diff options
Diffstat (limited to 'drivers/net/wireless/iwlwifi/iwl-agn-lib.c')
-rw-r--r-- | drivers/net/wireless/iwlwifi/iwl-agn-lib.c | 1113 |
1 files changed, 1113 insertions, 0 deletions
diff --git a/drivers/net/wireless/iwlwifi/iwl-agn-lib.c b/drivers/net/wireless/iwlwifi/iwl-agn-lib.c new file mode 100644 index 0000000..c465c85 --- /dev/null +++ b/drivers/net/wireless/iwlwifi/iwl-agn-lib.c @@ -0,0 +1,1113 @@ +/****************************************************************************** + * + * 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/etherdevice.h> +#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-io.h" +#include "iwl-helpers.h" +#include "iwl-agn-hw.h" +#include "iwl-agn.h" + +static inline u32 iwlagn_get_scd_ssn(struct iwl5000_tx_resp *tx_resp) +{ + return le32_to_cpup((__le32 *)&tx_resp->status + + tx_resp->frame_count) & MAX_SN; +} + +static int iwlagn_tx_status_reply_tx(struct iwl_priv *priv, + struct iwl_ht_agg *agg, + struct iwl5000_tx_resp *tx_resp, + int txq_id, u16 start_idx) +{ + u16 status; + struct agg_tx_status *frame_status = &tx_resp->status; + struct ieee80211_tx_info *info = NULL; + struct ieee80211_hdr *hdr = NULL; + u32 rate_n_flags = le32_to_cpu(tx_resp->rate_n_flags); + int i, sh, idx; + u16 seq; + + if (agg->wait_for_ba) + IWL_DEBUG_TX_REPLY(priv, "got tx response w/o block-ack\n"); + + agg->frame_count = tx_resp->frame_count; + agg->start_idx = start_idx; + agg->rate_n_flags = rate_n_flags; + agg->bitmap = 0; + + /* # frames attempted by Tx command */ + if (agg->frame_count == 1) { + /* Only one frame was attempted; no block-ack will arrive */ + status = le16_to_cpu(frame_status[0].status); + idx = start_idx; + + /* FIXME: code repetition */ + IWL_DEBUG_TX_REPLY(priv, "FrameCnt = %d, StartIdx=%d idx=%d\n", + agg->frame_count, agg->start_idx, idx); + + info = IEEE80211_SKB_CB(priv->txq[txq_id].txb[idx].skb[0]); + info->status.rates[0].count = tx_resp->failure_frame + 1; + info->flags &= ~IEEE80211_TX_CTL_AMPDU; + info->flags |= iwl_tx_status_to_mac80211(status); + iwlagn_hwrate_to_tx_control(priv, rate_n_flags, info); + + /* FIXME: code repetition end */ + + IWL_DEBUG_TX_REPLY(priv, "1 Frame 0x%x failure :%d\n", + status & 0xff, tx_resp->failure_frame); + IWL_DEBUG_TX_REPLY(priv, "Rate Info rate_n_flags=%x\n", rate_n_flags); + + agg->wait_for_ba = 0; + } else { + /* Two or more frames were attempted; expect block-ack */ + u64 bitmap = 0; + int start = agg->start_idx; + + /* Construct bit-map of pending frames within Tx window */ + for (i = 0; i < agg->frame_count; i++) { + u16 sc; + status = le16_to_cpu(frame_status[i].status); + seq = le16_to_cpu(frame_status[i].sequence); + idx = SEQ_TO_INDEX(seq); + txq_id = SEQ_TO_QUEUE(seq); + + if (status & (AGG_TX_STATE_FEW_BYTES_MSK | + AGG_TX_STATE_ABORT_MSK)) + continue; + + IWL_DEBUG_TX_REPLY(priv, "FrameCnt = %d, txq_id=%d idx=%d\n", + agg->frame_count, txq_id, idx); + + hdr = iwl_tx_queue_get_hdr(priv, txq_id, idx); + if (!hdr) { + IWL_ERR(priv, + "BUG_ON idx doesn't point to valid skb" + " idx=%d, txq_id=%d\n", idx, txq_id); + return -1; + } + + sc = le16_to_cpu(hdr->seq_ctrl); + if (idx != (SEQ_TO_SN(sc) & 0xff)) { + IWL_ERR(priv, + "BUG_ON idx doesn't match seq control" + " idx=%d, seq_idx=%d, seq=%d\n", + idx, SEQ_TO_SN(sc), + hdr->seq_ctrl); + return -1; + } + + IWL_DEBUG_TX_REPLY(priv, "AGG Frame i=%d idx %d seq=%d\n", + i, idx, SEQ_TO_SN(sc)); + + sh = idx - start; + if (sh > 64) { + sh = (start - idx) + 0xff; + bitmap = bitmap << sh; + sh = 0; + start = idx; + } else if (sh < -64) + sh = 0xff - (start - idx); + else if (sh < 0) { + sh = start - idx; + start = idx; + bitmap = bitmap << sh; + sh = 0; + } + bitmap |= 1ULL << sh; + IWL_DEBUG_TX_REPLY(priv, "start=%d bitmap=0x%llx\n", + start, (unsigned long long)bitmap); + } + + agg->bitmap = bitmap; + agg->start_idx = start; + IWL_DEBUG_TX_REPLY(priv, "Frames %d start_idx=%d bitmap=0x%llx\n", + agg->frame_count, agg->start_idx, + (unsigned long long)agg->bitmap); + + if (bitmap) + agg->wait_for_ba = 1; + } + return 0; +} + +void iwl_check_abort_status(struct iwl_priv *priv, + u8 frame_count, u32 status) +{ + if (frame_count == 1 && status == TX_STATUS_FAIL_RFKILL_FLUSH) { + IWL_ERR(priv, "TODO: Implement Tx flush command!!!\n"); + } +} + +static void iwlagn_rx_reply_tx(struct iwl_priv *priv, + struct iwl_rx_mem_buffer *rxb) +{ + struct iwl_rx_packet *pkt = rxb_addr(rxb); + u16 sequence = le16_to_cpu(pkt->hdr.sequence); + int txq_id = SEQ_TO_QUEUE(sequence); + int index = SEQ_TO_INDEX(sequence); + struct iwl_tx_queue *txq = &priv->txq[txq_id]; + struct ieee80211_tx_info *info; + struct iwl5000_tx_resp *tx_resp = (void *)&pkt->u.raw[0]; + u32 status = le16_to_cpu(tx_resp->status.status); + int tid; + int sta_id; + int freed; + + if ((index >= txq->q.n_bd) || (iwl_queue_used(&txq->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, txq->q.n_bd, txq->q.write_ptr, + txq->q.read_ptr); + return; + } + + info = IEEE80211_SKB_CB(txq->txb[txq->q.read_ptr].skb[0]); + memset(&info->status, 0, sizeof(info->status)); + + tid = (tx_resp->ra_tid & IWL50_TX_RES_TID_MSK) >> IWL50_TX_RES_TID_POS; + sta_id = (tx_resp->ra_tid & IWL50_TX_RES_RA_MSK) >> IWL50_TX_RES_RA_POS; + + if (txq->sched_retry) { + const u32 scd_ssn = iwlagn_get_scd_ssn(tx_resp); + struct iwl_ht_agg *agg = NULL; + + agg = &priv->stations[sta_id].tid[tid].agg; + + iwlagn_tx_status_reply_tx(priv, agg, tx_resp, txq_id, index); + + /* check if BAR is needed */ + if ((tx_resp->frame_count == 1) && !iwl_is_tx_success(status)) + info->flags |= IEEE80211_TX_STAT_AMPDU_NO_BACK; + + if (txq->q.read_ptr != (scd_ssn & 0xff)) { + index = iwl_queue_dec_wrap(scd_ssn & 0xff, txq->q.n_bd); + IWL_DEBUG_TX_REPLY(priv, "Retry scheduler reclaim " + "scd_ssn=%d idx=%d txq=%d swq=%d\n", + scd_ssn , index, txq_id, txq->swq_id); + + freed = iwlagn_tx_queue_reclaim(priv, txq_id, index); + iwl_free_tfds_in_queue(priv, sta_id, tid, freed); + + if (priv->mac80211_registered && + (iwl_queue_space(&txq->q) > txq->q.low_mark) && + (agg->state != IWL_EMPTYING_HW_QUEUE_DELBA)) { + if (agg->state == IWL_AGG_OFF) + iwl_wake_queue(priv, txq_id); + else + iwl_wake_queue(priv, txq->swq_id); + } + } + } else { + BUG_ON(txq_id != txq->swq_id); + + info->status.rates[0].count = tx_resp->failure_frame + 1; + info->flags |= iwl_tx_status_to_mac80211(status); + iwlagn_hwrate_to_tx_control(priv, + le32_to_cpu(tx_resp->rate_n_flags), + info); + + IWL_DEBUG_TX_REPLY(priv, "TXQ %d status %s (0x%08x) rate_n_flags " + "0x%x retries %d\n", + txq_id, + iwl_get_tx_fail_reason(status), status, + le32_to_cpu(tx_resp->rate_n_flags), + tx_resp->failure_frame); + + freed = iwlagn_tx_queue_reclaim(priv, txq_id, index); + iwl_free_tfds_in_queue(priv, sta_id, tid, freed); + + if (priv->mac80211_registered && + (iwl_queue_space(&txq->q) > txq->q.low_mark)) + iwl_wake_queue(priv, txq_id); + } + + iwlagn_txq_check_empty(priv, sta_id, tid, txq_id); + + iwl_check_abort_status(priv, tx_resp->frame_count, status); +} + +void iwlagn_rx_handler_setup(struct iwl_priv *priv) +{ + /* init calibration handlers */ + priv->rx_handlers[CALIBRATION_RES_NOTIFICATION] = + iwlagn_rx_calib_result; + priv->rx_handlers[CALIBRATION_COMPLETE_NOTIFICATION] = + iwlagn_rx_calib_complete; + priv->rx_handlers[REPLY_TX] = iwlagn_rx_reply_tx; +} + +void iwlagn_setup_deferred_work(struct iwl_priv *priv) +{ + /* in agn, the tx power calibration is done in uCode */ + priv->disable_tx_power_cal = 1; +} + +int iwlagn_hw_valid_rtc_data_addr(u32 addr) +{ + return (addr >= IWLAGN_RTC_DATA_LOWER_BOUND) && + (addr < IWLAGN_RTC_DATA_UPPER_BOUND); +} + +int iwlagn_send_tx_power(struct iwl_priv *priv) +{ + struct iwl5000_tx_power_dbm_cmd tx_power_cmd; + u8 tx_ant_cfg_cmd; + + /* half dBm need to multiply */ + tx_power_cmd.global_lmt = (s8)(2 * priv->tx_power_user_lmt); + + if (priv->tx_power_lmt_in_half_dbm && + priv->tx_power_lmt_in_half_dbm < tx_power_cmd.global_lmt) { + /* + * For the newer devices which using enhanced/extend tx power + * table in EEPROM, the format is in half dBm. driver need to + * convert to dBm format before report to mac80211. + * By doing so, there is a possibility of 1/2 dBm resolution + * lost. driver will perform "round-up" operation before + * reporting, but it will cause 1/2 dBm tx power over the + * regulatory limit. Perform the checking here, if the + * "tx_power_user_lmt" is higher than EEPROM value (in + * half-dBm format), lower the tx power based on EEPROM + */ + tx_power_cmd.global_lmt = priv->tx_power_lmt_in_half_dbm; + } + tx_power_cmd.flags = IWL50_TX_POWER_NO_CLOSED; + tx_power_cmd.srv_chan_lmt = IWL50_TX_POWER_AUTO; + + if (IWL_UCODE_API(priv->ucode_ver) == 1) + tx_ant_cfg_cmd = REPLY_TX_POWER_DBM_CMD_V1; + else + tx_ant_cfg_cmd = REPLY_TX_POWER_DBM_CMD; + + return iwl_send_cmd_pdu_async(priv, tx_ant_cfg_cmd, + sizeof(tx_power_cmd), &tx_power_cmd, + NULL); +} + +void iwlagn_temperature(struct iwl_priv *priv) +{ + /* store temperature from statistics (in Celsius) */ + priv->temperature = le32_to_cpu(priv->statistics.general.temperature); + iwl_tt_handler(priv); +} + +u16 iwlagn_eeprom_calib_version(struct iwl_priv *priv) +{ + struct iwl_eeprom_calib_hdr { + u8 version; + u8 pa_type; + u16 voltage; + } *hdr; + + hdr = (struct iwl_eeprom_calib_hdr *)iwl_eeprom_query_addr(priv, + EEPROM_5000_CALIB_ALL); + return hdr->version; + +} + +/* + * EEPROM + */ +static u32 eeprom_indirect_address(const struct iwl_priv *priv, u32 address) +{ + u16 offset = 0; + + if ((address & INDIRECT_ADDRESS) == 0) + return address; + + switch (address & INDIRECT_TYPE_MSK) { + case INDIRECT_HOST: + offset = iwl_eeprom_query16(priv, EEPROM_5000_LINK_HOST); + break; + case INDIRECT_GENERAL: + offset = iwl_eeprom_query16(priv, EEPROM_5000_LINK_GENERAL); + break; + case INDIRECT_REGULATORY: + offset = iwl_eeprom_query16(priv, EEPROM_5000_LINK_REGULATORY); + break; + case INDIRECT_CALIBRATION: + offset = iwl_eeprom_query16(priv, EEPROM_5000_LINK_CALIBRATION); + break; + case INDIRECT_PROCESS_ADJST: + offset = iwl_eeprom_query16(priv, EEPROM_5000_LINK_PROCESS_ADJST); + break; + case INDIRECT_OTHERS: + offset = iwl_eeprom_query16(priv, EEPROM_5000_LINK_OTHERS); + break; + default: + IWL_ERR(priv, "illegal indirect type: 0x%X\n", + address & INDIRECT_TYPE_MSK); + break; + } + + /* translate the offset from words to byte */ + return (address & ADDRESS_MSK) + (offset << 1); +} + +const u8 *iwlagn_eeprom_query_addr(const struct iwl_priv *priv, + size_t offset) +{ + u32 address = eeprom_indirect_address(priv, offset); + BUG_ON(address >= priv->cfg->eeprom_size); + return &priv->eeprom[address]; +} + +struct iwl_mod_params iwlagn_mod_params = { + .amsdu_size_8K = 1, + .restart_fw = 1, + /* the rest are 0 by default */ +}; + +void iwlagn_rx_queue_reset(struct iwl_priv *priv, struct iwl_rx_queue *rxq) +{ + unsigned long flags; + int i; + spin_lock_irqsave(&rxq->lock, flags); + INIT_LIST_HEAD(&rxq->rx_free); + INIT_LIST_HEAD(&rxq->rx_used); + /* Fill the rx_used queue with _all_ of the Rx buffers */ + for (i = 0; i < RX_FREE_BUFFERS + RX_QUEUE_SIZE; i++) { + /* In the reset function, these buffers may have been allocated + * to an SKB, so we need to unmap and free potential storage */ + if (rxq->pool[i].page != NULL) { + pci_unmap_page(priv->pci_dev, rxq->pool[i].page_dma, + PAGE_SIZE << priv->hw_params.rx_page_order, + PCI_DMA_FROMDEVICE); + __iwl_free_pages(priv, rxq->pool[i].page); + rxq->pool[i].page = NULL; + } + list_add_tail(&rxq->pool[i].list, &rxq->rx_used); + } + + for (i = 0; i < RX_QUEUE_SIZE; i++) + rxq->queue[i] = NULL; + + /* Set us so that we have processed and used all buffers, but have + * not restocked the Rx queue with fresh buffers */ + rxq->read = rxq->write = 0; + rxq->write_actual = 0; + rxq->free_count = 0; + spin_unlock_irqrestore(&rxq->lock, flags); +} + +int iwlagn_rx_init(struct iwl_priv *priv, struct iwl_rx_queue *rxq) +{ + u32 rb_size; + const u32 rfdnlog = RX_QUEUE_SIZE_LOG; /* 256 RBDs */ + u32 rb_timeout = 0; /* FIXME: RX_RB_TIMEOUT for all devices? */ + + if (!priv->cfg->use_isr_legacy) + rb_timeout = RX_RB_TIMEOUT; + + if (priv->cfg->mod_params->amsdu_size_8K) + rb_size = FH_RCSR_RX_CONFIG_REG_VAL_RB_SIZE_8K; + else + rb_size = FH_RCSR_RX_CONFIG_REG_VAL_RB_SIZE_4K; + + /* Stop Rx DMA */ + iwl_write_direct32(priv, FH_MEM_RCSR_CHNL0_CONFIG_REG, 0); + + /* Reset driver's Rx queue write index */ + iwl_write_direct32(priv, FH_RSCSR_CHNL0_RBDCB_WPTR_REG, 0); + + /* Tell device where to find RBD circular buffer in DRAM */ + iwl_write_direct32(priv, FH_RSCSR_CHNL0_RBDCB_BASE_REG, + (u32)(rxq->dma_addr >> 8)); + + /* Tell device where in DRAM to update its Rx status */ + iwl_write_direct32(priv, FH_RSCSR_CHNL0_STTS_WPTR_REG, + rxq->rb_stts_dma >> 4); + + /* Enable Rx DMA + * FH_RCSR_CHNL0_RX_IGNORE_RXF_EMPTY is set because of HW bug in + * the credit mechanism in 5000 HW RX FIFO + * Direct rx interrupts to hosts + * Rx buffer size 4 or 8k + * RB timeout 0x10 + * 256 RBDs + */ + iwl_write_direct32(priv, FH_MEM_RCSR_CHNL0_CONFIG_REG, + FH_RCSR_RX_CONFIG_CHNL_EN_ENABLE_VAL | + FH_RCSR_CHNL0_RX_IGNORE_RXF_EMPTY | + FH_RCSR_CHNL0_RX_CONFIG_IRQ_DEST_INT_HOST_VAL | + FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK | + rb_size| + (rb_timeout << FH_RCSR_RX_CONFIG_REG_IRQ_RBTH_POS)| + (rfdnlog << FH_RCSR_RX_CONFIG_RBDCB_SIZE_POS)); + + /* Set interrupt coalescing timer to default (2048 usecs) */ + iwl_write8(priv, CSR_INT_COALESCING, IWL_HOST_INT_TIMEOUT_DEF); + + return 0; +} + +int iwlagn_hw_nic_init(struct iwl_priv *priv) +{ + unsigned long flags; + struct iwl_rx_queue *rxq = &priv->rxq; + int ret; + + /* nic_init */ + spin_lock_irqsave(&priv->lock, flags); + priv->cfg->ops->lib->apm_ops.init(priv); + + /* Set interrupt coalescing calibration timer to default (512 usecs) */ + iwl_write8(priv, CSR_INT_COALESCING, IWL_HOST_INT_CALIB_TIMEOUT_DEF); + + spin_unlock_irqrestore(&priv->lock, flags); + + ret = priv->cfg->ops->lib->apm_ops.set_pwr_src(priv, IWL_PWR_SRC_VMAIN); + + priv->cfg->ops->lib->apm_ops.config(priv); + + /* Allocate the RX queue, or reset if it is already allocated */ + if (!rxq->bd) { + ret = iwl_rx_queue_alloc(priv); + if (ret) { + IWL_ERR(priv, "Unable to initialize Rx queue\n"); + return -ENOMEM; + } + } else + iwlagn_rx_queue_reset(priv, rxq); + + iwlagn_rx_replenish(priv); + + iwlagn_rx_init(priv, rxq); + + spin_lock_irqsave(&priv->lock, flags); + + rxq->need_update = 1; + iwl_rx_queue_update_write_ptr(priv, rxq); + + spin_unlock_irqrestore(&priv->lock, flags); + + /* Allocate or reset and init all Tx and Command queues */ + if (!priv->txq) { + ret = iwlagn_txq_ctx_alloc(priv); + if (ret) + return ret; + } else + iwlagn_txq_ctx_reset(priv); + + set_bit(STATUS_INIT, &priv->status); + + return 0; +} + +/** + * iwlagn_dma_addr2rbd_ptr - convert a DMA address to a uCode read buffer ptr + */ +static inline __le32 iwlagn_dma_addr2rbd_ptr(struct iwl_priv *priv, + dma_addr_t dma_addr) +{ + return cpu_to_le32((u32)(dma_addr >> 8)); +} + +/** + * iwlagn_rx_queue_restock - refill RX queue from pre-allocated pool + * + * If there are slots in the RX queue that need to be restocked, + * and we have free pre-allocated buffers, fill the ranks as much + * as we can, pulling from rx_free. + * + * This moves the 'write' index forward to catch up with 'processed', and + * also updates the memory address in the firmware to reference the new + * target buffer. + */ +void iwlagn_rx_queue_restock(struct iwl_priv *priv) +{ + struct iwl_rx_queue *rxq = &priv->rxq; + struct list_head *element; + struct iwl_rx_mem_buffer *rxb; + unsigned long flags; + + spin_lock_irqsave(&rxq->lock, flags); + while ((iwl_rx_queue_space(rxq) > 0) && (rxq->free_count)) { + /* The overwritten rxb must be a used one */ + rxb = rxq->queue[rxq->write]; + BUG_ON(rxb && rxb->page); + + /* Get next free Rx buffer, remove from free list */ + element = rxq->rx_free.next; + rxb = list_entry(element, struct iwl_rx_mem_buffer, list); + list_del(element); + + /* Point to Rx buffer via next RBD in circular buffer */ + rxq->bd[rxq->write] = iwlagn_dma_addr2rbd_ptr(priv, + rxb->page_dma); + rxq->queue[rxq->write] = rxb; + rxq->write = (rxq->write + 1) & RX_QUEUE_MASK; + rxq->free_count--; + } + spin_unlock_irqrestore(&rxq->lock, flags); + /* If the pre-allocated buffer pool is dropping low, schedule to + * refill it */ + if (rxq->free_count <= RX_LOW_WATERMARK) + queue_work(priv->workqueue, &priv->rx_replenish); + + + /* If we've added more space for the firmware to place data, tell it. + * Increment device's write pointer in multiples of 8. */ + if (rxq->write_actual != (rxq->write & ~0x7)) { + spin_lock_irqsave(&rxq->lock, flags); + rxq->need_update = 1; + spin_unlock_irqrestore(&rxq->lock, flags); + iwl_rx_queue_update_write_ptr(priv, rxq); + } +} + +/** + * iwlagn_rx_replenish - Move all used packet from rx_used to rx_free + * + * When moving to rx_free an SKB is allocated for the slot. + * + * Also restock the Rx queue via iwl_rx_queue_restock. + * This is called as a scheduled work item (except for during initialization) + */ +void iwlagn_rx_allocate(struct iwl_priv *priv, gfp_t priority) +{ + struct iwl_rx_queue *rxq = &priv->rxq; + struct list_head *element; + struct iwl_rx_mem_buffer *rxb; + struct page *page; + unsigned long flags; + gfp_t gfp_mask = priority; + + while (1) { + spin_lock_irqsave(&rxq->lock, flags); + if (list_empty(&rxq->rx_used)) { + spin_unlock_irqrestore(&rxq->lock, flags); + return; + } + spin_unlock_irqrestore(&rxq->lock, flags); + + if (rxq->free_count > RX_LOW_WATERMARK) + gfp_mask |= __GFP_NOWARN; + + if (priv->hw_params.rx_page_order > 0) + gfp_mask |= __GFP_COMP; + + /* Alloc a new receive buffer */ + page = alloc_pages(gfp_mask, priv->hw_params.rx_page_order); + if (!page) { + if (net_ratelimit()) + IWL_DEBUG_INFO(priv, "alloc_pages failed, " + "order: %d\n", + priv->hw_params.rx_page_order); + + if ((rxq->free_count <= RX_LOW_WATERMARK) && + net_ratelimit()) + IWL_CRIT(priv, "Failed to alloc_pages with %s. Only %u free buffers remaining.\n", + priority == GFP_ATOMIC ? "GFP_ATOMIC" : "GFP_KERNEL", + rxq->free_count); + /* We don't reschedule replenish work here -- we will + * call the restock method and if it still needs + * more buffers it will schedule replenish */ + return; + } + + spin_lock_irqsave(&rxq->lock, flags); + + if (list_empty(&rxq->rx_used)) { + spin_unlock_irqrestore(&rxq->lock, flags); + __free_pages(page, priv->hw_params.rx_page_order); + return; + } + element = rxq->rx_used.next; + rxb = list_entry(element, struct iwl_rx_mem_buffer, list); + list_del(element); + + spin_unlock_irqrestore(&rxq->lock, flags); + + BUG_ON(rxb->page); + rxb->page = page; + /* Get physical address of the RB */ + rxb->page_dma = pci_map_page(priv->pci_dev, page, 0, + PAGE_SIZE << priv->hw_params.rx_page_order, + PCI_DMA_FROMDEVICE); + /* dma address must be no more than 36 bits */ + BUG_ON(rxb->page_dma & ~DMA_BIT_MASK(36)); + /* and also 256 byte aligned! */ + BUG_ON(rxb->page_dma & DMA_BIT_MASK(8)); + + spin_lock_irqsave(&rxq->lock, flags); + + list_add_tail(&rxb->list, &rxq->rx_free); + rxq->free_count++; + priv->alloc_rxb_page++; + + spin_unlock_irqrestore(&rxq->lock, flags); + } +} + +void iwlagn_rx_replenish(struct iwl_priv *priv) +{ + unsigned long flags; + + iwlagn_rx_allocate(priv, GFP_KERNEL); + + spin_lock_irqsave(&priv->lock, flags); + iwlagn_rx_queue_restock(priv); + spin_unlock_irqrestore(&priv->lock, flags); +} + +void iwlagn_rx_replenish_now(struct iwl_priv *priv) +{ + iwlagn_rx_allocate(priv, GFP_ATOMIC); + + iwlagn_rx_queue_restock(priv); +} + +/* Assumes that the skb field of the buffers in 'pool' is kept accurate. + * If an SKB has been detached, the POOL needs to have its SKB set to NULL + * This free routine walks the list of POOL entries and if SKB is set to + * non NULL it is unmapped and freed + */ +void iwlagn_rx_queue_free(struct iwl_priv *priv, struct iwl_rx_queue *rxq) +{ + int i; + for (i = 0; i < RX_QUEUE_SIZE + RX_FREE_BUFFERS; i++) { + if (rxq->pool[i].page != NULL) { + pci_unmap_page(priv->pci_dev, rxq->pool[i].page_dma, + PAGE_SIZE << priv->hw_params.rx_page_order, + PCI_DMA_FROMDEVICE); + __iwl_free_pages(priv, rxq->pool[i].page); + rxq->pool[i].page = NULL; + } + } + + dma_free_coherent(&priv->pci_dev->dev, 4 * RX_QUEUE_SIZE, rxq->bd, + rxq->dma_addr); + dma_free_coherent(&priv->pci_dev->dev, sizeof(struct iwl_rb_status), + rxq->rb_stts, rxq->rb_stts_dma); + rxq->bd = NULL; + rxq->rb_stts = NULL; +} + +int iwlagn_rxq_stop(struct iwl_priv *priv) +{ + + /* stop Rx DMA */ + iwl_write_direct32(priv, FH_MEM_RCSR_CHNL0_CONFIG_REG, 0); + iwl_poll_direct_bit(priv, FH_MEM_RSSR_RX_STATUS_REG, + FH_RSSR_CHNL0_RX_STATUS_CHNL_IDLE, 1000); + + return 0; +} + +int iwlagn_hwrate_to_mac80211_idx(u32 rate_n_flags, enum ieee80211_band band) +{ + int idx = 0; + int band_offset = 0; + + /* HT rate format: mac80211 wants an MCS number, which is just LSB */ + if (rate_n_flags & RATE_MCS_HT_MSK) { + idx = (rate_n_flags & 0xff); + return idx; + /* Legacy rate format, search for match in table */ + } else { + if (band == IEEE80211_BAND_5GHZ) + band_offset = IWL_FIRST_OFDM_RATE; + for (idx = band_offset; idx < IWL_RATE_COUNT_LEGACY; idx++) + if (iwl_rates[idx].plcp == (rate_n_flags & 0xFF)) + return idx - band_offset; + } + + return -1; +} + +/* Calc max signal level (dBm) among 3 possible receivers */ +static inline int iwlagn_calc_rssi(struct iwl_priv *priv, + struct iwl_rx_phy_res *rx_resp) +{ + return priv->cfg->ops->utils->calc_rssi(priv, rx_resp); +} + +#ifdef CONFIG_IWLWIFI_DEBUG +/** + * iwlagn_dbg_report_frame - dump frame to syslog during debug sessions + * + * You may hack this function to show different aspects of received frames, + * including selective frame dumps. + * group100 parameter selects whether to show 1 out of 100 good data frames. + * All beacon and probe response frames are printed. + */ +static void iwlagn_dbg_report_frame(struct iwl_priv *priv, + struct iwl_rx_phy_res *phy_res, u16 length, + struct ieee80211_hdr *header, int group100) +{ + u32 to_us; + u32 print_summary = 0; + u32 print_dump = 0; /* set to 1 to dump all frames' contents */ + u32 hundred = 0; + u32 dataframe = 0; + __le16 fc; + u16 seq_ctl; + u16 channel; + u16 phy_flags; + u32 rate_n_flags; + u32 tsf_low; + int rssi; + + if (likely(!(iwl_get_debug_level(priv) & IWL_DL_RX))) + return; + + /* MAC header */ + fc = header->frame_control; + seq_ctl = le16_to_cpu(header->seq_ctrl); + + /* metadata */ + channel = le16_to_cpu(phy_res->channel); + phy_flags = le16_to_cpu(phy_res->phy_flags); + rate_n_flags = le32_to_cpu(phy_res->rate_n_flags); + + /* signal statistics */ + rssi = iwlagn_calc_rssi(priv, phy_res); + tsf_low = le64_to_cpu(phy_res->timestamp) & 0x0ffffffff; + + to_us = !compare_ether_addr(header->addr1, priv->mac_addr); + + /* if data frame is to us and all is good, + * (optionally) print summary for only 1 out of every 100 */ + if (to_us && (fc & ~cpu_to_le16(IEEE80211_FCTL_PROTECTED)) == + cpu_to_le16(IEEE80211_FCTL_FROMDS | IEEE80211_FTYPE_DATA)) { + dataframe = 1; + if (!group100) + print_summary = 1; /* print each frame */ + else if (priv->framecnt_to_us < 100) { + priv->framecnt_to_us++; + print_summary = 0; + } else { + priv->framecnt_to_us = 0; + print_summary = 1; + hundred = 1; + } + } else { + /* print summary for all other frames */ + print_summary = 1; + } + + if (print_summary) { + char *title; + int rate_idx; + u32 bitrate; + + if (hundred) + title = "100Frames"; + else if (ieee80211_has_retry(fc)) + title = "Retry"; + else if (ieee80211_is_assoc_resp(fc)) + title = "AscRsp"; + else if (ieee80211_is_reassoc_resp(fc)) + title = "RasRsp"; + else if (ieee80211_is_probe_resp(fc)) { + title = "PrbRsp"; + print_dump = 1; /* dump frame contents */ + } else if (ieee80211_is_beacon(fc)) { + title = "Beacon"; + print_dump = 1; /* dump frame contents */ + } else if (ieee80211_is_atim(fc)) + title = "ATIM"; + else if (ieee80211_is_auth(fc)) + title = "Auth"; + else if (ieee80211_is_deauth(fc)) + title = "DeAuth"; + else if (ieee80211_is_disassoc(fc)) + title = "DisAssoc"; + else + title = "Frame"; + + rate_idx = iwl_hwrate_to_plcp_idx(rate_n_flags); + if (unlikely((rate_idx < 0) || (rate_idx >= IWL_RATE_COUNT))) { + bitrate = 0; + WARN_ON_ONCE(1); + } else { + bitrate = iwl_rates[rate_idx].ieee / 2; + } + + /* print frame summary. + * MAC addresses show just the last byte (for brevity), + * but you can hack it to show more, if you'd like to. */ + if (dataframe) + IWL_DEBUG_RX(priv, "%s: mhd=0x%04x, dst=0x%02x, " + "len=%u, rssi=%d, chnl=%d, rate=%u,\n", + title, le16_to_cpu(fc), header->addr1[5], + length, rssi, channel, bitrate); + else { + /* src/dst addresses assume managed mode */ + IWL_DEBUG_RX(priv, "%s: 0x%04x, dst=0x%02x, src=0x%02x, " + "len=%u, rssi=%d, tim=%lu usec, " + "phy=0x%02x, chnl=%d\n", + title, le16_to_cpu(fc), header->addr1[5], + header->addr3[5], length, rssi, + tsf_low - priv->scan_start_tsf, + phy_flags, channel); + } + } + if (print_dump) + iwl_print_hex_dump(priv, IWL_DL_RX, header, length); +} +#endif + +static u32 iwlagn_translate_rx_status(struct iwl_priv *priv, u32 decrypt_in) +{ + u32 decrypt_out = 0; + + if ((decrypt_in & RX_RES_STATUS_STATION_FOUND) == + RX_RES_STATUS_STATION_FOUND) + decrypt_out |= (RX_RES_STATUS_STATION_FOUND | + RX_RES_STATUS_NO_STATION_INFO_MISMATCH); + + decrypt_out |= (decrypt_in & RX_RES_STATUS_SEC_TYPE_MSK); + + /* packet was not encrypted */ + if ((decrypt_in & RX_RES_STATUS_SEC_TYPE_MSK) == + RX_RES_STATUS_SEC_TYPE_NONE) + return decrypt_out; + + /* packet was encrypted with unknown alg */ + if ((decrypt_in & RX_RES_STATUS_SEC_TYPE_MSK) == + RX_RES_STATUS_SEC_TYPE_ERR) + return decrypt_out; + + /* decryption was not done in HW */ + if ((decrypt_in & RX_MPDU_RES_STATUS_DEC_DONE_MSK) != + RX_MPDU_RES_STATUS_DEC_DONE_MSK) + return decrypt_out; + + switch (decrypt_in & RX_RES_STATUS_SEC_TYPE_MSK) { + + case RX_RES_STATUS_SEC_TYPE_CCMP: + /* alg is CCM: check MIC only */ + if (!(decrypt_in & RX_MPDU_RES_STATUS_MIC_OK)) + /* Bad MIC */ + decrypt_out |= RX_RES_STATUS_BAD_ICV_MIC; + else + decrypt_out |= RX_RES_STATUS_DECRYPT_OK; + + break; + + case RX_RES_STATUS_SEC_TYPE_TKIP: + if (!(decrypt_in & RX_MPDU_RES_STATUS_TTAK_OK)) { + /* Bad TTAK */ + decrypt_out |= RX_RES_STATUS_BAD_KEY_TTAK; + break; + } + /* fall through if TTAK OK */ + default: + if (!(decrypt_in & RX_MPDU_RES_STATUS_ICV_OK)) + decrypt_out |= RX_RES_STATUS_BAD_ICV_MIC; + else + decrypt_out |= RX_RES_STATUS_DECRYPT_OK; + break; + }; + + IWL_DEBUG_RX(priv, "decrypt_in:0x%x decrypt_out = 0x%x\n", + decrypt_in, decrypt_out); + + return decrypt_out; +} + +static void iwlagn_pass_packet_to_mac80211(struct iwl_priv *priv, + struct ieee80211_hdr *hdr, + u16 len, + u32 ampdu_status, + struct iwl_rx_mem_buffer *rxb, + struct ieee80211_rx_status *stats) +{ + struct sk_buff *skb; + __le16 fc = hdr->frame_control; + + /* We only process data packets if the interface is open */ + if (unlikely(!priv->is_open)) { + IWL_DEBUG_DROP_LIMIT(priv, + "Dropping packet while interface is not open.\n"); + return; + } + + /* In case of HW accelerated crypto and bad decryption, drop */ + if (!priv->cfg->mod_params->sw_crypto && + iwl_set_decrypted_flag(priv, hdr, ampdu_status, stats)) + return; + + skb = dev_alloc_skb(128); + if (!skb) { + IWL_ERR(priv, "dev_alloc_skb failed\n"); + return; + } + + skb_add_rx_frag(skb, 0, rxb->page, (void *)hdr - rxb_addr(rxb), len); + + iwl_update_stats(priv, false, fc, len); + memcpy(IEEE80211_SKB_RXCB(skb), stats, sizeof(*stats)); + + ieee80211_rx(priv->hw, skb); + priv->alloc_rxb_page--; + rxb->page = NULL; +} + +/* Called for REPLY_RX (legacy ABG frames), or + * REPLY_RX_MPDU_CMD (HT high-throughput N frames). */ +void iwlagn_rx_reply_rx(struct iwl_priv *priv, + struct iwl_rx_mem_buffer *rxb) +{ + struct ieee80211_hdr *header; + struct ieee80211_rx_status rx_status; + struct iwl_rx_packet *pkt = rxb_addr(rxb); + struct iwl_rx_phy_res *phy_res; + __le32 rx_pkt_status; + struct iwl4965_rx_mpdu_res_start *amsdu; + u32 len; + u32 ampdu_status; + u32 rate_n_flags; + + /** + * REPLY_RX and REPLY_RX_MPDU_CMD are handled differently. + * REPLY_RX: physical layer info is in this buffer + * REPLY_RX_MPDU_CMD: physical layer info was sent in separate + * command and cached in priv->last_phy_res + * + * Here we set up local variables depending on which command is + * received. + */ + if (pkt->hdr.cmd == REPLY_RX) { + phy_res = (struct iwl_rx_phy_res *)pkt->u.raw; + header = (struct ieee80211_hdr *)(pkt->u.raw + sizeof(*phy_res) + + phy_res->cfg_phy_cnt); + + len = le16_to_cpu(phy_res->byte_count); + rx_pkt_status = *(__le32 *)(pkt->u.raw + sizeof(*phy_res) + + phy_res->cfg_phy_cnt + len); + ampdu_status = le32_to_cpu(rx_pkt_status); + } else { + if (!priv->_agn.last_phy_res_valid) { + IWL_ERR(priv, "MPDU frame without cached PHY data\n"); + return; + } + phy_res = &priv->_agn.last_phy_res; + amsdu = (struct iwl4965_rx_mpdu_res_start *)pkt->u.raw; + header = (struct ieee80211_hdr *)(pkt->u.raw + sizeof(*amsdu)); + len = le16_to_cpu(amsdu->byte_count); + rx_pkt_status = *(__le32 *)(pkt->u.raw + sizeof(*amsdu) + len); + ampdu_status = iwlagn_translate_rx_status(priv, + le32_to_cpu(rx_pkt_status)); + } + + if ((unlikely(phy_res->cfg_phy_cnt > 20))) { + IWL_DEBUG_DROP(priv, "dsp size out of range [0,20]: %d/n", + phy_res->cfg_phy_cnt); + return; + } + + if (!(rx_pkt_status & RX_RES_STATUS_NO_CRC32_ERROR) || + !(rx_pkt_status & RX_RES_STATUS_NO_RXE_OVERFLOW)) { + IWL_DEBUG_RX(priv, "Bad CRC or FIFO: 0x%08X.\n", + le32_to_cpu(rx_pkt_status)); + return; + } + + /* This will be used in several places later */ + rate_n_flags = le32_to_cpu(phy_res->rate_n_flags); + + /* rx_status carries information about the packet to mac80211 */ + rx_status.mactime = le64_to_cpu(phy_res->timestamp); + rx_status.freq = + ieee80211_channel_to_frequency(le16_to_cpu(phy_res->channel)); + rx_status.band = (phy_res->phy_flags & RX_RES_PHY_FLAGS_BAND_24_MSK) ? + IEEE80211_BAND_2GHZ : IEEE80211_BAND_5GHZ; + rx_status.rate_idx = + iwlagn_hwrate_to_mac80211_idx(rate_n_flags, rx_status.band); + rx_status.flag = 0; + + /* TSF isn't reliable. In order to allow smooth user experience, + * this W/A doesn't propagate it to the mac80211 */ + /*rx_status.flag |= RX_FLAG_TSFT;*/ + + priv->ucode_beacon_time = le32_to_cpu(phy_res->beacon_time_stamp); + + /* Find max signal strength (dBm) among 3 antenna/receiver chains */ + rx_status.signal = iwlagn_calc_rssi(priv, phy_res); + +#ifdef CONFIG_IWLWIFI_DEBUG + /* Set "1" to report good data frames in groups of 100 */ + if (unlikely(iwl_get_debug_level(priv) & IWL_DL_RX)) + iwlagn_dbg_report_frame(priv, phy_res, len, header, 1); +#endif + iwl_dbg_log_rx_data_frame(priv, len, header); + IWL_DEBUG_STATS_LIMIT(priv, "Rssi %d, TSF %llu\n", + rx_status.signal, (unsigned long long)rx_status.mactime); + + /* + * "antenna number" + * + * It seems that the antenna field in the phy flags value + * is actually a bit field. This is undefined by radiotap, + * it wants an actual antenna number but I always get "7" + * for most legacy frames I receive indicating that the + * same frame was received on all three RX chains. + * + * I think this field should be removed in favor of a + * new 802.11n radiotap field "RX chains" that is defined + * as a bitmask. + */ + rx_status.antenna = + (le16_to_cpu(phy_res->phy_flags) & RX_RES_PHY_FLAGS_ANTENNA_MSK) + >> RX_RES_PHY_FLAGS_ANTENNA_POS; + + /* set the preamble flag if appropriate */ + if (phy_res->phy_flags & RX_RES_PHY_FLAGS_SHORT_PREAMBLE_MSK) + rx_status.flag |= RX_FLAG_SHORTPRE; + + /* Set up the HT phy flags */ + if (rate_n_flags & RATE_MCS_HT_MSK) + rx_status.flag |= RX_FLAG_HT; + if (rate_n_flags & RATE_MCS_HT40_MSK) + rx_status.flag |= RX_FLAG_40MHZ; + if (rate_n_flags & RATE_MCS_SGI_MSK) + rx_status.flag |= RX_FLAG_SHORT_GI; + + iwlagn_pass_packet_to_mac80211(priv, header, len, ampdu_status, + rxb, &rx_status); +} + +/* Cache phy data (Rx signal strength, etc) for HT frame (REPLY_RX_PHY_CMD). + * This will be used later in iwl_rx_reply_rx() for REPLY_RX_MPDU_CMD. */ +void iwlagn_rx_reply_rx_phy(struct iwl_priv *priv, + struct iwl_rx_mem_buffer *rxb) +{ + struct iwl_rx_packet *pkt = rxb_addr(rxb); + priv->_agn.last_phy_res_valid = true; + memcpy(&priv->_agn.last_phy_res, pkt->u.raw, + sizeof(struct iwl_rx_phy_res)); +} |