A few updates to the ipc code, mostly some very needed cleanups (yes, we still
have loads more to do, but it's a start). Nothing particularly interesting,
except perhaps for the last two patches, which include a locking optimization
and barrier documentation for message queues. Absolutely no functional changes.
Applies on top of linux-next + Manfred's 'whitespace cleanup' patch. Tested
with LTP.
Thanks!
Davidlohr Bueso (7):
ipc: standardize code comments
ipc: remove braces for single statements
ipc: remove useless return statement
ipc: simplify sysvipc_proc_open return
ipc: delete seq_max field in struct ipc_ids
ipc: share ids rwsem when possible in ipcget_public
ipc,msg: document barriers
include/linux/ipc_namespace.h | 1 -
ipc/compat.c | 18 +--
ipc/mqueue.c | 6 +-
ipc/msg.c | 19 ++-
ipc/sem.c | 32 ++---
ipc/shm.c | 1 -
ipc/util.c | 328 ++++++++++++++++++++----------------------
ipc/util.h | 1 +
8 files changed, 198 insertions(+), 208 deletions(-)
--
1.8.1.4
Only found in ipc_rmid().
Signed-off-by: Davidlohr Bueso <[email protected]>
---
ipc/util.c | 4 ----
1 file changed, 4 deletions(-)
diff --git a/ipc/util.c b/ipc/util.c
index cfbd8fa..7afe7de 100644
--- a/ipc/util.c
+++ b/ipc/util.c
@@ -439,12 +439,8 @@ void ipc_rmid(struct ipc_ids *ids, struct kern_ipc_perm *ipcp)
int lid = ipcid_to_idx(ipcp->id);
idr_remove(&ids->ipcs_idr, lid);
-
ids->in_use--;
-
ipcp->deleted = true;
-
- return;
}
/**
--
1.8.1.4
Get rid of silly/useless label jumping.
Signed-off-by: Davidlohr Bueso <[email protected]>
---
ipc/util.c | 9 ++++-----
1 file changed, 4 insertions(+), 5 deletions(-)
diff --git a/ipc/util.c b/ipc/util.c
index 7afe7de..cecb46e 100644
--- a/ipc/util.c
+++ b/ipc/util.c
@@ -911,8 +911,10 @@ static int sysvipc_proc_open(struct inode *inode, struct file *file)
goto out;
ret = seq_open(file, &sysvipc_proc_seqops);
- if (ret)
- goto out_kfree;
+ if (ret) {
+ kfree(iter);
+ goto out;
+ }
seq = file->private_data;
seq->private = iter;
@@ -921,9 +923,6 @@ static int sysvipc_proc_open(struct inode *inode, struct file *file)
iter->ns = get_ipc_ns(current->nsproxy->ipc_ns);
out:
return ret;
-out_kfree:
- kfree(iter);
- goto out;
}
static int sysvipc_proc_release(struct inode *inode, struct file *file)
--
1.8.1.4
... and rewrite the function. For scenarios where the key is found and we end
up just doing different routinary checks, we can downgrade the ids->rwsem and
share it among concurrent readers. These checks include the following, which
are all safe to share the lock:
ops->more_checks() >> sem_more_checks(), shm_more_checks()
ipc_check_perms() >> ipcperms(),ops->associate() >> [lsm]_[ipctype]_associate()
Signed-off-by: Davidlohr Bueso <[email protected]>
---
ipc/util.c | 60 +++++++++++++++++++++++++++++++++---------------------------
1 file changed, 33 insertions(+), 27 deletions(-)
diff --git a/ipc/util.c b/ipc/util.c
index e1b4c6d..fc5c655 100644
--- a/ipc/util.c
+++ b/ipc/util.c
@@ -183,7 +183,7 @@ void __init ipc_init_proc_interface(const char *path, const char *header,
* ipc_findkey - find a key in an ipc identifier set
* @ids: ipc identifier set
* @key: key to find
- *
+ *
* Returns the locked pointer to the ipc structure if found or NULL
* otherwise. If key is found ipc points to the owning ipc structure
*
@@ -375,48 +375,54 @@ static int ipc_check_perms(struct ipc_namespace *ns,
* On success, the ipc id is returned.
*/
static int ipcget_public(struct ipc_namespace *ns, struct ipc_ids *ids,
- struct ipc_ops *ops, struct ipc_params *params)
+ struct ipc_ops *ops, struct ipc_params *params)
{
struct kern_ipc_perm *ipcp;
int flg = params->flg;
- int err;
+ int err = 0;
- /*
- * Take the lock as a writer since we are potentially going to add
- * a new entry + read locks are not "upgradable"
- */
down_write(&ids->rwsem);
ipcp = ipc_findkey(ids, params->key);
- if (ipcp == NULL) {
+
+ if (!ipcp) {
/* key not used */
if (!(flg & IPC_CREAT))
err = -ENOENT;
- else
+ else /* create new ipc object */
err = ops->getnew(ns, params);
- } else {
- /* ipc object has been locked by ipc_findkey() */
-
- if (flg & IPC_CREAT && flg & IPC_EXCL)
- err = -EEXIST;
- else {
- err = 0;
- if (ops->more_checks)
- err = ops->more_checks(ipcp, params);
- if (!err)
- /*
- * ipc_check_perms returns the IPC id on
- * success
- */
- err = ipc_check_perms(ns, ipcp, ops, params);
- }
+
+ goto done_write;
+ }
+
+ if ((flg & IPC_CREAT) && (flg & IPC_EXCL)) {
+ /* ipc object was locked by successful ipc_findkey() lookup */
ipc_unlock(ipcp);
+ err = -ENOENT;
+
+ goto done_write;
}
- up_write(&ids->rwsem);
+ /*
+ * The key was found, so we will just perform routinary checks on
+ * ipc the object. Share the lock among other readers.
+ */
+ downgrade_write(&ids->rwsem);
+
+ if (ops->more_checks)
+ err = ops->more_checks(ipcp, params);
+ if (!err)
+ /* returns the IPC id on success */
+ err = ipc_check_perms(ns, ipcp, ops, params);
+
+ ipc_unlock(ipcp);
+
+ up_read(&ids->rwsem);
+ return err;
+done_write:
+ up_write(&ids->rwsem);
return err;
}
-
/**
* ipc_rmid - remove an ipc identifier
* @ids: ipc identifier set
--
1.8.1.4
This field is only used to reset the ids seq number if it exceeds the smaller of
INT_MAX/SEQ_MULTIPLIER and USHRT_MAX, and can therefore be moved out of the
structure and into its own macro. Since each ipc_namespace contains a table of
3 pointers to struct ipc_ids we can save space in instruction text:
text data bss dec hex filename
56232 2348 24 58604 e4ec ipc/built-in.o
56216 2348 24 58588 e4dc ipc/built-in.o-after
Reviewed-by: Jonathan Gonzalez <[email protected]>
Signed-off-by: Davidlohr Bueso <[email protected]>
---
include/linux/ipc_namespace.h | 1 -
ipc/util.c | 13 ++-----------
ipc/util.h | 1 +
3 files changed, 3 insertions(+), 12 deletions(-)
diff --git a/include/linux/ipc_namespace.h b/include/linux/ipc_namespace.h
index f6c82de..e7831d2 100644
--- a/include/linux/ipc_namespace.h
+++ b/include/linux/ipc_namespace.h
@@ -21,7 +21,6 @@ struct user_namespace;
struct ipc_ids {
int in_use;
unsigned short seq;
- unsigned short seq_max;
struct rw_semaphore rwsem;
struct idr ipcs_idr;
int next_id;
diff --git a/ipc/util.c b/ipc/util.c
index cecb46e..e1b4c6d 100644
--- a/ipc/util.c
+++ b/ipc/util.c
@@ -139,19 +139,10 @@ __initcall(ipc_init);
*/
void ipc_init_ids(struct ipc_ids *ids)
{
- init_rwsem(&ids->rwsem);
-
ids->in_use = 0;
ids->seq = 0;
ids->next_id = -1;
- {
- int seq_limit = INT_MAX/SEQ_MULTIPLIER;
- if (seq_limit > USHRT_MAX)
- ids->seq_max = USHRT_MAX;
- else
- ids->seq_max = seq_limit;
- }
-
+ init_rwsem(&ids->rwsem);
idr_init(&ids->ipcs_idr);
}
@@ -304,7 +295,7 @@ int ipc_addid(struct ipc_ids *ids, struct kern_ipc_perm *new, int size)
if (next_id < 0) {
new->seq = ids->seq++;
- if (ids->seq > ids->seq_max)
+ if (ids->seq > IPCID_SEQ_MAX)
ids->seq = 0;
} else {
new->seq = ipcid_to_seqx(next_id);
diff --git a/ipc/util.h b/ipc/util.h
index d64db3e..9c47d6f 100644
--- a/ipc/util.h
+++ b/ipc/util.h
@@ -100,6 +100,7 @@ void __init ipc_init_proc_interface(const char *path, const char *header,
#define ipcid_to_idx(id) ((id) % SEQ_MULTIPLIER)
#define ipcid_to_seqx(id) ((id) / SEQ_MULTIPLIER)
+#define IPCID_SEQ_MAX min_t(int, INT_MAX/SEQ_MULTIPLIER, USHRT_MAX)
/* must be called with ids->rwsem acquired for writing */
int ipc_addid(struct ipc_ids *, struct kern_ipc_perm *, int);
--
1.8.1.4
Both expunge_all() and pipeline_send() rely on both a nil msg value and a
full barrier to guarantee the correct ordering when waking up a task. While
its counter part at the receiving end is well documented for the lockless
recv algorithm, we still need to document these specific smp_mb() calls.
Signed-off-by: Davidlohr Bueso <[email protected]>
---
ipc/msg.c | 19 +++++++++++++++++--
1 file changed, 17 insertions(+), 2 deletions(-)
diff --git a/ipc/msg.c b/ipc/msg.c
index 4377f4a..8c1cc76 100644
--- a/ipc/msg.c
+++ b/ipc/msg.c
@@ -253,8 +253,14 @@ static void expunge_all(struct msg_queue *msq, int res)
struct msg_receiver *msr, *t;
list_for_each_entry_safe(msr, t, &msq->q_receivers, r_list) {
- msr->r_msg = NULL;
+ msr->r_msg = NULL; /* initialize expunge ordering */
wake_up_process(msr->r_tsk);
+ /*
+ * Ensure that the wakeup is visible before setting r_msg as
+ * the receiving end depends on it: either spinning on a nil,
+ * or dealing with -EAGAIN cases. See lockless reveice part 1
+ * and 2 in do_msgrcv().
+ */
smp_mb();
msr->r_msg = ERR_PTR(res);
}
@@ -638,15 +644,22 @@ static inline int pipelined_send(struct msg_queue *msq, struct msg_msg *msg)
list_del(&msr->r_list);
if (msr->r_maxsize < msg->m_ts) {
+ /* initialize pipelined send ordering */
msr->r_msg = NULL;
wake_up_process(msr->r_tsk);
- smp_mb();
+ smp_mb(); /* see barrier comment below */
msr->r_msg = ERR_PTR(-E2BIG);
} else {
msr->r_msg = NULL;
msq->q_lrpid = task_pid_vnr(msr->r_tsk);
msq->q_rtime = get_seconds();
wake_up_process(msr->r_tsk);
+ /*
+ * Ensure that the wakeup is visible before
+ * setting r_msg, as the receiving end depends
+ * on it. See lockless reveice part 1 and 2 in
+ * do_msgrcv().
+ */
smp_mb();
msr->r_msg = msg;
@@ -654,6 +667,7 @@ static inline int pipelined_send(struct msg_queue *msq, struct msg_msg *msg)
}
}
}
+
return 0;
}
@@ -716,6 +730,7 @@ long do_msgsnd(int msqid, long mtype, void __user *mtext,
goto out_unlock0;
}
+ /* enqueue the sender and prepare to block */
ss_add(msq, &s);
if (!ipc_rcu_getref(msq)) {
--
1.8.1.4
IPC commenting style is all over the place, *specially* in util.c. This patch
orders things a bit.
Signed-off-by: Davidlohr Bueso <[email protected]>
---
ipc/sem.c | 24 +++----
ipc/shm.c | 1 -
ipc/util.c | 239 +++++++++++++++++++++++++++++--------------------------------
3 files changed, 125 insertions(+), 139 deletions(-)
diff --git a/ipc/sem.c b/ipc/sem.c
index c40876b..1b1acdb 100644
--- a/ipc/sem.c
+++ b/ipc/sem.c
@@ -225,7 +225,7 @@ static void unmerge_queues(struct sem_array *sma)
}
/**
- * merge_queues - Merge single semop queues into global queue
+ * merge_queues - merge single semop queues into global queue
* @sma: semaphore array
*
* This function merges all per-semaphore queues into the global queue.
@@ -474,7 +474,6 @@ static inline void sem_rmid(struct ipc_namespace *ns, struct sem_array *s)
*
* Called with sem_ids.rwsem held (as a writer)
*/
-
static int newary(struct ipc_namespace *ns, struct ipc_params *params)
{
int id;
@@ -682,7 +681,7 @@ static void wake_up_sem_queue_prepare(struct list_head *pt,
}
/**
- * wake_up_sem_queue_do(pt) - do the actual wake-up
+ * wake_up_sem_queue_do - do the actual wake-up
* @pt: list of tasks to be woken up
*
* Do the actual wake-up.
@@ -748,7 +747,7 @@ static int check_restart(struct sem_array *sma, struct sem_queue *q)
}
/**
- * wake_const_ops(sma, semnum, pt) - Wake up non-alter tasks
+ * wake_const_ops - wake up non-alter tasks
* @sma: semaphore array.
* @semnum: semaphore that was modified.
* @pt: list head for the tasks that must be woken up.
@@ -798,15 +797,14 @@ static int wake_const_ops(struct sem_array *sma, int semnum,
}
/**
- * do_smart_wakeup_zero(sma, sops, nsops, pt) - wakeup all wait for zero tasks
+ * do_smart_wakeup_zero - wakeup all wait for zero tasks
* @sma: semaphore array
* @sops: operations that were performed
* @nsops: number of operations
* @pt: list head of the tasks that must be woken up.
*
- * do_smart_wakeup_zero() checks all required queue for wait-for-zero
- * operations, based on the actual changes that were performed on the
- * semaphore array.
+ * Checks all required queue for wait-for-zero operations, based
+ * on the actual changes that were performed on the semaphore array.
* The function returns 1 if at least one operation was completed successfully.
*/
static int do_smart_wakeup_zero(struct sem_array *sma, struct sembuf *sops,
@@ -850,7 +848,7 @@ static int do_smart_wakeup_zero(struct sem_array *sma, struct sembuf *sops,
/**
- * update_queue(sma, semnum): Look for tasks that can be completed.
+ * update_queue - look for tasks that can be completed.
* @sma: semaphore array.
* @semnum: semaphore that was modified.
* @pt: list head for the tasks that must be woken up.
@@ -920,7 +918,7 @@ again:
}
/**
- * set_semotime(sma, sops) - set sem_otime
+ * set_semotime - set sem_otime
* @sma: semaphore array
* @sops: operations that modified the array, may be NULL
*
@@ -938,7 +936,7 @@ static void set_semotime(struct sem_array *sma, struct sembuf *sops)
}
/**
- * do_smart_update(sma, sops, nsops, otime, pt) - optimized update_queue
+ * do_smart_update - optimized update_queue
* @sma: semaphore array
* @sops: operations that were performed
* @nsops: number of operations
@@ -1647,7 +1645,7 @@ static struct sem_undo *lookup_undo(struct sem_undo_list *ulp, int semid)
}
/**
- * find_alloc_undo - Lookup (and if not present create) undo array
+ * find_alloc_undo - lookup (and if not present create) undo array
* @ns: namespace
* @semid: semaphore array id
*
@@ -1737,7 +1735,7 @@ out:
/**
- * get_queue_result - Retrieve the result code from sem_queue
+ * get_queue_result - retrieve the result code from sem_queue
* @q: Pointer to queue structure
*
* Retrieve the return code from the pending queue. If IN_WAKEUP is found in
diff --git a/ipc/shm.c b/ipc/shm.c
index 88c59c1e..7645961 100644
--- a/ipc/shm.c
+++ b/ipc/shm.c
@@ -477,7 +477,6 @@ static const struct vm_operations_struct shm_vm_ops = {
*
* Called with shm_ids.rwsem held as a writer.
*/
-
static int newseg(struct ipc_namespace *ns, struct ipc_params *params)
{
key_t key = params->key;
diff --git a/ipc/util.c b/ipc/util.c
index fdf03fa..af976fc6 100644
--- a/ipc/util.c
+++ b/ipc/util.c
@@ -110,15 +110,15 @@ static struct notifier_block ipc_memory_nb = {
};
/**
- * ipc_init - initialise IPC subsystem
+ * ipc_init - initialise ipc subsystem
*
- * The various system5 IPC resources (semaphores, messages and shared
- * memory) are initialised
- * A callback routine is registered into the memory hotplug notifier
- * chain: since msgmni scales to lowmem this callback routine will be
- * called upon successful memory add / remove to recompute msmgni.
+ * The various sysv ipc resources (semaphores, messages and shared
+ * memory) are initialised.
+ *
+ * A callback routine is registered into the memory hotplug notifier
+ * chain: since msgmni scales to lowmem this callback routine will be
+ * called upon successful memory add / remove to recompute msmgni.
*/
-
static int __init ipc_init(void)
{
sem_init();
@@ -131,13 +131,12 @@ static int __init ipc_init(void)
__initcall(ipc_init);
/**
- * ipc_init_ids - initialise IPC identifiers
- * @ids: Identifier set
+ * ipc_init_ids - initialise ipc identifiers
+ * @ids: ipc identifier set
*
- * Set up the sequence range to use for the ipc identifier range (limited
- * below IPCMNI) then initialise the ids idr.
+ * Set up the sequence range to use for the ipc identifier range (limited
+ * below IPCMNI) then initialise the ids idr.
*/
-
void ipc_init_ids(struct ipc_ids *ids)
{
init_rwsem(&ids->rwsem);
@@ -159,11 +158,11 @@ void ipc_init_ids(struct ipc_ids *ids)
#ifdef CONFIG_PROC_FS
static const struct file_operations sysvipc_proc_fops;
/**
- * ipc_init_proc_interface - Create a proc interface for sysipc types using a seq_file interface.
- * @path: Path in procfs
- * @header: Banner to be printed at the beginning of the file.
- * @ids: ipc id table to iterate.
- * @show: show routine.
+ * ipc_init_proc_interface - create a proc interface for sysipc types using a seq_file interface.
+ * @path: Path in procfs
+ * @header: Banner to be printed at the beginning of the file.
+ * @ids: ipc id table to iterate.
+ * @show: show routine.
*/
void __init ipc_init_proc_interface(const char *path, const char *header,
int ids, int (*show)(struct seq_file *, void *))
@@ -191,16 +190,15 @@ void __init ipc_init_proc_interface(const char *path, const char *header,
#endif
/**
- * ipc_findkey - find a key in an ipc identifier set
- * @ids: Identifier set
- * @key: The key to find
+ * ipc_findkey - find a key in an ipc identifier set
+ * @ids: ipc identifier set
+ * @key: key to find
*
- * Requires ipc_ids.rwsem locked.
- * Returns the LOCKED pointer to the ipc structure if found or NULL
- * if not.
- * If key is found ipc points to the owning ipc structure
+ * Returns the locked pointer to the ipc structure if found or NULL
+ * otherwise. If key is found ipc points to the owning ipc structure
+ *
+ * Called with ipc_ids.rwsem held.
*/
-
static struct kern_ipc_perm *ipc_findkey(struct ipc_ids *ids, key_t key)
{
struct kern_ipc_perm *ipc;
@@ -227,12 +225,11 @@ static struct kern_ipc_perm *ipc_findkey(struct ipc_ids *ids, key_t key)
}
/**
- * ipc_get_maxid - get the last assigned id
- * @ids: IPC identifier set
+ * ipc_get_maxid - get the last assigned id
+ * @ids: ipc identifier set
*
- * Called with ipc_ids.rwsem held.
+ * Called with ipc_ids.rwsem held.
*/
-
int ipc_get_maxid(struct ipc_ids *ids)
{
struct kern_ipc_perm *ipc;
@@ -258,17 +255,17 @@ int ipc_get_maxid(struct ipc_ids *ids)
}
/**
- * ipc_addid - add an IPC identifier
- * @ids: IPC identifier set
- * @new: new IPC permission set
- * @size: limit for the number of used ids
+ * ipc_addid - add an ipc identifier
+ * @ids: ipc identifier set
+ * @new: new ipc permission set
+ * @size: limit for the number of used ids
*
- * Add an entry 'new' to the IPC ids idr. The permissions object is
- * initialised and the first free entry is set up and the id assigned
- * is returned. The 'new' entry is returned in a locked state on success.
- * On failure the entry is not locked and a negative err-code is returned.
+ * Add an entry 'new' to the ipc ids idr. The permissions object is
+ * initialised and the first free entry is set up and the id assigned
+ * is returned. The 'new' entry is returned in a locked state on success.
+ * On failure the entry is not locked and a negative err-code is returned.
*
- * Called with writer ipc_ids.rwsem held.
+ * Called with writer ipc_ids.rwsem held.
*/
int ipc_addid(struct ipc_ids *ids, struct kern_ipc_perm *new, int size)
{
@@ -320,14 +317,14 @@ int ipc_addid(struct ipc_ids *ids, struct kern_ipc_perm *new, int size)
}
/**
- * ipcget_new - create a new ipc object
- * @ns: namespace
- * @ids: IPC identifer set
- * @ops: the actual creation routine to call
- * @params: its parameters
- *
- * This routine is called by sys_msgget, sys_semget() and sys_shmget()
- * when the key is IPC_PRIVATE.
+ * ipcget_new - create a new ipc object
+ * @ns: ipc namespace
+ * @ids: ipc identifer set
+ * @ops: the actual creation routine to call
+ * @params: its parameters
+ *
+ * This routine is called by sys_msgget, sys_semget() and sys_shmget()
+ * when the key is IPC_PRIVATE.
*/
static int ipcget_new(struct ipc_namespace *ns, struct ipc_ids *ids,
struct ipc_ops *ops, struct ipc_params *params)
@@ -341,19 +338,19 @@ static int ipcget_new(struct ipc_namespace *ns, struct ipc_ids *ids,
}
/**
- * ipc_check_perms - check security and permissions for an IPC
- * @ns: IPC namespace
- * @ipcp: ipc permission set
- * @ops: the actual security routine to call
- * @params: its parameters
+ * ipc_check_perms - check security and permissions for an ipc object
+ * @ns: ipc namespace
+ * @ipcp: ipc permission set
+ * @ops: the actual security routine to call
+ * @params: its parameters
*
- * This routine is called by sys_msgget(), sys_semget() and sys_shmget()
- * when the key is not IPC_PRIVATE and that key already exists in the
- * ids IDR.
+ * This routine is called by sys_msgget(), sys_semget() and sys_shmget()
+ * when the key is not IPC_PRIVATE and that key already exists in the
+ * ds IDR.
*
- * On success, the IPC id is returned.
+ * On success, the ipc id is returned.
*
- * It is called with ipc_ids.rwsem and ipcp->lock held.
+ * It is called with ipc_ids.rwsem and ipcp->lock held.
*/
static int ipc_check_perms(struct ipc_namespace *ns,
struct kern_ipc_perm *ipcp,
@@ -374,18 +371,18 @@ static int ipc_check_perms(struct ipc_namespace *ns,
}
/**
- * ipcget_public - get an ipc object or create a new one
- * @ns: namespace
- * @ids: IPC identifer set
- * @ops: the actual creation routine to call
- * @params: its parameters
- *
- * This routine is called by sys_msgget, sys_semget() and sys_shmget()
- * when the key is not IPC_PRIVATE.
- * It adds a new entry if the key is not found and does some permission
- * / security checkings if the key is found.
- *
- * On success, the ipc id is returned.
+ * ipcget_public - get an ipc object or create a new one
+ * @ns: ipc namespace
+ * @ids: ipc identifer set
+ * @ops: the actual creation routine to call
+ * @params: its parameters
+ *
+ * This routine is called by sys_msgget, sys_semget() and sys_shmget()
+ * when the key is not IPC_PRIVATE.
+ * It adds a new entry if the key is not found and does some permission
+ * / security checkings if the key is found.
+ *
+ * On success, the ipc id is returned.
*/
static int ipcget_public(struct ipc_namespace *ns, struct ipc_ids *ids,
struct ipc_ops *ops, struct ipc_params *params)
@@ -431,14 +428,13 @@ static int ipcget_public(struct ipc_namespace *ns, struct ipc_ids *ids,
/**
- * ipc_rmid - remove an IPC identifier
- * @ids: IPC identifier set
- * @ipcp: ipc perm structure containing the identifier to remove
+ * ipc_rmid - remove an ipc identifier
+ * @ids: ipc identifier set
+ * @ipcp: ipc perm structure containing the identifier to remove
*
- * ipc_ids.rwsem (as a writer) and the spinlock for this ID are held
- * before this function is called, and remain locked on the exit.
+ * ipc_ids.rwsem (as a writer) and the spinlock for this ID are held
+ * before this function is called, and remain locked on the exit.
*/
-
void ipc_rmid(struct ipc_ids *ids, struct kern_ipc_perm *ipcp)
{
int lid = ipcid_to_idx(ipcp->id);
@@ -453,13 +449,12 @@ void ipc_rmid(struct ipc_ids *ids, struct kern_ipc_perm *ipcp)
}
/**
- * ipc_alloc - allocate ipc space
- * @size: size desired
+ * ipc_alloc - allocate ipc space
+ * @size: size desired
*
- * Allocate memory from the appropriate pools and return a pointer to it.
- * NULL is returned if the allocation fails
+ * Allocate memory from the appropriate pools and return a pointer to it.
+ * NULL is returned if the allocation fails
*/
-
void *ipc_alloc(int size)
{
void *out;
@@ -471,14 +466,13 @@ void *ipc_alloc(int size)
}
/**
- * ipc_free - free ipc space
- * @ptr: pointer returned by ipc_alloc
- * @size: size of block
+ * ipc_free - free ipc space
+ * @ptr: pointer returned by ipc_alloc
+ * @size: size of block
*
- * Free a block created with ipc_alloc(). The caller must know the size
- * used in the allocation call.
+ * Free a block created with ipc_alloc(). The caller must know the size
+ * used in the allocation call.
*/
-
void ipc_free(void *ptr, int size)
{
if (size > PAGE_SIZE)
@@ -488,11 +482,11 @@ void ipc_free(void *ptr, int size)
}
/**
- * ipc_rcu_alloc - allocate ipc and rcu space
- * @size: size desired
+ * ipc_rcu_alloc - allocate ipc and rcu space
+ * @size: size desired
*
- * Allocate memory for the rcu header structure + the object.
- * Returns the pointer to the object or NULL upon failure.
+ * Allocate memory for the rcu header structure + the object.
+ * Returns the pointer to the object or NULL upon failure.
*/
void *ipc_rcu_alloc(int size)
{
@@ -534,17 +528,16 @@ void ipc_rcu_free(struct rcu_head *head)
}
/**
- * ipcperms - check IPC permissions
- * @ns: IPC namespace
- * @ipcp: IPC permission set
- * @flag: desired permission set.
+ * ipcperms - check ipc permissions
+ * @ns: ipc namespace
+ * @ipcp: ipc permission set
+ * @flag: desired permission set
*
- * Check user, group, other permissions for access
- * to ipc resources. return 0 if allowed
+ * Check user, group, other permissions for access
+ * to ipc resources. return 0 if allowed
*
- * @flag will most probably be 0 or S_...UGO from <linux/stat.h>
+ * @flag will most probably be 0 or S_...UGO from <linux/stat.h>
*/
-
int ipcperms(struct ipc_namespace *ns, struct kern_ipc_perm *ipcp, short flag)
{
kuid_t euid = current_euid();
@@ -572,15 +565,13 @@ int ipcperms(struct ipc_namespace *ns, struct kern_ipc_perm *ipcp, short flag)
*/
/**
- * kernel_to_ipc64_perm - convert kernel ipc permissions to user
- * @in: kernel permissions
- * @out: new style IPC permissions
+ * kernel_to_ipc64_perm - convert kernel ipc permissions to user
+ * @in: kernel permissions
+ * @out: new style ipc permissions
*
- * Turn the kernel object @in into a set of permissions descriptions
- * for returning to userspace (@out).
+ * Turn the kernel object @in into a set of permissions descriptions
+ * for returning to userspace (@out).
*/
-
-
void kernel_to_ipc64_perm(struct kern_ipc_perm *in, struct ipc64_perm *out)
{
out->key = in->key;
@@ -593,14 +584,13 @@ void kernel_to_ipc64_perm(struct kern_ipc_perm *in, struct ipc64_perm *out)
}
/**
- * ipc64_perm_to_ipc_perm - convert new ipc permissions to old
- * @in: new style IPC permissions
- * @out: old style IPC permissions
+ * ipc64_perm_to_ipc_perm - convert new ipc permissions to old
+ * @in: new style ipc permissions
+ * @out: old style ipc permissions
*
- * Turn the new style permissions object @in into a compatibility
- * object and store it into the @out pointer.
+ * Turn the new style permissions object @in into a compatibility
+ * object and store it into the @out pointer.
*/
-
void ipc64_perm_to_ipc_perm(struct ipc64_perm *in, struct ipc_perm *out)
{
out->key = in->key;
@@ -635,8 +625,8 @@ struct kern_ipc_perm *ipc_obtain_object(struct ipc_ids *ids, int id)
}
/**
- * ipc_lock - Lock an ipc structure without rwsem held
- * @ids: IPC identifier set
+ * ipc_lock - lock an ipc structure without rwsem held
+ * @ids: ipc identifier set
* @id: ipc id to look for
*
* Look for an id in the ipc ids idr and lock the associated ipc object.
@@ -693,11 +683,11 @@ out:
/**
* ipcget - Common sys_*get() code
- * @ns : namsepace
- * @ids : IPC identifier set
- * @ops : operations to be called on ipc object creation, permission checks
- * and further checks
- * @params : the parameters needed by the previous operations.
+ * @ns: namsepace
+ * @ids: ipc identifier set
+ * @ops: operations to be called on ipc object creation, permission checks
+ * and further checks
+ * @params: the parameters needed by the previous operations.
*
* Common routine called by sys_msgget(), sys_semget() and sys_shmget().
*/
@@ -711,7 +701,7 @@ int ipcget(struct ipc_namespace *ns, struct ipc_ids *ids,
}
/**
- * ipc_update_perm - update the permissions of an IPC.
+ * ipc_update_perm - update the permissions of an ipc object
* @in: the permission given as input.
* @out: the permission of the ipc to set.
*/
@@ -732,7 +722,7 @@ int ipc_update_perm(struct ipc64_perm *in, struct kern_ipc_perm *out)
/**
* ipcctl_pre_down_nolock - retrieve an ipc and check permissions for some IPC_XXX cmd
- * @ns: the ipc namespace
+ * @ns: ipc namespace
* @ids: the table of ids where to look for the ipc
* @id: the id of the ipc to retrieve
* @cmd: the cmd to check
@@ -779,14 +769,13 @@ err:
/**
- * ipc_parse_version - IPC call version
- * @cmd: pointer to command
+ * ipc_parse_version - ipc call version
+ * @cmd: pointer to command
*
- * Return IPC_64 for new style IPC and IPC_OLD for old style IPC.
- * The @cmd value is turned from an encoding command and version into
- * just the command code.
+ * Return IPC_64 for new style IPC and IPC_OLD for old style IPC.
+ * The @cmd value is turned from an encoding command and version into
+ * just the command code.
*/
-
int ipc_parse_version(int *cmd)
{
if (*cmd & IPC_64) {
--
1.8.1.4
Deal with checkpatch messages:
WARNING: braces {} are not necessary for single statement blocks
Signed-off-by: Davidlohr Bueso <[email protected]>
---
ipc/compat.c | 18 +++++++++---------
ipc/mqueue.c | 6 +++---
ipc/sem.c | 8 +++-----
ipc/util.c | 3 +--
4 files changed, 16 insertions(+), 19 deletions(-)
diff --git a/ipc/compat.c b/ipc/compat.c
index ed0530b..f71e962 100644
--- a/ipc/compat.c
+++ b/ipc/compat.c
@@ -288,11 +288,11 @@ static long do_compat_semctl(int first, int second, int third, u32 pad)
break;
case IPC_SET:
- if (version == IPC_64) {
+ if (version == IPC_64)
err = get_compat_semid64_ds(&s64, compat_ptr(pad));
- } else {
+ else
err = get_compat_semid_ds(&s64, compat_ptr(pad));
- }
+
up64 = compat_alloc_user_space(sizeof(s64));
if (copy_to_user(up64, &s64, sizeof(s64)))
err = -EFAULT;
@@ -515,11 +515,11 @@ long compat_sys_msgctl(int first, int second, void __user *uptr)
break;
case IPC_SET:
- if (version == IPC_64) {
+ if (version == IPC_64)
err = get_compat_msqid64(&m64, uptr);
- } else {
+ else
err = get_compat_msqid(&m64, uptr);
- }
+
if (err)
break;
p = compat_alloc_user_space(sizeof(m64));
@@ -702,11 +702,11 @@ long compat_sys_shmctl(int first, int second, void __user *uptr)
case IPC_SET:
- if (version == IPC_64) {
+ if (version == IPC_64)
err = get_compat_shmid64_ds(&s64, uptr);
- } else {
+ else
err = get_compat_shmid_ds(&s64, uptr);
- }
+
if (err)
break;
p = compat_alloc_user_space(sizeof(s64));
diff --git a/ipc/mqueue.c b/ipc/mqueue.c
index bb36aae..ccf1f9f 100644
--- a/ipc/mqueue.c
+++ b/ipc/mqueue.c
@@ -1303,11 +1303,11 @@ retry:
out_fput:
fdput(f);
out:
- if (sock) {
+ if (sock)
netlink_detachskb(sock, nc);
- } else if (nc) {
+ else if (nc)
dev_kfree_skb(nc);
- }
+
return ret;
}
diff --git a/ipc/sem.c b/ipc/sem.c
index 1b1acdb..bee5554 100644
--- a/ipc/sem.c
+++ b/ipc/sem.c
@@ -492,9 +492,9 @@ static int newary(struct ipc_namespace *ns, struct ipc_params *params)
size = sizeof(*sma) + nsems * sizeof(struct sem);
sma = ipc_rcu_alloc(size);
- if (!sma) {
+ if (!sma)
return -ENOMEM;
- }
+
memset(sma, 0, size);
sma->sem_perm.mode = (semflg & S_IRWXUGO);
@@ -1967,10 +1967,8 @@ sleep_again:
* If queue.status != -EINTR we are woken up by another process.
* Leave without unlink_queue(), but with sem_unlock().
*/
-
- if (error != -EINTR) {
+ if (error != -EINTR)
goto out_unlock_free;
- }
/*
* If an interrupt occurred we have to clean up the queue
diff --git a/ipc/util.c b/ipc/util.c
index af976fc6..cfbd8fa 100644
--- a/ipc/util.c
+++ b/ipc/util.c
@@ -183,9 +183,8 @@ void __init ipc_init_proc_interface(const char *path, const char *header,
NULL, /* parent dir */
&sysvipc_proc_fops,
iface);
- if (!pde) {
+ if (!pde)
kfree(iface);
- }
}
#endif
--
1.8.1.4
On Fri, 2014-01-10 at 10:48 -0800, Davidlohr Bueso wrote:
> Both expunge_all() and pipeline_send() rely on both a nil msg value and a
> full barrier to guarantee the correct ordering when waking up a task. While
> its counter part at the receiving end is well documented for the lockless
> recv algorithm, we still need to document these specific smp_mb() calls.
>
> Signed-off-by: Davidlohr Bueso <[email protected]>
> ---
> ipc/msg.c | 19 +++++++++++++++++--
> 1 file changed, 17 insertions(+), 2 deletions(-)
>
> diff --git a/ipc/msg.c b/ipc/msg.c
> index 4377f4a..8c1cc76 100644
> --- a/ipc/msg.c
> +++ b/ipc/msg.c
> @@ -253,8 +253,14 @@ static void expunge_all(struct msg_queue *msq, int res)
> struct msg_receiver *msr, *t;
>
> list_for_each_entry_safe(msr, t, &msq->q_receivers, r_list) {
> - msr->r_msg = NULL;
> + msr->r_msg = NULL; /* initialize expunge ordering */
> wake_up_process(msr->r_tsk);
> + /*
> + * Ensure that the wakeup is visible before setting r_msg as
> + * the receiving end depends on it: either spinning on a nil,
> + * or dealing with -EAGAIN cases. See lockless reveice part 1
^^^^^^^ lysdexic fingers
> + * and 2 in do_msgrcv().
> + */
> smp_mb();
> msr->r_msg = ERR_PTR(res);
> }
> @@ -638,15 +644,22 @@ static inline int pipelined_send(struct msg_queue *msq, struct msg_msg *msg)
>
> list_del(&msr->r_list);
> if (msr->r_maxsize < msg->m_ts) {
> + /* initialize pipelined send ordering */
> msr->r_msg = NULL;
> wake_up_process(msr->r_tsk);
> - smp_mb();
> + smp_mb(); /* see barrier comment below */
> msr->r_msg = ERR_PTR(-E2BIG);
> } else {
> msr->r_msg = NULL;
> msq->q_lrpid = task_pid_vnr(msr->r_tsk);
> msq->q_rtime = get_seconds();
> wake_up_process(msr->r_tsk);
> + /*
> + * Ensure that the wakeup is visible before
> + * setting r_msg, as the receiving end depends
> + * on it. See lockless reveice part 1 and 2 in
^^^^^^^ ditto
Hi David,
On 01/10/2014 07:48 PM, Davidlohr Bueso wrote:
> A few updates to the ipc code, mostly some very needed cleanups (yes, we still
> have loads more to do, but it's a start). Nothing particularly interesting,
> except perhaps for the last two patches, which include a locking optimization
> and barrier documentation for message queues. Absolutely no functional changes.
>
> Applies on top of linux-next + Manfred's 'whitespace cleanup' patch. Tested
> with LTP.
Now you confuse me.
In your other mail, you write that my patch doesn't apply to linux-next:
On 01/07/2014 06:12 PM, Davidlohr Bueso wrote:
> This patch doesn't apply on top of linux-next (which now includes
> Rafael's changes), could you please resend? I'm planning some more
> cleanups on top of this so, if Andrew agrees, I can include these
> changes to my patchset. Thanks, Davidlohr
Perhaps: Could you include (and update, if required) my patch into your
patchset?
It's probably simpler if we have only one set of ipc cleanups around.
> Thanks!
>
> Davidlohr Bueso (7):
> ipc: standardize code comments
> ipc: remove braces for single statements
> ipc: remove useless return statement
> ipc: simplify sysvipc_proc_open return
> ipc: delete seq_max field in struct ipc_ids
Acked-by: Manfred Spraul <[email protected]>
> ipc: share ids rwsem when possible in ipcget_public
This patch is a real change. Could you at least make it the last one in
the series?
I will try to review it, but I don't know yet when I will find the time.
> ipc,msg: document barriers
Acked-by: Manfred Spraul <[email protected]>
--
Manfred
On Sat, 2014-01-11 at 10:02 +0100, Manfred Spraul wrote:
> Hi David,
>
> On 01/10/2014 07:48 PM, Davidlohr Bueso wrote:
> > A few updates to the ipc code, mostly some very needed cleanups (yes, we still
> > have loads more to do, but it's a start). Nothing particularly interesting,
> > except perhaps for the last two patches, which include a locking optimization
> > and barrier documentation for message queues. Absolutely no functional changes.
> >
> > Applies on top of linux-next + Manfred's 'whitespace cleanup' patch. Tested
> > with LTP.
> Now you confuse me.
> In your other mail, you write that my patch doesn't apply to linux-next:
Just ignore that email, Andrew then applied (and fixed the merge
conflict) your whitespace cleanup patch and was already in linux-next by
the time I sent this patchset, so no need to worry/coordinate.
>
> On 01/07/2014 06:12 PM, Davidlohr Bueso wrote:
> > This patch doesn't apply on top of linux-next (which now includes
> > Rafael's changes), could you please resend? I'm planning some more
> > cleanups on top of this so, if Andrew agrees, I can include these
> > changes to my patchset. Thanks, Davidlohr
>
> Perhaps: Could you include (and update, if required) my patch into your
> patchset?
> It's probably simpler if we have only one set of ipc cleanups around.
>
> > Thanks!
> >
> > Davidlohr Bueso (7):
> > ipc: standardize code comments
> > ipc: remove braces for single statements
> > ipc: remove useless return statement
> > ipc: simplify sysvipc_proc_open return
> > ipc: delete seq_max field in struct ipc_ids
> Acked-by: Manfred Spraul <[email protected]>
> > ipc: share ids rwsem when possible in ipcget_public
> This patch is a real change. Could you at least make it the last one in
> the series?
> I will try to review it, but I don't know yet when I will find the time.
Thanks, I definitely want your ack/review for this one.
>
> > ipc,msg: document barriers
> Acked-by: Manfred Spraul <[email protected]>
>
> --
> Manfred