kernel-fxtec-pro1x/drivers/net/wireless/iwlwifi/iwl-agn-tx.c

1391 lines
41 KiB
C
Raw Normal View History

/******************************************************************************
*
* GPL LICENSE SUMMARY
*
* Copyright(c) 2008 - 2010 Intel Corporation. All rights reserved.
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of version 2 of the GNU General Public License as
* published by the Free Software Foundation.
*
* This program is distributed in the hope that it will be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110,
* USA
*
* The full GNU General Public License is included in this distribution
* in the file called LICENSE.GPL.
*
* Contact Information:
* Intel Linux Wireless <ilw@linux.intel.com>
* Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
*
*****************************************************************************/
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/init.h>
#include <linux/sched.h>
#include "iwl-dev.h"
#include "iwl-core.h"
#include "iwl-sta.h"
#include "iwl-io.h"
#include "iwl-helpers.h"
#include "iwl-agn-hw.h"
#include "iwl-agn.h"
/*
* mac80211 queues, ACs, hardware queues, FIFOs.
*
* Cf. http://wireless.kernel.org/en/developers/Documentation/mac80211/queues
*
* Mac80211 uses the following numbers, which we get as from it
* by way of skb_get_queue_mapping(skb):
*
* VO 0
* VI 1
* BE 2
* BK 3
*
*
* Regular (not A-MPDU) frames are put into hardware queues corresponding
* to the FIFOs, see comments in iwl-prph.h. Aggregated frames get their
* own queue per aggregation session (RA/TID combination), such queues are
* set up to map into FIFOs too, for which we need an AC->FIFO mapping. In
* order to map frames to the right queue, we also need an AC->hw queue
* mapping. This is implemented here.
*
* Due to the way hw queues are set up (by the hw specific modules like
* iwl-4965.c, iwl-5000.c etc.), the AC->hw queue mapping is the identity
* mapping.
*/
static const u8 tid_to_ac[] = {
/* this matches the mac80211 numbers */
2, 3, 3, 2, 1, 1, 0, 0
};
static inline int get_ac_from_tid(u16 tid)
{
if (likely(tid < ARRAY_SIZE(tid_to_ac)))
return tid_to_ac[tid];
/* no support for TIDs 8-15 yet */
return -EINVAL;
}
static inline int get_fifo_from_tid(struct iwl_rxon_context *ctx, u16 tid)
{
if (likely(tid < ARRAY_SIZE(tid_to_ac)))
return ctx->ac_to_fifo[tid_to_ac[tid]];
/* no support for TIDs 8-15 yet */
return -EINVAL;
}
/**
* iwlagn_txq_update_byte_cnt_tbl - Set up entry in Tx byte-count array
*/
void iwlagn_txq_update_byte_cnt_tbl(struct iwl_priv *priv,
struct iwl_tx_queue *txq,
u16 byte_cnt)
{
struct iwlagn_scd_bc_tbl *scd_bc_tbl = priv->scd_bc_tbls.addr;
int write_ptr = txq->q.write_ptr;
int txq_id = txq->q.id;
u8 sec_ctl = 0;
u8 sta_id = 0;
u16 len = byte_cnt + IWL_TX_CRC_SIZE + IWL_TX_DELIMITER_SIZE;
__le16 bc_ent;
WARN_ON(len > 0xFFF || write_ptr >= TFD_QUEUE_SIZE_MAX);
if (txq_id != priv->cmd_queue) {
sta_id = txq->cmd[txq->q.write_ptr]->cmd.tx.sta_id;
sec_ctl = txq->cmd[txq->q.write_ptr]->cmd.tx.sec_ctl;
switch (sec_ctl & TX_CMD_SEC_MSK) {
case TX_CMD_SEC_CCM:
len += CCMP_MIC_LEN;
break;
case TX_CMD_SEC_TKIP:
len += TKIP_ICV_LEN;
break;
case TX_CMD_SEC_WEP:
len += WEP_IV_LEN + WEP_ICV_LEN;
break;
}
}
bc_ent = cpu_to_le16((len & 0xFFF) | (sta_id << 12));
scd_bc_tbl[txq_id].tfd_offset[write_ptr] = bc_ent;
if (write_ptr < TFD_QUEUE_SIZE_BC_DUP)
scd_bc_tbl[txq_id].
tfd_offset[TFD_QUEUE_SIZE_MAX + write_ptr] = bc_ent;
}
void iwlagn_txq_inval_byte_cnt_tbl(struct iwl_priv *priv,
struct iwl_tx_queue *txq)
{
struct iwlagn_scd_bc_tbl *scd_bc_tbl = priv->scd_bc_tbls.addr;
int txq_id = txq->q.id;
int read_ptr = txq->q.read_ptr;
u8 sta_id = 0;
__le16 bc_ent;
WARN_ON(read_ptr >= TFD_QUEUE_SIZE_MAX);
if (txq_id != priv->cmd_queue)
sta_id = txq->cmd[read_ptr]->cmd.tx.sta_id;
bc_ent = cpu_to_le16(1 | (sta_id << 12));
scd_bc_tbl[txq_id].tfd_offset[read_ptr] = bc_ent;
if (read_ptr < TFD_QUEUE_SIZE_BC_DUP)
scd_bc_tbl[txq_id].
tfd_offset[TFD_QUEUE_SIZE_MAX + read_ptr] = bc_ent;
}
static int iwlagn_tx_queue_set_q2ratid(struct iwl_priv *priv, u16 ra_tid,
u16 txq_id)
{
u32 tbl_dw_addr;
u32 tbl_dw;
u16 scd_q2ratid;
scd_q2ratid = ra_tid & IWL_SCD_QUEUE_RA_TID_MAP_RATID_MSK;
tbl_dw_addr = priv->scd_base_addr +
IWLAGN_SCD_TRANSLATE_TBL_OFFSET_QUEUE(txq_id);
tbl_dw = iwl_read_targ_mem(priv, tbl_dw_addr);
if (txq_id & 0x1)
tbl_dw = (scd_q2ratid << 16) | (tbl_dw & 0x0000FFFF);
else
tbl_dw = scd_q2ratid | (tbl_dw & 0xFFFF0000);
iwl_write_targ_mem(priv, tbl_dw_addr, tbl_dw);
return 0;
}
static void iwlagn_tx_queue_stop_scheduler(struct iwl_priv *priv, u16 txq_id)
{
/* Simply stop the queue, but don't change any configuration;
* the SCD_ACT_EN bit is the write-enable mask for the ACTIVE bit. */
iwl_write_prph(priv,
IWLAGN_SCD_QUEUE_STATUS_BITS(txq_id),
(0 << IWLAGN_SCD_QUEUE_STTS_REG_POS_ACTIVE)|
(1 << IWLAGN_SCD_QUEUE_STTS_REG_POS_SCD_ACT_EN));
}
void iwlagn_set_wr_ptrs(struct iwl_priv *priv,
int txq_id, u32 index)
{
iwl_write_direct32(priv, HBUS_TARG_WRPTR,
(index & 0xff) | (txq_id << 8));
iwl_write_prph(priv, IWLAGN_SCD_QUEUE_RDPTR(txq_id), index);
}
void iwlagn_tx_queue_set_status(struct iwl_priv *priv,
struct iwl_tx_queue *txq,
int tx_fifo_id, int scd_retry)
{
int txq_id = txq->q.id;
int active = test_bit(txq_id, &priv->txq_ctx_active_msk) ? 1 : 0;
iwl_write_prph(priv, IWLAGN_SCD_QUEUE_STATUS_BITS(txq_id),
(active << IWLAGN_SCD_QUEUE_STTS_REG_POS_ACTIVE) |
(tx_fifo_id << IWLAGN_SCD_QUEUE_STTS_REG_POS_TXF) |
(1 << IWLAGN_SCD_QUEUE_STTS_REG_POS_WSL) |
IWLAGN_SCD_QUEUE_STTS_REG_MSK);
txq->sched_retry = scd_retry;
IWL_DEBUG_INFO(priv, "%s %s Queue %d on FIFO %d\n",
active ? "Activate" : "Deactivate",
scd_retry ? "BA" : "AC/CMD", txq_id, tx_fifo_id);
}
int iwlagn_txq_agg_enable(struct iwl_priv *priv, int txq_id,
int tx_fifo, int sta_id, int tid, u16 ssn_idx)
{
unsigned long flags;
u16 ra_tid;
int ret;
if ((IWLAGN_FIRST_AMPDU_QUEUE > txq_id) ||
(IWLAGN_FIRST_AMPDU_QUEUE + priv->cfg->num_of_ampdu_queues
<= txq_id)) {
IWL_WARN(priv,
"queue number out of range: %d, must be %d to %d\n",
txq_id, IWLAGN_FIRST_AMPDU_QUEUE,
IWLAGN_FIRST_AMPDU_QUEUE +
priv->cfg->num_of_ampdu_queues - 1);
return -EINVAL;
}
ra_tid = BUILD_RAxTID(sta_id, tid);
/* Modify device's station table to Tx this TID */
ret = iwl_sta_tx_modify_enable_tid(priv, sta_id, tid);
if (ret)
return ret;
spin_lock_irqsave(&priv->lock, flags);
/* Stop this Tx queue before configuring it */
iwlagn_tx_queue_stop_scheduler(priv, txq_id);
/* Map receiver-address / traffic-ID to this queue */
iwlagn_tx_queue_set_q2ratid(priv, ra_tid, txq_id);
/* Set this queue as a chain-building queue */
iwl_set_bits_prph(priv, IWLAGN_SCD_QUEUECHAIN_SEL, (1<<txq_id));
/* enable aggregations for the queue */
iwl_set_bits_prph(priv, IWLAGN_SCD_AGGR_SEL, (1<<txq_id));
/* Place first TFD at index corresponding to start sequence number.
* Assumes that ssn_idx is valid (!= 0xFFF) */
priv->txq[txq_id].q.read_ptr = (ssn_idx & 0xff);
priv->txq[txq_id].q.write_ptr = (ssn_idx & 0xff);
iwlagn_set_wr_ptrs(priv, txq_id, ssn_idx);
/* Set up Tx window size and frame limit for this queue */
iwl_write_targ_mem(priv, priv->scd_base_addr +
IWLAGN_SCD_CONTEXT_QUEUE_OFFSET(txq_id) +
sizeof(u32),
((SCD_WIN_SIZE <<
IWLAGN_SCD_QUEUE_CTX_REG2_WIN_SIZE_POS) &
IWLAGN_SCD_QUEUE_CTX_REG2_WIN_SIZE_MSK) |
((SCD_FRAME_LIMIT <<
IWLAGN_SCD_QUEUE_CTX_REG2_FRAME_LIMIT_POS) &
IWLAGN_SCD_QUEUE_CTX_REG2_FRAME_LIMIT_MSK));
iwl_set_bits_prph(priv, IWLAGN_SCD_INTERRUPT_MASK, (1 << txq_id));
/* Set up Status area in SRAM, map to Tx DMA/FIFO, activate the queue */
iwlagn_tx_queue_set_status(priv, &priv->txq[txq_id], tx_fifo, 1);
spin_unlock_irqrestore(&priv->lock, flags);
return 0;
}
int iwlagn_txq_agg_disable(struct iwl_priv *priv, u16 txq_id,
u16 ssn_idx, u8 tx_fifo)
{
if ((IWLAGN_FIRST_AMPDU_QUEUE > txq_id) ||
(IWLAGN_FIRST_AMPDU_QUEUE + priv->cfg->num_of_ampdu_queues
<= txq_id)) {
IWL_ERR(priv,
"queue number out of range: %d, must be %d to %d\n",
txq_id, IWLAGN_FIRST_AMPDU_QUEUE,
IWLAGN_FIRST_AMPDU_QUEUE +
priv->cfg->num_of_ampdu_queues - 1);
return -EINVAL;
}
iwlagn_tx_queue_stop_scheduler(priv, txq_id);
iwl_clear_bits_prph(priv, IWLAGN_SCD_AGGR_SEL, (1 << txq_id));
priv->txq[txq_id].q.read_ptr = (ssn_idx & 0xff);
priv->txq[txq_id].q.write_ptr = (ssn_idx & 0xff);
/* supposes that ssn_idx is valid (!= 0xFFF) */
iwlagn_set_wr_ptrs(priv, txq_id, ssn_idx);
iwl_clear_bits_prph(priv, IWLAGN_SCD_INTERRUPT_MASK, (1 << txq_id));
iwl_txq_ctx_deactivate(priv, txq_id);
iwlagn_tx_queue_set_status(priv, &priv->txq[txq_id], tx_fifo, 0);
return 0;
}
/*
* Activate/Deactivate Tx DMA/FIFO channels according tx fifos mask
* must be called under priv->lock and mac access
*/
void iwlagn_txq_set_sched(struct iwl_priv *priv, u32 mask)
{
iwl_write_prph(priv, IWLAGN_SCD_TXFACT, mask);
}
/*
* handle build REPLY_TX command notification.
*/
static void iwlagn_tx_cmd_build_basic(struct iwl_priv *priv,
struct sk_buff *skb,
struct iwl_tx_cmd *tx_cmd,
struct ieee80211_tx_info *info,
struct ieee80211_hdr *hdr,
u8 std_id)
{
__le16 fc = hdr->frame_control;
__le32 tx_flags = tx_cmd->tx_flags;
tx_cmd->stop_time.life_time = TX_CMD_LIFE_TIME_INFINITE;
if (!(info->flags & IEEE80211_TX_CTL_NO_ACK)) {
tx_flags |= TX_CMD_FLG_ACK_MSK;
if (ieee80211_is_mgmt(fc))
tx_flags |= TX_CMD_FLG_SEQ_CTL_MSK;
if (ieee80211_is_probe_resp(fc) &&
!(le16_to_cpu(hdr->seq_ctrl) & 0xf))
tx_flags |= TX_CMD_FLG_TSF_MSK;
} else {
tx_flags &= (~TX_CMD_FLG_ACK_MSK);
tx_flags |= TX_CMD_FLG_SEQ_CTL_MSK;
}
if (ieee80211_is_back_req(fc))
tx_flags |= TX_CMD_FLG_ACK_MSK | TX_CMD_FLG_IMM_BA_RSP_MASK;
else if (info->band == IEEE80211_BAND_2GHZ &&
priv->cfg->advanced_bt_coexist &&
(ieee80211_is_auth(fc) || ieee80211_is_assoc_req(fc) ||
ieee80211_is_reassoc_req(fc) ||
skb->protocol == cpu_to_be16(ETH_P_PAE)))
tx_flags |= TX_CMD_FLG_IGNORE_BT;
tx_cmd->sta_id = std_id;
if (ieee80211_has_morefrags(fc))
tx_flags |= TX_CMD_FLG_MORE_FRAG_MSK;
if (ieee80211_is_data_qos(fc)) {
u8 *qc = ieee80211_get_qos_ctl(hdr);
tx_cmd->tid_tspec = qc[0] & 0xf;
tx_flags &= ~TX_CMD_FLG_SEQ_CTL_MSK;
} else {
tx_flags |= TX_CMD_FLG_SEQ_CTL_MSK;
}
priv->cfg->ops->utils->tx_cmd_protection(priv, info, fc, &tx_flags);
tx_flags &= ~(TX_CMD_FLG_ANT_SEL_MSK);
if (ieee80211_is_mgmt(fc)) {
if (ieee80211_is_assoc_req(fc) || ieee80211_is_reassoc_req(fc))
tx_cmd->timeout.pm_frame_timeout = cpu_to_le16(3);
else
tx_cmd->timeout.pm_frame_timeout = cpu_to_le16(2);
} else {
tx_cmd->timeout.pm_frame_timeout = 0;
}
tx_cmd->driver_txop = 0;
tx_cmd->tx_flags = tx_flags;
tx_cmd->next_frame_len = 0;
}
#define RTS_DFAULT_RETRY_LIMIT 60
static void iwlagn_tx_cmd_build_rate(struct iwl_priv *priv,
struct iwl_tx_cmd *tx_cmd,
struct ieee80211_tx_info *info,
__le16 fc)
{
u32 rate_flags;
int rate_idx;
u8 rts_retry_limit;
u8 data_retry_limit;
u8 rate_plcp;
/* Set retry limit on DATA packets and Probe Responses*/
if (ieee80211_is_probe_resp(fc))
data_retry_limit = 3;
else
data_retry_limit = IWLAGN_DEFAULT_TX_RETRY;
tx_cmd->data_retry_limit = data_retry_limit;
/* Set retry limit on RTS packets */
rts_retry_limit = RTS_DFAULT_RETRY_LIMIT;
if (data_retry_limit < rts_retry_limit)
rts_retry_limit = data_retry_limit;
tx_cmd->rts_retry_limit = rts_retry_limit;
/* DATA packets will use the uCode station table for rate/antenna
* selection */
if (ieee80211_is_data(fc)) {
tx_cmd->initial_rate_index = 0;
tx_cmd->tx_flags |= TX_CMD_FLG_STA_RATE_MSK;
return;
}
/**
* If the current TX rate stored in mac80211 has the MCS bit set, it's
* not really a TX rate. Thus, we use the lowest supported rate for
* this band. Also use the lowest supported rate if the stored rate
* index is invalid.
*/
rate_idx = info->control.rates[0].idx;
if (info->control.rates[0].flags & IEEE80211_TX_RC_MCS ||
(rate_idx < 0) || (rate_idx > IWL_RATE_COUNT_LEGACY))
rate_idx = rate_lowest_index(&priv->bands[info->band],
info->control.sta);
/* For 5 GHZ band, remap mac80211 rate indices into driver indices */
if (info->band == IEEE80211_BAND_5GHZ)
rate_idx += IWL_FIRST_OFDM_RATE;
/* Get PLCP rate for tx_cmd->rate_n_flags */
rate_plcp = iwl_rates[rate_idx].plcp;
/* Zero out flags for this packet */
rate_flags = 0;
/* Set CCK flag as needed */
if ((rate_idx >= IWL_FIRST_CCK_RATE) && (rate_idx <= IWL_LAST_CCK_RATE))
rate_flags |= RATE_MCS_CCK_MSK;
/* Set up antennas */
if (priv->cfg->advanced_bt_coexist && priv->bt_full_concurrent) {
/* operated as 1x1 in full concurrency mode */
priv->mgmt_tx_ant = iwl_toggle_tx_ant(priv, priv->mgmt_tx_ant,
first_antenna(priv->hw_params.valid_tx_ant));
} else
priv->mgmt_tx_ant = iwl_toggle_tx_ant(priv, priv->mgmt_tx_ant,
priv->hw_params.valid_tx_ant);
rate_flags |= iwl_ant_idx_to_flags(priv->mgmt_tx_ant);
/* Set the rate in the TX cmd */
tx_cmd->rate_n_flags = iwl_hw_set_rate_n_flags(rate_plcp, rate_flags);
}
static void iwlagn_tx_cmd_build_hwcrypto(struct iwl_priv *priv,
struct ieee80211_tx_info *info,
struct iwl_tx_cmd *tx_cmd,
struct sk_buff *skb_frag,
int sta_id)
{
struct ieee80211_key_conf *keyconf = info->control.hw_key;
switch (keyconf->cipher) {
case WLAN_CIPHER_SUITE_CCMP:
tx_cmd->sec_ctl = TX_CMD_SEC_CCM;
memcpy(tx_cmd->key, keyconf->key, keyconf->keylen);
if (info->flags & IEEE80211_TX_CTL_AMPDU)
tx_cmd->tx_flags |= TX_CMD_FLG_AGG_CCMP_MSK;
IWL_DEBUG_TX(priv, "tx_cmd with AES hwcrypto\n");
break;
case WLAN_CIPHER_SUITE_TKIP:
tx_cmd->sec_ctl = TX_CMD_SEC_TKIP;
ieee80211_get_tkip_key(keyconf, skb_frag,
IEEE80211_TKIP_P2_KEY, tx_cmd->key);
IWL_DEBUG_TX(priv, "tx_cmd with tkip hwcrypto\n");
break;
case WLAN_CIPHER_SUITE_WEP104:
tx_cmd->sec_ctl |= TX_CMD_SEC_KEY128;
/* fall through */
case WLAN_CIPHER_SUITE_WEP40:
tx_cmd->sec_ctl |= (TX_CMD_SEC_WEP |
(keyconf->keyidx & TX_CMD_SEC_MSK) << TX_CMD_SEC_SHIFT);
memcpy(&tx_cmd->key[3], keyconf->key, keyconf->keylen);
IWL_DEBUG_TX(priv, "Configuring packet for WEP encryption "
"with key %d\n", keyconf->keyidx);
break;
default:
IWL_ERR(priv, "Unknown encode cipher %x\n", keyconf->cipher);
break;
}
}
/*
* start REPLY_TX command process
*/
int iwlagn_tx_skb(struct iwl_priv *priv, struct sk_buff *skb)
{
struct ieee80211_hdr *hdr = (struct ieee80211_hdr *)skb->data;
struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb);
struct ieee80211_sta *sta = info->control.sta;
struct iwl_station_priv *sta_priv = NULL;
struct iwl_tx_queue *txq;
struct iwl_queue *q;
struct iwl_device_cmd *out_cmd;
struct iwl_cmd_meta *out_meta;
struct iwl_tx_cmd *tx_cmd;
struct iwl_rxon_context *ctx = &priv->contexts[IWL_RXON_CTX_BSS];
int swq_id, txq_id;
dma_addr_t phys_addr;
dma_addr_t txcmd_phys;
dma_addr_t scratch_phys;
u16 len, len_org, firstlen, secondlen;
u16 seq_number = 0;
__le16 fc;
u8 hdr_len;
u8 sta_id;
u8 wait_write_ptr = 0;
u8 tid = 0;
u8 *qc = NULL;
unsigned long flags;
if (info->control.vif)
ctx = iwl_rxon_ctx_from_vif(info->control.vif);
spin_lock_irqsave(&priv->lock, flags);
if (iwl_is_rfkill(priv)) {
IWL_DEBUG_DROP(priv, "Dropping - RF KILL\n");
goto drop_unlock;
}
fc = hdr->frame_control;
#ifdef CONFIG_IWLWIFI_DEBUG
if (ieee80211_is_auth(fc))
IWL_DEBUG_TX(priv, "Sending AUTH frame\n");
else if (ieee80211_is_assoc_req(fc))
IWL_DEBUG_TX(priv, "Sending ASSOC frame\n");
else if (ieee80211_is_reassoc_req(fc))
IWL_DEBUG_TX(priv, "Sending REASSOC frame\n");
#endif
hdr_len = ieee80211_hdrlen(fc);
/* Find index into station table for destination station */
sta_id = iwl_sta_id_or_broadcast(priv, ctx, info->control.sta);
if (sta_id == IWL_INVALID_STATION) {
IWL_DEBUG_DROP(priv, "Dropping - INVALID STATION: %pM\n",
hdr->addr1);
goto drop_unlock;
}
IWL_DEBUG_TX(priv, "station Id %d\n", sta_id);
if (sta)
sta_priv = (void *)sta->drv_priv;
if (sta_priv && sta_priv->asleep) {
WARN_ON(!(info->flags & IEEE80211_TX_CTL_PSPOLL_RESPONSE));
/*
* This sends an asynchronous command to the device,
* but we can rely on it being processed before the
* next frame is processed -- and the next frame to
* this station is the one that will consume this
* counter.
* For now set the counter to just 1 since we do not
* support uAPSD yet.
*/
iwl_sta_modify_sleep_tx_count(priv, sta_id, 1);
}
/*
* Send this frame after DTIM -- there's a special queue
* reserved for this for contexts that support AP mode.
*/
if (info->flags & IEEE80211_TX_CTL_SEND_AFTER_DTIM) {
txq_id = ctx->mcast_queue;
/*
* The microcode will clear the more data
* bit in the last frame it transmits.
*/
hdr->frame_control |=
cpu_to_le16(IEEE80211_FCTL_MOREDATA);
} else
txq_id = ctx->ac_to_queue[skb_get_queue_mapping(skb)];
/* irqs already disabled/saved above when locking priv->lock */
spin_lock(&priv->sta_lock);
if (ieee80211_is_data_qos(fc)) {
qc = ieee80211_get_qos_ctl(hdr);
tid = qc[0] & IEEE80211_QOS_CTL_TID_MASK;
if (WARN_ON_ONCE(tid >= MAX_TID_COUNT)) {
spin_unlock(&priv->sta_lock);
goto drop_unlock;
}
seq_number = priv->stations[sta_id].tid[tid].seq_number;
seq_number &= IEEE80211_SCTL_SEQ;
hdr->seq_ctrl = hdr->seq_ctrl &
cpu_to_le16(IEEE80211_SCTL_FRAG);
hdr->seq_ctrl |= cpu_to_le16(seq_number);
seq_number += 0x10;
/* aggregation is on for this <sta,tid> */
if (info->flags & IEEE80211_TX_CTL_AMPDU &&
priv->stations[sta_id].tid[tid].agg.state == IWL_AGG_ON) {
txq_id = priv->stations[sta_id].tid[tid].agg.txq_id;
}
}
txq = &priv->txq[txq_id];
swq_id = txq->swq_id;
q = &txq->q;
if (unlikely(iwl_queue_space(q) < q->high_mark)) {
spin_unlock(&priv->sta_lock);
goto drop_unlock;
}
if (ieee80211_is_data_qos(fc)) {
priv->stations[sta_id].tid[tid].tfds_in_queue++;
if (!ieee80211_has_morefrags(fc))
priv->stations[sta_id].tid[tid].seq_number = seq_number;
}
spin_unlock(&priv->sta_lock);
/* Set up driver data for this TFD */
memset(&(txq->txb[q->write_ptr]), 0, sizeof(struct iwl_tx_info));
txq->txb[q->write_ptr].skb = skb;
txq->txb[q->write_ptr].ctx = ctx;
/* Set up first empty entry in queue's array of Tx/cmd buffers */
out_cmd = txq->cmd[q->write_ptr];
out_meta = &txq->meta[q->write_ptr];
tx_cmd = &out_cmd->cmd.tx;
memset(&out_cmd->hdr, 0, sizeof(out_cmd->hdr));
memset(tx_cmd, 0, sizeof(struct iwl_tx_cmd));
/*
* Set up the Tx-command (not MAC!) header.
* Store the chosen Tx queue and TFD index within the sequence field;
* after Tx, uCode's Tx response will return this value so driver can
* locate the frame within the tx queue and do post-tx processing.
*/
out_cmd->hdr.cmd = REPLY_TX;
out_cmd->hdr.sequence = cpu_to_le16((u16)(QUEUE_TO_SEQ(txq_id) |
INDEX_TO_SEQ(q->write_ptr)));
/* Copy MAC header from skb into command buffer */
memcpy(tx_cmd->hdr, hdr, hdr_len);
/* Total # bytes to be transmitted */
len = (u16)skb->len;
tx_cmd->len = cpu_to_le16(len);
if (info->control.hw_key)
iwlagn_tx_cmd_build_hwcrypto(priv, info, tx_cmd, skb, sta_id);
/* TODO need this for burst mode later on */
iwlagn_tx_cmd_build_basic(priv, skb, tx_cmd, info, hdr, sta_id);
iwl_dbg_log_tx_data_frame(priv, len, hdr);
iwlagn_tx_cmd_build_rate(priv, tx_cmd, info, fc);
iwl_update_stats(priv, true, fc, len);
/*
* Use the first empty entry in this queue's command buffer array
* to contain the Tx command and MAC header concatenated together
* (payload data will be in another buffer).
* Size of this varies, due to varying MAC header length.
* If end is not dword aligned, we'll have 2 extra bytes at the end
* of the MAC header (device reads on dword boundaries).
* We'll tell device about this padding later.
*/
len = sizeof(struct iwl_tx_cmd) +
sizeof(struct iwl_cmd_header) + hdr_len;
len_org = len;
firstlen = len = (len + 3) & ~3;
if (len_org != len)
len_org = 1;
else
len_org = 0;
/* Tell NIC about any 2-byte padding after MAC header */
if (len_org)
tx_cmd->tx_flags |= TX_CMD_FLG_MH_PAD_MSK;
/* Physical address of this Tx command's header (not MAC header!),
* within command buffer array. */
txcmd_phys = pci_map_single(priv->pci_dev,
&out_cmd->hdr, len,
PCI_DMA_BIDIRECTIONAL);
dma_unmap_addr_set(out_meta, mapping, txcmd_phys);
dma_unmap_len_set(out_meta, len, len);
/* Add buffer containing Tx command and MAC(!) header to TFD's
* first entry */
priv->cfg->ops->lib->txq_attach_buf_to_tfd(priv, txq,
txcmd_phys, len, 1, 0);
if (!ieee80211_has_morefrags(hdr->frame_control)) {
txq->need_update = 1;
} else {
wait_write_ptr = 1;
txq->need_update = 0;
}
/* Set up TFD's 2nd entry to point directly to remainder of skb,
* if any (802.11 null frames have no payload). */
secondlen = len = skb->len - hdr_len;
if (len) {
phys_addr = pci_map_single(priv->pci_dev, skb->data + hdr_len,
len, PCI_DMA_TODEVICE);
priv->cfg->ops->lib->txq_attach_buf_to_tfd(priv, txq,
phys_addr, len,
0, 0);
}
scratch_phys = txcmd_phys + sizeof(struct iwl_cmd_header) +
offsetof(struct iwl_tx_cmd, scratch);
len = sizeof(struct iwl_tx_cmd) +
sizeof(struct iwl_cmd_header) + hdr_len;
/* take back ownership of DMA buffer to enable update */
pci_dma_sync_single_for_cpu(priv->pci_dev, txcmd_phys,
len, PCI_DMA_BIDIRECTIONAL);
tx_cmd->dram_lsb_ptr = cpu_to_le32(scratch_phys);
tx_cmd->dram_msb_ptr = iwl_get_dma_hi_addr(scratch_phys);
IWL_DEBUG_TX(priv, "sequence nr = 0X%x\n",
le16_to_cpu(out_cmd->hdr.sequence));
IWL_DEBUG_TX(priv, "tx_flags = 0X%x\n", le32_to_cpu(tx_cmd->tx_flags));
iwl_print_hex_dump(priv, IWL_DL_TX, (u8 *)tx_cmd, sizeof(*tx_cmd));
iwl_print_hex_dump(priv, IWL_DL_TX, (u8 *)tx_cmd->hdr, hdr_len);
/* Set up entry for this TFD in Tx byte-count array */
if (info->flags & IEEE80211_TX_CTL_AMPDU)
priv->cfg->ops->lib->txq_update_byte_cnt_tbl(priv, txq,
le16_to_cpu(tx_cmd->len));
pci_dma_sync_single_for_device(priv->pci_dev, txcmd_phys,
len, PCI_DMA_BIDIRECTIONAL);
trace_iwlwifi_dev_tx(priv,
&((struct iwl_tfd *)txq->tfds)[txq->q.write_ptr],
sizeof(struct iwl_tfd),
&out_cmd->hdr, firstlen,
skb->data + hdr_len, secondlen);
/* Tell device the write index *just past* this latest filled TFD */
q->write_ptr = iwl_queue_inc_wrap(q->write_ptr, q->n_bd);
iwl_txq_update_write_ptr(priv, txq);
spin_unlock_irqrestore(&priv->lock, flags);
/*
* At this point the frame is "transmitted" successfully
* and we will get a TX status notification eventually,
* regardless of the value of ret. "ret" only indicates
* whether or not we should update the write pointer.
*/
/* avoid atomic ops if it isn't an associated client */
if (sta_priv && sta_priv->client)
atomic_inc(&sta_priv->pending_frames);
if ((iwl_queue_space(q) < q->high_mark) && priv->mac80211_registered) {
if (wait_write_ptr) {
spin_lock_irqsave(&priv->lock, flags);
txq->need_update = 1;
iwl_txq_update_write_ptr(priv, txq);
spin_unlock_irqrestore(&priv->lock, flags);
} else {
iwl_stop_queue(priv, txq->swq_id);
}
}
return 0;
drop_unlock:
spin_unlock_irqrestore(&priv->lock, flags);
return -1;
}
static inline int iwlagn_alloc_dma_ptr(struct iwl_priv *priv,
struct iwl_dma_ptr *ptr, size_t size)
{
ptr->addr = dma_alloc_coherent(&priv->pci_dev->dev, size, &ptr->dma,
GFP_KERNEL);
if (!ptr->addr)
return -ENOMEM;
ptr->size = size;
return 0;
}
static inline void iwlagn_free_dma_ptr(struct iwl_priv *priv,
struct iwl_dma_ptr *ptr)
{
if (unlikely(!ptr->addr))
return;
dma_free_coherent(&priv->pci_dev->dev, ptr->size, ptr->addr, ptr->dma);
memset(ptr, 0, sizeof(*ptr));
}
/**
* iwlagn_hw_txq_ctx_free - Free TXQ Context
*
* Destroy all TX DMA queues and structures
*/
void iwlagn_hw_txq_ctx_free(struct iwl_priv *priv)
{
int txq_id;
/* Tx queues */
if (priv->txq) {
for (txq_id = 0; txq_id < priv->hw_params.max_txq_num; txq_id++)
if (txq_id == priv->cmd_queue)
iwl_cmd_queue_free(priv);
else
iwl_tx_queue_free(priv, txq_id);
}
iwlagn_free_dma_ptr(priv, &priv->kw);
iwlagn_free_dma_ptr(priv, &priv->scd_bc_tbls);
/* free tx queue structure */
iwl_free_txq_mem(priv);
}
/**
* iwlagn_txq_ctx_alloc - allocate TX queue context
* Allocate all Tx DMA structures and initialize them
*
* @param priv
* @return error code
*/
int iwlagn_txq_ctx_alloc(struct iwl_priv *priv)
{
int ret;
int txq_id, slots_num;
unsigned long flags;
/* Free all tx/cmd queues and keep-warm buffer */
iwlagn_hw_txq_ctx_free(priv);
ret = iwlagn_alloc_dma_ptr(priv, &priv->scd_bc_tbls,
priv->hw_params.scd_bc_tbls_size);
if (ret) {
IWL_ERR(priv, "Scheduler BC Table allocation failed\n");
goto error_bc_tbls;
}
/* Alloc keep-warm buffer */
ret = iwlagn_alloc_dma_ptr(priv, &priv->kw, IWL_KW_SIZE);
if (ret) {
IWL_ERR(priv, "Keep Warm allocation failed\n");
goto error_kw;
}
/* allocate tx queue structure */
ret = iwl_alloc_txq_mem(priv);
if (ret)
goto error;
spin_lock_irqsave(&priv->lock, flags);
/* Turn off all Tx DMA fifos */
priv->cfg->ops->lib->txq_set_sched(priv, 0);
/* Tell NIC where to find the "keep warm" buffer */
iwl_write_direct32(priv, FH_KW_MEM_ADDR_REG, priv->kw.dma >> 4);
spin_unlock_irqrestore(&priv->lock, flags);
/* Alloc and init all Tx queues, including the command queue (#4/#9) */
for (txq_id = 0; txq_id < priv->hw_params.max_txq_num; txq_id++) {
slots_num = (txq_id == priv->cmd_queue) ?
TFD_CMD_SLOTS : TFD_TX_CMD_SLOTS;
ret = iwl_tx_queue_init(priv, &priv->txq[txq_id], slots_num,
txq_id);
if (ret) {
IWL_ERR(priv, "Tx %d queue init failed\n", txq_id);
goto error;
}
}
return ret;
error:
iwlagn_hw_txq_ctx_free(priv);
iwlagn_free_dma_ptr(priv, &priv->kw);
error_kw:
iwlagn_free_dma_ptr(priv, &priv->scd_bc_tbls);
error_bc_tbls:
return ret;
}
void iwlagn_txq_ctx_reset(struct iwl_priv *priv)
{
int txq_id, slots_num;
unsigned long flags;
spin_lock_irqsave(&priv->lock, flags);
/* Turn off all Tx DMA fifos */
priv->cfg->ops->lib->txq_set_sched(priv, 0);
/* Tell NIC where to find the "keep warm" buffer */
iwl_write_direct32(priv, FH_KW_MEM_ADDR_REG, priv->kw.dma >> 4);
spin_unlock_irqrestore(&priv->lock, flags);
/* Alloc and init all Tx queues, including the command queue (#4) */
for (txq_id = 0; txq_id < priv->hw_params.max_txq_num; txq_id++) {
slots_num = txq_id == priv->cmd_queue ?
TFD_CMD_SLOTS : TFD_TX_CMD_SLOTS;
iwl_tx_queue_reset(priv, &priv->txq[txq_id], slots_num, txq_id);
}
}
/**
* iwlagn_txq_ctx_stop - Stop all Tx DMA channels
*/
void iwlagn_txq_ctx_stop(struct iwl_priv *priv)
{
int ch;
unsigned long flags;
/* Turn off all Tx DMA fifos */
spin_lock_irqsave(&priv->lock, flags);
priv->cfg->ops->lib->txq_set_sched(priv, 0);
/* Stop each Tx DMA channel, and wait for it to be idle */
for (ch = 0; ch < priv->hw_params.dma_chnl_num; ch++) {
iwl_write_direct32(priv, FH_TCSR_CHNL_TX_CONFIG_REG(ch), 0x0);
if (iwl_poll_direct_bit(priv, FH_TSSR_TX_STATUS_REG,
FH_TSSR_TX_STATUS_REG_MSK_CHNL_IDLE(ch),
1000))
IWL_ERR(priv, "Failing on timeout while stopping"
" DMA channel %d [0x%08x]", ch,
iwl_read_direct32(priv, FH_TSSR_TX_STATUS_REG));
}
spin_unlock_irqrestore(&priv->lock, flags);
}
/*
* Find first available (lowest unused) Tx Queue, mark it "active".
* Called only when finding queue for aggregation.
* Should never return anything < 7, because they should already
* be in use as EDCA AC (0-3), Command (4), reserved (5, 6)
*/
static int iwlagn_txq_ctx_activate_free(struct iwl_priv *priv)
{
int txq_id;
for (txq_id = 0; txq_id < priv->hw_params.max_txq_num; txq_id++)
if (!test_and_set_bit(txq_id, &priv->txq_ctx_active_msk))
return txq_id;
return -1;
}
int iwlagn_tx_agg_start(struct iwl_priv *priv, struct ieee80211_vif *vif,
struct ieee80211_sta *sta, u16 tid, u16 *ssn)
{
int sta_id;
int tx_fifo;
int txq_id;
int ret;
unsigned long flags;
struct iwl_tid_data *tid_data;
tx_fifo = get_fifo_from_tid(iwl_rxon_ctx_from_vif(vif), tid);
if (unlikely(tx_fifo < 0))
return tx_fifo;
IWL_WARN(priv, "%s on ra = %pM tid = %d\n",
__func__, sta->addr, tid);
sta_id = iwl_sta_id(sta);
if (sta_id == IWL_INVALID_STATION) {
IWL_ERR(priv, "Start AGG on invalid station\n");
return -ENXIO;
}
if (unlikely(tid >= MAX_TID_COUNT))
return -EINVAL;
if (priv->stations[sta_id].tid[tid].agg.state != IWL_AGG_OFF) {
IWL_ERR(priv, "Start AGG when state is not IWL_AGG_OFF !\n");
return -ENXIO;
}
txq_id = iwlagn_txq_ctx_activate_free(priv);
if (txq_id == -1) {
IWL_ERR(priv, "No free aggregation queue available\n");
return -ENXIO;
}
spin_lock_irqsave(&priv->sta_lock, flags);
tid_data = &priv->stations[sta_id].tid[tid];
*ssn = SEQ_TO_SN(tid_data->seq_number);
tid_data->agg.txq_id = txq_id;
priv->txq[txq_id].swq_id = iwl_virtual_agg_queue_num(get_ac_from_tid(tid), txq_id);
spin_unlock_irqrestore(&priv->sta_lock, flags);
ret = priv->cfg->ops->lib->txq_agg_enable(priv, txq_id, tx_fifo,
sta_id, tid, *ssn);
if (ret)
return ret;
spin_lock_irqsave(&priv->sta_lock, flags);
tid_data = &priv->stations[sta_id].tid[tid];
if (tid_data->tfds_in_queue == 0) {
IWL_DEBUG_HT(priv, "HW queue is empty\n");
tid_data->agg.state = IWL_AGG_ON;
ieee80211_start_tx_ba_cb_irqsafe(vif, sta->addr, tid);
} else {
IWL_DEBUG_HT(priv, "HW queue is NOT empty: %d packets in HW queue\n",
tid_data->tfds_in_queue);
tid_data->agg.state = IWL_EMPTYING_HW_QUEUE_ADDBA;
}
spin_unlock_irqrestore(&priv->sta_lock, flags);
return ret;
}
int iwlagn_tx_agg_stop(struct iwl_priv *priv, struct ieee80211_vif *vif,
struct ieee80211_sta *sta, u16 tid)
{
int tx_fifo_id, txq_id, sta_id, ssn;
struct iwl_tid_data *tid_data;
int write_ptr, read_ptr;
unsigned long flags;
tx_fifo_id = get_fifo_from_tid(iwl_rxon_ctx_from_vif(vif), tid);
if (unlikely(tx_fifo_id < 0))
return tx_fifo_id;
sta_id = iwl_sta_id(sta);
if (sta_id == IWL_INVALID_STATION) {
IWL_ERR(priv, "Invalid station for AGG tid %d\n", tid);
return -ENXIO;
}
spin_lock_irqsave(&priv->sta_lock, flags);
tid_data = &priv->stations[sta_id].tid[tid];
ssn = (tid_data->seq_number & IEEE80211_SCTL_SEQ) >> 4;
txq_id = tid_data->agg.txq_id;
switch (priv->stations[sta_id].tid[tid].agg.state) {
case IWL_EMPTYING_HW_QUEUE_ADDBA:
/*
* This can happen if the peer stops aggregation
* again before we've had a chance to drain the
* queue we selected previously, i.e. before the
* session was really started completely.
*/
IWL_DEBUG_HT(priv, "AGG stop before setup done\n");
goto turn_off;
case IWL_AGG_ON:
break;
default:
IWL_WARN(priv, "Stopping AGG while state not ON or starting\n");
}
write_ptr = priv->txq[txq_id].q.write_ptr;
read_ptr = priv->txq[txq_id].q.read_ptr;
/* The queue is not empty */
if (write_ptr != read_ptr) {
IWL_DEBUG_HT(priv, "Stopping a non empty AGG HW QUEUE\n");
priv->stations[sta_id].tid[tid].agg.state =
IWL_EMPTYING_HW_QUEUE_DELBA;
spin_unlock_irqrestore(&priv->sta_lock, flags);
return 0;
}
IWL_DEBUG_HT(priv, "HW queue is empty\n");
turn_off:
priv->stations[sta_id].tid[tid].agg.state = IWL_AGG_OFF;
/* do not restore/save irqs */
spin_unlock(&priv->sta_lock);
spin_lock(&priv->lock);
/*
* the only reason this call can fail is queue number out of range,
* which can happen if uCode is reloaded and all the station
* information are lost. if it is outside the range, there is no need
* to deactivate the uCode queue, just return "success" to allow
* mac80211 to clean up it own data.
*/
priv->cfg->ops->lib->txq_agg_disable(priv, txq_id, ssn,
tx_fifo_id);
spin_unlock_irqrestore(&priv->lock, flags);
ieee80211_stop_tx_ba_cb_irqsafe(vif, sta->addr, tid);
return 0;
}
int iwlagn_txq_check_empty(struct iwl_priv *priv,
int sta_id, u8 tid, int txq_id)
{
struct iwl_queue *q = &priv->txq[txq_id].q;
u8 *addr = priv->stations[sta_id].sta.sta.addr;
struct iwl_tid_data *tid_data = &priv->stations[sta_id].tid[tid];
struct iwl_rxon_context *ctx;
ctx = &priv->contexts[priv->stations[sta_id].ctxid];
lockdep_assert_held(&priv->sta_lock);
switch (priv->stations[sta_id].tid[tid].agg.state) {
case IWL_EMPTYING_HW_QUEUE_DELBA:
/* We are reclaiming the last packet of the */
/* aggregated HW queue */
if ((txq_id == tid_data->agg.txq_id) &&
(q->read_ptr == q->write_ptr)) {
u16 ssn = SEQ_TO_SN(tid_data->seq_number);
int tx_fifo = get_fifo_from_tid(ctx, tid);
IWL_DEBUG_HT(priv, "HW queue empty: continue DELBA flow\n");
priv->cfg->ops->lib->txq_agg_disable(priv, txq_id,
ssn, tx_fifo);
tid_data->agg.state = IWL_AGG_OFF;
ieee80211_stop_tx_ba_cb_irqsafe(ctx->vif, addr, tid);
}
break;
case IWL_EMPTYING_HW_QUEUE_ADDBA:
/* We are reclaiming the last packet of the queue */
if (tid_data->tfds_in_queue == 0) {
IWL_DEBUG_HT(priv, "HW queue empty: continue ADDBA flow\n");
tid_data->agg.state = IWL_AGG_ON;
ieee80211_start_tx_ba_cb_irqsafe(ctx->vif, addr, tid);
}
break;
}
return 0;
}
static void iwlagn_tx_status(struct iwl_priv *priv, struct iwl_tx_info *tx_info)
{
struct ieee80211_hdr *hdr = (struct ieee80211_hdr *) tx_info->skb->data;
struct ieee80211_sta *sta;
struct iwl_station_priv *sta_priv;
rcu_read_lock();
sta = ieee80211_find_sta(tx_info->ctx->vif, hdr->addr1);
if (sta) {
sta_priv = (void *)sta->drv_priv;
/* avoid atomic ops if this isn't a client */
if (sta_priv->client &&
atomic_dec_return(&sta_priv->pending_frames) == 0)
ieee80211_sta_block_awake(priv->hw, sta, false);
}
rcu_read_unlock();
ieee80211_tx_status_irqsafe(priv->hw, tx_info->skb);
}
int iwlagn_tx_queue_reclaim(struct iwl_priv *priv, int txq_id, int index)
{
struct iwl_tx_queue *txq = &priv->txq[txq_id];
struct iwl_queue *q = &txq->q;
struct iwl_tx_info *tx_info;
int nfreed = 0;
struct ieee80211_hdr *hdr;
if ((index >= q->n_bd) || (iwl_queue_used(q, index) == 0)) {
IWL_ERR(priv, "Read index for DMA queue txq id (%d), index %d, "
"is out of range [0-%d] %d %d.\n", txq_id,
index, q->n_bd, q->write_ptr, q->read_ptr);
return 0;
}
for (index = iwl_queue_inc_wrap(index, q->n_bd);
q->read_ptr != index;
q->read_ptr = iwl_queue_inc_wrap(q->read_ptr, q->n_bd)) {
tx_info = &txq->txb[txq->q.read_ptr];
iwlagn_tx_status(priv, tx_info);
hdr = (struct ieee80211_hdr *)tx_info->skb->data;
if (hdr && ieee80211_is_data_qos(hdr->frame_control))
nfreed++;
tx_info->skb = NULL;
if (priv->cfg->ops->lib->txq_inval_byte_cnt_tbl)
priv->cfg->ops->lib->txq_inval_byte_cnt_tbl(priv, txq);
priv->cfg->ops->lib->txq_free_tfd(priv, txq);
}
return nfreed;
}
/**
* iwlagn_tx_status_reply_compressed_ba - Update tx status from block-ack
*
* Go through block-ack's bitmap of ACK'd frames, update driver's record of
* ACK vs. not. This gets sent to mac80211, then to rate scaling algo.
*/
static int iwlagn_tx_status_reply_compressed_ba(struct iwl_priv *priv,
struct iwl_ht_agg *agg,
struct iwl_compressed_ba_resp *ba_resp)
{
int i, sh, ack;
u16 seq_ctl = le16_to_cpu(ba_resp->seq_ctl);
u16 scd_flow = le16_to_cpu(ba_resp->scd_flow);
iwlwifi: parse block ack responses correctly Compressed BlockAck frames store the ACKs/NACKs in a 64-bit bitmap that starts at the sequence number of the first frame sent in the aggregated batch. Note that this is a selective ACKnowledgement following selective retransmission; e.g., if frames 1,4-5 in a batch are ACKed then the next transmission will include frames 2-3,6-10 (7 frames). In this latter case, the Compressed BlockAck will not have all meaningful information in the low order bits -- the semantically meaningful bits of the BA will be 0x1f3 (where the low-order frame is seq 2). The driver code originally just looked at the lower (in this case, 7) bits of the BlockAck. In this case, the lower 7 bits of 0x1f3 => only 5 packets, maximum, could ever be ACKed. In reality it should be looking at all of the bits, filtered by those corresponding to packets that were actually sent. This flaw meant that the number of correctly ACked packets could be significantly underreported and might result in asynchronous state between TX and RX sides as well as driver and uCode. Fix this and also add a shortcut that doesn't require the code to loop through all 64 bits of the bitmap but rather stops when no higher packets are ACKed. In my experiments this fix greatly reduces throughput swing, making throughput stable and high. It is also likely related to some of the stalls observed in aggregation mode and maybe some of the buffer underruns observed, e.g., http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=1968 http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=2098 http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=2018 Signed-off-by: Daniel Halperin <dhalperi@cs.washington.edu> Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
2010-05-24 19:41:30 -06:00
u64 bitmap, sent_bitmap;
int successes = 0;
struct ieee80211_tx_info *info;
if (unlikely(!agg->wait_for_ba)) {
IWL_ERR(priv, "Received BA when not expected\n");
return -EINVAL;
}
/* Mark that the expected block-ack response arrived */
agg->wait_for_ba = 0;
IWL_DEBUG_TX_REPLY(priv, "BA %d %d\n", agg->start_idx, ba_resp->seq_ctl);
/* Calculate shift to align block-ack bits with our Tx window bits */
sh = agg->start_idx - SEQ_TO_INDEX(seq_ctl >> 4);
if (sh < 0) /* tbw something is wrong with indices */
sh += 0x100;
/* don't use 64-bit values for now */
bitmap = le64_to_cpu(ba_resp->bitmap) >> sh;
if (agg->frame_count > (64 - sh)) {
IWL_DEBUG_TX_REPLY(priv, "more frames than bitmap size");
return -1;
}
/* check for success or failure according to the
* transmitted bitmap and block-ack bitmap */
iwlwifi: parse block ack responses correctly Compressed BlockAck frames store the ACKs/NACKs in a 64-bit bitmap that starts at the sequence number of the first frame sent in the aggregated batch. Note that this is a selective ACKnowledgement following selective retransmission; e.g., if frames 1,4-5 in a batch are ACKed then the next transmission will include frames 2-3,6-10 (7 frames). In this latter case, the Compressed BlockAck will not have all meaningful information in the low order bits -- the semantically meaningful bits of the BA will be 0x1f3 (where the low-order frame is seq 2). The driver code originally just looked at the lower (in this case, 7) bits of the BlockAck. In this case, the lower 7 bits of 0x1f3 => only 5 packets, maximum, could ever be ACKed. In reality it should be looking at all of the bits, filtered by those corresponding to packets that were actually sent. This flaw meant that the number of correctly ACked packets could be significantly underreported and might result in asynchronous state between TX and RX sides as well as driver and uCode. Fix this and also add a shortcut that doesn't require the code to loop through all 64 bits of the bitmap but rather stops when no higher packets are ACKed. In my experiments this fix greatly reduces throughput swing, making throughput stable and high. It is also likely related to some of the stalls observed in aggregation mode and maybe some of the buffer underruns observed, e.g., http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=1968 http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=2098 http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=2018 Signed-off-by: Daniel Halperin <dhalperi@cs.washington.edu> Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
2010-05-24 19:41:30 -06:00
sent_bitmap = bitmap & agg->bitmap;
/* For each frame attempted in aggregation,
* update driver's record of tx frame's status. */
iwlwifi: parse block ack responses correctly Compressed BlockAck frames store the ACKs/NACKs in a 64-bit bitmap that starts at the sequence number of the first frame sent in the aggregated batch. Note that this is a selective ACKnowledgement following selective retransmission; e.g., if frames 1,4-5 in a batch are ACKed then the next transmission will include frames 2-3,6-10 (7 frames). In this latter case, the Compressed BlockAck will not have all meaningful information in the low order bits -- the semantically meaningful bits of the BA will be 0x1f3 (where the low-order frame is seq 2). The driver code originally just looked at the lower (in this case, 7) bits of the BlockAck. In this case, the lower 7 bits of 0x1f3 => only 5 packets, maximum, could ever be ACKed. In reality it should be looking at all of the bits, filtered by those corresponding to packets that were actually sent. This flaw meant that the number of correctly ACked packets could be significantly underreported and might result in asynchronous state between TX and RX sides as well as driver and uCode. Fix this and also add a shortcut that doesn't require the code to loop through all 64 bits of the bitmap but rather stops when no higher packets are ACKed. In my experiments this fix greatly reduces throughput swing, making throughput stable and high. It is also likely related to some of the stalls observed in aggregation mode and maybe some of the buffer underruns observed, e.g., http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=1968 http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=2098 http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=2018 Signed-off-by: Daniel Halperin <dhalperi@cs.washington.edu> Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
2010-05-24 19:41:30 -06:00
i = 0;
while (sent_bitmap) {
ack = sent_bitmap & 1ULL;
successes += ack;
IWL_DEBUG_TX_REPLY(priv, "%s ON i=%d idx=%d raw=%d\n",
ack ? "ACK" : "NACK", i, (agg->start_idx + i) & 0xff,
agg->start_idx + i);
iwlwifi: parse block ack responses correctly Compressed BlockAck frames store the ACKs/NACKs in a 64-bit bitmap that starts at the sequence number of the first frame sent in the aggregated batch. Note that this is a selective ACKnowledgement following selective retransmission; e.g., if frames 1,4-5 in a batch are ACKed then the next transmission will include frames 2-3,6-10 (7 frames). In this latter case, the Compressed BlockAck will not have all meaningful information in the low order bits -- the semantically meaningful bits of the BA will be 0x1f3 (where the low-order frame is seq 2). The driver code originally just looked at the lower (in this case, 7) bits of the BlockAck. In this case, the lower 7 bits of 0x1f3 => only 5 packets, maximum, could ever be ACKed. In reality it should be looking at all of the bits, filtered by those corresponding to packets that were actually sent. This flaw meant that the number of correctly ACked packets could be significantly underreported and might result in asynchronous state between TX and RX sides as well as driver and uCode. Fix this and also add a shortcut that doesn't require the code to loop through all 64 bits of the bitmap but rather stops when no higher packets are ACKed. In my experiments this fix greatly reduces throughput swing, making throughput stable and high. It is also likely related to some of the stalls observed in aggregation mode and maybe some of the buffer underruns observed, e.g., http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=1968 http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=2098 http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=2018 Signed-off-by: Daniel Halperin <dhalperi@cs.washington.edu> Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
2010-05-24 19:41:30 -06:00
sent_bitmap >>= 1;
++i;
}
info = IEEE80211_SKB_CB(priv->txq[scd_flow].txb[agg->start_idx].skb);
memset(&info->status, 0, sizeof(info->status));
info->flags |= IEEE80211_TX_STAT_ACK;
info->flags |= IEEE80211_TX_STAT_AMPDU;
info->status.ampdu_ack_len = successes;
info->status.ampdu_len = agg->frame_count;
iwlagn_hwrate_to_tx_control(priv, agg->rate_n_flags, info);
IWL_DEBUG_TX_REPLY(priv, "Bitmap %llx\n", (unsigned long long)bitmap);
return 0;
}
/**
* translate ucode response to mac80211 tx status control values
*/
void iwlagn_hwrate_to_tx_control(struct iwl_priv *priv, u32 rate_n_flags,
struct ieee80211_tx_info *info)
{
struct ieee80211_tx_rate *r = &info->control.rates[0];
info->antenna_sel_tx =
((rate_n_flags & RATE_MCS_ANT_ABC_MSK) >> RATE_MCS_ANT_POS);
if (rate_n_flags & RATE_MCS_HT_MSK)
r->flags |= IEEE80211_TX_RC_MCS;
if (rate_n_flags & RATE_MCS_GF_MSK)
r->flags |= IEEE80211_TX_RC_GREEN_FIELD;
if (rate_n_flags & RATE_MCS_HT40_MSK)
r->flags |= IEEE80211_TX_RC_40_MHZ_WIDTH;
if (rate_n_flags & RATE_MCS_DUP_MSK)
r->flags |= IEEE80211_TX_RC_DUP_DATA;
if (rate_n_flags & RATE_MCS_SGI_MSK)
r->flags |= IEEE80211_TX_RC_SHORT_GI;
r->idx = iwlagn_hwrate_to_mac80211_idx(rate_n_flags, info->band);
}
/**
* iwlagn_rx_reply_compressed_ba - Handler for REPLY_COMPRESSED_BA
*
* Handles block-acknowledge notification from device, which reports success
* of frames sent via aggregation.
*/
void iwlagn_rx_reply_compressed_ba(struct iwl_priv *priv,
struct iwl_rx_mem_buffer *rxb)
{
struct iwl_rx_packet *pkt = rxb_addr(rxb);
struct iwl_compressed_ba_resp *ba_resp = &pkt->u.compressed_ba;
struct iwl_tx_queue *txq = NULL;
struct iwl_ht_agg *agg;
int index;
int sta_id;
int tid;
unsigned long flags;
/* "flow" corresponds to Tx queue */
u16 scd_flow = le16_to_cpu(ba_resp->scd_flow);
/* "ssn" is start of block-ack Tx window, corresponds to index
* (in Tx queue's circular buffer) of first TFD/frame in window */
u16 ba_resp_scd_ssn = le16_to_cpu(ba_resp->scd_ssn);
if (scd_flow >= priv->hw_params.max_txq_num) {
IWL_ERR(priv,
"BUG_ON scd_flow is bigger than number of queues\n");
return;
}
txq = &priv->txq[scd_flow];
sta_id = ba_resp->sta_id;
tid = ba_resp->tid;
agg = &priv->stations[sta_id].tid[tid].agg;
if (unlikely(agg->txq_id != scd_flow)) {
/*
* FIXME: this is a uCode bug which need to be addressed,
* log the information and return for now!
* since it is possible happen very often and in order
* not to fill the syslog, don't enable the logging by default
*/
IWL_DEBUG_TX_REPLY(priv,
"BA scd_flow %d does not match txq_id %d\n",
scd_flow, agg->txq_id);
return;
}
/* Find index just before block-ack window */
index = iwl_queue_dec_wrap(ba_resp_scd_ssn & 0xff, txq->q.n_bd);
spin_lock_irqsave(&priv->sta_lock, flags);
IWL_DEBUG_TX_REPLY(priv, "REPLY_COMPRESSED_BA [%d] Received from %pM, "
"sta_id = %d\n",
agg->wait_for_ba,
(u8 *) &ba_resp->sta_addr_lo32,
ba_resp->sta_id);
IWL_DEBUG_TX_REPLY(priv, "TID = %d, SeqCtl = %d, bitmap = 0x%llx, scd_flow = "
"%d, scd_ssn = %d\n",
ba_resp->tid,
ba_resp->seq_ctl,
(unsigned long long)le64_to_cpu(ba_resp->bitmap),
ba_resp->scd_flow,
ba_resp->scd_ssn);
IWL_DEBUG_TX_REPLY(priv, "DAT start_idx = %d, bitmap = 0x%llx\n",
agg->start_idx,
(unsigned long long)agg->bitmap);
/* Update driver's record of ACK vs. not for each frame in window */
iwlagn_tx_status_reply_compressed_ba(priv, agg, ba_resp);
/* Release all TFDs before the SSN, i.e. all TFDs in front of
* block-ack window (we assume that they've been successfully
* transmitted ... if not, it's too late anyway). */
if (txq->q.read_ptr != (ba_resp_scd_ssn & 0xff)) {
/* calculate mac80211 ampdu sw queue to wake */
int freed = iwlagn_tx_queue_reclaim(priv, scd_flow, index);
iwl_free_tfds_in_queue(priv, sta_id, tid, freed);
if ((iwl_queue_space(&txq->q) > txq->q.low_mark) &&
priv->mac80211_registered &&
(agg->state != IWL_EMPTYING_HW_QUEUE_DELBA))
iwl_wake_queue(priv, txq->swq_id);
iwlagn_txq_check_empty(priv, sta_id, tid, scd_flow);
}
spin_unlock_irqrestore(&priv->sta_lock, flags);
}