hemant Agrawal, Sachin Saxena, Tyler Retzlaff, Mattias Rönnblom, Jerin Jacob, Kiran Kumar K, Nithin Dabilpuram, Zhirun Yan, Maxime Coquelin, Chenbo Xia Convert spinlock (and as a consequence seqlock) to the clang capability annotations. Signed-off-by: David Marchand <david.marchand@redhat.com> --- drivers/bus/dpaa/base/qbman/qman.c | 4 ++-- drivers/net/fm10k/fm10k_ethdev.c | 4 ++-- lib/eal/include/generic/rte_spinlock.h | 14 +++++++------- lib/eal/include/rte_eal_memconfig.h | 4 ++-- lib/eal/include/rte_seqlock.h | 4 ++-- lib/graph/graph_private.h | 4 ++-- lib/vhost/vdpa.c | 2 +- 7 files changed, 18 insertions(+), 18 deletions(-) diff --git a/drivers/bus/dpaa/base/qbman/qman.c b/drivers/bus/dpaa/base/qbman/qman.c index f92b25343a..11fabcaff5 100644 --- a/drivers/bus/dpaa/base/qbman/qman.c +++ b/drivers/bus/dpaa/base/qbman/qman.c @@ -45,7 +45,7 @@ static inline int fq_isset(struct qman_fq *fq, u32 mask) } static inline void fq_lock(struct qman_fq *fq) - __rte_exclusive_lock_function(&fq->fqlock) + __rte_acquire_capability(&fq->fqlock) __rte_no_thread_safety_analysis { if (fq_isset(fq, QMAN_FQ_FLAG_LOCKED)) @@ -53,7 +53,7 @@ static inline void fq_lock(struct qman_fq *fq) } static inline void fq_unlock(struct qman_fq *fq) - __rte_unlock_function(&fq->fqlock) + __rte_release_capability(&fq->fqlock) __rte_no_thread_safety_analysis { if (fq_isset(fq, QMAN_FQ_FLAG_LOCKED)) diff --git a/drivers/net/fm10k/fm10k_ethdev.c b/drivers/net/fm10k/fm10k_ethdev.c index 7b490bea17..747042d621 100644 --- a/drivers/net/fm10k/fm10k_ethdev.c +++ b/drivers/net/fm10k/fm10k_ethdev.c @@ -116,7 +116,7 @@ fm10k_mbx_initlock(struct fm10k_hw *hw) static void fm10k_mbx_lock(struct fm10k_hw *hw) - __rte_exclusive_lock_function(FM10K_DEV_PRIVATE_TO_MBXLOCK(hw->back)) + __rte_acquire_capability(FM10K_DEV_PRIVATE_TO_MBXLOCK(hw->back)) { while (!rte_spinlock_trylock(FM10K_DEV_PRIVATE_TO_MBXLOCK(hw->back))) rte_delay_us(FM10K_MBXLOCK_DELAY_US); @@ -124,7 +124,7 @@ fm10k_mbx_lock(struct fm10k_hw *hw) static void fm10k_mbx_unlock(struct fm10k_hw *hw) - __rte_unlock_function(FM10K_DEV_PRIVATE_TO_MBXLOCK(hw->back)) + __rte_release_capability(FM10K_DEV_PRIVATE_TO_MBXLOCK(hw->back)) { rte_spinlock_unlock(FM10K_DEV_PRIVATE_TO_MBXLOCK(hw->back)); } diff --git a/lib/eal/include/generic/rte_spinlock.h b/lib/eal/include/generic/rte_spinlock.h index c2980601b2..c907d4e45c 100644 --- a/lib/eal/include/generic/rte_spinlock.h +++ b/lib/eal/include/generic/rte_spinlock.h @@ -32,7 +32,7 @@ extern "C" { /** * The rte_spinlock_t type. */ -typedef struct __rte_lockable { +typedef struct __rte_capability("spinlock") { volatile RTE_ATOMIC(int) locked; /**< lock status 0 = unlocked, 1 = locked */ } rte_spinlock_t; @@ -61,7 +61,7 @@ rte_spinlock_init(rte_spinlock_t *sl) */ static inline void rte_spinlock_lock(rte_spinlock_t *sl) - __rte_exclusive_lock_function(sl); + __rte_acquire_capability(sl); #ifdef RTE_FORCE_INTRINSICS static inline void @@ -87,7 +87,7 @@ rte_spinlock_lock(rte_spinlock_t *sl) */ static inline void rte_spinlock_unlock(rte_spinlock_t *sl) - __rte_unlock_function(sl); + __rte_release_capability(sl); #ifdef RTE_FORCE_INTRINSICS static inline void @@ -109,7 +109,7 @@ rte_spinlock_unlock(rte_spinlock_t *sl) __rte_warn_unused_result static inline int rte_spinlock_trylock(rte_spinlock_t *sl) - __rte_exclusive_trylock_function(1, sl); + __rte_try_acquire_capability(true, sl); #ifdef RTE_FORCE_INTRINSICS static inline int @@ -158,7 +158,7 @@ static inline int rte_tm_supported(void); */ static inline void rte_spinlock_lock_tm(rte_spinlock_t *sl) - __rte_exclusive_lock_function(sl); + __rte_acquire_capability(sl); /** * Commit hardware memory transaction or release the spinlock if @@ -169,7 +169,7 @@ rte_spinlock_lock_tm(rte_spinlock_t *sl) */ static inline void rte_spinlock_unlock_tm(rte_spinlock_t *sl) - __rte_unlock_function(sl); + __rte_release_capability(sl); /** * Try to execute critical section in a hardware memory transaction, @@ -190,7 +190,7 @@ rte_spinlock_unlock_tm(rte_spinlock_t *sl) __rte_warn_unused_result static inline int rte_spinlock_trylock_tm(rte_spinlock_t *sl) - __rte_exclusive_trylock_function(1, sl); + __rte_try_acquire_capability(true, sl); /** * The rte_spinlock_recursive_t type. diff --git a/lib/eal/include/rte_eal_memconfig.h b/lib/eal/include/rte_eal_memconfig.h index 0b1d0d4ff0..55d78de334 100644 --- a/lib/eal/include/rte_eal_memconfig.h +++ b/lib/eal/include/rte_eal_memconfig.h @@ -132,14 +132,14 @@ rte_mcfg_mempool_write_unlock(void) */ void rte_mcfg_timer_lock(void) - __rte_exclusive_lock_function(rte_mcfg_timer_get_lock()); + __rte_acquire_capability(rte_mcfg_timer_get_lock()); /** * Unlock the internal EAL Timer Library lock for exclusive access. */ void rte_mcfg_timer_unlock(void) - __rte_unlock_function(rte_mcfg_timer_get_lock()); + __rte_release_capability(rte_mcfg_timer_get_lock()); /** * If true, pages are put in single files (per memseg list), diff --git a/lib/eal/include/rte_seqlock.h b/lib/eal/include/rte_seqlock.h index e0e94900d1..670a680443 100644 --- a/lib/eal/include/rte_seqlock.h +++ b/lib/eal/include/rte_seqlock.h @@ -197,7 +197,7 @@ rte_seqlock_read_retry(const rte_seqlock_t *seqlock, uint32_t begin_sn) */ static inline void rte_seqlock_write_lock(rte_seqlock_t *seqlock) - __rte_exclusive_lock_function(&seqlock->lock) + __rte_acquire_capability(&seqlock->lock) { /* To synchronize with other writers. */ rte_spinlock_lock(&seqlock->lock); @@ -219,7 +219,7 @@ rte_seqlock_write_lock(rte_seqlock_t *seqlock) */ static inline void rte_seqlock_write_unlock(rte_seqlock_t *seqlock) - __rte_unlock_function(&seqlock->lock) + __rte_release_capability(&seqlock->lock) { rte_seqcount_write_end(&seqlock->count); diff --git a/lib/graph/graph_private.h b/lib/graph/graph_private.h index da48d73587..813dd78b9d 100644 --- a/lib/graph/graph_private.h +++ b/lib/graph/graph_private.h @@ -221,7 +221,7 @@ graph_spinlock_get(void); * Take a lock on the graph internal spin lock. */ void graph_spinlock_lock(void) - __rte_exclusive_lock_function(graph_spinlock_get()); + __rte_acquire_capability(graph_spinlock_get()); /** * @internal @@ -229,7 +229,7 @@ void graph_spinlock_lock(void) * Release a lock on the graph internal spin lock. */ void graph_spinlock_unlock(void) - __rte_unlock_function(graph_spinlock_get()); + __rte_release_capability(graph_spinlock_get()); /* Graph operations */ /** diff --git a/lib/vhost/vdpa.c b/lib/vhost/vdpa.c index 8abb073675..329dd572b7 100644 --- a/lib/vhost/vdpa.c +++ b/lib/vhost/vdpa.c @@ -32,7 +32,7 @@ static struct vdpa_device_list * const vdpa_device_list static struct rte_vdpa_device * __vdpa_find_device_by_name(const char *name) - __rte_exclusive_locks_required(&vdpa_device_list_lock) + __rte_requires_capability(&vdpa_device_list_lock) { struct rte_vdpa_device *dev, *ret = NULL; -- 2.47.0 ^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [RFC 2/3] eal: enhance lock annotations for spinlock and seqlock 2024-12-02 12:53 ` [RFC 2/3] eal: enhance lock annotations for spinlock and seqlock David Marchand @ 2024-12-05 6:18 ` Mattias Rönnblom 0 siblings, 0 replies; 10+ messages in thread From: Mattias Rönnblom @ 2024-12-05 6:18 UTC (permalink / raw) To: David Marchand, dev Cc: thomas, Hemant Agrawal, Sachin Saxena, Tyler Retzlaff, Mattias Rönnblom, Jerin Jacob, Kiran Kumar K, Nithin Dabilpuram, Zhirun Yan, Maxime Coquelin, Chenbo Xia On 2024-12-02 13:53, David Marchand wrote: > Convert spinlock (and as a consequence seqlock) to the clang > capability annotations. > Acked-by: Mattias Rönnnblom <mattias.ronnblom@ericsson.com> > Signed-off-by: David Marchand <david.marchand@redhat.com> > --- > drivers/bus/dpaa/base/qbman/qman.c | 4 ++-- > drivers/net/fm10k/fm10k_ethdev.c | 4 ++-- > lib/eal/include/generic/rte_spinlock.h | 14 +++++++------- > lib/eal/include/rte_eal_memconfig.h | 4 ++-- > lib/eal/include/rte_seqlock.h | 4 ++-- > lib/graph/graph_private.h | 4 ++-- > lib/vhost/vdpa.c | 2 +- > 7 files changed, 18 insertions(+), 18 deletions(-) > > diff --git a/drivers/bus/dpaa/base/qbman/qman.c b/drivers/bus/dpaa/base/qbman/qman.c > index f92b25343a..11fabcaff5 100644 > --- a/drivers/bus/dpaa/base/qbman/qman.c > +++ b/drivers/bus/dpaa/base/qbman/qman.c > @@ -45,7 +45,7 @@ static inline int fq_isset(struct qman_fq *fq, u32 mask) > } > > static inline void fq_lock(struct qman_fq *fq) > - __rte_exclusive_lock_function(&fq->fqlock) > + __rte_acquire_capability(&fq->fqlock) > __rte_no_thread_safety_analysis > { > if (fq_isset(fq, QMAN_FQ_FLAG_LOCKED)) > @@ -53,7 +53,7 @@ static inline void fq_lock(struct qman_fq *fq) > } > > static inline void fq_unlock(struct qman_fq *fq) > - __rte_unlock_function(&fq->fqlock) > + __rte_release_capability(&fq->fqlock) > __rte_no_thread_safety_analysis > { > if (fq_isset(fq, QMAN_FQ_FLAG_LOCKED)) > diff --git a/drivers/net/fm10k/fm10k_ethdev.c b/drivers/net/fm10k/fm10k_ethdev.c > index 7b490bea17..747042d621 100644 > --- a/drivers/net/fm10k/fm10k_ethdev.c > +++ b/drivers/net/fm10k/fm10k_ethdev.c > @@ -116,7 +116,7 @@ fm10k_mbx_initlock(struct fm10k_hw *hw) > > static void > fm10k_mbx_lock(struct fm10k_hw *hw) > - __rte_exclusive_lock_function(FM10K_DEV_PRIVATE_TO_MBXLOCK(hw->back)) > + __rte_acquire_capability(FM10K_DEV_PRIVATE_TO_MBXLOCK(hw->back)) > { > while (!rte_spinlock_trylock(FM10K_DEV_PRIVATE_TO_MBXLOCK(hw->back))) > rte_delay_us(FM10K_MBXLOCK_DELAY_US); > @@ -124,7 +124,7 @@ fm10k_mbx_lock(struct fm10k_hw *hw) > > static void > fm10k_mbx_unlock(struct fm10k_hw *hw) > - __rte_unlock_function(FM10K_DEV_PRIVATE_TO_MBXLOCK(hw->back)) > + __rte_release_capability(FM10K_DEV_PRIVATE_TO_MBXLOCK(hw->back)) > { > rte_spinlock_unlock(FM10K_DEV_PRIVATE_TO_MBXLOCK(hw->back)); > } > diff --git a/lib/eal/include/generic/rte_spinlock.h b/lib/eal/include/generic/rte_spinlock.h > index c2980601b2..c907d4e45c 100644 > --- a/lib/eal/include/generic/rte_spinlock.h > +++ b/lib/eal/include/generic/rte_spinlock.h > @@ -32,7 +32,7 @@ extern "C" { > /** > * The rte_spinlock_t type. > */ > -typedef struct __rte_lockable { > +typedef struct __rte_capability("spinlock") { > volatile RTE_ATOMIC(int) locked; /**< lock status 0 = unlocked, 1 = locked */ > } rte_spinlock_t; > > @@ -61,7 +61,7 @@ rte_spinlock_init(rte_spinlock_t *sl) > */ > static inline void > rte_spinlock_lock(rte_spinlock_t *sl) > - __rte_exclusive_lock_function(sl); > + __rte_acquire_capability(sl); > > #ifdef RTE_FORCE_INTRINSICS > static inline void > @@ -87,7 +87,7 @@ rte_spinlock_lock(rte_spinlock_t *sl) > */ > static inline void > rte_spinlock_unlock(rte_spinlock_t *sl) > - __rte_unlock_function(sl); > + __rte_release_capability(sl); > > #ifdef RTE_FORCE_INTRINSICS > static inline void > @@ -109,7 +109,7 @@ rte_spinlock_unlock(rte_spinlock_t *sl) > __rte_warn_unused_result > static inline int > rte_spinlock_trylock(rte_spinlock_t *sl) > - __rte_exclusive_trylock_function(1, sl); > + __rte_try_acquire_capability(true, sl); > > #ifdef RTE_FORCE_INTRINSICS > static inline int > @@ -158,7 +158,7 @@ static inline int rte_tm_supported(void); > */ > static inline void > rte_spinlock_lock_tm(rte_spinlock_t *sl) > - __rte_exclusive_lock_function(sl); > + __rte_acquire_capability(sl); > > /** > * Commit hardware memory transaction or release the spinlock if > @@ -169,7 +169,7 @@ rte_spinlock_lock_tm(rte_spinlock_t *sl) > */ > static inline void > rte_spinlock_unlock_tm(rte_spinlock_t *sl) > - __rte_unlock_function(sl); > + __rte_release_capability(sl); > > /** > * Try to execute critical section in a hardware memory transaction, > @@ -190,7 +190,7 @@ rte_spinlock_unlock_tm(rte_spinlock_t *sl) > __rte_warn_unused_result > static inline int > rte_spinlock_trylock_tm(rte_spinlock_t *sl) > - __rte_exclusive_trylock_function(1, sl); > + __rte_try_acquire_capability(true, sl); > > /** > * The rte_spinlock_recursive_t type. > diff --git a/lib/eal/include/rte_eal_memconfig.h b/lib/eal/include/rte_eal_memconfig.h > index 0b1d0d4ff0..55d78de334 100644 > --- a/lib/eal/include/rte_eal_memconfig.h > +++ b/lib/eal/include/rte_eal_memconfig.h > @@ -132,14 +132,14 @@ rte_mcfg_mempool_write_unlock(void) > */ > void > rte_mcfg_timer_lock(void) > - __rte_exclusive_lock_function(rte_mcfg_timer_get_lock()); > + __rte_acquire_capability(rte_mcfg_timer_get_lock()); > > /** > * Unlock the internal EAL Timer Library lock for exclusive access. > */ > void > rte_mcfg_timer_unlock(void) > - __rte_unlock_function(rte_mcfg_timer_get_lock()); > + __rte_release_capability(rte_mcfg_timer_get_lock()); > > /** > * If true, pages are put in single files (per memseg list), > diff --git a/lib/eal/include/rte_seqlock.h b/lib/eal/include/rte_seqlock.h > index e0e94900d1..670a680443 100644 > --- a/lib/eal/include/rte_seqlock.h > +++ b/lib/eal/include/rte_seqlock.h > @@ -197,7 +197,7 @@ rte_seqlock_read_retry(const rte_seqlock_t *seqlock, uint32_t begin_sn) > */ > static inline void > rte_seqlock_write_lock(rte_seqlock_t *seqlock) > - __rte_exclusive_lock_function(&seqlock->lock) > + __rte_acquire_capability(&seqlock->lock) > { > /* To synchronize with other writers. */ > rte_spinlock_lock(&seqlock->lock); > @@ -219,7 +219,7 @@ rte_seqlock_write_lock(rte_seqlock_t *seqlock) > */ > static inline void > rte_seqlock_write_unlock(rte_seqlock_t *seqlock) > - __rte_unlock_function(&seqlock->lock) > + __rte_release_capability(&seqlock->lock) > { > rte_seqcount_write_end(&seqlock->count); > > diff --git a/lib/graph/graph_private.h b/lib/graph/graph_private.h > index da48d73587..813dd78b9d 100644 > --- a/lib/graph/graph_private.h > +++ b/lib/graph/graph_private.h > @@ -221,7 +221,7 @@ graph_spinlock_get(void); > * Take a lock on the graph internal spin lock. > */ > void graph_spinlock_lock(void) > - __rte_exclusive_lock_function(graph_spinlock_get()); > + __rte_acquire_capability(graph_spinlock_get()); > > /** > * @internal > @@ -229,7 +229,7 @@ void graph_spinlock_lock(void) > * Release a lock on the graph internal spin lock. > */ > void graph_spinlock_unlock(void) > - __rte_unlock_function(graph_spinlock_get()); > + __rte_release_capability(graph_spinlock_get()); > > /* Graph operations */ > /** > diff --git a/lib/vhost/vdpa.c b/lib/vhost/vdpa.c > index 8abb073675..329dd572b7 100644 > --- a/lib/vhost/vdpa.c > +++ b/lib/vhost/vdpa.c > @@ -32,7 +32,7 @@ static struct vdpa_device_list * const vdpa_device_list > > static struct rte_vdpa_device * > __vdpa_find_device_by_name(const char *name) > - __rte_exclusive_locks_required(&vdpa_device_list_lock) > + __rte_requires_capability(&vdpa_device_list_lock) > { > struct rte_vdpa_device *dev, *ret = NULL; > ^ permalink raw reply [flat|nested] 10+ messages in thread
* [RFC 3/3] eal: enhance lock annotations for rwlock 2024-12-02 12:53 [RFC 0/3] Improve lock annotations David Marchand 2024-12-02 12:53 ` [RFC 1/3] eal: add enhanced " David Marchand 2024-12-02 12:53 ` [RFC 2/3] eal: enhance lock annotations for spinlock and seqlock David Marchand @ 2024-12-02 12:53 ` David Marchand 2024-12-12 16:00 ` [PATCH v2 0/3] Improve lock annotations David Marchand 3 siblings, 0 replies; 10+ messages in thread From: David Marchand @ 2024-12-02 12:53 UTC (permalink / raw) To: dev Cc: thomas, Anatoly Burakov, Tyler Retzlaff, Ferruh Yigit, Andrew Rybchenko, Yipeng Wang, Sameh Gobriel, Bruce Richardson, Vladimir Medvedkin, Maxime Coquelin, Chenbo Xia Convert rwlock to the clang capability annotations. Signed-off-by: David Marchand <david.marchand@redhat.com> --- lib/eal/common/eal_memalloc.h | 2 +- lib/eal/common/eal_private.h | 2 +- lib/eal/include/generic/rte_rwlock.h | 23 ++-- lib/eal/include/rte_eal_memconfig.h | 24 ++-- lib/ethdev/ethdev_driver.c | 4 +- lib/ethdev/ethdev_private.h | 4 +- lib/ethdev/rte_ethdev.c | 4 +- lib/hash/rte_cuckoo_hash.c | 8 +- lib/vhost/iotlb.h | 8 +- lib/vhost/vhost.c | 10 +- lib/vhost/vhost.h | 24 ++-- lib/vhost/vhost_crypto.c | 14 +-- lib/vhost/virtio_net.c | 170 +++++++++++++-------------- lib/vhost/virtio_net_ctrl.c | 2 +- 14 files changed, 150 insertions(+), 149 deletions(-) diff --git a/lib/eal/common/eal_memalloc.h b/lib/eal/common/eal_memalloc.h index 286ffb7633..0c267066d9 100644 --- a/lib/eal/common/eal_memalloc.h +++ b/lib/eal/common/eal_memalloc.h @@ -92,7 +92,7 @@ eal_memalloc_get_seg_fd_offset(int list_idx, int seg_idx, size_t *offset); int eal_memalloc_init(void) - __rte_shared_locks_required(rte_mcfg_mem_get_lock()); + __rte_requires_shared_capability(rte_mcfg_mem_get_lock()); int eal_memalloc_cleanup(void); diff --git a/lib/eal/common/eal_private.h b/lib/eal/common/eal_private.h index bb315dab04..89bc05ecc5 100644 --- a/lib/eal/common/eal_private.h +++ b/lib/eal/common/eal_private.h @@ -124,7 +124,7 @@ int rte_eal_memseg_init(void); * 0 on success, negative on error */ int rte_eal_memory_init(void) - __rte_shared_locks_required(rte_mcfg_mem_get_lock()); + __rte_requires_shared_capability(rte_mcfg_mem_get_lock()); /** * Configure timers diff --git a/lib/eal/include/generic/rte_rwlock.h b/lib/eal/include/generic/rte_rwlock.h index ac0474466a..197b245d11 100644 --- a/lib/eal/include/generic/rte_rwlock.h +++ b/lib/eal/include/generic/rte_rwlock.h @@ -23,6 +23,7 @@ */ #include <errno.h> +#include <stdbool.h> #include <rte_branch_prediction.h> #include <rte_common.h> @@ -57,7 +58,7 @@ extern "C" { /* Writer is waiting or has lock */ #define RTE_RWLOCK_READ 0x4 /* Reader increment */ -typedef struct __rte_lockable { +typedef struct __rte_capability("rwlock") { RTE_ATOMIC(int32_t) cnt; } rte_rwlock_t; @@ -90,7 +91,7 @@ rte_rwlock_init(rte_rwlock_t *rwl) */ static inline void rte_rwlock_read_lock(rte_rwlock_t *rwl) - __rte_shared_lock_function(rwl) + __rte_acquire_shared_capability(rwl) __rte_no_thread_safety_analysis { int32_t x; @@ -127,7 +128,7 @@ rte_rwlock_read_lock(rte_rwlock_t *rwl) */ static inline int rte_rwlock_read_trylock(rte_rwlock_t *rwl) - __rte_shared_trylock_function(0, rwl) + __rte_try_acquire_shared_capability(false, rwl) __rte_no_thread_safety_analysis { int32_t x; @@ -160,7 +161,7 @@ rte_rwlock_read_trylock(rte_rwlock_t *rwl) */ static inline void rte_rwlock_read_unlock(rte_rwlock_t *rwl) - __rte_unlock_function(rwl) + __rte_release_shared_capability(rwl) __rte_no_thread_safety_analysis { rte_atomic_fetch_sub_explicit(&rwl->cnt, RTE_RWLOCK_READ, rte_memory_order_release); @@ -178,7 +179,7 @@ rte_rwlock_read_unlock(rte_rwlock_t *rwl) */ static inline int rte_rwlock_write_trylock(rte_rwlock_t *rwl) - __rte_exclusive_trylock_function(0, rwl) + __rte_try_acquire_capability(false, rwl) __rte_no_thread_safety_analysis { int32_t x; @@ -200,7 +201,7 @@ rte_rwlock_write_trylock(rte_rwlock_t *rwl) */ static inline void rte_rwlock_write_lock(rte_rwlock_t *rwl) - __rte_exclusive_lock_function(rwl) + __rte_acquire_capability(rwl) __rte_no_thread_safety_analysis { int32_t x; @@ -238,7 +239,7 @@ rte_rwlock_write_lock(rte_rwlock_t *rwl) */ static inline void rte_rwlock_write_unlock(rte_rwlock_t *rwl) - __rte_unlock_function(rwl) + __rte_release_capability(rwl) __rte_no_thread_safety_analysis { rte_atomic_fetch_sub_explicit(&rwl->cnt, RTE_RWLOCK_WRITE, rte_memory_order_release); @@ -276,7 +277,7 @@ rte_rwlock_write_is_locked(rte_rwlock_t *rwl) */ static inline void rte_rwlock_read_lock_tm(rte_rwlock_t *rwl) - __rte_shared_lock_function(rwl); + __rte_acquire_shared_capability(rwl); /** * Commit hardware memory transaction or release the read lock if the lock is used as a fall-back @@ -286,7 +287,7 @@ rte_rwlock_read_lock_tm(rte_rwlock_t *rwl) */ static inline void rte_rwlock_read_unlock_tm(rte_rwlock_t *rwl) - __rte_unlock_function(rwl); + __rte_release_shared_capability(rwl); /** * Try to execute critical section in a hardware memory transaction, if it @@ -303,7 +304,7 @@ rte_rwlock_read_unlock_tm(rte_rwlock_t *rwl) */ static inline void rte_rwlock_write_lock_tm(rte_rwlock_t *rwl) - __rte_exclusive_lock_function(rwl); + __rte_acquire_capability(rwl); /** * Commit hardware memory transaction or release the write lock if the lock is used as a fall-back @@ -313,7 +314,7 @@ rte_rwlock_write_lock_tm(rte_rwlock_t *rwl) */ static inline void rte_rwlock_write_unlock_tm(rte_rwlock_t *rwl) - __rte_unlock_function(rwl); + __rte_release_capability(rwl); #ifdef __cplusplus } diff --git a/lib/eal/include/rte_eal_memconfig.h b/lib/eal/include/rte_eal_memconfig.h index 55d78de334..c3056021a0 100644 --- a/lib/eal/include/rte_eal_memconfig.h +++ b/lib/eal/include/rte_eal_memconfig.h @@ -48,84 +48,84 @@ rte_mcfg_ethdev_get_lock(void); */ void rte_mcfg_mem_read_lock(void) - __rte_shared_lock_function(rte_mcfg_mem_get_lock()); + __rte_acquire_shared_capability(rte_mcfg_mem_get_lock()); /** * Unlock the internal EAL shared memory configuration for shared access. */ void rte_mcfg_mem_read_unlock(void) - __rte_unlock_function(rte_mcfg_mem_get_lock()); + __rte_release_shared_capability(rte_mcfg_mem_get_lock()); /** * Lock the internal EAL shared memory configuration for exclusive access. */ void rte_mcfg_mem_write_lock(void) - __rte_exclusive_lock_function(rte_mcfg_mem_get_lock()); + __rte_acquire_capability(rte_mcfg_mem_get_lock()); /** * Unlock the internal EAL shared memory configuration for exclusive access. */ void rte_mcfg_mem_write_unlock(void) - __rte_unlock_function(rte_mcfg_mem_get_lock()); + __rte_release_capability(rte_mcfg_mem_get_lock()); /** * Lock the internal EAL TAILQ list for shared access. */ void rte_mcfg_tailq_read_lock(void) - __rte_shared_lock_function(rte_mcfg_tailq_get_lock()); + __rte_acquire_shared_capability(rte_mcfg_tailq_get_lock()); /** * Unlock the internal EAL TAILQ list for shared access. */ void rte_mcfg_tailq_read_unlock(void) - __rte_unlock_function(rte_mcfg_tailq_get_lock()); + __rte_release_shared_capability(rte_mcfg_tailq_get_lock()); /** * Lock the internal EAL TAILQ list for exclusive access. */ void rte_mcfg_tailq_write_lock(void) - __rte_exclusive_lock_function(rte_mcfg_tailq_get_lock()); + __rte_acquire_capability(rte_mcfg_tailq_get_lock()); /** * Unlock the internal EAL TAILQ list for exclusive access. */ void rte_mcfg_tailq_write_unlock(void) - __rte_unlock_function(rte_mcfg_tailq_get_lock()); + __rte_release_capability(rte_mcfg_tailq_get_lock()); /** * Lock the internal EAL Mempool list for shared access. */ void rte_mcfg_mempool_read_lock(void) - __rte_shared_lock_function(rte_mcfg_mempool_get_lock()); + __rte_acquire_shared_capability(rte_mcfg_mempool_get_lock()); /** * Unlock the internal EAL Mempool list for shared access. */ void rte_mcfg_mempool_read_unlock(void) - __rte_unlock_function(rte_mcfg_mempool_get_lock()); + __rte_release_shared_capability(rte_mcfg_mempool_get_lock()); /** * Lock the internal EAL Mempool list for exclusive access. */ void rte_mcfg_mempool_write_lock(void) - __rte_exclusive_lock_function(rte_mcfg_mempool_get_lock()); + __rte_acquire_capability(rte_mcfg_mempool_get_lock()); /** * Unlock the internal EAL Mempool list for exclusive access. */ void rte_mcfg_mempool_write_unlock(void) - __rte_unlock_function(rte_mcfg_mempool_get_lock()); + __rte_release_capability(rte_mcfg_mempool_get_lock()); /** * Lock the internal EAL Timer Library lock for exclusive access. diff --git a/lib/ethdev/ethdev_driver.c b/lib/ethdev/ethdev_driver.c index 9afef06431..13d328d97b 100644 --- a/lib/ethdev/ethdev_driver.c +++ b/lib/ethdev/ethdev_driver.c @@ -48,7 +48,7 @@ eth_dev_allocated(const char *name) static uint16_t eth_dev_find_free_port(void) - __rte_exclusive_locks_required(rte_mcfg_ethdev_get_lock()) + __rte_requires_capability(rte_mcfg_ethdev_get_lock()) { uint16_t i; @@ -65,7 +65,7 @@ eth_dev_find_free_port(void) static struct rte_eth_dev * eth_dev_get(uint16_t port_id) - __rte_exclusive_locks_required(rte_mcfg_ethdev_get_lock()) + __rte_requires_capability(rte_mcfg_ethdev_get_lock()) { struct rte_eth_dev *eth_dev = &rte_eth_devices[port_id]; diff --git a/lib/ethdev/ethdev_private.h b/lib/ethdev/ethdev_private.h index 0d36b9c30f..b07b1b4c42 100644 --- a/lib/ethdev/ethdev_private.h +++ b/lib/ethdev/ethdev_private.h @@ -70,9 +70,9 @@ void eth_dev_fp_ops_setup(struct rte_eth_fp_ops *fpo, void *eth_dev_shared_data_prepare(void) - __rte_exclusive_locks_required(rte_mcfg_ethdev_get_lock()); + __rte_requires_capability(rte_mcfg_ethdev_get_lock()); void eth_dev_shared_data_release(void) - __rte_exclusive_locks_required(rte_mcfg_ethdev_get_lock()); + __rte_requires_capability(rte_mcfg_ethdev_get_lock()); void eth_dev_rxq_release(struct rte_eth_dev *dev, uint16_t qid); void eth_dev_txq_release(struct rte_eth_dev *dev, uint16_t qid); diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_ethdev.c index 6413c54e3b..355d5c2df1 100644 --- a/lib/ethdev/rte_ethdev.c +++ b/lib/ethdev/rte_ethdev.c @@ -420,7 +420,7 @@ rte_eth_dev_is_valid_port(uint16_t port_id) static int eth_is_valid_owner_id(uint64_t owner_id) - __rte_exclusive_locks_required(rte_mcfg_ethdev_get_lock()) + __rte_requires_capability(rte_mcfg_ethdev_get_lock()) { if (owner_id == RTE_ETH_DEV_NO_OWNER || eth_dev_shared_data->next_owner_id <= owner_id) @@ -471,7 +471,7 @@ rte_eth_dev_owner_new(uint64_t *owner_id) static int eth_dev_owner_set(const uint16_t port_id, const uint64_t old_owner_id, const struct rte_eth_dev_owner *new_owner) - __rte_exclusive_locks_required(rte_mcfg_ethdev_get_lock()) + __rte_requires_capability(rte_mcfg_ethdev_get_lock()) { struct rte_eth_dev *ethdev = &rte_eth_devices[port_id]; struct rte_eth_dev_owner *port_owner; diff --git a/lib/hash/rte_cuckoo_hash.c b/lib/hash/rte_cuckoo_hash.c index 9575e8aa0c..00cbdf89d6 100644 --- a/lib/hash/rte_cuckoo_hash.c +++ b/lib/hash/rte_cuckoo_hash.c @@ -620,7 +620,7 @@ rte_hash_count(const struct rte_hash *h) /* Read write locks implemented using rte_rwlock */ static inline void __hash_rw_writer_lock(const struct rte_hash *h) - __rte_exclusive_lock_function(&h->readwrite_lock) + __rte_acquire_capability(&h->readwrite_lock) __rte_no_thread_safety_analysis { if (h->writer_takes_lock && h->hw_trans_mem_support) @@ -631,7 +631,7 @@ __hash_rw_writer_lock(const struct rte_hash *h) static inline void __hash_rw_reader_lock(const struct rte_hash *h) - __rte_shared_lock_function(&h->readwrite_lock) + __rte_acquire_shared_capability(&h->readwrite_lock) __rte_no_thread_safety_analysis { if (h->readwrite_concur_support && h->hw_trans_mem_support) @@ -642,7 +642,7 @@ __hash_rw_reader_lock(const struct rte_hash *h) static inline void __hash_rw_writer_unlock(const struct rte_hash *h) - __rte_unlock_function(&h->readwrite_lock) + __rte_release_capability(&h->readwrite_lock) __rte_no_thread_safety_analysis { if (h->writer_takes_lock && h->hw_trans_mem_support) @@ -653,7 +653,7 @@ __hash_rw_writer_unlock(const struct rte_hash *h) static inline void __hash_rw_reader_unlock(const struct rte_hash *h) - __rte_unlock_function(&h->readwrite_lock) + __rte_release_shared_capability(&h->readwrite_lock) __rte_no_thread_safety_analysis { if (h->readwrite_concur_support && h->hw_trans_mem_support) diff --git a/lib/vhost/iotlb.h b/lib/vhost/iotlb.h index 81ca04df21..72232b0dcf 100644 --- a/lib/vhost/iotlb.h +++ b/lib/vhost/iotlb.h @@ -11,28 +11,28 @@ static __rte_always_inline void vhost_user_iotlb_rd_lock(struct vhost_virtqueue *vq) - __rte_shared_lock_function(&vq->iotlb_lock) + __rte_acquire_shared_capability(&vq->iotlb_lock) { rte_rwlock_read_lock(&vq->iotlb_lock); } static __rte_always_inline void vhost_user_iotlb_rd_unlock(struct vhost_virtqueue *vq) - __rte_unlock_function(&vq->iotlb_lock) + __rte_release_shared_capability(&vq->iotlb_lock) { rte_rwlock_read_unlock(&vq->iotlb_lock); } static __rte_always_inline void vhost_user_iotlb_wr_lock(struct vhost_virtqueue *vq) - __rte_exclusive_lock_function(&vq->iotlb_lock) + __rte_acquire_capability(&vq->iotlb_lock) { rte_rwlock_write_lock(&vq->iotlb_lock); } static __rte_always_inline void vhost_user_iotlb_wr_unlock(struct vhost_virtqueue *vq) - __rte_unlock_function(&vq->iotlb_lock) + __rte_release_capability(&vq->iotlb_lock) { rte_rwlock_write_unlock(&vq->iotlb_lock); } diff --git a/lib/vhost/vhost.c b/lib/vhost/vhost.c index 5a50a06f8d..2f34d3500e 100644 --- a/lib/vhost/vhost.c +++ b/lib/vhost/vhost.c @@ -390,7 +390,7 @@ cleanup_device(struct virtio_net *dev, int destroy) static void vhost_free_async_mem(struct vhost_virtqueue *vq) - __rte_exclusive_locks_required(&vq->access_lock) + __rte_requires_capability(&vq->access_lock) { if (!vq->async) return; @@ -439,7 +439,7 @@ free_device(struct virtio_net *dev) static __rte_always_inline int log_translate(struct virtio_net *dev, struct vhost_virtqueue *vq) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { if (likely(!(vq->ring_addrs.flags & (1 << VHOST_VRING_F_LOG)))) return 0; @@ -488,7 +488,7 @@ translate_log_addr(struct virtio_net *dev, struct vhost_virtqueue *vq, static int vring_translate_split(struct virtio_net *dev, struct vhost_virtqueue *vq) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint64_t req_size, size; @@ -527,7 +527,7 @@ vring_translate_split(struct virtio_net *dev, struct vhost_virtqueue *vq) static int vring_translate_packed(struct virtio_net *dev, struct vhost_virtqueue *vq) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint64_t req_size, size; @@ -1772,7 +1772,7 @@ rte_vhost_extern_callback_register(int vid, static __rte_always_inline int async_channel_register(struct virtio_net *dev, struct vhost_virtqueue *vq) - __rte_exclusive_locks_required(&vq->access_lock) + __rte_requires_capability(&vq->access_lock) { struct vhost_async *async; int node = vq->numa_node; diff --git a/lib/vhost/vhost.h b/lib/vhost/vhost.h index 1f4192f5d1..0a7815832f 100644 --- a/lib/vhost/vhost.h +++ b/lib/vhost/vhost.h @@ -576,7 +576,7 @@ vhost_virtqueue_reconnect_log_packed(struct vhost_virtqueue *vq) static inline void vq_assert_lock__(struct virtio_net *dev, struct vhost_virtqueue *vq, const char *func) - __rte_assert_exclusive_lock(&vq->access_lock) + __rte_assert_capability(&vq->access_lock) { if (unlikely(!rte_rwlock_write_is_locked(&vq->access_lock))) rte_panic("VHOST_CONFIG: (%s) %s() called without access lock taken.\n", @@ -627,14 +627,14 @@ void __vhost_log_cache_write(struct virtio_net *dev, void __vhost_log_cache_write_iova(struct virtio_net *dev, struct vhost_virtqueue *vq, uint64_t iova, uint64_t len) - __rte_shared_locks_required(&vq->iotlb_lock); + __rte_requires_shared_capability(&vq->iotlb_lock); void __vhost_log_cache_sync(struct virtio_net *dev, struct vhost_virtqueue *vq); void __vhost_log_write(struct virtio_net *dev, uint64_t addr, uint64_t len); void __vhost_log_write_iova(struct virtio_net *dev, struct vhost_virtqueue *vq, uint64_t iova, uint64_t len) - __rte_shared_locks_required(&vq->iotlb_lock); + __rte_requires_shared_capability(&vq->iotlb_lock); static __rte_always_inline void vhost_log_write(struct virtio_net *dev, uint64_t addr, uint64_t len) @@ -684,7 +684,7 @@ vhost_log_used_vring(struct virtio_net *dev, struct vhost_virtqueue *vq, static __rte_always_inline void vhost_log_cache_write_iova(struct virtio_net *dev, struct vhost_virtqueue *vq, uint64_t iova, uint64_t len) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { if (likely(!(dev->features & (1ULL << VHOST_F_LOG_ALL)))) return; @@ -698,7 +698,7 @@ vhost_log_cache_write_iova(struct virtio_net *dev, struct vhost_virtqueue *vq, static __rte_always_inline void vhost_log_write_iova(struct virtio_net *dev, struct vhost_virtqueue *vq, uint64_t iova, uint64_t len) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { if (likely(!(dev->features & (1ULL << VHOST_F_LOG_ALL)))) return; @@ -904,24 +904,24 @@ void vhost_backend_cleanup(struct virtio_net *dev); uint64_t __vhost_iova_to_vva(struct virtio_net *dev, struct vhost_virtqueue *vq, uint64_t iova, uint64_t *len, uint8_t perm) - __rte_shared_locks_required(&vq->iotlb_lock); + __rte_requires_shared_capability(&vq->iotlb_lock); void *vhost_alloc_copy_ind_table(struct virtio_net *dev, struct vhost_virtqueue *vq, uint64_t desc_addr, uint64_t desc_len) - __rte_shared_locks_required(&vq->iotlb_lock); + __rte_requires_shared_capability(&vq->iotlb_lock); int vring_translate(struct virtio_net *dev, struct vhost_virtqueue *vq) - __rte_exclusive_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock); + __rte_requires_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock); uint64_t translate_log_addr(struct virtio_net *dev, struct vhost_virtqueue *vq, uint64_t log_addr) - __rte_shared_locks_required(&vq->iotlb_lock); + __rte_requires_shared_capability(&vq->iotlb_lock); void vring_invalidate(struct virtio_net *dev, struct vhost_virtqueue *vq) - __rte_exclusive_locks_required(&vq->access_lock); + __rte_requires_capability(&vq->access_lock); static __rte_always_inline uint64_t vhost_iova_to_vva(struct virtio_net *dev, struct vhost_virtqueue *vq, uint64_t iova, uint64_t *len, uint8_t perm) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { if (!(dev->features & (1ULL << VIRTIO_F_IOMMU_PLATFORM))) return rte_vhost_va_from_guest_pa(dev->mem, iova, len); diff --git a/lib/vhost/vhost_crypto.c b/lib/vhost/vhost_crypto.c index 7caf6d9afa..3dc41a3bd5 100644 --- a/lib/vhost/vhost_crypto.c +++ b/lib/vhost/vhost_crypto.c @@ -491,7 +491,7 @@ static __rte_always_inline struct virtio_crypto_inhdr * reach_inhdr(struct vhost_crypto_data_req *vc_req, struct vhost_crypto_desc *head, uint32_t max_n_descs) - __rte_shared_locks_required(&vc_req->vq->iotlb_lock) + __rte_requires_shared_capability(&vc_req->vq->iotlb_lock) { struct virtio_crypto_inhdr *inhdr; struct vhost_crypto_desc *last = head + (max_n_descs - 1); @@ -538,7 +538,7 @@ static __rte_always_inline void * get_data_ptr(struct vhost_crypto_data_req *vc_req, struct vhost_crypto_desc *cur_desc, uint8_t perm) - __rte_shared_locks_required(&vc_req->vq->iotlb_lock) + __rte_requires_shared_capability(&vc_req->vq->iotlb_lock) { void *data; uint64_t dlen = cur_desc->len; @@ -555,7 +555,7 @@ get_data_ptr(struct vhost_crypto_data_req *vc_req, static __rte_always_inline uint32_t copy_data_from_desc(void *dst, struct vhost_crypto_data_req *vc_req, struct vhost_crypto_desc *desc, uint32_t size) - __rte_shared_locks_required(&vc_req->vq->iotlb_lock) + __rte_requires_shared_capability(&vc_req->vq->iotlb_lock) { uint64_t remain; uint64_t addr; @@ -586,7 +586,7 @@ static __rte_always_inline int copy_data(void *data, struct vhost_crypto_data_req *vc_req, struct vhost_crypto_desc *head, struct vhost_crypto_desc **cur_desc, uint32_t size, uint32_t max_n_descs) - __rte_shared_locks_required(&vc_req->vq->iotlb_lock) + __rte_requires_shared_capability(&vc_req->vq->iotlb_lock) { struct vhost_crypto_desc *desc = *cur_desc; uint32_t left = size; @@ -670,7 +670,7 @@ prepare_write_back_data(struct vhost_crypto_data_req *vc_req, uint32_t offset, uint64_t write_back_len, uint32_t max_n_descs) - __rte_shared_locks_required(&vc_req->vq->iotlb_lock) + __rte_requires_shared_capability(&vc_req->vq->iotlb_lock) { struct vhost_crypto_writeback_data *wb_data, *head; struct vhost_crypto_desc *desc = *cur_desc; @@ -791,7 +791,7 @@ prepare_sym_cipher_op(struct vhost_crypto *vcrypto, struct rte_crypto_op *op, struct virtio_crypto_cipher_data_req *cipher, struct vhost_crypto_desc *head, uint32_t max_n_descs) - __rte_shared_locks_required(&vc_req->vq->iotlb_lock) + __rte_requires_shared_capability(&vc_req->vq->iotlb_lock) { struct vhost_crypto_desc *desc = head; struct vhost_crypto_writeback_data *ewb = NULL; @@ -945,7 +945,7 @@ prepare_sym_chain_op(struct vhost_crypto *vcrypto, struct rte_crypto_op *op, struct virtio_crypto_alg_chain_data_req *chain, struct vhost_crypto_desc *head, uint32_t max_n_descs) - __rte_shared_locks_required(&vc_req->vq->iotlb_lock) + __rte_requires_shared_capability(&vc_req->vq->iotlb_lock) { struct vhost_crypto_desc *desc = head, *digest_desc; struct vhost_crypto_writeback_data *ewb = NULL, *ewb2 = NULL; diff --git a/lib/vhost/virtio_net.c b/lib/vhost/virtio_net.c index d764d4bc6a..00be575e07 100644 --- a/lib/vhost/virtio_net.c +++ b/lib/vhost/virtio_net.c @@ -55,7 +55,7 @@ is_valid_virt_queue_idx(uint32_t idx, int is_tx, uint32_t nr_vring) static inline void vhost_queue_stats_update(const struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts, uint16_t count) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { struct virtqueue_stats *stats = &vq->stats; int i; @@ -90,7 +90,7 @@ static __rte_always_inline int64_t vhost_async_dma_transfer_one(struct virtio_net *dev, struct vhost_virtqueue *vq, int16_t dma_id, uint16_t vchan_id, uint16_t flag_idx, struct vhost_iov_iter *pkt) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { struct async_dma_vchan_info *dma_info = &dma_copy_track[dma_id].vchans[vchan_id]; uint16_t ring_mask = dma_info->ring_mask; @@ -140,7 +140,7 @@ static __rte_always_inline uint16_t vhost_async_dma_transfer(struct virtio_net *dev, struct vhost_virtqueue *vq, int16_t dma_id, uint16_t vchan_id, uint16_t head_idx, struct vhost_iov_iter *pkts, uint16_t nr_pkts) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { struct async_dma_vchan_info *dma_info = &dma_copy_track[dma_id].vchans[vchan_id]; int64_t ret, nr_copies = 0; @@ -221,7 +221,7 @@ vhost_async_dma_check_completed(struct virtio_net *dev, int16_t dma_id, uint16_t static inline void do_data_copy_enqueue(struct virtio_net *dev, struct vhost_virtqueue *vq) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { struct batch_copy_elem *elem = vq->batch_copy_elems; uint16_t count = vq->batch_copy_nb_elems; @@ -426,7 +426,7 @@ static __rte_always_inline void vhost_async_shadow_enqueue_packed_batch(struct vhost_virtqueue *vq, uint64_t *lens, uint16_t *ids) - __rte_exclusive_locks_required(&vq->access_lock) + __rte_requires_capability(&vq->access_lock) { uint16_t i; struct vhost_async *async = vq->async; @@ -443,7 +443,7 @@ vhost_async_shadow_enqueue_packed_batch(struct vhost_virtqueue *vq, static __rte_always_inline void vhost_async_shadow_dequeue_packed_batch(struct vhost_virtqueue *vq, uint16_t *ids) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { uint16_t i; struct vhost_async *async = vq->async; @@ -605,7 +605,7 @@ vhost_async_shadow_enqueue_packed(struct vhost_virtqueue *vq, uint16_t *id, uint16_t *count, uint16_t num_buffers) - __rte_exclusive_locks_required(&vq->access_lock) + __rte_requires_capability(&vq->access_lock) { uint16_t i; struct vhost_async *async = vq->async; @@ -627,7 +627,7 @@ vhost_shadow_enqueue_single_packed(struct virtio_net *dev, uint16_t *id, uint16_t *count, uint16_t num_buffers) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { vhost_shadow_enqueue_packed(vq, len, id, count, num_buffers); @@ -719,7 +719,7 @@ static __rte_always_inline int map_one_desc(struct virtio_net *dev, struct vhost_virtqueue *vq, struct buf_vector *buf_vec, uint16_t *vec_idx, uint64_t desc_iova, uint64_t desc_len, uint8_t perm) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t vec_id = *vec_idx; @@ -757,7 +757,7 @@ fill_vec_buf_split(struct virtio_net *dev, struct vhost_virtqueue *vq, uint32_t avail_idx, uint16_t *vec_idx, struct buf_vector *buf_vec, uint16_t *desc_chain_head, uint32_t *desc_chain_len, uint8_t perm) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t idx = vq->avail->ring[avail_idx & (vq->size - 1)]; uint16_t vec_id = *vec_idx; @@ -841,7 +841,7 @@ reserve_avail_buf_split(struct virtio_net *dev, struct vhost_virtqueue *vq, uint64_t size, struct buf_vector *buf_vec, uint16_t *num_buffers, uint16_t avail_head, uint16_t *nr_vec) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t cur_idx; uint16_t vec_idx = 0; @@ -892,7 +892,7 @@ fill_vec_buf_packed_indirect(struct virtio_net *dev, struct vhost_virtqueue *vq, struct vring_packed_desc *desc, uint16_t *vec_idx, struct buf_vector *buf_vec, uint32_t *len, uint8_t perm) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t i; uint32_t nr_descs; @@ -951,7 +951,7 @@ fill_vec_buf_packed(struct virtio_net *dev, struct vhost_virtqueue *vq, uint16_t avail_idx, uint16_t *desc_count, struct buf_vector *buf_vec, uint16_t *vec_idx, uint16_t *buf_id, uint32_t *len, uint8_t perm) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { bool wrap_counter = vq->avail_wrap_counter; struct vring_packed_desc *descs = vq->desc_packed; @@ -1017,7 +1017,7 @@ static __rte_noinline void copy_vnet_hdr_to_desc(struct virtio_net *dev, struct vhost_virtqueue *vq, struct buf_vector *buf_vec, struct virtio_net_hdr_mrg_rxbuf *hdr) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint64_t len; uint64_t remain = dev->vhost_hlen; @@ -1120,8 +1120,8 @@ static __rte_always_inline int async_fill_seg(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf *m, uint32_t mbuf_offset, uint64_t buf_iova, uint32_t cpy_len, bool to_desc) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { struct vhost_async *async = vq->async; uint64_t mapped_len; @@ -1162,7 +1162,7 @@ static __rte_always_inline void sync_fill_seg(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf *m, uint32_t mbuf_offset, uint64_t buf_addr, uint64_t buf_iova, uint32_t cpy_len, bool to_desc) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { struct batch_copy_elem *batch_copy = vq->batch_copy_elems; @@ -1200,8 +1200,8 @@ static __rte_always_inline int mbuf_to_desc(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf *m, struct buf_vector *buf_vec, uint16_t nr_vec, uint16_t num_buffers, bool is_async) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint32_t vec_idx = 0; uint32_t mbuf_offset, mbuf_avail; @@ -1330,8 +1330,8 @@ vhost_enqueue_single_packed(struct virtio_net *dev, struct rte_mbuf *pkt, struct buf_vector *buf_vec, uint16_t *nr_descs) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t nr_vec = 0; uint16_t avail_idx = vq->last_avail_idx; @@ -1392,8 +1392,8 @@ vhost_enqueue_single_packed(struct virtio_net *dev, static __rte_noinline uint32_t virtio_dev_rx_split(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts, uint32_t count) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint32_t pkt_idx = 0; uint16_t num_buffers; @@ -1452,7 +1452,7 @@ virtio_dev_rx_sync_batch_check(struct virtio_net *dev, struct rte_mbuf **pkts, uint64_t *desc_addrs, uint64_t *lens) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { bool wrap_counter = vq->avail_wrap_counter; struct vring_packed_desc *descs = vq->desc_packed; @@ -1556,7 +1556,7 @@ virtio_dev_rx_batch_packed_copy(struct virtio_net *dev, struct rte_mbuf **pkts, uint64_t *desc_addrs, uint64_t *lens) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint32_t buf_offset = sizeof(struct virtio_net_hdr_mrg_rxbuf); struct virtio_net_hdr_mrg_rxbuf *hdrs[PACKED_BATCH_SIZE]; @@ -1604,7 +1604,7 @@ static __rte_always_inline int virtio_dev_rx_sync_batch_packed(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint64_t desc_addrs[PACKED_BATCH_SIZE]; uint64_t lens[PACKED_BATCH_SIZE]; @@ -1626,8 +1626,8 @@ static __rte_always_inline int16_t virtio_dev_rx_single_packed(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf *pkt) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { struct buf_vector buf_vec[BUF_VECTOR_MAX]; uint16_t nr_descs = 0; @@ -1652,8 +1652,8 @@ virtio_dev_rx_packed(struct virtio_net *dev, struct vhost_virtqueue *__rte_restrict vq, struct rte_mbuf **__rte_restrict pkts, uint32_t count) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint32_t pkt_idx = 0; @@ -1767,7 +1767,7 @@ rte_vhost_enqueue_burst(int vid, uint16_t queue_id, static __rte_always_inline uint16_t async_get_first_inflight_pkt_idx(struct vhost_virtqueue *vq) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { struct vhost_async *async = vq->async; @@ -1796,8 +1796,8 @@ store_dma_desc_info_split(struct vring_used_elem *s_ring, struct vring_used_elem static __rte_noinline uint32_t virtio_dev_rx_async_submit_split(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts, uint32_t count, int16_t dma_id, uint16_t vchan_id) - __rte_exclusive_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { struct buf_vector buf_vec[BUF_VECTOR_MAX]; uint32_t pkt_idx = 0; @@ -1907,8 +1907,8 @@ vhost_enqueue_async_packed(struct virtio_net *dev, struct buf_vector *buf_vec, uint16_t *nr_descs, uint16_t *nr_buffers) - __rte_exclusive_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t nr_vec = 0; uint16_t avail_idx = vq->last_avail_idx; @@ -1967,8 +1967,8 @@ vhost_enqueue_async_packed(struct virtio_net *dev, static __rte_always_inline int16_t virtio_dev_rx_async_packed(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf *pkt, uint16_t *nr_descs, uint16_t *nr_buffers) - __rte_exclusive_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { struct buf_vector buf_vec[BUF_VECTOR_MAX]; @@ -1991,8 +1991,8 @@ virtio_dev_rx_async_packed_batch_enqueue(struct virtio_net *dev, struct rte_mbuf **pkts, uint64_t *desc_addrs, uint64_t *lens) - __rte_exclusive_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint32_t buf_offset = sizeof(struct virtio_net_hdr_mrg_rxbuf); struct virtio_net_hdr_mrg_rxbuf *hdrs[PACKED_BATCH_SIZE]; @@ -2053,8 +2053,8 @@ virtio_dev_rx_async_packed_batch(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts, int16_t dma_id, uint16_t vchan_id) - __rte_exclusive_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint64_t desc_addrs[PACKED_BATCH_SIZE]; uint64_t lens[PACKED_BATCH_SIZE]; @@ -2070,7 +2070,7 @@ virtio_dev_rx_async_packed_batch(struct virtio_net *dev, static __rte_always_inline void dma_error_handler_packed(struct vhost_virtqueue *vq, uint16_t slot_idx, uint32_t nr_err, uint32_t *pkt_idx) - __rte_exclusive_locks_required(&vq->access_lock) + __rte_requires_capability(&vq->access_lock) { uint16_t descs_err = 0; uint16_t buffers_err = 0; @@ -2102,8 +2102,8 @@ dma_error_handler_packed(struct vhost_virtqueue *vq, uint16_t slot_idx, static __rte_noinline uint32_t virtio_dev_rx_async_submit_packed(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts, uint32_t count, int16_t dma_id, uint16_t vchan_id) - __rte_exclusive_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint32_t pkt_idx = 0; uint16_t n_xfer; @@ -2176,7 +2176,7 @@ virtio_dev_rx_async_submit_packed(struct virtio_net *dev, struct vhost_virtqueue static __rte_always_inline void write_back_completed_descs_split(struct vhost_virtqueue *vq, uint16_t n_descs) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { struct vhost_async *async = vq->async; uint16_t nr_left = n_descs; @@ -2209,7 +2209,7 @@ write_back_completed_descs_split(struct vhost_virtqueue *vq, uint16_t n_descs) static __rte_always_inline void write_back_completed_descs_packed(struct vhost_virtqueue *vq, uint16_t n_buffers) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { struct vhost_async *async = vq->async; uint16_t from = async->last_buffer_idx_packed; @@ -2274,7 +2274,7 @@ write_back_completed_descs_packed(struct vhost_virtqueue *vq, static __rte_always_inline uint16_t vhost_poll_enqueue_completed(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts, uint16_t count, int16_t dma_id, uint16_t vchan_id) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { struct vhost_async *async = vq->async; struct async_inflight_info *pkts_info = async->pkts_info; @@ -2884,8 +2884,8 @@ desc_to_mbuf(struct virtio_net *dev, struct vhost_virtqueue *vq, struct buf_vector *buf_vec, uint16_t nr_vec, struct rte_mbuf *m, struct rte_mempool *mbuf_pool, bool legacy_ol_flags, uint16_t slot_idx, bool is_async) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint32_t buf_avail, buf_offset, buf_len; uint64_t buf_addr, buf_iova; @@ -3092,8 +3092,8 @@ static uint16_t virtio_dev_tx_split(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count, bool legacy_ol_flags) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t i; uint16_t avail_entries; @@ -3188,8 +3188,8 @@ static uint16_t virtio_dev_tx_split_legacy(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { return virtio_dev_tx_split(dev, vq, mbuf_pool, pkts, count, true); } @@ -3199,8 +3199,8 @@ static uint16_t virtio_dev_tx_split_compliant(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { return virtio_dev_tx_split(dev, vq, mbuf_pool, pkts, count, false); } @@ -3212,7 +3212,7 @@ vhost_reserve_avail_batch_packed(struct virtio_net *dev, uint16_t avail_idx, uintptr_t *desc_addrs, uint16_t *ids) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { bool wrap = vq->avail_wrap_counter; struct vring_packed_desc *descs = vq->desc_packed; @@ -3356,7 +3356,7 @@ virtio_dev_tx_batch_packed(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts, bool legacy_ol_flags) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t avail_idx = vq->last_avail_idx; uint32_t buf_offset = sizeof(struct virtio_net_hdr_mrg_rxbuf); @@ -3403,8 +3403,8 @@ vhost_dequeue_single_packed(struct virtio_net *dev, uint16_t *buf_id, uint16_t *desc_count, bool legacy_ol_flags) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { struct buf_vector buf_vec[BUF_VECTOR_MAX]; uint32_t buf_len; @@ -3453,8 +3453,8 @@ virtio_dev_tx_single_packed(struct virtio_net *dev, struct rte_mempool *mbuf_pool, struct rte_mbuf *pkts, bool legacy_ol_flags) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t buf_id, desc_count = 0; @@ -3514,8 +3514,8 @@ virtio_dev_tx_packed(struct virtio_net *dev, struct rte_mbuf **__rte_restrict pkts, uint32_t count, bool legacy_ol_flags) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint32_t pkt_idx = 0; @@ -3565,8 +3565,8 @@ static uint16_t virtio_dev_tx_packed_legacy(struct virtio_net *dev, struct vhost_virtqueue *__rte_restrict vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **__rte_restrict pkts, uint32_t count) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { return virtio_dev_tx_packed(dev, vq, mbuf_pool, pkts, count, true); } @@ -3576,8 +3576,8 @@ static uint16_t virtio_dev_tx_packed_compliant(struct virtio_net *dev, struct vhost_virtqueue *__rte_restrict vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **__rte_restrict pkts, uint32_t count) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { return virtio_dev_tx_packed(dev, vq, mbuf_pool, pkts, count, false); } @@ -3696,7 +3696,7 @@ static __rte_always_inline uint16_t async_poll_dequeue_completed(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts, uint16_t count, int16_t dma_id, uint16_t vchan_id, bool legacy_ol_flags) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { uint16_t start_idx, from, i; uint16_t nr_cpl_pkts = 0; @@ -3744,8 +3744,8 @@ static __rte_always_inline uint16_t virtio_dev_tx_async_split(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count, int16_t dma_id, uint16_t vchan_id, bool legacy_ol_flags) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { static bool allocerr_warned; bool dropped = false; @@ -3897,8 +3897,8 @@ virtio_dev_tx_async_split_legacy(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count, int16_t dma_id, uint16_t vchan_id) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { return virtio_dev_tx_async_split(dev, vq, mbuf_pool, pkts, count, dma_id, vchan_id, true); @@ -3910,8 +3910,8 @@ virtio_dev_tx_async_split_compliant(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count, int16_t dma_id, uint16_t vchan_id) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { return virtio_dev_tx_async_split(dev, vq, mbuf_pool, pkts, count, dma_id, vchan_id, false); @@ -3920,7 +3920,7 @@ virtio_dev_tx_async_split_compliant(struct virtio_net *dev, static __rte_always_inline void vhost_async_shadow_dequeue_single_packed(struct vhost_virtqueue *vq, uint16_t buf_id, uint16_t count) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { struct vhost_async *async = vq->async; uint16_t idx = async->buffer_idx_packed; @@ -3942,8 +3942,8 @@ virtio_dev_tx_async_single_packed(struct virtio_net *dev, struct rte_mbuf *pkts, uint16_t slot_idx, bool legacy_ol_flags) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { int err; uint16_t buf_id, desc_count = 0; @@ -3995,8 +3995,8 @@ virtio_dev_tx_async_packed_batch(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts, uint16_t slot_idx, uint16_t dma_id, uint16_t vchan_id) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t avail_idx = vq->last_avail_idx; uint32_t buf_offset = sizeof(struct virtio_net_hdr_mrg_rxbuf); @@ -4053,8 +4053,8 @@ static __rte_always_inline uint16_t virtio_dev_tx_async_packed(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count, uint16_t dma_id, uint16_t vchan_id, bool legacy_ol_flags) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint32_t pkt_idx = 0; uint16_t slot_idx = 0; @@ -4167,8 +4167,8 @@ static uint16_t virtio_dev_tx_async_packed_legacy(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count, uint16_t dma_id, uint16_t vchan_id) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { return virtio_dev_tx_async_packed(dev, vq, mbuf_pool, pkts, count, dma_id, vchan_id, true); @@ -4179,8 +4179,8 @@ static uint16_t virtio_dev_tx_async_packed_compliant(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count, uint16_t dma_id, uint16_t vchan_id) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { return virtio_dev_tx_async_packed(dev, vq, mbuf_pool, pkts, count, dma_id, vchan_id, false); diff --git a/lib/vhost/virtio_net_ctrl.c b/lib/vhost/virtio_net_ctrl.c index b8ee94018e..999e84db7c 100644 --- a/lib/vhost/virtio_net_ctrl.c +++ b/lib/vhost/virtio_net_ctrl.c @@ -26,7 +26,7 @@ struct virtio_net_ctrl_elem { static int virtio_net_ctrl_pop(struct virtio_net *dev, struct vhost_virtqueue *cvq, struct virtio_net_ctrl_elem *ctrl_elem) - __rte_shared_locks_required(&cvq->iotlb_lock) + __rte_requires_shared_capability(&cvq->iotlb_lock) { uint16_t avail_idx, desc_idx, n_descs = 0; uint64_t desc_len, desc_addr, desc_iova, data_len = 0; -- 2.47.0 ^ permalink raw reply [flat|nested] 10+ messages in thread
* [PATCH v2 0/3] Improve lock annotations 2024-12-02 12:53 [RFC 0/3] Improve lock annotations David Marchand ` (2 preceding siblings ...) 2024-12-02 12:53 ` [RFC 3/3] eal: enhance lock annotations for rwlock David Marchand @ 2024-12-12 16:00 ` David Marchand 2024-12-12 16:00 ` [PATCH v2 1/3] eal: add enhanced " David Marchand ` (2 more replies) 3 siblings, 3 replies; 10+ messages in thread From: David Marchand @ 2024-12-12 16:00 UTC (permalink / raw) To: dev; +Cc: thomas A recent bug (see 22aa9a9c7099 ("vhost: fix deadlock in Rx async path")) made more visible a gap in the clang thread safety annotations that DPDK uses: no distinction is made between releasing a read lock and releasing a write lock. Clang 3.6 and later offers improved thread safety checks. Marking objects as "lockable" has evolved into flagging some named "capability". clang reports the capability name when an error is reported (making this report a bit easier to understand). For example, a spinlock is now flagged as: typedef struct __rte_capability("spinlock") { volatile RTE_ATOMIC(int) locked; } rte_spinlock_t; For "exclusive" locking (spinlocks / write locks), the conversion is: - exclusive_lock_function -> acquire_capability - exclusive_trylock_function -> try_acquire_capability - unlock_function -> release_capability ... For "shared" locking (read locks): - shared_lock_function -> acquire_shared_capability - shared_trylock_function -> try_acquire_shared_capability - unlock_function -> release_shared_capability ... This series proposes to use those annotations (sticking to the convention of simply prefixing the compiler attributes with __rte_). The existing "old" annotations macros are left in place in case users started to rely on them. Note: DPDK requirements state that clang version must be >= 3.6 (following use of C11 standard). Comments welcome. -- David Marchand David Marchand (3): eal: add enhanced lock annotations eal: enhance lock annotations for spinlock and seqlock eal: enhance lock annotations for rwlock doc/api/doxy-api.conf.in | 12 ++ drivers/bus/dpaa/base/qbman/qman.c | 4 +- drivers/net/fm10k/fm10k_ethdev.c | 4 +- lib/eal/common/eal_memalloc.h | 2 +- lib/eal/common/eal_private.h | 2 +- lib/eal/include/generic/rte_rwlock.h | 23 ++-- lib/eal/include/generic/rte_spinlock.h | 14 +- lib/eal/include/rte_eal_memconfig.h | 28 ++-- lib/eal/include/rte_lock_annotations.h | 53 ++++++++ lib/eal/include/rte_seqlock.h | 4 +- lib/ethdev/ethdev_driver.c | 4 +- lib/ethdev/ethdev_private.h | 4 +- lib/ethdev/rte_ethdev.c | 4 +- lib/graph/graph_private.h | 4 +- lib/hash/rte_cuckoo_hash.c | 8 +- lib/vhost/iotlb.h | 8 +- lib/vhost/vdpa.c | 2 +- lib/vhost/vhost.c | 10 +- lib/vhost/vhost.h | 24 ++-- lib/vhost/vhost_crypto.c | 14 +- lib/vhost/virtio_net.c | 170 ++++++++++++------------- lib/vhost/virtio_net_ctrl.c | 2 +- 22 files changed, 233 insertions(+), 167 deletions(-) -- 2.47.0 ^ permalink raw reply [flat|nested] 10+ messages in thread
* [PATCH v2 1/3] eal: add enhanced lock annotations 2024-12-12 16:00 ` [PATCH v2 0/3] Improve lock annotations David Marchand @ 2024-12-12 16:00 ` David Marchand 2024-12-12 16:00 ` [PATCH v2 2/3] eal: enhance lock annotations for spinlock and seqlock David Marchand 2024-12-12 16:00 ` [PATCH v2 3/3] eal: enhance lock annotations for rwlock David Marchand 2 siblings, 0 replies; 10+ messages in thread From: David Marchand @ 2024-12-12 16:00 UTC (permalink / raw) To: dev; +Cc: thomas, Tyler Retzlaff Clang 3.6+ offers enhanced lock annotations when it comes to shared vs exclusive capability/lock release. Introduce macros for those new function attributes. Signed-off-by: David Marchand <david.marchand@redhat.com> --- Changes since RFC: - fixed build with clang < 11, --- doc/api/doxy-api.conf.in | 12 ++++++ lib/eal/include/rte_lock_annotations.h | 53 ++++++++++++++++++++++++++ 2 files changed, 65 insertions(+) diff --git a/doc/api/doxy-api.conf.in b/doc/api/doxy-api.conf.in index d23352d300..ac490e7631 100644 --- a/doc/api/doxy-api.conf.in +++ b/doc/api/doxy-api.conf.in @@ -103,6 +103,18 @@ PREDEFINED = __DOXYGEN__ \ __rte_shared_trylock_function(x)= \ __rte_assert_shared_lock(x)= \ __rte_unlock_function(x)= \ + __rte_capability(x)= \ + __rte_requires_capability(x)= \ + __rte_acquire_capability(x)= \ + __rte_try_acquire_capability(x)= \ + __rte_release_capability(x)= \ + __rte_assert_capability(x)= \ + __rte_requires_shared_capability(x)= \ + __rte_acquire_shared_capability(x)= \ + __rte_try_acquire_shared_capability(x)= \ + __rte_release_shared_capability(x)= \ + __rte_assert_shared_capability(x)= \ + __rte_exclude_capability(x)= \ __attribute__(x)= OPTIMIZE_OUTPUT_FOR_C = YES diff --git a/lib/eal/include/rte_lock_annotations.h b/lib/eal/include/rte_lock_annotations.h index 2456a69352..102eda9c51 100644 --- a/lib/eal/include/rte_lock_annotations.h +++ b/lib/eal/include/rte_lock_annotations.h @@ -43,6 +43,41 @@ extern "C" { #define __rte_locks_excluded(...) \ __attribute__((locks_excluded(__VA_ARGS__))) + +#if defined(__clang__) && __clang_major__ >= 11 +#define __rte_capability(...) \ + __attribute__((capability(__VA_ARGS__))) +#else +#define __rte_capability(...) \ + __attribute__((capability("mutex"))) +#endif + +#define __rte_requires_capability(...) \ + __attribute__((requires_capability(__VA_ARGS__))) +#define __rte_acquire_capability(...) \ + __attribute__((acquire_capability(__VA_ARGS__))) +#define __rte_try_acquire_capability(ret, ...) \ + __attribute__((try_acquire_capability(ret, __VA_ARGS__))) +#define __rte_release_capability(...) \ + __attribute__((release_capability(__VA_ARGS__))) +#define __rte_assert_capability(...) \ + __attribute__((assert_capability(__VA_ARGS__))) + +#define __rte_requires_shared_capability(...) \ + __attribute__((requires_shared_capability(__VA_ARGS__))) +#define __rte_acquire_shared_capability(...) \ + __attribute__((acquire_shared_capability(__VA_ARGS__))) +#define __rte_try_acquire_shared_capability(ret, ...) \ + __attribute__((try_acquire_shared_capability(ret, __VA_ARGS__))) +#define __rte_release_shared_capability(...) \ + __attribute__((release_shared_capability(__VA_ARGS__))) +#define __rte_assert_shared_capability(...) \ + __attribute__((assert_shared_capability(__VA_ARGS__))) + +#define __rte_exclude_capability(...) \ + __attribute__((exclude_capability(__VA_ARGS__))) + + #define __rte_no_thread_safety_analysis \ __attribute__((no_thread_safety_analysis)) @@ -67,6 +102,24 @@ extern "C" { #define __rte_locks_excluded(...) + +#define __rte_capability(...) + +#define __rte_requires_capability(...) +#define __rte_acquire_capability(...) +#define __rte_try_acquire_capability(...) +#define __rte_release_capability(...) +#define __rte_assert_capability(...) + +#define __rte_requires_shared_capability(...) +#define __rte_acquire_shared_capability(...) +#define __rte_try_acquire_shared_capability(...) +#define __rte_release_shared_capability(...) +#define __rte_assert_shared_capability(...) + +#define __rte_exclude_capability(...) + + #define __rte_no_thread_safety_analysis #endif /* RTE_ANNOTATE_LOCKS */ -- 2.47.0 ^ permalink raw reply [flat|nested] 10+ messages in thread
* [PATCH v2 2/3] eal: enhance lock annotations for spinlock and seqlock 2024-12-12 16:00 ` [PATCH v2 0/3] Improve lock annotations David Marchand 2024-12-12 16:00 ` [PATCH v2 1/3] eal: add enhanced " David Marchand @ 2024-12-12 16:00 ` David Marchand 2024-12-12 16:00 ` [PATCH v2 3/3] eal: enhance lock annotations for rwlock David Marchand 2 siblings, 0 replies; 10+ messages in thread From: David Marchand @ 2024-12-12 16:00 UTC (permalink / raw) To: dev Cc: thomas, Mattias Rönnnblom, Hemant Agrawal, Sachin Saxena, Tyler Retzlaff, Jerin Jacob, Kiran Kumar K, Nithin Dabilpuram, Zhirun Yan, Maxime Coquelin, Chenbo Xia Convert spinlock (and as a consequence seqlock) to the clang capability annotations. Signed-off-by: David Marchand <david.marchand@redhat.com> Acked-by: Mattias Rönnnblom <mattias.ronnblom@ericsson.com> --- drivers/bus/dpaa/base/qbman/qman.c | 4 ++-- drivers/net/fm10k/fm10k_ethdev.c | 4 ++-- lib/eal/include/generic/rte_spinlock.h | 14 +++++++------- lib/eal/include/rte_eal_memconfig.h | 4 ++-- lib/eal/include/rte_seqlock.h | 4 ++-- lib/graph/graph_private.h | 4 ++-- lib/vhost/vdpa.c | 2 +- 7 files changed, 18 insertions(+), 18 deletions(-) diff --git a/drivers/bus/dpaa/base/qbman/qman.c b/drivers/bus/dpaa/base/qbman/qman.c index f92b25343a..11fabcaff5 100644 --- a/drivers/bus/dpaa/base/qbman/qman.c +++ b/drivers/bus/dpaa/base/qbman/qman.c @@ -45,7 +45,7 @@ static inline int fq_isset(struct qman_fq *fq, u32 mask) } static inline void fq_lock(struct qman_fq *fq) - __rte_exclusive_lock_function(&fq->fqlock) + __rte_acquire_capability(&fq->fqlock) __rte_no_thread_safety_analysis { if (fq_isset(fq, QMAN_FQ_FLAG_LOCKED)) @@ -53,7 +53,7 @@ static inline void fq_lock(struct qman_fq *fq) } static inline void fq_unlock(struct qman_fq *fq) - __rte_unlock_function(&fq->fqlock) + __rte_release_capability(&fq->fqlock) __rte_no_thread_safety_analysis { if (fq_isset(fq, QMAN_FQ_FLAG_LOCKED)) diff --git a/drivers/net/fm10k/fm10k_ethdev.c b/drivers/net/fm10k/fm10k_ethdev.c index 7b490bea17..747042d621 100644 --- a/drivers/net/fm10k/fm10k_ethdev.c +++ b/drivers/net/fm10k/fm10k_ethdev.c @@ -116,7 +116,7 @@ fm10k_mbx_initlock(struct fm10k_hw *hw) static void fm10k_mbx_lock(struct fm10k_hw *hw) - __rte_exclusive_lock_function(FM10K_DEV_PRIVATE_TO_MBXLOCK(hw->back)) + __rte_acquire_capability(FM10K_DEV_PRIVATE_TO_MBXLOCK(hw->back)) { while (!rte_spinlock_trylock(FM10K_DEV_PRIVATE_TO_MBXLOCK(hw->back))) rte_delay_us(FM10K_MBXLOCK_DELAY_US); @@ -124,7 +124,7 @@ fm10k_mbx_lock(struct fm10k_hw *hw) static void fm10k_mbx_unlock(struct fm10k_hw *hw) - __rte_unlock_function(FM10K_DEV_PRIVATE_TO_MBXLOCK(hw->back)) + __rte_release_capability(FM10K_DEV_PRIVATE_TO_MBXLOCK(hw->back)) { rte_spinlock_unlock(FM10K_DEV_PRIVATE_TO_MBXLOCK(hw->back)); } diff --git a/lib/eal/include/generic/rte_spinlock.h b/lib/eal/include/generic/rte_spinlock.h index c2980601b2..c907d4e45c 100644 --- a/lib/eal/include/generic/rte_spinlock.h +++ b/lib/eal/include/generic/rte_spinlock.h @@ -32,7 +32,7 @@ extern "C" { /** * The rte_spinlock_t type. */ -typedef struct __rte_lockable { +typedef struct __rte_capability("spinlock") { volatile RTE_ATOMIC(int) locked; /**< lock status 0 = unlocked, 1 = locked */ } rte_spinlock_t; @@ -61,7 +61,7 @@ rte_spinlock_init(rte_spinlock_t *sl) */ static inline void rte_spinlock_lock(rte_spinlock_t *sl) - __rte_exclusive_lock_function(sl); + __rte_acquire_capability(sl); #ifdef RTE_FORCE_INTRINSICS static inline void @@ -87,7 +87,7 @@ rte_spinlock_lock(rte_spinlock_t *sl) */ static inline void rte_spinlock_unlock(rte_spinlock_t *sl) - __rte_unlock_function(sl); + __rte_release_capability(sl); #ifdef RTE_FORCE_INTRINSICS static inline void @@ -109,7 +109,7 @@ rte_spinlock_unlock(rte_spinlock_t *sl) __rte_warn_unused_result static inline int rte_spinlock_trylock(rte_spinlock_t *sl) - __rte_exclusive_trylock_function(1, sl); + __rte_try_acquire_capability(true, sl); #ifdef RTE_FORCE_INTRINSICS static inline int @@ -158,7 +158,7 @@ static inline int rte_tm_supported(void); */ static inline void rte_spinlock_lock_tm(rte_spinlock_t *sl) - __rte_exclusive_lock_function(sl); + __rte_acquire_capability(sl); /** * Commit hardware memory transaction or release the spinlock if @@ -169,7 +169,7 @@ rte_spinlock_lock_tm(rte_spinlock_t *sl) */ static inline void rte_spinlock_unlock_tm(rte_spinlock_t *sl) - __rte_unlock_function(sl); + __rte_release_capability(sl); /** * Try to execute critical section in a hardware memory transaction, @@ -190,7 +190,7 @@ rte_spinlock_unlock_tm(rte_spinlock_t *sl) __rte_warn_unused_result static inline int rte_spinlock_trylock_tm(rte_spinlock_t *sl) - __rte_exclusive_trylock_function(1, sl); + __rte_try_acquire_capability(true, sl); /** * The rte_spinlock_recursive_t type. diff --git a/lib/eal/include/rte_eal_memconfig.h b/lib/eal/include/rte_eal_memconfig.h index 0b1d0d4ff0..55d78de334 100644 --- a/lib/eal/include/rte_eal_memconfig.h +++ b/lib/eal/include/rte_eal_memconfig.h @@ -132,14 +132,14 @@ rte_mcfg_mempool_write_unlock(void) */ void rte_mcfg_timer_lock(void) - __rte_exclusive_lock_function(rte_mcfg_timer_get_lock()); + __rte_acquire_capability(rte_mcfg_timer_get_lock()); /** * Unlock the internal EAL Timer Library lock for exclusive access. */ void rte_mcfg_timer_unlock(void) - __rte_unlock_function(rte_mcfg_timer_get_lock()); + __rte_release_capability(rte_mcfg_timer_get_lock()); /** * If true, pages are put in single files (per memseg list), diff --git a/lib/eal/include/rte_seqlock.h b/lib/eal/include/rte_seqlock.h index e0e94900d1..670a680443 100644 --- a/lib/eal/include/rte_seqlock.h +++ b/lib/eal/include/rte_seqlock.h @@ -197,7 +197,7 @@ rte_seqlock_read_retry(const rte_seqlock_t *seqlock, uint32_t begin_sn) */ static inline void rte_seqlock_write_lock(rte_seqlock_t *seqlock) - __rte_exclusive_lock_function(&seqlock->lock) + __rte_acquire_capability(&seqlock->lock) { /* To synchronize with other writers. */ rte_spinlock_lock(&seqlock->lock); @@ -219,7 +219,7 @@ rte_seqlock_write_lock(rte_seqlock_t *seqlock) */ static inline void rte_seqlock_write_unlock(rte_seqlock_t *seqlock) - __rte_unlock_function(&seqlock->lock) + __rte_release_capability(&seqlock->lock) { rte_seqcount_write_end(&seqlock->count); diff --git a/lib/graph/graph_private.h b/lib/graph/graph_private.h index da48d73587..813dd78b9d 100644 --- a/lib/graph/graph_private.h +++ b/lib/graph/graph_private.h @@ -221,7 +221,7 @@ graph_spinlock_get(void); * Take a lock on the graph internal spin lock. */ void graph_spinlock_lock(void) - __rte_exclusive_lock_function(graph_spinlock_get()); + __rte_acquire_capability(graph_spinlock_get()); /** * @internal @@ -229,7 +229,7 @@ void graph_spinlock_lock(void) * Release a lock on the graph internal spin lock. */ void graph_spinlock_unlock(void) - __rte_unlock_function(graph_spinlock_get()); + __rte_release_capability(graph_spinlock_get()); /* Graph operations */ /** diff --git a/lib/vhost/vdpa.c b/lib/vhost/vdpa.c index 8abb073675..329dd572b7 100644 --- a/lib/vhost/vdpa.c +++ b/lib/vhost/vdpa.c @@ -32,7 +32,7 @@ static struct vdpa_device_list * const vdpa_device_list static struct rte_vdpa_device * __vdpa_find_device_by_name(const char *name) - __rte_exclusive_locks_required(&vdpa_device_list_lock) + __rte_requires_capability(&vdpa_device_list_lock) { struct rte_vdpa_device *dev, *ret = NULL; -- 2.47.0 ^ permalink raw reply [flat|nested] 10+ messages in thread
* [PATCH v2 3/3] eal: enhance lock annotations for rwlock 2024-12-12 16:00 ` [PATCH v2 0/3] Improve lock annotations David Marchand 2024-12-12 16:00 ` [PATCH v2 1/3] eal: add enhanced " David Marchand 2024-12-12 16:00 ` [PATCH v2 2/3] eal: enhance lock annotations for spinlock and seqlock David Marchand @ 2024-12-12 16:00 ` David Marchand 2 siblings, 0 replies; 10+ messages in thread From: David Marchand @ 2024-12-12 16:00 UTC (permalink / raw) To: dev Cc: thomas, Anatoly Burakov, Tyler Retzlaff, Ferruh Yigit, Andrew Rybchenko, Yipeng Wang, Sameh Gobriel, Bruce Richardson, Vladimir Medvedkin, Maxime Coquelin, Chenbo Xia Convert rwlock to the clang capability annotations. Signed-off-by: David Marchand <david.marchand@redhat.com> --- lib/eal/common/eal_memalloc.h | 2 +- lib/eal/common/eal_private.h | 2 +- lib/eal/include/generic/rte_rwlock.h | 23 ++-- lib/eal/include/rte_eal_memconfig.h | 24 ++-- lib/ethdev/ethdev_driver.c | 4 +- lib/ethdev/ethdev_private.h | 4 +- lib/ethdev/rte_ethdev.c | 4 +- lib/hash/rte_cuckoo_hash.c | 8 +- lib/vhost/iotlb.h | 8 +- lib/vhost/vhost.c | 10 +- lib/vhost/vhost.h | 24 ++-- lib/vhost/vhost_crypto.c | 14 +-- lib/vhost/virtio_net.c | 170 +++++++++++++-------------- lib/vhost/virtio_net_ctrl.c | 2 +- 14 files changed, 150 insertions(+), 149 deletions(-) diff --git a/lib/eal/common/eal_memalloc.h b/lib/eal/common/eal_memalloc.h index 286ffb7633..0c267066d9 100644 --- a/lib/eal/common/eal_memalloc.h +++ b/lib/eal/common/eal_memalloc.h @@ -92,7 +92,7 @@ eal_memalloc_get_seg_fd_offset(int list_idx, int seg_idx, size_t *offset); int eal_memalloc_init(void) - __rte_shared_locks_required(rte_mcfg_mem_get_lock()); + __rte_requires_shared_capability(rte_mcfg_mem_get_lock()); int eal_memalloc_cleanup(void); diff --git a/lib/eal/common/eal_private.h b/lib/eal/common/eal_private.h index bb315dab04..89bc05ecc5 100644 --- a/lib/eal/common/eal_private.h +++ b/lib/eal/common/eal_private.h @@ -124,7 +124,7 @@ int rte_eal_memseg_init(void); * 0 on success, negative on error */ int rte_eal_memory_init(void) - __rte_shared_locks_required(rte_mcfg_mem_get_lock()); + __rte_requires_shared_capability(rte_mcfg_mem_get_lock()); /** * Configure timers diff --git a/lib/eal/include/generic/rte_rwlock.h b/lib/eal/include/generic/rte_rwlock.h index ac0474466a..197b245d11 100644 --- a/lib/eal/include/generic/rte_rwlock.h +++ b/lib/eal/include/generic/rte_rwlock.h @@ -23,6 +23,7 @@ */ #include <errno.h> +#include <stdbool.h> #include <rte_branch_prediction.h> #include <rte_common.h> @@ -57,7 +58,7 @@ extern "C" { /* Writer is waiting or has lock */ #define RTE_RWLOCK_READ 0x4 /* Reader increment */ -typedef struct __rte_lockable { +typedef struct __rte_capability("rwlock") { RTE_ATOMIC(int32_t) cnt; } rte_rwlock_t; @@ -90,7 +91,7 @@ rte_rwlock_init(rte_rwlock_t *rwl) */ static inline void rte_rwlock_read_lock(rte_rwlock_t *rwl) - __rte_shared_lock_function(rwl) + __rte_acquire_shared_capability(rwl) __rte_no_thread_safety_analysis { int32_t x; @@ -127,7 +128,7 @@ rte_rwlock_read_lock(rte_rwlock_t *rwl) */ static inline int rte_rwlock_read_trylock(rte_rwlock_t *rwl) - __rte_shared_trylock_function(0, rwl) + __rte_try_acquire_shared_capability(false, rwl) __rte_no_thread_safety_analysis { int32_t x; @@ -160,7 +161,7 @@ rte_rwlock_read_trylock(rte_rwlock_t *rwl) */ static inline void rte_rwlock_read_unlock(rte_rwlock_t *rwl) - __rte_unlock_function(rwl) + __rte_release_shared_capability(rwl) __rte_no_thread_safety_analysis { rte_atomic_fetch_sub_explicit(&rwl->cnt, RTE_RWLOCK_READ, rte_memory_order_release); @@ -178,7 +179,7 @@ rte_rwlock_read_unlock(rte_rwlock_t *rwl) */ static inline int rte_rwlock_write_trylock(rte_rwlock_t *rwl) - __rte_exclusive_trylock_function(0, rwl) + __rte_try_acquire_capability(false, rwl) __rte_no_thread_safety_analysis { int32_t x; @@ -200,7 +201,7 @@ rte_rwlock_write_trylock(rte_rwlock_t *rwl) */ static inline void rte_rwlock_write_lock(rte_rwlock_t *rwl) - __rte_exclusive_lock_function(rwl) + __rte_acquire_capability(rwl) __rte_no_thread_safety_analysis { int32_t x; @@ -238,7 +239,7 @@ rte_rwlock_write_lock(rte_rwlock_t *rwl) */ static inline void rte_rwlock_write_unlock(rte_rwlock_t *rwl) - __rte_unlock_function(rwl) + __rte_release_capability(rwl) __rte_no_thread_safety_analysis { rte_atomic_fetch_sub_explicit(&rwl->cnt, RTE_RWLOCK_WRITE, rte_memory_order_release); @@ -276,7 +277,7 @@ rte_rwlock_write_is_locked(rte_rwlock_t *rwl) */ static inline void rte_rwlock_read_lock_tm(rte_rwlock_t *rwl) - __rte_shared_lock_function(rwl); + __rte_acquire_shared_capability(rwl); /** * Commit hardware memory transaction or release the read lock if the lock is used as a fall-back @@ -286,7 +287,7 @@ rte_rwlock_read_lock_tm(rte_rwlock_t *rwl) */ static inline void rte_rwlock_read_unlock_tm(rte_rwlock_t *rwl) - __rte_unlock_function(rwl); + __rte_release_shared_capability(rwl); /** * Try to execute critical section in a hardware memory transaction, if it @@ -303,7 +304,7 @@ rte_rwlock_read_unlock_tm(rte_rwlock_t *rwl) */ static inline void rte_rwlock_write_lock_tm(rte_rwlock_t *rwl) - __rte_exclusive_lock_function(rwl); + __rte_acquire_capability(rwl); /** * Commit hardware memory transaction or release the write lock if the lock is used as a fall-back @@ -313,7 +314,7 @@ rte_rwlock_write_lock_tm(rte_rwlock_t *rwl) */ static inline void rte_rwlock_write_unlock_tm(rte_rwlock_t *rwl) - __rte_unlock_function(rwl); + __rte_release_capability(rwl); #ifdef __cplusplus } diff --git a/lib/eal/include/rte_eal_memconfig.h b/lib/eal/include/rte_eal_memconfig.h index 55d78de334..c3056021a0 100644 --- a/lib/eal/include/rte_eal_memconfig.h +++ b/lib/eal/include/rte_eal_memconfig.h @@ -48,84 +48,84 @@ rte_mcfg_ethdev_get_lock(void); */ void rte_mcfg_mem_read_lock(void) - __rte_shared_lock_function(rte_mcfg_mem_get_lock()); + __rte_acquire_shared_capability(rte_mcfg_mem_get_lock()); /** * Unlock the internal EAL shared memory configuration for shared access. */ void rte_mcfg_mem_read_unlock(void) - __rte_unlock_function(rte_mcfg_mem_get_lock()); + __rte_release_shared_capability(rte_mcfg_mem_get_lock()); /** * Lock the internal EAL shared memory configuration for exclusive access. */ void rte_mcfg_mem_write_lock(void) - __rte_exclusive_lock_function(rte_mcfg_mem_get_lock()); + __rte_acquire_capability(rte_mcfg_mem_get_lock()); /** * Unlock the internal EAL shared memory configuration for exclusive access. */ void rte_mcfg_mem_write_unlock(void) - __rte_unlock_function(rte_mcfg_mem_get_lock()); + __rte_release_capability(rte_mcfg_mem_get_lock()); /** * Lock the internal EAL TAILQ list for shared access. */ void rte_mcfg_tailq_read_lock(void) - __rte_shared_lock_function(rte_mcfg_tailq_get_lock()); + __rte_acquire_shared_capability(rte_mcfg_tailq_get_lock()); /** * Unlock the internal EAL TAILQ list for shared access. */ void rte_mcfg_tailq_read_unlock(void) - __rte_unlock_function(rte_mcfg_tailq_get_lock()); + __rte_release_shared_capability(rte_mcfg_tailq_get_lock()); /** * Lock the internal EAL TAILQ list for exclusive access. */ void rte_mcfg_tailq_write_lock(void) - __rte_exclusive_lock_function(rte_mcfg_tailq_get_lock()); + __rte_acquire_capability(rte_mcfg_tailq_get_lock()); /** * Unlock the internal EAL TAILQ list for exclusive access. */ void rte_mcfg_tailq_write_unlock(void) - __rte_unlock_function(rte_mcfg_tailq_get_lock()); + __rte_release_capability(rte_mcfg_tailq_get_lock()); /** * Lock the internal EAL Mempool list for shared access. */ void rte_mcfg_mempool_read_lock(void) - __rte_shared_lock_function(rte_mcfg_mempool_get_lock()); + __rte_acquire_shared_capability(rte_mcfg_mempool_get_lock()); /** * Unlock the internal EAL Mempool list for shared access. */ void rte_mcfg_mempool_read_unlock(void) - __rte_unlock_function(rte_mcfg_mempool_get_lock()); + __rte_release_shared_capability(rte_mcfg_mempool_get_lock()); /** * Lock the internal EAL Mempool list for exclusive access. */ void rte_mcfg_mempool_write_lock(void) - __rte_exclusive_lock_function(rte_mcfg_mempool_get_lock()); + __rte_acquire_capability(rte_mcfg_mempool_get_lock()); /** * Unlock the internal EAL Mempool list for exclusive access. */ void rte_mcfg_mempool_write_unlock(void) - __rte_unlock_function(rte_mcfg_mempool_get_lock()); + __rte_release_capability(rte_mcfg_mempool_get_lock()); /** * Lock the internal EAL Timer Library lock for exclusive access. diff --git a/lib/ethdev/ethdev_driver.c b/lib/ethdev/ethdev_driver.c index 9afef06431..13d328d97b 100644 --- a/lib/ethdev/ethdev_driver.c +++ b/lib/ethdev/ethdev_driver.c @@ -48,7 +48,7 @@ eth_dev_allocated(const char *name) static uint16_t eth_dev_find_free_port(void) - __rte_exclusive_locks_required(rte_mcfg_ethdev_get_lock()) + __rte_requires_capability(rte_mcfg_ethdev_get_lock()) { uint16_t i; @@ -65,7 +65,7 @@ eth_dev_find_free_port(void) static struct rte_eth_dev * eth_dev_get(uint16_t port_id) - __rte_exclusive_locks_required(rte_mcfg_ethdev_get_lock()) + __rte_requires_capability(rte_mcfg_ethdev_get_lock()) { struct rte_eth_dev *eth_dev = &rte_eth_devices[port_id]; diff --git a/lib/ethdev/ethdev_private.h b/lib/ethdev/ethdev_private.h index 0d36b9c30f..b07b1b4c42 100644 --- a/lib/ethdev/ethdev_private.h +++ b/lib/ethdev/ethdev_private.h @@ -70,9 +70,9 @@ void eth_dev_fp_ops_setup(struct rte_eth_fp_ops *fpo, void *eth_dev_shared_data_prepare(void) - __rte_exclusive_locks_required(rte_mcfg_ethdev_get_lock()); + __rte_requires_capability(rte_mcfg_ethdev_get_lock()); void eth_dev_shared_data_release(void) - __rte_exclusive_locks_required(rte_mcfg_ethdev_get_lock()); + __rte_requires_capability(rte_mcfg_ethdev_get_lock()); void eth_dev_rxq_release(struct rte_eth_dev *dev, uint16_t qid); void eth_dev_txq_release(struct rte_eth_dev *dev, uint16_t qid); diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_ethdev.c index 6413c54e3b..355d5c2df1 100644 --- a/lib/ethdev/rte_ethdev.c +++ b/lib/ethdev/rte_ethdev.c @@ -420,7 +420,7 @@ rte_eth_dev_is_valid_port(uint16_t port_id) static int eth_is_valid_owner_id(uint64_t owner_id) - __rte_exclusive_locks_required(rte_mcfg_ethdev_get_lock()) + __rte_requires_capability(rte_mcfg_ethdev_get_lock()) { if (owner_id == RTE_ETH_DEV_NO_OWNER || eth_dev_shared_data->next_owner_id <= owner_id) @@ -471,7 +471,7 @@ rte_eth_dev_owner_new(uint64_t *owner_id) static int eth_dev_owner_set(const uint16_t port_id, const uint64_t old_owner_id, const struct rte_eth_dev_owner *new_owner) - __rte_exclusive_locks_required(rte_mcfg_ethdev_get_lock()) + __rte_requires_capability(rte_mcfg_ethdev_get_lock()) { struct rte_eth_dev *ethdev = &rte_eth_devices[port_id]; struct rte_eth_dev_owner *port_owner; diff --git a/lib/hash/rte_cuckoo_hash.c b/lib/hash/rte_cuckoo_hash.c index 9575e8aa0c..00cbdf89d6 100644 --- a/lib/hash/rte_cuckoo_hash.c +++ b/lib/hash/rte_cuckoo_hash.c @@ -620,7 +620,7 @@ rte_hash_count(const struct rte_hash *h) /* Read write locks implemented using rte_rwlock */ static inline void __hash_rw_writer_lock(const struct rte_hash *h) - __rte_exclusive_lock_function(&h->readwrite_lock) + __rte_acquire_capability(&h->readwrite_lock) __rte_no_thread_safety_analysis { if (h->writer_takes_lock && h->hw_trans_mem_support) @@ -631,7 +631,7 @@ __hash_rw_writer_lock(const struct rte_hash *h) static inline void __hash_rw_reader_lock(const struct rte_hash *h) - __rte_shared_lock_function(&h->readwrite_lock) + __rte_acquire_shared_capability(&h->readwrite_lock) __rte_no_thread_safety_analysis { if (h->readwrite_concur_support && h->hw_trans_mem_support) @@ -642,7 +642,7 @@ __hash_rw_reader_lock(const struct rte_hash *h) static inline void __hash_rw_writer_unlock(const struct rte_hash *h) - __rte_unlock_function(&h->readwrite_lock) + __rte_release_capability(&h->readwrite_lock) __rte_no_thread_safety_analysis { if (h->writer_takes_lock && h->hw_trans_mem_support) @@ -653,7 +653,7 @@ __hash_rw_writer_unlock(const struct rte_hash *h) static inline void __hash_rw_reader_unlock(const struct rte_hash *h) - __rte_unlock_function(&h->readwrite_lock) + __rte_release_shared_capability(&h->readwrite_lock) __rte_no_thread_safety_analysis { if (h->readwrite_concur_support && h->hw_trans_mem_support) diff --git a/lib/vhost/iotlb.h b/lib/vhost/iotlb.h index 81ca04df21..72232b0dcf 100644 --- a/lib/vhost/iotlb.h +++ b/lib/vhost/iotlb.h @@ -11,28 +11,28 @@ static __rte_always_inline void vhost_user_iotlb_rd_lock(struct vhost_virtqueue *vq) - __rte_shared_lock_function(&vq->iotlb_lock) + __rte_acquire_shared_capability(&vq->iotlb_lock) { rte_rwlock_read_lock(&vq->iotlb_lock); } static __rte_always_inline void vhost_user_iotlb_rd_unlock(struct vhost_virtqueue *vq) - __rte_unlock_function(&vq->iotlb_lock) + __rte_release_shared_capability(&vq->iotlb_lock) { rte_rwlock_read_unlock(&vq->iotlb_lock); } static __rte_always_inline void vhost_user_iotlb_wr_lock(struct vhost_virtqueue *vq) - __rte_exclusive_lock_function(&vq->iotlb_lock) + __rte_acquire_capability(&vq->iotlb_lock) { rte_rwlock_write_lock(&vq->iotlb_lock); } static __rte_always_inline void vhost_user_iotlb_wr_unlock(struct vhost_virtqueue *vq) - __rte_unlock_function(&vq->iotlb_lock) + __rte_release_capability(&vq->iotlb_lock) { rte_rwlock_write_unlock(&vq->iotlb_lock); } diff --git a/lib/vhost/vhost.c b/lib/vhost/vhost.c index 5a50a06f8d..2f34d3500e 100644 --- a/lib/vhost/vhost.c +++ b/lib/vhost/vhost.c @@ -390,7 +390,7 @@ cleanup_device(struct virtio_net *dev, int destroy) static void vhost_free_async_mem(struct vhost_virtqueue *vq) - __rte_exclusive_locks_required(&vq->access_lock) + __rte_requires_capability(&vq->access_lock) { if (!vq->async) return; @@ -439,7 +439,7 @@ free_device(struct virtio_net *dev) static __rte_always_inline int log_translate(struct virtio_net *dev, struct vhost_virtqueue *vq) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { if (likely(!(vq->ring_addrs.flags & (1 << VHOST_VRING_F_LOG)))) return 0; @@ -488,7 +488,7 @@ translate_log_addr(struct virtio_net *dev, struct vhost_virtqueue *vq, static int vring_translate_split(struct virtio_net *dev, struct vhost_virtqueue *vq) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint64_t req_size, size; @@ -527,7 +527,7 @@ vring_translate_split(struct virtio_net *dev, struct vhost_virtqueue *vq) static int vring_translate_packed(struct virtio_net *dev, struct vhost_virtqueue *vq) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint64_t req_size, size; @@ -1772,7 +1772,7 @@ rte_vhost_extern_callback_register(int vid, static __rte_always_inline int async_channel_register(struct virtio_net *dev, struct vhost_virtqueue *vq) - __rte_exclusive_locks_required(&vq->access_lock) + __rte_requires_capability(&vq->access_lock) { struct vhost_async *async; int node = vq->numa_node; diff --git a/lib/vhost/vhost.h b/lib/vhost/vhost.h index 1f4192f5d1..0a7815832f 100644 --- a/lib/vhost/vhost.h +++ b/lib/vhost/vhost.h @@ -576,7 +576,7 @@ vhost_virtqueue_reconnect_log_packed(struct vhost_virtqueue *vq) static inline void vq_assert_lock__(struct virtio_net *dev, struct vhost_virtqueue *vq, const char *func) - __rte_assert_exclusive_lock(&vq->access_lock) + __rte_assert_capability(&vq->access_lock) { if (unlikely(!rte_rwlock_write_is_locked(&vq->access_lock))) rte_panic("VHOST_CONFIG: (%s) %s() called without access lock taken.\n", @@ -627,14 +627,14 @@ void __vhost_log_cache_write(struct virtio_net *dev, void __vhost_log_cache_write_iova(struct virtio_net *dev, struct vhost_virtqueue *vq, uint64_t iova, uint64_t len) - __rte_shared_locks_required(&vq->iotlb_lock); + __rte_requires_shared_capability(&vq->iotlb_lock); void __vhost_log_cache_sync(struct virtio_net *dev, struct vhost_virtqueue *vq); void __vhost_log_write(struct virtio_net *dev, uint64_t addr, uint64_t len); void __vhost_log_write_iova(struct virtio_net *dev, struct vhost_virtqueue *vq, uint64_t iova, uint64_t len) - __rte_shared_locks_required(&vq->iotlb_lock); + __rte_requires_shared_capability(&vq->iotlb_lock); static __rte_always_inline void vhost_log_write(struct virtio_net *dev, uint64_t addr, uint64_t len) @@ -684,7 +684,7 @@ vhost_log_used_vring(struct virtio_net *dev, struct vhost_virtqueue *vq, static __rte_always_inline void vhost_log_cache_write_iova(struct virtio_net *dev, struct vhost_virtqueue *vq, uint64_t iova, uint64_t len) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { if (likely(!(dev->features & (1ULL << VHOST_F_LOG_ALL)))) return; @@ -698,7 +698,7 @@ vhost_log_cache_write_iova(struct virtio_net *dev, struct vhost_virtqueue *vq, static __rte_always_inline void vhost_log_write_iova(struct virtio_net *dev, struct vhost_virtqueue *vq, uint64_t iova, uint64_t len) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { if (likely(!(dev->features & (1ULL << VHOST_F_LOG_ALL)))) return; @@ -904,24 +904,24 @@ void vhost_backend_cleanup(struct virtio_net *dev); uint64_t __vhost_iova_to_vva(struct virtio_net *dev, struct vhost_virtqueue *vq, uint64_t iova, uint64_t *len, uint8_t perm) - __rte_shared_locks_required(&vq->iotlb_lock); + __rte_requires_shared_capability(&vq->iotlb_lock); void *vhost_alloc_copy_ind_table(struct virtio_net *dev, struct vhost_virtqueue *vq, uint64_t desc_addr, uint64_t desc_len) - __rte_shared_locks_required(&vq->iotlb_lock); + __rte_requires_shared_capability(&vq->iotlb_lock); int vring_translate(struct virtio_net *dev, struct vhost_virtqueue *vq) - __rte_exclusive_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock); + __rte_requires_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock); uint64_t translate_log_addr(struct virtio_net *dev, struct vhost_virtqueue *vq, uint64_t log_addr) - __rte_shared_locks_required(&vq->iotlb_lock); + __rte_requires_shared_capability(&vq->iotlb_lock); void vring_invalidate(struct virtio_net *dev, struct vhost_virtqueue *vq) - __rte_exclusive_locks_required(&vq->access_lock); + __rte_requires_capability(&vq->access_lock); static __rte_always_inline uint64_t vhost_iova_to_vva(struct virtio_net *dev, struct vhost_virtqueue *vq, uint64_t iova, uint64_t *len, uint8_t perm) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { if (!(dev->features & (1ULL << VIRTIO_F_IOMMU_PLATFORM))) return rte_vhost_va_from_guest_pa(dev->mem, iova, len); diff --git a/lib/vhost/vhost_crypto.c b/lib/vhost/vhost_crypto.c index 7caf6d9afa..3dc41a3bd5 100644 --- a/lib/vhost/vhost_crypto.c +++ b/lib/vhost/vhost_crypto.c @@ -491,7 +491,7 @@ static __rte_always_inline struct virtio_crypto_inhdr * reach_inhdr(struct vhost_crypto_data_req *vc_req, struct vhost_crypto_desc *head, uint32_t max_n_descs) - __rte_shared_locks_required(&vc_req->vq->iotlb_lock) + __rte_requires_shared_capability(&vc_req->vq->iotlb_lock) { struct virtio_crypto_inhdr *inhdr; struct vhost_crypto_desc *last = head + (max_n_descs - 1); @@ -538,7 +538,7 @@ static __rte_always_inline void * get_data_ptr(struct vhost_crypto_data_req *vc_req, struct vhost_crypto_desc *cur_desc, uint8_t perm) - __rte_shared_locks_required(&vc_req->vq->iotlb_lock) + __rte_requires_shared_capability(&vc_req->vq->iotlb_lock) { void *data; uint64_t dlen = cur_desc->len; @@ -555,7 +555,7 @@ get_data_ptr(struct vhost_crypto_data_req *vc_req, static __rte_always_inline uint32_t copy_data_from_desc(void *dst, struct vhost_crypto_data_req *vc_req, struct vhost_crypto_desc *desc, uint32_t size) - __rte_shared_locks_required(&vc_req->vq->iotlb_lock) + __rte_requires_shared_capability(&vc_req->vq->iotlb_lock) { uint64_t remain; uint64_t addr; @@ -586,7 +586,7 @@ static __rte_always_inline int copy_data(void *data, struct vhost_crypto_data_req *vc_req, struct vhost_crypto_desc *head, struct vhost_crypto_desc **cur_desc, uint32_t size, uint32_t max_n_descs) - __rte_shared_locks_required(&vc_req->vq->iotlb_lock) + __rte_requires_shared_capability(&vc_req->vq->iotlb_lock) { struct vhost_crypto_desc *desc = *cur_desc; uint32_t left = size; @@ -670,7 +670,7 @@ prepare_write_back_data(struct vhost_crypto_data_req *vc_req, uint32_t offset, uint64_t write_back_len, uint32_t max_n_descs) - __rte_shared_locks_required(&vc_req->vq->iotlb_lock) + __rte_requires_shared_capability(&vc_req->vq->iotlb_lock) { struct vhost_crypto_writeback_data *wb_data, *head; struct vhost_crypto_desc *desc = *cur_desc; @@ -791,7 +791,7 @@ prepare_sym_cipher_op(struct vhost_crypto *vcrypto, struct rte_crypto_op *op, struct virtio_crypto_cipher_data_req *cipher, struct vhost_crypto_desc *head, uint32_t max_n_descs) - __rte_shared_locks_required(&vc_req->vq->iotlb_lock) + __rte_requires_shared_capability(&vc_req->vq->iotlb_lock) { struct vhost_crypto_desc *desc = head; struct vhost_crypto_writeback_data *ewb = NULL; @@ -945,7 +945,7 @@ prepare_sym_chain_op(struct vhost_crypto *vcrypto, struct rte_crypto_op *op, struct virtio_crypto_alg_chain_data_req *chain, struct vhost_crypto_desc *head, uint32_t max_n_descs) - __rte_shared_locks_required(&vc_req->vq->iotlb_lock) + __rte_requires_shared_capability(&vc_req->vq->iotlb_lock) { struct vhost_crypto_desc *desc = head, *digest_desc; struct vhost_crypto_writeback_data *ewb = NULL, *ewb2 = NULL; diff --git a/lib/vhost/virtio_net.c b/lib/vhost/virtio_net.c index d764d4bc6a..00be575e07 100644 --- a/lib/vhost/virtio_net.c +++ b/lib/vhost/virtio_net.c @@ -55,7 +55,7 @@ is_valid_virt_queue_idx(uint32_t idx, int is_tx, uint32_t nr_vring) static inline void vhost_queue_stats_update(const struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts, uint16_t count) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { struct virtqueue_stats *stats = &vq->stats; int i; @@ -90,7 +90,7 @@ static __rte_always_inline int64_t vhost_async_dma_transfer_one(struct virtio_net *dev, struct vhost_virtqueue *vq, int16_t dma_id, uint16_t vchan_id, uint16_t flag_idx, struct vhost_iov_iter *pkt) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { struct async_dma_vchan_info *dma_info = &dma_copy_track[dma_id].vchans[vchan_id]; uint16_t ring_mask = dma_info->ring_mask; @@ -140,7 +140,7 @@ static __rte_always_inline uint16_t vhost_async_dma_transfer(struct virtio_net *dev, struct vhost_virtqueue *vq, int16_t dma_id, uint16_t vchan_id, uint16_t head_idx, struct vhost_iov_iter *pkts, uint16_t nr_pkts) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { struct async_dma_vchan_info *dma_info = &dma_copy_track[dma_id].vchans[vchan_id]; int64_t ret, nr_copies = 0; @@ -221,7 +221,7 @@ vhost_async_dma_check_completed(struct virtio_net *dev, int16_t dma_id, uint16_t static inline void do_data_copy_enqueue(struct virtio_net *dev, struct vhost_virtqueue *vq) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { struct batch_copy_elem *elem = vq->batch_copy_elems; uint16_t count = vq->batch_copy_nb_elems; @@ -426,7 +426,7 @@ static __rte_always_inline void vhost_async_shadow_enqueue_packed_batch(struct vhost_virtqueue *vq, uint64_t *lens, uint16_t *ids) - __rte_exclusive_locks_required(&vq->access_lock) + __rte_requires_capability(&vq->access_lock) { uint16_t i; struct vhost_async *async = vq->async; @@ -443,7 +443,7 @@ vhost_async_shadow_enqueue_packed_batch(struct vhost_virtqueue *vq, static __rte_always_inline void vhost_async_shadow_dequeue_packed_batch(struct vhost_virtqueue *vq, uint16_t *ids) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { uint16_t i; struct vhost_async *async = vq->async; @@ -605,7 +605,7 @@ vhost_async_shadow_enqueue_packed(struct vhost_virtqueue *vq, uint16_t *id, uint16_t *count, uint16_t num_buffers) - __rte_exclusive_locks_required(&vq->access_lock) + __rte_requires_capability(&vq->access_lock) { uint16_t i; struct vhost_async *async = vq->async; @@ -627,7 +627,7 @@ vhost_shadow_enqueue_single_packed(struct virtio_net *dev, uint16_t *id, uint16_t *count, uint16_t num_buffers) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { vhost_shadow_enqueue_packed(vq, len, id, count, num_buffers); @@ -719,7 +719,7 @@ static __rte_always_inline int map_one_desc(struct virtio_net *dev, struct vhost_virtqueue *vq, struct buf_vector *buf_vec, uint16_t *vec_idx, uint64_t desc_iova, uint64_t desc_len, uint8_t perm) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t vec_id = *vec_idx; @@ -757,7 +757,7 @@ fill_vec_buf_split(struct virtio_net *dev, struct vhost_virtqueue *vq, uint32_t avail_idx, uint16_t *vec_idx, struct buf_vector *buf_vec, uint16_t *desc_chain_head, uint32_t *desc_chain_len, uint8_t perm) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t idx = vq->avail->ring[avail_idx & (vq->size - 1)]; uint16_t vec_id = *vec_idx; @@ -841,7 +841,7 @@ reserve_avail_buf_split(struct virtio_net *dev, struct vhost_virtqueue *vq, uint64_t size, struct buf_vector *buf_vec, uint16_t *num_buffers, uint16_t avail_head, uint16_t *nr_vec) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t cur_idx; uint16_t vec_idx = 0; @@ -892,7 +892,7 @@ fill_vec_buf_packed_indirect(struct virtio_net *dev, struct vhost_virtqueue *vq, struct vring_packed_desc *desc, uint16_t *vec_idx, struct buf_vector *buf_vec, uint32_t *len, uint8_t perm) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t i; uint32_t nr_descs; @@ -951,7 +951,7 @@ fill_vec_buf_packed(struct virtio_net *dev, struct vhost_virtqueue *vq, uint16_t avail_idx, uint16_t *desc_count, struct buf_vector *buf_vec, uint16_t *vec_idx, uint16_t *buf_id, uint32_t *len, uint8_t perm) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { bool wrap_counter = vq->avail_wrap_counter; struct vring_packed_desc *descs = vq->desc_packed; @@ -1017,7 +1017,7 @@ static __rte_noinline void copy_vnet_hdr_to_desc(struct virtio_net *dev, struct vhost_virtqueue *vq, struct buf_vector *buf_vec, struct virtio_net_hdr_mrg_rxbuf *hdr) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint64_t len; uint64_t remain = dev->vhost_hlen; @@ -1120,8 +1120,8 @@ static __rte_always_inline int async_fill_seg(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf *m, uint32_t mbuf_offset, uint64_t buf_iova, uint32_t cpy_len, bool to_desc) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { struct vhost_async *async = vq->async; uint64_t mapped_len; @@ -1162,7 +1162,7 @@ static __rte_always_inline void sync_fill_seg(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf *m, uint32_t mbuf_offset, uint64_t buf_addr, uint64_t buf_iova, uint32_t cpy_len, bool to_desc) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { struct batch_copy_elem *batch_copy = vq->batch_copy_elems; @@ -1200,8 +1200,8 @@ static __rte_always_inline int mbuf_to_desc(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf *m, struct buf_vector *buf_vec, uint16_t nr_vec, uint16_t num_buffers, bool is_async) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint32_t vec_idx = 0; uint32_t mbuf_offset, mbuf_avail; @@ -1330,8 +1330,8 @@ vhost_enqueue_single_packed(struct virtio_net *dev, struct rte_mbuf *pkt, struct buf_vector *buf_vec, uint16_t *nr_descs) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t nr_vec = 0; uint16_t avail_idx = vq->last_avail_idx; @@ -1392,8 +1392,8 @@ vhost_enqueue_single_packed(struct virtio_net *dev, static __rte_noinline uint32_t virtio_dev_rx_split(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts, uint32_t count) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint32_t pkt_idx = 0; uint16_t num_buffers; @@ -1452,7 +1452,7 @@ virtio_dev_rx_sync_batch_check(struct virtio_net *dev, struct rte_mbuf **pkts, uint64_t *desc_addrs, uint64_t *lens) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { bool wrap_counter = vq->avail_wrap_counter; struct vring_packed_desc *descs = vq->desc_packed; @@ -1556,7 +1556,7 @@ virtio_dev_rx_batch_packed_copy(struct virtio_net *dev, struct rte_mbuf **pkts, uint64_t *desc_addrs, uint64_t *lens) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint32_t buf_offset = sizeof(struct virtio_net_hdr_mrg_rxbuf); struct virtio_net_hdr_mrg_rxbuf *hdrs[PACKED_BATCH_SIZE]; @@ -1604,7 +1604,7 @@ static __rte_always_inline int virtio_dev_rx_sync_batch_packed(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint64_t desc_addrs[PACKED_BATCH_SIZE]; uint64_t lens[PACKED_BATCH_SIZE]; @@ -1626,8 +1626,8 @@ static __rte_always_inline int16_t virtio_dev_rx_single_packed(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf *pkt) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { struct buf_vector buf_vec[BUF_VECTOR_MAX]; uint16_t nr_descs = 0; @@ -1652,8 +1652,8 @@ virtio_dev_rx_packed(struct virtio_net *dev, struct vhost_virtqueue *__rte_restrict vq, struct rte_mbuf **__rte_restrict pkts, uint32_t count) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint32_t pkt_idx = 0; @@ -1767,7 +1767,7 @@ rte_vhost_enqueue_burst(int vid, uint16_t queue_id, static __rte_always_inline uint16_t async_get_first_inflight_pkt_idx(struct vhost_virtqueue *vq) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { struct vhost_async *async = vq->async; @@ -1796,8 +1796,8 @@ store_dma_desc_info_split(struct vring_used_elem *s_ring, struct vring_used_elem static __rte_noinline uint32_t virtio_dev_rx_async_submit_split(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts, uint32_t count, int16_t dma_id, uint16_t vchan_id) - __rte_exclusive_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { struct buf_vector buf_vec[BUF_VECTOR_MAX]; uint32_t pkt_idx = 0; @@ -1907,8 +1907,8 @@ vhost_enqueue_async_packed(struct virtio_net *dev, struct buf_vector *buf_vec, uint16_t *nr_descs, uint16_t *nr_buffers) - __rte_exclusive_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t nr_vec = 0; uint16_t avail_idx = vq->last_avail_idx; @@ -1967,8 +1967,8 @@ vhost_enqueue_async_packed(struct virtio_net *dev, static __rte_always_inline int16_t virtio_dev_rx_async_packed(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf *pkt, uint16_t *nr_descs, uint16_t *nr_buffers) - __rte_exclusive_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { struct buf_vector buf_vec[BUF_VECTOR_MAX]; @@ -1991,8 +1991,8 @@ virtio_dev_rx_async_packed_batch_enqueue(struct virtio_net *dev, struct rte_mbuf **pkts, uint64_t *desc_addrs, uint64_t *lens) - __rte_exclusive_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint32_t buf_offset = sizeof(struct virtio_net_hdr_mrg_rxbuf); struct virtio_net_hdr_mrg_rxbuf *hdrs[PACKED_BATCH_SIZE]; @@ -2053,8 +2053,8 @@ virtio_dev_rx_async_packed_batch(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts, int16_t dma_id, uint16_t vchan_id) - __rte_exclusive_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint64_t desc_addrs[PACKED_BATCH_SIZE]; uint64_t lens[PACKED_BATCH_SIZE]; @@ -2070,7 +2070,7 @@ virtio_dev_rx_async_packed_batch(struct virtio_net *dev, static __rte_always_inline void dma_error_handler_packed(struct vhost_virtqueue *vq, uint16_t slot_idx, uint32_t nr_err, uint32_t *pkt_idx) - __rte_exclusive_locks_required(&vq->access_lock) + __rte_requires_capability(&vq->access_lock) { uint16_t descs_err = 0; uint16_t buffers_err = 0; @@ -2102,8 +2102,8 @@ dma_error_handler_packed(struct vhost_virtqueue *vq, uint16_t slot_idx, static __rte_noinline uint32_t virtio_dev_rx_async_submit_packed(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts, uint32_t count, int16_t dma_id, uint16_t vchan_id) - __rte_exclusive_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint32_t pkt_idx = 0; uint16_t n_xfer; @@ -2176,7 +2176,7 @@ virtio_dev_rx_async_submit_packed(struct virtio_net *dev, struct vhost_virtqueue static __rte_always_inline void write_back_completed_descs_split(struct vhost_virtqueue *vq, uint16_t n_descs) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { struct vhost_async *async = vq->async; uint16_t nr_left = n_descs; @@ -2209,7 +2209,7 @@ write_back_completed_descs_split(struct vhost_virtqueue *vq, uint16_t n_descs) static __rte_always_inline void write_back_completed_descs_packed(struct vhost_virtqueue *vq, uint16_t n_buffers) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { struct vhost_async *async = vq->async; uint16_t from = async->last_buffer_idx_packed; @@ -2274,7 +2274,7 @@ write_back_completed_descs_packed(struct vhost_virtqueue *vq, static __rte_always_inline uint16_t vhost_poll_enqueue_completed(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts, uint16_t count, int16_t dma_id, uint16_t vchan_id) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { struct vhost_async *async = vq->async; struct async_inflight_info *pkts_info = async->pkts_info; @@ -2884,8 +2884,8 @@ desc_to_mbuf(struct virtio_net *dev, struct vhost_virtqueue *vq, struct buf_vector *buf_vec, uint16_t nr_vec, struct rte_mbuf *m, struct rte_mempool *mbuf_pool, bool legacy_ol_flags, uint16_t slot_idx, bool is_async) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint32_t buf_avail, buf_offset, buf_len; uint64_t buf_addr, buf_iova; @@ -3092,8 +3092,8 @@ static uint16_t virtio_dev_tx_split(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count, bool legacy_ol_flags) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t i; uint16_t avail_entries; @@ -3188,8 +3188,8 @@ static uint16_t virtio_dev_tx_split_legacy(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { return virtio_dev_tx_split(dev, vq, mbuf_pool, pkts, count, true); } @@ -3199,8 +3199,8 @@ static uint16_t virtio_dev_tx_split_compliant(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { return virtio_dev_tx_split(dev, vq, mbuf_pool, pkts, count, false); } @@ -3212,7 +3212,7 @@ vhost_reserve_avail_batch_packed(struct virtio_net *dev, uint16_t avail_idx, uintptr_t *desc_addrs, uint16_t *ids) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { bool wrap = vq->avail_wrap_counter; struct vring_packed_desc *descs = vq->desc_packed; @@ -3356,7 +3356,7 @@ virtio_dev_tx_batch_packed(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts, bool legacy_ol_flags) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t avail_idx = vq->last_avail_idx; uint32_t buf_offset = sizeof(struct virtio_net_hdr_mrg_rxbuf); @@ -3403,8 +3403,8 @@ vhost_dequeue_single_packed(struct virtio_net *dev, uint16_t *buf_id, uint16_t *desc_count, bool legacy_ol_flags) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { struct buf_vector buf_vec[BUF_VECTOR_MAX]; uint32_t buf_len; @@ -3453,8 +3453,8 @@ virtio_dev_tx_single_packed(struct virtio_net *dev, struct rte_mempool *mbuf_pool, struct rte_mbuf *pkts, bool legacy_ol_flags) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t buf_id, desc_count = 0; @@ -3514,8 +3514,8 @@ virtio_dev_tx_packed(struct virtio_net *dev, struct rte_mbuf **__rte_restrict pkts, uint32_t count, bool legacy_ol_flags) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint32_t pkt_idx = 0; @@ -3565,8 +3565,8 @@ static uint16_t virtio_dev_tx_packed_legacy(struct virtio_net *dev, struct vhost_virtqueue *__rte_restrict vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **__rte_restrict pkts, uint32_t count) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { return virtio_dev_tx_packed(dev, vq, mbuf_pool, pkts, count, true); } @@ -3576,8 +3576,8 @@ static uint16_t virtio_dev_tx_packed_compliant(struct virtio_net *dev, struct vhost_virtqueue *__rte_restrict vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **__rte_restrict pkts, uint32_t count) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { return virtio_dev_tx_packed(dev, vq, mbuf_pool, pkts, count, false); } @@ -3696,7 +3696,7 @@ static __rte_always_inline uint16_t async_poll_dequeue_completed(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts, uint16_t count, int16_t dma_id, uint16_t vchan_id, bool legacy_ol_flags) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { uint16_t start_idx, from, i; uint16_t nr_cpl_pkts = 0; @@ -3744,8 +3744,8 @@ static __rte_always_inline uint16_t virtio_dev_tx_async_split(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count, int16_t dma_id, uint16_t vchan_id, bool legacy_ol_flags) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { static bool allocerr_warned; bool dropped = false; @@ -3897,8 +3897,8 @@ virtio_dev_tx_async_split_legacy(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count, int16_t dma_id, uint16_t vchan_id) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { return virtio_dev_tx_async_split(dev, vq, mbuf_pool, pkts, count, dma_id, vchan_id, true); @@ -3910,8 +3910,8 @@ virtio_dev_tx_async_split_compliant(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count, int16_t dma_id, uint16_t vchan_id) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { return virtio_dev_tx_async_split(dev, vq, mbuf_pool, pkts, count, dma_id, vchan_id, false); @@ -3920,7 +3920,7 @@ virtio_dev_tx_async_split_compliant(struct virtio_net *dev, static __rte_always_inline void vhost_async_shadow_dequeue_single_packed(struct vhost_virtqueue *vq, uint16_t buf_id, uint16_t count) - __rte_shared_locks_required(&vq->access_lock) + __rte_requires_shared_capability(&vq->access_lock) { struct vhost_async *async = vq->async; uint16_t idx = async->buffer_idx_packed; @@ -3942,8 +3942,8 @@ virtio_dev_tx_async_single_packed(struct virtio_net *dev, struct rte_mbuf *pkts, uint16_t slot_idx, bool legacy_ol_flags) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { int err; uint16_t buf_id, desc_count = 0; @@ -3995,8 +3995,8 @@ virtio_dev_tx_async_packed_batch(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mbuf **pkts, uint16_t slot_idx, uint16_t dma_id, uint16_t vchan_id) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint16_t avail_idx = vq->last_avail_idx; uint32_t buf_offset = sizeof(struct virtio_net_hdr_mrg_rxbuf); @@ -4053,8 +4053,8 @@ static __rte_always_inline uint16_t virtio_dev_tx_async_packed(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count, uint16_t dma_id, uint16_t vchan_id, bool legacy_ol_flags) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { uint32_t pkt_idx = 0; uint16_t slot_idx = 0; @@ -4167,8 +4167,8 @@ static uint16_t virtio_dev_tx_async_packed_legacy(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count, uint16_t dma_id, uint16_t vchan_id) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { return virtio_dev_tx_async_packed(dev, vq, mbuf_pool, pkts, count, dma_id, vchan_id, true); @@ -4179,8 +4179,8 @@ static uint16_t virtio_dev_tx_async_packed_compliant(struct virtio_net *dev, struct vhost_virtqueue *vq, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count, uint16_t dma_id, uint16_t vchan_id) - __rte_shared_locks_required(&vq->access_lock) - __rte_shared_locks_required(&vq->iotlb_lock) + __rte_requires_shared_capability(&vq->access_lock) + __rte_requires_shared_capability(&vq->iotlb_lock) { return virtio_dev_tx_async_packed(dev, vq, mbuf_pool, pkts, count, dma_id, vchan_id, false); diff --git a/lib/vhost/virtio_net_ctrl.c b/lib/vhost/virtio_net_ctrl.c index b8ee94018e..999e84db7c 100644 --- a/lib/vhost/virtio_net_ctrl.c +++ b/lib/vhost/virtio_net_ctrl.c @@ -26,7 +26,7 @@ struct virtio_net_ctrl_elem { static int virtio_net_ctrl_pop(struct virtio_net *dev, struct vhost_virtqueue *cvq, struct virtio_net_ctrl_elem *ctrl_elem) - __rte_shared_locks_required(&cvq->iotlb_lock) + __rte_requires_shared_capability(&cvq->iotlb_lock) { uint16_t avail_idx, desc_idx, n_descs = 0; uint64_t desc_len, desc_addr, desc_iova, data_len = 0; -- 2.47.0 ^ permalink raw reply [flat|nested] 10+ messages in thread
end of thread, other threads:[~2024-12-12 16:02 UTC | newest] Thread overview: 10+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2024-12-02 12:53 [RFC 0/3] Improve lock annotations David Marchand 2024-12-02 12:53 ` [RFC 1/3] eal: add enhanced " David Marchand 2024-12-02 16:13 ` Stephen Hemminger 2024-12-02 12:53 ` [RFC 2/3] eal: enhance lock annotations for spinlock and seqlock David Marchand 2024-12-05 6:18 ` Mattias Rönnblom 2024-12-02 12:53 ` [RFC 3/3] eal: enhance lock annotations for rwlock David Marchand 2024-12-12 16:00 ` [PATCH v2 0/3] Improve lock annotations David Marchand 2024-12-12 16:00 ` [PATCH v2 1/3] eal: add enhanced " David Marchand 2024-12-12 16:00 ` [PATCH v2 2/3] eal: enhance lock annotations for spinlock and seqlock David Marchand 2024-12-12 16:00 ` [PATCH v2 3/3] eal: enhance lock annotations for rwlock David Marchand
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).