| Message ID | 20230724094354.90817-24-zhengqi.arch@bytedance.com (mailing list archive) |
|---|---|
| State | Handled Elsewhere, archived |
| Headers | show |
| Series | use refcount+RCU method to implement lockless slab shrink | expand |
On 2023/7/24 17:43, Qi Zheng wrote: > In preparation for implementing lockless slab shrink, use new APIs to > dynamically allocate the drm-msm_gem shrinker, so that it can be freed > asynchronously using kfree_rcu(). Then it doesn't need to wait for RCU > read-side critical section when releasing the struct msm_drm_private. > > Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com> Reviewed-by: Muchun Song <songmuchun@bytedance.com> A nit bellow. > --- > drivers/gpu/drm/msm/msm_drv.c | 4 ++- > drivers/gpu/drm/msm/msm_drv.h | 4 +-- > drivers/gpu/drm/msm/msm_gem_shrinker.c | 36 ++++++++++++++++---------- > 3 files changed, 28 insertions(+), 16 deletions(-) > > diff --git a/drivers/gpu/drm/msm/msm_drv.c b/drivers/gpu/drm/msm/msm_drv.c > index 891eff8433a9..7f6933be703f 100644 > --- a/drivers/gpu/drm/msm/msm_drv.c > +++ b/drivers/gpu/drm/msm/msm_drv.c > @@ -461,7 +461,9 @@ static int msm_drm_init(struct device *dev, const struct drm_driver *drv) > if (ret) > goto err_msm_uninit; > > - msm_gem_shrinker_init(ddev); > + ret = msm_gem_shrinker_init(ddev); > + if (ret) > + goto err_msm_uninit; > > if (priv->kms_init) { > ret = priv->kms_init(ddev); > diff --git a/drivers/gpu/drm/msm/msm_drv.h b/drivers/gpu/drm/msm/msm_drv.h > index e13a8cbd61c9..84523d4a1e58 100644 > --- a/drivers/gpu/drm/msm/msm_drv.h > +++ b/drivers/gpu/drm/msm/msm_drv.h > @@ -217,7 +217,7 @@ struct msm_drm_private { > } vram; > > struct notifier_block vmap_notifier; > - struct shrinker shrinker; > + struct shrinker *shrinker; > > struct drm_atomic_state *pm_state; > > @@ -279,7 +279,7 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data, > unsigned long msm_gem_shrinker_shrink(struct drm_device *dev, unsigned long nr_to_scan); > #endif > > -void msm_gem_shrinker_init(struct drm_device *dev); > +int msm_gem_shrinker_init(struct drm_device *dev); > void msm_gem_shrinker_cleanup(struct drm_device *dev); > > int msm_gem_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma); > diff --git a/drivers/gpu/drm/msm/msm_gem_shrinker.c b/drivers/gpu/drm/msm/msm_gem_shrinker.c > index f38296ad8743..7daab1298c11 100644 > --- a/drivers/gpu/drm/msm/msm_gem_shrinker.c > +++ b/drivers/gpu/drm/msm/msm_gem_shrinker.c > @@ -34,8 +34,7 @@ static bool can_block(struct shrink_control *sc) > static unsigned long > msm_gem_shrinker_count(struct shrinker *shrinker, struct shrink_control *sc) > { > - struct msm_drm_private *priv = > - container_of(shrinker, struct msm_drm_private, shrinker); > + struct msm_drm_private *priv = shrinker->private_data; > unsigned count = priv->lru.dontneed.count; > > if (can_swap()) > @@ -100,8 +99,7 @@ active_evict(struct drm_gem_object *obj) > static unsigned long > msm_gem_shrinker_scan(struct shrinker *shrinker, struct shrink_control *sc) > { > - struct msm_drm_private *priv = > - container_of(shrinker, struct msm_drm_private, shrinker); > + struct msm_drm_private *priv = shrinker->private_data; > struct { > struct drm_gem_lru *lru; > bool (*shrink)(struct drm_gem_object *obj); > @@ -148,10 +146,11 @@ msm_gem_shrinker_shrink(struct drm_device *dev, unsigned long nr_to_scan) > struct shrink_control sc = { > .nr_to_scan = nr_to_scan, > }; > - int ret; > + unsigned long ret = SHRINK_STOP; > > fs_reclaim_acquire(GFP_KERNEL); > - ret = msm_gem_shrinker_scan(&priv->shrinker, &sc); > + if (priv->shrinker) > + ret = msm_gem_shrinker_scan(priv->shrinker, &sc); > fs_reclaim_release(GFP_KERNEL); > > return ret; > @@ -210,16 +209,27 @@ msm_gem_shrinker_vmap(struct notifier_block *nb, unsigned long event, void *ptr) > * > * This function registers and sets up the msm shrinker. > */ > -void msm_gem_shrinker_init(struct drm_device *dev) > +int msm_gem_shrinker_init(struct drm_device *dev) > { > struct msm_drm_private *priv = dev->dev_private; > - priv->shrinker.count_objects = msm_gem_shrinker_count; > - priv->shrinker.scan_objects = msm_gem_shrinker_scan; > - priv->shrinker.seeks = DEFAULT_SEEKS; > - WARN_ON(register_shrinker(&priv->shrinker, "drm-msm_gem")); > + > + priv->shrinker = shrinker_alloc(0, "drm-msm_gem"); > + if (!priv->shrinker) { Just "if (WARN_ON(!priv->shrinker))" > + WARN_ON(1); > + return -ENOMEM; > + } > + > + priv->shrinker->count_objects = msm_gem_shrinker_count; > + priv->shrinker->scan_objects = msm_gem_shrinker_scan; > + priv->shrinker->seeks = DEFAULT_SEEKS; > + priv->shrinker->private_data = priv; > + > + shrinker_register(priv->shrinker); > > priv->vmap_notifier.notifier_call = msm_gem_shrinker_vmap; > WARN_ON(register_vmap_purge_notifier(&priv->vmap_notifier)); > + > + return 0; > } > > /** > @@ -232,8 +242,8 @@ void msm_gem_shrinker_cleanup(struct drm_device *dev) > { > struct msm_drm_private *priv = dev->dev_private; > > - if (priv->shrinker.nr_deferred) { > + if (priv->shrinker) { > WARN_ON(unregister_vmap_purge_notifier(&priv->vmap_notifier)); > - unregister_shrinker(&priv->shrinker); > + shrinker_unregister(priv->shrinker); > } > }
On 2023/7/26 15:24, Muchun Song wrote: > > > On 2023/7/24 17:43, Qi Zheng wrote: >> In preparation for implementing lockless slab shrink, use new APIs to >> dynamically allocate the drm-msm_gem shrinker, so that it can be freed >> asynchronously using kfree_rcu(). Then it doesn't need to wait for RCU >> read-side critical section when releasing the struct msm_drm_private. >> >> Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com> > > Reviewed-by: Muchun Song <songmuchun@bytedance.com> > > A nit bellow. > >> --- >> drivers/gpu/drm/msm/msm_drv.c | 4 ++- >> drivers/gpu/drm/msm/msm_drv.h | 4 +-- >> drivers/gpu/drm/msm/msm_gem_shrinker.c | 36 ++++++++++++++++---------- >> 3 files changed, 28 insertions(+), 16 deletions(-) >> >> diff --git a/drivers/gpu/drm/msm/msm_drv.c >> b/drivers/gpu/drm/msm/msm_drv.c >> index 891eff8433a9..7f6933be703f 100644 >> --- a/drivers/gpu/drm/msm/msm_drv.c >> +++ b/drivers/gpu/drm/msm/msm_drv.c >> @@ -461,7 +461,9 @@ static int msm_drm_init(struct device *dev, const >> struct drm_driver *drv) >> if (ret) >> goto err_msm_uninit; >> - msm_gem_shrinker_init(ddev); >> + ret = msm_gem_shrinker_init(ddev); >> + if (ret) >> + goto err_msm_uninit; >> if (priv->kms_init) { >> ret = priv->kms_init(ddev); >> diff --git a/drivers/gpu/drm/msm/msm_drv.h >> b/drivers/gpu/drm/msm/msm_drv.h >> index e13a8cbd61c9..84523d4a1e58 100644 >> --- a/drivers/gpu/drm/msm/msm_drv.h >> +++ b/drivers/gpu/drm/msm/msm_drv.h >> @@ -217,7 +217,7 @@ struct msm_drm_private { >> } vram; >> struct notifier_block vmap_notifier; >> - struct shrinker shrinker; >> + struct shrinker *shrinker; >> struct drm_atomic_state *pm_state; >> @@ -279,7 +279,7 @@ int msm_ioctl_gem_submit(struct drm_device *dev, >> void *data, >> unsigned long msm_gem_shrinker_shrink(struct drm_device *dev, >> unsigned long nr_to_scan); >> #endif >> -void msm_gem_shrinker_init(struct drm_device *dev); >> +int msm_gem_shrinker_init(struct drm_device *dev); >> void msm_gem_shrinker_cleanup(struct drm_device *dev); >> int msm_gem_prime_mmap(struct drm_gem_object *obj, struct >> vm_area_struct *vma); >> diff --git a/drivers/gpu/drm/msm/msm_gem_shrinker.c >> b/drivers/gpu/drm/msm/msm_gem_shrinker.c >> index f38296ad8743..7daab1298c11 100644 >> --- a/drivers/gpu/drm/msm/msm_gem_shrinker.c >> +++ b/drivers/gpu/drm/msm/msm_gem_shrinker.c >> @@ -34,8 +34,7 @@ static bool can_block(struct shrink_control *sc) >> static unsigned long >> msm_gem_shrinker_count(struct shrinker *shrinker, struct >> shrink_control *sc) >> { >> - struct msm_drm_private *priv = >> - container_of(shrinker, struct msm_drm_private, shrinker); >> + struct msm_drm_private *priv = shrinker->private_data; >> unsigned count = priv->lru.dontneed.count; >> if (can_swap()) >> @@ -100,8 +99,7 @@ active_evict(struct drm_gem_object *obj) >> static unsigned long >> msm_gem_shrinker_scan(struct shrinker *shrinker, struct >> shrink_control *sc) >> { >> - struct msm_drm_private *priv = >> - container_of(shrinker, struct msm_drm_private, shrinker); >> + struct msm_drm_private *priv = shrinker->private_data; >> struct { >> struct drm_gem_lru *lru; >> bool (*shrink)(struct drm_gem_object *obj); >> @@ -148,10 +146,11 @@ msm_gem_shrinker_shrink(struct drm_device *dev, >> unsigned long nr_to_scan) >> struct shrink_control sc = { >> .nr_to_scan = nr_to_scan, >> }; >> - int ret; >> + unsigned long ret = SHRINK_STOP; >> fs_reclaim_acquire(GFP_KERNEL); >> - ret = msm_gem_shrinker_scan(&priv->shrinker, &sc); >> + if (priv->shrinker) >> + ret = msm_gem_shrinker_scan(priv->shrinker, &sc); >> fs_reclaim_release(GFP_KERNEL); >> return ret; >> @@ -210,16 +209,27 @@ msm_gem_shrinker_vmap(struct notifier_block *nb, >> unsigned long event, void *ptr) >> * >> * This function registers and sets up the msm shrinker. >> */ >> -void msm_gem_shrinker_init(struct drm_device *dev) >> +int msm_gem_shrinker_init(struct drm_device *dev) >> { >> struct msm_drm_private *priv = dev->dev_private; >> - priv->shrinker.count_objects = msm_gem_shrinker_count; >> - priv->shrinker.scan_objects = msm_gem_shrinker_scan; >> - priv->shrinker.seeks = DEFAULT_SEEKS; >> - WARN_ON(register_shrinker(&priv->shrinker, "drm-msm_gem")); >> + >> + priv->shrinker = shrinker_alloc(0, "drm-msm_gem"); >> + if (!priv->shrinker) { > > Just "if (WARN_ON(!priv->shrinker))" As suggested by Steven Pric in patch #24, this warning is unnecessary, so I will remove it in the next version. > >> + WARN_ON(1); >> + return -ENOMEM; >> + } >> + >> + priv->shrinker->count_objects = msm_gem_shrinker_count; >> + priv->shrinker->scan_objects = msm_gem_shrinker_scan; >> + priv->shrinker->seeks = DEFAULT_SEEKS; >> + priv->shrinker->private_data = priv; >> + >> + shrinker_register(priv->shrinker); >> priv->vmap_notifier.notifier_call = msm_gem_shrinker_vmap; >> WARN_ON(register_vmap_purge_notifier(&priv->vmap_notifier)); >> + >> + return 0; >> } >> /** >> @@ -232,8 +242,8 @@ void msm_gem_shrinker_cleanup(struct drm_device *dev) >> { >> struct msm_drm_private *priv = dev->dev_private; >> - if (priv->shrinker.nr_deferred) { >> + if (priv->shrinker) { >> WARN_ON(unregister_vmap_purge_notifier(&priv->vmap_notifier)); >> - unregister_shrinker(&priv->shrinker); >> + shrinker_unregister(priv->shrinker); >> } >> } >
diff --git a/drivers/gpu/drm/msm/msm_drv.c b/drivers/gpu/drm/msm/msm_drv.c index 891eff8433a9..7f6933be703f 100644 --- a/drivers/gpu/drm/msm/msm_drv.c +++ b/drivers/gpu/drm/msm/msm_drv.c @@ -461,7 +461,9 @@ static int msm_drm_init(struct device *dev, const struct drm_driver *drv) if (ret) goto err_msm_uninit; - msm_gem_shrinker_init(ddev); + ret = msm_gem_shrinker_init(ddev); + if (ret) + goto err_msm_uninit; if (priv->kms_init) { ret = priv->kms_init(ddev); diff --git a/drivers/gpu/drm/msm/msm_drv.h b/drivers/gpu/drm/msm/msm_drv.h index e13a8cbd61c9..84523d4a1e58 100644 --- a/drivers/gpu/drm/msm/msm_drv.h +++ b/drivers/gpu/drm/msm/msm_drv.h @@ -217,7 +217,7 @@ struct msm_drm_private { } vram; struct notifier_block vmap_notifier; - struct shrinker shrinker; + struct shrinker *shrinker; struct drm_atomic_state *pm_state; @@ -279,7 +279,7 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data, unsigned long msm_gem_shrinker_shrink(struct drm_device *dev, unsigned long nr_to_scan); #endif -void msm_gem_shrinker_init(struct drm_device *dev); +int msm_gem_shrinker_init(struct drm_device *dev); void msm_gem_shrinker_cleanup(struct drm_device *dev); int msm_gem_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma); diff --git a/drivers/gpu/drm/msm/msm_gem_shrinker.c b/drivers/gpu/drm/msm/msm_gem_shrinker.c index f38296ad8743..7daab1298c11 100644 --- a/drivers/gpu/drm/msm/msm_gem_shrinker.c +++ b/drivers/gpu/drm/msm/msm_gem_shrinker.c @@ -34,8 +34,7 @@ static bool can_block(struct shrink_control *sc) static unsigned long msm_gem_shrinker_count(struct shrinker *shrinker, struct shrink_control *sc) { - struct msm_drm_private *priv = - container_of(shrinker, struct msm_drm_private, shrinker); + struct msm_drm_private *priv = shrinker->private_data; unsigned count = priv->lru.dontneed.count; if (can_swap()) @@ -100,8 +99,7 @@ active_evict(struct drm_gem_object *obj) static unsigned long msm_gem_shrinker_scan(struct shrinker *shrinker, struct shrink_control *sc) { - struct msm_drm_private *priv = - container_of(shrinker, struct msm_drm_private, shrinker); + struct msm_drm_private *priv = shrinker->private_data; struct { struct drm_gem_lru *lru; bool (*shrink)(struct drm_gem_object *obj); @@ -148,10 +146,11 @@ msm_gem_shrinker_shrink(struct drm_device *dev, unsigned long nr_to_scan) struct shrink_control sc = { .nr_to_scan = nr_to_scan, }; - int ret; + unsigned long ret = SHRINK_STOP; fs_reclaim_acquire(GFP_KERNEL); - ret = msm_gem_shrinker_scan(&priv->shrinker, &sc); + if (priv->shrinker) + ret = msm_gem_shrinker_scan(priv->shrinker, &sc); fs_reclaim_release(GFP_KERNEL); return ret; @@ -210,16 +209,27 @@ msm_gem_shrinker_vmap(struct notifier_block *nb, unsigned long event, void *ptr) * * This function registers and sets up the msm shrinker. */ -void msm_gem_shrinker_init(struct drm_device *dev) +int msm_gem_shrinker_init(struct drm_device *dev) { struct msm_drm_private *priv = dev->dev_private; - priv->shrinker.count_objects = msm_gem_shrinker_count; - priv->shrinker.scan_objects = msm_gem_shrinker_scan; - priv->shrinker.seeks = DEFAULT_SEEKS; - WARN_ON(register_shrinker(&priv->shrinker, "drm-msm_gem")); + + priv->shrinker = shrinker_alloc(0, "drm-msm_gem"); + if (!priv->shrinker) { + WARN_ON(1); + return -ENOMEM; + } + + priv->shrinker->count_objects = msm_gem_shrinker_count; + priv->shrinker->scan_objects = msm_gem_shrinker_scan; + priv->shrinker->seeks = DEFAULT_SEEKS; + priv->shrinker->private_data = priv; + + shrinker_register(priv->shrinker); priv->vmap_notifier.notifier_call = msm_gem_shrinker_vmap; WARN_ON(register_vmap_purge_notifier(&priv->vmap_notifier)); + + return 0; } /** @@ -232,8 +242,8 @@ void msm_gem_shrinker_cleanup(struct drm_device *dev) { struct msm_drm_private *priv = dev->dev_private; - if (priv->shrinker.nr_deferred) { + if (priv->shrinker) { WARN_ON(unregister_vmap_purge_notifier(&priv->vmap_notifier)); - unregister_shrinker(&priv->shrinker); + shrinker_unregister(priv->shrinker); } }
In preparation for implementing lockless slab shrink, use new APIs to dynamically allocate the drm-msm_gem shrinker, so that it can be freed asynchronously using kfree_rcu(). Then it doesn't need to wait for RCU read-side critical section when releasing the struct msm_drm_private. Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com> --- drivers/gpu/drm/msm/msm_drv.c | 4 ++- drivers/gpu/drm/msm/msm_drv.h | 4 +-- drivers/gpu/drm/msm/msm_gem_shrinker.c | 36 ++++++++++++++++---------- 3 files changed, 28 insertions(+), 16 deletions(-)