madwifi/ath/if_athvar.h

1001 lines
37 KiB
C
Raw Normal View History

/*-
* Copyright (c) 2002-2005 Sam Leffler, Errno Consulting
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer,
without modification.
* 2. Redistributions in binary form must reproduce at minimum a disclaimer
* similar to the "NO WARRANTY" disclaimer below ("Disclaimer") and any
* redistribution must be conditioned upon including a substantially
* similar Disclaimer requirement for further binary redistribution.
* 3. Neither the names of the above-listed copyright holders nor the names
* of any contributors may be used to endorse or promote products derived
* from this software without specific prior written permission.
*
* Alternatively, this software may be distributed under the terms of the
* GNU General Public License ("GPL") version 2 as published by the Free
* Software Foundation.
*
* NO WARRANTY
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
* LIMITED TO, THE IMPLIED WARRANTIES OF NONINFRINGEMENT, MERCHANTIBILITY
* AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL
* THE COPYRIGHT HOLDERS OR CONTRIBUTORS BE LIABLE FOR SPECIAL, EXEMPLARY,
* OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
* INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER
* IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
* ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF
* THE POSSIBILITY OF SUCH DAMAGES.
*
* $Id$
*/
/*
* Defintions for the Atheros Wireless LAN controller driver.
*/
#ifndef _DEV_ATH_ATHVAR_H
#define _DEV_ATH_ATHVAR_H
#include "ah.h"
#include "ah_desc.h"
#include "ah_os.h"
#include "if_athioctl.h"
#include "net80211/ieee80211.h" /* XXX for WME_NUM_AC */
#include <asm/io.h>
#include <linux/list.h>
#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,15)
# include <asm/bitops.h>
#endif
/*
* Deduce if tasklets are available. If not then
* fall back to using the immediate work queue.
*/
#include <linux/interrupt.h>
#ifdef DECLARE_TASKLET /* native tasklets */
#define ATH_TQ_STRUCT tasklet_struct
#define ATH_INIT_TQUEUE(a,b,c) tasklet_init((a), (b), (unsigned long)(c))
#define ATH_SCHEDULE_TQUEUE(a,b) tasklet_schedule((a))
typedef unsigned long TQUEUE_ARG;
#define mark_bh(a) do {} while (0)
#else /* immediate work queue */
#define ATH_TQ_STRUCT tq_struct
#define ATH_INIT_TQUEUE(a,b,c) INIT_TQUEUE(a,b,c)
#define ATH_SCHEDULE_TQUEUE(a,b) do { \
*(b) |= queue_task((a), &tq_immediate); \
} while (0)
typedef void *TQUEUE_ARG;
#define tasklet_disable(t) do { (void) t; local_bh_disable(); } while (0)
#define tasklet_enable(t) do { (void) t; local_bh_enable(); } while (0)
#endif /* !DECLARE_TASKLET */
#include <linux/sched.h>
#if LINUX_VERSION_CODE < KERNEL_VERSION(2,5,41)
#include <linux/tqueue.h>
#define work_struct tq_struct
#define schedule_work(t) schedule_task((t))
#define flush_scheduled_work() flush_scheduled_tasks()
#define ATH_INIT_WORK(t, f) do { \
memset((t), 0, sizeof(struct tq_struct)); \
(t)->routine = (void (*)(void*)) (f); \
(t)->data=(void *) (t); \
} while (0)
#else
#include <linux/workqueue.h>
#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,20)
#define ATH_INIT_WORK(_t, _f) INIT_WORK((_t), (void (*)(void *))(_f), (_t));
#else
#define ATH_INIT_WORK(_t, _f) INIT_WORK((_t), (_f));
#endif
#endif /* KERNEL_VERSION < 2.5.41 */
/*
* Guess how the interrupt handler should work.
*/
#if !defined(IRQ_NONE)
typedef void irqreturn_t;
#define IRQ_NONE
#define IRQ_HANDLED
#endif /* !defined(IRQ_NONE) */
#ifndef SET_MODULE_OWNER
#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
#define SET_MODULE_OWNER(dev) do { \
dev->owner = THIS_MODULE; \
} while (0)
#else
#define SET_MODULE_OWNER(dev) do { } while (0)
#endif
#endif
#ifndef SET_NETDEV_DEV
#define SET_NETDEV_DEV(ndev, pdev)
#endif
#ifdef to_net_dev
#define ATH_GET_NETDEV_DEV(ndev) ((ndev)->dev.parent)
#else
#define ATH_GET_NETDEV_DEV(ndev) ((ndev)->class_dev.dev)
#endif
#if LINUX_VERSION_CODE < KERNEL_VERSION(2,4,23)
static inline struct net_device *_alloc_netdev(int sizeof_priv, const char *mask,
void (*setup)(struct net_device *))
{
struct net_device *dev;
int alloc_size;
/* ensure 32-byte alignment of the private area */
alloc_size = sizeof (*dev) + sizeof_priv + 31;
dev = (struct net_device *) kmalloc (alloc_size, GFP_KERNEL);
if (dev == NULL)
{
printk(KERN_ERR "alloc_dev: Unable to allocate device memory.\n");
return NULL;
}
memset(dev, 0, alloc_size);
if (sizeof_priv)
dev->priv = (void *) (((long)(dev + 1) + 31) & ~31);
setup(dev);
strcpy(dev->name, mask);
return dev;
}
/* Avoid name collision - some vendor kernels backport alloc_netdev() */
#undef alloc_netdev
#define alloc_netdev(s,m,d) _alloc_netdev(s, m, d)
/* Some vendors backport PDE, so make it a macro here */
#undef PDE
#define PDE(inode) ((struct proc_dir_entry *)(inode)->u.generic_ip)
#endif
/*
* Macro to expand scalars to 64-bit objects
*/
#define ito64(x) (sizeof(x)==8) ? (((unsigned long long int)(x)) & (0xff)) : \
(sizeof(x)==16) ? (((unsigned long long int)(x)) & 0xffff) : \
((sizeof(x)==32) ? (((unsigned long long int)(x)) & 0xffffffff): (unsigned long long int)(x))
/*
* Deal with the sysctl handler api changing.
*/
#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,8)
#define ATH_SYSCTL_DECL(f, ctl, write, filp, buffer, lenp, ppos) \
f(ctl_table *ctl, int write, struct file *filp, \
void __user *buffer, size_t *lenp)
#define ATH_SYSCTL_PROC_DOINTVEC(ctl, write, filp, buffer, lenp, ppos) \
proc_dointvec(ctl, write, filp, buffer, lenp)
#else /* LINUX_VERSION_CODE < KERNEL_VERSION(2,6,8) */
#define ATH_SYSCTL_DECL(f, ctl, write, filp, buffer, lenp, ppos) \
f(ctl_table *ctl, int write, struct file *filp, \
void __user *buffer, size_t *lenp, loff_t *ppos)
#define ATH_SYSCTL_PROC_DOINTVEC(ctl, write, filp, buffer, lenp, ppos) \
proc_dointvec(ctl, write, filp, buffer, lenp, ppos)
#endif
#define ATH_TIMEOUT 1000
#define ATH_DFS_WAIT_MIN_PERIOD 60 /* DFS wait is 60 seconds, per
* FCC/ETSI regulations. */
#define ATH_DFS_WAIT_SHORT_POLL_PERIOD 2 /* 2 seconds, for consecutive
* waits if not done yet. */
#define ATH_DFS_AVOID_MIN_PERIOD 1800 /* 30 minutes, per FCC/ETSI
* regulations */
#define ATH_DFS_TEST_RETURN_PERIOD 15 /* 15 seconds -- for mute test
* only */
#define ATH_LONG_CALINTERVAL_SECS 30 /* 30 seconds between calibrations */
#define ATH_SHORT_CALINTERVAL_SECS 1 /* 1 second between calibrations */
/*
* Maximum acceptable MTU
* MAXFRAMEBODY - WEP - QOS - RSN/WPA:
* 2312 - 8 - 2 - 12 = 2290
*/
#define ATH_MAX_MTU 2290
#define ATH_MIN_MTU 32
/* number of RX buffers */
#define ATH_RXBUF 100
/* number of TX buffers */
#define ATH_TXBUF 300
/* minimum number of beacon buffers */
#define ATH_MAXVAPS_MIN 2
/* maximum number of beacon buffers */
#define ATH_MAXVAPS_MAX 64
/* default number of beacon buffers */
#define ATH_MAXVAPS_DEFAULT 4
This patch augments the current reference counting code with: * Counters for total outstanding instances for each resource type (skb, ath_node and ath_buf) * One pair of acquisition/release functions per resource type in unlocked and one in locked * Adds some more _debug versions of functions in the call chain that acquire/release resources so that the original func/line in the driver as well as the func/line that affected the resource use can be shown in the trace. Intermediate stack frames aren't necessary to trace the leaks. * Changes naming convention for "lock-required" functions to suffix _locked for the versions that expect locking, to be consistent with some other places in the code. * Consolidate debug messages to the helper functions that actually affect the reference count or acquire/release a resource * Additional sanity checks and leak detection (esp for detecting node ref leaks through skb) * skb references are nulled out by the new sbk unref/free function. I've tested these changes extensively and found lots of cases where we didn't get enough node references when cloning skbuff, and where the kernel drops packets due to performance issues and leaks our node references. With these changes and the tracing enabled I have verified that: * TX BUF: tx buffers always go down to zero when the tx queue is done, and you can watch tx queue usage ratio go up and down again as the driver is working. There are no leaks here at the moment, although there *are* some in the madwifi-dfs branch during CAC at the moment. * skbuff leaks in all the common flows are fixed. We were leaking node references in a lot of places where kernel was dropping skb's due to congestion and we were failing to increment node references when cloning skbuffs. These are now detected, as are skbuffs that are reaped by the kernel while still holding a node reference. * the ath_node count works correctly and on an idle system we get about 5 references per station table node, with 2 node instances per VAP. One for the bss and one for the node in the station table, I believe. The ath_node count goes up and down but always lands back at the stable number based on the vaps you have configured and the number of actual stations in the station table. The point here is that it's pretty constant what you will see over time, despite excessive node creation/release in our code during input (esp input_all). Thank god for the slab allocator. git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2902 0192ed92-7a03-0410-a25b-9323aeb14dbd
2007-11-21 23:14:11 +03:00
/* number of TX buffers reserved for mgt frames */
#define ATH_TXBUF_MGT_RESERVED 20
/*
* dynamic turbo specific macros.
*/
#define ATH_TURBO_UP_THRESH 750000 /* bytes/sec */
#define ATH_TURBO_DN_THRESH 1000000 /* bytes/sec */
#define ATH_TURBO_PERIOD_HOLD 1 /* in seconds */
/*
* The only case where we see skbuff chains is due to FF aggregation in
* the driver.
*/
#ifdef ATH_SUPERG_FF
#define ATH_TXDESC 2 /* number of descriptors per buffer */
#else
#define ATH_TXDESC 1 /* number of descriptors per buffer */
#endif
#define ATH_TXMAXTRY 11 /* max number of transmit attempts */
/* Compress settings */
#define ATH_COMP_THRESHOLD 256 /* no compression for frames
longer than this threshold */
#define ATH_COMP_PROC_NO_COMP_NO_CCS 3
#define ATH_COMP_PROC_NO_COMP_ADD_CCS 2
#define ATH_COMP_PROC_COMP_NO_OPTIAML 1
#define ATH_COMP_PROC_COMP_OPTIMAL 0
#define ATH_DEFAULT_COMP_PROC ATH_COMP_PROC_COMP_OPTIMAL
#define INVALID_DECOMP_INDEX 0xFFFF
#define WEP_IV_FIELD_SIZE 4 /* wep IV field size */
#define WEP_ICV_FIELD_SIZE 4 /* wep ICV field size */
#define AES_ICV_FIELD_SIZE 8 /* AES ICV field size */
#define EXT_IV_FIELD_SIZE 4 /* ext IV field size */
/* XR specific macros */
#define XR_DEFAULT_GRPPOLL_RATE_STR "0.25 1 1 3 3 6 6 20"
#define GRPPOLL_RATE_STR_LEN 64
#define XR_SLOT_DELAY 30 /* in usec */
#define XR_AIFS 0
#define XR_NUM_RATES 5
#define XR_NUM_SUP_RATES 8
/* XR uplink should have same cwmin/cwmax value */
#define XR_CWMIN_CWMAX 7
#define XR_DATA_AIFS 3
#define XR_DATA_CWMIN 31
#define XR_DATA_CWMAX 1023
/* pick the threshold so that we meet most of the regulatory constraints */
#define XR_FRAGMENTATION_THRESHOLD 540
#define XR_TELEC_FRAGMENTATION_THRESHOLD 442
#define XR_MAX_GRP_POLL_PERIOD 1000 /* Maximum Group Poll Periodicity */
#define XR_DEFAULT_POLL_INTERVAL 100
#define XR_MIN_POLL_INTERVAL 30
#define XR_MAX_POLL_INTERVAL 1000
#define XR_DEFAULT_POLL_COUNT 32
#define XR_MIN_POLL_COUNT 16
#define XR_MAX_POLL_COUNT 64
#define XR_POLL_UPDATE_PERIOD 10 /* number of XR beacons */
#define XR_GRPPOLL_PERIOD_FACTOR 5 /* factor used in calculating grp poll interval */
#define XR_4MS_FRAG_THRESHOLD 128 /* fragmentation threshold for 4msec frame limit */
/*
* Maximum Values in ms for group poll periodicty
*/
#define GRP_POLL_PERIOD_NO_XR_STA_MAX 100
#define GRP_POLL_PERIOD_XR_STA_MAX 30
enum {
CCA_BG = 15,
CCA_A = 4,
CCA_PUREG = 4, /* pure G */
};
/*
* Percentage of the configured poll periodicity
*/
#define GRP_POLL_PERIOD_FACTOR_XR_STA 30 /* When XR Stations associated freq is 30% higher */
#define A_MAX(a,b) ((a) > (b) ? (a) : (b))
/*
* Macros to obtain the Group Poll Periodicity in various situations
*
* Curerntly there are the two cases
* (a) When there are no XR STAs associated
* (b) When there is atleast one XR STA associated
*/
#define GRP_POLL_PERIOD_NO_XR_STA(sc) (sc->sc_xrpollint)
#define GRP_POLL_PERIOD_XR_STA(sc) \
A_MAX(GRP_POLL_PERIOD_FACTOR_XR_STA * (sc->sc_xrpollint / 100), GRP_POLL_PERIOD_XR_STA_MAX)
/*
* When there are no XR STAs and a valid double chirp is received then the Group Polls are
* transmitted for 10 seconds from the time of the last valid double double-chirp
*/
#define NO_XR_STA_GRPPOLL_TX_DUR 10000
/*
* The key cache is used for h/w cipher state and also for
* tracking station state such as the current tx antenna.
* We also setup a mapping table between key cache slot indices
* and station state to short-circuit node lookups on rx.
* Different parts have different size key caches. We handle
* up to ATH_KEYMAX entries (could dynamically allocate state).
*/
#define ATH_KEYMAX 128 /* max key cache size we handle */
#define ATH_KEYBYTES (ATH_KEYMAX / NBBY) /* storage space in bytes */
#ifdef ATH_REVERSE_ENGINEERING
#define MIN_REGISTER_ADDRESS 0x0000 /* PCI register addresses are taken as releative to the appropriate BAR */
#define MAX_REGISTER_ADDRESS 0xc000 /* AR5212/AR5213 seems to have a 48k address range */
#define MAX_REGISTER_NAME_LEN 32 /* Maximum length of register nicknames in debug output */
#endif /* #ifdef ATH_REVERSE_ENGINEERING */
/*
* Convert from net80211 layer values to Ath layer values. Hopefully this will
* be optimised away when the two constants are the same.
*/
typedef unsigned int ath_keyix_t;
#define ATH_KEY(_keyix) ((_keyix == IEEE80211_KEYIX_NONE) ? HAL_TXKEYIX_INVALID : _keyix)
#define ATH_MIN_FF_RATE 12000 /* min rate for ff aggregation in kbps */
#define ATH_MIN_FF_RATE 12000 /* min rate for ff aggregation in kbps */
struct ath_buf;
typedef STAILQ_HEAD(, ath_buf) ath_bufhead;
/* driver-specific node state */
struct ath_node {
struct ieee80211_node an_node; /* base class */
u_int16_t an_decomp_index; /* decompression mask index */
u_int32_t an_avgrssi; /* average rssi over all rx frames */
u_int8_t an_prevdatarix; /* rate ix of last data frame */
u_int16_t an_minffrate; /* min rate in kbps for ff to aggregate */
struct ath_buf *an_tx_ffbuf[WME_NUM_AC]; /* ff staging area */
ath_bufhead an_uapsd_q; /* U-APSD delivery queue */
int an_uapsd_qdepth; /* U-APSD delivery queue depth */
ath_bufhead an_uapsd_overflowq; /* U-APSD overflow queue (for > MaxSp frames) */
int an_uapsd_overflowqdepth; /* U-APSD overflow queue depth */
spinlock_t an_uapsd_lock; /* U-APSD delivery queue lock */
/* variable-length rate control state follows */
};
#define ATH_NODE(_n) ((struct ath_node *)(_n))
#define ATH_NODE_CONST(ni) ((const struct ath_node *)(ni))
#define ATH_NODE_UAPSD_LOCK_INIT(_an) spin_lock_init(&(_an)->an_uapsd_lock)
#define ATH_NODE_UAPSD_LOCK_IRQ(_an) do { \
unsigned long __an_uapsd_lockflags; \
ATH_NODE_UAPSD_LOCK_CHECK(_an); \
spin_lock_irqsave(&(_an)->an_uapsd_lock, __an_uapsd_lockflags);
#define ATH_NODE_UAPSD_UNLOCK_IRQ(_an) \
ATH_NODE_UAPSD_LOCK_ASSERT(_an); \
spin_unlock_irqrestore(&(_an)->an_uapsd_lock, __an_uapsd_lockflags); \
} while (0)
#define ATH_NODE_UAPSD_LOCK_IRQ_INSIDE(_an) do { \
ATH_NODE_UAPSD_LOCK_CHECK(_an); \
spin_lock(&(_an)->an_uapsd_lock); \
} while(0)
#define ATH_NODE_UAPSD_UNLOCK_IRQ_INSIDE(_an) do { \
ATH_NODE_UAPSD_LOCK_ASSERT(_an); \
spin_unlock(&(_an)->an_uapsd_lock); \
} while (0)
#define ATH_NODE_UAPSD_UNLOCK_IRQ_EARLY(_an) \
ATH_NODE_UAPSD_LOCK_ASSERT(_an); \
spin_unlock_irqrestore(&(_an)->an_uapsd_lock, __an_uapsd_lockflags);
#if (defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)) && defined(spin_is_locked)
#define ATH_NODE_UAPSD_LOCK_ASSERT(_an) \
KASSERT(spin_is_locked(&(_an)->an_uapsd_lock), ("uapsd not locked!"))
#if (defined(ATH_DEBUG_SPINLOCKS))
#define ATH_NODE_UAPSD_LOCK_CHECK(_an) do { \
if (spin_is_locked(&(_an)->an_uapsd_lock)) \
printk(KERN_DEBUG "%s:%d - about to block on uapsd lock!\n", __func__, __LINE__); \
} while(0)
#else /* #if (defined(ATH_DEBUG_SPINLOCKS)) */
#define ATH_NODE_UAPSD_LOCK_CHECK(_an)
#endif /* #if (defined(ATH_DEBUG_SPINLOCKS)) */
#else
#define ATH_NODE_UAPSD_LOCK_ASSERT(_an)
#define ATH_NODE_UAPSD_LOCK_CHECK(_an)
#endif
#define ATH_RSSI_LPF_LEN 10
#define ATH_RSSI_DUMMY_MARKER 0x127
#define ATH_EP_MUL(x, mul) ((x) * (mul))
#define ATH_RSSI_IN(x) (ATH_EP_MUL((x), HAL_RSSI_EP_MULTIPLIER))
#define ATH_LPF_RSSI(x, y, len) \
((x != ATH_RSSI_DUMMY_MARKER) ? (((x) * ((len) - 1) + (y)) / (len)) : (y))
#define ATH_RSSI_LPF(x, y) do { \
if ((y) >= -20) \
x = ATH_LPF_RSSI((x), ATH_RSSI_IN((y)), ATH_RSSI_LPF_LEN); \
} while (0)
#define ATH_ANTENNA_DIFF 2 /* Num frames difference in
* tx to flip default recv
* antenna
*/
struct ath_buf {
/* FFXXX: convert both list types to TAILQ to save a field? */
STAILQ_ENTRY(ath_buf) bf_list;
#ifdef ATH_SUPERG_FF
TAILQ_ENTRY(ath_buf) bf_stagelist; /* fast-frame staging list */
#endif
struct ath_desc *bf_desc; /* virtual addr of desc */
struct ath_desc_status bf_dsstatus; /* tx/rx descriptor status */
dma_addr_t bf_daddr; /* physical addr of desc */
struct sk_buff *bf_skb; /* skbuff for buf */
dma_addr_t bf_skbaddr; /* physical addr of skb data - always used by one desc*/
u_int32_t bf_status; /* status flags */
u_int16_t bf_flags; /* tx descriptor flags */
u_int64_t bf_tsf;
int16_t bf_channoise;
#ifdef ATH_SUPERG_FF
/* XXX: combine this with bf_skbaddr if it ever changes to accommodate
* multiple segments.
*/
u_int16_t bf_numdescff; /* number of descs used for FF (these are extra) */
u_int32_t bf_queueage; /* "age" of txq when this buffer placed on stageq */
dma_addr_t bf_skbaddrff[ATH_TXDESC - 1]; /* extra addrs for FF */
#endif
int bf_taken_at_line; /* XXX: Want full alloc backtrace */
const char* bf_taken_at_func;
};
/* XXX: only managed for rx at the moment */
#define ATH_BUFSTATUS_RXDESC_DONE 0x00000001 /* rx descriptor processing complete, desc processed by hal */
#define ATH_BUFSTATUS_RADAR_DONE 0x00000002 /* marker to indicate a PHYERR for radar pulse
has already been handled. We may receive
multiple interrupts before the rx_tasklet
clears the queue */
#define ATH_BUFSTATUS_RXTSTAMP 0x00000004 /* RX timestamps needs to be adjusted */
/* DMA state for tx/rx descriptors. */
struct ath_descdma {
const char *dd_name;
struct ath_desc *dd_desc; /* descriptors */
dma_addr_t dd_desc_paddr; /* physical addr of dd_desc */
size_t dd_desc_len; /* size of dd_desc */
unsigned int dd_ndesc;
unsigned int dd_nbuf;
struct ath_buf *dd_bufptr; /* associated buffers */
};
struct ath_hal;
struct ath_desc;
struct ath_ratectrl;
struct ath_tx99;
struct proc_dir_entry;
/*
* Data transmit queue state. One of these exists for each
* hardware transmit queue. Packets sent to us from above
* are assigned to queues based on their priority. Not all
* devices support a complete set of hardware transmit queues.
* For those devices the array sc_ac2q will map multiple
* priorities to fewer hardware queues (typically all to one
* hardware queue).
*/
struct ath_txq {
u_int axq_qnum; /* hardware q number */
u_int32_t *axq_link; /* link ptr in last TX desc */
STAILQ_HEAD(, ath_buf) axq_q; /* transmit queue */
spinlock_t axq_lock; /* lock on q and link */
int axq_depth; /* queue depth */
u_int32_t axq_totalqueued; /* total ever queued */
u_int axq_intrcnt; /* count to determine if descriptor
* should generate int on this txq.
*/
/*
* Staging queue for frames awaiting a fast-frame pairing.
*/
TAILQ_HEAD(axq_headtype, ath_buf) axq_stageq;
/* scratch compression buffer */
char *axq_compbuf; /* scratch comp buffer */
dma_addr_t axq_compbufp; /* scratch comp buffer (phys)*/
u_int axq_compbufsz; /* scratch comp buffer size */
};
/* driver-specific vap state */
struct ath_vap {
struct ieee80211vap av_vap; /* base class */
int (*av_newstate)(struct ieee80211vap *, enum ieee80211_state, int);
/* XXX beacon state */
struct ath_buf *av_bcbuf; /* beacon buffer */
struct ieee80211_beacon_offsets av_boff;/* dynamic update state */
int av_bslot; /* beacon slot index */
struct ath_txq av_mcastq; /* multicast transmit queue */
#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,15)
atomic_t av_beacon_alloc; /* set to 1 when the next beacon needs
to be recomputed */
#else
unsigned int av_beacon_alloc;
#endif
};
#define ATH_VAP(_v) ((struct ath_vap *)(_v))
#define ATH_BEACON_AIFS_DEFAULT 0 /* Default aifs for ap beacon q */
#define ATH_BEACON_CWMIN_DEFAULT 0 /* Default cwmin for ap beacon q */
#define ATH_BEACON_CWMAX_DEFAULT 0 /* Default cwmax for ap beacon q */
#define ATH_TXQ_INTR_PERIOD 5 /* axq_intrcnt period for intr gen */
#define ATH_TXQ_LOCK_INIT(_tq) spin_lock_init(&(_tq)->axq_lock)
#define ATH_TXQ_LOCK_DESTROY(_tq)
#define ATH_TXQ_LOCK_IRQ(_tq) do { \
unsigned long __axq_lockflags; \
ATH_TXQ_LOCK_CHECK(_tq); \
spin_lock_irqsave(&(_tq)->axq_lock, __axq_lockflags);
#define ATH_TXQ_UNLOCK_IRQ(_tq) \
ATH_TXQ_LOCK_ASSERT(_tq); \
spin_unlock_irqrestore(&(_tq)->axq_lock, __axq_lockflags); \
} while (0)
#define ATH_TXQ_UNLOCK_IRQ_EARLY(_tq) \
ATH_TXQ_LOCK_ASSERT(_tq); \
spin_unlock_irqrestore(&(_tq)->axq_lock, __axq_lockflags);
#define ATH_TXQ_LOCK_IRQ_INSIDE(_tq) do { \
ATH_TXQ_LOCK_CHECK(_tq); \
spin_lock(&(_tq)->axq_lock); \
} while(0)
#define ATH_TXQ_UNLOCK_IRQ_INSIDE(_tq) do { \
ATH_TXQ_LOCK_ASSERT(_tq); \
spin_unlock(&(_tq)->axq_lock); \
} while(0)
#if (defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)) && defined(spin_is_locked)
#define ATH_TXQ_LOCK_ASSERT(_tq) \
KASSERT(spin_is_locked(&(_tq)->axq_lock), ("txq not locked!"))
#if (defined(ATH_DEBUG_SPINLOCKS))
#define ATH_TXQ_LOCK_CHECK(_tq) do { \
if (spin_is_locked(&(_tq)->axq_lock)) \
printk(KERN_DEBUG "%s:%d - about to block on txq lock!\n", __func__, __LINE__); \
} while(0)
#else /* #if (defined(ATH_DEBUG_SPINLOCKS)) */
#define ATH_TXQ_LOCK_CHECK(_tq)
#endif /* #if (defined(ATH_DEBUG_SPINLOCKS)) */
#else
#define ATH_TXQ_LOCK_ASSERT(_tq)
#define ATH_TXQ_LOCK_CHECK(_tq)
#endif
#define ATH_TXQ_INSERT_TAIL(_tq, _elm, _field) do { \
STAILQ_INSERT_TAIL(&(_tq)->axq_q, (_elm), _field); \
(_tq)->axq_depth++; \
(_tq)->axq_totalqueued++; \
} while (0)
#define ATH_TXQ_REMOVE_HEAD(_tq, _field) do { \
STAILQ_REMOVE_HEAD(&(_tq)->axq_q, _field); \
(_tq)->axq_depth--; \
} while (0)
/* move buffers from MCASTQ to CABQ */
#define ATH_TXQ_MOVE_MCASTQ(_tqs,_tqd) do { \
(_tqd)->axq_depth += (_tqs)->axq_depth; \
(_tqd)->axq_totalqueued += (_tqs)->axq_totalqueued; \
(_tqd)->axq_link = (_tqs)->axq_link; \
STAILQ_CONCAT(&(_tqd)->axq_q, &(_tqs)->axq_q); \
(_tqs)->axq_depth=0; \
(_tqs)->axq_totalqueued = 0; \
(_tqs)->axq_link = NULL; \
} while (0)
/*
* concat buffers from one queue to other
*/
#define ATH_TXQ_MOVE_Q(_tqs,_tqd) ATH_TXQ_MOVE_MCASTQ(_tqs,_tqd)
#define BSTUCK_THRESH 10 /* # of stuck beacons before resetting NB: this is a guess*/
struct ath_rp {
struct list_head list;
u_int64_t rp_tsf;
u_int8_t rp_rssi;
u_int8_t rp_width;
int rp_index;
int rp_allocated;
int rp_analyzed;
};
struct ath_softc {
struct ieee80211com sc_ic; /* NB: must be first */
struct net_device *sc_dev;
void __iomem *sc_iobase; /* address of the device */
struct semaphore sc_lock; /* dev-level lock */
struct net_device_stats sc_devstats; /* device statistics */
struct ath_stats sc_stats; /* private statistics */
int devid;
int sc_debug;
int sc_default_ieee80211_debug; /* default debug flags for new VAPs */
void (*sc_recv_mgmt)(struct ieee80211vap *, struct ieee80211_node *,
struct sk_buff *, int, int, u_int64_t);
This patch augments the current reference counting code with: * Counters for total outstanding instances for each resource type (skb, ath_node and ath_buf) * One pair of acquisition/release functions per resource type in unlocked and one in locked * Adds some more _debug versions of functions in the call chain that acquire/release resources so that the original func/line in the driver as well as the func/line that affected the resource use can be shown in the trace. Intermediate stack frames aren't necessary to trace the leaks. * Changes naming convention for "lock-required" functions to suffix _locked for the versions that expect locking, to be consistent with some other places in the code. * Consolidate debug messages to the helper functions that actually affect the reference count or acquire/release a resource * Additional sanity checks and leak detection (esp for detecting node ref leaks through skb) * skb references are nulled out by the new sbk unref/free function. I've tested these changes extensively and found lots of cases where we didn't get enough node references when cloning skbuff, and where the kernel drops packets due to performance issues and leaks our node references. With these changes and the tracing enabled I have verified that: * TX BUF: tx buffers always go down to zero when the tx queue is done, and you can watch tx queue usage ratio go up and down again as the driver is working. There are no leaks here at the moment, although there *are* some in the madwifi-dfs branch during CAC at the moment. * skbuff leaks in all the common flows are fixed. We were leaking node references in a lot of places where kernel was dropping skb's due to congestion and we were failing to increment node references when cloning skbuffs. These are now detected, as are skbuffs that are reaped by the kernel while still holding a node reference. * the ath_node count works correctly and on an idle system we get about 5 references per station table node, with 2 node instances per VAP. One for the bss and one for the node in the station table, I believe. The ath_node count goes up and down but always lands back at the stable number based on the vaps you have configured and the number of actual stations in the station table. The point here is that it's pretty constant what you will see over time, despite excessive node creation/release in our code during input (esp input_all). Thank god for the slab allocator. git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2902 0192ed92-7a03-0410-a25b-9323aeb14dbd
2007-11-21 23:14:11 +03:00
#ifdef IEEE80211_DEBUG_REFCNT
void (*sc_node_cleanup_debug)(struct ieee80211_node *, const char* func, int line);
void (*sc_node_free_debug)(struct ieee80211_node *, const char* func, int line);
#else /* #ifdef IEEE80211_DEBUG_REFCNT */
void (*sc_node_cleanup)(struct ieee80211_node *);
void (*sc_node_free)(struct ieee80211_node *);
This patch augments the current reference counting code with: * Counters for total outstanding instances for each resource type (skb, ath_node and ath_buf) * One pair of acquisition/release functions per resource type in unlocked and one in locked * Adds some more _debug versions of functions in the call chain that acquire/release resources so that the original func/line in the driver as well as the func/line that affected the resource use can be shown in the trace. Intermediate stack frames aren't necessary to trace the leaks. * Changes naming convention for "lock-required" functions to suffix _locked for the versions that expect locking, to be consistent with some other places in the code. * Consolidate debug messages to the helper functions that actually affect the reference count or acquire/release a resource * Additional sanity checks and leak detection (esp for detecting node ref leaks through skb) * skb references are nulled out by the new sbk unref/free function. I've tested these changes extensively and found lots of cases where we didn't get enough node references when cloning skbuff, and where the kernel drops packets due to performance issues and leaks our node references. With these changes and the tracing enabled I have verified that: * TX BUF: tx buffers always go down to zero when the tx queue is done, and you can watch tx queue usage ratio go up and down again as the driver is working. There are no leaks here at the moment, although there *are* some in the madwifi-dfs branch during CAC at the moment. * skbuff leaks in all the common flows are fixed. We were leaking node references in a lot of places where kernel was dropping skb's due to congestion and we were failing to increment node references when cloning skbuffs. These are now detected, as are skbuffs that are reaped by the kernel while still holding a node reference. * the ath_node count works correctly and on an idle system we get about 5 references per station table node, with 2 node instances per VAP. One for the bss and one for the node in the station table, I believe. The ath_node count goes up and down but always lands back at the stable number based on the vaps you have configured and the number of actual stations in the station table. The point here is that it's pretty constant what you will see over time, despite excessive node creation/release in our code during input (esp input_all). Thank god for the slab allocator. git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2902 0192ed92-7a03-0410-a25b-9323aeb14dbd
2007-11-21 23:14:11 +03:00
#endif /* #ifdef IEEE80211_DEBUG_REFCNT */
void *sc_bdev; /* associated bus device */
struct ath_hal *sc_ah; /* Atheros HAL */
spinlock_t sc_hal_lock; /* hardware access lock */
struct ath_ratectrl *sc_rc; /* tx rate control support */
struct ath_tx99 *sc_tx99; /* tx99 support */
void (*sc_setdefantenna)(struct ath_softc *, u_int);
unsigned int sc_invalid:1; /* being detached */
unsigned int sc_mrretry:1; /* multi-rate retry support */
unsigned int sc_softled:1; /* enable LED gpio status */
unsigned int sc_splitmic:1; /* split TKIP MIC keys */
unsigned int sc_needmib:1; /* enable MIB stats intr */
unsigned int sc_hasdiversity:1; /* rx diversity available */
unsigned int sc_diversity:1; /* enable rx diversity */
unsigned int sc_olddiversity:1; /* diversity setting before XR enable */
unsigned int sc_hasveol:1; /* tx VEOL support */
unsigned int sc_hastpc:1; /* per-packet TPC support */
unsigned int sc_dturbo:1; /* dynamic turbo capable */
unsigned int sc_dturbo_switch:1; /* turbo switch mode*/
unsigned int sc_dturbo_hold:1; /* dynamic turbo hold state */
unsigned int sc_rate_recn_state:1; /* dynamic turbo state recmded by ratectrl */
unsigned int sc_ignore_ar:1; /* ignore AR during transition */
unsigned int sc_ledstate:1; /* LED on/off state */
unsigned int sc_blinking:1; /* LED blink operation active */
unsigned int sc_beacons:1; /* beacons running */
unsigned int sc_hasbmask:1; /* bssid mask support */
unsigned int sc_mcastkey:1; /* mcast key cache search */
unsigned int sc_hastsfadd:1; /* tsf adjust support */
unsigned int sc_scanning:1; /* scanning active */
unsigned int sc_nostabeacons:1; /* no beacons for station */
unsigned int sc_xrgrppoll:1; /* xr group polls are active */
unsigned int sc_syncbeacon:1; /* sync/resync beacon timers */
unsigned int sc_hasclrkey:1; /* CLR key supported */
unsigned int sc_stagbeacons:1; /* use staggered beacons */
unsigned int sc_dfswait:1; /* waiting on channel for radar detect */
unsigned int sc_ackrate:1; /* send acks at high bitrate */
unsigned int sc_dfs_cac:1; /* waiting on channel for radar detect */
unsigned int sc_hasintmit:1; /* Interference mitigation */
This update uses the recently added HAL extensions for controlling noise immunity settings in order to correctly re-implement disabling of interference mitigation. A module parameter and sysctl parameter are provided for changing whether interference mitigation is enabled or disabled. When interference mitigation is disabled, we work around a HAL defect where the interference mitigation auto-tuning algorithm still starts and/or sets some initially high mitigation levels. With this fix, disabling interference mitigation with the current HAL behaves like it did in prior HALs. Far greater receive sensitivity and increased range is supported with this disabled. This is especially useful for long distance point-to-point links. As a part of this fix, a severe bug that was originally a workaround for the HAL issue has been corrected. When interference mitigation is enabled, we NEVER want to eat or throttle the MIB interrupts as the hardware counter callbacks to the HAL are what drives the interference mitigation calibration state machine. Conversely, if interferference mitigation is being blocked by our driver but the hAL may still be enabling the HAL_INT_MIB in the IMR, then we want to force the interrupt OFF in the mask and eat the interrupt. The failure case was where the interrupt would fire continually and never get properly handled because the HAL wasn't configured to handle interfernece mitigation - now we mask the interrupt OFF. With the 'throttling' hack, we didn't fix hte problem but made it worse - when interfernce mitigation was enabled we just blocked the necessary signals to get the counters updated and stop the interrupt from continuing to fire. The timer to re-enable the MIB interrupt after it fired was also wrong cause it would make sure the interrupt could never be disabled by the HAL or the driver. git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@3505 0192ed92-7a03-0410-a25b-9323aeb14dbd
2008-04-10 07:21:59 +04:00
unsigned int sc_useintmit:1; /* Interference mitigation enabled? */
unsigned int sc_txcont:1; /* Is continuous transmit enabled? */
unsigned int sc_dfs_testmode:1; /* IF this is on, AP vaps will stay in
* 'channel availability check' indefinately,
* reporting radar and interference detections.
*/
unsigned int sc_txcont_power; /* Continuous transmit power in 0.5dBm units */
unsigned int sc_txcont_rate; /* Continuous transmit rate in Mbps */
/* rate tables */
const HAL_RATE_TABLE *sc_rates[IEEE80211_MODE_MAX];
const HAL_RATE_TABLE *sc_currates; /* current rate table */
const HAL_RATE_TABLE *sc_xr_rates; /* XR rate table */
const HAL_RATE_TABLE *sc_half_rates; /* half rate table */
const HAL_RATE_TABLE *sc_quarter_rates; /* quarter rate table */
HAL_OPMODE sc_opmode; /* current hal operating mode */
enum ieee80211_phymode sc_curmode; /* current phy mode */
u_int16_t sc_curtxpow; /* current tx power limit */
u_int16_t sc_curaid; /* current association id */
HAL_CHANNEL sc_curchan; /* current h/w channel */
u_int8_t sc_curbssid[IEEE80211_ADDR_LEN];
u_int8_t sc_rixmap[256]; /* IEEE to h/w rate table ix */
struct {
u_int8_t ieeerate; /* IEEE rate */
u_int8_t flags; /* radiotap flags */
u_int16_t ledon; /* softled on time */
u_int16_t ledoff; /* softled off time */
} sc_hwmap[32]; /* h/w rate ix mappings */
u_int8_t sc_minrateix; /* min h/w rate index */
u_int8_t sc_protrix; /* protection rate index */
u_int8_t sc_mcastantenna; /* Multicast antenna number */
u_int8_t sc_txantenna; /* data tx antenna (fixed or auto) */
u_int16_t sc_nvaps; /* # of active virtual APs */
u_int8_t sc_nstavaps; /* # of active station VAPs */
u_int8_t sc_nmonvaps; /* # of monitor VAPs */
u_int8_t sc_nbcnvaps; /* # of vaps sending beacons */
u_int sc_fftxqmin; /* aggregation threshold */
HAL_INT sc_imask; /* interrupt mask copy */
u_int sc_keymax; /* size of key cache */
u_int8_t sc_keymap[ATH_KEYBYTES]; /* key use bit map */
struct ieee80211_node *sc_keyixmap[ATH_KEYMAX];/* key ix->node map */
u_int8_t sc_bssidmask[IEEE80211_ADDR_LEN];
u_int sc_ledpin; /* GPIO pin for driving LED */
u_int sc_ledon; /* pin setting for LED on */
u_int sc_ledidle; /* idle polling interval */
long unsigned int sc_ledevent; /* time of last LED event */
u_int8_t sc_rxrate; /* current rx rate for LED */
u_int8_t sc_txrate; /* current tx rate for LED */
u_int16_t sc_ledoff; /* off time for current blink */
struct timer_list sc_ledtimer; /* led off timer */
struct ATH_TQ_STRUCT sc_fataltq; /* fatal error intr tasklet */
int sc_rxbufsize; /* rx size based on mtu */
struct ath_descdma sc_rxdma; /* RX descriptors */
ath_bufhead sc_rxbuf; /* receive buffer */
struct ath_buf *sc_rxbufcur; /* current rx buffer */
u_int32_t *sc_rxlink; /* link ptr in last RX desc */
spinlock_t sc_rxbuflock;
struct ATH_TQ_STRUCT sc_rxtq; /* rx intr tasklet */
struct ATH_TQ_STRUCT sc_rxorntq; /* rxorn intr tasklet */
u_int8_t sc_defant; /* current default antenna */
u_int8_t sc_rxotherant; /* RXs on non-default antenna */
u_int16_t sc_cachelsz; /* cache line size */
struct ath_descdma sc_txdma; /* TX descriptors */
ath_bufhead sc_txbuf; /* TX buffers pool */
atomic_t sc_txbuf_counter; /* number of available TX
* buffers */
spinlock_t sc_txbuflock; /* txbuf lock */
u_int sc_txqsetup; /* h/w queues setup */
u_int sc_txintrperiod; /* tx interrupt batching */
struct ath_txq sc_txq[HAL_NUM_TX_QUEUES];
struct ath_txq *sc_ac2q[WME_NUM_AC]; /* WME AC -> h/w qnum */
struct ATH_TQ_STRUCT sc_txtq; /* tx intr tasklet */
u_int8_t sc_grppoll_str[GRPPOLL_RATE_STR_LEN];
struct ath_descdma sc_bdma; /* beacon descriptors */
ath_bufhead sc_bbuf; /* beacon buffers */
spinlock_t sc_bbuflock; /* beacon buffers lock */
u_int sc_bhalq; /* HAL q for outgoing beacons */
u_int sc_bmisscount; /* missed beacon transmits */
u_int32_t sc_ant_tx[8]; /* recent tx frames/antenna */
struct ath_txq *sc_cabq; /* tx q for cab frames */
struct ath_txq sc_grpplq; /* tx q for XR group polls */
struct ath_txq *sc_xrtxq; /* tx q for XR data */
struct ath_descdma sc_grppolldma; /* TX descriptors for grppoll */
ath_bufhead sc_grppollbuf; /* transmit buffers for grouppoll */
spinlock_t sc_grppollbuflock; /* grouppoll lock */
u_int16_t sc_xrpollint; /* xr poll interval */
u_int16_t sc_xrpollcount; /* xr poll count */
struct ath_txq *sc_uapsdq; /* tx q for uapsd */
struct ATH_TQ_STRUCT sc_bmisstq; /* bmiss intr tasklet */
struct ATH_TQ_STRUCT sc_bstucktq; /* beacon stuck intr tasklet */
enum {
OK, /* no change needed */
UPDATE, /* update pending */
COMMIT /* beacon sent, commit change */
} sc_updateslot; /* slot time update fsm */
int sc_slotupdate; /* slot to next advance fsm */
struct ieee80211vap **sc_bslot; /* beacon xmit slots */
int sc_bnext; /* next slot for beacon xmit */
int sc_beacon_cal; /* use beacon timer for calibration */
long unsigned int sc_calinterval_sec; /* current interval for calibration (in seconds) */
long unsigned int sc_lastcal; /* last time the calibration was performed */
struct timer_list sc_cal_ch; /* calibration timer */
HAL_NODE_STATS sc_halstats; /* station-mode rssi stats */
struct ctl_table_header *sc_sysctl_header;
struct ctl_table *sc_sysctls;
struct timer_list sc_mib_enable;
#ifdef ATH_REVERSE_ENGINEERING
u_int8_t register_snapshot[MAX_REGISTER_ADDRESS];
#endif /* #ifdef ATH_REVERSE_ENGINEERING */
#ifdef ATH_SUPERG_DYNTURBO
struct timer_list sc_dturbo_switch_mode;/* AP scan timer */
u_int32_t sc_dturbo_tcount; /* beacon intval count */
u_int32_t sc_dturbo_hold_max; /* hold count before switching to base*/
u_int16_t sc_dturbo_hold_count; /* hold count before switching to base*/
u_int16_t sc_dturbo_turbo_tmin; /* min turbo count */
u_int32_t sc_dturbo_bytes; /* bandwidth stats */
u_int32_t sc_dturbo_base_tmin; /* min time in base */
u_int32_t sc_dturbo_turbo_tmax; /* max time in turbo */
u_int32_t sc_dturbo_bw_base; /* bandwidth threshold */
u_int32_t sc_dturbo_bw_turbo; /* bandwidth threshold */
#endif
u_int sc_slottimeconf; /* manual override for slottime */
u_int sc_acktimeoutconf; /* manual override for ack timeout */
u_int sc_ctstimeoutconf; /* manual override for cts timeout */
struct timer_list sc_dfs_excl_timer; /* mark expiration timer task */
struct timer_list sc_dfs_cac_timer; /* dfs wait timer */
u_int32_t sc_dfs_cac_period; /* DFS wait time before accessing a
* channel (in seconds). FCC
* requires 60s. */
u_int32_t sc_dfs_excl_period; /* DFS channel non-occupancy limit
* after radar is detected (in seconds).
* FCC requires 30m. */
u_int64_t sc_rp_lasttsf; /* TSF at last detected radar pulse */
struct ath_rp *sc_rp; /* radar pulse circular array */
struct list_head sc_rp_list;
int sc_rp_num;
int sc_rp_min;
HAL_BOOL (*sc_rp_analyse)(struct ath_softc *sc);
struct ATH_TQ_STRUCT sc_rp_tq;
int sc_rp_ignored; /* if set, we ignored all
* received pulses */
int sc_radar_ignored; /* if set, we ignored all
* detected radars */
u_int32_t sc_nexttbtt;
u_int64_t sc_last_tsf;
};
typedef void (*ath_callback) (struct ath_softc *);
#define ATH_TXQ_SETUP(sc, i) ((sc)->sc_txqsetup & (1 << i))
#define ATH_TXBUF_LOCK_INIT(_sc) spin_lock_init(&(_sc)->sc_txbuflock)
#define ATH_TXBUF_LOCK_DESTROY(_sc)
#define ATH_TXBUF_LOCK_IRQ(_sc) do { \
unsigned long __txbuflockflags; \
ATH_TXBUF_LOCK_CHECK(_sc); \
spin_lock_irqsave(&(_sc)->sc_txbuflock, __txbuflockflags);
#define ATH_TXBUF_UNLOCK_IRQ(_sc) \
ATH_TXBUF_LOCK_ASSERT(_sc); \
spin_unlock_irqrestore(&(_sc)->sc_txbuflock, __txbuflockflags); \
} while (0)
#define ATH_TXBUF_UNLOCK_IRQ_EARLY(_sc) \
ATH_TXBUF_LOCK_ASSERT(_sc); \
spin_unlock_irqrestore(&(_sc)->sc_txbuflock, __txbuflockflags);
#if (defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)) && defined(spin_is_locked)
#define ATH_TXBUF_LOCK_ASSERT(_sc) \
KASSERT(spin_is_locked(&(_sc)->sc_txbuflock), ("txbuf not locked!"))
#if (defined(ATH_DEBUG_SPINLOCKS))
#define ATH_TXBUF_LOCK_CHECK(_sc) do { \
if (spin_is_locked(&(_sc)->sc_txbuflock)) \
printk(KERN_DEBUG "%s:%d - about to block on txbuf lock!\n", __func__, __LINE__); \
} while(0)
#else /* #if (defined(ATH_DEBUG_SPINLOCKS)) */
#define ATH_TXBUF_LOCK_CHECK(_sc)
#endif /* #if (defined(ATH_DEBUG_SPINLOCKS)) */
#else
#define ATH_TXBUF_LOCK_ASSERT(_sc)
#define ATH_TXBUF_LOCK_CHECK(_sc)
#endif
#define ATH_RXBUF_LOCK_INIT(_sc) spin_lock_init(&(_sc)->sc_rxbuflock)
#define ATH_RXBUF_LOCK_DESTROY(_sc)
#define ATH_RXBUF_LOCK_IRQ(_sc) do { \
unsigned long __rxbuflockflags; \
ATH_RXBUF_LOCK_CHECK(_sc); \
spin_lock_irqsave(&(_sc)->sc_rxbuflock, __rxbuflockflags);
#define ATH_RXBUF_UNLOCK_IRQ(_sc) \
ATH_RXBUF_LOCK_ASSERT(_sc); \
spin_unlock_irqrestore(&(_sc)->sc_rxbuflock, __rxbuflockflags); \
} while (0)
#define ATH_RXBUF_UNLOCK_IRQ_EARLY(_sc) \
ATH_RXBUF_LOCK_ASSERT(_sc); \
spin_unlock_irqrestore(&(_sc)->sc_rxbuflock, __rxbuflockflags);
#if (defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)) && defined(spin_is_locked)
#define ATH_RXBUF_LOCK_ASSERT(_sc) \
KASSERT(spin_is_locked(&(_sc)->sc_rxbuflock), ("rxbuf not locked!"))
#if (defined(ATH_DEBUG_SPINLOCKS))
#define ATH_RXBUF_LOCK_CHECK(_sc) do { \
if (spin_is_locked(&(_sc)->sc_rxbuflock)) \
printk(KERN_DEBUG "%s:%d - about to block on rxbuf lock!\n", __func__, __LINE__); \
} while(0)
#else /* #if (defined(ATH_DEBUG_SPINLOCKS)) */
#define ATH_RXBUF_LOCK_CHECK(_sc)
#endif /* #if (defined(ATH_DEBUG_SPINLOCKS)) */
#else
#define ATH_RXBUF_LOCK_ASSERT(_sc)
#define ATH_RXBUF_LOCK_CHECK(_sc)
#endif
#define ATH_BBUF_LOCK_INIT(_sc) spin_lock_init(&(_sc)->sc_bbuflock)
#define ATH_BBUF_LOCK_DESTROY(_sc)
#define ATH_BBUF_LOCK_IRQ(_sc) do { \
unsigned long __bbuflockflags; \
ATH_BBUF_LOCK_CHECK(_sc); \
spin_lock_irqsave(&(_sc)->sc_bbuflock, __bbuflockflags);
#define ATH_BBUF_UNLOCK_IRQ(_sc) \
ATH_BBUF_LOCK_ASSERT(_sc); \
spin_unlock_irqrestore(&(_sc)->sc_bbuflock, __bbuflockflags); \
} while (0)
#define ATH_BBUF_UNLOCK_IRQ_EARLY(_sc) \
ATH_BBUF_LOCK_ASSERT(_sc); \
spin_unlock_irqrestore(&(_sc)->sc_bbuflock, __bbuflockflags);
#if (defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)) && defined(spin_is_locked)
#define ATH_BBUF_LOCK_ASSERT(_sc) \
KASSERT(spin_is_locked(&(_sc)->sc_bbuflock), ("bbuf not locked!"))
#if (defined(ATH_DEBUG_SPINLOCKS))
#define ATH_BBUF_LOCK_CHECK(_sc) do { \
if (spin_is_locked(&(_sc)->sc_bbuflock)) \
printk(KERN_DEBUG "%s:%d - about to block on bbuf lock!\n", __func__, __LINE__); \
} while(0)
#else /* #if (defined(ATH_DEBUG_SPINLOCKS)) */
#define ATH_BBUF_LOCK_CHECK(_sc)
#endif /* #if (defined(ATH_DEBUG_SPINLOCKS)) */
#else
#define ATH_BBUF_LOCK_ASSERT(_sc)
#define ATH_BBUF_LOCK_CHECK(_sc)
#endif
#define ATH_GBUF_LOCK_INIT(_sc) spin_lock_init(&(_sc)->sc_grppollbuflock)
#define ATH_GBUF_LOCK_DESTROY(_sc)
#define ATH_GBUF_LOCK_IRQ(_sc) do { \
unsigned long __grppollbuflockflags; \
ATH_GBUF_LOCK_CHECK(_sc); \
spin_lock_irqsave(&(_sc)->sc_grppollbuflock, __grppollbuflockflags);
#define ATH_GBUF_UNLOCK_IRQ(_sc) \
ATH_GBUF_LOCK_ASSERT(_sc); \
spin_unlock_irqrestore(&(_sc)->sc_grppollbuflock, __grppollbuflockflags); \
} while (0)
#define ATH_GBUF_UNLOCK_IRQ_EARLY(_sc) \
ATH_GBUF_LOCK_ASSERT(_sc); \
spin_unlock_irqrestore(&(_sc)->sc_grppollbuflock, __grppollbuflockflags);
#if (defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)) && defined(spin_is_locked)
#define ATH_GBUF_LOCK_ASSERT(_sc) \
KASSERT(spin_is_locked(&(_sc)->sc_grppollbuflock), ("grppollbuf not locked!"))
#if (defined(ATH_DEBUG_SPINLOCKS))
#define ATH_GBUF_LOCK_CHECK(_sc) do { \
if (spin_is_locked(&(_sc)->sc_grppollbuflock)) \
printk(KERN_DEBUG "%s:%d - about to block on grppollbuf lock!\n", __func__, __LINE__); \
} while(0)
#else /* #if (defined(ATH_DEBUG_SPINLOCKS)) */
#define ATH_GBUF_LOCK_CHECK(_sc)
#endif /* #if (defined(ATH_DEBUG_SPINLOCKS)) */
#else
#define ATH_GBUF_LOCK_ASSERT(_sc)
#define ATH_GBUF_LOCK_CHECK(_sc)
#endif
/* Protects the device from concurrent accesses */
#define ATH_LOCK_INIT(_sc) init_MUTEX(&(_sc)->sc_lock)
#define ATH_LOCK_DESTROY(_sc)
#define ATH_LOCK(_sc) down(&(_sc)->sc_lock)
#define ATH_UNLOCK(_sc) up(&(_sc)->sc_lock)
int ath_attach(u_int16_t, struct net_device *, HAL_BUS_TAG);
int ath_detach(struct net_device *);
void ath_resume(struct net_device *);
void ath_suspend(struct net_device *);
#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,19)
irqreturn_t ath_intr(int, void *);
#else
irqreturn_t ath_intr(int, void *, struct pt_regs *);
#endif
int ath_ioctl_ethtool(struct ath_softc *, int, void __user *);
void bus_read_cachesize(struct ath_softc *, u_int8_t *);
void ath_sysctl_register(void);
void ath_sysctl_unregister(void);
int ar_device(int devid);
#define DEV_NAME(_d) \
((NULL == _d || NULL == _d->name || 0 == strncmp(_d->name, "wifi%d", 6)) ? \
"MadWifi" : \
_d->name)
#define VAP_DEV_NAME(_v) \
((NULL == _v) ? \
"MadWifi" : \
DEV_NAME(_v->iv_dev))
#define SC_DEV_NAME(_sc) \
((NULL == _sc) ? \
"MadWifi" : \
DEV_NAME(_sc->sc_dev))
#define VAP_IC_DEV_NAME(_v) \
((NULL == _v || NULL == _v->iv_ic) ? \
"MadWifi" : \
DEV_NAME(_v->iv_ic->ic_dev))
void ath_radar_detected(struct ath_softc *sc, const char* message);
#endif /* _DEV_ATH_ATHVAR_H */