diff options
author | Tejun Heo <tj@kernel.org> | 2019-11-04 15:54:30 -0800 |
---|---|---|
committer | Tejun Heo <tj@kernel.org> | 2019-11-12 08:18:04 -0800 |
commit | 743210386c0354a2f8ef3d697353c7d8477fa81d (patch) | |
tree | ac8071135c3e67ff425391540eb88c7dc343afb7 /kernel | |
parent | 40430452fd5da1509177ac597b394614cd3a121f (diff) | |
download | linux-stable-743210386c0354a2f8ef3d697353c7d8477fa81d.tar.gz linux-stable-743210386c0354a2f8ef3d697353c7d8477fa81d.tar.bz2 linux-stable-743210386c0354a2f8ef3d697353c7d8477fa81d.zip |
cgroup: use cgrp->kn->id as the cgroup ID
cgroup ID is currently allocated using a dedicated per-hierarchy idr
and used internally and exposed through tracepoints and bpf. This is
confusing because there are tracepoints and other interfaces which use
the cgroupfs ino as IDs.
The preceding changes made kn->id exposed as ino as 64bit ino on
supported archs or ino+gen (low 32bits as ino, high gen). There's no
reason for cgroup to use different IDs. The kernfs IDs are unique and
userland can easily discover them and map them back to paths using
standard file operations.
This patch replaces cgroup IDs with kernfs IDs.
* cgroup_id() is added and all cgroup ID users are converted to use it.
* kernfs_node creation is moved to earlier during cgroup init so that
cgroup_id() is available during init.
* While at it, s/cgroup/cgrp/ in psi helpers for consistency.
* Fallback ID value is changed to 1 to be consistent with root cgroup
ID.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Diffstat (limited to 'kernel')
-rw-r--r-- | kernel/bpf/helpers.c | 2 | ||||
-rw-r--r-- | kernel/bpf/local_storage.c | 2 | ||||
-rw-r--r-- | kernel/cgroup/cgroup.c | 76 | ||||
-rw-r--r-- | kernel/trace/blktrace.c | 4 |
4 files changed, 29 insertions, 55 deletions
diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c index 912e761cd17a..cada974c9f4e 100644 --- a/kernel/bpf/helpers.c +++ b/kernel/bpf/helpers.c @@ -317,7 +317,7 @@ BPF_CALL_0(bpf_get_current_cgroup_id) { struct cgroup *cgrp = task_dfl_cgroup(current); - return cgrp->kn->id; + return cgroup_id(cgrp); } const struct bpf_func_proto bpf_get_current_cgroup_id_proto = { diff --git a/kernel/bpf/local_storage.c b/kernel/bpf/local_storage.c index 5d867f6d7204..2ba750725cb2 100644 --- a/kernel/bpf/local_storage.c +++ b/kernel/bpf/local_storage.c @@ -569,7 +569,7 @@ void bpf_cgroup_storage_link(struct bpf_cgroup_storage *storage, return; storage->key.attach_type = type; - storage->key.cgroup_inode_id = cgroup->kn->id; + storage->key.cgroup_inode_id = cgroup_id(cgroup); map = storage->map; diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c index b5dcbee5aa6c..c12dcf7dc432 100644 --- a/kernel/cgroup/cgroup.c +++ b/kernel/cgroup/cgroup.c @@ -1308,10 +1308,7 @@ static void cgroup_exit_root_id(struct cgroup_root *root) void cgroup_free_root(struct cgroup_root *root) { - if (root) { - idr_destroy(&root->cgroup_idr); - kfree(root); - } + kfree(root); } static void cgroup_destroy_root(struct cgroup_root *root) @@ -1917,7 +1914,6 @@ void init_cgroup_root(struct cgroup_fs_context *ctx) atomic_set(&root->nr_cgrps, 1); cgrp->root = root; init_cgroup_housekeeping(cgrp); - idr_init(&root->cgroup_idr); root->flags = ctx->flags; if (ctx->release_agent) @@ -1938,12 +1934,6 @@ int cgroup_setup_root(struct cgroup_root *root, u16 ss_mask) lockdep_assert_held(&cgroup_mutex); - ret = cgroup_idr_alloc(&root->cgroup_idr, root_cgrp, 1, 2, GFP_KERNEL); - if (ret < 0) - goto out; - root_cgrp->id = ret; - root_cgrp->ancestor_ids[0] = ret; - ret = percpu_ref_init(&root_cgrp->self.refcnt, css_release, 0, GFP_KERNEL); if (ret) @@ -1976,6 +1966,8 @@ int cgroup_setup_root(struct cgroup_root *root, u16 ss_mask) goto exit_root_id; } root_cgrp->kn = root->kf_root->kn; + WARN_ON_ONCE(cgroup_id(root_cgrp) != 1); + root_cgrp->ancestor_ids[0] = cgroup_id(root_cgrp); ret = css_populate_dir(&root_cgrp->self); if (ret) @@ -3552,22 +3544,22 @@ static int cpu_stat_show(struct seq_file *seq, void *v) #ifdef CONFIG_PSI static int cgroup_io_pressure_show(struct seq_file *seq, void *v) { - struct cgroup *cgroup = seq_css(seq)->cgroup; - struct psi_group *psi = cgroup->id == 1 ? &psi_system : &cgroup->psi; + struct cgroup *cgrp = seq_css(seq)->cgroup; + struct psi_group *psi = cgroup_id(cgrp) == 1 ? &psi_system : &cgrp->psi; return psi_show(seq, psi, PSI_IO); } static int cgroup_memory_pressure_show(struct seq_file *seq, void *v) { - struct cgroup *cgroup = seq_css(seq)->cgroup; - struct psi_group *psi = cgroup->id == 1 ? &psi_system : &cgroup->psi; + struct cgroup *cgrp = seq_css(seq)->cgroup; + struct psi_group *psi = cgroup_id(cgrp) == 1 ? &psi_system : &cgrp->psi; return psi_show(seq, psi, PSI_MEM); } static int cgroup_cpu_pressure_show(struct seq_file *seq, void *v) { - struct cgroup *cgroup = seq_css(seq)->cgroup; - struct psi_group *psi = cgroup->id == 1 ? &psi_system : &cgroup->psi; + struct cgroup *cgrp = seq_css(seq)->cgroup; + struct psi_group *psi = cgroup_id(cgrp) == 1 ? &psi_system : &cgrp->psi; return psi_show(seq, psi, PSI_CPU); } @@ -4987,9 +4979,6 @@ static void css_release_work_fn(struct work_struct *work) tcgrp->nr_dying_descendants--; spin_unlock_irq(&css_set_lock); - cgroup_idr_remove(&cgrp->root->cgroup_idr, cgrp->id); - cgrp->id = -1; - /* * There are two control paths which try to determine * cgroup from dentry without going through kernfs - @@ -5154,10 +5143,12 @@ err_free_css: * it isn't associated with its kernfs_node and doesn't have the control * mask applied. */ -static struct cgroup *cgroup_create(struct cgroup *parent) +static struct cgroup *cgroup_create(struct cgroup *parent, const char *name, + umode_t mode) { struct cgroup_root *root = parent->root; struct cgroup *cgrp, *tcgrp; + struct kernfs_node *kn; int level = parent->level + 1; int ret; @@ -5177,15 +5168,13 @@ static struct cgroup *cgroup_create(struct cgroup *parent) goto out_cancel_ref; } - /* - * Temporarily set the pointer to NULL, so idr_find() won't return - * a half-baked cgroup. - */ - cgrp->id = cgroup_idr_alloc(&root->cgroup_idr, NULL, 2, 0, GFP_KERNEL); - if (cgrp->id < 0) { - ret = -ENOMEM; + /* create the directory */ + kn = kernfs_create_dir(parent->kn, name, mode, cgrp); + if (IS_ERR(kn)) { + ret = PTR_ERR(kn); goto out_stat_exit; } + cgrp->kn = kn; init_cgroup_housekeeping(cgrp); @@ -5195,7 +5184,7 @@ static struct cgroup *cgroup_create(struct cgroup *parent) ret = psi_cgroup_alloc(cgrp); if (ret) - goto out_idr_free; + goto out_kernfs_remove; ret = cgroup_bpf_inherit(cgrp); if (ret) @@ -5219,7 +5208,7 @@ static struct cgroup *cgroup_create(struct cgroup *parent) spin_lock_irq(&css_set_lock); for (tcgrp = cgrp; tcgrp; tcgrp = cgroup_parent(tcgrp)) { - cgrp->ancestor_ids[tcgrp->level] = tcgrp->id; + cgrp->ancestor_ids[tcgrp->level] = cgroup_id(tcgrp); if (tcgrp != cgrp) { tcgrp->nr_descendants++; @@ -5249,12 +5238,6 @@ static struct cgroup *cgroup_create(struct cgroup *parent) cgroup_get_live(parent); /* - * @cgrp is now fully operational. If something fails after this - * point, it'll be released via the normal destruction path. - */ - cgroup_idr_replace(&root->cgroup_idr, cgrp, cgrp->id); - - /* * On the default hierarchy, a child doesn't automatically inherit * subtree_control from the parent. Each is configured manually. */ @@ -5267,8 +5250,8 @@ static struct cgroup *cgroup_create(struct cgroup *parent) out_psi_free: psi_cgroup_free(cgrp); -out_idr_free: - cgroup_idr_remove(&root->cgroup_idr, cgrp->id); +out_kernfs_remove: + kernfs_remove(cgrp->kn); out_stat_exit: if (cgroup_on_dfl(parent)) cgroup_rstat_exit(cgrp); @@ -5305,7 +5288,6 @@ fail: int cgroup_mkdir(struct kernfs_node *parent_kn, const char *name, umode_t mode) { struct cgroup *parent, *cgrp; - struct kernfs_node *kn; int ret; /* do not accept '\n' to prevent making /proc/<pid>/cgroup unparsable */ @@ -5321,27 +5303,19 @@ int cgroup_mkdir(struct kernfs_node *parent_kn, const char *name, umode_t mode) goto out_unlock; } - cgrp = cgroup_create(parent); + cgrp = cgroup_create(parent, name, mode); if (IS_ERR(cgrp)) { ret = PTR_ERR(cgrp); goto out_unlock; } - /* create the directory */ - kn = kernfs_create_dir(parent->kn, name, mode, cgrp); - if (IS_ERR(kn)) { - ret = PTR_ERR(kn); - goto out_destroy; - } - cgrp->kn = kn; - /* * This extra ref will be put in cgroup_free_fn() and guarantees * that @cgrp->kn is always accessible. */ - kernfs_get(kn); + kernfs_get(cgrp->kn); - ret = cgroup_kn_set_ugid(kn); + ret = cgroup_kn_set_ugid(cgrp->kn); if (ret) goto out_destroy; @@ -5356,7 +5330,7 @@ int cgroup_mkdir(struct kernfs_node *parent_kn, const char *name, umode_t mode) TRACE_CGROUP_PATH(mkdir, cgrp); /* let's create and online css's */ - kernfs_activate(kn); + kernfs_activate(cgrp->kn); ret = 0; goto out_unlock; diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index a7dac5b63f3f..475e29498bca 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -171,7 +171,7 @@ void __trace_note_message(struct blk_trace *bt, struct blkcg *blkcg, blkcg = NULL; #ifdef CONFIG_BLK_CGROUP trace_note(bt, 0, BLK_TN_MESSAGE, buf, n, - blkcg ? cgroup_get_kernfs_id(blkcg->css.cgroup) : 0); + blkcg ? cgroup_id(blkcg->css.cgroup) : 1); #else trace_note(bt, 0, BLK_TN_MESSAGE, buf, n, 0); #endif @@ -759,7 +759,7 @@ static u64 blk_trace_bio_get_cgid(struct request_queue *q, struct bio *bio) if (!bio->bi_blkg) return 0; - return cgroup_get_kernfs_id(bio_blkcg(bio)->css.cgroup); + return cgroup_id(bio_blkcg(bio)->css.cgroup); } #else u64 blk_trace_bio_get_cgid(struct request_queue *q, struct bio *bio) |