linux/drivers/net/wireless/libertas/main.c

1310 lines
31 KiB
C
Raw Normal View History

/*
* This file contains the major functions in WLAN
* driver. It includes init, exit, open, close and main
* thread etc..
*/
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
#include <linux/module.h>
#include <linux/delay.h>
#include <linux/etherdevice.h>
#include <linux/hardirq.h>
#include <linux/netdevice.h>
#include <linux/if_arp.h>
#include <linux/kthread.h>
#include <linux/kfifo.h>
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 08:04:11 +00:00
#include <linux/slab.h>
#include <net/cfg80211.h>
#include "host.h"
#include "decl.h"
#include "dev.h"
#include "cfg.h"
#include "debugfs.h"
#include "cmd.h"
#include "mesh.h"
#define DRIVER_RELEASE_VERSION "323.p0"
const char lbs_driver_version[] = "COMM-USB8388-" DRIVER_RELEASE_VERSION
#ifdef DEBUG
"-dbg"
#endif
"";
/* Module parameters */
unsigned int lbs_debug;
EXPORT_SYMBOL_GPL(lbs_debug);
module_param_named(libertas_debug, lbs_debug, int, 0644);
unsigned int lbs_disablemesh;
EXPORT_SYMBOL_GPL(lbs_disablemesh);
module_param_named(libertas_disablemesh, lbs_disablemesh, int, 0644);
/*
* This global structure is used to send the confirm_sleep command as
* fast as possible down to the firmware.
*/
struct cmd_confirm_sleep confirm_sleep;
/*
* the table to keep region code
*/
u16 lbs_region_code_to_index[MRVDRV_MAX_REGION_CODE] =
{ 0x10, 0x20, 0x30, 0x31, 0x32, 0x40 };
/*
* FW rate table. FW refers to rates by their index in this table, not by the
* rate value itself. Values of 0x00 are
* reserved positions.
*/
static u8 fw_data_rates[MAX_RATES] =
{ 0x02, 0x04, 0x0B, 0x16, 0x00, 0x0C, 0x12,
0x18, 0x24, 0x30, 0x48, 0x60, 0x6C, 0x00
};
/**
* lbs_fw_index_to_data_rate - use index to get the data rate
*
* @idx: The index of data rate
* returns: data rate or 0
*/
u32 lbs_fw_index_to_data_rate(u8 idx)
{
if (idx >= sizeof(fw_data_rates))
idx = 0;
return fw_data_rates[idx];
}
/**
* lbs_data_rate_to_fw_index - use rate to get the index
*
* @rate: data rate
* returns: index or 0
*/
u8 lbs_data_rate_to_fw_index(u32 rate)
{
u8 i;
if (!rate)
return 0;
for (i = 0; i < sizeof(fw_data_rates); i++) {
if (rate == fw_data_rates[i])
return i;
}
return 0;
}
int lbs_set_iface_type(struct lbs_private *priv, enum nl80211_iftype type)
{
int ret = 0;
switch (type) {
case NL80211_IFTYPE_MONITOR:
ret = lbs_set_monitor_mode(priv, 1);
break;
case NL80211_IFTYPE_STATION:
if (priv->wdev->iftype == NL80211_IFTYPE_MONITOR)
ret = lbs_set_monitor_mode(priv, 0);
if (!ret)
ret = lbs_set_snmp_mib(priv, SNMP_MIB_OID_BSS_TYPE, 1);
break;
case NL80211_IFTYPE_ADHOC:
if (priv->wdev->iftype == NL80211_IFTYPE_MONITOR)
ret = lbs_set_monitor_mode(priv, 0);
if (!ret)
ret = lbs_set_snmp_mib(priv, SNMP_MIB_OID_BSS_TYPE, 2);
break;
default:
ret = -ENOTSUPP;
}
return ret;
}
int lbs_start_iface(struct lbs_private *priv)
{
struct cmd_ds_802_11_mac_address cmd;
int ret;
if (priv->power_restore) {
ret = priv->power_restore(priv);
if (ret)
return ret;
}
cmd.hdr.size = cpu_to_le16(sizeof(cmd));
cmd.action = cpu_to_le16(CMD_ACT_SET);
memcpy(cmd.macadd, priv->current_addr, ETH_ALEN);
ret = lbs_cmd_with_response(priv, CMD_802_11_MAC_ADDRESS, &cmd);
if (ret) {
lbs_deb_net("set MAC address failed\n");
goto err;
}
ret = lbs_set_iface_type(priv, priv->wdev->iftype);
if (ret) {
lbs_deb_net("set iface type failed\n");
goto err;
}
lbs_update_channel(priv);
priv->iface_running = true;
return 0;
err:
if (priv->power_save)
priv->power_save(priv);
return ret;
}
/**
* lbs_dev_open - open the ethX interface
*
* @dev: A pointer to &net_device structure
* returns: 0 or -EBUSY if monitor mode active
*/
static int lbs_dev_open(struct net_device *dev)
{
struct lbs_private *priv = dev->ml_priv;
int ret = 0;
lbs_deb_enter(LBS_DEB_NET);
if (!priv->iface_running) {
ret = lbs_start_iface(priv);
if (ret)
goto out;
}
spin_lock_irq(&priv->driver_lock);
netif_carrier_off(dev);
if (!priv->tx_pending_len)
netif_wake_queue(dev);
spin_unlock_irq(&priv->driver_lock);
out:
lbs_deb_leave_args(LBS_DEB_NET, "ret %d", ret);
return ret;
}
static bool lbs_command_queue_empty(struct lbs_private *priv)
{
unsigned long flags;
bool ret;
spin_lock_irqsave(&priv->driver_lock, flags);
ret = priv->cur_cmd == NULL && list_empty(&priv->cmdpendingq);
spin_unlock_irqrestore(&priv->driver_lock, flags);
return ret;
}
int lbs_stop_iface(struct lbs_private *priv)
{
unsigned long flags;
int ret = 0;
lbs_deb_enter(LBS_DEB_MAIN);
spin_lock_irqsave(&priv->driver_lock, flags);
priv->iface_running = false;
kfree_skb(priv->currenttxskb);
priv->currenttxskb = NULL;
priv->tx_pending_len = 0;
spin_unlock_irqrestore(&priv->driver_lock, flags);
cancel_work_sync(&priv->mcast_work);
del_timer_sync(&priv->tx_lockup_timer);
/* Disable command processing, and wait for all commands to complete */
lbs_deb_main("waiting for commands to complete\n");
wait_event(priv->waitq, lbs_command_queue_empty(priv));
lbs_deb_main("all commands completed\n");
if (priv->power_save)
ret = priv->power_save(priv);
lbs_deb_leave(LBS_DEB_MAIN);
return ret;
}
/**
* lbs_eth_stop - close the ethX interface
*
* @dev: A pointer to &net_device structure
* returns: 0
*/
static int lbs_eth_stop(struct net_device *dev)
{
struct lbs_private *priv = dev->ml_priv;
lbs_deb_enter(LBS_DEB_NET);
if (priv->connect_status == LBS_CONNECTED)
lbs_disconnect(priv, WLAN_REASON_DEAUTH_LEAVING);
spin_lock_irq(&priv->driver_lock);
netif_stop_queue(dev);
spin_unlock_irq(&priv->driver_lock);
lbs_update_mcast(priv);
cancel_delayed_work_sync(&priv->scan_work);
libertas: ensure we clean up a scan request properly Commit 2e30168b ("libertas: terminate scan when stopping interface") adds cleanup code to lbs_eth_stop to call cfg80211_scan_done if there's an outstanding cfg80211_scan_request. However, it assumes that the scan request was allocated via the cfg80211 stack. Libertas has its own internal allocation method, kept track of with priv->internal_scan. This doesn't set scan_req->wiphy, amongst other things, which results in hitting a BUG() when we call cfg80211_scan_done on the request. This provides a function to take care of the low-level scan_req cleanup details. We simply call that to deal with finishing up scan requests. The bug we were hitting was: [ 964.321495] kernel BUG at net/wireless/core.h:87! [ 964.329970] Unable to handle kernel NULL pointer dereference at virtual address 00000000 [ 964.341963] pgd = dcf80000 ... [ 964.849998] 9fe0: 00000000 beb417b8 4018e280 401e822c 60000010 00000004 00000000 00000000 [ 964.865007] [<c003104c>] (__bug+0x1c/0x28) from [<c0384ffc>] (cfg80211_scan_done+0x54/0x6c) [ 964.895324] [<c0384ffc>] (cfg80211_scan_done+0x54/0x6c) from [<bf028bac>] (lbs_eth_stop+0x10c/0x188 [libertas]) [ 964.895324] [<bf028bac>] (lbs_eth_stop+0x10c/0x188 [libertas]) from [<c03002a0>] (__dev_close_many+0x94/0xc4) [ 964.918995] [<c03002a0>] (__dev_close_many+0x94/0xc4) from [<c030037c>] (dev_close_many+0x78/0xe0) Signed-off-by: Andres Salomon <dilinger@queued.net> Acked-by: Dan Williams <dcbw@redhat.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2011-10-26 17:19:26 +00:00
if (priv->scan_req)
lbs_scan_done(priv);
netif_carrier_off(priv->dev);
if (!lbs_iface_active(priv))
lbs_stop_iface(priv);
lbs_deb_leave(LBS_DEB_NET);
return 0;
}
void lbs_host_to_card_done(struct lbs_private *priv)
{
unsigned long flags;
lbs_deb_enter(LBS_DEB_THREAD);
spin_lock_irqsave(&priv->driver_lock, flags);
del_timer(&priv->tx_lockup_timer);
priv->dnld_sent = DNLD_RES_RECEIVED;
/* Wake main thread if commands are pending */
if (!priv->cur_cmd || priv->tx_pending_len > 0) {
if (!priv->wakeup_dev_required)
wake_up(&priv->waitq);
}
spin_unlock_irqrestore(&priv->driver_lock, flags);
lbs_deb_leave(LBS_DEB_THREAD);
}
EXPORT_SYMBOL_GPL(lbs_host_to_card_done);
int lbs_set_mac_address(struct net_device *dev, void *addr)
{
int ret = 0;
struct lbs_private *priv = dev->ml_priv;
struct sockaddr *phwaddr = addr;
lbs_deb_enter(LBS_DEB_NET);
/*
* Can only set MAC address when all interfaces are down, to be written
* to the hardware when one of them is brought up.
*/
if (lbs_iface_active(priv))
return -EBUSY;
/* In case it was called from the mesh device */
dev = priv->dev;
memcpy(priv->current_addr, phwaddr->sa_data, ETH_ALEN);
memcpy(dev->dev_addr, phwaddr->sa_data, ETH_ALEN);
if (priv->mesh_dev)
memcpy(priv->mesh_dev->dev_addr, phwaddr->sa_data, ETH_ALEN);
lbs_deb_leave_args(LBS_DEB_NET, "ret %d", ret);
return ret;
}
static inline int mac_in_list(unsigned char *list, int list_len,
unsigned char *mac)
{
while (list_len) {
if (!memcmp(list, mac, ETH_ALEN))
return 1;
list += ETH_ALEN;
list_len--;
}
return 0;
}
static int lbs_add_mcast_addrs(struct cmd_ds_mac_multicast_adr *cmd,
struct net_device *dev, int nr_addrs)
{
int i = nr_addrs;
struct netdev_hw_addr *ha;
int cnt;
if ((dev->flags & (IFF_UP|IFF_MULTICAST)) != (IFF_UP|IFF_MULTICAST))
return nr_addrs;
netif_addr_lock_bh(dev);
cnt = netdev_mc_count(dev);
netdev_for_each_mc_addr(ha, dev) {
if (mac_in_list(cmd->maclist, nr_addrs, ha->addr)) {
lbs_deb_net("mcast address %s:%pM skipped\n", dev->name,
ha->addr);
cnt--;
continue;
}
if (i == MRVDRV_MAX_MULTICAST_LIST_SIZE)
break;
memcpy(&cmd->maclist[6*i], ha->addr, ETH_ALEN);
lbs_deb_net("mcast address %s:%pM added to filter\n", dev->name,
ha->addr);
i++;
cnt--;
}
netif_addr_unlock_bh(dev);
if (cnt)
return -EOVERFLOW;
return i;
}
void lbs_update_mcast(struct lbs_private *priv)
{
struct cmd_ds_mac_multicast_adr mcast_cmd;
int dev_flags = 0;
int nr_addrs;
int old_mac_control = priv->mac_control;
lbs_deb_enter(LBS_DEB_NET);
if (netif_running(priv->dev))
dev_flags |= priv->dev->flags;
if (priv->mesh_dev && netif_running(priv->mesh_dev))
dev_flags |= priv->mesh_dev->flags;
if (dev_flags & IFF_PROMISC) {
priv->mac_control |= CMD_ACT_MAC_PROMISCUOUS_ENABLE;
priv->mac_control &= ~(CMD_ACT_MAC_ALL_MULTICAST_ENABLE |
CMD_ACT_MAC_MULTICAST_ENABLE);
goto out_set_mac_control;
} else if (dev_flags & IFF_ALLMULTI) {
do_allmulti:
priv->mac_control |= CMD_ACT_MAC_ALL_MULTICAST_ENABLE;
priv->mac_control &= ~(CMD_ACT_MAC_PROMISCUOUS_ENABLE |
CMD_ACT_MAC_MULTICAST_ENABLE);
goto out_set_mac_control;
}
/* Once for priv->dev, again for priv->mesh_dev if it exists */
nr_addrs = lbs_add_mcast_addrs(&mcast_cmd, priv->dev, 0);
if (nr_addrs >= 0 && priv->mesh_dev)
nr_addrs = lbs_add_mcast_addrs(&mcast_cmd, priv->mesh_dev, nr_addrs);
if (nr_addrs < 0)
goto do_allmulti;
if (nr_addrs) {
int size = offsetof(struct cmd_ds_mac_multicast_adr,
maclist[6*nr_addrs]);
mcast_cmd.action = cpu_to_le16(CMD_ACT_SET);
mcast_cmd.hdr.size = cpu_to_le16(size);
mcast_cmd.nr_of_adrs = cpu_to_le16(nr_addrs);
lbs_cmd_async(priv, CMD_MAC_MULTICAST_ADR, &mcast_cmd.hdr, size);
priv->mac_control |= CMD_ACT_MAC_MULTICAST_ENABLE;
} else
priv->mac_control &= ~CMD_ACT_MAC_MULTICAST_ENABLE;
priv->mac_control &= ~(CMD_ACT_MAC_PROMISCUOUS_ENABLE |
CMD_ACT_MAC_ALL_MULTICAST_ENABLE);
out_set_mac_control:
if (priv->mac_control != old_mac_control)
lbs_set_mac_control(priv);
lbs_deb_leave(LBS_DEB_NET);
}
static void lbs_set_mcast_worker(struct work_struct *work)
{
struct lbs_private *priv = container_of(work, struct lbs_private, mcast_work);
lbs_update_mcast(priv);
}
void lbs_set_multicast_list(struct net_device *dev)
{
struct lbs_private *priv = dev->ml_priv;
schedule_work(&priv->mcast_work);
}
/**
* lbs_thread - handles the major jobs in the LBS driver.
* It handles all events generated by firmware, RX data received
* from firmware and TX data sent from kernel.
*
* @data: A pointer to &lbs_thread structure
* returns: 0
*/
static int lbs_thread(void *data)
{
struct net_device *dev = data;
struct lbs_private *priv = dev->ml_priv;
wait_queue_t wait;
lbs_deb_enter(LBS_DEB_THREAD);
init_waitqueue_entry(&wait, current);
for (;;) {
int shouldsleep;
u8 resp_idx;
lbs_deb_thread("1: currenttxskb %p, dnld_sent %d\n",
priv->currenttxskb, priv->dnld_sent);
add_wait_queue(&priv->waitq, &wait);
set_current_state(TASK_INTERRUPTIBLE);
spin_lock_irq(&priv->driver_lock);
if (kthread_should_stop())
shouldsleep = 0; /* Bye */
else if (priv->surpriseremoved)
shouldsleep = 1; /* We need to wait until we're _told_ to die */
else if (priv->psstate == PS_STATE_SLEEP)
shouldsleep = 1; /* Sleep mode. Nothing we can do till it wakes */
else if (priv->cmd_timed_out)
shouldsleep = 0; /* Command timed out. Recover */
else if (!priv->fw_ready)
shouldsleep = 1; /* Firmware not ready. We're waiting for it */
else if (priv->dnld_sent)
shouldsleep = 1; /* Something is en route to the device already */
else if (priv->tx_pending_len > 0)
shouldsleep = 0; /* We've a packet to send */
else if (priv->resp_len[priv->resp_idx])
shouldsleep = 0; /* We have a command response */
else if (priv->cur_cmd)
shouldsleep = 1; /* Can't send a command; one already running */
else if (!list_empty(&priv->cmdpendingq) &&
!(priv->wakeup_dev_required))
shouldsleep = 0; /* We have a command to send */
else if (kfifo_len(&priv->event_fifo))
shouldsleep = 0; /* We have an event to process */
else
shouldsleep = 1; /* No command */
if (shouldsleep) {
lbs_deb_thread("sleeping, connect_status %d, "
"psmode %d, psstate %d\n",
priv->connect_status,
priv->psmode, priv->psstate);
spin_unlock_irq(&priv->driver_lock);
schedule();
} else
spin_unlock_irq(&priv->driver_lock);
lbs_deb_thread("2: currenttxskb %p, dnld_send %d\n",
priv->currenttxskb, priv->dnld_sent);
set_current_state(TASK_RUNNING);
remove_wait_queue(&priv->waitq, &wait);
lbs_deb_thread("3: currenttxskb %p, dnld_sent %d\n",
priv->currenttxskb, priv->dnld_sent);
if (kthread_should_stop()) {
lbs_deb_thread("break from main thread\n");
break;
}
if (priv->surpriseremoved) {
lbs_deb_thread("adapter removed; waiting to die...\n");
continue;
}
lbs_deb_thread("4: currenttxskb %p, dnld_sent %d\n",
priv->currenttxskb, priv->dnld_sent);
/* Process any pending command response */
spin_lock_irq(&priv->driver_lock);
resp_idx = priv->resp_idx;
if (priv->resp_len[resp_idx]) {
spin_unlock_irq(&priv->driver_lock);
lbs_process_command_response(priv,
priv->resp_buf[resp_idx],
priv->resp_len[resp_idx]);
spin_lock_irq(&priv->driver_lock);
priv->resp_len[resp_idx] = 0;
}
spin_unlock_irq(&priv->driver_lock);
/* Process hardware events, e.g. card removed, link lost */
spin_lock_irq(&priv->driver_lock);
while (kfifo_len(&priv->event_fifo)) {
u32 event;
if (kfifo_out(&priv->event_fifo,
(unsigned char *) &event, sizeof(event)) !=
sizeof(event))
break;
spin_unlock_irq(&priv->driver_lock);
lbs_process_event(priv, event);
spin_lock_irq(&priv->driver_lock);
}
spin_unlock_irq(&priv->driver_lock);
if (priv->wakeup_dev_required) {
lbs_deb_thread("Waking up device...\n");
/* Wake up device */
if (priv->exit_deep_sleep(priv))
lbs_deb_thread("Wakeup device failed\n");
continue;
}
/* command timeout stuff */
if (priv->cmd_timed_out && priv->cur_cmd) {
struct cmd_ctrl_node *cmdnode = priv->cur_cmd;
netdev_info(dev, "Timeout submitting command 0x%04x\n",
le16_to_cpu(cmdnode->cmdbuf->command));
lbs_complete_command(priv, cmdnode, -ETIMEDOUT);
if (priv->reset_card)
priv->reset_card(priv);
}
priv->cmd_timed_out = 0;
if (!priv->fw_ready)
continue;
/* Check if we need to confirm Sleep Request received previously */
if (priv->psstate == PS_STATE_PRE_SLEEP &&
!priv->dnld_sent && !priv->cur_cmd) {
if (priv->connect_status == LBS_CONNECTED) {
lbs_deb_thread("pre-sleep, currenttxskb %p, "
"dnld_sent %d, cur_cmd %p\n",
priv->currenttxskb, priv->dnld_sent,
priv->cur_cmd);
lbs_ps_confirm_sleep(priv);
} else {
/* workaround for firmware sending
* deauth/linkloss event immediately
* after sleep request; remove this
* after firmware fixes it
*/
priv->psstate = PS_STATE_AWAKE;
netdev_alert(dev,
"ignore PS_SleepConfirm in non-connected state\n");
}
}
/* The PS state is changed during processing of Sleep Request
* event above
*/
if ((priv->psstate == PS_STATE_SLEEP) ||
(priv->psstate == PS_STATE_PRE_SLEEP))
continue;
if (priv->is_deep_sleep)
continue;
/* Execute the next command */
if (!priv->dnld_sent && !priv->cur_cmd)
lbs_execute_next_command(priv);
spin_lock_irq(&priv->driver_lock);
if (!priv->dnld_sent && priv->tx_pending_len > 0) {
int ret = priv->hw_host_to_card(priv, MVMS_DAT,
priv->tx_pending_buf,
priv->tx_pending_len);
if (ret) {
lbs_deb_tx("host_to_card failed %d\n", ret);
priv->dnld_sent = DNLD_RES_RECEIVED;
} else {
mod_timer(&priv->tx_lockup_timer,
jiffies + (HZ * 5));
}
priv->tx_pending_len = 0;
if (!priv->currenttxskb) {
/* We can wake the queues immediately if we aren't
waiting for TX feedback */
if (priv->connect_status == LBS_CONNECTED)
netif_wake_queue(priv->dev);
if (priv->mesh_dev &&
netif_running(priv->mesh_dev))
netif_wake_queue(priv->mesh_dev);
}
}
spin_unlock_irq(&priv->driver_lock);
}
del_timer(&priv->command_timer);
del_timer(&priv->tx_lockup_timer);
del_timer(&priv->auto_deepsleep_timer);
lbs_deb_leave(LBS_DEB_THREAD);
return 0;
}
/**
* lbs_setup_firmware - gets the HW spec from the firmware and sets
* some basic parameters
*
* @priv: A pointer to &struct lbs_private structure
* returns: 0 or -1
*/
static int lbs_setup_firmware(struct lbs_private *priv)
{
int ret = -1;
s16 curlevel = 0, minlevel = 0, maxlevel = 0;
lbs_deb_enter(LBS_DEB_FW);
/* Read MAC address from firmware */
memset(priv->current_addr, 0xff, ETH_ALEN);
ret = lbs_update_hw_spec(priv);
if (ret)
goto done;
/* Read power levels if available */
ret = lbs_get_tx_power(priv, &curlevel, &minlevel, &maxlevel);
if (ret == 0) {
priv->txpower_cur = curlevel;
priv->txpower_min = minlevel;
priv->txpower_max = maxlevel;
}
/* Send cmd to FW to enable 11D function */
ret = lbs_set_snmp_mib(priv, SNMP_MIB_OID_11D_ENABLE, 1);
lbs_set_mac_control(priv);
done:
lbs_deb_leave_args(LBS_DEB_FW, "ret %d", ret);
return ret;
}
int lbs_suspend(struct lbs_private *priv)
{
int ret;
lbs_deb_enter(LBS_DEB_FW);
if (priv->is_deep_sleep) {
ret = lbs_set_deep_sleep(priv, 0);
if (ret) {
netdev_err(priv->dev,
"deep sleep cancellation failed: %d\n", ret);
return ret;
}
priv->deep_sleep_required = 1;
}
ret = lbs_set_host_sleep(priv, 1);
netif_device_detach(priv->dev);
if (priv->mesh_dev)
netif_device_detach(priv->mesh_dev);
lbs_deb_leave_args(LBS_DEB_FW, "ret %d", ret);
return ret;
}
EXPORT_SYMBOL_GPL(lbs_suspend);
int lbs_resume(struct lbs_private *priv)
{
int ret;
lbs_deb_enter(LBS_DEB_FW);
ret = lbs_set_host_sleep(priv, 0);
netif_device_attach(priv->dev);
if (priv->mesh_dev)
netif_device_attach(priv->mesh_dev);
if (priv->deep_sleep_required) {
priv->deep_sleep_required = 0;
ret = lbs_set_deep_sleep(priv, 1);
if (ret)
netdev_err(priv->dev,
"deep sleep activation failed: %d\n", ret);
}
if (priv->setup_fw_on_resume)
ret = lbs_setup_firmware(priv);
lbs_deb_leave_args(LBS_DEB_FW, "ret %d", ret);
return ret;
}
EXPORT_SYMBOL_GPL(lbs_resume);
/**
* lbs_cmd_timeout_handler - handles the timeout of command sending.
* It will re-send the same command again.
*
* @data: &struct lbs_private pointer
*/
static void lbs_cmd_timeout_handler(unsigned long data)
{
struct lbs_private *priv = (struct lbs_private *)data;
unsigned long flags;
lbs_deb_enter(LBS_DEB_CMD);
spin_lock_irqsave(&priv->driver_lock, flags);
if (!priv->cur_cmd)
goto out;
netdev_info(priv->dev, "command 0x%04x timed out\n",
le16_to_cpu(priv->cur_cmd->cmdbuf->command));
priv->cmd_timed_out = 1;
/*
* If the device didn't even acknowledge the command, reset the state
* so that we don't block all future commands due to this one timeout.
*/
if (priv->dnld_sent == DNLD_CMD_SENT)
priv->dnld_sent = DNLD_RES_RECEIVED;
wake_up(&priv->waitq);
out:
spin_unlock_irqrestore(&priv->driver_lock, flags);
lbs_deb_leave(LBS_DEB_CMD);
}
/**
* lbs_tx_lockup_handler - handles the timeout of the passing of TX frames
* to the hardware. This is known to frequently happen with SD8686 when
* waking up after a Wake-on-WLAN-triggered resume.
*
* @data: &struct lbs_private pointer
*/
static void lbs_tx_lockup_handler(unsigned long data)
{
struct lbs_private *priv = (struct lbs_private *)data;
unsigned long flags;
lbs_deb_enter(LBS_DEB_TX);
spin_lock_irqsave(&priv->driver_lock, flags);
netdev_info(priv->dev, "TX lockup detected\n");
if (priv->reset_card)
priv->reset_card(priv);
priv->dnld_sent = DNLD_RES_RECEIVED;
wake_up_interruptible(&priv->waitq);
spin_unlock_irqrestore(&priv->driver_lock, flags);
lbs_deb_leave(LBS_DEB_TX);
}
/**
* auto_deepsleep_timer_fn - put the device back to deep sleep mode when
* timer expires and no activity (command, event, data etc.) is detected.
* @data: &struct lbs_private pointer
* returns: N/A
*/
static void auto_deepsleep_timer_fn(unsigned long data)
{
struct lbs_private *priv = (struct lbs_private *)data;
lbs_deb_enter(LBS_DEB_CMD);
if (priv->is_activity_detected) {
priv->is_activity_detected = 0;
} else {
if (priv->is_auto_deep_sleep_enabled &&
(!priv->wakeup_dev_required) &&
(priv->connect_status != LBS_CONNECTED)) {
struct cmd_header cmd;
lbs_deb_main("Entering auto deep sleep mode...\n");
memset(&cmd, 0, sizeof(cmd));
cmd.size = cpu_to_le16(sizeof(cmd));
lbs_cmd_async(priv, CMD_802_11_DEEP_SLEEP, &cmd,
sizeof(cmd));
}
}
mod_timer(&priv->auto_deepsleep_timer , jiffies +
(priv->auto_deep_sleep_timeout * HZ)/1000);
lbs_deb_leave(LBS_DEB_CMD);
}
int lbs_enter_auto_deep_sleep(struct lbs_private *priv)
{
lbs_deb_enter(LBS_DEB_SDIO);
priv->is_auto_deep_sleep_enabled = 1;
if (priv->is_deep_sleep)
priv->wakeup_dev_required = 1;
mod_timer(&priv->auto_deepsleep_timer ,
jiffies + (priv->auto_deep_sleep_timeout * HZ)/1000);
lbs_deb_leave(LBS_DEB_SDIO);
return 0;
}
int lbs_exit_auto_deep_sleep(struct lbs_private *priv)
{
lbs_deb_enter(LBS_DEB_SDIO);
priv->is_auto_deep_sleep_enabled = 0;
priv->auto_deep_sleep_timeout = 0;
del_timer(&priv->auto_deepsleep_timer);
lbs_deb_leave(LBS_DEB_SDIO);
return 0;
}
static int lbs_init_adapter(struct lbs_private *priv)
{
int ret;
lbs_deb_enter(LBS_DEB_MAIN);
memset(priv->current_addr, 0xff, ETH_ALEN);
priv->connect_status = LBS_DISCONNECTED;
priv->channel = DEFAULT_AD_HOC_CHANNEL;
priv->mac_control = CMD_ACT_MAC_RX_ON | CMD_ACT_MAC_TX_ON;
priv->radio_on = 1;
priv->psmode = LBS802_11POWERMODECAM;
priv->psstate = PS_STATE_FULL_POWER;
priv->is_deep_sleep = 0;
priv->is_auto_deep_sleep_enabled = 0;
priv->deep_sleep_required = 0;
priv->wakeup_dev_required = 0;
init_waitqueue_head(&priv->ds_awake_q);
init_waitqueue_head(&priv->scan_q);
priv->authtype_auto = 1;
priv->is_host_sleep_configured = 0;
priv->is_host_sleep_activated = 0;
init_waitqueue_head(&priv->host_sleep_q);
mutex_init(&priv->lock);
setup_timer(&priv->command_timer, lbs_cmd_timeout_handler,
(unsigned long)priv);
setup_timer(&priv->tx_lockup_timer, lbs_tx_lockup_handler,
(unsigned long)priv);
setup_timer(&priv->auto_deepsleep_timer, auto_deepsleep_timer_fn,
(unsigned long)priv);
INIT_LIST_HEAD(&priv->cmdfreeq);
INIT_LIST_HEAD(&priv->cmdpendingq);
spin_lock_init(&priv->driver_lock);
/* Allocate the command buffers */
if (lbs_allocate_cmd_buffer(priv)) {
pr_err("Out of memory allocating command buffers\n");
ret = -ENOMEM;
goto out;
}
priv->resp_idx = 0;
priv->resp_len[0] = priv->resp_len[1] = 0;
/* Create the event FIFO */
ret = kfifo_alloc(&priv->event_fifo, sizeof(u32) * 16, GFP_KERNEL);
kfifo: move struct kfifo in place This is a new generic kernel FIFO implementation. The current kernel fifo API is not very widely used, because it has to many constrains. Only 17 files in the current 2.6.31-rc5 used it. FIFO's are like list's a very basic thing and a kfifo API which handles the most use case would save a lot of development time and memory resources. I think this are the reasons why kfifo is not in use: - The API is to simple, important functions are missing - A fifo can be only allocated dynamically - There is a requirement of a spinlock whether you need it or not - There is no support for data records inside a fifo So I decided to extend the kfifo in a more generic way without blowing up the API to much. The new API has the following benefits: - Generic usage: For kernel internal use and/or device driver. - Provide an API for the most use case. - Slim API: The whole API provides 25 functions. - Linux style habit. - DECLARE_KFIFO, DEFINE_KFIFO and INIT_KFIFO Macros - Direct copy_to_user from the fifo and copy_from_user into the fifo. - The kfifo itself is an in place member of the using data structure, this save an indirection access and does not waste the kernel allocator. - Lockless access: if only one reader and one writer is active on the fifo, which is the common use case, no additional locking is necessary. - Remove spinlock - give the user the freedom of choice what kind of locking to use if one is required. - Ability to handle records. Three type of records are supported: - Variable length records between 0-255 bytes, with a record size field of 1 bytes. - Variable length records between 0-65535 bytes, with a record size field of 2 bytes. - Fixed size records, which no record size field. - Preserve memory resource. - Performance! - Easy to use! This patch: Since most users want to have the kfifo as part of another object, reorganize the code to allow including struct kfifo in another data structure. This requires changing the kfifo_alloc and kfifo_init prototypes so that we pass an existing kfifo pointer into them. This patch changes the implementation and all existing users. [akpm@linux-foundation.org: fix warning] Signed-off-by: Stefani Seibold <stefani@seibold.net> Acked-by: Greg Kroah-Hartman <gregkh@suse.de> Acked-by: Mauro Carvalho Chehab <mchehab@redhat.com> Acked-by: Andi Kleen <ak@linux.intel.com> Acked-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-12-21 22:37:26 +00:00
if (ret) {
pr_err("Out of memory allocating event FIFO buffer\n");
goto out;
}
out:
lbs_deb_leave_args(LBS_DEB_MAIN, "ret %d", ret);
return ret;
}
static void lbs_free_adapter(struct lbs_private *priv)
{
lbs_deb_enter(LBS_DEB_MAIN);
lbs_free_cmd_buffer(priv);
kfifo: move struct kfifo in place This is a new generic kernel FIFO implementation. The current kernel fifo API is not very widely used, because it has to many constrains. Only 17 files in the current 2.6.31-rc5 used it. FIFO's are like list's a very basic thing and a kfifo API which handles the most use case would save a lot of development time and memory resources. I think this are the reasons why kfifo is not in use: - The API is to simple, important functions are missing - A fifo can be only allocated dynamically - There is a requirement of a spinlock whether you need it or not - There is no support for data records inside a fifo So I decided to extend the kfifo in a more generic way without blowing up the API to much. The new API has the following benefits: - Generic usage: For kernel internal use and/or device driver. - Provide an API for the most use case. - Slim API: The whole API provides 25 functions. - Linux style habit. - DECLARE_KFIFO, DEFINE_KFIFO and INIT_KFIFO Macros - Direct copy_to_user from the fifo and copy_from_user into the fifo. - The kfifo itself is an in place member of the using data structure, this save an indirection access and does not waste the kernel allocator. - Lockless access: if only one reader and one writer is active on the fifo, which is the common use case, no additional locking is necessary. - Remove spinlock - give the user the freedom of choice what kind of locking to use if one is required. - Ability to handle records. Three type of records are supported: - Variable length records between 0-255 bytes, with a record size field of 1 bytes. - Variable length records between 0-65535 bytes, with a record size field of 2 bytes. - Fixed size records, which no record size field. - Preserve memory resource. - Performance! - Easy to use! This patch: Since most users want to have the kfifo as part of another object, reorganize the code to allow including struct kfifo in another data structure. This requires changing the kfifo_alloc and kfifo_init prototypes so that we pass an existing kfifo pointer into them. This patch changes the implementation and all existing users. [akpm@linux-foundation.org: fix warning] Signed-off-by: Stefani Seibold <stefani@seibold.net> Acked-by: Greg Kroah-Hartman <gregkh@suse.de> Acked-by: Mauro Carvalho Chehab <mchehab@redhat.com> Acked-by: Andi Kleen <ak@linux.intel.com> Acked-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-12-21 22:37:26 +00:00
kfifo_free(&priv->event_fifo);
del_timer(&priv->command_timer);
del_timer(&priv->tx_lockup_timer);
del_timer(&priv->auto_deepsleep_timer);
lbs_deb_leave(LBS_DEB_MAIN);
}
static const struct net_device_ops lbs_netdev_ops = {
.ndo_open = lbs_dev_open,
.ndo_stop = lbs_eth_stop,
.ndo_start_xmit = lbs_hard_start_xmit,
.ndo_set_mac_address = lbs_set_mac_address,
.ndo_set_rx_mode = lbs_set_multicast_list,
.ndo_change_mtu = eth_change_mtu,
.ndo_validate_addr = eth_validate_addr,
};
/**
* lbs_add_card - adds the card. It will probe the
* card, allocate the lbs_priv and initialize the device.
*
* @card: A pointer to card
* @dmdev: A pointer to &struct device
* returns: A pointer to &struct lbs_private structure
*/
struct lbs_private *lbs_add_card(void *card, struct device *dmdev)
{
struct net_device *dev;
struct wireless_dev *wdev;
struct lbs_private *priv = NULL;
lbs_deb_enter(LBS_DEB_MAIN);
/* Allocate an Ethernet device and register it */
wdev = lbs_cfg_alloc(dmdev);
if (IS_ERR(wdev)) {
pr_err("cfg80211 init failed\n");
goto done;
}
wdev->iftype = NL80211_IFTYPE_STATION;
priv = wdev_priv(wdev);
priv->wdev = wdev;
if (lbs_init_adapter(priv)) {
pr_err("failed to initialize adapter structure\n");
goto err_wdev;
}
dev = alloc_netdev(0, "wlan%d", ether_setup);
if (!dev) {
dev_err(dmdev, "no memory for network device instance\n");
goto err_adapter;
}
dev->ieee80211_ptr = wdev;
dev->ml_priv = priv;
SET_NETDEV_DEV(dev, dmdev);
wdev->netdev = dev;
priv->dev = dev;
dev->netdev_ops = &lbs_netdev_ops;
dev->watchdog_timeo = 5 * HZ;
dev->ethtool_ops = &lbs_ethtool_ops;
dev->flags |= IFF_BROADCAST | IFF_MULTICAST;
priv->card = card;
strcpy(dev->name, "wlan%d");
lbs_deb_thread("Starting main thread...\n");
init_waitqueue_head(&priv->waitq);
priv->main_thread = kthread_run(lbs_thread, dev, "lbs_main");
if (IS_ERR(priv->main_thread)) {
lbs_deb_thread("Error creating main thread.\n");
goto err_ndev;
}
priv->work_thread = create_singlethread_workqueue("lbs_worker");
INIT_WORK(&priv->mcast_work, lbs_set_mcast_worker);
priv->wol_criteria = EHS_REMOVE_WAKEUP;
priv->wol_gpio = 0xff;
priv->wol_gap = 20;
priv->ehs_remove_supported = true;
goto done;
err_ndev:
free_netdev(dev);
err_adapter:
lbs_free_adapter(priv);
err_wdev:
lbs_cfg_free(priv);
priv = NULL;
done:
lbs_deb_leave_args(LBS_DEB_MAIN, "priv %p", priv);
return priv;
}
EXPORT_SYMBOL_GPL(lbs_add_card);
void lbs_remove_card(struct lbs_private *priv)
{
struct net_device *dev = priv->dev;
lbs_deb_enter(LBS_DEB_MAIN);
lbs_remove_mesh(priv);
lbs_scan_deinit(priv);
/* worker thread destruction blocks on the in-flight command which
* should have been cleared already in lbs_stop_card().
*/
lbs_deb_main("destroying worker thread\n");
destroy_workqueue(priv->work_thread);
lbs_deb_main("done destroying worker thread\n");
if (priv->psmode == LBS802_11POWERMODEMAX_PSP) {
priv->psmode = LBS802_11POWERMODECAM;
lbs_set_ps_mode(priv, PS_MODE_ACTION_EXIT_PS, true);
}
if (priv->is_deep_sleep) {
priv->is_deep_sleep = 0;
wake_up_interruptible(&priv->ds_awake_q);
}
priv->is_host_sleep_configured = 0;
priv->is_host_sleep_activated = 0;
wake_up_interruptible(&priv->host_sleep_q);
/* Stop the thread servicing the interrupts */
priv->surpriseremoved = 1;
kthread_stop(priv->main_thread);
lbs_free_adapter(priv);
lbs_cfg_free(priv);
free_netdev(dev);
lbs_deb_leave(LBS_DEB_MAIN);
}
EXPORT_SYMBOL_GPL(lbs_remove_card);
int lbs_rtap_supported(struct lbs_private *priv)
{
if (MRVL_FW_MAJOR_REV(priv->fwrelease) == MRVL_FW_V5)
return 1;
/* newer firmware use a capability mask */
return ((MRVL_FW_MAJOR_REV(priv->fwrelease) >= MRVL_FW_V10) &&
(priv->fwcapinfo & MESH_CAPINFO_ENABLE_MASK));
}
int lbs_start_card(struct lbs_private *priv)
{
struct net_device *dev = priv->dev;
int ret = -1;
lbs_deb_enter(LBS_DEB_MAIN);
/* poke the firmware */
ret = lbs_setup_firmware(priv);
if (ret)
goto done;
if (!lbs_disablemesh)
lbs_init_mesh(priv);
else
pr_info("%s: mesh disabled\n", dev->name);
if (lbs_cfg_register(priv)) {
pr_err("cannot register device\n");
goto done;
}
if (lbs_mesh_activated(priv))
lbs_start_mesh(priv);
lbs_debugfs_init_one(priv, dev);
netdev_info(dev, "Marvell WLAN 802.11 adapter\n");
ret = 0;
done:
lbs_deb_leave_args(LBS_DEB_MAIN, "ret %d", ret);
return ret;
}
EXPORT_SYMBOL_GPL(lbs_start_card);
void lbs_stop_card(struct lbs_private *priv)
{
struct net_device *dev;
lbs_deb_enter(LBS_DEB_MAIN);
if (!priv)
goto out;
dev = priv->dev;
netif_stop_queue(dev);
netif_carrier_off(dev);
lbs_debugfs_remove_one(priv);
lbs_deinit_mesh(priv);
unregister_netdev(dev);
out:
lbs_deb_leave(LBS_DEB_MAIN);
}
EXPORT_SYMBOL_GPL(lbs_stop_card);
void lbs_queue_event(struct lbs_private *priv, u32 event)
{
unsigned long flags;
lbs_deb_enter(LBS_DEB_THREAD);
spin_lock_irqsave(&priv->driver_lock, flags);
if (priv->psstate == PS_STATE_SLEEP)
priv->psstate = PS_STATE_AWAKE;
kfifo_in(&priv->event_fifo, (unsigned char *) &event, sizeof(u32));
wake_up(&priv->waitq);
spin_unlock_irqrestore(&priv->driver_lock, flags);
lbs_deb_leave(LBS_DEB_THREAD);
}
EXPORT_SYMBOL_GPL(lbs_queue_event);
void lbs_notify_command_response(struct lbs_private *priv, u8 resp_idx)
{
lbs_deb_enter(LBS_DEB_THREAD);
if (priv->psstate == PS_STATE_SLEEP)
priv->psstate = PS_STATE_AWAKE;
/* Swap buffers by flipping the response index */
BUG_ON(resp_idx > 1);
priv->resp_idx = resp_idx;
wake_up(&priv->waitq);
lbs_deb_leave(LBS_DEB_THREAD);
}
EXPORT_SYMBOL_GPL(lbs_notify_command_response);
/**
* lbs_get_firmware - Retrieves two-stage firmware
*
* @dev: A pointer to &device structure
* @user_helper: User-defined helper firmware file
* @user_mainfw: User-defined main firmware file
* @card_model: Bus-specific card model ID used to filter firmware table
* elements
* @fw_table: Table of firmware file names and device model numbers
* terminated by an entry with a NULL helper name
* @helper: On success, the helper firmware; caller must free
* @mainfw: On success, the main firmware; caller must free
*
* returns: 0 on success, non-zero on failure
*/
int lbs_get_firmware(struct device *dev, const char *user_helper,
const char *user_mainfw, u32 card_model,
const struct lbs_fw_table *fw_table,
const struct firmware **helper,
const struct firmware **mainfw)
{
const struct lbs_fw_table *iter;
int ret;
BUG_ON(helper == NULL);
BUG_ON(mainfw == NULL);
/* Try user-specified firmware first */
if (user_helper) {
ret = request_firmware(helper, user_helper, dev);
if (ret) {
dev_err(dev, "couldn't find helper firmware %s\n",
user_helper);
goto fail;
}
}
if (user_mainfw) {
ret = request_firmware(mainfw, user_mainfw, dev);
if (ret) {
dev_err(dev, "couldn't find main firmware %s\n",
user_mainfw);
goto fail;
}
}
if (*helper && *mainfw)
return 0;
/* Otherwise search for firmware to use. If neither the helper or
* the main firmware were specified by the user, then we need to
* make sure that found helper & main are from the same entry in
* fw_table.
*/
iter = fw_table;
while (iter && iter->helper) {
if (iter->model != card_model)
goto next;
if (*helper == NULL) {
ret = request_firmware(helper, iter->helper, dev);
if (ret)
goto next;
/* If the device has one-stage firmware (ie cf8305) and
* we've got it then we don't need to bother with the
* main firmware.
*/
if (iter->fwname == NULL)
return 0;
}
if (*mainfw == NULL) {
ret = request_firmware(mainfw, iter->fwname, dev);
if (ret && !user_helper) {
/* Clear the helper if it wasn't user-specified
* and the main firmware load failed, to ensure
* we don't have mismatched firmware pairs.
*/
release_firmware(*helper);
*helper = NULL;
}
}
if (*helper && *mainfw)
return 0;
next:
iter++;
}
fail:
/* Failed */
if (*helper) {
release_firmware(*helper);
*helper = NULL;
}
if (*mainfw) {
release_firmware(*mainfw);
*mainfw = NULL;
}
return -ENOENT;
}
EXPORT_SYMBOL_GPL(lbs_get_firmware);
static int __init lbs_init_module(void)
{
lbs_deb_enter(LBS_DEB_MAIN);
memset(&confirm_sleep, 0, sizeof(confirm_sleep));
confirm_sleep.hdr.command = cpu_to_le16(CMD_802_11_PS_MODE);
confirm_sleep.hdr.size = cpu_to_le16(sizeof(confirm_sleep));
confirm_sleep.action = cpu_to_le16(PS_MODE_ACTION_SLEEP_CONFIRMED);
lbs_debugfs_init();
lbs_deb_leave(LBS_DEB_MAIN);
return 0;
}
static void __exit lbs_exit_module(void)
{
lbs_deb_enter(LBS_DEB_MAIN);
lbs_debugfs_remove();
lbs_deb_leave(LBS_DEB_MAIN);
}
module_init(lbs_init_module);
module_exit(lbs_exit_module);
MODULE_DESCRIPTION("Libertas WLAN Driver Library");
MODULE_AUTHOR("Marvell International Ltd.");
MODULE_LICENSE("GPL");