2009-02-24 08:30:20 -07:00
|
|
|
/*
|
|
|
|
* Copyright (c) 2007 Oracle. All rights reserved.
|
|
|
|
*
|
|
|
|
* This software is available to you under a choice of one of two
|
|
|
|
* licenses. You may choose to be licensed under the terms of the GNU
|
|
|
|
* General Public License (GPL) Version 2, available from the file
|
|
|
|
* COPYING in the main directory of this source tree, or the
|
|
|
|
* OpenIB.org BSD license below:
|
|
|
|
*
|
|
|
|
* Redistribution and use in source and binary forms, with or
|
|
|
|
* without modification, are permitted provided that the following
|
|
|
|
* conditions are met:
|
|
|
|
*
|
|
|
|
* - Redistributions of source code must retain the above
|
|
|
|
* copyright notice, this list of conditions and the following
|
|
|
|
* disclaimer.
|
|
|
|
*
|
|
|
|
* - Redistributions in binary form must reproduce the above
|
|
|
|
* copyright notice, this list of conditions and the following
|
|
|
|
* disclaimer in the documentation and/or other materials
|
|
|
|
* provided with the distribution.
|
|
|
|
*
|
|
|
|
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
|
|
|
|
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
|
|
|
|
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
|
|
|
|
* NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
|
|
|
|
* BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
|
|
|
|
* ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
|
|
|
|
* CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
|
|
|
* SOFTWARE.
|
|
|
|
*
|
|
|
|
*/
|
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 02:04:11 -06:00
|
|
|
#include <linux/slab.h>
|
2009-02-24 08:30:20 -07:00
|
|
|
#include <linux/types.h>
|
|
|
|
#include <linux/rbtree.h>
|
2011-03-23 17:41:45 -06:00
|
|
|
#include <linux/bitops.h>
|
2011-07-15 09:47:34 -06:00
|
|
|
#include <linux/export.h>
|
2009-03-02 02:49:28 -07:00
|
|
|
|
2009-02-24 08:30:20 -07:00
|
|
|
#include "rds.h"
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This file implements the receive side of the unconventional congestion
|
|
|
|
* management in RDS.
|
|
|
|
*
|
|
|
|
* Messages waiting in the receive queue on the receiving socket are accounted
|
|
|
|
* against the sockets SO_RCVBUF option value. Only the payload bytes in the
|
|
|
|
* message are accounted for. If the number of bytes queued equals or exceeds
|
|
|
|
* rcvbuf then the socket is congested. All sends attempted to this socket's
|
|
|
|
* address should return block or return -EWOULDBLOCK.
|
|
|
|
*
|
|
|
|
* Applications are expected to be reasonably tuned such that this situation
|
|
|
|
* very rarely occurs. An application encountering this "back-pressure" is
|
|
|
|
* considered a bug.
|
|
|
|
*
|
|
|
|
* This is implemented by having each node maintain bitmaps which indicate
|
|
|
|
* which ports on bound addresses are congested. As the bitmap changes it is
|
|
|
|
* sent through all the connections which terminate in the local address of the
|
|
|
|
* bitmap which changed.
|
|
|
|
*
|
|
|
|
* The bitmaps are allocated as connections are brought up. This avoids
|
|
|
|
* allocation in the interrupt handling path which queues messages on sockets.
|
|
|
|
* The dense bitmaps let transports send the entire bitmap on any bitmap change
|
|
|
|
* reasonably efficiently. This is much easier to implement than some
|
|
|
|
* finer-grained communication of per-port congestion. The sender does a very
|
|
|
|
* inexpensive bit test to test if the port it's about to send to is congested
|
|
|
|
* or not.
|
|
|
|
*/
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Interaction with poll is a tad tricky. We want all processes stuck in
|
|
|
|
* poll to wake up and check whether a congested destination became uncongested.
|
|
|
|
* The really sad thing is we have no idea which destinations the application
|
|
|
|
* wants to send to - we don't even know which rds_connections are involved.
|
|
|
|
* So until we implement a more flexible rds poll interface, we have to make
|
|
|
|
* do with this:
|
|
|
|
* We maintain a global counter that is incremented each time a congestion map
|
|
|
|
* update is received. Each rds socket tracks this value, and if rds_poll
|
|
|
|
* finds that the saved generation number is smaller than the global generation
|
|
|
|
* number, it wakes up the process.
|
|
|
|
*/
|
|
|
|
static atomic_t rds_cong_generation = ATOMIC_INIT(0);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Congestion monitoring
|
|
|
|
*/
|
|
|
|
static LIST_HEAD(rds_cong_monitor);
|
|
|
|
static DEFINE_RWLOCK(rds_cong_monitor_lock);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Yes, a global lock. It's used so infrequently that it's worth keeping it
|
|
|
|
* global to simplify the locking. It's only used in the following
|
|
|
|
* circumstances:
|
|
|
|
*
|
|
|
|
* - on connection buildup to associate a conn with its maps
|
|
|
|
* - on map changes to inform conns of a new map to send
|
|
|
|
*
|
|
|
|
* It's sadly ordered under the socket callback lock and the connection lock.
|
|
|
|
* Receive paths can mark ports congested from interrupt context so the
|
|
|
|
* lock masks interrupts.
|
|
|
|
*/
|
|
|
|
static DEFINE_SPINLOCK(rds_cong_lock);
|
|
|
|
static struct rb_root rds_cong_tree = RB_ROOT;
|
|
|
|
|
|
|
|
static struct rds_cong_map *rds_cong_tree_walk(__be32 addr,
|
|
|
|
struct rds_cong_map *insert)
|
|
|
|
{
|
|
|
|
struct rb_node **p = &rds_cong_tree.rb_node;
|
|
|
|
struct rb_node *parent = NULL;
|
|
|
|
struct rds_cong_map *map;
|
|
|
|
|
|
|
|
while (*p) {
|
|
|
|
parent = *p;
|
|
|
|
map = rb_entry(parent, struct rds_cong_map, m_rb_node);
|
|
|
|
|
|
|
|
if (addr < map->m_addr)
|
|
|
|
p = &(*p)->rb_left;
|
|
|
|
else if (addr > map->m_addr)
|
|
|
|
p = &(*p)->rb_right;
|
|
|
|
else
|
|
|
|
return map;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (insert) {
|
|
|
|
rb_link_node(&insert->m_rb_node, parent, p);
|
|
|
|
rb_insert_color(&insert->m_rb_node, &rds_cong_tree);
|
|
|
|
}
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* There is only ever one bitmap for any address. Connections try and allocate
|
|
|
|
* these bitmaps in the process getting pointers to them. The bitmaps are only
|
|
|
|
* ever freed as the module is removed after all connections have been freed.
|
|
|
|
*/
|
|
|
|
static struct rds_cong_map *rds_cong_from_addr(__be32 addr)
|
|
|
|
{
|
|
|
|
struct rds_cong_map *map;
|
|
|
|
struct rds_cong_map *ret = NULL;
|
|
|
|
unsigned long zp;
|
|
|
|
unsigned long i;
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
map = kzalloc(sizeof(struct rds_cong_map), GFP_KERNEL);
|
2010-01-12 12:56:44 -07:00
|
|
|
if (!map)
|
2009-02-24 08:30:20 -07:00
|
|
|
return NULL;
|
|
|
|
|
|
|
|
map->m_addr = addr;
|
|
|
|
init_waitqueue_head(&map->m_waitq);
|
|
|
|
INIT_LIST_HEAD(&map->m_conn_list);
|
|
|
|
|
|
|
|
for (i = 0; i < RDS_CONG_MAP_PAGES; i++) {
|
|
|
|
zp = get_zeroed_page(GFP_KERNEL);
|
|
|
|
if (zp == 0)
|
|
|
|
goto out;
|
|
|
|
map->m_page_addrs[i] = zp;
|
|
|
|
}
|
|
|
|
|
|
|
|
spin_lock_irqsave(&rds_cong_lock, flags);
|
|
|
|
ret = rds_cong_tree_walk(addr, map);
|
|
|
|
spin_unlock_irqrestore(&rds_cong_lock, flags);
|
|
|
|
|
2010-01-12 12:56:44 -07:00
|
|
|
if (!ret) {
|
2009-02-24 08:30:20 -07:00
|
|
|
ret = map;
|
|
|
|
map = NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
out:
|
|
|
|
if (map) {
|
|
|
|
for (i = 0; i < RDS_CONG_MAP_PAGES && map->m_page_addrs[i]; i++)
|
|
|
|
free_page(map->m_page_addrs[i]);
|
|
|
|
kfree(map);
|
|
|
|
}
|
|
|
|
|
|
|
|
rdsdebug("map %p for addr %x\n", ret, be32_to_cpu(addr));
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Put the conn on its local map's list. This is called when the conn is
|
|
|
|
* really added to the hash. It's nested under the rds_conn_lock, sadly.
|
|
|
|
*/
|
|
|
|
void rds_cong_add_conn(struct rds_connection *conn)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
rdsdebug("conn %p now on map %p\n", conn, conn->c_lcong);
|
|
|
|
spin_lock_irqsave(&rds_cong_lock, flags);
|
|
|
|
list_add_tail(&conn->c_map_item, &conn->c_lcong->m_conn_list);
|
|
|
|
spin_unlock_irqrestore(&rds_cong_lock, flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
void rds_cong_remove_conn(struct rds_connection *conn)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
rdsdebug("removing conn %p from map %p\n", conn, conn->c_lcong);
|
|
|
|
spin_lock_irqsave(&rds_cong_lock, flags);
|
|
|
|
list_del_init(&conn->c_map_item);
|
|
|
|
spin_unlock_irqrestore(&rds_cong_lock, flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
int rds_cong_get_maps(struct rds_connection *conn)
|
|
|
|
{
|
|
|
|
conn->c_lcong = rds_cong_from_addr(conn->c_laddr);
|
|
|
|
conn->c_fcong = rds_cong_from_addr(conn->c_faddr);
|
|
|
|
|
2010-01-12 12:56:44 -07:00
|
|
|
if (!(conn->c_lcong && conn->c_fcong))
|
2009-02-24 08:30:20 -07:00
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
void rds_cong_queue_updates(struct rds_cong_map *map)
|
|
|
|
{
|
|
|
|
struct rds_connection *conn;
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
spin_lock_irqsave(&rds_cong_lock, flags);
|
|
|
|
|
|
|
|
list_for_each_entry(conn, &map->m_conn_list, c_map_item) {
|
|
|
|
if (!test_and_set_bit(0, &conn->c_map_queued)) {
|
|
|
|
rds_stats_inc(s_cong_update_queued);
|
2015-02-10 11:33:37 -07:00
|
|
|
/* We cannot inline the call to rds_send_xmit() here
|
|
|
|
* for two reasons (both pertaining to a TCP transport):
|
|
|
|
* 1. When we get here from the receive path, we
|
|
|
|
* are already holding the sock_lock (held by
|
|
|
|
* tcp_v4_rcv()). So inlining calls to
|
|
|
|
* tcp_setsockopt and/or tcp_sendmsg will deadlock
|
|
|
|
* when it tries to get the sock_lock())
|
|
|
|
* 2. Interrupts are masked so that we can mark the
|
|
|
|
* the port congested from both send and recv paths.
|
|
|
|
* (See comment around declaration of rdc_cong_lock).
|
|
|
|
* An attempt to get the sock_lock() here will
|
|
|
|
* therefore trigger warnings.
|
|
|
|
* Defer the xmit to rds_send_worker() instead.
|
|
|
|
*/
|
|
|
|
queue_delayed_work(rds_wq, &conn->c_send_w, 0);
|
2009-02-24 08:30:20 -07:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
spin_unlock_irqrestore(&rds_cong_lock, flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
void rds_cong_map_updated(struct rds_cong_map *map, uint64_t portmask)
|
|
|
|
{
|
|
|
|
rdsdebug("waking map %p for %pI4\n",
|
|
|
|
map, &map->m_addr);
|
|
|
|
rds_stats_inc(s_cong_update_received);
|
|
|
|
atomic_inc(&rds_cong_generation);
|
|
|
|
if (waitqueue_active(&map->m_waitq))
|
|
|
|
wake_up(&map->m_waitq);
|
|
|
|
if (waitqueue_active(&rds_poll_waitq))
|
|
|
|
wake_up_all(&rds_poll_waitq);
|
|
|
|
|
|
|
|
if (portmask && !list_empty(&rds_cong_monitor)) {
|
|
|
|
unsigned long flags;
|
|
|
|
struct rds_sock *rs;
|
|
|
|
|
|
|
|
read_lock_irqsave(&rds_cong_monitor_lock, flags);
|
|
|
|
list_for_each_entry(rs, &rds_cong_monitor, rs_cong_list) {
|
|
|
|
spin_lock(&rs->rs_lock);
|
|
|
|
rs->rs_cong_notify |= (rs->rs_cong_mask & portmask);
|
|
|
|
rs->rs_cong_mask &= ~portmask;
|
|
|
|
spin_unlock(&rs->rs_lock);
|
|
|
|
if (rs->rs_cong_notify)
|
|
|
|
rds_wake_sk_sleep(rs);
|
|
|
|
}
|
|
|
|
read_unlock_irqrestore(&rds_cong_monitor_lock, flags);
|
|
|
|
}
|
|
|
|
}
|
2009-08-21 06:28:32 -06:00
|
|
|
EXPORT_SYMBOL_GPL(rds_cong_map_updated);
|
2009-02-24 08:30:20 -07:00
|
|
|
|
|
|
|
int rds_cong_updated_since(unsigned long *recent)
|
|
|
|
{
|
|
|
|
unsigned long gen = atomic_read(&rds_cong_generation);
|
|
|
|
|
|
|
|
if (likely(*recent == gen))
|
|
|
|
return 0;
|
|
|
|
*recent = gen;
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We're called under the locking that protects the sockets receive buffer
|
|
|
|
* consumption. This makes it a lot easier for the caller to only call us
|
|
|
|
* when it knows that an existing set bit needs to be cleared, and vice versa.
|
|
|
|
* We can't block and we need to deal with concurrent sockets working against
|
|
|
|
* the same per-address map.
|
|
|
|
*/
|
|
|
|
void rds_cong_set_bit(struct rds_cong_map *map, __be16 port)
|
|
|
|
{
|
|
|
|
unsigned long i;
|
|
|
|
unsigned long off;
|
|
|
|
|
|
|
|
rdsdebug("setting congestion for %pI4:%u in map %p\n",
|
|
|
|
&map->m_addr, ntohs(port), map);
|
|
|
|
|
|
|
|
i = be16_to_cpu(port) / RDS_CONG_MAP_PAGE_BITS;
|
|
|
|
off = be16_to_cpu(port) % RDS_CONG_MAP_PAGE_BITS;
|
|
|
|
|
2016-04-14 11:43:27 -06:00
|
|
|
set_bit_le(off, (void *)map->m_page_addrs[i]);
|
2009-02-24 08:30:20 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
void rds_cong_clear_bit(struct rds_cong_map *map, __be16 port)
|
|
|
|
{
|
|
|
|
unsigned long i;
|
|
|
|
unsigned long off;
|
|
|
|
|
|
|
|
rdsdebug("clearing congestion for %pI4:%u in map %p\n",
|
|
|
|
&map->m_addr, ntohs(port), map);
|
|
|
|
|
|
|
|
i = be16_to_cpu(port) / RDS_CONG_MAP_PAGE_BITS;
|
|
|
|
off = be16_to_cpu(port) % RDS_CONG_MAP_PAGE_BITS;
|
|
|
|
|
2016-04-14 11:43:27 -06:00
|
|
|
clear_bit_le(off, (void *)map->m_page_addrs[i]);
|
2009-02-24 08:30:20 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
static int rds_cong_test_bit(struct rds_cong_map *map, __be16 port)
|
|
|
|
{
|
|
|
|
unsigned long i;
|
|
|
|
unsigned long off;
|
|
|
|
|
|
|
|
i = be16_to_cpu(port) / RDS_CONG_MAP_PAGE_BITS;
|
|
|
|
off = be16_to_cpu(port) % RDS_CONG_MAP_PAGE_BITS;
|
|
|
|
|
2011-03-23 17:42:05 -06:00
|
|
|
return test_bit_le(off, (void *)map->m_page_addrs[i]);
|
2009-02-24 08:30:20 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
void rds_cong_add_socket(struct rds_sock *rs)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
write_lock_irqsave(&rds_cong_monitor_lock, flags);
|
|
|
|
if (list_empty(&rs->rs_cong_list))
|
|
|
|
list_add(&rs->rs_cong_list, &rds_cong_monitor);
|
|
|
|
write_unlock_irqrestore(&rds_cong_monitor_lock, flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
void rds_cong_remove_socket(struct rds_sock *rs)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
struct rds_cong_map *map;
|
|
|
|
|
|
|
|
write_lock_irqsave(&rds_cong_monitor_lock, flags);
|
|
|
|
list_del_init(&rs->rs_cong_list);
|
|
|
|
write_unlock_irqrestore(&rds_cong_monitor_lock, flags);
|
|
|
|
|
|
|
|
/* update congestion map for now-closed port */
|
|
|
|
spin_lock_irqsave(&rds_cong_lock, flags);
|
|
|
|
map = rds_cong_tree_walk(rs->rs_bound_addr, NULL);
|
|
|
|
spin_unlock_irqrestore(&rds_cong_lock, flags);
|
|
|
|
|
|
|
|
if (map && rds_cong_test_bit(map, rs->rs_bound_port)) {
|
|
|
|
rds_cong_clear_bit(map, rs->rs_bound_port);
|
|
|
|
rds_cong_queue_updates(map);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
int rds_cong_wait(struct rds_cong_map *map, __be16 port, int nonblock,
|
|
|
|
struct rds_sock *rs)
|
|
|
|
{
|
|
|
|
if (!rds_cong_test_bit(map, port))
|
|
|
|
return 0;
|
|
|
|
if (nonblock) {
|
|
|
|
if (rs && rs->rs_cong_monitor) {
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
/* It would have been nice to have an atomic set_bit on
|
|
|
|
* a uint64_t. */
|
|
|
|
spin_lock_irqsave(&rs->rs_lock, flags);
|
|
|
|
rs->rs_cong_mask |= RDS_CONG_MONITOR_MASK(ntohs(port));
|
|
|
|
spin_unlock_irqrestore(&rs->rs_lock, flags);
|
|
|
|
|
|
|
|
/* Test again - a congestion update may have arrived in
|
|
|
|
* the meantime. */
|
|
|
|
if (!rds_cong_test_bit(map, port))
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
rds_stats_inc(s_cong_send_error);
|
|
|
|
return -ENOBUFS;
|
|
|
|
}
|
|
|
|
|
|
|
|
rds_stats_inc(s_cong_send_blocked);
|
|
|
|
rdsdebug("waiting on map %p for port %u\n", map, be16_to_cpu(port));
|
|
|
|
|
|
|
|
return wait_event_interruptible(map->m_waitq,
|
|
|
|
!rds_cong_test_bit(map, port));
|
|
|
|
}
|
|
|
|
|
|
|
|
void rds_cong_exit(void)
|
|
|
|
{
|
|
|
|
struct rb_node *node;
|
|
|
|
struct rds_cong_map *map;
|
|
|
|
unsigned long i;
|
|
|
|
|
|
|
|
while ((node = rb_first(&rds_cong_tree))) {
|
|
|
|
map = rb_entry(node, struct rds_cong_map, m_rb_node);
|
|
|
|
rdsdebug("freeing map %p\n", map);
|
|
|
|
rb_erase(&map->m_rb_node, &rds_cong_tree);
|
|
|
|
for (i = 0; i < RDS_CONG_MAP_PAGES && map->m_page_addrs[i]; i++)
|
|
|
|
free_page(map->m_page_addrs[i]);
|
|
|
|
kfree(map);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Allocate a RDS message containing a congestion update.
|
|
|
|
*/
|
|
|
|
struct rds_message *rds_cong_update_alloc(struct rds_connection *conn)
|
|
|
|
{
|
|
|
|
struct rds_cong_map *map = conn->c_lcong;
|
|
|
|
struct rds_message *rm;
|
|
|
|
|
|
|
|
rm = rds_message_map_pages(map->m_page_addrs, RDS_CONG_MAP_BYTES);
|
|
|
|
if (!IS_ERR(rm))
|
|
|
|
rm->m_inc.i_hdr.h_flags = RDS_FLAG_CONG_BITMAP;
|
|
|
|
|
|
|
|
return rm;
|
|
|
|
}
|