blob: 7a18c28d5161d11680ce45ba33e6e5058521fb39 [file] [log] [blame]
/*
* Copyright (c) 2011-2017 The Linux Foundation. All rights reserved.
*
* Previously licensed under the ISC license by Qualcomm Atheros, Inc.
*
*
* Permission to use, copy, modify, and/or distribute this software for
* any purpose with or without fee is hereby granted, provided that the
* above copyright notice and this permission notice appear in all
* copies.
*
* THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL
* WARRANTIES WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED
* WARRANTIES OF MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE
* AUTHOR BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL
* DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
* PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
* TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
* PERFORMANCE OF THIS SOFTWARE.
*/
/*
* This file was originally distributed by Qualcomm Atheros, Inc.
* under proprietary terms before Copyright ownership was assigned
* to the Linux Foundation.
*/
/* OS abstraction libraries */
#include <qdf_nbuf.h> /* qdf_nbuf_t, etc. */
#include <qdf_atomic.h> /* qdf_atomic_read, etc. */
#include <qdf_util.h> /* qdf_unlikely */
/* APIs for other modules */
#include <htt.h> /* HTT_TX_EXT_TID_MGMT */
#include <ol_htt_tx_api.h> /* htt_tx_desc_tid */
/* internal header files relevant for all systems */
#include <ol_txrx_internal.h> /* TXRX_ASSERT1 */
#include <ol_tx_desc.h> /* ol_tx_desc */
#include <ol_tx_send.h> /* ol_tx_send */
#include <ol_txrx.h>
/* internal header files relevant only for HL systems */
#include <ol_tx_classify.h> /* ol_tx_classify, ol_tx_classify_mgmt */
#include <ol_tx_queue.h> /* ol_tx_enqueue */
#include <ol_tx_sched.h> /* ol_tx_sched */
/* internal header files relevant only for specific systems (Pronto) */
#include <ol_txrx_encap.h> /* OL_TX_ENCAP, etc */
#include <ol_tx.h>
#ifdef WLAN_FEATURE_FASTPATH
#include <hif.h> /* HIF_DEVICE */
#include <htc_api.h> /* Layering violation, but required for fast path */
#include <htt_internal.h>
#include <htt_types.h> /* htc_endpoint */
#include <cdp_txrx_peer_ops.h>
#include <cdp_txrx_ipa.h>
#include <cdp_txrx_handle.h>
int ce_send_fast(struct CE_handle *copyeng, qdf_nbuf_t msdu,
unsigned int transfer_id, uint32_t download_len);
#endif /* WLAN_FEATURE_FASTPATH */
/*
* The TXRX module doesn't accept tx frames unless the target has
* enough descriptors for them.
* For LL, the TXRX descriptor pool is sized to match the target's
* descriptor pool. Hence, if the descriptor allocation in TXRX
* succeeds, that guarantees that the target has room to accept
* the new tx frame.
*/
static struct ol_tx_desc_t *
ol_tx_prepare_ll(ol_txrx_vdev_handle vdev,
qdf_nbuf_t msdu,
struct ol_txrx_msdu_info_t *msdu_info)
{
struct ol_tx_desc_t *tx_desc;
struct ol_txrx_pdev_t *pdev = vdev->pdev;
(msdu_info)->htt.info.frame_type = pdev->htt_pkt_type;
tx_desc = ol_tx_desc_ll(pdev, vdev, msdu, msdu_info);
if (qdf_unlikely(!tx_desc)) {
/*
* If TSO packet, free associated
* remaining TSO segment descriptors
*/
if (qdf_nbuf_is_tso(msdu))
ol_free_remaining_tso_segs(
vdev, msdu_info, true);
TXRX_STATS_MSDU_LIST_INCR(
pdev, tx.dropped.host_reject, msdu);
return NULL;
}
return tx_desc;
}
#if defined(FEATURE_TSO)
void ol_free_remaining_tso_segs(ol_txrx_vdev_handle vdev,
struct ol_txrx_msdu_info_t *msdu_info,
bool is_tso_seg_mapping_done)
{
struct qdf_tso_seg_elem_t *next_seg;
struct qdf_tso_seg_elem_t *free_seg = msdu_info->tso_info.curr_seg;
struct ol_txrx_pdev_t *pdev;
bool is_last_seg = false;
if (qdf_unlikely(!vdev)) {
QDF_TRACE(QDF_MODULE_ID_TXRX, QDF_TRACE_LEVEL_ERROR,
"%s:vdev is null", __func__);
return;
} else {
pdev = vdev->pdev;
if (qdf_unlikely(!pdev)) {
QDF_TRACE(QDF_MODULE_ID_TXRX, QDF_TRACE_LEVEL_ERROR,
"%s:pdev is null", __func__);
return;
}
}
if (is_tso_seg_mapping_done) {
/*
* TSO segment are mapped already, therefore,
* 1. unmap the tso segments,
* 2. free tso num segment if it is a last segment, and
* 3. free the tso segments.
*/
struct qdf_tso_num_seg_elem_t *tso_num_desc =
msdu_info->tso_info.tso_num_seg_list;
if (qdf_unlikely(tso_num_desc == NULL)) {
QDF_TRACE(QDF_MODULE_ID_TXRX, QDF_TRACE_LEVEL_ERROR,
"%s %d TSO common info is NULL!",
__func__, __LINE__);
return;
}
while (free_seg) {
qdf_spin_lock_bh(&pdev->tso_seg_pool.tso_mutex);
tso_num_desc->num_seg.tso_cmn_num_seg--;
is_last_seg = (tso_num_desc->num_seg.tso_cmn_num_seg ==
0) ? true : false;
qdf_nbuf_unmap_tso_segment(pdev->osdev, free_seg,
is_last_seg);
qdf_spin_unlock_bh(&pdev->tso_seg_pool.tso_mutex);
if (is_last_seg) {
ol_tso_num_seg_free(pdev,
msdu_info->tso_info.tso_num_seg_list);
msdu_info->tso_info.tso_num_seg_list = NULL;
}
next_seg = free_seg->next;
ol_tso_free_segment(pdev, free_seg);
free_seg = next_seg;
}
} else {
/*
* TSO segment are not mapped therefore,
* free the tso segments only.
*/
while (free_seg) {
next_seg = free_seg->next;
ol_tso_free_segment(pdev, free_seg);
free_seg = next_seg;
}
}
}
/**
* ol_tx_prepare_tso() - Given a jumbo msdu, prepare the TSO
* related information in the msdu_info meta data
* @vdev: virtual device handle
* @msdu: network buffer
* @msdu_info: meta data associated with the msdu
*
* Return: 0 - success, >0 - error
*/
static inline uint8_t ol_tx_prepare_tso(ol_txrx_vdev_handle vdev,
qdf_nbuf_t msdu, struct ol_txrx_msdu_info_t *msdu_info)
{
msdu_info->tso_info.curr_seg = NULL;
if (qdf_nbuf_is_tso(msdu)) {
int num_seg = qdf_nbuf_get_tso_num_seg(msdu);
struct qdf_tso_num_seg_elem_t *tso_num_seg;
msdu_info->tso_info.tso_num_seg_list = NULL;
msdu_info->tso_info.tso_seg_list = NULL;
msdu_info->tso_info.num_segs = num_seg;
while (num_seg) {
struct qdf_tso_seg_elem_t *tso_seg =
ol_tso_alloc_segment(vdev->pdev);
if (tso_seg) {
qdf_tso_seg_dbg_record(tso_seg,
TSOSEG_LOC_PREPARETSO);
tso_seg->next =
msdu_info->tso_info.tso_seg_list;
msdu_info->tso_info.tso_seg_list
= tso_seg;
num_seg--;
} else {
/* Free above alocated TSO segements till now */
msdu_info->tso_info.curr_seg =
msdu_info->tso_info.tso_seg_list;
ol_free_remaining_tso_segs(vdev, msdu_info,
false);
return 1;
}
}
tso_num_seg = ol_tso_num_seg_alloc(vdev->pdev);
if (tso_num_seg) {
tso_num_seg->next = msdu_info->tso_info.
tso_num_seg_list;
msdu_info->tso_info.tso_num_seg_list = tso_num_seg;
} else {
/* Free the already allocated num of segments */
msdu_info->tso_info.curr_seg =
msdu_info->tso_info.tso_seg_list;
ol_free_remaining_tso_segs(vdev, msdu_info, false);
return 1;
}
if (qdf_unlikely(!qdf_nbuf_get_tso_info(vdev->pdev->osdev,
msdu, &(msdu_info->tso_info)))) {
/* Free the already allocated num of segments */
msdu_info->tso_info.curr_seg =
msdu_info->tso_info.tso_seg_list;
ol_free_remaining_tso_segs(vdev, msdu_info, false);
return 1;
}
msdu_info->tso_info.curr_seg =
msdu_info->tso_info.tso_seg_list;
num_seg = msdu_info->tso_info.num_segs;
} else {
msdu_info->tso_info.is_tso = 0;
msdu_info->tso_info.num_segs = 1;
}
return 0;
}
#endif
/**
* ol_tx_data() - send data frame
* @vdev: virtual device handle
* @skb: skb
*
* Return: skb/NULL for success
*/
qdf_nbuf_t ol_tx_data(void *data_vdev, qdf_nbuf_t skb)
{
struct ol_txrx_pdev_t *pdev;
qdf_nbuf_t ret;
ol_txrx_vdev_handle vdev = data_vdev;
if (qdf_unlikely(!vdev)) {
QDF_TRACE(QDF_MODULE_ID_TXRX, QDF_TRACE_LEVEL_DEBUG,
"%s:vdev is null", __func__);
return skb;
}
pdev = vdev->pdev;
if (qdf_unlikely(!pdev)) {
QDF_TRACE(QDF_MODULE_ID_TXRX, QDF_TRACE_LEVEL_DEBUG,
"%s:pdev is null", __func__);
return skb;
}
if ((ol_cfg_is_ip_tcp_udp_checksum_offload_enabled(pdev->ctrl_pdev))
&& (qdf_nbuf_get_protocol(skb) == htons(ETH_P_IP))
&& (qdf_nbuf_get_ip_summed(skb) == CHECKSUM_PARTIAL))
qdf_nbuf_set_ip_summed(skb, CHECKSUM_COMPLETE);
/* Terminate the (single-element) list of tx frames */
qdf_nbuf_set_next(skb, NULL);
ret = OL_TX_SEND(vdev, skb);
if (ret) {
ol_txrx_dbg("%s: Failed to tx", __func__);
return ret;
}
return NULL;
}
#ifdef IPA_OFFLOAD
qdf_nbuf_t ol_tx_send_ipa_data_frame(struct cdp_vdev *vdev, qdf_nbuf_t skb)
{
struct ol_txrx_pdev_t *pdev = cds_get_context(QDF_MODULE_ID_TXRX);
qdf_nbuf_t ret;
if (qdf_unlikely(!pdev)) {
ol_txrx_err("%s: pdev is NULL", __func__);
return skb;
}
if ((ol_cfg_is_ip_tcp_udp_checksum_offload_enabled(pdev->ctrl_pdev))
&& (qdf_nbuf_get_protocol(skb) == htons(ETH_P_IP))
&& (qdf_nbuf_get_ip_summed(skb) == CHECKSUM_PARTIAL))
qdf_nbuf_set_ip_summed(skb, CHECKSUM_COMPLETE);
/* Terminate the (single-element) list of tx frames */
qdf_nbuf_set_next(skb, NULL);
ret = OL_TX_SEND((struct ol_txrx_vdev_t *)vdev, skb);
if (ret) {
ol_txrx_dbg("%s: Failed to tx", __func__);
return ret;
}
return NULL;
}
#endif
#if defined(FEATURE_TSO)
/**
* ol_tx_tso_update_stats() - update TSO stats
* @pdev: pointer to ol_txrx_pdev_t structure
* @msdu_info: tso msdu_info for the msdu
* @msdu: tso mdsu for which stats are updated
* @tso_msdu_idx: stats index in the global TSO stats array where stats will be
* updated
*
* Return: None
*/
static inline void ol_tx_tso_update_stats(struct ol_txrx_pdev_t *pdev,
struct qdf_tso_info_t *tso_info,
qdf_nbuf_t msdu,
uint32_t tso_msdu_idx)
{
TXRX_STATS_TSO_HISTOGRAM(pdev, tso_info->num_segs);
TXRX_STATS_TSO_GSO_SIZE_UPDATE(pdev, tso_msdu_idx,
qdf_nbuf_tcp_tso_size(msdu));
TXRX_STATS_TSO_TOTAL_LEN_UPDATE(pdev,
tso_msdu_idx, qdf_nbuf_len(msdu));
TXRX_STATS_TSO_NUM_FRAGS_UPDATE(pdev, tso_msdu_idx,
qdf_nbuf_get_nr_frags(msdu));
}
/**
* ol_tx_tso_get_stats_idx() - retrieve global TSO stats index and increment it
* @pdev: pointer to ol_txrx_pdev_t structure
*
* Retrieve the current value of the global variable and increment it. This is
* done in a spinlock as the global TSO stats may be accessed in parallel by
* multiple TX streams.
*
* Return: The current value of TSO stats index.
*/
static uint32_t ol_tx_tso_get_stats_idx(struct ol_txrx_pdev_t *pdev)
{
uint32_t msdu_stats_idx = 0;
qdf_spin_lock_bh(&pdev->stats.pub.tx.tso.tso_stats_lock);
msdu_stats_idx = pdev->stats.pub.tx.tso.tso_info.tso_msdu_idx;
pdev->stats.pub.tx.tso.tso_info.tso_msdu_idx++;
pdev->stats.pub.tx.tso.tso_info.tso_msdu_idx &=
NUM_MAX_TSO_MSDUS_MASK;
qdf_spin_unlock_bh(&pdev->stats.pub.tx.tso.tso_stats_lock);
TXRX_STATS_TSO_RESET_MSDU(pdev, msdu_stats_idx);
return msdu_stats_idx;
}
#endif
#if defined(FEATURE_TSO)
qdf_nbuf_t ol_tx_ll(ol_txrx_vdev_handle vdev, qdf_nbuf_t msdu_list)
{
qdf_nbuf_t msdu = msdu_list;
struct ol_txrx_msdu_info_t msdu_info;
uint32_t tso_msdu_stats_idx = 0;
msdu_info.htt.info.l2_hdr_type = vdev->pdev->htt_pkt_type;
msdu_info.htt.action.tx_comp_req = 0;
/*
* The msdu_list variable could be used instead of the msdu var,
* but just to clarify which operations are done on a single MSDU
* vs. a list of MSDUs, use a distinct variable for single MSDUs
* within the list.
*/
while (msdu) {
qdf_nbuf_t next;
struct ol_tx_desc_t *tx_desc = NULL;
int segments = 1;
msdu_info.htt.info.ext_tid = qdf_nbuf_get_tid(msdu);
msdu_info.peer = NULL;
if (qdf_unlikely(ol_tx_prepare_tso(vdev, msdu, &msdu_info))) {
qdf_print("ol_tx_prepare_tso failed\n");
TXRX_STATS_MSDU_LIST_INCR(vdev->pdev,
tx.dropped.host_reject, msdu);
return msdu;
}
segments = msdu_info.tso_info.num_segs;
if (msdu_info.tso_info.is_tso) {
tso_msdu_stats_idx =
ol_tx_tso_get_stats_idx(vdev->pdev);
msdu_info.tso_info.msdu_stats_idx = tso_msdu_stats_idx;
ol_tx_tso_update_stats(vdev->pdev,
&(msdu_info.tso_info),
msdu, tso_msdu_stats_idx);
}
/*
* The netbuf may get linked into a different list inside the
* ol_tx_send function, so store the next pointer before the
* tx_send call.
*/
next = qdf_nbuf_next(msdu);
/* init the current segment to the 1st segment in the list */
while (segments) {
if (msdu_info.tso_info.curr_seg)
QDF_NBUF_CB_PADDR(msdu) =
msdu_info.tso_info.curr_seg->
seg.tso_frags[0].paddr;
segments--;
tx_desc = ol_tx_prepare_ll(vdev, msdu, &msdu_info);
if (!tx_desc)
return msdu;
/*
* If this is a jumbo nbuf, then increment the number
* of nbuf users for each additional segment of the msdu
* This will ensure that the skb is freed only after
* receiving tx completion for all segments of an nbuf.
*/
if (segments)
qdf_nbuf_inc_users(msdu);
TXRX_STATS_MSDU_INCR(vdev->pdev, tx.from_stack, msdu);
/*
* If debug display is enabled, show the meta-data being
* downloaded to the target via the HTT tx descriptor.
*/
htt_tx_desc_display(tx_desc->htt_tx_desc);
ol_tx_send(vdev->pdev, tx_desc, msdu, vdev->vdev_id);
if (msdu_info.tso_info.curr_seg) {
msdu_info.tso_info.curr_seg =
msdu_info.tso_info.curr_seg->next;
}
qdf_nbuf_reset_num_frags(msdu);
if (msdu_info.tso_info.is_tso) {
TXRX_STATS_TSO_INC_SEG(vdev->pdev,
tso_msdu_stats_idx);
TXRX_STATS_TSO_INC_SEG_IDX(vdev->pdev,
tso_msdu_stats_idx);
}
} /* while segments */
msdu = next;
} /* while msdus */
return NULL; /* all MSDUs were accepted */
}
#else /* TSO */
qdf_nbuf_t ol_tx_ll(ol_txrx_vdev_handle vdev, qdf_nbuf_t msdu_list)
{
qdf_nbuf_t msdu = msdu_list;
struct ol_txrx_msdu_info_t msdu_info;
msdu_info.htt.info.l2_hdr_type = vdev->pdev->htt_pkt_type;
msdu_info.htt.action.tx_comp_req = 0;
msdu_info.tso_info.is_tso = 0;
/*
* The msdu_list variable could be used instead of the msdu var,
* but just to clarify which operations are done on a single MSDU
* vs. a list of MSDUs, use a distinct variable for single MSDUs
* within the list.
*/
while (msdu) {
qdf_nbuf_t next;
struct ol_tx_desc_t *tx_desc = NULL;
msdu_info.htt.info.ext_tid = qdf_nbuf_get_tid(msdu);
msdu_info.peer = NULL;
tx_desc = ol_tx_prepare_ll(vdev, msdu, &msdu_info);
if (!tx_desc)
return msdu;
TXRX_STATS_MSDU_INCR(vdev->pdev, tx.from_stack, msdu);
/*
* If debug display is enabled, show the meta-data being
* downloaded to the target via the HTT tx descriptor.
*/
htt_tx_desc_display(tx_desc->htt_tx_desc);
/*
* The netbuf may get linked into a different list inside the
* ol_tx_send function, so store the next pointer before the
* tx_send call.
*/
next = qdf_nbuf_next(msdu);
ol_tx_send(vdev->pdev, tx_desc, msdu, vdev->vdev_id);
msdu = next;
}
return NULL; /* all MSDUs were accepted */
}
#endif /* TSO */
#ifdef WLAN_FEATURE_FASTPATH
/**
* ol_tx_prepare_ll_fast() Alloc and prepare Tx descriptor
*
* Allocate and prepare Tx descriptor with msdu and fragment descritor
* inforamtion.
*
* @pdev: pointer to ol pdev handle
* @vdev: pointer to ol vdev handle
* @msdu: linked list of msdu packets
* @pkt_download_len: packet download length
* @ep_id: endpoint ID
* @msdu_info: Handle to msdu_info
*
* Return: Pointer to Tx descriptor
*/
static inline struct ol_tx_desc_t *
ol_tx_prepare_ll_fast(struct ol_txrx_pdev_t *pdev,
ol_txrx_vdev_handle vdev, qdf_nbuf_t msdu,
uint32_t pkt_download_len, uint32_t ep_id,
struct ol_txrx_msdu_info_t *msdu_info)
{
struct ol_tx_desc_t *tx_desc = NULL;
uint32_t *htt_tx_desc;
void *htc_hdr_vaddr;
u_int32_t num_frags, i;
enum extension_header_type type;
tx_desc = ol_tx_desc_alloc_wrapper(pdev, vdev, msdu_info);
if (qdf_unlikely(!tx_desc))
return NULL;
tx_desc->netbuf = msdu;
if (msdu_info->tso_info.is_tso) {
tx_desc->tso_desc = msdu_info->tso_info.curr_seg;
qdf_tso_seg_dbg_setowner(tx_desc->tso_desc, tx_desc);
qdf_tso_seg_dbg_record(tx_desc->tso_desc,
TSOSEG_LOC_TXPREPLLFAST);
tx_desc->tso_num_desc = msdu_info->tso_info.tso_num_seg_list;
tx_desc->pkt_type = OL_TX_FRM_TSO;
TXRX_STATS_MSDU_INCR(pdev, tx.tso.tso_pkts, msdu);
} else {
tx_desc->pkt_type = OL_TX_FRM_STD;
}
htt_tx_desc = tx_desc->htt_tx_desc;
/* Make sure frags num is set to 0 */
/*
* Do this here rather than in hardstart, so
* that we can hopefully take only one cache-miss while
* accessing skb->cb.
*/
/* HTT Header */
/* TODO : Take care of multiple fragments */
type = ol_tx_get_ext_header_type(vdev, msdu);
/* TODO: Precompute and store paddr in ol_tx_desc_t */
/* Virtual address of the HTT/HTC header, added by driver */
htc_hdr_vaddr = (char *)htt_tx_desc - HTC_HEADER_LEN;
if (qdf_unlikely(htt_tx_desc_init(pdev->htt_pdev, htt_tx_desc,
tx_desc->htt_tx_desc_paddr, tx_desc->id, msdu,
&msdu_info->htt, &msdu_info->tso_info,
NULL, type))) {
/*
* HTT Tx descriptor initialization failed.
* therefore, free the tx desc
*/
ol_tx_desc_free(pdev, tx_desc);
return NULL;
}
num_frags = qdf_nbuf_get_num_frags(msdu);
/* num_frags are expected to be 2 max */
num_frags = (num_frags > QDF_NBUF_CB_TX_MAX_EXTRA_FRAGS)
? QDF_NBUF_CB_TX_MAX_EXTRA_FRAGS
: num_frags;
#if defined(HELIUMPLUS)
/*
* Use num_frags - 1, since 1 frag is used to store
* the HTT/HTC descriptor
* Refer to htt_tx_desc_init()
*/
htt_tx_desc_num_frags(pdev->htt_pdev, tx_desc->htt_frag_desc,
num_frags - 1);
#else /* ! defined(HELIUMPLUS) */
htt_tx_desc_num_frags(pdev->htt_pdev, tx_desc->htt_tx_desc,
num_frags-1);
#endif /* defined(HELIUMPLUS) */
if (msdu_info->tso_info.is_tso) {
htt_tx_desc_fill_tso_info(pdev->htt_pdev,
tx_desc->htt_frag_desc, &msdu_info->tso_info);
TXRX_STATS_TSO_SEG_UPDATE(pdev,
msdu_info->tso_info.msdu_stats_idx,
msdu_info->tso_info.curr_seg->seg);
} else {
for (i = 1; i < num_frags; i++) {
qdf_size_t frag_len;
qdf_dma_addr_t frag_paddr;
frag_len = qdf_nbuf_get_frag_len(msdu, i);
frag_paddr = qdf_nbuf_get_frag_paddr(msdu, i);
if (type != EXT_HEADER_NOT_PRESENT) {
frag_paddr +=
sizeof(struct htt_tx_msdu_desc_ext_t);
frag_len -=
sizeof(struct htt_tx_msdu_desc_ext_t);
}
#if defined(HELIUMPLUS)
htt_tx_desc_frag(pdev->htt_pdev, tx_desc->htt_frag_desc,
i - 1, frag_paddr, frag_len);
#if defined(HELIUMPLUS_DEBUG)
qdf_print("%s:%d: htt_fdesc=%p frag=%d frag_paddr=0x%0llx len=%zu",
__func__, __LINE__, tx_desc->htt_frag_desc,
i-1, frag_paddr, frag_len);
ol_txrx_dump_pkt(netbuf, frag_paddr, 64);
#endif /* HELIUMPLUS_DEBUG */
#else /* ! defined(HELIUMPLUS) */
htt_tx_desc_frag(pdev->htt_pdev, tx_desc->htt_tx_desc,
i - 1, frag_paddr, frag_len);
#endif /* defined(HELIUMPLUS) */
}
}
/*
* Do we want to turn on word_stream bit-map here ? For linux, non-TSO
* this is not required. We still have to mark the swap bit correctly,
* when posting to the ring
*/
/* Check to make sure, data download length is correct */
/*
* TODO : Can we remove this check and always download a fixed length ?
*/
if (QDF_NBUF_CB_TX_EXTRA_FRAG_FLAGS_EXT_HEADER(msdu))
pkt_download_len += sizeof(struct htt_tx_msdu_desc_ext_t);
if (qdf_unlikely(qdf_nbuf_len(msdu) < pkt_download_len))
pkt_download_len = qdf_nbuf_len(msdu);
/* Fill the HTC header information */
/*
* Passing 0 as the seq_no field, we can probably get away
* with it for the time being, since this is not checked in f/w
*/
/* TODO : Prefill this, look at multi-fragment case */
HTC_TX_DESC_FILL(htc_hdr_vaddr, pkt_download_len, ep_id, 0);
return tx_desc;
}
#if defined(FEATURE_TSO)
/**
* ol_tx_ll_fast() Update metadata information and send msdu to HIF/CE
*
* @vdev: handle to ol_txrx_vdev_t
* @msdu_list: msdu list to be sent out.
*
* Return: on success return NULL, pointer to nbuf when it fails to send.
*/
qdf_nbuf_t
ol_tx_ll_fast(ol_txrx_vdev_handle vdev, qdf_nbuf_t msdu_list)
{
qdf_nbuf_t msdu = msdu_list;
struct ol_txrx_pdev_t *pdev = vdev->pdev;
uint32_t pkt_download_len =
((struct htt_pdev_t *)(pdev->htt_pdev))->download_len;
uint32_t ep_id = HTT_EPID_GET(pdev->htt_pdev);
struct ol_txrx_msdu_info_t msdu_info;
uint32_t tso_msdu_stats_idx = 0;
msdu_info.htt.info.l2_hdr_type = vdev->pdev->htt_pkt_type;
msdu_info.htt.action.tx_comp_req = 0;
/*
* The msdu_list variable could be used instead of the msdu var,
* but just to clarify which operations are done on a single MSDU
* vs. a list of MSDUs, use a distinct variable for single MSDUs
* within the list.
*/
while (msdu) {
qdf_nbuf_t next;
struct ol_tx_desc_t *tx_desc;
int segments = 1;
msdu_info.htt.info.ext_tid = qdf_nbuf_get_tid(msdu);
msdu_info.peer = NULL;
if (qdf_unlikely(ol_tx_prepare_tso(vdev, msdu, &msdu_info))) {
ol_txrx_dbg("ol_tx_prepare_tso failed\n");
TXRX_STATS_MSDU_LIST_INCR(vdev->pdev,
tx.dropped.host_reject, msdu);
return msdu;
}
segments = msdu_info.tso_info.num_segs;
if (msdu_info.tso_info.is_tso) {
tso_msdu_stats_idx =
ol_tx_tso_get_stats_idx(vdev->pdev);
msdu_info.tso_info.msdu_stats_idx = tso_msdu_stats_idx;
ol_tx_tso_update_stats(vdev->pdev,
&(msdu_info.tso_info),
msdu, tso_msdu_stats_idx);
}
/*
* The netbuf may get linked into a different list
* inside the ce_send_fast function, so store the next
* pointer before the ce_send call.
*/
next = qdf_nbuf_next(msdu);
/* init the current segment to the 1st segment in the list */
while (segments) {
if (msdu_info.tso_info.curr_seg)
QDF_NBUF_CB_PADDR(msdu) = msdu_info.tso_info.
curr_seg->seg.tso_frags[0].paddr;
segments--;
msdu_info.htt.info.frame_type = pdev->htt_pkt_type;
msdu_info.htt.info.vdev_id = vdev->vdev_id;
msdu_info.htt.action.cksum_offload =
qdf_nbuf_get_tx_cksum(msdu);
switch (qdf_nbuf_get_exemption_type(msdu)) {
case QDF_NBUF_EXEMPT_NO_EXEMPTION:
case QDF_NBUF_EXEMPT_ON_KEY_MAPPING_KEY_UNAVAILABLE:
/* We want to encrypt this frame */
msdu_info.htt.action.do_encrypt = 1;
break;
case QDF_NBUF_EXEMPT_ALWAYS:
/* We don't want to encrypt this frame */
msdu_info.htt.action.do_encrypt = 0;
break;
default:
msdu_info.htt.action.do_encrypt = 1;
qdf_assert(0);
break;
}
tx_desc = ol_tx_prepare_ll_fast(pdev, vdev, msdu,
pkt_download_len, ep_id,
&msdu_info);
TXRX_STATS_MSDU_INCR(pdev, tx.from_stack, msdu);
if (qdf_likely(tx_desc)) {
/*
* if this is a jumbo nbuf, then increment the
* number of nbuf users for each additional
* segment of the msdu. This will ensure that
* the skb is freed only after receiving tx
* completion for all segments of an nbuf.
*/
if (segments)
qdf_nbuf_inc_users(msdu);
DPTRACE(qdf_dp_trace_ptr(msdu,
QDF_DP_TRACE_TXRX_FAST_PACKET_PTR_RECORD,
QDF_TRACE_DEFAULT_PDEV_ID,
qdf_nbuf_data_addr(msdu),
sizeof(qdf_nbuf_data(msdu)),
tx_desc->id, vdev->vdev_id));
/*
* If debug display is enabled, show the meta
* data being downloaded to the target via the
* HTT tx descriptor.
*/
if (QDF_NBUF_CB_TX_EXTRA_FRAG_FLAGS_EXT_HEADER
(msdu))
pkt_download_len +=
sizeof(struct htt_tx_msdu_desc_ext_t);
htt_tx_desc_display(tx_desc->htt_tx_desc);
if ((0 == ce_send_fast(pdev->ce_tx_hdl, msdu,
ep_id, pkt_download_len))) {
struct qdf_tso_info_t *tso_info =
&msdu_info.tso_info;
/*
* If TSO packet, free associated
* remaining TSO segment descriptors
*/
if (tx_desc->pkt_type ==
OL_TX_FRM_TSO) {
tso_info->curr_seg =
tso_info->curr_seg->next;
ol_free_remaining_tso_segs(vdev,
&msdu_info, true);
}
/*
* The packet could not be sent.
* Free the descriptor, return the
* packet to the caller.
*/
ol_tx_desc_frame_free_nonstd(pdev,
tx_desc,
htt_tx_status_download_fail);
return msdu;
}
if (msdu_info.tso_info.curr_seg) {
msdu_info.tso_info.curr_seg =
msdu_info.tso_info.curr_seg->next;
}
if (msdu_info.tso_info.is_tso) {
qdf_nbuf_reset_num_frags(msdu);
TXRX_STATS_TSO_INC_SEG(vdev->pdev,
tso_msdu_stats_idx);
TXRX_STATS_TSO_INC_SEG_IDX(vdev->pdev,
tso_msdu_stats_idx);
}
} else {
/*
* If TSO packet, free associated
* remaining TSO segment descriptors
*/
if (qdf_nbuf_is_tso(msdu))
ol_free_remaining_tso_segs(vdev,
&msdu_info, true);
TXRX_STATS_MSDU_LIST_INCR(
pdev, tx.dropped.host_reject, msdu);
/* the list of unaccepted MSDUs */
return msdu;
}
} /* while segments */
msdu = next;
} /* while msdus */
return NULL; /* all MSDUs were accepted */
}
#else
qdf_nbuf_t
ol_tx_ll_fast(ol_txrx_vdev_handle vdev, qdf_nbuf_t msdu_list)
{
qdf_nbuf_t msdu = msdu_list;
struct ol_txrx_pdev_t *pdev = vdev->pdev;
uint32_t pkt_download_len =
((struct htt_pdev_t *)(pdev->htt_pdev))->download_len;
uint32_t ep_id = HTT_EPID_GET(pdev->htt_pdev);
struct ol_txrx_msdu_info_t msdu_info;
msdu_info.htt.info.l2_hdr_type = vdev->pdev->htt_pkt_type;
msdu_info.htt.action.tx_comp_req = 0;
msdu_info.tso_info.is_tso = 0;
/*
* The msdu_list variable could be used instead of the msdu var,
* but just to clarify which operations are done on a single MSDU
* vs. a list of MSDUs, use a distinct variable for single MSDUs
* within the list.
*/
while (msdu) {
qdf_nbuf_t next;
struct ol_tx_desc_t *tx_desc;
msdu_info.htt.info.ext_tid = qdf_nbuf_get_tid(msdu);
msdu_info.peer = NULL;
msdu_info.htt.info.frame_type = pdev->htt_pkt_type;
msdu_info.htt.info.vdev_id = vdev->vdev_id;
msdu_info.htt.action.cksum_offload =
qdf_nbuf_get_tx_cksum(msdu);
switch (qdf_nbuf_get_exemption_type(msdu)) {
case QDF_NBUF_EXEMPT_NO_EXEMPTION:
case QDF_NBUF_EXEMPT_ON_KEY_MAPPING_KEY_UNAVAILABLE:
/* We want to encrypt this frame */
msdu_info.htt.action.do_encrypt = 1;
break;
case QDF_NBUF_EXEMPT_ALWAYS:
/* We don't want to encrypt this frame */
msdu_info.htt.action.do_encrypt = 0;
break;
default:
msdu_info.htt.action.do_encrypt = 1;
qdf_assert(0);
break;
}
tx_desc = ol_tx_prepare_ll_fast(pdev, vdev, msdu,
pkt_download_len, ep_id,
&msdu_info);
TXRX_STATS_MSDU_INCR(pdev, tx.from_stack, msdu);
if (qdf_likely(tx_desc)) {
DPTRACE(qdf_dp_trace_ptr(msdu,
QDF_DP_TRACE_TXRX_FAST_PACKET_PTR_RECORD,
QDF_TRACE_DEFAULT_PDEV_ID,
qdf_nbuf_data_addr(msdu),
sizeof(qdf_nbuf_data(msdu)), tx_desc->id,
vdev->vdev_id));
/*
* If debug display is enabled, show the meta-data being
* downloaded to the target via the HTT tx descriptor.
*/
if (QDF_NBUF_CB_TX_EXTRA_FRAG_FLAGS_EXT_HEADER(msdu))
pkt_download_len +=
sizeof(struct htt_tx_msdu_desc_ext_t);
htt_tx_desc_display(tx_desc->htt_tx_desc);
/*
* The netbuf may get linked into a different list
* inside the ce_send_fast function, so store the next
* pointer before the ce_send call.
*/
next = qdf_nbuf_next(msdu);
if ((0 == ce_send_fast(pdev->ce_tx_hdl, msdu,
ep_id, pkt_download_len))) {
/*
* The packet could not be sent
* Free the descriptor, return the packet to the
* caller
*/
ol_tx_desc_free(pdev, tx_desc);
return msdu;
}
msdu = next;
} else {
TXRX_STATS_MSDU_LIST_INCR(
pdev, tx.dropped.host_reject, msdu);
return msdu; /* the list of unaccepted MSDUs */
}
}
return NULL; /* all MSDUs were accepted */
}
#endif /* FEATURE_TSO */
#endif /* WLAN_FEATURE_FASTPATH */
#ifdef WLAN_FEATURE_FASTPATH
/**
* ol_tx_ll_wrapper() wrapper to ol_tx_ll
*
*/
qdf_nbuf_t
ol_tx_ll_wrapper(ol_txrx_vdev_handle vdev, qdf_nbuf_t msdu_list)
{
struct hif_opaque_softc *hif_device =
(struct hif_opaque_softc *)cds_get_context(QDF_MODULE_ID_HIF);
if (qdf_likely(hif_device && hif_is_fastpath_mode_enabled(hif_device)))
msdu_list = ol_tx_ll_fast(vdev, msdu_list);
else
msdu_list = ol_tx_ll(vdev, msdu_list);
return msdu_list;
}
#else
qdf_nbuf_t
ol_tx_ll_wrapper(ol_txrx_vdev_handle vdev, qdf_nbuf_t msdu_list)
{
return ol_tx_ll(vdev, msdu_list);
}
#endif /* WLAN_FEATURE_FASTPATH */
#ifdef QCA_LL_LEGACY_TX_FLOW_CONTROL
#define OL_TX_VDEV_PAUSE_QUEUE_SEND_MARGIN 400
#define OL_TX_VDEV_PAUSE_QUEUE_SEND_PERIOD_MS 5
static void ol_tx_vdev_ll_pause_queue_send_base(struct ol_txrx_vdev_t *vdev)
{
int max_to_accept;
qdf_spin_lock_bh(&vdev->ll_pause.mutex);
if (vdev->ll_pause.paused_reason) {
qdf_spin_unlock_bh(&vdev->ll_pause.mutex);
return;
}
/*
* Send as much of the backlog as possible, but leave some margin
* of unallocated tx descriptors that can be used for new frames
* being transmitted by other vdevs.
* Ideally there would be a scheduler, which would not only leave
* some margin for new frames for other vdevs, but also would
* fairly apportion the tx descriptors between multiple vdevs that
* have backlogs in their pause queues.
* However, the fairness benefit of having a scheduler for frames
* from multiple vdev's pause queues is not sufficient to outweigh
* the extra complexity.
*/
max_to_accept = vdev->pdev->tx_desc.num_free -
OL_TX_VDEV_PAUSE_QUEUE_SEND_MARGIN;
while (max_to_accept > 0 && vdev->ll_pause.txq.depth) {
qdf_nbuf_t tx_msdu;
max_to_accept--;
vdev->ll_pause.txq.depth--;
tx_msdu = vdev->ll_pause.txq.head;
if (tx_msdu) {
vdev->ll_pause.txq.head = qdf_nbuf_next(tx_msdu);
if (NULL == vdev->ll_pause.txq.head)
vdev->ll_pause.txq.tail = NULL;
qdf_nbuf_set_next(tx_msdu, NULL);
QDF_NBUF_UPDATE_TX_PKT_COUNT(tx_msdu,
QDF_NBUF_TX_PKT_TXRX_DEQUEUE);
tx_msdu = ol_tx_ll_wrapper(vdev, tx_msdu);
/*
* It is unexpected that ol_tx_ll would reject the frame
* since we checked that there's room for it, though
* there's an infinitesimal possibility that between the
* time we checked the room available and now, a
* concurrent batch of tx frames used up all the room.
* For simplicity, just drop the frame.
*/
if (tx_msdu) {
qdf_nbuf_unmap(vdev->pdev->osdev, tx_msdu,
QDF_DMA_TO_DEVICE);
qdf_nbuf_tx_free(tx_msdu, QDF_NBUF_PKT_ERROR);
}
}
}
if (vdev->ll_pause.txq.depth) {
qdf_timer_stop(&vdev->ll_pause.timer);
qdf_timer_start(&vdev->ll_pause.timer,
OL_TX_VDEV_PAUSE_QUEUE_SEND_PERIOD_MS);
vdev->ll_pause.is_q_timer_on = true;
if (vdev->ll_pause.txq.depth >= vdev->ll_pause.max_q_depth)
vdev->ll_pause.q_overflow_cnt++;
}
qdf_spin_unlock_bh(&vdev->ll_pause.mutex);
}
static qdf_nbuf_t
ol_tx_vdev_pause_queue_append(struct ol_txrx_vdev_t *vdev,
qdf_nbuf_t msdu_list, uint8_t start_timer)
{
qdf_spin_lock_bh(&vdev->ll_pause.mutex);
while (msdu_list &&
vdev->ll_pause.txq.depth < vdev->ll_pause.max_q_depth) {
qdf_nbuf_t next = qdf_nbuf_next(msdu_list);
QDF_NBUF_UPDATE_TX_PKT_COUNT(msdu_list,
QDF_NBUF_TX_PKT_TXRX_ENQUEUE);
DPTRACE(qdf_dp_trace(msdu_list,
QDF_DP_TRACE_TXRX_QUEUE_PACKET_PTR_RECORD,
QDF_TRACE_DEFAULT_PDEV_ID,
qdf_nbuf_data_addr(msdu_list),
sizeof(qdf_nbuf_data(msdu_list)), QDF_TX));
vdev->ll_pause.txq.depth++;
if (!vdev->ll_pause.txq.head) {
vdev->ll_pause.txq.head = msdu_list;
vdev->ll_pause.txq.tail = msdu_list;
} else {
qdf_nbuf_set_next(vdev->ll_pause.txq.tail, msdu_list);
}
vdev->ll_pause.txq.tail = msdu_list;
msdu_list = next;
}
if (vdev->ll_pause.txq.tail)
qdf_nbuf_set_next(vdev->ll_pause.txq.tail, NULL);
if (start_timer) {
qdf_timer_stop(&vdev->ll_pause.timer);
qdf_timer_start(&vdev->ll_pause.timer,
OL_TX_VDEV_PAUSE_QUEUE_SEND_PERIOD_MS);
vdev->ll_pause.is_q_timer_on = true;
}
qdf_spin_unlock_bh(&vdev->ll_pause.mutex);
return msdu_list;
}
/*
* Store up the tx frame in the vdev's tx queue if the vdev is paused.
* If there are too many frames in the tx queue, reject it.
*/
qdf_nbuf_t ol_tx_ll_queue(ol_txrx_vdev_handle vdev, qdf_nbuf_t msdu_list)
{
uint16_t eth_type;
uint32_t paused_reason;
if (msdu_list == NULL)
return NULL;
paused_reason = vdev->ll_pause.paused_reason;
if (paused_reason) {
if (qdf_unlikely((paused_reason &
OL_TXQ_PAUSE_REASON_PEER_UNAUTHORIZED) ==
paused_reason)) {
eth_type = (((struct ethernet_hdr_t *)
qdf_nbuf_data(msdu_list))->
ethertype[0] << 8) |
(((struct ethernet_hdr_t *)
qdf_nbuf_data(msdu_list))->ethertype[1]);
if (ETHERTYPE_IS_EAPOL_WAPI(eth_type)) {
msdu_list = ol_tx_ll_wrapper(vdev, msdu_list);
return msdu_list;
}
}
msdu_list = ol_tx_vdev_pause_queue_append(vdev, msdu_list, 1);
} else {
if (vdev->ll_pause.txq.depth > 0 ||
vdev->pdev->tx_throttle.current_throttle_level !=
THROTTLE_LEVEL_0) {
/*
* not paused, but there is a backlog of frms
* from a prior pause or throttle off phase
*/
msdu_list = ol_tx_vdev_pause_queue_append(
vdev, msdu_list, 0);
/*
* if throttle is disabled or phase is "on",
* send the frame
*/
if (vdev->pdev->tx_throttle.current_throttle_level ==
THROTTLE_LEVEL_0 ||
vdev->pdev->tx_throttle.current_throttle_phase ==
THROTTLE_PHASE_ON) {
/*
* send as many frames as possible
* from the vdevs backlog
*/
ol_tx_vdev_ll_pause_queue_send_base(vdev);
}
} else {
/*
* not paused, no throttle and no backlog -
* send the new frames
*/
msdu_list = ol_tx_ll_wrapper(vdev, msdu_list);
}
}
return msdu_list;
}
/*
* Run through the transmit queues for all the vdevs and
* send the pending frames
*/
void ol_tx_pdev_ll_pause_queue_send_all(struct ol_txrx_pdev_t *pdev)
{
int max_to_send; /* tracks how many frames have been sent */
qdf_nbuf_t tx_msdu;
struct ol_txrx_vdev_t *vdev = NULL;
uint8_t more;
if (NULL == pdev)
return;
if (pdev->tx_throttle.current_throttle_phase == THROTTLE_PHASE_OFF)
return;
/* ensure that we send no more than tx_threshold frames at once */
max_to_send = pdev->tx_throttle.tx_threshold;
/* round robin through the vdev queues for the given pdev */
/*
* Potential improvement: download several frames from the same vdev
* at a time, since it is more likely that those frames could be
* aggregated together, remember which vdev was serviced last,
* so the next call this function can resume the round-robin
* traversing where the current invocation left off
*/
do {
more = 0;
TAILQ_FOREACH(vdev, &pdev->vdev_list, vdev_list_elem) {
qdf_spin_lock_bh(&vdev->ll_pause.mutex);
if (vdev->ll_pause.txq.depth) {
if (vdev->ll_pause.paused_reason) {
qdf_spin_unlock_bh(&vdev->ll_pause.
mutex);
continue;
}
tx_msdu = vdev->ll_pause.txq.head;
if (NULL == tx_msdu) {
qdf_spin_unlock_bh(&vdev->ll_pause.
mutex);
continue;
}
max_to_send--;
vdev->ll_pause.txq.depth--;
vdev->ll_pause.txq.head =
qdf_nbuf_next(tx_msdu);
if (NULL == vdev->ll_pause.txq.head)
vdev->ll_pause.txq.tail = NULL;
qdf_nbuf_set_next(tx_msdu, NULL);
tx_msdu = ol_tx_ll_wrapper(vdev, tx_msdu);
/*
* It is unexpected that ol_tx_ll would reject
* the frame, since we checked that there's
* room for it, though there's an infinitesimal
* possibility that between the time we checked
* the room available and now, a concurrent
* batch of tx frames used up all the room.
* For simplicity, just drop the frame.
*/
if (tx_msdu) {
qdf_nbuf_unmap(pdev->osdev, tx_msdu,
QDF_DMA_TO_DEVICE);
qdf_nbuf_tx_free(tx_msdu,
QDF_NBUF_PKT_ERROR);
}
}
/*check if there are more msdus to transmit */
if (vdev->ll_pause.txq.depth)
more = 1;
qdf_spin_unlock_bh(&vdev->ll_pause.mutex);
}
} while (more && max_to_send);
vdev = NULL;
TAILQ_FOREACH(vdev, &pdev->vdev_list, vdev_list_elem) {
qdf_spin_lock_bh(&vdev->ll_pause.mutex);
if (vdev->ll_pause.txq.depth) {
qdf_timer_stop(&pdev->tx_throttle.tx_timer);
qdf_timer_start(
&pdev->tx_throttle.tx_timer,
OL_TX_VDEV_PAUSE_QUEUE_SEND_PERIOD_MS);
qdf_spin_unlock_bh(&vdev->ll_pause.mutex);
return;
}
qdf_spin_unlock_bh(&vdev->ll_pause.mutex);
}
}
void ol_tx_vdev_ll_pause_queue_send(void *context)
{
struct ol_txrx_vdev_t *vdev = (struct ol_txrx_vdev_t *)context;
struct ol_txrx_pdev_t *pdev = vdev->pdev;
if (pdev->tx_throttle.current_throttle_level != THROTTLE_LEVEL_0 &&
pdev->tx_throttle.current_throttle_phase == THROTTLE_PHASE_OFF)
return;
ol_tx_vdev_ll_pause_queue_send_base(vdev);
}
#endif /* QCA_LL_LEGACY_TX_FLOW_CONTROL */
static inline int ol_txrx_tx_is_raw(enum ol_tx_spec tx_spec)
{
return
tx_spec &
(OL_TX_SPEC_RAW | OL_TX_SPEC_NO_AGGR | OL_TX_SPEC_NO_ENCRYPT);
}
static inline uint8_t ol_txrx_tx_raw_subtype(enum ol_tx_spec tx_spec)
{
uint8_t sub_type = 0x1; /* 802.11 MAC header present */
if (tx_spec & OL_TX_SPEC_NO_AGGR)
sub_type |= 0x1 << HTT_TX_MSDU_DESC_RAW_SUBTYPE_NO_AGGR_S;
if (tx_spec & OL_TX_SPEC_NO_ENCRYPT)
sub_type |= 0x1 << HTT_TX_MSDU_DESC_RAW_SUBTYPE_NO_ENCRYPT_S;
if (tx_spec & OL_TX_SPEC_NWIFI_NO_ENCRYPT)
sub_type |= 0x1 << HTT_TX_MSDU_DESC_RAW_SUBTYPE_NO_ENCRYPT_S;
return sub_type;
}
static qdf_nbuf_t
ol_tx_non_std_ll(struct ol_txrx_vdev_t *vdev,
enum ol_tx_spec tx_spec,
qdf_nbuf_t msdu_list)
{
qdf_nbuf_t msdu = msdu_list;
htt_pdev_handle htt_pdev = vdev->pdev->htt_pdev;
struct ol_txrx_msdu_info_t msdu_info;
msdu_info.htt.info.l2_hdr_type = vdev->pdev->htt_pkt_type;
msdu_info.htt.action.tx_comp_req = 0;
/*
* The msdu_list variable could be used instead of the msdu var,
* but just to clarify which operations are done on a single MSDU
* vs. a list of MSDUs, use a distinct variable for single MSDUs
* within the list.
*/
while (msdu) {
qdf_nbuf_t next;
struct ol_tx_desc_t *tx_desc = NULL;
msdu_info.htt.info.ext_tid = qdf_nbuf_get_tid(msdu);
msdu_info.peer = NULL;
msdu_info.tso_info.is_tso = 0;
tx_desc = ol_tx_prepare_ll(vdev, msdu, &msdu_info);
if (!tx_desc)
return msdu;
/*
* The netbuf may get linked into a different list inside the
* ol_tx_send function, so store the next pointer before the
* tx_send call.
*/
next = qdf_nbuf_next(msdu);
if (tx_spec != OL_TX_SPEC_STD) {
if (tx_spec & OL_TX_SPEC_NO_FREE) {
tx_desc->pkt_type = OL_TX_FRM_NO_FREE;
} else if (tx_spec & OL_TX_SPEC_TSO) {
tx_desc->pkt_type = OL_TX_FRM_TSO;
} else if (tx_spec & OL_TX_SPEC_NWIFI_NO_ENCRYPT) {
uint8_t sub_type =
ol_txrx_tx_raw_subtype(tx_spec);
htt_tx_desc_type(htt_pdev, tx_desc->htt_tx_desc,
htt_pkt_type_native_wifi,
sub_type);
} else if (ol_txrx_tx_is_raw(tx_spec)) {
/* different types of raw frames */
uint8_t sub_type =
ol_txrx_tx_raw_subtype(tx_spec);
htt_tx_desc_type(htt_pdev, tx_desc->htt_tx_desc,
htt_pkt_type_raw, sub_type);
}
}
/*
* If debug display is enabled, show the meta-data being
* downloaded to the target via the HTT tx descriptor.
*/
htt_tx_desc_display(tx_desc->htt_tx_desc);
ol_tx_send(vdev->pdev, tx_desc, msdu, vdev->vdev_id);
msdu = next;
}
return NULL; /* all MSDUs were accepted */
}
#ifdef QCA_SUPPORT_SW_TXRX_ENCAP
static inline int ol_tx_encap_wrapper(struct ol_txrx_pdev_t *pdev,
ol_txrx_vdev_handle vdev,
struct ol_tx_desc_t *tx_desc,
qdf_nbuf_t msdu,
struct ol_txrx_msdu_info_t *tx_msdu_info)
{
if (OL_TX_ENCAP(vdev, tx_desc, msdu, tx_msdu_info) != A_OK) {
qdf_atomic_inc(&pdev->tx_queue.rsrc_cnt);
ol_tx_desc_frame_free_nonstd(pdev, tx_desc, 1);
if (tx_msdu_info->peer) {
/* remove the peer reference added above */
OL_TXRX_PEER_UNREF_DELETE(tx_msdu_info->peer);
}
return -EINVAL;
}
return 0;
}
#else
static inline int ol_tx_encap_wrapper(struct ol_txrx_pdev_t *pdev,
ol_txrx_vdev_handle vdev,
struct ol_tx_desc_t *tx_desc,
qdf_nbuf_t msdu,
struct ol_txrx_msdu_info_t *tx_msdu_info)
{
/* no-op */
return 0;
}
#endif
/* tx filtering is handled within the target FW */
#define TX_FILTER_CHECK(tx_msdu_info) 0 /* don't filter */
/**
* parse_ocb_tx_header() - Function to check for OCB
* @msdu: Pointer to OS packet (qdf_nbuf_t)
* @tx_ctrl: TX control header on a packet and extract it if present
*
* Return: true if ocb parsing is successful
*/
#define OCB_HEADER_VERSION 1
static bool parse_ocb_tx_header(qdf_nbuf_t msdu,
struct ocb_tx_ctrl_hdr_t *tx_ctrl)
{
struct ether_header *eth_hdr_p;
struct ocb_tx_ctrl_hdr_t *tx_ctrl_hdr;
/* Check if TX control header is present */
eth_hdr_p = (struct ether_header *)qdf_nbuf_data(msdu);
if (eth_hdr_p->ether_type != QDF_SWAP_U16(ETHERTYPE_OCB_TX))
/* TX control header is not present. Nothing to do.. */
return true;
/* Remove the ethernet header */
qdf_nbuf_pull_head(msdu, sizeof(struct ether_header));
/* Parse the TX control header */
tx_ctrl_hdr = (struct ocb_tx_ctrl_hdr_t *)qdf_nbuf_data(msdu);
if (tx_ctrl_hdr->version == OCB_HEADER_VERSION) {
if (tx_ctrl)
qdf_mem_copy(tx_ctrl, tx_ctrl_hdr,
sizeof(*tx_ctrl_hdr));
} else {
/* The TX control header is invalid. */
return false;
}
/* Remove the TX control header */
qdf_nbuf_pull_head(msdu, tx_ctrl_hdr->length);
return true;
}
#if defined(CONFIG_HL_SUPPORT) && defined(CONFIG_TX_DESC_HI_PRIO_RESERVE)
/**
* ol_tx_hl_desc_alloc() - Allocate and initialize a tx descriptor
* for a HL system.
* @pdev: the data physical device sending the data
* @vdev: the virtual device sending the data
* @msdu: the tx frame
* @msdu_info: the tx meta data
*
* Return: the tx decriptor
*/
static inline
struct ol_tx_desc_t *ol_tx_hl_desc_alloc(struct ol_txrx_pdev_t *pdev,
struct ol_txrx_vdev_t *vdev,
qdf_nbuf_t msdu,
struct ol_txrx_msdu_info_t *msdu_info)
{
struct ol_tx_desc_t *tx_desc = NULL;
if (qdf_atomic_read(&pdev->tx_queue.rsrc_cnt) >
TXRX_HL_TX_DESC_HI_PRIO_RESERVED) {
tx_desc = ol_tx_desc_hl(pdev, vdev, msdu, msdu_info);
} else if (qdf_nbuf_is_ipv4_pkt(msdu) == true) {
if ((QDF_NBUF_CB_GET_PACKET_TYPE(msdu) ==
QDF_NBUF_CB_PACKET_TYPE_DHCP) ||
(QDF_NBUF_CB_GET_PACKET_TYPE(msdu) ==
QDF_NBUF_CB_PACKET_TYPE_EAPOL)) {
tx_desc = ol_tx_desc_hl(pdev, vdev, msdu, msdu_info);
ol_txrx_info("Provided tx descriptor from reserve pool for DHCP/EAPOL\n");
}
}
return tx_desc;
}
#else
static inline
struct ol_tx_desc_t *ol_tx_hl_desc_alloc(struct ol_txrx_pdev_t *pdev,
struct ol_txrx_vdev_t *vdev,
qdf_nbuf_t msdu,
struct ol_txrx_msdu_info_t *msdu_info)
{
struct ol_tx_desc_t *tx_desc = NULL;
tx_desc = ol_tx_desc_hl(pdev, vdev, msdu, msdu_info);
return tx_desc;
}
#endif
#if defined(CONFIG_HL_SUPPORT)
/**
* ol_txrx_mgmt_tx_desc_alloc() - Allocate and initialize a tx descriptor
* for management frame
* @pdev: the data physical device sending the data
* @vdev: the virtual device sending the data
* @tx_mgmt_frm: the tx managment frame
* @tx_msdu_info: the tx meta data
*
* Return: the tx decriptor
*/
static inline
struct ol_tx_desc_t *
ol_txrx_mgmt_tx_desc_alloc(
struct ol_txrx_pdev_t *pdev,
struct ol_txrx_vdev_t *vdev,
qdf_nbuf_t tx_mgmt_frm,
struct ol_txrx_msdu_info_t *tx_msdu_info)
{
struct ol_tx_desc_t *tx_desc;
tx_msdu_info->htt.action.tx_comp_req = 1;
tx_desc = ol_tx_desc_hl(pdev, vdev, tx_mgmt_frm, tx_msdu_info);
return tx_desc;
}
/**
* ol_txrx_mgmt_send_frame() - send a management frame
* @vdev: virtual device sending the frame
* @tx_desc: tx desc
* @tx_mgmt_frm: management frame to send
* @tx_msdu_info: the tx meta data
* @chanfreq: download change frequency
*
* Return:
* 0 -> the frame is accepted for transmission, -OR-
* 1 -> the frame was not accepted
*/
static inline
int ol_txrx_mgmt_send_frame(
struct ol_txrx_vdev_t *vdev,
struct ol_tx_desc_t *tx_desc,
qdf_nbuf_t tx_mgmt_frm,
struct ol_txrx_msdu_info_t *tx_msdu_info,
uint16_t chanfreq)
{
struct ol_txrx_pdev_t *pdev = vdev->pdev;
struct ol_tx_frms_queue_t *txq;
/*
* 1. Look up the peer and queue the frame in the peer's mgmt queue.
* 2. Invoke the download scheduler.
*/
txq = ol_tx_classify_mgmt(vdev, tx_desc, tx_mgmt_frm, tx_msdu_info);
if (!txq) {
/* TXRX_STATS_MSDU_LIST_INCR(vdev->pdev, tx.dropped.no_txq,
* msdu);
*/
qdf_atomic_inc(&pdev->tx_queue.rsrc_cnt);
ol_tx_desc_frame_free_nonstd(vdev->pdev, tx_desc,
1 /* error */);
if (tx_msdu_info->peer) {
/* remove the peer reference added above */
OL_TXRX_PEER_UNREF_DELETE(tx_msdu_info->peer);
}
return 1; /* can't accept the tx mgmt frame */
}
/* Initialize the HTT tx desc l2 header offset field.
* Even though tx encap does not apply to mgmt frames,
* htt_tx_desc_mpdu_header still needs to be called,
* to specifiy that there was no L2 header added by tx encap,
* so the frame's length does not need to be adjusted to account for
* an added L2 header.
*/
htt_tx_desc_mpdu_header(tx_desc->htt_tx_desc, 0);
if (qdf_unlikely(htt_tx_desc_init(
pdev->htt_pdev, tx_desc->htt_tx_desc,
tx_desc->htt_tx_desc_paddr,
ol_tx_desc_id(pdev, tx_desc),
tx_mgmt_frm,
&tx_msdu_info->htt, &tx_msdu_info->tso_info, NULL, 0)))
return 1;
htt_tx_desc_display(tx_desc->htt_tx_desc);
htt_tx_desc_set_chanfreq(tx_desc->htt_tx_desc, chanfreq);
ol_tx_enqueue(vdev->pdev, txq, tx_desc, tx_msdu_info);
if (tx_msdu_info->peer) {
/* remove the peer reference added above */
OL_TXRX_PEER_UNREF_DELETE(tx_msdu_info->peer);
}
ol_tx_sched(vdev->pdev);
return 0;
}
#else
static inline
struct ol_tx_desc_t *
ol_txrx_mgmt_tx_desc_alloc(
struct ol_txrx_pdev_t *pdev,
struct ol_txrx_vdev_t *vdev,
qdf_nbuf_t tx_mgmt_frm,
struct ol_txrx_msdu_info_t *tx_msdu_info)
{
struct ol_tx_desc_t *tx_desc;
/* For LL tx_comp_req is not used so initialized to 0 */
tx_msdu_info->htt.action.tx_comp_req = 0;
tx_desc = ol_tx_desc_ll(pdev, vdev, tx_mgmt_frm, tx_msdu_info);
/* FIX THIS -
* The FW currently has trouble using the host's fragments table
* for management frames. Until this is fixed, rather than
* specifying the fragment table to the FW, specify just the
* address of the initial fragment.
*/
#if defined(HELIUMPLUS)
/* ol_txrx_dump_frag_desc("ol_txrx_mgmt_send(): after ol_tx_desc_ll",
* tx_desc);
*/
#endif /* defined(HELIUMPLUS) */
if (tx_desc) {
/*
* Following the call to ol_tx_desc_ll, frag 0 is the
* HTT tx HW descriptor, and the frame payload is in
* frag 1.
*/
htt_tx_desc_frags_table_set(
pdev->htt_pdev,
tx_desc->htt_tx_desc,
qdf_nbuf_get_frag_paddr(tx_mgmt_frm, 1),
0, 0);
#if defined(HELIUMPLUS) && defined(HELIUMPLUS_DEBUG)
ol_txrx_dump_frag_desc(
"after htt_tx_desc_frags_table_set",
tx_desc);
#endif /* defined(HELIUMPLUS) */
}
return tx_desc;
}
static inline
int ol_txrx_mgmt_send_frame(
struct ol_txrx_vdev_t *vdev,
struct ol_tx_desc_t *tx_desc,
qdf_nbuf_t tx_mgmt_frm,
struct ol_txrx_msdu_info_t *tx_msdu_info,
uint16_t chanfreq)
{
struct ol_txrx_pdev_t *pdev = vdev->pdev;
htt_tx_desc_set_chanfreq(tx_desc->htt_tx_desc, chanfreq);
QDF_NBUF_CB_TX_PACKET_TRACK(tx_desc->netbuf) =
QDF_NBUF_TX_PKT_MGMT_TRACK;
ol_tx_send_nonstd(pdev, tx_desc, tx_mgmt_frm,
htt_pkt_type_mgmt);
return 0;
}
#endif
/**
* ol_tx_hl_base() - send tx frames for a HL system.
* @vdev: the virtual device sending the data
* @tx_spec: indicate what non-standard transmission actions to apply
* @msdu_list: the tx frames to send
* @tx_comp_req: tx completion req
*
* Return: NULL if all MSDUs are accepted
*/
static inline qdf_nbuf_t
ol_tx_hl_base(
ol_txrx_vdev_handle vdev,
enum ol_tx_spec tx_spec,
qdf_nbuf_t msdu_list,
int tx_comp_req)
{
struct ol_txrx_pdev_t *pdev = vdev->pdev;
qdf_nbuf_t msdu = msdu_list;
struct ol_txrx_msdu_info_t tx_msdu_info;
struct ocb_tx_ctrl_hdr_t tx_ctrl;
htt_pdev_handle htt_pdev = pdev->htt_pdev;
tx_msdu_info.tso_info.is_tso = 0;
/*
* The msdu_list variable could be used instead of the msdu var,
* but just to clarify which operations are done on a single MSDU
* vs. a list of MSDUs, use a distinct variable for single MSDUs
* within the list.
*/
while (msdu) {
qdf_nbuf_t next;
struct ol_tx_frms_queue_t *txq;
struct ol_tx_desc_t *tx_desc = NULL;
qdf_mem_zero(&tx_ctrl, sizeof(tx_ctrl));
tx_msdu_info.peer = NULL;
/*
* The netbuf will get stored into a (peer-TID) tx queue list
* inside the ol_tx_classify_store function or else dropped,
* so store the next pointer immediately.
*/
next = qdf_nbuf_next(msdu);
tx_desc = ol_tx_hl_desc_alloc(pdev, vdev, msdu, &tx_msdu_info);
if (!tx_desc) {
/*
* If we're out of tx descs, there's no need to try
* to allocate tx descs for the remaining MSDUs.
*/
TXRX_STATS_MSDU_LIST_INCR(pdev, tx.dropped.host_reject,
msdu);
return msdu; /* the list of unaccepted MSDUs */
}
/* OL_TXRX_PROT_AN_LOG(pdev->prot_an_tx_sent, msdu);*/
if (tx_spec != OL_TX_SPEC_STD) {
#if defined(FEATURE_WLAN_TDLS)
if (tx_spec & OL_TX_SPEC_NO_FREE) {
tx_desc->pkt_type = OL_TX_FRM_NO_FREE;
} else if (tx_spec & OL_TX_SPEC_TSO) {
#else
if (tx_spec & OL_TX_SPEC_TSO) {
#endif
tx_desc->pkt_type = OL_TX_FRM_TSO;
}
if (ol_txrx_tx_is_raw(tx_spec)) {
/* CHECK THIS: does this need
* to happen after htt_tx_desc_init?
*/
/* different types of raw frames */
u_int8_t sub_type =
ol_txrx_tx_raw_subtype(
tx_spec);
htt_tx_desc_type(htt_pdev,
tx_desc->htt_tx_desc,
htt_pkt_type_raw,
sub_type);
}
}
tx_msdu_info.htt.info.ext_tid = qdf_nbuf_get_tid(msdu);
tx_msdu_info.htt.info.vdev_id = vdev->vdev_id;
tx_msdu_info.htt.info.frame_type = htt_frm_type_data;
tx_msdu_info.htt.info.l2_hdr_type = pdev->htt_pkt_type;
tx_msdu_info.htt.action.tx_comp_req = tx_comp_req;
/* If the vdev is in OCB mode,
* parse the tx control header.
*/
if (vdev->opmode == wlan_op_mode_ocb) {
if (!parse_ocb_tx_header(msdu, &tx_ctrl)) {
/* There was an error parsing
* the header.Skip this packet.
*/
goto MSDU_LOOP_BOTTOM;
}
}
txq = ol_tx_classify(vdev, tx_desc, msdu,
&tx_msdu_info);
if ((!txq) || TX_FILTER_CHECK(&tx_msdu_info)) {
/* drop this frame,
* but try sending subsequent frames
*/
/*TXRX_STATS_MSDU_LIST_INCR(pdev,
tx.dropped.no_txq,
msdu);*/
qdf_atomic_inc(&pdev->tx_queue.rsrc_cnt);
ol_tx_desc_frame_free_nonstd(pdev, tx_desc, 1);
if (tx_msdu_info.peer) {
/* remove the peer reference
* added above */
OL_TXRX_PEER_UNREF_DELETE(
tx_msdu_info.peer);
}
goto MSDU_LOOP_BOTTOM;
}
if (tx_msdu_info.peer) {
/*
* If the state is not associated then drop all
* the data packets received for that peer
*/
if (tx_msdu_info.peer->state ==
OL_TXRX_PEER_STATE_DISC) {
qdf_atomic_inc(
&pdev->tx_queue.rsrc_cnt);
ol_tx_desc_frame_free_nonstd(pdev,
tx_desc,
1);
OL_TXRX_PEER_UNREF_DELETE(
tx_msdu_info.peer);
msdu = next;
continue;
} else if (tx_msdu_info.peer->state !=
OL_TXRX_PEER_STATE_AUTH) {
if (tx_msdu_info.htt.info.ethertype !=
ETHERTYPE_PAE &&
tx_msdu_info.htt.info.ethertype
!= ETHERTYPE_WAI) {
qdf_atomic_inc(
&pdev->tx_queue.
rsrc_cnt);
ol_tx_desc_frame_free_nonstd(
pdev,
tx_desc, 1);
OL_TXRX_PEER_UNREF_DELETE(
tx_msdu_info.peer);
msdu = next;
continue;
}
}
}
/*
* Initialize the HTT tx desc l2 header offset field.
* htt_tx_desc_mpdu_header needs to be called to
* make sure, the l2 header size is initialized
* correctly to handle cases where TX ENCAP is disabled
* or Tx Encap fails to perform Encap
*/
htt_tx_desc_mpdu_header(tx_desc->htt_tx_desc, 0);
/*
* Note: when the driver is built without support for
* SW tx encap,the following macro is a no-op.
* When the driver is built with support for SW tx
* encap, it performs encap, and if an error is
* encountered, jumps to the MSDU_LOOP_BOTTOM label.
*/
if (ol_tx_encap_wrapper(pdev, vdev, tx_desc, msdu,
&tx_msdu_info))
goto MSDU_LOOP_BOTTOM;
/* initialize the HW tx descriptor */
htt_tx_desc_init(
pdev->htt_pdev, tx_desc->htt_tx_desc,
tx_desc->htt_tx_desc_paddr,
ol_tx_desc_id(pdev, tx_desc),
msdu,
&tx_msdu_info.htt,
&tx_msdu_info.tso_info,
&tx_ctrl,
vdev->opmode == wlan_op_mode_ocb);
/*
* If debug display is enabled, show the meta-data
* being downloaded to the target via the
* HTT tx descriptor.
*/
htt_tx_desc_display(tx_desc->htt_tx_desc);
ol_tx_enqueue(pdev, txq, tx_desc, &tx_msdu_info);
if (tx_msdu_info.peer) {
OL_TX_PEER_STATS_UPDATE(tx_msdu_info.peer,
msdu);
/* remove the peer reference added above */
OL_TXRX_PEER_UNREF_DELETE(tx_msdu_info.peer);
}
MSDU_LOOP_BOTTOM:
msdu = next;
}
ol_tx_sched(pdev);
return NULL; /* all MSDUs were accepted */
}
qdf_nbuf_t
ol_tx_hl(ol_txrx_vdev_handle vdev, qdf_nbuf_t msdu_list)
{
struct ol_txrx_pdev_t *pdev = vdev->pdev;
int tx_comp_req = pdev->cfg.default_tx_comp_req;
return ol_tx_hl_base(vdev, OL_TX_SPEC_STD, msdu_list, tx_comp_req);
}
static qdf_nbuf_t
ol_tx_non_std_hl(struct ol_txrx_vdev_t *vdev,
enum ol_tx_spec tx_spec,
qdf_nbuf_t msdu_list)
{
struct ol_txrx_pdev_t *pdev = vdev->pdev;
int tx_comp_req = pdev->cfg.default_tx_comp_req;
if (!tx_comp_req) {
if ((tx_spec == OL_TX_SPEC_NO_FREE) &&
(pdev->tx_data_callback.func))
tx_comp_req = 1;
}
return ol_tx_hl_base(vdev, tx_spec, msdu_list, tx_comp_req);
}
qdf_nbuf_t
ol_tx_non_std(struct cdp_vdev *pvdev,
enum ol_tx_spec tx_spec, qdf_nbuf_t msdu_list)
{
struct ol_txrx_vdev_t *vdev = (struct ol_txrx_vdev_t *)pvdev;
if (vdev->pdev->cfg.is_high_latency)
return ol_tx_non_std_hl(vdev, tx_spec, msdu_list);
else
return ol_tx_non_std_ll(vdev, tx_spec, msdu_list);
}
void
ol_txrx_data_tx_cb_set(struct cdp_vdev *pvdev,
ol_txrx_data_tx_cb callback, void *ctxt)
{
struct ol_txrx_vdev_t *vdev = (struct ol_txrx_vdev_t *)pvdev;
struct ol_txrx_pdev_t *pdev = vdev->pdev;
pdev->tx_data_callback.func = callback;
pdev->tx_data_callback.ctxt = ctxt;
}
void
ol_txrx_mgmt_tx_cb_set(struct cdp_pdev *ppdev,
uint8_t type,
ol_txrx_mgmt_tx_cb download_cb,
ol_txrx_mgmt_tx_cb ota_ack_cb, void *ctxt)
{
struct ol_txrx_pdev_t *pdev = (struct ol_txrx_pdev_t *)ppdev;
TXRX_ASSERT1(type < OL_TXRX_MGMT_NUM_TYPES);
pdev->tx_mgmt.callbacks[type].download_cb = download_cb;
pdev->tx_mgmt.callbacks[type].ota_ack_cb = ota_ack_cb;
pdev->tx_mgmt.callbacks[type].ctxt = ctxt;
}
#if defined(HELIUMPLUS)
void ol_txrx_dump_frag_desc(char *msg, struct ol_tx_desc_t *tx_desc)
{
uint32_t *frag_ptr_i_p;
int i;
qdf_print("OL TX Descriptor 0x%p msdu_id %d\n",
tx_desc, tx_desc->id);
qdf_print("HTT TX Descriptor vaddr: 0x%p paddr: %pad",
tx_desc->htt_tx_desc, &tx_desc->htt_tx_desc_paddr);
qdf_print("%s %d: Fragment Descriptor 0x%p (paddr=%pad)",
__func__, __LINE__, tx_desc->htt_frag_desc,
&tx_desc->htt_frag_desc_paddr);
/*
* it looks from htt_tx_desc_frag() that tx_desc->htt_frag_desc
* is already de-referrable (=> in virtual address space)
*/
frag_ptr_i_p = tx_desc->htt_frag_desc;
/* Dump 6 words of TSO flags */
print_hex_dump(KERN_DEBUG, "MLE Desc:TSO Flags: ",
DUMP_PREFIX_NONE, 8, 4,
frag_ptr_i_p, 24, true);
frag_ptr_i_p += 6; /* Skip 6 words of TSO flags */
i = 0;
while (*frag_ptr_i_p) {
print_hex_dump(KERN_DEBUG, "MLE Desc:Frag Ptr: ",
DUMP_PREFIX_NONE, 8, 4,
frag_ptr_i_p, 8, true);
i++;
if (i > 5) /* max 6 times: frag_ptr0 to frag_ptr5 */
break;
/* jump to next pointer - skip length */
frag_ptr_i_p += 2;
}
}
#endif /* HELIUMPLUS */
int
ol_txrx_mgmt_send_ext(struct cdp_vdev *pvdev,
qdf_nbuf_t tx_mgmt_frm,
uint8_t type, uint8_t use_6mbps, uint16_t chanfreq)
{
struct ol_txrx_vdev_t *vdev =
(struct ol_txrx_vdev_t *)pvdev;
struct ol_txrx_pdev_t *pdev = vdev->pdev;
struct ol_tx_desc_t *tx_desc;
struct ol_txrx_msdu_info_t tx_msdu_info;
int result = 0;
tx_msdu_info.tso_info.is_tso = 0;
tx_msdu_info.htt.action.use_6mbps = use_6mbps;
tx_msdu_info.htt.info.ext_tid = HTT_TX_EXT_TID_MGMT;
tx_msdu_info.htt.info.vdev_id = vdev->vdev_id;
tx_msdu_info.htt.action.do_tx_complete =
pdev->tx_mgmt.callbacks[type].ota_ack_cb ? 1 : 0;
/*
* FIX THIS: l2_hdr_type should only specify L2 header type
* The Peregrine/Rome HTT layer provides the FW with a "pkt type"
* that is a combination of L2 header type and 802.11 frame type.
* If the 802.11 frame type is "mgmt", then the HTT pkt type is "mgmt".
* But if the 802.11 frame type is "data", then the HTT pkt type is
* the L2 header type (more or less): 802.3 vs. Native WiFi
* (basic 802.11).
* (Or the header type can be "raw", which is any version of the 802.11
* header, and also implies that some of the offloaded tx data
* processing steps may not apply.)
* For efficiency, the Peregrine/Rome HTT uses the msdu_info's
* l2_hdr_type field to program the HTT pkt type. Thus, this txrx SW
* needs to overload the l2_hdr_type to indicate whether the frame is
* data vs. mgmt, as well as 802.3 L2 header vs. 802.11 L2 header.
* To fix this, the msdu_info's l2_hdr_type should be left specifying
* just the L2 header type. For mgmt frames, there should be a
* separate function to patch the HTT pkt type to store a "mgmt" value
* rather than the L2 header type. Then the HTT pkt type can be
* programmed efficiently for data frames, and the msdu_info's
* l2_hdr_type field won't be confusingly overloaded to hold the 802.11
* frame type rather than the L2 header type.
*/
/*
* FIX THIS: remove duplication of htt_frm_type_mgmt and
* htt_pkt_type_mgmt
* The htt module expects a "enum htt_pkt_type" value.
* The htt_dxe module expects a "enum htt_frm_type" value.
* This needs to be cleaned up, so both versions of htt use a
* consistent method of specifying the frame type.
*/
#ifdef QCA_SUPPORT_INTEGRATED_SOC
/* tx mgmt frames always come with a 802.11 header */
tx_msdu_info.htt.info.l2_hdr_type = htt_pkt_type_native_wifi;
tx_msdu_info.htt.info.frame_type = htt_frm_type_mgmt;
#else
tx_msdu_info.htt.info.l2_hdr_type = htt_pkt_type_mgmt;
tx_msdu_info.htt.info.frame_type = htt_pkt_type_mgmt;
#endif
tx_msdu_info.peer = NULL;
tx_desc = ol_txrx_mgmt_tx_desc_alloc(pdev, vdev, tx_mgmt_frm,
&tx_msdu_info);
if (!tx_desc)
return -EINVAL; /* can't accept the tx mgmt frame */
TXRX_STATS_MSDU_INCR(pdev, tx.mgmt, tx_mgmt_frm);
TXRX_ASSERT1(type < OL_TXRX_MGMT_NUM_TYPES);
tx_desc->pkt_type = type + OL_TXRX_MGMT_TYPE_BASE;
result = ol_txrx_mgmt_send_frame(vdev, tx_desc, tx_mgmt_frm,
&tx_msdu_info, chanfreq);
return 0; /* accepted the tx mgmt frame */
}
qdf_nbuf_t ol_tx_reinject(struct ol_txrx_vdev_t *vdev,
qdf_nbuf_t msdu, uint16_t peer_id)
{
struct ol_tx_desc_t *tx_desc = NULL;
struct ol_txrx_msdu_info_t msdu_info;
msdu_info.htt.info.l2_hdr_type = vdev->pdev->htt_pkt_type;
msdu_info.htt.info.ext_tid = HTT_TX_EXT_TID_INVALID;
msdu_info.peer = NULL;
msdu_info.htt.action.tx_comp_req = 0;
msdu_info.tso_info.is_tso = 0;
tx_desc = ol_tx_prepare_ll(vdev, msdu, &msdu_info);
if (!tx_desc)
return msdu;
HTT_TX_DESC_POSTPONED_SET(*((uint32_t *) (tx_desc->htt_tx_desc)), true);
htt_tx_desc_set_peer_id(tx_desc->htt_tx_desc, peer_id);
ol_tx_send(vdev->pdev, tx_desc, msdu, vdev->vdev_id);
return NULL;
}
#if defined(FEATURE_TSO)
/**
* ol_tso_seg_list_init() - function to initialise the tso seg freelist
* @pdev: the data physical device sending the data
* @num_seg: number of segments needs to be intialised
*
* Return: none
*/
void ol_tso_seg_list_init(struct ol_txrx_pdev_t *pdev, uint32_t num_seg)
{
int i = 0;
struct qdf_tso_seg_elem_t *c_element;
/* Host should not allocate any c_element. */
if (num_seg <= 0) {
ol_txrx_err("%s: ERROR: Pool size passed is 0",
__func__);
QDF_BUG(0);
pdev->tso_seg_pool.pool_size = i;
qdf_spinlock_create(&pdev->tso_seg_pool.tso_mutex);
return;
}
c_element = qdf_mem_malloc(sizeof(struct qdf_tso_seg_elem_t));
pdev->tso_seg_pool.freelist = c_element;
for (i = 0; i < (num_seg - 1); i++) {
if (qdf_unlikely(!c_element)) {
ol_txrx_err("%s: ERROR: c_element NULL for seg %d",
__func__, i);
QDF_BUG(0);
pdev->tso_seg_pool.pool_size = i;
pdev->tso_seg_pool.num_free = i;
qdf_spinlock_create(&pdev->tso_seg_pool.tso_mutex);
return;
}
/* set the freelist bit and magic cookie*/
c_element->on_freelist = 1;
c_element->cookie = TSO_SEG_MAGIC_COOKIE;
#ifdef TSOSEG_DEBUG
c_element->dbg.txdesc = NULL;
c_element->dbg.cur = -1; /* history empty */
qdf_tso_seg_dbg_record(c_element, TSOSEG_LOC_INIT1);
#endif /* TSOSEG_DEBUG */
c_element->next =
qdf_mem_malloc(sizeof(struct qdf_tso_seg_elem_t));
c_element = c_element->next;
}
/*
* NULL check for the last c_element of the list or
* first c_element if num_seg is equal to 1.
*/
if (qdf_unlikely(!c_element)) {
ol_txrx_err("%s: ERROR: c_element NULL for seg %d",
__func__, i);
QDF_BUG(0);
pdev->tso_seg_pool.pool_size = i;
pdev->tso_seg_pool.num_free = i;
qdf_spinlock_create(&pdev->tso_seg_pool.tso_mutex);
return;
}
c_element->on_freelist = 1;
c_element->cookie = TSO_SEG_MAGIC_COOKIE;
#ifdef TSOSEG_DEBUG
c_element->dbg.txdesc = NULL;
c_element->dbg.cur = -1; /* history empty */
qdf_tso_seg_dbg_record(c_element, TSOSEG_LOC_INIT2);
#endif /* TSOSEG_DEBUG */
c_element->next = NULL;
pdev->tso_seg_pool.pool_size = num_seg;
pdev->tso_seg_pool.num_free = num_seg;
qdf_spinlock_create(&pdev->tso_seg_pool.tso_mutex);
}
/**
* ol_tso_seg_list_deinit() - function to de-initialise the tso seg freelist
* @pdev: the data physical device sending the data
*
* Return: none
*/
void ol_tso_seg_list_deinit(struct ol_txrx_pdev_t *pdev)
{
int i;
struct qdf_tso_seg_elem_t *c_element;
struct qdf_tso_seg_elem_t *temp;
/* pool size 0 implies that tso seg list is not initialised*/
if (pdev->tso_seg_pool.freelist == NULL &&
pdev->tso_seg_pool.pool_size == 0)
return;
qdf_spin_lock_bh(&pdev->tso_seg_pool.tso_mutex);
c_element = pdev->tso_seg_pool.freelist;
i = pdev->tso_seg_pool.pool_size;
pdev->tso_seg_pool.freelist = NULL;
pdev->tso_seg_pool.num_free = 0;
pdev->tso_seg_pool.pool_size = 0;
qdf_spin_unlock_bh(&pdev->tso_seg_pool.tso_mutex);
qdf_spinlock_destroy(&pdev->tso_seg_pool.tso_mutex);
while (i-- > 0 && c_element) {
temp = c_element->next;
if (c_element->on_freelist != 1) {
qdf_tso_seg_dbg_bug("this seg already freed (double?)");
return;
} else if (c_element->cookie != TSO_SEG_MAGIC_COOKIE) {
qdf_print("this seg cookie is bad (memory corruption?)");
QDF_BUG(0);
return;
}
/* free this seg, so reset the cookie value*/
c_element->cookie = 0;
qdf_mem_free(c_element);
c_element = temp;
}
}
/**
* ol_tso_num_seg_list_init() - function to initialise the freelist of elements
* use to count the num of tso segments in jumbo
* skb packet freelist
* @pdev: the data physical device sending the data
* @num_seg: number of elements needs to be intialised
*
* Return: none
*/
void ol_tso_num_seg_list_init(struct ol_txrx_pdev_t *pdev, uint32_t num_seg)
{
int i = 0;
struct qdf_tso_num_seg_elem_t *c_element;
/* Host should not allocate any c_element. */
if (num_seg <= 0) {
ol_txrx_err("%s: ERROR: Pool size passed is 0",
__func__);
QDF_BUG(0);
pdev->tso_num_seg_pool.num_seg_pool_size = i;
qdf_spinlock_create(&pdev->tso_num_seg_pool.tso_num_seg_mutex);
return;
}
c_element = qdf_mem_malloc(sizeof(struct qdf_tso_num_seg_elem_t));
pdev->tso_num_seg_pool.freelist = c_element;
for (i = 0; i < (num_seg - 1); i++) {
if (qdf_unlikely(!c_element)) {
ol_txrx_err("%s: ERROR: c_element NULL for num of seg %d",
__func__, i);
QDF_BUG(0);
pdev->tso_num_seg_pool.num_seg_pool_size = i;
pdev->tso_num_seg_pool.num_free = i;
qdf_spinlock_create(&pdev->tso_num_seg_pool.
tso_num_seg_mutex);
return;
}
c_element->next =
qdf_mem_malloc(sizeof(struct qdf_tso_num_seg_elem_t));
c_element = c_element->next;
}
/*
* NULL check for the last c_element of the list or
* first c_element if num_seg is equal to 1.
*/
if (qdf_unlikely(!c_element)) {
ol_txrx_err("%s: ERROR: c_element NULL for num of seg %d",
__func__, i);
QDF_BUG(0);
pdev->tso_num_seg_pool.num_seg_pool_size = i;
pdev->tso_num_seg_pool.num_free = i;
qdf_spinlock_create(&pdev->tso_num_seg_pool.tso_num_seg_mutex);
return;
}
c_element->next = NULL;
pdev->tso_num_seg_pool.num_seg_pool_size = num_seg;
pdev->tso_num_seg_pool.num_free = num_seg;
qdf_spinlock_create(&pdev->tso_num_seg_pool.tso_num_seg_mutex);
}
/**
* ol_tso_num_seg_list_deinit() - function to de-initialise the freelist of
* elements use to count the num of tso segment
* in a jumbo skb packet freelist
* @pdev: the data physical device sending the data
*
* Return: none
*/
void ol_tso_num_seg_list_deinit(struct ol_txrx_pdev_t *pdev)
{
int i;
struct qdf_tso_num_seg_elem_t *c_element;
struct qdf_tso_num_seg_elem_t *temp;
/* pool size 0 implies that tso num seg list is not initialised*/
if (pdev->tso_num_seg_pool.freelist == NULL &&
pdev->tso_num_seg_pool.num_seg_pool_size == 0)
return;
qdf_spin_lock_bh(&pdev->tso_num_seg_pool.tso_num_seg_mutex);
c_element = pdev->tso_num_seg_pool.freelist;
i = pdev->tso_num_seg_pool.num_seg_pool_size;
pdev->tso_num_seg_pool.freelist = NULL;
pdev->tso_num_seg_pool.num_free = 0;
pdev->tso_num_seg_pool.num_seg_pool_size = 0;
qdf_spin_unlock_bh(&pdev->tso_num_seg_pool.tso_num_seg_mutex);
qdf_spinlock_destroy(&pdev->tso_num_seg_pool.tso_num_seg_mutex);
while (i-- > 0 && c_element) {
temp = c_element->next;
qdf_mem_free(c_element);
c_element = temp;
}
}
#endif /* FEATURE_TSO */