From 51f620fcc419913315623e81b69a4f2bc59100a8 Mon Sep 17 00:00:00 2001 From: Shakeel Butt Date: Sat, 19 Dec 2020 20:46:08 -0800 Subject: [PATCH] inotify, memcg: account inotify instances to kmemcg [ Upstream commit ac7b79fd190b02e7151bc7d2b9da692f537657f3 ] Currently the fs sysctl inotify/max_user_instances is used to limit the number of inotify instances on the system. For systems running multiple workloads, the per-user namespace sysctl max_inotify_instances can be used to further partition inotify instances. However there is no easy way to set a sensible system level max limit on inotify instances and further partition it between the workloads. It is much easier to charge the underlying resource (i.e. memory) behind the inotify instances to the memcg of the workload and let their memory limits limit the number of inotify instances they can create. With inotify instances charged to memcg, the admin can simply set max_user_instances to INT_MAX and let the memcg limits of the jobs limit their inotify instances. Link: https://lore.kernel.org/r/20201220044608.1258123-1-shakeelb@google.com Reviewed-by: Amir Goldstein Signed-off-by: Shakeel Butt Signed-off-by: Jan Kara Signed-off-by: Chuck Lever Signed-off-by: Sasha Levin --- fs/notify/fanotify/fanotify_user.c | 2 +- fs/notify/group.c | 25 ++++++++++++++++++++----- fs/notify/inotify/inotify_user.c | 4 ++-- include/linux/fsnotify_backend.h | 1 + 4 files changed, 24 insertions(+), 8 deletions(-) diff --git a/fs/notify/fanotify/fanotify_user.c b/fs/notify/fanotify/fanotify_user.c index 84de9f97bbc0..3e905b2e1b9c 100644 --- a/fs/notify/fanotify/fanotify_user.c +++ b/fs/notify/fanotify/fanotify_user.c @@ -976,7 +976,7 @@ SYSCALL_DEFINE2(fanotify_init, unsigned int, flags, unsigned int, event_f_flags) f_flags |= O_NONBLOCK; /* fsnotify_alloc_group takes a ref. Dropped in fanotify_release */ - group = fsnotify_alloc_group(&fanotify_fsnotify_ops); + group = fsnotify_alloc_user_group(&fanotify_fsnotify_ops); if (IS_ERR(group)) { free_uid(user); return PTR_ERR(group); diff --git a/fs/notify/group.c b/fs/notify/group.c index a4a4b1c64d32..ffd723ffe46d 100644 --- a/fs/notify/group.c +++ b/fs/notify/group.c @@ -111,14 +111,12 @@ void fsnotify_put_group(struct fsnotify_group *group) } EXPORT_SYMBOL_GPL(fsnotify_put_group); -/* - * Create a new fsnotify_group and hold a reference for the group returned. - */ -struct fsnotify_group *fsnotify_alloc_group(const struct fsnotify_ops *ops) +static struct fsnotify_group *__fsnotify_alloc_group( + const struct fsnotify_ops *ops, gfp_t gfp) { struct fsnotify_group *group; - group = kzalloc(sizeof(struct fsnotify_group), GFP_KERNEL); + group = kzalloc(sizeof(struct fsnotify_group), gfp); if (!group) return ERR_PTR(-ENOMEM); @@ -139,8 +137,25 @@ struct fsnotify_group *fsnotify_alloc_group(const struct fsnotify_ops *ops) return group; } + +/* + * Create a new fsnotify_group and hold a reference for the group returned. + */ +struct fsnotify_group *fsnotify_alloc_group(const struct fsnotify_ops *ops) +{ + return __fsnotify_alloc_group(ops, GFP_KERNEL); +} EXPORT_SYMBOL_GPL(fsnotify_alloc_group); +/* + * Create a new fsnotify_group and hold a reference for the group returned. + */ +struct fsnotify_group *fsnotify_alloc_user_group(const struct fsnotify_ops *ops) +{ + return __fsnotify_alloc_group(ops, GFP_KERNEL_ACCOUNT); +} +EXPORT_SYMBOL_GPL(fsnotify_alloc_user_group); + int fsnotify_fasync(int fd, struct file *file, int on) { struct fsnotify_group *group = file->private_data; diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index b564a32403aa..ad8fb4bca6dc 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -632,11 +632,11 @@ static struct fsnotify_group *inotify_new_group(unsigned int max_events) struct fsnotify_group *group; struct inotify_event_info *oevent; - group = fsnotify_alloc_group(&inotify_fsnotify_ops); + group = fsnotify_alloc_user_group(&inotify_fsnotify_ops); if (IS_ERR(group)) return group; - oevent = kmalloc(sizeof(struct inotify_event_info), GFP_KERNEL); + oevent = kmalloc(sizeof(struct inotify_event_info), GFP_KERNEL_ACCOUNT); if (unlikely(!oevent)) { fsnotify_destroy_group(group); return ERR_PTR(-ENOMEM); diff --git a/include/linux/fsnotify_backend.h b/include/linux/fsnotify_backend.h index a2e42d3cd87c..e5409b83e731 100644 --- a/include/linux/fsnotify_backend.h +++ b/include/linux/fsnotify_backend.h @@ -470,6 +470,7 @@ static inline void fsnotify_update_flags(struct dentry *dentry) /* create a new group */ extern struct fsnotify_group *fsnotify_alloc_group(const struct fsnotify_ops *ops); +extern struct fsnotify_group *fsnotify_alloc_user_group(const struct fsnotify_ops *ops); /* get reference to a group */ extern void fsnotify_get_group(struct fsnotify_group *group); /* drop reference on a group from fsnotify_alloc_group */