Commit 40689ac4 authored by Linus Torvalds's avatar Linus Torvalds

Merge tag 'dlm-3.7' of git://git.kernel.org/pub/scm/linux/kernel/git/teigland/linux-dlm

Pull dlm updates from David Teigland:
 "There are two main patches in this set, both related to the userland
  dlm_controld daemon.

  The first fixes a deadlock between dlm_controld and the dlm_send
  workqueue when both access configfs data simultaneously.

  The second reworks some code to get around a long standing, but
  intentional, unlock balance warning.  The userland daemon no longer
  takes a lock that is later released from the kernel.

  The other commits are minor fixes and changes."

* tag 'dlm-3.7' of git://git.kernel.org/pub/scm/linux/kernel/git/teigland/linux-dlm:
  dlm: check the maximum size of a request from user
  dlm: cleanup send_to_sock routine
  dlm: convert add_sock routine return value type to void
  dlm: remove redundant variable assignments
  dlm: fix unlock balance warnings
  dlm: fix uninitialized spinlock
  dlm: fix deadlock between dlm_send and dlm_controld
parents cc150a28 2b75bc91
......@@ -16,7 +16,7 @@
#include "user.h"
static uint64_t dlm_cb_seq;
static spinlock_t dlm_cb_seq_spin;
static DEFINE_SPINLOCK(dlm_cb_seq_spin);
static void dlm_dump_lkb_callbacks(struct dlm_lkb *lkb)
{
......
......@@ -750,6 +750,7 @@ static ssize_t comm_local_write(struct dlm_comm *cm, const char *buf,
static ssize_t comm_addr_write(struct dlm_comm *cm, const char *buf, size_t len)
{
struct sockaddr_storage *addr;
int rv;
if (len != sizeof(struct sockaddr_storage))
return -EINVAL;
......@@ -762,6 +763,13 @@ static ssize_t comm_addr_write(struct dlm_comm *cm, const char *buf, size_t len)
return -ENOMEM;
memcpy(addr, buf, len);
rv = dlm_lowcomms_addr(cm->nodeid, addr, len);
if (rv) {
kfree(addr);
return rv;
}
cm->addr[cm->addr_count++] = addr;
return len;
}
......@@ -878,34 +886,7 @@ static void put_space(struct dlm_space *sp)
config_item_put(&sp->group.cg_item);
}
static int addr_compare(struct sockaddr_storage *x, struct sockaddr_storage *y)
{
switch (x->ss_family) {
case AF_INET: {
struct sockaddr_in *sinx = (struct sockaddr_in *)x;
struct sockaddr_in *siny = (struct sockaddr_in *)y;
if (sinx->sin_addr.s_addr != siny->sin_addr.s_addr)
return 0;
if (sinx->sin_port != siny->sin_port)
return 0;
break;
}
case AF_INET6: {
struct sockaddr_in6 *sinx = (struct sockaddr_in6 *)x;
struct sockaddr_in6 *siny = (struct sockaddr_in6 *)y;
if (!ipv6_addr_equal(&sinx->sin6_addr, &siny->sin6_addr))
return 0;
if (sinx->sin6_port != siny->sin6_port)
return 0;
break;
}
default:
return 0;
}
return 1;
}
static struct dlm_comm *get_comm(int nodeid, struct sockaddr_storage *addr)
static struct dlm_comm *get_comm(int nodeid)
{
struct config_item *i;
struct dlm_comm *cm = NULL;
......@@ -919,19 +900,11 @@ static struct dlm_comm *get_comm(int nodeid, struct sockaddr_storage *addr)
list_for_each_entry(i, &comm_list->cg_children, ci_entry) {
cm = config_item_to_comm(i);
if (nodeid) {
if (cm->nodeid != nodeid)
continue;
found = 1;
config_item_get(i);
break;
} else {
if (!cm->addr_count || !addr_compare(cm->addr[0], addr))
continue;
found = 1;
config_item_get(i);
break;
}
}
mutex_unlock(&clusters_root.subsys.su_mutex);
......@@ -995,7 +968,7 @@ int dlm_config_nodes(char *lsname, struct dlm_config_node **nodes_out,
int dlm_comm_seq(int nodeid, uint32_t *seq)
{
struct dlm_comm *cm = get_comm(nodeid, NULL);
struct dlm_comm *cm = get_comm(nodeid);
if (!cm)
return -EEXIST;
*seq = cm->seq;
......@@ -1003,28 +976,6 @@ int dlm_comm_seq(int nodeid, uint32_t *seq)
return 0;
}
int dlm_nodeid_to_addr(int nodeid, struct sockaddr_storage *addr)
{
struct dlm_comm *cm = get_comm(nodeid, NULL);
if (!cm)
return -EEXIST;
if (!cm->addr_count)
return -ENOENT;
memcpy(addr, cm->addr[0], sizeof(*addr));
put_comm(cm);
return 0;
}
int dlm_addr_to_nodeid(struct sockaddr_storage *addr, int *nodeid)
{
struct dlm_comm *cm = get_comm(0, addr);
if (!cm)
return -EEXIST;
*nodeid = cm->nodeid;
put_comm(cm);
return 0;
}
int dlm_our_nodeid(void)
{
return local_comm ? local_comm->nodeid : 0;
......
......@@ -46,8 +46,6 @@ void dlm_config_exit(void);
int dlm_config_nodes(char *lsname, struct dlm_config_node **nodes_out,
int *count_out);
int dlm_comm_seq(int nodeid, uint32_t *seq);
int dlm_nodeid_to_addr(int nodeid, struct sockaddr_storage *addr);
int dlm_addr_to_nodeid(struct sockaddr_storage *addr, int *nodeid);
int dlm_our_nodeid(void);
int dlm_our_addr(struct sockaddr_storage *addr, int num);
......
......@@ -604,6 +604,7 @@ struct dlm_ls {
struct idr ls_recover_idr;
spinlock_t ls_recover_idr_lock;
wait_queue_head_t ls_wait_general;
wait_queue_head_t ls_recover_lock_wait;
struct mutex ls_clear_proc_locks;
struct list_head ls_root_list; /* root resources */
......@@ -616,15 +617,40 @@ struct dlm_ls {
char ls_name[1];
};
#define LSFL_WORK 0
#define LSFL_RUNNING 1
#define LSFL_RECOVERY_STOP 2
#define LSFL_RCOM_READY 3
#define LSFL_RCOM_WAIT 4
#define LSFL_UEVENT_WAIT 5
#define LSFL_TIMEWARN 6
#define LSFL_CB_DELAY 7
#define LSFL_NODIR 8
/*
* LSFL_RECOVER_STOP - dlm_ls_stop() sets this to tell dlm recovery routines
* that they should abort what they're doing so new recovery can be started.
*
* LSFL_RECOVER_DOWN - dlm_ls_stop() sets this to tell dlm_recoverd that it
* should do down_write() on the in_recovery rw_semaphore. (doing down_write
* within dlm_ls_stop causes complaints about the lock acquired/released
* in different contexts.)
*
* LSFL_RECOVER_LOCK - dlm_recoverd holds the in_recovery rw_semaphore.
* It sets this after it is done with down_write() on the in_recovery
* rw_semaphore and clears it after it has released the rw_semaphore.
*
* LSFL_RECOVER_WORK - dlm_ls_start() sets this to tell dlm_recoverd that it
* should begin recovery of the lockspace.
*
* LSFL_RUNNING - set when normal locking activity is enabled.
* dlm_ls_stop() clears this to tell dlm locking routines that they should
* quit what they are doing so recovery can run. dlm_recoverd sets
* this after recovery is finished.
*/
#define LSFL_RECOVER_STOP 0
#define LSFL_RECOVER_DOWN 1
#define LSFL_RECOVER_LOCK 2
#define LSFL_RECOVER_WORK 3
#define LSFL_RUNNING 4
#define LSFL_RCOM_READY 5
#define LSFL_RCOM_WAIT 6
#define LSFL_UEVENT_WAIT 7
#define LSFL_TIMEWARN 8
#define LSFL_CB_DELAY 9
#define LSFL_NODIR 10
/* much of this is just saving user space pointers associated with the
lock that we pass back to the user lib with an ast */
......@@ -667,7 +693,7 @@ static inline int dlm_locking_stopped(struct dlm_ls *ls)
static inline int dlm_recovery_stopped(struct dlm_ls *ls)
{
return test_bit(LSFL_RECOVERY_STOP, &ls->ls_flags);
return test_bit(LSFL_RECOVER_STOP, &ls->ls_flags);
}
static inline int dlm_no_directory(struct dlm_ls *ls)
......
......@@ -582,8 +582,6 @@ static int new_lockspace(const char *name, const char *cluster,
INIT_LIST_HEAD(&ls->ls_root_list);
init_rwsem(&ls->ls_root_sem);
down_write(&ls->ls_in_recovery);
spin_lock(&lslist_lock);
ls->ls_create_count = 1;
list_add(&ls->ls_list, &lslist);
......@@ -597,13 +595,24 @@ static int new_lockspace(const char *name, const char *cluster,
}
}
/* needs to find ls in lslist */
init_waitqueue_head(&ls->ls_recover_lock_wait);
/*
* Once started, dlm_recoverd first looks for ls in lslist, then
* initializes ls_in_recovery as locked in "down" mode. We need
* to wait for the wakeup from dlm_recoverd because in_recovery
* has to start out in down mode.
*/
error = dlm_recoverd_start(ls);
if (error) {
log_error(ls, "can't start dlm_recoverd %d", error);
goto out_callback;
}
wait_event(ls->ls_recover_lock_wait,
test_bit(LSFL_RECOVER_LOCK, &ls->ls_flags));
ls->ls_kobj.kset = dlm_kset;
error = kobject_init_and_add(&ls->ls_kobj, &dlm_ktype, NULL,
"%s", ls->ls_name);
......
This diff is collapsed.
......@@ -16,10 +16,12 @@
int dlm_lowcomms_start(void);
void dlm_lowcomms_stop(void);
void dlm_lowcomms_exit(void);
int dlm_lowcomms_close(int nodeid);
void *dlm_lowcomms_get_buffer(int nodeid, int len, gfp_t allocation, char **ppc);
void dlm_lowcomms_commit_buffer(void *mh);
int dlm_lowcomms_connect_node(int nodeid);
int dlm_lowcomms_addr(int nodeid, struct sockaddr_storage *addr, int len);
#endif /* __LOWCOMMS_DOT_H__ */
......@@ -17,6 +17,7 @@
#include "user.h"
#include "memory.h"
#include "config.h"
#include "lowcomms.h"
static int __init init_dlm(void)
{
......@@ -78,6 +79,7 @@ static void __exit exit_dlm(void)
dlm_config_exit();
dlm_memory_exit();
dlm_lockspace_exit();
dlm_lowcomms_exit();
dlm_unregister_debugfs();
}
......
......@@ -616,13 +616,13 @@ int dlm_ls_stop(struct dlm_ls *ls)
down_write(&ls->ls_recv_active);
/*
* Abort any recovery that's in progress (see RECOVERY_STOP,
* Abort any recovery that's in progress (see RECOVER_STOP,
* dlm_recovery_stopped()) and tell any other threads running in the
* dlm to quit any processing (see RUNNING, dlm_locking_stopped()).
*/
spin_lock(&ls->ls_recover_lock);
set_bit(LSFL_RECOVERY_STOP, &ls->ls_flags);
set_bit(LSFL_RECOVER_STOP, &ls->ls_flags);
new = test_and_clear_bit(LSFL_RUNNING, &ls->ls_flags);
ls->ls_recover_seq++;
spin_unlock(&ls->ls_recover_lock);
......@@ -642,12 +642,16 @@ int dlm_ls_stop(struct dlm_ls *ls)
* when recovery is complete.
*/
if (new)
down_write(&ls->ls_in_recovery);
if (new) {
set_bit(LSFL_RECOVER_DOWN, &ls->ls_flags);
wake_up_process(ls->ls_recoverd_task);
wait_event(ls->ls_recover_lock_wait,
test_bit(LSFL_RECOVER_LOCK, &ls->ls_flags));
}
/*
* The recoverd suspend/resume makes sure that dlm_recoverd (if
* running) has noticed RECOVERY_STOP above and quit processing the
* running) has noticed RECOVER_STOP above and quit processing the
* previous recovery.
*/
......@@ -709,7 +713,8 @@ int dlm_ls_start(struct dlm_ls *ls)
kfree(rv_old);
}
dlm_recoverd_kick(ls);
set_bit(LSFL_RECOVER_WORK, &ls->ls_flags);
wake_up_process(ls->ls_recoverd_task);
return 0;
fail:
......
......@@ -581,7 +581,7 @@ void dlm_receive_rcom(struct dlm_ls *ls, struct dlm_rcom *rc, int nodeid)
spin_lock(&ls->ls_recover_lock);
status = ls->ls_recover_status;
stop = test_bit(LSFL_RECOVERY_STOP, &ls->ls_flags);
stop = test_bit(LSFL_RECOVER_STOP, &ls->ls_flags);
seq = ls->ls_recover_seq;
spin_unlock(&ls->ls_recover_lock);
......
......@@ -41,6 +41,7 @@ static int enable_locking(struct dlm_ls *ls, uint64_t seq)
set_bit(LSFL_RUNNING, &ls->ls_flags);
/* unblocks processes waiting to enter the dlm */
up_write(&ls->ls_in_recovery);
clear_bit(LSFL_RECOVER_LOCK, &ls->ls_flags);
error = 0;
}
spin_unlock(&ls->ls_recover_lock);
......@@ -262,7 +263,7 @@ static void do_ls_recovery(struct dlm_ls *ls)
rv = ls->ls_recover_args;
ls->ls_recover_args = NULL;
if (rv && ls->ls_recover_seq == rv->seq)
clear_bit(LSFL_RECOVERY_STOP, &ls->ls_flags);
clear_bit(LSFL_RECOVER_STOP, &ls->ls_flags);
spin_unlock(&ls->ls_recover_lock);
if (rv) {
......@@ -282,26 +283,34 @@ static int dlm_recoverd(void *arg)
return -1;
}
down_write(&ls->ls_in_recovery);
set_bit(LSFL_RECOVER_LOCK, &ls->ls_flags);
wake_up(&ls->ls_recover_lock_wait);
while (!kthread_should_stop()) {
set_current_state(TASK_INTERRUPTIBLE);
if (!test_bit(LSFL_WORK, &ls->ls_flags))
if (!test_bit(LSFL_RECOVER_WORK, &ls->ls_flags) &&
!test_bit(LSFL_RECOVER_DOWN, &ls->ls_flags))
schedule();
set_current_state(TASK_RUNNING);
if (test_and_clear_bit(LSFL_WORK, &ls->ls_flags))
if (test_and_clear_bit(LSFL_RECOVER_DOWN, &ls->ls_flags)) {
down_write(&ls->ls_in_recovery);
set_bit(LSFL_RECOVER_LOCK, &ls->ls_flags);
wake_up(&ls->ls_recover_lock_wait);
}
if (test_and_clear_bit(LSFL_RECOVER_WORK, &ls->ls_flags))
do_ls_recovery(ls);
}
if (test_bit(LSFL_RECOVER_LOCK, &ls->ls_flags))
up_write(&ls->ls_in_recovery);
dlm_put_lockspace(ls);
return 0;
}
void dlm_recoverd_kick(struct dlm_ls *ls)
{
set_bit(LSFL_WORK, &ls->ls_flags);
wake_up_process(ls->ls_recoverd_task);
}
int dlm_recoverd_start(struct dlm_ls *ls)
{
struct task_struct *p;
......
......@@ -14,7 +14,6 @@
#ifndef __RECOVERD_DOT_H__
#define __RECOVERD_DOT_H__
void dlm_recoverd_kick(struct dlm_ls *ls);
void dlm_recoverd_stop(struct dlm_ls *ls);
int dlm_recoverd_start(struct dlm_ls *ls);
void dlm_recoverd_suspend(struct dlm_ls *ls);
......
......@@ -503,6 +503,13 @@ static ssize_t device_write(struct file *file, const char __user *buf,
#endif
return -EINVAL;
#ifdef CONFIG_COMPAT
if (count > sizeof(struct dlm_write_request32) + DLM_RESNAME_MAXLEN)
#else
if (count > sizeof(struct dlm_write_request) + DLM_RESNAME_MAXLEN)
#endif
return -EINVAL;
kbuf = kzalloc(count + 1, GFP_NOFS);
if (!kbuf)
return -ENOMEM;
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment